2014 National Monitoring Programs
Annual Report
(UATMP, NATTS, and CSATAM)
Final Report
EPA Contract No. EP-D-09-048 and 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
February 2017
-------
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-09-048 and 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 xv
List of Figures xvi
List of Tables xxix
List of Acronyms xxxvi
Abstract xxxviii
1.0 Introduction 1-1
1.1 Background 1-1
1.2 The Report 1-2
2.0 The 2014 National Monitoring Programs Network 2-1
2.1 Monitoring Locations 2-1
2.2 Analytical Methods and Pollutants Targeted for Monitoring 2-11
2.2.1 VOC and SNMOC Concurrent Sampling and Analytical Methods ... 2-13
2.2.2 Carbonyl Compound Sampling and Analytical Method 2-17
2.2.3 PAH Sampling and Analytical Method 2-18
2.2.4 Metals Sampling and Analytical Method 2-19
2.2.5 Hexavalent Chromium Sampling and Analytical Method 2-21
2.3 Sample Collection Schedules 2-21
2.4 Completeness 2-27
3.0 Summary of the 2014 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 3-4
3.3 Additional Program-Level Analyses of the 2014 National Monitoring
Programs Dataset 3-7
3.4 Additional Site-Specific Analyses 3-8
3.4.1 Site Characterization 3-8
3.4.2 Meteorological Analysis 3-9
in
-------
TABLE OF CONTENTS (Continued)
Page
3.4.3 Preliminary Risk-Based Screening and Pollutants of Interest 3-10
3.4.3.1 Site-Specific Comparison to Program-level Average
Concentrations 3-10
3.4.3.2 Site Trends Analysis 3-11
3.4.3.3 Cancer Risk and Noncancer Hazard Approximations 3-12
3.4.3.4 Risk-Based Emissions Assessment 3-14
4.0 Summary of the 2014 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 4-14
4.1.3 Central Tendency 4-14
4.2 Preliminary Risk-Based Screening and Pollutants of Interest 4-16
4.2.1 Concentrations of the Pollutants of Interest 4-21
4.2.2 Variability Analysis for the Pollutants of Interest 4-28
4.2.2.1 Inter-site Variability 4-28
4.2.2.2 Quarterly Variability Analysis 4-49
5.0 Sites in Arizona 5-1
5.1 Site Characterization 5-1
5.2 Meteorological Characterization 5-6
5.2.1 Meteorological Summary 5-6
5.2.2 Wind Rose Comparison 5-8
5.3 Pollutants of Interest 5-11
5.4 Concentrations 5-13
5.4.1 2014 Concentration Averages 5-13
5.4.2 Concentration Comparison 5-17
5.4.3 Concentration Trends 5-25
5.5 Additional Risk-Based Screening Evaluations 5-44
5.5.1 Cancer Risk and Noncancer Hazard Approximations 5-44
5.5.2 Risk-Based Emissions Assessment 5-46
5.6 Summary of the 2014 Monitoring Data for PXSS and SPAZ 5-51
iv
-------
TABLE OF CONTENTS (Continued)
Page
6.0 Sites in California 6-1
6.1 Site Characterization 6-1
6.2 Meteorological Characterization 6-10
6.2.1 Meteorological Summary 6-11
6.2.2 Wind Rose Comparison 6-12
6.3 Pollutants of Interest 6-16
6.4 Concentrations 6-17
6.4.1 2014 Concentration Averages 6-17
6.4.2 Concentration Comparison 6-20
6.4.3 Concentration Trends 6-24
6.5 Additional Risk-Based Screening Evaluations 6-31
6.5.1 Cancer Risk and Noncancer Hazard Approximations 6-31
6.5.2 Risk-Based Emissions Assessment 6-32
6.6 Summary of the 2014 Monitoring Data for the California Monitoring Sites .... 6-38
7.0 Sites in Colorado 7-1
7.1 Site Characterization 7-1
7.2 Meteorological Characterization 7-14
7.2.1 Meteorological Summary 7-14
7.2.2 Wind Rose Comparison 7-16
7.3 Pollutants of Interest 7-22
7.4 Concentrations 7-24
7.4.1 2014 Concentration Averages 7-25
7.4.2 Concentration Comparison 7-30
7.4.3 Concentration Trends 7-41
7.5 Additional Risk-Based Screening Evaluations 7-67
7.5.1 Cancer Risk and Noncancer Hazard Approximations 7-67
7.5.2 Risk-Based Emissions Assessment 7-71
7.6 Summary of the 2014 Monitoring Data for the Colorado Monitoring Sites 7-79
v
-------
TABLE OF CONTENTS (Continued)
Page
8.0 Site in the District of Columbia 8-1
8.1 Site Characterization 8-1
8.2 Meteorological Characterization 8-5
8.2.1 Meteorological Summary 8-5
8.2.2 Wind Rose Comparison 8-6
8.3 Pollutants of Interest 8-8
8.4 Concentrations 8-9
8.4.1 2014 Concentration Averages 8-9
8.4.2 Concentration Comparison 8-10
8.4.3 Concentration Trends 8-11
8.5 Additional Risk-Based Screening Evaluations 8-13
8.5.1 Cancer Risk and Noncancer Hazard Approximations 8-13
8.5.2 Risk-Based Emissions Assessment 8-14
8.6 Summary of the 2014 Monitoring Data for WADC 8-17
9.0 Sites in Florida 9-1
9.1 Site Characterization 9-1
9.2 Meteorological Characterization 9-13
9.2.1 Meteorological Summary 9-13
9.2.2 Wind Rose Comparison 9-15
9.3 Pollutants of Interest 9-21
9.4 Concentrations 9-22
9.4.1 2014 Concentration Averages 9-23
9.4.2 Concentration Comparison 9-27
9.4.3 Concentration Trends 9-31
9.5 Additional Risk-Based Screening Evaluations 9-43
9.5.1 Cancer Risk and Noncancer Hazard Approximations 9-43
9.5.2 Risk-Based Emissions Assessment 9-45
9.6 Summary of the 2014 Monitoring Data for the Florida Monitoring Sites 9-53
vi
-------
TABLE OF CONTENTS (Continued)
Page
10.0 Sites in Illinois 10-1
10.1 Site Characterization 10-1
10.2 Meteorological Characterization 10-9
10.2.1 Meteorological Summary 10-9
10.2.2 Wind Rose Comparison 10-11
10.3 Pollutants of Interest 10-15
10.4 Concentrations 10-18
10.4.1 2014 Concentration Averages 10-18
10.4.2 Concentration Comparison 10-25
10.4.3 Concentration Trends 10-36
10.5 Additional Risk-Based Screening Evaluations 10-59
10.5.1 Cancer Risk and Noncancer Hazard Approximations 10-59
10.5.2 Risk-Based Emissions Assessment 10-62
10.6 Summary of the 2014 Monitoring Data for NBIL, SPIL, and ROIL 10-69
11.0 Sites in Indiana 11-1
11.1 Site Characterization 11-1
11.2 Meteorological Characterization 11-8
11.2.1 Meteorological Summary 11-8
11.2.2 Wind Rose Comparison 11-9
11.3 Pollutants of Interest 11-12
11.4 Concentrations 11-13
11.4.1 2014 Concentration Averages 11-13
11.4.2 Concentration Comparison 11-15
11.4.3 Concentration Trends 11-17
11.5 Additional Risk-Based Screening Evaluations 11-23
11.5.1 Cancer Risk and Noncancer Hazard Approximations 11 -23
11.5.2 Risk-Based Emissions Assessment 11-24
11.6 Summary of the 2014 Monitoring Data for INDEM and WPIN 11-28
vii
-------
TABLE OF CONTENTS (Continued)
Page
12.0 Sites in Kentucky 12-1
12.1 Site Characterization 12-1
12.2 Meteorological Characterization 12-20
12.2.1 Meteorological Summary 12-21
12.2.2 Wind Rose Comparison 12-24
12.3 Pollutants of Interest 12-32
12.4 Concentrations 12-39
12.4.1 2014 Concentration Averages 12-39
12.4.2 Concentration Comparison 12-50
12.4.3 Concentration Trends 12-68
12.5 Additional Risk-Based Screening Evaluations 12-73
12.5.1 Cancer Risk and Noncancer Hazard Approximations 12-73
12.5.2 Risk-Based Emissions Assessment 12-80
12.6 Summary of the 2014 Monitoring Data for the Kentucky Monitoring Sites ... 12-95
13.0 Site in Massachusetts 13-1
13.1 Site Characterization 13-1
13.2 Meteorological Characterization 13-5
13.2.1 Meteorological Summary 13-6
13.2.2 Wind Rose Comparison 13-7
13.3 Pollutants of Interest 13-8
13.4 Concentrations 13-9
13.4.1 2014 Concentration Averages 13-9
13.4.2 Concentration Comparison 13-11
13.4.3 Concentration Trends 13-13
13.5 Additional Risk-Based Screening Evaluations 13-17
13.5.1 Cancer Risk and Noncancer Hazard Approximations 13-17
13.5.2 Risk-Based Emissions Assessment 13-18
13.6 Summary of the 2014 Monitoring Data for BOMA 13-21
viii
-------
TABLE OF CONTENTS (Continued)
Page
14.0 Site in Michigan 14-1
14.1 Site Characterization 14-1
14.2 Meteorological Characterization 14-5
14.2.1 Meteorological Summary 14-6
14.2.2 Wind Rose Comparison 14-7
14.3 Pollutants of Interest 14-8
14.4 Concentrations 14-9
14.4.1 2014 Concentration Averages 14-9
14.4.2 Concentration Comparison 14-13
14.4.3 Concentration Trends 14-18
14.5 Additional Risk-Based Screening Evaluations 14-30
14.5.1 Cancer Risk and Noncancer Hazard Approximations 14-30
14.5.2 Risk-Based Emissions Assessment 14-31
14.6 Summary of the 2014 Monitoring Data for DEMI 14-35
15.0 Site in Missouri 15-1
15.1 Site Characterization 15-1
15.2 Meteorological Characterization 15-5
15.2.1 Meteorological Summary 15-5
15.2.2 Wind Rose Comparison 15-7
15.3 Pollutants of Interest 15-8
15.4 Concentrations 15-10
15.4.1 2014 Concentration Averages 15-10
15.4.2 Concentration Comparison 15-15
15.4.3 Concentration Trends 15-24
15.5 Additional Risk-Based Screening Evaluations 15-40
15.5.1 Cancer Risk and Noncancer Hazard Approximations 15-40
15.5.2 Risk-Based Emissions Assessment 15-42
15.6 Summary of the 2014 Monitoring Data for S4MO 15-46
IX
-------
TABLE OF CONTENTS (Continued)
Page
16.0 Sites in New Jersey 16-1
16.1 Site Characterization 16-1
16.2 Meteorological Characterization 16-12
16.2.1 Meteorological Summary 16-12
16.2.2 Wind Rose Comparison 16-13
16.3 Pollutants of Interest 16-17
16.4 Concentrations 16-20
16.4.1 2014 Concentration Averages 16-21
16.4.2 Concentration Comparison 16-28
16.4.3 Concentration Trends 16-37
16.5 Additional Risk-Based Screening Evaluations 16-66
16.5.1 Cancer Risk and Noncancer Hazard Approximations 16-66
16.5.2 Risk-Based Emissions Assessment 16-70
16.6 Summary of the 2014 Monitoring Data for the New Jersey Monitoring
Sites 16-77
17.0 Sites in New York 17-1
17.1 Site Characterization 17-1
17.2 Meteorological Characterization 17-8
17.2.1 Meteorological Summary 17-8
17.2.2 Wind Rose Comparison 17-9
17.3 Pollutants of Interest 17-12
17.4 Concentrations 17-13
17.4.1 2014 Concentration Averages 17-13
17.4.2 Concentration Comparison 17-17
17.4.3 Concentration Trends 17-22
17.5 Additional Risk-Based Screening Evaluations 17-25
17.5.1 Cancer Risk and Noncancer Hazard Approximations 17-26
17.5.2 Risk-Based Emissions Assessment 17-27
17.6 Summary of the 2014 Monitoring Data for BXNY and ROCH 17-31
x
-------
TABLE OF CONTENTS (Continued)
Page
18.0 Sites in Oklahoma 18-1
18.1 Site Characterization 18-1
18.2 Meteorological Characterization 18-12
18.2.1 Meteorological Summary 18-12
18.2.2 Wind Rose Comparison 18-14
18.3 Pollutants of Interest 18-20
18.4 Concentrations 18-24
18.4.1 2014 Concentration Averages 18-24
18.4.2 Concentration Comparison 18-33
18.4.3 Concentration Trends 18-46
18.5 Additional Risk-Based Screening Evaluations 18-77
18.5.1 Cancer Risk and Noncancer Hazard Approximations 18-77
18.5.2 Risk-Based Emissions Assessment 18-81
18.6 Summary of the 2014 Monitoring Data for the Oklahoma Monitoring Sites.. 18-89
19.0 Site in Rhode Island 19-1
19.1 Site Characterization 19-1
19.2 Meteorological Characterization 19-5
19.2.1 Meteorological Summary 19-5
19.2.2 Wind Rose Comparison 19-6
19.3 Pollutants of Interest 19-8
19.4 Concentrations 19-9
19.4.1 2014 Concentration Averages 19-9
19.4.2 Concentration Comparison 19-10
19.4.3 Concentration Trends 19-11
19.5 Additional Risk-Based Screening Evaluations 19-13
19.5.1 Cancer Risk and Noncancer Hazard Approximations 19-13
19.5.2 Risk-Based Emissions Assessment 19-14
19.6 Summary of the 2014 Monitoring Data for PRRI 19-18
XI
-------
TABLE OF CONTENTS (Continued)
Page
20.0 Site in Utah 20-1
20.1 Site Characterization 20-1
20.2 Meteorological Characterization 20-5
20.2.1 Meteorological Summary 20-5
20.2.2 Wind Rose Comparison 20-7
20.3 Pollutants of Interest 20-8
20.4 Concentrations 20-9
20.4.1 2014 Concentration Averages 20-10
20.4.2 Concentration Comparison 20-14
20.4.3 Concentration Trends 20-23
20.5 Additional Risk-Based Screening Evaluations 20-38
20.5.1 Cancer Risk and Noncancer Hazard Approximations 20-38
20.5.2 Risk-Based Emissions Assessment 20-42
20.6 Summary of the 2014 Monitoring Data for BTUT 20-46
21.0 Site in Vermont 21-1
21.1 Site Characterization 21-1
21.2 Meteorological Characterization 21-5
21.2.1 Meteorological Summary 21-5
21.2.2 Wind Rose Comparison 21-7
21.3 Pollutants of Interest 21-8
21.4 Concentrations 21-9
21.4.1 2014 Concentration Averages 21-9
21.4.2 Concentration Comparison 21-11
21.4.3 Concentration Trends 21-12
21.5 Additional Risk-Based Screening Evaluations 21-13
21.5.1 Cancer Risk and Noncancer Hazard Approximations 21-13
21.5.2 Risk-Based Emissions Assessment 21-14
21.6 Summary of the 2014 Monitoring Data for the Vermont Monitoring Site 21-18
xii
-------
TABLE OF CONTENTS (Continued)
Page
22.0 Site in Virginia 22-1
22.1 Site Characterization 22-1
22.2 Meteorological Characterization 22-5
22.2.1 Meteorological Summary 22-6
22.2.2 Wind Rose Comparison 22-7
22.3 Pollutants of Interest 22-8
22.4 Concentrations 22-9
22.4.1 2014 Concentration Averages 22-9
22.4.2 Concentration Comparison 22-10
22.4.3 Concentration Trends 22-11
22.5 Additional Risk-Based Screening Evaluations 22-13
22.5.1 Cancer Risk and Noncancer Hazard Approximations 22-13
22.5.2 Risk-Based Emissions Assessment 22-14
22.6 Summary of the 2014 Monitoring Data for RIVA 22-17
23.0 Site in Washington 23-1
23.1 Site Characterization 23-1
23.2 Meteorological Characterization 23-5
23.2.1 Meteorological Summary 23-5
23.2.2 Wind Rose Comparison 23-7
23.3 Pollutants of Interest 23-8
23.4 Concentrations 23-9
23.4.1 2014 Concentration Averages 23-9
23.4.2 Concentration Comparison 23-13
23.4.3 Concentration Trends 23-19
23.5 Additional Risk-Based Screening Evaluations 23-29
23.5.1 Cancer Risk and Noncancer Hazard Approximations 23-29
23.5.2 Risk-Based Emissions Assessment 23-31
23.6 Summary of the 2014 Monitoring Data for SEW A 23-35
Xlll
-------
TABLE OF CONTENTS (Continued)
Page
24.0 Data Quality 24-1
24.1 Completeness 24-1
24.2 Method Precision 24-2
24.2.1 VOC Method Precision 24-5
24.2.2 SNMOC Method Precision 24-14
24.2.3 Carbonyl Compound Method Precision 24-17
24.2.4 PAH Method Precision 24-21
24.2.5 Metals Method Precision 24-24
24.2.6 Hexavalent Chromium Method Precision 24-26
24.3 Analytical Precision 24-26
24.3.1 VOC Analytical Precision 24-28
24.3.2 SNMOC Analytical Precision 24-39
24.3.3 Carbonyl Compound Analytical Precision 24-46
24.3.4 PAH Analytical Precision 24-52
24.3.5 Metals Analytical Precision 24-56
24.3.6 Hexavalent Chromium Analytical Precision 24-60
24.4 Accuracy 24-61
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 25-18
25.1.4 Data Quality Results Summary 25-23
25.2 Conclusions 25-24
25.3 Recommendations 25-26
26.0 References 26-1
xiv
-------
List of Appendices
Appendix A
AQS Site Descriptions for the 2014 NMP Monitoring Sites
Appendix B
Program Method Detection Limits (MDLs)
Appendix C
2014 VOC Raw Data
Appendix D
2014 SNMOC Raw Data
Appendix E
2014 Carbonyl Compounds Raw Data
Appendix F
2014 PAH and PAH Raw Data
Appendix G
2014 Metals Raw Data
Appendix H
2014 Hexavalent Chromium Raw Data
Appendix I
Summary of Invalidated 2014 Samples
Appendix J
2014 Summary Statistics for VOC Monitoring
Appendix K
2014 Summary Statistics for SNMOC Monitoring
Appendix L
2014 Summary Statistics for Carbonyl Compounds Monitoring
Appendix M
2014 Summary Statistics for PAH and PAH Monitoring
Appendix N
2014 Summary Statistics for Metals Monitoring
Appendix 0
2014 Summary Statistics for Hexavalent Chromium Monitoring
Appendix P
Glossary of Terms
Appendix Q
Risk Factors Used Throughout the 2014 NMP Report
Appendix R
National Weather Service Maps
XV
-------
LIST OF FIGURES
Page
2-1 Locations of the 2014 National Monitoring Programs Monitoring Sites 2-3
4-1 Inter-Site Variability for Acetaldehyde 4-34
4-2 Inter-Site Variability for Arsenic 4-35
4-3a Inter-Site Variability for Benzene - Method TO-15 4-36
4-3b Inter-Site Variability for Benzene - SNMOC 4-37
4-4a Inter-Site Variability for 1,3-Butadiene - Method TO-15 4-38
4-4b Inter-Site Variability for 1,3-Butadiene - SNMOC 4-39
4-5 Inter-Site Variability for Carbon Tetrachloride 4-40
4-6 Inter-Site Variability for/?-Dichlorobenzene 4-41
4-7 Inter-Site Variability for 1,2-Dichloroethane 4-42
4-8a Inter-Site Variability for Ethylbenzene - Method TO-15 4-43
4-8b Inter-Site Variability for Ethylbenzene - SNMOC 4-44
4-9 Inter-Site Variability for Formaldehyde 4-45
4-10 Inter-Site Variability for Hexachloro-l,3-butadiene 4-46
4-11 Inter-Site Variability for Naphthalene 4-47
4-12 Inter-Site Variability for Nickel 4-48
4-13 Comparison of Average Quarterly Acetaldehyde Concentrations 4-54
4-14a Comparison of Average Quarterly Arsenic (PMio) Concentrations 4-55
4-14b Comparison of Average Quarterly Arsenic (TSP) Concentrations 4-56
4-15a Comparison of Average Quarterly Benzene (Method TO-15) Concentrations 4-57
4-15b Comparison of Average Quarterly Benzene (SNMOC) Concentrations 4-58
4-16a Comparison of Average Quarterly 1,3-Butadiene (Method TO-15) Concentrations .... 4-59
4-16b Comparison of Average Quarterly 1,3-Butadiene (SNMOC) Concentrations 4-60
4-17 Comparison of Average Quarterly Carbon Tetrachloride Concentrations 4-61
4-18 Comparison of Average Quarterly /?-Dichlorobenzene Concentrations 4-62
4-19 Comparison of Average Quarterly 1,2-Dichloroethane Concentrations 4-63
4-20a Comparison of Average Quarterly Ethylbenzene (Method TO-15) Concentrations 4-64
4-20b Comparison of Average Quarterly Ethylbenzene (SNMOC) Concentrations 4-65
4-21 Comparison of Average Quarterly Formaldehyde Concentrations 4-66
4-22 Comparison of Average Quarterly Hexachloro-l,3-butadiene Concentrations 4-67
4-23 Comparison of Average Quarterly Naphthalene Concentrations 4-68
4-24a Comparison of Average Quarterly Nickel (PMio) Concentrations 4-69
4-24b Comparison of Average Quarterly Nickel (TSP) Concentrations 4-70
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 Wind Roses for the Wind Data Collected at PXSS 5-9
5-5 Wind Roses for the Wind Data Collected at SPAZ 5-10
5-6 Program vs. Site-Specific Average Acetaldehyde Concentration 5-17
5-7 Program vs. Site-Specific Average Arsenic (PMio) Concentration 5-18
5-8 Program vs. Site-Specific Average Benzene Concentrations 5-18
5-9 Program vs. Site-Specific Average 1,3-Butadiene Concentrations 5-19
5-10 Program vs. Site-Specific Average Carbon Tetrachloride Concentrations 5-20
xvi
-------
LIST OF FIGURES (Continued)
Page
5-11 Program vs. Site-Specific Average/>-Dichlorobenzene Concentrations 5-21
5-12 Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations 5-22
5-13 Program vs. Site-Specific Average Ethylbenzene Concentrations 5-23
5-14 Program vs. Site-Specific Average Formaldehyde Concentration 5-23
5-15 Program vs. Site-Specific Average Hexachloro-1,3-butadiene Concentration 5-24
5-16 Program vs. Site-Specific Average Naphthalene Concentration 5-25
5-17 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at PXSS 5-26
5-18 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at PXSS 5-27
5-19 Yearly Statistical Metrics for Benzene Concentrations Measured at PXSS 5-28
5-20 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at PXSS 5-29
5-21 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
PXSS 5-30
5-22 Yearly Statistical Metrics for /;-Dichlorobenzene Concentrations Measured at
PXSS 5-31
5-23 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
PXSS 5-32
5-24 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at PXSS 5-33
5-25 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at PXSS 5-34
5-26 Yearly Statistical Metrics for Hexachloro-1,3-butadiene Concentrations Measured
at PXSS 5-35
5-27 Yearly Statistical Metrics for Naphthalene Concentrations Measured at PXSS 5-36
5-28 Yearly Statistical Metrics for Benzene Concentrations Measured at SPAZ 5-37
5-29 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at SPAZ 5-38
5-30 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
SPAZ 5-40
5-31 Yearly Statistical Metrics for /;-Dichlorobenzene Concentrations Measured at
SPAZ 5-41
5-32 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
SPAZ 5-42
5-33 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at SPAZ 5-43
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 Wind Roses for the Wind Data Collected at CELA 6-13
6-8 Wind Roses for the Wind Data Collected at RUCA 6-14
6-9 Wind Roses for the Wind Data Collected at SJJCA 6-15
6-10 Program vs. Site-Specific Average Arsenic (PMio) Concentration 6-21
6-11 Program vs. Site-Specific Average Benzo(a)pyrene Concentration 6-21
6-12 Program vs. Site-Specific Average Naphthalene Concentrations 6-22
6-13 Program vs. Site-Specific Average Nickel (PMio) Concentration 6-23
6-14 Yearly Statistical Metrics for Naphthalene Concentrations Measured at CELA 6-24
6-15 Yearly Statistical Metrics for Naphthalene Concentrations Measured at RUCA 6-25
xvii
-------
LIST OF FIGURES (Continued)
Page
6-16 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at SJJCA 6-27
6-17 Yearly Statistical Metrics for Benzo(a)pyrene Concentrations Measured at SJJCA 6-28
6-18 Yearly Statistical Metrics for Naphthalene Concentrations Measured at SJJCA 6-29
6-19 Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at SJJCA 6-30
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
7-8 Carbondale, Colorado (RFCO) Monitoring Site 7-9
7-9 NEI Point Sources Located Within 10 Miles of RFCO 7-10
7-10 Wind Roses for the Wind Data Collected at GPCO 7-17
7-11 Wind Roses for the Wind Data Collected at BMCO 7-18
7-12 Wind Roses for the Wind Data Collected at the Garfield County Regional Airport
Weather Station 7-19
7-13 Wind Roses for the Wind Data Collected at the Aspen-Pitkin County Airport Weather
Station near RFCO 7-21
7-14 Program vs. Site-Specific Average Acenaphthene Concentration 7-31
7-15 Program vs. Site-Specific Average Acetaldehyde Concentrations 7-32
7-16 Program vs. Site-Specific Average Arsenic (PMio) Concentration 7-33
7-17a Program vs. Site-Specific Average Benzene (Method TO-15) Concentration 7-33
7-17b Program vs. Site-Specific Average Benzene (SNMOC) Concentrations 7-34
7-18a Program vs. Site-Specific Average 1,3-Butadiene (Method TO-15) Concentration 7-35
7-18b Program vs. Site-Specific Average 1,3-Butadiene (SNMOC) Concentrations 7-35
7-19 Program vs. Site-Specific Average Carbon Tetrachloride Concentration 7-37
7-20 Program vs. Site-Specific Average 1,2-Dichloroethane Concentration 7-37
7-2la Program vs. Site-Specific Average Ethylbenzene (Method TO-15) Concentration 7-38
7-2lb Program vs. Site-Specific Average Ethylbenzene (SNMOC) Concentration 7-38
7-22 Program vs. Site-Specific Average Formaldehyde Concentrations 7-39
7-23 Program vs. Site-Specific Average Hexachloro-1,3-butadiene Concentration 7-40
7-24 Program vs. Site-Specific Average Naphthalene Concentration 7-41
7-25 Yearly Statistical Metrics for Acenaphthene Concentrations Measured at GPCO 7-42
7-26 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at GPCO 7-43
7-27 Yearly Statistical Metrics for Benzene Concentrations Measured at GPCO 7-44
7-28 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at GPCO 7-45
7-29 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
GPCO 7-46
7-30 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
GPCO 7-47
7-31 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at GPCO 7-48
7-32 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at GPCO 7-50
7-33 Yearly Statistical Metrics for Hexachloro-1,3-butadiene Concentrations Measured
at GPCO 7-51
xviii
-------
LIST OF FIGURES (Continued)
Page
7-34 Yearly Statistical Metrics for Naphthalene Concentrations Measured at GPCO 7-52
7-35 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at BRCO 7-53
7-36 Yearly Statistical Metrics for Benzene Concentrations Measured at BRCO 7-54
7-37 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at BRCO 7-55
7-38 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at PACO 7-56
7-39 Yearly Statistical Metrics for Benzene Concentrations Measured at PACO 7-57
7-40 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at PACO 7-58
7-41 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at PACO 7-60
7-42 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at RICO 7-61
7-43 Yearly Statistical Metrics for Benzene Concentrations Measured at RICO 7-62
7-44 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at RICO 7-63
7-45 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at RICO 7-65
7-46 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at RICO 7-66
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 Wind Roses for the Wind Data Collected at WADC 8-7
8-4 Program vs. Site-Specific Average Naphthalene Concentration 8-11
8-5 Yearly Statistical Metrics for Naphthalene Concentrations Measured at WADC 8-12
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 Wind Roses for the St. Petersburg/Whitted Airport Weather Station near AZFL 9-16
9-10 Wind Roses for the St. Petersburg/Clearwater International Airport Weather
Station near SKFL 9-17
9-11 Wind Roses for the Wind Data Collected at SYFL 9-18
9-12 Wind Roses for the Orlando Executive Airport Weather Station near ORFL 9-19
9-13 Wind Roses for the Orlando Executive Airport Weather Station near PAFL 9-20
9-14 Program vs. Site-Specific Average Acetaldehyde Concentrations 9-28
9-15 Program vs. Site-Specific Average Arsenic (PMio) Concentration 9-29
9-16 Program vs. Site-Specific Average Formaldehyde Concentrations 9-30
9-17 Program vs. Site-Specific Average Naphthalene Concentration 9-31
9-18 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at AZFL 9-32
9-19 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at AZFL 9-33
9-20 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SKFL 9-34
9-21 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at SKFL 9-35
9-22 Yearly Statistical Metrics for Naphthalene Concentrations Measured at SKFL 9-36
9-23 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SYFL 9-37
9-24 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at SYFL 9-38
9-25 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at ORFL 9-40
xix
-------
LIST OF FIGURES (Continued)
Page
9-26 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at ORFL 9-41
9-27 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at PAFL 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 Wind Roses for the Chicago Executive Airport Weather Station near NBIL 10-12
10-7 Wind Roses for the Wind Data Collected at SPIL 10-13
10-8 Wind Roses for the Lambert/St. Louis International Airport Weather Station near
ROIL 10-14
10-9 Program vs. Site-Specific Average Acenaphthene Concentration 10-25
10-10 Program vs. Site-Specific Average Acetaldehyde Concentrations 10-26
10-11 Program vs. Site-Specific Average Benzene Concentrations 10-27
10-12 Program vs. Site-Specific Average 1,3-Butadiene Concentrations 10-28
10-13 Program vs. Site-Specific Average Carbon Tetrachloride Concentrations 10-29
10-14 Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations 10-30
10-15 Program vs. Site-Specific Average Ethylbenzene Concentration 10-31
10-16 Program vs. Site-Specific Average Fluoranthene Concentration 10-31
10-17 Program vs. Site-Specific Average Fluorene Concentration 10-32
10-18 Program vs. Site-Specific Average Formaldehyde Concentrations 10-33
10-19 Program vs. Site-Specific Average Hexachloro-1,3-butadiene Concentrations 10-34
10-20 Program vs. Site-Specific Average Naphthalene Concentration 10-35
10-21 Program vs. Site-Specific Average Trichloroethylene Concentration 10-35
10-22 Yearly Statistical Metrics for Acenaphthene Concentrations Measured at NBIL 10-37
10-23 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at NBIL 10-38
10-24 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at NBIL 10-39
10-25 Yearly Statistical Metrics for Benzene Concentrations Measured at NBIL 10-41
10-26 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at NBIL 10-42
10-27 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
NBIL 10-44
10-28 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
NBIL 10-45
10-29 Yearly Statistical Metrics for Fluoranthene Concentrations Measured at NBIL 10-46
10-30 Yearly Statistical Metrics for Fluorene Concentrations Measured at NBIL 10-47
10-31 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at NBIL 10-48
10-32 Yearly Statistical Metrics for Hexachloro-1,3-butadiene Concentrations Measured
at NBIL 10-49
10-33 Yearly Statistical Metrics for Naphthalene Concentrations Measured at NBIL 10-50
10-34 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SPIL 10-51
10-35 Yearly Statistical Metrics for Benzene Concentrations Measured at SPIL 10-52
10-36 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at SPIL 10-53
10-37 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
SPIL 10-54
xx
-------
LIST OF FIGURES (Continued)
Page
10-38 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
SPII. 10-55
10-39 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at SPIL 10-56
10-40 Yearly Statistical Metrics for Hexachloro-1,3-butadiene Concentrations Measured
at SPIL 10-57
10-41 Yearly Statistical Metrics for Trichloroethylene Concentrations Measured at SPIL.. 10-58
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 Wind Roses for the Wind Data Collected at INDEM 11-10
11-6 Wind Roses for the Wind Data Collected at WPIN 11-11
11-7 Program vs. Site-Specific Average Acetaldehyde Concentrations 11-16
11-8 Program vs. Site-Specific Average Formaldehyde Concentrations 11-17
11-9 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at INDEM.... 11-18
11-10 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at INDEM... 11-19
11-11 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at WPIN 11-21
11-12 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at WPIN 11-22
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
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 (CCKY) Monitoring Site 12-11
12-11 Calvert City, Kentucky (LAKY) Monitoring Site 12-12
12-12 Calvert City, Kentucky (TVKY) Monitoring Site 12-13
12-13 NEI Point Sources Located Within 10 Miles of ATKY, BLKY, CCKY, LAKY,
and TVKY 12-14
12-14 Lexington, Kentucky (LEKY) Monitoring Site 12-15
12-15 NEI Point Sources Located Within 10 Miles of LEKY 12-16
12-16 Wind Roses for the Wind Data Collected at ASKY 12-24
12-17 Wind Roses for the Tri-State/M. J. Ferguson Field Airport Weather Station near
ASKY-M 12-25
12-18 Wind Roses for the Wind Data Collected at GLKY 12-26
12-19 Wind Roses for the Evansville Regional Airport Weather Station near BAKY 12-27
12-20 Wind Roses for the Wind Data Collected at CCKY 12-28
12-21 Wind Roses for the Wind Data Collected at BLKY 12-29
12-22 Wind Roses for the Barkley Regional Airport Weather Station near ATKY, LAKY,
and TVKY 12-30
12-23 Wind Roses for the Blue Grass Airport Weather Station near LEKY 12-31
xxi
-------
LIST OF FIGURES (Continued)
Page
12-24 Program vs. Site-Specific Average Acetaldehyde Concentrations 12-51
12-25 Program vs. Site-Specific Average Arsenic (PMio) Concentrations 12-52
12-26 Program vs. Site-Specific Average Benzene Concentrations 12-54
12-27 Program vs. Site-Specific Average 1,3-Butadiene Concentrations 12-56
12-28 Program vs. Site-Specific Average Cadmium Concentration 12-58
12-29 Program vs. Site-Specific Average Carbon Tetrachloride Concentrations 12-59
12-30 Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations 12-61
12-31 Program vs. Site-Specific Average Formaldehyde Concentrations 12-63
12-32 Program vs. Site-Specific Average Hexachloro-1,3-butadiene Concentrations 12-64
12-33 Program vs. Site-Specific Average Manganese (PMio) Concentration 12-65
12-34 Program vs. Site-Specific Average Nickel (PMio) Concentration 12-65
12-35 Program vs. Site-Specific Average 1,1,2-Trichloroethane Concentration 12-66
12-36 Program vs. Site-Specific Average Vinyl Chloride Concentrations 12-67
12-37 Yearly Statistical Metrics for Benzene Concentrations Measured at GLKY 12-69
12-38 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at GLKY 12-70
12-39 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
GLKY 12-71
12-40 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
GLKY 12-72
12-41 Pollution Roses for 1,2-Dichloroethane Concentrations Measured at TVKY 12-79
13-1 Boston, Massachusetts (BOMA) Monitoring Site 13-2
13-2 NEI Point Sources Located Within 10 Miles of BOMA 13-3
13-3 Wind Roses for the Wind Data Collected at BOMA 13-7
13-4 Program vs. Site-Specific Average Arsenic (PMio) Concentration 13-11
13-5 Program vs. Site-Specific Average Naphthalene Concentration 13-12
13-6 Program vs. Site-Specific Average Nickel (PMio) Concentration 13-13
13-7 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at BOMA... 13-14
13-8 Yearly Statistical Metrics for Naphthalene Concentrations Measured at BOMA 13-15
13-9 Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at BOMA.... 13-16
14-1 Dearborn, Michigan (DEMI) Monitoring Site 14-2
14-2 NEI Point Sources Located Within 10 Miles of DEMI 14-3
14-3 Wind Roses for the Wind Data Collected at DEMI 14-7
14-4 Program vs. Site-Specific Average Acetaldehyde Concentration 14-13
14-5 Program vs. Site-Specific Average Benzene Concentration 14-13
14-6 Program vs. Site-Specific Average 1,3-Butadiene Concentration 14-14
14-7 Program vs. Site-Specific Average Carbon Tetrachloride Concentration 14-15
14-8 Program vs. Site-Specific Average 1,2-Dichloroethane Concentration 14-15
14-9 Program vs. Site-Specific Average Ethylbenzene Concentration 14-16
14-10 Program vs. Site-Specific Average Fluorene Concentration 14-17
14-11 Program vs. Site-Specific Average Formaldehyde Concentration 14-17
14-12 Program vs. Site-Specific Average Naphthalene Concentration 14-18
14-13 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at DEMI 14-19
14-14 Yearly Statistical Metrics for Benzene Concentrations Measured at DEMI 14-20
14-15 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at DEMI 14-21
xxii
-------
LIST OF FIGURES (Continued)
Page
14-16 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
DEMI 14-22
14-17 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
DEMI 14-24
14-18 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at DEMI 14-25
14-19 Yearly Statistical Metrics for Fluorene Concentrations Measured at DEMI 14-26
14-20 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at DEMI 14-27
14-21 Yearly Statistical Metrics for Naphthalene Concentrations Measured at DEMI 14-29
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 Wind Roses for the Wind Data Collected at S4MO 15-7
15-4 Program vs. Site-Specific Average Acenaphthene Concentration 15-15
15-5 Program vs. Site-Specific Average Acetaldehyde Concentration 15-16
15-6 Program vs. Site-Specific Average Arsenic (PMio) Concentration 15-16
15-7 Program vs. Site-Specific Average Benzene Concentration 15-17
15-8 Program vs. Site-Specific Average 1,3-Butadiene Concentration 15-17
15-9 Program vs. Site-Specific Average Cadmium (PMio) Concentration 15-18
15-10 Program vs. Site-Specific Average Carbon Tetrachloride Concentration 15-19
15-11 Program vs. Site-Specific Average/?-Dichlorobenzene Concentration 15-19
15-12 Program vs. Site-Specific Average 1,2-Dichloroethane Concentration 15-20
15-13 Program vs. Site-Specific Average Ethylbenzene Concentration 15-21
15-14 Program vs. Site-Specific Average Fluorene Concentration 15-21
15-15 Program vs. Site-Specific Average Formaldehyde Concentration 15-22
15-16 Program vs. Site-Specific Average Hexachloro-1,3-butadiene Concentration 15-22
15-17 Program vs. Site-Specific Average Naphthalene Concentration 15-23
15-18 Yearly Statistical Metrics for Acenaphthene Concentrations Measured at S4MO 15-25
15-19 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at S4MO 15-26
15-20 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at S4MO.... 15-27
15-21 Yearly Statistical Metrics for Benzene Concentrations Measured at S4MO 15-28
15-22 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at S4MO 15-29
15-23 Yearly Statistical Metrics for Cadmium (PMio) Concentrations Measured at
S4MO 15-30
15-24 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
S4MO 15-31
15-25 Yearly Statistical Metrics for /;-Dichlorobenzene Concentrations Measured at
S4MO 15-32
15-26 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
S4MO 15-34
15-27 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at S4MO 15-35
15-28 Yearly Statistical Metrics for Fluorene Concentrations Measured at S4MO 15-36
15-29 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at S4MO 15-37
15-30 Yearly Statistical Metrics for Hexachloro-1,3-butadiene Concentrations Measured
at S4MO 15-38
15-31 Yearly Statistical Metrics for Naphthalene Concentrations Measured at S4MO 15-39
XXlll
-------
LIST OF FIGURES (Continued)
Page
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 NEI Point Sources Located Within 10 Miles of ELNJ and NBNJ 16-8
16-8 Wind Roses for the Philadelphia International Airport Weather Station near CSNJ.. 16-14
16-9 Wind Roses for the Somerville-Somerset Airport Weather Station near CHNJ and
NBNJ 16-15
16-10 Wind Roses for the Newark International Airport Weather Station near ELNJ 16-16
16-11 Program vs. Site-Specific Average Acetaldehyde Concentrations 16-29
16-12 Program vs. Site-Specific Average Benzene Concentrations 16-30
16-13 Program vs. Site-Specific Average 1,3-Butadiene Concentrations 16-31
16-14 Program vs. Site-Specific Average Carbon Tetrachloride Concentrations 16-32
16-15 Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations 16-33
16-16 Program vs. Site-Specific Average Ethylbenzene Concentrations 16-34
16-17 Program vs. Site-Specific Average Formaldehyde Concentrations 16-35
16-18 Program vs. Site-Specific Average Hexachloro-1,3-butadiene Concentrations 16-36
16-19 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at CHNJ 16-38
16-20 Yearly Statistical Metrics for Benzene Concentrations Measured at CHNJ 16-39
16-21 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at CHNJ 16-41
16-22 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
CHNJ 16-42
16-23 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
CHNJ 16-43
16-24 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at CHNJ 16-44
16-25 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at ELNJ 16-46
16-26 Yearly Statistical Metrics for Benzene Concentrations Measured at ELNJ 16-47
16-27 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at ELNJ 16-49
16-28 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
ELNJ 16-50
16-29 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
ELNJ 16-51
16-30 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at ELNJ 16-52
16-31 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at ELNJ 16-53
16-32 Yearly Statistical Metrics for Hexachloro-1,3-butadiene Concentrations Measured
at ELNJ 16-54
16-33 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at NBNJ 16-55
16-34 Yearly Statistical Metrics for Benzene Concentrations Measured at NBNJ 16-56
16-35 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at NBNJ 16-58
16-36 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
NBNJ 16-59
16-37 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
NBNJ 16-61
16-38 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at NBNJ 16-62
xxiv
-------
LIST OF FIGURES (Continued)
Page
16-39 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at NBNJ 16-63
16-40 Yearly Statistical Metrics for Hexachloro-1,3-butadiene Concentrations Measured
at NBNJ 16-65
17-1 New York City, 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 Wind Roses for the La Guardia Airport Weather Station near BXNY 17-10
17-6 Wind Roses for the Wind Data Collected at ROCH 17-11
17-7 Program vs. Site-Specific Average Acenaphthene Concentrations 17-18
17-8 Program vs. Site-Specific Average Benzo(a)pyrene Concentration 17-19
17-9 Program vs. Site-Specific Average Fluoranthene Concentration 17-19
17-10 Program vs. Site-Specific Average Fluorene Concentrations 17-20
17-11 Program vs. Site-Specific Average Naphthalene Concentrations 17-21
17-12 Yearly Statistical Metrics for Acenaphthene Concentrations Measured at ROCH 17-23
17-13 Yearly Statistical Metrics for Fluorene Concentrations Measured at ROCH 17-24
17-14 Yearly Statistical Metrics for Naphthalene Concentrations Measured at ROCH 17-25
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 Yukon, Oklahoma (YUOK) Monitoring Site 18-7
18-7 NEI Point Sources Located Within 10 Miles of OCOK and YUOK 18-8
18-8 Wind Roses for the Wind Data Collected at TOOK 18-15
18-9 Wind Roses for the Wind Data Collected at TMOK 18-16
18-10 Wind Roses for the Wind Data Collected at TROK 18-17
18-11 Wind Roses for the Wind Data Collected at OCOK 18-18
18-12 Wind Roses for the Wind Data Collected at YUOK 18-19
18-13 Program vs. Site-Specific Average Acetaldehyde Concentrations 18-34
18-14 Program vs. Site-Specific Average Arsenic (TSP) Concentrations 18-35
18-15 Program vs. Site-Specific Average Benzene Concentrations 18-36
18-16 Program vs. Site-Specific Average 1,3-Butadiene Concentrations 18-37
18-17 Program vs. Site-Specific Average Carbon Tetrachloride Concentrations 18-39
18-18 Program vs. Site-Specific Average/?-Dichlorobenzene Concentrations 18-40
18-19 Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations 18-41
18-20 Program vs. Site-Specific Average Ethylbenzene Concentrations 18-42
18-21 Program vs. Site-Specific Average Formaldehyde Concentrations 18-43
18-22 Program vs. Site-Specific Average Hexachloro-1,3-butadiene Concentrations 18-44
18-23 Program vs. Site-Specific Average Manganese (TSP) Concentration 18-45
18-24 Program vs. Site-Specific Average Nickel (TSP) Concentrations 18-45
18-25 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at TOOK 18-47
18-26 Yearly Statistical Metrics for Arsenic (TSP) Concentrations Measured at TOOK 18-48
18-27 Yearly Statistical Metrics for Benzene Concentrations Measured at TOOK 18-49
xxv
-------
LIST OF FIGURES (Continued)
Page
18-28 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at TOOK 18-50
18-29 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
TOOK 18-51
18-30 Yearly Statistical Metrics for /;-Dichlorobenzene Concentrations Measured at
TOOK 18-52
18-31 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
TOOK 18-53
18-32 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at TOOK 18-54
18-33 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at TOOK 18-55
18-34 Yearly Statistical Metrics for Hexachloro-1,3-butadiene Concentrations Measured
at TOOK 18-57
18-35 Yearly Statistical Metrics for Manganese (TSP) Concentrations Measured at
TOOK 18-58
18-36 Yearly Statistical Metrics for Nickel (TSP) Concentrations Measured at TOOK 18-59
18-37 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at TMOK 18-60
18-38 Yearly Statistical Metrics for Arsenic (TSP) Concentrations Measured at TMOK .... 18-61
18-39 Yearly Statistical Metrics for Benzene Concentrations Measured at TMOK 18-62
18-40 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at TMOK 18-63
18-41 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
TMOK 18-64
18-42 Yearly Statistical Metrics for /;-Dichlorobenzene Concentrations Measured at
TMOK 18-65
18-43 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
TMOK 18-66
18-44 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at
TMOK 18-67
18-45 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at TMOK.... 18-68
18-46 Yearly Statistical Metrics for Hexachloro-1,3-butadiene Concentrations Measured
at TMOK 18-69
18-47 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at OCOK 18-70
18-48 Yearly Statistical Metrics for Arsenic (TSP) Concentrations Measured at OCOK 18-71
18-49 Yearly Statistical Metrics for Benzene Concentrations Measured at OCOK 18-72
18-50 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at OCOK 18-73
18-51 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
OCOK 18-74
18-52 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
OCOK 18-75
18-53 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at OCOK 18-76
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 Wind Roses for the Wind Data Collected at PRRI 19-7
19-4 Program vs. Site-Specific Average Naphthalene Concentration 19-11
19-5 Yearly Statistical Metrics for Naphthalene Concentrations Measured at PRRI 19-12
20-1 Bountiful, Utah (BTUT) Monitoring Site 20-2
xxvi
-------
LIST OF FIGURES (Continued)
Page
20-2 NEI Point Sources Located Within 10 Miles of BTUT 20-3
20-3 Wind Roses for the Wind Data Collected at BTUT 20-7
20-4 Program vs. Site-Specific Average Acetaldehyde Concentration 20-15
20-5 Program vs. Site-Specific Average Arsenic (PMio) Concentration 20-15
20-6 Program vs. Site-Specific Average Benzene Concentration 20-16
20-7 Program vs. Site-Specific Average 1,3-Butadiene Concentration 20-17
20-8 Program vs. Site-Specific Average Carbon Tetrachloride Concentration 20-17
20-9 Program vs. Site-Specific Average 1,2-Dichloroethane Concentration 20-18
20-10 Program vs. Site-Specific Average Dichloromethane Concentration 20-19
20-11 Program vs. Site-Specific Average Ethylbenzene Concentration 20-20
20-12 Program vs. Site-Specific Average Formaldehyde Concentration 20-20
20-13 Program vs. Site-Specific Average Hexachloro-1,3-butadiene Concentration 20-21
20-14 Program vs. Site-Specific Average Naphthalene Concentration 20-22
20-15 Program vs. Site-Specific Average Nickel (PMio) Concentration 20-22
20-16 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at BTUT 20-23
20-17 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at BTUT.... 20-24
20-18 Yearly Statistical Metrics for Benzene Concentrations Measured at BTUT 20-26
20-19 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at BTUT 20-27
20-20 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
BTUT 20-28
20-21 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
BTUT 20-29
20-22 Yearly Statistical Metrics for Dichloromethane Concentrations Measured at
BTUT 20-30
20-23 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at BTUT 20-31
20-24 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at BTUT 20-33
20-25 Yearly Statistical Metrics for Hexachloro-1,3-butadiene Concentrations Measured
at BTUT 20-35
20-26 Yearly Statistical Metrics for Naphthalene Concentrations Measured at BTUT 20-36
20-27 Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at BTUT 20-37
20-28 Pollution Rose for Formaldehyde Concentrations Measured at BTUT 20-41
21-1 Underhill, Vermont (UNVT) Monitoring Site 21-2
21-2 NEI Point Sources Located Within 10 Miles of UNVT 21-3
21-3 Wind Roses for the Wind Data Collected at UNVT 21-7
21-4 Program vs. Site-Specific Average Arsenic (PMio) Concentration 21-11
21-5 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at UNVT ... 21-12
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 Wind Roses for the Wind Data Collected at RIVA 22-7
22-4 Program vs. Site-Specific Average Naphthalene Concentration 22-11
22-5 Yearly Statistical Metrics for Naphthalene Concentrations Measured at RIVA 22-12
23-1 Seattle, Washington (SEWA) Monitoring Site 23-2
23-2 NEI Point Sources Located Within 10 Miles of SEWA 23-3
xxvii
-------
LIST OF FIGURES (Continued)
Page
23-3 Wind Roses for the Wind Data Collected at SEWA 23-7
23-4 Program vs. Site-Specific Average Acetaldehyde Concentration 23-13
23-5 Program vs. Site-Specific Average Arsenic (PMio) Concentration 23-14
23-6 Program vs. Site-Specific Average Benzene Concentration 23-14
23-7 Program vs. Site-Specific Average 1,3-Butadiene Concentration 23-15
23-8 Program vs. Site-Specific Average Carbon Tetrachloride Concentration 23-16
23-9 Program vs. Site-Specific Average 1,2-Dichloroethane Concentration 23-16
23-10 Program vs. Site-Specific Average Formaldehyde Concentration 23-17
23-11 Program vs. Site-Specific Average Naphthalene Concentration 23-18
23-12 Program vs. Site-Specific Average Nickel (PMio) Concentration 23-18
23-13 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SEWA 23-19
23-14 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at
SEWA 23-20
23-15 Yearly Statistical Metrics for Benzene Concentrations Measured at SEWA 23-21
23-16 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at SEWA 23-22
23-17 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
SEWA 23-23
23-18 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
SEWA 23-24
23-19 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at SEWA 23-25
23-20 Yearly Statistical Metrics for Naphthalene Concentrations Measured at SEWA 23-26
23-21 Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at SEWA 23-28
xxviii
-------
LIST OF TABLES
Page
1-1 Organization of the 2014 National Monitoring Programs Report 1-4
2-1 2014 National Monitoring Programs Sites and Past Program Participation 2-4
2-2 Site Characterizing Information for the 2014 National Monitoring Programs Sites 2-7
2-3 2014 VOC Method Detection Limits 2-15
2-4 2014 SNMOC Method Detection Limits 2-16
2-5 2014 Carbonyl Compound Method Detection Limits 2-18
2-6 2014 PAH Method Detection Limits 2-19
2-7 2014 Metals Method Detection Limits 2-20
2-8 2014 Hexavalent Chromium Method Detection Limit 2-21
2-9 2014 Sampling Schedules and Completeness Rates 2-23
2-10 Method Completeness Rates for 2014 2-28
3-1 Overview and Layout of Data Presented 3-1
3-2 NATTS MQO Core Analytes 3-7
3 -3 POM Groups for PA I Is 3-15
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 Carbonyl Compound Concentrations 4-10
4-4 Statistical Summaries of the PAH Concentrations 4-11
4-5 Statistical Summaries of the Metals Concentrations 4-12
4-6 Statistical Summary of the Hexavalent Chromium Concentrations 4-13
4-7 Results of the Program-Level Preliminary Risk-Based Screening Process 4-17
4-8 Site-Specific Risk-Based Screening Comparison 4-19
4-9 Annual Average Concentration Comparison of the VOC/SNMOC Pollutants of
Interest 4-22
4-10 Annual Average Concentration Comparison of the Carbonyl Compound Pollutants
of Interest 4-23
4-11 Annual Average Concentration Comparison of the PAH Pollutant of Interest 4-23
4-12 Annual Average Concentration Comparison of the Metals Pollutants of Interest 4-24
5-1 Geographical Information for the Arizona Monitoring Sites 5-5
5-2 Average Meteorological Conditions near the Arizona Monitoring Sites 5-7
5-3 Risk-Based Screening Results for the Arizona Monitoring Sites 5-12
5-4 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Arizona Monitoring Sites 5-14
5-5 Risk Approximations for the Arizona Monitoring Sites 5-45
5-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Arizona Monitoring Sites 5-47
5-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Arizona Monitoring
Sites 5-48
XXIX
-------
LIST OF TABLES (Continued)
Page
6-1 Geographical Information for the California Monitoring Sites 6-8
6-2 Average Meteorological Conditions near the California Monitoring Sites 6-12
6-3 Risk-Based Screening Results for the California Monitoring Sites 6-16
6-4 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
California Monitoring Sites 6-18
6-5 Risk Approximations for the California Monitoring Sites 6-32
6-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the California Monitoring Sites 6-34
6-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the California Monitoring
Sites 6-36
7-1 Geographical Information for the Colorado Monitoring Sites 7-11
7-2 Average Meteorological Conditions near the Colorado Monitoring Sites 7-15
7-3 Risk-Based Screening Results for the Colorado Monitoring Sites 7-22
7-4 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Colorado Monitoring Sites 7-26
7-5 Risk Approximations for the Colorado Monitoring Sites 7-68
7-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Colorado Monitoring Sites 7-72
7-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Colorado Monitoring
Sites 7-75
8-1 Geographical Information for the Washington, D.C. Monitoring Site 8-4
8-2 Average Meteorological Conditions near the Washington, D.C. Monitoring Site 8-6
8-3 Risk-Based Screening Results for the Washington, D.C. Monitoring Site 8-8
8-4 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Washington, D.C. Monitoring Site 8-10
8-5 Risk Approximations for the Washington, D.C. Monitoring Site 8-13
8-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Washington, D.C. Monitoring Site 8-15
8-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Washington, D.C.
Monitoring Site 8-16
9-1 Geographical Information for the Florida Monitoring Sites 9-10
9-2 Average Meteorological Conditions near the Florida Monitoring Sites 9-14
9-3 Risk-Based Screening Results for the Florida Monitoring Sites 9-21
9-4 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Florida Monitoring Sites 9-24
9-5 Risk Approximations for the Florida Monitoring Sites 9-44
9-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Florida Monitoring Sites 9-46
XXX
-------
LIST OF TABLES (Continued)
Page
9-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Florida Monitoring
Sites 9-49
10-1 Geographical Information for the Illinois Monitoring Sites 10-7
10-2 Average Meteorological Conditions near the Illinois Monitoring Sites 10-10
10-3 Risk-Based Screening Results for the Illinois Monitoring Sites 10-15
10-4 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Illinois Monitoring Sites 10-19
10-5 Risk Approximations for the Illinois Monitoring Sites 10-60
10-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Illinois Monitoring Sites 10-63
10-7 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 Average Meteorological Conditions near the Indiana Monitoring Sites 11-9
11-3 Risk-Based Screening Results for the Indiana Monitoring Sites 11-12
11-4 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Indiana Monitoring Sites 11-14
11-5 Risk Approximations for the Indiana Monitoring Sites 11-23
11-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Indiana Monitoring Sites 11-25
11-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Indiana Monitoring
Sites 11-26
12-1 Geographical Information for the Kentucky Monitoring Sites 12-17
12-2 Average Meteorological Conditions near the Kentucky Monitoring Sites 12-22
12-3 Overview of Pollutant Groups Sampled for at the Kentucky Monitoring Sites 12-33
12-4 Risk-Based Screening Results for the Kentucky Monitoring Sites 12-34
12-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Kentucky Monitoring Sites 12-40
12-6 Risk Approximations for the Kentucky Monitoring Sites 12-73
12-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Kentucky Monitoring Sites 12-81
12-8 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Kentucky Monitoring
Sites 12-86
13-1 Geographical Information for the Massachusetts Monitoring Site 13-4
13-2 Average Meteorological Conditions near the Massachusetts Monitoring Site 13-6
13-3 Risk-Based Screening Results for the Massachusetts Monitoring Site 13-8
13-4 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Massachusetts Monitoring Site 13-10
xxxi
-------
LIST OF TABLES (Continued)
Page
13-5 Risk Approximations for the Massachusetts Monitoring Site 13-18
13-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Massachusetts Monitoring Site 13-19
13-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Massachusetts
Monitoring Site 13-20
14-1 Geographical Information for the Michigan Monitoring Site 14-4
14-2 Average Meteorological Conditions near the Michigan Monitoring Site 14-6
14-3 Risk-Based Screening Results for the Michigan Monitoring Site 14-8
14-4 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Michigan Monitoring Site 14-10
14-5 Risk Approximations for the Michigan Monitoring Site 14-31
14-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Michigan Monitoring Site 14-32
14-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Michigan Monitoring
Site 14-33
15-1 Geographical Information for the Missouri Monitoring Site 15-4
15-2 Average Meteorological Conditions near the Missouri Monitoring Site 15-6
15-3 Risk-Based Screening Results for the Missouri Monitoring Site 15-9
15-4 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Missouri Monitoring Site 15-11
15-5 Risk Approximations for the Missouri Monitoring Site 15-41
15-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Missouri Monitoring Site 15-43
15-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Missouri Monitoring
Site 15-44
16-1 Geographical Information for the New Jersey Monitoring Sites 16-9
16-2 Average Meteorological Conditions near the New Jersey Monitoring Sites 16-13
16-3 Risk-Based Screening Results for the New Jersey Monitoring Sites 16-18
16-4 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
New Jersey Monitoring Sites 16-22
16-5 Risk Approximations for the New Jersey Monitoring Sites 16-67
16-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the New Jersey Monitoring Sites 16-71
16-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the New Jersey Monitoring
Sites 16-73
17-1 Geographical Information for the New York Monitoring Sites 17-6
17-2 Average Meteorological Conditions near the New York Monitoring Sites 17-9
17-3 Risk-Based Screening Results for the New York Monitoring Sites 17-12
XXXll
-------
LIST OF TABLES (Continued)
Page
17-4 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
New York Monitoring Sites 17-14
17-5 Risk Approximations for the New York Monitoring Sites 17-26
17-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the New York Monitoring Sites 17-28
17-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the New York Monitoring
Sites 17-29
18-1 Geographical Information for the Oklahoma Monitoring Sites 18-9
18-2 Average Meteorological Conditions near the Oklahoma Monitoring Sites 18-13
18-3 Risk-Based Screening Results for the Oklahoma Monitoring Sites 18-20
18-4 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Oklahoma Monitoring Sites 18-25
18-5 Risk Approximations for the Oklahoma Monitoring Sites 18-78
18-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Oklahoma Monitoring Sites 18-82
18-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Oklahoma Monitoring
Sites 18-85
19-1 Geographical Information for the Rhode Island Monitoring Site 19-4
19-2 Average Meteorological Conditions near the Rhode Island Monitoring Site 19-6
19-3 Risk-Based Screening Results for the Rhode Island Monitoring Site 19-8
19-4 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Rhode Island Monitoring Site 19-10
19-5 Risk Approximations for the Rhode Island Monitoring Site 19-13
19-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Rhode Island Monitoring Site 19-15
19-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Rhode Island
Monitoring Site 19-16
20-1 Geographical Information for the Utah Monitoring Site 20-4
20-2 Average Meteorological Conditions near the Utah Monitoring Site 20-6
20-3 Risk-Based Screening Results for the Utah Monitoring Site 20-8
20-4 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Utah Monitoring Site 20-11
20-5 Risk Approximations for the Utah Monitoring Site 20-39
20-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Utah Monitoring Site 20-43
20-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Utah Monitoring
Site 20-44
XXXlll
-------
LIST OF TABLES (Continued)
Page
21-1 Geographical Information for the Vermont Monitoring Site 21-4
21-2 Average Meteorological Conditions near the Vermont Monitoring Site 21-6
21-3 Risk-Based Screening Results for the Vermont Monitoring Site 21-8
21-4 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Vermont Monitoring Site 21-10
21-5 Risk Approximations for the Vermont Monitoring Site 21-14
21-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Vermont Monitoring Site 21-15
21-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Vermont
Monitoring Site 21-16
22-1 Geographical Information for the Virginia Monitoring Site 22-4
22-2 Average Meteorological Conditions near the Virginia Monitoring Site 22-6
22-3 Risk-Based Screening Results for the Virginia Monitoring Site 22-8
22-4 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Virginia Monitoring Site 22-10
22-5 Risk Approximations for the Virginia Monitoring Site 22-13
22-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Virginia Monitoring Site 22-15
22-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Virginia Monitoring
Site 22-16
23-1 Geographical Information for the Washington Monitoring Site 23-4
23-2 Average Meteorological Conditions near the Washington Monitoring Site 23-6
23-3 Risk-Based Screening Results for the Washington Monitoring Site 23-8
23-4 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Washington Monitoring Site 23-10
23-5 Risk Approximations for the Washington Monitoring Site 23-30
23-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Washington Monitoring Site 23-32
23-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Washington
Monitoring Site 23-33
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 Duplicate Samples
by Site and Pollutant 24-14
24-4 Carbonyl Compound Method Precision: Coefficient of Variation Based on
Duplicate and Collocated Samples by Site and Pollutant 24-18
24-5 PAH Method Precision: Coefficient of Variation Based on Collocated Samples
by Site and Pollutant 24-23
xxxiv
-------
LIST OF TABLES (Continued)
Page
24-6 Metals Method Precision: Coefficient of Variation Based on Collocated Samples
by Site and Pollutant 24-25
24-7 Hexavalent Chromium Method Precision: Coefficient of Variation Based on
Collocated Samples by Site 24-26
24-8 Analytical Precision by Analytical Method 24-27
24-9 VOC Analytical Precision: Coefficient of Variation Based on Replicate Analyses
by Site and Pollutant 24-29
24-10 SNMOC Analytical Precision: Coefficient of Variation Based on Replicate
Analyses by Site and Pollutant 24-40
24-11 Carbonyl Compound Analytical Precision: Coefficient of Variation Based on
Replicate Analyses by Site and Pollutant 24-47
24-12 PAH Analytical Precision: Coefficient of Variation Based on Replicate Analyses
by Site and Pollutant 24-53
24-13 Metals Analytical Precision: Coefficient of Variation Based on Replicate Analyses
by Site and Pollutant 24-57
24-14 Hexavalent Chromium Analytical Precision: Coefficient of Variation Based on
Replicate Analyses by Site 24-61
24-15 TO-15 NATTS PT Audit Samples 24-63
24-16 I O-l 1A NATTS PT Audit Samples 24-63
24-17 TO-13A NATTS PT Audit Samples 24-64
24-18 Metals NATTS PT Audit Samples 24-64
24-19 Hexavalent Chromium NATTS PT Audit Samples 24-64
24-20 Lead NAAQS Quarterly Audit Samples 24-65
25-1 Summary of Site-Specific Pollutants of Interest 25-19
xxxv
-------
LIST OF ACRONYMS
AADT
Annual Average Daily Traffic
AQS
Air Quality System
ASE
Accelerated Solvent Extractor
CBS A
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.
F
Fahrenheit
FAC
Federal Advisory Committee
FEM
Federal Equivalent Method
GC/MS-FID
Gas Chromatography/Mass Spectrometry and 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
In Hg
Inches of Mercury
kt
Knots
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
NCDC
National Climatic Data Center
ND
Non-detect
NEI
National Emissions Inventory
ng/m3
Nanograms per cubic meter
NMOC
Non-Methane Organic Compound(s)
NMP
National Monitoring Programs
NWS
National Weather Service
PAH
Polycyclic Aromatic Hydrocarbon(s)
PAMS
Photochemical Assessment Monitoring Stations
PM
Particulate Matter
xxxvi
-------
LIST OF ACRONYMS (Continued)
PMio
Particulate Matter less than 10 microns
POM
Polycyclic Organic Matter
ppbC
Parts per billion carbon
ppbv
Parts per billion by volume
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
SIP
State Implementation Plan(s)
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
yiL
Microliter
URE
Unit Risk Estimate(s)
UY
Ultraviolet
VOC
Volatile Organic Compound(s)
WBAN
Weather Bureau/Army/Navy ID
xxxvii
-------
Abstract
This report presents the results and conclusions from the ambient air monitoring conducted
as part of the 2014 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 2014 NMP includes data from
samples collected at 51 monitoring sites that collected 24-hour air samples, typically on a l-in-6
or l-in-12 day schedule, and analyzed by the national contract laboratory. Twenty-seven sites
sampled for 59 volatile organic compounds (VOCs); 32 sites sampled for 15 carbonyl
compounds; seven sites sampled for 80 speciated nonmethane organic compounds (SNMOCs);
19 sites sampled for 22 polycyclic aromatic hydrocarbons (PAHs); 21 sites sampled for 11
metals; and 2 sites sampled for hexavalent chromium. Nearly 225,000 ambient air concentrations
were measured during the 2014 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 and from season-to-season.
The ambient air monitoring data collected during the 2014 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 51 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.
xxxviii
-------
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. To achieve this goal, EPA sponsors the National Monitoring Programs (NMP), which
include the Photochemical Assessment Monitoring Stations (PAMS) network, 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 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, 2009a/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/ttnamti 1/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/ttnamti 1/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,
2009a). 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
1-1
-------
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
decrease over a period of time) may be identified. 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, 2016a).
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, 2009b). Currently, 27 NATTS sites are strategically placed across the country (EPA,
2016a).
The CSATAM Program was initiated in 2004 and is intended to support state, local, and
tribal agencies in conducting discreet, investigative projects of approximately 2-year durations
via periodic grant competitions. The objectives of the CSATAM Program include identifying
and profiling air toxics sources; developing and evaluating emerging measurement methods;
characterizing the degree and extent of local air toxics problems; and tracking progress of air
toxics reduction activities (EPA, 2009a).
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 2014. 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 contract with ERG for a special air toxics
1-2
-------
monitoring study in which their data are included in the report as well. The purpose of this report
is to summarize those data generated by the contract laboratory over the 2014 sampling effort.
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, 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 51 monitoring sites around the country. The
51 sites whose data are included in this report are located in or near 29 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 characterizations for each of the 51 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 present public health concerns, to
identify which emissions sources contribute most to air pollution, or to forecast whether
proposed pollution control initiatives could significantly improve 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?
1-3
-------
• 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?
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
sampling. Although many types of 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 2014 UATMP, NATTS, and CSATAM monitoring data,
henceforth referred to as NMP data, the complete set of measured concentrations 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, 2016b).
This report is organized into 26 sections and 18 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 overview,
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 2014 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 the NMP (specifically, the UATMP,
NATTS, and CSATAM Programs).
2
The 2014 National Monitoring
Programs Network
This section provides an overview on the 2014 NMP
monitoring effort, including:
• Monitoring locations
• Pollutants selected for monitoring
• Sampling and analytical methods
• Sampling schedules
• Completeness of the air monitoring programs.
1-4
-------
Table 1-1. Organization of the 2014 National Monitoring Programs Report (Continued)
Report
Section
Section Title
Overview of Contents
3
Summary of the 2014 National
Monitoring Programs Data
Treatments and Methods
This section presents and discusses the data treatments applied
to the 2014 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 2014 National
Monitoring Programs Data
This section presents and discusses the results of the data
treatments from the 2014 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 CBS A (CELA), the Riverside-San Bernardino-
Ontario, CA CBS A (RUCA), and the San Jose-Sunnyvale-Santa
Clara, CA CBSA (SJJCA)
7
Sites in Colorado
Monitoring results for the sites in the Grand Junction. CO
CBSA (GPCO) and the Glenwood Springs, CO CBSA (BMCO,
BRCO, 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 CBSA (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, CCKY, LAKY, 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 CB SA (CHNJ, ELNJ, and NBNJ) 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
(OCOK and YUOK)
19
Site in Rhode Island
Monitoring results for the site in the Providence-Warwick, RI-
MA CBSA (PRRI)
1-5
-------
Table 1-1. Organization of the 2014 National Monitoring Programs Report (Continued)
Report
Section
Section Title
Overview of Contents
20
Site in Utah
Monitoring results for the site in the Ogden-Clearfield, UT
CBS A (BTUT)
21
Site in Vermont
Monitoring results for the site in the Burlington-South
Burlington VT CBS A (UNVT)
22
Site in Virginia
Monitoring results for the site in the Richmond, VA CBS A
(RIVA)
23
Site in Washington
Monitoring results for the site in the Seattle-Tacoma-Bellevue,
WA CBS A (SEW A)
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 2014 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.
1-6
-------
2.0 The 2014 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. Data from
51 monitoring sites that collected 24-hour integrated ambient air samples for up to 12 months, at
l-in-6 or l-in-12 day sampling intervals, and sent them to ERG for analysis are included in this
report. 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-13A,
• trace metals from filters using EPA
Compendium Method IO-3.5/Federal Equivalent
EQL-0512-202, and
• hexavalent chromium from sodium bicarbonate-coated filters using ASTM D7614.
Section 2.2 provides additional information regarding each of the sampling methodologies used
to collect and analyze samples.
The following sections review the monitoring locations, pollutants selected for
monitoring, sampling and analytical methods, collection schedules, and completeness of the
2014 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
programs select the monitoring locations based on specific siting criteria and study needs.
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.
Methods (FEM) EQL-0512-201 or
2-1
-------
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 51 monitoring sites participating in the 2014
programs under the national contract, which encompass 29 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, 2013a). 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, 2013b). Table 2-1 lists the
respective monitoring program and the years of program participation under the national contract
for the 51 monitoring sites. Each of the 51 monitoring sites has been included in at least one
previous NMP annual report.
As Figure 2-1 and Table 2-1 show, the 2014 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 in order to distinguish this program from the other programs. Table 2-2
shows that the locations of the monitoring sites vary significantly, depending on the individual
program 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 2014 National Monitoring Programs Monitoring Sites1
K)
u>
Underbillj VT
[Boston, MA
Rochester, NY
Bountiful; UT
/ rRifle CO, I
/Parachute, CO / silt
Batt,e/entMesa,CO^arCb°ndale
/ Grand Junction, CO
^ Roxana, IL^ 1
St. Louis, MO'? Baskett, KY
Lexington, KY
Tulsa, OK (3)
Y^kon, OK -.1 \r
Los Angeles, CA
Calvert City, KY (5)'
Rubidoux, CA
Phoenix, AZ (2)
Oklahoma City, OK-
Winter Park, FL
>- Orlando, FL
Valrico, FL
Legend
Pinellas Park, FL
St. Petersburg, FL
Program
O NATTS
o UATMP
~1 CBSA
1 Includes monitoring sites participating under the NMP with the national contract laboratory
-------
Table 2-1. 2014 National Monitoring Programs Sites and Past Program Participation1
Monitoring Location
and Site Name
Program
2003 and Earlier
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
Ashland. KY (ASKY)
UATMP
~
~
Ashland, KY (ASKY-M)
UATMP
~
~
Baskett, KY (BAKY)
UATMP
~
~
Battlement Mesa, CO (BMCO)
UATMP
~
~
~
~
Boston MA (BOMA)
NATTS
2003
~
~
~
~
~
~
~
~
~
~
Bountiful, UT (BTUT)
NATTS
2003
~
~
~
~
~
~
~
~
~
~
Calvert City, KY (ATKY)
UATMP
~
~
Calvert City, KY (CCKY)
UATMP
~
~
Calvert City, KY (LAKY)
UATMP
~
~
Calvert City, KY (TVKY)
UATMP
~
~
Camden, NJ (CSNJ)
UATMP
~
Carbondale, CO (RFCO)
UATMP
~
~
Chester, NJ (CHNJ)
UATMP
2001-2003
~
~
~
~
~
~
~
~
~
~
Dearborn MI (DEMI)
NATTS
2001-2003
~
~
~
~
~
~
~
~
~
~
East Highland Park, VA (RIVA)
NATTS
~
~
~
~
~
~
Elizabeth, NJ (ELNJ)
UATMP
1999-2003
~
~
~
~
~
~
~
~
~
~
Gary, IN (INDEM)
UATMP
~
~
~
~
~
~
~
~
~
~
BOLD ITALICS = EPA-designated NATTS site
1 Includes monitoring sites participating under the NMP with the national contract laboratory
-------
Table 2-1. 2014 National Monitoring Programs Sites and Past Program Participation1 (Continued)
Monitoring Location
and Site Name
Program
2003 and Earlier
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
Grand Junction CO (GPCO)
NATTS
~
~
~
~
~
~
~
~
~
~
Grayson KY (GLKY)
NATTS
~
~
~
~
~
~
Indianapolis, IN (WPIN)
UATMP
~
~
~
~
~
~
~
~
Lexington, KY (LEKY)
UATMP
~
~
Los Angeles, CA (CELA)
NATTS
~
~
~
~
~
~
~
New York, NY (BXNY)
NATTS
~
~
~
~
~
~
~
North Brunswick, NJ (NBNJ)
UATMP
2001-2003
~
~
~
~
~
~
~
~
~
~
Northbrook, IL (NBIL)
NATTS
2003
~
~
~
~
~
~
~
~
~
~
Oklahoma City, OK (OCOK)
UATMP
~
~
~
~
~
Orlando, FL (PAFL)
UATMP
~
~
~
~
~
~
Parachute, CO (PACO)
UATMP
~
~
~
~
~
~
Phoenix, AZ (PXSS)
NATTS
2001-2003
~
~
~
~
~
~
~
~
~
Phoenix, AZ (SPAZ)
UATMP
2001
~
~
~
~
~
~
~
Pinellas Park, FL (SKFL)
NATTS
~
~
~
~
~
~
~
~
~
~
Providence, RI (PRRI)
NATTS
~
~
~
~
~
~
~
~
~
Rifle, CO (RICO)
UATMP
~
~
~
~
~
~
Rochester, NY (ROCH)
NATTS
~
~
~
~
~
~
~
~
BOLD ITALICS = EPA-designated NATTS site
1 Includes monitoring sites participating under the NMP with the national contract laboratory
-------
Table 2-1. 2014 National Monitoring Programs Sites and Past Program Participation1 (Continued)
Monitoring Location
and Site Name
Program
2003 and Earlier
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
Roxana, IL (ROIL)
Special
Study
~
~
Rubidoux, CA (RUCA)
NATTS
~
~
~
~
~
~
~
San Jose, CA (SJJCA)
NATTS
~
~
~
~
~
~
Schiller Park, IL (SPIL)
UATMP
2003
~
~
~
~
~
~
~
~
~
~
Seattle, WA (SEWA)
NATTS
~
~
~
~
~
~
~
~
~
Silt, CO (BRCO)
UATMP
~
~
~
~
~
~
Sinitliland, KY (BLKY)
UATMP
~
~
St. Louis, MO (S-IMO)
NATTS
2002, 2003
~
~
~
~
~
~
~
~
~
~
St. Petersburg, FL (AZFL)
UATMP
1991-1992, 2001-
2003
~
~
~
~
~
~
~
~
~
~
Tulsa, OK (TMOK)
UATMP
~
~
~
~
~
Tulsa, OK (TOOK)
UATMP
~
~
~
~
~
~
~
~
Tulsa, OK (TROK)
UATMP
~
Underhill, VT (UNVT)
NATTS
2002
~
~
~
~
~
~
~
~
~
Valrico, FL (SYFL)
NATTS
~
~
~
~
~
~
~
~
~
~
Wasliington, D.C. (WADC)
NATTS
~
~
~
~
~
~
~
~
~
Winter Park, FL (ORFL)
UATMP
1990-1991,2003
~
~
~
~
~
~
~
~
~
~
Yukon OK (YUOK)
UATMP
~
BOLD ITALICS = EPA-designated NATTS site
1 Includes monitoring sites participating under the NMP with the national contract laboratory
-------
Table 2-2. Site Characterizing Information for the 2014 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)
262.71
172.53
ASKY-M
21-019-0002
Ashland, KY
Industrial
Urban/City Center
12,842
(2012)
262.71
172.53
ATKY
21-157-0016
Calvert City, KY
Industrial
Suburban
3,262
(2012)
1,119.74
476.37
AZFL
12-103-0018
St. Petersburg, FL
Residential
Suburban
40,000
(2014)
2,132.17
3,217.48
BAKY
21-101-0014
Baskett, KY
Commercial
Rural
922
(2012)
397.98
268.40
BLKY
21-139-0004
Smithland, KY
Agricultural
Rural
2,510
(2013)
32.24
136.07
BMCO
08-045-0019
Battlement Mesa, CO
Commercial
Suburban
1,880
(2014)
3,787.70
327.61
BOMA
25-025-0042
Boston, MA
Commercial
Urban/City Center
27,654
(2010)
851.81
1,015.72
BRCO
08-045-0009
Silt, CO
Agricultural
Rural
1,182
(2014)
3,787.70
327.61
BTUT
49-011-0004
Bountiful, UT
Residential
Suburban
130,950
(2013)
1,163.85
930.74
BXNY
36-005-0110
New York, NY
Residential
Urban/City Center
98,298
(2013)
3,796.74
840.39
CCKY
21-157-0018
Calvert City, KY
Residential
Suburban
4,050
(2013)
1,119.74
476.37
CELA
06-037-1103
Los Angeles, CA
Residential
Urban/City Center
230,000
(2014)
21,804.55
14,773.30
CHNJ
34-027-3001
Chester, NJ
Agricultural
Rural
11,215
(2012)
680.93
1,278.46
CSNJ
34-007-0002
Camden NJ
Industrial
Urban/City Center
3,231
(2012)
577.27
953.66
BOLD ITALICS = EPA-designated NATTS site
individual references provided in each state section.
bReference: 2011 NEI version2 (EPA, 2015a)
°GPCO's metals sampler is at a separate, but adjacent, location; thus, this site lias two AQS codes.
d S4MO's county-level emissions are city-level data.
-------
Table 2-2. Site Characterizing Information for the 2014 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
96,870
(2014)
7,118.74
4,563.35
ELNJ
34-039-0004
Elizabeth, NJ
Industrial
Suburban
250,000
(2006)
814.19
1,017.46
GLKY
21-043-0500
Grayson KY
Residential
Rural
303
(2012)
75.96
145.24
GPCOc
08-077-0017
08-077-0018
Grand Junction CO
Commercial
Urban/City Center
12,000
(2014)
659.65
664.73
INDEM
18-089-0022
Gary, IN
Industrial
Urban/City Center
34,754
(2011)
1,603.10
1,607.33
LAKY
21-157-0019
Calvert City, KY
Residential
Suburban
1,189
(2012)
1,119.74
476.37
LEKY
21-067-0012
Lexington, KY
Residential
Suburban
18,993
(2014)
764.77
1,116.04
NBIL
17-031-4201
Northbrook, IL
Residential
Suburban
115,100
(2014)
15,663.06
8,882.46
NBNJ
34-023-0006
North Brunswick, NJ
Agricultural
Rural
114,322
(2010)
1,038.26
1,577.17
OCOK
40-109-1037
Oklahoma City, OK
Residential
Suburban
52,400
(2014)
2,156.08
3,425.17
ORFL
12-095-2002
Winter Park, FL
Commercial
Urban/City Center
31,500
(2014)
2,774.25
4,121.46
PACO
08-045-0005
Parachute, CO
Residential
Urban/City Center
16,000
(2014)
3,787.70
327.61
PAFL
12-095-1004
Orlando, FL
Commercial
Suburban
49,000
(2013)
2,774.25
4,121.46
PRRI
44-007-0022
Providence, RI
Residential
Urban/City Center
136,800
(2009)
1,362.28
1,350.29
PXSS
04-013-9997
Phoenix, AZ
Residential
Urban/City Center
35,103
(2010)
7,792.15
9,915.84
BOLD ITALICS = EPA-designated NATTS site
individual references provided in each state section.
bReference: 2011 NEI version2 (EPA, 2015a)
°GPCO's metals sampler is at a separate, but adjacent, location; thus, this site lias two AQS codes.
d S4MO's county-level emissions are city-level data.
-------
Table 2-2. Site Characterizing Information for the 2014 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)
RFCO
08-045-0018
Carbondale, CO
Residential
Rural
16,000
(2014)
3,787.70
327.61
RICO
08-045-0007
Rifle, CO
Commercial
Urban/City Center
17,000
(2014)
3,787.70
327.61
RIVA
51-087-0014
East Highland Park, VA
Residential
Suburban
72,000
(2013)
888.54
746.37
ROCH
36-055-1007
Rochester, NY
Residential
Urban/City Center
85,417
(2013)
2,959.44
1,742.27
ROIL
17-119-9010
Roxana, IL
Industrial
Suburban
7,750
(2013)
1,359.86
815.08
RUCA
06-065-8001
Rubidoux, CA
Residential
Suburban
158,000
(2014)
3,826.19
3,244.32
S4MO
29-510-0085
St. Louis, MO
Residential
Urban/City Center
100,179
(2013)
939.84d
611.09d
SEWA
53-033-0080
Seattle, WA
Residential
Urban/City Center
178,000
(2014)
7,310.24
6,890.17
SJJCA
06-085-0005
San Jose, CA
Commercial
Urban/City Center
124,000
(2014)
4,177.14
3,634.86
SKFL
12-103-0026
Pinellas Park, FL
Residential
Suburban
36,500
(2014)
2,132.17
3,217.48
SPAZ
04-013-4003
Phoenix, AZ
Residential
Urban/City Center
25,952
(2011)
7,792.15
9,915.84
SPIL
17-031-3103
Schiller Park, IL
Mobile
Suburban
193,800
(2013)
15,663.06
8,882.46
SYFL
12-057-3002
Valrico, FL
Residential
Rural
3,800
(2014)
3,155.70
4,260.15
TMOK
40-143-1127
Tulsa, OK
Residential
Urban/City Center
4,200
(2014)
1,902.81
4,149.89
TOOK
40-143-0235
Tulsa, OK
Industrial
Urban/City Center
65,800
(2014)
1,902.81
4,149.89
BOLD ITALICS = EPA-designated NATTS site
individual references provided in each state section.
bReference: 2011 NEI version2 (EPA, 2015a)
°GPCO's metals sampler is at a separate, but adjacent, location; thus, this site lias two AQS codes.
d S4MO's county-level emissions are city-level data.
-------
Table 2-2. Site Characterizing Information for the 2014 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)
TROK
40-143-0179
Tulsa, OK
Industrial
Urban/City Center
53,300
(2014)
1,902.81
4,149.89
TVKY
21-157-0014
Calvert City, KY
Industrial
Suburban
1,458
(2014)
1,119.74
476.37
UNVT
50-007-0007
Underhill, VT
Forest
Rural
1,100
(2011)
432.40
477.55
WADC
11-001-0043
Washington D.C.
Commercial
Urban/City Center
8,700
(2013)
933.45
829.76
WPIN
18-097-0078
Indianapolis, IN
Residential
Suburban
24,661
(2011)
2,627.90
4,042.65
YUOK
40-017-0101
Yukon OK
Commercial
Suburban
41,000
(2014)
680.10
447.57
BOLD ITALICS = EPA-designated NATTS site
individual references provided in each state section.
bReference: 2011 NEI version2 (EPA, 2015a)
°GPCO's metals sampler is at a separate, but adjacent, location; thus, this site lias two AQS codes.
d S4MO's county-level emissions are city-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 2 of the 2011 National
Emissions Inventory (NEI). (The 2014 NEI was published near the end of the
production of the 2014 NMP report and will be utilized in the 2015 NMP report.)
This information is discussed in further detail in the individual state sections.
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 was can be performed concurrently with Method TO-15.
• Compendium Method TO-11A was used to measure ambient air concentrations of
15 carbonyl compounds.
• Compendium Method TO-13A was used to measure ambient air concentrations of
22 PAHs.
• A combination of Compendium Method 10-3.5 and EPA Federal Equivalency
Methods (FEM) EQL-0512-201 or EQL-0512-202 was used to measure ambient air
concentrations of 11 metals.
• ASTMMethodD7614 was used to measure ambient air concentrations of hexavalent
chromium.
2-11
-------
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 sampler 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 reliably quantify concentrations of selected pollutants to a specific
confidence level. If a pollutant's concentration in ambient air is below 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 below the MDL is possible, the measurement reliability is lower. Therefore, when
pollutants are present at concentrations below 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 below the corresponding detection limits.
MDLs are determined annually at the ERG laboratory using 40 CFR, Part 136
Appendix B procedures (EPA, 1986) in accordance with the specifications presented in the
NATTS Technical Assistance Document (TAD) (EPA, 2009b). 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.
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 procedure
2-12
-------
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.
Tables 2-3 through 2-8 identify the specific target pollutants for each analytical method
and their corresponding MDLs, as determined for 2014. For the VOC and SNMOC analyses, the
experimentally determined MDLs do not change within a given year unless the sample was
diluted. The 2014 VOC and SNMOC MDLs are presented in Tables 2-3 and 2-4, respectively.
For the rest of the analytical methods, the MDLs vary due to the actual volume pulled through
the sample or if the sample was diluted. For these analyses, the range and average MDL is
presented for each pollutant in Tables 2-5 through 2-8, based on valid samples. If the MDLs
presented in Tables 2-5 through 2-8 include an MDL for a diluted sample, the MDL may appear
elevated. Dilutions cause the MDL to increase by a factor of the dilution; MDLs affected by
dilution are denoted in the tables. ERG's published pollutant-specific MDLs are also presented in
Appendix B.
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
2012a; ASTM, 2012; ASTM, 2013).
2.2.1 VOC and SNMOC Concurrent Sampling and Analytical Methods
VOC and SNMOC sampling and analysis can be performed concurrently using a
combined 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. When referring to SNMOC analysis, this report
may refer to this method as the "concurrent SNMOC method" or "concurrent SNMOC analysis"
because both methods can be employed at the same time to analyze the same sample. 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
2-13
-------
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 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 m-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 C and D.
Table 2-3 presents the MDLs for the laboratory analysis of VOC samples with
Method TO-15 and Table 2-4 presents the MDLs for the analysis of SNMOC samples. The MDL
for every VOC is less than or equal to 0.095 parts per billion by volume (ppbv). SNMOC
detection limits are expressed in parts per billion Carbon (ppbC). All SNMOC MDLs are less
than or equal to 1.58 ppbC.
2-14
-------
Table 2-3. 2014 VOC Method Detection Limits
Pollutant
2014
MDL
(ppbv)
Pollutant
2014
MDL
(ppbv)
Acetonitrile
0.045
Dichloromethane
0.014
Acetylene
0.013
1,2-Dichloropropane
0.015
Acrolein
0.094
cis-1,3 -Dichloropropene
0.018
Acrylonitrile
0.016
trans-1,3 -Dichloropropene
0.020
fcrt-Amyl Methyl Ether
0.011
Dichlorotetrafluoroethane
0.013
Benzene
0.015
Ethyl Acrylate
0.018
Bromochloro methane
0.015
Ethyl tort-Butyl Ether
0.016
Bromodichloromethane
0.017
Ethylbenzene
0.013
Bromoform
0.013
Hexachloro-1,3 -Butadiene
0.027
Bromomethane
0.012
Methyl Isobutyl Ketone
0.019
1.3 -Butadiene
0.013
Methyl Methacrylate
0.018
Carbon Disulfide
0.013
Methyl tert-Butyl Ether
0.018
Carbon Tetrachloride
0.017
n-Octane
0.012
Chlorobenzene
0.015
Propylene
0.030
Chloroethane
0.015
Styrene
0.014
Cliloroform
0.015
1,1,2,2-Tetrachloroethane
0.014
Cliloromethane
0.014
T etracliloroethylene
0.013
Cliloroprene
0.016
Toluene
0.013
Dibromochloro methane
0.014
1,2,4-Trichlorobenzene
0.050
1,2-Dibromoetliane
0.013
1.1.1 -T ricliloroetliane
0.015
«/-Dichlorobenzene
0.014
1,1,2-Trichloroethane
0.018
o-Dichlorobenzene
0.015
T ricliloroethylene
0.017
/?-Dichlorobcnzcnc
0.015
T richlorofluoro methane
0.015
Diclilorodifluorometliane
0.014
T riclilorotrifluoroetliane
0.016
1,1 -Dichloroethane
0.018
1,2,4-Trimethylbenzene
0.018
1,2-Dichloroethane
0.014
1.3.5 -T rimethy lbenzene
0.017
1,1 -Dichloroethene
0.014
Vinyl Chloride
0.012
cis-1,2 -Dichloroethy lene
0.014
w/./j-Xvlcne1
0.023
trans-1,2-Dichloroethylene
0.012
o-Xylene
0.012
1 Because w;-xylene and /^-xylene elute from the GC column at the same time, the
VOC analytical method reports the sum of -xylene and /^-xylene concentrations
and not concentrations of the individual isomers.
2-15
-------
Table 2-4. 2014 SNMOC Method Detection Limits
2014
2014
2014
MDL
MDL
MDL
Pollutant
(ppbC)1
Pollutant
(ppbC)1
Pollutant
(ppbC)1
Acetylene
0.12
1-Heptene
0.26
1-Pentene
0.16
Benzene
0.23
n-Hexane
0.35
6/.v-2-Pentcne
0.13
1.3 -Butadiene
0.19
1-Hexene
0.25
trans-2 -Pentene
0.13
//-Butane
0.25
67.Y-2-Hc\cnc
0.21
«-Pincnc
0.27
1-Butene
0.35
tr <7/75-2 -Hexene
0.23
/>-Pi nc nc
0.82
67.Y-2-Butcnc
0.08
Isobutane
0.15
Propane
0.57
;ra«.v-2-B lite nc
0.09
Isobutylene
0.15
/7-Propylbenzene
0.29
Cyclohexane
0.23
Isopentane
0.17
Propylene
0.11
Cyclopentane
0.15
Isoprene
0.14
Propyne
0.06
Cyclopentene
0.88
Isopropylbenzene
0.20
Styrene
0.36
w-Decane
0.39
2-Methyl-1 -Butene
0.13
Toluene
0.30
1-Decene
1.07
3 -Methyl-1 -Butene
0.18
/7-Tridecane
0.65
///-Diethylbenzene
0.67
2-Methyl-1 -Pentene
0.21
1-Tridecene
0.91
/j-Dictlivlbcnzcne
0.43
4-Methyl-1 -Pentene
0.15
1,2,3 -Trimethylbenzene
0.23
2,2 -Dimethy lbutane
0.21
2 -Methy 1-2 -B utene
0.20
1,2,4-Trimethylbenzene
1.07
2,3-Dimethylbutane
0.18
Methylcyclohexane
0.24
1,3,5 -T rimethylbenzene
0.37
2,3 -Dimethy lpentane
0.25
Methylcyclopentane
0.19
2,2,3 -T rimethy lpentane
0.29
2,4 -Dimethy lpentane
0.25
2-Methylheptane
0.36
2,2,4 -T rimethy lpentane
0.28
n-Dodecane
0.91
3-Methylheptane
0.33
2,3,4 -T rimethy lpentane
0.28
1-Dodecene
1.57
2-Methylhexane
0.23
n-Undecane
0.75
Ethane
1.52
3-Methylhexane
0.37
1-Undecene
1.20
2-Ethyl-l-butene
0.34
2-Methylpentane
0.36
/w-Xylene/p-Xylene2
0.45
Ethylbenzene
0.24
3-Methy lpentane
0.16
o-Xylene
0.23
Ethylene
0.21
n-Nonane
0.26
Sum of Knowns
NA
///-Ethyltoluene
0.38
1-Nonene
0.31
Sum of Unknowns
NA
o-Ethyltoluene
0.38
n-Octane
0.36
TNMOC
NA
p-Ethyltoluene
0.28
1-Octene
0.33
//-Heptane
0.24
n-Pentane
0.18
1 Concentration in ppbC = concentration in ppbv * number of carbon atoms in the compound.
2 Because ///-xylene and /^-xylene elute from the GC column at the same time, the SNMOC analytical method
reports the sum concentration for these two isomers and not concentrations of the individual isomers.
NA = Not applicable
2-16
-------
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 elute from the HPLC column at the same time, the
carbonyl compound analytical method reports only the sum concentration for these isomers and
not the separate concentrations for each isomer. Raw data for Method TO-11A are presented in
Appendix E.
Table 2-5 lists the MDLs reported by the ERG laboratory for measuring concentrations of
15 carbonyl compounds. Although the sensitivity varies from pollutant-to-pollutant and sample-
to-sample due to different volumes pulled through the samples, the average detection limit for
valid samples reported by the ERG laboratory for every carbonyl compound is less than or equal
to 0.016 ppbv.
2-17
-------
Table 2-5. 2014 Carbonyl Compound Method Detection Limits
Pollutant
Minimum
MDL
(ppbv)
Maximum
MDL2
(ppbv)
Average
MDL
(ppbv)
Acetaldehyde
0.003
0.014
0.008
Acetone
0.010
0.0402
0.016
Benzaldehyde
0.001
0.006
0.003
2-Butanone
0.001
0.006
0.002
Butyraldehyde
0.001
0.006
0.003
Crotonaldehyde
0.003
0.011
0.004
2,5-Dimethylbenzaldehyde
0.001
0.006
0.002
Formaldehyde
0.007
0.1792
0.013
Hexaldehyde
0.001
0.003
0.002
Isovaleraldehyde
0.001
0.006
0.002
Propionaldehyde
0.001
0.006
0.004
Tolualdehydes1
0.002
0.009
0.004
Valeraldehyde
0.001
0.006
0.003
1 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 Indicates that sample dilution was required to perform analysis.
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 are installed in
a high volume sampler by the site operators and allowed to sample for 24 hours. 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 2-6 lists the MDLs for the 22 PAH target pollutants. PAH detection limits are
expressed in nanograms per cubic meter (ng/m3). Although the sensitivity varies from pollutant-
to-pollutant and from sample-to-sample due to the different volumes pulled through the samples,
2-18
-------
the average detection limit for valid samples reported by the ERG laboratory for every PAH is
less than 0.485 ng/m3.
Table 2-6. 2014 PAH Method Detection Limits
Pollutant
Minimum
MDL
(ng/m3)
Maximum
MDL1
(ng/m3)
Average
MDL
(ng/m3)
Acenaphthene
0.068
0.414
0.128
Acenaphthylene
0.010
0.063
0.019
Anthracene
0.033
0.202
0.062
Benzo(a)anthracene
0.014
0.083
0.026
Benzo(a)pyrene
0.020
0.120
0.037
Benzo(b)fluoranthene
0.019
0.116
0.036
Benzo(e)pyrene
0.019
0.115
0.035
Benzo(g,h,i)perylene
0.018
0.107
0.033
Benzo(k)fluoranthene
0.021
0.128
0.039
Chrysene
0.017
0.101
0.031
Coronene
0.017
0.105
0.032
Cyclopenta[cd]pyrene
0.025
0.148
0.046
Dibenz(a,h)anthracene
0.018
0.111
0.034
Fluoranthene
0.017
0.105
0.032
Fluorene
0.053
0.321
0.099
9-Fluorenone
0.018
0.111
0.034
Indeno( 1,2,3 -cd)pyrene
0.020
0.121
0.037
Naphthalene
0.255
4.410
0.484
Perylene
0.024
0.145
0.045
Phenanthrene
0.044
0.263
0.081
Pyrene
0.017
0.102
0.032
Retene
0.013
0.081
0.025
1 A single sample required a dilution for all pollutants listed.
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
samplers, whereas the 8" x 10" quartz filter is used for high-volume samplers. 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 returned the filters,
along with the COC forms and all associated documentation, to the ERG laboratory for analysis.
2-19
-------
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, 2012a). Upon receipt at the
laboratory, the whole filters (47mm Teflon®) or filter strips (8" x 10" quartz) were digested using
a dilute nitric acid, hydrochloric acid, and/or hydrofluoric acid (Teflon® only) solution. The
digestate was then quantified using Inductively Coupled Plasma/Mass Spectrometry (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 2-7 lists the MDLs for the analysis of metals samples. Due to the difference in
sample volume/filter collection media, there are two sets of MDLs listed in Table 2-7, one for
each filter type. Although the sensitivity varies from pollutant-to-pollutant and from sample-to-
sample due to the different volumes pulled through the samples, the average detection limit for
valid samples reported by the ERG laboratory for every metal is less than 2.25 ng/m3 for the
quartz filters and less than or equal to 14.0 ng/m3 for the Teflon® filters.
Table 2-7. 2014 Metals Method Detection Limits
Pollutant
Minimum
MDL
(ng/m3)
Maximum
MDL
(ng/m3)
Average
MDL
(ng/m3)
Pollutant
Minimum
MDL
(ng/m3)
Maximum
MDL
(ng/m3)
Average
MDL
(ng/m3)
8" X 10" Quartz Filters
47mm Teflon™ Filters
Antimony
0.009
0.015
0.012
Antimony
0.090
0.120
0.101
Arsenic
0.042
0.074
0.057
Arsenic
0.190
0.260
0.221
Beryllium
0.001
0.003
0.002
Beryllium
0.010
0.020
0.020
Cadmium
0.003
0.005
0.004
Cadmium
0.010
0.010
0.010
Chromium
1.65
2.91
2.24
Chromium
12.1
16.5
14.0
Cobalt
0.017
0.029
0.022
Cobalt
0.010
0.020
0.020
Lead
0.054
0.9331
0.076
Lead
0.030
0.040
0.040
Manganese
0.144
0.253
0.194
Manganese
0.120
0.170
0.141
Mercury
0.004
0.0621
0.005
Mercury
0.030
0.040
0.031
Nickel
0.515
0.907
0.696
Nickel
0.160
0.220
0.181
Selenium
0.019
0.033
0.025
Selenium
0.250
0.350
0.292
indicates that sample dilution was required to perform analysis.
2-20
-------
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 sampler 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 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.
Although the sensitivity varies due to the different volumes pulled through the samples,
Table 2-8 presents the range and average detection limit (0.0035 ng/m3) for valid hexavalent
chromium samples reported by the ERG laboratory across the program.
Table 2-8. 2014 Hexavalent Chromium Method Detection Limit
Pollutant
Minimum
MDL
(ng/m3)
Maximum
MDL
(ng/m3)
Average
MDL
(ng/m3)
Hexavalent Chromium
0.0031
0.0036
0.0035
2.3 Sample Collection Schedules
Table 2-9 presents the first and last date upon which sample collection occurred for each
monitoring site sampling under the NMP in 2014. The first sample date for each site is generally
at the beginning of January and sampling continued through the end of December, although there
was an exception. Sampling at CCKY was discontinued at the beginning of October 2014. The
metals sampler at CCKY was redeployed at BLKY, where metals sampling reconvened at the
end of October 2014.
2-21
-------
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:
• 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, PACO, and RICO.
Sampling at RFCO was conducted on a l-in-12 day schedule for both methods.
• 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.
Table 2-9 shows the following:
• 27 sites collected VOC samples.
• 32 sites collected carbonyl compound samples.
• 7 sites collected SNMOC samples.
• 19 sites collected PAH samples.
• 21 sites collected metals samples.
• 2 sites collected hexavalent chromium samples.
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) samplers were available. Field blanks were collected once a 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 locations periodically strayed from the l-in-6 or l-in-12 day sampling schedule.
2-22
-------
Table 2-9. 2014 Sampling Schedules and Completeness Rates
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
1/5/14
12/31/14
61
61
100
61
61
100
ASKY-M
1/5/14
12/31/14
59
61
97
ATKY
1/7/14
12/31/14
61
61
100
AZFL
1/5/14
12/31/14
56
61
92
BAKY
1/5/14
12/31/14
58
61
95
BLKY
1/7/14
12/31/14
60
61
98
12
13
92
BMCO
1/5/14
12/31/14
27
31
872
51
61
84
BOMA
1/5/14
12/31/14
58
61
95
57
61
93
BRCO
1/5/14
12/31/14
25
31
812
50
51
82
BTUT
1/5/14
12/31/14
58
61
95
55
61
90
57
61
93
55
61
90
58
61
95
BXNY
1/5/14
12/31/14
57
61
93
CCKY
1/5/14
10/2/14
46
46
100
41
46
89
CELA
1/5/14
12/31/14
56
61
92
CHNJ
1/11/14
12/31/14
60
61
98
61
61
100
CSNJ
1/8/14
12/31/14
60
61
98
61
61
100
__
__
__
__
__
__
__
__
__
__
__
__
A = Number of valid samples collected.
B = Number of valid samples that should be collected in 2014 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.
BOLD ITALICS = EPA-designated NATTS site.
Orange shading indicates that completeness is below the MQO of 85 percent.
-------
Table 2-9. 2014 Sampling Schedules and Completeness Rates (Continued)
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
DEMI
1/5/14
12/31/14
61
61
100
60
61
98
60
61
98
ELNJ
1/8/14
12/31/14
61
61
100
59
61
97
GLKY
1/5/14
12/31/14
61
61
100
56
61
92
59
61
97
58
61
95
GPCO
1/5/14
12/31/14
58
61
95
57
61
93
59
61
97
60
61
98
INDEM
1/5/14
12/31/14
57
61
93
LAKY
1/7/14
12/31/14
56
61
92
LEKY
1/5/14
12/31/14
55
61
90
58
61
95
56
61
92
NBIL
1/5/14
12/31/14
55
61
90
56
61
92
53
61
87
56
61
92
55
61
90
NBNJ
1/8/14
12/31/14
20
61
33
60
61
98
OCOK
1/5/14
12/31/14
60
61
98
60
61
98
59
61
97
ORFL
1/5/14
12/31/14
60
61
98
PACO
1/5/14
12/31/14
25
31
812
57
61
93
PAFL2
1/5/14
12/31/14
30
31
97
PRRI
1/5/14
12/31/14
58
61
95
PXSS
1/5/14
12/31/14
61
61
100
61
61
100
__
__
__
60
61
98
__
__
__
59
61
97
A = Number of valid samples collected.
B = Number of valid samples that should be collected in 2014 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.
BOLD ITALICS = EPA-designated NATTS site.
Orange shading indicates that completeness is below the MQO of 85 percent.
-------
Table 2-9. 2014 Sampling Schedules and Completeness Rates (Continued)
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
RFCO2
1/5/14
12/31/14
26
31
84
28
31
90
RICO
1/5/14
12/31/14
27
31
872
54
61
89
RIVA
1/5/14
12/31/14
61
61
100
57
61
93
ROCH
1/5/14
12/31/14
57
61
93
ROIL
1/5/14
12/31/14
60
61
98
58
61
95
RUCA
1/5/14
12/31/14
59
61
97
S4MO
1/5/14
12/31/14
60
61
98
59
61
97
31
31
100
61
61
100
57
61
93
SEWA
1/5/14
12/31/14
61
61
100
60
61
98
60
61
98
61
61
100
SJJCA
1/5/14
12/31/14
61
61
100
59
61
97
SKFL
1/5/14
12/31/14
50
61
82
58
61
95
SPAZ2
1/11/14
12/25/14
30
30
100
SPIL
1/5/14
12/31/14
59
61
97
55
61
90
SYFL
1/5/14
12/25/14
55
61
90
TMOK
1/5/14
12/31/14
62
61
>100
62
61
>100
58
61
95
TOOK
1/5/14
12/31/14
61
61
100
61
61
100
__
__
__
62
61
102
__
__
__
__
__
__
A = Number of valid samples collected.
B = Number of valid samples that should be collected in 2014 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.
BOLD ITALICS = EPA-designated NATTS site.
Orange shading indicates that completeness is below the MQO of 85 percent.
-------
Table 2-9. 2014 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
1/5/14
12/31/14
61
61
100
61
61
100
59
61
97
TVKY
1/7/14
12/31/14
61
61
100
UNVT
1/5/14
12/31/14
61
61
100
59
61
97
WADC
1/5/14
12/31/14
60
61
98
WPIN
1/5/14
12/31/14
54
61
89
YUOK
1/5/14
12/31/14
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 2014 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.
BOLD ITALICS = EPA-designated NATTS site.
Orange shading indicates that completeness is below the MQO of 85 percent.
-------
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 efficiency with which the program is managed.
The completeness for each monitoring site and method sampled is presented in Table 2-9.
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, 2013). The data in Table 2-9 show that seven datasets from a total of 108
datasets from the 2014 NMP monitoring effort did not meet this MQO (orange shaded cells in
Table 2-9):
• BMCO - SNMOC
• BRCO - carbonyl compounds and SNMOC
• NBNJ - carbonyl compounds
• PACO - carbonyl compounds
• RFCO - carbonyl compounds
• SKFL - carbonyl compounds
The percent completeness for each of these datasets is just less than the MQO of 85 percent
(between 80 percent and 85 percent for each), with the exception of NBNJ.
2-27
-------
Appendix I identifies samples that were invalidated and lists the reason for invalidation,
based on the applied AQS null code. A defective sampler was discovered at SKFL and resulted
in the invalidation of carbonyl compound samples collected between July 22, 2014 and
September 20, 2014. A new sampler was installed on September 26, 2014 and sampling resumed
on September 27, 2014. Similarly, carbonyl compound results were invalidated between
May 5, 2014 and December 31, 2014 for NBNJ, after which a new sampler was installed.
Also of note, chromium and nickel concentrations measured at UNVT after July 4, 2014
were invalidated after the Vermont Department of Environmental Conservation determined that
the filters were contaminated by a new weighing and equilibration chamber at their laboratory.
As this affected only two of the 11 metals for which measurements were collected at this site,
this invalidation is not reflected in Table 2-9.
Table 2-10 presents method-specific completeness. Method-specific completeness was
greater than 85 percent for all methods performed under the 2014 NMP and ranged from
88 percent for SNMOCs to 100 percent for hexavalent chromium.
Table 2-10. Method Completeness Rates for 2014
Method
# of
Valid
Samples
# of
Samples
Scheduled
Method
Completeness
(%)
Minimum
Site-Specific
Completeness
(%)
Maximum
Site-Specific
Completeness
(%)
VOCs
1,556
1,601
97.2
90
(2 sites)
>100
(TMOK)
SNMOCs
351
397
88.4
82
(BRCO)
93
(PACO)
Carbonyl Compounds
1,678
1,802
93.1
33
(NBNJ)
>100
(TMOK)
PAHs
1,105
1,159
95.3
90
(.NBIL)
100
(SEWA)
Metals Analysis
1,144
1,188
96.3
87
(NBIL)
>100
(2 sites)
Hexavalent Chromium1
92
92
100
100
(RIVA and S4MO)
BOLD ITALICS = EPA-designated NATTS site.
1 Hexavalent chromium was sampled for at only two sites in 2014.
2-28
-------
3.0 Summary of the 2014 National Monitoring Programs Data Treatment and Methods
for Data Analysis
This section summarizes the data treatment
employed and approaches used to analyze the data
generated from samples collected during the 2014
NMP sampling year. These data were analyzed on
a program-wide basis as well as a site-specific
basis.
A total of 224,685 valid air toxics concentrations (including non-detects, duplicate
analyses, replicate analyses, and analyses for collocated samples) were produced from 7,819
valid samples collected at 51 monitoring sites during the 2014 reporting year. A tabular
presentation of the raw data and statistical summaries are found in Appendices C through O, as
presented in Table 3-1. Appendix P serves as the glossary for the NMP report and many of the
terms discussed and defined throughout the report are provided there.
Table 3-1. Overview and Layout of Data Presented
Pollutant Group
Number
of Sites
A
ppendix
Raw Data
Statistical Summary
VOCs
27
C
J
SNMOCs
7
D
K
Carbonyl Compounds
32
E
L
PAHs
19
F
M
Metals
21
G
N
Hexavalent Chromium
2
H
O
3.1 Approach to Data Treatment
This section examines the various statistical tools employed to characterize and analyze
the data collected during the 2014 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 concentration
data were treated. The following paragraphs describe techniques used to prepare this large
quantity of concentration data for data analysis.
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.
3-1
-------
For each monitoring site, the primary, duplicate (or collocated), and replicate
measurements were averaged together for each pollutant in order to calculate a single
concentration per sample date and method. This is referred to as thepreprocesseddaily
measurement.
Concentrations of m,p-x.ylene and o-xylene 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
statistics for VOCs and SNMOCs, respectively, present the xylenes results retained as
m,p-xylene and o-xylene species. Data for the isomers are also presented individually in the Data
Quality section (Section 24).
For the 2014 NMP, where statistical parameters are calculated based on the preprocessed
daily measurements, zeros have been substituted for non-detect results. This approach is
consistent with how data are loaded into AQS per the NATTS TAD (EPA, 2009b) as well as
other EPA air toxics monitoring programs, such as the School Air Toxics Monitoring Program
(SATMP) (EPA, 201 la), and other associated reports, including the NATTS Network
Assessment (EPA, 2012b). 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 4-6, 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.
3-2
-------
In addition, this report presents various time-based averages to summarize the
measurements for a specific site. Where applicable, quarterly and annual averages were
calculated for each site. The quarterly average concentration of a particular pollutant is simply
the average 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 in 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 (2014). 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.
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 above and below the concentration average (EPA, 201 la).
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 to
1.50 |ig/m3.
3-3
-------
3.2 Human Health Risk and the Pollutants of Interest
A practical approach to making an assessment on a large number of 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 2014 NMP
report. The following paragraphs provide an overview of health risk terms 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, 2015b). 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. 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, 2016c).
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 include
conditions such as asthma; noncancer health risks are presented as a hazard quotient, the value
below which no adverse health effects are expected (EPA, 2015b). Cancer risk is presented as a
probability while the hazard quotient is a ratio and thus, a unitless value.
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 air toxics 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
3-4
-------
adaption 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 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 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. The purpose of this data treatment is to have one concentration per
pollutant for each sample day.
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 nickel in
Table 4-7 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.
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, 2015c, 2015d).
3-5
-------
A note regarding measurements of acetonitrile, acrylonitrile, carbon disulfide, and
acrolein: Laboratory analysts have indicated that acetonitrile concentrations may be artificially
high (or non-existent) due to site conditions and potential cross-contamination with concurrent
sampling of carbonyl compounds using Method TO-11 A. Similarly, laboratory analysts have
also indicated that acrylonitrile and carbon disulfide concentrations may also be artificially high
due to potential contamination of the samplers using Method TO-15. Additionally, questions
about the consistency and reliability of acrolein measurements have been raised during other
monitoring projects, such as SAMP (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 also excluded from
program-wide and site-specific data analyses related to risk.
The NATTS TAD (EPA, 2009b) identifies 19 pollutants ("MQO Core Analytes") that
participating sites are required to sample and analyze for under the NATTS program. Table 3-2
presents these 19 NATTS MQO Core Analytes. Monitoring for these pollutants is required
because they are major health risk drivers according to EPA (EPA, 2009b). Many of the
pollutants listed in Table 3-2 are identified as pollutants of interest via the risk-based screening
process. Note that beginning in July 2013, hexavalent chromium was removed from the list of
required pollutants for which to sample under the NATTS program. As a result, many NATTS
sites discontinued sampling hexavalent chromium. In 2014, two NATTS sites sampled for this
pollutant, RIVA and S4MO, although S4MO also discontinued sampling in July 2014.
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 andEQL-0512-
201/202
Manganese
Lead
Nickel
Hexavalent chromium1
Metals/ASTM D7614
1 Hexavalent chromium was removed from the Core Analytes list in July 2013,
although a few NATTS sites continued to sample for it in 2014.
3.3 Additional Program-Level Analyses of the 2014 National Monitoring Programs
Dataset
This section summarizes additional analyses performed on the 2014 NMP dataset at the
program level. Additional program4evel analyses include a review of how concentrations vary
among the sites themselves and from quarter-to-quarter. The results of these 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 analysis, the annual average concentration for each site is 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.
3-7
-------
In order to further this analysis, the program-level average concentrations, as presented in
Section 4.1 in Tables 4-1 through 4-6, are plotted against the site-specific annual 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 through 4-3 are presented in method-specific units, but have
been converted to a common unit of measurement (|ig/m3) for the purposes of this 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 the 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 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 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
analyses are presented in the individual state-specific sections (Sections 5 through 23).
3.4.1 Site Characterization
For each site participating in the 2014 NMP, a site characterization was performed. This
characterization includes a review of the nearby area surrounding the monitoring site; plotting of
emissions sources surrounding the monitoring site; and obtaining traffic data and other
characterizing information. For the 2014 NMP report, the locations of point sources located near
the monitoring sites were obtained from Version 2 of the 2011 NEI (EPA, 2015a). Sources for
other site-characterizing data are provided in the individual state sections.
3-8
-------
3.4.2 Meteorological Analysis
Several site-specific meteorological analyses were performed in order to help readers
determine which meteorological factors may play a role in a given site's air quality. First, the
average (or mean) for several meteorological parameters (such as temperature, pressure, and
wind speed) are provided. Two averages are presented for each parameter, one average
representing all of 2014 and one average representing sample days only. These two averages
provide an indication of how meteorological conditions on sample days varied from typical
conditions experienced throughout the year.
The way in which these meteorological parameters were developed is different for the
2014 report compared to previous reports. Previously, these averages were based on hourly
meteorological observations collected from the National Weather Service (NWS) weather station
nearest each site and obtained from the National Climatic Data Center (NCDC). Beginning with
the 2014 report, where meteorological data collected at each monitoring site are available in
AQS, these data were used to calculate the full-year and sample day averages. This change was
made to better represent the meteorological conditions at the individual monitoring sites. Where
no site-specific data were available, or where data is missing from AQS, NWS data was
substituted.
Wind roses were constructed for each site in order to further characterize the meteorology
at or near each monitoring site. A wind rose shows the frequency at which a given wind speed
and direction are measured near the monitoring site, capturing day-to-day fluctuations at the
surface while allowing the predominant direction from which the wind blows to be identified.
Thus, a wind rose is often used in determining where to install an ambient monitoring site when
trying to capture emissions from an upwind source. A wind rose may also be useful in
determining whether high concentrations correlate with a specific wind direction. A wind rose
shows the frequency of wind directions as petals positioned around a 16-point compass, and uses
color or shading to represent wind speeds. Wind roses are constructed by uploading hourly
surface wind data obtained from either data collected at the individual monitoring site, where
possible, or the nearest NWS weather station, into a wind rose software program, WRPLOT
(Lakes, 2011). For each site, two wind roses were constructed. First, 2014 data were used to
construct a wind rose presenting wind data for the entire calendar year; then, a wind rose was
constructed to present wind data for sample days only. These wind roses are used to determine if
3-9
-------
the meteorological conditions on sample days were representative of conditions experienced
throughout the sampling year near each site.
The NWS defines calm winds as those less than 3 mph; thus, after converting and
formatting for the WRPLOT program, wind speeds of 0 knots to 2 knots are considered "calm"
(NOAA, 1998). But wind speed data collected at the individual sites is often reported at levels
less than this threshold. As such, the threshold has been lowered to 1 knot for the wind roses.
3.4.3 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
• The calculation of cancer risk and noncancer hazard approximations in relation to
cancer and noncancer health effects, including the emission tracer analysis
• Risk-based emissions assessment.
3.4.3.1 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 4-6, 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 analysis is an extension of the 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.
3-10
-------
The box plots used in this 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 white circle plotted on top of the bar and the horizontal
lines extending outward from the white circle 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 through 4-3 are presented in method-specific units, but have been converted to a
common unit of measurement (|ig/m3) for the purposes of this 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.
3.4.3.2 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. Thirty-five
of the 51 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).
The trends figures and analyses are presented as 1-year statistical metrics. The following
criteria were used to calculate valid statistical metrics:
• Analysis must have been performed under the NMP by the contract laboratory.
• There must be a minimum of at least 5 years of consecutive data.
Five individual statistical metrics were calculated for this 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
3-11
-------
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 blue line, and top of the box, respectively); and the
average (or mean) concentration (as denoted by the orange diamond). Each of the five 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, 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,
2016b) in order to ensure the use of the most up-to-date data available. Recall that non-detects
are uploaded into AQS as zeros (EPA, 2009b). Similar to other analyses presented in this report,
zeros representing these 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 for each site, reflecting the data treatment described in Section 3.1.
For 2014, 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.
3.4.3.3 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, 2009c). 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
3-12
-------
equal to one indicates that adverse noncancer effects are not likely to occur, and thus can be
considered to have negligible hazard." (EPA, 2015b).
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, 2015c/2015d). 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 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 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.
There are, however, limitations to this analysis. Wind data 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.
3-13
-------
3.4.3.4 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.
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 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.
3-14
-------
b. To apply weight based on noncancer toxicity, divide the emissions of each
pollutant by its noncancer RfC.
The PAHs and/or phenols 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 analysis. Table 3-3 presents the 22 PAHs
measured by Method TO-13A and their associated POM Groups, if applicable.
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
Benzo(g,h,i)perylene
Group 2
Group 2b
PAH 880E5
Benzo(k)fluoranthene
Group 6
PAH 176E4
Chrysene
Group 7
PAH 176E5
Coronene
NA
Cyclopenta[cd]pyrene
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
Phenanthrene
NA
PAH 000E0
Pyrene
NA
PAH 000E0
Retene
NA
* Emissions for these pollutants are reported to the NEI individually; therefore, they
are not included in one of the POM Groups.
1 Reference = EPA 2015c
NA = no POM Group assigned.
3-15
-------
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 analysis, but both names are provided in Table 3-3. Note
the following in regards 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 by Method TO-13A 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.
3-16
-------
4.0 Summary of the 2014 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.
4.1.1 Target Pollutant Detection Rates
There is an experimentally determined MDL for every target pollutant, as described in
Section 2.2. Quantification below 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 below the corresponding MDLs. 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 4-6 summarize the number of times each target pollutant was detected
out of the number of valid samples collected and analyzed. Approximately 55 percent of the
reported measurements (based on the preprocessed daily measurements) were equal to or greater
than the MDLs across the program. The following list provides the percentage of measurements
that were above the MDLs for each of the target pollutant groups:
• 43 percent for VOCs
• 42 percent for SNMOCs
• 83 percent for carbonyl compounds
• 73 percent for PAHs
• 79 percent for metals
4-1
-------
• 53 percent for hexavalent chromium.
Some pollutants were detected in every valid sample collected while others were
infrequently detected or not detected at all. Among the carbonyl compounds, formaldehyde and
acetaldehyde had the greatest number of measured detections (1,678), based on the preprocessed
daily measurements. These pollutants were reported in every valid carbonyl compound sample
collected (1,678). Eleven VOCs (acetylene, benzene, carbon tetrachloride, chloromethane,
dichlorodifluoromethane, dichlorotetrafluoroethane, ethylbenzene, m,p-y.ylene, propylene,
toluene, and trichlorofluoromethane) were detected in every valid VOC sample collected (1,556).
Eleven pollutants (acetylene, //-butane, ethane, ethylene, //-hexane, isobutene, 2-methylpentane,
//-pentane, propane, propylene, and toluene) were detected in every valid SNMOC sample
collected (351). 9-Fluorenone, fluoranthene, naphthalene, phenanthrene, and pyrene were
detected in every valid PAH sample collected (1,105). Antimony, cadmium, cobalt, lead, and
manganese were detected in every valid speciated metals sample collected (1,144). Hexavalent
chromium was detected in 49 samples (out of 92 valid samples collected).
BTUT and NBIL have the greatest number of measured detections by a considerable
margin (6,712 for BTUT and 6,174 for NBIL). They are two of only three NMP sites that
collected samples for at least five analytical methods/pollutant groups and the only two NMP site
to sample both VOCs and SNMOCs. However, the detection rates for these sites (67 percent and
62 percent, respectively) were not as high as other sites. Detection rates for sites that sampled
suites of pollutants that are frequently detected tended to be higher (refer to the list of method-
specific percentages of measurements above the MDL listed above). For example, metals were
rarely reported as non-detects. As a result, sites that sampled only metals (such as PAFL) would
be expected to have higher detection rates. PAFL's detection rate is 100 percent. Conversely,
VOCs had one of the lowest percentages of concentrations greater than the MDLs (43 percent).
A site measuring only VOCs would be expected to have lower detection rates, such as SPAZ
(49 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)
# of
#
# of
Measured
Arithmetic
First
Third
Standard
of Non-
Measured
Detections
Minimum2
Maximum
Mean
Median
Quartile
Quartile
Deviation
Pollutant
Detects1
Detections1
-------
Table 4-1. Statistical Summaries of the VOC 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
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
-Pincnc
347
4
0
1.87
4.82
0.041
0
0
0
0.402
Propane
0
351
0
1.65
167
27.1
18.4
7.78
35.4
26.0
/7-Propylbenzene
185
166
87
0.081
3.77
0.163
0
0
0.264
0.280
Propylene
0
351
0
0.417
19.2
1.19
1.01
0.719
1.31
1.12
Propyne
351
0
0
Not Detected
Styrene3
183
70
5
0.141
22.6
1.63
0
0
0.463
3.96
Toluene
0
351
0
0.558
67.0
5.37
4.08
2.64
6.59
5.30
//-Tridccanc
215
136
125
0.087
22.3
0.180
0
0
0.151
1.21
1-Tridecene
351
0
0
Not Detected
1,2,3 -T rimethy lbenzene
184
167
72
0.100
1.16
0.141
0
0
0.242
0.188
1,2,4-Trimethy lbenzene
7
344
132
0.120
8.90
0.899
0.678
0.407
1.13
0.807
1.3.5 -T rimethy lbenzene
139
212
114
0.097
1.54
0.244
0.190
0
0.414
0.274
2,2,3-Trimethylpentane
291
60
31
0.086
1.81
0.058
0
0
0
0.162
2,2,4-Trimethy lpentane3
174
175
39
0.107
12.1
0.452
0.107
0
0.527
0.981
2,3,4 -T rimethy lpentane
97
254
140
0.068
3.21
0.281
0.181
0
0.354
0.368
/7-Undecane
47
304
282
0.087
21.1
0.373
0.240
0.147
0.401
1.15
1-Undecene
328
23
23
0.086
0.718
0.017
0
0
0
0.075
w;-Xylene//?-Xylene
4
347
27
0.201
11.4
1.74
1.41
0.751
2.37
1.37
1 Out of 351 valid samples.
2 Excludes zeros for non-detects.
3 The total number of concentrations may not add up to 351 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)
Pollutant
#
of Non-
Detects1
# of
Measured
Detections1
# of
Measured
Detections
-------
Table 4-3. 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 4-4. 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 4-5. Statistical Summaries of the Metals Concentrations
4^
to
# 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 Summary of the Hexavalent Chromium Concentrations
Pollutant
#
of Non-
Detects1
# of
Measured
Detections1
# of
Measured
Detections
-------
4.1.2 Concentration Range and Data Distribution
The concentrations measured during the 2014 NMP exhibit a wide range of variability.
The minimum and maximum concentration measured (excluding zeros substituted for non-
detects) for each target pollutant are presented in Tables 4-1 through 4-6 (in respective pollutant
group units). Some pollutants, such as dichloromethane, had a wide range of concentrations
measured, while other pollutants, such as dichlorotetrafluoroethane, did not, even though they
were both detected frequently. For each method-specific pollutant group, the pollutant with the
largest range in measured concentrations is as follows:
• For VOCs, dichloromethane (0.056 ppbv to 2,420 ppbv)
• For SNMOCs, ethane (1.11 ppbC to 250 ppbC)
• For carbonyl compounds, formaldehyde (0.010 ppbv to 21.0 ppbv)
• For PAHs, naphthalene (0.780 ng/m3 to 568 ng/m3)
• For metals in PMio, manganese (0.270 ng/m3 to 136 ng/m3)
• For metals in TSP, manganese (2.59 ng/m3 to 53.9 ng/m3)
• For hexavalent chromium, 0.0036 ng/m3 to 0.640 ng/m3.
4.1.3 Central Tendency
In addition to the number of measured detections and the concentration ranges,
Tables 4-1 through 4-6 also present a number of central tendency and data distribution statistics
(arithmetic mean or average, median, first and third quartiles, and standard deviation) for each of
the pollutants sampled during the 2014 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 (15.7 ±2.91 ppbv)
• Di chl oromethane (4.04 ± 3.3 8 ppbv).
• Acetylene (1.13 ± 0.13 ppbv)
4-14
-------
The three SNMOCs with the highest average concentrations, as presented in Table 4-2,
are:
• Ethane (43.5 ± 4.50 ppbC)
• Propane (27.1 ± 2.73 ppbC)
• «-Butane (11.2 ± 1.09 ppbC).
The three carbonyl compounds with the highest average concentrations, as presented
in Table 4-3, are:
• Formaldehyde (2.25 ± 0.08 ppbv)
• Acetone (1.12 ± 0.04 ppbv).
• Acetaldehyde (0.974 ± 0.030 ppbv).
The three PAHs with the highest average concentrations, as presented in Tables 4-4, are:
• Naphthalene (66.5 ± 3.34 ng/m3)
• Phenanthrene (9.51 ± 0.866 ng/m3)
• Acenaphthene (4.89 ± 0.673 ng/m3).
The three metals with the highest average concentrations for both PMio and TSP
fractions, as presented in Table 4-5, are;
• Manganese (PMio = 8.02 ± 0.61 ng/m3, TSP = 17.7 ±1.13 ng/m3)
• Total chromium (PMio = 4.86 ± 0.22 ng/m3, TSP = 2.79 ± 0.17 ng/m3)
• Lead (PMio = 2.93 ± 0.21 ng/m3, TSP = 3.37 ± 0.32 ng/m3).
The average concentration of hexavalent chromium, as presented in Table 4-6, is
0.020 ± 0.014 ng/m3.
Appendices J through O present statistical calculations on a site-specific basis, similar to
those presented in Tables 4-1 through 4-6.
4-15
-------
4.2 Preliminary Risk-Based Screening and Pollutants of Interest
Based on the preliminary risk-based screening process described in Section 3.2, Table 4-7
identifies the pollutants that failed at least one screen; summarizes each pollutant's 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 Table 4-7 are listed in descending order by number of screens failed.
Table 4-7 shows that benzene failed the greatest number of screens (1,794), although carbon
tetrachloride, formaldehyde, acetaldehyde, 1,2-dichloroethane, and 1,3-butadiene each failed
greater than 1,000 screens. These pollutants were also among those with the greatest number of
measured detections among pollutants shown in Table 4-7. Conversely, three pollutants listed in
Table 4-7 failed only one screen each (1,1-dichloroethane, antimony, and tetrachloroethylene).
The number of measured detections for these pollutants varied significantly. Antimony and
tetrachloroethylene were detected in greater than 1,000 samples each, while 1,1-dichloroethane
was detected infrequently (61 out of 1,556 valid samples). Four pollutants exhibited a failure rate
of 100 percent (benzene, 1,2-dichloroethane, 1,2-dibromoethane, and chloroprene). Benzene and
1,2-dichloroetahne were detected in 100 percent and 93 percent of samples, respectively, while
1,2-dibromoethane and chloroprene were detected in 3 percent and less than 1 percent of samples
collected, respectively. Thus, 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.
4-16
-------
Table 4-7. 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
1,794
1,794
100.00
14.05
14.05
Formaldehyde
0.077
1,674
1,678
99.76
13.11
27.15
Acetaldehyde
0.45
1,592
1,678
94.87
12.46
39.62
Carbon Tetrachloride
0.17
1,551
1,556
99.68
12.14
51.76
1,2-Dichloroethane
0.038
1,441
1,441
100.00
11.28
63.04
1.3 -Butadiene
0.03
1,412
1,492
94.64
11.06
74.10
Arsenic
0.00023
934
1,111
84.07
7.31
81.41
Naphthalene
0.029
826
1,105
74.75
6.47
87.88
Hexachloro-1,3 -butadiene
0.045
339
359
94.43
2.65
90.53
Ethylbenzene
0.4
275
1,759
15.63
2.15
92.69
p-Dichlorobenzene
0.091
197
681
28.93
1.54
94.23
Nickel
0.0021
126
1,106
11.39
0.99
95.22
Vinyl chloride
0.11
102
248
41.13
0.80
96.01
Acenaphthene
0.011
96
1,039
9.24
0.75
96.77
Fluorene
0.011
82
814
10.07
0.64
97.41
Manganese
0.03
63
1,144
5.51
0.49
97.90
1,2-Dibromoethane
0.0017
40
40
100.00
0.31
98.21
T richloroethylene
0.2
30
315
9.52
0.23
98.45
Benzo(a)pyrene
0.00057
29
851
3.41
0.23
98.68
Cadmium
0.00056
29
1,144
2.53
0.23
98.90
Dichloromethane
60
28
1,504
1.86
0.22
99.12
Propionaldehyde
0.8
27
1,668
1.62
0.21
99.33
1,1,2 -T richloroethane
0.0625
26
28
92.86
0.20
99.54
Fluoranthene
0.011
26
1,105
2.35
0.20
99.74
Lead
0.015
15
1,144
1.31
0.12
99.86
Bromomethane
0.5
4
1,252
0.32
0.03
99.89
Hexavalent Chromium
0.000083
3
49
6.12
0.02
99.91
Acenaphthylene
0.011
2
586
0.34
0.02
99.93
Chloroform
9.8
2
1,408
0.14
0.02
99.95
Chloroprene
0.0021
2
2
100.00
0.02
99.96
Xylenes
10
2
1,796
0.11
0.02
99.98
1,1 -Dichloroethane
0.625
1
61
1.64
0.01
99.98
Antimony
0.02
1
1,144
0.09
0.01
99.99
T etracliloroethylene
3.8
1
1,272
0.08
0.01
100.00
Total
12,772
34,374
37.16
4-17
-------
The program-level pollutants of interest, as indicated by the shading in Table 4-7, are
identified as follows:
• Acetaldehyde
• 1,2-Dichloroethane
• Arsenic
• Ethylbenzene
• Benzene
• Formaldehyde
• 1,3-Butadiene
• Hexachloro-1,3-butadiene
• Carbon Tetrachloride
• Naphthalene
• /;-Dichlorobenzene
• Nickel.
The pollutants of interest identified via the preliminary risk-based screening approach for
2014 are similar to the pollutants identified in previous years. Acenaphthene is the only pollutant
that was a program-wide pollutant of interest for 2013 but is not on the list for 2014.
Acenaphthene is the second pollutant, behind vinyl chloride, just outside the 95 percent criteria,
as shown in Table 4-7, and therefore is not a pollutant of interest for 2014.
Of the 71 pollutants sampled for under the NMP that have corresponding screening
values, concentrations of 34 pollutants failed at least one screen. Of these, a total of 12,772
concentrations out of 34,374 concentrations (or 37 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 (12,772 of 56,119, or nearly 23 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 4-8 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 4-8 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
4-18
-------
failed screens and the total number of screens conducted (based on applicable measured
detections) and is also provided in Table 4-8.
Table 4-8. Site-Specific Risk-Based Screening Comparison
Site
# of
Failed
Screens
Total # of
Measured
Detections1
%of
Failed
Screens
# of Pollutant
Groups
Analyzed
PXSS
563
2,482
22.68
4
S4MO
561
2,655
21.13
5
TOOK
537
1,764
30.44
3
GPCO
502
2,514
19.97
4
BTUT
497
2,387
20.82
5
NBIL
495
2,473
20.02
5
TROK
489
1,704
28.70
3
TMOK
482
1,727
27.91
3
DEMI
473
2,024
23.37
3
SEWA
436
2,279
19.13
4
YUOK
421
1,660
25.36
3
ELNJ
409
1,213
33.72
2
LEKY
407
1,561
26.07
3
CSNJ
400
1,213
32.98
2
OCOK
400
1,613
24.80
3
SPIL
388
1,124
34.52
2
CHNJ
376
1,185
31.73
2
ROIL
373
1,060
35.19
2
ASKY
363
1,055
34.41
2
GLKY
360
2,009
17.92
4
NBNJ
315
1,068
29.49
2
TVKY
306
1,076
28.44
1
ATKY
279
1,001
27.87
1
BLKY
260
1,032
25.19
2
LAKY
252
893
28.22
1
CCKY
233
1,127
20.67
2
SPAZ
153
450
34.00
1
SKFL
143
901
15.87
2
RICO
142
418
33.97
2
ORFL
120
180
66.67
1
INDEM
113
171
66.08
1
PACO
113
402
28.11
2
AZFL
111
168
66.07
1
SYFL
110
165
66.67
1
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
4-19
-------
Table 4-8. Site-Specific Risk-Based Screening Comparison (Continued)
# of
Total # of
%of
# of Pollutant
Failed
Measured
Failed
Groups
Site
Screens
Detections1
Screens
Analyzed
WPIN
108
162
66.67
1
SJJCA
102
1,251
8.15
2
BOMA
98
1,367
7.17
2
ASKY-M
94
575
16.35
1
BRCO
92
311
29.58
2
BMCO
87
329
26.44
2
ROCH
86
813
10.58
1
RFCO
78
229
34.06
2
BXNY
77
833
9.24
1
WADC
57
751
7.59
1
BAKY
56
568
9.86
1
CELA
56
701
7.99
1
RIVA
50
668
7.49
2
RUCA
50
689
7.26
1
PRRI
46
833
5.52
1
PAFL
29
300
9.67
1
UNVT
24
985
2.44
2
Total
12,772
56,119
22.76
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, PXSS has the largest number of failed screens (563), followed by S4MO (561)
and TOOK (537); conversely, PRRI, PAFL, and UNVT failed relatively few screens each. Every
NMP site had at least one failed 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 4-8. For
example, sites sampling four or five pollutant groups tended to have a higher number of failed
screens due to the large number of pollutants sampled. For sites sampling only one or two
pollutant groups, it depends on the pollutant group sampled as the number of compounds
analyzed varies from one (hexavalent chromium) to 80 (SNMOCs). Although ORFL, SYFL, 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 4-7.
Thus, sites sampling only carbonyl compounds tend to have relatively high 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. These sites both
4-20
-------
sampled four pollutant groups and have a failure rate just 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 among the sites. Tables 4-9 through 4-12 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 is the average concentration of all
measured detections and zeros substituted for non-detects for a given year. Further, an annual
average is only calculated where at least three quarterly averages could be calculated and where
the site-specific method completeness is at least 85 percent. The annual average concentrations
in Tables 4-9 and 4-10, for VOC/SNMOCs and carbonyl compounds, respectively, are reported
in [j,g/m3 while the annual average concentrations for PAHs and metals, in Tables 4-11 and 4-12,
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 4-9 through 4-12 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 4-12 for each metal (TSP) pollutant of interest shown.
4-21
-------
Table 4-9. Annual Average Concentration Comparison of the VOC/SNMOC Pollutants of Interest
Carbon
P-
1,2-
Hexachloro-1,3-
Benzene
1,3-Butadiene
Tetrachloride
Dichlorobenzene
Dichloroethane
Ethylbenzene
Butadiene
Rank
frig/m3)
frig/m3)
frig/m3)
frig/m3)
frig/m3)
frig/m3)
frig/m3)
PACO
TVKY
TVKY
SPAZ
TVKY
SPAZ
BTUT
1
1.49 ±0.14
0.38 ±0.22
0.87 ±0.13
0.21 ±0.05
3.54 ± 1.66
0.60 ±0.13
0.05 ±0.01
ROIL
PXSS
BLKY
PXSS
LAKY
PXSS
TVKY
2
1.22 ±0.21
0.20 ±0.05
0.73 ±0.09
0.17 ±0.03
0.97 ±0.47
0.57 ±0.10
0.03 ±0.01
RICO
SPAZ
LAKY
S4MO
BLKY
GPCO
TOOK
3
1.09 ±0.14
0.19 ±0.06
0.71 ±0.06
0.13 ±0.05
0.81 ±0.43
0.45 ± 0.06
0.03 ±0.01
SPAZ
BLKY
CCKY
TOOK
ATKY
TOOK
ELNJ
4
1.09 ±0.24
0.17 ± 0.16
0.70 ±0.03
0.07 ±0.01
0.58 ±0.22
0.39 ±0.06
0.02 ±0.01
PXSS
GPCO
ATKY
TMOK
CCKY
TROK
TROK
5
1.05 ±0.20
0.17 ±0.03
0.68 ±0.03
0.07 ±0.01
0.49 ±0.16
0.37 ±0.05
0.02 ±0.01
TVKY
SPIL
SEWA
TROK
BTUT
DEMI
LAKY
6
1.04 ±0.35
0.13 ±0.02
0.67 ±0.02
0.06 ±0.01
0.11 ±0.01
0.37 ±0.09
0.02 ±0.01
TOOK
ELNJ
DEMI
BTUT
TOOK
ELNJ
NBNJ
7
1.03 ±0.11
0.12 ±0.01
0.67 ±0.02
0.05 ±0.04
0.09 ±0.01
0.36 ±0.05
0.02 ±0.01
GPCO
LAKY
ASKY
ELNJ
ELNJ
TMOK
CHNJ
8
0.99 ±0.12
0.11 ±0.04
0.65 ±0.02
0.04 ±0.01
0.09 ±0.01
0.34 ±0.05
0.02 ±0.01
ASKY
RICO
YUOK
CSNJ
TMOK
CSNJ
NBIL
9
0.87 ±0.39
0.10 ±0.03
0.63 ±0.02
0.04 ±0.01
0.08 ±0.01
0.33 ±0.11
0.02 ±0.01
TMOK
DEMI
NBNJ
LEKY
S4MO
ROIL
TMOK
10
0.81 ±0.08
0.10 ±0.02
0.63 ±0.02
0.03 ±0.01
0.08 ±0.01
0.32 ±0.10
0.02 ±0.01
BOLD ITALICS = EPA-designated NATTS Site
-------
Table 4-10. Annual Average Concentration Comparison of the
Carbonyl Compound Pollutants of Interest
Rank
Acet aldehyde
frig/m3)
Formaldehyde
frig/m3)
BTUT
BTUT
1
3.33 ±0.34
5.92 ±0.73
GPCO
CSNJ
2
2.80 ±0.25
4.48 ±0.52
ELNJ
ELNJ
3
2.78 ±0.21
4.44 ±0.52
SPIL
GPCO
4
2.52 ±0.50
3.90 ±0.35
PXSS
PXSS
5
2.52 ±0.29
3.46 ±0.24
CSNJ
S4MO
6
2.49 ±0.22
3.45 ±0.48
NBIL
TMOK
7
2.36 ±0.45
3.41 ±0.39
S4MO
DEMI
8
2.08 ±0.22
3.25 ±0.35
ORFL
LEKY
9
2.01 ±0.25
3.15 ±0.90
TOOK
SPIL
10
1.97 ±0.22
3.12 ±0.35
BOLD ITALICS = EPA-designated NATTS Site
Table 4-11. Annual Average Concentration Comparison of the PAH Pollutant of Interest
Rank
Naphthalene
(ng/m3)
1
DEMI
116.80 ± 18.59
2
NBIL
109.13 ±28.14
3
BXNY
101.09 ± 10.72
4
GPCO
100.03 ± 13.48
5
CELA
88.54 ± 13.86
6
S4MO
81.79 ± 12.61
7
PXSS
78.25 ± 15.71
8
RUCA
75.23 ± 15.07
9
WADC
67.34 ± 10.18
10
RIVA
62.57 ±8.89
BOLD ITALICS = EPA-designated NATTS Site
4-23
-------
Table 4-12. Annual Average Concentration Comparison of the Metals Pollutants of Interest
Arsenic
Arsenic
Nickel
Nickel
(PMio)
(TSP)
(PMio)
(TSP)
Rank
(ng/m3)
(ng/m3)
(ng/m3)
(ng/m3)
ASKY-M
TROK
ASKY-M
TOOK
1
1.14 ±0.36
0.77 ±0.12
2.19 ±0.56
2.25 ±0.31
S4MO
TOOK
BOMA
TROK
2
0.90 ±0.15
0.76 ±0.08
1.99 ±0.42
1.70 ±0.21
BAKY
TMOK
SEWA
TMOK
3
0.85 ±0.17
0.67 ±0.09
1.74 ±0.40
1.52 ±0.14
PAFL
OCOK
PXSS
YUOK
4
0.81 ±0.16
0.48 ±0.07
1.42 ±0.21
0.96 ±0.10
BTUT
YUOK
SJJCA
OCOK
5
0.79 ±0.26
0.44 ±0.05
1.42 ±0.34
0.91 ±0.09
LEKY
BTUT
6
0.67 ±0.11
1.38 ±0.24
SEWA
S4MO
7
0.60 ±0.11
1.00 ±0.18
PXSS
PAFL
8
0.55 ±0.11
0.97 ±0.10
CCKY
BAKY
9
0.55 ±0.12
0.72 ±0.12
SJJCA
LEKY
10
0.44 ±0.07
0.58 ±0.08
BOLD ITALICS = EPA-designated NATTS Site
Observations from Tables 4-9 through 4-12 include the following:
• The highest annual average concentration among the program-wide pollutants of
interest was calculated for formaldehyde for BTUT (5.92 ± 0.73 |ig/m3). This was
also true for BTUT in 2012 and 2013, although the concentration for 2014 is
considerably less than it was for 2013. Formaldehyde accounts for 10 of the 12
annual average concentrations greater than 3.0 |ig/m3 shown in Tables 4-9 through
4-12 (with acetaldehyde and 1,2-dichloroethane each accounting for one).
• Among the VOCs shown in Table 4-9, the highest annual average concentration was
calculated for 1,2-dichloroethane for TVKY (3.54 ± 1.66 |ig/m3). However, no other
NMP site sampling this pollutant has an annual average concentration greater than
1 |ig/m3. The range of annual average concentrations shown for 1,2-dichloroethane
varies considerably, spreading across two orders of magnitude. While the Calvert
City, Kentucky sites (ATKY, BLKY, CCKY, LAKY, and TVKY) account for the
five highest annual average concentrations of this pollutant in Table 4-9, their
averages are also rather variable.
• Seven of the 10 annual average concentrations of benzene shown in Table 4-9 are
greater than 1 |ig/m3. PACO has the highest annual average benzene concentration
(1.49 ± 0.14 |ig/m3) among sites sampling benzene, with three of the six Colorado
sites ranking among the 10 highest. Other sites ranking among the highest annual
average benzene concentrations include ROIL, the two Phoenix sites (SPAZ and
PXSS), TVKY, and. TOOK. Note that the sites with the largest confidence intervals
shown in Table 4-9 were calculated for TVKY and ASKY. The two highest benzene
4-24
-------
concentrations measured across the program were measured at ASKY (12.4 |ig/m3)
and TVKY (9.92 |ig/m3).
The highest annual average concentration of 1,3-butadiene (0.38 ± 0.22 |ig/m3) was
calculated for TVKY for 2014. No other site has an annual average concentration of
1,3-butadiene greater than 0.20 |ig/m3. Three of the five Calvert City, Kentucky sites
appear among those with the highest annual average concentrations of 1,3-butadiene
(TVKY, BLKY, and LAKY). Note the relatively large confidence intervals
associated with the annual averages for TVKY and BLKY. These two sites have the
highest measurements of 1,3-butadiene across the program; of the six 1,3-butadiene
concentrations greater than 1 |ig/m3 measured across the program, five were
measured at TVKY and one at BLKY. The two Phoenix, Arizona sites rank second
and third for their annual average concentrations of 1,3-butadiene and have similar
concentrations (PXSS, 0.20 ± 0.05 |ig/m3 and SPAZ, 0.19 ± 0.06 |ig/m3).
Calvert City, Kentucky sites account for five of the 10 highest annual average
concentrations of carbon tetrachloride. Most of the annual average concentrations of
carbon tetrachloride do not vary significantly across the sites; less than 0.10 |ig/m3
separates most of the annual average carbon tetrachloride concentrations. Only annual
average concentrations calculated for the Calvert City, Kentucky sites are greater than
or equal to 0.70 |ig/m3 (ATKY being the exception). Measurements of carbon
tetrachloride collected at Calvert City sites account for 23 of the 25 highest carbon
tetrachloride concentrations measured across the program, including 16 of the 17
measurements greater than 1 |ig/m3.
Similar to previous years, the two Phoenix, Arizona sites have the two highest annual
averages concentrations of />dichlorobenzene among NMP sites and account for two
of the three annual average concentrations greater than 0.1 |ig/m3 shown in Table 4-9
(SPAZ - 0.21 ± 0.05 |ig/m3 and PXSS - 0.17 ± 0.03 |ig/m3). S4MO is the other NMP
site with an annual average concentration greater than 0.1 |ig/m3 (0.13 ± 0.05 |ig/m3).
Note the relatively large confidence interval shown for BTUT, which has the seventh
highest annual average concentration (0.05 ± 0.04 |ig/m3). Only two concentrations of
p-dichlorobenzene greater than 1 |ig/m3 were measured at NMP sites in 2014, one at
S4MO (1.14 |ig/m3) and one at BTUT (1.03 |ig/m3); all other measurements are less
than 0.60 |ig/m3.
The Calvert City, Kentucky sites also account for the five highest annual average
concentrations of 1,2-dichloroethane, although the averages vary significantly among
them, ranging from 3.54 ± 1.66 |ig/m3 for TVKY to 0.49 ± 0.16 |ig/m3 for CCKY.
These five sites account for the highest 125 measurements of 1,2-dichloroethane
measured across the program, ranging from 0.304 |ig/m3 to 27.4 |ig/m3, with the top
seven (those greater than 12 |ig/m3) measured at TVKY. All other NMP sites have
annual averages are less than 0.10 |ig/m3 (except BTUT, whose annual average
concentration is 0.11 ± 0.01 |ig/m3).
The Phoenix, Arizona sites also have the two highest annual average concentrations
of ethylbenzene (SPAZ, 0.60 ± 0.13 |ig/m3 and PXSS, 0.57 ± 0.10 |ig/m3) across the
program, followed by GPCO (0.45 ± 0.06 |ig/m3); all other annual average
4-25
-------
concentrations shown in Table 4-9 are less than 0.4 |ig/m3. Note that the confidence
intervals for the sites with the ninth and tenth highest annual averages (CSNJ and
ROIL) are similar to the sites with the highest annual average ethylbenzene
concentrations shown in Table 4-9. CSNJ and ROIL are the only NMP sites at which
ethylbenzene concentrations greater than 3 |ig/m3 were measured.
Hexachloro-1,3-butadiene is the only VOC in Table 4-9 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 (0.05 ± 0.01 |ig/m3), although the range
of annual average concentrations for hexachloro-l,3-butadiene is relatively small,
varying by 0.03 |ig/m3 across the sites shown and by 0.05 |ig/m3 across all NMP sites.
TVKY, TOOK, TMOK, and ELNJ each appear in Table 4-9 a total of five times and
PXSS, SPAZ, and LAKY each appear four times. All three Tulsa, Oklahoma sites
appear in Table 4-9 for their annual average concentrations of />dichlorobenzene,
ethylbenzene, and hexachloro-l,3-butadiene. Both Phoenix, Arizona sites appear in
Table 4-9 for their annual average concentrations of benzene, 1,3-butadiene,
/;-dichlorobenzene, and ethylbenzene.
BTUT has the highest annual average concentrations of both acetaldehyde and
formaldehyde shown in Table 4-10. Seven of the NMP sites shown appear for both
pollutants (BTUT, GPCO, ELNJ, SPIL, PXSS, CSNJ, and S4MO), although their
order varies.
Annual average acetaldehyde concentrations shown in Table 4-10 vary from
3.33 ± 0.34 |ig/m3 for BTUT to 1.97 ± 0.221 |ig/m3 for TOOK. The maximum
acetaldehyde concentration was measured at NBIL (9.17 |ig/m3), which ranks seventh
for its annual average concentration, although a similar concentration was also
measured at BTUT (9.15 |ig/m3). Twenty-four additional acetaldehyde concentrations
greater than 5 |ig/m3 were measured across the program, with the most measured at
SPIL (8), which ranks fourth for its annual average acetaldehyde concentration.
As shown in Table 4-10, three NMP sites have annual average formaldehyde
concentrations greater than 4 |ig/m3 (BTUT, CSNJ, and ELNJ) and all 10 sites shown
in Table 4-10 have annual average concentrations of formaldehyde greater than
3 |ig/m3.
Although BTUT has the highest annual average concentration of formaldehyde
(5.92 ± 0.73 |ig/m3), the highest concentrations measured across the program were
measured at other sites. The maximum concentration of formaldehyde was measured
at LEKY (25.9 |ig/m3), which ranks ninth for its annual average concentration (and
has the largest confidence interval associated with its annual average). Other higher
formaldehyde concentrations were measured at NBNJ and AZFL, in addition to sites
shown in Table 4-10. Of the 20 formaldehyde concentrations greater than 10 |ig/m3
measured across the program, nine were measured at BTUT, five at NBNJ, two at
ELNJ, and one each at ROIL, LEKY, CSNJ, and AZFL.
4-26
-------
Naphthalene is the only PAH program-wide pollutant of interest. Table 4-11 shows
that the range of the 10 highest annual average concentrations of naphthalene vary
considerably, from 116.80 ± 18.59 ng/m3 for DEMI to 62.57 ± 8.89 ng/m3 forRIVA,
with four NMP sites having annual averages greater than 100 ng/m3. The maximum
naphthalene concentration was measured at NBIL (568 ng/m3), and is the only
measurement of this pollutant greater than 400 ng/m3. Two additional concentrations
greater than 300 ng/m3 were also measured NBIL; two were also measured at DEMI
and one was measured at PXSS. With the exception of RIVA, which ranks 10th in
Table 4-11, each of the NMP sites shown in Table 4-11 measured at least one
naphthalene concentration greater than 200 ng/m3 (with the most measured at NBIL,
10).
ASKY-M has the highest annual average concentration for both of the program-wide
PMio metals pollutants of interest, similar to 2013. Four of the five Kentucky sites
sampling PMio metals (and where annual average concentrations could be calculated)
appear in Table 4-12 for arsenic (GLKY is the exception) while three appear in
Table 4-12 for nickel. S4MO, PAFL, and BTUT round out the top five for arsenic.
Annual averages of arsenic for S4MO consistently rank among the highest in past
annual reports. Aside from ASKY-M, NATTS sites have the highest ranking annual
averages for nickel. For the last several years, the annual average nickel concentration
for SEWA has been at or near the top.
The maximum arsenic concentration was measured at ASKY-M (10.1 ng/m3), with
the next highest arsenic concentration nearly half as high (5.04 ng/m3, measured at
BTUT). Arsenic concentrations greater than 2 ng/m3 were measured at seven NMP
sampling PMio metals; ASKY-M has the greatest number of arsenic concentrations
greater than 2 ng/m3 (6), followed by BTUT (4), S4MO and BAKY (3), and NBIL
(2), with PXSS and PAFL each measuring one.
Among the Oklahoma sites sampling TSP metals, TROK has the highest annual
average concentration of arsenic (0.77 ± 0.12 ng/m3), similar to 2013, although the
annual average concentration for TOOK is similar (0.76 ± 0.08 ng/m3). The other
Tulsa site, TMOK, ranks third (0.67 ± 0.09 ng/m3) while the OCOK and YUOK sites
have significantly lower annual average concentrations of arsenic (0.48 ± 0.07 ng/m3
and 0.44 ± 0.05 ng/m3, respectively).
The range of annual average concentrations for nickel is relatively large, ranging
from 2.19 ± 0.56 ng/m3 for ASKY-M to 0.58 ± 0.08 ng/m3 for LEKY. Three nickel
concentrations greater than 9 ng/m3 were measured in 2014, each at a different site
(SJJCA, ASKY-M, and BOMA); these sites rank fifth, first, and second for their
annual average nickel concentrations, respectively, as shown in Table 4-12. ASKY-M
has the greatest number of nickel concentrations greater than 5 ng/m3 among NMP
sampling PMio metals (7), followed by SEWA (4), and BOMA (3), with SJJCA and
BTUT each measuring one.
Among the Oklahoma sites sampling TSP metals, the three Tulsa sites ranked highest
for nickel while the Oklahoma City sites have significantly lower annual average
concentrations of nickel. All but one of TOOK's 62 nickel concentrations are greater
4-27
-------
than 1 ng/m3 (with the exception at 0.989 ng/m3), with slightly fewer measured at the
other Tulsa sites (TMOK at 51 and TROK at 50). The number of nickel
concentrations greater than 1 ng/m3 measured at the two Oklahoma City sites is
considerably less (YUOK, 23 and OCOK, 20).
• PXSS appears on the top 10 list for nine of the 12 program-level pollutants of interest
shown in Tables 4-9 through 4-12; TOOK and TMOK appear in these tables for eight
of the 12 program-level pollutants of interest; ELNJ, S4MO, and BTUT appear in the
tables for seven of the 12 program-level pollutants of interest; and GPCO appears in
the tables six times.
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 (in
gray) overlain on the program-level averages (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 (i.e., 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 from the ERG laboratory is also plotted on the graph as an indication of the
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 by two methods (VOC and
SNMOC) and identified as program-level pollutants of interest, two graphs are presented, one for
each method. Note that while the Garfield County, Colorado sites have their canister samples
analyzed using the SNMOC method, BTUT and NBIL have their canister samples analyzed
4-28
-------
using both methods. While both results are shown in this section, only the VOCs results are
discussed throughout the remainder of this report, as described in Section 3.2.
Observations from Figures 4-1 through 4-12 include the following:
• The program-level average concentration of acetaldehyde is 1.76 |ig/m3, as shown in
purple in Figure 4-1. Site-specific annual average concentrations range from
0.42 |ig/m3 (BMCO) to 3.33 |ig/m3 (BTUT). The annual average concentration for
BTUT is nearly twice the program-level average concentration for acetaldehyde.
Other sites with annual average concentrations greater than the program-level average
include CSNJ, DEMI, ELNJ, GPCO, NBIL, ORFL, PXSS, ROIL, S4MO, SPIL,
TMOK, TOOK, and YUOK. SPIL and NBIL 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) other than BMCO
include GLKY, RICO, and SEWA. Annual averages could not be calculated for
BRCO, NBNJ, PACO, RFCO, and SKFL.
• Figure 4-2 presents the inter-site variability graph for arsenic, which also includes a
comparison of PMio results and TSP results. Note that only sites from Oklahoma are
using TSP samplers. The program-level average concentration of arsenic in PMio is
similar to the program average concentration of arsenic in TSP, with a PMio average
of 0.61 ng/m3 and a TSP average of 0.63 ng/m3. There is more variability across the
program associated with the PMio measurements than the TSP measurements, as
indicated by the range of annual average concentrations as well as confidence
intervals shown. Site-specific annual average arsenic concentrations range from
0.21 ng/m3 (UNVT) to 1.14 ng/m3 (ASKY-M) for PMio and 0.44 ng/m3 (YUOK) to
0.77 ng/m3 (TROK) for TSP. Annual averages could not be calculated for BLKY and
NBIL. ASKY-M has the most variability in the PMio measurements, while TROK has
the most variability in the TSP measurements, although the confidence intervals
calculated for ASKY-M are three times larger than those for TROK.
• Figure 4-3a is the inter-site variability graph for benzene, as measured with
Method TO-15. The program-level average concentration of benzene is 0.74 |ig/m3.
Site-specific annual average concentrations range from 0.42 |ig/m3 (GLKY) to
1.22 |ig/m3 (ROIL). Other sites with annual average concentrations greater than
1 |ig/m3 include PXSS, SPAZ, TOOK, and TVKY. Sites with relatively low annual
average concentrations (less than 0.5 |ig/m3) other than GLKY include CHNJ and
NBIL. ASKY and TVKY have the most variability associated with the measurements
collected, as indicated by the confidence intervals shown in Figure 4-3a.
• Figure 4-3b is the inter-site variability graph for benzene, as measured with the
concurrent SNMOC method. Canister samples collected at seven sites are analyzed
with this method. The program-level average concentration of benzene (SNMOC
only) is 0.86 |ig/m3. Site-specific annual average concentrations range from
0.46 |ig/m3 (RFCO) to 1.49 |ig/m3 (PACO). The annual average concentrations for
PACO and RICO are greater than the program-level average; the annual average
concentrations for BTUT, NBIL, and RFCO are less than the program-level average;
4-29
-------
and annual average concentrations for BMCO and BRCO could not be calculated.
Note that canisters from BTUT and NBIL are analyzed using both methods and their
annual average benzene concentrations are similar although slightly higher using the
VOC method.
• Figure 4-4a is the inter-site variability graph for 1,3-butadiene, as measured with
Method TO-15. The program-level average concentration of 1,3-butadiene is
0.10 |ig/m3. Site-specific annual average concentrations span an order of magnitude,
ranging from 0.037 |ig/m3 (GLKY) to 0.38 |ig/m3 (TVKY). While many sites' annual
average concentrations are less than the program-level average, including some
whose annual averages are just greater than the MDL, the annual average
concentrations forBLKY, ELNJ, GPCO, LAKY, PXSS, SPAZ, SPIL, and TVKY are
greater than the program-level average concentration. The sites with the most
variability associated with the measurements collected are TVKY and BLKY, as
indicated by the confidence intervals shown in Figure 4-4a.
• Figure 4-4b is the inter-site variability graph for 1,3-butadiene, as measured with the
concurrent SNMOC method. Canister samples collected at seven sites are analyzed
with this method. The program-level average concentration of 1,3-butadiene
(SNMOC only) is 0.035 |ig/m3. Site-specific annual average concentrations range
from 0.015 |ig/m3 (NBIL) to 0.103 |ig/m3 (RICO). The annual average concentrations
for BTUT and RICO are greater than the program-level average, with the annual
average for RICO three times the program-level average concentration, while the
remaining annual average concentrations are less than the program-level average
concentration (where they could be calculated). Note that the annual average
concentrations for NBIL, PACO, and RFCO are less than the MDL for 1,3-butadiene
with the SNMOC method. This means that the annual average concentrations shown
incorporate data containing many zeroes substituted for non-detects, many
concentrations that are less than the MDL, or a combination of both. The MDL for
1,3-butadiene is considerably higher for the SNMOC method (0.121 |ig/m3) than the
TO-15 Method (0.029 |ig/m3). Because so many of the results are less than the MDL
or non-detects, there is less certainty associated with the SNMOC results for this
pollutant.
• The program-level average concentration of carbon tetrachloride is 0.64 |ig/m3, as
shown in blue in Figure 4-5. For most sites, the annual average concentration is 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 not unexpected. 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,
2016d). 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 0.58 |ig/m3 to 0.68 |ig/m3, with only Calvert City sites
4-30
-------
falling outside this range. In addition, the confidence intervals shown for these sites
are larger than most sites, particularly for TVKY, indicating a higher level of
variability in the measurements compared to other NMP sites.
Figure 4-6 presents the program-level and annual average concentrations of
p-dichlorobenzene. This figure shows that the program-level average concentration
(0.041 |ig/m3) and most of the site-specific annual average concentrations are less
than the MDL for this pollutant (0.088 |ig/m3), as indicated by the dashed blue line.
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 2014 measurements of
p-dichlorobenzene are non-detects and of the measured detections, nearly 70 percent
were less than the MDL. Only three NMP sites have annual average concentrations
greater than the MDL for this pollutant: PXSS, SPAZ, and S4MO. PXSS has the
greatest number of />dichlorobenzene measurements greater than the MDL (46), with
25 each for SPAZ and S4MO (as well as TOOK). The two Arizona sites have had the
two highest annual average concentrations of this pollutant for the last several NMP
reports. Other sites with a higher number of measurements greater than the MDL
include TMOK (19), TROK (12), and ELNJ (11). The highest/>-dichlorobenzene
concentrations measured across the program were measured at S4MO (1.14 |ig/m3)
and BTUT (1.03 |ig/m3). These concentrations are roughly twice the third highest
concentration measured in 2014 and helps explain, at least partially, why the
confidence intervals are so large for these sites.
Figure 4-7 shows that 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 0.06 |ig/m3 (SPAZ) to 0.11 |ig/m3
(BTUT), which are all similar to or just greater than the MDL for this pollutant
(0.056 |ig/m3). The annual average concentrations of 1,2-dichloroethane for the five
Calvert City sites range from 0.49 |ig/m3 (CCKY) to 3.54 |ig/m3 (TVKY). The
confidence intervals for these annual average concentrations are relatively large,
indicating there is considerable variability in the measurements collected at these
sites. Concentrations measured at these sites are driving the program-level average
concentration (0.31 |ig/m3), which was a similar finding in the 2012 and 2013 NMP
reports. Without the Calvert City sites, the program-level average concentration
would be 0.08 |ig/m3.
Figure 4-8a is the inter-site variability graph for ethylbenzene, as measured with
Method TO-15. The program-level average concentration of ethylbenzene is
0.25 |ig/m3. Site-specific annual average concentrations range from 0.08 |ig/m3
(BLKY) to 0.60 |ig/m3 (SPAZ). PXSS is the only other NMP site with an annual
average concentration of ethylbenzene greater than 0.5 |ig/m3. Sites with relatively
low annual average ethylbenzene concentrations (less than 0.15 |ig/m3) other than
BLKY include ATKY, CCKY, CHNJ, GLKY, and TVKY.
Figure 4-8b is the inter-site variability graph for ethylbenzene, as measured with the
concurrent SNMOC method. Canister samples collected at seven sites are analyzed
with this method. The program-level average concentration of ethylbenzene (SNMOC
4-31
-------
only) is 0.18 |ig/m3. Site-specific annual average concentrations range from
0.11 |ig/m3 (RFCO) to 0.32 |ig/m3 (RICO). The annual average concentrations for
BTUT and RICO are greater than the program-level average; the annual average
concentrations for PACO is similar to the program-level average; and the annual
average concentrations for remaining sites are less than the program-level average
concentration (where they could be calculated). Note that canister samples collected
at BTUT and NBIL are analyzed using both methods and their annual averages are
very similar between the two methods.
• The program-level average concentration of formaldehyde is 2.77 |ig/m3, as shown in
purple in Figure 4-9. Site-specific annual average concentrations of formaldehyde
range from 0.60 |ig/m3 (SEWA) to 5.92 |ig/m3 (BTUT). This is the fourth year in a
row that BTUT has had the highest annual average concentration of formaldehyde
among NMP sites. The annual average concentration for BTUT is more twice the
program-level average for formaldehyde, with all other NMP sites having annual
average concentrations less than 4.5 |ig/m3 (although the annual averages for CSNJ
and ELNJ are 4.48 |ig/m3 and 4.44 |ig/m3, respectively). Sites with relatively low
annual average concentrations (less than 1 |ig/m3) other than SEWA include BMCO
and RICO. Annual averages could not be calculated for BRCO, NBNJ, PACO,
RFCO, and SKFL.
• Figure 4-10 presents the program-level and site-specific annual average
concentrations of hexachloro-l,3-butadiene. This figure shows that the program-level
average concentration (0.018 |ig/m3) and all of the site-specific annual average
concentrations are considerably less than the MDL for this pollutant (0.29 |ig/m3), as
indicated by the dashed blue line. Only one of the hexachloro-l,3-butadiene
measurements collected in 2014 is greater than the detection limit, as indicated in
Table 4-1. Of the 1,556 valid VOC samples collected, only 359 (or 23 percent)
included measured detections of hexachloro-1,3-butadiene. This indicates that a large
number of substituted zeroes are included in the program-level and annual average
concentrations shown in Figure 4-10, which generally pull the averages down. The
maximum hexachloro-l,3-butadiene concentration measured across the program (and
the single concentration greater than the MDL) was measured at OCOK (0.61 |ig/m3),
explaining the large confidence interval shown for this site.
• Figure 4-11 presents the program-level and site-specific annual average
concentrations of naphthalene. The program-level average concentration
(66.46 ng/m3), as well as all of the annual average concentrations, where they could
be calculated, are considerably greater than the MDL for this pollutant. The site-
specific annual averages varied considerably, from 9.58 ng/m3 (UNVT) to
116.80 ng/m3 (DEMI). Sites with annual average concentrations greater than
100 ng/m3 besides DEMI include NBIL, BXNY, and GPCO; sites with annual
average concentrations less than 50 ng/m3 besides UNVT includes SEWA, BOMA,
BTUT, 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 range from 17.5 ng/m3 to 568 ng/m3.
4-32
-------
• Figure 4-12 shows the inter-site variability graph for nickel, which also includes a
comparison of PMio results and TSP results. Note that only sites from Oklahoma are
using TSP samplers. The program-level average concentration of nickel (PMio) is
1.11 ng/m3, while the program-level average concentration of nickel (TSP) is
1.47 ng/m3. Site-specific annual average nickel concentrations range from 0.40 ng/m3
(GLKY) to 2.19 ng/m3 (ASKY-M) for PMio and 0.91 ng/m3 (OCOK) to 2.25 ng/m3
(TOOK) for TSP. Similar observations were made in the 2013 NMP report. Annual
averages could not be calculated for BLKY and NBIL. ASKY-M has the most
variability in the nickel measurements collected, with concentrations of nickel
ranging from 0.26 ng/m3 to 9.64 ng/m3, although other sites, such as BOMA, SEW A,
SJJCA, and TOOK also have relatively large confidence intervals.
4-33
-------
Figure 4-1. Inter-Site Variability for Acetaldehyde
T
J-
-¦
—
¦¦
T
-1-
T
T
¦I-
¦I-
..^L
rfi
rh
rii
rfi
T
T
±
^ ^ ^ ^ ^ ^ ^ -^VVVVV'
MonitoringSite
I I Program Auaraga I I ^ita-^p ai-ifii- Auaraga MD L (0.009 Hg/m3) Risk (0.45 Hg/m3)
-------
Figure 4-2. Inter-Site Variability for Arsenic
1.60
1.40
1.20
1.00
ao
c
c
o
4—'
nj
2 0.80
-------
1.6
Figure 4-3a. Inter-Site Variability for Benzene - Method TO-15
LtJ
On
1.4
1.2
E 1.0
0.8
o
a 0.6
<
0.4
0.2
0.0
ASKY ATKY BLKY BTUT CCKY CHNJ CSNJ DEMI ELNJ GLKY GPCO LAKY LEKY NBIL NBNJ OCOK PXSS ROIL S4MOSEWA SPAZ SPIL TMOKTOOK TROK TVKY YUOK
MonitoringSite
MDL (0.048 Hg/m3) Risk (0.13 ng/m3)
I Program Average
] Site-Specific Average
-------
Figure 4-3b. Inter-Site Variability for Benzene - SNMOC
1.80
1.60
1.40
1.20
m
CUD
o 1-00
*¦4-'
CO
u
° 0.80
(D
00
nj
CD
1
0.60
0.40
0.20
0.00
BMCO BRCO BTUT NBIL PACO RFCO RICO
MonitoringSite
I I Program Auaraga I I ^ita-^p ai-ifii- Auaraga MD L (0.12 Hg/m3) Risk (0.13 Hg/m3)
-------
Figure 4-4a. Inter-Site Variability for 1,3-Butadiene - Method TO-15
LtJ
00
T
T
T
¦¦
¦¦
-
T
¦¦
¦¦
fil
¦¦
¦¦
" rh
-1-
rfi
+
ft
^irh
"
¦¦
n
ASKY ATKY BLKY BTUT CCKY CHNJ CSNJ DEMI ELNJ GLKY GPCO LAKY LEKY NBIL NBNJ OCOK PXSS ROIL S4M0SEWA SPAZ SPIL TMOKTOOK TROK TVKY YUOK
MonitoringSite
I I Program Auaraga I I ^ita-^p ai-ifii- Auaraga MD L (0.029 Hg/m3) Risk (0.03 Hg/m3)
-------
0.16
Figure 4-4b. Inter-Site Variability for 1,3-Butadiene - SNMOC
0.14
0.12
E 0.10
0.08
O
u
g 0.06
<
0.04
0.02
0.00
BMCO
BRCO
I Program Average
BTUT NBIL
MonitoringSite
' Site-Specific Average
PACO RFCO RICO
MDL (0.121 ng/m3) Risk (0.03 ng/
-------
1.2
Figure 4-5. Inter-Site Variability for Carbon Tetrachloride
o
o
(D
I
0.6
0.4
0.2
ASKY ATKY BLKY BTUT CCKY CHNJ CSNJ DEMI ELNJ GLKY GPCO LAKY LEKY NBIL NBNJ OCOK PXSS ROIL S4MOSEWA SPAZ SPIL TMOKTOOK TROK TVKY YUOK
MonitoringSite
MDL(0.106 ng/m3) Risk (0.17 ng/m3)
I Program Average
I Site-Specific Average
-------
Figure 4-6. Inter-Site Variability for /7-Dichlorobenzene
T
¦¦
¦1"
T
j
-r
T
T
¦1"
± 1
ASKY ATKY BLKY BTUT CCKY CHNJ CSNJ DEMI ELNJ GLKY GPCO LAKY LEKY NBIL NBNJ OCOK PXSS ROIL S4MOSEWA SPAZ SPIL TMOKTOOKTROK TVKY YUOK
MonitoringSite
] Program Average I^Z^ZI Site-Specific Average MDL (0.088 ng/m3) Risk (0.091 ng/m3)
-------
Figure 4-7. Inter-Site Variability for 1,2-Dichloroethane
to
T
Average Concentration for
TVKY is 3.5411.66 ng/m3
¦¦
¦¦
T
¦¦
¦¦
|-i-|
T -s-
"
ASKY ATKY BLKY BTUT CCKY CHNJ CSNJ DEMI ELNJ GLKY GPCO LAKY LEKY NBIL NBNJ OCOK PXSS ROIL S4MOSEWA SPAZ SPIL TMOKTOOKTROK TVKY YUOK
MonitoringSite
] Program Average I^Z^ZI Site-Specific Average MDL(0.056 ng/m3) Risk (0.038 ng/m3)
-------
Figure 4-8a. Inter-Site Variability for Ethylbenzene - Method TO-15
T
T
T
T
¦¦
T
T
¦¦
¦I"
-
-1-
_
-L
rii
rti
1
rfi
T
T
T
¦1"
T
¦I"
¦L
Bn
ASKY ATKY BLKY BTUT CCKY CHNJ CSNJ DEMI ELNJ GLKY GPCO LAKY LEKY NBIL NBNJ OCOK PXSS ROIL S4MOSEWA SPAZ SPIL TMOKTOOK TROK TVKY YUOK
MonitoringSite
I I Program Auaraga I I ^ita-^p ai-ifii- Auaraga MD L (0.056 Hg/m3) Risk (0.4 Hg/m3)
-------
0.45
Figure 4-8b. Inter-Site Variability for Ethylbenzene - SNMOC
0.40
0.35
0.30
o 0.25
(D
I
0.15
0.10
0.05
0.00
BMCO
BRCO
I Program Average
BTUT NBIL
MonitoringSite
' Site-Specific Average
PACO
RFCO
MDL (o.i3i ng/m3)
RICO
Risk (0.4 ng/
-------
Figure 4-9. Inter-Site Variability for Formaldehyde
A ^ ^ °*v° & <*vy # *
' 1 Program Average
MonitoringSite
~ Site-Specific Average
MDL (0.012 Hg/m3) Risk (0.077 ng/m3)
-------
Figure 4-10. Inter-Site Variability for Hexachloro-l,3-butadiene
0.35
0.3
0.25
0.2
On
0.15
0.1
0.05
rn rn nh
ft ^l1!
T^i
ASKY ATKY BLKY BTUT CCKY CHNJ CSNJ DEMI ELNJ GLKY GPCO LAKY LEKY NBIL NBNJ OCOK PXSS ROIL S4MOSEWA SPAZ SPIL TMOKTOOKTROK TVKY YUOK
Monitoring Site
i i Program Average i i Site-Specific Average MDL (0.29 ng/m3) Risk (0.045 ng/m3)
-------
160
140
120
100
80
60
40
20
0
Figure 4-11. Inter-Site Variability for Naphthalene
BOMA BTUT BXNY CELA DEMI GLKY GPCO NBIL PRRI PXSS RIVA ROCH RUCA S4MO SEWA SJJCA SKFL UNVT WADC
Monitoring Site
I I Program Auaraga I I ^ita-^p ai-ifii- Auaraga MD L (0.413 ng/m3) Risk (29 ng/m3)
-------
Figure 4-12. Inter-Site Variability for Nickel
00
3.00
2.50
0.50
0.00
ASKY-M BAKY BLKY BOMA BTUT CCKY GLKY GPCO LEKY NBIL PAFL PXSS S4M0 SEWA SJJCA UNVT OCOK TMOK TOOK TROK YUOK
MonitoringSite
i i Program PM10 Average i i Program TSP Average i i Site-Specific Average Teflon MDL —— —Quartz MDL Risk
(0.183 ng/m3) (0.713 ng/m3) (2.1 ng/m3)
-------
4.2.2.2 Quarterly Variability Analysis
Figures 4-13 through 4-24b 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. The published MDL
from the ERG laboratory is also plotted on each graph, similar to the inter-site variability graphs.
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 same is also true for the air toxics measured by both Method
TO-15 and the concurrent SNMOC method.
"Missing" quarterly average concentrations in the figures for the pollutants of interest can
be attributed to several reasons. First, some of the program-wide pollutants of interest were
infrequently detected in some quarters and may have a quarterly average concentration of zero,
as a result of the substitution of zeros for non-detects. Thus, the quarterly average concentration
is not missing but rather not visible on the graph. Another reason for missing quarterly averages
in the figures 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. Lastly, 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.
Most of the program-level pollutants of interest were detected year-round. Few were
detected less frequently. For example, hexachloro-l,3-butadiene was not detected at every site,
as shown in Figure 4-22. This pollutant was not detected at DEMI or SEW A, and was detected in
only two quarters at GLKY and SPAZ. However, comparing the quarterly average
concentrations for sites with four valid quarterly averages in a year may reveal a temporal trend
for other 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).
4-49
-------
The quarterly average concentration 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. The quarterly average graphs may also reveal if
concentrations measured at a particular site are significantly lower than other sites. These graphs
may also reveal when there is very little variability in the quarterly averages across other sites.
Inter-state trends may also be revealed.
Observations from Figures 4-13 through 4-24b include the following:
• Figure 4-13 presents the site-specific quarterly average concentrations of
acetaldehyde. Twenty-five of the 32 sites sampling this pollutant have four quarterly
average concentrations available. For many of these sites, the quarterly average
concentrations for the second (7) and third (9) quarters are higher than the other
quarterly averages. These can be seen by the red and green bars extending higher in
Figure 4-13 than the others; examples include CSNJ, ELNJ, GPCO, and the Tulsa,
Oklahoma sites. This figure also shows that the highest quarterly average
concentrations were calculated for BTUT (fourth quarter) and SPIL (first quarter).
Other sites with quarterly average acetaldehyde concentrations greater than 3.0 |ig/m3
include ELNJ, GPCO, NBIL, NBNJ, and PXSS. Note that three of BTUT's four
quarterly average concentrations are greater than 3 |ig/m3. Sites with quarterly
average acetaldehyde concentrations less than 0.5 |ig/m3 include BMCO, BRCO,
PACO, RFCO, RICO, and SEW A; five of these sites are located in Garfield County,
Colorado.
• Figures 4-14a and 4-14b present the quarterly average concentrations of arsenic for
sites sampling speciated metals, first for PMio then for TSP. Figure 4-14b shows that
for each of the five sites sampling TSP metals, the third quarter average concentration
was the highest quarterly average concentration for this pollutant. A similar
observation can be made from Figure 4-14a for the sites sampling PMio metals, but to
a lesser extent. Of the 16 sites shown, 13 have four quarterly average concentrations
of arsenic available. For seven of these 13 sites, the third quarter average
concentration was the highest quarterly average concentration. These figures show
that ASKY-M's third quarter average concentration (1.91 ng/m3) is considerably
higher than most of the others shown; only five other sites have quarterly average
arsenic concentrations greater than 1 ng/m3 (BAKY, NBIL, PAFL, S4MO, and
TROK).
• Figures 4-15a and 4-15b present the quarterly average concentrations for sites
sampling benzene, first for Method TO-15 then for SNMOC. Of the 32 sites sampling
benzene with these methods, 24 have four quarterly average concentrations available.
The first quarter average is mostly commonly the highest quarterly average
concentration among the sites, followed by the third quarter average concentration.
TVKY's third quarter average concentration (1.87 |ig/m3) is the highest quarterly
average concentration shown, although PXSS, SPAZ, and PACO each have at least
one quarterly average concentration greater than 1.5 |ig/m3.
4-50
-------
• Figures 4-16a and 4-16b present the quarterly average concentrations for sites
sampling 1,3-butadiene, first for Method TO-15 then for SNMOC. For sites sampling
this pollutant with the SNMOC method, there are few quarterly average
concentrations shown; some of these are due non-detects (and substituted zeros)
rather than not being able to calculated them. For example, PACO is the only site
with four quarterly averages of 1,3-butadiene, but two of them are zero due to non-
detects. For sites sampling with Method TO-15, there are no zero quarterly averages,
only a few sites for which quarterly averages could not be calculated (such as CCKY
for the fourth quarter). Note, though, the MDLs between the two methods are quite
different (0.029 |ig/m3 for TO-15 and 0.105 |ig/m3 for SNMOC). Of the 24 sites for
which four quarterly average concentrations of 1,3-butadiene could be calculated, the
third quarter average is mostly commonly the highest quarterly average concentration
among the sites (11). This is a little different from previous reports, where 1,3-
butadiene tended to be higher during the colder months of the year. For 2014, just as
many sites have their highest quarterly average 1,3-butadiene concentration for the
first or fourth quarters (12) as they do the second or third quarters (12) of the year.
Yet for most sites, the differences among the quarterly averages are relatively small
(less than 0.1 |ig/m3). The site with the highest quarterly average concentration of
1,3-butadiene is TVKY (0.86 |ig/m3, for the third quarter). No other sites have
quarterly average concentrations of this pollutant greater than 0.5 |ig/m3 and only a
few sites have quarterly averages greater than 0.25 |ig/m3.
• Concentrations of some pollutants had a tendency to be higher in one quarter over the
others but the differences among the quarters were so small, it makes little difference.
For instance, Figure 4-17 shows that the second and third quarter average
concentrations are highest for most of the sites sampling carbon tetrachloride. Of the
23 sites with four available quarterly average concentrations of carbon tetrachloride,
21 have their highest quarterly average concentration calculated for either the second
or third quarters of 2014 (with only two highest for the first quarter, LAKY and
TVKY, and none highest for the fourth quarter). However, the quarterly average
concentrations for most monitoring sites vary by less than 0.15 |ig/m3. The site with
the largest difference in its quarterly average concentrations of carbon tetrachloride is
BLKY, as its second quarter average concentration is considerably higher than the
others (0.90 |ig/m3). Only two other NMP sites have quarterly average concentrations
greater 0.75 |ig/m3 (LAKY and TVKY); all four of TVKY's quarterly average
concentrations of carbon tetrachloride are greater than 0.75 |ig/m3.
• Figure 4-18 presents the quarterly average concentrations for /;-dichlorobenzene.
Note that most of the quarterly average concentrations shown are well below the
MDL shown in the figure. Recall from the previous section that the detection rate for
this pollutant is relatively low, around 44 percent, and most of the measured
detections are less than the MDL. Only five sites have at least one quarterly average
concentration greater than the MDL. For BTUT and TOOK, the third quarter average
is greater than the MDL. For the other three sites (S4MO, PXSS, and SPAZ), all four
quarterly average concentrations are well above the MDL. Of the 23 sites for which
four quarterly average concentrations could be calculated, the averages for the third
quarter are most frequently the highest (10), but of varying degrees of difference. For
example, LEKY's third quarter average concentration of p-dichlorobenzene is more
4-51
-------
than twice its next highest quarterly average, while the quarterly averages vary by
very little for ATKY.
As shown in Figure 4-19, most of the quarterly average concentrations for NMP sites
measuring 1,2-dichloroethane are similar to the MDL for this pollutant (0.056 |ig/m3).
The exceptions to this are all for the Calvert City, Kentucky sites. For the Calvert
City sites, most of the quarterly average concentrations of 1,2-dichloroethane fall
between 0.25 |ig/m3 and 1.25 |ig/m3, while all but one of TVKY's are greater than
this range, particularly the first quarter average concentration (6.81 |ig/m3).
Figures 4-20a and 4-20b present the quarterly average concentrations for sites
sampling ethylbenzene, first for Method TO-15 then for SNMOC. Of the 24 sites for
which four quarterly average concentrations could be calculated, the third quarter
averages were the highest quarterly average for 15 of them. The sites with the highest
quarterly average concentrations of ethylbenzene are PXSS and SPAZ, the only two
sites with quarterly averages greater than 0.75 |ig/m3 (for the first and fourth quarters
of 2014 for both sites). Only four additional sites (DEMI, GPCO, TOOK, and TROK)
have at least one quarterly average concentration greater than 0.5 |ig/m3. All of the
quarterly average concentrations shown for sites sampling with Method TO-15 are
greater than the MDL for this pollutant while several of the quarterly averages shown
for sites sampling with the SNMOC method are less than the MDL. Note, however,
that the MDL for the SNMOC method (0.131 |ig/m3) is more than twice the MDL for
Method TO-15 (0.056 |ig/m3).
Figure 4-21 presents the quarterly average concentrations of formaldehyde. Quarterly
average concentrations of formaldehyde tend to be highest for the summer months,
based on previous reports. Figure 4-21 shows that 25 of the 32 sites sampling
formaldehyde have four quarterly average concentrations available. Of these, 17
exhibited the highest quarterly average concentration for the third quarter (which
includes samples collected from July through September, and is shown in green).
Another five sites exhibited their highest quarterly formaldehyde average for the
second quarter (from April through June), which is shown in red. Thus, it appears that
formaldehyde concentrations tended to be highest during the warmer months of 2014
too, although there are exceptions. Several sites have quarterly average concentrations
greater than 5 |ig/m3, although BTUT is the only site with more than one (second,
third, and fourth quarters).
Figure 4-22 presents the quarterly average concentrations of hexachloro-1,3-
butadiene. The MDL for this pollutant is 0.29 |ig/m3 and all quarterly average
concentrations shown are less than 0.06 |ig/m3. Recall from the previous section that
only one concentration of hexachloro-1,3-butadiene measured in 2014 is greater than
the detection limit and that the detection rate is 23 percent. This indicates that a large
number of substituted zeroes are included in the quarterly average calculations,
including some sites where this pollutant was not detected at all (DEMI and SEW A).
Of the 19 sites sampling PAHs, four quarterly average concentrations of naphthalene
could be calculated for 15 of them. Of these, the first quarter average concentration
tended to be the highest (10 sites). However, the highest quarterly average
4-52
-------
concentrations shown in Figure 4-23 were calculated for DEMI's and NBIL's third
quarter average concentrations (203.35 ng/m3 and 164.80 ng/m3, respectively). The
quarterly average naphthalene concentrations shown vary considerably, from
UNVT's third quarter average concentration (4.49 ng/m3) to NBIL's third quarter
average concentration (203.35 ng/m3). GLKY and UNVT are the only two sites with
quarterly averages less than 25 ng/m3 while NBIL and DEMI are the only two with
quarterly averages greater than 150 ng/m3. Note that several NMP sites do not have a
second quarter average concentration for naphthalene shown in Figure 4-23. This is a
result of the invalidation of several PAH samples in April and early May resulting
from laboratory equipment issues. While this issue affected all sites that sampled
naphthalene, some sites had enough valid samples for the second quarter to still meet
the minimum criteria for calculating a quarterly average concentration while others
did not. The number of samples affected varies from two samples to four samples.
• Figures 4-24a and 4-24b present the quarterly average concentrations of nickel for
sites sampling speciated metals, first for PMio then for TSP. The quarterly average
concentrations of nickel do not exhibit an identifiable seasonal trend. Some sites'
quarterly average concentrations vary considerably (SEWA) while others do not
(OCOK). SEWA is the only site with a quarterly average concentration of nickel
greater than 3 ng/m3, although three additional sites have quarterly averages greater
than 2.5 ng/m3 (ASKY-M, BOMA, and TOOK). Several sites have quarterly average
concentrations less than 0.5 ng/m3, although all of these are sampling PMio metals.
4-53
-------
Figure 4-13. Comparison of Average Quarterly Acetaldehyde Concentrations
5.0
4.5
4.0
ST 3.5
M
=L
e-V ^ / **V0VV° & 4?4?4? * f/////
Monitoring Site
1st Quarter 2nd Quarter 3rd Quarter 4th Quarter MDL
(0.009 ng/m3)
-------
Figure 4-14a. Comparison of Average Quarterly Arsenic (PMio) Concentrations
2.00
ASKY-M BAKY BLKY BOMA BTUT CCKY GLKY GPCO LEKY NBIL PAFL PXSS S4MO SEWA SJJCA UNVT
Monitoring Site
1st Quarter 2nd Quarter 3rd Quarter 4th Quarter Quartz MDL Teflon MDL
(0.058 ng/m3) (0.221 ng/m3)
-------
Figure 4-14b. Comparison of Average Quarterly Arsenic (TSP) Concentrations
2.00
1.75
1.50
m
OCOK
TMOK
TOOK
TROK
YUOK
3rd Quarter
1st Quarter
Quartz MDL
(0.058 ng/m3)
-------
Figure 4-15a. Comparison of Average Quarterly Benzene (Method TO-15) Concentrations
2.0
ASKY ATKY BLKY BTUT CCKY CHNJ CSNJ DEMI ELNJ GLKY GPCO LAKY LEKY NBIL NBNJ OCOK PXSS ROIL S4MOSEWA SPAZ SPIL TMOKTOOK TROK TVKY YUOK
Monitoring Site
1st Quarter 2nd Quarter 3rd Quarter 4th Quarter MDL
(0.048 ng/m3)
-------
Figure 4-15b. Comparison of Average Quarterly Benzene (SNMOC) Concentrations
00
2.5
2.0
m
BMCO
BRCO
PACO
3rd Quarter
1st Quarter
(0.121 ng/m3)
-------
Figure 4-16a. Comparison of Average Quarterly 1,3-Butadiene (Method TO-15) Concentrations
1.0
0.9
0.8
0.7
£ 0.6
VO
0.4
0.3
0.2
0.1
0.0
L
ASKY ATKY BLKY BTUT CCKY CHNJ CSNJ DEMI ELNJ GLKY GPCO LAKY LEKY NBIL NBNJ OCOK PXSS ROIL S4MOSEWA SPAZ SPIL TMOKTOOK TROK TVKY YUOK
Monitoring Site
11st Quarter
I 2nd Quarter
I 3rd Quarter
I 4th Quarter
MDL
(0.029 ng/m3)
-------
Figure 4-16b. Comparison of Average Quarterly 1,3-Butadiene (SNMOC) Concentrations
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
BMCO
1st Quarter
I
BRCO
I 2nd Quarter
PACO
Monitoring Site
3rd Quarter
RFCO
RICO
I 4th Quarter
MDL
(0.105 ug/m3'
-------
Figure 4-17. Comparison of Average Quarterly Carbon Tetrachloride Concentrations
On
1.25
1.00
ASKY ATKY BLKY BTUT CCKY CHNJ CSNJ DEMI ELNJ GLKY GPCO LAKY LEKY NBIL NBNJ OCOK PXSS ROIL S4MOSEWA SPAZ SPIL TMOKTOOK TROK TVKY YUOK
3rd Quarter
1st Quarter
(0.106 ng/m3)
-------
Figure 4-18. Comparison of Average Quarterly />-Dichlorobenzene Concentrations
On
to
0.35
0.30
0.25
m
ASKY ATKY BLKY BTUT CCKY CHNJ CSNJ DEMI ELNJ GLKY GPCO LAKY LEKY NBIL NBNJ OCOK PXSS ROIL S4MOSEWA SPAZ SPIL TMOKTOOK TROK TVKY YUOK
Monitoring Site
1st Quarter 2nd Quarter 3rd Quarter 4th Quarter MDL
(0.088 ng/m3)
-------
Figure 4-19. Comparison of Average Quarterly 1,2-Dichloroethane Concentrations
On
LtJ
o
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
1st Quarter Average
Concentration for
TVKY is 6.81 ng/m3.
Jl
U
ASKY ATKY BLKY BTUT CCKY CHNJ CSNJ DEMI ELNJ GLKY GPCO LAKY LEKY NBIL NBNJ OCOK PXSS ROIL S4MOSEWA SPAZ SPIL TMOKTOOK TROK TVKY YUOK
Monitoring Site
11st Quarter
I 2nd Quarter
I 3rd Quarter
I 4th Quarter
MDL
(0.056 ng/m3)
-------
Figure 4-20a. Comparison of Average Quarterly Ethylbenzene (Method TO-15) Concentrations
On
1.25
1.00
0.75
o
0.50
0.25
0.00
ASKY ATKY BLKY BTUT CCKY CHNJ CSNJ DEMI ELNJ GLKY GPCO LAKY LEKY NBIL NBNJ OCOK PXSS ROIL S4MOSEWA SPAZ SPIL TMOKTOOK TROK TVKY YUOK
Monitoring Site
11st Quarter
I 2nd Quarter
I 3rd Quarter
I 4th Quarter
MDL
(0.056 ng/m3)
-------
Figure 4-20b. Comparison of Average Quarterly Ethylbenzene (SNMOC) Concentrations
1.25
1.00
0.75
On
0.50
0.25
0.00
I
BMCO
11st Quarter
BRCO
I 2nd Quarter
PACO
Monitoring Site
3rd Quarter
RFCO
I 4th Quarter
RICO
MDL
(0.131 ng/m3;
-------
Figure 4-21. Comparison of Average Quarterly Formaldehyde Concentrations
10.0
9.0
8.0
ST" 7.0
CUD
=L
Monitori ngSite
1st Quarter 2nd Quarter 3rd Quarter 4th Quarter MDL
(0.012 ng/m3)
-------
Figure 4-22. Comparison of Average Quarterly Hexachloro-l,3-butadiene Concentrations
On
-J
0.35
0.30
0.25
0.20
a 0.15
<
a
o.io
li.uLhllllh^ JjLlnhhil
ASKY ATKY BLKY BTUT CCKY CHNJ CSNJ DEMI ELNJ GLKY GPCO LAKY LEKY NBIL NBNJ OCOK PXSS ROIL S4MOSEWA SPAZ SPIL TMOKTOOK TROK TVKY YUOK
Monitoring Site
1st Quarter 2nd Quarter 3rd Quarter 4th Quarter MDL
(0.290 ng/m3)
-------
Figure 4-23. Comparison of Average Quarterly Naphthalene Concentrations
On
00
250
200
£
"cJ>
150
o
u
-------
Figure 4-24a. Comparison of Average Quarterly Nickel (PMio) Concentrations
On
VO
3.50
ASKY-M BAKY BLKY BOMA BTUT CCKY GLKY GPCO LEKY NBIL PAFL PXSS S4MO SEWA SJJCA UNVT
Monitoring Site
1st Quarter 2nd Quarter 3rd Quarter 4th Quarter Quartz MDL Teflon MDL
(0.713 ng/m3) (0.183 ng/m3)
-------
Figure 4-24b. Comparison of Average Quarterly Nickel (TSP) Concentrations
o
3.50
3.00
OCOK
TMOK
TOOK
TROK
YUOK
3rd Quarter
1st Quarter
Quartz MDL
(0.713 ng/m3)
-------
5.0 Sites in Arizona
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS and UATMP sites in Arizona, and integrates these
concentrations with emissions, meteorological, and risk information. Data generated by sources
other than ERG are not included in the data analyses contained in this report. Readers are
encouraged to refer to Sections 1 through 4 and the glossary (Appendix P) for detailed
discussions and definitions regarding the various data analyses presented below.
5.1 Site Characterization
This section characterizes the Arizona monitoring sites by providing geographical and
physical information about the location of the sites and the surrounding areas. 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 Arizona monitoring sites are located in Phoenix, Arizona. Figures 5-1 and 5-2 are
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 2011 NEI for point sources, version 2. 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 in order to
emphasize emissions sources within the boundary. Table 5-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates.
5-1
-------
Figure 5-1. Phoenix, Arizona (PXSS) Monitoring Site
W, Highland'Avei
-W Highland Ave
, i*-
w Turney Ave
-------
Figure 5-2. South Phoenix, Arizona (SPAZ) Monitoring Site
4-» ' I
^ I
00 ~'i
.W Marguerite Ave
~r
W-Tamarisk Ave
W .Wler. Ave
¦
- E Atlan ta Ave-~«aaw|
I |br« p» »
|i'-vTcWo;
£>/r ^ »
IA/ D.- ... ... 0.1
¦WAtlanta-Ave**
g e o n a rff £ iPtfV^Ave
W.Sun'and Ave
-------
Figure 5-3. NEI Point Sources Located Within 10 Miles of PXSS and SPAZ
112-25'0-W
Source Category Group (No. of Facilities)
* Aerospace/Aircraft Manufacturing Facility (2)
"1" Airport/Airline/Airport Support Operations (36)
$ Asphalt Production/Hot Mix Asphalt Plant (2)
X Battery Manufacturing Facility (1)
B Bulk Terminals/Bulk Plants (3)
C Chemical Manufacturing Facility (4)
6 Electrical Equipment Manufacturing Facility (6)
f Electricity Generation via Combustion (4)
E Electroplating, Plating, Polishing, Anodizing, and Coloring (1)
I Foundries, Iron and Steel (1)
A Foundries, Non-ferrous (1)
(B) Metal Can, Box, and Other Metal Container Manufacturing (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 (6)
R Plastic, Resin, or Rubber Products Plant (5)
X Rail Yard/Rail Line Operations (3)
W Woodwork, Furniture, Millwork & Wood Preserving Facility (2)
Note: Due to facility density and collocation, the total facilities
Legend displayed may not represent all facilities within the area of interest.
PXSS NATTS site SPAZ UATMP site O 10 mile radius County boundary
/
~~i^ ~e-
i
it
I Maricopa
/ County
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
25,952
Central Ave, south of
W Tamarisk Ave
1AADT reflects 2010 data for PXSS and 2011 data for SPAZ (AZ DOT, 2016)
BOLD ITALICS = EPA-designated NATTS Site
-------
PXSS is located in central Phoenix. Figure 5-1 shows that PXSS is located in a highly
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 Avenue 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, CBSA, 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 effect concentrations measured at a given monitoring site. PXSS
experiences a higher traffic volume compared to SPAZ, although the traffic volumes near both of
these 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 the monitoring sites (West
Camelback Road and Central Avenue).
5.2 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
sites in Arizona on sample days, as well as over the course of the year.
5.2.1 Meteorological Summary
In order to provide an overview of the meteorological conditions experienced at each
monitoring site, hourly meteorological data for 2014 were retrieved and sample day and full-year
averages developed for temperature, dew point, relative humidity, pressure, and wind speed.
5-6
-------
Weather data collected from the actual monitoring site(s) were obtained from AQS, where
available. If site-specific weather data were not available in AQS, then data were obtained from
NCDC for the NWS weather station located closest to the monitoring site(s), as described in
Section 3.4.2. For the Arizona sites, site-specific data were available for some, but not all, of the
parameters in Table 5-2. For PXSS, pressure, humidity, and wind information was available in
AQS; for SPAZ, only wind data was available. Data from the NWS weather station at Phoenix
Sky Harbor International Airport (WBAN 23183) was used for the remaining parameters. The
Phoenix Sky Harbor weather station is located 7.5 miles southeast of PXSS and 4.5 miles east-
northeast of SPAZ. A map showing the distance between each Arizona monitoring site and the
closest NWS weather station is provided in Appendix R. These data were used to determine how
meteorological conditions on sample days vary from conditions experienced throughout the year.
Table 5-2. Average Meteorological Conditions near the Arizona Monitoring Sites
Average
Type1
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(in Hg)
Average
Station
Pressure
(in Hg)
Prevailing
Wind
Direction
Average
Scalar Wind
Speed
(kt)
Phoenix, Arizona - PXSS2
Sample
Days
74.8
37.0
35.2
29.86
28.73
2.7
(72)
±0.8
±0.8
± 1.0
±0.01
±0.01
wsw
±0.1
75.3
37.9
35.4
29.87
28.74
2.5
2014
±0.4
±0.3
±0.4
± <0.01
±<0.01
wsw
±<0.1
South Phoenix, Arizona - SPAZ3
Sample
Days
76.8
38.0
31.2
29.85
28.67
2.4
(30)
± 1.1
± 1.3
± 1.6
±0.01
±0.01
w
±0.1
77.2
37.9
30.0
29.87
28.74
2.2
2014
±0.3
±0.3
±0.4
± <0.01
±<0.01
w
±<0.1
'Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
2Temperature, humidity, and wind parameters were measured at PXSS. Hie remaining information was obtained from the
closest NWS weather station located at Sky Harbor International Airport, WBAN 23183.
3Only wind parameters were measured at SPAZ. The remaining information was obtained from the closest NWS weather station
located at Sky Harbor International Airport, WBAN 23183.
5-7
-------
Table 5-2 presents average temperature, average dew point temperature, average relative
humidity, average station and sea level pressure, and wind information (average scalar wind
speed and prevailing wind direction) for days on which samples were collected and for all of
2014. Also included in Table 5-2 is the 95 percent confidence interval for each parameter. As
shown in Table 5-2, average meteorological conditions on sample days were representative of
average weather conditions experienced throughout the year at each site. The greatest difference
between the sample day and full-year averages was calculated for average relative humidity for
SPAZ, although the difference is not statistically significant.
The number of sample days for each site is provided in Table 5-2. Samples were
collected on a l-in-6 day schedule at PXSS while samples were collected on a l-in-12 day
schedule at SPAZ, yielding roughly half the number of collection events; thus, the number of
observations included in each sample day calculation for SPAZ is less. The difference in the
number of sample days is reflected in the larger confidence intervals for SPAZ (the fewer
observations, generally the larger the confidence intervals).
These sites experienced the warmest temperatures among NMP sites in 2014, based on
both the full-year and sample day average temperatures. These sites also experienced the lowest
relative humidity levels among all NMP sites in 2014, based on both the full-year and sample
day average relative humidity levels.
5.2.2 Wind Rose Comparison
Hourly surface wind data were also uploaded into a wind rose software program to
produce customized wind roses, as described in Section 3.4.2. A wind rose shows the frequency
of wind directions using "petals" positioned around a 16-point compass, and uses different colors
to represent wind speeds. Figure 5-4 presents two wind roses for the PXSS monitoring site. The
first is a wind rose representing wind observations for all of 2014 and the second is a wind rose
representing wind observations for days on which samples were collected in 2014. These are
used to identify the predominant wind speed and direction for 2014 and to determine if wind
observations on sample days were representative of conditions experienced over the entire year.
Figure 5-5 presents the full-year and sample day wind roses for SPAZ.
5-8
-------
Figure 5-4. Wind Roses for the Wind Data Collected at PXSS
2014 Wind Rose Sample Day Wind Rose
NORTH
NORTH
WEST
;WEST
WIND SPEED
(Knots)
WIND SPEED
(Knots)
SOUTH
SOUTH
Calms: 1.70%
Calms: 0.93%
Observations from Figure 5-4 for PXSS include the following:
• The 2014 wind rose shows that wind speeds were generally light near PXSS, with
wind speeds mostly in the 1 knot to 4 knot range, although the calm rate was less than
2 percent. West-southwesterly winds were the most commonly observed wind
direction at PXSS, although easterly and westerly winds were also observed
frequently. Winds from the northwest quadrant and those with from the south-
southeast to south-southwest were infrequently observed near PXSS.
• The sample day wind rose resembles the full-year wind rose, exhibiting both light
winds and a similar wind direction pattern, indicating that winds on sample days were
representative of those observed throughout the year near PXSS, even though west-
southwesterly winds were observed even more frequently on sample days.
5-9
-------
Figure 5-5. Wind Roses for Wind Data Collected at SPAZ
2014 Wind Rose Sample Day Wind Rose
WEST
8%. \
t
: east;
WIND SPEED
(Knots)
¦I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 10.23%
WIND SPEED
(Knots)
¦I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 9.19%
Observations from Figure 5-5 for SPAZ include the following:
• The 2014 wind rose shows that wind speeds were generally light near SPAZ, with a
calm rate was of 10 percent. West winds were the most commonly observed wind
direction, although easterly and east-southeasterly winds were also observed
frequently. Conversely, winds with a northerly component were infrequently
observed near SPAZ.
The sample day wind rose resembles the full-year wind rose, exhibiting both light
winds and a similar wind direction pattern, indicating that winds on sample days were
representative of those observed throughout the year near SPAZ.
5-10
-------
5.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each
Arizona monitoring 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. 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 5-3. 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 5-3. It is important to note which pollutants 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.
Observations from Table 5-3 include the following:
• The number of pollutants failing screens varied significantly between the two
monitoring sites, which is expected given the difference in pollutants measured at
each site.
• Concentrations of 15 pollutants failed at least one screen for PXSS; 72 percent of
concentrations for these 15 pollutants were greater than their associated risk screening
value (or failed screens).
• Concentrations of 11 pollutants contributed to 95 percent of failed screens for PXSS
and therefore were identified as pollutants of interest for PXSS. These 11 include two
carbonyl compounds, seven VOCs, one PMio metal, and one PAH.
• PXSS failed the highest number of screens (563) among all NMP sites (refer to
Table 4-8 of Section 4.2). However, the failure rate for PXSS, when incorporating all
pollutants with screening values, is relatively low, at 23 percent. This is due primarily
to the relatively high number of pollutants sampled for at this site, as discussed in
Section 4.2 and above.
• Concentrations of six pollutants failed screens for SPAZ; approximately 91 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 60 percent failure rate for ethylbenzene to a
100 percent failure rate for four pollutants; for PXSS, the percentage of screens failed
for each individual pollutant is more varied, ranging from 4 percent for
benzo(a)pyrene to 100 percent for four pollutants.
5-11
-------
• All 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 VOCs measured at these sites, benzene was detected in all valid samples and
failed 100 percent of screens for each site. Other VOCs, such as carbon tetrachloride,
1,2-dichloroethane, and 1,3-butadiene were detected frequently and also failed the
majority of screens. Formaldehyde was detected in all of the valid samples collected
at PXSS and also failed 100 percent of screens for this site. Acetaldehyde was also
detected in all carbonyl compound samples collected and failed all but one screen.
Table 5-3. Risk-Based Screening Results for the Arizona Monitoring Sites
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Phoenix, Arizona - PXSS
Benzene
0.13
61
61
100.00
10.83
10.83
Formaldehyde
0.077
61
61
100.00
10.83
21.67
Acetaldehyde
0.45
60
61
98.36
10.66
32.33
Carbon Tetrachloride
0.17
60
61
98.36
10.66
42.98
1.3 -Butadiene
0.03
59
60
98.33
10.48
53.46
1,2-Dichloroethane
0.038
53
53
100.00
9.41
62.88
Naphthalene
0.029
53
59
89.83
9.41
72.29
Arsenic (PMio)
0.00023
50
60
83.33
8.88
81.17
/?-Dichlorobcnzcnc
0.091
42
57
73.68
7.46
88.63
Ethylbenzene
0.4
29
61
47.54
5.15
93.78
Hexachloro-1,3 -butadiene
0.045
16
17
94.12
2.84
96.63
Nickel (PMio)
0.0021
11
60
18.33
1.95
98.58
Manganese (PMio)
0.03
5
60
8.33
0.89
99.47
Benzo(a)pyrene
0.00057
2
46
4.35
0.36
99.82
1,2-Dibromoethane
0.0017
1
1
100.00
0.18
100.00
Total
563
778
72.37
South Phoenix, Arizona - SPAZ
Benzene
0.13
30
30
100.00
19.61
19.61
Carbon Tetrachloride
0.17
30
30
100.00
19.61
39.22
1.3 -Butadiene
0.03
29
29
100.00
18.95
58.17
/?-Dichlorobcnzcnc
0.091
25
28
89.29
16.34
74.51
1,2-Dichloroethane
0.038
21
21
100.00
13.73
88.24
Ethylbenzene
0.4
18
30
60.00
11.76
100.00
Total
153
168
91.07
5-12
-------
5.4 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.
• 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 are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at PXSS and SPAZ are provided in Appendices J, L, M, and N.
5.4.1 2014 Concentration Averages
Quarterly and annual concentration averages 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 the entire year of sampling. Annual averages were
calculated for pollutants where three valid quarterly averages could be calculated 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 5-4, 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.
5-13
-------
Table 5-4. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Arizona Monitoring Sites
# of
Measured
1st
2nd
3rd
4th
Detections
Total
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs.
# >MDL
# of
Samples
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Phoenix, Arizona - PXSS
3.39
1.88
1.96
2.91
2.52
Acetaldehyde
61/61
61
±0.67
±0.38
±0.49
±0.54
±0.29
1.53
0.49
0.55
1.60
1.05
Benzene
61/61
61
±0.40
±0.11
±0.16
±0.42
±0.20
0.34
0.07
0.07
0.31
0.20
1.3 -Butadiene
60/59
61
±0.12
±0.02
±0.02
±0.10
±0.05
0.58
0.64
0.66
0.57
0.61
Carbon Tetrachloride
61/61
61
±0.04
±0.04
±0.03
±0.06
±0.02
0.26
0.10
0.11
0.20
0.17
/?-Dichlorobcnzcnc
57/46
61
±0.07
±0.03
±0.04
±0.05
±0.03
0.07
0.09
0.05
0.10
0.08
1,2-Dichloroethane
53/48
61
±0.03
±0.01
±0.01
±0.02
±0.01
0.82
0.30
0.35
0.78
0.57
Ethylbenzene
61/60
61
±0.25
±0.09
±0.11
±0.20
±0.10
3.80
2.99
3.70
3.43
3.46
Formaldehyde
61/61
61
±0.57
±0.39
±0.43
±0.49
±0.24
0.02
0.02
0.02
0.02
0.02
Hexachloro-1,3 -butadiene
17/0
61
±0.02
±0.02
±0.02
±0.02
±0.01
0.72
0.43
0.40
0.64
0.55
Arsenic (PMi0)a
60/52
60
±0.28
±0.25
±0.10
±0.25
±0.11
126.33
42.09
42.08
96.46
78.25
Naphthalene3
59/59
59
± 44.42
± 11.79
± 11.24
±22.50
± 15.71
South Phoenix, Arizona
-SPAZ
1.49
0.56
0.67
1.63
1.09
Benzene
30/30
30
±0.54
±0.21
±0.26
±0.36
±0.24
0.31
0.08
0.09
0.28
0.19
1.3 -Butadiene
29/29
30
±0.17
±0.04
±0.04
±0.08
±0.06
0.56
0.62
0.63
0.58
0.60
Carbon Tetrachloride
30/30
30
±0.11
±0.04
±0.05
±0.06
±0.03
0.26
0.12
0.18
0.28
0.21
/?-Dichlorobcnzcnc
28/25
30
±0.10
±0.07
±0.12
±0.08
±0.05
0.06
0.07
0.02
0.06
0.06
1,2-Dichloroethane
21/19
30
±0.04
±0.01
±0.02
±0.04
±0.01
0.79
0.39
0.42
0.81
0.60
Ethylbenzene
30/30
30
±0.30
±0.22
±0.17
±0.23
±0.13
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
5-14
-------
Observations for PXSS from Table 5-4 include the following:
• The pollutants of interest with the highest annual average concentrations for PXSS
are formaldehyde (3.46 ± 0.24 |ig/m3), acetaldehyde (2.52 ± 0.29 |ig/m3), and
benzene (1.05 ± 0.20 |ig/m3). These are the only pollutants of interest with annual
average concentrations greater than 1 |ig/m3 for this site.
• The first and fourth quarter average concentrations for benzene and 1,3-butadiene are
significantly greater than the second and third quarter average concentrations,
indicating 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 NMP report. A review of the benzene data shows that of the 14
benzene concentrations greater than 1.50 |ig/m3 measured at PXSS, all 14 were
measured during the first or fourth quarters of 2014 and the five benzene
concentrations greater than 2.50 |ig/m3 were measured at PXSS in either January or
December. For 1,3-butadiene, the 22 highest concentrations were measured during
the first or fourth quarters of 2014, with the four highest measured at PXSS in either
January or December.
• The quarterly averages for /;-dichlorobenzene and ethylbenzene exhibit a similar
seasonal tendency. Acetaldehyde's quarterly average concentrations reflect a similar
tendency, although the differences are less significant.
• The quarterly average concentrations of hexachloro-l,3-butadiene appear to be the
same across Table 5-4 (0.02 ± 0.02 |ig/m3). Increasing the number of decimal places
shows these quarterly averages vary between 0.017 |ig/m3 and 0.022 |ig/m3. Note
however, that the number of measured detections is only 17 out of 61 valid samples
and that none of these are greater than the MDL for this pollutant.
• Concentrations of naphthalene are similar to several of the VOCs in that the
concentrations measured are higher during the first and fourth quarters of 2014, as
shown by the quarterly average concentrations in Table 5-4. All but one of the 13
naphthalene concentrations greater than 100 ng/m3 were measured at PXSS during
the first or fourth quarters of 2014; conversely, all but three of the 26 naphthalene
concentrations less than 50 ng/m3, all but three were measured during the second and
third quarters of 2014.
• Arsenic is the only metal pollutant of interest for PXSS. The first and fourth quarter
average concentrations of arsenic are greater than the other two quarterly average
concentrations, although not significantly so. Of the seven concentrations of arsenic
greater than 1 ng/m3 measured at PXSS, only one was not measured in either January
or December (a measurement tied for second highest was measured in May).
5-15
-------
Observations for SPAZ from Table 5-4 include the following:
• The pollutant of interest with the highest annual average concentration for SPAZ is
benzene (1.09 ± 0.24 |ig/m3), which is the only pollutant of interest with an annual
average concentration greater than 1 |ig/m3. The annual average concentration of
benzene for SPAZ is similar to the annual average benzene concentration for PXSS.
• Similar to PXSS, benzene and 1,3-butadiene concentrations were highest during the
first and fourth quarters of 2014 at SPAZ. This is also true for ethylbenzene and
/;-dichlorobenzene. However, the confidence intervals calculated for these averages
indicate that the difference in the quarterly averages are not significant.
• The first quarter average concentration for most of the pollutants of interest for SPAZ
have relatively large confidence intervals, indicating measurements of these
pollutants were more variable during the first three months of 2014. The maximum
concentrations of benzene, 1,3-butadiene, ethylbenzene, and 1,2-dichloroethane were
all measured on January 11, 2014. The maximum />dichlorobenzene concentration
was measured at SPAZ on February 16, 2014, along with the second highest
ethylbenzene and 1,3-butadiene concentrations and the third highest benzene
concentration.
• Table 5-4 shows that of the 30 valid VOC samples collected at SPAZ,
1,2-dichloroethane was detected 21 times, of which 19 were greater than the MDL for
this pollutant, and non-detects were reported for nine samples. Seven of these non-
detects were for consecutive sample days between August 15, 2014 and October 26,
2014.
Tables 4-9 through 4-12 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 4-9 through 4-12 a total of 13 times.
• SPAZ and PXSS have the highest annual average concentrations of
/;-dichlorobenzene among all NMP sites sampling VOCs, similar to previous years.
These annual average concentrations of p-dichlorobenzene are roughly twice the next
highest concentration shown in Table 4-9. While the highest concentrations of
/;-dichlorobenzene were not measured at these two sites, SPAZ and PXSS do account
for 24 of the 34 highest concentrations measured across the program (those greater
than 0.25 |ig/m3), with each site measuring 12 each. By comparison, the next highest
site had eight (S4MO) and TVKY and BTUT each measured one.
• These two sites also have the highest and second highest annual average
concentration of ethylbenzene among NMP sites. Similarly, the highest
concentrations of ethylbenzene across the program were not measured at SPAZ or
PXSS, but these sites have the highest number of ethylbenzene measurements greater
than 1 |ig/m3, 12 for PXSS and 6 for SPAZ, while no other site measured more than
three and most NMP sites measured none.
5-16
-------
• These two sites also have the second and third highest annual average concentrations
of 1,3-butadiene and the fourth and fifth highest annual average concentrations of
benzene among NMP sites sampling these pollutants.
• PXSS has the fifth highest annual average concentrations of acetaldehyde and
formaldehyde among NMP sites sampling carbonyl compounds.
• The annual average concentration of naphthalene for PXSS ranks seventh among
NMP sites sampling PAHs, similar to 2013.
• PXSS ranks fourth for its annual average concentration of nickel and eighth for its
annual average concentration of arsenic among NMP sites sampling PMio metals.
5.4.2 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
in Table 5-4 for PXSS and SPAZ. Figures 5-6 through 5-16 overlay the 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 3.4.3.1, and are
discussed below.
Figure 5-6. Program vs. Site-Specific Average Acetaldehyde Concentration
L
O i
1
0123456789 10
Concentration (ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 5-6 presents the box plot for acetaldehyde for PXSS and shows the following:
• PXSS's annual average concentration is greater than the program-level average
concentration as well as the program-level third quartile. Recall from the previous
section that PXSS has the fifth highest annual average acetaldehyde concentration
among 27 NMP sites sampling this pollutant and where annual average
concentrations could be calculated. Acetaldehyde concentrations measured at PXSS
range from 0.113 ng/m3 to 5.16 ng/m3.
5-17
-------
Figure 5-7. Program vs. Site-Specific Average Arsenic (PMio) Concentration
Program Max Concentration = 10.1 ng/m3
.
VJ
0
1 2
3
Concentration {ng/m3)
4
5
6
Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i
Site: Site Average
o
Site Concentration Range
Figure 5-7 presents the box plot for arsenic for PXSS and shows the following:
• The program-level maximum arsenic concentration (10.1 ng/m3) is not shown directly
on the box plot in Figure 5-7 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 annual average arsenic (PMio) concentration for PXSS falls between the
program-level average concentration and the program-level median concentration.
Arsenic concentrations measured at PXSS in 2014 range from 0.07 ng/m3 to
2.22 ng/m3.
Figure 5-8. Program vs. Site-Specific Average Benzene Concentrations
Program Max Concentration = 12.4 ng/m3
¦+
Program Max Concentration = 12.4 ng/m3
0
2
4 6
Concentration (ng/m3)
8
10
Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i
Site: Site Average
o
Site Concentration Range
5-18
-------
Figure 5-8 presents the box plots for benzene for both sites and shows the following:
• The program-level maximum benzene concentration (12.4 |ig/m3) is not shown
directly on the box plots in Figure 5-8 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.
• 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. Recall from the previous section that SPAZ and
PXSS have the fourth and fifth highest annual average concentrations of benzene
among the 30 NMP sites sampling this pollutant and where annual average
concentrations could be calculated. The annual average benzene concentrations for
these sites are similar to each other, although the range of measurements is slightly
greater for PXSS.
Figure 5-9. Program vs. Site-Specific Average 1,3-Butadiene Concentrations
PXSS
Program Max Concentration = 5.90 ng/m3
1
-
u
1 1
Proeram Max Concentration = 5.90 ue/m3
-
W
0 0.2 0.4 0.6 0.8 1
Concentration (ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 5-9 presents the box plots for 1,3-butadiene for both sites and shows the
following:
• Similar to benzene, the program-level maximum 1,3-butadiene concentration
(5.90 |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 to 1 |ig/m3.
• The maximum 1,3-butadiene concentration measured at PXSS is slightly higher than
the maximum concentration measured at SPAZ, although both are an order of
magnitude less than the maximum concentration measured across the program.
5-19
-------
• The annual average concentrations for these two sites are similar to each other, and
both are greater than the program-level average concentration.
• A single non-detect of 1,3-butadiene was measured at each Arizona monitoring site.
Figure 5-10. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations
Program Max Concentration = 3.06 ng/m3
Program Max Concentration = 3.06 ng/m3
0 0.5 1 1.5 2 2.5
Concentration (ng/m3)
Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 5-10 presents the box plots for carbon tetrachloride 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. Note that the
program-level median and average concentrations 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.
• The annual average concentrations of carbon tetrachloride for the Arizona sites are
also similar to each other and both are just less than the program-level average
concentration of 0.64 |ig/m3.
5-20
-------
Figure 5-11. Program vs. Site-Specific Average />-Dichlorobenzene Concentrations
PXSS
O
O i
KJ 1
i i i i i
0 0.2 0.4 0.6 0.8 1 1.2
Concentration (ng/m3)
Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 5-11 presents the box plots for p-dichlorobenzene for both sites and shows the
following:
• The program-level first and second quartiles are both zero and therefore not visible on
the box plots.
• SPAZ and PXSS have the highest annual average concentrations of
/;-dichlorobenzene among the 27 NMP sites sampling VOCs. The annual average
concentrations for SPAZ and PXSS are roughly four and five times the program-level
average concentration (0.04 |ig/m3), respectively.
• Although the maximum concentrations measured at these sites are considerably less
than the maximum concentration measured across the program, these two sites
account for four of the 10 highest concentrations measured across the program.
• Two non-detects of p-dichlorobenzene were measured at SPAZ while four non-
detects were measured at PXSS; no other NMP sites has fewer than 10 non-detects of
this pollutant.
5-21
-------
Figure 5-12. Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations
¦
Program Max Concentration = 27.4 ng/m3
Program Max Concentration = 27.4 |ig/m3
0.4 0.6
Concentration {[j.g/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 5-12 presents the box plots for 1,2-dichloroethane for both sites and shows the
following:
• The scale of the box plots in Figure 5-12 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 (27.4 |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 0.31 |ig/m3, which is being driven by
the measurements at the upper end of the concentration range.
• The annual average concentration for PXSS is similar to the program-level median
concentration (0.081 |ig/m3) while the annual average concentration for SPAZ is less
than the program-level first quartile (0.069 |ig/m3).
5-22
-------
Figure 5-13. Program vs. Site-Specific Average Ethylbenzene Concentrations
b
O i
VJ 1
O i
KJ 1
0.5
l
1.5 2
Concentration (ng/m3)
2.5
3
Program:
1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site:
Site Average
o
Site Concentration Range
Figure 5-13 presents the box plots for ethylbenzene for both sites and shows the
following:
• The range of ethylbenzene concentrations measured at PXSS is greater than the range
of concentrations measured at SPAZ. While the maximum concentration measured at
PXSS is roughly half the maximum concentration measured across the program,
PXSS accounts for four of the 10 highest ethylbenzene concentrations measured
across the program.
• The annual average concentrations of ethylbenzene for these two sites are both more
than twice the program-level average; recall from the previous section that these sites
have the two highest annual average concentrations of ethylbenzene among NMP
sites sampling this pollutant.
Figure 5-14. Program vs. Site-Specific Average Formaldehyde Concentration
¦
9 12 15
Concentration (ng/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
5-23
-------
Figure 5-14 presents the box plot for formaldehyde for PXSS and shows the following:
• The range of formaldehyde concentrations measured at PXSS falls within a relatively
small range (1.59 |ig/m3 to 5.30 |ig/m3) compared to the range of concentrations
measured across the program. Yet, the annual average concentration for PXSS is
greater than the program-level average concentration and similar to the program-level
third quartile. Recall from the previous section that this site has the fifth highest
annual average concentration of formaldehyde among NMP sites sampling carbonyl
compounds.
• The minimum formaldehyde concentration measured at PXSS is greater than the
program-level first quartile and is among the highest minimum concentrations
measured at a given site.
Figure 5-15. Program vs. Site-Specific Average Hexachloro-l,3-butadiene Concentration
0.1 0.2
0.3 0.4
Concentration (ng/m3)
0.5
0.6
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site: Site Average
o
Site Concentration Range
Figure 5-15 presents the box plot for hexachloro-1,3-butadiene for PXSS and shows the
following:
• The program-level first, second, and third quartiles for hexachloro-1,3-butadiene are
zero and therefore not visible on the box plot.
• The annual average concentration of hexachloro-1,3-butadiene for PXSS is similar to
the program-level average concentration (0.018 |ig/m3). While the maximum
concentration measured at PXSS (0.118 |ig/m3) is considerably less than the
maximum concentration measured across the program (0.609 |ig/m3), it is the fourth
highest measurement of this pollutant (although concentrations of the same
magnitude were measured at several other sites). Forty-four non-detects of
hexachloro-1,3-butadiene were measured at PXSS and none of the measurements
were greater than the MDL for this pollutant.
5-24
-------
Figure 5-16. Program vs. Site-Specific Average Naphthalene Concentration
¦
O i
¦
kJ 1
i i i i i
0 100 200 300 400 500 600
Concentration {ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 5-16 presents the box plot for naphthalene for PXSS and shows the following:
• The annual average naphthalene concentration for PXSS falls between the program-
level average concentration (66.46 ng/m3) and the program-level third quartile
(84.10 ng/m3).
• The range of naphthalene concentrations measured at PXSS is among the larger
ranges measured and PXSS is one of only three NMP sites at which a naphthalene
concentration greater than 300 ng/m3 was measured (DEMI and NBIL are the other
two).
5.4.3 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 3.4.3.2.
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
2007. Thus, Figures 5-17 through 5-32 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.
5-25
-------
Figure 5-17. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at PXSS
o
5th Percentile
— Minimum
2010 2011
Year
ledian — Maximum
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.
Observations from Figure 5-17 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 sampler
maintenance issues on the primary sampler. 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.
• The maximum acetaldehyde concentration (6.21 |ig/m3) was measured on
January 1, 2009, although this measurement is not significantly higher than the
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 of the 1-year average
concentrations could not be calculated. However, 1-year averages shown vary by less
than 0.5 |ig/m3, ranging from 2.52 |ig/m3 (2014) to 2.90 |ig/m3 (2012).
• The minimum concentration has decreased slightly every year through 2013,
particularly from 2012 to 2013. This minimum acetaldehyde concentration
5-26
-------
(0.20 |ig/m3) was measured on July 21, 2013 and is unusually low for PXSS. Only
seven acetaldehyde concentrations less than 1 |ig/m3 have been measured at PXSS,
of which were measured in 2011 or later. The minimum concentration for 2014 is
similar to the minimum concentration measured in 2013, and 2014 has the most
acetaldehyde concentrations less than 1 |ig/m3 of any year (three).
all
Figure 5-18. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at PXSS
2010
Year
5th Percentile
— Minimum
- Median
— Maximum
O 95th Percentile
Observations from Figure 5-18 for arsenic concentrations measured at PXSS include the
following:
• PXSS began sampling arsenic under the NMP in January 2006.
• The maximum arsenic concentration (6.73 ng/m3) was measured on
December 26, 2007 and is more than twice the next highest concentration
(3.05 ng/m3), measured on August 19, 2011. In total, 16 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 the most measured in 2008 (four).
• After several years of decreasing slightly, the 1-year average concentration increased
significantly from 2010 to 2011, after which additional decreasing is shown through
2013. The 1-year average concentration is at a minimum for 2013 (0.49 ng/m3). The
maximum concentration and 95th percentile are also at a minimum for 2013.
5-27
-------
• With the exception of the median concentration, all of the statistical metrics increased
at least slightly from 2013 to 2014. The median concentration changed very little
between the two years (from 0.43 ng/m3 to 0.42 ng/m3).
Figure 5-19. Yearly Statistical Metrics for Benzene Concentrations Measured at PXSS
T
J-
LT"
o
I
o
--o...
f-
2010 2011
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 2007.
Observations from Figure 5-19 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
(5.21 |ig/m3). Four additional measurements greater than 4 |ig/m3 have been
measured at this site (during 2007, 2008, 2009, and 2011).
• The 15 highest benzene concentrations (those greater than 3.5 |ig/m3) were all
measured during the first or fourth quarter of any given year. Further, of the 110
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 quarters.
• 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
5-28
-------
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, with the largest change shown from 2009 to 2010. For 2010, the
number of benzene concentrations greater than 2 |ig/m3 decreased to 12, with the
number ranging from nine (2013) 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 from 2013 to 2014. The median
concentration is also at a minimum for 2014 (0.74 |ig/m3).
Figure 5-20. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at PXSS
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 1,3-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
concentration was measured. All but two of the 115 1,3-butadiene concentrations
greater than 0.30 |ig/m3 were measured during the first or fourth quarters. The two
not measured during the first or fourth quarters were 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 (2014) to 0.23 |ig/m3 (both 2009 and
5-29
-------
2011). The median concentration varies between 0.15 |ig/m3 and 0.17 |ig/m3 for the
years between 2008 and 2012, then fell to 0.13 |ig/m3 and 0.011 |ig/m3 for 2013 and
2014, respectively.
• There have been 10 non-detects of 1,3-butadiene 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. For 2011, the
minimum and 5th percentile were both equal to zero. None of the 10 non-detects of
1,3-butadiene were measured during the first or fourth quarters of any given year.
Figure 5-21. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at PXSS
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 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.
• For 2007, 2010, 2011, and 2014, the box and whisker plots for this pollutant 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).
5-30
-------
• All of the carbon tetrachloride measurements from 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 in 2011 is 23; the number of concentrations less
than 0.6 |ig/m3 in 2012 is five.
• All of the statistical parameters for carbon tetrachloride exhibit a decrease for 2013.
This is also true for 2014, with the exception of the median concentration, which did
not change between the two years. The 1-year average concentration also changed
little between 2013 and 2014.
Figure 5-22. Yearly Statistical Metrics for />-Dichlorobenzene Concentrations Measured at PXSS
2008 2009
2010 2011 2012 2013 2014
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 2007.
Observations from Figure 5-22 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.
5-31
-------
• The maximum, 95th percentile, 1-year average, and median concentrations all exhibit
a significant decreasing trend through 2010. Even the minimum concentration and 5th
percentile decreased each year from 2008 through 2010. Each of the statistical
parameters increased for 2011, with the exception 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.
• Prior to 2010, a single non-detect was measured; for 2010, nine non-detects were
measured, explaining the decrease in the minimum and 5th percentile shown from
2009 to 2010. The number of non-detects has varied between one (2012) and six
(2011) in the years following 2010.
Figure 5-23. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at PXSS
o
2010 2011
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 2007.
Observations from Figure 5-23 for 1,2-dichloroethane concentrations measured at PXSS
include the following:
• There were no measured detections of 1,2-dichloroethane in 2007, one measured
detection in 2008, seven in 2009, nine in 2010, 12 in 2011, 47 in 2012, 38 in 2013,
and 53 in 2014.
5-32
-------
• The median concentration is zero for each year until 2012, 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 difference between the 1-year average and median concentration for 2014 is at its
lowest since 2008, when there was only one measured detection.
Figure 5-24. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at PXSS
o
¦O"'
o
2010 2011
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 2007.
Observations from Figure 5-24 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 benzene concentration
was 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).
5-33
-------
• Similar to benzene and 1,3-butadiene, the highest ethylbenzene concentrations were
measured most often during the first and fourth quarters of the years. Ninety of the
100 highest concentrations (those greater than 1.0 |ig/m3) were measured between
January and March or October and December of any given year. The exceptions were
measured April (1), May (1), July (1), and September (7).
• The median ethylbenzene concentration has a decreasing trend 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 measurements greater than 1 |ig/m3 were measured in 2011 (20) than 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 2014, with the 1-year average concentration at a minimum for 2014
(0.57 |ig/m3). The median concentration is also at a minimum for 2014 (0.38 |ig/m3),
as are the minimum concentration and the 95th percentile.
Figure 5-25. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at PXSS
Maximum
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.
Observations from Figure 5-25 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 for 2007 is not presented,
5-34
-------
although the range of measurements is provided. In addition, much of the data
between February 2010 and March 2011 was invalidated due to sampler maintenance
issues on the primary sampler. 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 concentration is not provided.
• The five highest formaldehyde concentrations (ranging from 6.28 |ig/m3 to
7.56 |ig/m3) were all measured at PXSS in 2007 and all but one of the 11
formaldehyde concentrations greater than 6 |ig/m3 were measured in either 2007 or
2011 at PXSS (with the exception measured in 2012).
• The median concentration for 2007 is nearly 5 |ig/m3. The median concentration for
the years that follow are all less than 4 |ig/m3.
• Only one formaldehyde concentration less than 1 |ig/m3 has been measured at PXSS
(2012) and only 18 less than 2 |ig/m3 have been measured at PXSS since 2007. One-
third of these were measured in 2014 (6) and no other year has more than three.
Figure 5-26. Yearly Statistical Metrics for Hexachloro-l,3-butadiene Concentrations
Measured at PXSS
T
i I T I
<
rn rn n
2007 1 2008 2009 2010 2011 2012 2013 2014
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.
5-35
-------
Observations from Figure 5-26 for hexachloro-1,3-butadiene concentrations measured at
PXSS include the following:
• The maximum concentration of hexachloro-l,3-butadiene was measured on
September 3, 2008 (0.85 |ig/m3). No other concentration greater than 0.15 |ig/m3 has
been measured at PXSS.
• The median concentration for each year shown is zero, indicating that at least
50 percent of the measurements are non-detects for each year. The percentage of non-
detects was greater than 90 percent for each of the first six years of sampling,
including 2009, when the percentage was 100 percent. The percentage of non-detects
decreased to 87 percent for 2013 when eight measured detections were measured. The
percentage of non-detects is at a minimum for 2014 (72 percent), when the number of
measured detections more than doubled to 17.
Figure 5-27. Yearly Statistical Metrics for Naphthalene Concentrations Measured at PXSS
450
400
350
300
lif 250
c
(O
£ 200
u
c
u
150
100
50
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 July 2007.
Observations from Figure 5-27 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.
5-36
-------
• The maximum naphthalene concentration was measured on December 20, 2008
(400 ng/m3), with a similar concentration measured 12 days later on January 1, 2009
(386 ng/m3). Two additional 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, or
midpoint, for 2009 is 107 ng/m3. The median concentrations for the other years are
less than 100 ng/m3, ranging from 55.30 ng/m3 (2014) to 84.1 ng/m3 (2010), and have
a steady decreasing trend after 2009. The 1-year average concentration is also at a
maximum for 2009 (120.17 ng/m3) and at a minimum for 2014 (78.25 ng/m3),
varying between roughly 90 ng/m3 and 100 ng/m3 in between.
Figure 5-28. Yearly Statistical Metrics for Benzene Concentrations Measured at SPAZ
2012
2013
Maximum
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 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. Five additional measurements greater than 4 |ig/m3 have been measured at this
site (one for each year of sampling prior to 2012).
5-37
-------
• Similar to PXSS, 52 of the 56 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 1.25 |ig/m3 (16) for the years
prior to 2013.
• After several years of increasing, both 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 1-year average concentrations changed little between 2009 and 2011, then
decreased from 2011 to 2012 and again for 2013, with little change for 2014. The
median concentration exhibits more variability during this time frame.
Figure 5-29. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at SPAZ
o
2007 1 2008 2009 2010 2011 2012 2013 2014
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.
5-38
-------
Observations from Figure 5-29 for 1,3-butadiene concentrations measured at SPAZ
include the following:
• The only 1,3-butadiene concentration greater than 1 |ig/m3 was measured at SPAZ on
January 27, 2011 (1.29 |ig/m3). Four additional 1,3-butadiene concentrations greater
than 0.75 |ig/m3 have been measured at SPAZ, one in 2007, two in 2010, and one in
2011.
• Seventy-three of the 77 concentrations greater than 0.25 |ig/m3 were measured at
SPAZ during the first or fourth quarters of any given year, similar to the trend seen in
PXSS 1,3-butadiene 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.
• 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. 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.
5-39
-------
Figure 5-30. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
SPAZ
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.
Observations from Figure 5-30 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
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 the
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 the years shown, reaching a minimum for 2014 (0.61 |ig/m3).
However, the differences represent an overall change of less than 0.12 |ig/m3.
• The range of concentrations measured is at a minimum for 2013, as is the difference
between the 1-year average and median for 2013 (0.001 |ig/m3), indicating the least
amount of variability in the measurements compared to other years. However, the
difference between the 1-year average and median concentrations is relatively low for
every year, with the difference for 2008 being the largest (0.04 |ig/m3).
5-40
-------
Figure 5-31. Yearly Statistical Metrics for />-Dichlorobenzene Concentrations Measured at SPAZ
I
I
I
o.
-
o
o
O
I
I
I
L^-
T
2010 2011
Year
5th Percentile
— Minimum
- 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 5-31 for /;-dichlorobenzene concentrations measured at SPAZ
include the following:
• The widest range of p-dichlorobenzene concentrations measured is shown for 2008
(non-detect to 0.90 |ig/m3), while the range of concentrations measured the following
year is roughly half as wide. A review of the data shows that the number of
/;-dichlorobenzene concentrations greater than 0.3 |ig/m3 decreased by half from 2008
(8) to 2009 (4). All of the statistical metric exhibit increases from 2009 to 2010, with
the number of p-dichlorobenzene concentrations greater than 0.3 |ig/m3 increasing
nearly four-fold (15).
• The 1-year average concentration decreased from 2008 to 2009, increased for 2010,
then decreased slightly each year between 2011 and 2014. However, confidence
intervals calculated for these averages indicate that the changes are not statistically
significant. The median concentrations exhibit larger fluctuations than the 1-year
average concentrations.
5-41
-------
Figure 5-32. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at SPAZ
0.16
0.14
I
c
£ 0.08
—
c
o
O"
(J
0.06
—
o
0.02
>
o
o.oo HUB
2007 1 2008
2009
2010
2011
2012
2013
2014
Year
O 5th Percentile — Minimum
- Median
"
Maximum
O 95th Percentile
¦ Average
1
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 5-32 for 1,2-dichloroethane concentrations measured at SPAZ
include the following:
• There were no measured detections of 1,2-dichloroethane in 2007, one measured
detection in 2008, three in 2009, four in 2010, seven in 2011, 26 in 2012, 19 in 2013,
and 20 in 2014.
• The median concentration is zero for all years 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 5-32.
• The median concentration is greater than the 1-year average concentration for 2012,
2013, and 2014. 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.
• Confidence intervals calculated for the last three years of sampling indicate that the
steady decrease in the 1-year average and median concentrations shown is not
statistically significant due to the variability in the concentrations measured.
5-42
-------
Figure 5-33. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at SPAZ
Maximum
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 5-33 for ethylbenzene concentrations measured at SPAZ
include the following:
• The maximum concentration of ethylbenzene measured at SPAZ (3.44 |ig/m3) was
measured in 2007. The only other concentration greater than 3.0 |ig/m3 was measured
at SPAZ on January 27, 2011 (3.06 |ig/m3). All eight concentrations between
2.0 |ig/m3 and 3.0 |ig/m3 were measured in either 2007 (four) or 2011 (four).
• The median concentration is at a maximum for 2007, after which the median
decreases by half. Recall that 2007 includes only half a year's worth of samples. The
downward trend continues through 2009, followed by an increase that continues
through 2011. The median decreases somewhat for 2012, with additional decreases
for 2013 and 2014. The 1-year average concentration has a similar pattern, although
no 1-year average concentration is presented for 2007. These patterns are similar to
the patterns shown for 1,3-butadiene in Figure 5-29 and the patterns shown in
Figure 5-24 for PXSS's ethylbenzene concentrations.
• The only non-detects of ethylbenzene were measured during the first two full-years of
sampling at SPAZ.
5-43
-------
5.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at each Arizona monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4
for definitions and explanations regarding the various toxicity factors, time frames, and
calculations associated with these risk-based screenings.
5.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Arizona 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 3.4.3.3 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 5-5, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
5-44
-------
Table 5-5. Risk Approximations for the Arizona Monitoring Sites
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Phoenix, Arizona - PXSS
Acetaldehyde
0.0000022
0.009
61/61
2.52
±0.29
5.54
0.28
Benzene
0.0000078
0.03
61/61
1.05
±0.20
8.20
0.04
1.3 -Butadiene
0.00003
0.002
60/61
0.20
±0.05
5.94
0.10
Carbon Tetrachloride
0.000006
0.1
61/61
0.61
±0.02
3.67
0.01
/?-Dichlorobcnzcnc
0.000011
0.8
57/61
0.17
±0.03
1.84
<0.01
1,2-Dichloroethane
0.000026
2.4
53/61
0.08
±0.01
2.00
<0.01
Ethylbenzene
0.0000025
1
61/61
0.57
±0.10
1.41
<0.01
Formaldehyde
0.000013
0.0098
61/61
3.46
±0.24
45.04
0.35
Hexachloro -1,3 -butadiene
0.000022
0.09
17/61
0.02
±0.01
0.44
<0.01
Arsenic (PMi0)a
0.0043
0.000015
60/60
0.55
±0.11
2.35
0.04
Naphthalene1
0.000034
0.003
59/59
78.25
± 15.71
2.66
0.03
South Phoenix, Arizona - SPAZ
Benzene
0.0000078
0.03
30/30
1.09
±0.24
8.48
0.04
1.3 -Butadiene
0.00003
0.002
29/30
0.19
±0.06
5.62
0.09
Carbon Tetrachloride
0.000006
0.1
30/30
0.60
±0.03
3.58
0.01
/?-Dichlorobcnzcnc
0.000011
0.8
28/30
0.21
±0.05
2.31
<0.01
1,2-Dichloroethane
0.000026
2.4
21/30
0.06
±0.01
1.43
<0.01
Ethylbenzene
0.0000025
1
30/30
0.60
±0.13
1.51
<0.01
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
Observations for PXSS from Table 5-5 include the following:
• The pollutants of interest with the highest annual average concentrations are
formaldehyde, acetaldehyde, and benzene, and are the only pollutants of interest with
annual average concentrations greater than 1 |ig/m3.
• Based on the annual averages and cancer UREs, formaldehyde has the highest cancer
risk approximation (45.04 in-a-million), followed by benzene (8.20 in-a-million),
1,3-butadiene (5.94 in-a-million), and acetaldehyde (5.54 in-a-million).
5-45
-------
• Formaldehyde's cancer risk approximation for PXSS is the sixth highest cancer risk
approximation among the site-specific pollutants of interest across the program.
• 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 PXSS is formaldehyde (0.35). This noncancer hazard
approximation is the seventh highest noncancer hazard approximation among all site-
specific pollutants of interest.
Observations for SPAZ from Table 5-5 include the following:
• The pollutants with the highest annual average concentrations for SPAZ are benzene,
ethylbenzene, and carbon tetrachloride. Only benzene has an annual average
concentration greater than 1 |ig/m3,
• Based on the annual averages and cancer UREs, benzene has the highest cancer risk
approximation for SPAZ (8.48 in-a-million), followed by 1,3-butadiene
(5.62 in-a-million), and carbon tetrachloride (3.58 in-a-million). These cancer risk
approximations are similar to the approximations calculated for these same pollutants
for PXSS.
• 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 (0.09).
5.5.2 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 5-6 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
that have cancer toxicity factors. Table 5-6 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 5-6 provides the pollutants of interest with the highest cancer risk approximations (in-a-
million) for each site, as presented in Table 5-5. The emissions, toxicity-weighted emissions, and
cancer risk approximations are shown in descending order in Table 5-6. Table 5-7 presents
similar information, but is limited to those pollutants with noncancer toxicity factors.
5-46
-------
Table 5-6. 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)
Cancer
Cancer Risk
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(in-a-million)
Phoenix, Arizona (Maricopa County) - PXSS
Benzene
1,313.94
Formaldehyde
1.48E-02
Formaldehyde
45.04
Formaldehyde
1,141.02
Benzene
1.02E-02
Benzene
8.20
Ethylbenzene
862.37
1,3-Butadiene
5.42E-03
1,3-Butadiene
5.94
Acetaldehyde
576.27
Naphthalene
3.02E-03
Acetaldehyde
5.54
1.3 -Butadiene
180.82
Ethylbenzene
2.16E-03
Carbon Tetrachloride
3.67
T etrachloroethylene
95.59
POM, Group 2b
1.48E-03
Naphthalene
2.66
Naphthalene
88.77
Acetaldehyde
1.27E-03
Arsenic (PMio)
2.35
POM, Group 2b
16.83
POM, Group 2d
1.19E-03
1,2-Dichloroethane
2.00
POM, Group 2d
13.53
Arsenic, PM
1.03E-03
/?-Dichlorobcnzcnc
1.84
Dichloro methane
12.34
POM, Group 5a
7.15E-04
Ethylbenzene
1.41
South Phoenix, Arizona (Maricopa County) - SPAZ
Benzene
1,313.94
Formaldehyde
1.48E-02
Benzene
8.48
Formaldehyde
1,141.02
Benzene
1.02E-02
1,3-Butadiene
5.62
Ethylbenzene
862.37
1,3-Butadiene
5.42E-03
Carbon Tetrachloride
3.58
Acetaldehyde
576.27
Naphthalene
3.02E-03
/?-Dichlorobcnzcnc
2.31
1,3-Butadiene
180.82
Ethylbenzene
2.16E-03
Ethylbenzene
1.51
T etrachloroethylene
95.59
POM, Group 2b
1.48E-03
1,2-Dichloroethane
1.43
Naphthalene
88.77
Acetaldehyde
1.27E-03
POM, Group 2b
16.83
POM, Group 2d
1.19E-03
POM, Group 2d
13.53
Arsenic, PM
1.03E-03
Dichloro methane
12.34
POM, Group 5a
7.15E-04
-------
Table 5-7. 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)
Noncancer
Noncancer
Hazard
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(HQ)
Phoenix, Arizona (Marico
)a County) - PXSS
Toluene
5,233.19
Acrolein
2,932,324.18
Formaldehyde
0.35
Xylenes
3,296.34
Formaldehyde
116,431.08
Acetaldehyde
0.28
Hexane
2,752.67
1,3-Butadiene
90,410.71
1,3-Butadiene
0.10
Methanol
2,399.14
Acetaldehyde
64,030.43
Arsenic
0.04
Benzene
1,313.94
Benzene
43,798.12
Benzene
0.04
Formaldehyde
1,141.02
Lead, PM
34,426.96
Naphthalene
0.03
Ethylene glycol
880.96
Xylenes
32,963.37
Carbon Tetrachloride
0.01
Ethylbenzene
862.37
Naphthalene
29,589.71
Ethylbenzene
<0.01
Acetaldehyde
576.27
Arsenic, PM
16,021.47
Hexachloro-1,3 -butadiene
<0.01
Methyl isobutyl ketone
326.41
Propionaldehyde
10,771.78
/?-Dichlorobcnzcnc
<0.01
South Phoenix, Arizona (Maricopa County) - SPAZ
Toluene
5,233.19
Acrolein
2,932,324.18
1,3-Butadiene
0.09
Xylenes
3,296.34
Formaldehyde
116,431.08
Benzene
0.04
Hexane
2,752.67
1,3-Butadiene
90,410.71
Carbon Tetrachloride
0.01
Methanol
2,399.14
Acetaldehyde
64,030.43
Ethylbenzene
<0.01
Benzene
1,313.94
Benzene
43,798.12
/?-Dichlorobcnzcnc
<0.01
Formaldehyde
1,141.02
Lead, PM
34,426.96
1,2-Dichloroethane
<0.01
Ethylene glycol
880.96
Xylenes
32,963.37
Ethylbenzene
862.37
Naphthalene
29,589.71
Acetaldehyde
576.27
Arsenic, PM
16,021.47
Methyl isobutyl ketone
326.41
Propionaldehyde
10,771.78
-------
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 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 5.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
Observations from Table 5-6 include the following:
• Benzene, formaldehyde, 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, benzene, and 1,3-butadiene.
• Eight of the highest emitted pollutants in Maricopa County also have the highest
toxicity-weighted emissions.
• Formaldehyde has the highest cancer risk approximation for PXSS; carbonyl
compounds were not sampled for at SPAZ, thus, a cancer risk approximation is not
available for this pollutant for SPAZ. Formaldehyde has the second highest emissions
and the highest toxicity-weighted emissions for Maricopa County. Acetaldehyde,
which has the fourth highest cancer risk approximation for PXSS, also appears on
both emissions-based list for Maricopa County.
• Among the VOCs, benzene, 1,3-butadiene, and carbon tetrachloride have highest
cancer risk approximations for PXSS and SPAZ. The cancer risk approximations for
these pollutants are 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 23rd for quantity emitted and 28th for it toxicity-weighted
emissions.
• Naphthalene is among the highest emitted pollutants (seventh), has one of the highest
toxicity-weighted emissions (fourth), and has one of the highest cancer risk
approximations for PXSS (sixth). POM, Group 2b is the eighth highest emitted
"pollutant" in Maricopa County and ranks sixth for 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 5a ranks tenth for its toxicity-weighted emissions for Maricopa County.
5-49
-------
This POM group includes benzo(a)pyrene, which failed two screens for PXSS but
was not identified as a pollutant of interest for this site.
• Arsenic has the seventh highest cancer risk approximation among the pollutants of
interest for PXSS. This pollutant ranks ninth for its toxicity-weighted emissions but
does not appear among the highest emitted pollutants in Maricopa County (it ranks
20th).
Observations from Table 5-7 include the following:
• Toluene, xylenes, and hexane 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, formaldehyde, and 1,3-butadiene.
• Four of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Maricopa County.
• 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 16th for Maricopa County.
• Formaldehyde and acetaldehyde have the highest noncancer hazard approximations
for PXSS (although considerably less than an HQ of 1.0), both of which appear
among those with the highest emissions and toxicity-weighted emissions for
Maricopa County.
• 1,3-Butadiene and benzene have the highest noncancer hazard approximations among
the VOCs for both PXSS and SPAZ and are similar in magnitude between the two
sites. Benzene ranks fifth for both its emissions and its toxicity-weighted emissions.
1,3-Butadiene has the third highest toxicity-weighted emissions for Maricopa County
but is not one of the highest emitted pollutants in Maricopa County (with a noncancer
RfC), as it ranks 11th.
• Arsenic has the fourth highest noncancer hazard approximation for PXSS. Arsenic
has the ninth highest toxicity-weighted emissions for Maricopa County but is not one
of the highest emitted pollutants in Maricopa County (with a noncancer RfC), as it
ranks 40th.
• Naphthalene is another pollutant of interest for PXSS that appears among the
pollutants with the highest toxicity-weighted emissions for Maricopa County but
whose actual emissions rank outside the top 10 emitted pollutants, ranking 13th.
5-50
-------
5.6 Summary of the 2014 Monitoring Data for PXSS and SPAZ
Results from several of the data analyses described in this section include the following:
~~~ Fifteen pollutants failed screens for PXSS; six pollutants failed screens for SPAZ. The
six pollutants that failed screens for SPAZ also failed screens for PXSS.
~~~ Of the site-specific pollutants of interest for PXSS, formaldehyde had the highest
annual average concentration. For SPAZ, benzene had the highest annual average
concentration 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 and PXSS have the highest and second highest annual average concentrations
of p-dichlorobenzene and ethylbenzene among NMP sites sampling VOCs. These sites
also rank second and third highest for 1,3-butadiene.
~~~ Concentrations of benzene and ethylbenzene appear to be decreasing at the Arizona
sites. The detection rate of 1,2-dichloroethane increased significantly during the later
years of sampling.
~~~ Formaldehyde has the highest cancer risk approximation of the pollutants of interest
for PXSS; benzene has the highest cancer risk approximation of 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.
5-51
-------
6.0 Sites in California
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at three NATTS sites in California, and integrates these concentrations
with emissions, meteorological, and risk information. Data generated by sources other than ERG
are not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) 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 geographical and
physical information about the locations of the sites and the surrounding areas. 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.
Three NATTS monitoring sites are located in California. Two are located in Southern
California, in Los Angeles (CELA) and Rubidoux (RUCA), and a third monitoring site is located
in Northern California, in San Jose (SJJCA). Figure 6-1 is the composite satellite image retrieved
from ArcGIS Explorer showing the Los Angeles monitoring site and its immediate surroundings.
Figure 6-2 identifies nearby point source emissions locations by source category, as reported in
the 2011 NEI for point sources, version 2. 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 in order to emphasize emissions
sources within the boundary. Figures 6-3 through 6-6 are 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.
6-1
-------
Figure 6-1. Los Angeles, California (CELA) Monitoring Site
_Stad i lot Sd*it7 t own ,Ga te
ManitOU Ave
Mcszai St
5tdte ™stafi
Darwin^Ave
¦* - \ % $: ¦5-'> /
— ~ csx up-i>7c
" .•••
¦ ounty
Irfiiuul
ON
to
-------
Figure 6-2. NEI Point Sources Located Within 10 Miles of CELA
ns^s'o-w ii8*2oww 118*15'CTW ii8*iPharmaceutical Manufacturing (1)
R Plastic, Resin, or Rubber Products Plant (6)
Printing, Coating & Dyeing of Fabrics Facility (4)
P Printing/Pubbshing/Paper Product Manufacturing Facility (23)
E Pulp and Paper Plant (2)
X Rail Yard/Rail Line Operations (4)
W V Steel Mill (1)
T Textile, Yam, or Carpet Plant (3)
Truck/Bus/T ransportation Operations (1)
fill Utilities/Pipeline Construction (1)
* Wastewater Treatment Facility (3)
& Water Treatment Facility (2)
W Woodwork. Furniture, Millwork & Wood Preserving Facility (10)
CELA NATTS site O 10 mile radius
County boundary
Pacific
Ocean
- -i _
I
I Los Angeles
County
6-3
-------
Figure 6-3. Rubidoux, California (RUCA) Monitoring Site
, ft
r'5td d*cxo o"**^
v *•••*» }
*5th«S£
Jensen-Alvarado Historic Ranch Park,
\ SourcVOfUSGS&V
Source: NASA. N G A USGS
V 2008 Microsoft Corp.
geoham e s> o r\ St.
irencia
ON
¦k
-------
Figure 6-4. NEI Point Sources Located Within 10 Miles of RUCA
County
Riverside
County
* Aerospace/Aircraft Manufacturing Facility (1) ^
"J" Airport/Airline/Airport Support Operations (12) ^
A Animal Feedlot or Farm (9)
Asphalt Production/Hot Mix Asphalt Plant (4) I
0 Auto Body Shop/Painters/Automotive Stores (9) IT
Automobile/Truck Manufacturing Facility (4)
Automotive/RV Dealership (2) O
Brick. Structural Clay, or Clay Ceramics Plant (1) ®
A Building/Construction (1) (|j)
B Bulk Terminals/Bulk Plants (3) A
C Chemical Manufacturing Facility (3) (•)
£ Compressor Station (1) A
0 Electrical Equipment Manufacturing Facility (2) X
Electricity Generation via Combustion (5)
Ethanol Biorefineries (1)
Food Processing/Agriculture Facility (7)
Foundries, Iron and Steel (1)
Gasoline/Diesel Service Station (1)
Industrial Machinery or Equipment Plant (5)
Institutional (school, hospital, prison, etc.) (9)
Landfill (5)
Metal Can, Box, and Other Metal Container Manufacturing (4)
Metal Coating. Engraving, and Allied Services to Manufacturers (1)
Metals Processing/Fabrication Facility (12)
Military Base/National Security Facility (1)
Mine/Quarry/Mineral Processing Facility (6)
f Miscellaneous Commercial/Industrial Facility (18)
[Ml Municipal Vfeste Combustor (1)
•
"~
R
7
P
Oil and/or Gas Production (2)
Paint and Coating Manufacturing Facility (4)
Plastic. Resin, or Rubber Products Plant (4)
Portland Cement Manufacturing (2)
Printing/Publishing/Paper Product Manufacturing Facility (6)
H Pulp and Paper Rant (1)
X Rail Yard/Rail Line Operations (3)
V Steel Mill (2)
I Vtestewater Treatment Facility (3)
£ Vteter Treatment Facility (6)
VV V\foodwork. Furniture, Millwork & V\food Preserving Facility (4)
Legend
~ RUCA NATTS site
117°25'0"W 117°20'0"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.
O 10 mile radius
County boundary
San Bernardino I
Source Category Group (No. of Facilities)
6-5
-------
Figure 6-5. San Jose, California (SJJCA) Monitoring Site
-------
Figure 6-6. NEI Point Sources Located Within 10 Miles of SJJCA
Miles
J—- i
122'5'0'W 122WW 12r55'0"W
Legend
SJJCA NATTS site O 10 mile radius
County boundary
Source Category Group (No. of Facilities)
Aerospace/Aircraft Manufacturing Facility (3)
"IT Airport/Airline/Airport Support Operations (18)
^ Asphalt Production/Hot Mix Asphalt Plant (3)
0 Auto Body Shop/Painters/Automotive Stores (219)
Automobile/Truck Manufacturing Facility (3)
Automotive/RV Dealership (8)
~ Brick. Structural Clay, or Clay Ceramics Plant (1)
ft Building/Construction (13)
B Bulk Terminals/Bulk Plants (7)
C Chemical Manufacturing Facility (12)
1 Compressor Station (2)
[Xbrematory - Animal/Human (5)
© Dry Cleaning Facility (110)
6 Electrical Equipment Manufacturing Facility (262)
^ Electricity Generation via Combustion (60)
E Electroplating. Plating, Polishing, Anodizing, and Coloring (13)
V Fertilizer Plant (1)
F Food Processing/Agriculture Facility (20)
Foundries. Non-ferrous (1)
Gasoline/Diesel Service Station (7)
Glass Plant (1)
Hotels/Motels/Lodging (8)
Industrial Machinery or Equipment Plant (22)
Institutional (school, hospital, prison, etc.) (138)
Landfill (4)
Leather and Leather Products Facility (1)
Metal Can, Box, and Other Metal Container Manufacturing (4)
Metal Coating, Engraving, and Allied Services to Manufacturers (27)
Metals Processing/Fabrication Facility (29)
Military Base/National Security Facility (3)
Mine/Quarry/Mineral Processing Facility (11)
Mineral Wool/Wool Fiberglass Manufacturing Facilty (1)
Miscellaneous Commercial/Industrial Facility (292)
Oil and/or Gas Production (1)
Q Paint and Coating Manufacturing Facility (3)
Petroleum Products Manufacturing (2)
CD Pharmaceutical Manufacturing (9)
R Plastic, Resin, or Rubber Products Plant (4)
V Port and Harbor Operations (1)
I* Printing, Coating & Dyeing of Fabrics Facility (4)
P Printing/Publishing/Paper Product Manufacturing Facility (34)
s Pulp and Paper Plant (3)
X Rail Yard/Rail Line Operations (2)
TT Telecommunications/Radio Facility (91)
Testing Laboratories (2)
T Textile, Yarn, or Carpet Plant (1)
M Tobacco Manufacturing (8)
Truck/Bus/Transportation Operations (7)
l^] Utilities/Pipeline Construction (1)
I Wastewater Treatment Facility (10)
& Water Treatment Facility (30)
W Woodwork, Furniture, Millwork & Wood Preserving Facility (36)
121°50'0"W 121*45WV 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.
/
Alameda / -
County i
/
/
/
Santa Clara
County
Santa Cruz
County
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
230,000
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
158,000
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
124,000
Rte 87 (Guadalupe Pkwy) between
Julian St and W Taylor St
lAADT reflects 2014 data (CADOT, 2014)
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 large number of emissions sources within 10 miles of CELA include institutions such as
schools, hospitals, and/or prisons; auto body shops, painters, and automotive stores; 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 carpet
plant, 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 44 miles east of CELA.
Figure 6-4 shows that fewer emissions sources surround RUCA than CELA. Most of the
emissions sources are located to the northeast and northwest of the site, in San Bernardino
County. The point source located closest to RUCA is Flabob Airport. Although the emissions
source categories are varied, the emissions source categories with the greatest number of sources
within 10 miles of RUCA include airport operations; metals processing and fabrication; auto
6-9
-------
body shops, painters, and automotive stores; animal feedlots or farms; and institutions such as
schools, hospitals, and/or prisons.
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 are electrical equipment manufacturing; auto body, paint, and automotive
shops; institutions such as schools, hospitals, and/or prisons; dry cleaning; and
telecommunications. Sources closest to SJJCA include a food processing facility and several
auto body shops.
In addition to providing city, county, CBSA, 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 effect concentrations measured at a given monitoring site. CELA
experiences a higher traffic volume compared to the other California sites, although the traffic
volumes near these sites are all greater than 100,000. Compared to other NMP sites, CELA has
the second highest traffic volume, RUCA ranks fifth, and SJJCA ranks eighth highest. These
traffic volumes for CELA, RUCA and SJJCA were obtained from heavily traveled highways.
6.2 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
sites in California on sample days, as well as over the course of the year.
6-10
-------
6.2.1 Meteorological Summary
In order to provide an overview of the meteorological conditions experienced at each
monitoring site, hourly meteorological data for 2014 were retrieved and sample day and full-year
averages developed for temperature, dew point, relative humidity, pressure, and wind speed.
Weather data from the actual monitoring site(s) were obtained from AQS, where available. If
site-specific weather data were not available in AQS, then data were obtained from NCDC for
the NWS weather station located closest to the monitoring site(s), as described in Section 3.4.2.
For the California sites, site-specific data were available for some, but not all, of the parameters
in Table 6-2. For CELA, data from the NWS weather station at Downtown L.A./USC Campus
(WBAN 93134) were used for meteorological parameters without data and/or as surrogates for
parameters without complete observation records. The Downtown L.A./USC Campus weather
station is located 4.7 miles southwest of CELA. For RUCA, data from the NWS weather station
at Riverside Municipal Airport (WBAN 03171) were used, where needed; the Riverside
Municipal Airport weather station is located 3.5 miles south-southwest of RUCA. For SJJCA,
data from the NWS weather station at San Jose International Airport (WBAN 23293) were used
as needed; the weather station at the San Jose International Airport is located 1.8 miles west-
northwest of SJJCA. A map showing the distance between each California monitoring site and
the closest NWS weather station is provided in Appendix R. These data were used to determine
how meteorological conditions on sample days vary from conditions experienced throughout the
year.
Table 6-2 presents average temperature, average dew point temperature, average relative
humidity, average station and sea level pressure, and wind information (average scalar wind
speed and prevailing wind direction) for days on which samples were collected and for all of
2014. Also included in Table 6-2 is the 95 percent confidence interval for each parameter. As
shown in Table 6-2, average meteorological conditions on sample days were representative of
average weather conditions experienced throughout the year at each site. The differences
between the sample day and full-year averages were greatest for relative humidity, although the
difference is not statistically significant.
6-11
-------
Table 6-2. Average Meteorological Conditions near the California Monitoring Sites
Average
Type1
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(in Hg)
Average
Station
Pressure
(in Hg)
Prevailing
Wind
Direction
Average
Scalar Wind
Speed
(kt)
Los Angeles, California - CELA2
Sample
Days
67.0
51.4
60.4
29.95
29.69
3.9
(63)
±0.5
±0.6
± 1.1
±0.01
±0.01
NE
±0.1
67.0
51.1
59.7
29.96
29.69
3.9
2014
±0.2
±0.2
±0.4
±<0.01
± <0.01
NE
±<0.1
Rubidoux, California - RUCA3
Sample
Days
68.5
45.1
51.1
29.93
29.12
3.2
(63)
±0.6
±0.8
± 1.3
±0.01
±0.01
W
±0.1
68.2
45.1
51.3
29.93
29.12
3.0
2014
±0.3
±0.3
±0.5
±<0.01
± <0.01
W
±0.1
San Jose, California - SJJCA4
Sample
Days
61.0
47.7
66.0
30.01
29.96
5.8
(64)
±0.5
±0.5
± 1.0
±0.01
±0.01
WNW
±0.2
61.1
48.3
66.9
30.01
29.96
5.6
2014
±0.2
±0.2
±0.4
±<0.01
± <0.01
WNW
±0.1
'Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
2Wind parameters, relative humidity, temperature, and station pressure were measured at CELA for part of the year. Hie remaining
portion was obtained from the closest NWS weather station located at Downtown L.A./USC Campus, WBAN 93134, and used as a
surrogate. Data for the remaining parameters are from the NWS station.
3Dew point temperature and sea level pressure were not measured at RUCA. This information was obtained
from the closest NWS weather station located at Riverside Municipal Airport, WBAN 03171.
4Pressure was not measured at S.T.TCA. This information was obtained from the NWS weather station located at San Jose
International Airport, WBAN 23293. Wind parameters and temperature were measured year-round at S.T.TCA. Dew point and relative
humidity measurements were also collected at S.T.TCA, but were less complete; thus, the remaining portion was obtained from the NWS
weather station and used as a surrogate.
As expected, conditions in 2014 were cooler near SJJCA than near the two sites located
farther south. For the two southern California sites, average temperatures tended to be slightly
higher for RUCA, which is farther inland than CELA. Wind speeds tended to be higher at SJJCA
than the other two sites.
6.2.2 Wind Rose Comparison
Hourly surface wind data were uploaded into a wind rose software program to produce
customized wind roses, as described in Section 3.4.2. A wind rose shows the frequency of wind
directions using "petals" positioned around a 16-point compass, and uses different colors to
represent wind speeds. Figure 6-7 presents two wind roses for the CELA monitoring site. The
first is a wind rose representing wind observations for all of 2014 and the second is a wind rose
6-12
-------
representing wind observations for days on which samples were collected in 2014. These can be
used to identify the predominant wind speed and direction for 2014 and to determine if wind
observations on sample days were representative of conditions experienced over the entire year.
Figures 6-8 and 6-9 present the full-year and sample day wind roses for RUCA and SJJCA. The
wind roses for the NATTS sites in California represent wind observations collected at each site,
as available in AQS.
Figure 6-7. Wind Roses for the Wind Data Collected at CELA
NORTH
NORTH
SOUTH
2014 Wind Rose Sample Day Wind Rose
WIND SPEED
(Knots)
I I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1
Calms: 0.85%
WIND SPEED
(Knots)
I I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 0.74%
Observations from Figure 6-7 for CELA include the following:
• The 2014 wind rose shows that northeasterly winds were observed the most over the
course of the year. However, winds from the south-southwest, west-southwest, and
west were also observed frequently. Calm winds were infrequently observed as were
wind speeds greater than 11 knots. Higher wind speeds were most often observed
with a west-southwesterly wind direction.
• The sample day wind rose resembles the full-year wind rose, exhibiting similar wind
speed and direction patterns, indicating that wind conditions on sample days were
representative of those observed throughout the year near CELA.
6-13
-------
Figure 6-8. Wind Roses for the Wind Data Collected at RUCA
2014 Wind Rose Sample Day Wind Rose
a 4% ;
1 1
wm
5^4%
: EAST
jV
\/ESf
¥
'¦ EAST
WIND SPEED
(Knots)
¦I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 13.28%
WIND SPEED
(Knots)
¦I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 11.91%
Observations from Figure 6-8 for RUCA include the following:
• The 2014 wind rose shows that westerly and west-northwesterly winds were observed
the most at RUCA over the course of the year. Winds from the west to northwest to
north account for the majority of the wind observations near RUCA while winds from
the northeast, southeast, and southwest quadrants were infrequently observed. Calm
winds accounted for 13 percent of the observations while wind speeds greater than
11 knots accounted for few observations. Higher wind speeds were most often
observed with a northerly wind direction.
• The sample day wind rose resembles the full-year wind rose, exhibiting similar wind
speed and direction patterns, indicating that winds on sample days were
representative of those observed throughout the year near RUCA.
6-14
-------
Figure 6-9. Wind Roses for the Wind Data Collected at SJJCA
2014 Wind Rose
Sample Day Wind Rose
WEST
WIND SPEED
(Knots)
¦I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 1.85%
WIND SPEED
(Knots)
¦I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 1.69%
Observations from Figure 6-9 for SJJCA include the following:
• Winds from the west-northwest and northwest account for more than 40 percent of
wind observations near SJJCA. Winds from the east-southeast and southeast make up
one quarter of wind observations. Winds from the northeast and southwest quadrants
were rarely observed. Calm winds account for less than 2 percent of the observations.
The wind patterns shown on the sample day wind rose resemble the wind patterns
shown on the full-year wind rose, indicating that conditions on sample days were
representative of those experienced over the entire year.
6-15
-------
6.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each
California monitoring 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. 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 6-3. 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 6-3. 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 6-3. Risk-Based Screening Results for the California Monitoring Sites
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Los Angeles, California - CELA
Naphthalene
0.029
55
56
98.21
98.21
98.21
Benzo(a)pyrene
0.00057
1
41
2.44
1.79
100.00
Total
56
97
57.73
Rubidoux, California - RUCA
Naphthalene
0.029
50
59
84.75
100.00
100.00
Total
50
59
84.75
San Jose, California - SJJCA
Arsenic (PMio)
0.00023
48
57
84.21
47.06
47.06
Naphthalene
0.029
42
59
71.19
41.18
88.24
Nickel (PMio)
0.0021
6
61
9.84
5.88
94.12
Benzo(a)pyrene
0.00057
4
25
16.00
3.92
98.04
Acenaphthene
0.011
1
58
1.72
0.98
99.02
Fluorene
0.011
1
45
2.22
0.98
100.00
Total
102
305
33.44
Observations from Table 6-3 include the following:
• Concentrations of naphthalene failed the majority of screens for CELA, accounting
for 55 of the 56 failed screens for this site, while benzo(a)pyrene concentrations failed
a single screen. Thus, naphthalene is the only pollutant identified as a pollutant of
interest for CELA.
6-16
-------
• Naphthalene was the only PAH to fail screens for RUCA; thus, naphthalene is the
only pollutant of interest for this site. Naphthalene was detected in all 59 valid PAH
samples collected at RUCA and failed screens for 50 of these, representing an
85 percent failure rate.
• SJJCA is the only site that sampled metals (PMio) in addition to PAHs. For SJJCA,
concentrations of arsenic account for the majority of failed screens for the site (48),
although naphthalene concentrations also contributed to a large number of the total
failed screens (42). Together, these two pollutants account for 88 percent of SJJCA's
total failed screens. Nickel accounts for another 6 percent of the total failed screens
for this site and benzo(a)pyrene accounts for another 4 percent. These four pollutants
contributed to more than 95 percent of failed screens for SJJCA and were therefore
identified as pollutants of interest for this site. Acenaphthene and fluorene also failed
a single screen each for SJCCA but were not identified as pollutants of interest.
6.4 Concentrations
This section presents various concentration averages used to characterize air toxics
pollution 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.
• 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 are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at the California monitoring sites are provided in Appendices M and N.
6.4.1 2014 Concentration Averages
Quarterly and annual concentration averages 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
6-17
-------
calculated. An annual average concentration includes all measured detections and substituted
zeros for non-detects for the entire year of sampling. Annual averages were calculated for
pollutants where three valid quarterly averages could be calculated 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 6-4, 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.
Table 6-4. Quarterly and Annual Average Concentrations of the Pollutants of Interest
for the California Monitoring Sites
# of
Measured
1st
2nd
3rd
4th
Detections
Total
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs.
# >MDL
# of
Samples
Average
(ng/m3)
Average
(ng/m3)
Average
(ng/m3)
Average
(ng/m3)
Average
(ng/m3)
Los Angeles, California - CELA
142.32
64.55
78.08
88.54
Naphthalene
56/56
56
±43.53
NA
± 11.78
± 15.58
± 13.86
Rubidoux, California - RUCA
120.46
61.57
45.56
72.74
75.23
Naphthalene
59/59
59
± 47.44
±27.32
± 10.36
± 13.41
± 15.07
San Jose, California - SJJCA
0.53
0.23
0.49
0.51
0.44
Arsenic (PMio)
57/52
61
±0.18
±0.10
±0.17
±0.11
±0.07
0.16
0.20
<0.01
0.08
0.11
Benzo(a)pyrene
25/20
59
±0.13
±0.30
±0.01
±0.07
±0.07
93.33
38.59
31.03
72.96
59.91
Naphthalene
59/59
59
±48.41
± 10.34
±7.72
± 20.62
± 14.59
1.24
2.24
1.18
1.02
1.42
Nickel (PMio)
61/60
61
±0.57
± 1.16
±0.17
±0.36
±0.34
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
Observations for the California monitoring sites from Table 6-4 include the following:
• Naphthalene was identified as a pollutant of interest for all three sites. The annual
average concentration of naphthalene is highest for CELA and lowest for SJJCA.
• For each site, naphthalene concentrations appear highest during the first quarter of
2014, based on the quarterly averages, and lowest during the third quarter. However,
the confidence intervals calculated for each of the first quarter averages are relatively
large, indicating that there is considerable variability in the measurements.
• Note that CELA does not have a second quarter average concentration for
naphthalene shown in Table 6-4. This is a result of the invalidation of several PAH
6-18
-------
samples in April and early May resulting from laboratory equipment issues, as
discussed in Section 4.2.2. While this issue affected all sites that sampled
naphthalene, some sites had enough valid samples for the second quarter to still meet
the minimum criteria for calculating a quarterly average concentration while others
did not. The number of samples affected varies from three samples (RUCA and
SJJCA) to four samples (CELA).
Naphthalene concentrations measured at CELA range from 18.1 ng/m3 to 254 ng/m3
with a median concentration of 73.40 ng/m3. All six of CELA's naphthalene
concentrations greater than 150 ng/m3 were measured in either January or February.
However, three of the concentrations measured in March are among the lowest 10
concentrations measured at CELA. This helps explain the large confidence interval
shown for CELA's first quarter naphthalene concentration.
Naphthalene concentrations measured at RUCA range from 16.3 ng/m3 to 281 ng/m3
with a median concentration of 64.10 ng/m3. The maximum naphthalene
concentration measured at RUCA is the eighth highest concentration of naphthalene
among NMP sites sampling this pollutant. All five of RUCA's naphthalene
concentrations greater than 150 ng/m3 were measured in either January or February.
However, the minimum concentration was also collected during the first quarter
(March) as was the third-lowest concentration measured at RUCA. This helps explain
the large confidence interval shown for RUCA's first quarter average naphthalene
concentration. Of the 25 naphthalene concentrations less than 50 ng/m3 measured at
RUCA, the third quarter has the most (10) and the fourth quarter has the fewest (3).
Only three concentrations greater than the median naphthalene concentration were
measured during the third quarter at RUCA.
Naphthalene concentrations measured at SJJCA range from 14.7 ng/m3 to 279 ng/m3;
the maximum naphthalene concentration measured at SJJCA is the 10th highest
concentration of naphthalene among NMP sites sampling this pollutant. Yet, the
median naphthalene concentration of 38.40 ng/m3 is roughly half the median
concentration of the other two sites. Four of the five naphthalene concentrations
greater than 150 ng/m3 were measured at SJJCA in January 2014. These four were
measured on the first four sample days of 2014, after which another naphthalene
concentration greater than 100 ng/m3 was not measured again until October. This is
reflected in SJJCA's quarterly average concentrations.
Benzo(a)pyrene is also a pollutant of interest for SJJCA. The quarterly average
concentrations exhibit considerably variability, both in the magnitude of the averages
as well as the confidence intervals shown. Of the 59 valid concentrations measured at
SJJCA, 34 of them were non-detects. The number of non-detects measured during
each quarter varies from five (fourth quarter) to 14 (third quarter), with only one
measured detection for the third quarter of 2014, explaining why the third quarter
average concentration is so low compared to the other quarterly averages. The
measured detections of benzo(a)pyrene measured at SJJCA range from 0.021 ng/m3
to 1.76 ng/m3, which is the third highest measurement across NMP sites sampling this
pollutant. Both the minimum and maximum measured detections of benzo(a)pyrene
were measured during the second quarter of 2014, along with nine non-detects,
6-19
-------
explaining the relatively large confidence interval calculated for the second quarter
average concentration.
• Arsenic and nickel are also pollutants of interest for SJJCA. There were four non-
detects of arsenic measured at SJJCA, with the remaining concentrations ranging
from 0.08 ng/m3 to 1.47 ng/m3. All four non-detects were measured during the second
quarter of 2014. In addition, six of the 10 arsenic concentrations less than 0.25 ng/m3
were measured during the second quarter, including the minimum concentration
(although a measurement of the same magnitude was also measured at the end of
February). This explains the relatively low average shown for the second quarter
compared to the other quarterly average concentrations of this pollutant.
• Concentrations of nickel measured at SJJCA range from 0.17 ng/m3 to 9.73 ng/m3,
which is the maximum nickel concentration measured across NMP sites sampling this
pollutant. The median nickel concentration for SJJCA is 1.19 ng/m3. A review of the
quarterly average concentrations shows that the second quarter average is
considerably higher than the others and has a relatively high confidence interval
associated with it. Six of the 10 highest nickel concentrations measured at SJJCA
were measured during the second quarter, including the maximum concentration.
Also, of the 24 nickel concentrations less than 1 ng/m3 measured at SJJCA, only one
was measured during the second quarter of 2014.
Tables 4-9 through 4-12 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 and RUCA both appear in Table 4-11 for naphthalene, ranking fifth and
eighth, respectively.
• SJJCA appears twice in Table 4-12 for PMio metals. SJJCA has the fifth highest
annual average concentration of nickel and tenth highest annual average
concentration of arsenic among NMP sites sampling PMio metals.
6.4.2 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 6-4 for CELA, RUCA, and SJJCA. Figures 6-10 through 6-13 overlay the 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 3.4.3.1,
and are discussed below.
6-20
-------
Figure 6-10. Program vs. Site-Specific Average Arsenic (PMio) Concentration
-
Program Max Concentration = 10.1 ng/m3
Concentration (ng/m3]
Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 6-10 presents the box plot for arsenic (PMio) for SJJCA and shows the following:
• The program-level maximum arsenic concentration (10.1 ng/m3) is not shown directly
on the box plot in Figure 6-10 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 plots has been reduced.
• The maximum arsenic concentration measured at SJJCA is an order of magnitude less
than the maximum concentration measured across the program.
• The annual average arsenic concentration for SJJCA is less than the program-level
average concentration and similar to the program-level median concentration of
arsenic.
• The minimum concentration measured at SJJCA is zero, indicating that at least one
non-detect of arsenic was measured at SJJCA. Four non-detects of arsenic were
measured at SJJCA.
Figure 6-11. Program vs. Site-Specific Average Benzo(a)pyrene Concentration
Program Max Concentration = 3.19 ng/m3
1
T
0 0.5 1 1.5 2 2.5
Concentration (ng/m3)
Program: IstQuartile
2ndQuartile 3rdQuartile
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
6-21
-------
Figure 6-11 presents the box plot for benzo(a)pyrene for SJJCA and shows the following:
• Similar to arsenic, the scale of the box plot in Figure 6-11 has also been reduced to
allow for the observation of data points at the lower end of the concentration range.
• While the maximum benzo(a)pyrene concentration measured at SJJCA is roughly half
the maximum concentration measured across the program, it is the third highest
measurement of this pollutant across the program.
• The annual average benzo(a)pyrene concentration for SJJCA is similar to the
program-level average concentration. Note that the majority of the benzo(a)pyrene
concentrations measured across the program, as indicated by the first, second
(median), and third quartiles, fall into a relatively small range of measurements (less
than or equal to 0.133 ng/m3). SJJCA is one of only two NMP sites for which
benzo(a)pyrene is a pollutant of interest (BXNY is the other).
• The minimum concentration measured at SJJCA is zero, indicating that at least one
non-detect was measured at SJJCA; 34 non-detects of benzo(a)pyrene were measured
at SJJCA.
Figure 6-12. Program vs. Site-Specific Average Naphthalene Concentrations
-
O i
vJ 1
-
kJ
U
m
KJ
i
0 100 200 300 400 500 600
Concentration (ng/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range
6-22
-------
Figure 6-12 presents the box plots for naphthalene for all three sites and shows the
following:
• The range of naphthalene measurements is similar across the California monitoring
sites.
• The annual average concentration is highest for CELA, followed by RUCA and then
SJJCA. The annual average naphthalene concentration for CELA is greater than the
program-level average concentration and just greater than the program-level third
quartile. RUCA's annual average is greater than the program-level average
concentration while SJJCA's annual average is less than the program-level average
concentration.
• There were no non-detects of naphthalene measured at CELA, RUCA, SJJCA, or
across the program.
Figure 6-13. Program vs. Site-Specific Average Nickel (PMio) Concentration
-
O i
1
0 2 4 6 8 10
Concentration {ng/m3)
Progra m: 1st Qua rti 1 e
¦
2nd Quartile 3rd Quartile
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 6-13 presents the box plot for nickel for SJJCA and shows the following:
• The maximum nickel concentration measured across the program was measured at
SJJCA.
• SJJCA's annual average nickel concentration is greater than the program-level
average concentration and just greater than the program-level third quartile. Recall
from the previous section that SJJCA has the fifth highest annual average
concentration of nickel among NMP sites sampling PMio metals.
• There were no non-detects of nickel measured at SJJCA.
6-23
-------
6.4.3 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 3.4.3.2.
Both CELA and RUCA began sampling PAHs under the NMP in 2007. SJJCA began sampling
PAHs and metals under the NMP in 2008. Thus, Figures 6-14 through 6-20 present the 1-year
statistical metrics for each of the pollutants of interest first for CELA, then for RUCA, 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 6-14. Yearly Statistical Metrics for Naphthalene Concentrations Measured at CELA
JL
o
o...
2010 2011
Year
o 5th Percentile - Minimum
— Maximum
o 95th Percentile Averege
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2007.
Observations from Figure 6-14 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.
6-24
-------
• The smallest range of concentrations was measured in 2007, although the statistical
metrics do not represent a full year of sampling. The minimum concentration
measured at CELA was measured in 2007 (1.30 ng/m3); in addition, 2007 is the only
year in which a concentration less than 15 ng/m3 was measured (there were five in
total). The range of naphthalene measurements increased considerably in 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, 95th percentile, 1-year average, and
median concentrations decrease from 2009 to 2010 and again for 2011.
• All of the statistical parameters shown in Figure 6-14 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 initial year 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 and, with the
exception of the 5th percentile, are at a minimum for 2014 since the first full year of
sampling. 2014 is the first year that the 1-year average concentration is less than
100 ng/m3.
Figure 6-15. Yearly Statistical Metrics for Naphthalene Concentrations Measured at RUCA
o
T
o
2010 2011
Year
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 May 2007.
6-25
-------
Observations from Figure 6-15 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 smallest range of measurements was collected in 2007, similar to the observation
made for CELA, although the statistical metrics do not represent a full year of
sampling.
• The maximum naphthalene concentration was measured at RUCA in 2009
(406 ng/m3), although another concentration of similar magnitude was also measured
at RUCA in 2013. Naphthalene concentrations greater than 300 ng/m3 have been
measured at least once every year since 2009 until 2014.
• The 1-year average concentration has an increasing trend through 2012, although
2010 was down slightly. The median concentration has a similar pattern. After 2012,
these parameters decrease somewhat. The 1-year average concentration for 2014 is
the lowest it has been since 2008.
• The range of concentrations measured at RUCA reflects a relatively high level of
variability in the concentrations measured. For 2009, 2012, and 2013, the maximum
concentration is twice the 95th percentile. For these years, more than 100 ng/m3
separates the maximum concentration and the next highest concentration measured.
6-26
-------
Figure 6-16. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at SJJCA
Maximum
Observations from Figure 6-16 for arsenic concentrations measured at SJJCA include the
following:
• The maximum concentration of arsenic (3.09 n/gm3) was measured on the first day of
sampling at SJJCA (January 1, 2008). The only other arsenic concentration greater
than 2 ng/m3 was measured in 2013. All but one of the seven arsenic concentrations
greater than 1.5 ng/m3 were measured in 2008 (two) or 2013 (four).
• The 1-year average arsenic concentration decreased from 2008 to 2009. Although this
is due in part to 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. In
2009, two non-detects were measured at SJJCA, compared to none in 2008. In
addition, seven arsenic concentrations less than 0.1 ng/m3 were measured in 2009
compared to only two in 2008.
• Between 2010 and 2012, the range of concentrations measured changed little and the
1-year average arsenic 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), all 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 and
third 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
6-27
-------
2013, the most for any year of sampling (none of the previous years had more than
six).
• For 2014, the range of arsenic concentrations measured returned to previous levels,
with a 1-year average concentration that falls between 2012 and 2013 levels
(0.44 ng/m3).
Figure 6-17. Yearly Statistical Metrics for Benzo(a)pyrene Concentrations Measured at SJJCA
2.0
1.8
1.6
1.4
1.2
E
>-2
j„
Q 0.8
0.6
0.4
0.2
0.0
2008 1 2009 2010 2011 2012 2013 2014
Year
O 5th Percentile - Minimum — Median - Maximum o 95th Percentile Average
Observations from Figure 6-17 for benzo(a)pyrene 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 median benzo(a)pyrene concentration is zero for all years of PAH sampling at
SJJCA, indicating that at least half of the measurements were non-detects. The
percentage of non-detects has ranged from 58 percent (2013 and 2014) to 83 percent
(2010).
• The maximum benzo(a)pyrene concentration measured in 2014 (1.76 ng/m3) is the
maximum concentration measured since the onset of sampling at SJJCA. No other
concentration greater than 1 ng/m3 has been measured at this site, although one has
come close. 2014 has the highest number of concentrations greater than 0.5 ng/m3
6-28
-------
(five) compared to previous years, with three in 2011 and one each in all other years
except 2010, when none were measured.
Figure 6-18. Yearly Statistical Metrics for Naphthalene Concentrations Measured at SJJCA
~
o
2011
Year
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 6-18 for naphthalene concentrations measured at SJJCA
include the following:
• 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.
• 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.
• The median concentration has 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). The 1-year average concentration
exhibits more variability, having an undulating pattern from year-to-year, ranging
from 63.44 ng/m3 (2010) to 81.04 ng/m3 (2009) through 2012, then increasing to
94.13 ng/m3 for 2013.
Both the 1-year average and median concentrations are at a minimum for 2014.
6-29
-------
Figure 6-19. Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at SJJCA
Maximum
Observations from Figure 6-19 for nickel concentrations measured at SJJCA include the
following:
• The maximum concentration of nickel was measured at SJJCA on April 17, 2014
(9.73 ng/m3) and is more than twice the next highest nickel concentration measured at
this site (4.66 ng/m3 measured in 2013). The maximum concentration of nickel has a
steady increasing trend between 2009 and 2014.
• Both the 1-year average and median concentrations have a 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 concentrations
measured in 2011 were higher than the preceding year as the number of nickel
concentrations greater than 1 ng/m3 more than doubled, accounting for more than half
of the measurements in 2011 (compared to a quarter of the measurements in 2010).
• 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. Despite the continued increase in the maximum
concentration and 95th percentile, little change is shown for 2014.
6-30
-------
6.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at each California monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4
for definitions and explanations regarding the various toxicity factors, time frames, and
calculations associated with these risk-based screenings.
6.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the California 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 3.4.3.3 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 6-5, 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 6-5 include the following:
• Naphthalene has the highest (or only) annual average concentration for each of the
California monitoring sites among the site-specific pollutants of interest, as discussed
in the previous section. The annual average concentration for CELA is the highest of
the three annual averages for naphthalene, followed by the annual average for RUCA
and then SJJCA.
• Naphthalene also has the highest cancer risk approximation among the site-specific
pollutants of interest for the California monitoring sites. The cancer risk
approximations for naphthalene range from 2.04 in-a-million for SJJCA to 3.01 in-a-
million for CELA.
• SJJCA is the only site with pollutants of interest other than naphthalene. Among the
remaining pollutants, arsenic is the only other pollutant of interest for SJJCA with a
cancer risk approximation greater than 1 in-a-million (1.89 in-a-million).
• All of the noncancer hazard approximations for the pollutants of interest for the
California monitoring sites are less than 1.0, where noncancer RfCs are available,
indicating that no adverse noncancer health effects are expected from these individual
pollutants.
6-31
-------
Table 6-5. Risk Approximations for the California Monitoring Sites
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Los Angeles, California - CELA
Naphthalene
0.000034
0.003
56/56
88.54
± 13.86
3.01
0.03
Rubidoux, California - RUCA
Naphthalene
0.000034
0.003
59/59
75.23
± 15.07
2.56
0.03
San Jose, California - SJJCA
Arsenic (PMio)
0.0043
0.000015
57/61
0.44
±0.07
1.89
0.03
Benzo(a)pyrene
0.00176
25/59
0.11
±0.07
0.19
Naphthalene
0.000034
0.003
59/59
59.91
± 14.59
2.04
0.02
Nickel (PMio)
0.00048
0.00009
61/61
1.42
±0.34
0.68
0.02
— = A Cancer URE or Noncancer RfC is not available.
6.5.2 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 6-6 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
that have cancer toxicity factors. Table 6-6 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 6-6 provides the pollutants of interest with the highest cancer risk approximations (in-a-
million) for each site, as presented in Table 6-5. The emissions, toxicity-weighted emissions, and
cancer risk approximations are shown in descending order in Table 6-6. Table 6-7 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 3.4.3.4. Similar to the cancer risk and
6-32
-------
noncancer hazard approximations provided in Section 6.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
Observations from Table 6-6 include the following:
• Formaldehyde and benzene are the highest emitted pollutants with cancer UREs in
Los Angeles and Riverside Counties while benzene is emitted in slightly higher
quantities than formaldehyde in Santa Clara County. The quantity of emissions 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 all three counties. POM, Group la, benzene, and 1,3-butadiene rank
behind formaldehyde for Los Angeles County; benzene, hexavalent chromium, and
POM, Group la rank behind formaldehyde for Riverside County; and benzene, POM,
Group 2b, and hexavalent chromium rank behind formaldehyde for Santa Clara
County.
• Six of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Los Angeles County, while there are five in common for Riverside and
Santa Clara Counties.
• Naphthalene has the highest (or only) cancer risk approximation for all three
California sites. Naphthalene appears on both emissions-based lists for all three
counties.
• Arsenic and nickel do not appear on either emissions-based list for Santa Clara
County (they rank lower than tenth). Hexavalent chromium is the only metal shown
for Santa Clara County, ranking fourth highest for its toxicity-weighted emissions.
• Several POM Groups appear among the pollutants with the highest toxicity-weighted
emissions for each county. POM, Group 2b includes acenaphthene and fluorene, both
of which failed screens for SJJCA but were not identified as pollutants of interest.
POM, Group 5a includes benzo(a)pyrene, which failed one screen for CELA and was
identified as a pollutant of interest for SJJCA. POM, Group 5a ranks fifth for its
toxicity-weighted emissions but is not among the highest emitted in Santa Clara
County. POM, Group la, which appears among each county's toxicity-weighted
emissions, includes unspeciated compounds.
6-33
-------
Table 6-6. 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
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)
Los Angeles, California (Los Angeles County) - CELA
Formaldehyde
2,221.45
Formaldehyde
2.89E-02
Naphthalene
3.01
Benzene
1,913.13
POM, Group la
1.49E-02
Dichloro methane
1,682.67
Benzene
1.49E-02
Ethylbenzene
1,101.33
1,3-Butadiene
9.87E-03
T etrachloroethylene
1,076.88
POM, Group 2b
7.27E-03
Acetaldehyde
962.00
POM, Group 5a
6.02E-03
/?-Dichlorobcnzcnc
339.36
POM, Group 2d
5.84E-03
1.3 -Butadiene
328.83
Naphthalene
5.27E-03
POM, Group la
169.60
/?-Dichlorobcnzcnc
3.73E-03
Naphthalene
154.91
Hexavalent Chromium
3.03E-03
Rubidoux, California (Riverside County) - RUCA
Formaldehyde
418.81
Formaldehyde
5.44E-03
Naphthalene
2.56
Benzene
317.30
Benzene
2.47E-03
T etrachloroethylene
214.39
Hexavalent Chromium
2.04E-03
Dichloro methane
200.68
POM, Group la
1.88E-03
Acetaldehyde
197.01
1,3-Butadiene
1.47E-03
Ethylbenzene
191.03
POM, Group 2b
1.45E-03
/?-Dichlorobcnzcnc
70.48
POM, Group 5a
1.20E-03
1.3 -Butadiene
48.84
Naphthalene
1.19E-03
Naphthalene
34.99
POM, Group 2d
1.09E-03
1,3 -Dichloropropene
29.57
/j-Dichlorobcnzcne
7.75E-04
-------
Table 6-6. 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
Cancer UREs
Top 10 Cancer Toxicity-Weighted
Emissions
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(County-Level)
(County-Level)
(Site-Specific)
Cancer
Cancer Risk
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(in-a-million)
San Jose, California (Santa Clara County) - SJJCA
Benzene
356.17
Formaldehyde
4.46E-03
Naphthalene
2.04
Formaldehyde
342.81
Benzene
2.78E-03
Arsenic
1.89
Ethylbenzene
232.74
POM, Group 2b
1.73E-03
Nickel
0.68
Dichloro methane
191.47
Hexavalent Chromium
1.67E-03
Benzo(a)pyrene
0.19
Acetaldehyde
171.62
POM, Group 5a
1.63E-03
T etrachloroethylene
110.40
1,3-Butadiene
1.35E-03
/?-Dichlorobcnzcnc
60.37
POM, Group 2d
1.32E-03
1.3 -Butadiene
45.07
Naphthalene
1.26E-03
Naphthalene
37.18
POM, Group la
1.21E-03
T richloroethylene
29.51
/?-Dichlorobcnzcnc
6.64E-04
-------
Table 6-7. 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 Ha
Based on Annual Av<
(Site-S
izard Approximations
;rage Concentrations
>ecific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Los Angeles, California (Los Angeles County) - CELA
Toluene
8,265.39
Acrolein
6,797,409.70
Naphthalene
0.03
1,1,1 -T richloroethane
6,903.37
Chlorine
230,010.81
Xylenes
4,970.97
Formaldehyde
226,678.82
Hexane
4,520.90
1,3-Butadiene
164,416.69
Formaldehyde
2,221.45
Acetaldehyde
106,888.65
Benzene
1,913.13
Benzene
63,771.13
Dichloromethane
1,682.67
Cyanide Compounds, PM
63,440.92
Ethylene glycol
1,465.20
T richloroethylene
56,352.54
Methanol
1,338.85
Naphthalene
51,636.02
Ethylbenzene
1,101.33
Xylenes
49,709.73
Rubidoux, California (Riverside County) - RUCA
Toluene
1,541.54
Acrolein
1,151,923.43
Naphthalene
0.03
Xylenes
1,037.06
Chlorine
71,489.03
Hexane
1,034.89
Formaldehyde
42,736.21
1,1,1 -T richloroethane
617.84
1,3-Butadiene
24,417.60
Formaldehyde
418.81
Acetaldehyde
21,889.50
Benzene
317.30
Bromo methane
13,246.82
Ethylene glycol
241.17
Naphthalene
11,663.14
Methanol
218.85
Lead, PM
11,143.30
T etrachloroethylene
214.39
Benzene
10,576.82
Dichloromethane
200.68
T richloroethylene
10,486.48
-------
Table 6-7. 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 Ha
Based on Annual Av<
(Site-S
izard Approximations
;rage Concentrations
>ecific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
San Jose, California (Santa Clara County) - SJJCA
Toluene
1,762.28
Acrolein
1,804,553.18
Arsenic
0.03
1,1,1 -T richloroethane
1,289.63
Chlorine
91,338.84
Naphthalene
0.02
Hexane
1,014.84
Formaldehyde
34,980.53
Nickel
0.02
Xylenes
987.31
1,3-Butadiene
22,537.16
Benzene
356.17
Acetaldehyde
19,068.78
Formaldehyde
342.81
T richloroethylene
14,754.18
Ethylene glycol
280.57
Naphthalene
12,392.06
Ethylbenzene
232.74
Benzene
11,872.49
Methanol
216.21
Xylenes
9,873.13
Dichloromethane
191.47
Lead, PM
9,571.88
-------
Observations from Table 6-7 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. 1,1,1-Trichloroethane is the second highest
emitted pollutant in Los Angeles and Santa Clara Counties but ranks fourth for
Riverside County. Xylenes are the second highest emitted pollutant in Riverside
County but ranks third and fourth for Los Angeles and Santa Clara Counties,
respectively. Hexane is also among the top four emitted pollutants in each of these
counties.
• Acrolein, chlorine, formaldehyde, 1,3-butadiene, and acetaldehyde are the five
pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) for all three counties. Although acrolein and chlorine rank highest
for toxicity-weighted emissions for each county, neither pollutant appears among the
highest emitted. This is also true for acetaldehyde and 1,3-butadiene. Conversely,
formaldehyde has the fifth highest emissions for Los Angeles and Riverside Counties
and ranks sixth for Santa Clara County.
• Three of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Los Angeles and Santa Clara Counties, while only two of the highest
emitted pollutants also have the highest toxicity-weighted emissions for Riverside
County.
• Naphthalene is the only pollutant of interest for all three sites. Naphthalene does not
appear among the highest emitted pollutants (of those with a noncancer RfC) for any
of the three counties. Naphthalene ranks seventh for its toxicity-weighted emissions
for Riverside and Santa Clara Counties and ninth for its toxicity-weighted emissions
for Los Angeles County.
• Arsenic and nickel are the only other pollutants of interest for SJJCA for which
noncancer hazard approximations could be calculated. Lead is the only metal that
appears on either emissions-based list for Santa Clara County in Table 6-7.
Concentrations of lead did not fail screens for SJJCA.
Summary of the 2014 Monitoring Data for the California Monitoring Sites
Results from several of the data analyses described in this section include the following:
~~~ Naphthalene failed screens for all three California sites. One additional PAHfailed
screens for CELA and three additional PAHs and two PMw metals failed screens for
SJJCA. Naphthalene was identified as a pollutant of interest for all three sites.
~~~ Naphthalene had the highest annual average concentration among the site-specific
pollutants of interest for each of the California monitoring sites. CELA has the fifth
highest annual average concentration of naphthalene among NMP sites sampling
PAHs. The highest concentrations of naphthalene were measured at these sites in
January and/or February.
6-38
-------
The maximum nickel concentration across the program was measured at SJJCA.
Naphthalene concentrations have a decreasing trend at CELA in recent years.
Progressively higher nickel concentrations have been measured at SJJCA over the
last several years of sampling.
Naphthalene has the highest cancer risk approximation of the pollutants of interest
for each site. None of the pollutants of interest for the California sites have
noncancer hazard approximations greater than an HQ of 1.0.
6-39
-------
7.0 Sites in Colorado
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS and UATMP sites in Colorado, and integrates these
concentrations with emissions, meteorological, and risk information. Data generated by sources
other than ERG are not included in the data analyses contained in this report. Readers are
encouraged to refer to Sections 1 through 4 and the glossary (Appendix P) 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 geographical and
physical information about the location of the sites and the surrounding areas. 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 NATTS site in Colorado is located in Grand Junction (GPCO) while the five
UATMP sites are located in neighboring Garfield County, in the towns of Battlement Mesa
(BMCO), Silt (BRCO), Parachute (PACO), Carbondale (RFCO), and Rifle (RICO). Figure 7-1
for GPCO is 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 2011 NEI for point sources, version 2.
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 in order to emphasize emissions sources within the boundary. Figures 7-3 through 7-9
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.
7-1
-------
Figure 7-1. Grand Junction, Colorado (GPCO) Monitoring Site
;OurayrAve'
Ouray'Ave
-.Ouray; Ave.
Grand Ave
Grand Ave
Grand Ave
.White Avo"
Rood 'Ave1
Rood Ave
W.-MaiTTSt
Colorado Ave
WiGp|or<
1C3|
"Pitkin Ave
Wiriters~Ave
. %m
-------
Figure 7-2. NEI Point Sources Located Within 10 Miles of GPCO
O 10 mile radius
County boundary
Miles
Legend
GPCO NATTS site
3°35'0"W 108°30'0"W 108°25'0"W 108°20,0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
Mesa County
Source Category Group (No. of Facilities)
"t" Airport/Airline/Airport Support Operations (9)
£ Asphalt Production/Hot Mix Asphalt Plant (2)
0 Auto Body Shop/Painters/Automotive Stores (4)
© Automotive/RV Dealership (1)
Brick, Structural Clay, or Clay Ceramics Plant (1)
B Bulk Terminals/Bulk Plants (4)
C Chemical Manufacturing Facility (3)
1 Compressor Station (1)
[XI Crematory - Animal/Human (4)
® Dry Cleaning Facility (3)
6 Electrical Equipment Manufacturing Facility (1)
E Electroplating, Plating. Polishing, Anodizing, and Coloring (2)
If Gasoline/Diesel Service Station (43)
Industrial Machinery or Equipment Plant (3)
O Institutional (school, hospital, prison, etc.) (4)
A Landfill (1)
A Metal Coating, Engraving, and Allied Services to Manufacturers (1)
<•> Metals Processing/Fabrication Facility (2)
X Mine/Quarry/Mineral Processing Facility (35)
? Miscellaneous Commercial/Industrial Facility (1)
• Oil and/or Gas Production (2)
R Plastic, Resin, or Rubber Products Plant (1)
X Rail Yard/Rail Line Operations (2)
T Textile. Yarn, or Carpet Plant (1)
* Wastewater Treatment Facility (1)
W Woodwork, Furniture, Millwork & Wood Preserving Facility (1)
7-3
-------
Figure 7-3. Battlement Mesa, Colorado (BMCO) Monitoring Site
-------
Figure 7-4. Silt, Colorado (BRCO) Monitoring Site
g e ifn'a m *
-------
Figure 7-5. Parachute, Colorado (PACO) Monitoring Site
¦pjraclwEJ
-------
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 108=10'0"W 108:5'0"W 108e0'0"W 107°55'0"W 107°50'0"W 107°45'0"W 107°40'0"W 107*35'0"W
Rio Blanco'
County l
Garfield
County
J$' Mesa I
j I County i
^ 1 ^ 1 ^ 1 1—*
108'10'0"W 108°5'0"W lOrO'CTW 107,55,0"W 107"50'0"W 107"45'0"W 107'40'0"W 107"35'0"W 107,30'0"W 107"25'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
BMCO UATMP site ^ BRCO UATMP site ^ PACO UATMP site ^ RICO UATMP site
O 10 mile radius | County boundary
Source Category Group (No. of Facilities)
*
Airport/Airline/Airport Support Operations (7)
r
Gasoline/Diesel Service Station (19)
Asphalt Production/Hot Mix Asphalt Plant (1)
a
Landfill (1)
B
Bulk Terminals/Bulk Plants (1)
X
Mine/Quarry/Mineral Processing Facility (10)
i
Compressor Station (20)
•>
Miscellaneous Commercial/Industrial Facility (1)
®
Dry Cleaning Facility (1)
•
Oil and/or Gas Production (1,057)
f
Electricity Generation via Combustion (1)
X
Rail Yard/Rail Line Operations (1)
Gas Plant (2)
TT
Telecommunications/Radio Facility (1)
7-8
-------
Figure 7-8. Carbondale, Colorado (RFCO) Monitoring Site
:GtenWoociAvei£ \
-------
Figure 7-9. NEI Point Sources Located Within 10 Miles of RI CO
107"30"0"W
107°20'0"W
107°15,0"W
Garfield
County
Eagle
County
Pitkin
County
RFCO UATMP site
O 10 mile radius
County boundary
107"10'0"W
Mesa
County
Miles
' i ; ¦
107=250"W 107°20,D"W
Legend
107°15'0"W 107°10'0"W 107°5'0"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)
t Airport/Airline/Airport Support Operations (6)
f Building/Construction (1)
1 Compressor Station (1)
XI Crematory - Animal/Human (1)
If Gasoline/Diesel Service Station (10)
O Institutional (school, hospital, prison, etc.) (1)
x Mine/Quarry/Mineral Processing Facility (3)
7-10
-------
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
16,000
1-70 near exit 75
RICO
08-045-0007
Rifle
Garfield
Glenwood Springs,
CO
39.531813,
-107.782298
Commercial
Urban/City
Center
17,000
Rte 13 connecting US-6 and 1-70
RFCO
08-045-0018
Carbondale
Garfield
Glenwood Springs,
CO
39.412278,
-107.230397
Residential
Rural
16,000
Rte 133 just south of Hwy 82
1AADT reflects 2014 data for GPCO, PACO, RFCO, and RICO (CO DOT, 2014) 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 samplers, with the PAH sampler
located just outside the shelter. The second location, which is on the roof of an adjacent 2-story
building, is comprised of the metals samplers. 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 roughly east-
west 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 are an industrial
machinery/equipment plant, a bulk terminal/bulk plant, a gasoline/diesel service station, and an
auto body shop.
Four of the five Garfield County monitoring sites are situated in towns located along 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 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. The site is
surrounded primarily by residential subdivisions. 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.
7-12
-------
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
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 1.6 miles from BMCO;
these are the two Garfield County sites 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 the site 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. There are more
than 1,000 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).
The RFCO monitoring site is the only site in Garfield County not located along the 1-70
corridor. This site is located in the southeast corner of Garfield County in Carbondale. The town
of Carbondale resides in a valley between the Roaring Fork and Crystal Rivers, north of Mt.
Sopris (Carbondale, 2016). 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 Hill, is just over one-third of a
mile north of RFCO and is visible in the top right-hand corner of Figure 7-8.
Because RFCO is 24 miles from the next closest Garfield County monitoring site, the
emissions sources surrounding RFCO are provided in a separate map in Figure 7-9. This figure
shows that the few emissions sources within 10 miles of RFCO are primarily gasoline and/or
diesel service stations. There is also a building/construction company, a compressor station, three
mine/quarry/mineral processing facilities, and an airport within a few miles of this site.
7-13
-------
In addition to providing city, county, CBSA, 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 effect concentrations measured at a given monitoring site. The traffic
volumes near RICO, RFCO, PACO, and GPCO are considerably higher than the traffic volumes
near BMCO and BRCO. Yet, the traffic volumes for all six Colorado sites rank in the bottom
half compared to the traffic volumes for other NMP sites. The traffic volume for BRCO is one of
the lowest among all NMP sites. However, this monitoring site is located in the most rural of
settings compared to the other Colorado sites.
7.2 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
sites in Colorado on sample days, as well as over the course of the year.
7.2.1 Meteorological Summary
In order to provide an overview of the meteorological conditions experienced at each
monitoring site, hourly meteorological data for 2014 were retrieved and sample day and full-year
averages developed for temperature, dew point, relative humidity, pressure, and wind speed.
Weather data from the actual monitoring site(s) were obtained from AQS, where available. If
site-specific weather data were not available in AQS, then data were obtained from NCDC for
the NWS weather station located closest to the monitoring site(s), as described in Section 3.4.2.
For the Colorado sites, site-specific data were available from GPCO and BMCO for some, but
not all, of the site-specific parameters in Table 7-2. For GPCO, data from the NWS weather
station at Walker Field/Grand Junction Regional Airport (WBAN 23066) were used for
meteorological parameters without data and/or as a surrogate for parameters without complete
observation records. The Walker Field/Grand Junction Regional Airport weather station is
located 5 miles north-northeast of GPCO. For BMCO, data from the NWS weather station at
Garfield County Regional Airport (WBAN 03016) were used where needed; the Garfield County
Regional Airport Regional Airport weather station is located 17.7 miles east-northeast of BMCO.
Meteorological observations were not available in AQS for the remaining Garfield County sites,
and therefore, weather data from the closest NWS station were used and included in Table 7-2. A
map showing the distance between the Colorado monitoring sites and the closest NWS weather
7-14
-------
station is provided in Appendix R. These data were used to determine how meteorological
conditions on sample days vary from conditions experienced throughout the year.
Table 7-2. Average Meteorological Conditions near the Colorado Monitoring Sites
Average
Type1
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(in Hg)
Average
Station
Pressure
(in Hg)
Prevailing
Wind
Direction
Average
Scalar Wind
Speed
(kt)
Grand Junction, Colorado - GPCO2
Sample
Days
56.5
29.1
45.3
29.97
25.16
3.1
(75)
±0.9
±0.6
± 1.2
±0.01
±0.01
E
±0.1
54.5
28.5
47.4
30.00
25.17
3.1
2014
±0.4
±0.3
±0.5
±0.01
±<0.01
ESE
±<0.1
Battlement Mesa, Colorado - BMCO3
Sample
Days
50.2
27.8
47.4
30.03
24.54
3.4
(59)
± 1.0
±0.7
± 1.2
±0.01
±0.01
WSW
±0.1
50.5
28.8
48.5
30.03
24.55
3.4
2014
±0.4
±0.3
±0.5
± <0.01
±<0.01
WSW
±0.1
Garfield County Regional Airport4
BRCO
48.9
27.6
52.5
30.02
24.53
4.6
(62)
± 1.1
±0.7
± 1.3
±0.01
±0.01
W
±0.3
PACO
48.5
28.4
53.7
30.03
24.54
4.6
(60)
± 1.0
±0.7
± 1.3
±0.01
±0.01
W
±0.3
RICO
48.5
28.3
53.8
30.03
24.53
4.5
(61)
± 1.1
±0.7
± 1.3
± <0.01
±0.01
W
±0.3
48.4
28.8
55.1
30.03
24.54
4.4
2014
±0.4
±0.3
±0.5
± <0.01
±<0.01
W
±0.1
Aspen-Pitkin County Regional Airport5
RFCO
42.9
24.8
54.7
29.97
22.62
5.2
(35)
± 1.3
±0.9
± 1.5
±0.01
±0.01
SSW
±0.3
42.4
25.2
57.0
29.98
22.62
5.0
2014
±0.4
±0.3
±0.5
± <0.01
±<0.01
SSW
±0.1
'Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
2Temperature and wind parameters were measured at GPCO, along with relative humidity measurements for part of the year.
Hie remaining portion was obtained from the closest NWS weather station located at Walker Field/Grand Junction Regional
Airport, WBAN 23066, and used as a surrogate. Data for the remaining parameters are from the NWS station.
3Dew point temperature and sea level pressure were not measured at BMCO. This information was obtained from the closest
NWS weather station located at Garfield County Regional Airport, WBAN 03016.
4Meteorological data for BRCO, PACO, and RICO were not available in AQS. This information was obtained from the NWS
weather station located at Garfield County Regional Airport, WBAN 03016.
'Meteorological data for RFCO were not available in AQS. This information was obtained from the NWS weather station
located at Aspen-Pitkin Comity Airport, WBAN 93073.
7-15
-------
Table 7-2 presents average temperature, average dew point temperature, average relative
humidity, average station and sea level pressure, and wind information (average scalar wind
speed and prevailing wind direction) for days on which samples were collected and for all of
2014. Also included in Table 7-2 is the 95 percent confidence interval for each parameter. As
shown in Table 7-2, average meteorological conditions on sample days were generally
representative of average weather conditions experienced throughout the year at each site. For
GPCO, the greatest difference between the sample day and full-year averages was calculated for
average relative humidity, although average temperature has a similar difference. A number of
make-up samples were collected at GPCO in 2014, particularly for PAHs. These were collected
primarily during the warmer months of the year, between April and September, and may
account for the differences shown for these parameters. Among the Garfield County sites, the
greatest differences between the sample day and full-year averages were also calculated for
relative humidity.
7.2.2 Wind Rose Comparison
Hourly surface wind data were uploaded into a wind rose software program to produce
customized wind roses, as described in Section 3.4.2. A wind rose shows the frequency of wind
directions using "petals" positioned around a 16-point compass, and uses different colors to
represent wind speeds. Figure 7-10 presents two wind roses for the GPCO monitoring site. The
first is a wind rose representing wind observations for all of 2014 and the second is a wind rose
representing wind observations for days on which samples were collected in 2014. These can be
used to identify the predominant wind speed and direction for 2014 and to determine if wind
observations on sample days were representative of conditions experienced over the entire year.
Figures 7-11 through 7-13 presents the full-year and sample day wind roses for the Garfield
County sites.
7-16
-------
Figure 7-10. Wind Roses for the Wind Data Collected at GPCO
2014 Wind Rose Sample Day Wind Rose
WIND SPEED
(Knots)
¦I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 3.00%
WIND SPEED
(Knots)
¦I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 3.12%
Observations from Figure 7-10 for GPCO include the following:
• The 2014 wind rose shows that winds from the east to southeast are the most
frequently observed wind directions at GPCO, together accounting for more than one-
third of observations. Winds from the west to northwest make up a secondary wind
grouping. Winds from the northeast and southwest quadrants were infrequently
observed. Wind speeds greater than 11 knots were rarely observed near GPCO, and
calm winds account for 3 percent of observations.
• The sample day wind patterns at GPCO resemble the full-year wind patterns,
indicating that wind conditions on sample days were representative of those
experienced over the entire year.
7-17
-------
Figure 7-11. Wind Roses for the Wind Data Collected at BMCO
2014 Wind Rose Sample Day Wind Rose
NORTH
NORTH
WEST
WEST
WIND SPEED
(Knots)
WIND SPEED
(Knots)
SOUTH
SOUTH
Calms: 3.16%
Calms: 3.34%
Observations from Figure 7-11 for BMCO include the following:
• The 2014 wind rose for BMCO shows that west-southwesterly winds were the most
frequently observed wind direction near BMCO (14.2 percent), although winds from
the east-southeast were observed nearly as frequently (13.9 percent). Winds from the
western quadrants were not observed as frequently as those from the eastern
quadrants, with winds from the south observed the least. Yet winds from the eastern
quadrants tended to be lighter than those from the western quadrants. Calm winds
accounted for about 3 percent of wind observations.
• The sample day wind patterns at BMCO resemble the full-year wind patterns,
indicating that wind conditions on sample days were representative of those
experienced over the entire year.
7-18
-------
Figure 7-12. Wind Roses for the Wind Data Collected at the Garfield County Regional
Airport Weather Station
2014 Wind Rose
BRCO Sample Day Wind Rose
WIND SPEED
(Knots)
¦ -22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
WIND SPEED
(Knots)
¦I >=22
¦ 17 -21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 36.72%
PACO Sample Day Wind Rose
WIND SPEED
(Knots)
HI >= 22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
H 1-4
Calms: 36.21%
RICO Sample Day Wind Rose
WIND SPEED
(Knots)
¦9 >=22
¦ 17-21
¦ 11-17
¦ 7-11
J 4-7
¦ 1-4
Calms: 36.91%
7-19
-------
Observations from Figure 7-12 for BRCO, PACO, and RICO include the following:
• Wind data was not collected at BRCO, PACO, and RICO (or were not available in
AQS); thus, Figure 7-12 presents the 2014 full-year wind rose for the weather station
located at the Garfield County Regional Airport and sample day wind roses based
these sites' individual sample days.
• The 2014 wind rose shows that westerly winds were the most frequently observed
wind direction at the Garfield County Airport. Winds from the east to south to west to
northwest were also observed, while northerly to northeasterly winds were
infrequently observed. Calm winds accounted for more than one-third of the
observations while the strongest winds tended to be those from the western quadrants.
• The sample day wind patterns based on each sites' sample days resemble the full-year
wind patterns, indicating that wind conditions on sample days were representative of
those experienced over the entire year.
7-20
-------
Figure 7-13. Wind Roses for the Wind Data Collected at the Aspen-Pitkin County Airport
Weather Station near RFCO
2014 Wind Rose Sample Day Wind Rose
WIND SPEED
(Knots)
I I >-22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1
Calms: 22.27%
WIND SPEED
(Knots)
I I >-22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1
Calms: 23.57%
SOUTH
SOUTH
Observations from Figures 7-13 for RFCO include the following:
• Wind data was not collected at RFCO (or were not available in AQS); thus,
Figure 7-13 presents the 2014 full-year wind rose for the weather station located at
the Aspen-Pitkin County Airport and the sample day wind roses based this site's
individual sample days.
• The 2014 wind rose shows that south-southwesterly winds were the most frequently
observed wind direction at the weather station closest to RFCO (20 percent), with
winds from the south and north each accounting for more than 12 percent of the
observations. Winds from the east or west were rarely observed. Calm winds
accounted for about 22 percent of wind observations while winds greater than
17 knots were rarely observed.
• The sample day wind patterns near RFCO generally resemble the full-year wind
patterns, although winds from the north and north-northwest were observed less often
and winds from the southeast to south were observed more often.
7-21
-------
7.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each
Colorado monitoring 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. 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 7-3. 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 7-3. 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 the Garfield County
sites.
Table 7-3. Risk-Based Screening Results for the Colorado Monitoring Sites
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Grand Junction, Colorado - GPCO
Acetaldehyde
0.45
58
58
100.00
11.55
11.55
Formaldehyde
0.077
58
58
100.00
11.55
23.11
Naphthalene
0.029
58
60
96.67
11.55
34.66
Benzene
0.13
57
57
100.00
11.35
46.02
1.3 -Butadiene
0.03
57
57
100.00
11.35
57.37
Carbon Tetrachloride
0.17
57
57
100.00
11.35
68.73
1,2-Dichloroethane
0.038
50
50
100.00
9.96
78.69
Arsenic (PMio)
0.00023
33
50
66.00
6.57
85.26
Ethylbenzene
0.4
31
57
54.39
6.18
91.43
Acenaphthene
0.011
12
60
20.00
2.39
93.82
Hexachloro-1,3 -butadiene
0.045
12
13
92.31
2.39
96.22
Benzo(a)pyrene
0.00057
7
50
14.00
1.39
97.61
Dichloromethane
60
7
57
12.28
1.39
99.00
1,2-Dibromoethane
0.0017
2
2
100.00
0.40
99.40
Fluorene
0.011
2
55
3.64
0.40
99.80
/?-Dichlorobcnzcnc
0.091
1
19
5.26
0.20
100.00
Total
502
760
66.05
7-22
-------
Table 7-3. Risk-Based Screening Results for the Colorado Monitoring Sites (Continued)
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Battlement Mesa, Colorado - BMCO
Benzene
0.13
51
51
100.00
58.62
58.62
Formaldehyde
0.077
25
27
92.59
28.74
87.36
Acetaldehyde
0.45
11
27
40.74
12.64
100.00
Total
87
105
82.86
Silt, Colorado - BRCO
Benzene
0.13
49
49
100.00
53.26
53.26
Formaldehyde
0.077
24
25
96.00
26.09
79.35
Acetaldehyde
0.45
16
25
64.00
17.39
96.74
1.3 -Butadiene
0.03
3
3
100.00
3.26
100.00
Total
92
102
90.20
Parachute, Colorado - PACO
Benzene
0.13
57
57
100.00
50.44
50.44
Formaldehyde
0.077
25
25
100.00
22.12
72.57
Acetaldehyde
0.45
14
25
56.00
12.39
84.96
1,3-Butadiene
0.03
14
14
100.00
12.39
97.35
Ethylbenzene
0.4
3
55
5.45
2.65
100.00
Total
113
176
64.20
Carbondale, Colorado - RFCO
Benzene
0.13
27
27
100.00
34.62
34.62
Formaldehyde
0.077
26
26
100.00
33.33
67.95
Acetaldehyde
0.45
18
26
69.23
23.08
91.03
1.3 -Butadiene
0.03
7
7
100.00
8.97
100.00
Total
78
86
90.70
Rifle, Colorado - RICO
Benzene
0.13
54
54
100.00
38.03
38.03
1,3-Butadiene
0.03
33
33
100.00
23.24
61.27
Formaldehyde
0.077
26
27
96.30
18.31
79.58
Acetaldehyde
0.45
16
27
59.26
11.27
90.85
Ethylbenzene
0.4
13
52
25.00
9.15
100.00
Total
142
193
73.58
Observations from Table 7-3 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.
7-23
-------
• Concentrations of 16 pollutants failed at least one screen for GPCO; 66 percent of the
concentrations for these 16 pollutants were greater than their associated risk screening
value (or failed screens).
• Eleven pollutants contributed to 95 percent of failed screens for GPCO and therefore
were identified as pollutants of interest for GPCO. These 11 include two carbonyl
compounds, six VOCs, two 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). The same three pollutants (benzene,
formaldehyde, and acetaldehyde) failed screens for each Garfield County site.
1,3-Butadiene also failed screens at four of the five sites and ethylbenzene also failed
screens for PACO and RICO.
• Benzene, formaldehyde, and acetaldehyde were identified as pollutants of interest for
all five Garfield County sites. 1,3-Butadiene was also identified as a pollutant of
interest for PACO, RFCO, and RICO. Ethylbenzene was also identified as a pollutant
of interest for RICO.
• Benzene failed 100 percent of screens for all six Colorado sites.
• Carbonyl compound samples were collected on a l-in-12 day sampling schedule at
BMCO, BRCO, PACO, and RICO, while SNMOC samples were collected on a 1-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.
Both carbonyl compound and SNMOC samples were collected on a l-in-12 day
sampling schedule at RFCO.
7.4 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.
• 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 are presented from previous years
of sampling in order to characterize concentration trends at each site.
7-24
-------
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at the Colorado monitoring sites are provided in Appendices J through N.
7.4.1 2014 Concentration Averages
Quarterly and annual concentration averages 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 the entire year of sampling. Annual averages were
calculated for pollutants where three valid quarterly averages could be calculated 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 7-4, 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.
7-25
-------
Table 7-4. Quarterly and Annual Average Concentrations of the Pollutants of Interest
for the Colorado Monitoring Sites
Pollutant
# of
Measured
Detections
vs. # >MDL
# of
Samples
1st
Quarter
Average
frig/m3)
2nd
Quarter
Average
frig/m3)
3rd
Quarter
Average
frig/m3)
4th
Quarter
Average
frig/m3)
Annual
Average
frig/m3)
Grand Junction, Colorado - GPCO
Acetaldehyde
58/58
58
2.47
±0.37
3.36
±0.56
2.89
±0.51
2.41
±0.51
2.80
±0.25
Benzene
57/57
57
1.26
±0.23
0.61
±0.09
0.73
±0.10
1.30
±0.21
0.99
±0.12
1.3 -Butadiene
57/57
57
0.21
±0.06
0.08
±0.02
0.10
±0.02
0.26
±0.06
0.17
±0.03
Carbon Tetrachloride
57/57
57
0.54
±0.09
0.60
±0.05
0.63
±0.02
0.56
±0.05
0.58
±0.03
1,2-Dichloroethane
50/49
57
0.08
±0.02
0.08
±0.01
0.06
±0.02
0.08
±0.02
0.07
±0.01
Ethylbenzene
57/57
57
0.50
±0.10
0.25
±0.05
0.39
±0.07
0.67
±0.14
0.45
±0.06
Formaldehyde
58/58
58
3.50
±0.39
3.42
±0.51
5.16
±0.82
3.45
±0.63
3.90
±0.35
Hexachloro -1,3 -butadiene
13/0
57
0.02
±0.02
0.01
±0.02
0.01
±0.02
0.02
±0.02
0.02
±0.01
Acenaphthene3
60/60
60
3.59
±0.55
NA
11.68
± 1.94
5.27
± 1.53
7.17
± 1.20
Arsenic (PMi0)a
50/33
59
0.41
±0.11
0.17
±0.10
0.15
±0.07
0.38
±0.11
0.28
±0.06
Naphthalene1
60/60
60
120.56
±39.18
NA
91.59
± 13.36
110.83
± 26.63
100.03
± 13.48
Battlement Mesa, Colorado - BMCO
Acetaldehyde
27/27
27
0.29
±0.08
NA
0.58
±0.30
0.36
±0.19
0.42
±0.11
Benzene
51/51
51
1.14
±0.21
NA
NA
0.84
±0.11
NA
Formaldehyde
27/27
27
0.52
±0.11
NA
1.18
±0.47
0.54
±0.25
0.77
±0.18
Silt, Colorado - BRCO
Acetaldehyde
25/25
25
0.30
±0.18
NA
NA
0.53
±0.16
NA
Benzene
49/49
50
0.88
±0.17
NA
NA
0.83
±0.13
NA
Formaldehyde
25/25
25
0.46
±0.26
NA
NA
0.75
±0.19
NA
Parachute, Colorado - PACO
Acetaldehyde
25/25
25
0.35
±0.18
0.68
±0.23
0.80
±0.26
NA
NA
Benzene
57/57
57
1.68
±0.36
1.39
±0.28
1.37
±0.28
1.52
±0.23
1.49
±0.14
1.3 -Butadiene
14/8
57
0.01
±0.01
0
0
0.10
±0.04
0.03
±0.01
Formaldehyde
25/25
25
0.67
±0.34
1.31
±0.41
1.80
±0.34
NA
NA
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
a Average concentrations provided for the pollutants below the blue line for GPCO are presented in ng/m3 for ease of viewing.
7-26
-------
Table 7-4. Quarterly and Annual Average Concentrations of the Pollutants of Interest
for the Colorado Monitoring Sites (Continued)
Pollutant
# of
Measured
Detections
vs. # >MDL
# of
Samples
1st
Quarter
Average
frig/m3)
2nd
Quarter
Average
frig/m3)
3rd
Quarter
Average
frig/m3)
4th
Quarter
Average
frig/m3)
Annual
Average
frig/m3)
Carbondale, Colorado - RFCO
Acetaldehyde
26/26
26
0.26
±0.15
0.67
±0.26
NA
0.60
±0.10
NA
Benzene
27/27
28
NA
0.37
±0.09
0.33
±0.09
0.56
±0.20
0.46
±0.09
1,3-Butadiene
7/3
28
NA
0
0
0.07
±0.05
0.03
±0.02
Formaldehyde
26/26
26
0.33
±0.16
0.89
±0.26
NA
0.72
±0.10
NA
Rifle, Colorado - RICO
Acetaldehyde
27/27
27
0.49
±0.21
0.54
±0.30
0.65
±0.27
0.38
±0.24
0.52
±0.12
Benzene
54/54
54
1.43
±0.34
NA
0.78
±0.09
1.43
±0.19
1.09
±0.14
1,3-Butadiene
33/20
54
0.12
±0.04
NA
0.03
±0.03
0.24
±0.06
0.10
±0.03
Ethylbenzene
52/51
54
0.33
±0.07
NA
0.30
±0.04
0.42
±0.10
0.32
±0.04
Formaldehyde
27/27
27
0.71
±0.32
0.69
±0.32
1.12
±0.41
0.48
±0.36
0.74
±0.17
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
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 7-4 include the following:
• The pollutants with the highest annual average concentrations for GPCO are
formaldehyde (3.90 ± 0.35 |ig/m3) and acetaldehyde (2.80 ± 0.25 |ig/m3). These are
also the only pollutants with annual average concentrations greater than 1 |ig/m3,
although benzene is very close (0.99 ± 0.12 |ig/m3). The annual average
concentrations for these carbonyl compounds are considerably lower for 2014 than
they were for 2013.
• The third quarter average concentration for formaldehyde is significantly higher than
the other quarterly averages and has a relatively large confidence interval associated
with it. A review of the data shows that all but 10 formaldehyde concentrations
greater than 5 |ig/m3 were measured in samples collected at GPCO during the third
quarter of 2014, including all five samples collected in July, two from August, and
three from September. These measurements range from 5.08 |ig/m3 to 7.38 |ig/m3.
• The second quarter average concentration for acetaldehyde appears considerably
higher than the other quarterly averages. A review of the data shows that five of the
seven acetaldehyde concentrations greater than 4 |ig/m3 were measured in samples
collected at GPCO between April and June (with the other two in early July). The
second quarter also has the fewest concentrations on the lower end of the range; of
the 12 acetaldehyde concentrations less than 2 |ig/m3, only one was measured at
7-27
-------
GPCO during the second quarter while three to four were measured in each of the
other calendar quarters.
• Concentrations of benzene and 1,3-butadiene appear highest during the colder months
of the year, based on the quarterly averages shown in Table 7-4. A review of the data
shows that 22 of GPCO's 23 benzene concentrations greater than 1 |ig/m3 were
measured during the first or fourth quarters of 2014 (and the one exception was in late
September). Conversely, the 10 lowest benzene concentrations were measured
between April and August, during the second or third quarters of 2014. Similarly, all
22 of GPCO's 1,3-butadiene concentrations greater than 0.15 |ig/m3 were measured
during the first or fourth quarters of 2014 while 10 of the 11 lowest concentrations
were measured between April and August. Ethylbenzene concentrations also appear
to exhibit this tendency, with the second quarter average significantly lower than the
other quarterly average concentrations. The highest ethylbenzene concentration
measured during the second quarter of 2014 is lower than the average concentration
for the year and all but one of the seven ethylbenzene concentrations less than
0.2 |ig/m3 were measured at GPCO during the second quarter.
• Second quarter average concentrations for acenaphthene and naphthalene are not
provided in Table 7-4. Several PAH samples collected in April were invalidated as a
result of laboratory equipment issues. In addition, the site experienced sampler issues
in June. Thus, the criteria for calculating a quarterly average concentration was not
met.
• Concentrations of acenaphthene appear considerably higher during the warmer
months of the year, based on the quarterly averages shown in Table 7-4. A review of
the data shows that of 16 acenaphthene concentrations greater than 10 ng/m3, all but
two were measured between July and September. In addition, none of the 17
acenaphthene concentrations less than 4 ng/m3 were measured during the third quarter
while between three (second quarter) and nine (first quarter) were measured during
the other calendar quarters.
• GPCO is one of only four NMP sites with an annual average concentration of
naphthalene greater than 100 ng/m3 (100.03 ± 13.48 ng/m3). Concentrations of
naphthalene measured at GPCO range from 26.6 ng/m3 to 245 ng/m3 with all three of
GPCO's naphthalene concentrations greater than 200 ng/m3 measured in January
2014. Twenty-five naphthalene concentrations measured at GPCO are greater than
100 ng/m3; only DEMI (29) has more individual naphthalene measurements greater
than 100 ng/m3 than GPCO, although BXNY ties GPCO with 25.
Observations for the Garfield County sites from Table 7-4 include the following:
• Although acetaldehyde and formaldehyde are pollutants of interest for each Garfield
County site, annual average concentrations of these carbonyl compounds could not be
calculated for BRCO, PACO, or RFCO because the completeness criteria for this
method is less than 85 percent for each site, as discussed in Section 2.4. In addition,
BMCO does not have a second quarter average concentration for these pollutants
because there were too few valid samples collected during this calendar quarter. Thus,
7-28
-------
RICO is the only Garfield County site with four quarterly averages and an annual
average presented in Table 7-4 for formaldehyde and acetaldehyde.
• The annual average concentrations of acetaldehyde and formaldehyde for BMCO and
RICO are fairly similar to each other and significantly less than the annual averages
for GPCO. In fact, RICO and BMCO have the two lowest annual average
concentrations of acetaldehyde and second and third lowest annual average
concentrations of formaldehyde among NMP sites sampling carbonyl compounds.
• BMCO and BRCO are missing several quarterly average benzene concentrations and
do not have an annual average concentration for benzene presented. These are a result
of summertime sampling issues at BMCO and a series of invalid samples in May and
August at BRCO. Among the other three Garfield County sites, the annual average
benzene concentration ranges from 0.46 ± 0.09 |ig/m3 (RFCO) to 1.49 ± 0.14 |ig/m3
(PACO). The annual average benzene concentration for PACO is the highest among
all NMP sites sampling this pollutant, with RICO's annual average ranking third
(1.09 ± 0.14 |ig/m3), A review of the data shows that PACO has the highest number
of benzene concentrations greater than 1 |ig/m3 among all NMP sites sampling
benzene (49, or 86 percent of samples collected, with the next highest site at 33).
Conversely, RFCO is tied for lowest, with only one benzene concentration greater
than 1 |ig/m3. This site has the second lowest annual average concentration of
benzene among NMP sites. The quarterly average concentrations of benzene reflect a
similar pattern in the magnitude of the benzene measurements. Note that the first and
fourth quarter average concentrations of benzene for RICO are significantly higher
than the third quarter average concentration (and a quarterly average concentration
could not be calculated for the second quarter of 2014). A review of this site's data
shows that all of the benzene concentrations measured between March and August
2014 are less than 1 |ig/m3, ranging from 0.289 |ig/m3 to 0.944 |ig/m3. Conversely,
most of the concentrations measured during the rest of 2014 are greater than 1 |ig/m3,
and range from 0.72 |ig/m3 to 2.51 |ig/m3.
• 1,3-Butadiene is also a pollutant of interest for PACO, RFCO, and RICO. The annual
average concentration for RICO is three times greater than the annual averages
calculated for PACO and RFCO. Note that the second and third quarter average
concentrations for 1,3-butadiene for PACO and RICO are both 0, indicating that this
pollutant was not detected at these sites during these two calendar quarters.
1,3-Butadiene was detected in only seven samples collected at RFCO and in
14 samples collected at PACO, compared to 33 samples collected at RICO. Only one
non-detect of 1,3-butadiene was measured at RICO during the first and fourth
quarters of 2014 while few measured detections were measured during the second and
third quarters of 2014. In fact, there were no measured detections of 1,3-butadiene
measured at RICO between mid-May and early September and only six measured
during the second and third quarters. Concentrations of 1,3-butadiene measured at
these three sites were highest during the fourth quarter and is reflected in the fourth
quarter average concentrations.
7-29
-------
• The only other pollutant of interest shown in Table 7-4 is ethylbenzene for RICO.
Concentrations of ethylbenzene also appear highest during the fourth quarter of 2014
at RICO, although the difference is not significant.
Tables 4-9 through 4-12 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 4-9 through
4-12 a total of nine times, with GPCO appearing the most (6).
• PACO has the highest annual average concentration of benzene among all NMP sites
sampling this pollutant, as indicated above, with RICO (third highest) and GPCO
(eighth highest) also appearing in Table 4-9.
• GPCO and RICO have the fifth and ninth highest annual average concentrations of
1,3-butadiene, respectively, among NMP sites sampling this pollutant.
• GPCO also ranks third for its annual average concentration of ethylbenzene.
• GPCO's annual average concentration of acetaldehyde ranks second highest among
NMP sites sampling carbonyl compounds, as shown in Table 4-10. GPCO also ranks
fourth for its annual average concentration of formaldehyde.
• GPCO has the fourth highest annual concentration of naphthalene among NMP sites
sampling PAHs, as shown in Table 4-11.
7.4.2 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 7-4 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) and one for samples collected and analyzed with the SNMOC method
(the Garfield County sites), where annual averages could be calculated. Figures 7-14 through
7-24 overlay the 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 3.4.3.1, and are discussed below.
7-30
-------
Figure 7-14. Program vs. Site-Specific Average Acenaphthene Concentration
O
Program Max Concentration = 198 ng/m3
40 60
Concentration {ng/m3)
Program: 1st Quartile
2nd Quartile 3rd Quartile
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 7-14 presents the box plot for acenaphthene for GPCO and shows the following:
• The program4evel maximum acenaphthene concentration (198 ng/m3) is not shown
directly on the box plot in Figure 7-14 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 to 100 ng/m3.
• GPCO's annual average concentration of acenaphthene is greater than the program-
level average concentration as well as the program-level third quartile. This site has
the fourth highest annual average concentration of acenaphthene among NMP sites
sampling PAHs. Yet, the range of concentrations shown in Figure 7-14 appears
relatively small as the maximum concentration measured at GPCO is considerably
less than the maximum concentration measured across the program. A review of the
data shows that all 35 acenaphthene measurements greater than GPCO's maximum
acenaphthene concentration were measured at only three other NMP sites (NBIL,
ROCH, and DEMI). Concentrations of acenaphthene measured at GPCO range from
0.66 ng/m3 to 24.1 ng/m3, with a median concentration of 5.98 ng/m3, which is the
third-highest median concentration among sites sampling this pollutant.
7-31
-------
Figure 7-15. Program vs. Site-Specific Average Acetaldehyde Concentrations
n
O
0123456789 10
Concentration {[j.g/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 7-15 presents the box plots for acetaldehyde and shows the following:
• Figure 7-15 presents the box plots for GPCO, BMCO, and RICO, the three Colorado
sites for which annual average concentrations for acetaldehyde could be calculated.
• The box plots show that the range of acetaldehyde concentrations measured at GPCO
is considerably larger than the range of concentrations measured at the two Garfield
County sites shown. GPCO has the highest annual average acetaldehyde
concentration among the Colorado sites, where they could be calculated. The annual
average concentration for GPCO is more than five times greater than the annual
average acetaldehyde concentrations for BMCO and RICO, and is the second highest
annual average among NMP sites sampling carbonyl compounds. The minimum
acetaldehyde concentration measured at GPCO is greater than the annual average
concentrations for BMCO and RICO, an observation that was also made in the 2013
and 2012 NMP reports. The entire range of acetaldehyde concentrations measured at
BMCO and RICO is less than the program-level average and median concentrations.
Recall from the previous section that these two sites have the lowest annual average
acetaldehyde concentrations among NMP sites sampling carbonyl compounds.
7-32
-------
Figure 7-16. Program vs. Site-Specific Average Arsenic (PMio) Concentration
Program Max Concentration = 10.1 ng/m3
i i i i i
0 1 2 3 4 5 6
Concentration {ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 7-16 presents the box plot for arsenic for GPCO and shows the following:
• The program-level maximum arsenic concentration (10.1 ng/m3) is not shown directly
on the box plot because 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.
• GPCO's maximum arsenic concentration is similar to the program-level third
quartile, and is the lowest maximum concentration among NMP sites sampling this
pollutant. The annual average concentration of arsenic for GPCO is just greater than
the program-level first quartile. This site has the second-lowest annual average
concentration of arsenic among NMP sites sampling arsenic.
Figure 7-17a. Program vs. Site-Specific Average Benzene (Method TO-15) Concentration
1-
Program Max Concentration = 12.4 ng/m3
0 2 4 6 8 10
Concentration (ng/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
7-33
-------
Figure 7-17b. Program vs. Site-Specific Average Benzene (SNMOC) Concentrations
/-\
(J
RFCO
^ 1
0 0.5 1 1.5 2 2.5 3
Concentration (ng/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
Figures 7-17a and 7-17b present the box plots for benzene and show the following:
• Figure 7-17a presents the minimum, maximum, and annual average concentration of
benzene for GPCO compared to the benzene concentrations measured across the
program for NMP sites sampling VOCs with Method TO-15; Figure 7-17b 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 are not the same in the
figures.
• The program-level maximum benzene concentration (12.4 ng/m3) is not shown
directly on the box plot in Figure 7-17a because 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.
• Figure 7-17a shows that the annual average benzene concentration for GPCO is
higher than the program-level average concentration and the program-level third
quartile. The maximum benzene concentration measured at GPCO is considerably
less than the maximum benzene concentration measured across the program. Even
though the range of benzene concentrations for GPCO appears relatively small, this
site has the sixth highest annual average benzene concentration among NMP sites
sampling benzene with Method TO-15.
7-34
-------
• Figure 7-17b includes a box plot for three of the five Garfield County sites. The
maximum benzene concentration measured at PACO is the maximum benzene
concentration measured among the seven NMP sites sampling SNMOCs
(2.93 |ig/m3). Of the Garfield County sites, PACO has the highest annual average
concentration of benzene, followed by RICO and RFCO. The range of benzene
concentrations measured at RFCO is considerably smaller than the ranges shown for
the other Garfield County sites. This site's annual average benzene concentration is
equivalent to the program-level first quartile.
Figure 7-18a. Program vs. Site-Specific Average 1,3-Butadiene (Method TO-15)
Concentration
¦-H
Program Max Concentration = 5.90 |ig/m3
0.4 0.6
Concentration {[j.g/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 7-18b. Program vs. Site-Specific Average 1,3-Butadiene (SNMOC) Concentrations
PACO
O
RFCO
O
o
0 0.1 0.2 0.3 0.4 0.5
Concentration (ng/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
7-35
-------
Figures 7-18a and 7-18b present the box plots for 1,3-butadiene and show the following:
• Similar to the box plots for benzene, Figure 7-18a presents the minimum, maximum,
and annual average concentration of 1,3-butadiene for GPCO compared to the
1,3-butadiene concentrations measured across the program for NMP sites sampling
VOCs with Method TO-15; Figure 7-18b presents the minimum, maximum, and
annual average 1,3-butadiene concentrations for the Garfield County sites compared
to the 1,3-butadiene concentrations measured across the program for NMP sites
sampling SNMOCs. Note that the scales are not the same in the figures.
• The program-level maximum 1,3-butadiene concentration (5.90 ng/m3) is not shown
directly on the box plot in Figure 7-18a because 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.
• Figure 7-18a shows that the maximum 1,3-butadiene concentration measured at
GPCO is an order of magnitude less than the program-level maximum concentration.
GPCO's annual average concentration of 1,3-butadiene is greater than the program-
level average concentration and the program-level third quartile and is the fifth
highest annual average concentration among NMP sites sampling 1,3-butadiene with
Method TO-15. The minimum concentration of 1,3-butadiene measured at GPCO is
similar to the program-level first quartile.
• Figure 7-18b includes a box plot for 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, and thus, not visible in Figure 7-18b, indicating
that at least half 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. The maximum 1,3-butadiene concentration
measured at RICO (0.460 |ig/m3) is at least twice the maximum concentration
measured among the remaining Garfield County sites. Of the Garfield County sites
shown, RICO has the highest annual average concentration of 1,3-butadiene,
followed by PACO and RFCO. The annual average concentrations for PACO and
RFCO are less the program-level average concentration while RICO's annual average
is three times the program-level average concentration (among sites sampling
SNMOCs).
7-36
-------
Figure 7-19. Program vs. Site-Specific Average Carbon Tetrachloride Concentration
Program Max Concentration = 3.06 ng/m3
0
0.5
1 1.5
Concentration (ng/m3)
2
2.5
Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site: Site Average
o
Site Concentration Range
Figure 7-19 presents the box plot for carbon tetrachloride for GPCO and shows the
following:
• The program-level maximum carbon tetrachloride concentration (3.06 ng/m3) is not
shown directly on the box plot in Figure 7-19 because 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.
• The maximum carbon tetrachloride concentration measured at GPCO is considerably
less than the program-level maximum concentration. The annual average carbon
tetrachloride concentration for GPCO is just less than the program-level first quartile.
The annual average carbon tetrachloride concentration for GPCO is the lowest annual
average among NMP sites sampling this pollutant. However, the variability in the
annual averages among NMP sites is rather small, with less than 0.1 |ig/m3 separating
most sites' annual averages.
Figure 7-20. Program vs. Site-Specific Average 1,2-Dichloroethane Concentration
I
Program Max Concentration = 27.4 ng/m3
0
0.2
0.4 0.6
Concentration {[j.g/m3)
0.8
l
Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site: Site Average
o
Site Concentration Range
7-37
-------
Figure 7-20 presents the box plot for 1,2-dichloroethane for GPCO and shows the
following:
• The program-level maximum 1,2-dichloroethane concentration (27.4 |ig/m3) is not
shown directly on the box plot in Figure 7-20 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 are less than the
program-level average concentration of 0.31 |ig/m3 and GPCO's maximum
concentration is half the magnitude of the program-level average concentration. The
annual average concentration for GPCO falls between program-level first quartile and
median concentration. Note that the program-level average concentration is being
driven by the measurements at the upper end of the concentration range.
Figure 7-21a. Program vs. Site-Specific Average Ethylbenzene (Method TO-15) Concentration
i
{J 1
t 1 1 1 1 r
0 0.5 1 1.5 2 2.5 3 3.5
Concentration (ng/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 7-21b. Program vs. Site-Specific Average Ethylbenzene (SNMOC) Concentration
, ¦
O i
¦
i i i i i
0 0.2 0.4 0.6 0.8 1 1.2
Concentration (ng/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
Figures 7-2la and 7-2lb present the box plots for ethylbenzene and show the following:
• Ethylbenzene is a pollutant of interest for GPCO and RICO; thus, Figure 7-2la
presents the minimum, maximum, and annual average concentration of ethylbenzene
for GPCO compared to the ethylbenzene concentrations measured across the program
7-38
-------
for NMP sites sampling VOCs with Method TO-15 while Figure 7-21b presents the
minimum, maximum, and annual average 1,3-butadiene concentrations for RICO
compared to the ethylbenzene concentrations measured across the program for NMP
sites sampling SNMOCs. Note that the scales are not the same in the figures.
• Figure 7-2la shows that the maximum ethylbenzene concentration measured at
GPCO is less than half the magnitude of the program-level maximum concentration.
GPCO's annual average concentration of ethylbenzene is greater than the program-
level average concentration and the program-level third quartile. GPCO has the third
highest annual average concentration among NMP sites sampling ethylbenzene. The
minimum concentration of ethylbenzene measured at GPCO is greater than the
program-level first quartile.
Figure 7-2lb shows that the annual average ethylbenzene concentration for RICO is
greater than the program-level average concentration and the program-level third
quartile for sites sampling SNMOCs. The range of ethylbenzene concentrations
measured at RICO ranges from 0.125 |ig/m3 to 0.691 |ig/m3, as well as two non-
detects.
Figure 7-22. Program vs. Site-Specific Average Formaldehyde Concentrations
1
1
*
n
12 15
Concentration {[j.g/m3]
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 7-22 presents the box plots for formaldehyde and shows the following:
• Figure 7-22 presents the box plots for GPCO, BMCO, and RICO, the three Colorado
sites for which annual average concentrations could be calculated. The box plots for
7-39
-------
formaldehyde exhibit similar concentration patterns to those shown in the box plots
for acetaldehyde in Figure 7-15.
• The box plots show that the range of formaldehyde concentrations measured at GPCO
is considerably larger than the range of concentrations measured at the two Garfield
County sites shown. GPCO has the highest annual average formaldehyde
concentration among the Colorado sites, where they could be calculated. The annual
average concentration for GPCO is nearly four times greater than the annual average
formaldehyde concentrations for BMCO and RICO, and is the fourth highest annual
average among NMP sites sampling formaldehyde. The minimum formaldehyde
concentration measured at GPCO is greater than the annual average concentrations
calculated for BMCO and RICO, an observation that was also made in the 2013 and
2012 NMP reports. The entire range of formaldehyde concentrations measured at
BMCO and RICO is less than the program-level average and median concentrations
for 2014. Recall from the previous section that these two sites have some of the
lowest annual average formaldehyde concentrations among NMP sites sampling
carbonyl compounds.
Figure 7-23. Program vs. Site-Specific Average Hexachloro-l,3-butadiene Concentration
0.3 0.4
Concentration {[j.g/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 7-23 presents the box plot for hexachloro-1,3-butadiene for GPCO and shows the
following:
• The program-level first, second (median), and third quartiles are all zero for
hexachloro-1,3-butadiene and therefore not visible on the box plot. This is due to the
large number of non-detects of this pollutant across the program (77 percent). Fifty-
seven valid VOC samples were collected at GPCO and of these, hexachloro-1,3-
butadiene was detected in only 13 of them. Thus, many zeroes are substituted into the
annual average concentration of this pollutant.
• The maximum hexachloro-l,3-butadiene concentration measured at GPCO is about
one-sixth the magnitude of the maximum hexachloro-1,3-butadiene concentration
measured across the program. The annual average concentration for GPCO is similar
to the program-level average concentration of hexachloro-l,3-butadiene.
7-40
-------
Figure 7-24. Program vs. Site-Specific Average Naphthalene Concentration
¦
r\ i
1
i i i i i
0 100 200 300 400 500 600
Concentration {ng/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 7-24 presents the box plot for naphthalene for GPCO and shows the following:
• GPCO is one of four NMP sites sampling naphthalene with an annual average
concentration greater than 100 ng/m3. Recall from the previous section that GPCO
has the fourth highest annual average naphthalene concentration among NMP sites
sampling PAHs. The annual average naphthalene concentration for GPCO is greater
than the program-level average concentration and the program-level third quartile.
The minimum concentration of naphthalene measured at GPCO is just less than the
program-level first quartile.
7.4.3 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 3.4.3.2.
GPCO has sampled carbonyl compounds and VOCs under the NMP since 2004, PAHs since
2008, and metals since 2014; thus, Figures 7-25 through 7-34 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
sampling SNMOCs and carbonyl compounds under the NMP in 2008, thus, Figures 7-35 through
7-46 present the 1-year statistical metrics for each of the pollutants of interest for these three
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 BMCO began sampling SNMOCs
and carbonyl compounds under the NMP in 2010, sampling did not begin until September 2010,
7-41
-------
which is less than the 6-month requirement; thus, the trends analysis was not conducted for this
site. RFCO began sampling in 2012, which is less than 5 years of sampling, excluding this site
from trends analysis as well.
Figure 7-25. Yearly Statistical Metrics for Acenaphthene Concentrations Measured at GPCO
X r
m
d • 5
3 t
- - i
5th Percentile
— Minimum
- Med "en
— 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 7-25 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.
• Five of the six highest concentrations of acenaphthene were measured at GPCO
during the spring of 2012 and ranged from 53.7 ng/m3 to 182 ng/m3. The only other
measurement greater than 50 ng/m3 collected at GPCO was measured in November
2008 (62.2 ng/m3).
• 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; for
example, nine of the 16 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 when none were measured). Even if the two highest
7-42
-------
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 7-25 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
1-year average, median, and maximum concentrations are at a minimum for 2014.
Figure 7-26. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at GPCO
Maximum Concentration
for 2004 is 93.0 |J.g/m3
-S-
rh
-2—
—
hr
iA.
UJ
r t
2008 2009 2010 2011 2012 2013 2014
Year
O 5th Percentile
O 95th Percentile
Observations from Figure 7-26 for acetaldehyde concentrations measured at GPCO
include the following:
• The maximum acetaldehyde concentration was measured at GPCO in 2004. The two
highest acetaldehyde concentrations (93.0 |ig/m3 and 54.9 |ig/m3) were both
measured in 2004. The maximum concentrations measured in subsequent years were
significantly lower as no additional concentrations greater than 20 |ig/m3 were
measured. 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, seven 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 1-year average and
median concentrations are both at a minimum for 2010, representing a statistically
7-43
-------
significant decrease from 2009. The 1-year average concentration increases steadily
between 2010 and 2013, with the 1-year average at a maximum since 2004. The
median concentration exhibits a similar pattern.
• Concentrations measured in 2014 return to levels near those shown for 2012,
although both the minimum and maximum concentrations for 2014 are among the
lowest measured at GPCO for each year shown.
Figure 7-27. Yearly Statistical Metrics for Benzene Concentrations Measured at GPCO
•o
o-
2004 2005 2006 2007 2008
2009 2010 2011 2012 2013 2014
Year
O 5th Percentile
— Minimum
— Maximum O 95th Percentile •••~¦*• Avera
Observations from Figure 7-27 for benzene concentrations measured at GPCO include
the following:
• The maximum benzene concentration (10.6 |ig/m3) was measured on June 8, 2011.
Only 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 and median concentrations. After a period of increasing for 2008
and 2009, a significant decrease is shown for 2010. This decreasing trend continues
through 2014, when several of the statistical metrics are at a minimum. This is also
true for the median concentration, except that the median increases slightly for 2014.
The range of benzene concentrations measured is at a minimum for 2014, with less
than 2 |ig/m3 separating the minimum and maximum concentrations, and the 1-year
average, 95th percentile, and maximum concentrations are at a minimum for the years
7-44
-------
shown. Yet, the median concentration exhibits a slight increase from 2013 to 2014.
The number of benzene concentrations falling between 1 |ig/m3 and 2 |ig/m3 doubled
from 2013 (10) to 2014 (20), accounting for more than one-third of the concentrations
measured in 2014.
Figure 7-28. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at GPCO
1.2
1.0
0.8
mE
S
c
o
4= 0.6
0.4
0.2
0.0
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year
o 5th Percentile - Minimum - Median - Maximum o 95th Percentile ...4.... Average
Observations from Figure 7-28 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
0.07 |ig/m3 over the years of sampling, ranging from 0.132 |ig/m3 (2010) 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, but there were also no
non-detects reported for 2012, while there were seven reported for 2011.
7-45
-------
• The largest year-to-year change in the median concentration is the decrease shown
from 2012 to 2013. Although non-detects were not measured in either year, the
number of 1,3-butadiene measurements less than 0.1 |ig/m3 nearly doubled from 2012
(17) to 2013 (31), thus, representing half of the measurements for 2013.
• Each of the statistical parameters exhibits a slight increase from 2013 to 2014.
Figure 7-29. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured
at GPCO
o
r—1—I
o
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year
O 5th Percentile
- Minimum
— Median
- Maximum o 95th Percentile
Observations from Figure 7-29 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
the difference between the minimum and 5th percentile for 2012 compared to other
years (where the 1-year average and/or median concentrations for a given year are
less than the 5th percentile for 2012).
7-46
-------
• 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.
• 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 steady decreasing trend is shown through 2010, with little change in the
measurements 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. While the central tendency statistics changed little for 2014,
the minimum concentration appears to have increased considerably. The lowest
concentration measured actually varies little, the difference is a result of a non-detect.
Figure 7-30. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations
Measured at GPCO
__
V
0
o
0
-
in
O
o
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year
O 5th Percentile - Minimum - Median - Maximum o 95th Percentile • Average
7-47
-------
Observations from Figure 7-30 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.
The percentage of measured detections decreased slightly for 2013 (74 percent) but
approached 90 percent again for 2014.
• As the number of measured detections increases, so do each of the corresponding
statistical metrics shown in Figure 7-30. The percentage of measured detections
increased by 63 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 2012,
2013, and 2014. This is because there were still non-detects (or zeros) factoring into
the 1-year average concentration for each year, which tend to pull the average down.
Excluding non-detects, the minimum concentration for 2012 would be 0.04 |ig/m3,
with a difference between the minimum and maximum concentrations measured for
2012 of less than 0.1 |ig/m3. This is also true for 2013 and 2014.
Figure 7-31. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at GPCO
"e
1
c
o
c
c
u
o
•o.
i
I
I
I
Lr
~f-1
JL.
-r
JZ.
-s-
Q
JL.
2004
2005
2006
2007
2008
2009
Yea
2010
2011
2012
2013
2014
1
o
5th Percentile
-
Minimun
-
Median
- Maxim urr
O
95th Percentile
¦ Average
1
7-48
-------
Observations from Figure 7-31 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
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
of 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 exhibits a
slight increasing trend beginning with 2009 and continuing through 2012.
• All of the statistical parameters exhibit a decrease from 2012 to 2013. The maximum
ethylbenzene concentration measured in 2013 is the lowest maximum concentration
for any given year of sampling shown in Figure 7-31. Although the range of
ethylbenzene concentrations measured increased just slightly for 2014, the 1-year
average concentration is at a minimum across the years of sampling (as is the 95th
percentile).
7-49
-------
Figure 7-32. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at GPCO
Maximum Concentration
for 2004 is 40.5 |Jg/m3
i.
|-S-|
LjJ
2004 2005
-S-
r—5—i
~
T~
LgJ
2010
2011
2012
T
2013 2014
O 5th Percentile
— Minimum
Maximum O 95th Percentile
Observations from Figure 7-32 for formaldehyde concentrations measured at GPCO
include the following:
• The trends graph for formaldehyde resembles the trends graph for acetaldehyde in
that 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 days in 2004
as the two highest acetaldehyde concentrations. The next eight highest formaldehyde
concentrations were measured at GPCO in 2013 and range from 13.9 |ig/m3 to
21.9 |ig/m3.
• 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 metrics 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
average concentration for 2013 is greater than the maximum concentrations measured
7-50
-------
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.
• 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.
Figure 7-33. Yearly Statistical Metrics for Hexachloro-l,3-butadiene Concentrations
Measured at GPCO
o
2004 2005
-
2006 2007 200S 2009
O
O 5th Percentile
- Minimum
— Median
- Maximum
O 95th Percentile
Observations from Figure 7-33 for hexachloro-1,3-butadiene concentrations measured at
GPCO include the following:
• The number of measured detections of hexachloro-1,3-butadiene for each year is very
low, from zero measured detections in 2004, 2008, and 2009 to 13 (or 23 percent) for
2014. This explains why the minimum, 5th percentile, and median concentrations
(and in some cases, the 1-year averages) are all zero for each year of sampling. The
detection rate has increased slightly over the last few years. Additional years of
sampling are needed to determine if this trend continues.
• The maximum hexachloro-l,3-butadiene concentration was measured in 2005
(0.26 |ig/m3). Although nine additional measurements greater than 0.20 |ig/m3 have
been measured at GPCO, all but one of these were measured between 2005 and 2007.
7-51
-------
• The large number of non-detects, and thus zeroes substituted into the calculations,
combined with few measured detections results in relatively low 1-year average
concentrations with relatively large confidence intervals.
Figure 7-34. Yearly Statistical Metrics for Naphthalene Concentrations Measured at GPCO
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 7-34 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 seven years of sampling and concentrations greater than 250 ng/m3 have been
measured in all years of sampling except 2014.
• The trends graph for naphthalene resembles the trends graphs for acenaphthene
shown in Figure 7-25. 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. This was followed by a
decrease in 2013 to levels less than those shown for 2011. The smallest range of
naphthalene concentrations was measured in 2014, with all of the statistical
parameters exhibiting decreases from 2013 to 2014. The 1-year average
concentration, the 95th percentile, and the maximum concentration for 2014 are at a
minimum across the years of sampling.
7-52
-------
Figure 7-35. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
BRCO
2010 1
20111
Year
o 5th Percentile
— Minimum
— 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 7-35 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. This is 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 27 acetaldehyde concentrations
greater than 1 |ig/m3 have been measured at BRCO since the onset of sampling.
• Concentrations of acetaldehyde measured at BRCO have a decreasing trend across
the years of sampling, and nearly all of the statistical parameters at a minimum for
2014.
7-53
-------
Figure 7-36. Yearly Statistical Metrics for Benzene Concentrations Measured at BRCO
T
HS-£ 0 E±a
—!
— •
2011
Year
O 5th Percentile
Maximum O 95th Percentile
1 A 1-year average is not presented due to low method completeness in 2014.
Observations from Figure 7-36 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.
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 decreased, from
1.05 |ig/m3 in 2008 to 0.65 |ig/m3 in 2012.
All of the statistical metrics exhibit an increase from 2012 to 2013, returning to
concentration levels similar to 2010. This is followed by a return to 2012 levels for
2014, based on the available statistical metrics.
7-54
-------
Figure 7-37. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
BRCO
12
nE
1
° g
c
c
u
2
T
I
1-L-i
<
—£— —
9 y S : e
2008 2009 2010 1 2011 1 2012 2013 20142
Year
o 5th Percentile — Minimum - Median — Maximum o 95th Percentile A--- Avercge
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 7-37 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 three times higher than the next highest concentration
measured at this site (3.11 |ig/m3, measured on August 31, 2012). Only three
additional formaldehyde concentrations greater than 2.0 |ig/m3 have been measured at
BRCO.
• 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 do not
include measurements for an entire year.
• Several statistical parameters exhibit a decreasing trend between 2012 and 2014,
including the maximum concentration, the 95th percentile, the 5th percentile, and the
minimum concentration.
7-55
-------
Figure 7-38. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
PACO
o
~
20111
Year
o 5th Percentile
— Minimum
— Maximum
o 95th Percentile
• Average
1 A 1-year average is not presented due to low method completeness in 2011.
2 A 1-year average is not presented due to low method completeness in 2014.
Observations from Figure 7-38 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. Note that carbonyl compounds are sampled on a l-in-12
sampling schedule at PACO.
• 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
across the years of sampling (although two of the seven years do not follow this
pattern, and are discussed in the bullets that follow). Several of the statistical
parameters are at a minimum for 2014, including the median, 95th percentile, and
maximum concentrations. Those not a minimum for 2014 are at a minimum for either
2012 or 2013.
• For 2011, fewer valid samples were collected but those greater than 1 |ig/m3 make up
a higher percentage of the measurements, resulting in a higher median concentration.
7-56
-------
In addition, there minimum concentration measured in 2011 was higher the others
years of sampling, with fewest concentrations less than 0.5 |ig/m3 since 2008.
• For 2013, both the 1-year average and median concentrations exhibit an increase. 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, indicating an
increase in the variability of the measurements.
Figure 7-39. Yearly Statistical Metrics for Benzene Concentrations Measured at PACO
1 A 1-year average is not presented due to low method completeness in 2012.
Observations from Figure 7-39 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 sampler issues.
• 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 but was less (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.
7-57
-------
• 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. The number of benzene
concentrations greater than 3 |ig/m3 increased from six in 2008 to 15 in 2009,
accounting for more than a quarter of the measurements in 2009.
• Concentrations of benzene exhibit a significant decreasing trend between 2009 and
2010, when the maximum and 95th percentile decreased by nearly half. This
decreasing trend continued into 2011 and 2012. Although a 1-year average
concentration could not be calculated for 2012, the maximum, 95th percentile, and
median concentrations are at a minimum for 2012. Benzene concentrations greater
than 3 |ig/m3 were not measured in 2012.
• All of the statistical parameters shown increased considerably from 2012 to 2013.
The range within which the majority of the measurements fall, indicated by the 5 th
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 for 2014,
although not quite returning to levels shown for 2012.
Figure 7-40. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at PACO
o
Maximum
Concentration for
2009 is 3.15 ug/m3
2011
Year
X
>0
2014
5th Percentile - Minimum
95th Percentile
1 A 1-year average is not presented due to low method completeness in 2012.
7-58
-------
Observations from Figure 7-40 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 (0.19 |ig/m3). Excluding the maximum concentration for
2009 would result is a 1-year average concentration of only 0.028 |ig/m3 (rather than
0.88 |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 highest concentration for this year was also measured in
December but was considerably less (0.16 |ig/m3). The 95th percentile for 2010 is
greater than the maximum concentration measured for all other years except 2009 and
more than tripled from 2009 to 2010. Even though half of the measurements in 2010
were non-detects, the December measurements for 2010 are driving the top-end
statistical parameters upward.
• 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 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. The maximum and 95th
percentile decreased considerably for 2013 and are at a minimum for the period of
sampling, as is the 1-year average concentration. Although slightly higher
concentrations were measured in 2014, non-detects still account for 75 percent of the
concentrations measured.
7-59
-------
Figure 7-41. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at PACO
"O—.
o
20111
Year
5th Percentile
- Minimum
95th Percentile
1 A 1-year average is not presented due to low method completeness in 2011.
2 A 1-year average is not presented due to low method completeness in 2014.
Observations from Figure 7-41 for formaldehyde concentrations measured at PACO
include the following:
• Only four formaldehyde concentrations greater than 3 |ig/m3 have been measured at
PACO (one is 2008, two in 2009, and one in 2010).
• 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 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. Nine formaldehyde measurements
less than 1 |ig/m3 were measured in 2012, accounting for one-third of concentrations
7-60
-------
measured in 2012. However, lower concentrations (those less than 1 |ig/m3)
continued to be measured in subsequent years, with seven measured in 2013 and 11
measured in 2014, the most of any year shown. All 14 formaldehyde concentrations
less than 0.5 |ig/m3 measured at PACO were measured in 2012 (5), 2013 (4), or 2014
(5).
Figure 7-42. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at RICO
2010 1
20111
Year
20131
o 5th Percentile
— Minimum
— Maximum
o 95th Percentile
¦ Average
1 A 1-year average is not presented due to low method completeness in 2010, 2011, and 2013.
Observations from Figure 7-42 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. 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 an overall decreasing trend, based on the decreases shown for
nearly all of the other statistical parameters. Acetaldehyde concentrations greater than
2.5 |ig/m3 were not measured after 2009 and concentrations greater than 2 |ig/m3
were not measured in 2014.
7-61
-------
• The minimum concentration and 5th percentile decreased considerably from 2011 to
2012, with additional decreases for 2013. Concentrations measured 2012 and later
account for 22 of the 23 concentrations less than 0.5 |ig/m3 measured at RICO (four
were measured in 2012, six in 2013, and 12 in 2014).
Figure 7-43. Yearly Statistical Metrics for Benzene Concentrations Measured at RICO
Observations from Figure 7-43 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 four highest benzene concentrations measured at RICO were all measured in
January 2009, with the next two highest also measured in 2009, but in different
months.
• All of the statistical metrics 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 explains 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.
7-62
-------
• 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 to 2013. 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 levels shown for 2012. The statistical metrics 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.
Figure 7-44. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at RICO
o
2011
Year
O 5th Percentile
o 95th Percentile
Observations from Figure 7-44 for 1,3-butadiene concentrations measured at RICO
include the following:
• The five highest 1,3-butadiene concentrations were all 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 zero for each year of sampling;
this indicates that at least 5 percent of the measurements were non-detects for each
7-63
-------
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.
Although the range of concentrations measured varies between 2010 and 2012, the
1-year average concentration decreases only slightly while the median concentration
increases steadily.
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
half. Overall, the 1,3-butadiene concentrations measured were lower in 2013. The
number of concentrations greater than 0.25 |ig/m3 decreased from 17 in 2012 to five
in 2013; further, the number of concentrations less than 0.1 |ig/m3 (including non-
detects) increased from 15 in 2012 to 32 in 2013, accounting for more than half of the
concentrations measured 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 in 2013 to 21 in 2014.
7-64
-------
Figure 7-45. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at RICO
j Maximum
! Concentration for
! 2009 is 25.7 ng/m3
T
Pn
, I
i i
£
—2—'
L-j-J
2
lt_
2008 2009 2010 2011 2012 2013 2014
Year
O 5th Percentile — Minimum — Median — Maximum O 95th Percentile Average
Observations from Figure 7-45 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 measured at RICO 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 metrics shown in Figure 7-45 increased from 2012 to 2013,
with several of them returning to levels similar to those calculated for 2011.
• The only two non-detects of ethylbenzene measured at RICO were both measured in
2014.
7-65
-------
Figure 7-46. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at RICO
Maximum
1 A 1-year average is not presented due to low method completeness in 2010, 2011, and 2013.
Observations from Figure 7-46 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 on August 26, 2013 (4.38 |ig/m3). Only three additional formaldehyde
concentrations measured at RICO are greater than 3 |ig/m3 (one each in 2008, 2010,
and 2011).
• Few 1-year average concentrations could be calculated for RICO; however, the
measurements appear to have an overall decreasing trend after 2010, despite a few
higher concentrations measured, based on the decreases shown for several of the
other statistical parameters.
• The minimum and 5th percentile decreased considerably from 2011 to 2012 and
continued into 2013 and 2014, similar to acetaldehyde. Twenty-four of the 25
measurements less than 0.75 |ig/m3 collected at RICO were measured between 2012
and 2014, with 2014 having the most (15).
7-66
-------
7.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at each Colorado monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4
for definitions and explanations regarding the various toxicity factors, time frames, and
calculations associated with these risk-based screenings.
7.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Colorado 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 3.4.3.3 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 7-5, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
7-67
-------
Table 7-5. Risk Approximations for the Colorado Monitoring Sites
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Grand Junction, Colorado - GPCO
Acetaldehyde
0.0000022
0.009
58/58
2.80
±0.25
6.16
0.31
Benzene
0.0000078
0.03
57/57
0.99
±0.12
7.72
0.03
1.3 -Butadiene
0.00003
0.002
57/57
0.17
±0.03
4.98
0.08
Carbon Tetrachloride
0.000006
0.1
57/57
0.58
±0.03
3.49
0.01
1,2 -Dichloroethane
0.000026
2.4
50/57
0.07
±0.01
1.92
<0.01
Ethylbenzene
0.0000025
1
57/57
0.45
±0.06
1.14
<0.01
Formaldehyde
0.000013
0.0098
58/58
3.90
±0.35
50.68
0.40
Hexachloro -1,3 -butadiene
0.000022
0.09
13/57
0.02
±0.01
0.42
<0.01
Acenaphthene3
0.000088
60/60
7.17
± 1.20
0.63
Arsenic (PMi0)a
0.0043
0.000015
50/59
0.28
±0.06
1.19
0.02
Naphthalene1
0.000034
0.003
60/60
100.03
± 13.48
3.40
0.03
Battlement Mesa, Colorado - BMCO
Acetaldehyde
0.0000022
0.009
27/27
0.42
±0.11
0.93
0.05
Benzene
0.0000078
0.03
51/51
NA
NA
NA
Formaldehyde
0.000013
0.0098
27/27
0.77
±0.18
9.95
0.08
Silt, Colorado - BRCO
Acetaldehyde
0.0000022
0.009
25/25
NA
NA
NA
Benzene
0.0000078
0.03
49/50
NA
NA
NA
Formaldehyde
0.000013
0.0098
25/25
NA
NA
NA
— = A Cancer URE or Noncancer RfC is not available.
NA = Not available due to the criteria for calculating an annual average.
a Average concentrations provided for the pollutants below the blue line for GPCO are presented in ng/m3 for ease of
viewing.
7-68
-------
Table 7-5. Risk Approximations for the Colorado Monitoring Sites (Continued)
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Parachute, Colorado - PACO
Acetaldehyde
0.0000022
0.009
25/25
NA
NA
NA
Benzene
0.0000078
0.03
57/57
1.49
±0.14
11.65
0.05
1.3 -Butadiene
0.00003
0.002
14/57
0.03
±0.01
0.88
0.01
Formaldehyde
0.000013
0.0098
25/25
NA
NA
NA
Carbondale, Colorado - RFCO
Acetaldehyde
0.0000022
0.009
26/26
NA
NA
NA
Benzene
0.0000078
0.03
27/28
0.46
±0.09
3.62
0.02
1,3-Butadiene
0.00003
0.002
7/28
0.03
±0.02
0.76
0.01
Formaldehyde
0.000013
0.0098
26/26
NA
NA
NA
Rifle, Colorado - RICO
Acetaldehyde
0.0000022
0.009
27/27
0.52
±0.12
1.13
0.06
Benzene
0.0000078
0.03
54/54
1.09
±0.14
8.50
0.04
1.3 -Butadiene
0.00003
0.002
33/54
0.10
±0.03
3.08
0.05
Ethylbenzene
0.0000025
1
52/54
0.32
±0.04
0.79
<0.01
Formaldehyde
0.000013
0.0098
27/27
0.74
±0.17
9.66
0.08
— = A Cancer URE or Noncancer RfC is not available.
NA = Not available due to the criteria for calculating an annual average.
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 7-5 include the following:
• Formaldehyde, acetaldehyde, and benzene have the highest annual average
concentrations among GPCO's pollutants of interest.
• Formaldehyde has the highest cancer risk approximation for this site (50.68 in-a-
million), followed by benzene (7.72 in-a-million), acetaldehyde (6.16 in-a-million),
and 1,3-butadiene (4.98 in-a-million). GPCO's cancer risk approximation for
formaldehyde is the fifth highest cancer risk approximation calculated across the
program for 2014 and the fourth highest cancer risk approximation calculated for
formaldehyde.
7-69
-------
• None of the pollutants of interest for GPCO have a noncancer hazard approximation
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 (0.40 and 0.31, respectively) among the pollutants
of interest for GPCO. The noncancer hazard approximation for formaldehyde for
GPCO is the fifth highest noncancer hazard approximation calculated across the
program for 2014 and the fourth highest noncancer hazard approximation calculated
for formaldehyde.
Observations for the Garfield County sites from Table 7-5 include the following:
• Benzene, acetaldehyde, and formaldehyde were identified as pollutants of interest for
each Garfield County site.
• Annual average benzene concentrations could be calculated for three of these sites
(PACO, RFCO, and RICO). Among these sites, the annual average concentrations of
benzene range from 0.46 ± 0.09 |ig/m3 (RFCO) to 1.49 ± 0.14 |ig/m3 (PACO). The
cancer risk approximations for benzene for these sites range from 3.62 in-a-million
(RFCO) to 11.65 in-a-million (PACO). The noncancer hazard approximations
calculated for benzene for the Garfield County sites with available annual average
concentrations of benzene are considerably less than 1.0 (all are 0.05 or less). This
indicates that no adverse noncancer health effects are expected from this individual
pollutant.
• Annual average formaldehyde concentrations could only be calculated for BMCO and
RICO; the annual averages for these two sites are similar to each other (0.74 ±0.17
|ig/m3 for RICO and 0.77 ± 0.18 |ig/m3 for BMCO). The cancer risk approximations
for these sites are also similar to each other (9.66 in-a-million for RICO and 9.95 in-
a-million for BMCO) as are the noncancer hazard approximations (both are 0.08).
• Similarly, annual average concentrations of acetaldehyde could only be calculated for
BMCO and RICO. The annual average acetaldehyde concentrations calculated for
BMCO and RICO are 0.42 ± 0.11 |ig/m3 and 0.52 ± 0.12 |ig/m3, respectively. The
cancer risk approximations for these sites are both around 1 in-a-million (1.13 in-a-
million for RICO and 0.93 in-a-million for BMCO). The noncancer hazard
approximation for BMCO is 0.05 and for RICO is 0.06, both considerably less than
the level of concern (1.0).
• 1,3-Butadiene was identified as a pollutant of interest for PACO, RFCO, and RICO;
the annual average concentrations of 1,3-butadiene for these sites range from
0.03 ± 0.01 |ig/m3 for PACO to 0.10 ± 0.03 |ig/m3 for RICO. The cancer risk
approximations for these sites for 1,3-butadiene range from 0.76 in-a-million (RFCO)
to 3.08 in-a-million (RICO). The noncancer hazard approximations calculated for
1,3-butadiene for these Garfield County sites are 0.05 or less.
7-70
-------
7.5.2 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 7-6 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
that have cancer toxicity factors. Table 7-6 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 7-6 provides the pollutants of interest with the highest cancer risk approximations (in-a-
million) for each site, as presented in Table 7-5. The emissions, toxicity-weighted emissions, and
cancer risk approximations are shown in descending order in Table 7-6. Table 7-7 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 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 7.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
7-71
-------
Table 7-6. 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
Top 10 Cancer Toxicity-Weighted
Emissions
Top 10 Cancer Risk Approximations
Based on Annual Average Concentrations
(County-Level)
(County-Level)
(Site-Specific)
Cancer
Cancer Risk
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(in-a-million)
Grand Junction, Colorado (Mesa County) - GPCO
Benzene
166.34
Formaldehyde
1.72E-03
Formaldehyde
50.68
Formaldehyde
131.94
Benzene
1.30E-03
Benzene
7.72
Ethylbenzene
55.92
1,3-Butadiene
4.48E-04
Acetaldehyde
6.16
Acetaldehyde
49.20
Naphthalene
2.34E-04
1,3-Butadiene
4.98
1.3 -Butadiene
14.93
POM, Group 2b
1.55E-04
Carbon Tetrachloride
3.49
Naphthalene
6.89
Ethylbenzene
1.40E-04
Naphthalene
3.40
Dichloromethane
5.44
Acetaldehyde
1.08E-04
1,2-Dichloroethane
1.92
T etrachloroethylene
1.86
POM, Group 2d
1.00E-04
Arsenic (PMio)
1.19
POM, Group 2b
1.76
POM, Group 5a
6.90E-05
Ethylbenzene
1.14
POM, Group 2d
1.14
Arsenic, PM
3.36E-05
Acenaphthene
0.63
Battlement Mesa, Colorado (Garfield County) - BMCO
Benzene
652.88
Formaldehyde
7.96E-03
Formaldehyde
9.95
Formaldehyde
612.56
Benzene
5.09E-03
Acetaldehyde
0.93
Acetaldehyde
112.59
1,3-Butadiene
3.78E-04
Ethylbenzene
67.74
Acetaldehyde
2.48E-04
1,3-Butadiene
12.62
Ethylbenzene
1.69E-04
Naphthalene
4.78
Naphthalene
1.62E-04
T etrachloroethylene
1.01
POM, Group 2b
7.72E-05
POM, Group 2b
0.88
POM, Group 2d
5.42E-05
POM, Group 2d
0.62
POM, Group 5a
3.89E-05
Dichloromethane
0.25
Arsenic, PM
3.28E-05
-------
Table 7-6. 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)
Cancer
Cancer Risk
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(in-a-million)
Silt, Colorado (Garfield County) - BRCO
Benzene
652.88
Formaldehyde
7.96E-03
Formaldehyde
612.56
Benzene
5.09E-03
Acetaldehyde
112.59
1,3-Butadiene
3.78E-04
Ethylbenzene
67.74
Acetaldehyde
2.48E-04
1.3 -Butadiene
12.62
Ethylbenzene
1.69E-04
Naphthalene
4.78
Naphthalene
1.62E-04
T etrachloroethylene
1.01
POM, Group 2b
7.72E-05
POM, Group 2b
0.88
POM, Group 2d
5.42E-05
POM, Group 2d
0.62
POM, Group 5a
3.89E-05
Dichloromethane
0.25
Arsenic, PM
3.28E-05
Parachute, Colorado (Garfield County) - PACO
Benzene
652.88
Formaldehyde
7.96E-03
Benzene
11.65
Formaldehyde
612.56
Benzene
5.09E-03
1,3-Butadiene
0.88
Acetaldehyde
112.59
1,3-Butadiene
3.78E-04
Ethylbenzene
67.74
Acetaldehyde
2.48E-04
1,3-Butadiene
12.62
Ethylbenzene
1.69E-04
Naphthalene
4.78
Naphthalene
1.62E-04
T etrachloroethylene
1.01
POM, Group 2b
7.72E-05
POM, Group 2b
0.88
POM, Group 2d
5.42E-05
POM, Group 2d
0.62
POM, Group 5a
3.89E-05
Dichloromethane
0.25
Arsenic, PM
3.28E-05
-------
Table 7-6. 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)
Cancer
Cancer Risk
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(in-a-million)
Carbondale, Colorado (Garfield
County) - RFCO
Benzene
652.88
Formaldehyde
7.96E-03
Benzene
3.62
Formaldehyde
612.56
Benzene
5.09E-03
1,3-Butadiene
0.76
Acetaldehyde
112.59
1,3-Butadiene
3.78E-04
Ethylbenzene
67.74
Acetaldehyde
2.48E-04
1.3 -Butadiene
12.62
Ethylbenzene
1.69E-04
Naphthalene
4.78
Naphthalene
1.62E-04
T etrachloroethylene
1.01
POM, Group 2b
7.72E-05
POM, Group 2b
0.88
POM, Group 2d
5.42E-05
POM, Group 2d
0.62
POM, Group 5a
3.89E-05
Dichloromethane
0.25
Arsenic, PM
3.28E-05
Rifle, Colorado (Garfield County) - RICO
Benzene
652.88
Formaldehyde
7.96E-03
Formaldehyde
9.66
Formaldehyde
612.56
Benzene
5.09E-03
Benzene
8.50
Acetaldehyde
112.59
1,3-Butadiene
3.78E-04
1,3-Butadiene
3.08
Ethylbenzene
67.74
Acetaldehyde
2.48E-04
Acetaldehyde
1.13
1,3-Butadiene
12.62
Ethylbenzene
1.69E-04
Ethylbenzene
0.79
Naphthalene
4.78
Naphthalene
1.62E-04
T etrachloroethylene
1.01
POM, Group 2b
7.72E-05
POM, Group 2b
0.88
POM, Group 2d
5.42E-05
POM, Group 2d
0.62
POM, Group 5a
3.89E-05
Dichloromethane
0.25
Arsenic, PM
3.28E-05
-------
Table 7-7. 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
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)
Noncancer
Noncancer
Hazard
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(HQ)
Grand Junction, Colorado (Mesa County) - GPCO
Toluene
381.86
Acrolein
550,555.59
Formaldehyde
0.40
Xylenes
274.58
Formaldehyde
13,463.29
Acetaldehyde
0.31
Benzene
166.34
1,3-Butadiene
7,464.46
1,3-Butadiene
0.08
Formaldehyde
131.94
Benzene
5,544.61
Naphthalene
0.03
Hexane
120.83
Acetaldehyde
5,466.88
Benzene
0.03
Methanol
102.01
Xylenes
2,745.81
Arsenic
0.02
Ethylbenzene
55.92
Naphthalene
2,298.28
Carbon Tetrachloride
0.01
Acetaldehyde
49.20
Antimony, PM
1,050.63
Ethylbenzene
<0.01
Ethylene glycol
29.13
Lead, PM
767.25
Hexachloro-1,3 -butadiene
<0.01
1.3 -Butadiene
14.93
Arsenic, PM
521.58
1,2-Dichloroethane
<0.01
Battlement Mesa, Colorado (Garfield County) - BMCO
Toluene
1,190.11
Acrolein
3,464,518.24
Formaldehyde
0.08
Xylenes
730.99
Formaldehyde
62,505.94
Acetaldehyde
0.05
Benzene
652.88
Benzene
21,762.81
Methanol
623.52
Acetaldehyde
12,509.99
Formaldehyde
612.56
Xylenes
7,309.95
Hexane
169.35
1,3-Butadiene
6,308.09
Acetaldehyde
112.59
Naphthalene
1,592.72
Acrolein
69.29
Propionaldehyde
567.82
Ethylbenzene
67.74
Cadmium, PM
526.47
1,3-Butadiene
12.62
Arsenic, PM
508.98
-------
Table 7-7. 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)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Silt, Colorado (Garfield County) - BRCO
Toluene
1,190.11
Acrolein
3,464,518.24
Xylenes
730.99
Formaldehyde
62,505.94
Benzene
652.88
Benzene
21,762.81
Methanol
623.52
Acetaldehyde
12,509.99
Formaldehyde
612.56
Xylenes
7,309.95
Hexane
169.35
1,3-Butadiene
6,308.09
Acetaldehyde
112.59
Naphthalene
1,592.72
Acrolein
69.29
Propionaldehyde
567.82
Ethylbenzene
67.74
Cadmium, PM
526.47
1.3 -Butadiene
12.62
Arsenic, PM
508.98
Parachute, Colorado (Garfield County) - PACO
Toluene
1,190.11
Acrolein
3,464,518.24
Benzene
0.05
Xylenes
730.99
Formaldehyde
62,505.94
1,3-Butadiene
0.01
Benzene
652.88
Benzene
21,762.81
Methanol
623.52
Acetaldehyde
12,509.99
Formaldehyde
612.56
Xylenes
7,309.95
Hexane
169.35
1,3-Butadiene
6,308.09
Acetaldehyde
112.59
Naphthalene
1,592.72
Acrolein
69.29
Propionaldehyde
567.82
Ethylbenzene
67.74
Cadmium, PM
526.47
1,3-Butadiene
12.62
Arsenic, PM
508.98
-------
Table 7-7. 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
Top 10 Noncancer Toxicity-Weighted
Emissions
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(County-Level)
(County-Level)
(Site-Specific)
Noncancer
Noncancer
Hazard
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(HQ)
Carbondale, Colorado (Garfield County) - RFCO
Toluene
1,190.11
Acrolein
3,464,518.24
Benzene
0.02
Xylenes
730.99
Formaldehyde
62,505.94
1,3-Butadiene
0.01
Benzene
652.88
Benzene
21,762.81
Methanol
623.52
Acetaldehyde
12,509.99
Formaldehyde
612.56
Xylenes
7,309.95
Hexane
169.35
1,3-Butadiene
6,308.09
Acetaldehyde
112.59
Naphthalene
1,592.72
Acrolein
69.29
Propionaldehyde
567.82
Ethylbenzene
67.74
Cadmium, PM
526.47
1.3 -Butadiene
12.62
Arsenic, PM
508.98
Rifle, Colorado (Garfield County) - RICO
Toluene
1,190.11
Acrolein
3,464,518.24
Formaldehyde
0.08
Xylenes
730.99
Formaldehyde
62,505.94
Acetaldehyde
0.06
Benzene
652.88
Benzene
21,762.81
1,3-Butadiene
0.05
Methanol
623.52
Acetaldehyde
12,509.99
Benzene
0.04
Formaldehyde
612.56
Xylenes
7,309.95
Ethylbenzene
<0.01
Hexane
169.35
1,3-Butadiene
6,308.09
Acetaldehyde
112.59
Naphthalene
1,592.72
Acrolein
69.29
Propionaldehyde
567.82
Ethylbenzene
67.74
Cadmium, PM
526.47
1,3-Butadiene
12.62
Arsenic, PM
508.98
-------
Observations from Table 7-6 include the following:
• The 10 highest emitted pollutants with cancer UREs in Mesa County are the highest
emitted pollutants in Garfield County, although not necessarily in the same order.
Benzene and formaldehyde top both lists, although the emissions are more than three
times higher for Garfield County than 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 the same pollutants listed for the pollutants with
the highest toxicity-weighted emissions.
• Eight of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Mesa County; the same eight pollutants have the highest emitted
pollutants and highest toxicity-weighted emissions for Garfield County.
• For GPCO, eight of the 10 pollutants with the highest cancer risk approximations also
appear among the pollutants with the highest toxicity-weighted emissions for Mesa
County (the exceptions are carbon tetrachloride and 1,2-dichloroethane). Note that
POM, Group 2b, which ranks fifth for toxicity-weighted emissions, includes several
PAHs sampled for at GPCO including acenaphthene.
• Each of the pollutants of interest identified for the Garfield County sites appear on
both emissions-based lists in Table 7-6.
Observations from Table 7-7 include the following:
• Toluene is the highest emitted pollutant with a noncancer RfC in both Mesa and
Garfield Counties, although the emissions are considerably higher in Garfield County.
These two counties have an additional eight 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 one county with an NMP site, rarely does it
appear among the highest emitted pollutants. Garfield County is one of only two
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. A similar observation was made in previous NMP reports.
• Five of the highest emitted pollutants in Mesa County 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 both counties, is not among those pollutants with the
7-78
-------
highest toxicity-weighted emissions. Several metals appear near the bottom of each
toxicity-weighted emissions list but do not appear among the highest emitted.
• Formaldehyde, acetaldehyde, benzene, and 1,3-butadiene are pollutants of interest for
GPCO that appear on all three lists in Table 7-7. Naphthalene appears among the
pollutants with the highest noncancer hazard approximations and highest toxicity-
weighted emissions, but is not among the highest emitted pollutants with a noncancer
RfC in Mesa County. This is also true for arsenic. Ethylbenzene appears among the
pollutants with the highest noncancer hazard approximations for GPCO and highest
emissions in Mesa County, but is not among those with the highest toxicity-weighted
emissions.
• Each of the pollutants of interest identified for the Garfield County sites appear on
both emissions-based lists in Table 7-7, 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.
7.6 Summary of the 2014 Monitoring Data for the Colorado Monitoring Sites
Results from several of the data analyses described in this section include the following:
~~~ Sixteen pollutants failed screens for GPCO. The number ofpollutants failing screens
for the Garfield County sites ranged from three to five.
~~~ Formaldehyde and acetaldehyde have highest annual average concentrations for
GPCO; these were the only pollutants with annual average concentrations greater
than 1 ng/m3.
~~~ RICO was the only Garfield County site for which annual average concentrations
could be calculatedfor each of its pollutants of interest.
~~~ PACO and RICO have the highest and third-highest annual average concentrations
of benzene among NMP sites, while GPCO's annual average concentration of
ethylbenzene ranks third-highest. GPCO also has the second highest annual average
concentration of acetaldehyde and the fourth highest annual average concentrations
of formaldehyde and naphthalene among all NMP sites sampling these pollutants.
~~~ GPCO and three Garfield County sites have sampled under the NMP for at least
5 years. Notable trends for these sites include: Benzene concentrations at GPCO
have an overall decreasing trend across the years of sampling while concentrations
of naphthalene have decreased in recent years. In addition, the detection rate of
1,2-dichloroethane at GPCO has increased during the last few years of sampling.
Concentrations of acetaldehyde andformaldehyde appear to have a decreasing trend
at RICO.
~~~ Formaldehyde has the highest cancer risk approximation of the pollutants of interest
for GPCO. Benzene andformaldehyde have the highest cancer risk approximations
for the five Garfield County sites, depending upon whether annual average
7-79
-------
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.
7-80
-------
8.0 Site in the District of Columbia
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS site in Washington, D.C., and integrates these
concentrations with emissions, meteorological, and risk information. Data generated by sources
other than ERG are not included in the data analyses contained in this report. Readers are
encouraged to refer to Sections 1 through 4 and the glossary (Appendix P) 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
geographical and physical information about the location of the site and the surrounding area.
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.
Figure 8-1 is 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 2011 NEI for point sources, version 2.
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 in order to emphasize emissions sources within the boundary. Table 8-1 provides
supplemental geographical information such as land use, location setting, and locational
coordinates.
8-1
-------
Figure 8-1. Washington, D.C. (WADC) Monitoring Site
i. "•'"¦'I'll'
li ''J.i
[siviLife^i^qiu^bia Rd ^3g3^#»,'an Ave
rSrti_y. ¦« ¦" ^SSSfc-. . j eVV^ * \
fflw 1 V, ,
¦ •.» «¦ -z-^iiB"'1,!?. i fp 1 < t
• ?rr«fc I « «£.5 •' Jffi J 'f n %
*\
f^lSh^Ei,l,mon,svNW. :xi. '^tT s \
Mfr: ' -I ¦fi" ^.4," l* | i U "**> » I II ~lr^5flM™TH s-Ti«* « ' n ¦ 1 : . •
Franklin St NE " U
V "* II _4t*> 'li jl1-
i / i - I i\M , |H-U
•ill:-
Evarts St NE ,
w nsm« »1|
mil 'JriMT,
:, ' 7^ l] « iff®
'*?» _'.. K
l.St.NW 1,
%rrr
Howard Univetxjty Jr*
-n*3xs
, nC3'
00
Is TOW
fesk-i
WffT^W]
M,L1\ . RrvantStNW
P IS k:
Bryant St Nw
LJfSw^rT"—", '
Bryant
;St"NW
f?T^J
. . A 2
. n|fi
:GhanniriqiSt^NWi;
¦ ••pw-im 'm ~©"*.!,*
I ;-t ^ B. ^jg M
i :§r;;.
Bryant St NW u
. ; pn^lMpPWr.9 jJ;*M
1,1. nv.M. ,! •.,:. '. .
Qvi
Lw*j » w
-Douglas St. NE—
journal1
[ * * ML' • z*
—t6¦# ' Z Dtxoc University H=piti/' V St NW 4 » ' . f f. L^lU V / f f^V fiL* , I ~
' 7 ' Dixon Unfeenrty ; J V St NNW^^T^l 1 » . J f. T
^ Bldg- Howard ^ ^ ,,L
^ sni^h^a2i®iSlfcmi
— V St NE
¦
'I tts'i.'/
kJP&—
^ g1 --—'a -
Q- i -T T, ¦ L :i *c a /
siqi : ^ J _ /
z' "—-I • • ' M
&- >1
-* ** » * *¦ 1 *
IWettmirwtpr St-NW4 ~|
-------
Figure 8-2. NEI Point Sources Located Within 10 Miles of W A DC
77-5-0-W
77 O'O'-W
76"55'0"W
DISTRICT OF
COLUMBIA
MARYLAND
VIRGINIA
Source Category Group (No. of Facilities)
*t" Airport/Airline/Airport Support Operations (26) x
i Asphalt Production/Hot Mix Asphalt Plant (5) ?
B Bulk Terminals/Bulk Plants (1) "ft
f Electricity Generation via Combustion (3) P
> Hotels/Motels/Lodging (5) X
o Institutional (school, hospital, prison, etc.) (19) *
A Military Base/National Security Facility (11)
Mine/Quarry/Mineral Processing Facility (1)
Miscellaneous Commercial/Industrial Facility (6)
Paint and Coating Manufacturing Facility (1)
Printing/Publishing/Paper Product Manufacturing Facility (6)
Rail Yard/Rail Line Operations (1)
Water Treatment Facility (2)
WADC NATTS site
O 10 mile radius
County boundary
Legend
1 l 1
77°5'0"W 77°0'0"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
8,700
First St between W St and
VSt
1AADT reflects 2013 data (DC DOT, 2014)
BOLD ITALICS = EPA-designated NATTS Site
00
-U
-------
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. Just to the northeast, a construction
project has commenced, which is part of the First Street Tunnel Project (DC WSA, 2016). 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 institutions source category,
which includes hospitals, schools, and prisons, etc.; 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, CBSA, 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 effect concentrations measured at a given monitoring site. The traffic
volume experienced near WADC is less than 9,000 vehicles and is in the bottom third compared
to other NMP sites. The traffic volume provided is for First Street, the closest roadway east of
the monitoring site, between W Street and V Street, three to four blocks south of the site.
Ongoing construction may affect typical traffic patterns in the area.
8.2 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
site in Washington, D.C. on sample days, as well as over the course of the year.
8.2.1 Meteorological Summary
In order to provide an overview of the meteorological conditions experienced at each
monitoring site, hourly meteorological data for 2014 were retrieved and sample day and full-year
averages developed for temperature, dew point, relative humidity, pressure, and wind speed.
Weather data from the actual monitoring site(s) were obtained from AQS, where available. If
site-specific weather data were not available in AQS, then data were obtained from NCDC for
the NWS weather station located closest to the monitoring site(s), as described in Section 3.4.2.
8-5
-------
For WADC, site-specific data were available for all the meteorological parameters except dew
point temperature and sea level pressure. Data for these parameters were obtained from the NWS
weather station at Ronald Reagan Washington National Airport (WBAN 13743). The Reagan
National weather station is located 5.2 miles south-southwest of WADC. A map showing the
distance between the WADC monitoring site and the closest NWS weather station is provided in
Appendix R. These data were used to determine how meteorological conditions on sample days
vary from conditions experienced throughout the year.
Table 8-2. Average Meteorological Conditions near the Washington, D.C. Monitoring Site
Average
Average
Average
Average
Average
Average
Dew Point
Relative
Sea Level
Station
Prevailing
Scalar Wind
Average
Temperature
Temperature
Humidity
Pressure
Pressure
Wind
Speed
Type1
(°F)
(°F)
(%)
(in Hg)
(in Hg)
Direction
(kt)
Washington, D.C. - WADC2
Sample
Days
56.7
42.3
58.8
30.06
29.91
7.4
(63)
±0.9
± 1.0
± 1.0
±0.01
±0.01
NW
±0.2
56.9
42.5
58.7
30.05
29.90
6.9
2014
±0.4
±0.4
±0.4
± <0.01
±<0.01
NW
±0.1
'Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
2Temperature, humidity, station pressure and wind parameters were measured at WADC. The remaining information was
obtained from the closest NWS weather station located at Reagan National Airport, WBAN 13743.
Table 8-2 presents average temperature, average dew point temperature, average relative
humidity, average station and sea level pressure, and wind information (average scalar wind
speed and prevailing wind direction) for days on which samples were collected and for all of
2014. Also included in Table 8-2 is the 95 percent confidence interval for each parameter. As
shown in Table 8-2, average meteorological conditions on sample days were representative of
average weather conditions experienced throughout the year near WADC.
8.2.2 Wind Rose Comparison
Hourly surface wind data were also uploaded into a wind rose software program to
produce customized wind roses, as described in Section 3.4.2. A wind rose shows the frequency
of wind directions using "petals" positioned around a 16-point compass, and uses different colors
to represent wind speeds. Figure 8-3 presents two wind roses for the WADC monitoring site. The
first is a wind rose representing wind observations for all of 2014 and the second is a wind rose
representing wind observations for days on which samples were collected in 2014. These are
8-6
-------
used to identify the predominant wind speed and direction for 2014 and to determine if wind
observations on sample days were representative of conditions experienced over the entire year.
Figure 8-3. Wind Roses for the Wind Data Collected at WADC
2014 Wind Rose Sample Day Wind Rose
^8%
EAST
WEST
\
EAST
WIND SPEED
(Knots)
I I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
H 1
Calms: 0.01%
WIND SPEED
(Knots)
I I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
H 1
Calms: 0.00%
Observations from Figure 8-3 for WADC include the following:
• In 2014, winds from the northwest accounted for the largest percentage of wind
observations at WADC. The strongest winds were observed with this wind direction.
Winds from the south to south-southwest together account for approximately one-
fifth percent of wind observations near WADC, while winds from due north were not
observed at WADC in 2014. Winds with an easterly component were observed more
often than those with a westerly component, excluding the primary directions already
discussed. Calm winds were rarely observed at WADC.
• The sample day wind patterns resemble those on the full-year wind rose, although
there are some differences. Northwesterly winds account for an even higher
percentage of wind observations on sample days in 2014 while winds from the south
and south-southwest accounted for fewer observations on sample days.
8-7
-------
8.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for the
Washington, D.C. monitoring 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 8-3.
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 8-3. It is
important to note which pollutants were sampled for at each site when reviewing the results of
this analysis. Only PAHs were sampled for at WADC in 2014.
Table 8-3. Risk-Based Screening Results for the Washington, D.C. Monitoring Site
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Washington, D.C. - WADC
Naphthalene
0.029
55
60
91.67
96.49
96.49
Benzo(a)pyrene
0.00057
1
52
1.92
1.75
98.25
Fluorene
0.011
1
41
2.44
1.75
100.00
Total
57
153
37.25
Observations from Table 8-3 include the following:
• Concentrations of three pollutants failed screens for WADC: naphthalene,
benzo(a)pyrene, and fluorene.
• Concentrations of naphthalene failed 92 percent of screens, while concentrations of
benzo(a)pyrene and fluorene failed a single screen each.
• Naphthalene accounted for more than 96 percent of the total failed screens for
WADC; thus, naphthalene is WADC's only pollutant of interest.
-------
8.4 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 site.
• 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 are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at WADC are provided in Appendix M.
8.4.1 2014 Concentration Averages
Quarterly and annual average concentrations 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 the entire year of sampling. Annual average concentrations
were calculated for pollutants where three valid quarterly averages could be calculated 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 8-4, 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.
8-9
-------
Table 8-4. Quarterly and Annual Average Concentrations of the Pollutants of Interest
for the Washington, D.C. Monitoring Site
Pollutant
# of
Measured
Detections
vs.
# >MDL
# of
Samples
1st
Quarter
Average
(ng/m3)
2nd
Quarter
Average
(ng/m3)
3rd
Quarter
Average
(ng/m3)
4th
Quarter
Average
(ng/m3)
Annual
Average
(ng/m3)
Washington, D.C. - WADC
Naphthalene
60/60
60
83.50
± 29.50
61.99
± 20.29
65.94
± 16.12
57.92
± 16.09
67.34
± 10.18
Observations for WADC from Table 8-4 include the following:
• Naphthalene was detected in every valid PAH sample collected at WADC.
• Concentrations of naphthalene measured at WADC range from 19.4 ng/m3 to
208 ng/m3.
• The first quarter average concentration of naphthalene is higher than the other
quarterly averages shown in Table 8-4, and the associated confidence intervals
indicate that there is considerably variability in the measurements, particularly for
those measured during the first quarter. The two highest concentrations of
naphthalene measured at WADC, 208 ng/m3 and 172 ng/m3, were both measured in
January 2014. The first quarter has the highest number of naphthalene concentrations
greater than 100 ng/m3 (four) and is the only quarter with a concentration greater than
200 ng/m3. Yet, the number of naphthalene concentrations less than 50 ng/m3
measured at WADC is fairly similar across the quarters (between five and seven were
measured during each quarter). Concentrations measured during the first quarter span
the largest concentration range.
• As shown in Table 4-11, WADC has the ninth highest annual average concentration
of naphthalene compared to other NMP sites sampling PAHs.
8.4.2 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 the site-specific pollutants of
interest, where applicable. Thus, a box plot was created for naphthalene for WADC. Figure 8-4
overlays the site's minimum, annual average, and maximum naphthalene concentrations onto the
program-level minimum, first quartile, median, average, third quartile, and maximum
concentrations, as described in Section 3.4.3.1, and are discussed below.
8-10
-------
Figure 8-4. Program vs. Site-Specific Average Naphthalene Concentration
¦
0
100
200
300
Concentration {ng/m3)
400
500
600
Program:
Site:
1st Qua rti le
¦
Site Average
o
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
Site Concentration Range
4th Qua rti le
~
Average
i
Figure 8-4 presents the box plot for naphthalene for WADC and shows the following:
• The maximum naphthalene concentration measured at WADC is considerably less
than the program-level maximum concentration (568 ng/m3).
• The annual average concentration of naphthalene for WADC (67.34 ± 10.18 ng/m3) is
similar to the program-level average concentration (66.5 ng/m3).
• There were no non-detects of naphthalene measured at WADC, or across the
program.
8.4.3 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 3.4.3.2.
WADC has sampled PAHs under the NMP since mid-2008. Thus, Figure 8-5 presents the 1-year
statistical metrics for naphthalene for WADC. 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.
8-11
-------
Figure 8-5. Yearly Statistical Metrics for Naphthalene Concentrations Measured at WADC
Maximum
1 A 1-year average is not presented because sampling under the NMP did not begin until late June 2008.
Observations from Figure 8-5 for naphthalene concentrations measured at WADC
include the following:
• WADC began sampling PAHs under the NMP in late June 2008.
• The maximum naphthalene concentration shown 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 have been measured in each year of sampling
except 2008 (which included only half a year's worth of samples), 2013 and 2014.
• The 1-year average concentrations exhibit an overall decreasing trend between 2009
and 2014. The 1-year average concentration is less than 100 ng/m3 for the first time in
2013. The 1-year average concentration has decreased by nearly half since 2009.
• The median concentration also has an overall decreasing trend, although the median
increased from 2011 to 2012 before exhibiting further decreases from 2012 to 2013
and again in 2014. (While the 1-year average also increased slightly during this time,
the difference is less than 2 ng/m3.) The median concentration is less than 100 ng/m3
for each year shown in Figure 8-5, and is at a minimum for 2014 (57.0 ng/m3).
• The difference between the 5th and 95th percentiles is at a minimum for 2014,
indicating that the majority of concentrations measured fell within a tighter range of
measurements than the previous years. With the exception of 2011, this is true for
each year following 2009.
8-12
-------
8.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at the WADC monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4 for
definitions and explanations regarding the various toxicity factors, time frames, and calculations
associated with these risk-based screenings.
8.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for WADC 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 3.4.3.3 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 8-5, where applicable. Cancer risk approximations are presented as
probabilities while the noncancer hazard approximations are ratios and thus, unitless values.
Table 8-5. Risk Approximations for the Washington, D.C. Monitoring Site
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections vs.
# of Samples
Annual
Average
(ng/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Washington, D.C. - WADC
Naphthalene
0.000034
0.003
60/60
67.34
± 10.18
2.29
0.02
Observations for WADC from Table 8-5 include the following:
• As discussed in Section 8.4.1, the annual average concentration of naphthalene for
WADC is the ninth highest annual average concentration compared to other NMP
sites sampling this pollutant.
• The cancer risk approximation for naphthalene is greater than 1.0 in-a-million
(2.29 in-a-million).
• The noncancer hazard approximation for naphthalene is significantly less than
1.0, indicating that no adverse noncancer health effects are expected from this
individual pollutant.
8-13
-------
8.5.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, Tables 8-6 and 8-7 present an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 8-6 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
that have cancer toxicity factors. Table 8-6 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 8-6 provides the cancer risk approximation (in-a-million) for the pollutant of interest for
WADC, as presented in Table 8-5. The emissions, toxicity-weighted emissions, and cancer risk
approximations are shown in descending order in Table 8-6. Table 8-7 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 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 8.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
8-14
-------
Table 8-6. 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)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Washington, D.C. (District of Columbia) - WADC
Benzene
110.18
Formaldehyde
1.21E-03
Naphthalene
2.29
Formaldehyde
92.82
Benzene
8.59E-04
Acetaldehyde
52.06
1,3-Butadiene
5.06E-04
Ethylbenzene
51.75
Naphthalene
2.78E-04
T etrachloroethylene
18.70
POM, Group 2b
2.21E-04
1.3 -Butadiene
16.86
Nickel, PM
1.51E-04
Naphthalene
8.18
POM, Group 2d
1.50E-04
POM, Group 2b
2.51
Ethylbenzene
1.29E-04
POM, Group 2d
1.71
Acetaldehyde
1.15E-04
Dichloromethane
0.82
POM, Group 5a
1.11E-04
-------
Table 8-7. 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
(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)
Washington, D.C. (District of Columbia) - WADC
Toluene
363.94
Acrolein
229,665.41
Naphthalene
0.02
Methanol
352.82
Formaldehyde
9,471.05
Hexane
217.66
1.3 -Butadiene
8,432.47
Xylenes
213.36
Acetaldehyde
5,784.35
Ethylene glycol
123.11
Benzene
3,672.70
Benzene
110.18
Nickel, PM
3,505.21
Formaldehyde
92.82
Chlorine
3,176.67
Acetaldehyde
52.06
Naphthalene
2,725.10
Ethylbenzene
51.75
Xylenes
2,133.58
Methyl isobutyl ketone
26.88
Arsenic, PM
1,691.85
-------
Observations from Table 8-6 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 seventh 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,
Group 2b includes several PAHs sampled for at WADC including fluorene, which
failed a single screen for WADC. POM, Group 2d does not include any PAHs
sampled for at WADC. POM, Group 5a includes benzo(a)pyrene, which also failed a
single screen.
Observations from Table 8-7 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 eighth 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 8-7.
Summary of the 2014 Monitoring Data for WADC
Results from several of the data analyses described in this section include the following:
~~~ Although concentrations of three PAHs failed screens, naphthalene failed the
majority of screens and was therefore the only pollutant of interest identified via the
risk screening process.
~~~ The annual average concentration of naphthalene for WADC ranks ninth highest
among NMP sites sampling this pollutant.
8-17
-------
~~~ Concentrations of naphthalene have an overall decreasing trend at WADC and are at
a minimum for 2014.
8-18
-------
9.0 Sites in Florida
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS and UATMP sites in Florida, and integrates these
concentrations with emissions, meteorological, and risk information. Data generated by sources
other than ERG are not included in the data analyses contained in this report. Readers are
encouraged to refer to Sections 1 through 4 and the glossary (Appendix P) 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 geographical and
physical information about the location of the sites and the surrounding areas. 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 five Florida sites are located in two separate urban areas. Three sites (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 are
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 2011 NEI
for point sources, version 2. 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 in order to emphasize emissions sources within the
boundaries. Figures 9-4 through 9-8 are 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.
9-1
-------
Figure 9-1. St. Petersburg, Florida (AZFL) Monitoring Site
-------
Figure 9-2. Pinellas Park, Florida (SKFL) Monitoring Site
vo
U>
)4th=Ave-N.
7*- Sird-Teirace.N-
,89thJerraceN]
l9th-AveN<
—88th Ave N
87th Ave N
,86th Ave. N.
>86thAyeN'
• 86th Ave N >86th Ave n-
i"Terrace,N.
^8SthTerrace N
85th Ave
84th Ave N-
>84th'Ave>N<
-84th,Ave.N.
l3rd
83rd Ave N
. - 82nd Ave N
»81st Terrace N
•BIst'Ave N-
[8l5t-A'veN
.80thAve-N.
i Ave N-
,.79ih Ave N
-------
Figure 9-3. NEI Point Sources Located Within 10 Miles of AZFL and SKFL
82"50'0"W
82°45'0"W
Pinellas
County
82"50'0"W
82° 35'0" W
———
Hillsborough *
County *
Gulf of
\ Mexico
\
\
\ _ _
I
V
1
\
I
\ m
\ Tampa Bay \
'/
I .
82°40'0"W 82°35'0"W
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.
site O 10 mile radius
County boundary
Legend
AZFL UATMP site SKFL NATTS
Source Category Group (No. of Facilities)
Aerospace/Aircraft Manufacturing Facility (1) A
"1" Airport/Airline/Airport Support Operations (9) <•>
it Asphalt Production/Hot Mix Asphalt Plant (1) ?
B Bulk Terminals/Bulk Plants (2) El
C Chemical Manufacturing Facility (2) '[]
ffi Dry Cleaning Facility (1) cz>
6 Electrical Equipment Manufacturing Facility (4) R
f Electricity Generation via Combustion (2) p
F Food Processing/Agriculture Facility (2) a.
Industrial Machinery or Equipment Plant (3) *
O Institutional (school, hospital, prison, etc.) (1) W
¦ Landfill (2)
Metal Coating, Engraving, and Allied Services to Manufacturers (1)
Metals Processing/Fabrication Facility (6)
Miscellaneous Commercial/Industrial Facility (6)
Municipal Waste Combustor (1)
Paint and Coating Manufacturing Facility (2)
Pharmaceutical Manufacturing (1)
Plastic, Resin, or Rubber Products Plant (4)
Printing/Publishing/Paper Product Manufacturing Facility (9)
Ship/Boat Manufacturing or Repair Facility (6)
Wastewater Treatment Facility (2)
Woodwork, Furniture, Millwork & Wood Preserving Facility (1)
9-4
-------
Figure 9-4. Valrico, Florida (SYFL) Monitoring Site
5
QJ
>
n
ydney Rd
' j %
Diamond Hill GoW &
-------
Figure 9-5. NEI Point Sources Located Within 10 Miles of SYFL
Hillsborough
County
pfFi
tor &
Tampa
Bay
82"25'0"W 82r20'0"W 82"15'0"W 82°10'0"W 82"5'0"W
Note: Due to facility density and collocation, the total facilities
Legend displayed may not represent all facilities within the area of interest.
~
SYFL NATTS site
O 10 mile radius
County boundary
Source Category Group (No. of Facilities)
Aerospace/Aircraft Manufacturing Facility (1)
"1" Airport/Airline/Airport Support Operations (7)
£ Asphalt Production/Hot Mix Asphalt Plant (3)
0 Auto Body Shop/Painters/Automotive Stores (1)
« Automobile/Truck Manufacturing Facility (1)
a Brick, Structural Clay, or Clay Ceramics Plant (1)
1 Compressor Station (1)
6 Electrical Equipment Manufacturing Facility (1)
V Fertilizer Plant (1)
F Food Processing/Agriculture Facility (3)
A Landfill (1)
(j§) Metal Can, Box, and Other Metal Container Manufacturing (1)
A Metal Coating, Engraving, and Allied Services to Manufacturers (1)
<•> Metals Processing/Fabrication Facility (5)
X Mine/Quarry/Mineral Processing Facility (2)
? Miscellaneous Commercial/Industrial Facility (3)
M Municipal Waste Combustor (1)
< Pesticide Manufacturing Plant (1)
A Petroleum Refinery (1)
R Plastic, Resin, or Rubber Products Plant (2)
P Printing/Publishing/Paper Product Manufacturing Facility (2)
X Rail Yard/Rail Line Operations (2)
W V\foodwork, Furniture, Millwork & V\food Preserving Facility (1)
9-6
-------
Figure 9-6. Winter Park, Florida (ORFL) Monitoring Site
.W.SwoopeAve-
i Dallas Ave¦«
Si - iJt K ,11
Fairbanks Ave WE3
hks Ave
:Holl Ave-'
Kentucky Ave
Publix
ir>gtonAve
-------
Figure 9-7. Orlando, Florida (PAFL) Monitoring Site
•
-------
Figure 9-8. NEI Point Sources Located Within 10 Miles of QRFL and PAFL
61 20'0"W
Seminole
County
\ Orange
^ County
\
8r25'0"W 81"20'0"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.
Legend
~ ORFL UATMP site ~ PAFL UATMP site O 10 mile radius
Source Category Group (No. of Facilities)
County boundary
f
Airport/Airline/Airport Support Operations (23)
¦
Landfill (1)
it
Asphalt Production/Hot Mix Asphalt Plant (5)
<8>
Metal Can, Box, and Other Metal Container Manufacturing (1)
0
Auto Body Shop/Painters/Automotive Stores (1)
A
Metal Coating, Engraving, and Allied Services to Manufacturers (2)
«
Automobile/Truck Manufacturing Facility (3)
Metals Processing/Fabrication Facility (2)
B
Bulk Terminals/Bulk Plants (1)
X
Mine/Quarry/Mineral Processing Facility (1)
1
Compressor Station (1)
•?
Miscellaneous Commercial/Industrial Facility (5)
e
Electrical Equipment Manufacturing Facility (2)
"0
Paint and Coating Manufacturing Facility (4)
f
Electricity Generation via Combustion (1)
R
Plastic, Resin, or Rubber Products Plant (1)
F
Food Processing/Agriculture Facility (5)
P
Printing/Publishing/Paper Product Manufacturing Facility (4)
*
Industrial Machinery or Equipment Plant (2)
X
Rail Yard/Rail Line Operations (2)
o
Institutional (school, hospital, prison, etc.) (6)
-------
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
40,000
66th St N, N of 9th St
SKFL
12-103-0026
Pinellas
Park
Pinellas
Tampa-St.
Petersburg-
Clearwater, FL
27.850348,
-82.714465
Residential
Suburban
36,500
66th St N, S of 102 Ave N
SYFL
12-057-3002
Valrico
Hillsborough
Tampa-St.
Petersburg-
Clearwater, FL
27.965650,
-82.230400
Residential
Rural
3,800
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
31,500
Orlando Ave, N of Morse Blvd
PAFL
12-095-1004
Orlando
Orange
Orlando-
Kissimmee-
Sanford, FL
28.550833,
-81.345556
Commercial
Suburban
49,000
Colonial/MLK Blvd, b/w
Primrose Rd & Bumby Ave
1AADT reflects 2013 data for PAFL and 2014 data for AZFL, SKFL, SYFL, and ORFL (FL DOT, 2014)
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 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; printing, publishing, and paper product manufacturing; metals processing and
fabrication; and ship/boat manufacturing or repair are the source categories with the greatest
number of emissions sources in the St. Petersburg area (based on the areas covered by the
10-mile radii). The emissions source closest to AZFL is a plastic, resin, or rubber products plant.
While the emissions source closest to SKFL falls into the miscellaneous commercial/industrial
facility source category, a plastic, resin, or rubber products plant, a metals processing/fabrication
facility, and a ship/boat manufacturing or repair facility are also located within 2 miles 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
portion of Figure 9-4) are tanks that are part of the local water treatment facility. This site serves
9-11
-------
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 include a number of sources categories. The
airport source category and metals processing and fabrication are the source categories with the
greatest number of emissions sources near SYFL. The closest source to SYFL with reportable air
emissions in the 2011 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 radii, 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 source to ORFL is a hospital, which falls into the
institutions category, and the heliport located at the hospital, which falls into the airport source
category.
9-12
-------
In addition to providing city, county, CBSA, 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
vehicles can significantly effect concentrations measured at a given monitoring site. The traffic
volume is lowest near SYFL and highest near PAFL, among the Florida sites, with the traffic
volumes for four of the five sites are greater than 30,000 vehicles. The traffic volume for PAFL
ranks 19th highest among other NMP sites, with the traffic volumes for AZFL, SKFL, and ORFL
in the middle of the range compared to other NMP sites (ranking 21st, 22nd, and 25th,
respectively). The traffic volume near SYFL is considerably less than the other Florida sites and
is in the bottom third compared to other NMP sites.
9.2 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
sites in Florida on sample days, as well as over the course of the year.
9.2.1 Meteorological Summary
In order to provide an overview of the meteorological conditions experienced at each
monitoring site, hourly meteorological data for 2014 were retrieved and sample day and full-year
averages developed for temperature, dew point, relative humidity, pressure, and wind speed.
Weather data from the actual monitoring site(s) were obtained from AQS, where available. If
site-specific weather data were not available in AQS, then data were obtained from NCDC for
the NWS weather station located closest to the monitoring site(s), as described in Section 3.4.2.
Limited site-specific data were available in AQS for the Florida sites. For SYFL, site-specific
wind information was available in AQS and is presented in Table 9-2. For the remaining sites
and parameters, data from the closest NWS weather stations are presented. A map showing the
distance between each Florida monitoring site and the closest NWS weather station is provided
in Appendix R. These data were used to determine how meteorological conditions on sample
days vary from conditions experienced throughout the year.
9-13
-------
Table 9-2. Average Meteorological Conditions near the Florida Monitoring Sites
Average
Type1
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(in Hg)
Average
Station
Pressure
(in Hg)
Prevailing
Wind
Direction
Average
Scalar Wind
Speed
(kt)
St. Petersburg, Florida - AZFL2
Sample
Days
73.3
64.2
74.6
30.04
30.03
7.1
(61)
±0.5
±0.5
±0.7
±0.01
±0.01
N
±0.2
73.4
64.4
74.9
30.04
30.03
7.1
2014
±0.2
±0.2
±0.3
± <0.01
±<0.01
N
±0.1
Pinellas Park, Florida - SKFL3
Sample
Days
71.9
62.7
74.7
30.05
30.05
6.6
(66)
±0.5
±0.6
±0.8
±0.01
±0.01
ENE
±0.2
72.4
63.3
74.9
30.05
30.04
6.5
2014
±0.2
±0.2
±0.3
± <0.01
±<0.01
N
±0.1
Valrico, Florida - SYFL
4
Sample
Days
71.2
60.5
72.3
30.03
4.9
(60)
±0.7
±0.7
± 1.1
NA
±0.01
N
±0.2
71.1
60.5
72.3
30.04
5.1
2014
±0.3
±0.3
±0.4
NA
±<0.01
NW
±0.1
Winter Park, Florida - ORFL5
Sample
Days
71.9
61.4
72.5
30.06
29.93
5.7
(60)
±0.6
±0.7
± 1.0
±0.01
±0.01
N
±0.2
72.3
62.3
73.3
30.06
29.94
5.7
2014
±0.2
±0.3
±0.4
± <0.01
±<0.01
N
±0.1
Orlando, Florida - PAFL6
Sample
Days
71.6
61.3
73.2
30.06
29.93
5.6
(30)
±0.8
±0.9
± 1.4
±0.01
±0.01
N
±0.3
72.3
62.3
73.3
30.06
29.94
5.7
2014
±0.2
±0.3
±0.4
± <0.01
±<0.01
N
±0.1
'Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages,
information was obtained from the closest NWS weather station located at St.Petersburg/Whitted Airport, WBAN 92806.
'Information was obtained from the closest NWS weather station located at St.Petersburg/Clearwater International Airport,
WBAN 12873.
4Only wind parameters were measured at S YFL year-round. The remaining information was obtained from the closest NWS
weather station located at Tampa Executive Airport (formerly Vandenburg Airport), WBAN 92816. Data completeness at this
station was between 84% and 90% for each parameter.
''"Information was obtained from the closest NWS weather station located at Orlando Executive Airport, WBAN 12841.
NA= Sea level pressure was not recorded at the Vandenberg Airport.
9-14
-------
Table 9-2 presents average temperature, average dew point temperature, average relative
humidity, average station and sea level pressure, and wind information (average scalar wind
speed and prevailing wind direction) for days on which samples were collected and for all of
2014. Also included in Table 9-2 is the 95 percent confidence interval for each parameter. As
shown in Table 9-2, average meteorological conditions at the Florida monitoring sites on sample
days in 2014 were representative of average weather conditions experienced throughout the
entire year. The largest difference between the 2014 averages and the sample day averages is
shown for PAFL for dew point temperature.
The Florida sites have some of the highest daily average temperatures among the NMP
sites, behind only the Arizona sites. The highest average dew point temperatures among NMP
sites were calculated for the Florida monitoring sites. The Florida sites also experienced some of
the highest relative humidity levels among NMP sites.
9.2.2 Wind Rose Comparison
Hourly surface wind data were also uploaded into a wind rose software program to
produce customized wind roses, as described in Section 3.4.2. A wind rose shows the frequency
of wind directions using "petals" positioned around a 16-point compass, and uses different colors
to represent wind speeds. Figure 9-9 presents two wind roses for the AZFL monitoring site. The
first is a wind rose representing wind observations for all of 2014 and the second is a wind rose
representing wind observations for days on which samples were collected in 2014. These are
used to identify the predominant wind speed and direction for 2014 and to determine if wind
observations on sample days were representative of conditions experienced over the entire year.
Figures 9-10 through 9-13 present the full-year and sample day wind roses for the remaining
Florida monitoring sites.
9-15
-------
Figure 9-9. Wind Roses for the St. Petersburg/Whitted Airport Weather Station near
AZFL
2014 Wind Rose
Sample Day Wind Rose
/ ^
WEST
WIND SPEED
(Knots)
I I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 8.10%
WIND SPEED
(Knots)
I I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 6.98%
Observations from Figure 9-9 for AZFL include the following:
• The weather station at St. Petersburg/Whitted Airport is located 7.1 miles east-
southeast of AZFL. Between them is most of the city of St. Petersburg. Note that the
Whitted Airport is located on the Tampa Bay coast while AZFL is on the west side of
the peninsula near the Boca Ciega Bay.
• The full-year wind rose shows that winds from the north, northeast quadrant, and east
were the most commonly observed wind directions near AZFL, accounting for more
than 40 percent of observations. Winds from the western quadrants were observed
less frequently than winds from the eastern quadrants. Calm winds account for
roughly 8 percent of the hourly wind measurements.
• The sample day wind patterns resemble the full-year wind patterns, with northerly
winds observed the most, along with those from the northeast quadrant and east. The
strongest winds on sample days were observed with easterly wind observations. A
lower percentage of calm winds were observed on sample days.
9-16
-------
Figure 9-10. Wind Roses for the St. Petersburg/Clearwater International Airport Weather
WIND SPEED
(Knots)
I I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 10.10%
Observations from Figure 9-10 for SKFL include the following:
• The weather station at St. Petersburg/Clearwater International Airport is located
4.5 miles north-northeast of SKFL. The St. Petersburg/Clearwater Airport is located
on Old Tampa Bay while SKFL is located farther inland.
• The full-year wind rose shows that winds from a variety of directions were observed
near SKFL. Winds from the northwest to north and northeast to east were the most
commonly observed wind directions while winds from the southwest quadrant were
observed the least. Calm winds accounted for less than 12 percent of the hourly wind
measurements.
• Winds from the northwest to north to east-northeast were frequently observed near
SKFL on sample days, with each of these wind directions accounting for a slightly
higher percentage of observations on sample days than in 2014 as a whole. Thus,
fewer winds with a southerly component were observed on sample days.
Station near SKFL
2014 Wind Rose
Sample Day Wind Rose
WEST
WIND SPEED
(Knots)
I I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 11.55%
9-17
-------
NORTH
WEST;
SOUTH
NORTH
WEST;
SOUTH
Figure 9-11. Wind Roses for the Wind Data Collected at SYFL
2014 Wind Rose Sample Day Wind Rose
WIND SPEED
(Knots)
I I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
WIND SPEED
(Knots)
Observations from Figure 9-11 for SYFL include the following:
• SYFL is the only Florida monitoring site in which wind observations from the site
were available in AQS.
• The full-year wind rose shows that winds from all directions were observed at SYFL,
with winds from the northwest and southeast observed slightly more often than winds
from other directions. Winds greater than 11 knots were observed most often with
winds from the southwest and northwest quadrants than with winds from other
directions. Calm winds account for less than 1 percent of observations collected at
SYFL in 2014.
• Winds from the north were observed most frequently at SYFL on sample days, with
winds from the north to northeast accounting for the directions with the most
observations. The highest wind speeds on sample days were observed with winds
from the southwest and northwest quadrants. Calm winds account for a similar
percentage of observations on sample days as the full-year.
9-18
-------
Figure 9-12. Wind Roses for the Orlando Executive Airport Weather Station near ORFL
2014 Wind Rose Sample Day Wind Rose
: EAST
WEST
WIND SPEED
(Knots)
¦I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 16.57%
WIND SPEED
(Knots)
¦I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 18.76%
calms
Observations from Figures 9-12 for ORFL include the following: Karla to revise for
The closest weather station to both ORFL and PAFL is the one located at the Orlando
Executive Airport. The weather station is located 4 miles south-southeast of ORFL
and less than 1 mile east-southeast of PAFL. Thus, the full-year wind roses presented
for these sites are identical.
Although winds from all directions were observed near these sites, northerly winds
were the most frequently observed near ORFL and PAFL in 2014, accounting for
nine percent of observations. Winds from the east and south each also accounted for
more than 6 percent of observations. Calm winds accounted for approximately
17 percent of observations in 2014.
The sample day wind rose for ORFL shows that winds from the north prevailed,
accounting for more than 10 percent of wind observations. Winds from the north-
northeast, east, and south each accounted for more than 6 percent of observations.
Calm winds accounted for approximately 19 percent of observations on sample days.
9-19
-------
Figure 9-13. Wind Roses for the Orlando Executive Airport Weather Station near PAFL
2014 Wind Rose Sample Day Wind Rose
WEST
W/Z -
EAST
iWEST
EAST.
WIND SPEED
(Knots)
¦I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 16.57%
WIND SPEED
(Knots)
¦I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 19.61%
Observations from Figures 9-13 for PAFL include the following:
• The closest weather station to both ORFL and PAFL is the one located at the Orlando
Executive Airport. The weather station is located less than 1 mile east-southeast of
PAFL, as PAFL is located on the edge of the Orlando Executive Airport property.
The distance between PAFL and the weather station at Orlando Executive Airport is
the shortest distance calculated between a weather station and an NMP site. The full-
year wind rose presented for PAFL is identical to the one presented for ORFL in
Figure 9-12.
• Although winds from all directions were observed near these sites, northerly winds
were the most frequently observed near ORFL and PAFL in 2014, accounting for
nine percent of observations. Winds from the east and south each also accounted for
more than 6 percent of observations. Calm winds accounted for approximately
17 percent of observations in 2014.
• The sample day wind rose for PAFL shares the northerly prominence of ORFL's
sample day wind rose but has a higher percentage of wind observations from the
west-northwest and west. The strongest winds were observed with these directions.
Note that although the sample days are fairly standardized, samples are collected at
PAFL on a l-in-12 day sampling schedule, leading to roughly half the sample days
included in the sample day wind rose as ORFL.
9-20
-------
9.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each Florida
monitoring 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. 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 9-3.
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 9-3. It is
important to note which pollutants were sampled for at each site when reviewing the results of
this analysis. Only carbonyl compounds were sampled for at AZFL, SYFL, and ORFL. PAHs
were sampled for in addition to carbonyl compounds at SKFL. Only PMio metals were sampled
for at PAFL.
Table 9-3. 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
56
56
100.00
50.45
50.45
Acetaldehyde
0.45
55
56
98.21
49.55
100.00
Total
111
112
99.11
Pinellas Park, Florida - SKFL
Acetaldehyde
0.45
50
50
100.00
34.97
34.97
Formaldehyde
0.077
50
50
100.00
34.97
69.93
Naphthalene
0.029
43
58
74.14
30.07
100.00
Total
143
158
90.51
Valrico, Florida - SYFL
Acetaldehyde
0.45
55
55
100.00
50.00
50.00
Formaldehyde
0.077
55
55
100.00
50.00
100.00
Total
110
110
100.00
Winter Park, Florida - ORFL
Acetaldehyde
0.45
60
60
100.00
50.00
50.00
Formaldehyde
0.077
60
60
100.00
50.00
100.00
Total
120
120
100.00
Orlando, Florida - PAFL
Arsenic (PMio)
0.00023
29
30
96.67
100.00
100.00
Total
29
30
96.67
9-21
-------
Observations from Table 9-3 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 and ORFL, formaldehyde and acetaldehyde failed the same number of
screens and contributed equally to the total number of failed screens. For AZFL,
acetaldehyde failed one less screen than formaldehyde. For all three sites,
formaldehyde failed 100 percent of screens.
• Three pollutants failed at least one screen for SKFL (acetaldehyde, formaldehyde,
and naphthalene). Acetaldehyde and formaldehyde failed the same number of screens
and contributed equally to the total number of failed screens, with both failing
100 percent of screens. Naphthalene failed a few less screens, but was still identified
as a pollutant of interest for this site.
• Arsenic is the only PMio metal to fail screens for PAFL. This pollutant was detected
in every metals sample collected at PAFL and failed all but one screen.
9.4 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.
• 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 are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at the Florida monitoring sites are provided in Appendices L, M, and N.
9-22
-------
9.4.1 2014 Concentration Averages
Quarterly and annual concentration averages 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 the entire year of sampling. Annual averages were
calculated for pollutants where three valid quarterly averages could be calculated 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 9-4, 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.
9-23
-------
Table 9-4. Quarterly and Annual Average Concentrations of the Pollutants of Interest
for the Florida Monitoring Sites
Pollutant
# of
Measured
Detections
vs.
# >MDL
# of
Samples
1st
Quarter
Average
frig/m3)
2nd
Quarter
Average
frig/m3)
3rd
Quarter
Average
frig/m3)
4th
Quarter
Average
frig/m3)
Annual
Average
frig/m3)
St. Petersburg, Florida - AZFL
Acetaldehyde
56/56
56
1.54
±0.29
1.51
±0.27
0.86
±0.17
1.26
±0.32
1.31
±0.15
Formaldehyde
56/56
56
1.54
±0.20
2.34
±0.32
3.13
± 1.98
2.12
±0.40
2.24
±0.44
Pinellas Park, Florida - SKFL
Acetaldehyde
50/50
50
0.89
±0.17
0.96
±0.11
NA
1.47
±0.32
NA
Formaldehyde
50/50
50
1.15
±0.14
2.25
±0.29
NA
2.14
±0.37
NA
Naphthalene1
58/58
58
76.65
± 30.40
66.38
±23.61
36.58
±7.73
40.88
± 12.31
54.29
± 10.42
Valrico, Florida - SYFL
Acetaldehyde
55/55
55
1.13
±0.15
1.31
±0.15
1.14
±0.24
1.54
±0.37
1.27
±0.12
Formaldehyde
55/55
55
1.53
±0.17
2.91
±0.39
1.73
±0.33
1.88
±0.40
2.03
±0.22
Winter Park, Florida - ORFL
Acetaldehyde
60/60
60
1.91
±0.48
2.04
±0.41
1.50
±0.19
2.56
±0.69
2.01
±0.25
Formaldehyde
60/60
60
1.48
±0.23
2.72
±0.78
2.75
±0.46
1.91
±0.46
2.20
±0.28
(
)rlando, Florida - PAFL
Arsenic (PMi0)a
30/30
30
0.76
±0.27
0.68
±0.23
0.59
±0.18
1.19
±0.42
0.81
±0.16
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 due to the criteria for calculating a quarterly and/or annual average.
Observations for AZFL from Table 9-4 include the following:
• The first quarter average concentrations of acetaldehyde and formaldehyde for AZFL
are equivalent; yet, the quarterly average concentrations of acetaldehyde decrease
during the warmer months of the year at AZFL while the quarterly average
concentrations of formaldehyde increase. The third quarter average acetaldehyde
concentration is roughly half the concentration shown for the first quarter while the
third quarter average concentration of formaldehyde is twice the first quarter average
concentration.
• Concentrations of acetaldehyde measured at AZFL range from 0.441 |ig/m3 to
2.74 |ig/m3, A review of the acetaldehyde data collected at AZFL shows that
acetaldehyde concentrations greater than 2 |ig/m3 were measured during each
calendar quarter except the third and the number of acetaldehyde concentrations less
9-24
-------
than 1 |ig/m3 measured during the third quarter is two to three times greater than the
number measured during the other calendar quarters.
• Concentrations of formaldehyde measured at AZFL range from 1.02 |ig/m3 to
13.4 |ig/m3. The confidence interval shown for the third quarter average
formaldehyde concentration is large and may indicate that the increase shown across
the calendar quarters is attributable, at least in part, to outlier(s). The maximum
formaldehyde concentration measured at AZFL is 13.4 |ig/m3, which is nearly four
times greater than the next highest concentration measured at this site (3.52 |ig/m3).
This concentration is also the fourth highest formaldehyde concentration measured at
an NMP site in 2014. If the maximum concentration was removed from the
calculation, the quarterly average concentration for the third quarter would fall
between the second and fourth quarter average concentrations.
Observations for SKFL from Table 9-4 include the following:
• Due to a defective sampler, carbonyl compound samples collected between
July 22, 2014 and September 20, 2014 were invalidated. Thus, third quarter and
annual average concentrations for the carbonyl compounds could not be calculated
for SKFL.
• Concentrations of acetaldehyde measured at SKFL range from 0.473 |ig/m3 to
3.05 |ig/m3. Looking at the quarterly averages, it appears that higher acetaldehyde
concentrations were measured after sampling resumed in September. All seven
acetaldehyde concentrations greater than 1.5 |ig/m3 were measured during the fourth
quarter of 2014. Of the 23 acetaldehyde concentrations greater than 1 |ig/m3
measured at SKFL, 11 were measured between January and June and the additional
12 were measured after sampling resumed at the end of September.
• Concentrations of formaldehyde do not follow this trend. Concentrations of
formaldehyde measured at SKFL range from 0.663 |ig/m3 to 3.20 |ig/m3, with all five
formaldehyde concentrations less than 1 |ig/m3 measured between January and
March. All of the formaldehyde concentrations measured during the first quarter are
less than 1.75 |ig/m3, with few concentrations less than 1.75 |ig/m3 measured during
the remaining calendar quarters (two each during the second and third quarter and
five measured during the fourth quarter). The first quarter formaldehyde average
concentration for SKFL is the lowest quarterly average concentration for this
pollutant among the Florida sites.
• Concentrations of naphthalene measured at SKFL range from 11.9 ng/m3 to
191 ng/m3. The first and second quarterly average concentrations of naphthalene are
greater than the third and fourth quarter averages and have larger confidence intervals
associated with them. All six naphthalene concentrations greater than 100 ng/m3 were
measured at SKFL during the first and second quarters of the year. Of the 25
naphthalene concentrations greater than 50 ng/m3 measured at SKFL, 17 were
measured between January and June 2014. Conversely, nine of the 13 lowest
concentrations (those less than 25 ng/m3) were measured during the second half of
2014.
9-25
-------
Observations for SYFL from Table 9-4 include the following:
• Concentrations of formaldehyde were higher than concentrations of acetaldehyde
measured at SYFL, based on the quarterly and annual average concentrations.
• Concentrations of acetaldehyde measured at SYFL range from 0.518 |ig/m3 to
2.55 |ig/m3. The quarterly average concentrations vary relatively little, with the fourth
quarter average the highest and exhibiting the most variability. Both the minimum
and maximum acetaldehyde concentrations were measured at SYFL during the fourth
quarter of 2014, as were three of the four acetaldehyde concentrations greater than
2 |ig/m3.
• Concentrations of formaldehyde measured at SYFL range from 0.752 |ig/m3 to
4.25 |ig/m3, with all six concentrations greater than 3 |ig/m3 measured in May or
June. This explains why, at least in part, the second quarter average concentration is
more than 1 |ig/m3 greater than the other quarterly averages for formaldehyde. The
number of formaldehyde concentrations greater than 2 |ig/m3 measured at SYFL
during the second quarter (14) is considerably greater than the number measured
during the other calendar quarters (zero during the first quarter, and five each during
the third and fourth quarters).
Observations for ORFL from Table 9-4 include the following:
• Concentrations of acetaldehyde measured at ORFL range from 0.756 |ig/m3 to
5.45 |ig/m3, ORFL has the highest annual average concentration of acetaldehyde
among the Florida sites (with available annual averages). Based on the quarterly
average concentrations, acetaldehyde concentrations were lowest at ORFL during the
third quarter of the year and highest (and most variable) during the fourth quarter.
Four of the five highest acetaldehyde concentrations were measured at ORFL
between October and December (those greater than 3.5 |ig/m3). Only one
acetaldehyde concentration greater than 2 |ig/m3 was measured at ORFL during the
third quarter, while six or more were measured during each of the other calendar
quarters (with nine measured during the fourth quarter).
• Concentrations of formaldehyde measured at ORFL range from 0.756 |ig/m3 to
6.54 |ig/m3, the maximum of which is the second highest formaldehyde concentration
measured at a Florida site in 2014. Although the second and third quarter average
concentrations are similar to each other, the confidence interval for the second quarter
is nearly twice the confidence interval for the third quarter average. Three of the four
highest formaldehyde concentrations measured at ORFL were measured in May or
June, with eight of the 10 concentrations greater than 3 |ig/m3 measured during the
second or third quarters of 2014.
9-26
-------
Observations for PAFL from Table 9-4 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.
• Arsenic is the only pollutant identified as a pollutant of interest for this site.
Concentrations of arsenic measured at PAFL span an order of magnitude, ranging
from 0.223 ng/m3 to 2.11 ng/m3.
• Concentrations of arsenic appear highest during the fourth quarter and lowest during
the third quarter, based on the quarterly average concentrations, as the fourth quarter
average is twice the third quarter average concentration. The four highest arsenic
concentrations measured at PAFL were measured between October and December.
The fourth quarter has the greatest number of arsenic measurements greater than
1 ng/m3 (five, including the only measurement greater than 2 ng/m3), while none were
measured between April and September and three were measured during the first
quarter. In addition, arsenic concentrations less than 0.5 ng/m3 were not measured
during the fourth quarter, while between two and three were measured during the
other calendar quarters.
Tables 4-9 through 4-12 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 4-10 for formaldehyde. ORFL is
the only Florida site with an annual average concentration of acetaldehyde greater
than 2 |ig/m3 and ranks ninth among NMP sites sampling this pollutant.
• SKFL does not appear among the sites with the highest annual average concentration
of naphthalene among NMP sites sampling this pollutant.
• The annual average concentration of arsenic for PAFL ranks fourth highest among
NMP sites sampling PMio metals.
9.4.2 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 9-4 for each of the Florida monitoring sites. Figures 9-14 through 9-17 overlay the sites'
minimum, annual average, and maximum concentrations onto the program-level minimum, first
9-27
-------
quartile, median, average, third quartile, and maximum concentrations, as described in Section
3.4.3.1, and are discussed below.
Figure 9-14. Program vs. Site-Specific Average Acetaldehyde Concentrations
¦
F-
0
0123456789 10
Concentration (ng/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 9-14 presents the box plots for acetaldehyde for AZFL, SYFL, and ORFL and
shows the following:
• The box plots show that the range of acetaldehyde concentrations measured is
smallest for SYFL and largest for ORFL. The maximum acetaldehyde concentration
measured at ORFL is twice the maximum concentrations measured at AZFL and
SYFL. The six highest acetaldehyde concentrations among the Florida sites were
measured at ORFL.
• The annual average concentrations calculated for AZFL and SYFL are similar to each
other and are less than both the program4evel average and median concentrations.
The annual average for ORFL is greater than the program-level median and average
concentrations.
• A box plot was not created for SKFL because this site does not have an annual
average concentration.
9-28
-------
Figure 9-15. Program vs. Site-Specific Average Arsenic (PMio) Concentration
~
O i
Program Max Concentration = 10.1 ng/m3
KJ 1
i i i i i
0 1 2 3 4 5 6
Concentration (ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 9-15 presents the box plot for arsenic for PAFL and shows the following:
• The program-level maximum arsenic (PMio) concentration (10.1 ng/m3) is not shown
directly on the box plot in Figure 9-15 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 plots has been reduced.
• The maximum arsenic concentration measured at PAFL is roughly one-fifth the
maximum arsenic (PMio) concentration measured across the program. The minimum
concentration of arsenic measured at PAFL is just less than the program-level first
quartile and is the highest minimum arsenic concentration among NMP sites
sampling arsenic (PMio) metals. These observations were also made in the 2013 NMP
report.
• The annual average concentration of arsenic for PAFL is greater than the program-
level average concentration (0.61 ng/m3) and similar to the program-level third
quartile (0.80 ng/m3). There were no non-detects of arsenic measured at PAFL.
9-29
-------
Figure 9-16. Program vs. Site-Specific Average Formaldehyde Concentrations
¦
¦4-
¦
9 12 15
Concentration {[ig/m3]
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 9-16 presents the box plots for formaldehyde for AZFL, SYFL, and ORFL and
shows the following:
• The box plots show that the range of formaldehyde concentrations measured is
smallest for SYFL and largest for AZFL. The maximum formaldehyde concentration
measured at AZFL is twice the maximum formaldehyde concentration measured at
ORFL and three times the maximum concentration measured at SYFL.
• All of the annual average concentrations of formaldehyde calculated for the Florida
sites (for which an annual average could be calculated) are less than the program-
level average concentration (2.77 |ig/m3) and just less than the program-level median
concentration (2.41 |ig/m3),
• A box plot was not created for SKFL because this site does not have an annual
average concentration.
9-30
-------
Figure 9-17. Program vs. Site-Specific Average Naphthalene Concentration
SKFL
100
200
300
Concentration {ng/m3)
400
500
Program:
1st Qua rti le
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i
Site:
Site Average
o
Site Concentration Range
Figure 9-17 presents the box plot for naphthalene for SKFL and shows the following:
• The maximum arsenic concentration measured at SKFL (191 ng/m3) is considerably
less than the maximum concentration measured across the program (568 ng/m3).
• The annual average concentration of naphthalene for SKFL is less than the program-
level average concentration (66.5 ng/m3) and just greater than the program-level
median concentration (50.7 ng/m3).
9.4.3 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 3.4.3.2.
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 9-18 through 9-27 present the 1-year statistical metrics for each of the
pollutants of interest for each of these Florida monitoring 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.
9-31
-------
Figure 9-18. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at AZFL
J I
o.o I
I
T
I
I
I
o
UJ-1
I
1
t
T
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year
5th Percentile — Minimum - Median — Maximum
O 95th Percentile
Observations from Figure 9-18 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 to
decrease significantly, a trend that continued through 2008.
• The 1-year average and median concentrations began to increase again in 2009. This
increase cannot be attributed to an outlier here or there because nearly all of the
statistical metrics 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. The 1-year average and
9-32
-------
median concentrations shown for 2014 are the second lowest since sampling began,
behind only those calculated for 2008.
Figure 9-19. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at AZFL
I
2
o
I A
.-•W
T
1
I
o
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year
O 5th Percentile - Minimum - Median - Maximum o 95th Percentile ¦¦•^•••Average
Observations from Figure 9-19 for formaldehyde concentrations measured at AZFL
include the following:
• The maximum formaldehyde concentration (16.1 |ig/m3) was measured in 2001, after
which the highest concentration measured in any given year decreased by nearly half,
until 2014. The maximum concentration measured in 2014 (13.4 |ig/m3) is just less
than the maximum concentration measured in 2001.
• The 1-year average and median formaldehyde concentrations 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.
• Although the increase in the maximum concentration measured in 2014 is apparent in
Figure 9-19, each of the statistical parameters also exhibits at least a slight increase
from 2013 to 2014. If the outlier was removed from the calculation, the 1-year
average concentration would decrease only slightly, and the other parameters would
change little. Five concentrations less than the minimum concentration measured in
2014 were measured in 2013. On the other end of the concentration range, the number
of formaldehyde concentrations greater than 2.5 |ig/m3 doubled from 2013 (8) to
9-33
-------
2014 (16). Thus, the slight increases shown for 2014 were not solely attributable to
the outlier concentration.
• The trends shown for formaldehyde in Figure 9-19 are almost the opposite of the
trends shown for acetaldehyde in Figure 9-18, particularly for the period between
2004 through 2008.
Figure 9-20. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SKFL
X
T
JL.
2005 2006
O 5th Percentile
2007 2008 2009 2010 2011 2012 2013 2014 4
Year
- Minimum
- Median - Maximum
O 95th Percentile
Observations from Figure 9-20 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 9-20 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 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,
reaching a maximum in 2010. A significant decrease is shown for 2011 and continued
9-34
-------
into 2012. Although the range of concentrations measured decreased by half for 2013,
the 1-year average concentration changed little.
• The majority of acetaldehyde concentrations measured at SKFL in 2014, as
determined by the difference between the 5th and 95th percentiles, are within the
smallest range since 2006. Although an annual average concentration could not be
calculated for 2014, the median concentrations is shown for all years of sampling.
The median concentration is at a minimum for 2014, and is less than 1 |ig/m3 for the
first time since the first full year of sampling.
Figure 9-21. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at SKFL
o
Concentration for
2005 is 91.7 ng/m3
2005 2006
2007 2008 2009 2010 2011 2012
Year
2013 20141
O 95th Percentile
Observations from Figure 9-21 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 second highest formaldehyde concentration was measured at SKFL
in 2012, and is considerably less (11.4 |ig/m3). No other formaldehyde concentrations
greater than 6 |ig/m3 have been measured at SKFL.
• 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 and median concentrations exhibit an overall decreasing trend
through 2010. The range of measurements is at a minimum for 2010 and the 1-year
9-35
-------
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,
indicating that concentrations of formaldehyde were higher overall at SKFL for 2012.
The 5th percentile for 2012 is greater than several of the central tendency statistics for
several of the previous years.
• All of the statistical parameters exhibit a decrease for 2013. Both the minimum and
5th percentile are at a minimum for 2013.
• The median concentration exhibits a slight increase for 2014, despite the smaller
range of formaldehyde measurements measured in 2014.
Figure 9-22. Yearly Statistical Metrics for Naphthalene Concentrations Measured at SKFL
2011
Year
o 5th Percentile
— Minimum
— Maximum
o 95th Percentile
Observations from Figure 9-22 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).
9-36
-------
• The range within which the majority of naphthalene concentrations fall changed little
through 2011. There is an increase shown for 2012 as this year has the greatest
number of measurements greater than 200 ng/m3 (seven). This increase is followed by
a considerable decrease for 2013, which has the fewest measurements greater than
200 ng/m3 (one) through 2013.
• A decreasing trend in naphthalene concentrations is shown after 2012, with both the
1-year average and median concentrations at a minimum for 2014; this is the first
year in which concentrations of naphthalene greater than 200 ng/m3 were not
measured at SKFL.
Figure 9-23. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SYFL
Sf 10
&
O
£ y e
—£—
2008 2009 2010 2011 2012 2013 2014
Year
O 5th Percentile
- Minimum
- Maximum o 95th Percentile
Observations from Figure 9-23 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
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 (2004).
• After a decreasing trend through 2006, all of the statistical parameters increased for
2007. Even if the two measurements of acetaldehyde discussed above were removed
9-37
-------
from the calculation, the 1-year average concentration for 2007 would still be nearly
twice the next highest 1-year average concentration. While every other year of
sampling has three or less, 2007 has the greatest number of acetaldehyde
concentrations greater than 3 |ig/m3 (16). Thus, it is not just the two highest
measurements driving this 1-year average concentration.
• With the exception of 2007, the 1-year average concentrations of acetaldehyde have
fluctuated between 1.03 |ig/m3 (2011) and 1.60 |ig/m3 (2004). Confidence intervals
calculated for the 1-year averages between 2009 and 2012 indicate that the year-to-
year changes are statistically significant, although the undulating pattern indicates no
specific trend.
• Little change is shown from 2013 to 2014, despite the decrease in the maximum
concentration measured. The difference between the 1-year average concentrations
for these two years is less than 0.025 |ig/m3.
Figure 9-24. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at SYFL
^ J
j}
o
*
t 0 S ?
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year
Observations from Figure 9-24 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.
9-38
-------
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.
Although the maximum concentration for 2007 is considerably less than the
maximum measured in 2005, the other statistical parameters exhibit significant
increases for 2007. In particular, the 95th percentile increased five-fold and the 1-year
average doubled from 2006 to 2007. These statistical parameters indicate that the
concentrations measured in 2007 were higher overall compared to other years. The
number of formaldehyde concentrations greater than 5 |ig/m3 is highest for 2007
(six), while every other year of sampling has two or less.
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 1-year average concentration for 2013 is the lowest since 2006 and has the
smallest range of measurements of any year shown.
Slight increases are shown for most of the parameters for 2014.
9-39
-------
Figure 9-25. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at ORFL
~
I
m
r
2008 2009
Year
O 5th Percentile
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 9-25 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.
• The maximum acetaldehyde concentration was measured in 2006 (9.55 |ig/m3).
Concentrations of at least 5 |ig/m3 have been measured most years of sampling,
although this wasn't the case between 2010 and 2012.
• 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
only 0.1 |ig/m3, despite the considerable increase in the median shown for 2009. The
undulating pattern returns between 2009 and 2012.
• 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. 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.
9-40
-------
• 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.
Figure 9-26. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at ORFL
T
£
|—5—|
KH
2"
2008 2009
Year
O 5th Percentile
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 9-26 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).
• The 1-year average concentrations exhibit an overall decreasing trend through 2011,
starting at 3.26 |ig/m3 for 2004 and reaching a minimum of 1.89 |ig/m3 for 2011. The
statistical metrics for 2007 are the exception to this. However, if the maximum
concentration measured in 2007 was excluded from the calculation, the 1-year
average concentration would exhibit a constant decreasing trend across the years
through 2011. The median concentrations have decreased as well, but exhibited a
considerable increase in 2009, with additional decreases through 2011.
• The central tendency statistics for formaldehyde hover around 2 |ig/m3 for the last
several years of sampling at ORFL.
9-41
-------
Figure 9-27. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at PAFL
Maximum
Observations from Figure 9-27 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.
• Four of the six arsenic concentrations greater than 2 ng/m3 were measured at PAFL in
2012, and ranged from 2.08 ng/m3 to 3.86 ng/m3. The others were measured in 2013
and 2014.
• The range of arsenic concentrations measured is at a minimum for 2010, increases for
2011, then doubles for 2012. The range within which the majority of concentrations
fall, as indicated by the difference between the 5th and 95th percentiles, nearly
doubles from 2010 to 2011 and again for 2012.
• The 1-year average concentration decreases slightly through 2010. After a slight
increase for 2011, the 1-year average increases substantially for 2012. This is the first
time the 1-year average concentration is greater than 1 ng/m3. The median
concentration exhibits a decreasing trend through 2011, even though the range of
measurements increases for 2011, then also increases for 2012. The number of
concentrations greater than 1 ng/m3 increased from two in 2010 to five in 2011 to
nine in 2012. 2012 is the first year arsenic concentrations greater than 2 ng/m3 were
measured at PAFL.
9-42
-------
• Most of the statistical parameters exhibit a decrease from 2012 to 2013, with the 95th
percentile decreasing by almost half from 2012 to 2013. The number of arsenic
measurements greater than 1 ng/m3 returned to five in 2013.
• Despite little change in the majority of arsenic concentrations measured, both the
1-year average and median concentrations exhibit increases for 2014, particularly the
median concentration. Arsenic concentrations greater than 0.5 ng/m3 account for just
over half of the measurements in 2013 and nearly three-quarters of the measurements
in 2014. Thus, the median concentration is at a maximum for 2014.
9.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at each Florida monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4 for
definitions and explanations regarding the various toxicity factors, time frames, and calculations
associated with these risk-based screenings.
9.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Florida 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 3.4.3.3 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 9-5, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
9-43
-------
Table 9-5. Risk Approximations for the Florida Monitoring Sites
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections vs.
# of Samples
Annual
Average
frig/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
St. Petersburg, Florida - AZFL
Acetaldehyde
0.0000022
0.009
56/56
1.31
±0.15
2.88
0.15
Formaldehyde
0.000013
0.0098
56/56
2.24
±0.44
29.06
0.23
Pinellas Park, Florida - SKFL
Acetaldehyde
0.0000022
0.009
50/50
NA
NA
NA
Formaldehyde
0.000013
0.0098
50/50
NA
NA
NA
Naphthalene1
0.000034
0.003
58/58
54.29
± 10.42
1.85
0.02
Valrico, Florida - SYFL
Acetaldehyde
0.0000022
0.009
55/55
1.27
±0.12
2.80
0.14
Formaldehyde
0.000013
0.0098
55/55
2.03
±0.22
26.45
0.21
Winter Park, Florida - ORFL
Acetaldehyde
0.0000022
0.009
60/60
2.01
±0.25
4.42
0.22
Formaldehyde
0.000013
0.0098
60/60
2.20
±0.28
28.61
0.22
Orlando, Florida -
PAFL
Arsenic (PMi0)a
0.0043
0.000015
30/30
0.81
±0.16
3.50
0.05
NA = Not available due to the criteria for calculating an annual average.
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of viewing.
Observations for the Florida sites from Table 9-5 include the following:
• Formaldehyde has the highest cancer risk approximations among the various
pollutants of interest for the Florida sites (where they could be calculated). These
cancer risk approximations span a relatively small range (26.45 in-a-million for SYFL
to 29.06 in-a-million for AZFL).
• The cancer risk approximations for acetaldehyde are an order of magnitude less than
the cancer risk approximations for formaldehyde, ranging from 2.80 in-a-million for
SKFL to 4.42 in-a-million for ORFL.
• For SKFL, naphthalene has a cancer risk approximation of 1.85 in-a-million.
• For PAFL, arsenic has a cancer risk approximation of 3.50 in-a-million.
• 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
9-44
-------
from these individual pollutants. The highest noncancer hazard approximation was
calculated for formaldehyde (0.23), based on the annual average concentration for
AZFL, although similar noncancer hazard approximations were also calculated for
ORFL and SYFL as well as ORFL's annual average concentration of acetaldehyde.
9.5.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, Tables 9-6 and 9-7 present an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 9-6 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
that have cancer toxicity factors. Table 9-6 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 9-6 provides the pollutants with the highest cancer risk approximations (in-a-million) for
each site, as presented in Table 9-5. The emissions, toxicity-weighted emissions, and cancer risk
approximations are shown in descending order in Table 9-6. Table 9-7 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 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 9.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
9-45
-------
Table 9-6. 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)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
St. Petersburg, Florida (Pinellas County) - AZFL
Benzene
423.95
Benzene
3.31E-03
Formaldehyde
29.06
Ethylbenzene
262.81
Formaldehyde
2.68E-03
Acetaldehyde
2.88
Formaldehyde
206.42
1,3-Butadiene
1.77E-03
Acetaldehyde
133.62
Naphthalene
6.91E-04
1.3 -Butadiene
58.94
Ethylbenzene
6.57E-04
Naphthalene
20.33
POM, Group 2b
3.19E-04
Dichloro methane
3.85
Acetaldehyde
2.94E-04
POM, Group 2b
3.63
POM, Group 2d
2.68E-04
POM, Group 2d
3.04
Arsenic, PM
2.28E-04
T etrachloroethylene
1.67
Nickel, PM
1.48E-04
Pinellas Park, Florida (Pinellas County) - SKFL
Benzene
423.95
Benzene
3.31E-03
Naphthalene
1.85
Ethylbenzene
262.81
Formaldehyde
2.68E-03
Formaldehyde
206.42
1,3-Butadiene
1.77E-03
Acetaldehyde
133.62
Naphthalene
6.91E-04
1,3-Butadiene
58.94
Ethylbenzene
6.57E-04
Naphthalene
20.33
POM, Group 2b
3.19E-04
Dichloro methane
3.85
Acetaldehyde
2.94E-04
POM, Group 2b
3.63
POM, Group 2d
2.68E-04
POM, Group 2d
3.04
Arsenic, PM
2.28E-04
T etrachloroethylene
1.67
Nickel, PM
1.48E-04
-------
Table 9-6. 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)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Valrico, Florida (Hillsborough County) - SYFL
Benzene
439.99
Formaldehyde
3.47E-03
Formaldehyde
26.45
Ethylbenzene
294.34
Benzene
3.43E-03
Acetaldehyde
2.80
Formaldehyde
266.66
1,3-Butadiene
1.89E-03
Acetaldehyde
166.39
Nickel, PM
1.47E-03
1.3 -Butadiene
63.16
Cadmium, PM
1.37E-03
Naphthalene
27.75
Arsenic, PM
1.23E-03
Methyl fcrt-biityl ether
7.67
Naphthalene
9.43E-04
POM, Group 2b
5.34
Ethylbenzene
7.36E-04
POM, Group 2d
4.24
Hexavalent Chromium
6.15E-04
Nickel, PM
3.07
POM, Group 2b
4.70E-04
Winter Park, Florida (Orange County) - ORFL
Benzene
557.93
Hexavalent Chromium
5.22E-03
Formaldehyde
28.61
Formaldehyde
373.01
Formaldehyde
4.85E-03
Acetaldehyde
4.42
Ethylbenzene
343.02
Benzene
4.35E-03
Acetaldehyde
198.71
1,3-Butadiene
2.41E-03
1,3-Butadiene
80.46
Naphthalene
1.03E-03
Naphthalene
30.26
Ethylbenzene
8.58E-04
POM, Group 2b
6.43
POM, Group 2b
5.65E-04
POM, Group 2d
4.63
Acetaldehyde
4.37E-04
T etrachloroethylene
2.91
POM, Group 2d
4.08E-04
Dichloro methane
1.09
Arsenic, PM
3.86E-04
-------
Table 9-6. 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)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Orlando, Florida (Orange County) - PAFL
Benzene
557.93
Hexavalent Chromium
5.22E-03
Arsenic
3.50
Formaldehyde
373.01
Formaldehyde
4.85E-03
Ethylbenzene
343.02
Benzene
4.35E-03
Acetaldehyde
198.71
1,3-Butadiene
2.41E-03
1.3 -Butadiene
80.46
Naphthalene
1.03E-03
Naphthalene
30.26
Ethylbenzene
8.58E-04
POM, Group 2b
6.43
POM, Group 2b
5.65E-04
POM, Group 2d
4.63
Acetaldehyde
4.37E-04
T etrachloroethylene
2.91
POM, Group 2d
4.08E-04
Dichloro methane
1.09
Arsenic, PM
3.86E-04
-------
Table 9-7. 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)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
St. Petersburg, Florida (Pinellas County) - AZFL
Toluene
1,691.40
Acrolein
529,170.22
Formaldehyde
0.23
Xylenes
1,112.81
1,3-Butadiene
29,468.37
Acetaldehyde
0.15
Hexane
837.02
Formaldehyde
21,063.17
Methanol
533.81
Acetaldehyde
14,846.98
Benzene
423.95
Benzene
14,131.55
Ethylbenzene
262.81
Xylenes
11,128.12
Formaldehyde
206.42
Naphthalene
6,776.34
Ethylene glycol
183.89
Lead, PM
4,834.15
Acetaldehyde
133.62
Arsenic, PM
3,541.06
Methyl isobutyl ketone
85.23
Nickel, PM
3,431.37
Pinellas Park, Florida (Pinellas County) - SKFL
Toluene
1,691.40
Acrolein
529,170.22
Naphthalene
0.02
Xylenes
1,112.81
1,3-Butadiene
29,468.37
Hexane
837.02
Formaldehyde
21,063.17
Methanol
533.81
Acetaldehyde
14,846.98
Benzene
423.95
Benzene
14,131.55
Ethylbenzene
262.81
Xylenes
11,128.12
Formaldehyde
206.42
Naphthalene
6,776.34
Ethylene glycol
183.89
Lead, PM
4,834.15
Acetaldehyde
133.62
Arsenic, PM
3,541.06
Methyl isobutyl ketone
85.23
Nickel, PM
3,431.37
-------
Table 9-7. 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)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Valrico, Florida (Hillsborough County) - SYFL
Toluene
1,908.71
Acrolein
839,881.94
Formaldehyde
0.21
Xylenes
1,141.28
Cadmium, PM
76,216.14
Acetaldehyde
0.14
Hexane
974.23
Nickel, PM
34,087.06
Methanol
723.09
1,3-Butadiene
31,578.65
Benzene
439.99
Formaldehyde
27,210.45
Hydrochloric acid
356.26
Arsenic, PM
19,144.38
Ethylbenzene
294.34
Acetaldehyde
18,488.27
Ethylene glycol
287.12
Hydrochloric acid
17,813.03
Formaldehyde
266.66
Benzene
14,666.49
Acetaldehyde
166.39
Xylenes
11,412.81
Winter Park, Florida (Orange County) - ORFL
Toluene
2,144.16
Acrolein
1,048,114.49
Formaldehyde
0.22
Xylenes
1,437.17
1,3-Butadiene
40,232.28
Acetaldehyde
0.22
Hexane
985.39
Formaldehyde
38,061.79
Methanol
678.41
Hexamethylene-l,6-diisocyanate, gas
30,043.31
Benzene
557.93
Acetaldehyde
22,079.12
Formaldehyde
373.01
Benzene
18,597.79
Ethylbenzene
343.02
Xylenes
14,371.73
Ethylene glycol
268.80
Naphthalene
10,085.90
Acetaldehyde
198.71
Arsenic, PM
5,985.06
Styrene
101.41
Hydrochloric acid
4,682.79
-------
Table 9-7. 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)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Orlando, Florida (Orange County) - PAFL
Toluene
2,144.16
Acrolein
1,048,114.49
Arsenic
0.05
Xylenes
1,437.17
1,3-Butadiene
40,232.28
Hexane
985.39
Formaldehyde
38,061.79
Methanol
678.41
Hexamethylene-l,6-diisocyanate, gas
30,043.31
Benzene
557.93
Acetaldehyde
22,079.12
Formaldehyde
373.01
Benzene
18,597.79
Ethylbenzene
343.02
Xylenes
14,371.73
Ethylene glycol
268.80
Naphthalene
10,085.90
Acetaldehyde
198.71
Arsenic, PM
5,985.06
Styrene
101.41
Hydrochloric acid
4,682.79
-------
Observations from Table 9-6 include the following:
• Benzene, ethylbenzene, formaldehyde, and acetaldehyde are the highest emitted
pollutants with cancer UREs in Pinellas, Hillsborough, and Orange Counties,
although not necessarily in that order.
• Benzene, formaldehyde, and 1,3-butadiene have the highest toxicity-weighted
emissions for Pinellas County. The same three pollutants have the highest toxicity-
weighted emissions for Hillsborough County but the order is different. Hexavalent
chromium has the highest toxicity-weighted emissions for Orange County, followed
by the other three pollutants.
• Eight of the highest emitted pollutants in Pinellas and Orange Counties also have the
highest toxicity-weighted emissions while seven of the highest emitted pollutants in
Hillsborough County also have the highest toxicity-weighted emissions.
• Formaldehyde, which has the highest cancer risk approximations for the sites
sampling carbonyl compounds and where annual averages could be calculated, 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 Pinellas
and Orange Counties, but acetaldehyde does not appear among those pollutants with
the highest toxicity-weighted emissions for Hillsborough County (it ranks 12th).
• 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 Pinellas
County and appear among the pollutants with the highest toxicity-weighted
emissions. POM, Groups 2b and 2d include several PAHs sampled for at SKFL, none
of which failed screens for these sites.
• Arsenic is the only pollutant of interest for PAFL. Arsenic ranks 10th for toxicity-
weighted emissions for Orange County, but is not among the highest emitted
pollutants, ranking 23rd for quantity emitted. This indicates that even a "low"
quantity of emissions may translate to a relatively "high" toxicity level. Arsenic also
appears among those with the highest toxicity-weighted emissions for the other two
Florida counties with NMP sites. Several metals appear among those with the highest
toxicity-weighted emissions for Pinellas and Hillsborough Counties, but metals were
not sampled there under the NMP.
Observations from Table 9-7 include the following:
• Toluene, xylenes, and hexane are the highest emitted pollutants with noncancer RfCs
in these three Florida counties.
9-52
-------
• 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. None of the Florida sites sampled VOCs under the NMP.
• Four of the highest emitted pollutants in Pinellas and Orange Counties also have the
highest toxicity-weighted emissions. Five of the highest emitted pollutants in
Hillsborough County also have the highest toxicity-weighted emissions. Four of these
pollutants are in common amongst the counties: formaldehyde, acetaldehyde,
benzene, and xylenes.
• Formaldehyde and acetaldehyde appear on both emissions-based lists for each
county.
• Naphthalene is among the pollutants with the highest toxicity-weighted emissions for
each county, except Hillsborough County, but is not among the highest emitted
pollutants (with a noncancer RfC) in any of these counties.
• Arsenic ranks ninth for toxicity-weighted emissions for Orange County, but is not
among the highest emitted pollutants, ranking 46th for quantity emitted. Arsenic is
the only metal that appears among the pollutants with the highest toxicity-weighted
emissions for Orange County. 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.
Metals were sampled for only at PAFL under the NMP.
9.6 Summary of the 2014 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 failed screens for PAFL.
~~~ Among the Florida sites, ORFL has the highest annual average acetaldehyde
concentration while concentrations of formaldehyde did not vary significantly among
these sites. Note that annual averages could not be calculated for SKFL for the
carbonyl compounds. Arsenic was the only metal identified as a pollutant of interest
for PAFL; its annual average rankedfourth highest among NMP sites sampling PMio
metals. PAHs were sampledfor at SKFL and naphthalene was the only PAH
identified as a pollutant of interest for this site. Concentrations of naphthalene
measured at this site were on the lower end compared to other sites sampling this
pollutant.
~~~ Concentrations of acetaldehyde and naphthalene have a decreasing trend at SKFL
while concentrations of acetaldehyde at ORFL have an increasing trend.
9-53
-------
~~~ Formaldehyde has the highest cancer risk approximation among the pollutants of
interest for each Florida site, where carbonyl compounds were sampled and annual
averages could be calculated. None of the pollutants of interest have noncancer
hazard approximations greater than an HQ of 1.0.
9-54
-------
10.0 Sites in Illinois
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS and UATMP sites in Illinois, and integrates these
concentrations with emissions, meteorological, and risk information. Data generated by sources
other than ERG are not included in the data analyses contained in this report. Readers are
encouraged to refer to Sections 1 through 4 and the glossary (Appendix P) 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 geographical and
physical information about the location of the sites and the surrounding areas. 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.
Two monitoring sites are located in northwestern suburbs of Greater Chicago. NBIL is
located in Northbrook and SPIL is located in Schiller Park. The third site (ROIL) is located in
Roxana, just north of the St. Louis CBS A. Figures 10-1 and 10-2 are 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 2011 NEI for point sources, version 2, 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 in order 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.
10-1
-------
Figure 10-1. Northbrook, Illinois (Mill,) Monitoring Site
Edens Expy SpurW^ g^ens gXpy Spur g
RosemaryLn.
Chicago Bounic Garden
Green Acres Country 1
Dundee Rd
a \Dundee Rd
Dundee Rd
~PPfe-ln.
<£ f Source: USGS 1
Source: NASA, NGA, USGS
2008 Microsoft Corp/i"'
o
N>
-------
Figure 10-2. Schiller Park, Illinois (SPIL) Monitoring Site
Iwinona=Ave-
Lawrenc
Lawrence Gt
.EastwoodAy
-------
Figure 10-3. NEI Point Sources Located Within 10 Miles of NBIL and SPIL
87C50'0"W
Lake
County
Lake
Michigan
Cook
County
DuPage ^
County
87°35'0"W
Legend
~ NBIL NATTS site
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
SPIL UATMP site O 10 mile radius
Source Category Group (No. of Facilities)
County boundary
T
Airport/Airline/Airport Support Operations (30)
i
Foundries. Iron and Steel (2)
<—>
Pharmaceutical Manufacturing (7)
i
Asphalt Production/Hot Mix Asphalt Plant (9)
A
Foundries, Non-ferrous (16)
R
Plastic. Resin, or Rubber Products Rant (34)
0
Auto Body Shop/Painters/Automotive Stores (1)
¦
Gasoline/Diesel Service Station (2)
t
Printing, Coating & Dyeing of Fabrics Facility (2)
s
Automobile/Truck Manufacturing Facility (6)
s
Glass Plant (3)
P
Printing/Publlshing/Paper Product Manufacturing Facility (84)
®
Automotive/RV Dealership (1)
>
Hotels/Motels/Lodging (2)
IB
Pulp and Paper Plant (1)
~
Brick. Structural Clay, or Clay Ceramics Plant (1)
*
Industrial Machinery or Equipment Plant (32)
X
Rail Yard/Rail Line Operations (6)
t
Building/Construction (4)
o
Institutional (school, hospital, pnson, etc.) (47)
V
Steel Mill (3)
B
Bulk Terminals/Bulk Plants (11)
¦
Landfill (8)
TT
Telecommunications/Radio Facility (21)
C
Chemical Manufacturing Facility (35)
®
Metal Can. Box. and Other Metal Container Manufacturing (2)
«
Testing Laboratories (1)
i
Compressor Station (7)
A
Metal Coating, Engraving, and Allied Services to Manufacturers (25)
T
Textile, Yarn, or Carpet Plant (1)
[Xj Crematory - Animal/Human (20)
®
Metals Processing/Fabrication Facility (77)
M
Tobacco Manufacturing (1)
Dry Cleaning Facility (63)
X
Mine/Quarry/Mineral Processing Facility (38)
y
Utilities/Pipeline Construction (2)
e
Electrical Equipment Manufacturing Facility (40)
¦?
Miscellaneous Commercial/Industrial Facility (103)
i
Wastewater Treatment Facility (6)
*
Electricity Generation via Combustion (7)
•
Oil and/or Gas Production (5)
Water Treatment Facility (4)
E
F
Electroplating, Plating, Polishing, Anodizing, and Coloring (57)
Food Processing/Agriculture Facility (45)
0
Paint and Coating Manufacturing Facility (11)
w
Woodwork, Furniture. Millwork & Wood Preserving Facility (11)
10-4
-------
Figure 10-4. Roxana, Illinois (ROIL) Monitoring Site
-------
Figure 10-5. NEI Point Sources Located Within 10 Miles of ROIL
90"15'0"W
90"10'0"W
90C0'0"W
89°55'0"W
89°50'0"W
Mississippi
River ,
ILLINOIS
St. Charles
County
MISSOURI
St. Louis
90C25'0"W
90°20'0"W
90°10"0"W
90C5'0"W
Madison '
County |
/ I
Missouri River _ — — — —
; ;* <
-v*~ * [ i
I 1. * !
St. Clair I
County
' t
St. Louis '
County |
Legend
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
ROIL UATMP site
S4MO NATTS site O 10 mile radius
County boundary
Source Category Group (No. of Facilities)
•5" Aerospace/Aircraft Manufacturing Facility (1)
T Airport/Airline/Airport Support Operations (7)
'i Asphalt Production/Hot Mix Asphalt Plant (4)
t=3Brick, Structural Clay, or Clay Ceramics Plant (1)
B Bulk Terminals/Bulk Plants (11)
C Chemical Manufacturing Facility (3)
i Compressor Station (7)
IXCrematory - Animal/Human (3)
© Dry Cleaning Facility (1)
f Electricity Generation via Combustion (3)
F Food Processing/Agriculture Facility (3)
A Foundries, Non-ferrous (1)
H Gasoline/Diesel Service Station (1)
"~"Industrial Machinery or Equipment Plant (1)
O Institutional (school, hospital, prison, etc.) (5)
» Landfill (2)
A Metal Coating, Engraving, and Allied Services to Manufacturers (2)
® Metals Processing/Fabrication Facility (3)
x Mine/Quarry/Mineral Processing Facility (12)
? Miscellaneous Commercial/Industrial Facility (12)
< Pesticide Manufacturing Plant (1)
~ Petroleum Products Manufacturing (1)
a Petroleum Refinery (1)
¥ Port and Harbor Operations (2)
P Printing/Publishing/Paper Product Manufacturing Facility (1)
!B Pulp and Paper Plant (1)
X Rail Yard/Rail Line Operations (1)
VSteel 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
7,750
S Central Ave at Hawthorne Ave
1 AADT reflects 2013 data for SPIL and ROIL and 2014 data for NBIL (IL DOT, 2013/2014)
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 are printing/publishing/paper product manufacturing;
metals processing/fabrication; dry cleaning; electroplating, plating, polishing, anodizing, and
coloring; institutions (schools, hospitals, prisons, etc.); and food processing/agriculture. 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
10-8
-------
(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.
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, CBSA, 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 effect concentrations measured at a given monitoring site. SPIL
experiences a higher traffic volume compared to NBIL, although the traffic volumes near these
sites are both significantly greater than the traffic volume near ROIL. SPIL's traffic volume is
the third highest among all NMP sites. The traffic volume for NBIL ranks ninth 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 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
sites in Illinois on sample days, as well as over the course of the year.
10.2.1 Meteorological Summary
In order to provide an overview of the meteorological conditions experienced at each
monitoring site, hourly meteorological data for 2014 were retrieved and sample day and full-year
averages developed for temperature, dew point, relative humidity, pressure, and wind speed.
Weather data from the actual monitoring site(s) were obtained from AQS, where available. If
site-specific weather data were not available in AQS, then data were obtained from NCDC for
10-9
-------
the NWS weather station located closest to the monitoring site(s), as described in Section 3.4.2.
For the Illinois sites, site-specific data were available for some, but not all, of the parameters in
Table 10-2. For NBIL, many of the meteorological instruments were down in 2014 and only
pressure information was available in AQS; for SPIL, only wind information was available in
AQS; weather data were not available in AQS for ROIL. Data from the closest NWS weather
station was used for the remaining parameters. The Chicago Executive Airport weather station is
located 5.6 miles west-southwest of NBIL; the O'Hare International Airport weather station is
located 3.6 miles northwest of SPIL; and the Lambert/St. Louis International Airport weather
station is located 17.4 miles west-southwest of ROIL. A map showing the distance between each
Illinois monitoring site and the closest NWS weather station is provided in Appendix R. These
data were used to determine how meteorological conditions on sample days vary from conditions
experienced throughout the year.
Table 10-2. Average Meteorological Conditions near the Illinois Monitoring Sites
Average
Average
Average
Average
Average
Average
Dew Point
Relative
Sea Level
Station
Prevailing
Scalar Wind
Average
Temperature
Temperature
Humidity
Pressure
Pressure
Wind
Speed
Type1
(°F)
(°F)
(%)
(in Hg)
(in Hg)
Direction
(kt)
Northbrook, Illinois - NBIL2
Sample
Days
46.3
34.3
65.3
30.08
29.45
6.5
(62)
± 1.1
± 1.1
±0.8
±0.01
±0.01
W
±0.2
47.8
36.6
67.5
30.03
29.39
7.0
2014
±0.5
±0.4
±0.3
±<0.01
± <0.01
s
±0.1
Schiller Park, Illinois - SPIL3
Sample
Days
47.0
33.7
62.1
30.06
29.33
6.2
(61)
± 1.2
± 1.1
±0.8
±0.01
±0.01
sw
±0.2
47.9
35.4
64.2
30.01
29.28
6.9
2014
±0.5
±0.4
±0.3
±<0.01
± <0.01
sw
±0.1
Roxana, Illinois - ROIL4
Sample
Days
54.3
40.2
61.4
30.09
29.32
6.9
(62)
± 1.1
± 1.1
±0.8
±0.01
±0.01
ESE
±0.2
55.9
42.2
62.6
30.04
29.28
7.4
2014
±0.5
±0.4
±0.4
±<0.01
± <0.01
s
±0.1
'Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
2Only station pressure was measured at NBIL in 2014. The remaining information was obtained from the closest NWS weather
station located at Chicago Executive Airport, WBAN 04838.
3Only wind parameters were measured at SPIL. The remaining information was obtained from the closest NWS weather station located
at O'Hare International Airport, WBAN 94846.
4This information was obtained from the NWS weather station located at Lambert/St. Louis International Airport, WBAN 13994.
10-10
-------
Table 10-2 presents average temperature, average dew point temperature, average relative
humidity, average station and sea level pressure, and wind information (average scalar wind
speed and prevailing wind direction) for days on which samples were collected and for all of
2014. Also included in Table 10-2 is the 95 percent confidence interval for each parameter. As
shown in Table 10-2, average meteorological conditions on sample days near NBIL, SPIL, and
ROIL were generally representative of average weather conditions experienced throughout the
year near these sites. The largest difference shown in Table 10-2 is for NBIL and the dew point
temperature, although dew point and/or relative humidity have the largest differences for each
site. Note the difference in the temperature parameters between the Chicago sites and ROIL.
These differences are expected, given the roughly 250 mile distance between these sites.
10.2.2 Wind Rose Comparison
Hourly surface wind data were also uploaded into a wind rose software program to
produce customized wind roses, as described in Section 3.4.2. A wind rose shows the frequency
of wind directions using "petals" positioned around a 16-point compass, and uses different colors
to represent wind speeds. Figure 10-6 presents two wind roses for the NBIL monitoring site. The
first is a wind rose representing wind observations for all of 2014 and the second is a wind rose
representing wind observations for days on which samples were collected in 2014. These are
used to identify the predominant wind speed and direction for 2014 and to determine if wind
observations on sample days were representative of conditions experienced over the entire year.
Figures 10-7 and 10-8 present the full-year and sample day wind roses for SPIL and ROIL.
10-11
-------
Figure 10-6. Wind Roses for the Chicago Executive Airport Weather Station near NBIL
2014 Wind Rose Sample Day Wind Rose
NORTH
NORTH
WEST
WEST
WIND SPEED
(Knots)
WIND SPEED
(Knots)
SOUTH
SOUTH
Calms: 15.36%
Calms: 15.79%
Observations from Figure 10-6 for NBIL include the following:
• The wind instruments at NBIL were down in 2014, so the wind data provided here is
from the Chicago Executive Airport weather station, which is located 5.6 miles west-
southwest of NBIL, and about four times as far from Lake Michigan as NBIL.
• The full-year wind rose shows that winds from a variety of directions were observed
near NBIL. Winds from the south, south-southwest, and west together account for
nearly one-third of wind observations while winds from the east-southeast and
southeast were observed the least. Calm were observed for 15 percent of the hourly
measurements while the strongest winds were most often out of the south.
• The sample day wind patterns generally resemble the full-year wind patterns in that
winds from the south, south-southwest, and west were observed the most and winds
from the east-southeast and southeast were observed the least. Westerly winds
accounted for a higher percentage of observations on sample days while fewer
southerly and south-southwesterly winds were observed compared to the full-year
wind rose. Fewer winds from the northwest quadrant were observed on sample days
while a greater percentage of winds from the northeast quadrant were observed. Also,
winds appear lighter on sample days; winds speeds greater than 11 knots account for
a fewer percentage of observations on sample days than throughout the year.
10-12
-------
Figure 10-7. Wind Roses for the Wind Data Collected at SPIL
2014 Wind Rose Sample Day Wind Rose
¦ " NORTH NORTH
6%
WEST
EAST
WEST
WIND SPEED
(Knots)
¦I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 0.16%
WIND SPEED
(Knots)
¦I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 0.55%
Observations from Figure 10-7 for SPIL include the following:
• SPIL is the only NMP site in Illinois for which 2014 wind data were available in
AQS.
• The 2014 wind rose shows that winds from the southwest quadrant account for the
majority of wind observations collected at SPIL in 2014, with very few wind
observations from the eastern quadrants. Few calm winds were observed at SPIL in
2014.
The sample day wind patterns resemble those of the full-year wind rose, with the
winds from the south to southwest to west accounting for the majority of observations
and a calm rate less than 1 percent.
10-13
-------
Figure 10-8. Wind Roses for the Lambert/St. Louis International Airport Weather Station
near ROIL
2014 Wind Rose
Sample Day Wind Rose
NORTH""--....
15%
| 15%
12%
!2%
9%.
m 9%.
6% \
7 1 [east]
jWEST; I
I EAST:
WIND SPEED
(Knots)
I I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 12.44%
WIND SPEED
(Knots)
I I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 15.12%
Observations from Figure 10-8 for ROIL include the following:
• The Lambert/St. Louis International Airport weather station is located 17.4 miles
west-southwest of ROIL. The airport lies on the northwest side of St. Louis and is
south of the Missouri River.
• The 2014 wind rose shows that winds from a variety of directions were observed near
ROIL, although winds from the northeast quadrant were rarely observed. Winds from
the south were observed the most in 2014, with winds from the west to northwest also
commonly observed. Calm winds were observed for 12 percent of the hourly
measurements while the strongest winds were most often observed with winds with a
westerly component.
Winds from the south and east-southeast were nearly equally observed near ROIL on
sample days. The remaining wind directions accounted for 6 percent of observations
or fewer. Winds from the west to northwest were observed slightly less on sample
days compared to the full year. The calm rate on sample days is higher (15 percent)
than for the full year's worth of observations (12 percent).
10-14
-------
10.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each Illinois
monitoring 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. 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 10-3.
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 10-3. 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 only VOCs and carbonyl compounds were sampled for at SPIL and ROIL.
Table 10-3. Risk-Based Screening Results for the Illinois Monitoring Sites
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Northbrook, Illinois - NBIL
Benzene
0.13
56
56
100.00
11.31
11.31
Carbon Tetrachloride
0.17
55
56
98.21
11.11
22.42
Formaldehyde
0.077
55
55
100.00
11.11
33.54
1,2-Dichloroethane
0.038
54
54
100.00
10.91
44.44
Acetaldehyde
0.45
53
55
96.36
10.71
55.15
Naphthalene
0.029
46
55
83.64
9.29
64.44
Arsenic (PMio)
0.00023
45
53
84.91
9.09
73.54
1.3 -Butadiene
0.03
35
43
81.40
7.07
80.61
Acenaphthene
0.011
27
55
49.09
5.45
86.06
Fluorene
0.011
25
49
51.02
5.05
91.11
Fluoranthene
0.011
18
55
32.73
3.64
94.75
Hexacliloro -1,3 -butadiene
0.045
17
19
89.47
3.43
98.18
Chloroform
9.8
2
56
3.57
0.40
98.59
/?-Dichlorobcnzcnc
0.091
2
22
9.09
0.40
98.99
Ethylbenzene
0.4
2
56
3.57
0.40
99.39
Benzo(a)pyrene
0.00057
1
55
1.82
0.20
99.60
Cadmium (PMio)
0.00056
1
53
1.89
0.20
99.80
Dichloromethane
60
1
55
1.82
0.20
100.00
Total
495
902
54.88
10-15
-------
Table 10-3. Risk-Based Screening Results for the Illinois Monitoring Sites (Continued)
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Schiller Park, Illinois - SPIL
Acetaldehyde
0.45
59
59
100.00
15.21
15.21
Formaldehyde
0.077
59
59
100.00
15.21
30.41
Benzene
0.13
55
55
100.00
14.18
44.59
1.3 -Butadiene
0.03
55
55
100.00
14.18
58.76
Carbon Tetrachloride
0.17
55
55
100.00
14.18
72.94
1,2-Dichloroethane
0.038
52
52
100.00
13.40
86.34
T richloroethylene
0.2
24
49
48.98
6.19
92.53
Hexacliloro -1,3 -butadiene
0.045
12
13
92.31
3.09
95.62
Ethylbenzene
0.4
7
55
12.73
1.80
97.42
/?-Dichlorobcnzcnc
0.091
4
25
16.00
1.03
98.45
Propionaldehyde
0.8
3
59
5.08
0.77
99.23
1,2-Dibromoethane
0.0017
2
2
100.00
0.52
99.74
T etrachloroethylene
3.8
1
55
1.82
0.26
100.00
Total
388
593
65.43
Roxana, Illinois - ROIL
Acetaldehyde
0.45
60
60
100.00
16.09
16.09
Formaldehyde
0.077
60
60
100.00
16.09
32.17
Benzene
0.13
58
58
100.00
15.55
47.72
Carbon Tetrachloride
0.17
58
58
100.00
15.55
63.27
1.3 -Butadiene
0.03
51
55
92.73
13.67
76.94
1,2-Dichloroethane
0.038
51
51
100.00
13.67
90.62
Hexacliloro -1,3 -butadiene
0.045
16
17
94.12
4.29
94.91
Ethylbenzene
0.4
14
58
24.14
3.75
98.66
1,2-Dibromoethane
0.0017
2
2
100.00
0.54
99.20
p-Dichlorobenzene
0.091
1
17
5.88
0.27
99.46
Propionaldehyde
0.8
1
60
1.67
0.27
99.73
Xylenes
10
1
58
1.72
0.27
100.00
Total
373
554
67.33
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.
• Eighteen pollutants failed at least one screen for NBIL; 55 percent of concentrations
for these 18 pollutants were greater than their associated risk screening value (or
failed screens).
10-16
-------
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 sixth highest number of screens (495) among NMP sites, as shown in
Table 4-8 of Section 4.2, and had the third highest number of pollutants whose
concentrations failed screens (18). However, the failure rate for NBIL, when
incorporating all pollutants with screening values, is relatively low, at 20 percent.
This is due primarily to the relatively high number of pollutants sampled for at this
site. NBIL is one of only three NMP sites sampling five pollutant groups and one of
only two sites to sample with both the TO-15 and SNMOC methods. Recall from
Section 3.2 that 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.
Thirteen pollutants failed screens for SPIL; approximately 65 percent of
concentrations for these 13 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.
Twelve pollutants failed screens for ROIL; approximately 67 percent of
concentrations for these 12 pollutants were greater than their associated risk screening
value (or failed screens).
Eight pollutants contributed to 95 percent of failed screens for ROIL and therefore
were identified as pollutants of interest for this site. These eight include two carbonyl
compounds and six VOCs.
The Illinois monitoring sites have seven pollutants of interest in common: two
carbonyl compounds (acetaldehyde and formaldehyde) and five VOCs (benzene,
1,3-butadiene, carbon tetrachloride, 1,2-dichloroethane, and hexachloro-1,3-
butadiene). Of these, benzene, formaldehyde, and 1,2-dichloroethane failed
100 percent of screens for each site.
10-17
-------
10.4 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.
• 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 are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at NBIL, SPIL, and ROIL are provided in Appendices J through N.
10.4.1 2014 Concentration Averages
Quarterly and annual concentration averages 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 the entire year of sampling. Annual averages were calculated for
pollutants where three valid quarterly averages could be calculated 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 10-4, 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.
10-18
-------
Table 10-4. Quarterly and Annual Average Concentrations of the Pollutants of Interest
for the Illinois Monitoring Sites
Pollutant
# of
Measured
Detections
vs.
# >MDL
# of
Samples
1st
Quarter
Average
frig/m3)
2nd
Quarter
Average
frig/m3)
3rd
Quarter
Average
frig/m3)
4th
Quarter
Average
frig/m3)
Annual
Average
frig/m3)
Northbrook, Illinois - NBIL
Acetaldehyde
55/55
55
2.35
±0.74
2.65
± 1.10
3.00
±0.76
1.29
±0.90
2.36
±0.45
Benzene
56/56
56
0.63
±0.09
0.34
±0.05
0.56
±0.16
NA
0.49
±0.06
1.3 -Butadiene
43/36
56
0.05
±0.04
0.02
±0.01
0.05
±0.02
NA
0.04
±0.01
Carbon Tetrachloride
56/55
56
0.48
±0.09
0.67
±0.02
0.65
±0.03
NA
0.60
±0.03
1,2-Dichloroethane
54/52
56
0.07
±0.01
0.08
±0.01
0.06
±0.01
NA
0.07
± <0.01
Formaldehyde
55/55
55
1.23
±0.39
1.65
±0.55
1.31
±0.36
0.91
±0.34
1.29
±0.21
Hexachloro-1,3 -butadiene
19/0
56
0.03
±0.02
0.02
±0.02
0.03
±0.02
NA
0.02
±0.01
Acenaphthene3
55/55
55
2.75
± 1.61
26.05
± 10.69
46.31
±24.01
6.31
±4.68
20.62
±8.20
Arsenic (PMio)3
53/53
53
NA
NA
1.07
±0.33
0.54
±0.17
NA
Fluoranthene3
55/55
55
1.74
±0.69
11.38
±4.59
16.58
±5.12
2.65
± 1.51
8.12
±2.37
Fluorene3
49/49
55
2.31
± 1.50
21.66
±8.73
38.81
± 19.27
5.78
±4.05
17.37
±6.69
Naphthalene3
55/55
55
45.44
± 19.67
132.42
± 44.09
203.35
± 70.64
51.91
±27.25
109.13
±28.14
Schiller Park, Illinois - SPIL
Acetaldehyde
59/59
59
3.92
± 1.30
1.53
±0.43
1.58
±0.22
2.91
± 1.09
2.52
±0.50
Benzene
55/55
55
0.79
±0.11
NA
0.92
±0.21
0.85
±0.24
0.79
±0.09
1.3 -Butadiene
55/55
55
0.14
±0.03
NA
0.16
±0.03
0.13
±0.03
0.13
±0.02
Carbon Tetrachloride
55/55
55
0.57
±0.05
NA
0.63
±0.02
0.58
±0.07
0.60
±0.02
1,2-Dichloroethane
52/52
55
0.08
±0.01
NA
0.06
±0.02
0.09
±0.02
0.08
±0.01
Formaldehyde
59/59
59
4.11
±0.80
2.87
±0.86
3.06
±0.43
2.43
±0.51
3.12
±0.35
Hexachloro-1,3 -butadiene
13/0
55
0.03
±0.02
NA
0.01
±0.01
0.03
±0.02
0.02
±0.01
T richloroethylene
49/37
55
0.34
±0.34
NA
0.72
±0.42
0.36
±0.38
0.44
±0.17
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
10-19
-------
Table 10-4. Quarterly and Annual Average Concentrations of the Pollutants of Interest
for the Illinois Monitoring Sites (Continued)
# of
Measured
1st
2nd
3rd
4th
Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs.
# >MDL
# of
Samples
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Roxana, Illinois - ROIL
1.61
1.94
2.33
1.74
1.91
Acetaldehyde
60/60
60
±0.30
±0.23
±0.41
±0.39
±0.18
1.39
1.30
1.09
1.11
1.22
Benzene
58/58
58
±0.78
±0.30
±0.29
±0.28
±0.21
0.05
0.06
0.08
0.10
0.07
1,3-Butadiene
55/51
58
±0.02
±0.01
±0.02
±0.04
±0.01
0.59
0.66
0.65
0.60
0.63
Carbon Tetrachloride
58/58
58
±0.06
±0.03
±0.03
±0.05
±0.02
0.10
0.09
0.07
0.06
0.08
1,2-Dichloroethane
51/51
58
±0.02
± <0.01
±0.02
±0.02
±0.01
0.43
0.31
0.30
0.26
0.32
Ethylbenzene
58/58
58
±0.45
±0.06
±0.08
±0.10
±0.10
1.87
3.52
4.82
1.98
3.05
Formaldehyde
60/60
60
±0.29
±0.65
± 1.11
±0.41
±0.45
0.02
0.01
0.02
0.03
0.02
Hexachloro-1,3 -butadiene
17/0
58
±0.02
±0.02
±0.02
±0.02
±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 due to the criteria for calculating a quarterly and/or annual average.
Observations for NBIL from Table 10-4 include the following:
• The pollutants with the highest annual average concentrations are acetaldehyde
(2.36 ± 0.45 |ig/m3) and formaldehyde (1.29 ± 0.21 |ig/m3). NBIL is one of only two
sites for which the annual average concentration of acetaldehyde is greater than the
annual average concentration of formaldehyde (SEWA is the other). A similar
observation was made in the 2013 NMP report. The annual average concentrations
for the remaining pollutants of interest are less than 1 |ig/m3, with carbon
tetrachloride as the next highest (0.60 ± 0.03 |ig/m3).
• Acetaldehyde concentrations measured at NBIL range from 0.159 |ig/m3 to
9.17 |ig/m3. The maximum acetaldehyde concentration measured at NBIL is the
maximum acetaldehyde concentration measured across the program. An acetaldehyde
concentration greater than 5 |ig/m3 was measured at NBIL during each calendar
quarter in 2014. The fourth quarter average acetaldehyde concentration is roughly
half the magnitude of the other quarter averages and has a relatively large confidence
interval associated with it. A review of the data shows that at least seven acetaldehyde
concentrations greater than 2 |ig/m3 were measured during each quarter except the
fourth quarter, when only one was measured. On the other end of the concentration
range, the number of acetaldehyde concentrations less than 1 |ig/m3 measured at
NBIL during the fourth quarter of 2014 (seven) is more than twice the number
measured throughout the rest of the year (three), including the only two less than
0.5 |ig/m3.
10-20
-------
The quarterly average concentrations of formaldehyde exhibit a similar pattern as the
quarterly averages of acetaldehyde. A review of the data shows that formaldehyde
concentrations measured atNBIL range from 0.108 |ig/m3 to 4.74 |ig/m3, with the
maximum concentration measured on the same day as the maximum acetaldehyde
concentration (April 11, 2014). The documentation indicated that there may have
been timer issues for this sample. At least two formaldehyde concentrations greater
than 2 |ig/m3 were measured during each calendar quarter except the fourth quarter,
when none were measured. On the other end of the concentration range, the number
of formaldehyde concentrations less than 0.5 |ig/m3 measured atNBIL during the
fourth quarter of 2014 (three) is more than the number measured throughout the rest
of the year (one), and includes the only two less than 0.25 |ig/m3.
Fourth quarter average concentrations for the VOCs in Table 10-4 could not be
calculated because there were too many invalid samples during this quarter to meet
the 75 percent criteria. Of the VOCs shown, carbon tetrachloride and benzene have
the only annual average concentrations greater than 0.1 |ig/m3.
The first quarter average concentration of carbon tetrachloride is significantly less
than the other available quarterly averages. A review of the data shows that all six
carbon tetrachloride measurements less than 0.5 |ig/m3 were measured during the first
quarter, including one less than 0.1 |ig/m3. The first quarter is the only calendar
quarter during which a concentration greater than 0.7 |ig/m3 was not measured.
The second quarter average concentration of benzene is significantly less than the
other available quarterly averages. A review of the data shows that benzene
concentrations measured at NBIL span an order of magnitude, ranging from
0.189 |ig/m3 to 1.39 |ig/m3. There were 23 benzene concentrations greater than
0.5 |ig/m3 measured atNBIL; of these, only one was measured during the second
quarter of 2014, compared to 11, seven, and four measured during the first, third, and
fourth quarters, respectively.
There were a number of invalid metals samples during the first and second quarters of
2014, resulting in these quarters not meeting the 75 percent criteria for quarterly
averages to be calculated and thus, no annual average is provided either. However, a
site-specific statistical summary for arsenic is provided in Appendix N. Arsenic
concentrations measured at NBIL in 2014 range from 0.117 ng/m3 to 2.16 ng/m3.
Table 10-4 shows that the third quarter average concentration of arsenic is twice the
fourth quarter average. Of the 10 arsenic concentrations greater than 1 ng/m3
measured at NBIL, seven were measured during the third quarter (compared to one
measured during the second quarter and two measured during the fourth quarter of
2014). Comparing just the two quarters with available quarterly average
concentrations, six concentrations were measured during the third quarter that are
greater than the maximum concentration measuring during the fourth quarter. On the
other end of the concentration range, the number of arsenic concentrations less than
0.5 ng/m3 measured during the fourth quarter (10) is five times greater than the
number measured during the third quarter of 2014 (two).
10-21
-------
• Of the PAHs shown, naphthalene has the highest annual average concentration
(109.13 ± 28.14 ng/m3). Concentrations of each of the PAH pollutants of interest
were significantly higher during the warmer months of the year and exhibit a
relatively large amount of variability, based on the confidence intervals.
Concentrations of naphthalene measured at NBIL range from 17.5 ng/m3 to
568 ng/m3. The maximum concentration measured at NBIL is the highest naphthalene
concentration measured across the program in 2014. This was also true in 2013. Only
two of the 22 naphthalene concentrations greater than 100 ng/m3 measured at NBIL
were measured outside the second and third quarters of the year; conversely, all but
two of the 21 measurements less than 50 ng/m3 were measured during the first or
fourth quarters of 2014.
• 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 0.491 ng/m3 to 198 ng/m3, accounting
for six of the 10 highest acenaphthene measurements across the program (those
greater than 50 ng/m3). Concentrations of fluorene measured at NBIL range from
0.782 ng/m3 to 161 ng/m3, including the maximum fluorene measurement across the
program. Concentrations of fluoranthene range from 0.514 ng/m3 to 36.8 ng/m3, with
all six fluoranthene concentrations greater than 20 ng/m3 across the program
measured at NBIL. Many of the higher PAH concentrations were measured on the
same days. For instance, the highest naphthalene, acenaphthene, and fluorene
concentrations were measured at NBIL on August 9, 2014, along with the second
highest fluoranthene concentration.
Observations for SPIL from Table 10-4 include the following:
• The pollutants with the highest annual average concentrations are formaldehyde
(3.12 ± 0.35 |ig/m3) and acetaldehyde (2.52 ± 0.50 |ig/m3). These are the only
pollutants of interest with annual average concentrations greater than 1 |ig/m3. Of the
VOCs, benzene (0.79 ± 0.09 |ig/m3) and carbon tetrachloride (0.60 ± 0.02 |ig/m3)
have the highest annual average concentrations for SPIL.
• Concentrations of formaldehyde measured at SPIL appear highest during the first
quarter of 2014, based on the quarterly average concentrations. Five of the seven
formaldehyde concentrations greater than or equal to 5 |ig/m3 were measured at SPIL
between January and March, while none of the 10 formaldehyde concentrations less
than 2 |ig/m3 were measured during the first quarter.
• Concentrations of acetaldehyde measured at SPIL span an order of magnitude,
ranging from 0.849 |ig/m3 to 8.99 |ig/m3. The maximum acetaldehyde concentration
measured at SPIL is the third highest acetaldehyde concentration measured across the
program. SPIL has the highest number of acetaldehyde concentrations greater than
5 |ig/m3 (eight) of any other NMP site (NBIL has four). The first and fourth quarter
average concentrations for SPIL are considerably higher than the other quarterly
averages, and the confidence intervals indicate that there is considerable variability
associated with these averages. All 12 acetaldehyde concentrations greater than
4 |ig/m3 were measured at SPIL during the first (seven) or fourth (five) quarters of
10-22
-------
2014. All but one of the acetaldehyde concentrations measured during the first quarter
are greater than the averages calculated for the second and third quarters of the year.
• Second quarter average concentrations for the VOCs in Table 10-4 could not be
calculated for SPIL because there were too many invalid samples during this quarter
to meet the 75 percent criteria.
• SPIL has a significantly higher annual average concentration of 1,3-butadiene than
NBIL. The maximum 1,3-butadiene concentration for each Chicago site was
measured on the same day, January 23, 2014 (0.308 |ig/m3 for NBIL and 0.288 |ig/m3
for SPIL). After these two measurements, concentrations measured at SPIL account
for the next 23 highest 1,3-butadiene concentrations measured at these two sites. Of
the 37 1,3-butadiene concentrations greater than 0.1 |ig/m3 measured at these two
sites, 35 were measured at SPIL. Benzene concentrations were also higher than SPIL
than NBIL. All but one of the 13 benzene concentrations greater than 1 |ig/m3
measured at these two sites were measured at SPIL.
• The first and fourth quarter average concentrations of trichloroethylene have
confidence intervals greater than or equal to the averages themselves. The second
quarter average concentration also has a relatively large confidence interval
associated with it. A review of the data shows that trichloroethylene was detected in
89 percent of the samples collected at SPIL, with measured detections ranging from
0.0485 |ig/m3 to 3.09 |ig/m3. Seven of the nine trichloroethylene concentrations
greater than 1 |ig/m3 measured across the program were measured at SPIL; further, 22
of the 25 highest trichloroethylene concentrations (those greater than 0.25 |ig/m3)
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, and 2013
NMP reports.
Observations for ROIL from Table 10-4 include the following:
• The pollutants with the highest annual average concentrations are formaldehyde
(3.05 ± 0.45 |ig/m3), acetaldehyde (1.91 ± 0.18 |ig/m3), and benzene
(1.22 ± 0.21 |ig/m3). These are the only pollutants of interest with annual average
concentrations greater than 1 |ig/m3. ROIL's annual average concentration of benzene
is significantly higher than the annual average concentrations of benzene for the
Chicago sites.
• The second and third quarter average concentrations for formaldehyde are
significantly higher than the first and fourth quarter averages. A review of the data
shows that formaldehyde concentrations measured at ROIL range from 1.11 |ig/m3 to
10.5 |ig/m3, ROIL is one of seven NMP sites at which formaldehyde concentrations
greater than 10 |ig/m3 were measured. All but one of the 24 formaldehyde
concentrations greater than 3 |ig/m3 measured at ROIL were measured between April
and September. At the other end of the concentration range, all but one of the 18
formaldehyde concentrations less than 2 |ig/m3 were measured during the first and
fourth quarters of 2014.
10-23
-------
• Concentrations of acetaldehyde also appear higher during the second and third quarter
of the year, although the difference among the quarterly averages is considerably less.
While acetaldehyde concentrations greater than 2 |ig/m3 were measured during each
calendar quarter, the majority of them (17 of 23) were measured between April and
September. Conversely, nine of the 10 lowest acetaldehyde concentrations were
measured at ROIL during the first or fourth quarters, primarily in January and
December.
• The confidence interval for the first quarter average concentration of benzene is two
to three times larger than the confidence intervals shown for the other quarterly
averages shown for ROIL. Benzene concentrations measured at ROIL range from
0.301 |ig/m3 to 5.25 |ig/m3, which is the third highest benzene concentration
measured at an NMP site in 2014. The two highest benzene concentrations measured
at ROIL were both from samples collected during the first quarter (5.25 |ig/m3 on
February 28, 2014 and 3.13 |ig/m3 on March 18, 2014). ROIL is one of only five
NMP sites at which benzene concentrations greater than 3 |ig/m3 were measured.
• Ethylbenzene is the only pollutant of interest for ROIL that is not a pollutant of
interest for each of the Chicago sites. The confidence interval for the first quarter
average concentration of ethylbenzene is greater than the quarterly average itself,
indicating the likely influence of outliers. Ethylbenzene concentrations measured at
ROIL range from 0.065 |ig/m3 to 3.01 |ig/m3. This maximum ethylbenzene
concentration was measured at ROIL on the same day as the maximum benzene
concentration, and is the second highest ethylbenzene concentration measured across
the program in 2014. All other ethylbenzene concentrations measured at ROIL in
2014 are less than 0.85 |ig/m3 and all other ethylbenzene concentrations measured
during the first quarter at this site are less than 0.5 |ig/m3, explaining the large
confidence interval associated with the first quarter average concentration.
Tables 4-9 through 4-12 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for NBIL, SPIL,
and ROIL from those tables include the following:
• The Illinois monitoring sites appear in Tables 4-9 through 4-12 a total of eight times,
with NBIL appearing three times, SPIL appearing three times, and ROIL appearing
twice.
• Table 4-9 shows that ROIL has the second highest annual average concentration of
benzene among NMP sites sampling this pollutant, with neither Chicago site
appearing in this table for benzene. ROIL also appears for ethylbenzene, ranking
tenth. SPIL's annual average concentration of 1,3-butadiene ranks sixth and NBIL's
annual average concentration of hexachloro-l,3-butadiene ranks ninth among NMP
sites sampling these pollutants.
• SPIL and NBIL both appear in Table 4-10 for their annual average concentrations of
acetaldehyde, ranking fourth and seventh, respectively (ROIL's annual average is just
outside the top 10, ranking 11th.) Note that the annual averages for SPIL and NBIL
10-24
-------
have the largest confidence intervals among the sites shown. SPIL ranks tenth for its
annual average concentration of formaldehyde, with ROIL again just outside the top
10. In contrast to the other sites, the annual average concentration of formaldehyde
for NBIL is one of the lowest among NMP sites sampling carbonyl compounds.
• NBIL ranks second for its annual average concentration of naphthalene among NMP
sites sampling PAHs, as shown in Table 4-11. The confidence interval associated
with NBIL's annual average is the largest among the averages shown, a reflection of
the variability within the measurements.
10.4.2 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 10-4 for NBIL, SPIL, and ROIL. Figures 10-9 through 10-21 overlay the 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 3.4.3.1,
and are discussed below. Because an annual average concentration could not be calculated for
arsenic, a box plot is not presented for NBIL.
Figure 10-9. Program vs. Site-Specific Average Acenaphthene Concentration
Program Max Concentration = 198 ng/m3
NBIL Max Concentration = 198 ng/m3
0 20 40 60 80 100
Concentration {ng/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 10-9 presents the box plot for acenaphthene for NBIL and shows the following:
• NBIL is the only Illinois site to sample PAHs under the NMP is 2014. The program-
level maximum concentration (198 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 to 100 ng/m3.
10-25
-------
• The maximum acenaphthene concentration measured at NBIL is the maximum
concentration measured across the program, although the next highest concentration
measured at NBIL is roughly one-third the magnitude (63.6 ng/m3). Only NBIL and
one other NMP site have individual acenaphthene concentrations greater than
40 ng/m3.
• NBIL's annual average acenaphthene concentration is more than four times the
program-level average concentration, with more than half of NBIL's acenaphthene
measurements greater than the program-level average concentration. 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 10-10. Program vs. Site-Specific Average Acetaldehyde Concentrations
NBIL
o
kJ
SPIL
1
0123456789 10
Concentration (ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 10-10 presents the box plots for acetaldehyde for all three sites and shows the
following:
• The range of acetaldehyde concentrations measured is largest for NBIL and smallest
for ROIL. The program-level maximum concentration of acetaldehyde was measured
at NBIL, although a similar concentration was also measured at SPIL.
• The annual average concentration of acetaldehyde for each Illinois monitoring site is
greater than the program-level average concentration, with the annual averages for
NBIL and SPIL also greater than the program-level third quartile
10-26
-------
• The minimum acetaldehyde concentration measured at NBIL is among the lowest
measured across the program.
Figure 10-11. Program vs. Site-Specific Average Benzene Concentrations
4
Program Max Concentration = 12.4 ng/m3
SPIL
n
Program Max Concentration = 12.4 ng/m3
I
-O-
Program Max Concentration = 12.4 |ig/m3
4 6
Concentration {[j.g/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 10-11 presents the box plots for benzene for all three sites and shows the
following:
• Similar to the box plot for acenaphthene, the program-level maximum benzene
concentration (12.4 |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 range of concentrations measured at these sites is smallest for NBIL and largest
for ROIL. Benzene concentrations greater than 2 |ig/m3 were not measured at NBIL
or SPIL while several were measured at ROIL.
• NBIL's annual average benzene concentration is less than the program-level median
concentration and is the fourth lowest among NMP sites sampling this pollutant.
SPIL's annual average benzene concentration is just greater than the program-level
average concentration while ROIL's annual average is greater than the program-level
average concentration and third quartile. Among NMP sites sampling benzene,
ROIL's annual average concentration ranks second.
10-27
-------
Figure 10-12. Program vs. Site-Specific Average 1,3-Butadiene Concentrations
Program Max Concentration = 5.90 ng/m3
¦-+
Program Max Concentration = 5.90 |ig/m3
1
O
Program Max Concentration = 5.90 ng/m3
,
°
0.4 0.6
Concentration {[j.g/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 10-12 presents the box plots for 1,3-butadiene for all three sites and shows the
following:
• The program-level maximum 1,3-butadiene concentration (5.90 |ig/m3) is not shown
directly on the box plots as the scale has been reduced to 1.0 |ig/m3 to allow for the
observation of data points at the lower end of the concentration range.
• The range of 1,3-butadiene concentrations is largest for NBIL and smallest for ROIL.
While non-detects of this pollutant were measured at NBIL and ROIL, the minimum
concentration measured at SPIL is greater than the program-level first quartile.
• The annual average concentration of 1,3-butadiene for NBIL is similar to the
program-level first quartile; ROIL's annual average concentration just greater than
the program-level median concentration; and SPIL's annual average is greater than
the program-level average concentration and third quartile.
10-28
-------
Figure 10-13. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations
NBIL
Program Max Concentration = 3.06 |-ig/m3
SPIL
Program Max Concentration = 3.06 \±g/m3
Program Max Concentration = 3.06 \±g/m3
0 0.5 1 1.5 2 2.5
Concentration (ng/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 10-13 presents the box plots for carbon tetrachloride for all three sites and shows
the following:
• The scale of these box plots have also been reduced to allow for the observation of
data points at the lower end of the concentration range.
• The maximum carbon tetrachloride concentrations measured at the Illinois sites are
relatively similar and considerably less than the program-level maximum
concentration. The minimum concentrations measured at ROIL and SPIL are fairly
similar to each other while the minimum concentration measured at NBIL is an order
of magnitude less (and the second lowest concentration of this pollutant across the
program).
• The annual average carbon tetrachloride concentrations for these three sites fall
between the program-level median and average concentrations, with the annual
average for ROIL just slightly greater than the annual averages for NBIL and SPIL.
10-29
-------
Figure 10-14. Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations
¦
Program Max Concentration = 27.4 ng/m3
¦
Program Max Concentration = 27.4 ng/m3
Program Max Concentration = 27.4 |ig/m3
0.4 0.6
Concentration {[j.g/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 10-14 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 (27.4 |ig/m3) is
considerably greater than the majority of measurements. This is another example of
measurements at the upper end of the concentration range driving the program-level
average concentration, as the program-level average is three times the program-level
third quartile.
• All of the concentrations of 1,2-dichloroethane measured at the Illinois sites are less
than the program-level average concentration of 0.31 |ig/m3.
• The annual average concentrations for SPIL and ROIL are similar to each other, and
both are just less than the program-level median concentration, with NBIL's annual
average concentration similar to the program-level first quartile.
10-30
-------
Figure 10-15. Program vs. Site-Specific Average Ethylbenzene Concentration
V
0 0.5 1 1.5 2 2.5 3 3.5
Concentration (ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 10-15 presents the box plot for ethylbenzene for ROIL and shows the following:
• ROIL is the only Illinois site for which ethylbenzene is a pollutant of interest.
• The second highest ethylbenzene concentration measured across the program was
measured at ROIL (3.01 |ig/m3). However, the next highest concentration measured
at ROIL is considerably less (0.835 |ig/m3) and only one other ethylbenzene
concentration is greater than 0.5 |ig/m3.
• The annual average concentration of ethylbenzene for ROIL is greater than the
program-level average concentration and just greater than the program-level third
quartile.
Figure 10-16. Program vs. Site-Specific Average Fluoranthene Concentration
O i
,
VwJ 1
0 5 10 15 20 25 30 35 40
Concentration (ng/m3)
Program: 1st Quartile
2nd Quartile 3rd Quartile
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 10-16 presents the box plot for fluoranthene for NBIL and shows the following:
• NBIL is the only Illinois site to sample PAHs under the NMP is 2014.
• The maximum fluoranthene concentration measured at NBIL is the maximum
concentration measured across the program, with the six highest fluoranthene
concentrations across the program measured at NBIL.
10-31
-------
• NBIL's annual average fluoranthene concentration is more than three times the
program-level average concentration. More than half of NBIL's fluoranthene
measurements are greater than the program-level average concentration.
Figure 10-17. Program vs. Site-Specific Average Fluorene Concentration
Program Max Concentration = 161 ng/m3
o
kJ
NBIL Max Concentration = 161 ng/m3
0 20 40 60 80 100
Concentration {ng/m3)
Program: 1st Quartile
2nd Quartile 3rd Quartile
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 10-17 presents the box plot for flourene for NBIL and shows the following:
• Fluorene is another PAH pollutant of interest for NBIL. The program-level maximum
concentration (161 ng/m3) of fluorene 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 to 100 ng/m3. Note
that the program-level first quartile is zero and thus, not visible on the box plot.
• The maximum fluorene concentration measured at NBIL is the maximum fluorene
concentration measured across the program. Only NBIL and one other NMP site
(ROCH) have individual fluorene concentrations greater than 100 ng/m3. All other
fluorene concentrations measured at NBIL are less than 50 ng/m3.
• NBIL's annual average fluorene concentration is nearly four times the program-level
average concentration; more than half of NBIL's fluorene measurements greater than
the program-level average concentration.
10-32
-------
Figure 10-18. Program vs. Site-Specific Average Formaldehyde Concentrations
¦
o ¦
¦
0
0 3 6 9 12 15 18 21 24 27
Concentration (ng/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 10-18 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 ROIL.
• The annual average concentrations of formaldehyde for SPIL and ROIL are similar to
each other and more than twice the annual average concentration for NBIL. The
annual average concentrations of formaldehyde for SPIL and ROIL are greater than
the program-level average concentration while NBIL's annual average is just less
than the program-level first quartile.
• The minimum concentration measured at NBIL is among the lowest measured across
the program and is an order of magnitude less than the minimum concentrations for
the other two Illinois sites.
10-33
-------
Figure 10-19. Program vs. Site-Specific Average Hexachloro-l,3-butadiene Concentrations
1
0
0.1
0.2
0.3 0.4
Concentration (ng/m3)
0.5
0.6
0.7
Program:
1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site:
Site Average
o
Site Concentration Range
Figure 10-19 presents the box plots for hexachloro-1,3-butadiene for all three sites and
shows the following:
• The program-level first, second (median), and third quartiles are all zero and therefore
not visible on the box plot. This is due to the large number of non-detects of this
pollutant across the program (77 percent).
• Between 55 and 60 valid VOC samples were collected at each of the Illinois sites; of
these, fewer than 20 measured detections were measured at each site. Thus, many
zeroes are substituted into the annual average concentrations for this pollutant. The
range of concentrations measured and the annual average concentrations are fairly
similar across the Illinois monitoring sites.
10-34
-------
Figure 10-20. Program vs. Site-Specific Average Naphthalene Concentration
-
LJ 1
0 100 200 300 400 500 600
Concentration {ng/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 10-20 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. NBIL is the
only NMP site at which a naphthalene concentration greater than 400 ng/m3 was
measured.
• NBIL's annual average naphthalene concentration is greater than the program-level
average concentration and program-level third quartile. Recall from the previous
section that NBIL's annual average is the second highest annual average
concentration of naphthalene among NMP sites sampling PAHs in 2014.
Figure 10-21. Program vs. Site-Specific Average Trichloroethylene Concentration
rs ,
kJ 1
0 0.5 1 1.5 2 2.5 3 3.5
Concentration (ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 10-21 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 all zero due to the large
number of non-detects; thus, only the fourth quartile is visible in Figure 10-21.
10-35
-------
• Although the maximum concentration of trichloroethylene across the program was
not measured at SPIL, a concentration of similar magnitude was measured at SPIL.
Among NMP sites sampling this pollutant, SPIL has the greatest number of measured
detections (49), with the next closest site at 22 (CSNJ). Concentrations measured at
SPIL account for nearly half (33) of the trichloroethylene measurements greater than
0.1 |ig/m3 measured across the program (75).
• The annual average concentration for SPIL (0.44 |ig/m3) is considerably higher than
the program-level average concentration (0.032 |ig/m3).
10.4.3 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 3.4.3.2.
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 10-22 through 10-41 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, a trends analysis
was not performed.
10-36
-------
Figure 10-22. Yearly Statistical Metrics for Acenaphthene Concentrations Measured at NBIL
2011
Year
5th Percentile
— Minimum
— Maximum
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 10-22 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 two additional acenaphthene concentrations greater than 100 ng/m3
measured at NBIL in 2013. Five of the seven acenaphthene concentrations greater
than 75 ng/m3 were measured at NBIL in 2013.
• 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. Recall, however, that 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
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
10-37
-------
percentile, and maximum concentration. Five acenaphthene concentrations measured
in 2013 are greater than the maximum concentrations measured in 2012. Also, 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. Yet the median concentration exhibits little change, indicating that the
changes shown are due to the measurements at the upper end of the concentration
range, as mentioned in previous bullets.
Figure 10-23. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at NBIL
X
LoJ
—c—
o
T
2005 A 2006 2007
2013 2014
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 2005.
Observations from Figure 10-23 for acetaldehyde concentrations measured at NBIL
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 40 acetaldehyde concentrations greater than 3 |ig/m3 measured at NBIL
10-38
-------
were measured after 2009 (one in 2010, three in 2011, six in 2012, and 15 were
measured in 2013 and again in 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). After 2009,
acetaldehyde concentrations measured at NBIL increase significantly as nearly all of
the statistical metrics exhibit an increase from 2009 to 2010 and again for each year
afterward. The 1-year average concentration for 2013 is greater than the maximum
concentrations measured for several of the early years of sampling and the 5th
percentile for 2013 is greater than the 1-year average concentrations for each of the
earlier years of sampling. 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.
Figure 10-24. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at NBIL
2005 2006 2007 2008 2009 2010 2011 2012 2013 20141
O 5th Percentile — Minimum — Median — Maximum O 95th Percentile Average
1 A 1-year average is not presented because the criteria for an annual average was not met in 2014.
10-39
-------
Observations from Figure 10-24 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 12, 2009,
although a similar concentration was also measured in 2010. Only four concentrations
greater than 3 ng/m3 have been measured at NBIL (one in 2006, one in 2009, 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 the course of sampling. The 1-year average
concentration increased from 2005 to 2006, changed little for 2007, decreased slightly
for 2008, after which the 1-year average concentration remained steady through 2012,
hovering 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 first year of sampling.
• Slight increases in the statistical parameters presented are shown for 2014, although a
1-year average concentration is not provided because a number of metals samples
were invalid during the first and second quarters of 2014, and the criteria for an
annual average to be calculated was not met.
10-40
-------
Figure 10-25. Yearly Statistical Metrics for Benzene Concentrations Measured at NBIL
o
-r
rh
LjJ
LjJ
2003 2004 2005
-f-
2008
-r
r~
o-
I
2013
2014
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 2003.
2 A 1-year average is not presented because there was a gap in sampling from late October 2004 until late
December 2004.
Observations from Figure 10-25 for benzene concentrations measured at NBIL include
the following:
• Although sampling for VOCs at NBIL began in 2003, sampling under the NMP did
not begin until April; 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. 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, as the
1-year average concentration decreased significantly from 2005 to 2006, with slight
decreases for 2007, 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. This decreasing continued into 2012 (although the
median concentration actually increased slightly) and 2013. Several of the statistical
10-41
-------
parameters are at a minimum for 2013, which is the first year the 1-year average
concentration is less than 0.5 |ig/m3,
• Although a few additional higher benzene concentrations were measured in 2014
compared to 2013, relatively little change in the statistical parameters is shown for
2014, with the 1-year average concentration still less than 0.5 |ig/m3.
Figure 10-26. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at NBIL
i Maximum
| Concentration for
; 2011 is 2.68 ng/m3
-
i L
o
X I ^ rL [
-0- Q .2. -
<
>¦ ....
X
jl.
B- c
3
2003 1 2004 2 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
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 until late
December 2004.
Observations from Figure 10-26 for 1,3-butadiene concentrations measured at NBIL
include the following:
• The maximum 1,3-butadiene concentration was measured 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
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, the minimum and 5th percentile are zero, indicating the
presence of non-detects (at least 5 percent of the measurements). 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
7 percent (2007), although the percentage of non-detects has been increasing slightly
10-42
-------
each year since, reaching 38 percent of measurements for 2013. This percentage
decreased somewhat for 2014 (23 percent).
The 1-year average concentration decreased slightly between 2005 and 2009,
although the changes are not significant. From 2009 to 2010, the 1-year average
doubled, and then nearly doubled again for 2011. However, there is a significant
amount of variability associated with these measurements, based on the confidence
intervals. Even with the relatively high concentrations measured in 2010 and 2011,
the 95th percentile changed only slightly, indicating that the majority of the
measurements were within the same 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, is at a minimum for 2013, as is the 1-year average
concentration, which decreased significantly from 2012 to 2013.
Slight increases are shown for most of the statistical parameters from 2013 to 2014,
although the majority of concentrations fell within the second smallest range since
2005, the first year with a full year's worth of samples.
10-43
-------
Figure 10-27. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at NBIL
X
Maximum
Concentration for
2014 is 4.98 ng/m3
X
r
1
-0
JL
¦O"
~o
r
_£L
1
1
1
rh
o-
T
w
20031 20042 2005
2008 2009 2010 2011 2012 2013 2014
Year
o 5th Percentile
— Minimum
— Maximum
o 95th Percentile
..4.... 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 there was a gap in sampling from late October 2004 until late
December 2004.
Observations from Figure 10-27 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.
• After decreasing slightly between 2005 and 2007, the 1-year average concentration
increased significantly for 2008. The 1-year average concentration exhibits a
significant decreasing trend after 2008 that continued through 2011, when the 1-year
average returned to 2007 levels. After exhibiting an increase for 2012, the 1-year
average concentration is at a minimum for 2013 (0.60 |ig/m3), with a negligible
change for 2014. The 1-year average concentrations presented range from 0.60 |ig/m3
10-44
-------
(2013 and 2014) to 0.83 |ig/m3 (2008). The median concentration exhibits a similar
pattern.
Figure 10-28. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations
Measured at NBIL
^ 0.10
I
2003 2004 2005 2006
o
O ^
2008 2009
Year
o
2012 2013 2014
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 2003.
2 A 1-year average is not presented because there was a gap in sampling from late October 2004 until late
December 2004.
Observations from Figure 10-28 for 1,2-dichloroethane concentrations measured at NBIL
include the following:
• There were no measured detections of 1,2-dichloroethane in 2003, 2004, or 2008. The
number 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 is 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 (10).
10-45
-------
• The 5th percentile is greater than zero for the first time for 2014, when only two non-
detects of 1,2-dichloroethane were measured at NBIL.
Figure 10-29. Yearly Statistical Metrics for Fluoranthene Concentrations Measured at NBIL
o
2011
Year
O 5th Percentile - Minimum - Med en - 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 10-29 for fluoranthene concentrations measured at NBIL
include the following:
• The maximum fluoranthene concentration was measured at NBIL on July 21, 2013
(43.7 ng/m3), although a similar concentration was also measured earlier in July as
well as in 2012, also in July. The maximum fluoranthene concentration measured
each year from 2011 forward were all measured during the first week of July.
• 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. Recall,
however, that 2008 does not include a full year's worth of sampling. The median
fluoranthene concentrations shown after 2009 vary little.
• The 1-year average concentration of fluoranthene increases between 2009 and 2011,
decreases slightly for 2012, then increases slightly for 2013 and 2014. Both the 1-year
average and median concentrations are at a maximum for 2014.However, confidence
intervals calculated for these averages indicate that the changes are not statistically
significant due to the relatively large amount of variability in the measurements.
10-46
-------
Figure 10-30. Yearly Statistical Metrics for Fluorene Concentrations Measured at NBIL
2011
Year
5th Percentile
— Minimum
— Maximum
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 10-30 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 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. Recall, however, that
2008 does not include a full year's worth of sampling. A steady increase in the
median concentration is shown after 2009.
• 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 a slight decrease is exhibited for
2014. Confidence intervals calculated for these averages indicate that there is a
relatively large amount of variability in these measurements. The range of fluorene
measurements spans two orders of magnitude for each year. For example, the
minimum and maximum concentrations for 2012 are 0.93 ng/m3 and 93.4 ng/m3,
respectively.
• The maximum concentration of fluorene measured at NBIL has increased each year
since 2009, more than doubling in magnitude, and exceeding 100 ng/m3 in 2014.
10-47
-------
Figure 10-31. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at NBIL
E
1 30
i Maximum
1 Concentration for «
| 2006 is 91.7 ng/m3 [
3 ^
¦3 i--""
0
k
-T y t i
2007 2008
2009 2010
Year
2011 2012
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 2005.
Observations from Figure 10-31 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 is greater than the 95th percentile for
2006.
• The statistical metrics for 2010 are also affected by the higher concentrations;
however, concentrations measured this year are higher overall, as indicated by seven-
fold increase in the 95th percentile. Although difficult to discern in Figure 10-31, 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 greater than most years.
• Although the maximum concentration measured in 2012 is less than the 95th
percentile for 2011, the 1-year average concentration did not change significantly for
2012 and the median concentration increased. This is because the number of
concentrations in the middle of the concentration range increased. The number of
10-48
-------
measurements between 2 |ig/m3 and 4 |ig/m3 nearly doubled from 2011 (15) to 2012
(29).
• The range of formaldehyde concentrations measured at NBIL after 2010 has a
decreasing trend and, for 2014, is at its smallest in five years. There is a significant
decrease in the 1-year average concentrations between 2010 and 2014, although the
1-year averages prior to 2010 are still lower.
Figure 10-32. Yearly Statistical Metrics for Hexachloro-l,3-butadiene Concentrations
Measured at NBIL
o
u
.Q..
i
O 5th Percentile
- Minimum
- 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 there was a gap in sampling from late October 2004 until late
December 2004.
Observations from Figure 10-32 for hexachloro-l,3-butadiene concentrations measured at
NBIL include the following:
• There were no measured detections of hexachloro-1,3-butadiene measured at NBIL
during the first 2 years of sampling. Non-detects made up 85 percent of
measurements in 2005, and between 2006 and 2013, the percentage of non-detects
was greater than 90 percent each year, including 2010, when again no measured
detections were measured.
• The number of non-detects fell to 66 percent for 2014. The number of measured
detections for 2014 is at 19, which is nearly greater than the number of measured
detections across the previous years of sampling combined (22). This explains the
increase in the 1-year average concentration shown for 2014.
10-49
-------
• The maximum hexachloro-l,3-butadiene concentration measured at NBIL was
measured on July 5, 2008 (0.299 |ig/m3) and is the only measurement of this pollutant
greater than 0.25 |ig/m3. Only 41 total measured detections have been measured at
NBIL since the onset of sampling. The effect of the non-detects (zeros) factored into
the statistical calculations can be seen in the scale of the trends graph and by noting
that none of the 1-year average concentrations shown are greater than 0.025 |ig/m3.
Figure 10-33. Yearly Statistical Metrics for Naphthalene Concentrations Measured at NBIL
2011
Year
O
2012
-r
2013
— Minimum
— Maximum
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 10-33 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 concentration of naphthalene increases from 2009 to 2010 then
decreases slightly for 2011 and 2012. The 1-year average concentration then doubles
for 2013, after which a slight decrease is exhibited for 2014. The median
concentration exhibits a similar pattern. The central tendency parameters for
naphthalene exhibit a similar pattern of changes as those shown on the trends graphs
for the other PAH pollutants of interest for NBIL.
10-50
-------
Figure 10-34. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SPIL
O 5th Percentile
O 95th Percentile
¦~•¦•Average
1 A 1-year average is not presented because consistent sampling did not begin until March 2005.
Observations from Figure 10-34 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-one of the 33 concentrations of
acetaldehyde greater than 5 |ig/m3 were measured after 2010 (eight in 2011, eight in
2012, seven in 2013, and eight in 2014), with the other two 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 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 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 profiles of acetaldehyde concentrations measured at SPIL in 2012, 2013, and
2014 are more similar to 2011 than other years of sampling. Yet, these measurements
10-51
-------
reflect considerable variability, based on the range of concentrations measured and
spread of the central tendency statistics.
Figure 10-35. Yearly Statistical Metrics for Benzene Concentrations Measured at SPIL
©
r r
1
_0_
t
o
L-J-.
O
t
©
T
liOn
f
2005 2006 2007
2008 2009
Year
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 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.
• The only two concentrations of benzene greater than 5 |ig/m3 were both measured in
2005.
• The 1-year average benzene concentration has a significant decreasing trend over the
years between 2004 and 2009. The 1-year average concentration increased
significantly from 2009 to 2010, after which the 1-year average benzene
concentration has an undulating pattern, varying between 0.74 |ig/m3 (2013) and
0.95 |ig/m3 (2012). The median concentration has a similar pattern through 2010 but
has a steady decreasing in the years that follow.
• 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
10-52
-------
exception of 2013, when the range of benzene concentrations is slightly smaller than
other recent years.
Figure 10-36. Yearly Statistical Metrics for 1,3-Butadiene Concentrations
Measured at SPIL
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 10-36 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, only seven 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 and 2013.
The detection rate for 1,3-butadiene has increased over time, increasing from
approximately 55 percent measured detections in 2003 and 2004 to a 100 percent
detection rate in 2008 and 2009. A single non-detect has been measured in each of the
following years after 2009.
The 1-year average concentrations of 1,3-butadiene changed little between 2004 and
2006, then decreased 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. A slight decreasing trend in the 1-year average concentration is
shown after 2011. Despite these changes, most of the 1-year average concentrations
10-53
-------
shown fall between 0.10 |ig/m3 and 0.15 |ig/m3, with only the minimum (0.08 |ig/m3
for 2009) and maximum (0.16 |ig/m3 for 2011) falling outside this range.
• The 5th and 95th percentiles indicate the range within which the majority of
concentrations fall. This range decreased considerably between 2004 and 2009,
increased for 2010 and 2011, then began to decrease again. The difference between
these two parameters is at a minimum for 2013, with just a slight increase shown for
2014.
Figure 10-37. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at SPIL
I
2003 1 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
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 2003.
Observations from Figure 10-37 for carbon tetrachloride concentrations measured at
SPIL include the following:
• The maximum concentration of carbon tetrachloride was measured three times, once
in 2005 and twice in 2008 (1.20 |ig/m3).
• 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 1-year average concentration exhibits a decreasing trend
after 2008 that continued through 2011, when the 1-year average is at a minimum
10-54
-------
(0.58 |ig/m3). The increase shown for 2012 brings the 1-year average carbon
tetrachloride concentration back to near 2010 levels.
• With the exception of the 5th percentile, most of the statistical parameters exhibit a
decrease for 2013 and all of them exhibit decreases for 2014.
Figure 10-38. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations
Measured at SPIL
Maximum
Concentration for
2003 is 0.75 Mg/m3
0.00 0
2003 1 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
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.
Observations from Figure 10-38 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.
For 2012, the percentage of non-detects decreased to 8 percent of samples collected
and was at a minimum of 5 percent for 2013, which is the first year that the 5th
percentile is greater than zero. The percentage of non-detects for 2014 is also
5 percent.
• 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
10-55
-------
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 on
December 25, 2014).
• 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 the 1-year average concentration shown for 2012. The 1-year
average concentrations vary by less than 0.01 |ig/m3 between 2012 and 2014.
Figure 10-39. Yearly Statistical Metrics for Formaldehyde Concentrations
Measured at SPIL
Maximum
Concentrati on for
2006 is 162 |Jg/m3
o..
15
I J
L J
,
y \i
2007 20
<3 i 6
08 2009 20
40 E
10 2011 20
L -i
W
12 20
•m K^i
L o
13 2014
±
i i
2007 2008 2009 2010 2011 2012 2013 2014
Year
O 5th Percentile — Minimum
— 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 10-39 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 any of the
other years shown in Figure 10-39 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.
• The 1-year average concentration for 2006 is 13.76 |ig/m3. After 2006, the 1-year
average concentration decreased each year, reaching a minimum of 1.85 |ig/m3 for
10-56
-------
2009. There is an increasing trend in the 1-year average concentration between 2009
and 2011, after which little change is shown. Between 2011 and 2014, the 1-year
average concentrations varied between 3.07 |ig/m3 (2012) and 3.31 |ig/m3 (2013).
Figure 10-40. Yearly Statistical Metrics for Hexachloro-l,3-butadiene Concentrations
Measured at SPIL
o
I
Maximum
Concentrati on for
2011 is 0.684 ng/m3
-6 ¦¦¦&¦¦¦¦
2008 2009
I
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 10-40 for hexachloro-l,3-butadiene concentrations measured at
SPIL include the following:
• The trends graph for hexachloro-1,3-butadiene measurements resembles the trends
graph for 1,2-dichloroethane in that the statistical parameters reflect that non-detects
make up the majority of measurements of this pollutant.
• There were no measured detections of hexachloro-1,3-butadiene measured at SPIL
during the first 2 years of sampling. Non-detects made up 83 percent of
measurements in 2005 and 93 percent in 2006. Between 2007 and 2010, the
percentage of non-detects was constant at 98 percent, with only a single measured
detection for each year. After 2010, the percentage of non-detects began to fall
slightly each year, reaching a minimum of 78 percent for 2014.
• The maximum hexachloro-l,3-butadiene concentration measured at SPIL was
measured on December 11, 2011 (0.684 |ig/m3) and is the only measurement of this
pollutant greater than 0.25 |ig/m3. Only 51 total measured detections have been
measured at SPIL since the onset of sampling. The effect of the non-detects (zeros)
10-57
-------
factored into the statistical calculations can be seen in the scale of the trends graph
and by noting that none of the 1-year average concentrations shown are greater than
0.025 |ig/m3.
Figure 10-41. Yearly Statistical Metrics for Trichloroethylene Concentrations
Measured at SPIL
Maximum
Concentration for
2003 is 110 m*/m3
Q
"3"
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year
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 2003.
Observations from Figure 10-41 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.
• 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.
10-58
-------
• 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.
10.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at each Illinois monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4 for
definitions and explanations regarding the various toxicity factors, time frames, and calculations
associated with these risk-based screenings.
10.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Illinois 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 3.4.3.3 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 10-5, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Observations for the Illinois sites from Table 10-5 include the following:
• Formaldehyde and acetaldehyde are the pollutants with the highest annual average
concentrations for all three sites.
• Formaldehyde has the highest cancer risk approximation for all three sites, ranging
from 16.82 in-a-million for NBIL to 40.52 in-a-million for SPIL. There were no other
pollutants for which a cancer risk approximation greater than 10 in-a-million was
calculated, although ROIL's benzene is close (9.50 in-a-million).
• None of the pollutants of interest for NBIL, SPIL, or ROIL 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 among the pollutants of interest for the Illinois sites
is formaldehyde (0.32 for SPIL).
10-59
-------
Table 10-5. Risk Approximations for the Illinois Monitoring Sites
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Northbrook, Illinois - NBIL
Acetaldehyde
0.0000022
0.009
55/55
2.36
±0.45
5.18
0.26
Benzene
0.0000078
0.03
56/56
0.49
±0.06
3.86
0.02
1.3 -Butadiene
0.00003
0.002
43/56
0.04
±0.01
1.25
0.02
Carbon Tetrachloride
0.000006
0.1
56/56
0.60
±0.03
3.62
0.01
1,2-Dichloroethane
0.000026
2.4
54/56
0.07
±<0.01
1.79
<0.01
Formaldehyde
0.000013
0.0098
55/55
1.29
±0.21
16.82
0.13
Hexachloro-1,3 -butadiene
0.000022
0.09
19/56
0.02
±0.01
0.48
<0.01
Acenaphthene3
0.000088
55/55
20.62
±8.20
1.81
Arsenic (PMio)3
0.0043
0.000015
53/53
NA
NA
NA
Fluoranthene3
0.000088
55/55
8.12
±2.37
0.71
Fluorene3
0.000088
49/55
17.37
±6.69
1.53
Naphthalene3
0.000034
0.003
55/55
109.13
±28.14
3.71
0.04
Schiller Park, Illinois - SPIL
Acetaldehyde
0.0000022
0.009
59/59
2.52
±0.50
5.55
0.28
Benzene
0.0000078
0.03
55/55
0.79
±0.09
6.20
0.03
1.3 -Butadiene
0.00003
0.002
55/55
0.13
±0.02
3.93
0.07
Carbon Tetrachloride
0.000006
0.1
55/55
0.60
±0.02
3.60
0.01
1,2-Dichloroethane
0.000026
2.4
52/55
0.08
±0.01
2.06
<0.01
Formaldehyde
0.000013
0.0098
59/59
3.12
±0.35
40.52
0.32
Hexachloro-1,3 -butadiene
0.000022
0.09
13/55
0.02
±0.01
0.38
<0.01
T richloroethylene
0.0000048
0.002
49/55
0.44
±0.17
2.13
0.22
— = 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.
10-60
-------
Table 10-5. Risk Approximations for the Illinois Monitoring Sites (Continued)
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Roxana, Illinois - ROIL
Acetaldehyde
0.0000022
0.009
60/60
1.91
±0.18
4.20
0.21
Benzene
0.0000078
0.03
58/58
1.22
±0.21
9.50
0.04
1.3 -Butadiene
0.00003
0.002
55/58
0.07
±0.01
2.21
0.04
Carbon Tetrachloride
0.000006
0.1
58/58
0.63
±0.02
3.76
0.01
1,2-Dichloroethane
0.000026
2.4
51/58
0.08
±0.01
2.02
<0.01
Ethylbenzene
0.0000025
1
58/58
0.32
±0.10
0.81
<0.01
Formaldehyde
0.000013
0.0098
60/60
3.05
±0.45
39.63
0.31
Hexachloro-1,3 -butadiene
0.000022
0.09
17/58
0.02
±0.01
0.46
<0.01
— = 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.
10-61
-------
10.5.2 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 10-6 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
that have cancer toxicity factors. Table 10-6 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 10-6 provides the pollutants with the highest cancer risk approximations (in-a-million) for
each Illinois site, as presented in Table 10-5. The emissions, toxicity-weighted emissions, and
cancer risk approximations are shown in descending order in Table 10-6. Table 10-7 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 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 10.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
10-62
-------
Table 10-6. 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)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Northbrook, Illinois (Cook County) - NBIL
Benzene
1,391.32
Formaldehyde
1.48E-02
Formaldehyde
16.82
Formaldehyde
1,135.39
Benzene
1.09E-02
Acetaldehyde
5.18
Ethylbenzene
756.81
1,3-Butadiene
6.47E-03
Benzene
3.86
Acetaldehyde
623.34
Hexavalent Chromium
4.02E-03
Naphthalene
3.71
1.3 -Butadiene
215.66
Naphthalene
3.60E-03
Carbon Tetrachloride
3.62
T etrachloroethylene
187.87
Arsenic, PM
2.64E-03
Acenaphthene
1.81
Naphthalene
105.84
Ethylbenzene
1.89E-03
1,2-Dichloroethane
1.79
T richloroethylene
99.56
POM, Group 2b
1.81E-03
Fluorene
1.53
Dichloromethane
35.41
Acetaldehyde
1.37E-03
1,3-Butadiene
1.25
POM, Group 2b
20.53
POM, Group 2d
1.19E-03
Fluoranthene
0.71
Schiller Park, Illinois (Cook County) - SPIL
Benzene
1,391.32
Formaldehyde
1.48E-02
Formaldehyde
40.52
Formaldehyde
1,135.39
Benzene
1.09E-02
Benzene
6.20
Ethylbenzene
756.81
1,3-Butadiene
6.47E-03
Acetaldehyde
5.55
Acetaldehyde
623.34
Hexavalent Chromium
4.02E-03
1,3-Butadiene
3.93
1.3 -Butadiene
215.66
Naphthalene
3.60E-03
Carbon Tetrachloride
3.60
T etrachloroethylene
187.87
Arsenic, PM
2.64E-03
T richloroethylene
2.13
Naphthalene
105.84
Ethylbenzene
1.89E-03
1,2-Dichloroethane
2.06
T richloroethylene
99.56
POM, Group 2b
1.81E-03
Hexachloro-1,3 -butadiene
0.38
Dichloromethane
35.41
Acetaldehyde
1.37E-03
POM, Group 2b
20.53
POM, Group 2d
1.19E-03
-------
Table 10-6. 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)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Roxana, Illinois (Madison County) - ROIL
Formaldehyde
117.39
Coke Oven Emissions, PM
1.58E-02
Formaldehyde
39.63
Benzene
116.81
Formaldehyde
1.53E-03
Benzene
9.50
Ethylbenzene
56.77
Hexavalent Chromium
1.29E-03
Acetaldehyde
4.20
Acetaldehyde
50.30
Arsenic, PM
1.03E-03
Carbon Tetrachloride
3.76
Coke Oven Emissions, PM
15.95
Benzene
9.11E-04
1,3-Butadiene
2.21
Naphthalene
14.00
Naphthalene
4.76E-04
1,2-Dichloroethane
2.02
1.3 -Butadiene
12.69
1,3-Butadiene
3.81E-04
Ethylbenzene
0.81
Dichloromethane
12.11
Nickel, PM
3.20E-04
Hexachloro-1,3 -butadiene
0.46
T etrachloroethylene
3.60
POM, Group 5a
2.42E-04
POM, Group 2b
1.85
POM, Group 2b
1.63E-04
-------
Table 10-7. 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
Top 10 Noncancer Toxicity-Weighted Emissions
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(County-Level)
(County-Level)
(Site-Specific)
Noncancer
Noncancer
Hazard
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(HQ)
Northbrook, Illinois (Cook County) - NBIL
Toluene
10,461.60
Acrolein
4,043,182.36
Acetaldehyde
0.26
Xylenes
3,369.01
Formaldehyde
115,856.06
Formaldehyde
0.13
Methanol
3,041.83
1,3-Butadiene
107,829.46
Naphthalene
0.04
Hexane
2,784.74
Cyanide Compounds, gas
86,974.16
1,3-Butadiene
0.02
Benzene
1,391.32
Acetaldehyde
69,259.50
Benzene
0.02
Formaldehyde
1,135.39
T richloroethy lene
49,780.32
Carbon Tetrachloride
0.01
Ethylene glycol
1,052.17
Benzene
46,377.32
Hexachloro-1,3 -butadiene
<0.01
Ethylbenzene
756.81
Arsenic, PM
40,902.71
1,2-Dichloroethane
<0.01
Acetaldehyde
623.34
Naphthalene
35,279.80
Methyl isobutyl ketone
342.65
Xylenes
33,690.12
Schiller Park, Illinois (Cook County) - SPIL
Toluene
10,461.60
Acrolein
4,043,182.36
Formaldehyde
0.32
Xylenes
3,369.01
Formaldehyde
115,856.06
Acetaldehyde
0.28
Methanol
3,041.83
1,3-Butadiene
107,829.46
T richloroethy lene
0.22
Hexane
2,784.74
Cyanide Compounds, gas
86,974.16
1,3-Butadiene
0.07
Benzene
1,391.32
Acetaldehyde
69,259.50
Benzene
0.03
Formaldehyde
1,135.39
T richloroethy lene
49,780.32
Carbon Tetrachloride
0.01
Ethylene glycol
1,052.17
Benzene
46,377.32
Hexachloro-1,3 -butadiene
<0.01
Ethylbenzene
756.81
Arsenic, PM
40,902.71
1,2-Dichloroethane
<0.01
Acetaldehyde
623.34
Naphthalene
35,279.80
Methyl isobutyl ketone
342.65
Xylenes
33,690.12
-------
Table 10-7. 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
Top 10 Noncancer Toxicity-Weighted Emissions
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(County-Level)
(County-Level)
(Site-Specific)
Noncancer
Noncancer
Hazard
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(HQ)
Roxana, Illinois (Madison County) - ROIL
Toluene
635.77
Acrolein
274,415.42
Formaldehyde
0.31
Xylenes
208.66
Chlorine
95,420.68
Acetaldehyde
0.21
Hexane
195.81
Hexamethylene-l,6-diisocyanate, gas
25,000.00
Benzene
0.04
Methanol
178.11
Manganese, PM
16,632.19
1,3-Butadiene
0.04
Hydrochloric acid
128.20
Arsenic, PM
16,022.05
Carbon Tetrachloride
0.01
Formaldehyde
117.39
Lead, PM
14,477.27
Ethylbenzene
<0.01
Benzene
116.81
Formaldehyde
11,978.64
Hexachloro-1,3 -butadiene
<0.01
Ethylbenzene
56.77
Cyanide Compounds, gas
7,490.07
1,2-Dichloroethane
<0.01
Ethylene glycol
53.93
Nickel, PM
7,414.81
Acetaldehyde
50.30
Hydrochloric acid
6,410.24
-------
Observations from Table 10-6 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 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, benzene, and 1,3-butadiene. Coke
oven emissions top Madison County's toxicity-weighted emissions, followed by
formaldehyde and hexavalent chromium.
• Seven of the highest emitted pollutants in Cook County also have the highest toxicity-
weighted emissions while six of the highest emitted pollutants in Madison County
also have the highest toxicity-weighted emissions.
• For NBIL and SPIL, formaldehyde is the pollutant with the highest cancer risk
approximation. This pollutant also has the highest toxicity-weighted emissions and
ranks second for quantity emitted in Cook County. Benzene, acetaldehyde, and 1,3-
butadiene also appear on all three lists for both sites. For ROIL, formaldehyde is also
the pollutant with the highest cancer risk approximation. This pollutant also has the
highest emissions in Madison County and the second highest toxicity-weighted
emissions. Benzene and 1,3-butadiene also appear on all three lists for ROIL.
• Carbon tetrachloride, which has the fifth highest cancer risk approximation for NBIL
and SPIL and the fourth highest cancer risk approximation for ROIL, does not appear
on either county's emissions-based list. Similarly, 1,2-dichloroethane appears on
neither emissions-based list, though it ranks among the pollutants with the highest
cancer risk approximations for all three sites.
• Naphthalene has the fourth highest cancer risk approximation for NBIL. This
pollutant also has the fifth highest toxicity-weighted emissions for Cook County and
ranks seventh for quantity emitted. POM, Group 2b ranks 10th for quantity emitted
and eighth 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. Arsenic, which is a pollutant of interest for NBIL but for which an annual
average concentration could not be calculated, is also among those with the highest
toxicity-weighted emissions for Cook County.
• Trichloroethylene has the sixth highest cancer risk approximation for SPIL and 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 13th).
• Ethylbenzene is a pollutant of interest for ROIL and ranks seventh for its cancer risk
approximation. This pollutant is the third highest emitted pollutant in Madison
County but does not appear among those with the highest toxicity-weighted emissions
(it ranks 11th).
10-67
-------
Observations from Table 10-7 include the following:
• Toluene and xylenes are the highest emitted pollutants with noncancer RfCs in both
Cook and Madison Counties, although 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 four of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Cook County (formaldehyde, benzene, xylenes, 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. Benzene is another pollutant of interest
for these sites and appears on both emissions-based lists for Cook County. These
three pollutants are also pollutants of interest for ROIL but only formaldehyde
appears on both emissions-based lists (acetaldehyde and benzene only appear among
the highest emitted, ranking 12th and 15th, respectively, for their toxicity-weighted
emissions).
• Naphthalene and 1,3-butadiene are also pollutants of interest for NBIL and are among
those with the highest toxicity-weighted emissions in Cook County but are not among
the highest emitted. This is also true for arsenic. Trichloroethylene and 1,3-butadiene
are pollutants of interest for SPIL and are among those with the highest toxicity-
weighted emissions but are not among the highest emitted in Cook County.
• Formaldehyde is the pollutant of interest with the highest noncancer hazard
approximation for ROIL (albeit less than an HQ of 1.0). Formaldehyde ranks sixth for
the quantity emitted in Madison County and seventh for its toxicity-weighted
emissions. This is the only pollutant to appear on all three lists on Table 10-7.
Acetaldehyde, benzene, and ethylbenzene are pollutants of interest for ROIL that are
among the highest emitted but none of these appear among those with the highest
toxicity-weighted emissions.
• Several metals appear among the pollutants with the highest toxicity-weighted
emissions in Madison County, although none are among the highest emitted. Metals
were not sampled for at ROIL under the NMP.
10-68
-------
10.6 Summary of the 2014 Monitoring Data for NBIL, SPIL, and ROIL
Results from several of the data analyses described in this section include the following:
~~~ Eighteen pollutants (two carbonyl compounds, nine VOCs,five PAHs, and two
speciated metals) failed screens for NBIL; 13 pollutants (three carbonyl compounds
and 10 VOCs) failed screens for SPIL; and 12 pollutants (three carbonyl compounds
and nine VOCsj failed screens for ROIL.
~~~ Formaldehyde had the highest annual average concentration among the pollutants of
interest for SPIL and ROIL, while acetaldehyde had the highest annual average
concentration among the pollutants of interest for NBIL.
~~~ The maximum concentrations of several pollutants across the program were
measured at NBIL (acetaldehyde, acenaphthene, fluoranthene, fluorene, and
naphthalene).
~~~ Concentrations of acetaldehyde have been increasing significantly in recent years at
NBIL, although little change is shown between 2013 and 2014, while concentrations
of formaldehyde have a decreasing trend in recent years. Like many NMP sites, a
significant decrease in the number of non-detects reportedfor 1,2-dichloroethane has
occurred at both Chicago sites.
~~~ Formaldehyde has the highest cancer risk approximation among the pollutants of
interest for all three 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 examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the UATMP sites in Indiana, and integrates these concentrations with
emissions, meteorological, and risk information. Data generated by sources other than ERG are
not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) 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 geographical and
physical information about the location of the sites and the surrounding areas. 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 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 are 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 2011 NEI for point sources, version 2. 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 in order to
emphasize emissions sources within the boundary. Table 11-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates.
11-1
-------
Figure 11-1. Gary, Indiana (INDEM) Monitoring Site
Indiana East-West
-------
Figure 11-2. NEI Point Sources Located Within 10 Miles of INDEM
87"25'0"W
87o20'0"W
87"15'0"W
87°10'0"W
Lake
Michigan
Porter \
County i
Lake
County
Legend
*t* Airport/Airline/Airport Support Operations (14)
£ Asphalt Production/Hot Mix Asphalt Plant (1)
! 1 Brick, Structural Clay, or Clay Ceramics Plant (1)
B Bulk Terminals/Bulk Plants (5)
C Chemical Manufacturing Facility (11)
H Coke Battery (2)
I Compressor Station (7)
# Electricity Generation via Combustion (6)
E Electroplating, Plating, Polishing, Anodizing, and Coloring (2)
A Landfill (2)
(¦) Metal Can, Box, and Other Metal Container Manufacturing (1)
A Metal Coating, Engraving, and Allied Services to Manufacturers (1)
<•> Metals Processing/Fabrication Facility (7)
X Mine/Quarry/Mineral Processing Facility (16)
? Miscellaneous Commercial/Industrial Facility (20)
[] Paint and Coating Manufacturing Facility (1)
0 Petroleum Products Manufacturing (2)
A Petroleum Refinery (2)
R Plastic, Resin, or Rubber Products Plant (1)
P Printing/Publishing/Paper Product Manufacturing Facility (1)
X Rail Yard/Rail Line Operations (6)
V Steel Mill (10)
' I
i - -4
1 ILLIN
I
S- I- -
Legend
87°25'0"W 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
O 10 mile radius
County boundary
11-3
-------
Figure 11-3. Indianapolis, Indiana (WPIN) Monitoring Site
,E 32nd St
Frank R Reckwith"
. ¦ Park
¦Efe8th»St.
.^Zth.St—J|
,E-27th.St
Roach Sti
-------
Figure 11-4. NEI Point Sources Located Within 10 Miles of WPIN
86 "10'0"W
Hamilton
County
Marion
County
Legend
— 85"55'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
I Hancock
County
Shelby
County
Hendricks
County
~ WPIN UATMP site O 10 mile radius | I County boundary
Source Category Group (No. of Facilities)
*
Aerospace/Aircraft Manufacturing Facility (2)
O
Institutional (school, hospital, prison, etc.) (2)
T
Airport/Airline/Airport Support Operations (26)
A
Landfill (1)
*
Asphalt Production/Hot Mix Asphalt Plant (1)
A
Metal Coating, Engraving, and Allied Services to Manufacturers (2)
«
Automobile/Truck Manufacturing Facility (2)
<•>
Metals Processing/Fabrication Facility (5)
B
Bulk Terminals/Bulk Plants (3)
o
Miscellaneous Commercial/Industrial Facility (4)
C
Chemical Manufacturing Facility (6)
0
Municipal Waste Combustor (1)
S
Coke Battery (1)
D
Paint and Coating Manufacturing Facility (2)
l
Compressor Station (1)
CD
Pharmaceutical Manufacturing (2)
e
Electrical Equipment Manufacturing Facility (2)
R
Plastic, Resin, or Rubber Products Plant (3)
f
Electricity Generation via Combustion (3)
X
Rail Yard/Rail Line Operations (1)
F
Food Processing/Agriculture Facility (2)
«
Wastewater Treatment Facility (1)
I
Foundries, Iron and Steel (1)
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
34,754
1-90 N of 1-65 Interchange
WPIN
18-097-0078
Indianapolis
Marion
Indianapolis-Carmel-
Anderson, IN
39.811097,
-86.114469
Residential
Suburban
24,611
Keystone Ave at 30th St
1AADT reflects 2011 data (IN DOT, 2011)
-------
INDEM is located in Gary, Indiana, approximately 11 miles east of the Indiana-Illinois
border and 25 miles southeast of Chicago. Gary is located 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 steel mills; aircraft operations, which includes airports and related operations as well as
small runways and heliports, such as those associated with hospitals or TV stations; chemical
manufacturing; and mine/quarry/mineral processing. The sources closest to INDEM include a
steel mill; an industrial complex that includes several facilities that fall into the miscellaneous
commercial/industrial category as well as two mines/quarries and another steel mill; and a
heliport at a police station and a hospital.
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 painting and coating manufacturer, a metals processing/fabrication facility,
and a heliport. Each of these facilities is located within 2 miles of WPIN.
In addition to providing city, county, CBSA, 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 effect 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
11-7
-------
traffic volumes were obtained for roadways fairly close to the monitoring sites (1-90 near 1-65
for INDEM and North Keystone Avenue at East 30th Street for WPIN).
11.2 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
sites in Indiana on sample days, as well as over the course of the year.
11.2.1 Meteorological Summary
In order to provide an overview of the meteorological conditions experienced at each
monitoring site, hourly meteorological data for 2014 were retrieved and sample day and full-year
averages developed for temperature, dew point, relative humidity, pressure, and wind speed.
Weather data from the actual monitoring site(s) were obtained from AQS, where available. If
site-specific weather data were not available in AQS, then data were obtained from NCDC for
the NWS weather station located closest to the monitoring site(s), as described in Section 3.4.2.
For INDEM, site-specific data were available for all of the parameters in Table 11-2 except sea
level pressure. For WPIN, AQS data were available but the instrumentation was down
occasionally, with a longer outage from mid-May through early July; thus, data from the NWS
weather station at Eagle Creek Airpark (WBAN 53842) were used as surrogates for missing data.
The Eagle Creek Airpark weather station is located 9.7 miles west of WPIN. A map showing the
distance between each monitoring site and the closest NWS weather station is provided in
Appendix R. These data were used to determine how meteorological conditions on sample days
vary from conditions experienced throughout the year.
Table 11-2 presents average temperature, average dew point temperature, average relative
humidity, average station and sea level pressure, and wind information (average scalar wind
speed and prevailing wind direction) for days on which samples were collected and for all of
2014. Also included in Table 11-2 is the 95 percent confidence interval for each parameter. As
shown in Table 11-2, average meteorological conditions on sample days at INDEM were
generally representative of average weather conditions experienced throughout the year at this
site. The greatest difference between the sample day and full-year averages was calculated for
average dew point temperature. The difference in several parameters may be attributable, at least
in part, to extra samples collected in December. For WPIN, the greatest difference was
calculated for average relative humidity, although relatively large differences were also
11-8
-------
calculated for average temperature and average dew point. A number of make-up samples were
collected throughout out the year, primarily in March, which may contribute to these differences.
Table 11-2. Average Meteorological Conditions near the Indiana Monitoring Sites
Average
Average
Average
Average
Average
Average
Dew Point
Relative
Sea Level
Station
Prevailing
Scalar Wind
Average
Temperature
Temperature
Humidity
Pressure
Pressure
Wind
Speed
Type1
(°F)
(°F)
(%)
(in Hg)
(in Hg)
Direction
(kt)
Gary, Indiana - INDEM2
Sample
Days
45.7
35.7
70.6
29.44
6.0
(62)
± 1.1
± 1.1
±0.9
NA
±0.01
ssw
±0.2
46.9
37.3
71.9
29.39
6.5
2014
±0.4
±0.4
±0.4
NA
±<0.01
s
±0.1
Indianapolis,
Indiana - WPIN3
Sample
Days
49.1
37.3
66.3
30.07
29.23
4.3
(68)
± 1.1
± 1.0
±0.9
±0.01
±0.01
WNW
±0.1
50.8
39.8
69.0
30.04
29.20
4.4
2014
±0.4
±0.4
±0.4
± <0.01
±<0.01
WNW
±0.1
'Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
2A11 parameters except sea level pressure were measured at INDEM. Sea level pressure was not collected at INDEM or the
closest NWS weather station and thus, is not presented here.
3A11 parameters except sea level pressure were measured at WPIN; however, the meteorological instruments were down part of
the year, so data from the closest NWS weather station located at Eagle Creek Airpark, WBAN 53842, were used as a surrogate.
NA= Sea level pressure was not recorded at the Lansing Municipal Airport.
11.2.2 Wind Rose Comparison
Hourly surface wind data were also uploaded into a wind rose software program to
produce customized wind roses, as described in Section 3.4.2. A wind rose shows the frequency
of wind directions using "petals" positioned around a 16-point compass, and uses different colors
to represent wind speeds. Figure 11-5 presents two wind roses for the INDEM monitoring site.
The first is a wind rose representing wind observations for all of 2014 and the second is a wind
rose representing wind observations for days on which samples were collected in 2014. These
are used to identify the predominant wind speed and direction for 2014 and to determine if wind
observations on sample days were representative of conditions experienced over the entire year.
Figure 11-6 presents the full-year and sample day wind roses for WPIN.
11-9
-------
Figure 11-5. Wind Roses for the Wind Data Collected at INDEM
2014 Wind rose Sample Day Wind Rose
NORTH
NORTH
WEST
;WEST
WIND SPEED
(Knots)
WIND SPEED
(Knots)
¦SOUTH
SOUTH
Calms: 0.25%
Calms: 0.27%
Observations from Figure 11-5 for INDEM include the following:
• The 2014 wind rose shows that light winds were infrequent and calm winds account
for less than 1 percent of the wind observations. Winds from the south-southeast to
west-southwest were commonly observed, together accounting for more than half of
wind observations. Winds from the north to northeast make up a secondary wind
grouping. Winds from the east-northeast to east-southeast were infrequently
observed, as were winds from the west-northwest to northwest.
• The sample day wind rose resembles the full-year wind rose, exhibiting similar wind
speeds and direction observation patterns. While winds from the south-southeast to
west-southwest still accounted for the majority of observations on sample days, the
percentage of observations was more evenly distributed on sample days than over the
course of the year.
11-10
-------
Figure 11-6. Wind Roses for the Wind Data Collected at WPIN
2014 Wind Rose Sample Day Wind Rose
NORTH
NORTH
WEST
WEST
WIND SPEED
(Knots)
WIND SPEED
(Knots)
SOUTH
SOUTH
Calms: 5.46%
Calms: 1.75%
Observations from Figure 11-6 for WPIN include the following:
• Winds from the south-southeast, south, the southwestern quadrant, west, and west-
northwest account for the majority (nearly 55 percent) of wind observations at WPIN
in 2014. With the exception of east winds, winds from each of the remaining
directions accounted for less than 5 percent of observations. Calm winds were
observed for roughly 5 percent of observations. Winds greater than 17 knots were
rarely observed at WPIN.
• The wind patterns on the sample day wind rose exhibit more variability than the full-
year wind rose. Winds from the south-southeast to west-northwest still account for
the majority of wind observations, but winds from the north-northwest and northeast
to east accounted for a higher percentage of the observations on sample days. The
calm rate was lower on sample days as well, with less than 2 percent of the
observations falling into this category.
• Recall from the previous section that wind sensors were down at WPIN for a portion
of 2014 and NWS data were used as a surrogate for missing data.
11-11
-------
11.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each
Indiana monitoring 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. 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 11-3. 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 11-3. 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 11-3. 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
Formaldehyde
0.077
57
57
100.00
50.44
50.44
Acetaldehyde
0.45
56
57
98.25
49.56
100.00
Total
113
114
99.12
Indianapolis, Indiana - WPIN
Acetaldehyde
0.45
54
54
100.00
50.00
50.00
Formaldehyde
0.077
54
54
100.00
50.00
100.00
Total
108
108
100.00
Observations from Table 11-3 include the following:
• Acetaldehyde and formaldehyde are the only pollutants to fail screens for INDEM
and WPIN.
• Formaldehyde failed 100 percent of screens for both sites. Acetaldehyde failed
100 percent of screens for WPIN and 98 percent of screens for INDEM.
• Both pollutants were identified as pollutants of interest for each site, contributing
equally or nearly equally to the number of failed screens for each site.
• Note that only three carbonyl compounds have risk screening values (formaldehyde,
acetaldehyde, and propionaldehyde).
11-12
-------
11.4 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.
• 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 are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at the Indiana sites are provided in Appendix L.
11.4.1 2014 Concentration Averages
Quarterly and annual concentration averages 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 the entire year of sampling. Annual averages were
calculated for pollutants where three valid quarterly averages could be calculated 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 11-4, 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.
11-13
-------
Table 11-4. Quarterly and Annual Average Concentrations of the Pollutants of Interest
for the Indiana Monitoring Sites
Pollutant
# of
Measured
Detections
vs. # of
Samples
Total
# of
Samples
1st
Quarter
Average
frig/m3)
2nd
Quarter
Average
frig/m3)
3rd
Quarter
Average
frig/m3)
4th
Quarter
Average
frig/m3)
Annual
Average
frig/m3)
Gary, Indiana - INDEM
Acetaldehyde
57/57
57
1.10
±0.19
1.08
±0.16
1.28
±0.22
1.17
±0.23
1.16
±0.10
Formaldehyde
57/57
57
1.86
±0.31
3.11
±0.53
3.43
±0.64
1.81
±0.46
2.59
±0.30
Indianapolis, Indiana - WPIN
Acetaldehyde
54/54
54
1.73
±0.18
NA
1.85
±0.26
1.39
±0.28
1.68
±0.14
Formaldehyde
54/54
54
2.58
±0.33
NA
3.38
±0.51
1.71
±0.51
2.68
±0.32
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
Observations for the Indiana sites from Table 11-4 include the following:
• For both sites, acetaldehyde and formaldehyde were detected in all of the valid
carbonyl compound samples collected.
• The annual average formaldehyde concentrations for these two sites are similar to
each other. The annual average concentration of formaldehyde for INDEM is more
than twice this site's annual average concentration of acetaldehyde. While the annual
average formaldehyde concentration for WPIN is also greater than this site's annual
average acetaldehyde concentration, the difference is less.
• The second and third quarter average concentrations of formaldehyde are
significantly higher than the first and fourth quarter averages for INDEM. A review
of the data shows that all but one of the 17 highest formaldehyde concentrations
(those greater than 3 |ig/m3) were measured between April and September and ranged
from 3.09 |ig/m3 to 6.37 |ig/m3; conversely, all but two of the 19 lowest
concentrations (those less than 2 |ig/m3) were measured between January and March
or October and December. This supports the trend identified in Section 4.4.2
indicating that formaldehyde concentrations tended to be higher during the warmer
months of the year. Acetaldehyde concentrations measured at INDEM do not exhibit
this trend. The quarterly average concentrations of acetaldehyde vary by only
0.2 |ig/m3.
• Power supply issues at WPIN resulted in samples not being collected between mid-
May and late June. As a result, no quarterly average concentrations could be
calculated for the second quarter.
• Even though the average formaldehyde concentration for the third quarter is twice the
average concentration for the fourth quarter, their confidence intervals are the same,
indicating more variability in the fourth quarter measurements at WPIN.
11-14
-------
Tables 4-9 through 4-12 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:
• INDEM does not appear in Table 4-10. Its annual average concentration of
formaldehyde ranks 17th and its annual average concentration of acetaldehyde ranks
23rd among NMP sites sampling carbonyl compounds.
• WPIN does not appear in Table 4-10 either. Its annual average concentration of
formaldehyde ranks 15th and its annual average concentration of acetaldehyde ranks
17th among NMP sites sampling carbonyl compounds.
• This is the first time in several years that WPIN does not appear in this table for
formaldehyde.
11.4.2 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 11-4 for INDEM and WPIN. Figures 11-7 and 11-8 overlay the 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 3.4.3.1, and are
discussed below.
11-15
-------
Figure 11-7. Program vs. Site-Specific Average Acetaldehyde Concentrations
INDEM
WPIN
12 3
4 5 6
Concentration (ng/m3)
7
8
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site: Site Average
o
Site Concentration Range
Figure 11-7 presents the box plots for acetaldehyde for both sites and shows the
following:
11-16
¦
The maximum concentration of acetaldehyde measured at INDEM is less than the
program-level third quartile. The annual average concentration for INDEM is less
than both the program-level average and median concentrations.
Acetaldehyde concentrations measured at WPIN are higher than those measured at
INDEM. WPIN's annual average concentration is just less than the program-level
average concentration and just greater than the program-level median concentration
The minimum concentration measured at WPIN is just less than the program-level
first quartile.
-------
Figure 11-8. Program vs. Site-Specific Average Formaldehyde Concentrations
INDEM
WPIN
0 3 6 9 12 15 18 21 24 27
Concentration (ng/m3)
Progra m: 1st Qua rti 1 e
2ndQuartile 3rdQuartile
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 11-8 presents the box plots for formaldehyde for both sites and shows the
following:
• The range of formaldehyde concentrations measured at INDEM is slightly larger than
the range measured at WPIN.
• The annual average concentrations for these two sites are similar to each other. Both
annual averages fall between the program-level average and program-level median
concentrations.
11.4.3 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 3.4.3.2.
INDEM and WPIN have sampled carbonyl compounds under the NMP since 2004 and 2007,
respectively. Thus, Figures 11-9 through 11-12 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.
I I I I I I I I I
11-17
-------
Figure 11-9. Yearly Statistical Metrics for Acetaldehyde Concentrations
Measured at INDEM
2005 1 2006
X I
2009
Year
o 5th Percentile
— Minimum
— Maximum
o 95th Percentile
..4.... Average
1 A 1-year average is not presented due to a gap in sampling between September 2005 and November 2005.
Observations from Figure 11-9 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 11-9 begins with 2004. However, 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.
• 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 all
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 for 2008).
11-18
-------
• 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 samplers were switched out
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.
• 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 similar.
• With the exception of the minimum concentration, each of the statistical parameters
increased slightly for 2014 compared to 2013, but are still in line with those shown
for recent years.
Figure 11-10. Yearly Statistical Metrics for Formaldehyde Concentrations
Measured at INDEM
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented due to a break in sampling between September 2005 and November
2005."
11-19
-------
Observations from Figure 11-10 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 11-10. There have been 38
concentrations of formaldehyde greater than 100 |ig/m3 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. There were 24 formaldehyde concentrations measured in 2008
that were less than the minimum concentration measured in 2007; those 24
measurements represent 40 percent of the concentrations measured in 2008. The last
"high" concentration was measured on August 4, 2008, after which no formaldehyde
concentrations greater than 4 |ig/m3 were measured that year.
• All the statistical metrics decreased significantly for 2009 and the years that follow,
with the 1-year average concentrations ranging from 2.14 |ig/m3 (2013) to 2.59 |ig/m3
(2014). The number of formaldehyde measurements greater than 4 |ig/m3 ranged
from two to nine for each year between 2009 and 2014 (with the most measured in
2014), compared to 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 sampler used at INDEM during the PAMS season was replaced in 2009 and
their 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
sampler 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.
11-20
-------
Figure 11-11. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at WPIN
8.0
7.0
6.0
5.0
E
If
0.0
2007 2008 2009 2010 2011 2012 2013 2014
Year
O 5th Percentile - Minimum - Med en — Maximum o 95th Percentile •¦•¦^•¦¦Average
Observations from Figure 11-11 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 11-11 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. Three additional concentrations greater than
5 |ig/m3 have been measured at WPIN (two in 2007 and one in 2012).
• 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 concentrations
greater than 3 |ig/m3 increased five-fold, from three measured in 2009 to 15 measured
in 2010.
• The 1-year average concentration has a decreasing trend again 2010, with all of the
statistical parameters at a minimum for 2014 over the years of sampling, with the
exception of the minimum concentration.
11-21
-------
Figure 11-12. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at WPIN
Maximum
Observations from Figure 11-12 for formaldehyde concentrations measured at WPIN
include the following:
• The maximum concentration of formaldehyde measured at WPIN was measured in
2011 (11.1 |ig/m3). The next three highest concentrations were measured at WPIN in
2012 and ranged from 9.87 |ig/m3 to 10.7 |ig/m3, although additional formaldehyde
concentrations greater than 9 |ig/m3 were also measured in 2007, 2009, and 2011.
• 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 data for 2011 and
2012 show that the number of concentrations in the 2 |ig/m3 to 4 |ig/m3 range
increased from 21 in 2011 to 29 in 2012 while the number of concentrations in the
4 |ig/m3 to 6 |ig/m3 range decreased by nearly half (from 20 in 2011 to 11 in 2012).
These changes explain the change in the median concentration while a few additional
measurements in the upper end of the concentration range explain the increase in the
95th percentile.
• Nearly all of the statistical parameters exhibit decreases for 2013, with additional
decreases shown for 2014. Each of the statistical parameters are at a minimum for
2014, with the 1-year average concentration less than 3 |ig/m3 for the first time.
11-22
-------
11.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at each Indiana monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.3 for
definitions and explanations regarding the various toxicity factors, time frames, and calculations
associated with these risk-based screenings.
11.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Indiana 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 3.4.3.3 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 11-5, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Table 11-5. Risk Approximations for the Indiana Monitoring Sites
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Gary, Indiana - INDEM
Acetaldehyde
0.0000022
0.009
57/57
1.16
±0.10
2.55
0.13
Formaldehyde
0.000013
0.0098
57/57
2.59
±0.30
33.73
0.26
Indianapolis, Indiana - WPIN
Acetaldehyde
0.0000022
0.009
54/54
1.68
±0.14
3.69
0.19
Formaldehyde
0.000013
0.0098
54/54
2.68
±0.32
34.81
0.27
Observations for the Indiana sites from Table 11-5 include the following:
• For both sites, the annual average concentration of formaldehyde is greater than the
annual average concentration of acetaldehyde. The annual average acetaldehyde
11-23
-------
concentration for WPIN is greater than the annual average for INDEM. While this is
also true for the annual averages of formaldehyde, the difference is less.
• The cancer risk approximation for formaldehyde is an order of magnitude higher than
the cancer risk approximation for acetaldehyde for both sites.
• The cancer risk approximation for formaldehyde for WPIN (34.81 in-a-million) is
similar to the cancer risk approximation for formaldehyde for INDEM (33.73 in-a-
million). There is a similar difference between the cancer risk approximations for
acetaldehyde (3.69 in-a-million for WPIN and 2.55 in-a-million for INDEM).
• 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.
11.5.2 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 11-6 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
that have cancer toxicity factors. Table 11-6 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 11-6 provides the pollutants with the highest cancer risk approximations (in-a-million) for
each site, as presented in Table 11-5. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 11-6. Table 11-7 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 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 11.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
11-24
-------
Table 11-6. 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)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Gary, Indiana (Lake County) - INDEM
Benzene
177.07
Coke Oven Emissions, PM
2.38E-03
Formaldehyde
33.73
Formaldehyde
145.37
POM, Group lb
1.92E-03
Acetaldehyde
2.55
Ethylbenzene
94.06
Formaldehyde
1.89E-03
Acetaldehyde
84.11
Benzene
1.38E-03
1.3 -Butadiene
27.28
Hexavalent Chromium
9.67E-04
POM, Group lb
21.84
1,3-Butadiene
8.18E-04
Naphthalene
13.35
Arsenic, PM
6.53E-04
T etrachloroethylene
9.35
Naphthalene
4.54E-04
POM, Group 2b
2.78
POM, Group 2b
2.45E-04
POM, Group 2d
2.68
Nickel, PM
2.38E-04
Indianapolis, Indiana (Marion County) - WPIN
Benzene
421.74
Formaldehyde
4.14E-03
Formaldehyde
34.81
Formaldehyde
318.24
Benzene
3.29E-03
Acetaldehyde
3.69
Ethylbenzene
268.73
1,3-Butadiene
1.87E-03
Acetaldehyde
189.64
Naphthalene
1.11E-03
1,3-Butadiene
62.21
Arsenic, PM
1.05E-03
T etrachloroethylene
33.59
Ethylbenzene
6.72E-04
Naphthalene
32.73
POM, Group 2b
6.51E-04
POM, Group 2b
7.40
Nickel, PM
5.08E-04
POM, Group 2d
5.22
POM, Group 2d
4.60E-04
Propylene oxide
4.72
Hexavalent Chromium
4.20E-04
-------
Table 11-7. 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)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Gary, Indiana (Lake County) - INDEM
Toluene
669.66
Acrolein
476,219.32
Formaldehyde
0.26
Xylenes
436.89
Lead, PM
52,690.40
Acetaldehyde
0.13
Hexane
427.58
Manganese, PM
22,492.64
Methanol
328.00
Hydrochloric acid
16,187.23
Hydrochloric acid
323.74
Formaldehyde
14,834.03
Benzene
177.07
1,3-Butadiene
13,641.59
Formaldehyde
145.37
Chlorine
12,016.67
Ethylene glycol
98.80
Arsenic, PM
10,126.34
Ethylbenzene
94.06
Acetaldehyde
9,345.56
Acetaldehyde
84.11
Benzene
5,902.45
Indianapolis, Indiana (Marion County) - WPIN
Toluene
1,660.99
Acrolein
1,224,556.10
Formaldehyde
0.27
Xylenes
1,008.89
Formaldehyde
32,473.78
Acetaldehyde
0.19
Hexane
773.82
1,3-Butadiene
31,104.81
Methanol
532.81
Hydrochloric acid
23,337.36
Hydrochloric acid
466.75
Acetaldehyde
21,070.71
Benzene
421.74
Arsenic, PM
16,282.89
Formaldehyde
318.24
Benzene
14,057.94
Ethylbenzene
268.73
Lead, PM
13,691.58
Ethylene glycol
203.01
Nickel, PM
11,766.86
Acetaldehyde
189.64
Naphthalene
10,909.09
-------
Observations from Table 11-6 include the following:
• Benzene, formaldehyde, and ethylbenzene are the three highest emitted pollutants
with cancer UREs in both Marion and Lake County, although the quantity emitted is
higher in Marion County.
• Coke oven emissions is the pollutant with the highest toxicity-weighted emissions (of
the pollutants with cancer UREs) for Lake County, followed by POM, Group lb and
formaldehyde. Formaldehyde, benzene, and 1,3-butadiene are the pollutants with the
highest toxicity-weighted emissions for Marion County.
• Six of the highest emitted pollutants in Lake County also have the highest toxicity-
weighted emissions; seven 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 and
WPIN. Acetaldehyde and formaldehyde appear among the highest emitted pollutants
for both counties, with only formaldehyde appearing among the pollutants with the
highest toxicity-weighted emissions for each county. Formaldehyde has the highest
toxicity-weighted emissions in Marion County and ranks third highest for Lake
County.
• While several metals (arsenic, nickel, and hexavalent chromium) are among the
pollutants with the highest toxicity-weighted emissions for both counties, none of
these are among the highest emitted pollutants for either county. This demonstrates
that a pollutant does not have to be emitted in large quantities to be a health hazard.
• Several POM Groups and naphthalene appear among the highest emitted pollutants
and the pollutants with the highest toxicity-weighted emissions for both counties.
Neither site sampled PAHs under the NMP.
Observations from Table 11-7 include the following:
• Toluene, xylenes, and hexane are the three highest emitted pollutants with noncancer
RfCs in both Marion and Lake County, although the quantity emitted is higher in
Marion County. The same 10 pollutants appear on each county's list of highest
emitted pollutants.
• Acrolein is the pollutant with the highest toxicity-weighted emissions (of the
pollutants with noncancer RfCs) for both counties. Lead and manganese rank second
and third for Lake County, while formaldehyde and 1,3-butadiene rank second and
third for Marion County.
• Four of the highest emitted pollutants in Lake County also have the highest toxicity-
weighted emissions (formaldehyde, acetaldehyde, benzene, and hydrochloric acid).
The same four pollutants appear on both emissions-based lists for Marion County.
11-27
-------
• Several metals are among the pollutants with the highest toxicity-weighted emissions
for Lake and Marion Counties, although none of these appear among the highest
emitted pollutants.
• Formaldehyde and acetaldehyde appear in all three columns in Table 11-7 for both
sites.
Summary of the 2014 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 2014. Acetaldehyde
andformaldehyde failed screens for each site and were identified as pollutants of
interest for each site.
~~~ The annual average concentration of formaldehyde is greater than the annual
average concentration of acetaldehyde for both sites. Concentrations offormaldehyde
exhibit a seasonal trend, with higher concentrations measured during the warmer
months of the year.
~~~ Concentrations of formaldehyde and acetaldehyde decreased significantly at INDEM
from 2008 to 2009; these changes may be at least partially explained by a sampler
replacement. Concentrations of formaldehyde and acetaldehyde both exhibit a
decreasing trend at WPIN in the last few years.
11-28
-------
12.0 Sites in Kentucky
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS and UATMP sites in Kentucky, and integrates these
concentrations with emissions, meteorological, and risk information. Data generated by sources
other than ERG are not included in the data analyses contained in this report. Readers are
encouraged to refer to Sections 1 through 4 and the glossary (Appendix P) for detailed
discussions and definitions regarding the various data analyses presented below.
12.1 Site Characterization
This section characterizes the Kentucky monitoring sites by providing geographical and
physical information about the location of the sites and the surrounding areas. 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.
Data from 10 monitoring sites in Kentucky are included in this section, one NATTS site
and nine predominantly "source-oriented" sites. Three monitoring sites are located in northeast
Kentucky, two in Ashland and one near Grayson Lake. One monitoring site is located south of
Evansville, Indiana in the town of Baskett. Five monitoring sites are located in or near the
Calvert City area, east of Paducah, Kentucky. The final monitoring site is located in Lexington,
in north-central Kentucky. A composite satellite image and facility map is provided for each site
in Figures 12-1 through 12-15. 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 2011 NEI for
point sources, version 2. 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 in order 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.
12-1
-------
Figure 12-1. Ashland, Kentucky (ASKY) Monitoring Site
ge.Rd ^°6o
Hart-St-
610ft
USG'S '
,Sou rcejASA. N G A . USGS
i 20 0 8 Microsoft Corp.
-------
Figure 12-2. Ashland, Kentucky (ASKY-M) Monitoring Site
-------
Figure 12-3. NEI Point Sources Located Within 10 Miles of ASKY and ASKY-M
82"50'0"W
82"35'0"W
82'25'0"W
Greenup
County
Ohio River
Boyd
County
ASKY UATMP site
ASKY-M UATMP site O 10 mile radius
County boundary
Legend
i
* Carter
' County
F *
i ~r
\
WEST *
VIRGINIA \
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 intere
Source Category Group (No. of Facilities)
•t" Airport/Airline/Airport Support Operations (6)
£ Asphalt Production/Hot Mix Asphalt Plant (1)
B Bulk Terminals/Bulk Plants (1)
c Chemical Manufacturing Facility (5)
S Coke Battery (1)
i Compressor Station (3)
f Electricity Generation via Combustion (4)
F Food Processing/Agriculture Facility (2)
if Gasoline/Diesel Service Station (1)
• Landfill (2)
© Metals Processing/Fabrication Facility (2)
x Mine/Quarry/Mineral Processing Facility (8)
? Miscellaneous Commercial/Industrial Facility (3)
0 Paint and Coating Manufacturing Facility (1)
< Pesticide Manufacturing Plant (1)
n Petroleum Refinery (1)
R Plastic, Resin, or Rubber Products Plant (2)
V Port and Harbor Operations (2)
x Rail Yard/Rail Line Operations (2)
A. Ship/Boat Manufacturing or Repair Facility (1)
V Steel Mill (2)
© Testing Laboratories (1)
12-4
-------
Figure 12-4. Grayson, Kentucky (GLKY) Monitoring Site
to
/Appalachian Mountains
\ ¦ I
BgfcjX ^
¥/ Lyr*
JB9KJ
/' V**1-
P5r m
s»*L
i# /
Source: USGS
Source: NASA, NGA, USGS
© 2008 Microsoft Corp.
-------
Figure 12-5. NEI Point Sources Located Within 10 Miles of GLKY
82"50 0 W
— (
Greenup 1
* County \ r Boyd
1 f \ / County
\ \
®mw&
Carter
County
Rowan
Elliot
County I
County
\yj Lawrence \
County
83°15"0°W 83°10'0"W 83°5'0"W 83°0'0"W 82°55'0-W 82350'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
Legend
^ GLKY NATTS site O 10 mile radius
County boundary
Source Category Group (No. of Facilities)
Asphalt Production/Hot Mix Asphalt Plant (1)
n Brick, Structural Clay, or Clay Ceramics Plant (2)
b Bulk Terminals/Bulk Plants (1)
F Food Processing/Agriculture Facility (1)
o Institutional (school, hospital, prison, etc.) (1)
x Mine/Quarry/Mineral Processing Facility (2)
12-6
-------
Figure 12-6, Baskett, Kentucky (BAKY) Monitoring Site
Source: USGS
iource: NASA. N G A . USGS
© 2008 Microsoft Corp.
-------
Figure 12-7. NEI Point Sources Located Within 10 Miles of BAKY
87°30'0"W
87C25'0"W
87°20'0"W
87°15'0"W
INDIANA \
Ohio River
Henderson
County
Source Category Group (No. of Facilities)
"t Airport/Airline/Airport Support Operations (6) x
i Asphalt Production/Hot Mix Asphalt Plant (1) ?
B Bulk Terminals/Bulk Plants (1)
C Chemical Manufacturing Facility (2) R
ffi Dry Cleaning Facility (2) E
f Electricity Generation via Combustion (4) X
E Electroplating, Plating, Polishing, Anodizing, and Coloring (1) A
F Food Processing/Agriculture Facility (3) ®
<•> Metals Processing/Fabrication Facility (6)
Mine/Quarry/Mineral Processing Facility (3)
Miscellaneous Commercial/Industrial Facility (3)
Paint and Coating Manufacturing Facility (1)
Plastic, Resin, or Rubber Products Plant (1)
Pulp and Paper Plant (2)
Rail Yard/Rail Line Operations (1)
Ship/Boat Manufacturing or Repair Facility (1)
Testing Laboratories (1)
BAKY UATMP site
O 10 mile radius
County boundary
Legend
87"30'0"W 87"25'0"W 87°20'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
Daviess
County
I
J
12-8
-------
Figure 12-8. Calvert City, Kentucky (ATKY) Monitoring Site
Johmonfoiley^Rd
-------
Figure 12-9. Smithland, Kentucky (BLKY) Monitoring Site
-------
Figure 12-10. Calvert City, Kentucky (CCKY) Monitoring Site
-------
Figure 12-11. Calvert City, Kentucky (LAKY) Monitoring Site
626ft
Source: USGS
Source: NASA, NGA, USGS
© 2008 Microsoft Corp.
to
K>
-------
Figure 12-12. Calvert City, Kentucky (TVKY) Monitoring Site
-------
Figure 12-13. NEI Point Sources Located Within 10 Miles of ATKY, BLKY, CCKY,
LAKY, and TVKY
Livingston
County
OHIO
Lyon
County
innessee
River
Lake
Barkley
Marshall
County
Legend
~ ATKY UATMP site ^
~ BLKY UATMP site ^
3°30'0"W 88"25'0"W 88°20'0"W 88"15'0"W 88*1 WW
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
CCKY UATMP site
LAKY UATMP site
^ TVKY UATMP site
O 10 mile radius
County boundary
Source Category Group (No. of Facilities)
"1" Airport/Airline/Airport Support Operations (1)
'i Asphalt Production/Hot Mix Asphalt Plant (1)
B Bulk Terminals/Bulk Plants (1)
C Chemical Manufacturing Facility (10)
i Compressor Station (1)
f Electricity Generation via Combustion (1)
+ Industrial Machinery or Equipment Plant (1)
<•> Metals Processing/Fabrication Facility (3)
X Mine/Quarry/Mineral Processing Facility (7)
? Miscellaneous Commercial/Industrial Facility (5)
R Plastic, Resin, or Rubber Products Plant (4)
•ii. Ship/Boat Manufacturing or Repair Facility (4)
V Steel Mill (1)
§
_ - B-
0
f ?l
I
I
- - - - r
i
McCracken |
County |
I
' Crittenden
* County
I
1 County
2.5 5
12-14
-------
Figure 12-14. Lexington, Kentucky (LEKY) Monitoring Site
Fred Douglass
-------
Figure 12-15. NET Point Sources Located Within 10 Miles of LEKY
84°35'0"W
84"30'0"W
84"25"0"W
Scott
County
Fayette
County
Jessamine
County
Source Category Group (No. of Facilities)
"t" Airport/Airline/Airport Support Operations (9)
A Animal Feedlot or Farm (1)
£ Asphalt Production/Hot Mix Asphalt Plant (4)
0 Auto Body Shop/Painters/Automotive Stores (2)
5 Automobile/Truck Manufacturing Facility (4)
B Bulk Terminals/Bulk Plants (3)
X Crematory - Animal/Human (1)
(3) Dry Cleaning Facility (1)
6 Electrical Equipment Manufacturing Facility (3)
# Electricity Generation via Combustion (1)
E Electroplating, Plating, Polishing, Anodizing, and Coloring (1)
F Food Processing/Agriculture Facility (2)
fV Glass Plant (1)
-Jfr- Industrial Machinery or Equipment Plant (6)
O Institutional (school, hospital, prison, etc.) (10)
A Landfill (2)
<•> Metals Processing/Fabrication Facility (2)
A Military Base/National Security Facility (1)
X Mine/Quarry/Mineral Processing Facility (13)
? Miscellaneous Commercial/Industrial Facility (4)
Q Paint and Coating Manufacturing Facility (1)
R Plastic, Resin, or Rubber Products Plant (3)
P Printing/Publishing/Paper Product Manufacturing Facility (3)
TT Telecommunications/Radio Facility (1)
M Tobacco Manufacturing (1)
W Woodwork, Furniture, Millwork & Wood Preserving Facility (4)
LEKY UATMP site
G 10 mile radius
County boundary
Miles
^ r—
84°45'0"W &4'40'0"W
Legend
84'35'0-W 84'30'0"W 84325'0"W 84520'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
I
Bourbon
County
I
I &
** Woodford
County
1
I
1
\
\
? I
12-16
-------
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
12,842
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
922
Rte 1078 N of Hwy 60
ATKY
21-157-0016
Calvert
City
Marshall
None
37.041760,
-88.354070
Industrial
Suburban
3,262
Main St (Rte 95), S of Johnson Rilev
Rd
BLKY
21-139-0004
Smithland
Livingston
Paducah, KY-IL
37.071510,
-88.333890
Agricultural
Rural
2,510
Rte 93/453, E of BloodworthRd
CCKY
21-157-0018
Calvert
City
Marshall
None
37.027020,
-88.343870
Residential
Suburban
4,050
Industrial Pkwy, S of E 5th Ave
LAKY
21-157-0019
Calvert
City
Marshall
None
37.037180,
-88.334110
Residential
Suburban
1,189
Rte 282 (Gilbertsville Hwy), E of
Industrial Lane
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 2014 data for ASKY, LEKY & TVKY; 2012 data for ASKY-M, GLKY, BAKY, ATKY, and LAKY; and 2013 data for BLKY and CCKY (KYTC, 2014)
BOLD ITALICS = EPA-designated NATTS Site
-------
There are two Kentucky monitoring sites 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 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 ASKY-M 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 metals processing/fabrication facility and coke
battery 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, 2016; ACE, 2016). 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 radius in Figure 12-5. Sources within 10 miles of GLKY are involved in
asphalt production, brick/structural clay/clay ceramics manufacturing, food processing, and
mining, among others.
12-18
-------
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, and 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.
There are five monitoring sites 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, 2016).
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. Approximately 1 mile east-southeast down Gilbertsville Highway is the LAKY
monitoring site. LAKY is located behind a mobile home park. Although located in a residential
area, industrial areas are located to the west, northwest, and north. Just over one-half mile north
of LAKY is the TVKY monitoring site. This monitoring site is located at a power substation just
south of another chemical manufacturing plant. The fourth monitoring site in Calvert City is
located at Calvert City Elementary School. The CCKY site is located behind the school, which
backs up to a forested area just south of the aforementioned rail line and to the south of most of
the industry. The BLKY site 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. The composite satellite images for these
sites are provided in alphabetical order by site in Figures 12-8 through 12-12.
Figure 12-13 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
12-19
-------
located between ATKY, LAKY, and the Tennessee River. Many of the emissions sources closest
to the Calvert City sites are in the chemical manufacturing source category. There are also
several plastic, resin, or rubber product plants located between these sites. 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.
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 12-14. LEKY is located just over a half-mile south of New Circle Road (4/421), a loop
encircling the city of Lexington. Figure 12-15 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 a food processing plant, the aforementioned electrical equipment manufacturing plant, a
crematory, a metals processing and fabrication facility, and an automobile/truck manufacturing
facility.
In addition to providing city, county, CBSA, 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 effect 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 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
sites in Kentucky on sample days, as well as over the course of the year.
12-20
-------
12.2.1 Meteorological Summary
In order to provide an overview of the meteorological conditions experienced at each
monitoring site, hourly meteorological data for 2014 were retrieved and sample day and full-year
averages developed for temperature, dew point, relative humidity, pressure, and wind speed.
Weather data from the actual monitoring site(s) were obtained from AQS, where available. If
site-specific weather data were not available in AQS, then data were obtained from NCDC for
the NWS weather station located closest to the monitoring site(s), as described in Section 3.4.2.
For the Kentucky sites, site-specific data were available for some, but not all, of the parameters
in Table 12-2 for half of the monitoring sites. Temperature and wind data were collected at
ASKY, BLKY, and CCKY; at GLKY, temperature, pressure, dew point, relative humidity, and
wind data were collected. For the remaining sites, meteorological observations were not
available in AQS. Weather data from the closest NWS station were used for meteorological
parameters without data and/or as a surrogate for parameters without complete observation
records. A map showing the distance between each Kentucky monitoring site and the closest
NWS weather station is provided in Appendix R. These data were used to determine how
meteorological conditions on sample days vary from conditions experienced throughout the year.
Table 12-2 presents average temperature, average dew point temperature, average relative
humidity, average station and sea level pressure, and wind information (average scalar wind
speed and prevailing wind direction) for days on which samples were collected and for all of
2014. Also included in Table 12-2 is the 95 percent confidence interval for each parameter. As
shown in Table 12-2, average meteorological conditions on sample days were generally
representative of average weather conditions experienced throughout the year at each site.
GLKY is the site with the smallest differences between the sample day averages and the full-year
averages, while the Calvert City sites have the largest, particularly for dew point. It should be
noted that even though sample days are generally standardized, the need for making up invalid
samples leads to additional sample days. This is why although the data are from the same
weather station, the sample day averages are often different from each other, such is the case
with ATKY, LAKY, and TVKY.
12-21
-------
Table 12-2. Average Meteorological Conditions near the Kentucky Monitoring Sites
Average
Type1
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(in Hg)
Average
Station
Pressure
(in Hg)
Prevailing
Wind
Direction
Average
Scalar Wind
Speed
(kt)
Health De
jartment, Ashland, Kentucky - ASKY2
Sample
Days
55.1
41.4
71.4
30.08
29.18
1.4
(62)
± 1.0
± 1.1
± 1.1
±0.01
±0.01
NW
±0.1
56.9
43.0
71.1
30.05
29.15
1.3
2014
±0.4
±0.4
±0.4
± <0.01
±<0.01
NW
±<0.1
21st and Greenup, Ashland, Kentucky - ASKY-M
3
Sample
Days
51.8
41.8
71.7
30.07
29.18
4.5
(62)
± 1.0
± 1.0
± 1.0
±0.01
±0.01
ssw
±0.2
53.5
43.0
71.1
30.05
29.15
4.6
2014
±0.4
±0.4
±0.4
± <0.01
±<0.01
ssw
±0.1
Grayson, Kentucky - GLKY4
Sample
Days
52.6
43.5
74.3
30.06
29.14
2.5
(73)
±0.9
±0.9
± 1.0
±0.01
±0.03
sw
±0.1
53.6
44.1
73.7
30.05
29.14
2.6
2014
±0.4
±0.4
±0.4
± <0.01
±0.01
sw
±0.1
Baskett, Kentucky - BAKY5
Sample
Days
53.2
41.4
66.8
30.10
29.68
5.5
(64)
± 1.1
± 1.0
±0.8
±0.01
±0.01
ssw
±0.2
55.3
43.9
68.0
30.06
29.64
5.9
2014
±0.4
±0.4
±0.4
± <0.01
±<0.01
ssw
±0.1
Calvert City,
Kentucky - CCKY6
Sample
Days
57.8
46.1
65.2
30.05
29.61
2.7
(50)
± 1.3
± 1.3
± 1.0
±0.01
±0.01
sw
±0.1
Jan-Oct
59.5
48.2
66.9
30.03
29.59
2.8
2014
±0.5
±0.5
±0.4
± <0.01
±<0.01
SE
±0.1
'Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
2Temperature and wind parameters were measured at ASKY. Data for the remaining parameters were obtained from the closest
NWS weather station located at Tri-State/M.J. Ferguson Field Airport, WBAN 03860.
'Meteorological data for ASKY-M were not available in AQS. This information was obtained from the NWS weather station
located at Tri-State/M.J. Ferguson Field Airport, WBAN 03860.
4A11 meteorological parameters except sea level pressure were measured at GLKY. Data for sea level pressure were obtained
from the closest NWS weather station located at Tri-State/M.J. Ferguson Field Airport, WBAN 03860.
'Meteorological data for BAKY were not available in AQS. This information was obtained from the NWS weather station
located at Evansville Regional Airport, WBAN 93817.
"Temperature and wind parameters were measured at CCKY through the end of the sampling period. Data for the remaining
parameters were obtained from the closest NWS weather station located at Barkley Regional Airport, WBAN 03816.
7Temperature and wind parameters were measured at BLKY. Data for the remaining parameters were obtained from the closest
NWS weather station located at Barkley Regional Airport, WBAN 03816.
8Meteorological data for ATKY, LAKY, and TVKY were not available in AQS. This information was obtained from the NWS
weather station located at Barkley Regional Airport, WBAN 03816.
'Meteorological data for LEKY were not available in AQS. This information was obtained from the NWS weather station
located at Blue Grass Airport, WBAN 93820.
12-22
-------
Table 12-2. Average Meteorological Conditions near the Kentucky Monitoring Sites (Continued)
Average
Type1
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(in Hg)
Average
Station
Pressure
(in Hg)
Prevailing
Wind
Direction
Average
Scalar Wind
Speed
(kt)
Smithland, Kentucky - BLKY7
Sample
Days
54.2
43.5
66.3
30.09
29.64
3.3
(62)
± 1.1
± 1.1
±0.9
± <0.01
±<0.01
ssw
±0.1
55.7
45.9
69.0
30.05
29.61
3.3
2014
±0.4
±0.4
±0.4
± <0.01
±<0.01
s
±0.1
Barkley Regional Airport8
ATKY
55.7
43.7
66.4
30.09
29.64
5.4
(61)
± 1.1
± 1.1
±0.9
±0.01
±0.01
s
±0.2
LAKY
55.6
43.6
66.7
30.09
29.65
5.3
(62)
± 1.1
± 1.1
±0.9
±0.01
±0.01
s
±0.2
TVKY
56.3
44.4
66.8
30.08
29.63
5.4
(63)
± 1.1
± 1.1
±0.9
±0.01
±0.01
s
±0.2
57.1
45.9
69.0
30.05
29.61
5.8
2014
±0.4
±0.4
±0.4
± <0.01
±<0.01
s
±0.1
Lexington, Kentucky - LEKY9
LEKY
53.6
42.0
67.4
30.08
29.03
6.7
(67)
± 1.0
± 1.0
±0.8
±0.01
±0.01
s
±0.2
54.7
43.6
68.5
30.05
29.00
7.1
2014
±0.4
±0.4
±0.4
± <0.01
±<0.01
s
±0.1
'Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
2Temperature and wind parameters were measured at ASKY. Data for the remaining parameters were obtained from the closest
NWS weather station located at Tri-State/M.J. Ferguson Field Airport, WBAN 03860.
'Meteorological data for ASKY-M were not available in AQS. This information was obtained from the NWS weather station
located at Tri-State/M.J. Ferguson Field Airport, WBAN 03860.
4A11 meteorological parameters except sea level pressure were measured at GLKY. Data for sea level pressure were obtained
from the closest NWS weather station located at Tri-State/M.J. Ferguson Field Airport, WBAN 03860.
'Meteorological data for BAKY were not available in AQS. This information was obtained from the NWS weather station
located at Evansville Regional Airport, WBAN 93817.
"Temperature and wind parameters were measured at CCKY through the end of the sampling period. Data for the remaining
parameters were obtained from the closest NWS weather station located at Barkley Regional Airport, WBAN 03816.
7Temperature and wind parameters were measured at BLKY. Data for the remaining parameters were obtained from the closest
NWS weather station located at Barkley Regional Airport, WBAN 03816.
8Meteorological data for ATKY, LAKY, and TVKY were not available in AQS. This information was obtained from the NWS
weather station located at Barkley Regional Airport, WBAN 03816.
'Meteorological data for LEKY were not available in AQS. This information was obtained from the NWS weather station
located at Blue Grass Airport, WBAN 93820.
12-23
-------
12.2.2 Wind Rose Comparison
Hourly surface wind data were uploaded into a wind rose software program to produce
customized wind roses, as described in Section 3.4.2. A wind rose shows the frequency of wind
directions using "petals" positioned around a 16-point compass, and uses different colors to
represent wind speeds. Figure 12-16 presents two wind roses for the ASKY monitoring site. The
first is a wind rose representing wind observations for all of 2014 and the second is a wind rose
representing wind observations for days on which samples were collected in 2014. These can be
used to identify the predominant wind speed and direction for 2014 and to determine if wind
observations on sample days were representative of conditions experienced over the entire year.
Figures 12-17 through 12-23 present the full-year and sample day wind roses for the remaining
Kentucky sites.
Figure 12-16. Wind Roses for the Wind Data Collected at ASKY
2014 Wind Rose Sample Day Wind Rose
WIND SPEED
(Knots)
I I >-22
¦ 17-21
11-17
¦ 7-11
~ 4-7
H 1
Cams: 35.97%
i2%*
1" Tea st '
W1NDSFEED
(Knots)
~ >-22
¦ 17-21
11-17
¦ 7-11
~ 4-7
1
Cairns: 34.21%
Observations from Figure 12-16 for ASKY include the following:
• The 2014 wind rose shows that calm winds accounted for more than one-third of
observations at ASKY. Winds from the western quadrants were more frequently
observed than those from the eastern quadrants. Northwesterly winds were observed
the most, although winds from the south-southeast to north-northwest each accounted
for more than 4 percent of the observations. Wind speeds less than 4 knots were
frequently observed near ASKY, while winds greater than 11 knots were rarely
observed, but most often observed with westerly winds.
NORTH
SOUTH
12-24
-------
• The sample day wind rose also exhibits light wind patterns. Northwesterly winds
were also prevalent on samples days. Fewer winds from the southwest quadrant were
observed on samples days while a higher percentage of winds from the northwest
quadrant were observed.
Figure 12-17. Wind Roses for the Tri-State/M.J. Ferguson Field Airport Weather Station
near ASKY-M
2014 Wind Rose
Sample Day Wind Rose
WIND SPEED
(Knots)
I I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 24.05%
WIND SPEED
(Knots)
I I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 25.27%
Observations from Figure 12-17 for ASKY-M include the following:
• The Tri-State/M.J. Ferguson Field weather station is the closest weather station to
ASKY-M. The weather station is located 8.7 miles south-southeast of ASKY-M. This
weather station is located in West Virginia, south of the Ohio River and east of the
Big Sandy River.
• The full-year wind rose shows that south-southwesterly winds were the most
frequently observed winds near ASKY-M. Winds from the southwest quadrant,
including winds from the south, accounted for the majority of observations. Calm
winds were observed for about one-quarter of the observations, and wind speeds
greater than 11 knots were more commonly observed with winds from the western
quadrants than the eastern quadrants.
• While south-southwesterly winds were also prevalent on sample days, they accounted
for fewer observations. Compared to the full-year, winds from the south to southwest
accounted for fewer observations on sample days while winds from the west-
northwest to north-northwest and northeast accounted for additional observations.
12-25
-------
Figure 12-18. Wind Roses for the Wind Data Collected at GLKY
2014 Wind Rose Sample Day Wind Rose
¦ " NORTH NORTH
WEST
EAST.
WIND SPEED
(Knots)
¦I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 21.91%
WIND SPEED
(Knots)
¦I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 23.81%
Observations from Figure 12-18 for GLKY include the following:
• The full-year wind rose for GLKY shows that southwesterly winds were prevalent at
GLKY, accounting for nearly 20 percent of the observations collected at this site.
Together, winds from the south-southwest to west-southwest account nearly
40 percent of observations. Winds from the north-northeast and northeast represent a
secondary group of wind observations. Winds from the northwest quadrant were
infrequently observed and winds from the southeast quadrant were rarely observed.
Calm winds accounted for nearly 22 percent of the wind observations and wind
speeds greater than 11 knots were observed most often with southwesterly winds.
• The wind patterns on the sample day wind rose generally resemble those on the full-
year wind rose, although there are a few differences. Fewer winds from the southwest
quadrant were observed on sample days, while a higher percentage of winds from the
northwest quadrant were observed (although still representing a relatively small
percentage of the observations).
12-26
-------
Figure 12-19. Wind Roses for the Evansville Regional Airport Weather Station near BAKY
2014 Wind Rose
Sample Day Wind Rose
WIND SPEED
(Knots)
¦I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 19.71%
WEST
WIND SPEED
(Knots)
¦I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 23.89%
Observations from Figure 12-19 for BAKY include the following:
• The Evansville Regional Airport weather station is located approximately 12 miles
north-northwest of BAKY. This weather station is in Ohio, with most of the city of
Evansville between the site and the station.
• The 2014 wind rose shows that winds from a variety of directions were observed near
BAKY, although winds from the western quadrants and due north and due south were
observed most often while winds from the eastern quadrants were observed less often.
Calm winds account for just less than 20 percent of the observations.
• The sample day wind rose for BAKY shares some similarities with the full-year wind
rose, but exhibits some differences as well. Although south-southwesterly winds were
still prevalent, northerly winds were observed nearly as often. In addition, winds from
the northwest to northeast accounted for a higher percentage of observations on
sample days while winds from the south and southwest to west were observed less
often. Calm winds accounted for nearly 24 percent of observations on sample days.
12-27
-------
Figure 12-20. Wind Roses for the Wind Data Collected at CCKY
Jan - Oct 2014 Wind Rose Sample Day Wind Rose
4%
EAST.
WIND SPEED
(Knots)
¦I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 12.85%
WIND SPEED
(Knots)
¦I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 14.29%
Observations from Figure 12-20 for CCKY include the following:
• The full-year wind rose for CCKY shows that winds from the southern quadrants
were observed more often at this site than those from the northern quadrants.
Southeasterly winds prevailed near CCKY, and along with winds from the south-
southeast to southwest, account for 40 percent of the wind observations. Conversely,
winds from the north-northeast to east were infrequently observed. Calm winds
accounted for just less than 13 percent of observations at CCKY.
• The sample day wind rose resembles the full-year wind rose in that winds from the
southern quadrants were observed more often than those from the northern quadrants.
However, winds from the southeast to southwest were more evenly distributed, and
winds from the southwest account for slightly higher percentage of observations on
sample days. In addition, a higher percentage of winds from the northeast to east is
shown on the sample day wind rose, while fewer winds from the northwest were
observed.
• Recall that sampling was discontinued at CCKY in October, thus, the wind roses
presented here include wind observations through the end of the sampling period
only.
12-28
-------
Figure 12-21. Wind Roses for the Wind Data Collected at BLKY
2014 Wind Rose
Sample Day Wind Rose
WIND SPEED
(Knots)
¦I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 13.34%
WIND SPEED
(Knots)
¦I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 14.17%
Observations from Figure 12-21 for BLKY include the following:
• The full-year wind rose shows that winds from all directions were observed at BLKY,
with winds from the south and south-southwest observed most often, accounting for
10 percent and 9 percent of observations, respectively. Winds from the northwest
quadrant, including north, were also commonly observed, each direction accounting
for at least 6 percent of observations. This is also true for southeasterly and
southwesterly winds. Winds from the east and east-southeast and south-southwest and
west were observed least often. Calm winds account for 13 percent of the hourly
measurements while the strongest winds were observed with winds from the south
and south-southwest.
• The wind patterns on the sample day wind rose generally resemble those on the full-
year wind rose, although there are some differences. For instance, winds from the
north to east-northeast account for a higher percentage of observations on sample
days.
12-29
-------
Figure 12-22. Wind Roses for the Barkley Regional Airport Weather Station near ATKY,
LAKY, and TVKY
2014 Wind Rose
Sample Day Wind Rose for ATKY
EAST
WIND SPEED
(Knots)
HI >= 22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
HI 1-4
Calms: 20.82%
EAST,
WIND SPEED
(Knots)
¦I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 24.11%
Sample Day Wind Rose for LAKY
Sample Day Wind Rose for TVKY
/ (A
iWEST
EAST
WIND SPEED
(Knots)
~ »22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
HI 1-4
Calms: 24.53%
WIND SPEED
(Knots)
m >=22
¦ 17-21
¦ 11-17
¦ 7-11
I I 4-7
¦ 1
Calms: 23.41%
12-30
-------
Observations from Figure 12-22 for ATKY, LAKY, and TVKY include the following:
• The Barkley Regional Airport weather station is the closest weather station to the
sites in and near Calvert City. The weather station is located between 23 miles and
25 miles west of the Calvert City monitoring sites and just west of the Paducah metro
area.
• The full-year wind rose shows that winds from the south, southwest quadrant, and
north account for the majority of wind observations near these sites, although calm
winds account for nearly 21 percent of the hourly measurements.
• The sample day wind roses for ATKY, LAKY, and TVKY resemble each other as
well as the full-year wind rose. For each site, the sample day wind rose shows that
southerly winds were prevalent on sample days near each site, with winds from the
south, southwest quadrant, and north accounting for the highest percentage of wind
observations. Calm winds accounted for 23 percent to 25 percent of the wind
observations on sample days near each site while the strongest winds were most often
observed with winds from the northwest quadrant.
Figure 12-23. Wind Roses for the Blue Grass Airport Weather Station near LEKY
2014 Wind Rose Sample Day Wind Rose
SOUTH
WIND SPEED
(Knots)
I I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 10.05%
WIND SPEED
(Knots)
I I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 12.38%
Observations from Figure 12-23 for LEKY include the following:
• The Blue Grass Airport weather station is located approximately 6 miles west-
southwest of the LEKY monitoring site. The airport is located on the western edge of
the Lexington metro area.
12-31
-------
• The full-year wind rose shows that winds from the south, southwest quadrant, and
west account for the majority of wind observations near LEKY, particularly winds
from the south, which account for approximately 14 percent of observations. Winds
from most of the other directions account for roughly 5 percent of wind observations
or less each. Calm winds accounted for 10 percent of the hourly measurements in
2014.
• The wind patterns on the sample day wind rose for LEKY resemble the wind patterns
shown on the full-year wind rose, although there are some differences. Winds from
the south and south-southwest accounted for fewer observations on sample days
while winds from the northwest quadrant and northeast accounted for a higher
percentage of observations.
12.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each
Kentucky monitoring 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. 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 12-3 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 12-4, with the pollutants of interest for each site shaded in gray.
12-32
-------
Table 12-3. Overview of Pollutant Groups Sampled for at the
Kentucky Monitoring Sites
Site
VOCs
Carbonyl
Compounds
PAHs
PMio
Metals
ASKY
—
—
ASKY-M
—
—
—
V
GLKY
V
BAKY
—
—
—
V
ATKY
—
—
—
BLKY1
V
—
—
CCKY1
V
—
—
LAKY
V
—
—
—
TVKY
V
—
—
—
LEKY
--
-- = This pollutant group was not sampled for at this site.
BOLD ITALICS = EPA-designated NATTS Site
1 Sampling at CCKY was discontinued in October 2014 and the metals instrumentation
moved to BLKY, where sampling resumed.
Observations from Table 12-3 include the following:
• Carbonyl compounds, VOCs, PAHs, and PMio metals were sampled for at GLKY
throughout 2014.
• Additional sites sampling PMio metals include ASKY-M, BAKY, BLKY, CCKY,
and LEKY.
• Additional sites sampling VOCs include ASKY, ATKY, BLKY, CCKY, LAKY,
TVKY, and LEKY.
• Additional sites sampling carbonyl compounds include ASKY and LEKY.
• No additional sites sampled PAHs.
• Sampling at the CCKY site was discontinued in early October 2014 and the
metals instrumentation was moved to BLKY site. The first metals sample was
collected at BLKY on October 20, 2014.
12-33
-------
Table 12-4. 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
Acetaldehyde
0.45
61
61
100.00
16.80
16.80
Benzene
0.13
61
61
100.00
16.80
33.61
Carbon Tetrachloride
0.17
61
61
100.00
16.80
50.41
Formaldehyde
0.077
61
61
100.00
16.80
67.22
1,2-Dichloroethane
0.038
58
58
100.00
15.98
83.20
1.3 -Butadiene
0.03
53
56
94.64
14.60
97.80
Hexacliloro -1,3 -butadiene
0.045
5
6
83.33
1.38
99.17
/?-Dichlorobcnzcnc
0.091
2
23
8.70
0.55
99.72
Ethylbenzene
0.4
1
61
1.64
0.28
100.00
Total
363
448
81.03
21st and Greenup, Ashland, Kentucky - ASKY-M
Arsenic (PMio)
0.00023
55
58
94.83
58.51
58.51
Nickel (PMio)
0.0021
18
59
30.51
19.15
77.66
Manganese (PMio)
0.03
12
59
20.34
12.77
90.43
Cadmium (PMio)
0.00056
5
59
8.47
5.32
95.74
Lead (PMio)
0.015
4
59
6.78
4.26
100.00
Total
94
294
31.97
Grayson, Kentucky - GLKY
Formaldehyde
0.077
61
61
100.00
16.94
16.94
Acetaldehyde
0.45
58
61
95.08
16.11
33.06
Benzene
0.13
56
56
100.00
15.56
48.61
Carbon Tetrachloride
0.17
56
56
100.00
15.56
64.17
1,2-Dichloroethane
0.038
48
48
100.00
13.33
77.50
1.3 -Butadiene
0.03
38
47
80.85
10.56
88.06
Arsenic (PMio)
0.00023
36
55
65.45
10.00
98.06
Hexacliloro -1,3 -butadiene
0.045
4
5
80.00
1.11
99.17
Naphthalene
0.029
3
58
5.17
0.83
100.00
Total
360
447
80.54
Baskett, Kentucky - BAKY
Arsenic (PMio)
0.00023
54
57
94.74
96.43
96.43
Nickel (PMio)
0.0021
2
57
3.51
3.57
100.00
Total
56
114
49.12
12-34
-------
Table 12-4. Risk-Based Screening Results for the Kentucky Monitoring Sites (Continued)
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Atmos Energy, Calvert City, Kentucky - ATKY
Benzene
0.13
61
61
100.00
21.86
21.86
Carbon Tetrachloride
0.17
61
61
100.00
21.86
43.73
1,2-Dichloroethane
0.038
60
60
100.00
21.51
65.23
1.3 -Butadiene
0.03
49
53
92.45
17.56
82.80
Vinyl chloride
0.11
28
36
77.78
10.04
92.83
Hexacliloro -1,3 -butadiene
0.045
14
14
100.00
5.02
97.85
1.1,2-Trichloroethane
0.0625
3
3
100.00
1.08
98.92
1,2-Dibromoethane
0.0017
2
2
100.00
0.72
99.64
T richloroethylene
0.2
1
17
5.88
0.36
100.00
Total
279
307
90.88
Smithland, Kentucky - BLKY
Benzene
0.13
60
60
100.00
23.08
23.08
Carbon Tetrachloride
0.17
59
60
98.33
22.69
45.77
1,2-Dichloroethane
0.038
57
57
100.00
21.92
67.69
1.3 -Butadiene
0.03
39
47
82.98
15.00
82.69
Vinyl chloride
0.11
19
33
57.58
7.31
90.00
Hexacliloro -1,3 -butadiene
0.045
13
13
100.00
5.00
95.00
Arsenic (PMio)
0.00023
8
11
72.73
3.08
98.08
1,1,2-Trichloroethane
0.0625
4
4
100.00
1.54
99.62
1,2-Dibromoethane
0.0017
1
1
100.00
0.38
100.00
Total
260
286
90.91
Calvert City Elementary School, Calvert City, Kentucky - CCKY
Benzene
0.13
46
46
100.00
19.74
19.74
Carbon Tetrachloride
0.17
46
46
100.00
19.74
39.48
1,2-Dichloroethane
0.038
46
46
100.00
19.74
59.23
1,3-Butadiene
0.03
38
41
92.68
16.31
75.54
Arsenic (PMio)
0.00023
35
39
89.74
15.02
90.56
Vinyl chloride
0.11
10
27
37.04
4.29
94.85
Hexacliloro -1,3 -butadiene
0.045
9
9
100.00
3.86
98.71
1,2-Dibromoethane
0.0017
1
1
100.00
0.43
99.14
Ethylbenzene
0.4
1
46
2.17
0.43
99.57
1.1,2-Trichloroethane
0.0625
1
1
100.00
0.43
100.00
Total
233
302
77.15
12-35
-------
Table 12-4. Risk-Based Screening Results for the Kentucky Monitoring Sites (Continued)
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Lazy Daze, Calvert City, Kentucky - LAKY
Benzene
0.13
56
56
100.00
22.22
22.22
Carbon Tetrachloride
0.17
56
56
100.00
22.22
44.44
1,2-Dichloroethane
0.038
56
56
100.00
22.22
66.67
1.3 -Butadiene
0.03
47
49
95.92
18.65
85.32
Vinyl chloride
0.11
16
33
48.48
6.35
91.67
Hexacliloro -1,3 -butadiene
0.045
15
15
100.00
5.95
97.62
1.1,2-Trichloroethane
0.0625
3
4
75.00
1.19
98.81
1,2-Dibromoethane
0.0017
2
2
100.00
0.79
99.60
/?-Dichlorobcnzcnc
0.091
1
19
5.26
0.40
100.00
Total
252
290
86.90
TVA Substation, Calvert City, Kentucky - TVKY
Benzene
0.13
61
61
100.00
19.93
19.93
Carbon Tetrachloride
0.17
61
61
100.00
19.93
39.87
1,2-Dichloroethane
0.038
61
61
100.00
19.93
59.80
1.3 -Butadiene
0.03
53
55
96.36
17.32
77.12
Vinyl chloride
0.11
28
42
66.67
9.15
86.27
Hexacliloro -1,3 -butadiene
0.045
19
20
95.00
6.21
92.48
1.1,2-Trichloroethane
0.0625
15
15
100.00
4.90
97.39
1,2-Dibromoethane
0.0017
4
4
100.00
1.31
98.69
/j-Dichlorobcnzcne
0.091
2
25
8.00
0.65
99.35
1,1 -Dichloroethane
0.625
1
15
6.67
0.33
99.67
T richloroethylene
0.2
1
18
5.56
0.33
100.00
Total
306
377
81.17
Lexin
gton, Kentucky - LEKY
Benzene
0.13
58
58
100.00
14.25
14.25
Carbon Tetrachloride
0.17
58
58
100.00
14.25
28.50
1,2-Dichloroethane
0.038
57
57
100.00
14.00
42.51
Acetaldehyde
0.45
55
55
100.00
13.51
56.02
Formaldehyde
0.077
55
55
100.00
13.51
69.53
1,3-Butadiene
0.03
54
56
96.43
13.27
82.80
Arsenic (PMio)
0.00023
50
56
89.29
12.29
95.09
Hexacliloro -1,3 -butadiene
0.045
12
12
100.00
2.95
98.03
/j-Dichlorobcnzcne
0.091
4
26
15.38
0.98
99.02
Ethylbenzene
0.4
3
58
5.17
0.74
99.75
Propionaldehyde
0.8
1
55
1.82
0.25
100.00
Total
407
546
74.54
12-36
-------
Observations for the Ashland sites from Table 12-4 include the following:
• The number of pollutants failing screens varied significantly among the monitoring
sites; this is expected given the different pollutants measured at each site, as shown in
Table 12-3. VOCs and carbonyl compounds were sampled for at ASKY while PMio
metals were sampled for at ASKY-M.
• Concentrations of nine pollutants failed at least one screen for ASKY, with 81 percent
of concentrations for these nine pollutants greater than their associated risk screening
value (or failing screens).
• Six pollutants contributed to 95 percent of failed screens for ASKY and therefore
were identified as pollutants of interest. These six include two carbonyl compounds
and four VOCs.
• Concentrations of five metals failed at least one screen for ASKY-M, with 32 percent
of concentrations for these five pollutants greater than their associated risk screening
value (or failing screens).
• Four metals contributed to 95 percent of failed screens for ASKY-M and therefore
were identified as pollutants 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 (S4MO is the other site).
Observations for GLKY from Table 12-4 include the following:
• All four pollutant groups shown in Table 12-3 were sampled for at GLKY.
• Concentrations of nine pollutants failed at least one screen for GLKY, with nearly
81 percent of concentrations for these nine 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 12-4 include the following:
• BAKY sampled for PMio metals only.
• Concentrations of arsenic and nickel failed at least one screen for BAKY, with
49 percent of concentrations for these two pollutants greater than their associated risk
screening value (or failing screens).
• Arsenic contributed to 96 percent of the failed screens for BAKY and therefore was
identified as BAKY's sole pollutant of interest.
12-37
-------
Observations for the Calvert City sites from Table 12-4 include the following:
• VOCs were sampled for at all five Calvert City sites. PMio metals were also sampled
for at CCKY until October and then at BLKY afterwards.
• The number of pollutants whose concentrations were greater than their associated risk
screening value varied from nine (3 sites) to 11 (TVKY).
• Concentrations of nine VOCs failed screens for ATKY, and six of these contributed
to 95 percent of failed screens for ATKY and thus, were identified as pollutants of
interest for this site.
• Concentrations of nine pollutants failed screens for BLKY, with six of the VOCs
contributed to 95 percent of failed screens for BLKY and thus, were identified as
pollutants of interest for this site.
• Concentrations of 10 pollutants failed screens for CCKY, and six VOCs and arsenic
contributing to 95 percent of failed screens for CCKY; thus, these seven pollutant
were identified as pollutants of interest for this site.
• Concentrations of nine VOCs failed screens for LAKY, and six of these contributed
to 95 percent of failed screens for LAKY and thus, were identified as pollutants of
interest for this site.
• Concentrations of 11 VOCs failed screens for TVKY, and seven of these 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, vinyl chloride, and
hexachloro-1,3-butadiene were identified as pollutants of interest for all five Calvert
City sites. These sites are the only NMP sites with vinyl chloride as a pollutant of
interest.
Observations for LEKY from Table 12-4 include the following:
• Carbonyl compounds, VOCs, and PMio metals were sampled for at LEKY.
• Concentrations of 11 pollutants failed at least one screen for LEKY, with nearly
75 percent of concentrations of these 11 pollutants greater than their associated risk
screening value (or failing screens).
• Seven pollutants contributed to 95 percent of failed screens for LEKY and therefore
were identified as pollutants of interest. These include two carbonyl compounds, four
VOCs, and one metal.
12-38
-------
12.4 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.
• 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 are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at the Kentucky monitoring sites are provided in Appendices J, L, M, and N.
12.4.1 2014 Concentration Averages
Quarterly and annual concentration averages were calculated 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 the entire year of sampling. Annual averages were calculated for
pollutants where three valid quarterly averages could be calculated 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 12-5, 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.
12-39
-------
Table 12-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Kentucky Monitoring Sites
Pollutant
# of
Measured
Detections
vs.
# >MDL
# of
Samples
1st
Quarter
Average
(jig/m3)
2nd
Quarter
Average
(jig/m3)
3rd
Quarter
Average
(jig/m3)
4th
Quarter
Average
(jig/m3)
Annual
Average
(jig/m3)
Health De
jartment, Ashland, Kentucky - ASKY
Acetaldehyde
61/61
61
1.23
±0.17
1.48
±0.27
1.19
±0.21
1.10
±0.19
1.25
±0.11
Benzene
61/61
61
0.85
±0.15
0.62
±0.15
1.47
± 1.62
0.56
±0.10
0.87
±0.39
1.3 -Butadiene
56/53
61
0.07
±0.02
0.04
±0.02
0.06
±0.02
0.06
±0.01
0.06
±0.01
Carbon Tetrachloride
61/61
61
0.63
±0.05
0.66
±0.03
0.67
±0.02
0.62
±0.03
0.65
±0.02
1,2-Dichloroethane
58/58
61
0.08
±0.01
0.08
±0.01
0.05
±0.02
0.08
±0.01
0.07
±0.01
Formaldehyde
61/61
61
1.71
±0.24
3.00
±0.63
3.26
±0.67
1.26
±0.23
2.29
±0.31
21st and Greenup, Ashland, Kentucky - ASKY-M
Arsenic (PMi0)a
58/55
59
0.78
±0.20
0.98
±0.38
1.91
± 1.29
0.85
±0.49
1.14
±0.36
Cadmium (PMi0)a
59/59
59
0.22
±0.06
0.26
±0.08
0.33
±0.13
0.13
±0.05
0.24
±0.04
Manganese (PMi0)a
59/59
59
15.73
±5.92
18.04
±5.98
23.42
±9.30
15.52
±7.32
18.22
±3.53
Nickel (PMi,;,)a
59/59
59
2.89
± 1.69
2.23
±0.89
2.23
±0.69
1.44
± 1.22
2.19
±0.56
Grayson, Kentucky - GLKY
Acetaldehyde
61/61
61
0.96
±0.18
0.99
±0.16
0.73
±0.10
0.77
±0.15
0.86
±0.08
Benzene
56/56
56
0.59
±0.08
NA
0.32
±0.04
0.43
±0.15
0.42
±0.05
1.3 -Butadiene
47/40
56
0.04
±0.01
NA
0.04
±0.01
0.04
±0.01
0.04
±0.01
Carbon Tetrachloride
56/56
56
0.58
±0.07
NA
0.65
±0.02
0.57
±0.04
0.61
±0.02
1,2-Dichloroethane
48/45
56
0.08
±<0.01
NA
0.03
±0.02
0.07
±0.01
0.06
±0.01
Formaldehyde
61/61
61
0.77
±0.15
1.79
±0.35
2.39
±0.45
0.95
±0.18
1.49
±0.22
Arsenic (PMi0)a
55/36
59
0.29
±0.13
0.36
±0.14
0.56
±0.21
0.36
±0.16
0.39
±0.08
Baskett, Kentucky - BAKY
Arsenic (PMi0)a
57/54
58
0.52
±0.14
0.91
±0.33
1.26
±0.33
0.74
±0.45
0.85
±0.17
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
12-40
-------
Table 12-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Kentucky Monitoring Sites (Continued)
Pollutant
# of
Measured
Detections
vs.
# >MDL
# of
Samples
1st
Quarter
Average
(jig/m3)
2nd
Quarter
Average
(jig/m3)
3rd
Quarter
Average
(jig/m3)
4th
Quarter
Average
(jig/m3)
Annual
Average
(jig/m3)
Atmos Energy, Calvert City, Kentucky - ATKY
Benzene
61/61
61
0.79
±0.21
0.38
±0.09
0.78
±0.17
0.52
±0.08
0.62
±0.08
1.3 -Butadiene
53/49
61
0.09
±0.09
0.09
±0.07
0.11
±0.05
0.06
±0.02
0.09
±0.03
Carbon Tetrachloride
61/61
61
0.69
±0.10
0.71
±0.08
0.70
±0.03
0.63
±0.02
0.68
±0.03
1,2-Dichloroethane
60/59
61
0.51
±0.60
0.52
±0.32
1.01
±0.52
0.30
±0.27
0.58
±0.22
Hexachloro-1,3 -butadiene
14/0
61
0.02
±0.02
0.01
±0.02
0.03
±0.02
0.02
±0.02
0.02
±0.01
Vinyl chloride
36/33
61
0.64
±0.74
1.29
± 1.36
0.72
±0.51
0.44
±0.49
0.77
±0.40
Smithland, Kentucky - BLKY
Benzene
60/60
60
0.75
±0.24
0.73
±0.44
0.40
±0.10
0.60
±0.22
0.62
±0.13
1.3 -Butadiene
47/41
60
0.08
±0.07
0.47
±0.67
0.04
±0.02
0.10
±0.08
0.17
±0.16
Carbon Tetrachloride
60/60
60
0.64
±0.08
0.90
±0.31
0.68
±0.03
0.71
±0.16
0.73
±0.09
1,2-Dichloroethane
57/57
60
1.29
± 1.61
0.96
±0.61
0.62
±0.74
0.42
±0.27
0.81
±0.43
Hexachloro-1,3 -butadiene
13/0
60
0.02
±0.02
0.01
±0.02
0.03
±0.03
0.01
±0.02
0.02
±0.01
Vinyl chloride
33/32
60
0.27
±0.28
0.16
±0.08
0.05
±0.05
0.10
±0.06
0.14
±0.07
Calvert City Elementary School, Calvert City, Kentucky - CCKY
Benzene
46/46
46
0.66
±0.11
0.37
±0.08
0.67
±0.17
NA
0.56
±0.08
1.3 -Butadiene
41/38
46
0.08
±0.03
0.04
±0.02
0.13
±0.06
NA
0.08
±0.03
Carbon Tetrachloride
46/46
46
0.69
±0.07
0.69
±0.03
0.72
±0.06
NA
0.70
±0.03
1,2-Dichloroethane
46/45
46
0.44
±0.28
0.24
±0.14
0.83
±0.34
NA
0.49
±0.16
Hexachloro-1,3 -butadiene
9/0
46
0.02
±0.02
0.01
±0.02
0.02
±0.02
NA
0.02
±0.01
Vinyl chloride
27/26
46
0.09
±0.08
0.03
±0.03
0.12
±0.08
NA
0.08
±0.04
Arsenic (PMi0)a
39/35
41
0.38
±0.16
0.49
±0.23
0.77
±0.21
NA
0.55
±0.12
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
12-41
-------
Table 12-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Kentucky Monitoring Sites (Continued)
Pollutant
# of
Measured
Detections
vs.
# >MDL
# of
Samples
1st
Quarter
Average
(jig/m3)
2nd
Quarter
Average
(jig/m3)
3rd
Quarter
Average
(jig/m3)
4th
Quarter
Average
(jig/m3)
Annual
Average
(jig/m3)
Lazy Daze, Calvert City, Kentucky - LAKY
Benzene
56/56
56
0.79
±0.32
0.59
±0.40
0.86
±0.22
0.59
±0.11
0.70
±0.13
1.3 -Butadiene
49/48
56
0.12
±0.10
0.05
±0.03
0.19
±0.12
0.07
±0.02
0.11
±0.04
Carbon Tetrachloride
56/56
56
0.77
±0.27
0.68
±0.05
0.74
±0.10
0.65
±0.02
0.71
±0.06
1,2-Dichloroethane
56/55
56
1.08
± 1.09
0.35
±0.26
1.35
±0.53
1.08
± 1.52
0.97
±0.47
Hexachloro-1,3 -butadiene
15/0
56
0.01
±0.02
0.02
±0.02
0.05
±0.03
0.01
±0.02
0.02
±0.01
Vinyl chloride
33/31
56
0.28
±0.28
0.04
±0.04
0.16
±0.08
0.06
±0.05
0.13
±0.07
TVA Substation, Calvert City, Kentucky - TVKY
Benzene
61/61
61
0.98
±0.35
0.42
±0.12
1.87
± 1.21
0.88
±0.44
1.04
±0.35
1.3 -Butadiene
55/53
61
0.21
±0.15
0.30
±0.36
0.86
±0.76
0.13
±0.09
0.38
±0.22
Carbon Tetrachloride
61/61
61
0.93
±0.32
0.76
±0.08
0.89
±0.27
0.91
±0.35
0.87
±0.13
1,2-Dichloroethane
61/61
61
6.81
±5.83
1.06
± 1.09
3.98
±2.93
2.64
±2.38
3.54
± 1.66
Hexachloro-1,3 -butadiene
20/0
61
0.04
±0.02
0.04
±0.03
0.03
±0.02
0.02
±0.02
0.03
±0.01
1.1,2 -T richloroethane
15/13
61
0.03
±0.03
0.01
±0.02
0.05
±0.03
0.02
±0.03
0.03
±0.01
Vinyl chloride
42/41
61
0.66
±0.81
1.57
±2.93
0.32
±0.17
0.17
±0.14
0.69
±0.76
Lexington, Kentucky - LEKY
Acetaldehyde
55/55
55
1.82
±0.27
1.64
±0.23
1.53
±0.26
1.20
±0.26
1.55
±0.13
Benzene
58/58
58
0.73
±0.08
0.53
±0.08
0.64
±0.11
0.58
±0.06
0.62
±0.04
1.3 -Butadiene
56/55
58
0.05
±0.02
0.05
±0.01
0.09
±0.01
0.07
±0.02
0.07
±0.01
Carbon Tetrachloride
58/58
58
0.58
±0.05
0.66
±0.03
0.60
±0.03
0.62
±0.02
0.61
±0.02
1,2-Dichloroethane
57/54
58
0.08
±<0.01
0.09
±0.02
0.07
±0.01
0.08
±0.01
0.08
±0.01
Formaldehyde
55/55
55
1.96
±0.44
4.98
±3.37
3.58
±0.59
1.86
±0.61
3.15
±0.90
Arsenic (PMi0)a
56/51
56
0.68
±0.15
0.47
±0.21
0.81
±0.26
0.73
±0.24
0.67
±0.11
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
12-42
-------
Observations for the Ashland sites from Table 12-5 include the following:
• VOCs and carbonyl compounds were sampled for at ASKY and PMio metals were
sampled for at ASKY-M. Thus, these sites have no pollutants of interest in common.
• With the exception of 1,3-butadiene and 1,2-dichloroethane, each of the pollutants of
interest for ASKY was detected in all the valid VOC samples collected. However,
even these two pollutants were detected in a majority of the samples collected.
• The pollutants of interest with the highest annual average concentrations for ASKY
are formaldehyde (2.29 ± 0.31 |ig/m3), acetaldehyde (1.25 ±0.11 |ig/m3), and
benzene (0.87 ± 0.39 |ig/m3). The annual average concentrations for the two carbonyl
compounds are similar to the annual averages for 2013 while the annual average
concentration of benzene for 2014 is approximately half of its 2013 annual average.
Note the relatively high confidence interval for the annual average concentration of
benzene.
• The second and third quarter average concentrations of formaldehyde are
significantly higher than the other quarterly averages for this pollutant for ASKY. A
review of the data shows that all 15 formaldehyde concentrations greater than or
equal to 3 |ig/m3 were measured at ASKY during the second and third quarters of
2014 while all but one of the 17 concentrations less than 1.5 |ig/m3 were all measured
during the first or fourth quarters of 2014 (and the exception was measured on the
first sample day in April).
• The third quarter average benzene concentration is two to three times greater than the
other quarterly averages and has a confidence interval greater than the average itself.
This indicates that outliers may be affecting this quarterly average. A review of the
data shows that the maximum benzene concentration was measured at ASKY on
July 22, 2014 (12.4 |ig/m3) and is the maximum concentration of benzene measured
across the program. The maximum benzene concentration for 2013 was also
measured at this site. This measurement is more than seven times greater than the
next highest benzene concentrations measured at this site (1.74 |ig/m3), which was
also measured during the third quarter. Only one other benzene concentration
measured at ASKY during the third quarter is greater than 1.0 |ig/m3, with the
remaining concentrations ranging from 0.295 |ig/m3 to 1.01 |ig/m3 and a median
concentration of 0.679 |ig/m3 for the quarter. This explains the large confidence
interval associated with this quarterly average concentration.
• Table 4-9 presents the NMP sites with the 10 highest annual average concentrations
for each of the program-level VOC pollutants of interest. This table shows that ASKY
has the ninth highest annual average concentration of benzene calculated across the
program. This site has the largest confidence interval among the sites shown,
indicating that this annual average is influenced by outliers while most of the other
annual averages likely run higher on a more consistent basis. Excluding the maximum
concentration from the calculation would result in an annual average concentration
for ASKY in the middle of the site-specific annual average concentrations of benzene
and a much smaller confidence interval.
12-43
-------
With the exception of arsenic, the metal pollutants of interest were detected in all of
the valid samples collected at ASKY-M. Arsenic was detected in all but one of
samples collected.
The pollutant of interest with the highest annual average concentration for ASKY-M
is manganese (18.22 ± 3.53 ng/m3), followed by nickel (2.19 ± 0.56 ng/m3) and
arsenic (1.14 ± 0.36 ng/m3), with the annual average concentration of cadmium
considerably lower (0.24 ± 0.04 ng/m3).
Several of the quarterly average concentrations for the pollutants of interest for
ASKY-M have relatively large confidence intervals, indicating the concentrations
measured at ASKY-M are highly variable. Concentrations of nickel range from
0.260 ng/m3 to 9.64 ng/m3, with a median concentration of 1.41 ng/m3, considerably
less than the annual average concentration. Both the first and fourth quarter averages
of nickel have relatively large confidence intervals associated with them. All but one
of the samples collected at ASKY-M during the fourth quarter were less than 2 ng/m3,
yet the maximum nickel concentration (9.64 ng/m3) was also measured during the
fourth quarter. This explains the relatively large confidence interval shown for the
fourth quarter. The first quarter average concentration has the second highest number
of nickel concentrations less than 1 ng/m3, but the second, third, and fourth highest
concentrations, each falling between 7 ng/m3 and 9 ng/m3, were also measured during
the first quarter of 2014.
Concentrations of manganese measured at ASKY-M range from 1.77 ng/m3 to
67.5 ng/m3, with a median concentration of 14.8 ng/m3. Several of the highest
manganese concentrations across the program were measured at ASKY-M, including
two measurements greater than 50 ng/m3, which were measured on back-to-back
sample days (September 26, 2014 and October 2, 2014; note that construction was
noted near the monitoring site on September 26, 2014). Several manganese
concentrations greater than 25 ng/m3 were measured at ASKY-M during each of the
calendar quarters, from two measured during the fourth quarter, three measured
during the first quarter, and five each measured during the second and third quarters.
Conversely, at least one manganese concentration less than 5 ng/m3 was also
measured during each calendar quarter. This explains the relatively large confidence
intervals shown for each quarterly average of manganese in Table 12-5.
Concentrations of arsenic measured at ASKY-M span two orders of magnitude,
ranging from 0.14 ng/m3 to 10.1 ng/m3 plus one non-detect, with a median
concentration of 0.71 ng/m3. The maximum arsenic concentration measured at
ASKY-M is the maximum arsenic concentration measured across the program and is
roughly three times higher than the second highest concentration measured at this site
(3.58 ng/m3). ASKY-M has the highest number of arsenic measurements greater than
2 ng/m3 compared to any other NMP site (6). The third quarter average concentration
of arsenic is considerably higher than the other quarterly averages and has a relatively
large confidence interval associated with it. Of the 20 arsenic concentrations greater
than 1 ng/m3, nine were measured during the third quarter, including three of the four
highest arsenic concentrations measured at this site.
12-44
-------
• Concentrations of cadmium measured at ASKY-M range from 0.010 ng/m3 to
0.79 ng/m3, with a median concentration of 0.175 ng/m3. The third quarter average
concentration of cadmium is greater than the other quarterly averages and the
associated confidence interval is the highest of the four. A review of the data shows
that four of the five highest cadmium concentrations (those greater than 0.5 ng/m3)
were measured at ASKY-M during the third quarter of 2014.
• It should be noted that the COCs for samples collected from mid-May through mid-
July at ASKY-M denoted that the samples were potentially influenced by a nearby
source, as indicated by the request to apply the "NS" flag to the data when uploaded
into AQS.
• Table 4-12 presents the NMP sites with the 10 highest annual average concentrations
for each of the program-level metal pollutants of interest. This table shows that the
highest annual average concentrations for arsenic and nickel across the program were
calculated for ASKY-M. A similar observation was made in the 2013 report.
Observations for GLKY from Table 12-5 include the following:
• GLKY sampled VOCs, carbonyl compounds, metals (PMio), and PAHs.
• The only pollutant of interest with an annual average concentration greater than
1 |ig/m3 is formaldehyde (1.49 ± 0.22 |ig/m3). However, this is one of the lowest
annual averages of formaldehyde calculated among NMP sites sampling carbonyl
compounds.
• Concentrations of formaldehyde were higher during the warmer months of the year,
based on the quarterly averages. All but one of the 24 formaldehyde concentrations
greater than 1.50 |ig/m3 were measured during the second or third quarters of 2014.
Conversely, all but one of the 24 concentrations less than 1.0 |ig/m3 were measured
during the first or fourth quarters of the year.
• Concentrations of acetaldehyde do not exhibit the same tendency as formaldehyde.
Concentrations of this pollutant were highest during the first and second quarters. Of
the 16 acetaldehyde concentrations greater than 1 |ig/m3, 12 were measured during
the first and second quarters, with only one measured during the third quarter and
three during the fourth quarter.
• A number of VOC samples collected between mid-May and mid-June were
invalidated due to a sampler issue; as a result, no second quarter average
concentrations are shown in Table 12-5 for GLKY.
• Based on the quarterly average concentrations shown, benzene concentrations appear
highest during the first quarter of 2014, yet the fourth quarter average has the highest
associated confidence interval. A review of the data shows that the maximum
benzene concentration was measured on November 22, 2014 (1.47 |ig/m3), and is the
only benzene concentration greater than 1 |ig/m3 measured at GLKY. The next eight
highest benzene concentrations were all measured at GLKY during the first quarter of
12-45
-------
the year. Benzene concentrations measured during the fourth quarter but before the
maximum concentration was measured were less than or equal to the median benzene
concentration for GLKY; all but one of the concentrations measured after the
maximum concentration was measured were greater than the median concentration.
• All of GLKY's 1,3-butadiene concentrations are less than 0.1 |ig/m3 and range from
0.020 |ig/m3 to 0.075 |ig/m3, including nine non-detects. There is little variability in
the quarterly average concentrations shown for this pollutant.
• The third quarter average concentration of 1,2-dichloroethane is less than the other
quarterly averages shown and has the most variability associated with it. All eight
non-detects of 1,2-dichloroethane were measured at GLKY during the third quarter of
2014. In addition, all measured detections from the third quarter were less than or
equal to the median 1,2-dichloroethane concentration.
• 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 0.04 ng/m3 to
1.53 ng/m3, plus four non-detects. The four highest arsenic concentrations measured
at GLKY were measured on consecutive sample days between September 14, 2014
and October 2, 2014.
• GLKY is not listed in Tables 4-9 through 4-12, which present the NMP sites with the
10 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. GLKY's annual average benzene
concentration is the lowest among all NMP sites sampling this pollutant.
Observations for BAKY from Table 12-5 include the following:
• Only speciated metals were sampled for at BAKY; only arsenic was identified as a
pollutant of interest for BAKY.
• Arsenic was measured in all but one of the 58 valid metals samples collected at
BAKY.
• Arsenic concentrations measured at BAKY range from 0.06 ng/m3 to 3.36 ng/m3, plus
the one non-detect, with a median concentration of 0.755 ng/m3. The maximum
arsenic concentration was measured on October 26, 2014 and is among the highest
arsenic concentrations measured across the program. Yet, this measurement is the
only arsenic concentration greater than 1 ng/m3 measured during the fourth quarter of
2014, explaining the variability shown by the confidence interval for the fourth
quarter average concentration. Most of the 15 arsenic concentrations greater than
1 ng/m3 were measured at BAKY during the third quarter of the year (nine).
• Among NMP sites sampling PMio metals, BAKY has the third highest annual
average concentration of arsenic, as shown in Table 4-12. A similar observation was
made in the 2013 NMP report.
12-46
-------
Observations for the Calvert City monitoring sites from Table 12-5 include the following:
• VOC samples were collected at all five Calvert City sites; PMio metals were sampled
for in addition to VOCs at CCKY through October 2, 2014, after which sampling at
this location was discontinued. Thus, fourth quarter average concentrations are not
available in table 12-5.
• Some of the highest concentrations of VOCs were measured at the Calvert City sites
and these data are reviewed in the bullets that follow.
• Vinyl chloride is an infrequently detected pollutant under the NMP in typical urban
atmospheres. Across the program, this pollutant was detected in 16 percent of the
total samples collected. Together, the five Calvert City sites account for nearly
70 percent of the 248 measured detections of this pollutant. The Calvert City sites
account for all 56 concentrations of vinyl chloride greater than 0.30 |ig/m3 measured
across the program, including the 19 measurements greater than 1 |ig/m3. The
maximum concentration of vinyl chloride across the program was measured at TVKY
(22.95 |ig/m3), although additional measurements greater than 5 |ig/m3 were also
measured at ATKY. A vinyl chloride concentration greater than 1 |ig/m3 was
measured at least once at four of the five Calvert City sites in 2014 (CCKY is the
exception).
• Vinyl chloride is a pollutant of interest for all five Calvert City sites. As shown in
Table 12-5, annual average concentrations for these sites range from
0.08 ± 0.04 |ig/m3 for CCKY to 0.77 ± 0.40 |ig/m3 for ATKY. All of the annual
average and quarterly average concentrations of vinyl chloride for these sites have
relatively large confidence intervals, indicating the relatively large amount of
variability associated with these measurements, including substitutions for non-
detects. The number of non-detects measured at each site range from 19 (TVKY and
CCKY) to 27 (BLKY).
• Another pollutant for which the highest concentrations program-wide were measured
at the Calvert City sites is 1,2-dichloroethane. The 124 highest concentrations of
1,2-dichloroethane across the program were measured at the Calvert City sites. This
includes all 78 measurements greater than 1 |ig/m3 and the nine greater than
10 |ig/m3.
• 1,2-Dichloroethane is a pollutant of interest for all five Calvert City sites. Annual
average concentrations for these sites range from 0.49 ± 0.16 |ig/m3 for CCKY to
3.54 ± 1.66 |ig/m3 for TVKY. All of the sites except CCKY have at least one
quarterly average concentration of 1,2-dichloroethane with an associated confidence
interval greater than the average itself, indicating the relatively large amount of
variability associated with these measurements.
• Some of the highest measurements of carbon tetrachloride were also measured at the
Calvert City sites. Of the 17 carbon tetrachloride concentrations greater than or equal
to 1 |ig/m3 measured across the program, 16 were measured at the Calvert City sites
(seven at TVKY, three at BLKY, and two each at ATKY, CCKY, and LAKY). The
maximum carbon tetrachloride concentration measured at TVKY on
12-47
-------
November 13, 2014 (3.06 |ig/m3), although a concentration of similar magnitude
(2.86 |ig/m3) was also measured at TVKY in September.
Carbon tetrachloride is a pollutant of interest for all five Calvert City sites. Annual
average concentrations range from 0.68 ± 0.03 |ig/m3 for ATKY to 0.87 ± 0.13 |ig/m3
for TVKY. Quarterly average concentrations for TVKY exhibit the most variability,
ranging from 0.76 ± 0.08 |ig/m3 for the second quarter of 2014 to 0.93 ± 0.32 |ig/m3
for the first quarter. Most of the quarterly average concentrations calculated for NMP
sites sampling carbon tetrachloride in 2014 fall between 0.55 |ig/m3 and 0.75 |ig/m3;
all four of TVKY's quarterly averages, BLKY's second quarter average, and LAKY's
first quarter are outside this range.
All 10 1,3-butadiene concentrations greater than 0.75 |ig/m3 measured across the
program were measured at the Calvert City sites (seven at TVKY, two at BLKY, and
one at LAKY). Concentrations of 1,3-butadiene greater than 0.5 |ig/m3 were
measured at only seven NMP sites in 2014, and concentrations measured at TVKY,
BLKY, LAKY, and ATKY account for 18 out of 28 of them. Annual average
concentrations of 1,3-butadiene range from 0.08 ± 0.03 |ig/m3 for CCKY to
0.38 ± 0.22 |ig/m3 for TVKY.
The concentrations of 1,3-butadiene for the Calvert City sites exhibit considerable
variability, particularly for TVKY, with quarterly average concentrations that range
from 0.13 ± 0.09 |ig/m3 for the fourth quarter of 2014 to 0.86 ± 0.76 |ig/m3 for the
third quarter. Many of the quarterly average concentrations for the Calvert City sites
have relatively large confidence intervals associated with them, a few of which are
greater than the average itself.
Hexachloro-1,3-butadiene is another infrequently detected pollutant that is a pollutant
of interest for all five Calvert City sites. Concentrations measured at the Calvert City
sites account for roughly one-fifth of the 359 measured detections of this pollutant
across the program. The number of times hexachloro-l,3-butadiene was detected in
samples collected at these sites ranges from nine (CCKY) to 20 (TVKY); thus, zeros
substituted for non-detects make up the majority of the measurements incorporated
into the quarterly and annual averages shown in Table 12-5. As a result, the annual
averages are not significantly different across the sites, ranging from
0.016 ± 0.010 |ig/m3 for CCKY to 0.029 ± 0.011 |ig/m3 for TVKY.
Benzene is the only other VOC that is a pollutant of interest across the Calvert City
sites. Annual average concentrations of benzene range from 0.56 ± 0.08 |ig/m3 for
CCKY to 1.04 ± 0.35 |ig/m3 for TVKY. Benzene concentrations measured at TVKY
exhibit the most variability, ranging from 0.20 |ig/m3 to 9.92 |ig/m3. The maximum
benzene concentration was measured at TVKY on July 28, 2014 and is the second
highest benzene concentration measured across the program, second only to the
maximum concentration measured at ASKY. Concentrations measured at TVKY
account for three of the eight benzene measurements greater than 3 |ig/m3 measured
across the program with LAKY accounting for a fourth. TVKY's third quarter
average concentration is the highest quarterly average concentration of benzene
12-48
-------
calculated for NMP sites sampling this pollutant. All but one of TVKY's benzene
concentrations measured in July and August are greater than 1 |ig/m3.
• 1,1,2-Trichloroethane is a pollutant of interest for TVKY and is the only NMP site for
which this is true. This pollutant was detected in 28 samples collected across the
program in 2014, with measurements from the Calvert City sites accounting for 27 of
them. This pollutant was detected 15 times at TVKY, four times each at BLKY and
LAKY, three times at ATKY, and once at CCKY. The program-level maximum
concentration of this pollutant (0.607 |ig/m3) was measured at LAKY on December 1,
2014 and concentrations of 1,1,2-trichloroethane greater than 0.2 |ig/m3 were not
measured at any other NMP site in 2014.
• Table 4-9 presents the NMP sites with the 10 highest annual average concentrations
for each of the program-level VOC pollutants of interest. This table shows that the
Calvert City sites account for the five highest annual average concentrations of
carbon tetrachloride and 1,2-dichloroethane across the program, with TVKY ranking
highest for each. Calvert City sites account for three of the five highest annual
average concentrations of 1,3-butadiene across the program (with ATKY and CCKY
as the exceptions). TVKY has the sixth highest annual average benzene concentration
among sites sampling this pollutant. TVKY and LAKY rank second and sixth highest,
respectively, among NMP sites for their annual average concentrations of hexachloro-
1,3-butadiene.
• Arsenic is the only non-VOC pollutant of interest for CCKY. Concentrations of
arsenic measured at CCKY range from 0.10 ng/m3 to 1.75 ng/m3, plus two non-
detects, with a median concentration of 0.43 ng/m3. Concentrations measured in 2014
appear highest during the warmer months of the year and lowest during the colder
months, based on the available quarterly average concentrations. The number of
arsenic concentrations greater than 0.5 ng/m3 measured at CCKY increases from three
during the first quarter to four during the second quarter to 10 during the third quarter.
Conversely, the number of arsenic concentrations less than 0.3 ng/m3 measured at
CCKY decreases from five during the first quarter to three during the second quarter
to one during the third quarter. This site has the ninth highest annual average
concentration of arsenic among NMP sites sampling PMio metals, as shown in
Table 4-12. A similar observation was made in the 2013 NMP report.
• It should be noted that during the first four months of 2014, construction was
occurring near the TVKY monitoring site.
Observations for LEKY from Table 12-5 include the following:
• VOC, carbonyl compound, and speciated metals samples were collected at LEKY in
2014.
• The annual average concentration for formaldehyde (3.15 ± 0.90 |ig/m3) is twice the
annual average concentration of acetaldehyde (1.55 ± 0.13 |ig/m3), the two carbonyl
compound pollutants of interest for LEKY and the only two pollutants with annual
average concentrations greater than 1 |ig/m3.
12-49
-------
• The second and third quarter average concentrations of formaldehyde are
considerably higher than the first and fourth quarter averages, indicating that
formaldehyde concentrations tended to be higher during the warmer months of the
year at this site. However, the confidence interval for the second quarter average is
five time larger than the confidence interval for the third quarter average, indicating
the presence of outliers. Concentrations of formaldehyde measured at LEKY range
from 0.766 |ig/m3 to 25.85 |ig/m3, with the maximum concentration measured on
May 23, 2014. This measurement is the maximum formaldehyde concentration
measured across the program. The next highest formaldehyde concentration measured
at LEKY (5.73 |ig/m3) is one-fifth the magnitude of the maximum concentration. If
the maximum formaldehyde concentration was excluded from the second quarter
average concentration, it would be similar to the third quarter average concentration.
• While not significantly different, the quarterly average concentrations of acetaldehyde
exhibit a decreasing trend across the year. The highest acetaldehyde concentrations
were measured on the last sample day of January and first sample day of February
while the lowest concentrations were measured in December. Of the 12 acetaldehyde
concentrations greater than 2 |ig/m3, four were measured during the first and second
quarters, three were measured during the third quarter, and only one was measured
during the fourth quarter. Looking at the lower end of the concentration range,
acetaldehyde concentrations less than 1 |ig/m3 were not measured during the first and
second quarters, while two were measured during the third quarter and five were
measured during the fourth quarter of 2014.
• Among the VOC pollutants of interest for LEKY, benzene has the highest annual
average concentration (0.62 ± 0.04 |ig/m3), although the annual average concentration
of carbon tetrachloride is very similar (0.61 ± 0.02 |ig/m3).
• Concentrations of arsenic measured at LEKY range from 0.10 ng/m3 to 1.88 ng/m3,
with a median concentration of 0.63 ng/m3, including 13 measurements greater than
1 ng/m3. Among NMP sites sampling PMio metals, LEKY has the sixth highest
annual average concentration of arsenic, as shown in Table 4-12.
• It should be noted that during the first quarter of 2014, demolition of a nearby
building was occurring near the monitoring site. Also, in June, paving and
construction were noted adjacent to the site.
12.4.2 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 the
pollutants of interest for each of the Kentucky monitoring sites. Figures 12-24 through 12-36
overlay the sites' minimum, annual average, and maximum concentrations onto the program-
level minimum, first quartile, median, average, third quartile, and maximum concentrations, as
12-50
-------
described in Section 3.4.3.1. Figures 12-24 through 12-36 and their associated observations are
discussed below.
Figure 12-24. Program vs. Site-Specific Average Acetaldehyde Concentrations
ASKY
GLKY
LEKY
0123456789 10
Concentration (ng/m3)
¦
¦
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 12-24 presents the box plots for acetaldehyde for ASKY, GLKY, and LEKY and
shows the following:
• The range of acetaldehyde concentrations measured at ASKY is similar to the range
measured at LEKY. For GLKY, all but one of the acetaldehyde concentrations
measured are less than the program-level average concentration.
• Among these three sites, GLKY has the lowest annual average concentration, which
is less than the program-level first quartile; LEKY has the highest of the three, with
an annual average similar to the program-level median concentration; ASKY's annual
average is between the two, although the annual average concentrations for all three
sites are less than the program-level average concentration.
12-51
-------
Figure 12-25. Program vs. Site-Specific Average Arsenic (PMio) Concentrations
-
ASKY-M
Program Max Concentration = 10.1 ng/m3
ASKY-M Max Concentration = 10.1 ng/m3
Program MaxConcentration = 10.1 ng/m3
E
CCKY
Program Max Concentration = 10.1 ng/m3
E
Program Max Concentration = 10.1 ng/m3
1
Program MaxConcentration = 10.1 ng/m3
0
1
2 3
Concentration (ng/m3)
4
5
6
Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i
Site: Site Average
o
Site Concentration Range
12-52
-------
Figure 12-25 presents the box plots for arsenic for the five Kentucky sites sampling PMio
metals and shows the following:
• The program-level maximum arsenic concentration (10.1 ng/m3) is not shown directly
on the box plots in Figure 12-25 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 box plots show that the range of arsenic concentrations measured is smallest for
GLKY and largest for ASKY-M and that the maximum arsenic concentration
measured across the program was measured at ASKY-M.
• The annual average concentrations of arsenic for ASKY-M, BAKY, and LEKY are
greater than the program-level average concentration while the annual average
concentrations for CCKY and GLKY are less than the program-level average
concentration. The annual average concentration for GLKY is also less than the
program-level median concentration.
• At least one non-detect of arsenic was measured at each site shown except LEKY.
12-53
-------
Figure 12-26. Program vs. Site-Specific Average Benzene Concentrations
Program Max Concentration - 12.4 ng/m3
b
p
ASKY Max Concentration = 12.4 ng/m3
¦H—¦
Program Max Concentration = 12.4 ng/m3
Program Max Concentration = 12.4 |ig/m;
ATKY
¦ o
Program Max Concentration = 12.4 ng/m3
°
¦o
Program Max Concentration = 12.4 ng/m3
1°
14 "
Program Max Concentration = 12.4 ng/m3
Program Max Concentration = 12.4 |ig/m3
_
—VJ 1
1 1
1
Program Max Concentration = 12.4 ng/m3
0 2 4 6 8 10
Concentration (ng/m3)
Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
12-54
-------
Figure 12-26 presents the box plots for benzene for the eight Kentucky sites sampling
VOCs and shows the following:
• The program-level maximum benzene concentration (12.4 |ig/m3) is not shown
directly on the box plots in Figure 12-26 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 box plots show that the maximum benzene concentration measured across the
program was measured at ASKY. While all other benzene concentrations measured at
the Kentucky sites fall within the range of benzene concentrations shown on the box
plots, the second highest benzene concentration across the program was measured at
TVKY. The range of benzene concentrations measured at the remaining sites are
considerably smaller, with the smallest range of measurements shown for LEKY.
• The annual average concentrations of benzene across the Kentucky sites range from
0.42 ± 0.05 |ig/m3 (GLKY) to 1.04 ± 0.35 |ig/m3 (TVKY); the annual averages for
ASKY and TVKY are greater than the program-level average concentration.
12-55
-------
Figure 12-27. Program vs. Site-Specific Average 1,3-Butadiene Concentrations
¦H-
Program Max Concentration = 5.90 ng/m3
Program Max Concentration = 5.90 ng/m3
B
Program Max Concentration = 5.90 ng/m3
Program Max Concentration = 5.90 ng/m3
BLKY Max Concentration = 4.92 ng/m3
Program Max Concentration = 5.90 ng/m3
Program Max Concentration = 5.90 ng/m3
Program Max Concentration = 5.90 ng/m3
-
1
r\
TVKY Max Concentration = 5.90 ng/m3
¦H-
Program Max Concentration = 5.90 |ig/m3
0.4 0.6
Concentration {[j.g/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range
12-56
-------
Figure 12-27 presents the box plots for 1,3-butadiene for the eight Kentucky sites
sampling VOCs and shows the following:
• The program-level maximum concentration (5.90 |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 plot has
been reduced to 1 |ig/m3. Also, since the maximum 1,3-butadiene concentration for
several sites is greater than the scale of the box plots, the site-specific maximum
concentrations are labeled for these sites.
• The maximum 1,3-butadiene concentration measured across the program was
measured at TVKY, although the maximum concentration measured at BLKY is also
greater than the scale of the box plot.
• The annual average concentration of 1,3-butadiene for TVKY is more than three
times the program-level average concentration; the annual average concentrations for
BLKY and LAKY are also greater than the program-level average. The annual
average concentrations for ATKY, CCKY, and LEKY fall between the program-level
median and average concentrations. The annual average concentration for ASKY is
just less than the program-level median concentration and the annual average for
GLKY is just less than the program-level first quartile. Note that 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.
12-57
-------
Figure 12-28. Program vs. Site-Specific Average Cadmium Concentration
\-
Program Max Concentration = 70.7 ng/m3
.
U 1
i i i i i
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Concentration {ng/m3)
Program: 1st Quartile
2nd Quartile 3rd Quartile
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 12-28 presents the box plot for cadmium for ASKY-M and shows the following:
• Similar to other pollutants, the program-level maximum cadmium concentration
(70.7 ng/m3) is not shown directly on the box plot as the scale of the box plot has
been reduced to 3 ng/m3 in order to allow for the observation of data points at the
lower end of the concentration range.
• The maximum concentration measured at ASKY-M (0.79 ng/m3) is an order of
magnitude less than the maximum concentration measured across the program.
However, the maximum concentration measured this site is among the higher
measurements (ninth highest).
• The annual average concentration of cadmium for ASKY-M is just greater than the
program-level average concentration. Note that the program-level average cadmium
concentration is nearly two times the third quartile and nearly three times the
program-level median concentration, indicating that the cadmium concentrations on
the upper end of the concentration range, particularly the maximum program-level
concentration, are driving the program-level average concentration upward, as the
second highest cadmium concentration measured across the program falls within the
scale of the box plot. ASKY-M is one of only two NMP sites sampling PMio metals
for which cadmium is a pollutant of interest.
12-58
-------
Figure 12-29. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations
ASKY
Program Max Concentration = 3.06 |ig/m3
GLKY
Program Max Concentration = 3.06 M-g/m3
ATKY
Program Max Concentration = 3.06 |-ig/m3
BLKY
Program Max Concentration = 3.06 ng/m3
"VJ 1
1 1
CCKY
Program Max Concentration = 3.06 |ig/m3
LAKY
Program Max Concentration = 3.06 ^ig/m3
TVKY
Program Max Concentration = 3.06 ng/m3
1
/-\
U
TVKY Max Concentration = 3.06 ue/m3
1
Program Max Concentration = 3.06 \±g/m3
0 0.5 1 1.5 2 2.5
Concentration (ng/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range
12-59
-------
Figure 12-29 presents the box plots for carbon tetrachloride for the eight Kentucky sites
sampling VOCs and shows the following:
• The program-level maximum carbon tetrachloride concentration (3.06 |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 plot has been reduced. Note that the program-level median and
average concentrations are similar and plotted nearly on top of each other.
• The two highest carbon tetrachloride concentrations measured across the program
were measured at TVKY, although concentrations greater than 1 |ig/m3 were also
measured atBLKY, LAKY, and ATKY (and one other NMP site, GPCO).
• The annual average concentrations for the five Calvert City sites are greater than the
program-level average and greater than or similar to the program-level third quartile.
For the remaining Kentucky sites sampling carbon tetrachloride, the annual average
concentration for ASKY is similar to the program-level average while the annual
averages for GLKY and LEKY fall between the program-level first quartile and
program-level median and average concentration.
12-60
-------
Figure 12-30. Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations
I
Program Max Concentration = 27.4 ng/m3
Program Max Concentration = 27.4 ng/m3
Program Max Concentration = 27.4 j-ig/m3
1
-
vJ
1
Program Max Concentration - 27.4 ng/m3
-
— u
BLKY Max Concentration = 11.1 ng/m3
Program Max Concentration = 27.4 ng/m3
1
CCKY Max Concentration = 2.01 ug/m3
^ 1
Program Max Concentration = 27.4 ng/m3
1
-
-
n—
LAKY Max Concentration = 11.3 ng/m3
1
Program Max Concentration - 27.4 ng/m3
1
TVKY Max Concentration = 27.41 ng/m3; Average Concentration = 3.54 ng/m3
Program Max Concentration = 27.4 |ig/m3
0.4 0.6
Concentration {[j.g/m3)
Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
12-61
-------
Figure 12-30 presents the box plots for 1,2-dichloroethane for the eight Kentucky sites
sampling VOCs and shows the following:
• Similar to other pollutants, the program-level maximum concentration (27.4 |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 |ig/m3 in order to allow for the observations 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 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.
• 1,2-Dichloroethane concentrations measured at ASKY, GLKY, and LEKY are all less
than the program-level average concentration, with all of GLKY's measurements less
the program-level third quartile. By comparison, the maximum concentration
measured at each of the Calvert City sites exceeds the scale of the box plots,
including TVKY's annual average concentration.
• Recall from the previous section that the annual average concentrations for the
Calvert City sites account for the five highest annual average concentrations of
1,2-dichloroethane among NMP sites sampling VOCs. Among the Calvert City sites,
CCKY has the lowest annual average concentration of 1,2-dichloroethane, although it
is still more than four times the annual average for the NMP monitoring site with the
next highest annual average (BTUT), as shown in Table 4-9.
12-62
-------
Figure 12-31. Program vs. Site-Specific Average Formaldehyde Concentrations
GLKY
¦
O i
¦
1
0 3 6 9 12 15 18 21 24 27
Concentration (ng/m3)
1H
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 12-31 presents the box plots for formaldehyde for ASKY, GLKY, and LEKY and
shows the following:
• The range of formaldehyde concentrations measured was smallest for GLKY and
largest for LEKY, where the program-level maximum formaldehyde concentration
was measured.
• Among these three sites, GLKY has the lowest annual average concentration while
LEKY has the highest. LEKY's annual average concentration is just greater than the
program-level average concentration; ASKY's annual average is just less the
program-level median concentration; and GLKY's annual average is just greater than
to the program-level first quartile.
12-63
-------
Figure 12-32. Program vs. Site-Specific Average Hexachloro-l,3-butadiene Concentrations
ATKY
BLKY
CCKY
LAKY
4
4
>
O i
1
i i i i i i
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Concentration (ng/m3)
Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 12-32 presents the box plots for hexachloro-1,3-butadiene for the five 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 Kentucky sites for which hexachloro-1,3-butadiene was identified as a pollutant
of interest are all Calvert City sites. The maximum hexachloro-l,3-butadiene
concentrations measured at these sites are fairly similar to each other, with less than
0.05 |ig/m3 separating them. The number of measured detections of this pollutant for
these sites ranged from nine (CCKY) to 20 (TVKY), and thus, hexachloro-1,3-
butadiene was detected in fewer than one-third of valid samples for these sites.
12-64
-------
• Even though the annual average hexachloro-l,3-butadiene concentrations for all of
the sites shown are similar to the program-level average concentration, the annual
averages of this pollutant for all NMP sites sampling VOCs fall within a relatively
tight range across the program (spanning 0.05 |ig/m3 for all NMP sites sampling
VOCs, including several sites at which hexachloro-l,3-butadiene was not detected).
Figure 12-33. Program vs. Site-Specific Average Manganese (PMio) Concentration
~
O i
\J 1
0
20
40
60 80
Concentration {ng/m3)
100
120
140
Program:
1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site:
Site Average
o
Site Concentration Range
Figure 12-33 presents the box plot for manganese for ASKY-M and shows the following:
• Although the maximum manganese concentration across the program was not
measured at ASKY-M, this site does have one of the higher measurements, including
two greater than 50 ng/m3.
• The annual average concentration of manganese for ASKY-M is more than twice the
program-level average concentration. Note that ASKY-M is the only NMP site
sampling PMio metals for which manganese is a pollutant of interest (and one of only
two NMP sites if the sites sampling TSP metals are included).
Figure 12-34. Program vs. Site-Specific Average Nickel (PMio) Concentration
4 6
Concentration {ng/m3)
Progra m: 1st Qua rti 1 e
¦
2nd Quartile 3rd Quartile
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
12-65
-------
Figure 12-34 presents the box plot for nickel for ASKY-M and shows the following:
• Although the maximum nickel concentration measured across the program was not
measured at ASKY-M, ASKY-M's maximum concentration does rank second
highest, with a difference less than 0.1 ng/m3.
• The annual average concentration of nickel for ASKY-M is nearly two times greater
than the program-level average concentration and is the highest annual average
concentration of nickel calculated among NMP sites sampling PMio metals.
Figure 12-35. Program vs. Site-Specific Average 1,1,2-Trichloroethane Concentration
0.1
0.2
0.3 0.4
Concentration (ng/m3)
0.5
0.6
Program:
1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site:
Site Average
o
Site Concentration Range
Figure 12-35 presents the box plot for 1,1,2-trichloroethane for TVKY, the only site 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 plot.
• The maximum 1,1,2-trichloroethane concentration across the program was measured
at LAKY and not at TVKY, and although the next two highest 1,1,2-trichloroethane
concentrations were measured at TVKY, they were considerably less. This pollutant
was detected 15 times at TVKY, or in less than a quarter of the samples collected at
this site, despite having the highest detection rate among NMP sites sampling VOCs.
TVKY is the only NMP site with 1,1,2-trichloroethane as a pollutant of interest.
12-66
-------
Figure 12-36. Program vs. Site-Specific Average Vinyl Chloride Concentrations
Program Max Concentration = 23.0 ng/m3
O
ATKY Max Concentration = 7.45 ng/m3
-O-
Program Max Concentration = 23.0 ng/m3
BLKY Max Concentration = 1.68 ng/m3
Program Max Concentration = 23.0 ng/m3
-o-
Program Max Concentration = 23.0 ng/m3
LAKY Max Concentration = 1.19 ng/m3
Program Max Concentration = 23.0 ng/m3
1
r»
TVKY Max Concentration = 23.0 ng/m3
1
0
0.2
0.4 0.6
Concentration {[j.g/m3)
0.8
l
Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i
Site: Site Average
o
Site Concentration Range
Figure 12-36 presents the box plots for vinyl chloride for the five 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.
• Similar to other pollutants, the program-level maximum concentration (23.0 |ig/m3) is
not shown directly on the box plots for vinyl chloride as the scale of the box plots has
been reduced. Also, since the maximum vinyl chloride concentration for several sites
is greater than the scale of the box plots, the site-specific maximum concentrations
are labeled for these sites.
12-67
-------
• The maximum vinyl chloride concentration measured at TVKY is the maximum
concentration measured across the program, although several concentrations greater
than the scale of the box plots were measured at four of the five Calvert City sites.
• The annual average vinyl chloride concentrations for these sites range from
0.08 ± 0.04 |ig/m3 for CCKY to 0.77 ± 0.40 |ig/m3 for ATKY, all of which are greater
than the program-level average concentration of 0.07 |ig/m3. Note that the annual
averages for ATKY and TVKY are more than five times greater than the annual
averages for the remaining three sites.
• The number of measured detections ranges from 27 for CCKY to 42 for TVKY, with
no other NMP site measuring more than eight measured detections of this pollutant.
12.4.3 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 3.4.3.2. The
only pollutant groups for which GLKY has sampled under the NMP since at least 2010 is VOCs
and PAHs (carbonyl compounds and PMio metals sampling began in 2011). Thus, Figures 12-37
through 12-40 present the 1-year statistical metrics for each of the VOC pollutants of interest for
GLKY (no PAHs were identified as pollutants of interest 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 trends analysis; in
these cases, a 1-year average concentration is not provided, although the range and percentiles
are still presented. The remaining Kentucky sites did not begin sampling under the NMP until
2012, and thus, no trends analysis was performed.
12-68
-------
Figure 12-37. Yearly Statistical Metrics for Benzene Concentrations Measured at GLKY
1 A 1-year average is not presented because sampling under the NMP did not begin until June 2010.
Observations from Figure 12-37 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), and is one of only three benzene concentrations greater than 2 |ig/m3
measured at this site.
• The 1-year median benzene concentration exhibits a steady decreasing trend at this
site, decreasing from 0.59 |ig/m3 for 2010 to 0.36 |ig/m3 for 2014. The 1- year
average concentration decreases from 0.58 |ig/m3 to 0.42 |ig/m3 between 2011 and
2014, with a slight increase shown from 2012 to 2013. If the maximum concentration
measured in 2013 was excluded from the dataset, the 1-year average would exhibit a
pattern similar to the 1-year median concentration.
12-69
-------
Figure 12-38. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
GLKY
20101
o 5th Percentile
. o
2012
Year
— Minimum
— Maximum
o 95th Percentile
..4.... Average
1 A 1-year average is not presented because sampling under the NMP did not begin until June 2010.
Observations from Figure 12-38 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; only two non-detects were measured in 2012, compared
to between nine (2014) 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 median also exhibits
an increase.
• Despite the decreases shown in the statistical parameters from 2012 to 2013, the
median concentration exhibited an additional increase, albeit slight. The actual
number of concentrations greater than 0.05 |ig/m3 changed very little between these
two years, with the number of measurements between 0.05 |ig/m3 and 0.075 |ig/m3
increasing from 12 in 2012 to 19 in 2013.
• Each of the statistical parameters exhibits a slight decrease from 2013 to 2014, with
the exception of the minimum and 5th percentile, which remained zero.
12-70
-------
Figure 12-39. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at GLKY
2012
Year
o 5th Percentile
— Minimum
— Maximum
o 95th Percentile
..4.... Average
1 A 1-year average is not presented because sampling under the NMP did not begin until June 2010.
Observations from Figure 12-39 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, with several parameters at a minimum for
2014, which is the first year without a measurement greater than 0.8 |ig/m3.
12-71
-------
Figure 12-40. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured
at GLKY
J
a o.io
2012
Year
o 5th Percentile
— Minimum
— Maximum
o 95th Percentile
..4.... Average
1 A 1-year average is not presented because sampling under the NMP did not begin until June 2010.
Observations from Figure 12-40 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, 54 in 2012, peaked in 2013 with 56,
then fell slightly to 48 in 2014.
• As the number of non-detects decreased and measured detections increased, the
1-year average and median 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 several non-detects (or zeros)
factoring into the 1-year average concentration for each year: 2012 (6), 2013 (5), and
2014 (8), which drive the 1-year average concentrations down in the same manner
that a maximum or outlier concentration can drive the average up.
• The maximum concentration, 95th percentile, median, and average concentrations for
2014 exhibit decreases from 2013. Eleven concentrations measured in 2013 are
greater than the maximum concentration measured in 2014. Further, the number of
concentrations greater than 0.8 |ig/m3 fell from 27 measured in 2013 to 14 in 2014.
12-72
-------
12.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at the Kentucky monitoring sites. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4
for definitions and explanations regarding the various toxicity factors, time frames, and
calculations associated with these risk-based screenings.
12.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Kentucky 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 3.4.3.3 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 12-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Table 12-6. Risk Approximations for the Kentucky Monitoring Sites
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Health Department, Ashland, Kentucky - ASKY
Acetaldehyde
0.0000022
0.009
61/61
1.25
±0.11
2.74
0.14
Benzene
0.0000078
0.03
61/61
0.87
±0.39
6.77
0.03
1,3-Butadiene
0.00003
0.002
56/61
0.06
±0.01
1.76
0.03
Carbon Tetrachloride
0.000006
0.1
61/61
0.65
±0.02
3.87
0.01
1,2-Dichloroethane
0.000026
2.4
58/61
0.07
±0.01
1.90
<0.01
Formaldehyde
0.000013
0.0098
61/61
2.29
±0.31
29.75
0.23
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of viewing.
- = A Cancer URE or Noncancer RfC is not available.
12-73
-------
Table 12-6. Risk Approximations for the Kentucky Monitoring Sites (Continued)
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
21st and Greenup, Ashland, Kentucky - ASKY-M
Arsenic (PMi0)a
0.0043
0.000015
58/59
1.14
±0.36
4.88
0.08
Cadmium (PMi0)a
0.0018
0.00001
59/59
0.24
±0.04
0.42
0.02
Manganese (PMi0)a
0.0003
59/59
18.22
±3.53
0.06
Nickel (PMi,:,)a
0.00048
0.00009
59/59
2.19
±0.56
1.05
0.02
Grayson, Kentucky - GLKY
Acetaldehyde
0.0000022
0.009
61/61
0.86
±0.08
1.89
0.10
Benzene
0.0000078
0.03
56/56
0.42
±0.05
3.31
0.01
1,3-Butadiene
0.00003
0.002
47/56
0.04
±0.01
1.12
0.02
Carbon Tetrachloride
0.000006
0.1
56/56
0.61
±0.02
3.64
0.01
1,2-Dichloroethane
0.000026
2.4
48/56
0.06
±0.01
1.63
<0.01
Formaldehyde
0.000013
0.0098
61/61
1.49
±0.22
19.41
0.15
Arsenic (PMi0)a
0.0043
0.000015
55/59
0.39
±0.08
1.69
0.03
Baskett, Kentucky - BAKY
Arsenic (PMi0)a
0.0043
0.000015
57/58
0.85
±0.17
3.67
0.06
Atmos Energy, Calvert City, Kentucky - ATKY
Benzene
0.0000078
0.03
61/61
0.62
±0.08
4.82
0.02
1,3-Butadiene
0.00003
0.002
53/61
0.09
±0.03
2.61
0.04
Carbon Tetrachloride
0.000006
0.1
61/61
0.68
±0.03
4.09
0.01
1,2-Dichloroethane
0.000026
2.4
60/61
0.58
±0.22
15.06
<0.01
Hexacliloro -1,3 -butadiene
0.000022
0.09
14/61
0.02
±0.01
0.44
<0.01
Vinyl chloride
0.0000088
0.1
36/61
0.77
±0.40
6.77
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.
- = A Cancer URE or Noncancer RfC is not available.
12-74
-------
Table 12-6. Risk Approximations for the Kentucky Monitoring Sites (Continued)
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Smithland, Kentucky - BLKY
Benzene
0.0000078
0.03
60/60
0.62
±0.13
4.82
0.02
1,3-Butadiene
0.00003
0.002
47/60
0.17
±0.16
5.20
0.09
Carbon Tetrachloride
0.000006
0.1
60/60
0.73
±0.09
4.40
0.01
1,2-Dichloroethane
0.000026
2.4
57/60
0.81
±0.43
20.98
<0.01
Hexacliloro -1,3 -butadiene
0.000022
0.09
13/60
0.02
±0.01
0.40
<0.01
Vinyl chloride
0.0000088
0.1
33/60
0.14
±0.07
1.23
<0.01
Calvert City Elementary School, Calvert City, Kentucky - CCKY
Benzene
0.0000078
0.03
46/46
0.56
±0.08
4.39
0.02
1,3-Butadiene
0.00003
0.002
41/46
0.08
±0.03
2.48
0.04
Carbon Tetrachloride
0.000006
0.1
46/46
0.70
±0.03
4.19
0.01
1,2-Dichloroethane
0.000026
2.4
46/46
0.49
±0.16
12.78
<0.01
Hexacliloro -1,3 -butadiene
0.000022
0.09
9/46
0.02
±0.01
0.35
<0.01
Vinyl chloride
0.0000088
0.1
27/46
0.08
±0.04
0.70
<0.01
Arsenic (PMi0)a
0.0043
0.000015
39/41
0.55
±0.12
2.34
0.04
Lazy Daze, Calvert City, Kentucky - LAKY
Benzene
0.0000078
0.03
56/56
0.70
±0.13
5.50
0.02
1,3-Butadiene
0.00003
0.002
49/56
0.11
±0.04
3.22
0.05
Carbon Tetrachloride
0.000006
0.1
56/56
0.71
±0.06
4.25
0.01
1,2-Dichloroethane
0.000026
2.4
56/56
0.97
±0.47
25.17
<0.01
Hexacliloro -1,3 -butadiene
0.000022
0.09
15/56
0.02
±0.01
0.52
<0.01
Vinyl chloride
0.0000088
0.1
33/56
0.13
±0.07
1.14
<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.
- = A Cancer URE or Noncancer RfC is not available.
12-75
-------
Table 12-6. Risk Approximations for the Kentucky Monitoring Sites (Continued)
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
TVA Substation, Calvert City, Kentucky - TVKY
Benzene
0.0000078
0.03
61/61
1.04
±0.35
8.13
0.03
1,3-Butadiene
0.00003
0.002
55/61
0.38
±0.22
11.55
0.19
Carbon Tetrachloride
0.000006
0.1
61/61
0.87
±0.13
5.22
0.01
1,2-Dichloroethane
0.000026
2.4
61/61
3.54
± 1.66
91.92
<0.01
Hexachloro -1,3 -butadiene
0.000022
0.09
20/61
0.03
±0.01
0.63
<0.01
1,1,2-Trichloroethane
0.000016
0.4
15/61
0.03
±0.01
0.48
<0.01
Vinyl chloride
0.0000088
0.1
42/61
0.69
±0.76
6.05
0.01
Lexington, Kentucky - LEKY
Acetaldehyde
0.0000022
0.009
55/55
1.55
±0.13
3.40
0.17
Benzene
0.0000078
0.03
58/58
0.62
±0.04
4.84
0.02
1,3-Butadiene
0.00003
0.002
56/58
0.07
±0.01
2.02
0.03
Carbon Tetrachloride
0.000006
0.1
58/58
0.61
±0.02
3.68
0.01
1,2-Dichloroethane
0.000026
2.4
57/58
0.08
±0.01
2.02
<0.01
Formaldehyde
0.000013
0.0098
55/55
3.15
±0.90
40.92
0.32
Arsenic (PMi0)a
0.0043
0.000015
56/56
0.67
±0.11
2.89
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.
- = A Cancer URE or Noncancer RfC is not available.
Observations for the Kentucky monitoring sites from Table 12-6 include the following:
• The pollutants with the highest annual average concentrations for ASKY are
formaldehyde, acetaldehyde, and benzene. Formaldehyde and benzene are the
pollutants with the highest cancer risk approximations for ASKY (29.75 in-a-million
and 6.77 in-a-million, respectively). All of the noncancer hazard approximations for
the pollutants of interest for ASKY are considerably less than an HQ of 1.0 (0.23 or
less), indicating that no adverse noncancer health effects are expected from these
individual pollutants.
12-76
-------
The pollutant of interest with the highest annual average concentration for ASKY-M
is manganese. Arsenic has the highest cancer risk approximation among ASKY-M's
pollutants of interest (4.88 in-a-million). All of the noncancer hazard approximations
for the pollutants of interest for ASKY-M are considerably less than an HQ of 1.0
(0.08 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 an annual average
concentration greater than 1 |ig/m3. This pollutant also has the only cancer risk
approximation greater than 10 in-a-million for GLKY (19.41 in-a-million). All of the
noncancer hazard approximations for the pollutants of interest for GLKY are
considerably less than an HQ of 1.0 (0.15 or less), indicating that no adverse
noncancer health effects are expected from these individual pollutants.
Arsenic is the only pollutant of interest for BAKY. Arsenic has a cancer risk
approximation greater than 1 in-a-million for BAKY (3.67 in-a-million). The
noncancer hazard approximation for arsenic for BAKY is considerably less than an
HQ of 1.0 (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 12.78 in-a-million (CCKY) to 91.92 in-a-million (TVKY). This cancer risk
approximation for TVKY is the highest one calculated in the 2014 NMP report. The
cancer risk approximation for 1,3-butadiene for TVKY is also greater than 10 in-a-
million (11.55 in-a-million), which is the highest approximation calculated for this
pollutant across the program. With the exception of CCKY, each Calvert City site has
at least one additional pollutant for which a cancer risk approximation greater than
5 in-a-million was calculated.
All of the noncancer hazard approximations for the pollutants of interest for the
Calvert City sites are less than an HQ of 1.0, indicating that no adverse noncancer
health effects are expected from these individual pollutants. For each of these sites,
the pollutant with the highest noncancer hazard approximation is 1,3-butadiene,
which ranged from 0.04 for ATKY and CCKY to 0.19 for TVKY. (Note that the
noncancer hazard approximation for arsenic for CCKY is the same as the one
calculated for 1,3-butadiene.)
Formaldehyde and acetaldehyde are the only pollutants of interest for LEKY with an
annual average concentration greater than 1 |ig/m3. The cancer risk approximation for
LEKY for formaldehyde (40.92 in-a-million) is an order of magnitude greater than
the cancer risk approximation for the pollutant with the next highest cancer risk
approximation for this site (benzene, 4.84 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 (0.32 or less), indicating that no adverse noncancer health effects are
expected from these individual pollutants.
12-77
-------
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 speed and 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, similar to the wind roses presented in Section 12.2.3. Thus,
high concentrations may be shown in relation to the direction of potential emissions sources.
Hourly NWS wind observations from the Barkley Regional Airport used in this analysis 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 3.4.3.3.
Figure 12-41 presents the pollution rose for all 61 1,2-dichloroethane concentrations
measured at TVKY. However, the magnitude of the higher concentrations is such that all of the
lower concentrations are plotted nearly on top of each other. As a result, two pollution roses were
created for TVKY, one that shows 1,2-dichloroethane measurements greater than 1 |ig/m3
(27 measurements) and one that shows 1,2-dichloroethane measurements less than 1 |ig/m3 (34
measurements).
Observations for Figure 12-41 include the following:
• Concentrations greater than 1 |ig/m3 are plotted on the first pollution rose in two
colors, with pink representing concentrations between 1 |ig/m3 and 10 |ig/m3 and blue
representing concentrations between 10 |ig/m3 and 30 |ig/m3. The pollution rose
shows that the concentrations greater than 10 |ig/m3 tended to be measured at TVKY
on days with an average wind direction between 315° (northwest) and 360° (north).
Concentrations between 1 |ig/m3 and 10 |ig/m3 tended to be measured on sample days
with an average wind direction between 315° (northwest) and 45° (northeast).
• Concentrations less than 1 |ig/m3 are plotted on the second pollution rose in two
colors, with yellow representing concentrations between 0.1 |ig/m3 and 1 |ig/m3 and
purple representing concentrations less than 0.1 |ig/m3. 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, although these lower
concentrations were measured most often on days with an average wind direction
from the southwest quadrant, between 180° (south) and 270° (west).
12-78
-------
Figure 12-41. Pollution Roses for 1,2-Dichloroethane Concentrations Measured at TVKY
Measurements greater than 1 |ig/m3
360/0
315.
270
225
135
180
01-10 ng/m3 ® 10-30 ng/m3
Measurements less than 1 |ig/m3
360/0
315
1 Ug/rrv
270
225
135
180
• <0.1 ng/m3 O 0.1-1.0 ng/m3
12-79
-------
12.5.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, Tables 12-7 and 12-8 present an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 12-7 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
that have cancer toxicity factors. Table 12-7 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 12-7 provides the pollutants with the highest cancer risk approximations (in-a-million) for
each site, as presented in Table 12-6. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 12-7. Table 12-8 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 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 12.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
12-80
-------
Table 12-7. 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)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Health Department, Ashland, Kentucky (Boyd County) - ASKY
Benzene
61.79
Coke Oven Emissions, PM
7.25E-03
Formaldehyde
29.75
Formaldehyde
20.35
Hexavalent Chromium
9.84E-04
Benzene
6.77
Ethylbenzene
13.36
Nickel, PM
6.71E-04
Carbon Tetrachloride
3.87
Acetaldehyde
11.59
Benzene
4.82E-04
Acetaldehyde
2.74
Coke Oven Emissions, PM
7.32
Formaldehyde
2.64E-04
1,2-Dichloroethane
1.90
1.3 -Butadiene
3.65
2,4-Dinitrotoluene
1.96E-04
1,3-Butadiene
1.76
2,4-Dinitrotoluene
2.20
1,3-Butadiene
1.10E-04
Tetrachloroethylene
2.00
Naphthalene
6.77E-05
Naphthalene
1.99
Cadmium, PM
5.96E-05
Nickel, PM
1.40
POM, Group 2b
4.51E-05
21st and Greenup, Ashland, Kentucky (Boyd County) - ASKY-M
Benzene
61.79
Coke Oven Emissions, PM
7.25E-03
Arsenic (PMio)
4.88
Formaldehyde
20.35
Hexavalent Chromium
9.84E-04
Nickel (PMio)
1.05
Ethylbenzene
13.36
Nickel, PM
6.71E-04
Cadmium (PMio)
0.42
Acetaldehyde
11.59
Benzene
4.82E-04
Coke Oven Emissions, PM
7.32
Formaldehyde
2.64E-04
1,3-Butadiene
3.65
2,4-Dinitrotoluene
1.96E-04
2,4-Dinitrotoluene
2.20
1,3-Butadiene
1.10E-04
Tetrachloroethylene
2.00
Naphthalene
6.77E-05
Naphthalene
1.99
Cadmium, PM
5.96E-05
Nickel, PM
1.40
POM, Group 2b
4.51E-05
-------
Table 12-7. 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)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Grayson, Kentucky (Carter County) - GLKY
Benzene
20.12
Formaldehyde
1.78E-04
Formaldehyde
19.41
Formaldehyde
13.70
Benzene
1.57E-04
Carbon Tetrachloride
3.64
Acetaldehyde
9.15
1,3-Butadiene
6.86E-05
Benzene
3.31
Ethylbenzene
9.14
Naphthalene
5.57E-05
Acetaldehyde
1.89
1.3 -Butadiene
2.29
POM, Group 2b
3.60E-05
Arsenic (PMio)
1.69
Naphthalene
1.64
POM, Group 2d
2.66E-05
1,2-Dichloroethane
1.63
POM, Group 2b
0.41
Ethylbenzene
2.29E-05
1,3-Butadiene
1.12
POM, Group 2d
0.30
POM, Group 5a
2.23E-05
POM, Group 6
0.04
Acetaldehyde
2.01E-05
Trichloroethylene
0.03
POM, Group 6
7.36E-06
Baskett, Kentucky (Henderson County) - BAKY
Formaldehyde
52.75
Formaldehyde
6.86E-04
Arsenic (PMio)
3.67
Benzene
42.14
Naphthalene
5.72E-04
Acetaldehyde
27.10
POM, Group 2d
3.74E-04
Naphthalene
16.81
Benzene
3.29E-04
Ethylbenzene
16.17
Hexavalent Chromium
2.83E-04
Tetrachloroethylene
6.71
Nickel, PM
2.73E-04
1,3-Butadiene
6.59
POM, Group 2b
2.52E-04
POM, Group 2d
4.25
1,3-Butadiene
1.98E-04
POM, Group 2b
2.87
Acetaldehyde
5.96E-05
Dichloro methane
0.83
Cadmium, PM
5.03E-05
-------
Table 12-7. 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)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Atmos Energy, Calvert City, Kentucky (Marshall County) - ATKY
Benzene
139.16
Benzene
1.09E-03
1,2-Dichloroethane
15.06
Ethylbenzene
100.03
Hexavalent Chromium
6.36E-04
Vinyl chloride
6.77
Formaldehyde
36.34
Formaldehyde
4.72E-04
Benzene
4.82
Acetaldehyde
33.61
1,3-Butadiene
4.26E-04
Carbon Tetrachloride
4.09
Vinyl chloride
30.93
Vinyl chloride
2.72E-04
1,3-Butadiene
2.61
1.3 -Butadiene
14.20
Ethylbenzene
2.50E-04
Hexachloro-1,3 -butadiene
0.44
1,2-Dichloroethane
9.25
1,2-Dichloroethane
2.41E-04
Naphthalene
3.45
POM, Group la
2.34E-04
POM, Group la
2.66
Naphthalene
1.17E-04
Carbon Tetrachloride
2.32
Nickel, PM
7.66E-05
Calvert City Elementary School, Calvert City, Kentucky (Marshall County) - CCKY
Benzene
139.16
Benzene
1.09E-03
1,2-Dichloroethane
12.78
Ethylbenzene
100.03
Hexavalent Chromium
6.36E-04
Benzene
4.39
Formaldehyde
36.34
Formaldehyde
4.72E-04
Carbon Tetrachloride
4.19
Acetaldehyde
33.61
1,3-Butadiene
4.26E-04
1,3-Butadiene
2.48
Vinyl chloride
30.93
Vinyl chloride
2.72E-04
Arsenic (PMio)
2.34
1.3 -Butadiene
14.20
Ethylbenzene
2.50E-04
Vinyl chloride
0.70
1,2-Dichloroethane
9.25
1,2-Dichloroethane
2.41E-04
Hexachloro-1,3 -butadiene
0.35
Naphthalene
3.45
POM, Group la
2.34E-04
POM, Group la
2.66
Naphthalene
1.17E-04
Carbon Tetrachloride
2.32
Nickel, PM
7.66E-05
-------
Table 12-7. 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)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Lazy Daze, Calvert City, Kentucky (Marshall County) - LAKY
Benzene
139.16
Benzene
1.09E-03
1,2-Dichloroethane
25.17
Ethylbenzene
100.03
Hexavalent Chromium
6.36E-04
Benzene
5.50
Formaldehyde
36.34
Formaldehyde
4.72E-04
Carbon Tetrachloride
4.25
Acetaldehyde
33.61
1,3-Butadiene
4.26E-04
1,3-Butadiene
3.22
Vinyl chloride
30.93
Vinyl chloride
2.72E-04
Vinyl chloride
1.14
1.3 -Butadiene
14.20
Ethylbenzene
2.50E-04
Hexachloro-1,3 -butadiene
0.52
1,2-Dichloroethane
9.25
1,2-Dichloroethane
2.41E-04
Naphthalene
3.45
POM, Group la
2.34E-04
POM, Group la
2.66
Naphthalene
1.17E-04
Carbon Tetrachloride
2.32
Nickel, PM
7.66E-05
TVA Substation, Calvert City, Kentucky (Marshall County) - TVKY
Benzene
139.16
Benzene
1.09E-03
1,2-Dichloroethane
91.92
Ethylbenzene
100.03
Hexavalent Chromium
6.36E-04
1,3-Butadiene
11.55
Formaldehyde
36.34
Formaldehyde
4.72E-04
Benzene
8.13
Acetaldehyde
33.61
1,3-Butadiene
4.26E-04
Vinyl chloride
6.05
Vinyl chloride
30.93
Vinyl chloride
2.72E-04
Carbon Tetrachloride
5.22
1.3 -Butadiene
14.20
Ethylbenzene
2.50E-04
Hexachloro-1,3 -butadiene
0.63
1,2-Dichloroethane
9.25
1,2-Dichloroethane
2.41E-04
1,1,2-Trichloroethane
0.48
Naphthalene
3.45
POM, Group la
2.34E-04
POM, Group la
2.66
Naphthalene
1.17E-04
Carbon Tetrachloride
2.32
Nickel, PM
7.66E-05
-------
Table 12-7. 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)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Smithland, Kentucky (Livingston County) - BLKY
Benzene
14.04
Formaldehyde
1.50E-04
1,2-Dichloroethane
20.98
Formaldehyde
11.52
Benzene
1.09E-04
1,3-Butadiene
5.20
Acetaldehyde
6.66
1,3-Butadiene
5.45E-05
Benzene
4.82
Ethylbenzene
5.39
Naphthalene
2.27E-05
Carbon Tetrachloride
4.40
1.3 -Butadiene
1.82
POM, Group 2b
1.63E-05
Vinyl chloride
1.23
Naphthalene
0.67
Acetaldehyde
1.47E-05
Hexachloro-1,3 -butadiene
0.40
POM, Group 2b
0.18
Ethylbenzene
1.35E-05
POM, Group 2d
0.15
POM, Group 2d
1.31E-05
POM, Group 6
0.03
Nickel, PM
1.16E-05
Trichloroethylene
0.03
POM, Group 5a
1.14E-05
Lexington, Kentucky (Fayette County) - LEKY
Benzene
135.46
Formaldehyde
1.20E-03
Formaldehyde
40.92
Formaldehyde
92.28
Benzene
1.06E-03
Benzene
4.84
Ethylbenzene
82.26
1,3-Butadiene
5.57E-04
Carbon Tetrachloride
3.68
Acetaldehyde
54.60
Naphthalene
3.50E-04
Acetaldehyde
3.40
1,3-Butadiene
18.57
POM, Group 2b
2.32E-04
Arsenic
2.89
Tetrachloroethylene
13.04
Ethylbenzene
2.06E-04
1,3-Butadiene
2.02
Naphthalene
10.31
POM, Group 2d
1.52E-04
1,2-Dichloroethane
2.02
POM, Group 2b
2.63
Hexavalent Chromium
1.34E-04
Trichloroethylene
1.94
Arsenic, PM
1.28E-04
POM, Group 2d
1.73
Acetaldehyde
1.20E-04
-------
Table 12-8. 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)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Health Department, Ashland, Kentucky (Boyd County) - ASKY
Toluene
89.31
Acrolein
63,687.70
Formaldehyde
0.23
Benzene
61.79
Chlorine
45,169.74
Acetaldehyde
0.14
Xylenes
52.93
Manganese, PM
33,849.17
1,3-Butadiene
0.03
Hexane
49.05
Nickel, PM
15,539.17
Benzene
0.03
Methanol
39.10
Lead, PM
11,227.89
Carbon Tetrachloride
0.01
Hydrochloric acid
27.65
Cadmium, PM
3,311.70
1,2-Dichloroethane
<0.01
Formaldehyde
20.35
Formaldehyde
2,076.07
Ethylbenzene
13.36
Benzene
2,059.62
Acetaldehyde
11.59
1,3-Butadiene
1,826.21
Manganese, PM
10.15
Hydrochloric acid
1,382.51
21st and Greenup, Ashland, Kentucky (Boyd County) - ASKY-M
Toluene
89.31
Acrolein
63,687.70
Arsenic (PMio)
0.08
Benzene
61.79
Chlorine
45,169.74
Manganese (PMio)
0.06
Xylenes
52.93
Manganese, PM
33,849.17
Nickel (PMio)
0.02
Hexane
49.05
Nickel, PM
15,539.17
Cadmium (PMio)
0.02
Methanol
39.10
Lead, PM
11,227.89
Hydrochloric acid
27.65
Cadmium, PM
3,311.70
Formaldehyde
20.35
Formaldehyde
2,076.07
Ethylbenzene
13.36
Benzene
2,059.62
Acetaldehyde
11.59
1,3-Butadiene
1,826.21
Manganese, PM
10.15
Hydrochloric acid
1,382.51
-------
Table 12-8. 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)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Grayson, Kentucky (Carter County) - GLKY
Toluene
58.15
Acrolein
45,189.54
Formaldehyde
0.15
Xylenes
35.59
Formaldehyde
1,397.58
Acetaldehyde
0.10
Hexane
25.95
Cyanide Compounds, gas
1,278.36
Arsenic (PMio)
0.03
Benzene
20.12
1.3 -Butadiene
1,144.01
1,3-Butadiene
0.02
Methanol
15.68
Acetaldehyde
1,016.31
Benzene
0.01
Formaldehyde
13.70
Benzene
670.52
Carbon Tetrachloride
0.01
Acetaldehyde
9.15
Naphthalene
545.72
1,2-Dichloroethane
<0.01
Ethylbenzene
9.14
Xylenes
355.95
Ethylene glycol
5.50
Arsenic, PM
91.58
1.3 -Butadiene
2.29
Propionaldehyde
84.60
Baskett, Kentucky (Henderson County) - BAKY
Carbonyl sulfide
128.78
Acrolein
76,864.06
Arsenic (PMio)
0.06
Toluene
112.00
Manganese, PM
7,205.03
Xylenes
78.62
Nickel, PM
6,326.84
Hexane
54.97
Naphthalene
5,604.26
Formaldehyde
52.75
Formaldehyde
5,383.11
Benzene
42.14
1,3-Butadiene
3,295.39
Methanol
28.37
Chlorine
3,245.91
Acetaldehyde
27.10
Acetaldehyde
3,010.82
Naphthalene
16.81
Cadmium, PM
2,795.50
Ethylbenzene
16.17
4,4'-Methylenediphenyl diisocyanate, gas
2,483.57
-------
Table 12-8. 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)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Atmos Energy, Calvert City, Kentucky (Marshall County) - ATKY
Methanol
677.58
Chlorine
210,803.93
1,3-Butadiene
0.04
Xylenes
522.49
Acrolein
125,961.44
Benzene
0.02
Toluene
480.91
1.3 -Butadiene
7,098.77
Vinyl chloride
0.01
Benzene
139.16
Xylenes
5,224.89
Carbon Tetrachloride
0.01
Hexane
100.70
Benzene
4,638.60
1,2-Dichloroethane
<0.01
Ethylbenzene
100.03
Hydrochloric acid
4,173.99
Hexachloro-1,3 -butadiene
<0.01
Hydrochloric acid
83.48
Acetaldehyde
3,734.64
Vinyl acetate
73.28
Formaldehyde
3,708.16
Formaldehyde
36.34
Acrylic acid
2,916.21
Acetaldehyde
33.61
Nickel, PM
1,773.75
Calvert City Elementary School, Calvert City, Kentucky (Marshall County) - CCKY
Methanol
677.58
Chlorine
210,803.93
1,3-Butadiene
0.04
Xylenes
522.49
Acrolein
125,961.44
Arsenic (PMio)
0.04
Toluene
480.91
1.3 -Butadiene
7,098.77
Benzene
0.02
Benzene
139.16
Xylenes
5,224.89
Carbon Tetrachloride
0.01
Hexane
100.70
Benzene
4,638.60
Vinyl chloride
<0.01
Ethylbenzene
100.03
Hydrochloric acid
4,173.99
1,2-Dichloroethane
<0.01
Hydrochloric acid
83.48
Acetaldehyde
3,734.64
Hexachloro-1,3 -butadiene
<0.01
Vinyl acetate
73.28
Formaldehyde
3,708.16
Formaldehyde
36.34
Acrylic acid
2,916.21
Acetaldehyde
33.61
Nickel, PM
1,773.75
-------
Table 12-8. 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)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Lazy Daze, Calvert City, Kentucky (Marshall County) - LAKY
Methanol
677.58
Chlorine
210,803.93
1,3-Butadiene
0.05
Xylenes
522.49
Acrolein
125,961.44
Benzene
0.02
Toluene
480.91
1.3 -Butadiene
7,098.77
Carbon Tetrachloride
0.01
Benzene
139.16
Xylenes
5,224.89
Vinyl chloride
<0.01
Hexane
100.70
Benzene
4,638.60
1,2-Dichloroethane
<0.01
Ethylbenzene
100.03
Hydrochloric acid
4,173.99
Hexachloro-1,3 -butadiene
<0.01
Hydrochloric acid
83.48
Acetaldehyde
3,734.64
Vinyl acetate
73.28
Formaldehyde
3,708.16
Formaldehyde
36.34
Acrylic acid
2,916.21
Acetaldehyde
33.61
Nickel, PM
1,773.75
TVA Substation, Calvert City, Kentucky (Marshall County) - TVKY
Methanol
677.58
Chlorine
210,803.93
1,3-Butadiene
0.19
Xylenes
522.49
Acrolein
125,961.44
Benzene
0.03
Toluene
480.91
1.3 -Butadiene
7,098.77
Carbon Tetrachloride
0.01
Benzene
139.16
Xylenes
5,224.89
Vinyl chloride
0.01
Hexane
100.70
Benzene
4,638.60
1,2-Dichloroethane
<0.01
Ethylbenzene
100.03
Hydrochloric acid
4,173.99
Hexachloro-1,3 -butadiene
<0.01
Hydrochloric acid
83.48
Acetaldehyde
3,734.64
1,1,2-Trichloroethane
<0.01
Vinyl acetate
73.28
Formaldehyde
3,708.16
Formaldehyde
36.34
Acrylic acid
2,916.21
Acetaldehyde
33.61
Nickel, PM
1,773.75
-------
Table 12-8. 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
Top 10 Noncancer Hazard Approximations
Noncancer RfCs
Top 10 Noncancer Toxicity-Weighted Emissions
Based on Annual Average Concentrations
(County-Level)
(County-Level)
(Site-Specific)
Noncancer
Noncancer
Hazard
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(HQ)
Smithland, Kentucky (Livingston County) - BLKY
Toluene
43.04
Acrolein
20,492.52
1,3-Butadiene
0.09
Xylenes
38.50
Formaldehyde
1,175.61
Benzene
0.02
Benzene
14.04
1,3-Butadiene
909.04
Carbon Tetrachloride
0.01
Hexane
12.08
Acetaldehyde
740.44
Vinyl chloride
<0.01
Formaldehyde
11.52
Cyanide Compounds, gas
527.46
1,2-Dichloroethane
<0.01
Acetaldehyde
6.66
Benzene
467.89
Hexachloro-1,3 -butadiene
<0.01
Ethylbenzene
5.39
Xylenes
384.98
Methanol
5.38
Nickel, PM
268.68
Ethylene glycol
1.89
Naphthalene
222.51
1.3 -Butadiene
1.82
Manganese, PM
201.76
Lexington, Kentucky (Fayette County) - LEKY
Toluene
487.75
Acrolein
277,725.18
Formaldehyde
0.32
Xylenes
315.94
Formaldehyde
9,416.72
Acetaldehyde
0.17
Hexane
246.40
1,3-Butadiene
9,286.39
Arsenic (PMio)
0.04
Methanol
176.71
Acetaldehyde
6,066.40
1,3-Butadiene
0.03
Benzene
135.46
Benzene
4,515.24
Benzene
0.02
Formaldehyde
92.28
Naphthalene
3,436.18
Carbon Tetrachloride
0.01
Ethylbenzene
82.26
Xylenes
3,159.41
1,2-Dichloroethane
<0.01
Ethylene glycol
59.08
Hexamethylene-1,6-diisocyanate, gas
2,051.30
Acetaldehyde
54.60
Arsenic, PM
1,982.63
Methyl isobutyl ketone
29.90
4,4'-Methylenediphenyl diisocyanate, gas
1,757.48
-------
Observations from Table 12-7 include the following:
• Among the Kentucky counties with monitoring sites, emissions (for pollutants with
cancer UREs) are highest in Fayette County (LEKY) and Marshall County (Calvert
City) and lowest in Livingston County (BLKY) and Carter County (GLKY).
• Benzene, formaldehyde, ethylbenzene are the highest emitted pollutants with cancer
UREs in Boyd County, where the Ashland sites are located. Coke oven emissions,
hexavalent chromium, and nickel 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.
• For ASKY, formaldehyde, benzene, and 1,3-butadiene are among the pollutants with
the highest cancer risk approximations and appear on both emissions-based lists.
Acetaldehyde, which has the fourth highest cancer risk approximation for ASKY, has
the fourth highest emissions for Boyd County but is not among the pollutants with the
highest toxicity-weighted emissions (acetaldehyde ranks 13th for toxicity-weighted
emissions). Carbon tetrachloride and 1,2-dichloroethane, the other pollutants of
interest for ASKY, 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. While cadmium ranks ninth in Boyd County for its
toxicity-weighted emissions, it is not among the highest emitted (ranking 18th).
Arsenic, which has the highest cancer risk approximation for ASKY-M, appears on
neither emissions-based list (ranking 24th for total emissions and 15th for toxicity-
weighted emissions).
• Benzene, formaldehyde, acetaldehyde, and ethylbenzene are the highest emitted
pollutants with cancer UREs in Carter County, where GLKY is located.
Formaldehyde, benzene, 1,3-butadiene, and naphthalene are the pollutants with the
highest toxicity-weighted emissions (of the pollutants with cancer UREs) for this
county. Nine of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Carter County (all but one of which are sampled for at GLKY).
• Formaldehyde has the highest cancer risk approximation for GLKY, and ranks first
for its toxicity-weighted emissions and second for its total emissions in Carter
County, as shown in Table 12-7. Benzene, 1,3-butadiene, and acetaldehyde also
appear on all three lists. The three remaining pollutants of interest appear on neither
emissions-based list.
• Three POM Groups appear among the highest emitted pollutants in Carter County
(POM, Groups 2b, 2d, and 6) 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.
• Formaldehyde, benzene, and acetaldehyde are the highest emitted pollutants with
cancer UREs in Henderson County, where BAKY is located. Formaldehyde,
12-91
-------
naphthalene, and POM Group 2d 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 22nd 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 four of the five Calvert City sites are
located. Benzene, hexavalent chromium, and formaldehyde 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 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 (30.93 tpy) is the highest emissions for this pollutant among NMP
counties and is considerably higher than the next highest emissions of this pollutant
(1.77 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 NMP site that has carbon
tetrachloride emissions greater than 1 tpy (2.32 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
(9.25 tpy) again ranks highest for emissions, with Los Angeles County the next
closest at 1.34 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 an exception, as this
pollutant appears among the highest emitted but not those with the highest toxicity-
weighted emissions (ranking 16th). Hexachloro-l,3-butadiene is a pollutant of
interest for all four sites in Marshall County but does not appear on either emissions-
based list. 1,1,2-Trichloroethane is a pollutant of interest for TVKY and appears on
neither emissions-based list. Emissions of this pollutant are highest in Marshall
County compared to other counties with NMP sites.
Arsenic is the only pollutant of interest among the speciated metals sampled for at
CCKY. Arsenic appears on neither emissions-based list for Marshall County (arsenic
ranks 25th for total emissions and 13th for toxicity-weighted emissions).
12-92
-------
• Benzene, formaldehyde, and acetaldehyde 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.
• Few of BLKY's pollutants of interest appear among the pollutants on the emissions-
based lists for Livingston County (only 1,3-butadiene and benzene).
• Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
cancer UREs in Fayette County, where LEKY 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 Fayette County.
• Formaldehyde has the highest cancer risk approximation among LEKY's pollutants
of interest. Formaldehyde, benzene, acetaldehyde, and 1,3-butadiene appear among
the highest emitted pollutants in Fayette County and appear among those with the
highest toxicity-weighted emissions. Arsenic, another pollutant of interest for LEKY,
appears among those with the highest toxicity-weighted emissions but ranks 22nd for
total emissions. The two remaining pollutants of interest, carbon tetrachloride and
1.2-dichloroethane, appear on neither emissions-based list.
Observations from Table 12-8 include the following:
• Among the Kentucky counties with monitoring sites, emissions (for pollutants with
noncancer RfCs) are highest in Marshall County (Calvert City) and Fayette County
(LEKY) and lowest in Carter County (GLKY) and Livingston County (BLKY).
• Toluene, benzene, 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. Four 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, two (formaldehyde and benzene) also appear
on both emissions-based lists. Acetaldehyde 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 two pollutants of interest for ASKY appear
on neither emissions-based list in Table 12-8.
12-93
-------
Manganese ranks tenth for its total emissions and has the third highest toxicity-
weighted emissions for Boyd County. Nickel 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. Arsenic is the only
pollutant of interest for ASKY-M that does not appear in either emissions-based list.
Toluene, xylenes, and hexane are the highest emitted pollutants with noncancer RfCs
in Carter County. Acrolein, formaldehyde, and cyanide compounds (gaseous) 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.
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, benzene, and 1,3-butadiene appear on all three lists for
GLKY. Arsenic is among the pollutants with the highest toxicity-weighted emissions
but is not among the highest emitted in Carter County (its emissions rank 32nd).
Carbon tetrachloride and 1,2-dichloroethane, the remaining two pollutants of interest
for GLKY, appear on neither emissions-based list for Carter County.
Carbonyl sulfide, toluene, and xylenes 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, manganese, and nickel 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 18th 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, xylenes, and toluene are the highest emitted pollutants with noncancer
RfCs in Marshall County. Chlorine, acrolein, and 1,3-butadiene are the pollutants
with the highest toxicity-weighted emissions (of the pollutants with noncancer RfCs)
for this county. This is the only county with an NMP site for which acrolein was not
the pollutant with the highest toxicity-weighted emissions. Five of the highest emitted
pollutants also have the highest toxicity-weighted emissions for Marshall County.
Benzene is the only pollutant of interest for the Calvert City sites to appear on all
three lists. 1,3-Butadiene has the highest nonancer hazard approximation for all four
12-94
-------
Calvert City sites located in Marshall County (as well as the one located in Livingston
County). This pollutant has the third highest toxicity-weighted emissions but is not
among the highest emitted (ranking 14th). None of the other VOC pollutants of
interest for the Calvert City sites appear on either emissions-based list for Marshall
County. This is also true for arsenic, a pollutant of interest for CCKY.
• 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.
• 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.
• 1,3-Butadiene and benzene have the highest noncancer hazard approximations for
BLKY. These pollutants appear on both emissions-based lists for Livingston County
but are the only pollutants of interest for BLKY to do so.
• Toluene, xylenes, and hexane are the highest emitted pollutants with noncancer RfCs
in Fayette 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. Four 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.
• Formaldehyde, acetaldehyde, and benzene appear on all three lists in Table 12-8 for
LEKY. 1,3-Butadiene and arsenic rank third and ninth, respectively, for their toxicity-
weighted emissions but are not among the highest emitted in Fayette County.
12.6 Summary of the 2014 Monitoring Data for the Kentucky Monitoring Sites
Results from several of the data analyses described in this section include the following:
~~~ Eight monitoring sites sampledfor VOCs; five monitoring sites sampledfor PMio
metals; three monitoring sites sampledfor carbonyl compounds; and PAHs were
sampledfor at GLKY. Sampling at the CCKY site was discontinued in October 2014
and the metals instrumentation was moved to BLKY, where sampling was initiated at
the end of October.
12-95
-------
The number ofpollutants failing screens for the Kentucky sites varies from two
(BAKY) to 11 (TVKY and LEKY).
ASKY-M has the highest annual average concentrations of arsenic and nickel among
NMP sites sampling PM 10 metals, similar to 2013. Three additional Kentucky sites
(BAKY, LEKY, and CCKY) are among the NMP sites with the highest annual average
concentrations of arsenic and BAKY and LEKY are also among the sites with the
highest annual average concentrations of nickel.
The two highest benzene concentrations measured across the program were
measured at ASKY and TVKY, which have the ninth and sixth highest annual average
concentration of benzene, respectively, among NMP sites sampling this pollutant.
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, and 1,2-dichloroethane among NMP sites
sampling VOCs. Further, the annual averages for all five Calvert City sites rank in
the top five among NMP sites for carbon tetrachloride and 1,2-dichloroethane.
The maximum formaldehyde concentration measured across the program was
measured at LEKY; this site ranks ninth highest for its annual average concentration
compared to other NMP sites sampling carbonyl compounds.
The cancer risk approximation for 1,2-dichloroethane for TVKY is the highest cancer
risk approximation calculated among site-specific pollutants of interest.
12-96
-------
13.0 Site in Massachusetts
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS site in Massachusetts, and integrates these concentrations
with emissions, meteorological, and risk information. Data generated by sources other than ERG
are not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) 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 geographical and
physical information about the location of the site and the surrounding area. 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 BOMA monitoring site is located in Boston. Figure 13-1 is 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 2011 NEI for point sources, version 2. 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 in order to
emphasize emissions sources within the boundary. Table 13-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates.
13-1
-------
Figure 13-1. Boston, Massachusetts (BOMA) Monitoring Site
Bostori'Medica!
Jim Rice Field
School \ .-A
lalcolm X Bivd
."John D
Obryant
SCHL
Math &
Sen nee
Roxbury^'yKetgl
Warren Pi
Highland
Park
£,76 ft
rf) Source: USGS . x-
jrc'eV NASA. NGA.^USG
20 0 8 MicrostffT^orp,
• John I
Connolly
to
-------
Figure 13-2. NEI Point Sources Located Within 10 Miles of BO MA
71°15'0"W
71°10'0"W
71"5'0"W
71C0'0"W
Essex
County
Middlesex
County
£ Norfolk
/""County
Suffolk
County
Norfolk
County
71°15'0*W
71°10'0"W
71°5'0"W
Middlesex
\
\
\
V
\
* i
\ « LO
* " \ ^
Boston ^
Harbor \
71°10,0"W 71 °5'0"W 71°0'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
Miles
71°25'0"W 71°20'0"W 71"15,0"W
Legend
~
BOMA NATTS site O 10 mile radius
County boundary
Source Category Group (No. of Facilities)
*
T
«
B
C
Aerospace/Aircraft Manufacturing Facility (1)
Airport/Airline/Airport Support Operations (18)
Asphalt Production/Hot Mix Asphalt Plant (2)
Automobile/Truck Manufacturing Facility (1)
Bulk Terminals/Bulk Plants (9)
Chemical Manufacturing Facility (3)
CX] Crematory-Animal/Human (1)
0 Dry Cleaning Facility (2)
G Electrical Equipment Manufacturing Facility (1)
f Electricity Generation via Combustion (8)
E Electroplating, Plating, Polishing, Anodizing, and Coloring (4)
F Food Processing/Agriculture Facility (3)
1 Foundries, Iron and Steel (1)
> Hotels/Motels/Lodging (1)
Industrial Machinery or Equipment Plant (1)
O Institutional (school, hospital, prison, etc.) (43)
A Metal Coating, Engraving, and Allied Services to Manufacturers (1)
Metals Processing/Fabrication Facility (2)
Military Base/National Security Facility (1)
Mine/Quarry/Mineral Processing Facility (2)
Miscellaneous Commercial/Industrial Facility (38)
Municipal Waste Combustor (1)
Oil and/or Gas Production (2)
Paint and Coating Manufacturing Facility (1)
Pharmaceutical Manufacturing (1)
Plastic, Resin, or Rubber Products Plant (1)
Printing/Publishing/Paper Product Manufacturing Facility (1)
Rail Yard/Rail Line Operations (1)
Ship/Boat Manufacturing or Repair Facility (1)
Telecommunications/Radio Facility (2)
Testing Laboratories (1)
Truck/Bus/Transportation Operations (3)
Wastewater Treatment Facility (3)
Water Treatment 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
3AADT reflects 2010 data (MA DOT, 2010)
BOLD ITALICS = EPA-designated NATTS Site
u>
-U
-------
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 a city bus terminal
located another block west of the monitoring site. In recent years, the buses servicing the area
were converted to compressed natural gas (CNG). The monitoring site is 1.3 miles south of 1-90
and 1 mile west 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; bulk terminals and bulk plants; and electricity generating units (via combustion).
Sources located within 1 mile of BOMA include several hospitals, a heliport at one of the
hospitals, a university, and a dry cleaning facility. Figure 13-2 shows that BOMA is located less
than 2 miles from the shoreline (Dorchester Bay).
In addition to providing city, county, CBSA, 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 effect 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 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
site in Massachusetts on sample days, as well as over the course of the year.
13-5
-------
13.2.1 Meteorological Summary
In order to provide an overview of the meteorological conditions experienced at each
monitoring site, hourly meteorological data for 2014 were retrieved and sample day and full-year
averages developed for temperature, dew point, relative humidity, pressure, and wind speed.
Weather data from the actual monitoring site(s) were obtained from AQS, where available. If
site-specific weather data were not available in AQS, then data were obtained from NCDC for
the NWS weather station located closest to the monitoring site(s), as described in Section 3.4.2.
For BOMA, site-specific data were available all of the parameters except dew point temperature
and sea level pressure. Data for these parameters were obtained from the NWS weather station at
Logan International Airport (WBAN 14739). The Logan Airport weather station is located 4.3
miles east-northeast of BOMA. A map showing the distance between the BOMA monitoring site
and the closest NWS weather station is provided in Appendix R. These data were used to
determine how meteorological conditions on sample days vary from conditions experienced
throughout the year.
Table 13-2. Average Meteorological Conditions near the Massachusetts Monitoring Site
Average
Type1
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(in Hg)
Average
Station
Pressure
(in Hg)
Prevailing
Wind
Direction
Average
Scalar Wind
Speed
(kt)
Boston, Massachusetts - BOMA2
Sample
Days
50.9
37.7
58.0
30.00
29.94
4.6
(62)
± 1.0
± 1.1
±0.9
±0.01
±0.01
SW
±0.1
51.4
38.4
58.6
30.00
29.94
4.5
2014
±0.4
±0.4
±0.4
±0.01
±0.01
sw
±<0.1
'Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
2Temperature, humidity, station pressure and wind parameters were measured at BOMA. The remaining information was
obtained from the closest NWS weather station located at Logan International Airport, WBAN 14739.
Table 13-2 presents average temperature, average dew point temperature, average relative
humidity, average station and sea level pressure, and wind information (average scalar wind
speed and prevailing wind direction) for days on which samples were collected and for all of
2014. Also included in Table 13-2 is the 95 percent confidence interval for each parameter. As
shown in Table 13-2, average meteorological conditions on sample days were representative of
average weather conditions experienced throughout the year at BOMA.
13-6
-------
13.2.2 Wind Rose Comparison
Hourly surface wind data were also uploaded into a wind rose software program to
produce customized wind roses, as described in Section 3.4.2. A wind rose shows the frequency
of wind directions using "petals" positioned around a 16-point compass, and uses different colors
to represent wind speeds. Figure 13-3 presents two wind roses for the BOMA monitoring site.
The first is a wind rose representing wind observations for all of 2014 and the second is a wind
rose representing wind observations for days on which samples were collected in 2014. These
are used to identify the predominant wind speed and direction for 2014 and to determine if wind
observations on sample days were representative of conditions experienced over the entire year.
Figure 13-3. Wind Roses for the Wind Data Collected at BOMA
2014 Wind Rose Sample Day Wind Rose
WEST
; east! ;WEST
1 EAST
WIND SPEED
(Knots)
HI >= 22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 0.00%
WIND SPEED
(Knots)
HI >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 0.00%
Observations from Figure 13-3 for BOMA include the following:
• Winds from the south-southwest to southwest account for the majority (about one-
third) of wind observations at BOMA in 2014. Winds from the southwest quadrant,
including west, and winds from the north to northeast are the only wind directions
accounting for more than 5 percent of observations. Calm winds were not measured at
BOMA.
• The sample day wind patterns generally resemble the full-year wind patterns. Winds
from the southwest to west account for an even higher percentage of winds
observations on sample days at BOMA.
13-7
-------
13.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for the
Massachusetts monitoring 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 13-3.
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 13-3. 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 13-3. Risk-Based Screening Results for the Massachusetts Monitoring Site
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
Boston, Massachusetts - BOMA
Arsenic (PMio)
0.00023
42
57
73.68
42.86
42.86
Naphthalene
0.029
40
57
70.18
40.82
83.67
Nickel (PMio)
0.0021
15
58
25.86
15.31
98.98
Benzo(a)pyrene
0.00057
1
57
1.75
1.02
100.00
Total
98
229
42.79
Observations from Table 13-3 include the following:
• Four pollutants failed at least one screen for BOMA; approximately 43 percent of
concentrations for these four pollutants were greater than their associated risk
screening value (or failed screens).
• Three pollutants contributed to 95 percent of failed screens for BOMA and therefore
were identified as pollutants of interest for this site. These include two PMio metals
(arsenic and nickel) and one PAH (naphthalene).
• Naphthalene and arsenic each account for more than 40 percent of the total failed
screens for BOMA while nickel accounts for 15 percent of the total failed screens.
13-8
-------
13.4 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.
• 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 are presented from previous years
of sampling in order to characterize concentration trends at the site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at BOMA are provided in Appendices M and N.
13.4.1 2014 Concentration Averages
Quarterly and annual concentration averages 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 the entire year of sampling. Annual averages were calculated for
pollutants where three valid quarterly averages could be calculated 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 13-4, 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.
13-9
-------
Table 13-4. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Massachusetts Monitoring Site
# of
Measured
1st
2nd
3rd
4th
Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs.
# >MDL
# of
Samples
Average
(ng/m3)
Average
(ng/m3)
Average
(ng/m3)
Average
(ng/m3)
Average
(ng/m3)
Boston, Massachusetts
-BOMA
0.45
0.32
0.50
0.29
0.39
Arsenic (PMio)
57/45
58
±0.14
±0.14
±0.09
±0.09
±0.06
53.05
43.18
52.97
28.70
44.55
Naphthalene
57/57
57
± 19.45
± 15.25
±9.22
±3.52
±6.72
2.25
2.84
1.60
1.28
1.99
Nickel (PMio)
58/58
58
±0.81
± 1.34
±0.38
±0.48
±0.42
Observations for BOMA from Table 13-4 include the following:
• Naphthalene is the pollutant with the highest annual average concentration
(44.55 ± 6.72 ng/m3) among BOMA's pollutants of interest. The annual average
concentrations for the remaining pollutants of interest are at least an order of
magnitude lower.
• Concentrations of naphthalene measured at BOMA range from 13.8 ng/m3 to
158 ng/m3. The maximum naphthalene concentration measured at BOMA was
measured on March 12, 2014 and is the only naphthalene concentration measured at
this site that is greater than 100 ng/m3.
• Concentrations of naphthalene appear more variable during the first half of the year,
based on the confidence intervals calculated for the quarterly averages. Both the
minimum and maximum concentrations were measured during the first half of the
year, about one month apart. Nine naphthalene concentrations greater than 50 ng/m3
were measured at BOMA during the first half of 2014, and nine were also measured
during the second half. For the first half of the year, these measurements were spread
out across the calendar quarters; for the second half of the year, all nine were
measured during the third quarter of 2014 (none were measured during the fourth
quarter).
• Concentrations of arsenic measured at BOMA in 2014 are all less than 1 ng/m3,
ranging from 0.007 ng/m3 to 0.98 ng/m3 and one non-detect. Both the maximum and
minimum (non-detect) concentrations of arsenic were measured during the second
quarter of 2014. The second and fourth quarter averages are more similar to each
other and the first and third quarterly averages are more similar to each other, with
confidence intervals that are rather large relative to the averages themselves. Only
three concentrations greater than 0.5 ng/m3 were measured during the second and
fourth quarters, compared to 14 measured during the other two (six were measured
during the first quarter and eight were measured during the third). In addition, all six
arsenic concentrations less than 0.1 ng/m3 were measured at BOMA during the
second and fourth quarters (three measured during each calendar quarter). Arsenic
13-10
-------
concentrations less than 0.1 ng/m3 were not measured during the first and third
quarters of 2014.
• Concentrations of nickel measured at BOMA range from 0.435 ng/m3 to 9.00 ng/m3.
The first and second quarter average concentrations are higher than the other
quarterly averages for 2014 and have more variability associated with their individual
measurements, as their confidence intervals are two to three times higher than the
confidence intervals for the third and fourth quarterly averages. The seven highest
concentrations (those greater than 3.75 ng/m3) were measured at BOMA between
February and June, and 12 of the 16 measurements greater than or equal to 2 ng/m3
were measured during the first half of the year. Several of the lowest nickel
concentrations were measured at BOMA during the fourth quarter, with
concentrations less than 1 ng/m3 accounting for half (seven) of the 14 concentrations
measured between October and December; no other quarter has more than three.
• Table 4-12 presents the NMP sites with the 10 highest annual average concentrations
for each of the program4evel speciated metals pollutants of interest. This table shows
that BOMA has the second highest annual average concentration of nickel among
NMP sites sampling PMio metals (and does not appear in the table for arsenic).
13.4.2 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 13-4 for BOMA. Figures 13-4 through 13-6 overlay the site's minimum, annual average,
and maximum concentrations onto the program-level minimum, first quartile, median, average,
third quartile, and maximum concentrations, as described in Section 3.4.3.1, and are discussed
below.
Figure 13-4. Program vs. Site-Specific Average Arsenic (PMio) Concentration
Program Max Concentration = 10.1 ng/m3
1 2
3
Concentration {ng/m3)
4
5
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site: Site Average
o
Site Concentration Range
13-11
-------
Figure 13-4 presents the box plot for arsenic for BOMA and shows the following:
• The program-level maximum arsenic concentration (10.1 ng/m3) is not shown directly
on the box plot in Figure 13-4 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 concentration of arsenic measured at BOMA is about an order of
magnitude less than the maximum concentration of arsenic measured across the
program.
• The annual average arsenic (PMio) concentration for BOMA is less than the program-
level average concentration (0.61 ng/m3) and the program-level median concentration
(0.45 ng/m3).
Figure 13-5. Program vs. Site-Specific Average Naphthalene Concentration
100
200 300
Concentration {ng/m3)
400
500
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i
Site: Site Average
o
Site Concentration Range
Figure 13-5 presents the box plot for naphthalene for BOMA and shows the following:
• The range of naphthalene concentrations measured at BOMA is much smaller than
the range of concentrations measured across the program.
• The annual average naphthalene concentration for BOMA is less than the program-
level average concentration (66.46 ng/m3) and the program-level median
concentration (50.70 ng/m3).
13-12
-------
Figure 13-6. Program vs. Site-Specific Average Nickel (PMio) Concentration
•
O i
0 2 4 6 8 10
Concentration (ng/m3)
Progra m: 1st Qua rti 1 e
2nd Quartile 3rd Quartile
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 13-6 presents the box plot for nickel for BOMA and shows the following:
• BOMA's maximum nickel concentration of 9.00 ng/m3 is the third highest nickel
concentration measured across the program. The minimum nickel concentration
measured at BOMA is just less than the program-level first quartile. Only two other
NMP sites have minimum nickel concentrations greater than BOMA's.
• BOMA's annual average concentration of nickel is nearly twice the program-level
average concentration and is also greater than the program-level third quartile. As
discussed in the previous section, BOMA has the second highest annual average
concentration of nickel, behind only ASKY-M.
13.4.3 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 3.4.3.2.
BOMA has sampled PMio metals under the NMP since 2003 and PAHs since 2008. Thus,
Figures 13-7 through 13-9 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.
13-13
-------
Figure 13-7. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at BOMA
Ji.
2005
2006 2007 2008
2009
Year
2011 2012 2013 2014
5th Percentile
— Minimum
— Maximum
O 95th Percentile
1 A 1-year average is not presented because there were breaks in sampling during portions of 2004.
Observations from Figure 13-7 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). Only two additional concentrations greater than 2 ng/m3 have been
measured at BOMA, one in 2004 and one in 2006. Arsenic concentrations greater
than 1.5 ng/m3 have not been measured after 2008.
• 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 fall from 0.61 ng/m3 to 0.53 ng/m3, making the changes in the 1-year averages
between 2007 and 2009 more subtle.
13-14
-------
• All of the statistical metrics exhibit a decrease from 2008 to 2009 and again for 2010.
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 continued to increase for
2014 (18) and a non-detect was measured for the first time. The maximum arsenic
concentration measured in 2014 at BOMA is less than 1 ng/m3 for the first time since
the onset of sampling.
Figure 13-8. Yearly Statistical Metrics for Naphthalene Concentrations Measured at BOMA
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 13-8 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.
13-15
-------
• The maximum naphthalene concentration was measured at BOMA on the very first
sample day (May 6, 2008), although a similar measurement was also collected in
2012. Only two additional concentrations greater than 200 ng/m3 have been measured
at BOMA (one each in 2008 and 2009).
• 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 for 2013 and reaching a minimum in 2014.
• With the exception of 2012, the 1-year average concentrations have a decreasing
trend at BOMA, decreasing from 70.33 ng/m3 for 2009 to 44.52 ng/m3 for 2014,
which is the first year the 1-year average is less than 50 ng/m3. (If the maximum
concentration measured in 2012 was excluded from the dataset, the 1-year average
concentration would exhibit virtually no change from 2011 to 2012.) The median
concentration also exhibits this decreasing trend after 2010.
Figure 13-9. Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at BOMA
-L
I
2009
Year
5th Percentile
— Minimum
— Maximum
O 95th Percentile
1 A 1-year average is not presented because there were breaks in sampling during portions of 2004.
Observations from Figure 13-9 for nickel concentrations measured at BOMA include the
following:
• The maximum nickel concentration was measured at BOMA in 2004 (17.2 ng/m3).
Nickel concentrations greater than 9 ng/m3 have not been measured at BOMA since
2005.
13-16
-------
• 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 (ranging from
1.38 ng/m3 for 2011 to 1.42 ng/m3 for 2013). Considerably increases, however, are
shown for 2014, as all of the statistical parameters, except the minimum
concentration, exhibit increases. The 1-year average concentration for 2014 is nearly
2 ng/m3; the 1-year average concentration hasn't been greater than 2 ng/m3 since
2007.
13.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at the BOMA monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4 for
definitions and explanations regarding the various toxicity factors, time frames, and calculations
associated with these risk-based screenings.
13.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for BOMA 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 3.4.3.3 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 13-5, 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 13-5 include the following:
• Among the pollutants of interest for BOMA, naphthalene has the highest annual
average concentration while arsenic has the lowest annual average concentration.
• Although the annual average concentration for naphthalene is two orders of
magnitude greater than the annual average concentration of arsenic, the cancer risk
approximations for these two pollutants are fairly similar (1.68 in-a-million for
arsenic and 1.51 in-a-million for naphthalene). This speaks to the relative toxicity of
one pollutant compared to the other.
13-17
-------
• 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.
Table 13-5. Risk Approximations for the Massachusetts Monitoring Site
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Boston, Massachusetts - BOMA
Arsenic (PMio)
0.0043
0.000015
57/58
0.39
±0.06
1.68
0.03
Naphthalene
0.000034
0.003
57/57
44.55
±6.72
1.51
0.01
Nickel (PMio)
0.00048
0.00009
58/58
1.99
±0.42
0.95
0.02
13.5.2 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 13-6 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
that have cancer toxicity factors. Table 13-6 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 13-6 provides the pollutants with the highest cancer risk approximations (in-a-million) for
BOMA, as presented in Table 13-5. The emissions, toxicity-weighted emissions, and cancer risk
approximations are shown in descending order in Table 13-6. Table 13-7 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 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 13.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
13-18
-------
Table 13-6. 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)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Boston, Massachusetts (Suffolk County) - BOMA
Formaldehyde
143.05
Formaldehyde
1.86E-03
Arsenic (PMio)
1.68
Benzene
137.55
Nickel, PM
1.22E-03
Naphthalene
1.51
Acetaldehyde
66.22
Benzene
1.07E-03
Nickel (PMio)
0.95
Ethylbenzene
64.30
1,3-Butadiene
7.34E-04
1.3 -Butadiene
24.47
Arsenic, PM
4.62E-04
T etrachloroethylene
19.26
Hexavalent Chromium
4.41E-04
Naphthalene
10.82
Naphthalene
3.68E-04
POM, Group 2b
3.41
POM, Group 2b
3.00E-04
Nickel, PM
2.53
Ethylbenzene
1.61E-04
POM, Group 2d
1.66
POM, Group 2d
1.46E-04
-------
Table 13-7. 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)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Boston, Massachusetts (Suffolk County) - BOMA
Toluene
511.53
Acrolein
501,247.48
Arsenic (PMio)
0.03
Hexane
399.62
Nickel, PM
28,149.17
Nickel (PMio)
0.02
Xylenes
271.73
Formaldehyde
14,596.91
Naphthalene
0.01
Formaldehyde
143.05
1.3 -Butadiene
12,234.00
Benzene
137.55
Acetaldehyde
7,357.30
Acetaldehyde
66.22
Arsenic, PM
7,162.73
Ethylbenzene
64.30
Benzene
4,584.90
Methyl isobutyl ketone
55.81
Naphthalene
3,605.47
1.3 -Butadiene
24.47
Cadmium, PM
3,035.10
T etrachloroethylene
19.26
Xylenes
2,717.32
-------
Observations from Table 13-6 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, nickel, and benzene.
• Eight of the highest emitted pollutants in Suffolk County also have the highest
toxicity-weighted emissions.
• All three of BOMA's pollutants of interest appear among the pollutants with the
highest toxicity-weighted emissions for Suffolk County. Nickel and naphthalene are
also among those with the highest total emissions in Suffolk County while arsenic is
not among the highest emitted (it ranks 16th).
• POM, Group 2b ranks eighth for both quantity emitted and its toxicity-weighted
emissions. POM, Group 2b includes several PAHs sampled for at BOMA including
acenaphthene and fluorene, none of which failed a single screen. POM, Group 2d
ranks tenth for both quantity emitted and its toxicity-weighted emissions. POM,
Group 2d does not include any PAHs sampled for at BOMA.
Observations from Table 13-7 include the following:
• Toluene, hexane, 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, nickel, and formaldehyde.
• Five of the highest emitted pollutants in Suffolk County also have the highest
toxicity-weighted emissions.
• All three 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). Cadmium, which
was also sampled for at BOMA but did not fail any screens, also appears among the
pollutants with the highest toxicity-weighted emissions for Suffolk County
13.6 Summary of the 2014 Monitoring Data for BOMA
Results from several of the data analyses described in this section include the following:
~~~ Four pollutants failed screens for BOMA, with naphthalene and arsenic accounting
for a majority of the failed screens.
~~~ Naphthalene had the highest annual average concentration among the pollutants of
interest for BOMA.
13-21
-------
~~~ BOMA has the second highest annual average concentration of nickel for 2014
among NMP sites sampling PMio metals.
~~~ Naphthalene concentrations have an overall decreasing trend at BOMA.
Concentrations of nickel decreased significantly early on during sampling at BOMA,
and, after a few years with little change, exhibited an increase for 2014.
13-22
-------
14.0 Site in Michigan
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS site in Michigan, and integrates these concentrations
with emissions, meteorological, and risk information. Data generated by sources other than ERG
are not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) for detailed discussions and definitions
regarding the various data analyses presented below.
14.1 Site Characterization
This section characterizes the monitoring site by providing geographical and physical
information about the location of the site and the surrounding area. 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 DEMI monitoring site is located in the Detroit-Warren-Dearborn, Michigan CBSA.
Figure 14-1 is 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 2011 NEI for point sources, version 2.
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 in order to emphasize the emissions sources within the boundary. Table 14-1 provides
supplemental geographical information such as land use, location setting, and locational
coordinates.
14-1
-------
Figure 14-1. Dearborn, Michigan (DEMI) Monitoring Site
mm
-------
Figure 14-2. NEI Point Sources Located Within 10 Miles of DEMI
83°20'0"W
83"10'0"W
83"5'0"W
83"0'0"W
82'55'0"W
Macomb '
County \
Oakland
County
Wayne
County
DEMI NATTS site O 10 mile radius
County boundary
Source Category Group (No. of Facilities)
"f Airport/Airline/Airport Support Operations (10)
'i Asphalt Production/Hot Mix Asphalt Plant (4)
5 Automobile/Truck Manufacturing Facility (5)
B Bulk Terminals/Bulk Plants (10)
C Chemical Manufacturing Facility (5)
i Compressor Station (2)
6 Electrical Equipment Manufacturing Facility (1)
# Electricity Generation via Combustion (8)
E Electroplating, Plating, Polishing, Anodizing, and Coloring (4)
F Food Processing/Agriculture Facility (2)
I Foundries, Iron and Steel (1)
Industrial Machinery or Equipment Plant (2)
O Institutional (school, hospital, prison, etc.) (9)
(g) Metal Can, Box, and Other Metal Container Manufacturing (1)
A Metal Coating, Engraving, and Allied Services to Manufacturers
<•> Metals Processing/Fabrication Facility (4)
X Mine/Quarry/Mineral Processing Facility (10)
? Miscellaneous Commercial/Industrial Facility (18)
[m] Municipal Waste Combustor (1)
~ Paint and Coating Manufacturing Facility (2)
Petroleum Products Manufacturing (1)
* Petroleum Refinery (1)
cd Pharmaceutical Manufacturing (1)
R Plastic, Resin, or Rubber Products Plant (3)
P Printing/Publishing/Paper Product Manufacturing Facility (1)
X Rail Yard/Rail Line Operations (6)
V Steel Mill (2)
© Testing Laboratories (4)
* Wastewater Treatment Facility (1)
(4)
CANADA
83°10'0"W 83°5'0"W 83o0'0"W 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.
Miles
¦T— , I
83°25'0"W 03 20'0"W 83°15'0"W
Legend
Lake
St. Clair
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
96,870
1-94 from Ford Plant to Rotunda Dr
1AADT reflects 2014 data (MI DOT, 2014)
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. Note that DEMI
is located approximately 3 miles from the Canadian border, and that no emission sources
information is provided for Canada.
In addition to providing city, county, CBSA, 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 effect concentrations measured at a given monitoring site. The traffic
volume near DEMI is nearly 100,000 and ranks 13th highest among NMP sites. Traffic for
DEMI is provided for 1-94, between the Ford Plant and Rotunda Drive.
14.2 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
site in Michigan on sample days, as well as over the course of the year.
14-5
-------
14.2.1 Meteorological Summary
In order to provide an overview of the meteorological conditions experienced at each
monitoring site, hourly meteorological data for 2014 were retrieved and sample day and full-year
averages developed for temperature, dew point, relative humidity, pressure, and wind speed.
Weather data from the actual monitoring site(s) were obtained from AQS, where available. If
site-specific weather data were not available in AQS, then data were obtained from NCDC for
the NWS weather station located closest to the monitoring site(s), as described in Section 3.4.2.
For the Michigan site, site-specific data were available for some, but not all, of the parameters in
Table 14-2. For DEMI, temperature, pressure, humidity, and wind information was available in
AQS. Data from the NWS weather station at Detroit City Airport (WBAN 14822) were used for
the remaining parameters (sea level pressure and dew point temperature). The Detroit City
Airport weather station is located 10 miles northeast of DEMI. A map showing the distance
between the monitoring site and the closest NWS weather station is provided in Appendix R.
These data were used to determine how meteorological conditions on sample days vary from
conditions experienced throughout the year.
Table 14-2. Average Meteorological Conditions near the Michigan Monitoring Site
Average
Type1
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(in Hg)
Average
Station
Pressure
(in Hg)
Prevailing
Wind
Direction
Average
Scalar Wind
Speed
(kt)
Dearborn, Michigan - DEMI2
Sample
Days
48.6
35.3
65.3
30.06
29.27
5.3
(68)
± 1.1
± 1.0
±0.8
±0.01
±0.01
WNW
±0.1
49.0
36.4
66.6
30.02
29.24
5.5
2014
±0.4
±0.4
±0.3
±0.01
±<0.01
WNW
±0.1
'Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
2Temperature, humidity, station pressure and wind parameters were measured at DEMI. The remaining information was
obtained from the closest NWS weather station located at Detroit City Airport, WBAN 14822.
Table 14-2 presents average temperature, average dew point temperature, average relative
humidity, average station and sea level pressure, and wind information (average scalar wind
speed and prevailing wind direction) for days on which samples were collected and for all of
2014. Also included in Table 14-2 is the 95 percent confidence interval for each parameter.
Average meteorological conditions on sample days near DEMI were fairly representative of
average weather conditions experienced throughout the year. The largest difference between the
parameters shown in Table 14-2 is for relative humidity.
14-6
-------
14.2.2 Wind Rose Comparison
Hourly surface wind data were also uploaded into a wind rose software program to
produce customized wind roses, as described in Section 3.4.2. A wind rose shows the frequency
of wind directions using "petals" positioned around a 16-point compass, and uses different colors
to represent wind speeds. Figure 14-3 presents two wind roses for the DEMI monitoring site. The
first is a wind rose representing wind observations for all of 2014 and the second is a wind rose
representing wind observations for days on which samples were collected in 2014. These are
used to identify the predominant wind speed and direction for 2014 and to determine if wind
observations on sample days were representative of conditions experienced over the entire year.
Figure 14-3. Wind Roses for the Wind Data Collected at DEMI
2014 Wind Rose Sample Day Wind Rose
WIND SPEED
(Knots)
HI >= 22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 0.75%
WIND SPEED
(Knots)
HI >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 0.81%
Observations from Figure 14-3 include the following:
• The full-year wind rose shows that winds from a variety of directions were observed
at DEMI, although winds from the south-southeast, southwest, and west-northwest
each account for nearly 10 percent of the observations. Winds from the western
quadrants were observed more often than those from the eastern quadrants. The
strongest winds were most often observed with winds from the western quadrants.
Calm winds were infrequently observed at DEMI in 2014.
• The sample day wind rose for DEMI bears some resemblance to the full-year wind
rose, although there are also differences. Winds from the west-northwest account for
a slightly greater number of observations on sample days while the number of south-
southeasterly and southwesterly wind observations was fewer. Fewer northwesterly
14-7
-------
winds were observed on sample days while the number of north-northwesterly and
northerly wind observations was higher compared to the full-year wind rose. Similar
to the full-year wind rose, winds from the western quadrants were observed more
often on sample days than those from the eastern quadrants. The percentage of calm
winds on the sample day wind rose is similar to the percentage shown on the full-year
wind rose.
14.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for DEMI 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 14-3. 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 14-3. 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 14-3. Risk-Based Screening Results for the Michigan Monitoring Site
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Dearborn, Michigan - DEMI
Acetaldehyde
0.45
61
61
100.00
12.90
12.90
Formaldehyde
0.077
61
61
100.00
12.90
25.79
Benzene
0.13
60
60
100.00
12.68
38.48
Carbon Tetrachloride
0.17
60
60
100.00
12.68
51.16
Naphthalene
0.029
60
60
100.00
12.68
63.85
1.3 -Butadiene
0.03
59
60
98.33
12.47
76.32
1,2-Dichloroethane
0.038
56
56
100.00
11.84
88.16
Ethylbenzene
0.4
20
60
33.33
4.23
92.39
Fluorene
0.011
13
52
25.00
2.75
95.14
Acenaphthene
0.011
12
59
20.34
2.54
97.67
/?-Dichlorobcnzcnc
0.091
4
14
28.57
0.85
98.52
Fluoranthene
0.011
4
60
6.67
0.85
99.37
Benzo(a)pyrene
0.00057
3
60
5.00
0.63
100.00
Total
473
723
65.42
14-8
-------
Observations from Table 14-3 for DEMI include the following:
• Thirteen pollutants failed at least one screen for DEMI; 65 percent of concentrations
for these 13 pollutants were greater than their associated risk screening value (or
failed screens).
• Nine pollutants contributed to 95 percent of failed screens for DEMI and therefore
were identified as pollutants of interest for this site. These nine include two carbonyl
compounds, five VOCs, and two PAHs.
• The first five pollutants listed in Table 14-3 each failed 100 percent of screens, with
each contributing to approximately 13 percent to the total number of failed screens;
together these five pollutants account for nearly 65 of the total failed screens. One
concentration of 1,3-butadiene was less than the screening value, and thus, this
pollutant only failed 98 percent of its screens. Concentrations of 1,2-dichloroethane
also failed 100 percent of screens. Thus, the first seven pollutants account for nearly
90 percent of the total failed screens. The failure rate was considerably lower for
concentrations of the remaining pollutants.
14.4 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.
• 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 are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each data analysis is performed where the data meet the applicable criteria specified in
the appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at DEMI are provided in Appendices J, L, and M.
14.4.1 2014 Concentration Averages
Quarterly and annual concentration averages were calculated for the pollutants of interest
for the Michigan 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
14-9
-------
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 the entire year of sampling. Annual averages were calculated for
pollutants where three valid quarterly averages could be calculated 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 14-4, 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.
Table 14-4. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Michigan Monitoring Site
# of
Measured
1st
2nd
3rd
4th
Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs.
# >MDL
# of
Samples
Average
(Ug/m3)
Average
(Ug/m3)
Average
(Ug/m3)
Average
(Ug/m3)
Average
(Ug/m3)
Dearborn, Michigan
-DEMI
1.99
2.10
1.99
1.36
1.85
Acetaldehyde
61/61
61
±0.29
±0.27
±0.35
±0.20
±0.15
0.75
0.69
1.02
0.49
0.73
Benzene
60/60
60
±0.12
±0.16
±0.33
±0.07
±0.10
0.10
0.08
0.17
0.07
0.10
1.3 -Butadiene
60/59
60
±0.03
±0.02
±0.07
±0.01
±0.02
0.64
0.70
0.71
0.64
0.67
Carbon Tetrachloride
60/60
60
±0.03
±0.03
±0.03
±0.03
±0.02
0.08
0.08
0.06
0.06
0.07
1,2-Dichloroethane
56/54
60
±0.01
±0.01
±0.02
±0.01
±0.01
0.24
0.37
0.60
0.27
0.37
Ethylbenzene
60/59
60
±0.07
±0.17
±0.22
±0.19
±0.09
2.77
3.76
4.40
2.15
3.25
Formaldehyde
61/61
61
±0.42
±0.78
±0.67
±0.39
±0.35
1.37
9.68
12.60
3.68
6.93
Fluorene3
52/52
60
±0.84
±4.43
±3.29
± 1.14
± 1.76
88.10
136.94
164.80
73.64
116.80
Naphthalene3
60/60
60
±21.92
±40.59
±42.51
± 22.69
± 18.59
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 14-4 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.
14-10
-------
Concentrations of acetaldehyde are significantly lower during the fourth quarter of
2014, compared to the rest of the year, based on the quarterly averages shown.
Concentrations of acetaldehyde measured at DEMI range from 0.809 |ig/m3 to
3.29 |ig/m3. Only one acetaldehyde concentration greater than 2 |ig/m3 was measured
at DEMI during the fourth quarter of 2014, compared to at least seven in each of the
other calendar quarters. Acetaldehyde concentrations less than 1.5 |ig/m3 account for
more than half of the samples collected between October and December (10),
compared to only four measured throughout the rest of the year.
The fourth quarter formaldehyde average is also the lowest of the four shown above,
although the concentrations of formaldehyde are more variable. Concentrations of
formaldehyde measured at DEMI range from 1.17 |ig/m3 to 6.93 |ig/m3. A review of
the data shows that 14 of the 15 formaldehyde concentrations greater than 4 |ig/m3
were measured during the second and third quarters of the year, the majority of which
were measured between June and August. At the other end of the concentration range,
12 of the 13 formaldehyde concentrations less than 2 |ig/m3 were measured at DEMI
during the first or fourth quarters of 2014, predominantly in November and
December.
The third quarter average concentration of benzene is higher than the other quarterly
averages and the associated confidence interval for it is two to three times higher than
the others shown. Benzene concentrations measured at DEMI range from
0.282 |ig/m3 to 2.22 |ig/m3. The three highest concentrations of benzene, all greater
than 1.5 |ig/m3, were measured in August and September. Of the nine benzene
concentrations greater than 1 |ig/m3 measured at DEMI, five were measured during
the third quarter, compared to one during the first quarter, three during the second,
and none during the fourth. The fourth quarter average benzene concentration is less
than half the third quarter average and has the smallest confidence interval among the
averages. Benzene concentrations greater than 1 |ig/m3 were not measured at DEMI
during the fourth quarter of 2014. Eighteen benzene concentrations greater than the
maximum concentration measured during the fourth quarter (0.755 |ig/m3) were
measured at DEMI throughout the rest of the year and are spread fairly evenly across
the quarters. Ten benzene concentrations less than 0.5 |ig/m3 were measured at DEMI
during the fourth quarter compared to six during the rest of the year.
The quarterly average concentrations of 1,3-butadiene have a similar pattern as the
quarterly average concentrations of benzene. The two highest 1,3-butadiene
concentrations (0.430 |ig/m3 and 0.359 |ig/m3) were measured on the same days as the
two highest benzene concentrations. In fact, the eight highest 1,3-butadiene
concentrations were measured on the same days as the eight highest benzene
concentrations, although the order varied a little.
The third quarter average concentration of ethylbenzene is approximately twice the
other quarterly averages and has the largest confidence interval associated with it. A
review of the data shows that the maximum concentration of this pollutant was
measured on December 13, 2014 (1.55 |ig/m3), although a similar concentration also
measured on July 7, 2014 (1.54 |ig/m3). These are the fifth and sixth highest
concentrations of ethylbenzene measured across the program. The number of
14-11
-------
ethylbenzene concentrations greater than 0.5 |ig/m3 measured during the third quarter
(8) is greater than the number measured during the rest of the year (6). In addition, no
concentrations less than 0.1 |ig/m3 were measured during the third quarter, compared
to 10 measured during other calendar quarters (three during the first quarter, one
during the second, and six during the fourth).
• Concentrations of both PAHs appear higher during the warmer months of the year,
although all of the quarterly average concentrations of naphthalene and fluorene have
relatively large confidence intervals associated with them, indicating that the
measurements are highly variable. Concentrations of naphthalene measured at DEMI
span an order of magnitude, ranging from 36.1 ng/m3 to 369 ng/m3. Some of the
highest naphthalene concentrations across the program were measured at DEMI,
including the second and fourth highest concentrations. Seven naphthalene
concentrations measured at DEMI are greater than 200 ng/m3, the second highest
among sites sampling PAHs. The two highest naphthalene concentrations were
measured at DEMI on the same days as the two highest benzene and 1,3-butadiene
concentrations (September 14th and September 26th). There is alignment among the
sampling dates of several of the highest concentrations for each of these three
compounds.
• Concentrations of fluorene measured at DEMI range from 1.63 ng/m3 to 31.4 ng/m3,
and include eight non-detects. The second and third quarter average concentrations
are significantly greater than the other quarterly average concentrations. The 10
highest fluorene concentrations measured at DEMI were measured between June and
September. At the other end of the concentration range, 13 of the 15 fluorene
concentrations less than 3 ng/m3 were measured during the first or fourth quarters,
with the other two measured in April. Further, all eight non-detects were measured
during the first quarter of 2014.
Tables 4-9 through 4-12 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 4-9 for VOCs three times. This site has the sixth highest
annual average concentration of ethylbenzene and the tenth highest annual average
concentration of 1,3-butadiene. This site also has the seventh highest annual average
concentration of carbon tetrachloride; however, with the exception of TVKY, the
difference among the annual average concentrations of this pollutant varies little.
• DEMI appears in Table 4-10 among the NMP sites with the highest annual average
concentration of formaldehyde, ranking eighth highest.
• DEMI has the highest annual average concentration of naphthalene among NMP sites
sampling PAHs, as shown in Table 4-11. DEMI has the most naphthalene
concentrations greater than 100 ng/m3 (29) among NMP sites sampling PAHs.
14-12
-------
14.4.2 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 14-4. Figures 14-4 through 14-12 overlay the Michigan site's minimum, annual average,
and maximum concentrations onto the program-level minimum, first quartile, median, average,
third quartile, and maximum concentrations, as described in Section 3.4.3.1, and are discussed
below.
Figure 14-4. Program vs. Site-Specific Average Acetaldehyde Concentration
4 5 6
Concentration {[j.g/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 14-4 presents the box plot for acetaldehyde for DEMI and shows the following:
• The range of acetaldehyde concentrations measured at DEMI is relatively small
compared to the range measured across the program.
• DEMI's annual average concentration of acetaldehyde is just greater than the
program-level average concentration.
Figure 14-5. Program vs. Site-Specific Average Benzene Concentration
Program Max Concentration = 12.4 |ig/m;
DEMI
o
2
4
6
8
10
Concentration {[j.g/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
14-13
-------
Figure 14-5 presents the box plot for benzene for DEMI and shows the following:
• The program-level maximum benzene concentration (12.4 |ig/m3) is not shown
directly on the box plot in Figure 14-5 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 benzene concentration measured at DEMI is considerably less than the
maximum benzene concentration measured across the program.
• The annual average benzene concentration for this site is similar to the program-level
average concentration.
Figure 14-6. Program vs. Site-Specific Average 1,3-Butadiene Concentration
6
Program Max Concentration = 5.90 ng/m3
.
?
t 1 1 r
0 0.2 0.4 0.6 0.8 1
Concentration (ng/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 14-6 presents the box plot for 1,3-butadiene for DEMI and shows the following:
• Similar to benzene, the program-level maximum 1,3-butadiene concentration
(5.90 |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 to 1 |ig/m3.
• The maximum 1,3-butadiene concentration measured at DEMI is 0.43 |ig/m3,
considerably less than the maximum concentration measured across the program. Yet,
the annual average concentration for DEMI is similar to the program-level average
concentration and both are similar to the program-level third quartile.
• Non-detects were not measured at the Michigan monitoring site.
14-14
-------
Figure 14-7. Program vs. Site-Specific Average Carbon Tetrachloride Concentration
Program Max Concentration = 3.06 |ig/m3
0 0.5 1 1.5 2 2.5
Concentration (ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 14-7 presents the box plot for carbon tetrachloride for DEMI and shows the
following:
• The scale of the box plot in Figure 14-7 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 median and average concentrations are similar and plotted nearly on
top of each other.
• The range of carbon tetrachloride concentrations measured at DEMI is the smallest
range for any NMP site sampling this pollutant.
• Despite this small range of measurements, the annual average concentration of carbon
tetrachloride for DEMI is just greater than the program-level average concentration of
0.64 |ig/m3 (and just less than the program-level third quartile).
Figure 14-8. Program vs. Site-Specific Average 1,2-Dichloroethane Concentration
DEMI
Program Max Concentration = 27.4 ng/m3
t 1 1 r
0 0.2 0.4 0.6 0.8 1
Concentration (ng/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
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
14-15
-------
program-level maximum 1,2-dichloroethane concentration (27.4 |ig/m3) is
considerably greater than the majority of measurements.
• All of the concentrations of 1,2-dichloroethane measured at DEMI are less than the
program-level average concentration of 0.31 |ig/m3, which is being driven by the
measurements at the upper end of the concentration range.
• The annual average concentration of 1,2-dichloroethane for DEMI is nearly
equivalent to the program-level first quartile (0.069 |ig/m3).
Figure 14-9. Program vs. Site-Specific Average Ethylbenzene Concentration
[
\j 1
0 0.5 1 1.5 2 2.5 3 3.5
Concentration (ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 14-9 presents the box plot for ethylbenzene for DEMI and shows the following:
• The maximum ethylbenzene concentration measured at DEMI is approximately half
the maximum concentration measured across the program, although DEMI's
maximum concentration is the fifth highest measurement across the program.
• The annual average concentration of ethylbenzene for DEMI is greater than both the
program-level average and third quartile; recall from the previous section that this site
has the sixth highest annual average concentration of ethylbenzene among NMP sites
sampling this pollutant.
14-16
-------
Figure 14-10. Program vs. Site-Specific Average Fluorene Concentration
O i
Program Max Concentration = 161 ng/m3
KJ 1
i i i i
0 20 40 60 80 100
Concentration {ng/m3)
Program: 1st Quartile
2nd Quartile 3rd Quartile
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 14-10 presents the box plot for fluorene for DEMI and shows the following:
• Like many other pollutants of interest, the scale of the box plot in Figure 14-10 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 is zero and therefore
not visible on the box plot.
• The maximum fluorene concentration measured at DEMI is about one-fifth the
maximum concentration measured across the program.
• The annual average fluorene concentration for DEMI is greater than the program-
level average concentration (6.32 ng/m3) and third quartile. DEMI's annual average is
the third highest among NMP sites sampling PAHs.
Figure 14-11. Program vs. Site-Specific Average Formaldehyde Concentration
I I I I I I 1 1 1
0 3 6 9 12 15 18 21 24 27
Concentration (ng/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 14-11 presents the box plot for formaldehyde for DEMI and shows the following:
• The range of formaldehyde concentrations measured at DEMI falls within a relatively
small range (1.17 |ig/m3 to 6.93 |ig/m3) compared to the range of concentrations
measured across the program. Yet, the annual average concentration for DEMI lies
between the program-level average concentration and the program-level third
14-17
-------
quartile. Recall from the previous section that this site has the eighth highest annual
average concentration of formaldehyde among NMP sites sampling carbonyl
compounds.
• The minimum formaldehyde concentration measured at DEMI is just less than the
program-level first quartile. DEMI is among nine NMP sites where the minimum
formaldehyde concentration measured is greater than 1 |ig/m3.
Figure 14-12. Program vs. Site-Specific Average Naphthalene Concentration
-
0
100
200
300
Concentration {ng/m3)
400
500
Program:
1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site:
Site Average
o
Site Concentration Range
Figure 14-12 presents the box plot for naphthalene for DEMI and shows the following:
• DEMI's maximum naphthalene concentration is the second highest naphthalene
measurement across the program.
• The minimum concentration of naphthalene measured at DEMI is greater than the
program-level first quartile. DEMI has the second highest minimum concentration
among NMP sites sampling this pollutant.
• The annual average concentration of naphthalene for DEMI is just less than twice the
program-level average concentration and is the highest annual average concentration
among NMP sites sampling naphthalene.
14.4.3 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 3.4.3.2.
DEMI has sampled VOCs and carbonyl compounds under the NMP since 2003 and PAHs since
2008. Thus, Figures 14-13 through 14-21 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.
14-18
-------
Figure 14-13. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at DEMI
E
*3 5.0
1
~
T
T
rL
TT
t
I
2004 2005 2006
2007 1 2008 1 2009 2010 2011 2012 2013 2014
Year
5th Percentile - Minimum - Median
95th Percentile
1 A 1-year average is not presented because data from March 2007 to March 2008 was invalidated.
Observations from Figure 14-13 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 from the primary sampler were invalidated between
March 13, 2007 and March 25, 2008 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). 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
median concentrations exhibit an increasing trend after 2009 that levels off for 2012,
with additional increases shown for 2013 and 2014 (1-year average only).
14-19
-------
• The 1-year average concentration for 2014 is at its highest since 2005.
Figure 14-14. Yearly Statistical Metrics for Benzene Concentrations Measured at DEMI
o
o
I
20031 2004 2005
o
n
pL
o
"ft-
A.
~r
'-t-'
2007
2008
2009
2010
2011
2012
4-
o
2013 2014
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented due to low completeness for 2003.
Observations from Figure 14-14 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. Only two other concentrations greater than 5 |ig/m3
have been measured at DEMI, one in 2003 and one in 2007.
• 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 and median concentrations are at a minimum for 2013.
14-20
-------
• Nearly all of the statistical metrics exhibit an increase for 2014, although the 1-year
average and median concentrations are both still less than these parameters for other
year except 2013.
Figure 14-15. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at DEMI
A,
T
I
JL.
"r
"T
20031 2004 2005
2008 2009
Year
2011 2012
2013 2014
O 5th Percentile
— Minimum
— Maximum o 95th Percentile
1 A 1-year average is not presented due to low completeness for 2003.
Observations from Figure 14-15 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 concentrations greater than 0.90 |ig/m3 were measured
in 2004 and 2006.
• 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
maximum 95th percentile was calculated for 2004. This indicates there is a high level
of variability within these measurements.
• There were fewer non-detects in 2005 and 2006, as indicated by the increase in the
median concentration, and even fewer in 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
14-21
-------
8 percent for the years that follow until 2014. There were no non-detects measured in
2014.
• Even as the number of non-detects decreased (and thus, the number of zeros factored
into the calculated 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.
• The 1-year average concentration decreased significantly from 2012 to 2013, as did
the median, both of which are at their lowest since 2009.
• All of the statistical metrics exhibit increases for 2014.
Figure 14-16. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured
at DEMI
1.6
1.4
1.2
1.0
mE
If
u
0.6
0.4
0.2
0.0
2003 1 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year
o 5th Percentile - Minimum - Median - Maximum o 95th Percentile Aversge
1 A 1-year average is not presented due to low completeness for 2003.
Observations from Figure 14-16 for carbon tetrachloride concentrations measured at
DEMI include the following:
• In 2003, the measured detections ranged from 0.32 |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 0.5 |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.
14-22
-------
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.
For 2012, the difference between the 5th and 95th percentiles is less than 0.25 |ig/m3,
even though an increase in the 1-year average and median concentrations is shown.
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.
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. Despite this tightening
of measurements, little change is shown in the central tendency statistics for 2014.
14-23
-------
Figure 14-17. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured
at DEMI
Maximum
Concentration for
2006 is 3.44 ng/m3
~
o
2003 1 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year
o 5th Percentile
— Minimum
— Maximum
o 95th Percentile
• Average
1 A 1-year average is not presented due to low completeness for 2003.
Observations from Figure 14-17 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
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 2013 and 2014.
• As the number of measured detections increase, so do each of the corresponding
statistical metrics shown in Figure 14-17.
• As the number of measured detections increased dramatically for 2012, and the years
following, the 1-year average and median concentrations increased correspondingly.
The median concentration is greater than the 1-year average concentration for each
year from 2012 forward. This is because there were still several non-detects (or zeros)
factoring into the 1-year average concentration for each year, which can pull down an
average in the same manner an outlier can drive an average upward.
14-24
-------
Figure 14-18. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at DEMI
5th Percentile
o
o
-9~
O
2005 2006 2007 2008 2009 2010
Year
2011
-t-
2014
— Minimum
— Maximum
O 95th Percentile
1 A 1-year average is not presented due to low completeness for 2003.
Observations from Figure 14-18 for ethylbenzene concentrations measured at DEMI
include the following:
• The maximum ethylbenzene concentration was measured at DEMI in September
2004 (4.35 |ig/m3). Only two other ethylbenzene concentrations greater than 3 |ig/m3
have been measured at DEMI (one each in 2011 and 2012). Only 11 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.
• The maximum concentration measured exhibits a steady increasing trend between
2008 and 2012, with all of the statistical parameters exhibiting increases for 2010,
and most continuing this increase for 2011.
• While the maximum concentration increased for 2012, the minimum concentration
decreased (and one non-detect was measured). The number of 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.
• For 2013, all of the statistical metrics exhibit decreases, with the exception of the
minimum concentration, as there were no non-detects measured in 2013.
14-25
-------
Concentrations less than 0.25 |ig/m3 account for an even greater percentage of the
measurements, accounting for 27 of the measurements (or more than 40 percent) for
2013. Additional decreases are shown for several of the parameters for 2014, with
concentrations less than 0.25 |ig/m3 accounting for more than half of the
measurements for the first time since 2009.
Figure 14-19. Yearly Statistical Metrics for Fluorene Concentrations Measured at DEMI
160
140
120
100
mE
| 80
c
c
o
u
60
40
20
0
20081 2009 2010 2011 2012 2013 2014
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 2008.
Observations from Figure 14-19 for fluorene 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 fluorene concentration (152 ng/m3) was measured at DEMI on
August 18, 2010. Two additional measurements greater than 100 ng/m3 have been
measured at DEMI (one in August 2008 and another in August 2010). All eight
concentrations greater than 50 ng/m3 were measured in June, July, or August of a
given year and all 40 concentrations greater than or equal to 20 ng/m3 were measured
at DEMI during the second or third quarters of the year (the warmer months of the
year).
• 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
1—
rh X
—
•<
<
—o— -
14-26
-------
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.
• 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
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. The median
concentrations have varied less than 2.25 ng/m3 over the years, ranging from
4.58 ng/m3 (2014) to 6.82 ng/m3 (2010). The 1-year average concentrations exhibit
more variability, ranging from 6.93 ng/m3 (2014) to 12.62 ng/m3 (2010).
Figure 14-20. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at DEMI
I
I
2005 2006
I
~
I
X r*i
i*i
2007 1 2008 1 2009 2010
Year
2011 2012 2013 2014
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because data from March 2007 to March 2008 was invalidated.
14-27
-------
Observations from Figure 14-20 for formaldehyde concentrations measured at DEMI
include the following:
• Recall that 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. Because less than 75 percent of
the samples were valid in 2008, 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.
• The decrease in the 1-year average concentration shown between 2005 and 2006 is
significant (from 5.35 |ig/m3 to 2.92 |ig/m3). The 1-year average concentrations for
the years following 2006 (where they could be calculated) did not vary significantly
through 2011.
• All of the statistical parameters exhibit increases for 2012. A review of the data
shows that the concentrations measured in 2012 were higher in general compared to
2011. For instance, there were seven measurements less than 1 |ig/m3 in 2011 and
only one in 2012. On the higher end of the concentration range, nine concentrations
greater than 4 |ig/m3 were measured in 2011 compared to 21 in 2012.
• While most of the statistical parameters exhibit decreases for 2013, the minimum
concentration measured in 2013 is at its highest in 10 years.
• Little change is shown in the statistical parameters for 2014.
14-28
-------
Figure 14-21. Yearly Statistical Metrics for Naphthalene Concentrations Measured at DEMI
Maximum
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2008.
Observations from Figure 14-21 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 year except 2013 and 2014).
• With the exception of the maximum concentration, all of the statistical parameters
exhibit increases from 2009 to 2010. 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.
• Although all of the statistical parameters exhibit increases for 2014, with the
exception of the minimum concentration, each is at its second-lowest, behind only
2013.
14-29
-------
14.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at the Michigan monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4
for definitions and explanations regarding the various toxicity factors, time frames, and
calculations associated with these risk-based screenings.
14.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Michigan site 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 3.4.3.3 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 14-5, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Observations from Table 14-5 include the following:
• Formaldehyde has the highest annual average concentration for DEMI, followed by
acetaldehyde, benzene, and carbon tetrachloride.
• These four pollutants also have the highest cancer risk approximations for this site,
although the order varies. Formaldehyde's cancer risk approximation is the highest
(42.30 in-a-million), with all other cancer risk approximations an order of magnitude
lower.
• 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 (0.33).
14-30
-------
Table 14-5. Risk Approximations for the Michigan Monitoring Site
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Dearborn, Michigan - DEMI
Acetaldehyde
0.0000022
0.009
61/61
1.85
±0.15
4.07
0.21
Benzene
0.0000078
0.03
60/60
0.73
±0.10
5.68
0.02
1,3-Butadiene
0.00003
0.002
60/60
0.10
±0.02
3.06
0.05
Carbon Tetrachloride
0.000006
0.1
60/60
0.67
±0.02
4.04
0.01
1,2 -Dichloroethane
0.000026
2.4
56/60
0.07
±0.01
1.80
<0.01
Ethylbenzene
0.0000025
1
60/60
0.37
±0.09
0.91
<0.01
Formaldehyde
0.000013
0.0098
61/61
3.25
±0.35
42.30
0.33
Fluorene3
0.000088
52/60
6.93
± 1.76
0.61
Naphthalene1
0.000034
0.003
60/60
116.80
± 18.59
3.97
0.04
- = 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.
14.5.2 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 14-6 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
that have cancer toxicity factors. Table 14-6 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 14-6 provides the 10 pollutants of interest with the highest cancer risk approximations (in-
a-million) for DEMI, as presented in Table 14-5. The emissions, toxicity-weighted emissions,
and cancer risk approximations are shown in descending order in Table 14-6. Table 14-7
presents similar information, but is limited to those pollutants with noncancer toxicity factors.
14-31
-------
Table 14-6. 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)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Dearborn, Michigan (Wayne County) - DEMI
Benzene
524.56
Coke Oven Emissions, PM
8.62E-03
Formaldehyde
42.30
Formaldehyde
438.33
Formaldehyde
5.70E-03
Benzene
5.68
Ethylbenzene
338.52
Benzene
4.09E-03
Acetaldehyde
4.07
Acetaldehyde
254.42
POM, Group 5a
3.22E-03
Carbon Tetrachloride
4.04
1.3 -Butadiene
79.05
Hexavalent Chromium
2.53E-03
Naphthalene
3.97
Naphthalene
45.78
1,3-Butadiene
2.37E-03
1,3-Butadiene
3.06
T etrachloroethylene
30.63
Arsenic, PM
2.06E-03
1,2-Dichloroethane
1.80
T richloroethylene
17.05
Naphthalene
1.56E-03
Ethylbenzene
0.91
Dichloromethane
10.97
Nickel, PM
9.22E-04
Fluorene
0.61
POM, Group 2b
9.34
Ethylbenzene
8.46E-04
-------
Table 14-7. 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)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Dearborn, Michigan (Wayne County) - DEMI
Hydrochloric acid
3,022.43
Acrolein
1,456,276.15
Formaldehyde
0.33
Toluene
2,046.58
Hydrochloric acid
151,121.26
Acetaldehyde
0.21
Hexane
1,276.18
Formaldehyde
44,727.33
1,3-Butadiene
0.05
Xylenes
1,255.32
1,3-Butadiene
39,523.56
Naphthalene
0.04
Methanol
1,113.64
Arsenic, PM
31,862.61
Benzene
0.02
Benzene
524.56
Acetaldehyde
28,268.80
Carbon Tetrachloride
0.01
Formaldehyde
438.33
Nickel, PM
21,350.40
Ethylbenzene
<0.01
Ethylene glycol
384.08
Manganese, PM
21,158.92
1,2-Dichloroethane
<0.01
Ethylbenzene
338.52
Benzene
17,485.46
Acetaldehyde
254.42
Naphthalene
15,259.58
-------
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 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 14.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
Observations from Table 14-6 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.
• Five of the highest emitted pollutants in Wayne County also have the highest toxicity-
weighted emissions.
• Formaldehyde has the highest cancer risk approximation for DEMI. This pollutant
also appears on both emissions-based lists, ranking second for both its quantity
emitted and its toxicity-weighted emissions. Benzene, naphthalene, 1,3-butadiene,
and ethylbenzene are also pollutants of interest that appear on both emissions-based
lists.
• Acetaldehyde has the third highest cancer risk approximation for DEMI and is one of
the highest emitted pollutants in Wayne County but does not appear among those
with the highest toxicity-weighted emissions. This is also true for fluorene, which is
included as part of POM, Group 2b in the NEI.
• Carbon tetrachloride and 1,2-dichloroethane, the two remaining pollutants of interest
shown in Table 14-6, do not appear on either emissions-based list.
Observations from Table 14-7 include the following:
• Hydrochloric acid, toluene, and hexane are the highest emitted pollutants with
noncancer RfCs in Wayne County. Wayne County is one of the few counties with an
NMP site where toluene is the not the highest emitted pollutant in the noncancer
table. The quantity of emissions for the highest ranking pollutants in Table 14-7 is an
order of magnitude higher than the quantity of emissions for the highest ranking
pollutants in Table 14-6.
14-34
-------
• 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 approximation for DEMI (although
none of the pollutants of interest have associated noncancer hazard approximations
greater than 1.0). Formaldehyde emissions rank seventh 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.
• Several metals appear among the pollutants with the highest toxicity-weighted
emissions for Wayne County. (This was also true for the pollutants with cancer UREs
in Table 14-6.) Speciated metals were not sampled for under the NMP through the
contract laboratory.
14.6 Summary of the 2014 Monitoring Data for DEMI
Results from several of the data analyses described in this section include the following:
~~~ Thirteen pollutants failed screens for DEMI, including two carbonyl compounds,
six 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 /ig/m3.
~~~ DEMI has the highest annual average concentration of naphthalene 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. The detection rate of 1,2-dichloroethane has increased significantly at
DEMI during the last few years of sampling.
~~~ Formaldehyde has the highest cancer risk approximation 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.
14-35
-------
15.0 Site in Missouri
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS site in Missouri, and integrates these concentrations with
emissions, meteorological, and risk information. Data generated by sources other than ERG are
not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) 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 geographical and
physical information about the location of the site and the surrounding area. 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 S4MO monitoring site is located in the St. Louis, MO-IL CBSA. Figure 15-1 is 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 2011 NEI for point sources, version 2. 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 in order to
emphasize emissions sources within the boundary. Table 15-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates.
15-1
-------
Figure 15-1. St. Louis, Missouri (S4MO) Monitoring Site
Hyde-Park
A JSouTce:lJLJSG
[So u rc®JNASA" N G A »USg's
*^1' 2bffa^Mcroifiit^Co rp. ^
-------
Figure 15-2. NEI Point Sources Located Within 10 Miles of S4MO
90°10'0"W
Missouri
River
MISSOURI
ILLINOIS
is&issippi
River >
St. Louis
County
St. Louis
City
90°20'0"W
Airport/Airtine/Airport Support Operations (21)
Asphalt Production/Hot Mix Asphalt Plant (6)
Y Breweries/Distilleries/Wineries (1)
f Building/Construction (1)
B Bulk Terminals/Bulk Plants (8)
C Chemical Manufacturing Facility (25)
(\j Coke Battery (2)
1 Compressor Station (2)
^>
-------
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
100,179
1-70 at 1-44 split (at bridge)
1AADT reflects 2013 data (MO DOT, 2013)
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; mines, quarries, and mineral processing facilities; and rail yard/rail line operations.
Within 1 mile of S4MO are a pharmaceutical manufacturing facility, a printing and publishing
facility, a leather products facility, a metals processing/fabrication facility, and a chemical
manufacturing facility.
In addition to providing city, county, CBSA, and land use/location setting information,
Table 15-1 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 effect concentrations measured at a given monitoring site. The
traffic volume experienced near S4MO is just greater than 100,000 and ranks 11th highest
among other NMP sites, which falls in the upper third of the range compared to other NMP sites.
The traffic estimate provided is for 1-70 near the split with 1-44 (at the bridge).
15.2 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
site in Missouri on sample days, as well as over the course of the year.
15.2.1 Meteorological Summary
In order to provide an overview of the meteorological conditions experienced at each
monitoring site, hourly meteorological data for 2014 were retrieved and sample day and full-year
averages developed for temperature, dew point, relative humidity, pressure, and wind speed.
Weather data from the actual monitoring site(s) were obtained from AQS, where available. If
site-specific weather data were not available in AQS, then data were obtained from NCDC for
the NWS weather station located closest to the monitoring site(s), as described in Section 3.4.2.
15-5
-------
For the Missouri site, site-specific data were available for some, but not all, of the parameters in
Table 15-2. For S4MO, temperature, pressure, humidity, and wind information was available in
AQS. Data from the NWS weather station at St. Louis Downtown Airport (WBAN 03960) were
used for the remaining parameters (sea level pressure and dew point temperature). The St. Louis
Downtown Airport weather station is located 6.3 miles south-southeast of S4MO. A map
showing the distance between the monitoring site and the closest NWS weather station is
provided in Appendix R. These data were used to determine how meteorological conditions on
sample days vary from conditions experienced throughout the year.
Table 15-2. Average Meteorological Conditions near the Missouri Monitoring Site
Average
Type1
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(in Hg)
Average
Station
Pressure
(in Hg)
Prevailing
Wind
Direction
Average
Scalar Wind
Speed
(kt)
St. Louis, Missouri - S4MO2
Sample
Days
54.9
41.9
65.8
30.08
29.55
3.1
(61)
± 1.1
± 1.1
±0.9
±0.01
±0.01
SE
±0.1
56.1
43.5
67.4
30.04
29.51
3.2
2014
±0.5
±0.4
±0.4
± <0.01
±<0.01
SE
±<0.1
'Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
2Temperature, humidity, station pressure, and wind parameters were measured at S4MO. The remaining information was
obtained from the closest NWS weather station located at St. Louis Downtown Airport, WBAN 03960.
Table 15-2 presents average temperature, average dew point temperature, average
relative humidity, average station and sea level pressure, and wind information (average scalar
wind speed and prevailing wind direction) for days on which samples were collected and for all
of 2014. Also included in Table 15-2 is the 95 percent confidence interval for each parameter.
Average meteorological conditions on sample days at S4MO were fairly representative of
average weather conditions experienced throughout the year. The difference between the full-
year averages and sample day averages is largest for relative humidity and dew point
temperature, as shown in Table 15-2.
15-6
-------
15.2.2 Wind Rose Comparison
Hourly surface wind data were also uploaded into a wind rose software program to
produce customized wind roses, as described in Section 3.4.2. A wind rose shows the frequency
of wind directions using "petals" positioned around a 16-point compass, and uses different colors
to represent wind speeds. Figure 15-3 presents two wind roses for the S4MO monitoring site.
The first is a wind rose representing wind observations for all of 2014 and the second is a wind
rose representing wind observations for days on which samples were collected in 2014. These
are used to identify the predominant wind speed and direction for 2014 and to determine if wind
observations on sample days were representative of conditions experienced over the entire year.
NORTH
SOUTH
Figure 15-3. Wind Roses for the Wind Data Collected at S4MO
2014 Wind Rose Sample Day Wind Rose
WIND SPEED
(Knots)
HI >= 22
¦ 17-21
WIND SPEED
(Knots)
HI >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
SOUTH
15-7
-------
Observations from Figure 15-3 for S4MO include the following:
• The full-year wind rose shows that winds from the southeast to south and west to
northwest were frequently observed at S4MO, with prevailing winds from the
southeast. Winds from these directions account for more than 50 percent of
observations. North-northwesterly winds and easterly winds are the only other winds
accounting for at least 6 percent of observations. Calm winds were observed for
approximately 5 percent of the hourly wind measurements. Wind speeds greater than
11 knots were rarely observed, but were out of the southeast or west when they were
measured.
• The wind patterns on the sample day wind rose mostly resemble the wind patterns on
the full-year wind rose, although there are a few differences. Fewer winds from the
south and northwest quadrant, including west, were observed on sample days while
winds from the north-northeast, east, and east-southeast were observed more
frequently. The calm rate was slightly higher on sample days.
15.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for the S4MO
monitoring 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 15-3.
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 15-3. It is
important to note which pollutants were sampled for at each site when reviewing the results of
this analysis. VOCs, PAHs, carbonyl compounds, metals (PMio), and hexavalent chromium were
sampled for at S4MO. Hexavalent chromium sampling at S4MO was discontinued after
July 4, 2014.
15-8
-------
Table 15-3. Risk-Based Screening Results for the Missouri Monitoring Site
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
St. Louis, Missouri - S4MO
Acetaldehyde
0.45
60
60
100.00
10.70
10.70
Formaldehyde
0.077
60
60
100.00
10.70
21.39
Benzene
0.13
59
59
100.00
10.52
31.91
Carbon Tetrachloride
0.17
59
59
100.00
10.52
42.42
1,2-Dichloroethane
0.038
58
58
100.00
10.34
52.76
Arsenic (PMio)
0.00023
57
61
93.44
10.16
62.92
1.3 -Butadiene
0.03
57
58
98.28
10.16
73.08
Naphthalene
0.029
49
57
85.96
8.73
81.82
p-Dichlorobenzene
0.091
24
45
53.33
4.28
86.10
Acenaphthene
0.011
12
57
21.05
2.14
88.24
Cadmium (PMio)
0.00056
12
61
19.67
2.14
90.37
Fluorene
0.011
11
53
20.75
1.96
92.34
Hexacliloro -1,3 -butadiene
0.045
11
12
91.67
1.96
94.30
Ethylbenzene
0.4
8
59
13.56
1.43
95.72
Lead (PMio)
0.015
6
61
9.84
1.07
96.79
Manganese (PMio)
0.03
4
61
6.56
0.71
97.50
Nickel (PMio)
0.0021
4
60
6.67
0.71
98.22
Benzo(a)pyrene
0.00057
3
56
5.36
0.53
98.75
1,2-Dibromoethane
0.0017
2
2
100.00
0.36
99.11
Hexavalent Chromium
0.000083
2
25
8.00
0.36
99.47
Acenaphthylene
0.011
1
32
3.13
0.18
99.64
Antimony (PMio)
0.02
1
61
1.64
0.18
99.82
Propionaldehyde
0.8
1
60
1.67
0.18
100.00
Total
561
1,177
47.66
Observations from Table 15-3 include the following:
• Concentrations of 23 pollutants failed at least one screen for S4MO; approximately
48 percent of concentrations for these 23 pollutants were greater than their associated
risk screening value (or failed screens). S4MO has the highest number of individual
pollutants failing screens.
• S4MO failed the second highest number of screens (561) among all NMP sites,
behind only PXSS (refer to Table 4-8 of Section 4.2). Yet, the failure rate for S4MO,
when incorporating all pollutants with screening values, is approximately 22 percent.
This is due primarily to the relatively large number of pollutants sampled for at this
site, as discussed in Section 4.2.
15-9
-------
• Fourteen pollutants contributed to 95 percent of failed screens for S4MO and
therefore were identified as pollutants of interest for this site. These 14 pollutants
include two carbonyl compounds, seven VOCs, two PMio metals, and three PAHs.
S4MO has the greatest number of pollutants of interest among NMP sites.
• Acetaldehyde, formaldehyde, benzene, carbon tetrachloride, 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 also failed 100 percent of screens but was
detected in only two VOC samples collected and is not a pollutant of interest for
S4MO.
• Cadmium was identified as a pollutant of interest for only two NMP sites sampling
metals: S4MO and ASKY-M.
15.4 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.
• 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 are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at S4MO are provided in Appendices J, L, M, N, and O.
15.4.1 2014 Concentration Averages
Quarterly and annual concentration averages 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
15-10
-------
zeros for non-detects for the entire year of sampling. Annual averages were calculated for
pollutants where at least three valid quarterly averages could be calculated 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 15-4, 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.
Table 15-4. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Missouri Monitoring Site
# of
Measured
1st
2nd
3rd
4th
Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs.
# >MDL
# of
Samples
Average
(Hg/m3)
Average
(Hg/m3)
Average
(Hg/m3)
Average
(Hg/m3)
Average
(ng/m3)
St. Louis, Missouri -
S4MO
2.38
2.71
1.88
1.42
2.08
Acetaldehyde
60/60
60
±0.48
±0.47
±0.30
±0.31
±0.22
0.80
0.46
0.81
0.75
0.71
Benzene
59/59
59
±0.16
±0.06
±0.16
±0.13
±0.07
0.09
0.05
0.12
0.11
0.09
1.3 -Butadiene
58/58
59
±0.03
±0.01
±0.03
±0.03
±0.02
0.59
0.65
0.65
0.60
0.62
Carbon Tetrachloride
59/59
59
±0.07
±0.03
±0.03
±0.06
±0.02
0.10
0.09
0.17
0.15
0.13
/?-Dichlorobcnzcnc
45/25
59
±0.09
±0.07
±0.05
±0.14
±0.05
0.08
0.07
0.08
0.09
0.08
1,2-Dichloroethane
58/58
59
±0.02
±0.01
±0.01
±0.01
±0.01
0.21
0.15
0.32
0.24
0.23
Ethylbenzene
59/59
59
±0.08
±0.04
±0.07
±0.07
±0.04
2.89
5.13
4.19
1.83
3.45
Formaldehyde
60/60
60
±0.49
±0.96
± 1.02
±0.34
±0.48
0.02
0.01
0.03
0.01
0.02
Hexachloro -1,3 -butadiene
12/0
59
±0.02
±0.01
±0.02
±0.02
±0.01
1.75
12.55
3.86
6.43
Acenaphthene3
57/56
57
±0.67
NA
±2.72
±2.06
± 1.56
0.71
0.68
1.29
0.93
0.90
Arsenic (PMi0)a
61/58
61
±0.27
±0.20
±0.39
±0.26
±0.15
0.31
0.54
0.33
0.16
0.33
Cadmium (PMi0)a
61/61
61
±0.12
±0.32
±0.12
±0.05
±0.09
2.47
12.10
4.22
6.65
Fluorene3
53/53
57
± 1.05
NA
±2.41
± 1.64
± 1.38
62.17
109.13
79.80
81.79
Naphthalene1
57/57
57
± 22.87
NA
±26.39
±23.88
± 12.61
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
15-11
-------
Observations for S4MO from Table 15-4 include the following:
• The pollutants with the highest annual average concentrations are formaldehyde
(3.45 ± 0.48 |ig/m3) and acetaldehyde (2.08 ± 0.22 |ig/m3). These are the only
pollutants of interest with annual averages greater than 1 |ig/m3.
• Concentrations of formaldehyde measured at S4MO range from 0.715 |ig/m3 to
9.02 |ig/m3. The quarterly average concentrations of formaldehyde vary considerably,
with the second quarter average nearly three times greater than the fourth quarter
average, and two of the four have relatively large confidence intervals. Formaldehyde
concentrations less than 2 |ig/m3 were not measured at S4MO during the second or
third quarters while 10 were measured during the fourth quarter (and two were
measured at S4MO during the first quarter). At the higher end of the concentration
range, formaldehyde concentrations greater than 5 |ig/m3 were not measured at
S4MO during the first or fourth quarters of the year while 11 were measured during
the second and third quarters (seven during the second and four during the third).
• Concentrations of acetaldehyde measured during the second half of the year appear
lower than those measured during the first half of the year, based on the quarterly
average concentrations shown. A review of the data shows that nine acetaldehyde
concentrations greater than 3 |ig/m3 were measured during the first half of 2014
compared to only 1 during the second half. At the other end of the concentration
range, only one acetaldehyde concentration less than 1 |ig/m3 was measured during
the first half of 2014 compared to six during the second half of the year.
• The second quarter average concentration of benzene is significantly less than the
quarterly averages shown for the remaining quarterly averages. Concentrations of
benzene measured at S4MO in 2014 range from 0.307 |ig/m3 to 1.47 |ig/m3. None of
the nine benzene concentrations greater than or equal to 1 |ig/m3 were measured
during the second quarter while three were measured during each of the other
calendar quarters. In addition, the number of benzene concentrations less than
0.5 |ig/m3 measured at S4MO during the second quarter (8) is greater than the number
measured throughout the rest of the year combined (5), with no more than two in any
of the other calendar quarters.
• Concentrations of 1,3-butadiene follow a similar pattern as benzene in that the second
quarter average is significantly lower than the other quarterly averages. A review of
the data shows that none of the 18 1,3-butadiene concentrations greater than
0.1 |ig/m3 were measured during the second quarter of 2014, and were predominantly
measured during the second half of the year (three were measured during the first
quarter, eight during the third, and seven during the fourth). The number of
concentrations measured each quarter that are greater than the maximum
concentration measured during the second quarter (0.0709 |ig/m3) ranged from seven
(first quarter) to 12 (third quarter). The one non-detect of 1,3-butadiene was measured
at S4MO in April.
• The quarterly average concentrations of />dichlorobenzene for the second half of
2014 are higher than those for the first half of the year, although the confidence
intervals shown are relatively large, particularly for the fourth quarter of 2014.
15-12
-------
Concentrations of/>dichlorobenzene measured at S4MO range from 0.0361 |ig/m3 to
1.14 |ig/m3 and include 14 non-detects. The maximum /?-dichlorobenzene
concentration measured at S4MO was measured in October and is the highest
/;-dichlorobenzene concentration measured across the program. The second highest
/;-dichlorobenzene concentration measured at S4MO was measured in March and is
half as high. At least one non-detect was measured during each calendar quarter,
although the number varies across the quarters (six during the first quarter, three
during the second, one during the third, and four during the fourth).
Concentrations of arsenic also appear higher during the second half of the year at
S4MO, particularly the third quarter of 2014. A review of the data shows that arsenic
concentrations measured at S4MO range from 0.133 ng/m3 to 2.96 ng/m3. Of the 23
arsenic concentrations greater than 1 ng/m3 measured at S4MO, 17 were measured
after July 1, including all three greater than 2 ng/m3.
The quarterly averages of cadmium exhibit considerably variability, with the second
quarter average concentration more than three times greater than the fourth quarter
average concentration. Concentrations of cadmium measured at S4MO span two
orders of magnitude, ranging from 0.02 ng/m3 to 1.98 ng/m3. The second and fourth
highest concentrations of cadmium across the program were both measured at S4MO
in April. S4MO is one of only four NMP sites where cadmium concentrations greater
than 1 ng/m3 were measured. The maximum cadmium concentration measured during
the fourth quarter is 0.305 ng/m3; at least six cadmium concentrations greater than
0.305 ng/m3 were measured during each of the other calendar quarters.
Naphthalene has the highest annual average concentration among the PAHs identified
as pollutants of interest for S4MO.
Laboratory instrument issues combined with sampler issues at the site resulted in too
many invalid samples for a quarterly average concentration to be calculated for the
second quarter of 2014 for the PAHs.
The quarterly average concentrations of naphthalene are highly variable, with the
third quarter average more than twice the first quarter average, and each of the
confidence intervals is relatively large. This indicates that the concentrations of
naphthalene measured at S4MO exhibit considerable variability. Concentrations of
naphthalene measured at S4MO range from 0.78 ng/m3 to 214 ng/m3. The minimum
concentration measured at S4MO is the lowest naphthalene concentration measured at
an NMP in 2014. It is the only measurement less than 7 ng/m3 at a non-rural
monitoring site and the only one less than 2 ng/m3 at any NMP site sampling this
pollutant. By comparison, the next lowest naphthalene concentration measured at
S4MO is 18.55 ng/m3. The number of naphthalene concentrations greater than
100 ng/m3 measured at S4MO is greatest for the third quarter (nine) and two to three
times greater than the number measured during each of the other calendar quarters
(between three and five were measured during each). Conversely, the number of
naphthalene concentrations less than 50 ng/m3 is greatest for the first quarter (eight)
and lowest for the third quarter (one), with the number for the remaining calendar
quarters falling in-between.
15-13
-------
• Concentrations of acenaphthene and fluorene appear to be highest during the third
quarter, based on the available quarterly average concentrations, although each of the
quarterly averages exhibits a considerable level of variability. A review of the data
shows that the seven highest concentrations of each pollutant were measured on the
same days at S4MO (although the exact order varies). Most of these days are in the
third quarter. For example, of the 17 concentrations of acenaphthene greater than
10 ng/m3, 11 were measured between July and September (with the others measured
in May, June, or October). Conversely, all eight acenaphthene concentrations less
than 1 ng/m3 were measured during the colder months of the year (January, February
November or December). Of the 15 fluorene concentrations greater than 10 ng/m3,
10 were measured between July and September (with the others measured in May,
June, or October). Conversely, all seven fluorene measurements less than 2 ng/m3
were also measured during the colder months of the year (January, February
November or December).
Tables 4-9 through 4-12 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 4-9 through 4-12 a total of seven times, appearing in each
table at least once.
• S4MO has the third highest annual average concentration of />dichlorobenzene and
the 10th highest annual average concentration of 1,2-dichloroethane among NMP
sites sampling VOCs, as shown in Table 4-9.
• S4MO appears in Table 4-10 for both formaldehyde and acetaldehyde, ranking eighth
and sixth, respectively, among NMP sites sampling carbonyl compounds.
• S4MO's annual average concentration of naphthalene ranks sixth highest among
NMP sites sampling PAHs.
• S4MO has the second highest annual average concentration of arsenic and the seventh
highest annual average concentration of nickel among NMP sites sampling PMio
metals.
15-14
-------
15.4.2 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 15-4 for S4MO. Figures 15-4 through 15-17 overlay the site's minimum, annual average,
and maximum concentrations onto the program-level minimum, first quartile, median, average,
third quartile, and maximum concentrations, as described in Section 3.4.3.1, and are discussed
below.
Figure 15-4. Program vs. Site-Specific Average Acenaphthene Concentration
o-
Program Max Concentration = 198 ng/m3
40 60
Concentration {ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 15-4 presents the box plot for acenaphthene for S4MO and shows the following:
• The program-level maximum acenaphthene concentration (198 ng/m3) is not shown
directly on the box plot in Figure 15-4 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. Note that the program-level average
concentration is being influenced by the measurements at the higher end of the
concentration range, as the program-level average concentration is greater than the
program-level third quartile.
• The maximum acenaphthene concentration measured at S4MO is considerably less
than the maximum acenaphthene concentration measured across the program. Yet, the
annual average acenaphthene concentration for this site is greater than the program-
level average concentration.
• There were no non-detects of acenaphthene measured at S4MO while non-detects
account for 6 percent of the measurements at the program level.
15-15
-------
Figure 15-5. Program vs. Site-Specific Average Acetaldehyde Concentration
L
m
^ 1
4 5 6
Concentration (ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 15-5 presents the box plot for acetaldehyde for S4MO and shows the following:
• The maximum acetaldehyde concentration measured at S4MO is roughly half the
maximum concentration measured across the program.
• The annual average concentration of acetaldehyde for S4MO is greater than the
program-level average concentration (1.76 |ig/m3) and less than the program-level
third quartile.
Figure 15-6. Program vs. Site-Specific Average Arsenic (PMio) Concentration
-O-
Program Max Concentration = 10.1 ng/m3
0
1 2
3
Concentration {ng/m3)
4
5
6
Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site: Site Average
o
Site Concentration Range
Figure 15-6 presents the box plot for arsenic (PMio) for S4MO and shows the following:
• The program-level maximum arsenic concentration (10.1 ng/m3) is not shown directly
on the box plot in Figure 15-6 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 arsenic (PMio) concentration measured at S4MO is about one-third the
maximum concentration measured across the program. Yet, S4MO's annual average
arsenic (PMio) concentration is greater than both the program-level average
concentration and third quartile. Recall from the previous section that this site has the
second highest annual average arsenic concentration among NMP sites sampling
15-16
-------
PMio metals. S4MO has the highest number of arsenic concentrations greater than or
equal to 1 ng/m3 among all NMP sites sampling arsenic (23).
Figure 15-7. Program vs. Site-Specific Average Benzene Concentration
Program Max Concentration = 12.4 ng/m3
0 2 4 6 8 10
Concentration (ng/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 15-7 presents the box plot for benzene for S4MO and shows the following:
• The program-level maximum benzene concentration (12.4 |ig/m3) is not shown
directly on the box plot in Figure 15-7 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 benzene concentrations measured at S4MO is relatively small compared
to the range measured at the program-lev el. In fact, the range of benzene
concentrations for S4MO is among the smallest compared to other NMP sites
sampling benzene with Method TO-15.
• The annual average benzene concentration for S4MO is less than the program-level
average concentration but greater than the program-level median concentration.
Figure 15-8. Program vs. Site-Specific Average 1,3-Butadiene Concentration
Program Max Concentration = 5.90 ng/m3
,
t 1 1 r
0 0.2 0.4 0.6 0.8 1
Concentration (ng/m3)
Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
15-17
-------
Figure 15-8 presents the box plot for 1,3-butadiene for S4MO and shows the following:
• Similar to other pollutants, the program-level maximum 1,3-butadiene concentration
(5.905 |ig/m3) is not shown directly on the box plot as the scale has been reduced in
to allow for the observation of data points at the lower end of the concentration range.
• The range of 1,3-butadiene concentrations measured at S4MO spans less than
0.3 |ig/m3 and includes a single non-detect.
• The annual average 1,3-butadiene concentration for S4MO is just less than the
program-level average concentration.
Figure 15-9. Program vs. Site-Specific Average Cadmium (PMio) Concentration
1
—u
t 1 1 1 r
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Concentration {ng/m3)
Program: 1st Quartile
2nd Quartile 3rd Quartile
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 15-9 presents the box plot for cadmium (PMio) for S4MO and shows the
following:
• The program-level maximum cadmium concentration (70.7 ng/m3) is not shown
directly on the box plot in Figure 15-9 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. Note that the second highest cadmium
concentration measured across the program is the maximum cadmium concentration
measured at S4MO (1.98 ng/m3).
• The program-level average cadmium concentration is being driven by the outlier
concentration measured at another site. The program-level average is higher than the
program-level third quartile.
• S4MO's annual average cadmium (PMio) concentration is greater than the program-
level average concentration and is nearly three times greater than the program-level
third quartile. S4MO has the highest number of cadmium concentrations greater than
0.5 ng/m3 among all NMP sites sampling this pollutant (12, with the next highest site
at five).
15-18
-------
Figure 15-10. Program vs. Site-Specific Average Carbon Tetrachloride Concentration
Program Max Concentration = 3.06 ng/m3
0
0.5
1 1.5
Concentration (ng/m3)
2
2.5
Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i
Site: Site Average
o
Site Concentration Range
Figure 15-10 presents the box plot for carbon tetrachloride 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. Note that the
program-level median and average concentrations are similar and plotted nearly on
top of each other.
• The majority of carbon tetrachloride concentrations measured at S4MO fall between
0.45 |ig/m3 and 0.75 |ig/m3; however, the minimum concentration measured at S4MO
(0.189 |ig/m3) is among the lower carbon tetrachloride concentrations measured
among NMP sites sampling this pollutant.
• The annual average concentration of carbon tetrachloride for S4MO is just less than
both the program-level median and average concentrations, although less than
0.01 |ig/m3 separates these two parameters.
Figure 15-11. Program vs. Site-Specific Average />-Dichlorobenzene Concentration
0
0.2 0.4
0.6
Concentration (ng/m3)
0.8
l
1.2
Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i
Site: Site Average
o
Site Concentration Range
15-19
-------
Figure 15-11 presents the box plot for p-dichlorobenzene for S4MO and shows the
following:
• 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 p-dichlorobenzene). Fourteen non-
detects were measured at S4MO.
• The maximum /;-dichlorom ethane concentration measured across the program was
measured at S4MO (1.14 |ig/m3) and is one of only two p-dichlorobenzene
concentrations greater than 1 |ig/m3 measured in 2014. Concentrations measured at
S4MO account for three of the five highest p-dichlorobenzene concentrations
measured in 2014. Recall from the previous section that this site has the third highest
annual average concentration of p-dichlorobenzene among NMP sites sampling
VOCs.
• S4MO is one of only three NMP sites with an annual average concentration of this
pollutant greater than 0.1 |ig/m3. S4MO's annual average concentration is more than
three times greater than the program-level average concentration.
Figure 15-12. Program vs. Site-Specific Average 1,2-Dichloroethane Concentration
h
Program Max Concentration = 27.4 ng/m3
°
0 0.2 0.4 0.6 0.8 1
Concentration (ng/m3)
Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 15-12 presents the box plot for 1,2-dichloroethane for S4MO and shows the
following:
• The scale of the box plot in Figure 15-12 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 (27.4 |ig/m3) is
considerably greater than the majority of measurements.
• All the concentrations of 1,2-dichloroethane measured at S4MO are less than the
program-level average concentration of 0.31 |ig/m3, which is being driven by the
measurements at the upper end of the concentration range.
• The annual average concentration for S4MO is similar to the program-level median
concentration (0.081 |ig/m3).
15-20
-------
Figure 15-13. Program vs. Site-Specific Average Ethylbenzene Concentration
H
0
0.5
l
1.5 2
Concentration (|ig/m3)
2.5
3
Program:
1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site:
Site Average
o
Site Concentration Range
Figure 15-13 presents the box plot for ethylbenzene for S4MO and shows the following:
• The maximum ethylbenzene concentration measured at S4MO is roughly one-fifth
the maximum concentration measured across the program.
• The annual average concentration of ethylbenzene for S4MO is just less than the
program-level average concentration.
Figure 15-14. Program vs. Site-Specific Average Fluorene Concentration
O
Program Max Concentration = 161 ng/m3
40 60
Concentration {ng/m3)
Program: 1st Quartile
¦
2nd Quartile 3rd Quartile
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 15-14 presents the box plot for fluorene for S4MO and shows the following:
• Similar to the box plot for acenaphthene, the program-level maximum fluorene
concentration (161 ng/m3) is not shown directly on the box plot in Figure 15-14 as the
scale of the box plot has been reduced. Note that the first quartile is zero and
therefore not visible on the graph for fluorene due to the number of non-detects.
• The maximum fluorene concentration measured at S4MO is considerably less than
the maximum fluorene concentration measured across the program. Yet, the annual
average fluorene concentration for this site is greater than both the program-level
average concentration and third quartile.
15-21
-------
• Four non-detects of fluorene were measured at S4MO while non-detects account for
more than 25 percent of the measurements at the program-lev el.
Figure 15-15. Program vs. Site-Specific Average Formaldehyde Concentration
¦+
9 12 15
Concentration {jig/m3]
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 15-15 presents the box plot for formaldehyde for S4MO and shows the following:
• The maximum formaldehyde concentration measured at S4MO is roughly one-third
the maximum concentration measured across the program.
• The annual average concentration for S4MO is greater than the program-level
average concentration and just less than the program-level third quartile. Recall from
the previous section that S4MO's annual average concentration ranks sixth highest
among NMP sites sampling formaldehyde.
Figure 15-16. Program vs. Site-Specific Average Hexachloro-l,3-butadiene Concentration
0
0.1
0.2
0.3 0.4
Concentration (ng/m3)
0.5
0.6
0.7
Program:
1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site:
Site Average
o
Site Concentration Range
15-22
-------
Figure 15-16 presents the box plot for hexachloro-1,3-butadiene for S4MO and shows the
following:
• The first, second, and third quartiles for this pollutant are zero and thus, not visible on
the box plot, due to the large number of non-detects (more than 76 percent of the
measurements are non-detects for hexachloro-1,3-butadiene). Forty-seven non-detects
were measured at S4MO, accounting for nearly 80 percent of the sample collected.
• The maximum concentration measured at S4MO is about one-sixth the program-level
maximum concentration.
• The annual average concentration of hexachloro-1,3-butadiene for S4MO is just
slightly less than the program-level average concentration, although most of the site-
specific annual average concentrations of this pollutant vary little.
Figure 15-17. Program vs. Site-Specific Average Naphthalene Concentration
r> i
.
1
i i i i i
0 100 200 300 400 500 600
Concentration {ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 15-17 presents the box plot for naphthalene for S4MO and shows the following:
• Non-detects of naphthalene were not measured in 2014, across the program or at
S4MO, despite the low minimum concentration shown on the box plot. As previously
mentioned, the minimum concentration of naphthalene across the program was
measured at S4MO and is an anomaly, for both this site and for the program. Since
being added to the NATTS program only two lower concentrations of naphthalene
have been measured at an NMP site.
• Despite the low minimum concentration, the annual average concentration of
naphthalene for S4MO is greater than the program-level average concentration and
just less than the program-level third quartile. Recall from the previous section that
S4MO's annual average concentration ranks sixth highest among NMP sites sampling
naphthalene.
15-23
-------
15.4.3 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 3.4.3.2.
S4MO has sampled VOCs and carbonyl compounds under the NMP since 2002, PMio metals
since 2003, and PAHs since 2008. Thus, Figures 15-18 through 15-31 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.
15-24
-------
Figure 15-18. Yearly Statistical Metrics for Acenaphthene Concentrations Measured at
S4MO
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 15-18 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 were 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,
13 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,
with years with lower averages alternating 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.
15-25
-------
Figure 15-19. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at S4MO
X
¦o.
-Si.
L-2-
O"
X
X
^ s - t
2003 2004 2005 2006
2007 2008 2009 2010 2011 2012 2013 2014
Year
5th Percentile
— Minimum
- Median — Maximum
O 95th Percentile
Observations from Figure 15-19 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
metrics 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).
• 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. The 1-year average
concentrations have ranged from 1.83 |ig/m3 (2008) and 4.10 |ig/m3 (2010).
• A significant decrease in the 1-year average concentration is shown from 2007 to
2008, which is followed by an increasing trend through 2010. A significant decrease
is again shown from 2010 through 2012, after which more subtle changes are shown.
A similar pattern is exhibited by the median concentrations. 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.
15-26
-------
Figure 15-20. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at S4MO
Maximum
Concentration for
2007 is 44.1 ng/m3
t y ? t s § i ^ ^
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year
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 2003.
Observations from Figure 15-20 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).
• This figure shows that years with little variability in the measurements seem to
alternate with years with significant variability, particularly between 2004 and 2010.
Less variability in the measurements is shown in the last few years of sampling.
• 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 all 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.
15-27
-------
Figure 15-21. Yearly Statistical Metrics for Benzene Concentrations Measured at S4MO
T
o
1
¦O-
T
pL
o
-A-
K>
±.
2004 2005
2007 2008 2009 2010 2011 2012
Year
2013 2014
5th Percentile — Minimum
- Median — Maximum
O 95th Percentile
Observations from Figure 15-21 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.
• Only one benzene concentration greater than 5 |ig/m3 has been measured at S4MO
(2003). Three benzene concentrations greater than 4 |ig/m3 have also been measured
(one each in 2003, 2006, and 2008).
• The 1-year average concentrations exhibit 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.
15-28
-------
• With the exception of the minimum concentration, all of the statistical parameters are
at a minimum for 2013. The change between 2003 and 2013 represents nearly a
60 percent decrease in the 1-year average concentration and an 82 percent decrease in
the median concentration.
• All of the statistical parameters exhibit slight increases for 2014.
Figure 15-22. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at S4MO
"-j-1
2007
"r
2008
JL
Lrl
2013 2014
o 5th Percentile
- Minimum
- Maximum
o 95th Percentile
Observations from Figure 15-22 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 and 2012. After 2006, no more than five non-
detects of 1,3-butadiene have been measured at S4MO for any given year.
Between 2004 and 2008, the 1-year average concentration changed very 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
15-29
-------
indicated by a minimum and 5th percentile of zero, such as 2009 and 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.
Figure 15-23. Yearly Statistical Metrics for Cadmium (PMio) Concentrations Measured at
S4MO
pj-
I ^ r
* ^ <
T T
—
-2
o
—St
""
o
I X r
H
20031 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
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 2003.
Observations from Figure 15-23 for cadmium concentrations measured at S4MO include
the following:
• The maximum cadmium concentration was measured in 2009 (9.71 ng/m3). The four
additional cadmium concentrations greater than 5 ng/m3 were measured at S4MO in
2004 (one), 2008 (two), and 2009 (one).
• A steady decreasing trend is shown in the 1-year average and median concentrations
through 2006. Even though the 1-year average concentration exhibits an increasing
trend between 2006 and 2009, the median concentration does not, and continued
decreasing during most of this time. This indicates that concentrations at the upper
end of the concentration range are driving the 1-year averages, particularly for 2008
and 2009, while the concentrations at the lower end of the concentration range are
accounting for a higher percentage of measurements.
• The range of concentrations measured decreased significantly from 2009 to 2010.
15-30
-------
• Even though the range of concentrations increased every year between 2010 and
2013, the 1-year average concentration changed little while the median exhibits a
slight decreasing trend, before leveling out for 2013.
• Each of the statistical parameters is at a minimum for 2014, with the 1-year average
concentration less than 0.5 ng/m3 and the median concentration less than 0.25 ng/m3
for the first time since the onset of sampling.
Figure 15-24. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
S4MO
i—£—i
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year
O 5th Percentile
o 95th Percentile
Observations from Figure 15-24 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 change 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 between 2005 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.
15-31
-------
• Both the median and 1-year average concentrations have a decreasing trend between
2008 and 2010, with the largest change shown for 2010.
• Between 2010 and 2012, the 1-year average concentrations have 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.
• 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, from 12 in 2012 to eight in 2013, and three in 2014.
Figure 15-25. Yearly Statistical Metrics for />-Dichlorobenzene Concentrations Measured at
S4MO
r
r
<
> <>• gg y L
i-4
4 L
b w 4 a.
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year
Observations from Figure 15-25 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
15-32
-------
of non-detects varies between 10 percent (2012) and 25 percent (2014) each year
following 2009. The percentage of non-detects for 2014 is the highest percentage
since 2005.
After little change in the early years, the 1-year average and median concentrations
exhibit a steady increasing trend 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 p-dichlorobenzene concentration was measured (6.18 (J,g/m3). The
difference between the median and 1-year average concentration is also an indicator
of this variability. For 2008, the 1-year average concentration was more than three
times greater than the median concentration.
The concentrations measured decreased considerably from 2008 to 2009 then
increased again in 2010. Concentrations measured in 2010 were higher and more
variable than those measured in 2009. Five concentrations measured in 2010 were
greater than the maximum concentration measured in 2009 and the number of
concentrations greater than 0.5 [j,g/m3 more than doubled, from four in 2009 to 10 for
2010. At the same time, the number of non-detects increased from three in 2009 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. While the 1-year average concentration exhibits a decrease
during this time, the median concentration increased slightly for 2011 and then did
not change for 2012.
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.
Most of the statistical parameters exhibit increases for 2014, even though more than a
quarter of the measurements are non-detects.
15-33
-------
Figure 15-26. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
S4MO
I
I.
~± b-
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year
o 5th Percentile
o 95th Percentile
• Average
Observations from Figure 15-26 for 1,2-dichloroethane concentrations measured at
S4MO include the following:
• With the exceptions of 2012, 2013, and 2014, the median concentration is zero for all
years, 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. Beginning in 2009, the
number of measured detections increased steadily, from five in 2009, to 10 in 2010,
18 in 2011, 56 in 2012, and 58 in both 2013 and 2014.
• As the number of measured detections increased in the later years of sampling, each
of the corresponding statistical metrics shown in Figure 15-26 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 1-year average concentrations shown between 2012 and 2014 vary little, from
0.082 [j,g/m3 for 2014 to 0.087 [j,g/m3 for 2013.
15-34
-------
Figure 15-27. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at S4MO
-o
T
T
O
-t-
-0
T
2004 2005
5th Percentile — Minimum
2007 2008 2009 2010 2011
Year
2013 2014
- Median
— Maximum
O 95th Percentile
Observations from Figure 15-27 for ethylbenzene concentrations measured at S4MO
include the following:
• The highest ethylbenzene concentrations were predominantly measured prior to 2008.
Six of the seven ethylbenzene concentrations greater than 2 [j,g/m3 were measured in
2007 or earlier, and four of these were measured in 2003. The exception was
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). Fifteen
concentrations measured in 2010 are greater than the maximum concentration
measured in 2009. The maximum concentration measured in 2009 in nearly
equivalent to the 1-year average concentration for 2010.
• A steady decreasing trend in the ethylbenzene concentrations measured at S4MO is
shown again for the years following 2010. Even though a few higher concentrations
were measured in 2014, most of the statistical parameters exhibit additional decreases
compared to 2013.
15-35
-------
Figure 15-28. Yearly Statistical Metrics for Fluorene Concentrations Measured at S4MO
2011
Year
O 5th Percentile - Minimum - Med en - Maximum o 95th Percentile •¦•~¦¦¦Average
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2008.
Observations from Figure 15-28 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 15-18.
• Two measurements 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
35 fluorene concentrations greater than 15 ng/m3, 27 were measured at S4MO
between June and August of any given year and none were measured between
December and March.
• Several of the statistical parameters shown exhibit decreases from 2008 to 2009.
From 2009 to 2010, the range of concentrations measured increased considerably but
the median concentration decreased, a trend that continued into 2011. A similar
observation was made for acenaphthene.
• With the exception of the maximum concentration, 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 concentrations greater than
10 ng/m3 increased from 13 to 22 from 2011 to 2012; conversely, the number of
concentrations less than 2 ng/m3 decreased from 11 to three from 2011 to 2012.
15-36
-------
• All of the statistical parameters exhibit decreases for 2013.
• The first, and only, non-detects (four) of fluorene were measured at S4MO in 2014.
Despite these non-detects, many of the statistical parameters exhibit slight increases
for 2014.
Figure 15-29. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at S4MO
£
2 25
I
10
5 [-
o
~
£
2005 2006
2007 2008 2009 2010 2011 2012
Year
2013 2014
5th Percentile - Minimum
- Med en - Maximum o 95th Percentile
Observations from Figure 15-29 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 all 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 did not change significantly between 2009 and
2010, even though the smallest range of concentrations was measured in 2010.
15-37
-------
• 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 2014.
Figure 15-30. Yearly Statistical Metrics for Hexachloro-l,3-butadiene Concentrations
Measured at S4MO
^ 0.50
2003 2004 2005 2006
2007 2008 2009 2010 2011 2012
Year
t&J
2013 2014
5th Percentile
— Minimum
- Med "en
— Maximum
O 95th Percentile
Observations from Figure 15-30 for hexachloro-l,3-butadiene concentrations measured at
S4MO include the following:
• The median concentration of hexachloro-1,3-butadiene for each year of sampling is
zero, indicating that at least 50 percent of the measurements were non-detects. For
2003, 2004, and 2007 through 2010, 100 percent of the measurements were non-
detects.
• For 2005 and 2006, the percentage of measured detections was less than 15 percent.
For 2011, measured detections accounted for 16 percent of the measurements. For
2012, that number increased to 22 percent and then up to 26 percent for 2013. The
detection rate fell slightly in 2014 to 20 percent.
• Over the last 4 years of sampling, the 1-year average concentration has varied from
0.015 |ig/m3 (2014) to 0.029 |ig/m3 (2011).
15-38
-------
Figure 15-31. Yearly Statistical Metrics for Naphthalene Concentrations Measured at S4MO
Maximum
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2008.
Observations from Figure 15-31 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 0.78 ng/m3 (2014) to 784 ng/m3
(2010).
• The 1-year average concentration has ranged from 72.07 ng/m3 (2013) to 135 ng/m3
(2010). The median varies less, ranging from 66.85 ng/m3 (2013) to 89.85 ng/m3
(2010). All of the statistical parameters, except the minimum concentration, are at a
minimum for 2013.
• 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
15-39
-------
15.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at the S4MO monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4 for
definitions and explanations regarding the various toxicity factors, time frames, and calculations
associated with these risk-based screenings.
15.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for S4MO 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 3.4.3.3 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 15-5, 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 15-5 include the following:
• The pollutants with the highest annual average concentrations for S4MO are
formaldehyde, acetaldehyde, and benzene.
• The same three pollutants have the highest cancer risk approximations for S4MO,
although the order is different. Formaldehyde's cancer risk approximation for S4MO
(44.91 in-a-million) is an order of magnitude higher than the cancer risk
approximations for these two other pollutants.
• Benzene has the highest cancer risk approximation for S4MO among the VOCs
(5.51 in-a-million); arsenic has the highest cancer risk approximation for S4MO
among the metals (3.88 in-a-million); and naphthalene has the highest cancer risk
approximation for S4MO among the PAHs (2.78 in-a-million).
• 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
approximation is formaldehyde (0.35).
15-40
-------
Table 15-5. Risk Approximations for the Missouri Monitoring Site
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
St. Louis, Missouri - S4MO
Acetaldehyde
0.0000022
0.009
60/60
2.08
±0.22
4.57
0.23
Benzene
0.0000078
0.03
59/59
0.71
±0.07
5.51
0.02
1,3-Butadiene
0.00003
0.002
58/59
0.09
±0.02
2.73
0.05
Carbon Tetrachloride
0.000006
0.1
59/59
0.62
±0.02
3.74
0.01
/?-Dichlorobcnzcnc
0.000011
0.8
45/59
0.13
±0.05
1.41
<0.01
1,2-Dichloroethane
0.000026
2.4
58/59
0.08
±0.01
2.15
<0.01
Ethylbenzene
0.0000025
1
59/59
0.23
±0.04
0.58
<0.01
Formaldehyde
0.000013
0.0098
60/60
3.45
±0.48
44.91
0.35
Hexachloro-1,3 -butadiene
0.000022
0.09
12/59
0.02
±0.01
0.33
<0.01
Acenaphthene3
0.000088
57/57
6.43
± 1.56
0.57
Arsenic (PMi0)a
0.0043
0.000015
61/61
0.90
±0.15
3.88
0.06
Cadmium (PMi0)a
0.0018
0.00001
61/61
0.33
±0.09
0.60
0.03
Fluorene3
0.000088
53/57
6.65
± 1.38
0.59
Naphthalene3
0.000034
0.003
57/57
81.79
± 12.61
2.78
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.
15-41
-------
15.5.2 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 15-6 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
that have cancer toxicity factors. Table 15-6 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 15-6 provides the pollutants of interest with the highest cancer risk approximations (in-a-
million) for S4MO, as presented in Table 15-5. The emissions, toxicity-weighted emissions, and
cancer risk approximations are shown in descending order in Table 15-6. Table 15-7 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 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 15.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
15-42
-------
Table 15-6. 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)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
St. Louis, Missouri (St. Louis City) - S4MO
Formaldehyde
86.19
Formaldehyde
1.12E-03
Formaldehyde
44.91
Benzene
85.02
Hexavalent Chromium
7.89E-04
Benzene
5.51
Ethylbenzene
48.46
Benzene
6.63E-04
Acetaldehyde
4.57
Acetaldehyde
46.53
1,3-Butadiene
3.78E-04
Arsenic (PMio)
3.88
T richloroethylene
15.45
Naphthalene
3.26E-04
Carbon Tetrachloride
3.74
1.3 -Butadiene
12.60
Arsenic, PM
2.49E-04
Naphthalene
2.78
Naphthalene
9.59
POM, Group 2b
1.80E-04
1,3-Butadiene
2.73
T etrachloroethylene
5.83
POM, Group 2d
1.44E-04
1,2-Dichloroethane
2.15
Dichloromethane
3.65
Ethylbenzene
1.21E-04
/?-Dichlorobcnzcnc
1.41
POM, Group 2b
2.05
Acetaldehyde
1.02E-04
Cadmium (PMio)
0.60
-------
Table 15-7. 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)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
St. Louis, Missouri (St. Louis City) - S4MO
Toluene
313.06
Acrolein
268,853.09
Formaldehyde
0.35
Hexane
226.91
Formaldehyde
8,795.13
Acetaldehyde
0.23
Methanol
208.08
T richloroethylene
7,726.86
Arsenic (PMio)
0.06
Xylenes
196.63
1.3 -Butadiene
6,302.24
1,3-Butadiene
0.05
Formaldehyde
86.19
Acetaldehyde
5,170.07
Cadmium (PMio)
0.03
Benzene
85.02
Arsenic, PM
3,864.62
Naphthalene
0.03
Hydrochloric acid
70.78
Hydrochloric acid
3,539.11
Benzene
0.02
Ethylene glycol
64.32
Cadmium, PM
3,474.08
Carbon Tetrachloride
0.01
Ethylbenzene
48.46
Lead, PM
3,349.14
Ethylbenzene
0.00
Acetaldehyde
46.53
Naphthalene
3,195.33
Hexachloro-1,3 -butadiene
0.00
-------
Observations from Table 15-6 include the following:
• Formaldehyde, benzene, and ethylbenzene are the highest emitted pollutants with
cancer UREs in the city of St. Louis.
• The pollutants with the highest toxicity-weighted emissions (of the pollutants with
cancer UREs) are formaldehyde, hexavalent chromium, and benzene.
• Seven 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 approximation. Benzene,
acetaldehyde, naphthalene, and 1,3-butadiene also appear on all three lists.
• Arsenic has the fourth highest cancer risk approximation for S4MO. While arsenic is
not one of the highest emitted pollutants in the city of St. Louis, it ranks sixth for its
toxicity-weighted emissions. Carbon tetrachloride, 1,2-dichloroethane,
/;-dichlorobenzene, and cadmium also appear among the pollutants of interest with
the highest cancer risk approximations for S4MO but none of these appear on either
emissions-based list.
• POM, Group 2b is the 10th highest emitted "pollutant" in the city of St. Louis and
ranks seventh 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 15-7 include the following:
• Toluene, hexane, and methanol are the highest emitted pollutants with noncancer
RfCs in the city of St. Louis.
• The pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) are acrolein, formaldehyde, and trichloroethylene. 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.
• Three of the highest emitted pollutants in the city of St. Louis also have the highest
toxicity-weighted emissions.
• Formaldehyde, the pollutant with highest noncancer hazard approximation for S4MO,
has the second highest toxicity-weighted emissions and the fifth highest total
emissions (of the pollutants with noncancer RfCs). Acetaldehyde also appears on all
three lists.
15-45
-------
• Arsenic and cadmium, both pollutants of interest for S4MO, appear among the
pollutants with the highest toxicity-weighted emissions, but are not among the highest
emitted.
15.6 Summary of the 2014 Monitoring Data for S4MO
Results from several of the data analyses described in this section include the following:
~~~ Twenty-three pollutants failed screens for S4MO. S4MO failed the second 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 second highest annual average concentration of arsenic among NMP
sites sampling PMio metals. S4MO also has the third highest annual average
concentration of p-dichlorobenzene among NMP sites sampling VOCs.
~~~ Concentrations of benzene, ethylbenzene, and cadmium have decreased significantly
since sampling began at S4MO.
~~~ Formaldehyde has the highest cancer risk approximation of the pollutants of interest
for S4MO. None of the pollutants of interest have noncancer hazard approximations
greater than an HQ of 1.0.
15-46
-------
16.0 Sites in New Jersey
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at UATMP sites in New Jersey, and integrates these concentrations
with emissions, meteorological, and risk information. Data generated by sources other than ERG
are not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) 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 geographical and
physical information about the location of the sites and the surrounding areas. 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 data.
One New Jersey monitoring site (CSNJ) is located in the Philadelphia-Camden-
Wilmington, PA-NJ-DE-MD CBSA while the other three New Jersey sites are located within the
New York-Newark-Jersey City, NY-NJ-PA CBSA. Figure 16-1 is 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 2011 NEI for point sources, version 2. Note that only sources within 10 miles
of the site 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 in order to
emphasize emissions sources within the boundary. Figures 16-3 through 16-7 are the composite
satellite maps and emissions source maps for CHNJ, ELNJ, and NBNJ. Table 16-1 provides
supplemental geographical information such as land use, location setting, and locational
coordinates.
16-1
-------
Figure 16-1. Camden, New Jersey (CSNJ) Monitoring Site
Chestnuts
,l
-------
Figure 16-2. NEI Point Sources Located Within 10 Miles of CSNJ
74"55'0"W
75°io,o"w
75°5'0"W
75°0'0"W
20'0"W
7531 5'0"W
Burlington
County
V
PENNSYLVANIA
Camden
, County 1
%
Gloucester
\ County
jTfyr[ Delaware
' River
75"25'0"W
Legend
75°20'0*W
75°15'0"W
75°10'0"W 75"5'0"W 75Do",0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
CSNJ UATMP site Q 10 mile radius
County boundary
T
i
S
•
B
C
f
F
*
o
A
®
A
x
Airport/Airline/Airport Support Operations (35)
?
Miscellaneous Commercial/Industrial Facility (21)
Asphalt Production/Hot Mix Asphalt Plant (3)
M
Municipal Waste Combustor (2)
Automobile/Truck Manufacturing Facility (1)
•
Oil and/or Gas Production (3)
Building/Construction (1)
Paint and Coating Manufacturing Facility (1)
Bulk Terminals/Bulk Plants (12)
Petroleum Products Manufacturing (2)
Chemical Manufacturing Facility (12)
a
Petroleum Refinery (5)
Compressor Station (5)
CD
Pharmaceutical Manufacturing (2)
Electricity Generation via Combustion (9)
R
Plastic. Resin, or Rubber Products Plant (4)
Food Processing/Agriculture Facility (2)
¥
Port and Harbor Operations (2)
Industrial Machinery or Equipment Plant (1)
P
Printing/Publishing/Paper Product Manufacturing Facility (14)
Institutional (school, hospital, prison, etc.) (36)
B
Pulp and Paper Plant (3)
Landfill (1)
A
Ship/Boat Manufacturing or Repair Facility (1)
Metal Can, Box. and Other Metal Container Manufacturing (1)
V
Steel Mill (1)
Metal Coating, Engraving, and Allied Services to Manufacturers (2)
TT
Telecommunications/Radio Facility (1)
Metals Processing/Fabrication Facility (8)
Truck/Bus/Transportation Operations (1)
Military Base/National Security Facility (3)
V
Wastewater Treatment Facility (4)
Mine/Quarry/Mineral Processing Facility (2)
16-3
-------
Figure 16-3. Chester, New Jersey (CHNJ) Monitoring Site
-------
Figure 16-4. NEI Point Sources Located Within 10 Miles of CHN.I
74°35'0"W
Sussex
County
Morris
County
Somerset
County
Hunterdon \
County i
Source Category Group (No. of Facilities)
<5< Aerospace/Aircraft Manufacturing Facility (1)
¦f Airport/Airline/Airport Support Operations (12)
Asphalt Production/Hot Mix Asphalt Plant (1)
C Chemical Manufacturing Facility (3)
e Electrical Equipment Manufacturing Facility (1)
F Food Processing/Agriculture Facility (1)
? Miscellaneous Commercial/Industrial Facility (2)
W Woodwork, Furniture, Millwork & Wood Preserving Facility (1)
Legend
74°45,0"W 74°40'0"W 74°35,0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
CHNJ UATMP site O 10 mile radius | | County boundary
V
\ Warren
\ County
16-5
-------
Figure 16-5. Elizabeth, New Jersey (ELNJ) Monitoring Site
-------
Figure 16-6. North Brunswick, New Jersey (NBNJ) Monitoring Site
-------
Figure 16-7. NEI Point Sources Located Within 10 Miles of ELNJ and NBNJ
74335'0"W 74°30'0"W 74°25'0"W 74o20'0"W 74°15-0"W 74°10,0"W 74'5'0"W 74°0'0"W 73°55'0"W
\ / *
/ Essex i ^ ^ fa JHudsonf ,,
, _ \ — —/ l4-C\jV County ft.
i County \ - o
Morris
County
NEW
\ \ Richmond \
\ YORK \
\ County
u Union \* +
County ^
Somerset \
I o County
f, ^ \
\ / Raritan
Monmouth \
County \
Middlesex \
County y
Q ^k?P
74°45'0"W 74°40"0"W 74°35,0"W 74"30'0"W 74°25'0"W 74*20,0"W 74°15,0"W 74°10,0"W 74'5'0'W
. Note: Due to facility density and collocation, the total facilities
Legend displayed may not represent all facilities within the area of interest.
~
ELNJ UATMP site
~
NBNJ UATMP site O 10 mile radius
County boundary
Source Category Group (No. of Facilities)
^ Airport/Airline/Airport Support Operations (46)
£
I
«
Asphalt Production/Hot Mix Asphalt Plant (6)
Breweries/Distilleries/Wineries (1)
Building/Construction (1)
B Bulk Terminals/Bulk Plants (26)
C Chemical Manufacturing Facility (34)
I Compressor Station (3)
© Electrical Equipment Manufacturing Facility (1)
^ Electricity Generation via Combustion (22)
E Electroplating. Plating. Polishing, Anodizing, and Coloring (3)
=)(= Ethan ol Biorefineries (1)
F Food Processing/Agriculture Facility (8)
A. Foundries. Non-ferrous (1)
Industrial Machinery or Equipment Plant (2)
O Institutional (school, hospital, prison, etc.) (16)
A Landfill (5)
¦ Metal Can. Box, and Other Metal Container Manufacturing (3)
/\ Metal Coating, Engraving, and Allied Services to Manufacturers (7)
(•) Metals Processing/Fabrication Facility (14)
X Mine/Quarry/Mineral Processing Facility (2)
f Miscellaneous Commercial/Industrial Facility (23)
[Ml Municipal Waste Combustor (2)
Q Paint and Coating Manufacturing Facility (11)
3 Petroleum Refinery (3)
C—> Pharmaceutical Manufacturing (8)
R Plastic, Resin, or Rubber Products Plant (9)
Port and Harbor Operations (2)
P Printing/Publishing/Paper Product Manufacturing Facility (16)
EH Pulp and Paper Plant (1)
^ Steel Mill (2)
Testing Laboratories (1)
W Wastewater Treatment Facility (7)
4 Water Treatment Facility (1)
VV Woodwork, Furniture, Millwork & Wood Preserving Facility (1)
16-8
-------
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
Between Exits 13 & 13AonI-95
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 Lane/617
3AADT for ELNJ reflects 2006 data fromNJ Department of Treasury (NJ DOTr, 2008); AADT reflects 2010 data for NBNJ and 2012 data for CSNJ and CHNJ from the NJ
DOT (NJ DOT, 2014).
On
-------
The CSNJ monitoring site is located just outside Philadelphia, across the state line, in
the city of Camden in southwest New Jersey. The monitoring site is 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 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; printing,
publishing, and paper product manufacturing; chemical manufacturing; and bulk terminals and
bulk plants. The sources closest to CSNJ include a metals processing and fabrication facility; a
mine/quarry/minerals processing facility; an airport/airport operations facility; and a metal can,
box, and other container manufacturing 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 and west, as well as on
the east side of the Turnpike, while residential neighborhoods are located to the north and
northwest of ELNJ.
16-10
-------
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, part of
which can be seen on the right-hand side of Figure 16-6.
Figure 16-7 shows that the outer portions of the 10-mile boundaries for ELNJ and NBNJ
intersect; these sites are located approximately 17 miles apart. Many emissions sources surround
these two sites. The majority of the emissions sources are located in northern 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,
and electricity generation via combustion source categories. The emissions sources in closest
proximity to the NBNJ monitoring site are involved in plastic, resin, or rubber products
manufacturing, airport and airport support operations, and pharmaceutical manufacturing.
In addition to providing city, county, CBSA, 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 effect concentrations measured at a given monitoring site. ELNJ
and NBNJ experience a significantly higher traffic volume than CHNJ and CSNJ. Traffic data
for ELNJ are provided for 1-95, between Exit 13 and 13 A; this is the highest traffic volume
among all NMP sites. Traffic data for NBNJ are provided for US-1, east of State Road 617
(Ryders Lane); 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.
16-11
-------
16.2 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
sites in New Jersey on sample days, as well as over the course of the year.
16.2.1 Meteorological Summary
In order to provide an overview of the meteorological conditions experienced at each
monitoring site, hourly meteorological data for 2014 were retrieved and sample day and full-year
averages developed for temperature, dew point, relative humidity, pressure, and wind speed.
Weather data from the actual monitoring site(s) were obtained from AQS, where available. If
site-specific weather data were not available in AQS, then data were obtained from NCDC for
the NWS weather station located closest to the monitoring site(s), as described in Section 3.4.2.
For the New Jersey sites, not enough site-specific data were available in AQS for the parameters
listed in Table 16-2; thus, data from the closest NWS weather station to each monitoring site was
used for this analysis. A map showing the distance between each New Jersey monitoring site and
the closest NWS weather station is provided in Appendix R. These data were used to determine
how meteorological conditions on sample days vary from conditions experienced throughout the
year.
Table 16-2 presents average temperature, average dew point temperature, average relative
humidity, average station and sea level pressure, and wind information (average scalar wind
speed and prevailing wind direction) for days on which samples were collected and for all of
2014. Also included in Table 16-2 is the 95 percent confidence interval for each parameter. Note
that the weather station at Somerville/Somerset Airport is the closest weather station to both
CHNJ and NBNJ. Even though sample days are mostly standardized, missed and/or invalid
samples and the need for "make-up" samples results in some differences in sample days among
the sites; thus, the sample day averages are not the same for these two sites.
As shown in Table 16-2, average meteorological conditions on sample days were
representative of average weather conditions experienced throughout the year near each site. The
greatest difference between the sample day and full-year averages was calculated for average
relative humidity for NBNJ.
16-12
-------
Table 16-2. Average Meteorological Conditions near the New Jersey Monitoring Sites
Average
Type1
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(in Hg)
Average
Station
Pressure
(in Hg)
Prevailing
Wind
Direction
Average
Scalar Wind
Speed
(kt)
Camden, New Jersey - CSNJ2
Sample
Days
54.0
38.4
58.1
30.06
30.04
8.4
(62)
± 1.0
± 1.1
±0.9
±0.01
±0.01
WNW
±0.2
55.2
40.3
60.1
30.04
30.01
7.7
2014
±0.4
±0.4
±0.4
± <0.01
±<0.01
SW
±0.1
Somerville, New Jersey/Somerset Airport3
CHNJ
51.0
38.5
65.6
30.02
29.91
3.5
(61)
± 1.0
± 1.1
± 1.0
± <0.01
±<0.01
NW
±0.2
NBNJ
51.1
38.4
64.9
30.01
29.91
3.5
(61)
± 1.0
± 1.1
± 1.0
±0.01
±0.01
NW
±0.2
51.3
39.3
67.2
30.02
29.91
3.1
2014
±0.4
±0.4
±0.4
± <0.01
±<0.01
NW
±0.1
Elizabeth, New Jersey - ELNJ4
Sample
Days
53.3
38.6
60.3
30.04
30.01
8.4
(61)
± 1.0
± 1.1
±0.9
±0.01
±0.01
W
±0.3
54.2
39.9
61.7
30.02
30.00
7.8
2014
±0.4
±0.4
±0.4
± <0.01
±<0.01
w
±0.1
'Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
2This information was obtained from the NWS weather station located at Philadelphia International Airport, WBAN 13739.
3This information was obtained from the NWS weather station located at Somerville, New Jersey /Somerset Airport, WBAN
54785.
4This information was obtained from the NWS weather station located at Newark International Airport, WBAN 14734.
16.2.2 Wind Rose Comparison
Hourly surface wind data were also uploaded into a wind rose software program to
produce customized wind roses, as described in Section 3.4.2. A wind rose shows the frequency
of wind directions using "petals" positioned around a 16-point compass, and uses different colors
to represent wind speeds. Figure 16-8 presents two wind roses for the CSNJ monitoring site. The
first is a wind rose representing wind observations for all of 2014 and the second is a wind rose
representing wind observations for days on which samples were collected in 2014. These are
used to identify the predominant wind speed and direction for 2014 and to determine if wind
observations on sample days were representative of conditions experienced over the entire year.
Figures 16-9 and 16-10 present the full-year and sample day wind roses for CHNJ/NBNJ, and
ELNJ.
16-13
-------
Figure 16-8. Wind Roses for the Philadelphia International Airport Weather Station near
CSNJ
2014 Wind Rose
Sample Day Wind Rose
WIND SPEED
(Knots)
I I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 7.13%
WIND SPEED
(Knots)
I I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 7.39%
Observations from Figure 16-8 for CSNJ include the following:
• The Philadelphia International Airport weather station is located 7.2 miles southwest
of CSNJ. Both the site and the weather station are located near the Delaware River,
which separates Pennsylvania from New Jersey, and runs along the east and south
sides of Philadelphia.
• The full-year wind rose shows that winds from a variety of directions were observed
near CSNJ, with winds from the southwestern and northwestern quadrants observed
more frequently than those from the eastern quadrants. Winds from the western
quadrants, and from due south and north account for approximately the same
percentage of observations, with each individual direction accounting for between
6 percent and 8 percent of observations. Calm winds account for 7 percent of
observations near CSNJ while the strongest winds were most often observed with
westerly to north-northwesterly winds.
• The even distribution of winds from the western quadrants is not shown on the
sample day wind rose. On sample days, winds from the west to northwest were
prevalent, with fewer winds from the south, southwest quadrant, and north-northwest.
East winds were also observed more frequently on sample days. The percentage of
north-northeast winds, winds from the southeast quadrant, and calm winds shown on
the sample day wind rose are similar to the percentages shown on the 2014 wind rose.
16-14
-------
Figure 16-9. Wind Roses for the Somerville-Somerset Airport Weather Station near CHNJ
and NUN J
2014 Wind Rose
CHNJ Sample Day Wind Rose
WIND SPEED
(Knots)
HI >= 22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
HI 1-4
Calms: 38.66%
WIND SPEED
(Knots)
m >s 22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 47.40%
NBNJ Sample Day Wind Rose
1
WIND SPEED
(Knots)
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
IH 1-4
Calms: 47.13%
SOUTH
NORTH
! EAST
16-15
-------
Observations from Figure 16-9 for CHNJ and NBNJ include the following:
• The weather station at Somerville/Somerset Airport is the closest weather station to
both CHNJ and NBNJ. The Somerville/Somerset Airport weather station is located
11.3 miles south of CHNJ and 16.7 miles northwest of NBNJ.
• The full-year wind rose for these sites shows that calm winds account for nearly
40 percent of observations. For the remaining observations, northwesterly winds were
observed most frequently, followed by winds from the north-northwest, north, and
west-northwest, together accounting for nearly 20 percent of the observations. With
the exception of south-southeasterly and southerly winds, winds from the other
quadrants were infrequently observed.
• The sample day wind roses for CHNJ and NBNJ resemble each other. Both wind
roses show that calm winds were prevalent near these sites, with calm winds
accounting for 47 percent of the observations on sample days. For the remaining
observations, northwesterly winds were prevalent on sample days, with winds from
the west-northwest to north accounting for the majority of observations.
Figure 16-10. Wind Roses for the Newark International Airport Weather Station near
ELNJ
2014 Wind Rose
Sample Day Wind Rose
WIND SPEED
(Knots)
I I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 7.75%
6%
K>
EAST
WIND SPEED
(Knots)
I I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 9.08%
16-16
-------
Observations from Figure 16-10 for ELNJ include the following:
• The Newark International Airport weather station is located 3.5 miles northeast of
ELNJ. Both the site and the weather station are located in close proximity to the New
Jersey Turnpike.
• The full-year wind rose shows that winds from a variety of directions were observed
near ELNJ, although winds from the northeast to east-southeast were observed
infrequently. Westerly winds were observed the most, accounting for nearly
12 percent of observations. Calm winds accounted for nearly 8 percent of
observations. The strongest winds were associated with west-southwesterly to
northerly winds.
• Westerly winds were also prevalent on sample days, accounting for a slightly higher
percentage of observations. Winds from the southwest to northwest, to north and
north-northeast accounted for the majority of observations, similar to what is shown
on the full-year wind rose. Calms winds accounted for 9 percent of observations on
sample days.
16.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each New
Jersey monitoring 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. 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 16-3. 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 16-3. 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 all four New
Jersey sites.
16-17
-------
Table 16-3. 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
Benzene
0.13
61
61
100.00
15.25
15.25
1.3 -Butadiene
0.03
61
61
100.00
15.25
30.50
Carbon Tetrachloride
0.17
61
61
100.00
15.25
45.75
Acetaldehyde
0.45
60
60
100.00
15.00
60.75
Formaldehyde
0.077
60
60
100.00
15.00
75.75
1,2-Dichloroethane
0.038
59
59
100.00
14.75
90.50
Hexachloro-1,3 -butadiene
0.045
12
12
100.00
3.00
93.50
Ethylbenzene
0.4
11
61
18.03
2.75
96.25
/?-Dichlorobcnzcnc
0.091
6
30
20.00
1.50
97.75
Bromomethane
0.5
3
53
5.66
0.75
98.50
Propionaldehyde
0.8
2
60
3.33
0.50
99.00
T richloroethylene
0.2
2
22
9.09
0.50
99.50
Vinyl cliloride
0.11
1
8
12.50
0.25
99.75
Xylenes
10
1
61
1.64
0.25
100.00
Total
400
669
59.79
Chester, New Jersey - CHNJ
Benzene
0.13
61
61
100.00
16.22
16.22
Carbon Tetrachloride
0.17
61
61
100.00
16.22
32.45
Acetaldehyde
0.45
60
60
100.00
15.96
48.40
1,2-Dichloroethane
0.038
60
60
100.00
15.96
64.36
Formaldehyde
0.077
60
60
100.00
15.96
80.32
1.3 -Butadiene
0.03
56
59
94.92
14.89
95.21
Hexachloro-1,3 -butadiene
0.045
14
16
87.50
3.72
98.94
1,2-Dibromoethane
0.0017
3
3
100.00
0.80
99.73
/?-Dichlorobcnzcnc
0.091
1
16
6.25
0.27
100.00
Total
376
396
94.95
16-18
-------
Table 16-3. Risk-Based Screening Results for the New Jersey Monitoring Sites (Continued)
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Elizabeth, New Jersey - ELNJ
Acetaldehyde
0.45
61
61
100.00
14.91
14.91
Formaldehyde
0.077
61
61
100.00
14.91
29.83
Benzene
0.13
59
59
100.00
14.43
44.25
1.3 -Butadiene
0.03
59
59
100.00
14.43
58.68
Carbon Tetrachloride
0.17
59
59
100.00
14.43
73.11
1,2-Dichloroethane
0.038
57
57
100.00
13.94
87.04
Ethylbenzene
0.4
19
59
32.20
4.65
91.69
Hexachloro-1,3 -butadiene
0.045
18
18
100.00
4.40
96.09
/?-Dichlorobcnzcnc
0.091
10
30
33.33
2.44
98.53
Propionaldehyde
0.8
3
61
4.92
0.73
99.27
1,2-Dibromoethane
0.0017
2
2
100.00
0.49
99.76
Chloroprene
0.0021
1
1
100.00
0.24
100.00
Total
409
527
77.61
North Brunswick, New Jersey - NBNJ
Benzene
0.13
60
60
100.00
19.05
19.05
Carbon Tetrachloride
0.17
60
60
100.00
19.05
38.10
1,2-Dichloroethane
0.038
58
58
100.00
18.41
56.51
1.3 -Butadiene
0.03
56
57
98.25
17.78
74.29
Acetaldehyde
0.45
20
20
100.00
6.35
80.63
Formaldehyde
0.077
20
20
100.00
6.35
86.98
Hexachloro-1,3 -butadiene
0.045
16
17
94.12
5.08
92.06
Ethylbenzene
0.4
11
60
18.33
3.49
95.56
/?-Dichlorobcnzcnc
0.091
5
25
20.00
1.59
97.14
Propionaldehyde
0.8
5
20
25.00
1.59
98.73
1,2-Dibromoethane
0.0017
3
3
100.00
0.95
99.68
Bromomethane
0.5
1
50
2.00
0.32
100.00
Total
315
450
70.00
16-19
-------
Observations from Table 16-3 include the following:
• Fourteen pollutants failed at least one screen for CSNJ; 60 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 CSNJ and therefore
were identified as pollutants of interest for this site. These eight include two carbonyl
compounds and six VOCs.
• Nine pollutants failed at least one screen for CHNJ; 96 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.
• Twelve pollutants failed at least one screen for ELNJ, with nearly 78 percent of
concentrations for these 12 pollutants greater than their associated risk screening
value (or failing screens).
• Eight pollutants contributed to 95 percent of failed screens for ELNJ and therefore
were identified as pollutants of interest for this site. These eight include two carbonyl
compounds and six VOCs.
• Twelve pollutants failed at least one screen for NBNJ, with 70 percent of
concentrations for these 12 pollutants greater than their associated risk screening
value (or failing screens).
• Eight pollutants contributed to 95 percent of failed screens for NBNJ and therefore
were identified as pollutants of interest for this site. These eight include two carbonyl
compounds and six VOCs.
• The New Jersey sites have six pollutants of interest in common: acetaldehyde,
formaldehyde, benzene, carbon tetrachloride, 1,3-butadiene, and 1,2-dichloroethane.
If CHNJ is excluded, the New Jersey sites would also have ethylbenzene and
hexachloro-1,3-butadiene in common.
16.4 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.
16-20
-------
• 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 are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at the New Jersey monitoring sites are provided in Appendices J and L.
16.4.1 2014 Concentration Averages
Quarterly and annual concentration averages 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 the entire year of sampling. Annual averages were calculated for
pollutants where three valid quarterly averages could be calculated 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 16-4, 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.
16-21
-------
Table 16-4. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the New Jersey Monitoring Sites
# of
Measured
1st
2nd
3rd
4th
Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs.
# >MDL
# of
Samples
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Camden, New Jersey -
CSNJ
2.20
2.84
2.99
1.94
2.49
Acetaldehyde
60/60
60
±0.29
±0.44
±0.55
±0.20
±0.22
0.97
0.54
0.86
0.66
0.76
Benzene
61/61
61
±0.14
±0.09
±0.25
±0.14
±0.09
0.11
0.06
0.10
0.11
0.09
1.3 -Butadiene
61/61
61
±0.03
±0.01
±0.02
±0.03
±0.01
0.53
0.64
0.66
0.60
0.61
Carbon Tetrachloride
61/61
61
±0.06
±0.03
±0.02
±0.06
±0.03
0.09
0.08
0.07
0.08
0.08
1,2-Dichloroethane
59/58
61
±0.01
±0.01
±0.02
±0.01
±0.01
0.30
0.22
0.49
0.29
0.33
Ethylbenzene
61/61
61
±0.11
±0.05
±0.43
±0.13
±0.11
3.36
4.87
6.33
3.35
4.48
Formaldehyde
60/60
60
±0.47
± 1.17
± 1.05
±0.53
±0.52
0.01
0.02
0.02
0.02
0.02
Hexachloro-1,3 -butadiene
12/0
61
±0.01
±0.02
±0.02
±0.02
±0.01
Chester, New Jersey -
CHNJ
1.34
1.27
1.18
1.22
1.25
Acetaldehyde
60/60
60
±0.34
±0.25
±0.17
±0.19
±0.12
0.69
0.44
0.36
0.40
0.47
Benzene
61/61
61
±0.06
±0.14
±0.05
±0.04
±0.05
0.07
0.05
0.07
0.06
0.06
1.3 -Butadiene
59/57
61
±0.02
±0.01
±0.01
±0.01
±0.01
0.57
0.66
0.64
0.55
0.60
Carbon Tetrachloride
61/61
61
±0.06
±0.03
±0.03
±0.05
±0.03
0.08
0.08
0.07
0.08
0.08
1,2-Dichloroethane
60/57
61
±0.01
±0.01
±0.01
±0.01
±<0.01
1.80
2.41
3.07
1.10
2.06
Formaldehyde
60/60
60
±0.50
±0.82
±0.78
±0.22
±0.34
NA = Not available due to the criteria for calculating a quarterly or annual average concentration.
16-22
-------
Table 16-4. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the New Jersey Monitoring Sites (Continued)
# of
Measured
1st
2nd
3rd
4th
Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs.
# >MDL
# of
Samples
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Elizabeth, New Jersey
ELNJ
2.50
3.21
3.02
2.40
2.78
Acetaldehyde
61/61
61
±0.35
±0.50
±0.49
±0.25
±0.21
1.15
0.63
0.68
0.66
0.78
Benzene
59/59
59
±0.27
±0.12
±0.10
±0.08
±0.09
0.15
0.09
0.12
0.11
0.12
1.3 -Butadiene
59/59
59
±0.04
±0.02
±0.02
±0.02
±0.01
0.57
0.64
0.64
0.62
0.62
Carbon Tetrachloride
59/59
59
±0.06
±0.03
±0.04
±0.06
±0.03
0.09
0.09
0.07
0.09
0.09
1,2-Dichloroethane
57/57
59
±0.02
±0.01
±0.01
±0.01
±0.01
0.41
0.32
0.39
0.33
0.36
Ethylbenzene
59/59
59
±0.13
±0.11
±0.06
±0.08
±0.05
3.74
4.70
6.38
3.03
4.44
Formaldehyde
61/61
61
±0.59
±0.98
± 1.29
±0.46
±0.52
0.03
0.02
0.03
0.02
0.02
Hexachloro-1,3 -butadiene
18/0
59
±0.02
±0.02
±0.02
±0.02
±0.01
North Brunswick, New Jersey - NBNJ
3.17
Acetaldehyde
20/20
20
±0.92
NA
NA
NA
NA
0.86
0.42
0.43
0.44
0.54
Benzene
60/60
60
±0.13
±0.06
±0.10
±0.08
±0.07
0.10
0.05
0.06
0.05
0.07
1.3 -Butadiene
57/56
60
±0.03
±0.01
±0.01
±0.02
±0.01
0.56
0.68
0.67
0.62
0.63
Carbon Tetrachloride
60/60
60
±0.06
±0.03
±0.02
±0.03
±0.02
0.09
0.09
0.07
0.08
0.08
1,2-Dichloroethane
58/55
60
±0.01
±0.01
±0.02
±0.01
±0.01
0.32
0.40
0.16
0.11
0.25
Ethylbenzene
60/59
60
±0.14
±0.15
±0.02
±0.03
±0.06
6.39
Formaldehyde
20/20
20
±2.57
NA
NA
NA
NA
0.03
0.03
0.03
<0.01
0.02
Hexachloro-1,3 -butadiene
17/0
60
±0.02
±0.03
±0.03
±0.01
±0.01
NA = Not available due to the criteria for calculating a quarterly or annual average concentration.
16-23
-------
Observations for CSNJ from Table 16-4 include the following:
• The pollutants of interest with the highest annual average concentrations are
formaldehyde (4.48 ± 0.52 |ig/m3) and acetaldehyde (2.49 ± 0.22 |ig/m3). These are
the only two pollutants with annual average concentrations greater than 1 |ig/m3. Of
the VOCs, benzene has the highest annual average concentration (0.76 ± 0.09 |ig/m3).
• Concentrations of formaldehyde appear highest during the second and third quarters
of 2014, based on the quarterly averages shown. A review of the data shows that
formaldehyde concentrations measured at CSNJ range from 2.01 |ig/m3 to
11.9 |ig/m3. All 12 formaldehyde concentrations greater than 6 |ig/m3 were measured
at CSNJ between June and September and all but four of the 25 highest
concentrations were measured during the second and third quarters (with those four
measured in either March and October).
• Concentrations of acetaldehyde also appear highest during the second and third
quarters of 2014, although the differences are less remarkable. Acetaldehyde
concentrations measured at CSNJ range from 1.02 |ig/m3 to 6.30 |ig/m3, with the
maximum concentration measured on the same day at CSNJ as the maximum
formaldehyde concentration (August 27, 2014). All but one of the 15 acetaldehyde
concentrations greater than 3 |ig/m3 were measured at CSNJ during the second and
third quarters of 2014. Conversely, all but two of the 15 concentrations less than
2 |ig/m3 were measured at CSNJ during the first or fourth quarters of 2014 (with the
two exceptions measured in April).
• The maximum concentration of benzene (2.12 |ig/m3) was also measured at CSNJ on
August 27, 2014 and is the only benzene concentration greater than 2 |ig/m3
measured at this site. Of the 12 benzene concentrations greater than 1 |ig/m3
measured at CSNJ, six were measured during the first quarter, five were measured
during the third quarter, one was measured during the fourth quarter, and none were
measured during the second quarter of 2014. In addition, all 12 benzene
concentrations less than 0.5 |ig/m3 were measured at CSNJ during the second or
fourth quarters, with the four lowest concentration measured in April and May. This
explains the differences shown in the quarterly average benzene concentrations, even
though the differences are not statistically significant.
• The third quarter average concentration of ethylbenzene (0.49 ± 0.43 |ig/m3) is
considerably higher than the other quarterly averages and has a confidence interval of
nearly the same magnitude. This indicates the likely influence of outliers. The
maximum ethylbenzene concentration (3.35 |ig/m3) was measured at CSNJ on
September 26, 2014 and is the maximum ethylbenzene concentration measured across
the program. The second highest concentration measured at CSNJ (1.07 |ig/m3) was
measured in October and all other concentrations measured at CSNJ are less than
0.7 |ig/m3.
16-24
-------
Observations for CHNJ from Table 16-4 include the following:
• The pollutants of interest with the highest annual average concentrations are
formaldehyde (2.06 ± 0.34 |ig/m3) and acetaldehyde (1.25 ± 0.12 |ig/m3). These are
the only two pollutants with annual average concentrations greater than 1 |ig/m3. Of
the VOCs, carbon tetrachloride has the highest annual average concentration
(0.60 ± 0.03 |ig/m3).
• Concentrations of formaldehyde appear highest during the second and third quarters
of 2014, with the third quarter average concentration nearly three times higher than
the fourth quarter average concentration. A review of the data shows that
formaldehyde concentrations measured at CHNJ range from 0.588 |ig/m3 to
6.07 |ig/m3. Similar to CSNJ, the maximum concentration of formaldehyde was
measured at CHNJ on August 27, 2014. All but one of the 12 formaldehyde
concentrations greater than 3 |ig/m3 were measured at CHNJ between June and
August, including the four concentrations greater than 5 |ig/m3. Conversely, the six
lowest formaldehyde concentrations were measured at CHNJ between November and
December. The quarterly average concentration for the fourth quarter is particularly
low, as formaldehyde concentrations greater than 2 |ig/m3 were not measured at
CHNJ during the fourth quarter, compared to five measured during the first quarter,
six during the second quarter, and 11 during the third quarter.
• The first quarter average concentration of benzene is significantly higher than the
other quarterly averages while there is little difference among the other quarterly
averages. A review of the data shows that benzene concentrations measured at CHNJ
range from 0.234 |ig/m3 to 1.32 |ig/m3. While the maximum benzene concentration
was measured at CHNJ in April, the next six highest benzene concentrations were
measured in either January or March. Of the 22 benzene concentrations greater than
0.5 |ig/m3 measured at CHNJ, 15 were measured during the first quarter, accounting
for all of the concentrations measured that quarter. Benzene concentrations less than
the annual average concentration were not measured during the first quarter of 2014.
Observations for ELNJ from Table 16-4 include the following:
• The pollutants of interest with the highest annual average concentrations are
formaldehyde (4.44 ± 0.52 |ig/m3) and acetaldehyde (2.78 ± 0.21 |ig/m3). These are
the only two pollutants with annual average concentrations greater than 1 |ig/m3. Of
the VOCs, benzene has the highest annual average concentration (0.78 ± 0.09 |ig/m3).
• Similar to CSNJ and CHNJ, concentrations of formaldehyde measured at ELNJ
tended to be higher during the warmer months of the year, as indicated by the second
and third quarter average concentrations. A review of the data shows that
formaldehyde concentrations measured at ELNJ range from 1.77 |ig/m3 to
11.6 |ig/m3, with the maximum concentration measured on August 27, 2014, the same
day the maximum formaldehyde concentrations at CSNJ and CHNJ were measured.
Of the 18 formaldehyde concentrations greater than 5 |ig/m3 measured at ELNJ, all
but one was measured between April and September, with the five highest
concentrations measured during the third quarter of 2014.
16-25
-------
• Similar to CHNJ, the first quarter average concentration of benzene for ELNJ is
significantly higher than the other quarterly averages while there is little difference
among the other quarterly averages. A review of the data shows that benzene
concentrations measured at ELNJ range from 0.371 |ig/m3 to 2.57 |ig/m3. The five
highest benzene concentrations measured at ELNJ were measured between January
and March; further, seven benzene concentrations greater than 1 |ig/m3 were
measured during the first quarter of 2014 while no other calendar quarter has more
than one.
Observations for NBNJ from Table 16-4 include the following:
• In regards to carbonyl compound sampling at NBNJ, a defective sampler was
identified at the NBNJ site and the results between May 5, 2014 and
December 31, 2014 were invalidated. A new sampler was installed in January 2015.
As a result, only a first quarter average concentration could be calculated for this site.
• Even with this invalidation, some of the highest formaldehyde concentrations across
the program were measured at NBNJ. Concentrations of formaldehyde measured at
NBNJ between January and May 2014 span an order of magnitude, ranging from
2.15 |ig/m3 to 21.8 |ig/m3, including five formaldehyde concentrations greater than
10 |ig/m3 as well as the second and third highest formaldehyde concentrations
measured across the program. Higher concentrations of acetaldehyde were also
measured at NBNJ. Concentrations of acetaldehyde measured at NBNJ between
January and May 2014 range from 2.02 |ig/m3 to 8.92 |ig/m3, which is the fourth
highest acetaldehyde concentration measured across the program. The maximum
acetaldehyde and maximum formaldehyde concentrations were both measured on
January 11, 2014.
• The VOCs with the highest annual average concentrations for NBNJ are carbon
tetrachloride (0.63 ± 0.02 |ig/m3) and benzene (0.54 ± 0.07 |ig/m3).
• Similar to CHNJ and ELNJ, the first quarter average concentration of benzene for
NBNJ is significantly higher than the other quarterly averages while there is little
difference among the other quarterly averages. A review of the data shows that
benzene concentrations measured at NBNJ range from 0.234 |ig/m3 to 1.44 |ig/m3.
The maximum benzene concentration was measured on the same day in January as
the maximum acetaldehyde and formaldehyde concentrations. Four benzene
concentrations greater than 1 |ig/m3 were measured at NBNJ between January and
March, while only one was measured the rest of the year; further, benzene
concentrations less than 0.5 |ig/m3 were not measured during the first quarter of 2014
while each other calendar quarter has more than 10.
• The quarterly average concentrations of 1,3-butadiene have a similar pattern as the
quarterly average concentrations of benzene, although the differences are not
significant. A review of the data shows that 1,3-butadiene concentrations measured at
NBNJ range from 0.022 |ig/m3 to 0.279 |ig/m3, plus three non-detects. The maximum
1,3-butadiene concentration was also measured on January 11, 2014. Four
1,3-butadiene concentrations greater than 0.1 |ig/m3 were measured at NBNJ between
16-26
-------
January and March, while only one was measured the rest of the year; further, the
first quarter includes the fewest 1,3-butadiene concentrations less than 0.05 |ig/m3.
• The first and second quarter average concentrations of ethylbenzene appear higher
than the other two quarterly averages and have larger confidence intervals associated
with them. All 16 ethylbenzene concentrations greater than 0.25 |ig/m3 measured at
NBNJ were measured between January and June. The maximum ethylbenzene
concentration was measured at NBNJ on May 11, 2014(1.18 |ig/m3), although a
similar concentration was also measured on January 11, 2014 (1.10 |ig/m3).
Conversely, all nine ethylbenzene concentrations less than 0.1 |ig/m3 were measured
during the fourth quarter of 2014, which has the lowest quarterly average
concentration.
Additional observations for the New Jersey sites from Table 16-4 include:
• Formaldehyde and acetaldehyde were the pollutants of interest with the highest
annual average concentrations for each New Jersey site (with the exception of NBNJ,
where annual averages could not be calculated). Concentrations of these pollutants
were higher at CSNJ and ELNJ than CHNJ. Even though annual averages could not
be calculated for NBNJ, some of the highest concentrations of formaldehyde and
acetaldehyde across the program were measured at this site. Formaldehyde
concentrations were higher during the warmer months of the year at each site, as
indicated by the quarterly averages.
• Benzene and carbon tetrachloride have the highest annual average concentrations of
the VOC pollutants of interest. Concentrations of benzene were also higher at CSNJ
and ELNJ compared to CHNJ and NBNJ while concentrations of carbon tetrachloride
varied little across the sites.
• The maximum concentrations of several pollutants of interest in common among the
New Jersey sites were measured on the same sample day. For instance, the maximum
concentrations of acetaldehyde, formaldehyde, benzene, 1,3-butadiene were all
measured on the same day at NBNJ (January 11, 2014). Further, the maximum
1,3-butadiene concentrations for all four sites were measured on January 11, 2014.
The maximum concentrations of formaldehyde measured at CSNJ, CHNJ, and ELNJ
were all measured on August 27, 2014. The maximum acetaldehyde concentrations
for CSNJ and ELNJ were also measured on this date in August.
Tables 4-9 through 4-12 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 4-9 for VOCs a total of 10 times (CSNJ, twice;
CHNJ, once; ELNJ, five times; and NBNJ, twice).
• Three New Jersey sites appear in Table 4-9 for hexachloro-1,3-butadiene, with ELNJ,
NBNJ, and CHNJ ranking fourth, seventh, and eighth, respectively, for this pollutant.
16-27
-------
CSNJ's annual average concentration is similar to the other sites, although it ranks
20th among NMP sites sampling VOCs. Note that annual averages of this pollutant
vary by only 0.05 |ig/m3 across the program.
• ELNJ and CSNJ both appear among those sites with the highest annual average
concentrations of ethylbenzene and />dichlorobenzene, neither of which rank higher
than seventh.
• ELNJ also ranks seventh and eighth, respectively, for the annual average
concentrations of 1,3-butadiene and 1,2-dichloroethane.
• NBNJ ranks tenth for its annual average concentration of carbon tetrachloride.
• CSNJ and ELNJ both appear in Table 4-10 for both carbonyl compounds. CSNJ has
the second highest annual average concentration of formaldehyde and the sixth
highest annual average concentration of acetaldehyde, among NMP sites sampling
these pollutants. ELNJ has the third highest annual average concentrations of both
acetaldehyde and formaldehyde among NMP sites sampling carbonyl compounds.
16.4.2 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 16-4 for each of the New Jersey sites. Figures 16-11 through 16-18 overlay the 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
3.4.3.1, and are discussed below.
16-28
-------
Figure 16-11. Program vs. Site-Specific Average Acetaldehyde Concentrations
h
—o
N-
—o
0123456789 10
Concentration (ng/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 16-11 presents the box plots for acetaldehyde for the New Jersey sites and shows
the following:
• This figure presents the boxplots for CSNJ, CHNJ, and ELNJ. Recall from the
previous section that annual average concentrations for the carbonyl compounds
could not be calculated for NBNJ, and thus, a box plot for this site is not provided.
• The range of acetaldehyde concentrations is largest for CSNJ and smallest for CHNJ.
The minimum concentrations measured at CSNJ and ELNJ are greater than the
program-level first quartile, with the minimum for ELNJ just less than the program-
level median concentration.
• Among these sites, ELNJ's annual average concentration is just greater than the
annual average concentration for CSNJ, both of which are twice the annual average
concentration for CHNJ. The annual averages for CSNJ and ELNJ are both greater
than the program-level average concentration as well as the program-level third
quartile while the annual average for CHNJ is less than the program-level average
and median concentrations. Recall from the previous section that ELNJ has the third
highest annual average concentration among the NMP sites sampling this pollutant
and CSNJ's annual average concentration ranks sixth.
16-29
-------
Figure 16-12. Program vs. Site-Specific Average Benzene Concentrations
h
k
Program Max Concentration = 12.4 ng/m3
?
*
4
Program Max Concentration = 12.4 ng/m3
Program Max Concentration = 12.4 ng/m3
1
14
Program Max Concentration = 12.4 ng/m3
l
0
2
4 6
Concentration (ng/m3)
8
10
Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i
Site: Site Average
o
Site Concentration Range
Figure 16-12 presents the box plots for benzene for the New Jersey sites and shows the
following:
• The program-level maximum benzene concentration (12.4 |ig/m3) is not shown
directly on the box plots in Figure 16-12 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.
• Compared to the maximum benzene concentration measured across the program, the
maximum concentrations measured at each New Jersey site are considerably less
(none were greater than 3 |ig/m3). The range of benzene concentrations measured was
largest for ELNJ and CSNJ and smallest at CHNJ and NBNJ.
The annual average concentrations of benzene for ELNJ and CSNJ are similar to the
average concentration across the program, while the annual averages for CHNJ and
NBNJ fall between the program-level first and second quartiles.
16-30
-------
Figure 16-13. Program vs. Site-Specific Average 1,3-Butadiene Concentrations
¦H
Program Max Concentration = 5.90 ng/m3
¦H-
Program Max Concentration = 5.90 ng/m3
Program Max Concentration = 5.90 ng/m3
B
Program Max Concentration = 5.90 ng/m3
0.2
0.4 0.6
Concentration (ng/m3)
0.8
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
i
Site: Site Average
Site Concentration Range
o
Figure 16-13 presents the box plots for 1,3-butadiene for the New Jersey sites and shows
the following:
• Similar to benzene, the program-level maximum 1,3-butadiene concentration
(5.90 |ig/m3) is not shown directly on the box plots in Figure 16-13 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 |ig/m3.
• All of the 1,3-butadiene concentrations measured at the New Jersey sites are less than
0.3 |ig/m3, A few non-detects were measured at NBNJ and CHNJ while none were
measured at the other two sites. The minimum concentration measured at ELNJ is
greater than the program-level first quartile.
The annual average concentrations of 1,3-butadiene for the New Jersey sites range
from 0.06 ± 0.01 |ig/m3 (CHNJ) to 0.12 ± 0.01 |ig/m3 (ELNJ), with only ELNJ's
annual average concentration greater than the program-level average.
16-31
-------
Figure 16-14. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations
CSNJ
Program Max Concentration = 3.06 |-ig/m3
CHNJ
Program Max Concentration = 3.06 i-ig/m3
ELNJ
Program Max Concentration = 3.06 \±g/m3
Program Max Concentration = 3.06 \±g/m3
0 0.5 1 1.5 2 2.5
Concentration (ng/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 16-14 presents the box plots for carbon tetrachloride for the New Jersey sites and
shows the following:
• The scale of the box plots in Figure 16-14 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 median and average concentrations are similar and plotted nearly on
top of each other.
• All of the carbon tetrachloride concentrations measured at these sites range from
0.25 |ig/m3 and 0.85 |ig/m3. The maximum concentrations measured at three of the
sites are similar to each other, with the maximum for ELNJ slightly higher than the
others. The minimum concentrations measured at each site are more variable.
• The annual average concentrations of carbon tetrachloride for the New Jersey sites
are similar to each other, ranging from 0.60 |ig/m3 and 0.63 |ig/m3, with each just less
than the program-level average concentration of 0.64 |ig/m3.
16-32
-------
Figure 16-15. Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations
¦
Program Max Concentration = 27.4 ng/m3
¦
Program Max Concentration = 27.4 ng/m3
¦
Program Max Concentration = 27.4 ng/m3
Program Max Concentration = 27.4 |ig/m3
0.4 0.6
Concentration {[j.g/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range
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 (27.4 |ig/m3) is
considerably greater than the majority of measurements.
• All of the concentrations of 1,2-dichloroethane measured at the New Jersey sites are
less than the program-level average concentration of 0.31 |ig/m3, which is being
driven by the measurements at the upper end of the concentration range. In fact, all of
the concentrations measured at the New Jersey sites are less than half the program-
level average concentration.
• The annual average concentrations for CSNJ, CHNJ, and NBNJ are similar to the
program-level median concentration (0.081 |ig/m3) while the annual average
concentration for ELNJ is just slightly higher (0.086 |ig/m3).
16-33
-------
Figure 16-16. Program vs. Site-Specific Average Ethylbenzene Concentrations
1
ELNJ
O i
{J 1
1
-4-
0 0.5 1 1.5 2 2.5 3 3.5
Concentration (ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 16-16 presents the box plots for ethylbenzene for the New Jersey sites and shows
the following:
• This figure presents the box plots for CSNJ, ELNJ, and NBNJ. Ethylbenzene was not
identified as a pollutant of interest for CHNJ and thus, a box plot is not presented for
this site.
• The maximum ethylbenzene concentration measured across the program (3.35 |ig/m3)
was measured at CSNJ; the next highest concentration measured at this site was
considerably less (1.07 |ig/m3). Two ethylbenzene concentrations greater than
1 |ig/m3 were also measured at NBNJ. The range of ethylbenzene concentrations
measured at ELNJ is smaller than the other sites, although the minimum
concentration measured at ELNJ is greater than the program-level first quartile.
• The annual average concentration for NBNJ is similar to the program-level average
concentration (0.25 |ig/m3) while the annual average concentrations for CSNJ and
ELNJ are greater than the program-level average concentration.
16-34
-------
Figure 16-17. Program vs. Site-Specific Average Formaldehyde Concentrations
N
—o ¦
K
h
O
0 3 6 9 12 15 18 21 24 27
Concentration {[j.g/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 16-17 presents the box plots for formaldehyde for the New Jersey sites and shows
the following:
• This figure presents the box plots for CSNJ, CHNJ, and ELNJ. Recall from the
previous section that annual average concentrations for the carbonyl compounds
could not be calculated for NBNJ, and thus, a box plot for this site is not provided.
• The range of formaldehyde concentrations measured at CSNJ is similar to the range
measured at ELNJ, with the smallest range measured at CHNJ. Although all
considerably less than the program-level maximum concentration, the maximum
concentration measured at CHNJ is roughly half the maximum concentrations
measured at CSNJ and ELNJ. The minimum concentrations measured at CSNJ and
ELNJ are greater than the program-level first quartile.
• The annual average concentrations of formaldehyde for CSNJ and ELNJ are similar
to each other and both are greater than the program-level average and third quartile.
Recall from the previous section that CSNJ and ELNJ have the second and third
highest annual average concentrations of formaldehyde, respectively, among the
NMP sites sampling this pollutant. The annual average for CHNJ is less than the
program-level average and median concentrations.
16-35
-------
Figure 16-18. Program vs. Site-Specific Average Hexachloro-l,3-butadiene Concentrations
4
0
0.1
0.2
0.3 0.4
Concentration (ng/m3)
0.5
0.6
0.7
Program:
1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site:
Site Average
o
Site Concentration Range
Figure 16-18 presents the box plots for hexchloro-1,3-butadiene for the New Jersey sites
and shows the following:
• This figure presents the box plots for hexchloro-1,3-butadiene for CSNJ, ELNJ,
and NBNJ, the three New Jersey sites for which this pollutant was identified as a
pollutant of interest. Note that the first, second, and third quartiles for hexchloro-
1,3-butadiene are zero at the program-level and therefore not visible on the box
plots due to the large number of non-detects.
• The range of hexchloro-1,3-butadiene concentrations measured at these sites are
similar to each other, with all measured detections less than 0.13 |ig/m3, and the
number of measured detections ranging from 12 (CSNJ) to 18 (ELNJ), although
none were greater than the MDL.
• The annual average hexchloro-1,3-butadiene concentrations for ELNJ and NBNJ
are just greater than the program-level average concentration, though by only a
small margin, while the annual average hexchloro-1,3-butadiene concentration for
CSNJ is just less than the program-lev el average concentration.
16-36
-------
16.4.3 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 3.4.3.2.
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 16-40 present the 1-year statistical metrics for each of the pollutants
of interest first for CHNJ, then for ELNJ and NBNJ. 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. CSNJ began sampling under the NMP is 2013; thus, a trends analysis was not
performed for this site.
16-37
-------
Figure 16-19. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at CHNJ
I
I
Maximum
Concentration for
2004 is 29.1 ng/m3
I
1
o
o
i
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year
O 5th Percentile
O 95th Percentile Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
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.
• 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. Only 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).
• An overall decreasing trend in the 1-year average and median concentrations is shown
though 2006, with the exception of 2004, when the maximum concentrations were
measured. Between 2006 and 2010, the 1-year average and median concentrations
changed little, with the 1-year average concentrations 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, the 95th percentile decreased
nearly 1 |ig/m3, 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. Additional decreases for all
16-38
-------
of the statistical parameters are shown for 2013. The range of concentrations
measured at CHNJ compressed even further for 2014, with the majority of the
measurements falling into the smallest range since the onset of sampling.
Figure 16-20. Yearly Statistical Metrics for Benzene Concentrations Measured at CHNJ
I
~
T
n
o.
o
o
T
2006
2007
2008
1
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 May 2001.
2 A 1-year average is not presented due to low completeness in 2005.
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). Only 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 decreasing trend through
2007, although a 1-year average concentration is not provided for 2001 or 2005.
16-39
-------
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 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 is at a minimum for the following year; 2010 has the smallest range
of benzene measurements of any year 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. Although the range of
concentrations measured is at a maximum for 2013, all of the statistical metrics
exhibit decreases for 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.
Despite the year-to-year increases or decreases shown in the 1-year average
concentrations of benzene between 2006 and 2014, the averages have varied by less
than 0.2 |ig/m3, ranging from 0.47 |ig/m3 (several years) to 0.64 |ig/m3 (2012).
16-40
-------
Figure 16-21. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at CHNJ
I
i ~ y-
4.
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year
5th Percentile
— Minimum
— 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. Only 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 all
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 reported 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 that follow.
While these changes do correspond with the changes in non-detects discussed above,
the measurement of concentrations on the highest end of the concentration range
became more frequent over the years, particularly in 2014. The number of
1,3-butadiene concentrations greater than 0.5 |ig/m3 ranged from five to 10 between
2006 and 2008, decreased to four for 2009 and two for 2010, then increased each year
afterward, reaching a maximum of 40 in 2014, and accounting for more than half of
16-41
-------
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.
Figure 16-22. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
CHNJ
I
l
T
l
l
1
20011 2002 2003 2004 20052 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year
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.
Observations from Figure 16-22 for carbon tetrachloride concentrations measured at
CHNJ include the following:
• The range of carbon tetrachloride concentrations measured appears to increase
significantly from 2001 to 2002, with fairly similar ranges measured between 2003
and 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 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
calculated for 2008. Thirteen concentrations measured in 2008 were greater than the
maximum concentration measured in 2007. 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.
16-42
-------
• 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
2014, 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 2014 vary by less than 0.10 |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
-5- *>
1 I xA.
i
o
o
I
20011 2002 2003 2004 2005 * 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year
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 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
later years of sampling. The number of measured detections decreased slightly for
2013 but still account for more than 85 percent of measurements. For, 2014, the
percentage of measured detections is at a maximum of 97 percent.
• 2012 is the first year that the median concentration and 5th percentile are greater than
zero. Aside from the three non-detects, the range of concentrations measured in 2012
is relatively small, ranging from 0.053 |ig/m3 to 0.121 |ig/m3. The 1-year average and
16-43
-------
median concentrations calculated for 2012 are less than 0.001 |ig/m3 apart, indicating
little variability associated with the concentrations measured in 2012.
• The 5th percentile returned to zero for 2013, as six additional non-detects were
measured in 2013. However, the 1-year average and median concentrations changed
little. The effects of the additional non-detects are balanced by the additional
concentrations measured at the upper end of the concentration range. The number of
1,2-dichloroethane concentrations greater than 0.1 |ig/m3 increased from four in 2012
to 10 in 2013.
• The statistical metrics for 2014 resemble those shown for 2012, and for the second
time, the 5th percentile is greater than zero.
Figure 16-24. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at CHNJ
I
—
T
Maximum
Concentration for
2004 is 57.2 ng/m3
pl-
.2.
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
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 2001.
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 almost four times the maximum concentrations shown for other
years.
16-44
-------
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 concentrations changed little through 2009.
Less than 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, when both
statistical parameters are at a minimum. 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 fell 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. The maximum formaldehyde
concentration measured at CHNJ in 2013 is the lowest maximum concentration for
any given year.
Relatively little change is shown in the range of concentrations measured in 2014
compared to 2013 and the 1-year average concentrations for these two years vary by
less than 0.1 |ig/m3.
16-45
-------
Figure 16-25. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at ELNJ
E
If 10
I
H
¦o-
o
jr
9
T
T
o
T
"r
s
!
I
2000 A 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year
5th Percentile
95th Percentile
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 prior to 2008.
• 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.
16-46
-------
• 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.
• The range of concentrations measured in 2014 is the smallest since the onset of
sampling at ELNJ. The slight increases in the 1-year average and median
concentrations shown for 2014 result from a decrease in the number of concentrations
at the lower end of the range. Nine concentrations measured in 2013 are less than the
minimum concentration measured in 2014 (1.48 |ig/m3); further, the number of
acetaldehyde concentrations less than 2 |ig/m3 decreased from 20 in 2013 to eight in
2014.
Figure 16-26. Yearly Statistical Metrics for Benzene Concentrations Measured at ELNJ
Maximum
Concentration for
2009 is 34.3 ng/m3
o
'-r
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009
Year
2010 2011 2012 2013 2014
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
16-47
-------
was measured in 2009. The third highest concentration was also measured in 2009. In
all, 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 concentrations is
shown through 2007.
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 nearly tripled from 2007 to 2008, with eight concentrations measured in
2007 less than the minimum concentration measured in 2008; in addition, the 5th
percentile increased as well, indicating that the outlier may not be 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.
Additional decreases are shown for 2013, as no benzene concentrations greater than
2 |ig/m3 were 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
2014 compared to 2013, the 1-year average benzene concentration is at a minimum
for 2014 (0.78 |ig/m3).
16-48
-------
Figure 16-27. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at ELNJ
I
I
->
T
1
Maximum
Concentration for
2009 is 2.57 ng/m3
X
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year
5th Percentile
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 was measured at ELNJ in 2009 and is
nearly two and a half times the next highest concentration (measured in 2001). These
are the only concentrations of 1,3-butadiene measured at ELNJ that are greater than
1 |ig/m3 and only 16 concentrations measured at ELNJ are greater than 0.5 |ig/m3.
• 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).
• There is a decreasing trend in the 1-year average concentration through 2004, after
which the 1-year average concentration remains fairly static. 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 2005 and 2014, the 1-year
average concentration has ranged from 0.11 |ig/m3 (2013) to 0.16 |ig/m3 (2006 and
2009).
• Concentrations of 1,3-butadiene measured at ELNJ have become less variable in
recent years, with concentrations measured in 2010, 2013, and 2014 exhibiting the
16-49
-------
least variability. These years have the smallest range of concentrations measured and
the smallest differences between the 5th and 95th percentiles (2010 and 2013 only),
the range within which the majority of concentrations fall.
Figure 16-28. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
ELNJ
16
1.4
12
1.0
T
nE
1
J 08
I
I
1
T T
c
u
0.6
0.4
I
o
-L
o
O'
¦o
-i
V
O
¦o
¦o
©
L^J
"V
r
0 2
¦¦
.
0.0
¦¦
T
2000 1 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year
1
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 16-28 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; 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 a decreasing trend after 2005 and is at a
minimum for 2013. Less than 0.25 |ig/m3 separates these parameters for 2013. A
slight widening of the range is shown for 2014.
16-50
-------
Figure 16-29. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
ELNJ
i
I
o.oo ^- o -S- ^
2000 1 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
o 95th Percentile
Average
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 measured detections were not measured 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. This explains the significant increase in the 1-year average
concentrations shown for the later years of sampling.
• 2012 is the first year that the median concentration is greater than zero. Aside from
the six non-detects, the range of concentrations measured in 2012 is relatively small,
ranging from 0.061 |ig/m3 to 0.150 |ig/m3. The 1-year average and median
concentrations calculated for 2012 are approximately 0.0015 |ig/m3 apart, indicating
relatively little variability associated with the concentrations measured in 2012.
• For 2013, the number of non-detects more than doubled (from six in 2012 to 14 in
2013), accounting for nearly one-quarter of the concentrations measured. Yet, the
1-year average concentration changed little and the median concentration increased.
Although the maximum concentration increased only slightly from 2012 to 2013, 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.
16-51
-------
• Only two non-detects were measured at ELNJ in 2014 and thus, is the first year that
the 5th percentile is greater than zero.
Figure 16-30. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at
ELNJ
I
T
r-J-i
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year
5th Percentile
- Minimum
O 95th Percentile
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:
• The trends graph for ELNJ's ethylbenzene concentrations resembles the trends graph
for ELNJ's benzene concentrations.
• There is an overall decreasing trend in the 1-year average and median concentrations
between 2001 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, meaning 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 measured in 2009 more closely resemble those collected in 2007
than 2008, with the exception of the maximum concentration measured.
16-52
-------
• The smallest range of ethylbenzene concentrations was measured in 2010, with all
concentrations measured spanning less than 0.75 |ig/m3.
• Between 2009 and 2014, the majority of concentrations fell within a fairly similar
range and the 1-year average concentrations did not change significantly, ranging
from 0.36 |ig/m3 (2014) to 0.51 |ig/m3 (2011).
Figure 16-31. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
ELNJ
T
I
i
I
"r
I
TT
T
:
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year
5th Percentile - Minimum
95th Percentile ¦ Average
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
(15.88 |ig/m3), A total of 15 concentrations greater than 10 |ig/m3 have been
measured at ELNJ, with the most measured in 2007 (three).
• 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 during this time, 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.
16-53
-------
• 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
shown through 2010, followed by a decrease for 2011, then another round of
increasing. The 1-year average concentration of formaldehyde for ELNJ 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).
Figure 16-32. Yearly Statistical Metrics for Hexachloro-l,3-butadiene Concentrations
Measured at ELNJ
¦ Maximum
j Concentration for
| 2006 is 0.534 (J.g/m3
T
o O- ^ ^
o-
o
"'•»> ^
-o
<
~
20001 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
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 16-32 for hexachloro-l,3-butadiene concentrations measured at
ELNJ include the following:
• There were no measured detections of hexachloro-1,3-butadiene measured at ELNJ
during the first 5 years of sampling.
• The number of measured detections increased to 13 for 2005, representing 22 percent
of measurements, then decreased to five for 2006. Between 2007 and 2010, a single
measured detection was measured (2008). Beginning in 2011, the number of
measured detections began increasing, from five for 2011 to seven for 2012, 13 in
2013, and is at a maximum of 18 for 2014.
16-54
-------
Figure 16-33. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at NBNJ
Maximum
Concentration for
2004islll|ig/m3
I
I
I
T
T
I
o
I
o
20011 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 20142
Year
5th Percentile
- Minimum
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 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.
• The maximum acetaldehyde concentration was measured in 2004 (111 |ig/m3). This
concentration is nearly seven times higher, and an order of magnitude higher, than the
next highest concentration (16.2 |ig/m3, measured in 2005).
• Of the 30 concentrations greater than 8 |ig/m3 measured at NBNJ, 28 were measured
at NBNJ in 2004 or 2005 (with one each measured in 2008 and 2014). 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.
• The 1-year average concentration decreases significantly between 2005 and 2007, as
do all of the other statistical parameters. This is followed by a significant increase in
the concentrations measured for 2008, with the range of concentrations measured
doubling.
16-55
-------
• 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 sampler issue resulted in the invalidation of carbonyl compound data
from May 2014 through the end of the year. While a 1-year average concentration is
not provided for 2014 in Figure 16-33, the range of concentrations is provided. 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.
Figure 16-34. Yearly Statistical Metrics for Benzene Concentrations Measured at NBNJ
4.5
4.0
3.5
3.0
mE
3 25
c
o
§ 2.0
c
u
1.5
1.0
0.5
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.
Observations from Figure 16-34 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.
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year
16-56
-------
The maximum benzene concentration was measured in 2012 (4.00 |ig/m3); aside from
this measurement, only three additional concentrations of benzene greater than
3 |ig/m3 have been measured at NBNJ.
Although a slight decreasing trend in the 1-year average concentration is shown
between 2002 and 2004, a significant decrease is shown between 2005 and 2007. The
median concentration is less than 0.5 |ig/m3 for the first time in 2007 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 fluctuation in the
range of concentrations measured.
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. The minimum and 5th percentile
increased considerably for 2012; there were 17 benzene concentrations measured in
2011 that 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 and median concentrations is shown after
2012 and are both at a minimum for 2014 (and just slightly less than the 1-year
average and median concentrations calculated for 2007).
16-57
-------
Figure 16-35. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at NBNJ
~
o
T
"r
a
T
T
T
2.
T
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year
5th Percentile
— Minimum
— Maximum
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 16-35 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.
• 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 almost 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. Three non-detects were
measured in 2013 and again in 2014.
• 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
16-58
-------
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, 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.
• The 1,3-butadiene concentrations measured in 2013 decreased from 2012 levels but
were still higher than those measured in the previous years. Despite a few higher
concentrations measured in 2014, slight decreases are also shown for the 1-year
average and median concentrations for 2014.
Figure 16-36. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
NBNJ
I
rln
I
X
I
20011 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year
O 5th Percentile - Minimum - Med en - Maximum o 95th Percentile Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
16-59
-------
Observations from Figure 16-36 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 both the lower and upper end of the concentration
range. 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.
• 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.
• 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.
• Carbon tetrachloride concentrations measured in 2013 exhibit the least amount of
variability in that the difference between the 5th and 95th percentiles is at a minimum
and the difference between the 1-year average and median concentrations is less than
0.005 |ig/m3.
• The smallest of range of carbon tetrachloride concentrations was measured in 2014.
16-60
-------
Figure 16-37. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
NBNJ
x
I
I
0.00 ^ JOta""
20011 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
a u
o 5th Percentile
— Minimum
— Maximum
o 95th Percentile
..4.... Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
Observations from Figure 16-37 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 one and four 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 a total of three in 2009, to 11 for
2010, 18 in 2011, 58 for 2012, 59 in 2013, and 57 in 2014. This increase in the
number of measured detections is very similar to what was exhibited by the
concentrations measured at CHNJ and ELNJ. This also explains the significant
increase in the 1-year average concentrations shown, particularly for the later years of
sampling.
• 2012 is the first year that the median concentration is greater than zero. Aside from
the two non-detects, the range of concentrations measured in 2012 is relatively small,
ranging from 0.053 |ig/m3 to 0.146 |ig/m3. The 1-year average and median
concentrations calculated for 2012 are 0.001 |ig/m3 apart, indicating relatively little
variability associated with the concentrations measured in 2012. Similar observations
can be made for 2013 and 2014.
16-61
-------
Figure 16-38. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at
NBNJ
I
o
I
Maximum
Concentration for
2011 is 11.5 Ui/m3
~ X
20011 2002 2003 2004 2005 2005 2007 2008 2009 2010 2011 2012 2013 2014
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.
Observations from Figure 16-38 for ethylbenzene concentrations measured at NBNJ
include the following:
• The maximum ethylbenzene concentration (36.3 |ig/m3) was measured at NBNJ on
May 25, 2001. All but one of the 23 concentrations of ethylbenzene greater than
5 |ig/m3 were measured at NBNJ in 2001, and were measured on nearly consecutive
sample days between May and October.
• All but five concentrations measured in 2001 are greater than the maximum
concentration measured in 2002. As a result, each of the statistical parameters exhibit
considerable decreases, with the median concentration decreasing from 7.29 |ig/m3 to
0.91 |ig/m3. Additional decreases shown for 2003.
• The slight increases shown in the 1-year average concentration between 2003 and
2005 is followed by a significant decrease for 2006. Between 2006 and 2010, the
1-year average concentration varied by less than 0.1 |ig/m3, ranging from 0.17 |ig/m3
(2009) to 0.26 |ig/m3 (2006). The concentrations measured during these years are
considerably less than those prior to 2006.
• While most of the ethylbenzene concentrations measured in 2011 fall into a similar
range as the previous years, the maximum concentration measured in 2011
(11.5 |ig/m3) is an order of magnitude greater than the second highest concentration
measured that year (1.01 |ig/m3). While this measurement is driving the 1-year
16-62
-------
average concentration for 2011, the other statistical metrics exhibit increases as well,
indicating concentrations were higher overall in 2011. The number of ethylbenzene
concentrations greater than 0.25 |ig/m3 accounted for more than half of the
concentrations measured at NBNJ in 2011, nearly doubling from 17 in 2010 to 33 in
2011.
• Ethylbenzene concentrations measured in 2012 and 2013 resemble those measured in
2010.
• Despite having similar 1-year average concentrations (all three are 0.25 |ig/m3), the
ethylbenzene concentrations measured in 2014 exhibit more variability than the
previous two years. The range of concentrations measured more than doubled, and the
range within which the majority of concentrations fall is at its largest since 2005.
Figure 16-39. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
NBNJ
Maximum
Concentration for
2004 is 96.1 ng/m3
I
o
O
—
"""9 s ^ -S -4
I
4-rS
20011 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 20142
Year
o 5th Percentile
— Minimum
95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
Observations from Figure 16-39 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). In all,
16-63
-------
concentrations greater than 20 |ig/m3 have been measured during five of the 14 years
shown, including 2014.
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
2003. If the maximum concentration was excluded from the calculations for 2004, the
1-year average concentration for 2004 would fall between those calculated for 2003
and 2005, exhibiting lesser increases, but an increase none the less. The number of
formaldehyde concentrations greater than 3 |ig/m3 doubled from 2003 to 2004, from
16 to 34. Outlier aside, a similar range of concentration to 2004 was measured in
2005, with the median concentration exhibiting another 1 |ig/m3 increase.
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. The measurements for 2011 exhibit a considerable
amount of variability compared to the rest of the years within this period. The 95th
percentile for 2011 is more than double the 95th percentile for the other years within
this period. Yet, the median concentrations are nearly the same for 2011 and 2012.
Several of the statistical parameters exhibit at least a slight increase for 2013. The
number of formaldehyde concentrations greater than 2 |ig/m3 measured at NBNJ in
2013 increased considerably, from 18 in 2012 to 33 in 2013, accounting for more than
half of the measurements in 2013.
A 1-year average concentration is not provided in Figure 16-39 for 2014, as a sampler
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 16-33. 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.
16-64
-------
Figure 16-40. Yearly Statistical Metrics for Hexachloro-l,3-butadiene Concentrations
Measured at NBNJ
o
o
0.00
1
20011 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Year
o 5th Percentile
— Minimum
— Maximum
o 95th Percentile
..4.... Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
Observations from Figure 16-40 for hexachloro-l,3-butadiene concentrations measured at
NBNJ include the following:
• There were no measured detections of hexachloro-1,3-butadiene measured during the
first 4 years of sampling at NBNJ.
• The number of measured detections increased to nine for 2005, representing
16 percent of measurements, then decreased to five for 2006. The number of
measured detections returned to zero between 2007 and 2009. A single measured
detection was reported for 2010, after which the number of measured detections has
increased each year (eight for 2011, 11 for 2012, 16 for 2013, and 17 for 2014).
• Even though measured detections accounted for more than one-quarter of the
concentrations measured during the last two years of sampling, none have been
greater than the detection limit for this pollutant.
16-65
-------
16.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at each New Jersey monitoring site. Refer to Sections 3.2, 3.4.3.3, and
3.4.3.4 for definitions and explanations regarding the various toxicity factors, time frames, and
calculations associated with these risk-based screenings.
16.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the New Jersey 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 3.4.3.3 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 16-5, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
16-66
-------
Table 16-5. Risk Approximations for the New Jersey Monitoring Sites
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Camden, New Jersey - CSNJ
Acetaldehyde
0.0000022
0.009
60/60
2.49
±0.22
5.48
0.28
Benzene
0.0000078
0.03
61/61
0.76
±0.09
5.90
0.03
1.3 -Butadiene
0.00003
0.002
61/61
0.09
±0.01
2.84
0.05
Carbon Tetrachloride
0.000006
0.1
61/61
0.61
±0.03
3.64
0.01
1,2-Dichloroethane
0.000026
2.4
59/61
0.08
±0.01
2.10
<0.01
Ethylbenzene
0.0000025
1
61/61
0.33
±0.11
0.81
<0.01
Formaldehyde
0.000013
0.0098
60/60
4.48
±0.52
58.20
0.46
Hexachloro-1,3 -butadiene
0.000022
0.09
12/61
0.02
±0.01
0.34
<0.01
Chester, New Jersey - CHNJ
Acetaldehyde
0.0000022
0.009
60/60
1.25
±0.12
2.75
0.14
Benzene
0.0000078
0.03
61/61
0.47
±0.05
3.69
0.02
1.3 -Butadiene
0.00003
0.002
59/61
0.06
±0.01
1.84
0.03
Carbon Tetrachloride
0.000006
0.1
61/61
0.60
±0.03
3.62
0.01
1,2-Dichloroethane
0.000026
2.4
60/61
0.08
± <0.01
2.01
<0.01
Formaldehyde
0.000013
0.0098
60/60
2.06
±0.34
26.79
0.21
— = A Cancer URE or Noncancer RfC is not available.
NA = Not available due to the criteria for calculating an annual average concentration.
16-67
-------
Table 16-5. Risk Approximations for the New Jersey Monitoring Sites (Continued)
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Elizabeth, New Jersey - ELNJ
Acetaldehyde
0.0000022
0.009
61/61
2.78
±0.21
6.11
0.31
Benzene
0.0000078
0.03
59/59
0.78
±0.09
6.10
0.03
1.3 -Butadiene
0.00003
0.002
59/59
0.12
±0.01
3.58
0.06
Carbon Tetrachloride
0.000006
0.1
59/59
0.62
±0.03
3.71
0.01
1,2-Dichloroethane
0.000026
2.4
57/59
0.09
±0.01
2.24
<0.01
Ethylbenzene
0.0000025
1
59/59
0.36
±0.05
0.91
<0.01
Formaldehyde
0.000013
0.0098
61/61
4.44
±0.52
57.73
0.45
Hexachloro-1,3 -butadiene
0.000022
0.09
18/59
0.02
±0.01
0.54
<0.01
North Brunswick, New Jersey - NBNJ
Acetaldehyde
0.0000022
0.009
20/20
NA
NA
NA
Benzene
0.0000078
0.03
60/60
0.54
±0.07
4.19
0.02
1.3 -Butadiene
0.00003
0.002
57/60
0.07
±0.01
1.97
0.03
Carbon Tetrachloride
0.000006
0.1
60/60
0.63
±0.02
3.79
0.01
1,2-Dichloroethane
0.000026
2.4
58/60
0.08
±0.01
2.10
<0.01
Ethylbenzene
0.0000025
1
60/60
0.25
±0.06
0.62
<0.01
Formaldehyde
0.000013
0.0098
20/20
NA
NA
NA
Hexachloro-1,3 -butadiene
0.000022
0.09
17/60
0.02
±0.01
0.51
<0.01
— = A Cancer URE or Noncancer RfC is not available.
NA = Not available due to the criteria for calculating an annual average concentration.
16-68
-------
Observations from Table 16-5 include the following:
• For CSNJ, the pollutants of interest with the highest annual average concentrations
are formaldehyde, acetaldehyde, and benzene. Formaldehyde has the highest cancer
risk approximation for this site (58.20 in-a-million), followed by benzene and
acetaldehyde. The 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 CSNJ. CSNJ's cancer risk approximation for formaldehyde is the highest
cancer risk approximation among the pollutants of interest for the New Jersey sites
and the third highest among all NMP sites. 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 approximation for
CSNJ (0.46).
• For CHNJ, the pollutants with the highest annual average concentrations are
formaldehyde, acetaldehyde, and carbon tetrachloride. Formaldehyde has the highest
cancer risk approximation for this site (26.79 in-a-million), followed by benzene and
carbon tetrachloride. The cancer risk approximation for formaldehyde is at least an
order of magnitude higher than the approximations for the other pollutants of interest
for CHNJ. None of the pollutants of interest for CHNJ 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 (0.21).
• 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
approximation for benzene is similar to the cancer risk approximation for
acetaldehyde. ELNJ's cancer risk approximation for formaldehyde (57.73 in-a-
million) is similar to the cancer risk approximation calculated for CSNJ and is the
fourth highest cancer risk approximation among all NMP 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 approximation for ELNJ (0.45).
• For NBNJ, the pollutants with the highest annual average concentrations are carbon
tetrachloride, benzene, and ethylbenzene. Recall, however, that annual average
concentrations could not be calculated for the carbonyl compounds. Benzene has the
highest cancer risk approximation for NBNJ (4.19 in-a-million), followed by carbon
tetrachloride (3.79 in-a-million). None of the pollutants of interest for NBNJ have
noncancer hazard approximations greater than 1.0, indicating that adverse noncancer
health effects are not expected from these individual pollutants. All of the noncancer
hazard approximation calculated for NBNJ are less than 0.05.
16-69
-------
16.5.2 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 16-6 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
that have cancer toxicity factors. Table 16-6 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 16-6 provides the pollutants with the highest cancer risk approximations (in-a-million) for
each New Jersey site, as presented in Table 16-5. The emissions, toxicity-weighted emissions,
and cancer risk approximations are shown in descending order in Table 16-6. Table 16-7
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 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 16.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
16-70
-------
Table 16-6. 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)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Camden, New Jersey (Camden County) - CSNJ
Benzene
130.73
Formaldehyde
1.24E-03
Formaldehyde
58.20
Formaldehyde
95.20
Benzene
1.02E-03
Benzene
5.90
Ethylbenzene
64.14
1,3-Butadiene
6.27E-04
Acetaldehyde
5.48
Acetaldehyde
55.11
Naphthalene
3.58E-04
Carbon Tetrachloride
3.64
1.3 -Butadiene
20.89
POM, Group 2b
2.36E-04
1,3-Butadiene
2.84
Tetrachloroethylene
11.72
Nickel, PM
2.02E-04
1,2-Dichloroethane
2.10
Naphthalene
10.52
POM, Group 2d
1.65E-04
Ethylbenzene
0.81
POM, Group 2b
2.68
Ethylbenzene
1.60E-04
Hexachloro-1,3 -butadiene
0.34
POM, Group 2d
1.87
Arsenic, PM
1.38E-04
Trichloroethylene
1.20
POM, Group 5a
1.23E-04
Chester, New Jersey (Morris County) - CHNJ
Benzene
161.55
Benzene
1.26E-03
Formaldehyde
26.79
Formaldehyde
95.57
Formaldehyde
1.24E-03
Benzene
3.69
Ethylbenzene
86.05
1,3-Butadiene
7.62E-04
Carbon Tetrachloride
3.62
Acetaldehyde
58.64
Naphthalene
3.43E-04
Acetaldehyde
2.75
1,3-Butadiene
25.41
Ethylbenzene
2.15E-04
1,2-Dichloroethane
2.01
Tetrachloroethylene
11.82
POM, Group 2b
2.05E-04
1,3-Butadiene
1.84
Naphthalene
10.09
Nickel, PM
1.97E-04
Dichloromethane
5.27
POM, Group 2d
1.45E-04
POM, Group 2b
2.33
POM, Group 5a
1.31E-04
POM, Group 2d
1.64
Arsenic, PM
1.30E-04
-------
Table 16-6. 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)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Elizabeth, New Jersey (Union County) - ELNJ
Benzene
138.53
Formaldehyde
1.30E-03
Formaldehyde
57.73
Formaldehyde
99.64
Benzene
1.08E-03
Acetaldehyde
6.11
Ethylbenzene
74.03
1,3-Butadiene
6.29E-04
Benzene
6.10
Acetaldehyde
59.30
Nickel, PM
4.27E-04
Carbon Tetrachloride
3.71
1.3 -Butadiene
20.96
Naphthalene
3.75E-04
1,3-Butadiene
3.58
Tetrachloroethylene
14.36
Arsenic, PM
2.03E-04
1,2-Dichloroethane
2.24
Naphthalene
11.04
Ethylbenzene
1.85E-04
Ethylbenzene
0.91
Dichloro methane
2.96
POM, Group 2b
1.84E-04
Hexachloro-1,3 -butadiene
0.54
POM, Group 2b
2.09
Hexavalent Chromium
1.55E-04
Trichloroethylene
1.77
POM, Group 2d
1.33E-04
North Brunswick, New Jersey (Middlesex County) - NBNJ
Benzene
213.63
Formaldehyde
1.81E-03
Benzene
4.19
Formaldehyde
139.48
Benzene
1.67E-03
Carbon Tetrachloride
3.79
Ethylbenzene
110.60
1,3-Butadiene
9.59E-04
1,2-Dichloroethane
2.10
Acetaldehyde
83.83
Naphthalene
5.42E-04
1,3-Butadiene
1.97
1.3 -Butadiene
31.96
Hydrazine
4.38E-04
Ethylbenzene
0.62
Tetrachloroethylene
24.38
POM, Group 2b
2.82E-04
Hexachloro-1,3 -butadiene
0.51
Naphthalene
15.95
Ethylbenzene
2.77E-04
POM, Group 2b
3.20
POM, Group 2d
2.03E-04
Trichloroethylene
3.19
Arsenic, PM
1.86E-04
Dichloro methane
3.03
POM, Group 5a
1.85E-04
-------
Table 16-7. 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)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Camden, New Jersey (Camden County) - CSNJ
Toluene
422.02
Acrolein
598,846.68
Formaldehyde
0.46
Hexane
272.19
1,3-Butadiene
10,445.69
Acetaldehyde
0.28
Xylenes
249.52
Formaldehyde
9,713.82
1,3-Butadiene
0.05
Benzene
130.73
Acetaldehyde
6,122.91
Benzene
0.03
Formaldehyde
95.20
Nickel, PM
4,680.35
Carbon Tetrachloride
0.01
Ethylbenzene
64.14
Benzene
4,357.53
Ethylbenzene
<0.01
Acetaldehyde
55.11
Naphthalene
3,506.47
Hexachloro-1,3 -butadiene
<0.01
Methyl isobutyl ketone
32.86
Xylenes
2,495.17
1,2-Dichloroethane
<0.01
Hydrochloric acid
29.17
Arsenic, PM
2,139.62
1.3 -Butadiene
20.89
Cadmium, PM
1,996.78
Chester, New Jersey (Morris County) - CHNJ
Toluene
528.02
Acrolein
251,595.35
Formaldehyde
0.21
Xylenes
342.26
1,3-Butadiene
12,707.28
Acetaldehyde
0.14
Hexane
314.43
Formaldehyde
9,751.59
1,3-Butadiene
0.03
Benzene
161.55
Acetaldehyde
6,515.84
Benzene
0.02
Formaldehyde
95.57
Benzene
5,385.08
Carbon Tetrachloride
0.01
Ethylbenzene
86.05
Nickel, PM
4,561.47
1,2-Dichloroethane
<0.01
Ethylene glycol
81.36
Xylenes
3,422.56
Acetaldehyde
58.64
Naphthalene
3,363.00
Methyl isobutyl ketone
44.35
Lead, PM
2,402.90
Methanol
38.69
Arsenic, PM
2,017.68
-------
Table 16-7. 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)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Elizabeth, New Jersey (Union County) - ELNJ
Toluene
482.50
Acrolein
306,476.99
Formaldehyde
0.45
Hexane
351.57
Cyanide Compounds, PM
37,500.01
Acetaldehyde
0.31
Xylenes
279.98
1,3-Butadiene
10,478.70
1,3-Butadiene
0.06
Benzene
138.53
Formaldehyde
10,167.13
Benzene
0.03
Formaldehyde
99.64
Nickel, PM
9,894.58
Carbon Tetrachloride
0.01
Ethylbenzene
74.03
Acetaldehyde
6,588.93
Ethylbenzene
<0.01
Acetaldehyde
59.30
Benzene
4,617.66
Hexachloro-1,3 -butadiene
<0.01
Ethylene glycol
45.18
Chlorine
4,370.00
1,2-Dichloroethane
<0.01
Methyl isobutyl ketone
44.98
Naphthalene
3,678.63
Cyanide Compounds, PM
30.00
Lead, PM
3,167.18
North Brunswick, New Jersey (Middlesex County) - NBNJ
Toluene
721.66
Acrolein
424,778.44
1,3-Butadiene
0.03
Hexane
499.90
1,3-Butadiene
15,980.24
Benzene
0.02
Xylenes
432.41
Formaldehyde
14,232.86
Carbon Tetrachloride
0.01
Benzene
213.63
Acetaldehyde
9,314.11
Hexachloro-1,3 -butadiene
<0.01
Formaldehyde
139.48
Benzene
7,120.91
Ethylbenzene
<0.01
Ethylbenzene
110.60
Naphthalene
5,317.45
1,2-Dichloroethane
<0.01
Acetaldehyde
83.83
Lead, PM
5,099.62
Methyl isobutyl ketone
58.79
Titanium tetrachloride
4,535.00
Ethylene glycol
35.26
Xylenes
4,324.07
1.3 -Butadiene
31.96
Arsenic, PM
2,886.75
-------
Observations from Table 16-6 include the following:
• Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
cancer UREs in all four New Jersey counties with NMP sites.
• Formaldehyde, benzene, and 1,3-butadiene are the pollutants with the highest
toxicity-weighted emissions (of the pollutants with cancer UREs) for all four New
Jersey counties, although the order is different for Morris County (CHNJ).
• Six of the 10 highest emitted pollutants in Union and Middlesex Counties also have
the highest toxicity-weighted emissions. Seven of the highest emitted pollutants in
Camden and Morris Counties also have the highest toxicity-weighted emissions.
• Formaldehyde, benzene, ethylbenzene, and 1,3-butadiene are among the pollutants
with the highest cancer risk approximations for CSNJ and also appear on both
emissions-based lists. Acetaldehyde is also among the pollutants with the highest
cancer risk approximations for CSNJ; this pollutant appears among the highest
emitted pollutants in Camden County but does not appear among those with the
highest toxicity-weighted emissions. These observations are also true for ELNJ.
• Formaldehyde, benzene, and 1,3-butadiene are among the pollutants with the highest
cancer risk approximations for CHNJ and also appear on both emissions-based lists.
Acetaldehyde is also among the pollutants with the highest cancer risk
approximations for CHNJ; this pollutant appears among the highest emitted pollutants
in Morris County but does not appear among those with the highest toxicity-weighted
emissions.
• Benzene, ethylbenzene, and 1,3-butadiene are among the pollutants with the highest
cancer risk approximations for NBNJ and also appear on both emissions-based lists.
• Carbon tetrachloride, 1,2-dichloroethane, and hexachloro-1,3-butadiene are additional
pollutants of interest for the New Jersey sites. These pollutants do not appear on
either emissions-based list for any of the four counties.
• Arsenic, nickel, and several POM Groups appear among the pollutants with the
highest toxicity-weighted emissions for the New Jersey counties with NMP sites.
Neither speciated metals nor PAHs were sampled for at these sites under the NMP.
Observations from Table 16-7 include the following:
• Toluene, hexane, and xylenes are the highest emitted pollutants with noncancer RfCs
in Camden, Union, and Middlesex Counties. In Morris County (CHNJ), toluene is
also the highest emitted pollutant, but the xylenes emissions are greater than the
hexane emissions.
• 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
16-75
-------
11th and 17th for these counties). Although acrolein was sampled for at all four 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. 1,3-Butadiene and
formaldehyde 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 1,3-butadiene and formaldehyde 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, acetaldehyde, 1,3-butadiene, and benzene are pollutants of interest for
CSNJ and each appears on both emissions-based lists for Camden County.
Ethylbenzene is another pollutant of interest for CSNJ and appears among the highest
emitted in Camden County but is not among those with the highest toxicity-weighted
emissions.
Formaldehyde, acetaldehyde, and benzene are pollutants of interest for CHNJ that
appear on both emissions-based lists for Morris County. 1,3-Butadiene is another
pollutant of interest for CHNJ and ranks second among those with the highest
toxicity-weighted emissions, but is not among the highest emitted in Morris County.
Formaldehyde, acetaldehyde, and benzene are pollutants of interest for ELNJ that
appear on both emissions-based lists for Union County. 1,3-Butadiene is another
pollutant of interest for ELNJ and ranks third among those with the highest toxicity-
weighted emissions, but is not among the highest emitted in Union County.
Ethylbenzene, also a pollutant of interest for ELNJ, appears among the highest
emitted pollutants (with a noncancer RfC) but is not among those with the highest
toxicity-weighted emissions.
Benzene and 1,3-butadiene are pollutants of interest for NBNJ that appear on both
emissions-based lists for Middlesex County. Ethylbenzene is also a pollutant of
interest for NBNJ and appears among the highest emitted pollutants but is not among
those with the highest toxicity-weighted emissions.
Carbon tetrachloride, 1,2-dichloroethane, and hexachloro-1,3-butadiene are additional
pollutants of interest for the New Jersey sites. These pollutants do not appear on
either emissions-based list for any of the four counties.
Several speciated metals and naphthalene appear among the pollutants with the
highest toxicity-weighted emissions for each New Jersey county with an NMP site.
Neither speciated metals nor PAHs were sampled for at the New Jersey sites under
the NMP.
16-76
-------
16.6 Summary of the 2014 Monitoring Data for the New Jersey Monitoring Sites
Results from several of the data analyses described in this section include the following:
~~~ Concentrations of 14 pollutants failed at least one screen for CSNJ; nine failed
screens for CHNJ; 12 failed screens for ELNJ; and 12 failed screens for NBNJ.
~~~ 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.
~~~ CSNJ has the second highest annual average concentration of formaldehyde and the
sixth highest annual average concentration of acetaldehyde among NMP sites
sampling carbonyl compounds. ELNJ has the third highest annual average
concentrations of both acetaldehyde andformaldehyde among sites sampling these
pollutants.
~~~ 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.
~~~ The detection rates of 1,2-dichloroethane and hexachloro-1,3-butadiene at CHNJ,
ELNJ, and NBNJ have been increasing steadily over the last few years of sampling.
Concentrations of 1,3-butadiene have an increasing trend at CHNJ over recent years.
~~~ Formaldehyde has the highest cancer risk approximation of the pollutants of interest
for CSNJ, CHNJ, and ELNJ; benzene has the highest cancer risk approximation of
the pollutants of interest for NBNJ (among those for which cancer risk
approximations could be calculated). None of the pollutants of interest for these sites
have noncancer hazard approximations greater than an HQ of 1.0.
16-77
-------
17.0 Sites in New York
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS sites in New York, and integrates these concentrations
with emissions, meteorological, and risk information. Data generated by sources other than ERG
are not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) 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 geographical and
physical information about the locations of the sites and the surrounding areas. 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 New York City (BXNY) and one is located
in Rochester (ROCH). Figure 17-1 is a composite satellite image retrieved from ArcGIS
Explorer showing the New York City monitoring site and its immediate surroundings.
Figure 17-2 identifies nearby point source emissions locations by source category, as reported in
the 2011 NEI for point sources, version 2. 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 in order 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.
17-1
-------
Figure 17-1. New York City, New York (BXNY) Monitoring Site
W /;7>h6-5/ -w*
? /7fSs>L/£
/f 7 *,? . r*—# . /., if * l Is. ^S«
'ffflrfs&rJ*'¥
£ff¥^$Fi
I /
At /
V\ /
Ajyy -x w
T . •
i5 £}/
s#- 7 ^
% ./v •»/
fw »>*«r
£> ¦ • J/ v
I##
4L '*»
frM"'?frd,s.
-—4
K>
s -4;
S/5
". t. ¦' " ss^. T^f-, : I ~. /.«¦
r$^ / tnrfz
I x M TrHLS /
L / /IJ «
• -w '.<,- > ^v"> /TyIpSS JBr. Miff Sr.,
jfc^VTX /s^ * ''p\
St Mtrys Park
*¦ ilvrC^s
>,?v
-------
Figure 17-2. NEI Point Sources Located Within 10 Miles of BXNY
73"50'0"W
Hudson
River
1 Westchester
\ County
Long
Island
Sound
O Bronx
County
NEW
JERSEY
' Nassau
\ County
East River
Queens
County
Kings \
County \.
74°5"0"W
74"0"0"W 73"55'0"W 73°50'0"W 73U45,0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
Legend
~ BXNY NATTS site
74"5'0"W
74"0'0"W
i
73"40'0"W
i
O 10 mile radius County boundary
Source Category Group (No. of Facilities)
T Airport/Airline/Airport Support Operations (25)
B Bulk Terminals/Bulk Plants (7)
C Chemical Manufacturing Facility (4)
i Compressor Station (2)
# Electricity Generation via Combustion (18)
E Electroplating, Plating, Polishing, Anodizing, and Coloring (1)
F Food Processing/Agriculture Facility (5)
> Hotels/Motels/Lodging (1)
O Institutional (school, hospital, prison, etc.) (27)
A Metal Coating, Engraving, and Allied Services to Manufacturers (2)
® Metals Processing/Fabrication Facility (4)
X Mine/Quarry/Mineral Processing Facility (1)
? Miscellaneous Commercial/Industrial Facility (33)
• Oil and/or Gas Production (2)
~ Paint and Coating Manufacturing Facility (6)
cd Pharmaceutical Manufacturing (2)
R Plastic, Resin, or Rubber Products Plant (3)
P Printing/Publishing/Paper Product Manufacturing Facility (15)
** Truck/Bus/Transportation Operations (2)
* Vtestewater Treatment Facility (5)
17-3
-------
Figure 17-3. Rochester, Mew York (ROCH) Monitoring Site
-------
Figure 17-4. NET Point Sources Located Within 10 Miles of ROCH
77°35'0"W
77 '25'0"W
Ontario
Wayne
County
\ Monroe
County
Ontario
County
77"50'0"W 77"45'0"W 77"40'0"W 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.
Legend
ROCH NATTS site O 10 mile radius
County boundary
Source Category Group (No. of Facilities)
f
Airport/Airline/Airport Support Operations (6)
¦
Landfill (1)
B
Bulk Terminals/Bulk Plants (4)
©
Metals Processing/Fabrication Facility (2)
C
Chemical Manufacturing Facility (4)
"?
Miscellaneous Commercial/Industrial Facility (3)
e
Electrical Equipment Manufacturing Facility (2)
cz>
Pharmaceutical Manufacturing (1)
s
Glass Plant (1)
R
Plastic, Resin, or Rubber Products Plant (1)
*
Industrial Machinery or Equipment Plant (1)
P
Printing/Publishing/Paper Product Manufacturing Facility (3)
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
98,298
1-278 between 1-87 & 1-895
ROCH
36-055-1007
Rochester
Monroe
Rochester, NY
43.146180,
-77.548170
Residential
Urban/City
Center
85,417
1-490 at 1-590
1AADT reflects 2013 data (NYS DOT, 2013)
BOLD ITALICS = EPA-designated NATTS Site
-------
BXNY is located on the property of Public School 52 (PS 52) in the Bronx Borough of
New York City, 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 less than 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 printing,
publishing, and paper product manufacturing. 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 side 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, and printing, publishing, and paper product manufacturers nearby,
to name a few. The closest source to ROCH is an electrical equipment manufacturer.
17-7
-------
In addition to providing city, county, CBSA, 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 effect concentrations measured at a given monitoring site.
Traffic volume is higher near BXNY than ROCH, which rank 12th and 14th, respectively,
among NMP sites, although both are in the upper third compared to other NMP sites. The traffic
data for BXNY is for 1-278 between 1-87 and 1-895; the traffic data for ROCH are provided for
1-490 at 1-590.
17.2 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
sites in New York on sample days, as well as over the course of the year.
17.2.1 Meteorological Summary
In order to provide an overview of the meteorological conditions experienced at each
monitoring site, hourly meteorological data for 2014 were retrieved and sample day and full-year
averages developed for temperature, dew point, relative humidity, pressure, and wind speed.
Weather data from the actual monitoring site(s) were obtained from AQS, where available. If
site-specific weather data were not available in AQS, then data were obtained from NCDC for
the NWS weather station located closest to the monitoring site(s), as described in Section 3.4.2.
For BXNY, site-specific data were not available; thus, data from the NWS weather station at La
Guardia Airport (WBAN 14732) was used for the parameters in Table 17-2. For ROCH,
temperature, pressure, humidity, and wind information was available in AQS while dew point
temperature and sea level pressure data were obtained from the NWS weather station at Greater
Rochester International Airport (WBAN 14768). Additionally, relative humidity observations for
the first 3 months of 2014 were not available in AQS; thus, relative humidity data from the NWS
station was used as a surrogate for the first part of the year for ROCH. A map showing the
distance between each New York monitoring site and the closest NWS weather station is
provided in Appendix R. These data were used to determine how meteorological conditions on
sample days vary from conditions experienced throughout the year.
17-8
-------
Table 17-2. Average Meteorological Conditions near the New York Monitoring Sites
Average
Type1
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(in Hg)
Average
Station
Pressure
(in Hg)
Prevailing
Wind
Direction
Average
Scalar Wind
Speed
(kt)
New York City,
New York - BXNY2
Sample
Days
53.0
38.0
59.1
30.03
30.00
9.6
(61)
± 1.0
± 1.0
±0.9
±0.01
±0.01
NW
±0.3
54.2
39.0
59.3
30.02
29.99
8.9
2014
±0.4
±0.4
±0.4
± <0.01
±<0.01
NW
±0.1
Rochester, New York - ROCH3
Sample
Days
49.5
36.2
66.1
30.05
29.44
3.5
(63)
± 1.0
± 1.0
±0.9
±0.01
±0.01
SSE
±0.1
50.5
37.3
66.4
30.02
29.40
3.4
2014
±0.4
±0.4
±0.4
± <0.01
±0.01
sw
±<0.1
'Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
2Data for BXNY were obtained from the closest NWS weather station located at La Guardia Airport, WBAN 14732.
3Sea level pressure and dew point temperature observations were not available in AQS for ROCH and were obtained from the
closest NWS weather station located at Greater Rochester International Airport, WBAN 14768. In addition, site-specific
relative humidity data for the first 3 months of 2014 were not available in AQS; thus, NWS data was used as a surrogate where
data were missing.
Table 17-2 presents average temperature, average dew point temperature, average relative
humidity, average station and sea level pressure, and wind information (average scalar wind
speed and prevailing wind direction) for days on which samples were collected and for all of
2014. Also included in Table 17-2 is the 95 percent confidence interval for each parameter. As
shown in Table 17-2, average meteorological conditions on sample days were representative of
average weather conditions experienced throughout the year at each site. Although the difference
between the sample day and full-year average wind speed for BXNY is statistically significant,
these averages are the highest wind speed averages among NMP sites.
17.2.2 Wind Rose Comparison
Hourly surface wind data were also uploaded into a wind rose software program to
produce customized wind roses, as described in Section 3.4.2. A wind rose shows the frequency
of wind directions using "petals" positioned around a 16-point compass, and uses different colors
to represent wind speeds. Figure 17-5 presents two wind roses for the BXNY monitoring site.
The first is a wind rose representing wind observations for all of 2014 and the second is a wind
rose representing wind observations for days on which samples were collected in 2014. These
are used to identify the predominant wind speed and direction for 2014 and to determine if wind
17-9
-------
observations on sample days were representative of conditions experienced over the entire year.
Figure 17-6 presents the full-year and sample day wind roses for ROCH.
Figure 17-5. Wind Roses for the La Guardia Airport Weather Station near BXNY
2014 Wind Rose Sample Day Wind Rose
WIND SPEED
(Knots)
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
H 1
Calms: 6.70%
EAST
WIND SPEED
(Knots)
I I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
H 1
Calms: 5.40%
Observations from Figure 17-5 for BXNY include the following:
• The weather station at La Guardia Airport is located 2.8 miles south-southeast of
BXNY. The East River and Rikers Island separate the site and the weather station.
• The full-year wind rose shows that winds from a variety of directions are observed
near BXNY, although winds from the southeast quadrant were rarely observed.
Winds from the west to northwest to north account for approximately 40 percent of
the wind observations. Winds from the northeast account for another 10 percent of
observations while winds from the south account for nearly 12 percent. Calm winds
were observed for nearly 7 percent of the hourly measurements near BXNY.
• The sample day wind rose shares many similarities with the full-year wind rose, such
as the prominence of winds from the northwest quadrant and the lack of winds from
the southeast quadrant. There are some differences, though. For example, winds from
the northwest account for a higher percentage on wind observations on sample days
while there were fewer southerly wind observations on sample days.
17-10
-------
Figure 17-6. Wind Roses for the Wind Data Collected at ROCH
2014 Wind Rose Sample Day Wind Rose
NORTH
20%
16%
12% \ \
8% V \
WEST
east'
vvy
WIND SPEED
(Knots)
¦ 17-21
¦ 11-17
¦I 7-11
~ 4.7
1-4
Calms: 280%
WIND SPEED
(Knots)
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 2.46%
Observations from Figure 17-6 for ROCH include the following:
• The full-year wind rose shows that winds from the east-southeast to south-southwest
were frequently observed, while winds from the northeast and northwest quadrants
were not observed. Light winds account for the majority of wind observations at this
site, although calm winds were observed for less than 3 percent of the hourly
measurements. Few wind speeds greater than 11 knots were observed at ROCH in
2014.
• The wind patterns shown on the sample day wind rose are similar to the full-year
wind patterns. The calm rate is also similar between the two wind roses. This
indicates that wind observations on sample days were representative of those
observed throughout the year at ROCH.
17-11
-------
17.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each New
York monitoring 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. 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 17-3. 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 17-3. 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.
Table 17-3. Risk-Based Screening Results for the New York Monitoring Sites
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
New York City, New York - BXNY
Naphthalene
0.029
57
57
100.00
74.03
74.03
Acenaphthene
0.011
8
57
14.04
10.39
84.42
Fluorene
0.011
7
53
13.21
9.09
93.51
Benzo(a)pyrene
0.00057
2
57
3.51
2.60
96.10
Fluoranthene
0.011
2
57
3.51
2.60
98.70
Acenaphthylene
0.011
1
44
2.27
1.30
100.00
Total
77
325
23.69
Rochester, New York - ROCH
Naphthalene
0.029
39
57
68.42
45.35
45.35
Acenaphthene
0.011
23
57
40.35
26.74
72.09
Fluorene
0.011
22
50
44.00
25.58
97.67
Fluoranthene
0.011
2
57
3.51
2.33
100.00
Total
86
221
38.91
Observations from Table 17-3 include the following:
• Concentrations of six pollutants failed screens for BXNY; 24 percent of
concentrations for these six pollutants were greater than their associated risk
screening value (or failed screens).
• Five of the six PAHs that failed screens were identified as pollutants of interest for
BXNY. Although the 95 percent criteria is met with benzo(a)pyrene, fluoranthene is
17-12
-------
also considered a pollutant of interest for BXNY because it failed the same number of
screens as benzo(a)pyrene, per the steps described in Section 3.2.
• Concentrations of four pollutants failed screens for ROCH; 39 percent of
concentrations for these four pollutants were greater than their associated risk
screening value (or failed screens).
• Three of these four pollutants contributed to 95 percent of failed screens for ROCH
and therefore were identified as pollutants of interest for this site.
• For both sites, naphthalene, acenaphthene, and fluorene were identified as pollutants
of interest. Naphthalene failed the majority of screens for each site, accounting for
74 percent of failed screens for BXNY and 45 percent of failed screens for ROCH.
Acenaphthene and fluorene together account for 15 failed screens for BXNY and
45 failed screens for ROCH. Thus, the number of failed screens of acenaphthene and
fluorene is three times greater for ROCH than BXNY.
17.4 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.
• 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 are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at BXNY and ROCH are provided in Appendix M.
17.4.1 2014 Concentration Averages
Quarterly and annual concentration averages 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
17-13
-------
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 the entire year of sampling. Annual averages were calculated for
pollutants where three valid quarterly averages could be calculated 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 17-4, 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.
Table 17-4. Quarterly and Annual Average Concentrations of the Pollutants of
Interest for the New York Monitoring Sites
# of
Measured
1st
2nd
3rd
4th
Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs.
# >MDL
# of
Samples
Average
(ng/m3)
Average
(ng/m3)
Average
(ng/m3)
Average
(ng/m3)
Average
(ng/m3)
New York City, New York - BXNY
2.83
7.74
11.98
3.44
6.28
Acenaphthene
57/57
57
± 1.37
±2.25
±3.07
±0.88
± 1.35
0.39
0.11
0.09
0.22
0.21
Benzo(a)pyrene
57/55
57
±0.30
±0.04
±0.02
±0.16
±0.09
3.82
4.40
5.88
3.58
4.38
Fluoranthene
57/57
57
± 1.59
± 1.23
± 1.08
± 1.62
±0.72
3.23
7.73
11.44
3.69
6.32
Fluorene
53/53
57
± 1.50
± 1.99
±2.58
±0.63
± 1.21
116.93
93.45
109.06
84.98
101.09
Naphthalene
57/57
57
±33.58
± 20.27
± 10.84
± 13.41
± 10.72
Rochester, New York - ROCH
1.81
35.57
17.05
18.17
Acenaphthene
57/57
57
±0.95
NA
±23.12
± 13.08
±7.93
1.37
24.58
10.97
12.57
Fluorene
50/50
57
±0.95
NA
± 12.55
±7.68
±4.64
27.42
89.93
51.11
57.53
Naphthalene
57/57
57
±7.54
NA
±35.78
±26.01
± 14.40
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
Observations for BXNY from Table 17-4 include the following:
• Acenaphthene, benzo(a)pyrene, fluoranthene, and naphthalene were detected in all
the valid PAH samples collected at BXNY, while four non-detects of fluorene were
measured.
• Of the pollutants of interest for BXNY, naphthalene has the highest annual average
concentration, benzo(a)pyrene has the lowest, and the annual averages for
acenaphthene and fluorene are similar to each other.
17-14
-------
Concentrations of naphthalene measured at BXNY range from 37.2 ng/m3 to
228 ng/m3. Concentrations of naphthalene appear highest during the first quarter,
although the quarterly averages of naphthalene are not significantly different from
each other. A review of the data shows that the three highest naphthalene
concentrations measured at BXNY were measured on back-to-back sample days in
January, and all three are greater than 200 ng/m3. Concentrations of naphthalene
greater than 100 ng/m3 were measured at BXNY during each calendar quarter,
ranging from two measured during the fourth quarter to nine measured during the
third quarter.
Concentrations of benzo(a)pyrene measured at BXNY span two orders of magnitude,
ranging from 0.0206 ng/m3 to 2.37 ng/m3. The maximum concentration measured at
BXNY is the second highest benzo(a)pyrene concentration measured across the
program. BXNY is one of only two NMP sites with more than one benzo(a)pyrene
concentration greater than 1 ng/m3. The first and fourth quarter average
concentrations are higher than the other quarterly averages and have relatively large
confidence intervals associated with them, particularly the first quarter average. A
review of the data shows that all 10 benzo(a)pyrene concentrations greater than
0.25 ng/m3 measured at BXNY were measured between January and March (eight)
and November (2). The two highest concentrations were measured on
February 22, 2014 (2.37 ng/m3) and November 19, 2014 (1.31 ng/m3).
The annual average and quarterly average concentrations of acenaphthene and
fluorene are similar to each other. Concentrations of acenaphthene range from
0.762 ng/m3 to 23.4 ng/m3 while concentrations of fluorene range from 1.57 ng/m3 to
22.5 ng/m3 plus four non-detects. For both pollutants, the second and third quarter
average concentrations 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 fluorene, all 13 concentrations greater than 10 ng/m3 were
measured between May and September and none of the 17 concentrations less than
3.5 ng/m3 were measured during the second or third quarters of 2014. For
acenaphthene, all but one of the 11 concentrations greater than 10 ng/m3 were
measured between May and September and none of the 18 concentrations less than
3.0 ng/m3 were measured during the second or third quarters.
Concentrations of fluoranthene measured at BXNY range from 1.22 ng/m3 to
14.9 ng/m3. Even though the third quarter average concentration is the highest of the
four, this quarter average exhibits the least variability among the quarterly average
concentrations. The first and fourth quarter averages exhibit the most variability
despite being lower in magnitude compared to the other quarterly averages.
Concentrations measured during the third quarter fall into the smallest range, with
each concentration between 3 ng/m3 and 9 ng/m3. Concentrations measured during
the fourth quarter have largest range, between 1 ng/m3 and 15 ng/m3.
17-15
-------
Observations for ROCH from Table 17-4 include the following:
• Acenaphthene and naphthalene were detected in all 57 valid PAH samples collected
at ROCH, while seven non-detects of fluorene were measured.
• Of the pollutants of interest for ROCH, naphthalene has the highest annual average
concentration, followed by acenaphthene and fluorene.
• Laboratory instrument issues combined with a sampler issue resulted in too many
invalid samples for quarterly average concentrations to be calculated for the second
quarter of 2014. The same laboratory issue also affected BXNY samples, but
quarterly average concentrations for the second quarter could still be calculated.
• Quarterly average concentrations of each of the pollutants of interest are highest for
the third quarter; these quarterly averages also have the largest confidence intervals,
indicating considerable variability.
• Concentrations of naphthalene measured at ROCH range from 10.0 ng/m3 to
227 ng/m3. The maximum naphthalene concentration was measured just at outside the
third quarter on October 2, 2014, although a similar concentration (224 ng/m3) was
measured on the previous sample day, on September 26, 2014. Four of the six
naphthalene concentrations greater than 100 ng/m3 were measured at ROCH during
the third quarter (with the exceptions measured in late June or early October). The
number of naphthalene concentrations greater than 50 ng/m3 measured during the
third quarter (10) is greater than the number measured during the remaining calendar
quarters combined (nine).
• Concentrations of acenaphthene range from 0.548 ng/m3 to 188 ng/m3 and
concentrations of fluorene range from 0.909 ng/m3 to 104 ng/m3 plus the seven non-
detects. The maximum concentration of each of these pollutants was measured on
September 20, 2014 and each is the second highest concentration measured across the
program among NMP sites sampling PAHs. For both pollutants, the next highest
concentration measured at ROCH is considerably less, although concentrations
measured at ROCH are among the highest across the program.
• The first quarter average concentrations of acenaphthene and fluorene are
significantly less than the other quarterly averages shown in Table 17-4 for these
pollutants.
• All six acenaphthene concentrations less than 1 ng/m3 were measured during the first
quarter of 2014. In addition, the number of acenaphthene concentrations less than
5 ng/m3 measured during the first quarter (13) is more than the number measured
during the other calendar quarters combined (10, with none measured during the third
quarter). Conversely, acenaphthene concentrations greater than 25 ng/m3 were
measured during each calendar quarter except the first (with two each measured
during the second and fourth quarters of 2014 and seven measured during the third).
17-16
-------
• Six of the seven non-detects of fluorene were measured during the first quarter of
2014, with the exception measured in June. One fluorene concentration greater than
5 ng/m3 was measured during the first quarter of 2014 while 30 were measured during
the other calendar quarters (eight during the second quarter, 15 during the third, and
seven during the fourth). Fluorene concentrations greater than 25 ng/m3 were
measured during each calendar quarter except the first (with two each measured
during the second and fourth quarters of 2014 and five measured during the third).
Tables 4-9 through 4-12 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:
• Naphthalene is the only PAH pollutant of interest at the program-level. BXNY has
the third highest annual average concentration of naphthalene among NMP sites
sampling PAHs, as shown in Table 4-11, and is one of only four NMP sites with an
annual average naphthalene concentration greater than 100 ng/m3.
• ROCH does not appear in Table 4-11 for naphthalene (it ranks 12th). The annual
average naphthalene concentration for ROCH is approximately half the annual
average concentration for BXNY.
17.4.2 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 each of the pollutants of
interest for BXNY and ROCH. Figures 17-7 through 17-11 overlay the site's minimum, annual
average, and maximum concentrations onto the program-level minimum, first quartile, median,
average, third quartile, and maximum concentrations for each pollutant, as described in
Section 3.4.3.1, and are discussed below.
17-17
-------
Figure 17-7. Program vs. Site-Specific Average Acenaphthene Concentrations
Program Max Concentration = 198 ng/m3
-o-
Program Max Concentration = 198 ng/m3
ROCH Max Concentration = 188 ng/m3
40 60
Concentration {ng/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 17-7 presents the box plots for acenaphthene for BXNY and ROCH and shows the
following:
• The program-level maximum concentration (198 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 to
100 ng/m3. In addition, the maximum acenaphthene concentration measured at ROCH
also exceeds the scale of the box plots and thus, has been denoted directly on
ROCH's box plot.
• The maximum acenaphthene concentration measured at ROCH is eight times greater
than the maximum concentration measured at BXNY and is the second highest
concentration measured across the program.
• The annual average concentrations for both sites are greater than the program-level
average concentration, although the annual average for ROCH is three times greater
the annual average concentration for BXNY.
• ROCH has the second highest annual average concentration of acenaphthene among
NMP sites sampling PAHs (behind only NBIL).
• Although non-detects of acenaphthene were measured at several NMP sites sampling
PAHS, none were measured at BXNY or ROCH.
17-18
-------
Figure 17-8. Program vs. Site-Specific Average Benzo(a)pyrene Concentration
Program Max Concentration = 3.19 ng/m3
1
'
—KJ '
0 0.5 1 1.5 2 2.5
Concentration (ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 17-8 presents the box plot for benzo(a)pyrene for BXNY and shows the following:
• The program-level maximum concentration (3.19 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 been reduced.
• BXNY is one of only two sites for which benzo(a)pyrene is a pollutant of interest
(SJJCA 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 second highest.
• The minimum benzo(a)pyrene concentration measured at BXNY is greater than the
program-level first quartile.
• The annual average concentration for BXNY is nearly twice the program-level
average concentration; this site has the second highest annual average concentration
of benzo(a)pyrene among NMP sites sampling PAHs.
Figure 17-9. Program vs. Site-Specific Average Fluoranthene Concentration
o
KJ
20 25
Concentration {ng/m3)
Program: 1st Quartile
¦
2nd Quartile 3rd Quartile
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
17-19
-------
Figure 17-9 presents the box plot for fluoranthene for BXNY and shows the following:
• BXNY is one of only two sites for which fluoranthene is a pollutant of interest (NBIL
is the other).
• The maximum fluoranthene concentration measured at BXNY is considerably less
than the maximum concentration measured across the program.
• The minimum concentration measured at BXNY is just less than the program-level
median concentration. This means that the minimum fluoranthene concentration
measured at BXNY is greater than half of the concentrations measured across the
program by NMP sites sampling PAHs.
• The annual average concentration for BXNY is nearly twice the program-level
average concentration (2.32 ng/m3).
Figure 17-10. Program vs. Site-Specific Average Fluorene Concentrations
BXNY
r\ ,
Program Max Concentration = 161 ng/m3
ROCH
ROCH Max Concentration = 104 ng/m;
Program Max Concentration = 161 ng/m;
40 60
Concentration {ng/m3)
Program: IstQuartile
¦
2ndQuartile 3rdQuartile
~ ~
4thQuartile
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 17-10 presents the box plots for fluorene for BXNY and ROCH and shows the
following:
• The program-level maximum concentration (161 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 to
100 ng/m3. In addition, the maximum fluorene concentration measured at ROCH also
exceeds the scale of the box plots and thus, has been denoted directly on ROCH's box
plot.
• The maximum fluorene concentration measured at ROCH is nearly four times greater
than the maximum concentration measured at BXNY and is the second highest
concentration measured across the program.
17-20
-------
• The annual average concentrations for both sites are greater than the program-level
average concentration, although the annual average for ROCH is two times greater
the annual average concentration for BXNY.
• ROCH has the second highest annual average concentration of fluorene among NMP
sites sampling PAHs (behind only NBIL).
Figure 17-11. Program vs. Site-Specific Average Naphthalene Concentrations
,
1
¦
0
100
200
300
Concentration {ng/m3)
400
500
Program:
1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site:
Site Average
o
Site Concentration Range
Figure 17-11 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 17-11 shows that the maximum naphthalene
concentrations measured at these sites are similar to each other. Yet, the minimum
concentrations measured at these sites are considerably different. The minimum
concentration measured at BXNY is greater than the program-level first quartile
and the highest minimum concentration measured among NMP sites sampling
naphthalene. A similar observation was made in the 2013 NMP report.
• The annual average naphthalene concentration for ROCH is nearly half the annual
average for BXNY and is less than the program-level average concentration but
greater than the program-level median concentration. The annual average
concentration for BXNY is greater than the program-level average and third
quartile. Recall that BXNY has the third highest annual average concentration of
naphthalene among NMP sites sampling PAHs.
17-21
-------
17.4.3 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 3.4.3.2.
Sampling for PAHs at BXNY began in 2008. However, in June 2010, the monitoring instruments
at BXNY were 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. A trends analysis was not performed for BXNY because
sampling did not occur consecutively at the same location.
Sampling for PAHs at ROCH began in July 2008, so a trends analysis was performed for
ROCH. However, due to the mid-year start, a 1-year average concentration for 2008 is not
presented, although the range of measurements is provided. In addition, 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, the range of measurements is provided
for 2009, although a 1-year average concentration is not provided 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. One-year average
concentrations are provided in Figures 17-12 through 17-14 beginning in 2011.
17-22
-------
Figure 17-12. Yearly Statistical Metrics for Acenaphthene Concentrations Measured at ROCH
_L [
J-
2011
Year
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 17-12 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 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, and in total, three
acenaphthene concentrations measured in 2014 are greater than the maximum
concentration measured in 2013. Despite these higher measurements, the median and
average concentrations changed little.
17-23
-------
Figure 17-13. Yearly Statistical Metrics for Fluorene Concentrations Measured at ROCH
2011
Year
O 5th Perc entile
— 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 17-13 for fluorene concentrations measured at ROCH include
the following:
• The trends graph for fluorene resembles the trends graph for acenaphthene.
• The range of fluorene concentrations measured at ROCH decreased from 2008 to 2009
and the median concentration decreased by more than half during this time frame. Recall,
though, that 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.
The median increased by 67 percent from 2012 to 2013 while the 1-year average
concentration increased by about half that percentage as the range of concentrations
increased from 2012 to 2013 (at both ends of the concentration range).
The maximum concentration of fluorene measured at ROCH was measured in 2014
(104 ng/m3). Only two other concentrations greater than 50 ng/m3 have been measured at
this site, one in 2013 (53.4 ng/m3) and one in 2014 (51.4 ng/m3). Despite the higher
concentrations measured, slight decreases 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). Non-detects have only been measured in 2013 (two) and 2014.
17-24
-------
Figure 17-14. Yearly Statistical Metrics for Naphthalene Concentrations Measured at ROCH
o
2011
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 17-14 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 plot only includes half a year's
worth of samples.
• Even though the maximum concentration has increased each year since 2011, 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, and the 1-year average for 2014
decreased slightly. 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.
17.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at the New York monitoring sites. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4
for definitions and explanations regarding the various toxicity factors, time frames, and
calculations associated with these risk-based screenings.
17-25
-------
17.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the New York 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 3.4.3.3 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 17-5, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Table 17-5. Risk Approximations for the New York Monitoring Sites
Pollutant
Cancer
URE
(Ug/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
New York City, New York - BXNY
Acenaphthene
0.000088
57/57
6.28
± 1.35
0.55
Benzo(a)pyrene
0.00176
57/57
0.21
±0.09
0.37
Fluoranthene
0.000088
57/57
4.38
±0.72
0.39
Fluorene
0.000088
53/57
6.32
± 1.21
0.56
Naphthalene
0.000034
0.003
57/57
101.09
± 10.72
3.44
0.03
Rochester, New York - ROCH
Acenaphthene
0.000088
57/57
18.17
±7.93
1.60
Fluorene
0.000088
50/57
12.57
±4.64
1.11
Naphthalene
0.000034
0.003
57/57
57.53
± 14.40
1.96
0.02
— = A Cancer URE or Noncancer RfC is not available.
17-26
-------
Observations for the New York sites from Table 17-5 include the following:
• Naphthalene has the highest annual average concentration among the pollutants of
interest for each site. Although the annual average concentration for BXNY is nearly
twice the annual average for ROCH, the confidence interval for ROCH's annual
average concentration is larger than the confidence interval for BXNY's annual
average.
• Naphthalene also has the highest cancer risk approximation for each site
(3.44 in-a-million for BXNY and 1.96 in-a-million for ROCH). The cancer risk
approximations for the other pollutants of interest for BXNY are all less than
1 in-a-million. For ROCH, the cancer risk approximations for each pollutant of
interest is between 1 in-a-million and 2 in-a-million.
• Naphthalene is the only site-specific pollutant of interest that has a noncancer RfC.
The noncancer hazard approximations for naphthalene for each site are both less than
0.05, considerably less than 1.0, indicating that no adverse noncancer health effects
are expected from this individual pollutant.
17.5.2 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 17-6 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
that have cancer toxicity factors. Table 17-6 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 17-6 provides the pollutants of interest with the highest cancer risk approximations (in-a-
million) for each site, as presented in Table 17-5. The emissions, toxicity-weighted emissions,
and cancer risk approximations are shown in descending order in Table 17-6. Table 17-7
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 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 17.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
17-27
-------
Table 17-6. 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)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
New York City, New York (Bronx County) - BXNY
Benzene
127.66
Formaldehyde
1.04E-03
Naphthalene
3.44
Ethylbenzene
92.28
Benzene
9.96E-04
Fluorene
0.56
Tetrachloroethylene
81.66
1,3-Butadiene
4.42E-04
Acenaphthene
0.55
Formaldehyde
80.26
Naphthalene
2.76E-04
Fluoranthene
0.39
Acetaldehyde
47.43
Ethylbenzene
2.31E-04
Benzo(a)pyrene
0.37
1.3 -Butadiene
14.74
Arsenic, PM
2.19E-04
Naphthalene
8.11
POM, Group 2b
1.53E-04
POM, Group 2b
1.74
Nickel, PM
1.40E-04
POM, Group 2d
1.53
POM, Group 2d
1.35E-04
Trichloroethylene
1.05
POM, Group 5a
1.11E-04
Rochester, New York (Monroe County) - ROCH
Benzene
257.25
Formaldehyde
2.24E-03
Naphthalene
1.96
Formaldehyde
172.39
Benzene
2.01E-03
Acenaphthene
1.60
Ethylbenzene
140.93
1,3-Butadiene
1.24E-03
Fluorene
1.11
Acetaldehyde
98.59
Naphthalene
6.88E-04
Dichloro methane
46.10
POM, Group 2b
5.08E-04
1,3-Butadiene
41.31
Arsenic, PM
3.81E-04
Tetrachloroethylene
24.16
Ethylbenzene
3.52E-04
Naphthalene
20.23
POM, Group 2d
3.16E-04
Trichloroethylene
6.40
Hexavalent Chromium
2.69E-04
POM, Group 2b
5.77
POM, Group 5a
2.69E-04
-------
Table 17-7. 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 Hazard Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer
Hazard
Approximation
(HQ)
New York City, New York (Bronx County) - BXNY
Toluene
2,161.13
Acrolein
203,787.77
Naphthalene
0.03
Methanol
793.11
Formaldehyde
8,190.17
Hexane
479.04
1,3-Butadiene
7,368.79
Xylenes
293.26
Acetaldehyde
5,270.29
Ethylene glycol
275.15
Benzene
4,255.29
Benzene
127.66
Cadmium, PM
3,946.26
Ethylbenzene
92.28
Arsenic, PM
3,399.31
Tetrachloroethylene
81.66
Nickel, PM
3,238.72
Formaldehyde
80.26
Xylenes
2,932.64
Methyl isobutyl ketone
63.81
Naphthalene
2,702.28
Rochester, New York (Monroe County) - ROCH
Toluene
1,679.94
Acrolein
492,322.38
Naphthalene
0.02
Methanol
510.18
1,3-Butadiene
20,653.74
Xylenes
507.26
Formaldehyde
17,591.01
Hexane
498.21
Acetaldehyde
10,954.81
Benzene
257.25
Hydrochloric acid
10,479.37
Hydrochloric acid
209.59
Cadmium, PM
9,067.59
Formaldehyde
172.39
Benzene
8,575.15
Ethylene glycol
149.53
Naphthalene
6,742.12
Ethylbenzene
140.93
Arsenic, PM
5,913.63
Acetaldehyde
98.59
Nickel, PM
5,849.56
-------
Observations from Table 17-6 include the following:
• Benzene, ethylbenzene, and tetrachloroethylene are the highest emitted pollutants
with cancer UREs in Bronx County while benzene, formaldehyde, and ethylbenzene
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 both New York
counties.
• Seven 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 approximation for each site, appears on both emissions-based lists for Bronx and
Monroe Counties.
• Emissions of several POM Groups rank among the highest emitted pollutants as well
as 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 2d also appears on both emissions-based lists for Bronx
County, although none of the PAHs sampled with Method TO-13A are included in
this group. 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, Groups 2b, 2d, and 5a also appear among the pollutants with the highest
toxicity-weighted emissions for Monroe County while only POM, Group 2b appears
among the highest emitted pollutants for Monroe County.
Observations from Table 17-7 include the following:
• Toluene and methanol are the highest emitted pollutants with noncancer RfCs in both
Bronx and Monroe Counties. The emissions of toluene are considerably higher than
the other pollutants listed for both Bronx and Monroe Counties.
• The pollutant with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) is acrolein for both counties. Formaldehyde and 1,3-butadiene round
out the top three for both counties, although the order varies.
• 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.
17-30
-------
• Naphthalene is the only pollutant of interest for each site for which a noncancer
hazard approximation 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.
17.6 Summary of the 2014 Monitoring Data for BXNY and ROCH
Results from several of the data analyses described in this section include the following:
~~~ Six pollutants failed screens for BXNY, of which five were identified as pollutants of
interest. Four pollutants failed screens for ROCH, of which three 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 concentration for both sites, although
the annual average for BXNY is nearly twice the annual average for ROCH.
~~~ Concentrations of acenaphthene andfluorene for BXNY were highest during the
warmer months of the year.
~~~ BXNY has the third highest annual average concentration of naphthalene among
NMP sites sampling PAHs. Some of the highest concentrations of acenaphthene and
fluorene across the program were measured at ROCH.
~~~ The highest concentrations of all three of ROCH's pollutants of interest, since the
onset of sampling at this site, were measured in 2014
~~~ Naphthalene has the highest cancer risk approximation among the pollutants of
interest for both BXNY and ROCH. None of the pollutants of interest have noncancer
hazard approximations greater than an HQ of 1.0.
17-31
-------
18.0 Sites in Oklahoma
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the UATMP sites in Oklahoma, and integrates these concentrations
with emissions, meteorological, and risk information. Data generated by sources other than ERG
are not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) 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 geographical and
physical information about the locations of the sites and the surrounding areas. 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.
Five monitoring sites are located in Oklahoma. Three sites (TOOK, TMOK, and TROK)
are located in Tulsa, Oklahoma. Another monitoring site is located in Oklahoma City and the
final one is located in Yukon, Oklahoma, just west of Oklahoma City (YUOK).
Figures 18-1 through 18-3 are 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 2011 NEI for
point sources, version 2. 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 in order to emphasize emissions sources within the
boundaries. Figures 18-5 through 18-7 are the composite satellite maps and emissions sources
map for the Oklahoma City sites. Table 18-1 provides supplemental geographical information
such as land use, location setting, and locational coordinates for each site.
18-1
-------
Figure 18-1. Public Works, Tulsa, Oklahoma (TOOK) Monitoring Site
.L22nd.
Woodward Blvd
-E<24th»St-S'
00
Festival Parte
*W 25th*St«S>
-------
Figure 18-2. Fire Station, Tulsa, Oklahoma (TIY1QK) Monitoring Site
-------
Figure 18-3. Riverside, Tulsa, Oklahoma (TROK) Monitoring Site
W-Edison-St
«W,Easton,Ri,
.W Easton PI--3>
Pc\
-v
Easton Sl
»W-Eastbn«St
Beady St
-W Archer.
2W Admiral Blvd
V<" wMvdW
har\esPa9eBW - 1
-------
Figure 18-4. NEI Point Sources Located Within 10 Miles of TMOK, TOOK, and TROK
96°15'0"W
96°5'0"W
County
Tulsa
County
Keystone
Lake
Arkansas
River
*
Aerospace/Aircraft Manufacturing Facility (4)
*
Glass Plant (2)
•
Oil and/or Gas Production (1)
*
Airport/Airline/Airport Support Operations (14)
*
Industrial Machinery or Equipment Plant (5)
~
Paint and Coating Manufacturing Facility (2)
Asphalt Production/Hot Mix Asphalt Plant (1)
O
Institutional (school, hospital, prison, etc.) (1)
»
Landfill (2)
Petroleum Products Manufacturing (1)
Automobile/Truck Manufacturing Facility (1)
1)
A
Petroleum Refinery (2)
nn
Brick. Structural Clay, or Clay Ceramics Plant (1)
Metal Can, Box, and Other Metal Container Manufacturing (2)
R
Plastic, Resin, or Rubber Products Plant (5)
B
Bulk Terminals/Bulk Plants (1)
A
Metal Coating, Engraving, and Allied Services to Manufacturers (2)
7
Portland Cement Manufacturing (1)
C
Chemical Manufacturing Facility (2)
<•>
Metals Processing/Fabrication Facility (2)
*
Printing, Coating & Dyeing of Fabrics Facility (1)
1
Compressor Station (2)
X
Mine/Quarry/Mineral Processing Facility (1)
X
Rail Yard/Rail Line Operations (1)
#
Electricity Generation via Combustion (2)
9
Miscellaneous Commercial/Industrial Facility (2)
Q
M
Railroad Engines/Parts Manufacturing Facility (1)
E
Electroplating, Plating, Polishing, Anodizing, and Coloring (1)
Municipal Waste Combustor (1)
V
Steel Mill (1)
Note: Due to facility density and collocation, the total facilities
Legend displayed may not represent all facilities within the area of interest.
^ TMOK UATMP site ^ TOOK UATMP site ^ TROK UATMP site
0 10 mile radius County boundary
Source Category Group (No. of Facilities)
Creek
County
Rogers
County
18-5
-------
Figure 18-5. Oklahoma City, Oklahoma (OCOK) Monitoring Site
^no-Circle;
Ave •
Rd
E Memorial Rd
E Memorial Rd
rGrey'Fox'Run«
Butternut PI
r k |
- .. - :
Source: NASA, NGA, USGS
3 2008 Microsoft Corp. -
"c,/-Gouniry.Rlace,Rd.
627ft
Woh-PI -jtRiniirvPI
00
-------
Figure 18-6. Yukon, Oklahoma (YUOK) Monitoring Site
oo
-------
Figure 18-7. NEI Point Sources Located Within 10 Miles of OCOK and YUOK
Lake
Hefner
North Canadian
River .
Lake
Overholser
Grady
County
97o50'0"W
Kingfisher
County
Logan
County
Source Category Group (No. of Facilities)
"f Airport/Airline/Airport Support Operations (21)
a Brick, Structural Clay, or Clay Ceramics Plant (2)
B Bulk Terminals/Bulk Plants (1)
C Chemical Manufacturing Facility (1)
1 Compressor Station (7)
f Electricity Generation via Combustion (2)
E Electroplating, Plating, Polishing, Anodizing, and Coloring (1)
F Food Processing/Agriculture Facility (1)
^ Industrial Machinery or Equipment Plant (1)
o Institutional (school, hospital, prison, etc.) (1)
¦ Landfill (3)
(§) Metal Can, Box, and Other Metal Container Manufacturing (1)
? Miscellaneous Commercial/Industrial Facility (2)
• Oil and/or Gas Production (14)
< Pesticide Manufacturing Plant (1)
R Plastic, Resin, or Rubber Products Plant (2)
~ OCOK UATMP site ~ YUOK UATMP site
O 10 mile radius County boundary
Miles
i » '
97°55,0"W 97"50'0"W 97°45'0"W
Legend
Canadian
97°40'0"W 97*35'0"W 97*30'0"W 97°25,0"W 97°20'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
~
Lake
Oklahoma
County
tanley
Drapery, -r.
I Canadian '
County |
I
1
Cleveland
County
18-8
-------
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
65,800
1-244, south of Arkansas River
TMOK
40-143-1127
Tulsa
Tulsa
Tulsa, OK
36.204902,
-95.976537
Residential
Urban/City
Center
4,200
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
53,300
64/51/412, west of 1-244
OCOK
40-109-1037
Oklahoma
City
Oklahoma
Oklahoma City,
OK
35.614131,
-97.475083
Residential
Suburban
52,433
US-77 north of 44 (Turnpike), before
bend
YUOK
40-017-0101
Yukon
Canadian
Oklahoma City,
OK
35.479215,
-97.751503
Commercial
Suburban
41,000
1-40 west of Hwy 4
(east of Exit 132)
1AADT reflects 2014 data (OK DOT, 2014)
00
vo
-------
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 just south of West 25th Street South. Another
refinery is located to the northwest of the site, on the other side of 1-244. 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, while there are no point sources within 2 miles of TMOK. 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 closest to TOOK include two petroleum refineries (including
one directly under the star symbol for TOOK); a rail yard; a municipal waste combustor; a
compressor station; a metal coating, engraving, and allied services to manufacturers facility; an
airport/airport support operation; and a facility generating electricity via combustion. The closest
point source to TROK is a refinery located on the other side of the Arkansas River, according to
18-10
-------
Figure 18-4. However, several industrial facilities are located between the site and river but are
not included in the NEI for point sources.
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 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 18-6. Yukon is a rapidly growing area, with both commercial and residential
development.
Figure 18-7 shows that YUOK is located about 18 miles southwest of OCOK. Most of
the point sources located within 10 miles of these sites are located in the center of Oklahoma
City (south of OCOK and east of YUOK). 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 brick, structural clay, or clay
ceramics. The source closest to YUOK is an oil and gas production facility, although a chemical
manufacturing facility is located roughly the same distance away.
In addition to providing city, county, CBSA, 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 effect 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 OCOK is higher than the traffic near YUOK. The traffic data for four of the five Oklahoma
18-11
-------
sites rank between 16th and 20th highest among NMP sites, while the traffic data for TMOK are
in the bottom third compared to other NMP sites.
18.2 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
sites in Oklahoma on sample days, as well as over the course of the year.
18.2.1 Meteorological Summary
In order to provide an overview of the meteorological conditions experienced at each
monitoring site, hourly meteorological data for 2014 were retrieved and sample day and full-year
averages developed for temperature, dew point, relative humidity, pressure, and wind speed.
Weather data from the actual monitoring site(s) were obtained from AQS, where available. If
site-specific weather data were not available in AQS, then data were obtained from NCDC for
the NWS weather station located closest to the monitoring site(s), as described in Section 3.4.2.
For the Oklahoma sites, site-specific data were available for some, but not all, of the parameters
in Table 18-2. For each site, temperature, humidity, and wind information was available in AQS.
For TOOK, meteorological data were not available for much of August and a few days in
September and November; thus, data from the NWS weather station at Richard Lloyd Jones Jr.
Airport (WBAN 53908) were used for meteorological parameters without data and/or as
surrogates for parameters without complete observation records. The Richard Lloyd Jones Jr.
Airport weather station is located 6.1 miles south of TOOK. For TMOK and TROK, data from
the NWS weather station at Tulsa International Airport (WBAN 13968) were used where
needed; the Tulsa International Airport weather station is located 5.0 miles east of TMOK and
7.8 miles east-northeast of TROK. RUCA. For OCOK and YUOK, data from the NWS weather
station at Wiley Post Airport (WBAN 03954) were used as needed; the weather station at Wiley
Post Airport is located 11.1 miles west-southwest of OCOK and 7.0 miles east-northeast of
YUOK. A map showing the distance between each Oklahoma monitoring site and the closest
NWS weather station is provided in Appendix R. These data were used to determine how
meteorological conditions on sample days vary from conditions experienced throughout the year.
18-12
-------
Table 18-2. Average Meteorological Conditions near the Oklahoma Monitoring Sites
Average
Type1
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(in Hg)
Average
Station
Pressure
(in Hg)
Prevailing
Wind
Direction
Average
Scalar Wind
Speed
(kt)
Public Works, Tulsa, Oklahoma
-TOOK2
Sample
Days
58.4
45.3
57.7
30.07
29.38
5.4
(62)
± 1.0
± 1.1
± 1.0
±0.01
±0.01
SSE
±0.2
58.6
46.4
59.6
30.05
29.36
5.2
2014
±0.4
±0.4
±0.4
± <0.01
±<0.01
SSE
±0.1
Fire Station, Tulsa, Oklahoma -
TMOK3
Sample
Days
57.9
44.1
58.5
30.04
29.32
4.8
(64)
± 1.1
± 1.1
± 1.1
± <0.01
±0.01
S
±0.2
59.2
46.1
60.6
30.01
29.30
4.8
2014
±0.4
±0.4
±0.5
± <0.01
±<0.01
S
±0.1
Riverside, Tulsa,
Oklahoma -
TROK3
Sample
Days
58.5
44.4
59.3
30.05
29.33
1.4
(63)
± 1.1
± 1.1
± 1.1
±0.01
±0.01
ssw
±0.1
59.7
46.1
60.8
30.01
29.30
1.4
2014
±0.4
±0.4
±0.4
± <0.01
±<0.01
ssw
±<0.1
Oklahoma City,
Oklahoma-OCOK4
Sample
Days
59.8
45.1
59.5
30.04
28.66
5.1
(62)
± 1.1
± 1.2
± 1.0
±0.01
±0.01
SSE
±0.2
60.2
46.1
60.7
30.00
28.64
5.0
2014
±0.4
±0.4
±0.4
± <0.01
±<0.01
S
±0.1
Yukon, Oklahoma - YUOK4
Sample
Days
57.3
44.4
57.8
30.05
28.67
7.0
(62)
± 1.1
± 1.2
± 1.0
±0.01
±0.01
S
±0.2
58.3
46.1
59.0
30.00
28.64
6.8
2014
±0.4
±0.4
±0.4
± <0.01
±<0.01
s
±0.1
'Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
2Temperature, humidity, and wind parameters were measured at TOOK. Where data are missing for these three parameters, or
for the parameters for which observations were not collected, data was obtained from the closest NWS weather station located
at Richard Lloyd Jones Jr. Airport, WBAN 53908.
'Temperature, humidity, and wind parameters were also measured at TMOK and TROK. Hie remaining information was
obtained from the closest NWS weather station located at Tulsa International Airport, WBAN 13968.
4Temperature, humidity, and wind parameters were also measured at OCOK and YUOK. The remaining information was
obtained from the closest NWS weather station located at Wiley Post Airport, WBAN 03954.
18-13
-------
Table 18-2 presents average temperature, average dew point temperature, average relative
humidity, average station and sea level pressure, and wind information (average scalar wind
speed and prevailing wind direction) for days on which samples were collected and for all of
2014. Also included in Table 18-2 is the 95 percent confidence interval for each parameter. As
shown in Table 18-2, average meteorological conditions on sample days were generally
representative of average weather conditions experienced throughout the year at each site. The
greatest differences between the sample day and full-year averages for each site were for average
dew point and average relative humidity, particularly for TMOK.
18.2.2 Wind Rose Comparison
Hourly surface wind data were also uploaded into a wind rose software program to
produce customized wind roses, as described in Section 3.4.2. A wind rose shows the frequency
of wind directions using "petals" positioned around a 16-point compass, and uses different colors
to represent wind speeds. Figure 18-8 presents two wind roses for the TOOK monitoring site.
The first is a wind rose representing wind observations for all of 2014 and the second is a wind
rose representing wind observations for days on which samples were collected in 2014. These
are used to identify the predominant wind speed and direction for 2014 and to determine if wind
observations on sample days were representative of conditions experienced over the entire year.
Figures 18-9 through 18-12 present the full-year and sample day wind roses for the remaining
Oklahoma monitoring sites.
18-14
-------
Figure 18-8. Wind Roses for the Wind Data Collected at TOOK
2014 Wind Rose Sample Day Wind Rose
¦ " NORTH NORTH
tW 10%
5%
WEST
\ 2
EAST.
mav 10%
5%
WEST
/
EAST.
WIND SPEED
(Knots)
¦I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 8.04%
WIND SPEED
(Knots)
¦I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 6.92%
Observations from Figure 18-8 for TOOK include the following:
• The full-year wind rose shows that south-southeasterly winds were prevalent at
TOOK in 2014, accounting for more than one-fifth of observations. Winds from the
southeast to south-southwest together accounted for nearly 50 percent of
observations. Winds from the north-northwest to north-northeast make up a
secondary wind grouping. Winds from the east and west were rarely observed. Calm
winds accounted for 8 percent of wind speed observations in 2014 while stronger
winds were most often observed with south-southeasterly to southerly winds.
• The sample day wind rose for TOOK resembles the full-year wind rose, with winds
from the southeast, south-southeast, and south accounting for the majority of
observations on sample days.
• Recall from the previous section that wind data were not available at TOOK for a
portion of 2014 (primarily during a 2-week stretch in August due to building/site
repair, as indicated in AQS) and NWS data were used as a surrogate for missing data.
18-15
-------
Figure 18-9. Wind Roses for the Wind Data Collected at TMOK
2014 Wind Rose Sample Day Wind Rose
NORTH
NORTH
WEST
WEST
WIND SPEED
(Knots)
WIND SPEED
(Knots)
SOUTH
SOUTH
Calms: 12.45%
Calms: 14.11%
Observations from Figure 18-9 for TMOK include the following:
• The 2014 wind rose shows that winds from the south were prevalent on sample days
at TMOK. Southerly winds accounted for more than 20 percent of observations in
2014, with winds from the south-southeast to south-southwest together accounting for
more than 40 percent of observations. Winds from the north were observed for more
than 10 percent of observations, with winds from the north-northwest to north-
northeast together accounting for another one-quarter of observations. Winds from
the east and west were rarely observed. Calm winds accounted for more than
12 percent of observations at TMOK in 2014.
• The sample day wind rose for TMOK resembles the full-year wind rose, with winds
from the south accounting for the highest percentage of wind observations on sample
days.
18-16
-------
Figure 18-10. Wind Roses for the Wind Data Collected at TROK
2014 Wind Rose Sample Day Wind Rose
NORTH
NORTH
WEST
;WEST
WIND SPEED
(Knots)
WIND SPEED
(Knots)
SOUTH
SOUTH
Calms: 27.75%
Calms: 26.28%
Observations from Figure 18-10 for TROK include the following:
• TROK's wind roses show that winds were considerably lighter near TROK than at
the other Tulsa sites. (This can also be seen in Table 18-2 in the wind speed column.)
The largest percentage of wind speed observations fall into the 1 knot to 4 knots
range, with very few observations exceeding 7 knots. Calm winds account for greater
than one-quarter of observations, both on sample days and throughout the year.
• The 2014 wind rose shows that winds from the south-southwest were the most
commonly observed, followed by southwesterly winds, north-northeasterly winds,
and northerly winds.
• South-southwesterly winds were also prevalent on sample days, while southerly and
southwesterly winds accounted for a similar percentage of observations. Northerly
winds accounted for a slightly higher percentage of winds on sample days than winds
from the north-northeast (while the opposite is true for the full year wind rose).
18-17
-------
Figure 18-11. Wind Roses for the Wind Data Collected at OCOK
2014 Wind Rose Sample Day Wind Rose
¦ " NORTH - . .. NORTH
*2
WEST
" 8%
EAST
WEST
WIND SPEED
(Knots)
¦I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 3.79%
WIND SPEED
(Knots)
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 3.08%
Observations from Figure 18-11 for OCOK include the following:
• The 2014 wind rose shows that winds from the southeast to south-southwest account
for nearly 50 percent of observations at OCOK, with southerly winds observed the
most. Winds from the north were the next most frequently observed wind direction,
with all other wind directions accounting for fewer than 6 percent of observations.
Calm winds were observed for less than 4 percent of observations while the strongest
winds were most often observed with winds with a northerly component.
• The sample day wind rose for OCOK shows that winds from the southeast to south-
southwest still account for the majority of observations, although south-southeasterly
winds were observed the most often on sample days. Winds from the north-northwest
to east-southeast are fairly evenly distributed on the sample day wind rose.
18-18
-------
Figure 18-12. Wind Roses for the Wind Data Collected at YUOK
2014 Wind Rose Sample Day Wind Rose
WEST
'SOUTH
NORTH
NORTH
;WEST
WIND SPEED
(Knots)
WIND SPEED
(Knots)
SOUTH
Calms: 0.65%
Calms: 0.67%
Observations from Figure 18-12 for YUOK include the following:
• The 2014 wind rose shows that southerly winds were prevalent at YUOK, with winds
from the southeast to south-southwest together accounting for more than 40 percent
of observations. Winds from the north and north-northeast also accounted for nearly
20 percent of observations in 2014. Less than 1 percent of wind observations for 2014
are classified as calm, the fewest of any Oklahoma monitoring site, while the
strongest winds were most often observed with south-southwesterly winds.
• Winds from the southeast to south-southwest account for the majority of observations
on sample days as well, although slightly fewer southerly and south-southwesterly
winds are offset by additional southeasterly and south-southeasterly winds compared
to the full-year wind rose. Winds from the north and north-northeast accounted for a
similar percentage of observations on sample days. Cam winds accounted for a
similar percentage of observations on sample days, and winds greater than 17 knots
were most often observed with south-southwesterly winds at YUOK.
18-19
-------
18.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each
Oklahoma monitoring 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. 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 18-3. 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 18-3. 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
each Oklahoma monitoring site.
Table 18-3. Risk-Based Screening Results for the Oklahoma Monitoring Sites
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Public Works, Tulsa, Oklahoma - TOOK
Arsenic (TSP)
0.00023
62
62
100.00
11.55
11.55
Acetaldehyde
0.45
61
61
100.00
11.36
22.91
Benzene
0.13
61
61
100.00
11.36
34.26
Carbon Tetrachloride
0.17
61
61
100.00
11.36
45.62
Formaldehyde
0.077
61
61
100.00
11.36
56.98
1.3 -Butadiene
0.03
57
59
96.61
10.61
67.60
1,2-Dichloroethane
0.038
50
50
100.00
9.31
76.91
/?-Dichlorobcnzcnc
0.091
24
51
47.06
4.47
81.38
Ethylbenzene
0.4
24
61
39.34
4.47
85.85
Nickel (TSP)
0.0021
24
62
38.71
4.47
90.32
Manganese (TSP)
0.03
23
62
37.10
4.28
94.60
Hexachloro -1,3 -butadiene
0.045
18
19
94.74
3.35
97.95
1,2-Dibromoethane
0.0017
3
3
100.00
0.56
98.51
Lead (TSP)
0.015
3
62
4.84
0.56
99.07
Propionaldehyde
0.8
3
61
4.92
0.56
99.63
Cadmium (TSP)
0.00056
2
62
3.23
0.37
100.00
Total
537
858
62.59
18-20
-------
Table 18-3. Risk-Based Screening Results for the Oklahoma Monitoring Sites (Continued)
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Fire Station, Tulsa, Oklahoma - TMOK
Acetaldehyde
0.45
62
62
100.00
12.86
12.86
Benzene
0.13
62
62
100.00
12.86
25.73
Formaldehyde
0.077
62
62
100.00
12.86
38.59
Carbon Tetrachloride
0.17
61
62
98.39
12.66
51.24
Arsenic (TSP)
0.00023
56
58
96.55
11.62
62.86
1.3 -Butadiene
0.03
56
58
96.55
11.62
74.48
1,2-Dichloroethane
0.038
54
54
100.00
11.20
85.68
/?-Dichlorobcnzcnc
0.091
19
49
38.78
3.94
89.63
Ethylbenzene
0.4
18
62
29.03
3.73
93.36
Hexacliloro -1,3 -butadiene
0.045
17
17
100.00
3.53
96.89
Nickel (TSP)
0.0021
9
58
15.52
1.87
98.76
Cadmium (TSP)
0.00056
2
58
3.45
0.41
99.17
Manganese (TSP)
0.03
2
58
3.45
0.41
99.59
1,2-Dibromoethane
0.0017
1
1
100.00
0.21
99.79
Propionaldehyde
0.8
1
62
1.61
0.21
100.00
Total
482
783
61.56
Riverside, Tulsa, Oklahoma - TROK
Acetaldehyde
0.45
61
61
100.00
12.47
12.47
Benzene
0.13
61
61
100.00
12.47
24.95
Carbon Tetrachloride
0.17
61
61
100.00
12.47
37.42
Formaldehyde
0.077
61
61
100.00
12.47
49.90
Arsenic (TSP)
0.00023
59
59
100.00
12.07
61.96
1.3 -Butadiene
0.03
59
60
98.33
12.07
74.03
1,2-Dichloroethane
0.038
50
50
100.00
10.22
84.25
Ethylbenzene
0.4
24
61
39.34
4.91
89.16
Hexacliloro -1,3 -butadiene
0.045
17
18
94.44
3.48
92.64
/?-Dichlorobcnzcnc
0.091
12
47
25.53
2.45
95.09
Nickel (TSP)
0.0021
11
59
18.64
2.25
97.34
Manganese (TSP)
0.03
8
59
13.56
1.64
98.98
Cadmium (TSP)
0.00056
3
59
5.08
0.61
99.59
1,2-Dibromoethane
0.0017
2
2
100.00
0.41
100.00
Total
489
718
68.11
18-21
-------
Table 18-3. Risk-Based Screening Results for the Oklahoma Monitoring Sites (Continued)
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Oklahoma City, Oklahoma - OCOK
Acetaldehyde
0.45
60
60
100.00
15.00
15.00
Benzene
0.13
60
60
100.00
15.00
30.00
Formaldehyde
0.077
60
60
100.00
15.00
45.00
Carbon Tetrachloride
0.17
59
60
98.33
14.75
59.75
Arsenic (TSP)
0.00023
53
59
89.83
13.25
73.00
1,2-Dichloroethane
0.038
53
53
100.00
13.25
86.25
1.3 -Butadiene
0.03
38
46
82.61
9.50
95.75
Hexacliloro -1,3 -butadiene
0.045
10
11
90.91
2.50
98.25
T richloroethylene
0.2
2
6
33.33
0.50
98.75
Cadmium (TSP)
0.00056
1
59
1.69
0.25
99.00
1,2-Dibromoethane
0.0017
1
1
100.00
0.25
99.25
/?-Dichlorobcnzcnc
0.091
1
18
5.56
0.25
99.50
Ethylbenzene
0.4
1
60
1.67
0.25
99.75
Nickel (TSP)
0.0021
1
59
1.69
0.25
100.00
Total
400
612
65.36
Yukon, Oklahoma - YUOK
Acetaldehyde
0.45
61
61
100.00
14.49
14.49
Benzene
0.13
61
61
100.00
14.49
28.98
Carbon Tetrachloride
0.17
61
61
100.00
14.49
43.47
Formaldehyde
0.077
61
61
100.00
14.49
57.96
1,2-Dichloroethane
0.038
55
55
100.00
13.06
71.02
Arsenic (TSP)
0.00023
49
61
80.33
11.64
82.66
1.3 -Butadiene
0.03
44
52
84.62
10.45
93.11
Hexacliloro -1,3 -butadiene
0.045
14
15
93.33
3.33
96.44
Manganese (TSP)
0.03
9
61
14.75
2.14
98.57
/?-Dichlorobcnzcnc
0.091
2
20
10.00
0.48
99.05
Propionaldehyde
0.8
2
61
3.28
0.48
99.52
Ethylbenzene
0.4
1
61
1.64
0.24
99.76
Nickel (TSP)
0.0021
1
61
1.64
0.24
100.00
Total
421
691
60.93
Observations from Table 18-3 include the following:
• Concentrations of 16 pollutants failed at least one screen for TOOK; nearly
63 percent of concentrations for these 16 pollutants were greater than their associated
risk screening value (or failed screens).
• Concentrations of 12 pollutants contributed to 95 percent of failed screens for TOOK
and therefore were identified as pollutants of interest for this site. These 12 include
18-22
-------
two carbonyl compounds, seven 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 15 pollutants failed at least one screen for TMOK; nearly
62 percent of concentrations for these 15 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 14 pollutants failed at least one screen for TROK; 68 percent of
concentrations for these 14 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, seven VOCs, and one TSP metal.
Concentrations of 14 pollutants failed at least one screen for OCOK; 65 percent of
concentrations for these 14 pollutants were greater than their associated risk screening
value (or failed screens).
Concentrations of seven pollutants contributed to 95 percent of failed screens for
OCOK and therefore were identified as pollutants of interest for this site. These seven
include two carbonyl compounds, four VOCs, and one TSP metal.
Concentrations of 13 pollutants failed at least one screen for YUOK; nearly
61 percent of concentrations for these 13 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.
The number of pollutants identified as pollutants of interest range from seven to 12
among the Oklahoma sites. The Tulsa sites have 10 pollutants of interest in common:
acetaldehyde and formaldehyde, arsenic, and seven VOCs. The only differences are
for TOOK, which has two additional TSP metals (manganese and nickel) compared
to the other two sites. The Oklahoma City sites have seven pollutants of interest in
common: acetaldehyde, arsenic, benzene, 1,3-butadiene, carbon tetrachloride,
1,2-dichloroethane, and formaldehyde.
Concentrations measured at TOOK failed the third highest number of screens among
NMP sites, with other Oklahoma sites ranking seventh (TMOK), eighth (TROK),
11th (YUOK), and 15th (OCOK) as shown in Table 4-8.
18-23
-------
18.4 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.
• 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 are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at the Oklahoma sites are provided in Appendices J, L, and N.
18.4.1 2014 Concentration Averages
Quarterly and annual concentration averages 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 the entire year of sampling. Annual averages were calculated for
pollutants where three valid quarterly averages could be calculated 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 18-4, 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.
18-24
-------
Table 18-4. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Oklahoma Monitoring Sites
# of
Measured
1st
2nd
3rd
4th
Detection
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
s vs.
# >MDL
# of
Samples
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Public Works, Tulsa, Oklahoma - TOOK
1.47
2.07
2.84
1.52
1.97
Acetaldehyde
61/61
61
±0.24
±0.28
±0.50
±0.37
±0.22
0.97
0.89
1.22
1.05
1.03
Benzene
61/61
61
±0.16
±0.17
±0.20
±0.30
±0.11
0.08
0.06
0.09
0.08
0.08
1.3 -Butadiene
59/57
61
±0.02
±0.01
±0.02
±0.03
±0.01
0.60
0.66
0.63
0.55
0.61
Carbon Tetrachloride
61/61
61
±0.05
±0.02
±0.03
±0.07
±0.03
0.04
0.07
0.12
0.07
0.07
/?-Dichlorobcnzcnc
51/25
61
±0.01
±0.03
±0.02
±0.02
±0.01
0.08
0.08
0.11
0.08
0.09
1,2-Dichloroethane
50/50
61
±0.02
±0.02
±0.02
±0.04
±0.01
0.35
0.34
0.58
0.31
0.39
Ethylbenzene
61/61
61
±0.09
±0.10
±0.16
±0.09
±0.06
1.97
3.33
4.85
1.71
2.95
Formaldehyde
61/61
61
±0.32
±0.51
±0.89
±0.47
±0.42
0.04
0.02
0.02
0.02
0.03
Hexachloro-1,3 -butadiene
19/0
61
±0.02
±0.02
±0.02
±0.02
±0.01
0.55
0.83
0.97
0.69
0.76
Arsenic (TSP)
62/62
62
±0.09
±0.11
±0.20
±0.19
±0.08
24.67
31.49
26.48
17.60
25.04
Manganese (TSP)
62/62
62
±5.41
±5.28
±6.85
±4.65
±2.92
2.60
2.36
2.15
1.89
2.25
Nickel (TSP)
62/62
62
± 1.16
±0.45
±0.38
±0.32
±0.31
Fire Station, Tulsa, Oklahoma - TMOK
1.56
2.00
2.22
1.49
1.81
Acetaldehyde
62/62
62
±0.27
±0.33
±0.39
±0.31
±0.17
0.89
0.67
0.91
0.77
0.81
Benzene
62/62
62
±0.20
±0.18
±0.17
±0.13
±0.08
0.10
0.06
0.12
0.09
0.09
1.3 -Butadiene
58/57
62
±0.04
±0.02
±0.02
±0.03
±0.01
0.57
0.67
0.65
0.58
0.62
Carbon Tetrachloride
62/61
62
±0.04
±0.03
±0.03
±0.08
±0.03
0.08
0.05
0.09
0.05
0.07
/?-Dichlorobcnzcnc
49/19
62
±0.02
±0.03
±0.02
±0.03
±0.01
0.08
0.09
0.08
0.09
0.08
1,2-Dichloroethane
54/53
62
±0.03
±0.02
±0.02
±0.02
±0.01
0.37
0.27
0.46
0.28
0.34
Ethylbenzene
62/62
62
±0.12
±0.07
±0.08
±0.08
±0.05
3.27
4.37
4.22
1.90
3.41
Formaldehyde
62/62
62
±0.47
±0.73
±0.85
±0.44
±0.39
0.04
0.01
0.02
0.01
0.02
Hexachloro-1,3 -butadiene
17/0
62
±0.02
±0.02
±0.02
±0.02
±0.01
0.47
0.69
0.85
0.65
0.67
Arsenic (TSP)a
58/58
58
±0.13
±0.12
±0.20
±0.23
±0.09
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
18-25
-------
Table 18-4. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Oklahoma Monitoring Sites (Continued)
# of
Measured
1st
2nd
3rd
4th
Detection
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
s vs.
# >MDL
# of
Samples
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Riverside, Tulsa, Oklahoma - TROK
1.34
1.82
2.24
1.54
1.73
Acetaldehyde
61/61
61
±0.26
±0.29
±0.30
±0.33
±0.16
0.77
0.73
0.85
0.85
0.80
Benzene
61/61
61
±0.09
±0.10
±0.13
±0.21
±0.07
0.07
0.06
0.09
0.08
0.07
1.3 -Butadiene
60/59
61
±0.01
±0.02
±0.02
±0.02
±0.01
0.60
0.66
0.64
0.58
0.62
Carbon Tetrachloride
61/61
61
±0.04
±0.02
±0.04
±0.03
±0.02
0.05
0.05
0.08
0.05
0.06
/?-Dichlorobcnzcnc
47/12
61
±0.02
±0.02
±0.02
±0.02
±0.01
0.08
0.08
0.09
0.07
0.08
1,2-Dichloroethane
50/50
61
±0.02
±0.02
±0.02
±0.03
±0.01
0.24
0.38
0.51
0.34
0.37
Ethylbenzene
61/61
61
±0.05
±0.11
±0.10
±0.08
±0.05
2.20
3.31
4.07
1.63
2.78
Formaldehyde
61/61
61
±0.44
±0.55
±0.77
±0.45
±0.36
0.03
0.02
0.02
0.02
0.02
Hexachloro-1,3 -butadiene
18/0
61
±0.02
±0.02
±0.02
±0.02
±0.01
0.59
0.69
1.04
0.77
0.77
Arsenic (TSP)a
59/59
59
±0.11
±0.16
±0.24
±0.33
±0.12
Oklahoma City, Oklahoma - OCOK
1.49
2.16
1.86
1.45
1.73
Acetaldehyde
60/60
60
±0.28
±0.57
±0.34
±0.31
±0.20
0.62
0.51
0.81
0.65
0.65
Benzene
60/60
60
±0.09
±0.14
±0.11
±0.11
±0.06
0.03
0.03
0.05
0.04
0.04
1.3 -Butadiene
46/38
60
±0.02
±0.02
±0.01
±0.01
±0.01
0.54
0.65
0.65
0.62
0.61
Carbon Tetrachloride
60/59
60
±0.10
±0.03
±0.02
±0.03
±0.03
0.08
0.07
0.05
0.08
0.07
1,2-Dichloroethane
53/49
60
±0.01
±0.01
±0.02
±0.01
±0.01
1.33
3.26
4.08
1.98
2.63
Formaldehyde
60/60
60
±0.26
±0.79
±0.64
±0.40
±0.38
0.38
0.54
0.63
0.41
0.48
Arsenic (TSP)a
59/59
59
±0.08
±0.10
±0.15
±0.11
±0.06
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
18-26
-------
Table 18-4. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Oklahoma Monitoring Sites (Continued)
# of
Measured
1st
2nd
3rd
4th
Detection
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
s vs.
# >MDL
# of
Samples
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Yukon, Oklahoma - YUOK
2.03
1.86
1.95
1.39
1.80
Acetaldehyde
61/61
61
±0.94
±0.46
±0.26
±0.30
±0.27
0.57
0.53
0.56
0.55
0.56
Benzene
61/61
61
±0.08
±0.13
±0.08
±0.08
±0.04
0.03
0.04
0.05
0.05
0.05
1,3-Butadiene
52/44
61
±0.02
±0.02
±0.01
±0.02
±0.01
0.59
0.68
0.66
0.61
0.63
Carbon Tetrachloride
61/61
61
±0.04
±0.03
±0.02
±0.03
±0.02
0.07
0.09
0.07
0.08
0.08
1,2-Dichloroethane
55/54
61
±0.02
±0.01
±0.01
±0.02
±0.01
2.37
3.37
4.20
1.83
2.92
Formaldehyde
61/61
61
±0.62
±0.64
±0.63
±0.43
±0.36
0.02
0.03
0.01
0.02
Hexachloro-1,3 -butadiene
15/0
61
±0.02
0
±0.02
±0.02
±0.01
0.32
0.53
0.56
0.37
0.44
Arsenic (TSP)a
61/61
61
±0.08
±0.06
±0.09
±0.10
±0.05
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
Observations for the Oklahoma sites from Table 18-4 include the following:
• Formaldehyde has the highest annual average concentration of the pollutants of
interest for each site, followed by acetaldehyde and benzene (with one exception).
With the exception of TOOK, acetaldehyde and formaldehyde were the only
pollutants of interest with annual average concentrations greater than 1 |ig/m3 for
each site. For TOOK, benzene also has an annual average concentration greater than
1 |ig/m3.
• Annual average concentrations of formaldehyde range from 2.63 ± 0.38 |ig/m3 for
OCOK to 3.41 ± 0.39 |ig/m3 for TMOK. The annual average concentrations of
acetaldehyde range from 1.73 ± 0.16 |ig/m3 for TROK to 1.97 ± 0.22 |ig/m3 for
TOOK. The annual average concentrations of benzene range from 0.56 ± 0.04 |ig/m3
for YUOK to 1.03 ± 0.11 |ig/m3 for TOOK. TOOK has had the highest annual
average benzene concentration among the Oklahoma sites for several years (and is
usually one of the highest across the program), including 2014, although the
difference is becoming less significant.
• Concentrations of the carbonyl compounds, formaldehyde in particular, tended to be
highest in the warmer months and lowest in the cooler months.
• The rest of this section discusses concentrations measured at the individual
monitoring sites in Oklahoma.
18-27
-------
Observations for TOOK from Table 18-4 include the following:
• Formaldehyde concentrations measured at TOOK were highest during the warmer
months of the year, as indicated by the quarterly average concentrations. All 18
formaldehyde concentrations greater than 3.5 |ig/m3 were measured at TOOK
between May and September, with the six measurements between 5 |ig/m3 and
8 |ig/m3 measured in July and August. Conversely, all but one of the 13 formaldehyde
concentrations less than 1.5 |ig/m3 were measured in January, February, November or
December (including all six sample days in December).
• While the highest acetaldehyde concentrations were also measured at TOOK during
July and August, there is more variability during the seasons in which the higher
concentrations were measured. For instance, 13 acetaldehyde concentrations greater
than 2.5 |ig/m3 were measured at TOOK in 2014: three were measured during the
second quarter, eight were measured during the third quarter, and two were measured
during the fourth quarter.
• Concentrations greater than 1 |ig/m3 account for just over half (32) of the benzene
concentrations measured at TOOK in 2014 and only two NMP sites have more
(PACO and ROIL). These measurements were spread throughout the year, with nine
measured during the first quarter of 2014, six during the second quarter, 11 during the
third quarter, and six during the fourth quarter (including the maximum concentration
(2.84 |ig/m3), the only benzene concentration greater than 2 |ig/m3 measured at
TOOK).
• Three ethylbenzene concentrations greater than 1 |ig/m3 were measured at TOOK in
2014, one each in July, August, and September, which is reflected in the third quarter
average concentrations shown in Table 18-4.
• The third quarter average concentration for p-dichlorobenzene is greater than the
other quarterly averages for TOOK. None of the 10 non-detects of this pollutant were
measured during the third quarter. In addition, 13 of the 15 concentrations measured
during the third quarter are greater than 0.1 |ig/m3, while the number for the
remaining calendar quarters ranges from zero (first quarter) to six (second quarter).
• Concentrations of arsenic measured at TOOK appear highest during the third quarter
and lowest during the first quarter, although the confidence intervals shown for the
third and the fourth quarters are similar to each other and twice those shown for the
other quarterly averages. The variability exhibited by the concentrations measured
during each quarter can be evaluated by comparing the differences between the
quarterly average and median concentration for each quarter. The differences between
the quarterly average and median concentration for the first, second, and third
quarters are less than 0.1 ng/m3 (0.06 ng/m3, 0.05 ng/m3, and 0.01 ng/m3,
respectively) while the difference for the fourth quarter is 0.20 ng/m3. Thus, this
quarter exhibits the most variability. Four of the 10 highest arsenic concentrations
measured at TOOK were measured during the fourth quarter while five of the 10
lowest arsenic concentrations measured at TOOK were also measured during the
fourth quarter.
18-28
-------
• The confidence interval for the first quarter average concentration of nickel is two to
three times larger than the confidence intervals for the remaining quarterly average
concentrations. Three of the four highest nickel concentrations were measured at
TOOK during the first quarter, including two greater than 7 ng/m3, which are among
the 10 highest nickel concentrations measured at NMP sites sampling nickel. Three of
the four lowest nickel concentrations measured at TOOK were measured during the
first quarter of 2014, including the minimum concentration for this site (0.989 ng/m3).
Thus, nickel concentrations measured at TOOK during the first quarter exhibit the
most variability.
Observations for TMOK from Table 18-4 include the following:
• Formaldehyde concentrations measured at TMOK were considerably lower during
the fourth quarter compared to the rest of the year. Concentrations measured at
TMOK range from 0.631 |ig/m3 to 8.01 |ig/m3, with nine of the 10 concentrations less
than 2 |ig/m3 measured during the fourth quarter. Twenty-one formaldehyde
concentrations greater than 4 |ig/m3 were measured at TMOK, all of which were
measured between February and August.
• Concentrations of acetaldehyde appear higher during the warmer months of the year,
based on the quarterly average concentrations, although the differences are not
significant. All four acetaldehyde concentrations greater than 3 |ig/m3 were measured
between May and August and 16 of the 25 concentrations greater than 2 |ig/m3 were
measured during the second and third quarters of 2014. Conversely, 11 of the 12
lowest concentrations were measured during the first and fourth quarters.
• The quarterly average concentrations of several VOCs are highest for the first and
third quarters of 2014 and lowest for the second and fourth quarters, although the
difference is not significant. The highest concentrations of benzene, ethylbenzene,
and 1,3-butadiene were all measured on the same days: February 16, 2014 and
August 3, 2014. A review of each of these pollutants' 10 highest concentrations
shows that more than half of these measurements were from samples collected during
the first and third quarters.
• Concentrations of hexachloro-l,3-butadiene appear highest for the first quarter of
2014. This is mostly due to the number of measured detections of this pollutant.
Hexachloro-1,3-butadiene was detected in 17 samples collected at TMOK; of these,
eight were measured during the first quarter, which is at least twice the number for
the other calendar quarters, which range from two (second quarter) to four (third
quarter). While the maximum hexachloro-1,3-butadiene concentration was also
measured during the first quarter (0.110 |ig/m3), so were several of the lowest
concentrations measured at this site; note, however, that all of the measurements of
this pollutant are less than the MDL.
• Arsenic concentrations appear highest during the third quarter and lowest for the first
quarter. The number of arsenic concentrations greater than 1 ng/m3 was highest for
the third quarter (4) and the lowest during the first quarter (0), with the number
between the two for the other calendar quarters. Conversely, only three arsenic
18-29
-------
concentrations less than 0.5 ng/m3 were measured at TMOK during the third quarter
compared to eight for the first quarter, including the three lowest concentrations
measured at this site. Note that the confidence interval is largest for the fourth quarter
average concentration. The maximum arsenic concentration was measured at TMOK
in December; December is also the most with the highest number of arsenic
concentrations less than 0.5 ng/m3 (9).
Observations for TROK from Table 18-4 include the following:
• The second and third quarter average concentrations of formaldehyde are
significantly higher than the other quarterly averages for TROK. The maximum
concentration of formaldehyde (6.39 |ig/m3) was measured on July 22, 2014 at
TROK, the same day the maximum concentration was measured at TOOK and
second highest concentration was measured at TMOK. The six concentrations of
formaldehyde greater than 5 |ig/m3 were measured at TROK between May and
August, and 20 of the 24 concentrations greater than 3 |ig/m3 were measured during
the second and third quarters of 2014 (with two each in the first and fourth quarters).
Conversely, all five concentrations of formaldehyde less than 1 |ig/m3 were measured
in November and December and all but two of the 22 concentrations less than
2 |ig/m3 were measured during the first and fourth quarters of 2014.
• The quarterly average concentrations of acetaldehyde have a similar pattern as those
for formaldehyde but to a lesser degree. Two of the three acetaldehyde concentrations
greater than 3 |ig/m3 were measured in August (with the other measured in May)
while none of the 21 concentrations less than 1.5 |ig/m3 were measured during the
third quarter of 2014 (compared to nine, four, and eight for the first, second, and
fourth quarters of the year).
• With the exception of two VOCs, the third quarter average concentration for each
pollutant of interest for TMOK is the highest among the quarterly averages shown in
Table 18-4 (although the statistical significance varies among the pollutants).
• The third and fourth quarter average benzene concentrations for TROK the same,
although the fourth quarter average has a larger confidence interval associated with it.
The two highest benzene concentrations were measured at TROK on
December 25, 2014 (2.09 |ig/m3) and October 20, 2014 (1.33 |ig/m3), with two other
benzene concentrations greater than 1 |ig/m3 also measured during the fourth quarter.
Three benzene concentrations greater than 1 |ig/m3 were also measured during the
third quarter, compared to one each during the first and second quarters.
• The third quarter average ethylbenzene concentration is the highest of the four
quarterly averages and is twice the first quarter average shown in Table 18-4. A
review of the data shows that there were no ethylbenzene concentrations greater than
0.5 |ig/m3 measured at TROK during the first quarter, compared to six measured
during the third quarter (and the number ranging from two to four for the remaining
calendar quarters). At the other end of the range, there were no ethylbenzene
concentrations less than 0.25 |ig/m3 measured at TROK during the third quarter,
18-30
-------
compared to nine measured during the first quarter (and the number ranging from
three to six for the remaining calendar quarters).
• The third quarter average arsenic concentration is the highest of the four quarterly
averages and is nearly twice the magnitude of the first quarter average concentration.
A review of the data shows that there were no arsenic concentrations greater than 1
ng/m3 measured at TROK during the first quarter, compared to seven measured
during the third quarter (and the number ranging from one to three for the other two
calendar quarters). Note that both the minimum (0.234 ng/m3) and maximum (2.22
ng/m3) arsenic concentrations were measured at TROK during the fourth quarter,
explaining the relatively large confidence interval shown for this quarterly average.
Observations for OCOK from Table 18-4 include the following:
• Similar to the Tulsa sites, the second and third quarter average concentrations of
formaldehyde are higher than the other quarterly averages for OCOK, and although
the third quarter average concentration is the highest, the confidence interval is
highest for the second quarter. A review of the data shows that all 16 formaldehyde
concentrations greater than 3.5 |ig/m3 were measured at OCOK during the second (7)
and third (9) quarters, with the maximum concentration measured on April 11, 2014
(6.90 |ig/m3). Formaldehyde concentrations less than 2 |ig/m3 were not measured
between May and September and only two of these were measured outside the first
and fourth quarters of 2014.
• The maximum acetaldehyde concentration was measured at OCOK on the same day
as the maximum formaldehyde concentration (April 11, 2014). The second highest
acetaldehyde concentration (3.18 |ig/m3) was also measured during the second
quarter, along with six additional measurements greater than 2 |ig/m3 (the most of any
calendar quarter).
• The maximum benzene and 1,3-butadiene concentrations were also measured at
OCOK on April 11, 2014. The carbon tetrachloride and arsenic concentrations
measured at OCOK on this date are not the maximums measured, but are among the
higher concentrations measured, ranking fifth and sixth, respectively for each
pollutant.
• The first quarter average concentration of carbon tetrachloride is the lowest of the
four shown for OCOK and the lowest among quarterly averages of carbon
tetrachloride calculated for the Oklahoma sites. A review of the data shows that the
five concentrations of carbon tetrachloride less than 0.5 |ig/m3 were measured at
OCOK during the first quarter of 2014, including the minimum concentration
measured across the program (0.0378 |ig/m3). The next lowest carbon tetrachloride
concentration measured at OCOK is an order of magnitude higher.
• Concentrations of arsenic measured at OCOK appear highest during the warmer
months and lowest during the cooler months. The only arsenic concentration greater
than 1 ng/m3 was measured at OCOK on August 3, 2014 (1.19 ng/m3), with the next
two highest arsenic concentrations measured in July and August. Sixteen arsenic
18-31
-------
concentrations greater than 0.5 ng/m3 were measured at OCOK between April and
September, compared to seven for the remainder of the year. On the other end of the
concentration scale, arsenic concentrations less than 0.25 ng/m3 were not measured at
OCOK between April and September, compared to eight for the remainder of the
year.
Observations for YUOK from Table 18-4 include the following:
• Although similar to the second and third quarter averages, the first quarter average
concentration of acetaldehyde has a confidence interval two to three times larger than
the other averages shown, indicating a high level of variability and/or potential
outliers. The maximum acetaldehyde concentration was measured at YUOK on
January 17, 2014 (7.80 |ig/m3) and is more than twice the next highest concentration
measured during the first quarter (3.79 |ig/m3), with only one other measurement
falling in between (4.28 |ig/m3, measured in April 11, 2014). YUOK is one of only
five NMP sites with an acetaldehyde measurement greater than 7 |ig/m3. All other
acetaldehyde concentrations measured at YUOK are less than 3 |ig/m3.
• Two formaldehyde concentrations of 6.15 |ig/m3 were measured at YUOK, one on
April 11, 2014, the same day as the highest formaldehyde concentration was
measured at OCOK, and one on August 15, 2014. Formaldehyde concentrations
greater than 4 |ig/m3 were measured in the first (2), second (3), and third (8) quarters
of 2014 while none were measured during the fourth quarter. Conversely, at least one
concentration less than 2 |ig/m3 was measured during each calendar quarter (seven
during the first quarter, two during the second quarter, one during the third quarter,
and 10 during the fourth quarter).
• The second quarterly average concentration of hexachloro-1,3-butadiene is zero,
indicating that all measurements were non-detects. Measured detections were not
measured at YOUK between March and June. This is also true for December. There
was at least one measured detection in all other months, ranging from one (January,
July, September, and October) to five (February). However, none of these were
greater than the MDL for this pollutant.
• The maximum arsenic concentration was measured on July 4, 2014 and is the only
measurement greater than 1 ng/m3 measured at this site (1.03 ng/m3). Higher arsenic
concentrations were measured more often during the warmer months of the year and
lower concentrations measured more often during the cooler months of the year. For
instance, arsenic concentrations greater than 0.5 ng/m3 were measured during each
calendar quarter, two during the first, 10 during the second, nine during the third, and
three during the fourth. Conversely, arsenic concentration less than 0.25 ng/m3 were
not measured during the second and third calendar quarters, compared to seven
during the first quarter and six during the fourth quarter.
18-32
-------
Tables 4-9 through 4-12 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 4-9 through 4-12 a total of 26 times. However,
because they are the only sites sampling TSP metals, each Oklahoma site appears for
each metal, accounting for 10 of the appearances.
• All three Tulsa sites appear in Table 4-9 among the sites with the highest annual
average concentrations of ethylbenzene, with TOOK ranking fourth, TROK ranking
fifth, and TMOK ranking eighth. These annual averages vary by only 0.05 |ig/m3.
The Tulsa sites also rank fourth through sixth for their annual averages of
/;-dichlorobenzene. These three sites also appear in Table 4-9 for their annual
averages of hexachloro-1,3-butadiene.
• TOOK has the seventh highest annual average of concentration of benzene, with
TMOK's annual average ranking 10th. These two sites also rank seventh and ninth,
respectively, for their annual average concentrations of 1,2-dichloroethane.
• YUOK appears only once in Table 4-9: YUOK has the ninth highest annual average
concentration of carbon tetrachloride among NMP sites sampling VOCs (although
only 0.02 |ig/m3 separates the Oklahoma sites' annual average concentrations of this
pollutant).
• OCOK does not appear in Table 4-9.
• The annual average concentration of acetaldehyde for TOOK ranks 10th among NMP
sites sampling this pollutant. The annual average concentration of formaldehyde for
TMOK ranks seventh among NMP sites. The remaining sites do not appear in
Table 4-10 for carbonyl compounds.
• The Tulsa sites rank higher than OCOK and YUOK for the two TSP metals shown in
Table 4-12 and there is a considerable decrease 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 while TOOK has the highest annual
average nickel concentration among the Oklahoma sites. Similar observations were
made in the 2013 NMP report.
18.4.2 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 18-4 for the five Oklahoma sites. Figures 18-13 through 18-24 overlay these sites'
minimum, annual average, and maximum concentrations onto the program-level minimum, first
18-33
-------
quartile, median, average, third quartile, and maximum concentrations, as described in Section
3.4.3.1, and are discussed below.
Figure 18-13. Program vs. Site-Specific Average Acetaldehyde Concentrations
h
—
-o
k
P <
cU
P <
0123456789 10
Concentration (ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 18-13 presents the box plots for acetaldehyde for all five sites and shows the
following:
• The range of acetaldehyde concentrations measured was smallest for TROK and
largest for YUOK, with the minimum concentrations measured similar across the
sites but the maximum concentrations varying from 3.30 |ig/m3 to 7.80 |ig/m3,
• TOOK has the highest annual average concentration of acetaldehyde among the
Oklahoma sites, which is greater than the program-level average concentration. The
18-34
-------
annual average concentrations for the remaining sites vary by less than 0.08 |ig/m3,
and are similar to the program-level average concentration.
Figure 18-14. Program vs. Site-Specific Average Arsenic (TSP) Concentrations
TOOK
vJ
TMOK
TROK
OCOK
YUOK
Concentration {ng/m3)
Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 18-14 presents the box plots for arsenic for all five sites and shows the following:
• Because the Oklahoma sites are the only sites sampling TSP metals, Figure 18-14
compares each Oklahoma site's arsenic data against the combined Oklahoma data.
• The range of arsenic concentrations measured was smallest for YUOK and largest for
TROK. Non-detects of arsenic were not measured at these sites.
• The annual average arsenic (TSP) concentration is greatest for TROK (although the
annual average for TOOK is similar) and lowest for YUOK. This figure also shows
18-35
-------
that arsenic concentrations were higher at the Tulsa sites, based on both the range of
measurements as well as the annual average concentrations.
Figure 18-15. Program vs. Site-Specific Average Benzene Concentrations
TOOK
I
-O-
Program Max Concentration = 12.4 |ig/m3
TMOK
Program Max Concentration = 12.4 ng/m3
TROK
Program Max Concentration = 12.4 M-g/m3
J '
OCOK
h
Program Max Concentration = 12.4 ng/m3
1
*
h
Program Max Concentration = 12.4 ng/m3
0 2 4 6 8 10
Concentration (ng/m3)
Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 18-15 presents the box plots for benzene for all five sites and shows the following:
• The program-level maximum benzene concentration (12.4 |ig/m3) is not shown
directly on the box plots in Figure 18-15 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 range of benzene concentrations was smaller at the Oklahoma City sites
compared to the Tulsa sites.
18-36
-------
• The annual average concentrations of benzene decrease across the sites as the
Figure 18-15 is viewed from top to bottom, with the annual average for TOOK nearly
twice the annual average for YUOK. TOOK's annual average concentration is greater
than the program-level average concentration and third quartile; TMOK and TROK's
annual average concentrations are similar to each other and fall between the program-
level average concentration and third quartile; OCOK's annual average is between the
program-level median and average concentrations; and YUOK's annual average is
just less than the program-level median concentration.
TOOK
TMOK
TROK
OCOK
YUOK
0 0.2 0.4 0.6 0.8 1
Concentration (ng/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 18-16. Program vs. Site-Specific Average 1,3-Butadiene Concentrations
Program Max Concentration = 5.90 ng/m3
n
Program Max Concentration = 5.90 ng/m3
Program Max Concentration = 5.90 ng/m3
Program Max Concentration = 5.90 ng/m3
1
Program Max Concentration = 5.90 ng/m3
18-37
-------
Figure 18-16 presents the box plot for 1,3-butadiene for all five sites and shows the
following:
• Similar to benzene, the program-level maximum 1,3-butadiene concentration
(5.90 |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 been reduced to 1 |ig/m3.
• The range of 1,3-butadiene concentrations measured at the Oklahoma sites are
considerably less than the range measured at the program-level, as all concentrations
measured at these sites are less than 0.25 |ig/m3.
• All of the annual average concentrations of 1,3-butadiene for the Oklahoma sites are
less than the program-level average concentration. The annual average concentration
of 1,3-butadiene is highest for TMOK and lowest for OCOK with approximately
0.05 |ig/m3 separating them. The annual average concentrations for the Tulsa sites are
greater than the annual averages for the Oklahoma City sites, with the annual
averages for the Tulsa sites greater than the program-level median concentration and
the annual averages for OCOK and YUOK less than the program-level median
concentration.
18-38
-------
Figure 18-17. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations
TOOK
Program Max Concentration = 3.06 |ig/m3
TMOK
Program Max Concentration = 3.06 |ig/m3
TROK
Program Max Concentration = 3.06 |ig/m3
OCOK
Program Max Concentration = 3.06 |ig/m3
Program Max Concentration = 3.06 (ig/m3
0 0.5 1 1.5 2 2.5
Concentration (ng/m3)
Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 18-17 presents the box plots for carbon tetrachloride for all five 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. Note that the
program-level median and average concentrations are similar and plotted nearly on
top of each other.
• Several of the lowest carbon tetrachloride concentrations across the program were
measured at Oklahoma sites, including the minimum concentration measured in 2014.
18-39
-------
• The annual average concentrations for these five sites vary little, ranging from
0.61 |ig/m3 and 0.63 |ig/m3 for each site, all of which are just less than the
program level average concentration of 0.64 |ig/m3.
Figure 18-18. Program vs. Site-Specific Average p-Dichlorobenzene Concentrations
TOOK
O i
KJ 1
TMOK
-O
—
o
i i i i i
0 0.2 0.4 0.6 0.8 1 1.2
Concentration (ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 18-18 presents the box plots for /;-dichlorobenzene for TOOK, TMOK, and
TROK, and shows the following:
• /;-Dichlorobenzene is a pollutant of interest for only the three Tulsa sites. Note that
the program-level first and second quartiles are both zero and therefore not visible on
the box plots.
• All /;-dichlorobenzene concentrations measured at these sites are less than 0.2 |ig/m3,
each an order of magnitude less than the maximum concentration measured across the
program.
• The annual average p-dichlorobenzene concentration for each Tulsa site is greater
than the program-level average concentration, with the annual averages for TOOK
and TMOK also greater than the program-level third quartile (and the annual average
for TROK similar to it).
• The number of non-detects measured at these sites ranges from 10 (TOOK) to 14
(TROK).
18-40
-------
Figure 18-19. Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations
TOOK
¦
D
Program Max Concentration = 27.4 ng/m3
TMOK
O
Program Max Concentration = 27.4 ng/m3
°
TROK
¦
Program Max Concentration = 27.4 ng/m3
OCOK
¦
Program Max Concentration = 27.4 ng/m3
Program Max Concentration = 27.4 |ig/m3
0.4 0.6
Concentration (ng/m3)
Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 18-19 presents the box plots for 1,2-dichloroethane for all five sites and shows the
following:
• The scale of the box plots in Figure 18-19 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 (27.4 |ig/m3) is
considerably greater than the majority of measurements.
• The program-level average concentration 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 (even the maximum concentration measured at TOOK, although
difficult to discern in Figure 18-19).
18-41
-------
• The annual average concentrations of 1,2-dichloroethane for these sites fall on either
side of the program-level median concentration, with less than 0.02 |ig/m3 separating
these sites' annual averages.
Figure 18-20. Program vs. Site-Specific Average Ethylbenzene Concentrations
O i
{J 1
TMOK
O i
KJ 1
Y
O i
{J 1
t 1 1 1 1 r
0 0.5 1 1.5 2 2.5 3 3.5
Concentration (ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 18-20 presents the box plots for ethylbenzene for TOOK, TMOK, and TROK, and
shows the following:
• Ethylbenzene is a pollutant of interest for only the three Tulsa sites.
• The range of ethylbenzene concentrations measured is largest for TOOK and smallest
for TROK.
• The annual average concentrations for each of the Tulsa sites are greater than the
program-level average concentration and third quartile. Approximately 0.05 |ig/m3
separates the annual average concentrations of ethylbenzene for the Tulsa sites.
18-42
-------
Figure 18-21. Program vs. Site-Specific Average Formaldehyde Concentrations
1+
¦
--o ¦
¦
t—'
¦
—'
¦
t—'
0 3 6 9 12 15 18 21 24 27
Concentration (ng/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 18-21 presents the box plots for formaldehyde for all five sites and shows the
following:
• The maximum formaldehyde concentrations measured at these sites are all
considerably less than the maximum concentration measured across the program,
with the highest concentration among the Oklahoma sites measured at TMOK
(although a similar concentration was also measured at TOOK).
• TMOK has the highest annual average concentration of formaldehyde among the
Oklahoma sites, which is the only one greater than 3 |ig/m3. The annual averages for
the Tulsa sites and YUOK are all greater than the program-level average
concentration (although the difference for TROK is negligible).
18-43
-------
Figure 18-22. Program vs. Site-Specific Average Hexachloro-l,3-butadiene Concentrations
TOOK
TMOK
TROK
YUOK
0.1
0.2
0.3
0.4
0.5
0.6
Concentration (ng/m3)
Program:
IstQuartile
¦
2nd Quartile
~
3rd Quartile
~
4th Quartile
~
Average
i
Site:
Site Average
o
Site Concentration Range
-O-
Figure 18-22 presents the box plots for hexachloro-1,3-butadiene for four of the five
Oklahoma sites, and shows the following:
• Hexachloro-1,3-butadiene was not identified as a pollutant of interest for OCOK, and
thus, this site has no box plot in Figure 18-22. Note that the program-level first,
second, and third quartiles for hexachloro-l,3-butadiene are zero and therefore not
visible on the box plot.
• The ranges of hexachloro-l,3-butadiene concentrations measured at the Tulsa sites
are similar to each other, with a slightly smaller the range measured at YUOK.
However, non-detects make up the majority of concentrations measured at these sites.
• The annual average concentrations of hexachloro-l,3-butadiene for the three Tulsa
sites are just slightly greater than the program-level average concentration, with the
annual average for YUOK just slightly less than the program-level average. Less than
0.01 |ig/m3 separates these annual averages.
18-44
-------
Figure 18-23. Program vs. Site-Specific Average Manganese (TSP) Concentration
^ 1
VJ 1
0 10 20 30 40 50 60
Concentration {ng/m3)
Program: 1st Quartile
2nd Quartile 3rd Quartile
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 18-23 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 18-23
compares the manganese concentrations measured at TOOK against the combined
Oklahoma data.
• The maximum manganese concentration among the Oklahoma sites was
measured at TOOK.
• The annual average manganese concentration for TOOK is greater than the
program-level manganese concentration and third quartile (TSP only). A similar
observations was made in the 2013 NMP report.
Figure 18-24. Program vs. Site-Specific Average Nickel (TSP) Concentrations
Concentration {ng/m3]
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 18-24 presents the box plot for nickel for TOOK and shows the following:
• Similar to manganese, TOOK is the only Oklahoma site for which nickel is a
pollutant of interest. Because the Oklahoma sites are the only sites sampling TSP
metals, Figure 18-24 compares the nickel concentrations measured at TOOK
against the combined Oklahoma data. Note that the majority of concentrations
18-45
-------
measured at the Oklahoma sites fall into a more compressed range for nickel than
for manganese, as indicated by the first, second, and third quartiles in the box
plot.
• The maximum nickel concentration among the Oklahoma sites was measured at
TOOK. The minimum nickel concentration measured at TOOK is greater than the
first quartile (TSP only).
• The annual average nickel concentration for TOOK is greater than the program-
level nickel concentration and third quartile (TSP only).
18.4.3 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 3.4.3.2.
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 18-25 through
18-53 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.
18-46
-------
Figure 18-25. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at TOOK
Maximum
1 A 1-year average is not presented because issues at the onset of sampling resulted in low completeness.
Observations from Figure 18-25 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 concentrations were all measured in 2011 or 2012. Of the 35 acetaldehyde
concentrations greater than 4 |ig/m3 measured at TOOK, 12 were measured in 2012,
eight were measured in 2011, five were measured in 2010, and three or fewer were
measured in each of the other years.
• The statistical metrics exhibit an increasing trend between 2009 and 2011, with little
change shown in the acetaldehyde measurements from 2011 to 2012. The 95th
percentiles for 2011 and 2012 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.
18-47
-------
Figure 18-26. Yearly Statistical Metrics for Arsenic (TSP) Concentrations Measured at TOOK
Maximum
Observations from Figure 18-26 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 18-26 excludes data from 2006 per the criteria specified in
Section 3.4.3.2.
• 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.
All eight concentrations of arsenic greater than 2 ng/m3 were measured in either 2007
or 2008.
• 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 increased for 2011, an increase that
continued into 2012.
• All of the statistical parameters exhibit decreases from 2012 to 2013.
• Little change is shown in the central tendency statistics between 2013 and 2014
despite the higher concentrations measured in 2014. The additional concentrations at
the upper end of the concentration range measured in 2014 (seven arsenic
concentrations greater than 1.25 ng/m3 were measured in 2014 compared to two in
18-48
-------
2013) are countered by a higher number of measurements at the lower end of the
concentration range (the number of arsenic concentrations less than 0.5 ng/m3
doubled from 2013 (nine) to 2014 (18)).
Figure 18-27. Yearly Statistical Metrics for Benzene Concentrations Measured at TOOK
1
T
T
0
rU rh
..... -<
I ± ^ i
2006 1 2007 2008 2009 2010 2011 2012 2013 2014
Year
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 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 of the 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 1-year average benzene concentration has fluctuated over the years. After a
significant decrease from 2008 to 2009, an increasing trend through 2011 occurred.
After 2011, a significant decrease trend in benzene concentrations is shown. Most of
the statistical parameters are at a minimum for 2014 (with the exceptions calculated
for 2013). The smallest range of benzene concentrations was measured in 2014, with
18-49
-------
the maximum concentration measured in 2014 less than 3 |ig/m3 and the 95th
percentile is less than 2 |ig/m3 for the first time since sampling began in 2006.
Figure 18-28. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at TOOK
0.40
0.35
0.30
0.25
"e
s
£ 0.20 T
! r-t
u
c
u
0.15
0.10
0.05
0.00
2006 1 2007 2008 2009 2010 2011 2012 2013 2014
Year
O 5th Percentile - Minimum - Median - Maximum o 95th Percentile Averege
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 1,3-butadiene concentrations measured at TOOK
include the following:
• The maximum concentration of 1,3-butadiene was measured in December 2011
(0.34 |ig/m3), although a similar concentration was also measured in 2007
(0.33 |ig/m3). Concentrations of 1,3-butadiene greater than 0.3 |ig/m3 have not been
measured at TOOK.
• 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 and 2012) 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.
18-50
-------
• With the exception of the maximum concentration, all of the statistical parameters
exhibit decreases for 2013. Additional decreases are shown for 2014 for the statistical
parameters representing the upper end of the concentration range while other
statistical parameters exhibit increases.
Figure 18-29. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
TOOK
£
If 1.0
2010
Year
- Minimum
- Maximum
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 carbon tetrachloride concentrations measured at TOOK
include the following:
• Similar to other compounds, 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 2012,
when the difference between the minimum and maximum concentrations is less than
0.5 |ig/m3.
The 1-year average concentration increased slightly from 2007 to 2008, after which
little change is shown through 2011. Between 2008 and 2011, the 1-year average
concentrations range from 0.61 |ig/m3 to 0.63 |ig/m3. A slight increase is shown for
2012 (0.66 |ig/m3), even though the measurements for this year exhibit the least
variability. After 2012, the 1-year average concentration of carbon tetrachloride
18-51
-------
returned to previous levels. Across the years of sampling, the 1-year average (and
median concentrations) have varied by only about 0.10 |ig/m3.
Figure 18-30. Yearly Statistical Metrics for />-Dichlorobenzene Concentrations Measured at
TOOK
1 <
~
> ^
> rh ph
Ets E
20061 2007 2008 2009 2010 2011 2012 2013 2014
Year
1 A 1-year average is not presented because issues at the onset of sampling resulted in low completeness.
Observations from Figure 18-30 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). Only 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.
• 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).
• After 2011, concentrations ofp-dichlorobenzene decreased significantly.
Concentrations greater than 0.2 |ig/m3 were not measured in 2012 or afterward. Aside
18-52
-------
from the minimum and 5th percentile, all of the statistical metrics are at a minimum
for 2013. There is relatively little difference in the p-dichlorobenzene concentrations
measured in 2012 and 2014.
• There were no non-detects of/?-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) then decreases to 11 for
2014.
Figure 18-31. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
TOOK
1
,o
o
. o
2006 1 2007 2008 2009 2010 2011 2012 2013 2014
Year
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 18-31 for 1,2-dichloroethane concentrations measured at
TOOK include the following:
• The median concentration for all years 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. Greater than 30 measured detections were measured in 2013, and in 2014, there
were 50 measured detections, the most of any year since the onset of sampling.
18-53
-------
• The 1-year average concentration for 2012 is less than the median concentration,
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. The 1-year average
concentration for 2012 is less than the median, indicating that 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). This is also true for 2013 and 2014, although the difference
between the two statistical parameters is less.
Figure 18-32. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at TOOK
o
<2
2007
2010
Year
- Minimum
- Maximum
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 18-32 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
18-54
-------
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 approaching 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. Most of the statistical
parameters exhibit additional decreases for 2014, albeit slight decreases.
Figure 18-33. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at TOOK
o
o
2007 2008 2009
2010
Year
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.
18-55
-------
Observations from Figure 18-33 for formaldehyde concentrations measured at TOOK
include the following:
• The maximum concentration of formaldehyde (12.8 |ig/m3) was measured at TOOK
on June 26, 2012. Only one other measurement greater than 10 |ig/m3 has been
measured at TOOK (10.2 |ig/m3 measured in 2011).
• All 84 formaldehyde concentrations greater than 5 |ig/m3 were measured at TOOK
during the second and third quarters, particularly the period between June and August
(accounting for 72 concentrations), regardless of year.
• The trends graph for formaldehyde resembles the graph for acetaldehyde, with an
increasing trend in the 1-year average concentration shown for formaldehyde between
2009 and 2011. The 1-year average increased by 1 |ig/m3 over this period (with
increases exhibited by the median concentration as well).
• Even though the maximum formaldehyde concentration was measured in 2012, all of
the other statistical parameters exhibit slight decreases. Further decreases are shown
for all of the statistical parameters for 2013.
• The range of formaldehyde concentrations measured at TOOK in 2014 is similar to
those measured in 2013, with subtle changes in the 1-year average and median
concentrations.
18-56
-------
Figure 18-34. Yearly Statistical Metrics for Hexachloro-l,3-butadiene Concentrations
Measured at TOOK
0.25
0.20
m~" 0.15
E
u 0.10
0.05
0.00
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 18-34 for hexachloro-l,3-butadiene concentrations measured at
TOOK include the following:
• The trends graphs for hexachloro-1,3-butadiene resembles the trends graph for
1,2-dichloroethane in that there were few measured detections in the first several
years of sampling at TOOK.
• The median concentration is zero for all years of sampling, indicating that at least half
of the measurements were non-detects for each year. Between 2006 and 2010, there
were a total of four measured detections. In 2011, five measured detections were
reported. This number doubled for 2012, increased to 13 for 2013, and is at a
maximum for 2014 (19).
, 1
>J 2
> <
.... -<
>
>
20061 2007 2008 2009 2010 2011 2012 2013 2014
18-57
-------
Figure 18-35. Yearly Statistical Metrics for Manganese (TSP) Concentrations Measured at
TOOK
Year
o 5th Percentile
o 95th Percentile
Observations from Figure 18-35 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 the 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 measurements 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 measurements 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.
18-58
-------
• The range of concentrations measured in 2014 is among the smallest measured at
TOOK. The 1-year average and median concentrations vary by less than 1 ng/m3 for
2014, indicating less variability in the manganese concentrations measured in 2014.
Figure 18-36. Yearly Statistical Metrics for Nickel (TSP) Concentrations Measured at TOOK
O 95th Percentile
Observations from Figure 18-36 for nickel (TSP) concentrations measured at TOOK
include the following:
• The trends graph for nickel resembles the trends graph for manganese in several
ways.
• The maximum concentration of nickel (12.8 ng/m3) was measured at TOOK on the
same day as the maximum concentration of manganese (October 18, 2012, the day of
a dust storm). Only one additional nickel concentrations greater than 10 ng/m3 has
been measured at TOOK (11.0 ng/m3 measured on July 3, 2013). Eight of the 10
nickel concentrations greater than 5 ng/m3 were measured at TOOK in either 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. The minimum concentration shown for 2012
is greater than the 5th percentile for the four previous years.
18-59
-------
• With the exception of the maximum concentration, the statistical metrics shown for
2012 through 2014 more closely resemble those shown for 2007 than the years in-
between.
Figure 18-37. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at TMOK
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 18-37 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, after which the range of measurements has decreased each year.
• A decreasing trend is shown in the 1-year average and median concentrations
between 2011 and 2014, with many of the statistical parameters at a minimum for
2014.
18-60
-------
Figure 18-38. Yearly Statistical Metrics for Arsenic (TSP) Concentrations Measured at TMOK
Maximum
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 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.
• The three highest arsenic concentrations measured at TMOK were all measured in
2009, and all but one of the six arsenic concentrations greater than 2 ng/m3 were
measured in 2009. 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 in 2012 (15) is more
than double the number measured in each of the previous years, with the exception of
2009 (16).
• Excluding 2009, the statistical metrics for arsenic concentrations measured at TMOK
resemble those shown in Figure 18-26 for TOOK.
18-61
-------
Figure 18-39. Yearly Statistical Metrics for Benzene Concentrations Measured at TMOK
4.5
4.0
3.5
3.0
m£
5 25
c
| 2.0
c
o
u
1.5
1.0
0.5
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 April 2009.
Observations from Figure 18-39 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 benzene concentrations greater than 3 |ig/m3 have been
measured in all years of sampling prior to 2013.
• The 1-year average and median benzene concentrations have a significant decreasing
trend between 2010 and 2014, with the largest decrease shown for 2013. The 1-year
average and median concentrations have both decreased by half since the onset of
sampling. The maximum concentration measured in 2014 is less than the 95th
percentile for the previous year of sampling. This is also true for 2013.
18-62
-------
Figure 18-40. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at TMOK
Maximum
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 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, with the range
of concentrations measured increasing each year through 2012. After 2012, the range
of measurements decreases each year.
• 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 (2014) 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).
18-63
-------
Figure 18-41. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
TMOK
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 2009.
Observations from Figure 18-41 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 highest number of carbon tetrachloride
concentrations greater than 0.6 |ig/m3 was measured in 2012, accounting for 51 of the
61 measurements (compared to between 30 and 40 for each 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
shown for 2014 is the third-lowest carbon tetrachloride concentration measured
across the program in 2014.
• All of the 1-year average carbon tetrachloride concentrations shown fall between
0.60 |ig/m3 and 0.70 |ig/m3. This is also true for the median concentration.
18-64
-------
Figure 18-42. Yearly Statistical Metrics for o-Dichlorobenzene Concentrations Measured at
TMOK
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 2009.
Observations from Figure 18-42 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 the concentrations of p-dichlorobenzene is shown through
2012. The median decreases by almost half during this period 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).
• The decreasing trend in p-dichlorobenzene concentrations measured at TMOK shown
prior to 2013 resumes in 2014.
18-65
-------
Figure 18-43. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
TMOK
0.20
0.18
0.16
0.14
fr o.i2
E
1
I 0.10
c
c
u 0.08
0.06
0.04
0.02
0.00
2009 1 2010 2011 2012 2013 2014
Year
o 5th Percentile — Minimum - Median — Maximum o 95th Percentile Aversge
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 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. In 2014,
measured detections account for 53 of the 62 valid samples collected, representing an
85 percent detection rate.
• 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 after 2012 is less than the median, indicating that concentrations on the
lower end of the concentration range are pulling the 1-year average downward (just
like a maximum or outlier concentration can drive the average upward).
18-66
-------
Figure 18-44. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at TMOK
Maximum
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2009.
Observations from Figure 18-44 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 0.05 |ig/m3
separates the median concentrations for these years and approximately 0.01 |ig/m3
separates the 1-year average concentrations during this period.
• With the exception of the maximum concentration, all of the statistical parameters
exhibit decreases for 2013. These decreases continue into 2014, with the 1-year
average concentration at a minimum (and less than 0.4 |ig/m3 for the first time).
18-67
-------
Figure 18-45. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
TMOK
2009 1
2011 2012
Year
o 5th Percentile
— Minimum
— Maximum
o 95th Percentile
..4.... Average
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2009.
Observations from Figure 18-45 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.
• The 1-year average concentration increased from 2010 to 2011, then decreases each
year through 2013, when the 1-year average is at a minimum. Slight increases are
shown for 2014. However, these changes are not statistically significant. The 1-year
average concentrations have ranged from 3.19 |ig/m3 (2013) to 3.94 |ig/m3 (2011).
The median concentration is also at a minimum for 2013, ranging from 2.63 |ig/m3
(2013) to 3.30 |ig/m3 (2014) across the years of sampling.
• The 1-year average concentrations for formaldehyde exhibit a similar pattern as the
1-year average concentrations for acetaldehyde for TMOK shown through 2013. For
2014, the acetaldehyde concentrations continue to decrease while the formaldehyde
concentrations exhibit a slight increase.
18-68
-------
Figure 18-46. Yearly Statistical Metrics for Hexachloro-l,3-butadiene Concentrations
Measured at TMOK
20091
2011 2012
Year
o 5th Percentile
— Minimum
— Maximum
o 95th Percentile
..4.... Average
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2009.
Observations from Figure 18-46 for hexachloro-l,3-butadiene concentrations measured at
TMOK include the following:
• 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. There were no measured detections in 2009, two in 2010, three in 2011, nine in
2012, 14 in 2013, and 16 were measured in 2014.
Although 2014 has the highest number of measured detections since the onset of
sampling, all of the measured detections are less than the MDL. This is true for all
years of sampling.
18-69
-------
Figure 18-47. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at OCOK
2011 2012
Year
O 5th Percentile - Minimum - Med en - Maximum o 95th Percentile ••¦A-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 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). Only 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 is shown from 2011 to 2012.
• All of the statistical parameters exhibit decreases from 2012 to 2013. Most of the
statistical parameters exhibit at least slight decreases for 2014 as well.
18-70
-------
Figure 18-48. Yearly Statistical Metrics for Arsenic (TSP) Concentrations Measured at OCOK
3.5
3.0
2.5
0.0
2009 1 2010 2011 2012 2013 2014
Year
O 5th Percentile - Minimum - Med en — 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 18-48 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).
The maximum concentration measured each year after 2009 has been steadily
decreasing, reaching a minimum for 2013 (1.03 ng/m3), with 2014 the only year that
does not follow this pattern. At the same time, the minimum concentration measured
increased each year through 2012, reaching a maximum of 0.21 ng/m3.
• Most of the 1-year average concentrations of arsenic fall between 0.40 ng/m3 and
0.50 ng/m3, with 2012 as the only exception (0.57 ng/m3). Nearly as many arsenic
concentrations greater than 1 ng/m3 were measured in 2012 (7) as all other years put
together (8).
18-71
-------
Figure 18-49. Yearly Statistical Metrics for Benzene Concentrations Measured at OCOK
Maximum
Concentration for
2013 is 9.38 ng/m3
2009 1
2010
2011
2012
Year
2013
2014
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 18-49 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). No other benzene concentrations greater than 3 |ig/m3 have been
measured at OCOK.
• With the exception of 2013, the 1-year average concentration has a decreasing trend
across the years of sampling. If the maximum concentration measured in 2013 was
excluded from the calculation, as no other benzene concentrations greater than
2 |ig/m3 were measured in 2013, the 1-year average concentration would have a
continuous decreasing trend through 2013, with virtually no change for 2014.
• Benzene concentrations measured at OCOK in 2014 exhibit the least amount of
variability (excluding 2009, which does not include a full year's worth 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.
18-72
-------
Figure 18-50. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at OCOK
Maximum
Concentration for
2011 is 10.0 Mi/m3
o
2011 2012
Year
5th Percentile
— Minimum
— Maximum
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 18-50 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.
The median concentrations shown between 2010 and 2014 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), despite the variation in the range of concentrations measured.
The range within which the majority of concentrations fall, as indicated by the
difference between the 5th and 95th percentiles, increased each year through 2012.
This is followed by a decreasing of the range in 2013 and again in 2014.
18-73
-------
Figure 18-51. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
OCOK
1.2
1.0
0.8
\
u
0.4
0.2
0.0
2009 1 2010 2011 2012 2013 2014
Year
o 5th Percentile — Minimum - Median — Maximum o 95th Percentile Aversge
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 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 concentrations
measured in other years are less than 0.90 |ig/m3.
• The range of carbon tetrachloride concentrations measured at OCOK has 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
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, five carbon tetrachloride
concentrations less than 0.1 |ig/m3 have been measured at OCOK, one each in 2009,
2010, and 2014, and two in 2011. This explains why the box and whisker plots for
carbon tetrachloride appear "inverted" for several years, with the minimum
_L
~
-
1
—
wm
rh
18-74
-------
concentration extending farther away from the majority of the measurements rather
than the maximum concentration, which is more common (see acetaldehyde as an
example).
Figure 18-52. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
OCOK
Concentration for
2011 is 0.706 ug/m3
o
2011 2012
Year
5th Percentile
- Minimum
- Maximum
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 18-52 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 to 11 for 2010 and 13 for 2011. For
2012, the number of measured detections increased by a factor of four (up to 52). The
number of measured detections is relatively constant for 2013 and 2014.
• The increase in the 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 2012, 2013, and 2014 and less than 0.005 |ig/m3
separates the median concentrations 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 2014, even as a greater number of
measured detections were measured.
18-75
-------
Figure 18-53. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
OCOK
«
2011 2012
Year
o 5th Percentile
— Minimum
— Maximum
o 95th Percentile
..4.... Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2009.
Observations from Figure 18-53 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). All 17 formaldehyde concentrations greater than
7 |ig/m3 were measured at OCOK in either 2011 or 2012.
• 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
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 are lower, as the statistical
parameters exhibit decreases in the years following 2011, particularly at the upper
end of the concentration range. The concentrations measured in 2013 are similar to
those measured in 2014 as the statistical parameters shown for these years exhibit
little change.
18-76
-------
18.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at each Oklahoma monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4
for definitions and explanations regarding the various toxicity factors, time frames, and
calculations associated with these risk-based screenings.
18.5.1 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 3.4.3.3 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 18-5, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Observations from Table 18-5 include the following:
• Formaldehyde and acetaldehyde have the highest annual average concentrations for
each site. Among the VOCs, benzene has the highest annual average concentration
for four of the five sites (YUOK is the exception, where carbon tetrachloride has the
highest annual average concentration among the VOCs). Arsenic is the only TSP
metal that was identified as a pollutant of interest for all five of the Oklahoma sites.
Annual average arsenic concentrations are all less than 1 ng/m3.
• Formaldehyde and benzene have the highest cancer risk approximations among the
pollutants of interest for each Oklahoma monitoring site. Cancer risk approximations
for formaldehyde range from 34.18 in-a-million for OCOK to 44.38 in-a-million for
TMOK. TMOK's cancer risk approximation for formaldehyde ranks eighth highest
among all cancer risk approximations program-wide. Benzene cancer risk
approximations for the Oklahoma monitoring sites range from 4.33 in-a-million for
YUOK to 8.07 in-a-million for TOOK.
• For arsenic, the cancer risk approximations range from 1.91 in-a-million for YUOK
to 3.32 in-a-million for TROK.
18-77
-------
• None of the pollutants of interest 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 was calculated for
formaldehyde for TMOK (0.35).
Table 18-5. Risk Approximations for the Oklahoma Monitoring Sites
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Public Works, Tulsa, Oklahoma - TOOK
Acetaldehyde
0.0000022
0.009
61/61
1.97
±0.22
4.33
0.22
Benzene
0.0000078
0.03
61/61
1.03
±0.11
8.07
0.03
1.3 -Butadiene
0.00003
0.002
59/61
0.08
±0.01
2.30
0.04
Carbon Tetrachloride
0.000006
0.1
61/61
0.61
±0.03
3.64
0.01
/?-Dichlorobcnzcnc
0.000011
0.8
51/61
0.07
±0.01
0.82
<0.01
1,2 -Dichloroethane
0.000026
2.4
50/61
0.09
±0.01
2.27
<0.01
Ethylbenzene
0.0000025
1
61/61
0.39
±0.06
0.99
<0.01
Formaldehyde
0.000013
0.0098
61/61
2.95
±0.42
38.30
0.30
Hexachloro -1,3 -butadiene
0.000022
0.09
19/61
0.03
±0.01
0.57
<0.01
Arsenic (TSP)a
0.0043
0.000015
62/62
0.76
±0.08
3.26
0.05
Manganese (TSP)a
0.0003
62/62
25.04
±2.92
0.08
Nickel (TSP)a
0.00048
0.00009
62/62
2.25
±0.31
1.08
0.02
— = 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.
18-78
-------
Table 18-5. Risk Approximations for the Oklahoma Monitoring Sites (Continued)
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Fire Station, Tulsa, Oklahoma - TMOK
Acetaldehyde
0.0000022
0.009
62/62
1.81
±0.17
3.98
0.20
Benzene
0.0000078
0.03
62/62
0.81
±0.08
6.32
0.03
1.3 -Butadiene
0.00003
0.002
58/62
0.09
±0.01
2.77
0.05
Carbon Tetrachloride
0.000006
0.1
62/62
0.62
±0.03
3.69
0.01
/?-Dichlorobcnzcnc
0.000011
0.8
49/62
0.07
±0.01
0.76
<0.01
1,2 -Dichloroethane
0.000026
2.4
54/62
0.08
±0.01
2.20
<0.01
Ethylbenzene
0.0000025
1
62/62
0.34
±0.05
0.86
<0.01
Formaldehyde
0.000013
0.0098
62/62
3.41
±0.39
44.38
0.35
Hexachloro -1,3 -butadiene
0.000022
0.09
17/62
0.02
±0.01
0.47
<0.01
Arsenic (TSP)a
0.0043
0.000015
58/58
0.67
±0.09
2.87
0.04
Riverside, Tulsa, Oklahoma - TROK
Acetaldehyde
0.0000022
0.009
61/61
1.73
±0.16
3.81
0.19
Benzene
0.0000078
0.03
61/61
0.80
±0.07
6.25
0.03
1,3-Butadiene
0.00003
0.002
60/61
0.07
±0.01
2.23
0.04
Carbon Tetrachloride
0.000006
0.1
61/61
0.62
±0.02
3.73
0.01
/?-Dichlorobcnzcnc
0.000011
0.8
47/61
0.06
±0.01
0.64
<0.01
1,2 -Dichloroethane
0.000026
2.4
50/61
0.08
±0.01
2.06
<0.01
Ethylbenzene
0.0000025
1
61/61
0.37
±0.05
0.92
<0.01
Formaldehyde
0.000013
0.0098
61/61
2.78
±0.36
36.20
0.28
Hexachloro -1,3 -butadiene
0.000022
0.09
18/61
0.02
±0.01
0.52
<0.01
Arsenic (TSP)a
0.0043
0.000015
59/59
0.77
±0.12
3.32
0.05
— = 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.
18-79
-------
Table 18-5. Risk Approximations for the Oklahoma Monitoring Sites (Continued)
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Oklahoma City, Oklahoma - OCOK
Acetaldehyde
0.0000022
0.009
60/60
1.73
±0.20
3.81
0.19
Benzene
0.0000078
0.03
60/60
0.65
±0.06
5.07
0.02
1.3 -Butadiene
0.00003
0.002
46/60
0.04
±0.01
1.16
0.02
Carbon Tetrachloride
0.000006
0.1
60/60
0.61
±0.03
3.68
0.01
1,2 -Dichloroethane
0.000026
2.4
53/60
0.07
±0.01
1.85
<0.01
Formaldehyde
0.000013
0.0098
60/60
2.63
±0.38
34.18
0.27
Arsenic (TSP)a
0.0043
0.000015
59/59
0.48
±0.06
2.08
0.03
Yukon, Oklahoma - YUOK
Acetaldehyde
0.0000022
0.009
61/61
1.80
±0.27
3.95
0.20
Benzene
0.0000078
0.03
61/61
0.56
±0.04
4.33
0.02
1.3 -Butadiene
0.00003
0.002
52/61
0.05
±0.01
1.38
0.02
Carbon Tetrachloride
0.000006
0.1
61/61
0.63
±0.02
3.81
0.01
1,2 -Dichloroethane
0.000026
2.4
55/61
0.08
±0.01
2.02
<0.01
Formaldehyde
0.000013
0.0098
61/61
2.92
±0.36
38.00
0.30
Hexachloro -1,3 -butadiene
0.000022
0.09
15/61
0.02
±0.01
0.37
<0.01
Arsenic (TSP)a
0.0043
0.000015
61/61
0.44
±0.05
1.91
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.
18-80
-------
18.5.2 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 18-6 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
that have cancer toxicity factors. Table 18-6 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 18-6 provides the pollutants with the highest cancer risk approximations (in-a-million) for
each site, as presented in Table 18-5. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 18-6. Table 18-7 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 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 18.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
18-81
-------
Table 18-6. 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
642.74
Benzene
5.01E-03
Formaldehyde
38.30
Ethylbenzene
397.71
Hexavalent Chromium
4.29E-03
Benzene
8.07
Formaldehyde
314.78
Formaldehyde
4.09E-03
Acetaldehyde
4.33
Acetaldehyde
183.16
1,3-Butadiene
2.69E-03
Carbon Tetrachloride
3.64
1.3 -Butadiene
89.52
Naphthalene
1.08E-03
Arsenic (TSP)
3.26
T etrachloroethylene
54.93
Ethylbenzene
9.94E-04
1,3-Butadiene
2.30
Naphthalene
31.71
POM, Group 2b
5.18E-04
1,2-Dichloroethane
2.27
T richloroethylene
16.89
POM, Group 2d
4.29E-04
Nickel (TSP)
1.08
Dichloromethane
8.60
Acetaldehyde
4.03E-04
Ethylbenzene
0.99
POM, Group 2b
5.89
Nickel, PM
3.15E-04
/j-Dichlorobcnzcne
0.82
Fire Station, Tulsa, Oklahoma (Tulsa County) - TMOK
Benzene
642.74
Benzene
5.01E-03
Formaldehyde
44.38
Ethylbenzene
397.71
Hexavalent Chromium
4.29E-03
Benzene
6.32
Formaldehyde
314.78
Formaldehyde
4.09E-03
Acetaldehyde
3.98
Acetaldehyde
183.16
1,3-Butadiene
2.69E-03
Carbon Tetrachloride
3.69
1.3 -Butadiene
89.52
Naphthalene
1.08E-03
Arsenic (TSP)
2.87
T etrachloroethylene
54.93
Ethylbenzene
9.94E-04
1,3-Butadiene
2.77
Naphthalene
31.71
POM, Group 2b
5.18E-04
1,2-Dichloroethane
2.20
T richloroethylene
16.89
POM, Group 2d
4.29E-04
Ethylbenzene
0.86
Dichloromethane
8.60
Acetaldehyde
4.03E-04
/j-Dichlorobcnzcne
0.76
POM, Group 2b
5.89
Nickel, PM
3.15E-04
Hexachloro-1,3 -butadiene
0.47
-------
Table 18-6. 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
642.74
Benzene
5.01E-03
Formaldehyde
36.20
Ethylbenzene
397.71
Hexavalent Chromium
4.29E-03
Benzene
6.25
Formaldehyde
314.78
Formaldehyde
4.09E-03
Acetaldehyde
3.81
Acetaldehyde
183.16
1,3-Butadiene
2.69E-03
Carbon Tetrachloride
3.73
1.3 -Butadiene
89.52
Naphthalene
1.08E-03
Arsenic (TSP)
3.32
T etrachloroethylene
54.93
Ethylbenzene
9.94E-04
1,3-Butadiene
2.23
Naphthalene
31.71
POM, Group 2b
5.18E-04
1,2-Dichloroethane
2.06
T richloroethylene
16.89
POM, Group 2d
4.29E-04
Ethylbenzene
0.92
Dichloromethane
8.60
Acetaldehyde
4.03E-04
/j-Dichlorobcnzcne
0.64
POM, Group 2b
5.89
Nickel, PM
3.15E-04
Hexachloro-1,3 -butadiene
0.52
Oklahoma City, Oklahoma (Oklahoma County) - OCOK
Benzene
469.97
Benzene
3.67E-03
Formaldehyde
34.18
Ethylbenzene
297.38
Formaldehyde
3.63E-03
Benzene
5.07
Formaldehyde
279.17
1,3-Butadiene
1.78E-03
Acetaldehyde
3.81
Acetaldehyde
149.46
Hexavalent Chromium
9.52E-04
Carbon Tetrachloride
3.68
1.3 -Butadiene
59.23
Naphthalene
8.47E-04
Arsenic (TSP)
2.08
T etrachloroethylene
48.47
Ethylbenzene
7.43E-04
1,2-Dichloroethane
1.85
Naphthalene
24.91
POM, Group 2b
4.40E-04
1,3-Butadiene
1.16
Dichloromethane
14.77
POM, Group 2d
3.52E-04
POM, Group 2b
5.01
Acetaldehyde
3.29E-04
POM, Group 2d
3.99
Arsenic, PM
2.40E-04
-------
Table 18-6. 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)
Yukon, Oklahoma (Canadian County) - YUOK
Formaldehyde
153.30
Formaldehyde
1.99E-03
Formaldehyde
38.00
Benzene
69.83
Benzene
5.45E-04
Benzene
4.33
Acetaldehyde
50.89
1,3-Butadiene
3.53E-04
Acetaldehyde
3.95
Ethylbenzene
34.53
Naphthalene
1.61E-04
Carbon Tetrachloride
3.81
1.3 -Butadiene
11.75
Acetaldehyde
1.12E-04
1,2-Dichloroethane
2.02
Naphthalene
4.72
Ethylbenzene
8.63E-05
Arsenic (TSP)
1.91
T etrachloroethylene
2.35
POM, Group 2b
8.61E-05
1,3-Butadiene
1.38
POM, Group 2b
0.98
POM, Group 2d
7.69E-05
Hexachloro-1,3 -butadiene
0.37
Dichloromethane
0.88
POM, Group 5a
5.54E-05
POM, Group 2d
0.87
Arsenic, PM
2.78E-05
-------
Table 18-7. 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
2,096.03
Acrolein
869,130.16
Formaldehyde
0.30
Xylenes
1,502.33
1,3-Butadiene
44,761.98
Acetaldehyde
0.22
Hexane
862.22
Formaldehyde
32,120.04
Manganese (TSP)
0.08
Benzene
642.74
Benzene
21,424.75
Arsenic (TSP)
0.05
Ethylbenzene
397.71
Acetaldehyde
20,350.75
1,3-Butadiene
0.04
Methanol
360.45
Xylenes
15,023.31
Benzene
0.03
Formaldehyde
314.78
Naphthalene
10,570.68
Nickel (TSP)
0.02
Acetaldehyde
183.16
T richloroethy lene
8,445.87
Carbon Tetrachloride
0.01
Ethylene glycol
120.05
Nickel, PM
7,292.24
Ethylbenzene
<0.01
1.3 -Butadiene
89.52
Lead, PM
5,904.21
Hexachloro-1,3 -butadiene
<0.01
Fire Station, Tulsa, Oklahoma (Tulsa County) - TMOK
Toluene
2,096.03
Acrolein
869,130.16
Formaldehyde
0.35
Xylenes
1,502.33
1,3-Butadiene
44,761.98
Acetaldehyde
0.20
Hexane
862.22
Formaldehyde
32,120.04
1,3-Butadiene
0.05
Benzene
642.74
Benzene
21,424.75
Arsenic (TSP)
0.04
Ethylbenzene
397.71
Acetaldehyde
20,350.75
Benzene
0.03
Methanol
360.45
Xylenes
15,023.31
Carbon Tetrachloride
0.01
Formaldehyde
314.78
Naphthalene
10,570.68
Ethylbenzene
<0.01
Acetaldehyde
183.16
T richloroethy lene
8,445.87
Hexachloro-1,3 -butadiene
<0.01
Ethylene glycol
120.05
Nickel, PM
7,292.24
/?-Dichlorobcnzcnc
<0.01
1,3-Butadiene
89.52
Lead, PM
5,904.21
1,2-Dichloroethane
<0.01
-------
Table 18-7. 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
2,096.03
Acrolein
869,130.16
Formaldehyde
0.28
Xylenes
1,502.33
1,3-Butadiene
44,761.98
Acetaldehyde
0.19
Hexane
862.22
Formaldehyde
32,120.04
Arsenic (TSP)
0.05
Benzene
642.74
Benzene
21,424.75
1,3-Butadiene
0.04
Ethylbenzene
397.71
Acetaldehyde
20,350.75
Benzene
0.03
Methanol
360.45
Xylenes
15,023.31
Carbon Tetrachloride
0.01
Formaldehyde
314.78
Naphthalene
10,570.68
Ethylbenzene
<0.01
Acetaldehyde
183.16
T richloroethylene
8,445.87
Hexachloro-1,3 -butadiene
<0.01
Ethylene glycol
120.05
Nickel, PM
7,292.24
/?-Dichlorobcnzcnc
<0.01
1.3 -Butadiene
89.52
Lead, PM
5,904.21
1,2-Dichloroethane
<0.01
Oklahoma City, Oklahoma (Oklahoma County) - OCOK
Toluene
1,716.89
Acrolein
825,550.98
Formaldehyde
0.27
Xylenes
1,179.06
1,3-Butadiene
29,617.41
Acetaldehyde
0.19
Hexane
800.82
Formaldehyde
28,486.41
Arsenic (TSP)
0.03
Benzene
469.97
Acetaldehyde
16,606.84
Benzene
0.02
Methanol
444.71
Benzene
15,665.51
1,3-Butadiene
0.02
Ethylbenzene
297.38
Xylenes
11,790.55
Carbon Tetrachloride
0.01
Formaldehyde
279.17
Naphthalene
8,303.68
1,2-Dichloroethane
<0.01
Ethylene glycol
202.30
Arsenic, PM
3,725.81
Acetaldehyde
149.46
Nickel, PM
3,115.38
Methyl isobutyl ketone
71.17
Propionaldehyde
2,411.31
-------
Table 18-7. 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)
Yukon, Oklahoma (Canadian County) - YUOK
Xylenes
270.61
Acrolein
960,699.47
Formaldehyde
0.30
Toluene
218.04
Formaldehyde
15,642.59
Acetaldehyde
0.20
Formaldehyde
153.30
1.3 -Butadiene
5,876.61
Arsenic (TSP)
0.03
Hexane
119.27
Acetaldehyde
5,654.44
1,3-Butadiene
0.02
Methanol
80.71
Xylenes
2,706.07
Benzene
0.02
Benzene
69.83
Benzene
2,327.52
Carbon Tetrachloride
0.01
Acetaldehyde
50.89
Naphthalene
1,574.09
Hexachloro-1,3 -butadiene
<0.01
Ethylbenzene
34.53
Cyanide Compounds, gas
1,510.86
1,2-Dichloroethane
<0.01
Ethylene glycol
22.94
Lead, PM
1,020.34
Acrolein
19.21
Arsenic, PM
430.77
-------
Observations from Table 18-6 include the following:
• Benzene is the highest emitted pollutant with a cancer URE in Tulsa and Oklahoma
Counties, followed by ethylbenzene and formaldehyde. The highest emitted
pollutants in Canadian County are formaldehyde, benzene, and acetaldehyde. The
quantity of emissions is highest in Tulsa County and lowest in Canadian County.
• The pollutant with the highest toxicity-weighted emissions (of the pollutants with
cancer UREs) for Tulsa County is benzene, followed by hexavalent chromium and
formaldehyde. The pollutant with the highest toxicity-weighted emissions (of the
pollutants with cancer UREs) for Oklahoma County is also benzene, followed by
formaldehyde and 1,3-butadiene. The pollutant with the highest toxicity-weighted
emissions (of the pollutants with cancer UREs) for Canadian County is 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. Eight of the highest
emitted pollutants in Canadian County also have the highest toxicity-weighted
emissions.
• 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 and 1,3-butadiene
also appear on all three lists for each site. Ethylbenzene is also a pollutant of interest
for all three Tulsa sites and appears on both emissions-based lists.
• Nickel is a pollutant of interest for TOOK and has one of the higher cancer risk
approximations for this site. Nickel has the 10th highest toxicity-weighted emissions
for Tulsa County but is not among the highest emitted (its emissions rank 12th).
Arsenic is a pollutant of interest for each Oklahoma site. Although this pollutant has
one of the higher cancer risk approximations for TOOK, TMOK, and TROK, arsenic
is not on either emissions-based list for Tulsa County (ranking 24th for its emissions
and 11th for its toxicity-weighted emissions). For both Oklahoma and Canadian
Counties, arsenic ranks 10th for its toxicity-weighted emissions, and 24th and 27th
for its total emissions, respectively.
• Carbon tetrachloride and 1,2-dichloroethane are pollutants of interest for each site and
have one of the higher cancer risk approximations for each site but do not appear on
either emissions-based site. />Dichlorobenzene is a pollutant of interest for all three
Tulsa sites and appears on neither emissions-based list. Hexachloro-l,3-butadiene is
another pollutant of interest for several sites but appears on neither emissions-based
list.
• Naphthalene and several POM Groups appear in Table 18-6 for quantity emitted and
toxicity-weighted emissions. PAHs were not sampled for under the NMP at the
Oklahoma sites.
18-88
-------
Observations from Table 18-7 include the following:
• Toluene and xylenes are the highest emitted pollutants with noncancer RfCs in Tulsa
and Oklahoma Counties, while the order was reversed for Canadian County.
Emissions were generally highest in Tulsa County and lowest in Canadian County.
• Acrolein is the pollutant with the highest toxicity-weighted emissions (of the
pollutants with noncancer RfCs) for all three counties. Acrolein has the highest
toxicity-weighted emissions for almost all counties with NMP sites but appears
among the highest emitted for only two. Canadian County is one of those counties,
with acrolein ranking 10th among those with the highest emissions. Compared to
other counties with NMP sites, Canadian County's acrolein emissions are not
exceedingly high (19.21 tpy), but are the 11th highest for counties with NMP sites
and are slightly higher than the emissions for Tulsa County (17.38 tpy) and Oklahoma
County (16.51 tpy). Acrolein was sampled for at all of the 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 Oklahoma County also have the highest
toxicity-weighted emissions; five of the highest emitted pollutants in Tulsa County
and Canadian County also have the highest toxicity-weighted emissions. Although
toluene is one of, if not the highest emitted pollutant in all three 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. These pollutants appear on both emissions-based lists for
each county. Benzene also appears on all three lists for each site. This is also true for
1,3-butadiene for the Tulsa sites. 1,3-Butadiene has one of the highest noncancer
hazard approximations for OCOK and YUOK, and has some of the highest toxicity-
weighted emissions for their counties, but is not one of the 10 highest emitted
pollutants in either county (but is just outside the list at 11th highest for each county).
• Several metals appear among the pollutants with the highest toxicity-weighted
emissions for each county but no metals are listed among the highest emitted
pollutants for any of the three counties. This speaks to the relative toxicity of the
speciated metals.
18.6 Summary of the 2014 Monitoring Data for the Oklahoma Monitoring Sites
Results from several of the data analyses described in this section include the following:
~~~ Sixteen pollutants failed at least one screen for TOOK; 15 pollutants failed screens
for TMOK; 14 pollutants failed screens for TROK; 14 pollutants failed screens for
OCOK; and 13 pollutants failed screens for YUOK
~~~ Formaldehyde and acetaldehyde had the highest annual average concentrations for
each site. Concentrations of formaldehyde tended to be higher during the warmer
months of the year.
18-89
-------
After several years of increasing, concentrations of acetaldehyde, ethylbenzene, and
manganese decreased at TOOK after 2012. Other pollutants exhibit this trend as well
but the difference is less significant. Benzene concentrations measured at TOOK have
been decreasing over the last few years. Benzene, acetaldehyde, and ethylbenzene
concentrations have also been decreasing at TMOK and concentrations of the
acetaldehyde andformaldehyde have been decreasing at OCOK In addition, the
detection rates of 1,2-dichloroethane andhexachloro-1,3-butadiene have been
increasing at TOOK and TMOK over the last few years of sampling, particularly for
1,2-dichloroethane. This is also true for 1,2-dichloroethane measurements at OCOK.
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.
18-90
-------
19.0 Site in Rhode Island
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS site in Rhode Island, and integrates these concentrations
with emissions, meteorological, and risk information. Data generated by sources other than ERG
are not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) 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 geographical
and physical information about the location of the site and the surrounding area. 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 PRRI monitoring site is located in south Providence. Figure 19-1 is 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 2011 NEI for point sources, version 2. 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 in order to
emphasize emissions sources within the boundary. Table 19-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates.
19-1
-------
Figure 19-1. Providence, Rhode Island (PRR1) Monitoring Site
-------
Figure 19-2. NET Point Sources Located Within 10 Miles of PRRI
71°35'0"W 71°30'0"W 71 C25'0"W 71°20,0"W 71°15'0"W 71°10'0"W
MASSACHUSETTS
\ Providence
\ County
Bristol
County
Providence
Harbor
Kent
County
^^—i
71°40'0"W 71°35'0"W 71"30'0"W 71"25'0"W 71°20'0"W
Note: Due to facility density and collocation, the total facilities
nd displayed may not represent all facilities within the area of interest.
County boundary
Source Category Group (No. of Facilities)
*
Airport/Airline/Airport Support Operations (11)
A
Foundries, Non-ferrous (1)
CD
Pharmaceutical Manufacturing (2)
§
Asphalt Production/Hot Mix Asphalt Plant (8)
nt
Glass Plant (1)
R
Plastic, Resin, or Rubber Products Plant (18)
0
Auto Body Shop/Painters/Automotive Stores (25)
>
Hotels/Motels/Lodging (1)
t
Printing, Coating & Dyeing of Fabrics Facility (2)
B
Bulk Terminals/Bulk Plants (5)
*
Industrial Machinery or Equipment Plant (11)
P
Printing/Publishing/Paper Product Manufacturing Facility (16)
C
Chemical Manufacturing Facility (10)
o
Institutional (school, hospital, prison, etc.) (25)
A
Ship/Boat Manufacturing or Repair Facility (4)
1
Compressor Station (2)
A
Landfill (1)
TT
Telecommunications/Radio Facility (4)
®
Dry Cleaning Facility (41)
Metals Processing/Fabrication Facility (24)
Testing Laboratories (1)
e
Electrical Equipment Manufacturing Facility (18)
X
Mine/Quarry/Mineral Processing Facility (2)
T
Textile, Yam, or Carpet Rant (9)
i
Electricity Generation via Combustion (4)
Miscellaneous Commercial/Industrial Facility (75)
¦rfV
Truck/Bus/Transportation Operations (1)
E
Electroplating, Plating, Polishing, Anodizing, and Coloring (25)
•
Oil and/or Gas Production (1)
¥
V\festewater Treatment Facility (1)
F
Food Processing/Agriculture Facility (9)
~
Paint and Coating Manufacturing Facility (3)
w
Vtoodwork, Furniture, Millwork & Wfood Preserving Facility (5)
I
Foundries, Iron and Steel (1)
IX
PRRI NATTS site O 10 mile radius
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
136,800
1-95 near 1-195
1AADT reflects 2009 data (RI DOT, 2009)
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, entering downtown Providence. The Providence Harbor is just on the
other side of 1-95 and can be seen 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; institutions (such as
schools, prisons, and hospitals); metals processing and fabrication facilities; electroplating,
plating, polishing, anodizing, and coloring facilities; and auto body shops, painters, and
automotive stores. Sources within one-half mile of PRRI include several hospitals, a heliport at a
hospital, a bulk terminal/bulk plant, an electroplating, plating, polishing, anodizing, and coloring
facility, and a facility that falls into the miscellaneous commercial and industrial source category.
In addition to providing city, county, CBSA, 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 effect concentrations measured at a given monitoring site. The traffic
volume experienced near PRRI is greater than 100,000 and is the sixth 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 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
site in Rhode Island on sample days, as well as over the course of the year.
19.2.1 Meteorological Summary
In order to provide an overview of the meteorological conditions experienced at each
monitoring site, hourly meteorological data for 2014 were retrieved and sample day and full-year
averages developed for temperature, dew point, relative humidity, pressure, and wind speed.
Weather data from the actual monitoring site(s) were obtained from AQS, where available. If
site-specific weather data were not available in AQS, then data were obtained from NCDC for
19-5
-------
the NWS weather station located closest to the monitoring site(s), as described in Section 3.4.2.
For PRRI, site-specific data were available all the parameters except dew point temperature and
sea level pressure. Data for these parameters were obtained from the NWS weather station at
T.F. Green Airport (WBAN 14765). The T.F. Green Airport weather station is located 6 miles
south of PRRI. A map showing the distance between the PRRI monitoring site and the closest
NWS weather station is provided in Appendix R. These data were used to determine how
meteorological conditions on sample days vary from conditions experienced throughout the year.
Table 19-2. Average Meteorological Conditions near the Rhode Island Monitoring Site
Average
Type1
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(in Hg)
Average
Station
Pressure
(in Hg)
Prevailing
Wind
Direction
Average
Scalar Wind
Speed
(kt)
Providence, Rhode Island - PRRI2
Sample
Days
51.1
38.9
65.0
30.01
29.94
5.5
(62)
± 1.0
± 1.1
± 1.0
±0.01
±0.01
NW
±0.1
51.3
39.1
65.7
30.01
29.95
5.3
2014
±0.4
±0.4
±0.4
± <0.01
±<0.01
NW
±0.1
'Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
2Temperature, humidity, and wind parameters were measured at PRRI. The remaining information was obtained from the
closest NWS weather station located at T.F. Green International Airport, WBAN 14765.
Table 19-2 presents average temperature, average dew point temperature, average relative
humidity, average station and sea level pressure, and wind information (average scalar wind
speed and prevailing wind direction) for days on which samples were collected and for all of
2014. Also included in Table 19-2 is the 95 percent confidence interval for each parameter. As
shown in Table 19-2, average meteorological conditions on sample days were representative of
average weather conditions experienced throughout the year near PRRI.
19.2.2 Wind Rose Comparison
Hourly surface wind data were also uploaded into a wind rose software program to
produce customized wind roses, as described in Section 3.4.2. A wind rose shows the frequency
of wind directions using "petals" positioned around a 16-point compass, and uses different colors
to represent wind speeds. Figure 19-3 presents two wind roses for the PRRI monitoring site. The
first is a wind rose representing wind observations for all of 2014 and the second is a wind rose
representing wind observations for days on which samples were collected in 2014. These are
19-6
-------
used to identify the predominant wind speed and direction for 2014 and to determine if wind
observations on sample days were representative of conditions experienced over the entire year.
Figure 19-3. Wind Roses for the Wind Data Collected at PRRI
2014 Wind Rose Sample Day Wind Rose
f 6%,
6%,
t3% i
y j
: EAST
jWESTi ™
V
: EAST;
WIND SPEED
(Knots)
I I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
H 1
Calms: 0.45%
WIND SPEED
(Knots)
I I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
H 1
Calms: 0.14%
Observations from Figure 19-3 for PRRI include the following:
• The full-year wind rose shows that northwesterly winds were observed most,
accounting for 12 percent of observations. Winds from the western quadrants,
including due north and due south, were often observed more frequently at PRRI than
winds from the eastern quadrants. Calm winds account for less than 1 percent of the
hourly measurements.
• The wind patterns shown on the sample day wind rose are similar to the full-year
wind patterns, as winds from the northwest were observed the most and winds from
the western quadrants were observed more often than those from the eastern
quadrants. Fewer winds from the south-southwest and southwest were observed on
sample days while a higher number of winds from the west-northwest and northwest
were observed at PRRI on sample days.
19-7
-------
19.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for PRRI 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 19-3. 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 19-3. It is important to note which
pollutants were sampled for at each site when reviewing the results of this analysis. Only PAHs
were sampled for at PRRI in 2014 under the NMP.
Table 19-3. Risk-Based Screening Results for the Rhode Island Monitoring Site
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Providence, Rhode Island - PRRI
Naphthalene
0.029
45
58
77.59
97.83
97.83
Benzo(a)pyrene
0.00057
1
58
1.72
2.17
100.00
Total
46
116
39.66
Observations from Table 19-3 include the following:
• Concentrations of two PAHs failed at least one screen for PRRI: naphthalene and
benzo(a)pyrene.
• Concentrations of naphthalene account for 45 of the 46 failed screens, with
benzo(a)pyrene failing a single screen.
• Naphthalene accounts for 98 percent of the total failed screens for PRRI. Thus,
naphthalene is the only pollutant of interest for PRRI.
19-8
-------
19.4 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.
• 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 are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at PRRI are provided in Appendix M.
19.4.1 2014 Concentration Averages
Quarterly and annual concentration averages 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 quarter for a quarterly average to be calculated.
An annual average concentration includes all measured detections and substituted zeros for non-
detects for the entire year of sampling. Annual averages were calculated for pollutants where
three valid quarterly averages could be calculated 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 19-4, 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.
19-9
-------
Table 19-4. Quarterly and Annual Average Concentrations of the Pollutants of
Interest for the Rhode Island Monitoring Site
Pollutant
# of
Measured
Detections
vs.
# >MDL
# of
Samples
1st
Quarter
Average
(ng/m3)
2nd
Quarter
Average
(ng/m3)
3rd
Quarter
Average
(ng/m3)
4th
Quarter
Average
(ng/m3)
Annual
Average
(ng/m3)
Providence, Rhode Island - PRRI
Naphthalene
58/58
58
73.57
±25.32
41.11
± 12.65
51.38
± 12.29
40.75
±9.08
52.06
±8.40
Observations for PRRI from Table 19-4 include the following:
• Naphthalene was detected in all of the valid PAH samples collected at PRRI.
• Concentrations of naphthalene measured at PRRI span an order of magnitude, ranging
from 14.7 ng/m3 to 163 ng/m3.
• The first quarter average concentration of naphthalene is greater than the other
quarterly averages and has a confidence interval two to three times larger than the
confidence intervals associated with the other quarterly average concentrations. This
indicates considerably variability in the measurements. Four of the five naphthalene
concentrations greater than 100 ng/m3 were measured between January and March
(the fifth was measured in September). The range of concentrations measured during
the first quarter is nearly twice the range measured during the other calendar quarters.
The median concentration for the first quarter measurements is approximately 57
ng/m3. The difference between first quarter average concentration and the median
concentration for this quarter is another indicator that the naphthalene concentrations
measured during the first quarter are highly variable.
19.4.2 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 19-4 for PRRI. Figure 19-4 overlays PRRI's minimum, annual average, and maximum
naphthalene concentrations onto the program-level minimum, first quartile, median, average,
third quartile, and maximum concentrations, as described in Section 3.4.3.1, and are discussed
below.
19-10
-------
Figure 19-4. Program vs. Site-Specific Average Naphthalene Concentration
E
0
100
200
300
Concentration {ng/m3)
400
500
600
Program:
Site:
1st Qua rti le
¦
Site Average
o
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
Site Concentration Range
4th Qua rti le
~
Average
i
Figure 19-4 presents the box plot for naphthalene for PRRI and shows the following:
• The maximum naphthalene concentration measured at PRRI is about one-third the
maximum concentration measured at the program-level.
• There were no non-detects of naphthalene measured at PRRI (or across the program).
• The annual average naphthalene concentration for PRRI is similar to the program-
level median concentration. PRRI's annual average concentration of naphthalene is in
the bottom-third of the range compared to other NMP sites sampling PAHs (ranking
14th).
19.4.3 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 3.4.3.2.
PRRI has sampled PAHs under the NMP since 2008. Thus, Figure 19-5 presents the 1-year
statistical metrics for the pollutant 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 Naphthalene Concentrations Measured at PRRI
o
2011
Year
O 5th Percentile — Minimum - Med en - Maximum o 95th Percentile •¦•¦^¦¦¦Average
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 naphthalene 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 maximum naphthalene concentration was measured at PRRI in 2011
(301 ng/m3). In total, 10 naphthalene concentrations greater than 200 ng/m3 have been
measured at PRRI, of which seven were measured in November of any given year. In
fact, the maximum concentration for all years between 2009 and 2013 was measured
in November. Of the 25 naphthalene concentrations greater than 150 ng/m3 measured
at PRRI, more than half (17) were measured during the fourth quarter of any given
year and 22 of these 25 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).
19-12
-------
• The concentrations of naphthalene measured at PRRI have a decreasing trend
between 2012 and 2014. Several of the statistical parameters are at a minimum for
2014, including the 1-year average concentration and the median concentration
(42.20 ng/m3). The median concentration is less than 50 ng/m3 for the first time since
the onset of sampling.
19.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at the PRRI monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4 for
definitions and explanations regarding the various toxicity factors, time frames, and calculations
associated with these risk-based screenings.
19.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Rhode Island monitoring site 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 3.4.3.3 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 19-5, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Table 19-5. Risk Approximations for the Rhode Island Monitoring Site
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections vs.
# of Samples
Annual
Average
(ng/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Providence, Rhode Island - PRRI
Naphthalene
0.000034
0.003
58/58
52.06
±8.40
1.77
0.02
19-13
-------
Observations for PRRI from Table 19-5 include the following:
• Naphthalene has both a cancer URE and a noncancer RfC.
• The cancer risk approximation for naphthalene is 1.77 in-a-million.
• The noncancer hazard approximation for naphthalene is negligible (0.02), indicating
that no adverse noncancer health effects are expected from this individual pollutant.
19.5.2 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 19-6 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
that have cancer toxicity factors. Table 19-6 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 19-6 provides the pollutants of interest with the highest cancer risk approximations (in-a-
million) for PRRI, as presented in Table 19-5. The emissions, toxicity-weighted emissions, and
cancer risk approximations are shown in descending order in Table 19-6. Table 19-7 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 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 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 19.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
19-14
-------
Table 19-6. 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)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Providence, Rhode Island (Providence County) - PRRI
Benzene
196.93
Formaldehyde
1.89E-03
Naphthalene
1.77
Formaldehyde
145.48
Benzene
1.54E-03
Ethylbenzene
94.74
1.3 -Butadiene
9.40E-04
Acetaldehyde
76.01
Naphthalene
5.27E-04
1.3 -Butadiene
31.32
POM, Group 2b
4.00E-04
Tetrachloroethylene
17.48
POM, Group 2d
2.39E-04
Naphthalene
15.50
Ethylbenzene
2.37E-04
Trichloroethylene
6.49
POM, Group 5a
2.30E-04
POM, Group 2b
4.54
Arsenic, PM
1.83E-04
Dichloro methane
4.12
Acetaldehyde
1.67E-04
-------
Table 19-7. 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)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Providence, Rhode Island (Providence County) - PRRI
Toluene
636.76
Acrolein
336,121.99
Naphthalene
0.02
Xylenes
390.67
1.3 -Butadiene
15,660.35
Methanol
386.43
Formaldehyde
14,844.73
Hexane
324.64
Acetaldehyde
8,445.40
Benzene
196.93
Benzene
6,564.42
Formaldehyde
145.48
Naphthalene
5,167.59
Ethylene glycol
130.02
Xylenes
3,906.66
Ethylbenzene
94.74
Nickel, PM
3,332.59
Acetaldehyde
76.01
T richloroethy lene
3,243.04
Methyl isobutyl ketone
41.59
Arsenic, PM
2,840.49
-------
Observations from Table 19-6 include the following:
• Benzene, formaldehyde, and ethylbenzene 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, which is the only pollutant of interest for PRRI, has the seventh highest
emissions and the fourth highest toxicity-weighted emissions for Providence County.
• Several POM Groups appear among the pollutants with the highest toxicity-weighted
emissions for Providence County. POM, Groups 2b and 2d rank fifth and sixth for
their toxicity-weighted emissions, respectively, and POM, Group 2b also ranks ninth
for its quantity emitted. POM, Group 2b includes several PAHs sampled for at PRRI,
although none of these pollutants failed screens.
• POM, Group 5a ranks eighth for toxicity-weighted emissions. POM, Group 5a
includes benzo(a)pyrene, which failed a single screen for PRRI. POM, Group 5a is
not among the highest emitted "pollutants" in Providence County.
Observations from Table 19-7 include the following:
• Toluene, xylenes, and methanol 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, 1,3-butadiene, 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 15th).
19-17
-------
19.6 Summary of the 2014 Monitoring Data for PRRI
Results from several of the data analyses described in this section include the following:
~~~ Naphthalene and benzo(a)pyrene each failed at least one screen for PRRI, with
concentrations of naphthalene accounting for 98 percent of the failed screens. As
such, naphthalene is PRRI's only pollutant of interest.
~~~ Concentrations of naphthalene measured at PRRI span an order of magnitude,
ranging from 14.7 ng/m3 to 163 ng/m3.
~~~ Concentrations of naphthalene measured at PRRI have a decreasing trend over the
last two years.
19-18
-------
20.0 Site in Utah
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS site in Utah, and integrates these concentrations with
emissions, meteorological, and risk information. Data generated by sources other than ERG are
not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) 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 geographical and
physical information about the location of the site and the surrounding area. 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 is 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 2011 NEI for point sources, version 2. 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 in order to
emphasize emissions sources within the boundary. Table 20-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates.
20-1
-------
Figure 20-1. Bountiful, Utah (BTUT) Monitoring Site
.W.l890.Nr
',W«700.S
West
Bountiful
Park
Elbet Ave
kil(e'Cornmbhs.Way<
SJ -Village>Square,RdJ
W Pages Ln
E-1200.S,
£•1250 N
Hillside
¦MapJewood Dr .
Viewmont Df
¦W32S
W 800 N
- *[ Source: USGS -
.Source: NASA, NG'A, USGS
' ®B008 Microsoft Corp.
q geonames.orgjB
K>
O
K>
-------
Figure 20-2. NEI Point Sources Located Within 10 Miles of BTUT
112WW
1ir55'0"W
Davis
County
Salt Lake
County
112#10'0"W
112"5'0"W
¦
111"50'0"W 111 r'45'0"W
Source Category Group (No. of Facilities)
¦f Airport/Airline/Airport Support Operations (8)
'i Asphalt Production/Hot Mix Asphalt Plant (2)
b Bulk Terminals/Bulk Plants (1)
* Electricity Generation via Combustion (2)
-*¦ Industrial Machinery or Equipment Plant (2)
o Institutional (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)
Rail Yard/Rail Line Operations (2)
BTUT NATTS site
O 10 mile radius
County boundary
i ~f~ _
112°5'0"W 112 WW
Legend
"i I 1
1ir55'0"W 111"50'0"W 111 °45'0"W 111o40,0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
- L
Morgan
County
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
130,950
1-15, North of Hwy 89 junction
1AADT reflects 2013 data (UT DOT, 2013)
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 surrounding 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 refineries 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, CBSA, 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 effect concentrations measured at a given monitoring site. The traffic
volume experienced near BTUT is nearly 131,000 and ranks in the top third compared to the
traffic volumes 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 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
site in Utah on sample days, as well as over the course of the year.
20.2.1 Meteorological Summary
In order to provide an overview of the meteorological conditions experienced at each
monitoring site, hourly meteorological data for 2014 were retrieved and sample day and full-year
averages developed for temperature, dew point, relative humidity, pressure, and wind speed.
Weather data from the actual monitoring site(s) were obtained from AQS, where available. If
20-5
-------
site-specific weather data were not available in AQS, then data were obtained from NCDC for
the NWS weather station located closest to the monitoring site(s), as described in Section 3.4.2.
For the Utah site, site-specific data were available for some, but not all, of the parameters in
Table 20-2. For BTUT, temperature, pressure, humidity, and wind information was available in
AQS. Data from the NWS weather station at Salt Lake City International Airport (WBAN
24127) were used for the remaining parameters (sea level pressure and dew point temperature).
The Salt Lake City International Airport weather station is located 9.7 miles south-southwest of
BTUT. A map showing the distance between the monitoring site and the closest NWS weather
station is provided in Appendix R. These data were used to determine how meteorological
conditions on sample days vary from conditions experienced throughout the year.
Table 20-2. Average Meteorological Conditions near the Utah Monitoring Site
Average
Type1
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(in Hg)
Average
Station
Pressure
(in Hg)
Prevailing
Wind
Direction
Average
Scalar Wind
Speed
(kt)
Bountiful, Utah - BTUT
2
Sample
Days
54.2
33.7
49.0
29.96
25.67
3.5
(71)
±0.9
±0.5
± 1.0
±0.01
±0.01
SE
±0.1
53.7
33.3
50.0
29.98
25.68
3.4
2014
±0.4
±0.2
±0.5
± <0.01
±<0.01
SE
±<0.1
'Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
2Temperature, humidity, station pressure, and wind parameters were measured at BTUT. The remaining information was
obtained from the closest NWS weather station located at Salt Lake City International, WBAN 24127.
Table 20-2 presents average temperature, average dew point temperature, average relative
humidity, average station and sea level pressure, and wind information (average scalar wind
speed and prevailing wind direction) for days on which samples were collected and for all of
2014. Also included in Table 20-2 is the 95 percent confidence interval for each parameter.
Average meteorological conditions on sample days at BTUT were representative of average
weather conditions experienced throughout the year.
20-6
-------
20.2.2 Wind Rose Comparison
Hourly surface wind data were also uploaded into a wind rose software program to
produce customized wind roses, as described in Section 3.4.2. A wind rose shows the frequency
of wind directions using "petals" positioned around a 16-point compass, and uses different colors
to represent wind speeds. Figure 20-3 presents two wind roses for the BTUT monitoring site.
The first is a wind rose representing wind observations for all of 2014 and the second is a wind
rose representing wind observations for days on which samples were collected in 2014. These
are used to identify the predominant wind speed and direction for 2014 and to determine if wind
observations on sample days were representative of conditions experienced over the entire year.
Figure 20-3. Wind Roses for the Wind Data Collected at BTUT
2014 Wind Rose Sample Day Wind Rose
ijSqife«4%
WEST
! EAST;
WEST
1 EAST
WIND SPEED
(Knots)
HI >= 22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 0.01%
WIND SPEED
(Knots)
HI >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 0.00%
Observations from Figure 20-3 for BTUT include the following:
• The full-year wind rose shows that winds from the southeast and northwest quadrants
were prevalent at BTUT in 2014, with southeasterly winds observed the most. Winds
from the northeast and southwest quadrants were infrequently observed. Winds were
generally light at BTUT although calm winds were rarely observed. The strongest
wind speeds were most often observed with south-southeasterly and southerly winds.
The wind patterns shown on the sample day wind rose are similar to the full-year
wind patterns, indicating that wind conditions in 2014 were similar to wind
conditions experienced on sample days at BTUT.
20-7
-------
20.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for the Utah
monitoring 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 20-3.
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 20-3. It is
important to note which pollutants each site sampled for when reviewing the results of this
analysis. VOCs, carbonyl compounds, SNMOCs, PAHs, and metals (PMio) were sampled for at
BTUT. BTUT is one of only three NMP sites sampling five suites of pollutants under the NMP
and one of only two NMP sites sampling both SNMOC and VOCs (NBIL is the other).
Table 20-3. 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
58
58
100.00
11.67
11.67
Formaldehyde
0.077
58
58
100.00
11.67
23.34
Benzene
0.13
55
55
100.00
11.07
34.41
Carbon Tetrachloride
0.17
55
55
100.00
11.07
45.47
1,2-Dichloroethane
0.038
54
54
100.00
10.87
56.34
1.3 -Butadiene
0.03
46
50
92.00
9.26
65.59
Arsenic (PMio)
0.00023
42
52
80.77
8.45
74.04
Naphthalene
0.029
41
58
70.69
8.25
82.29
Hexacliloro -1,3 -butadiene
0.045
28
29
96.55
5.63
87.93
Dichloromethane
60
20
55
36.36
4.02
91.95
Ethylbenzene
0.4
11
55
20.00
2.21
94.16
Nickel (PMio)
0.0021
8
57
14.04
1.61
95.77
1,2-Dibromoethane
0.0017
6
6
100.00
1.21
96.98
/?-Dichlorobcnzcnc
0.091
5
26
19.23
1.01
97.99
Propionaldehyde
0.8
5
58
8.62
1.01
98.99
Lead (PMio)
0.015
2
57
3.51
0.40
99.40
Benzo(a)pyrene
0.00057
1
24
4.17
0.20
99.60
Cadmium (PMio)
0.00056
1
57
1.75
0.20
99.80
Chloroprene
0.0021
1
1
100.00
0.20
100.00
Total
497
865
57.46
20-8
-------
Observations from Table 20-3 include the following:
• Concentrations of 19 pollutants failed at least one screen for BTUT; approximately
57 percent of concentrations for these 19 pollutants were greater than their associated
risk screening value (or failed screens). BTUT has the second highest number of
individual pollutants failing screens.
• Concentrations for 12 pollutants contributed to 95 percent of failed screens for BTUT
and therefore were identified as pollutants of interest for this site. These 12 include
two carbonyl compounds, seven VOCs, two PMio metals, and one PAH.
• Acetaldehyde, formaldehyde, benzene and carbon tetrachloride 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.
• Recall from Section 3.2 that 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.
20.4 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.
• 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 are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at BTUT are provided in Appendix J through Appendix N.
20-9
-------
20.4.1 2014 Concentration Averages
Quarterly and annual concentration averages 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 the entire year of sampling. Annual averages were calculated for
pollutants where three valid quarterly averages could be calculated 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 20-4, where
applicable. Note that concentrations of the PAHs and PMio metals are presented in ng/m3 in
Table 20-4 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.
20-10
-------
Table 20-4. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Utah Monitoring Site
# of
Measured
1st
2nd
3rd
4th
Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs.
# >MDL
# of
Samples
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Bountiful, Utah - BTUT
2.39
3.06
3.42
4.27
3.33
Acetaldehyde
58/58
58
±0.44
±0.30
±0.90
±0.56
±0.34
0.86
0.39
0.68
0.98
0.73
Benzene
55/55
55
±0.25
±0.12
±0.20
±0.24
±0.12
0.10
0.03
0.05
0.10
0.07
1.3 -Butadiene
50/48
55
±0.03
±0.02
±0.01
±0.04
±0.01
0.54
0.64
0.60
0.60
0.60
Carbon Tetrachloride
55/55
55
±0.06
±0.03
±0.03
±0.07
±0.03
0.17
0.10
0.09
0.09
0.11
1,2-Dichloroethane
54/54
55
±0.02
±0.01
±0.01
±0.02
±0.01
830.76
493.12
58.37
1.81
314.23
Dichloromethane
55/55
55
± 1442.97
±415.95
±55.20
±0.85
±316.90
0.34
0.17
0.33
0.35
0.30
Ethylbenzene
55/54
55
±0.13
±0.06
±0.12
±0.10
±0.05
3.35
5.46
5.51
8.87
5.92
Formaldehyde
58/58
58
±0.80
±0.55
±0.91
± 1.72
±0.73
0.06
0.04
0.05
0.03
0.05
Hexachloro -1,3 -butadiene
29/0
55
±0.03
±0.02
±0.02
±0.02
±0.01
0.84
0.55
0.82
0.94
0.79
Arsenic (PMi0)a
52/42
57
±0.97
±0.19
±0.22
±0.48
±0.26
50.13
35.34
35.37
51.07
42.98
Naphthalene1
58/58
58
±21.23
±6.48
±5.65
± 13.90
±6.51
1.42
1.25
1.33
1.54
1.38
Nickel (PMi,;,)a
57/57
57
±0.84
±0.26
±0.29
±0.47
±0.24
a Average concentrations provided for the pollutant below the blue line are presented in ng/m3 for ease of viewing.
Observations for BTUT from Table 20-4 include the following:
• The pollutants with the highest annual average concentrations are dichloromethane,
formaldehyde, acetaldehyde, and benzene, consistent with the last several years of
sampling.
• Dichloromethane has the highest annual average concentration for BTUT again for
2014, and is similar to its annual average calculated for 2013. The annual average
concentration for 2014 has a very large confidence interval associated it, indicating
the likely presence of outliers, as do the quarterly average concentrations. A review of
the data shows that concentrations of dichloromethane measured at BTUT in 2014
range from 0.491 |ig/m3 to 8,423 |ig/m3. The maximum concentration of this
pollutant was measured on March 30, 2014 and is one of four dichloromethane
concentrations greater than 1,000 |ig/m3 measured at this site. Fifteen of the 21
dichloromethane concentrations greater than 100 |ig/m3 measured across the program
were measured at BTUT (with the other six measured at GPCO). The median
concentration of dichloromethane for BTUT is 21.10 |ig/m3, which is greater than all
but one of the other NMP sites' annual average dichloromethane concentrations,
20-11
-------
indicating that the statistics for this site are not being thrown off just by one or two
outliers. The four highest dichloromethane concentrations measured at BTUT were
measured between March 30, 2014 and May 11, 2014; concentrations greater than
250 |ig/m3 were not measured after July; and concentrations greater than 100 |ig/m3
were not measured after the third quarter of 2014. In fact, dichloromethane
concentrations greater than 10 |ig/m3 were not measured during the fourth quarter of
2014, while between six and 13 were measured during each of the other calendar
quarters. The quarterly average concentrations shown in Table 20-4 reflect these
variations in concentrations measured each quarter.
• The quarterly average concentrations of formaldehyde show that concentrations of
this pollutant were highest during the fourth quarter of 2014 and lowest during the
first quarter of 2014. The fourth quarter average concentration is more than twice the
first quarter average, with the second and third quarter averages in-between the two.
Formaldehyde concentrations measured at BTUT in 2014 range from 2.01 |ig/m3 to
12.8 |ig/m3. All but one of the nine formaldehyde concentrations greater than
10 |ig/m3 were measured between November and December. Conversely, six of the
seven lowest formaldehyde concentrations were measured at BTUT during the first
quarter of 2014. Similar observations can be made for the quarterly average
concentrations of acetaldehyde. Acetaldehyde concentrations measured at BTUT in
2014 range from 1.37 |ig/m3 to 9.15 |ig/m3. All but one of the 15 acetaldehyde
concentrations greater than 4 |ig/m3 were measured during the second half of the
year, with the majority measured during the fourth quarter (10).
• Concentrations of benzene appear lowest during the second quarter and highest
during the fourth quarter, based on the quarterly average concentrations shown in
Table 20-4. A review of the data shows that concentrations of benzene measured at
BTUT range from 0.131 |ig/m3 to 2.05 |ig/m3, with all seven benzene measurements
less than 0.3 |ig/m3 measured during the second quarter of 2014. In fact, benzene
concentrations greater than 1 |ig/m3 were not measured during the second quarter, the
only quarter for which this is true, with three measured during the first quarter, two
during the third, and seven during the fourth. Concentrations of ethylbenzene also
appear lowest during the second quarter of 2014. Ethylbenzene concentrations greater
than 0.4 |ig/m3 were not measured at BTUT during the second quarter while at least
three concentrations greater than 0.4 |ig/m3 were measured during each of the other
calendar quarters. In addition, all four ethylbenzene measurements less than
0.1 |ig/m3 were measured at BTUT during the second quarter.
• Concentrations of 1,3-butadiene were higher during the first and fourth quarters of
2014, or during the colder months of the year, based on the quarterly averages shown
in Table 20-4. All 14 1,3-butadiene concentrations greater than 0.1 |ig/m3 measured at
BTUT were measured during the first or fourth quarters, specifically in January (2),
February (4), October (1), November (4), or December (3). The maximum
1,3-butadiene concentration (0.266 |ig/m3) was measured on November 19, 2014, the
same day as the maximum benzene concentration. Four of the five non-detects of
1,3-butadiene were measured during the second quarter, with the fifth measured in
December.
20-12
-------
The first quarter average concentration of 1,2-dichloroethane is significantly greater
than the other the three quarterly average concentrations of this pollutant. All seven
I,2-dichloroethane concentrations greater than 0.15 |ig/m3 were measured between
January and March. In addition, only one measurement of this pollutant less than
0.1 |ig/m3 was measured during the first quarter compared to eight, 11, and nine
measured during the following calendar quarters, respectively.
The first quarter average arsenic concentration has a confidence interval larger than
its associated quarterly average, indicating that concentrations measured during this
quarter are highly variable. A review of the data shows that concentrations of arsenic
measured at BTUT range from 0.03 ng/m3 to 5.04 ng/m3, and include five non-
detects. Both the highest and lowest arsenic concentrations were measured at BTUT
during the first quarter. All five non-detects were measured during the first quarter,
with three measured in February and one each in January and March. Additionally,
the two highest concentration (5.04 ng/m3 and 4.86 ng/m3) were measured in January.
These two measurements are the second and third highest arsenic concentrations
measured among NMP sites sampling metals. Three of the seven arsenic
concentrations greater than 3 ng/m3 measured across the program were measured at
BTUT.
Concentrations of nickel measured at BTUT range from 0.27 ng/m3 to 5.76 ng/m3,
with the maximum concentration measured on the same day as the maximum arsenic
concentration (January 17, 2014). BTUT is one of five NMP sites sampling PMio
metals with a nickel concentrations greater than 5 ng/m3. Less than 0.3 ng/m3
separates the quarterly average concentrations of nickel for BTUT but the first quarter
average has a confidence interval that is two to three times greater than the
confidence intervals for the other quarterly average concentrations. The range of
concentrations measured during the first quarter is two to three times the range of
concentrations measured during the other calendar quarters. Although the number of
nickel concentrations greater than 1 ng/m3 measured during the first quarter (6) is less
than the number measured during each of the other quarters (between 9 and 11), the
magnitude of the maximum concentration compared to the other concentrations is
driving the average upward.
Concentrations of naphthalene appear highest during the first and fourth quarters of
the year, and exhibit more variability, based on the quarterly average concentrations
shown in Table 20-4. Concentrations of naphthalene measured at BTUT range from
II.6 ng/m3 to 147 ng/m3, with the maximum concentration of naphthalene measured
on January 17, 2014 (which is the same day as the maximum nickel and arsenic
concentrations were measured). All 11 concentrations of naphthalene greater than
60 ng/m3 were measured at BTUT in January (4), October (2), November (3), or
December (2).
20-13
-------
Tables 4-9 through 4-12 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 appears in Table 4-9 through 4-12 a total of seven times for the program-level
pollutants of interest.
• BTUT is listed for three of the program-level VOC pollutants of interest shown in
Table 4-9. BTUT has the highest annual average concentration for hexachloro-1,3-
butadiene among NMP sites sampling this pollutant. BTUT is the only NMP site for
which more than half of the hexachloro-l,3-butadiene measurements were measured
detections (as opposed to non-detects). BTUT also ranks sixth for 1,2-dichloroethane
and seventh for /;-dichlorobenzene.
• For the fourth year in a row, BTUT has the highest annual average concentration of
formaldehyde among NMP sites sampling carbonyl compounds, as shown in
Table 4-10, with a statistically significant difference shown between the first and
second ranking monitoring sites. BTUT is the only site for which the annual average
concentration is greater than 5 |ig/m3. For the second year in a row, BTUT also ranks
highest for its annual average concentration of acetaldehyde, with the only annual
average concentration greater than 3 |ig/m3 among NMP sites.
• BTUT does not appear in Table 4-11 for PAHs. This site's annual average
concentration of the naphthalene is among the lower averages for sites sampling
PAHs. A similar observation was made in the 2013 report.
• BTUT ranks fifth highest for its annual average concentration of arsenic (PMio), as
shown in Table 4-12. BTUT's annual average concentration ranks sixth highest for
nickel.
20.4.2 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 20-4 for BTUT. Figures 20-4 through 20-15 overlay the site's minimum, annual average,
and maximum concentrations onto the program-level minimum, first quartile, median, average,
third quartile, and maximum concentrations, as described in Section 3.4.3.1, and are discussed
below.
20-14
-------
Figure 20-4. Program vs. Site-Specific Average Acetaldehyde Concentration
O i
KJ 1
0123456789 10
Concentration (ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 20-4 presents the box plot for acetaldehyde for BTUT and shows the following:
• The maximum acetaldehyde concentration measured at BTUT is also the maximum
acetaldehyde concentration measured at the program-level.
• The minimum acetaldehyde concentration measured at BTUT (1.37 |ig/m3) is greater
than the program-level first quartile but less than the program-level median
concentration. BTUT is one of four NMP sites whose minimum acetaldehyde
concentration is greater than 1 |ig/m3.
• The annual average acetaldehyde concentration for BTUT is nearly twice the
program-level average concentration, is greater than the program-level third quartile,
and is the highest annual average concentration among NMP sites sampling carbonyl
compounds.
Figure 20-5. Program vs. Site-Specific Average Arsenic (PMio) Concentration
-o-
Program Max Concentration = 10.1 ng/m3
Concentration {ng/m3]
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 20-5 presents the box plot for arsenic (PMio) for BTUT and shows the following:
• The program-level maximum arsenic (PMio) concentration (10.1 ng/m3) is not shown
directly on the box plot in Figure 20-5 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 6 ng/m3.
20-15
-------
• Although BTUT's maximum arsenic concentration is not the maximum arsenic
concentration measured at the program-level, it is the second highest arsenic
concentration measured across the program.
• The annual average concentration of arsenic calculated for BTUT is greater than the
program-level average concentration and is similar to the program-level third quartile.
Recall from the previous section that BTUT has the fifth highest annual average
concentration of arsenic among NMP sites sampling PMio metals.
• Five non-detects of arsenic were measured at BTUT.
Figure 20-6. Program vs. Site-Specific Average Benzene Concentration
14
Program Max Concentration = 12.4 ng/m3
0 2 4 6 8 10
Concentration (ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 20-6 presents the box plot for benzene for BTUT and shows the following:
• The program-level maximum benzene concentration (12.4 |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 plots has been reduced.
• The maximum benzene concentration measured at BTUT is about one-sixth the
magnitude of the maximum benzene concentration measured across the program.
• The minimum benzene concentration measured across the program was measured at
BTUT (0.131 |ig/m3).
• The annual average benzene concentration for BTUT is similar to the program-level
average concentration (0.74 |ig/m3).
20-16
-------
Figure 20-7. Program vs. Site-Specific Average 1,3-Butadiene Concentration
Program Max Concentration = 5.90 |-ig/m3
,
i i i i
0 0.2 0.4 0.6 0.8 1
Concentration (ng/m3)
Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 20-7 presents the box plot for 1,3-butadiene for BTUT and shows the following:
• Similar to benzene, the program-level maximum 1,3-butadiene concentration
(5.90 |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 to 1 |ig/m3.
• The maximum 1,3-butadiene concentration measured at BTUT is considerably less
than the maximum concentration measured across the program.
• The annual average concentration of 1,3-butadiene for BTUT is just greater than the
program-level median concentration.
• Five non-detects were measured at BTUT.
Figure 20-8. Program vs. Site-Specific Average Carbon Tetrachloride Concentration
Program Max Concentration = 3.06 \±g/m3
0 0.5 1 1.5 2 2.5
Concentration (ng/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 20-8 presents the box plot for carbon tetrachloride for BTUT and shows the
following:
• The scale of the box plot for carbon tetrachloride has also been reduced to allow for
the observation of data points at the lower end of the concentration range. Note that
20-17
-------
the program-level median and average concentrations are similar and plotted nearly
on top of each other.
• The range of carbon tetrachloride concentrations measured at BTUT span about
0.5 |ig/m3.
• The annual average concentration of carbon tetrachloride for BTUT is similar to the
program-level first quartile and is the second-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 20-9. Program vs. Site-Specific Average 1,2-Dichloroethane Concentration
, ¦
Program Max Concentration = 27.4 ng/m3
¦
t 1 1 r
0 0.2 0.4 0.6 0.8 1
Concentration (ng/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 20-9 presents the box plot for 1,2-dichloroethane for BTUT and shows the
following:
• The scale of the box plot in Figure 20-9 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 (27.4 |ig/m3) is
considerably greater than the majority of measurements.
• All of the concentrations of 1,2-dichloroethane measured at BTUT are less than the
program-level average concentration of 0.31 |ig/m3, which is being driven by the
measurements at the upper end of the concentration range.
• The annual average concentration for BTUT is just greater than the program-level
third quartile and ranks fifth highest among NMP sites sampling this pollutant. BTUT
is the only non-Calvert City, Kentucky site for which the annual average
concentration of 1,2-dichloroethane is greater than 0.1 |ig/m3, albeit only slightly.
• A single non-detect of 1,2-dichloroethane was measured at BTUT.
20-18
-------
Figure 20-10. Program vs. Site-Specific Average Dichloromethane Concentration
BTUT Max Concentration = 8,423 (-ig/m3
Program Max Concentration =8,423 |ig/m3
0 50 100 150 200 250 300 350 400 450 500
Concentration (ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 20-10 presents the box plot for dichloromethane for BTUT and shows the
following:
• The maximum dichloromethane concentration across the program was measured at
BTUT (8,423 |ig/m3) and is plotted on the box plot in Figure 20-10, as the box plot
was reduced to 500 |ig/m3 in order to allow both the program-level average
concentration and BTUT's annual average to appear on the figure. While the first,
second, and third quartiles are still not visible on the box plot, this does indicate that a
high percentage of the dichloromethane concentrations measured across the program
fall below the concentration levels shown on the box plot and those measured at
BTUT. The minimum dichloromethane concentration measured at BTUT
(0.49 |ig/m3) is greater than the program-level median concentration (0.44 |ig/m3). As
discussed in the previous section, dichloromethane concentrations measured at BTUT
account for 15 of the 21 measurements greater than 100 |ig/m3 across the program.
The maximum dichloromethane concentration measured at BTUT is more than 15
times greater than the next highest concentration measured at another NMP site.
Concentrations of dichloromethane measured at BTUT typically run high compared
to other NMP sites, but concentrations measured in 2014 are particularly high
compared to past years. This was also true in 2013.
• The program-level average concentration (14.05 |ig/m3) is an order of magnitude
greater than third quartile (0.91 |ig/m3), indicating that while most of the
dichloromethane concentrations measured across the program are less than 1 |ig/m3,
the concentrations at the upper end of the range are driving that program-level
average. BTUT is the only site for which dichloromethane is a pollutant of interest
and has the highest annual average concentration of dichloromethane among sites
sampling this pollutant (the annual average concentration for BTUT is nearly eight
times greater than the next highest annual average for an NMP sites sampling
di chl oromethane).
20-19
-------
Figure 20-11. Program vs. Site-Specific Average Ethylbenzene Concentration
n°
0.5 1
1.5 2
Concentration (ng/m3)
2.5
3
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site: Site Average
o
Site Concentration Range
Figure 20-11 presents the box plot for ethylbenzene for BTUT and shows the following:
• Ethylbenzene concentrations measured at BTUT span about 1 |ig/m3.
• The annual average ethylbenzene concentration for BTUT lies between the program-
level average concentration and the program-level third quartile.
• Non-detects of ethylbenzene were not measured at BTUT or at any of the NMP sites
sampling this pollutant with Method TO-15.
Figure 20-12. Program vs. Site-Specific Average Formaldehyde Concentration
,
O i
{J 1
0 3 6 9 12 15 18 21 24 27
Concentration {[j.g/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 20-12 presents the box plot for formaldehyde for BTUT and shows the following:
• The minimum formaldehyde concentration measured at BTUT (2.01 |ig/m3) is greater
than the program-level first quartile but less than the program-level median
concentration. BTUT is one of only three NMP sites at which the minimum
formaldehyde concentration measured is greater than 2 |ig/m3.
20-20
-------
• The annual average formaldehyde concentration for BTUT is more than two times
greater than the program-level average concentration and, as discussed in the previous
section, is the highest annual average formaldehyde concentration among NMP sites
sampling carbonyl compounds.
• Even though the maximum formaldehyde concentration was not measured at BTUT,
this site has the greatest number of formaldehyde concentrations greater than
10 |ig/m3 among NMP sites sampling carbonyl compounds (nine, compared five for
NBNJ and two or less for five additional sites).
Figure 20-13. Program vs. Site-Specific Average Hexachloro-l,3-butadiene Concentration
BTUT
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Concentration (ng/m3)
Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 20-13 presents the box plot for hexachloro-1,3-butadiene for BTUT and shows the
following:
• The program-level first, second, and third quartiles for hexachloro-1,3-butadiene are
zero and therefore not visible on the box plot.
• Twenty-six non-detects of hexachloro-1,3-butadiene were measured at BTUT, which
is the least number of non-detects among NMP sites sampling this pollutant. None of
the remaining measurements were greater than the MDL for this pollutant.
• While the maximum concentration measured at BTUT (0.139 |ig/m3) is considerably
less than the maximum concentration measured across the program (0.609 |ig/m3), it
is the second highest measurement of this pollutant (although the same concentration
was also measured at one other NMP site).
• The annual average concentration of hexachloro-1,3-butadiene for BTUT is nearly
three times greater than the program-level average concentration (0.018 |ig/m3) and is
the highest annual average concentration among NMP sites sampling this pollutant.
20-21
-------
Figure 20-14. Program vs. Site-Specific Average Naphthalene Concentration
mo
0
100
200
300
Concentration {ng/m3)
400
500
600
Program:
1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site:
Site Average
o
Site Concentration Range
Figure 20-14 presents the box plot for naphthalene for BTUT and shows the following:
• The annual average naphthalene concentration for BTUT falls between the program-
level first quartile (28.90 ng/m3) and the program-level median concentration
(50.70 ng/m3).
• The annual average concentration of naphthalene for BTUT ranks 17th among the 19
sites sampling this pollutant.
Figure 20-15. Program vs. Site-Specific Average Nickel (PMio) Concentration
-
i
{J 1
0 2 4 6 8 10
Concentration {ng/m3)
Progra m: 1st Qua rti 1 e
¦
2nd Quartile 3rd Quartile
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
Figure 20-15 presents the box plot for nickel (PMio) for BTUT and shows the following:
• The maximum concentration of nickel measured at BTUT is the 12th highest nickel
concentration measured across the program.
• The annual average concentration of nickel for BTUT is greater than the program-
level average concentration and just greater than the program-level third quartile.
BTUT's annual average concentration ranks sixth among other NMP sites sampling
PMio metals.
20-22
-------
20.4.3 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 3.4.3.2.
BTUT has sampled carbonyl compounds, VOCs, metals, and SNMOCs under the NMP since
2003 and PAHs since 2008. Thus, Figures 20-16 through 20-28 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.
Figure 20-16. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at BTUT
i—J-H
I
2004 2005 2006 2007
2009
Year
2010 2011 2012 2013
o 5th Percentile
— Minimum
- Median
— Maximum
o 95th Percentile
Observations from Figure 20-16 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 20-16 excludes
data from 2003.
20-23
-------
• 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), and another in 2004 (10.8 |ig/m3).
Acetaldehyde concentrations greater than 10 |ig/m3 have not been measured at BTUT
since 2008.
• 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 nearly tripled from 2012 (11) to 2013 (32). 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. Although all of the
parameters except the maximum concentration exhibit decreases for 2014, the
parameters are still at higher levels than they have been at in several years.
Figure 20-17. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at BTUT
Maximum
Concentration for
2004 is 32-99 ng/m3
o
2004 2005 2006
2009
Year
o 5th Percentile
— Minimum
— Maximum
o 95th Percentile
¦ Aver ege
20-24
-------
Observations from Figure 20-17 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 20-17 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 in 2004.
• Of the 40 highest arsenic concentrations measured at BTUT (those greater than
3 ng/m3), 36 were measured during the colder months of the year, with 18 measured
during the first quarter of the calendar year and 18 measured during the fourth quarter
of the calendar year, suggesting a seasonality in the measurements.
• 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 statistical parameters 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. Although difficult to
discern in Figure 20-17, the median arsenic concentration has increased slightly each
year since 2011.
20-25
-------
Figure 20-18. Yearly Statistical Metrics for Benzene Concentrations Measured at BTUT
I
A.
T
I
2004 2005
5th Percentile
2009
Year
— Minimum
- Median
— Maximum
\-0_\
2
O 95th Percentile
Observations from Figure 20-18 for benzene concentrations measured at BTUT include
the following:
• Sampling for VOCs under the NMP began at BTUT in late July 2003. Because this
represents less than half of the sampling year, Figure 20-18 excludes data from 2003.
• 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 50 of the 54 highest concentrations (those greater than 2.50 |ig/m3) were measured
during the first (28) or fourth (22) 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 2014 (0.73 |ig/m3), the first year only one
benzene concentration greater than 2 |ig/m3 was been measured. The median
concentration is also at a minimum for 2014 (0.65 |ig/m3) and exhibits a similar trend
as the 1-year average, except it did not exhibit the same increase for 2009 as the
1-year average concentration.
20-26
-------
Figure 20-19. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at BTUT
o
2009
Year
5th Percentile
— Minimum
- Median
— Maximum
O 95th Percentile
Observations from Figure 20-19 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 and then the 5th percentile for 2007. 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. There was a
single non-detect of this pollutant in 2012, three in 2013, and five in 2014.
• 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 2011, the 1-year average concentration
hovers around 0.1 |ig/m3, ranging from 0.099 |ig/m3 (2011) to 0.116 |ig/m3 (2006).
The median concentration varies a little more, ranging from 0.044 |ig/m3 (2005) to
0.089 |ig/m3 (2006), although the median concentration varies less for the remaining
years in this period.
20-27
-------
• With the exception of the minimum concentration, all of the statistical parameters
exhibit slight increases for 2012. Although not a significant change, the 1-year
average concentration is at a maximum for 2012 (0.117 |ig/m3),
• A decreasing trend in the 1,3-butadiene concentrations measured at BTUT is shown
between 2012 and 2014, the first year that measurements greater than 0.3 |ig/m3 were
not measured and the 1-year average concentration is at a minimum.
Figure 20-20. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
BTUT
2004 2005 2006 2007 2008 2009 2010 2011
Year
2012 2013
O 5th Percentile
o 95th Percentile
Observations from Figure 20-20 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. After decreasing slightly 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.
20-28
-------
• 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. The difference
between the 5th and 95th percentiles, or the range within which a majority of
concentrations fall, is also at a minimum for 2013.
• Each of the statistical parameters exhibits a slight increase for 2014, with the
exception of the 5th percentile, which did not change.
Figure 20-21. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
BTUT
x.
I
T
N
-O
O
2009
Year
O 5th Percentile
— Minimum
— Median
— Maximum
O 95th Percentile
Observations from Figure 20-21 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 in 2009, seven in 2010, 15 in 2011, 47 in 2012, 37 in 2013, and 54 in 2014.
This explains the increases shown in the 1-year average concentrations (as well as
other statistical parameters) for 2010 through 2014. The first year with a median
concentration greater than zero is 2012. This indicates that there were more measured
detections than non-detects for the first time since the onset of sampling.
20-29
-------
• 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 one-quarter of the 73
1,2-dichloroethane concentrations greater than 0.1 |ig/m3 measured at BTUT since
the onset of sampling. Little change is shown in the central tendency statistics for
2014, during which the largest number of 1,2-dichloroethane concentrations greater
than 0.1 |ig/m3 was measured at BTUT (accounting for the highest percentage thus
far at nearly 40 percent).
Figure 20-22. Yearly Statistical Metrics for Dichloromethane Concentrations Measured at
BTUT
3 Concentrations in
2010 > 1,000 ng/m3
o
1 Concentration in
2011 > 1,000 |ig/m3
3 Concentrations in
2013 > 1,000 ng/m3
4 Concentrations in
2014 > 1,000 ng/m3
~
o
2011 2012 2013
5th Percentile
- Minimum
- Maximum
O 95th Percentile
Observations from Figure 20-22 for dichloromethane concentrations measured at BTUT
include the following:
• Prior to 2008, the maximum concentration of dichloromethane measured at BTUT
was 1.64 |ig/m3 (in 2005). The statistical parameters for the early years of sampling
are provided in the inset in Figure 20-22.
Beginning in 2008, "higher" concentrations of dichloromethane began to be measured
at BTUT. In 2008, the first concentration greater than 100 |ig/m3 was measured
(202 |ig/m3). In 2009, four concentrations greater than 100 |ig/m3 were measured. In
2010, three dichloromethane concentrations greater than 1,000 |ig/m3 were measured,
along with six more greater than 100 |ig/m3. For 2011, one concentration greater than
20-30
-------
1,000 |ig/m3 was measured, along with four more greater than 100 |ig/m3. For 2012,
only one concentration greater than 100 |ig/m3 was measured. For 2013, three
concentrations greater than 1,000 |ig/m3 were measured, along with eight more
greater than 100 |ig/m3. The maximum dichloromethane concentration was measured
at BTUT in 2014 (8,423 |ig/m3) along with three others greater than 1,000 |ig/m3 and
10 others greater than 100 |ig/m3.
• There does not appear to be a pattern in the time of year that these higher
concentrations are measured. Of the 45 concentrations measured at BTUT greater
than 100 |ig/m3, at least one has been measured in each month of the year. January
and September have with the greatest number of these higher measurements (7 each),
with October having the fewest (1). Of the 11 concentrations measured at BTUT
greater than 1,000 |ig/m3, five were measured during the spring months (April, May,
or June), which is the most for any calendar quarter.
• Most of the statistical parameters are at a maximum for 2014. The largest range of
concentrations was measured in 2014 and range from 0.49 |ig/m3 to 8,423 |ig/m3. The
1-year average concentration increased by 35 percent and the median concentration
increased by a factor of three. This indicates that concentrations measured in 2014
were higher overall compared to previous years, as the median concentration is driven
less by outliers than an average may be.
Figure 20-23. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at BTUT
I
.2.
JL
T
Y
2004
2005
o
2009
Year
O 5th Percentile
- Minimum
— Median
- Maximum
O 95th Percentile
20-31
-------
Observations from Figure 20-23 for ethylbenzene concentrations measured at BTUT
include the following:
• The maximum concentration of ethylbenzene measured at BTUT was measured in
2013 (5.53 |ig/m3) and is the only concentration greater than 5 |ig/m3 measured at this
site. In total, only seven concentrations greater than 2 |ig/m3 have been measured at
BTUT (of which three were measured in 2013).
• A steady decreasing trend in the 1-year average concentration is shown from 2004
through 2008, representing just less than a 50 percent decrease (from 0.70 |ig/m3 for
2004 to 0.38 |ig/m3 for 2008). However, most of the change is realized between 2004
and 2006.
• Between 2007 and 2009, little change is shown in the concentrations measured, with
the 1-year average concentrations varying by less than 0.012 |ig/m3.
• Nearly all of the statistical parameters exhibit increases for 2010, particularly the
maximum concentration. However, removing the maximum concentration measured
in 2010 from the data set would result in a 1-year average concentration similar to
those shown for 2007 through 2009. This is also true for 2011.
• The range of ethylbenzene concentrations measured in 2012 is considerably smaller
than the two preceding years, and is the smallest since the onset of sampling at
BTUT. This is followed by the largest range of ethylbenzene concentrations measured
in 2013, with the 1-year average concentration at its highest since 2005. The range
within which the majority of concentrations fall, as indicated by the 5th and 95th
percentiles is also at its largest for 2013, yet the median concentration is at its lowest
since sampling began. Even with the higher concentrations measured, 2013 has the
fewest number of measurements greater than 0.25 |ig/m3.
• The smallest range of ethylbenzene concentrations was measured at BTUT in 2014
and both the 1-year average and median concentrations are at a minimum since the
onset of sampling.
20-32
-------
Figure 20-24. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at BTUT
2004 2005
2006 2007
2008 2009 2010 2011 2012 2013 2014
Year
5th Percentile
— Minimum
- Median
— Maximum
O 95th Percentile
Observations from Figure 20-24 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 as the highest acetaldehyde concentration. This
measurement is more than twice the next highest concentration (19.9 |ig/m3),
measured in 2011. Concentrations greater than 15 |ig/m3 were measured 12 times
between 2004 and 2007, plus three additional times in 2011.
• 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 ran 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, the 1-year average concentration began to decrease, reaching a minimum
for 2008 (2.44 |ig/m3). 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.
20-33
-------
• Little change is shown in the 1-year average concentration between 2011 and 2012
and the range of concentrations measured is smaller for 2012, yet 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 statistics for 2013. The
1-year average concentration nearly doubled and the median concentration increased
by 159 percent from 2012. The number of formaldehyde concentrations greater than
10 |ig/m3 is highest for 2013 (16) and concentrations greater than 5 |ig/m3 account 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 the central tendency statistics for 2014. Although
the range of measurements did not change much, the 1-year average concentration
decreased by more than 2 |ig/m3 and the median concentration decreased by more
than 3 |ig/m3 from 2013. Fewer concentrations at the top of the concentration range
were measured in 2014 (nine formaldehyde concentrations greater than 10 |ig/m3
were measured in 2014 compared to 16 in 2013). At the same time, the number of
formaldehyde concentrations less than 5 |ig/m3 increased from 13 in 2013 to 24 in
2014. Yet, both the 1-year average and median concentrations for 2014 are still
greater than 5 |ig/m3.
20-34
-------
Figure 20-25. Yearly Statistical Metrics for Hexachloro-l,3-butadiene Concentrations
Measured at BTUT
T
O
o
—
-
L -i
~
fi
h *
o
o 5th Percentile
o 95th Percentile
• Average
Observations from Figure 20-25 for hexachloro-l,3-butadiene concentrations measured at
BTUT include the following:
• Two hexachloro-1,3-butadiene 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 three or fewer measured detections in 2006, 2009, 2010,
2012, and 2013. Hexachloro-1,3-butadiene was detected 12 times in 2005 and 29
times in 2014.
• As 2014 is the first year that non-detects account for fewer than half of the
measurements, a significant increase in the central tendency statistics is shown.
20-35
-------
Figure 20-26. Yearly Statistical Metrics for Naphthalene Concentrations Measured at BTUT
Maximum
Observations from Figure 20-26 for naphthalene concentrations measured at BTUT
include the following:
• Although PAH sampling began at BTUT in April 2008, complications with the
sampler lead to a 6-month lapse in sampling until mid-October. Thus, Figure 20-27
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 steady 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.
• Concentrations increased slightly for 2013, with the 95th percentile for 2013 greater
than the maximum concentrations measured for the two previous years.
• The majority of naphthalene concentrations, as indicated by the 5th and 95th
percentiles, fell into their smallest range in 2014, with the 95th percentile less than
20-36
-------
100 ng/m3 for the first time since sampling for this pollutant began at BTUT. Both the
1-year average and median concentrations are at a minimum for 2014.
• Concentrations of naphthalene exhibit seasonality. Of the 47 naphthalene
concentrations greater than 100 ng/m3 measured at BTUT since 2009, all but three
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 (16), November (11), or
December (14). Naphthalene concentrations greater than 100 ng/m3 have not been
measured at BTUT between April and August.
Figure 20-27. Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at BTUT
1 h x L
? "--i & y s y s
-i-l ... <
"
y -
2004 2005 2006 2007
2009
Year
2011 2012 2013 2014
O 95th Percentile
Observations from Figure 20-27 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
2.6 ng/m3 (2010) or up to nearly 30 ng/m3 (2005). This variability is reflected in the
20-37
-------
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 2014 exhibit less variability. The
1-year average concentrations calculated for each year during this period fall on either
side of 1.4 ng/m3, with less than 0.06 ng/m3 separating the three of them. Less than
0.15 ng/m3 separates the median concentrations calculated for each of these years.
20.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at the BTUT monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4 for
definitions and explanations regarding the various toxicity factors, time frames, and calculations
associated with these risk-based screenings.
20.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for BTUT 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 3.4.3.3 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 20-5, where applicable. Cancer risk approximations are presented as
probabilities while the noncancer hazard approximations are ratios and thus, unitless values.
20-38
-------
Table 20-5. Risk Approximations for the Utah Monitoring Site
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Bountiful, Utah - BTUT
Acetaldehyde
0.0000022
0.009
58/58
3.33
±0.34
7.33
0.37
Benzene
0.0000078
0.03
55/55
0.73
±0.12
5.70
0.02
1,3-Butadiene
0.00003
0.002
50/55
0.07
±0.01
2.14
0.04
Carbon Tetrachloride
0.000006
0.1
55/55
0.60
±0.03
3.57
0.01
1,2 -Dichloroethane
0.000026
2.4
54/55
0.11
±0.01
2.82
<0.01
Dichloromethane
0.000000016
0.6
55/55
314.23
±316.90
5.03
0.52
Ethylbenzene
0.0000025
1
55/55
0.30
±0.05
0.75
<0.01
Formaldehyde
0.000013
0.0098
58/58
5.92
±0.73
76.95
0.60
Hexachloro -1,3 -butadiene
0.000022
0.09
29/55
0.05
±0.01
0.99
<0.01
Arsenic (PMi0)a
0.0043
0.000015
52/57
0.79
±0.26
3.38
0.05
Naphthalene1
0.000034
0.003
58/58
42.98
±6.51
1.46
0.01
Nickel (PMi,;,)a
0.00048
0.00009
57/57
1.38
±0.24
0.66
0.02
- = 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.
Observations for BTUT from Table 20-5 include the following:
• The pollutants with the highest annual average concentrations are dichloromethane,
formaldehyde, and acetaldehyde, as discussed in Section 20.4.1.
• The pollutants with the highest cancer risk approximations are formaldehyde,
acetaldehyde, and benzene. The cancer risk approximation for formaldehyde for
BTUT (76.95 in-a-million) is the second highest cancer risk approximation across the
program, and the highest cancer risk approximation for formaldehyde. The remaining
cancer risk approximations calculated for BTUT are all less than 10 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 (0.60), which is the highest noncancer hazard approximation calculated
among the site-specific pollutants of interest with noncancer toxicity factors.
20-39
-------
• Dichloromethane's high annual average concentration for BTUT does not translate
into high risk approximations. This is an indication of the toxicity potential of
dichloromethane concentrations in ambient air.
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 speed and 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, similar to the wind roses presented in Section 20.2.2. Thus,
high concentrations may be shown in relation to the direction of potential emissions sources.
Wind observations collected at BTUT and obtained from AQS 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 3.4.3.3. Figure 20-28 presents the
pollution rose for all 58 formaldehyde concentrations measured at BTUT.
Observations from Figure 20-28 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 samples
days with an average wind direction from the eastern quadrants. Of these, more
measurements are associated with an average wind direction from the southeastern
quadrant. Relatively few measurements were measured on sample days with an
average wind direction from the southwest quadrant.
• For each concentration range shown on the pollution rose, the largest number of
concentrations were associated with average winds from the southeast quadrant.
Among the nine formaldehyde concentrations measured at BTUT greater than
10 |ig/m3 (as indicated by the cluster of blue dots), several were measured on a
sample day with an average wind direction roughly from the east (including those
between 75° and 115°).
• 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.
20-40
-------
• 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 east. If the formaldehyde concentrations are averaged by compass direction using
an 8-point compass, the highest average concentrations are calculated for west and
east. However, the westerly direction only includes three formaldehyde
concentrations while the easterly direction includes 13. Other wind directions, such as
southeast, incorporate many concentrations of varying magnitude.
• 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, 2017).
Figure 20-28. Pollution Rose for Formaldehyde Concentrations Measured at BTUT
360/0
315/
o
\
/ \
/o ° \
oO
O-
270
o
of V
-Qj
4-
\°/"
\ /
/ 'o..
/
O-
X
\ /
225 \
\ 45
Q O
qX°
?y" 8 ^
:dOOo04-l.
o
i10o
VP
o°oo^H
au,°^' o
o
o
i 15 Jig/m3
-i 90
I CO
o
---"o x
QO o\ /
° ° V
/
o
135
0<5 |ig/m3
180
0 5-10 |ig/m3
O > 10 |ig/m3
20-41
-------
20.5.2 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 20-6 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
that have cancer toxicity factors. Table 20-6 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 20-6 provides the pollutants of interest with the highest cancer risk approximations (in-a-
million) for BTUT, as presented in Table 20-5. The emissions, toxicity-weighted emissions, and
cancer risk approximations are shown in descending order in Table 20-6. Table 20-7 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 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 20.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
20-42
-------
Table 20-6. 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)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Bountiful, Utah (Davis County) - BTUT
Benzene
120.37
Benzene
9.39E-04
Formaldehyde
76.95
Formaldehyde
68.30
Formaldehyde
8.88E-04
Acetaldehyde
7.33
Ethylbenzene
67.10
Hexavalent Chromium
6.26E-04
Benzene
5.70
Dichloromethane
46.51
1,3-Butadiene
4.62E-04
Dichloromethane
5.03
Acetaldehyde
41.70
Naphthalene
2.84E-04
Carbon Tetrachloride
3.57
1.3 -Butadiene
15.40
POM, Group 2b
1.79E-04
Arsenic (PMio)
3.38
Naphthalene
8.35
Ethylbenzene
1.68E-04
1,2-Dichloroethane
2.82
T etrachloroethylene
6.26
POM, Group 2d
1.23E-04
1,3-Butadiene
2.14
POM, Group 2b
2.03
POM, Group 5a
9.95E-05
Naphthalene
1.46
POM, Group 2d
1.39
Acetaldehyde
9.17E-05
Hexachloro-1,3 -butadiene
0.99
-------
Table 20-7. 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)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Bountiful, Utah (Davis County) - BTUT
Toluene
539.04
Acrolein
192,602.57
Formaldehyde
0.60
Hexane
370.73
1.3 -Butadiene
7,700.14
Dichloromethane
0.52
Xylenes
286.60
Formaldehyde
6,969.87
Acetaldehyde
0.37
Methanol
205.85
Acetaldehyde
4,633.66
Arsenic (PMio)
0.05
Ethylene glycol
121.88
Benzene
4,012.18
1,3-Butadiene
0.04
Benzene
120.37
Xylenes
2,865.98
Benzene
0.02
Formaldehyde
68.30
Naphthalene
2,782.33
Nickel (PMio)
0.02
Ethylbenzene
67.10
Lead, PM
982.29
Naphthalene
0.01
Methyl isobutyl ketone
51.39
Arsenic, PM
703.33
Carbon Tetrachloride
0.01
Dichloro methane
46.51
Hexane
529.62
Hexachloro-1,3 -butadiene
0.00
-------
Observations from Table 20-6 include the following:
• Benzene, formaldehyde, ethylbenzene, and dichloromethane 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 benzene, formaldehyde, hexavalent chromium, and 1,3-butadiene.
• Eight of the highest emitted pollutants also have the highest toxicity-weighted
emissions in Davis County.
• Formaldehyde, which has the highest cancer risk approximation for BTUT, ranks
second for both emissions-based lists, behind benzene. Acetaldehyde, 1,3-butadiene,
naphthalene, and ethylbenzene also appear on all three lists in Table 20-6.
Dichloromethane, which has the highest annual average concentration and the fourth
highest cancer risk approximation for BTUT, ranks fourth for quantity of emissions in
Davis County but is not among those with the highest toxicity-weighted emissions (it
ranks 22nd). Arsenic, carbon tetrachloride, hexachloro-l,3-butadiene, and
1,2-dichloroethane, the remaining pollutants of interest listed for BTUT, appear on
neither emissions-based list.
• POM, Group 2b is the ninth highest emitted "pollutant" in Davis County and ranks
sixth for toxicity-weighted emissions. POM, Group 2b includes several PAHs
sampled for at BTUT including acenaphthylene, fluoranthene, and perylene. None of
the PAHs included in POM, Group 2b failed screens for BTUT. POM, Group 5a
ranks ninth for its toxicity-weighted emissions. POM, Group 5a includes
benzo(a)pyrene, which failed a single screen for BTUT.
Observations from Table 20-7 include the following:
• Toluene, hexane, and xylenes 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, 1,3-butadiene, 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.
• Formaldehyde, dichloromethane, and acetaldehyde have the highest noncancer hazard
approximations for BTUT (although all are less than 1.0). Formaldehyde and benzene
are the only listed pollutants of interest to appear on both emissions-based lists.
Acetaldehyde, arsenic, 1,3-butadiene, and naphthalene are pollutants of interest for
BTUT that rank among the pollutants with the highest toxicity-weighted emissions in
Davis County but do not appear among those with the highest total emissions.
20-45
-------
Dichloromethane ranks 10th for its quantity emitted in Davis County but does not
appear among those highest toxicity-weighted emissions. Hexachloro-1,3-butadiene,
nickel, and carbon tetrachloride are pollutants of interest for BTUT that do not appear
on either emissions-based list in Table 20-7.
20.6 Summary of the 2014 Monitoring Data for BTUT
Results from several of the data analyses described in this section include the following:
~~~ Nineteen pollutants failed at least one screen for BTUT.
~~~ Dichloromethane had the highest annual average concentration among the pollutants
of interest for BTUT, followed by formaldehyde and acetaldehyde.
~~~ For the fourth year in a row, BTUT has the highest annual average formaldehyde
concentration among NMP sites sampling this pollutant. BTUT also has the highest
annual average concentrations of acetaldehyde and hexachloro-1,3-butadiene among
other NMP sites.
~~~ Concentrations of benzene have an overall decreasing trend at BTUT; the 1-year
average concentration for 2014 is the lowest 1-year average concentration of
benzene calculated since the onset of sampling at BTUT. Concentrations of
1,3-butadiene have also been decreasing in recent years. Concentrations of both
acetaldehyde andformaldehyde increased significantly for 2013 then decreasedfor
2014. The detection rates of both 1,2-dichloroethane and hexachloro-1,3-butadiene is
at a maximum for 2014.
~~~ Formaldehyde has the highest cancer risk approximation among the pollutants of
interest for BTUT (which is the second highest cancer risk approximation across the
program). None of the pollutants of interest have noncancer hazard approximations
greater than an HQ of 1.0.
20-46
-------
21.0 Site in Vermont
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS site in Vermont, and integrates these concentrations with
emissions, meteorological, and risk information. Data generated by sources other than ERG are
not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) 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 geographical and
physical information about the location of the site and the surrounding area. 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 Vermont NATTS site (UNVT) is located in the town of Underhill, in northwest
Vermont, in the Burlington-South Burlington, VT CBSA. Figure 21-1 is the 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 2011 NEI for point sources, version 2. 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 in order 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.
21-1
-------
Figure 21-1. Underbill, Vermont (UNVT) Monitoring Site
-------
Figure 21-2. NEI Point Sources Located Within 10 Miles of UNVT
72°50"0"W
72°45'0"W
73°10'0"W
73°5'0"W
73"0'0"W
72"55'0"W
72"40,0"W
72"35'0"W
i Franklin
\ County
Lamoille
County
Chittenden
County
T\ Qf \ K ® *
Lake _ \
Champlain \
V
\
\
Washington
County
\
^ - *
" \
\
73°20"0"W 73"15'0"W 73°10'0*W 73°5'0"W 73'WW 72°55,0"W 72°50'0"W 72°45'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
Legend
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
1,100
Pleasant Valley Rd, North of Harvey
Rd
1AADT reflects 2011 data forUNVT (CCRPC, 2016)
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,
2016). 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 background site for the region for
trends assessment, standards compliance, and long-range transport assessment.
Most of the emissions sources near UNVT are located to the east and southeast 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. The 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, CBSA, 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 effect concentrations measured at a given monitoring site. The traffic
volume near UNVT is relatively light, with approximately 1,100 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 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
site in Vermont on sample days, as well as over the course of the year.
21.2.1 Meteorological Summary
In order to provide an overview of the meteorological conditions experienced at each
monitoring site, hourly meteorological data for 2014 were retrieved and sample day and full-year
averages developed for temperature, dew point, relative humidity, pressure, and wind speed.
Weather data from the actual monitoring site(s) were obtained from AQS, where available. If
site-specific weather data were not available in AQS, then data were obtained from NCDC for
21-5
-------
the NWS weather station located closest to the monitoring site(s), as described in Section 3.4.2.
For the Vermont site, site-specific data were available for some, but not all, of the parameters in
Table 21-2. For UNVT, temperature, pressure, humidity, and wind information was available in
AQS. Data from the NWS weather station at Morrisville-Stowe State Airport (WBAN 54771)
were used for the remaining parameters (sea level pressure and dew point temperature). The
Morrisville-Stowe State Airport weather station is located 13 miles east of UNVT. In addition,
the UNVT meteorological station was down for two weeks at the end of September through early
October for repairs at the site and experienced a malfunction during the second half of
December. Thus, NWS data was used as a surrogate here as well. A map showing the distance
between the monitoring site and the closest NWS weather station is provided in Appendix R.
These data were used to determine how meteorological conditions on sample days vary from
conditions experienced throughout the year.
Table 21-2. Average Meteorological Conditions near the Vermont Monitoring Site
Average
Type1
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(in Hg)
Average
Station
Pressure
(in Hg)
Prevailing
Wind
Direction
Average
Scalar Wind
Speed
(kt)
Underhill, Vermont - UNVT2
Sample
Days
41.7
31.8
64.1
30.01
28.60
2.2
(69)
± 1.1
± 1.1
±0.8
±0.01
±0.01
NNW
±0.1
43.7
33.3
63.5
30.00
28.59
2.3
2014
±0.4
±0.5
±0.3
±<0.01
±0.01
NNW
±<0.1
'Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
2Temperature, humidity, station pressure, and wind parameters were measured at UNVT. Hie remaining information was
obtained from the closest NWS weather station located at Morrisville-Stowe State Airport, WBAN 54771. In addition, while
the meteorological station at UNVT was down for parts of the year, NWS data was used as a surrogate.
Table 21-2 presents average temperature, average dew point temperature, average
relative humidity, average station and sea level pressure, and wind information (average scalar
wind speed and prevailing wind direction) for days on which samples were collected and for all
of 2014. Also included in Table 21-2 is the 95 percent confidence interval for each parameter.
Average meteorological conditions on sample days at UNVT were fairly representative of
average weather conditions experienced throughout the year, as shown in Table 21-2. The
difference between a full-year average and sample day average is largest for average
temperature. Compared to other NMP sites, the Vermont site experiences some of the coldest
21-6
-------
temperatures, as this site has the lowest average sample day temperature. UNVT also has the
lowest average wind speed.
21.2.2 Wind Rose Comparison
Hourly surface wind data were also uploaded into a wind rose software program to
produce customized wind roses, as described in Section 3.4.2. A wind rose shows the frequency
of wind directions using "petals" positioned around a 16-point compass, and uses different colors
to represent wind speeds. Figure 21-3 presents two wind roses for the UNVT monitoring site.
The first is a wind rose representing wind observations for all of 2014 and the second is a wind
rose representing wind observations for days on which samples were collected in 2014. These
are used to identify the predominant wind speed and direction for 2014 and to determine if wind
observations on sample days were representative of conditions experienced over the entire year.
Figure 21-3. Wind Roses for the Wind Data Collected at UNVT
2014 Wind Rose Sample Day Wind Rose
WIND SPEED
(Knots)
I I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 5.43%
6%
3^ '
EAST.
WIND SPEED
(Knots)
I I >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
¦ 1-4
Calms: 6.61%
21-7
-------
Observations from Figure 21-3 for UNVT include the following:
• The full-year rose shows that light winds were prevalent near UNVT, as winds speeds
greater than 4 knots were infrequently observed. Winds from the north-northwest
were observed the most, although winds from the south-southeast and south account
for a similar percentage of observations (each of these three wind directions were
observed for approximately 11 percent of the wind observations). Winds from the
northeast and southwest quadrants account for relatively few observations.
• The wind patterns shown on the sample day wind rose are similar to those shown on
the full-year wind rose. Light winds prevailed and winds from the north-northwest
account for the highest percentage of wind directions observed (greater than
13 percent). The increase in the percentage of north-northwesterly winds is offset by
slightly fewer south-southwesterly and southerly wind observations on sample days.
• Recall from the previous section that wind sensors were down at UNVT for a portion
of 2014 and NWS data were used as a surrogate for missing data.
21.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for the UNVT
monitoring 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 21-3.
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 21-3. It is
important to note which pollutants were sampled for at each site when reviewing the results of
this analysis. PAHs and metals (PMio) were sampled for under the NMP at UNVT in 2014.
Table 21-3. Risk-Based Screening Results for the Vermont Monitoring Site
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Underhill, Vermont - UNVT
Arsenic (PMio)
0.00023
24
56
42.86
100.00
100.00
Total
24
56
42.86
21-8
-------
Observations from Table 21-3 include the following:
• Arsenic is the only pollutant whose concentrations failed at least one screen for
UNVT. Thus, arsenic is UNVT's only pollutant of interest.
• Approximately 43 percent of arsenic concentrations for were greater than their
associated risk screening value (or failed screens).
• Concentrations of the remaining metals and PAHs sampled for at UNVT did not fail
any screens. UNVT is the only NMP site sampling PAHs for which naphthalene did
not fail any screens.
• It should be noted, however, that the Vermont Department of Environmental
Conservation invalidated all of its nickel and total chromium concentrations for the
second half of 2014. This is due to a contamination issue related to a new weighing
and equilibration chamber at their laboratory, as discussed in Section 2.4.
21.4 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.
• 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 are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at UNVT are provided in Appendices M and N.
21.4.1 2014 Concentration Averages
Quarterly and annual concentration averages 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
21-9
-------
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 the entire year of sampling. Annual averages were calculated for
pollutants where three valid quarterly averages could be calculated 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 21-4, 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.
Table 21-4. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Vermont Monitoring Site
Pollutant
# of
Measured
Detections
vs. # >MDL
# of
Samples
1st
Quarter
Average
(ng/m3)
2nd
Quarter
Average
(ng/m3)
3rd
Quarter
Average
(ng/m3)
4th
Quarter
Average
(ng/m3)
Annual
Average
(ng/m3)
Underhill, Vermont - UNVT
Arsenic (PMio)
56/25
61
0.16
±0.07
0.20
±0.06
0.32
±0.10
0.18
±0.11
0.21
±0.04
Observations from Table 21-4 include the following:
• Arsenic was detected in 56 of the 61 valid PMio metals samples collected at UNVT in
2014. Of these, 25 were greater than the MDL associated with arsenic sampling.
• All of the arsenic concentrations measured at UNVT are less than 1 ng/m3. Arsenic
measurements range from 0.003 ng/m3 to 0.83 ng/m3 and include five non-detects.
Non-detects of arsenic were measured during the first (2) and fourth (3) quarters of
the year.
• Among NMP sites sampling arsenic, UNVT has the lowest annual average
concentration of this pollutant (0.21 ± 0.04 ng/m3).
• Based on the quarterly average concentrations of arsenic calculated for UNVT,
concentrations appear highest during the third quarter. However, the differences are
not statistically significant.
21-10
-------
21.4.2 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 21-4 for UNVT. Figure 21-4 overlays the site's minimum, annual average, and maximum
concentrations onto the program-level minimum, first quartile, median, average, third quartile,
and maximum concentrations, as described in Section 3.4.3.1, and are discussed below.
Figure 21-4. Program vs. Site-Specific Average Arsenic (PMio) Concentration
-
Program Max Concentration = 10.1 ng/m3
i i i i i
0 1 2 3 4 5 6
Concentration {ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 21-4 presents the box plot for arsenic for UNVT and shows the following:
• The program-level maximum arsenic concentration (10.1 ng/m3) is not shown directly
on the box plot in Figure 21-4 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 plots has been reduced.
• The maximum arsenic concentration measured at UNVT (0.83 ng/m3) is just greater
than the program-level third quartile (0.77 ng/m3) and is the second-lowest site-
specific maximum concentration among NMP sites sampling arsenic (PMio).
• The annual average arsenic concentration for UNVT is just less than the program-
level first quartile (25th percentile). As discussed previously, the annual average
concentration of arsenic for UNVT is the lowest annual average arsenic concentration
among NMP sites sampling this pollutant.
• Five of the 33 non-detects of arsenic measured across the program were measured at
UNVT.
21-11
-------
21.4.3 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 3.4.3.2.
UNVT has sampled PMio metals under the NMP since 2008. Thus, Figure 21-5 presents the
annual statistical metrics for the pollutant 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.
Figure 21-5. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at
UNVT
2011
Year
5th Percentile
- Minimum
- Med en
- Maximum
O 95th Percentile
Observations from Figure 21-5 for arsenic concentrations measured at UNVT include the
following:
• Arsenic concentrations greater than 1 ng/m3 have not been measured at UNVT since
the onset of sampling in 2008. The maximum arsenic concentration was measured at
UNVT in 2012 (0.90 ng/m3).
• With the exception of the 95th percentile, each of the statistical parameters exhibits a
decrease from 2008 to 2009 and again for 2010. The minimum concentration
21-12
-------
measured in 2008 is 0.05 ng/m3, which decreased to 0.02 ng/m3 for 2009, and the first
non-detects were measured in 2010 (three). Between three and six non-detects were
measured each year following 2010.
• Overall, a similar range of arsenic concentrations has been measured at UNVT from
year-to-year. The 1-year average concentrations of arsenic for UNVT have changed
little over the years of sampling, ranging from 0.21 ng/m3 (2014) to 0.28 ng/m3
(2013). Likewise, the median concentration has ranged from 0.18 ng/m3 (2010) to
0.26 ng/m3 (2013).
• The change in the 1-year average concentration is largest between 2013 and 2014.
While not statistically significant, this difference can be primarily attributed to
concentrations at the lower end of the concentration range. Concentrations less than
0.1 ng/m3, excluding non-detects, account for only four measurements in 2013
compared to 13 in 2014. The number of non-detects of arsenic measured each year
varied by only one (four in 2013 and five in 2014).
21.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at the Vermont monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4 for
definitions and explanations regarding the various toxicity factors, time frames, and calculations
associated with these risk-based screenings.
21.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Vermont monitoring site 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 3.4.3.3 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 21-5, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
21-13
-------
Table 21-5. Risk Approximations for the Vermont Monitoring Site
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Underhill, Vermont - UNVT
Arsenic (PMio)
0.0043
0.000015
56/61
0.21
±0.04
0.92
0.01
Observations from Table 21-5 include the following:
• The annual average arsenic concentration for UNVT is the lowest annual average for
this pollutant among NMP sites sampling metals.
• The cancer risk approximation for UNVT is less than 1 in-a-million and is the lowest
cancer risk approximation for arsenic among NMP sites (0.92 in-a-million).
• The noncancer hazard approximation for arsenic for UNVT is considerably less than
1.0 (0.01), indicating that no adverse noncancer health effects are expected from this
individual pollutant.
21.5.2 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 21-6 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
that have cancer toxicity factors. Table 21-6 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 21-6 provides the pollutants of interest with the highest cancer risk approximations (in-a-
million) for UNVT, as presented in Table 21-5. The emissions, toxicity-weighted emissions, and
cancer risk approximations are shown in descending order in Table 21-6. Table 21-7 presents
similar information, but is limited to those pollutants with noncancer toxicity factors.
21-14
-------
Table 21-6. 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)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Underhill, Vermont (Chittenden County) - UNVT
Benzene
103.48
Formaldehyde
8.77E-04
Arsenic (PMio)
0.92
Formaldehyde
67.43
Benzene
8.07E-04
Acetaldehyde
37.96
1,3-Butadiene
4.06E-04
Ethylbenzene
37.92
Arsenic, PM
3.13E-04
1.3 -Butadiene
13.53
Naphthalene
2.29E-04
Naphthalene
6.75
POM, Group 2b
1.52E-04
Dichloromethane
2.55
Hexavalent Chromium
1.19E-04
T etrachloroethylene
2.22
POM, Group 5a
1.06E-04
POM, Group 2b
1.73
Nickel, PM
9.73E-05
POM, Group 2d
1.02
Ethylbenzene
9.48E-05
-------
Table 21-7. 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)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Underhill, Vermont (Chittenden County) - UNVT
Toluene
250.92
Acrolein
546,915.43
Arsenic (PMio)
0.01
Xylenes
174.70
Chlorine
12,098.33
Hexane
106.23
Manganese, PM
9,934.56
Benzene
103.48
Formaldehyde
6,880.35
Methanol
90.73
1.3 -Butadiene
6,767.01
Formaldehyde
67.43
Arsenic, PM
4,859.91
Acetaldehyde
37.96
Acetaldehyde
4,218.14
Ethylbenzene
37.92
Benzene
3,449.29
Hydrochloric acid
35.41
Cadmium, PM
2,474.68
Ethylene glycol
31.16
Nickel, PM
2,252.18
-------
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 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 21.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
Observations from Table 21-6 include the following:
• Benzene, formaldehyde, acetaldehyde, and ethylbenzene are the highest emitted
pollutants with cancer UREs in Chittenden County.
• Formaldehyde is the pollutant with the highest toxicity-weighted emissions (of the
pollutants with cancer UREs) in Chittenden County, followed by benzene,
1,3-butadiene, and arsenic (PM).
• Six of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Chittenden County. Benzene and formaldehyde are at or near the top of
the emissions-based lists.
• Arsenic is UNVT's only pollutant of interest for 2014. Arsenic has the fourth highest
toxicity-weighted emissions, but is not one of the highest emitted in Chittenden
County (it ranks 15th). Nickel also appears among the pollutants with the highest
toxicity-weighted emissions, ranking ninth, while its quantity emitted in Chittenden
County rank 12th. Nickel was sampled for at UNVT, but concentrations measured
during the second half of the year were invalidated due to contamination of the filters.
Observations from Table 21-7 include the following:
• Toluene, xylenes, and hexane 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).
• Three of the highest emitted pollutants for Chittenden County also have the highest
toxicity-weighted emissions.
• Four metals sampled for at UNVT appear among the pollutants with the highest
toxicity-weighted emissions, including arsenic, although none appear among the
21-17
-------
highest emitted (manganese ranks 19th and is the highest ranking among the four
metals listed). This speaks to the relative toxicity potential of pollutants emitted in
small quantities.
21.6 Summary of the 2014 Monitoring Data for the Vermont Monitoring Site
Results from several of the data analyses described in this section include the following:
~~~ Metals (PMw) and PAHs were sampledfor at UNVT.
~~~ Concentrations of nickel and total chromium sampled during the second half of 2014
were invalidated due to a filter contamination issue at the Vermont laboratory.
~~~ Arsenic is the only pollutant of interest for UNVT. All of the arsenic concentrations
measured at UNVT in 2014 are less than 1 ng/m3.
~~~ The annual average arsenic concentration for UNVT is the lowest annual average for
this pollutant among NMP sites sampling metals.
21-18
-------
22.0 Site in Virginia
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS site in Virginia, and integrates these concentrations with
emissions, meteorological, and risk information. Data generated by sources other than ERG are
not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) 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 geographical and
physical information about the location of the site and the surrounding area. 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 is 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 2011 NEI for point sources,
version 2. 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 in order to emphasize emissions sources within the
boundary. Table 22-1 provides supplemental geographical information such as land use, location
setting, and locational coordinates.
22-1
-------
Figure 22-1. East Highland Park, Virginia (RIVA) Monitoring Site
-------
Figure 22-2. NEI Point Sources Located Within 10 Miles of RIVA
Hanover
County
Henrico
County
Goochland
"County j
Richmond
City
James
River,
77C30'0*W 77C25'0"W 77C20'0"W 77°15trW
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)
T
Airport/Airline/Airport Support Operations (10)
•>
*
Asphalt Production/Hot Mix Asphalt Plant (1)
1
B
Bulk Terminals/Bulk Plants (6)
R
C
Chemical Manufacturing Facility (2)
P
t
Electricity Generation via Combustion (4)
E
F
Food Processing/Agriculture Facility (1)
X
o
Institutional (school, hospital, prison, etc.) (2)
¦
Landfill (1)
M
®
Metals Processing/Fabrication Facility (1)
W
A
Military Base/National Security Facility (1)
Miscellaneous Commercial/Industrial Facility (1)
Paint and Coating Manufacturing Facility (2)
Plastic, Resin, or Rubber Products Plant (3)
Printing/Publishing/Paper Product Manufacturing Facility (4)
Pulp and Paper Plant (1)
Rail Yard/Rail Line Operations (4)
Testing Laboratories (1)
Tobacco Manufacturing (1)
Woodwork, Furniture, Millwork & Wood Preserving Facility (1)
RIVA NATTS site
O 10 mile radius
County boundary
t .
King William '
County \
Miles
r— ;
77'40'0"W 77°35"0"W
Legend
77°25'0"W
\ Chesterfield ^
\ County \
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
72,000
1-64 at Mechanicsville Turnpike
1AADT reflects 2013 data (VADOT, 2013)
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 to the south and southwest of the site and within the city of Richmond. The sources
closest to RIVA are a metals processing and fabrication facility and a heliport at the Medical
College of Virginia. 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; bulk terminals and bulk plants; printing, publishing, and paper product manufacturers;
rail yard and rail line operations; and facilities generating electricity via combustion.
In addition to providing city, county, CBSA, 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 effect concentrations measured at a given monitoring site. The traffic
volume experienced near RIVA is 72,000, 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 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
site in Virginia on sample days, as well as over the course of the year.
22-5
-------
22.2.1 Meteorological Summary
In order to provide an overview of the meteorological conditions experienced at each
monitoring site, hourly meteorological data for 2014 were retrieved and sample day and full-year
averages developed for temperature, dew point, relative humidity, pressure, and wind speed.
Weather data from the actual monitoring site(s) were obtained from AQS, where available. If
site-specific weather data were not available in AQS, then data were obtained from NCDC for
the NWS weather station located closest to the monitoring site(s), as described in Section 3.4.2.
For the Virginia site, site-specific data were available for some, but not all, of the parameters in
Table 22-2. For RIVA, temperature, pressure, and wind information was available in AQS. Data
from the NWS weather station at Richmond International Airport (WBAN 13740) were used for
the remaining parameters (relative humidity, sea level pressure, and dew point temperature). The
Richmond International Airport weather station is located 5.7 miles southeast of RIVA. A map
showing the distance between the monitoring site and the closest NWS weather station is
provided in Appendix R. These data were used to determine how meteorological conditions on
sample days vary from conditions experienced throughout the year.
Table 22-2. Average Meteorological Conditions near the Virginia Monitoring Site
Average
Average
Average
Average
Average
Average
Dew Point
Relative
Sea Level
Station
Prevailing
Scalar Wind
Average
Temperature
Temperature
Humidity
Pressure
Pressure
Wind
Speed
Type1
(°F)
(°F)
(%)
(in Hg)
(in Hg)
Direction
(kt)
East
Highland Park, Virginia -
RIVA2
Sample
Days
57.9
44.9
63.6
30.06
29.85
3.4
(67)
±0.9
±0.9
± 1.0
±0.01
±0.01
S
±0.1
57.9
44.6
63.2
30.06
29.85
3.4
2014
±0.4
±0.4
±0.4
± <0.01
±<0.01
S
±<0.1
'Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
2Temperature, station pressure, and wind parameters were measured at RIVA in 2014. Hie remaining information was obtained
from the closest NWS weather station located at Richmond International Airport, WBAN 13740.
Table 22-2 presents average temperature, average dew point temperature, average relative
humidity, average station and sea level pressure, and wind information (average scalar wind
speed and prevailing wind direction) for days on which samples were collected and for all of
2014. Also included in Table 22-2 is the 95 percent confidence interval for each parameter. As
shown, average meteorological conditions on sample days at RIVA were representative of
average weather conditions experienced throughout the year.
22-6
-------
22.2.2 Wind Rose Comparison
Hourly surface wind data were also uploaded into a wind rose software program to
produce customized wind roses, as described in Section 3.4.2. A wind rose shows the frequency
of wind directions using "petals" positioned around a 16-point compass, and uses different colors
to represent wind speeds. Figure 22-3 presents two wind roses for the RIVA monitoring site. The
first is a wind rose representing wind observations for all of 2014 and the second is a wind rose
representing wind observations for days on which samples were collected in 2014. These are
used to identify the predominant wind speed and direction for 2014 and to determine if wind
observations on sample days were representative of conditions experienced over the entire year.
Figure 22-3. Wind Roses for the Wind Data Collected at RIVA
2014 Wind Rose Sample Day Wind Rose
WIND SPEED
(Knots)
HI >= 22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
WIND SPEED
(Knots)
HI >=22
¦ 17-21
¦ 11-17
¦ 7-11
~ 4-7
Observations from Figure 22-3 for RIVA include the following:
• The 2014 wind rose shows that wind blows predominantly from the southern
quadrants at RIVA and rarely from the northern quadrants. Although winds from the
south were observed the most (13 percent of observations), winds from the southeast
to southwest account for majority of observations. Winds were light at RIVA, with
winds greater than 17 knots not observed at this site, although calm winds were also
not observed at RIVA. Wind speeds greater than 11 knots were most often observed
with west-northwesterly winds.
• The wind patterns on the sample day wind rose resemble the wind patterns on the
full-year wind rose, indicating that wind observations on sample days were
representative of those observed throughout the year.
22-7
-------
22.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for the Virginia
monitoring 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 22-3.
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 22-3. It is
important to note which pollutants were sampled for at each site when reviewing the results of
this analysis. PAHs and hexavalent chromium were sampled for year-round at RIVA. RIVA is
the only NATTS site to sample hexavalent chromium year-round in 2014.
Table 22-3. Risk-Based Screening Results for the Virginia Monitoring Site
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
East Highland Park, Virginia - RIVA
Naphthalene
0.029
48
57
84.21
96.00
96.00
Benzo(a)pyrene
0.00057
1
36
2.78
2.00
98.00
Hexavalent Chromium
0.000083
1
24
4.17
2.00
100.00
Total
50
117
42.74
Observations from Table 22-3 include the following:
• Concentrations of two PAHs, naphthalene and benzo(a)pyrene, and hexavalent
chromium failed screens for RIVA.
• Concentrations of naphthalene failed 48 screens, which represents an 84 percent
failure rate. Concentrations of the other two pollutants failed a single screen each.
• Concentrations of naphthalene account for 48 of the 50 failed screens for RIVA,
accounting for 96 percent of failed screens. Thus, naphthalene is RIVA's only
pollutant of interest.
22-8
-------
22.4 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.
• 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 are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at RIVA are provided in Appendices M and O.
22.4.1 2014 Concentration Averages
Quarterly and annual concentration averages 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 the entire year of sampling. Annual averages were calculated for
pollutants where three valid quarterly averages could be calculated 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 22-4, 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.
22-9
-------
Table 22-4. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Virginia Monitoring Site
Pollutant
# of
Measured
Detections
vs. # >MDL
# of
Samples
1st
Quarter
Average
(ng/m3)
2nd
Quarter
Average
(ng/m3)
3rd
Quarter
Average
(ng/m3)
4th
Quarter
Average
(ng/m3)
Annual
Average
(ng/m3)
East Highland Park, Virginia - RIVA
Naphthalene
57/57
57
55.57
± 17.52
57.85
± 10.28
69.26
± 22.46
67.41
±20.81
62.57
±8.89
Observations for RIVA from Table 22-4 include the following:
• Concentrations of naphthalene measured at RIVA range from 21.3 ng/m3 to
178 ng/m3.
• The first quarter average concentration of naphthalene is the lowest and the third
quarter average is the highest, although the differences among the quarterly averages
are not statistically significant and each quarterly average concentration exhibits
considerable variability, as indicated by the confidence intervals. Concentrations
greater than 100 ng/m3 were measured during each calendar quarter except the second
quarter while concentrations less than 35 ng/m3 were measured during each calendar
quarter in 2014.
• Compared to other NMP sites sampling PAHs, RIVA has the 10th highest annual
average concentration of naphthalene, as shown in Table 4-11 of Section 4.
22.4.2 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 22-4 for RIVA. Figure 22-4 overlays the site's minimum, annual average, and maximum
concentrations onto the program-level minimum, first quartile, median, average, third quartile,
and maximum concentrations, as described in Section 3.4.3.1, and are discussed below.
22-10
-------
Figure 22-4. Program vs. Site-Specific Average Naphthalene Concentration
¦
0
100
200
300
Concentration {ng/m3)
400
500
600
Program:
Site:
1st Quartile
¦
Site Average
o
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
Site Concentration Range
4th Quartile
~
Average
i
Figure 22-4 presents the box plot for naphthalene for RIVA and shows the following:
• The maximum naphthalene concentration measured at RIVA (178 ng/m3) is
roughly one-third the program-level maximum concentration (568 ng/m3).
• There were no non-detects of naphthalene measured at RIVA, or across the
program (although difficult to discern in Figure 22-4).
• The annual average concentration of naphthalene for RIVA (62.57 ng/m3) is just
less than the program-level average concentration (66.46 ng/m3).
22.4.3 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 3.4.3.2.
RIVA began sampling PAHs under the NMP in October 2008. Thus, Figure 22-5 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.
22-11
-------
Figure 22-5. Yearly Statistical Metrics for Naphthalene Concentrations Measured at RIVA
Maximum
Observations from Figure 22-5 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 22-5 begins with 2009.
• The 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.
• 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).
• Most of the statistical parameters are at a minimum for 2014. Since 2009, the median
concentration decreased by 33 percent and the 1-year average concentration
decreased by 44 percent.
22-12
-------
22.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at the RIVA monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4 for
definitions and explanations regarding the various toxicity factors, time frames, and calculations
associated with these risk-based screenings.
22.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for RIVA 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 3.4.3.3 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 22-5, where applicable. Cancer risk approximations are presented as
probabilities while the noncancer hazard approximations are ratios and thus, unitless values.
Table 22-5. Risk Approximations for the Virginia Monitoring Site
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections vs.
# of Samples
Annual
Average
(ng/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
East Highland Park, Virginia - RIVA
Naphthalene
0.000034
0.003
57/57
62.57
±8.89
2.13
0.02
Observations for RIVA from Table 22-5 include the following:
• The annual average concentration of naphthalene for RIVA is 62.57 ± 8.89 ng/m3.
• The cancer risk approximation for naphthalene based on RIVA's annual average
concentration is 2.13 in-a-million.
• The noncancer hazard approximation for naphthalene is considerably less than 1.0
(0.02), indicating that no adverse noncancer health effects are expected from this
individual pollutant.
22-13
-------
22.5.2 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 22-6 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
that have cancer toxicity factors. Table 22-6 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 22-6 provides the pollutants of interest with the highest cancer risk approximations (in-a-
million) for RIVA, as presented in Table 22-5. The emissions, toxicity-weighted emissions, and
cancer risk approximations are shown in descending order in Table 22-6. Table 22-7 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 average concentrations to be
calculated. A more in-depth discussion of this analysis is provided in Section 3.4.3.4. Similar to
the cancer risk and noncancer hazard approximations provided in Section 22.5.1, this analysis
may help policy-makers prioritize their air monitoring activities.
22-14
-------
Table 22-6. 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)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
East Highland Park, Virginia (Henrico County) - RIVA
Benzene
102.27
Formaldehyde
1.12E-03
Naphthalene
2.13
Formaldehyde
86.37
Benzene
7.98E-04
Acetaldehyde
50.16
1,3-Butadiene
5.48E-04
Ethylbenzene
48.29
Naphthalene
2.84E-04
1.3 -Butadiene
18.27
POM, Group 2b
2.22E-04
Tetrachloroethylene
17.17
POM, Group 2d
1.26E-04
Naphthalene
8.35
Ethylbenzene
1.21E-04
POM, Group 2b
2.52
Acetaldehyde
1.10E-04
POM, Group 2d
1.43
POM, Group 5a
8.51E-05
Trichloroethylene
0.85
Arsenic, PM
6.88E-05
-------
Table 22-7. 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)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
East Highland Park, Virginia (Henrico County) - RIVA
Toluene
542.35
Acrolein
276,867.54
Naphthalene
0.02
Hexane
196.44
1.3 -Butadiene
9,132.57
Xylenes
193.08
Formaldehyde
8,812.86
Methanol
181.20
Acetaldehyde
5,572.79
Benzene
102.27
Benzene
3,408.91
Formaldehyde
86.37
Naphthalene
2,783.60
Ethylene glycol
62.63
Xylenes
1,930.83
Acetaldehyde
50.16
Arsenic, PM
1,067.34
Ethylbenzene
48.29
Lead, PM
808.19
Methyl isobutyl ketone
24.42
Propionaldehyde
556.24
-------
Observations from Table 22-6 include the following:
• Benzene, formaldehyde, and acetaldehyde 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 seventh highest
emissions and the fourth highest toxicity-weighted emissions for Henrico County.
• POM, Group 2b is the eighth 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. Benzo(a)pyrene
failed a single screen for RIVA but was not identified as a pollutant of interest.
Observations from Table 22-7 include the following:
• Toluene, hexane, 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, 1,3-butadiene, and formaldehyde.
• Four 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 13th).
22.6 Summary of the 2014 Monitoring Data for RIVA
Results from several of the data analyses described in this section include the following:
~~~ Concentrations of two PAHS and hexavalent chromium failed at least one screen,
although naphthalene was the only pollutant identified as a pollutant of interest for
RIVA.
~~~ RIVA has the 10th highest annual average concentration of naphthalene among NMP
sites sampling PAHs.
22-17
-------
~~~ Concentrations of naphthalene have a decreasing trend at RIVA since the onset of
PAH sampling at this site.
22-18
-------
23.0 Site in Washington
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS site in Washington, and integrates these concentrations
with emissions, meteorological, and risk information. Data generated by sources other than ERG
are not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) 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 geographical and
physical information about the location of the site and the surrounding area. 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 NATTS site in Washington is located in Seattle. Figure 23-1 is 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 2011 NEI for point sources, version 2. 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 in order to
emphasize emissions sources within the boundary. Table 23-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates.
23-1
-------
Figure 23-1. Seattle, Washington (SEWA) Monitoring Site
»S*horton,Sl
,^-SiHorton£t«
Wa>
SKll.b St
indsSt
S Spokane St
•SpokanciStM
JW i I
S Char'estown St
Jefferson Park'
Center
S Dakota St
S Nevada St
SOregon &
VAPuget
-wmd " .
« Health
Care
Sytffrm V
OSGA
irce: NASA, N G
'2 0 0 8^H 1 cr.o s.-q'AI
orpl
to
L.J
tb
-------
Figure 23-2. NEI Point Sources Located Within 10 Miles of SEW A
122°30'0"W
122"25'0"W
122°20'0"W
Laki
Washington
Lake
Sammamish
King
County
King
County
122°30'0"W
Legend
~ SEWA NATTS site
O 10 mile radius
County boundary
122°10'0"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.
A
/
Kitsap '
County /
/
/
" ^ -L .
/
/
/
/ -r
/
i
/
Source Category Group (No. of Facilities)
*
Aerospace/Aircraft Manufacturing Facility (2)
A
Metal Coating, Engraving, and Allied Services to Manufacturers (1)
t
Airport/Airline/Airport Support Operations (27)
<•>
Metals Processing/Fabrication Facility (4)
«
Automobile/Truck Manufacturing Facility (3)
Miscellaneous Commercial/Industrial Facility (1)
en
Brick, Structural Clay, or Clay Ceramics Plant (2)
0
Paint and Coating Manufacturing Facility (1)
B
Bulk Terminals/Bulk Plants (2)
7
Portland Cement Manufacturing (1)
e
Electrical Equipment Manufacturing Facility (1)
X
Rail Yard/Rail Line Operations (2)
F
Food Processing/Agriculture Facility (2)
Ship/Boat Manufacturing or Repair Facility (1)
n*
Glass Plant (1)
w
Steel Mill (1)
o
Institutional (school, hospital, prison, etc.) (1)
I
WastewaterTreatment 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
178,000
1-5 S at Spokane St Viaduct
1AADT reflects 2014 data (WS DOT, 2014)
BOLD ITALICS = EPA-designated NATTS Site
to
-L
-------
The SEWA monitoring site is located in Seattle, at the southeast corner of the Beacon
Hill Reservoir. With the reservoir covered, the entire area is part of Jefferson Park (Seattle,
2016). 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 less than 1 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 and a food processing facility, as shown in Figure 23-2.
In addition to providing city, county, CBSA, 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 effect concentrations measured at a given monitoring site. The traffic
volume experienced near SEWA is 178,000, 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 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
site in Washington on sample days, as well as over the course of the year.
23.2.1 Meteorological Summary
In order to provide an overview of the meteorological conditions experienced at each
monitoring site, hourly meteorological data for 2014 were retrieved and sample day and full-year
averages developed for temperature, dew point, relative humidity, pressure, and wind speed.
Weather data from the actual monitoring site(s) were obtained from AQS, where available. If
site-specific weather data were not available in AQS, then data were obtained from NCDC for
the NWS weather station located closest to the monitoring site(s), as described in Section 3.4.2.
23-5
-------
For the Washington site, site-specific data were available for some, but not all, of the parameters
in Table 23-2. For SEW A, temperature, station pressure, humidity, and wind information was
available in AQS. Data from the NWS weather station at Boeing Field/King County International
Airport (WBAN 24234) were used for the remaining parameters (sea level pressure and dew
point temperature). The King County International Airport weather station is located 2.7 miles
south of SEWA. In addition, pressure and humidity measurements at SEWA from May 2014
were not in AQS; thus, data from Seattle Airport was used as a surrogate. A map showing the
distance between the monitoring site and the closest NWS weather station is provided in
Appendix R. These data were used to determine how meteorological conditions on sample days
vary from conditions experienced throughout the year.
Table 23-2. Average Meteorological Conditions near the Washington Monitoring Site
Average
Type1
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(in Hg)
Average
Station
Pressure
(in Hg)
Prevailing
Wind
Direction
Average
Scalar Wind
Speed
(kt)
Seattle, Washington - SEWA2
Sample
Days
52.6
43.6
77.6
30.05
29.72
3.2
(76)
±0.5
±0.5
±0.8
±0.01
±0.01
S
±0.1
53.4
44.4
77.4
30.03
29.71
3.2
2014
±0.2
±0.2
±0.4
± <0.01
±<0.01
S
±<0.1
'Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
2Temperature, humidity, station pressure, and wind parameters were measured at SEWA. Hie remaining information was
obtained from the closest NWS weather station located at King Comity International Airport, WBAN 24234. In addition,
station pressure and relative humidity data from the NWS station were used as a surrogate where meteorological data were not
available in AQS.
Table 23-2 presents average temperature, average dew point temperature, average relative
humidity, average station and sea level pressure, and wind information (average scalar wind
speed and prevailing wind direction) for days on which samples were collected and for all of
2014. Also included in Table 23-2 is the 95 percent confidence interval for each parameter.
Average meteorological conditions on sample days at SEWA were fairly representative of
average weather conditions experienced throughout the year. The largest differences between the
parameters shown in Table 23-2 are for average temperature and average dew point temperature.
23-6
-------
23.2.2 Wind Rose Comparison
Hourly surface wind data were also uploaded into a wind rose software program to
produce customized wind roses, as described in Section 3.4.2. A wind rose shows the frequency
of wind directions using "petals" positioned around a 16-point compass, and uses different colors
to represent wind speeds. Figure 23-3 presents two wind roses for the SEW A monitoring site.
The first is a wind rose representing wind observations for all of 2014 and the second is a wind
rose representing wind observations for days on which samples were collected in 2014. These
are used to identify the predominant wind speed and direction for 2014 and to determine if wind
observations on sample days were representative of conditions experienced over the entire year.
Figure 23-3. Wind Roses for Wind Data Collected at SEWA
2014 Wind Rose Sample Day Wind Rose
"HnORTH-- NORTH
¦15% / _ f _ ' 15%
;WEST
WIND SPED
(Knots)
WIND SPEED
(Knots)
SOUTH
SOUTH
Calms: 1.13% Calms: 0.99%
Observations from Figure 23-3 for SEWA include the following:
• The 2014 wind rose shows that light winds prevailed at SEWA. Winds from the
southeast to south-southwest were observed the most, along with west-northwesterly
winds, while winds from the west-southwest, west, and north-northwest were
observed the least.
• The wind patterns shown on the sample day wind rose are similar to the wind patterns
shown on the full-year wind rose, indicating that wind conditions in on sample days
were representative of those experienced throughout the year.
23-7
-------
23.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for SEWA 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 23-3. 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 23-3. It is important to note which
pollutants were sampled for at the site when reviewing the results of this analysis. PMio metals,
VOCs, PAHs, and carbonyl compounds were sampled for at SEWA.
Table 23-3. Risk-Based Screening Results for the Washington Monitoring Site
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Seattle, Washington - SEWA
Formaldehyde
0.077
61
61
100.00
13.99
13.99
Benzene
0.13
60
60
100.00
13.76
27.75
Carbon Tetrachloride
0.17
60
60
100.00
13.76
41.51
1,2-Dichloroethane
0.038
53
53
100.00
12.16
53.67
1.3 -Butadiene
0.03
50
53
94.34
11.47
65.14
Arsenic (PMio)
0.00023
47
60
78.33
10.78
75.92
Naphthalene
0.029
42
61
68.85
9.63
85.55
Acetaldehyde
0.45
38
61
62.30
8.72
94.27
Nickel (PMio)
0.0021
16
60
26.67
3.67
97.94
Ethylbenzene
0.4
5
60
8.33
1.15
99.08
Cadmium (PMio)
0.00056
2
60
3.33
0.46
99.54
Acenaphthene
0.011
1
60
1.67
0.23
99.77
Benzo(a)pyrene
0.00057
1
40
2.50
0.23
100.00
Total
436
749
58.21
Observations from Table 23-3 for SEWA include the following:
• Concentrations of 13 pollutants failed at least one screen for SEWA; 58 percent of
concentrations for these 13 pollutants were greater than their associated risk screening
value (or failed screens).
23-8
-------
• Concentrations of nine pollutants contributed to 95 percent of failed screens for
SEWA and therefore were identified as pollutants of interest for the site. These nine
include two carbonyl compounds, four VOCs, two PMio metals, and one PAH.
• Benzene, carbon tetrachloride, and formaldehyde were detected in every valid VOC
or carbonyl compound sample collected at SEWA and failed 100 percent of screens.
1,2-Dichloroethane also failed 100 percent of screens, but was not detected in every
sample collected.
23.4 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 site.
• 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 are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at SEWA are provided in Appendices J, L, M, and N.
23.4.1 2014 Concentration Averages
Quarterly and annual concentration averages 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 the entire year of sampling. Annual averages were calculated for
pollutants where three valid quarterly averages could be calculated and where method
completeness was greater than or equal to 85 percent, as presented in Section 2.4. Quarterly and
23-9
-------
annual average concentrations for the pollutants of interest for the Washington monitoring site
are presented in Table 23-4, 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.
Table 23-4. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Washington Monitoring Site
# of
Measured
1st
2nd
3rd
4th
Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs.
# >MDL
# of
Samples
Average
(Hg/m3)
Average
(Hg/m3)
Average
(Hg/m3)
Average
(Hg/m3)
Average
(ng/m3)
Seattle, Washington
- SEWA
0.38
0.44
1.04
0.85
0.69
Acetaldehyde
61/61
61
±0.10
±0.10
±0.28
±0.31
±0.13
0.70
0.32
0.36
0.68
0.50
Benzene
60/60
60
±0.19
±0.05
±0.06
±0.14
±0.07
0.09
0.03
0.06
0.10
0.07
1,3-Butadiene
53/52
60
±0.05
±0.01
±0.02
±0.03
±0.01
0.67
0.70
0.69
0.63
0.67
Carbon Tetrachloride
60/60
60
±0.07
±0.02
±0.02
±0.02
±0.02
0.08
0.07
0.04
0.07
0.06
1,2-Dichloroethane
53/49
60
±0.01
±<0.01
±0.01
±0.01
±0.01
0.38
0.42
0.89
0.69
0.60
Formaldehyde
61/61
61
±0.12
±0.07
±0.20
±0.22
±0.10
0.69
0.36
0.59
0.76
0.60
Arsenic (PMi0)a
60/50
60
±0.29
±0.20
±0.15
±0.23
±0.11
43.11
34.44
59.70
55.16
49.25
Naphthalene1
61/61
61
± 13.01
± 13.74
± 14.84
± 16.48
±7.46
0.88
1.25
3.17
1.64
1.74
Nickel (PMi,;,)a
60/59
60
±0.28
±0.50
±0.97
±0.85
±0.40
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
Observations from Table 23-4 include the following:
• 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 (0.69 ±0.13 |ig/m3), carbon tetrachloride (0.67 ± 0.02 |ig/m3),
formaldehyde (0.60 ±0.10 |ig/m3), and benzene (0.50 ± 0.07 |ig/m3). These are
similar to the annual average concentrations calculated for 2012 and 2013.
• Even though acetaldehyde has the highest annual average concentration among
SEWA's pollutants of interest, this annual average is one of the lowest among NMP
sites sampling acetaldehyde. In addition, SEWA's annual average concentration of
formaldehyde is the lowest among all NMP sites. Few NMP sites have annual
average concentrations of these two pollutants less than 1 |ig/m3, and the annual
23-10
-------
averages for both acetaldehyde and formaldehyde for SEWA are less than 1 |ig/m3.
Similar observations were made in previous NMP reports.
The quarterly average concentrations of acetaldehyde for the first and second quarters
of the year are significantly lower than the quarterly average concentrations for the
third and fourth quarters of the year. Only one acetaldehyde concentration greater
than 1 |ig/m3 was measured at SEWA during the first half of 2014 while 13 were
measured during the second half of the year. In addition, seven concentrations
measured between January and June are less than the minimum concentration
measured during the second half of the year. A similar observation can be made for
formaldehyde, although the difference is less dramatic. Formaldehyde concentrations
greater than 1 |ig/m3 were not measured at SEWA during the first half of 2014 while
seven were measured during the second half of the year. In addition, seven of the nine
formaldehyde concentrations less than 0.3 |ig/m3 were measured between January and
June while only two were measured during the second half of the year. The three
lowest concentrations of acetaldehyde and formaldehyde were measured at SEWA on
the same days in January and February while the three highest concentrations of each
pollutant were measured on the same days in November, September, and August.
The first and fourth quarter average benzene concentrations shown in Table 23-4 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 range from 0.138 |ig/m3 to 1.71 |ig/m3 and that all 17 benzene
concentrations greater than 0.6 |ig/m3 were measured at SEWA during the first (9) or
fourth (8) quarters of 2014, including two greater than 1 |ig/m3 (January 5, 2014 and
November 19, 2014). Conversely, none of the 17 concentrations of benzene less than
0.35 |ig/m3 were measured at SEWA during the first or fourth quarters of 2014. A
similar observation can be made for 1,3-butadiene. All 10 1,3-butadiene
concentrations greater than 0.1 |ig/m3 were measured in January, March, November,
or December. The maximum concentrations of both benzene and 1,3-butadiene were
measured on January 5, 2014. Several of the highest concentrations of these
compounds were measured at SEWA on the same days.
Concentrations of 1,2-dichloroethane measured during the third quarter appear lower
than those measured during the rest of the year, based on the quarterly average
concentrations shown in Table 23-4. A review of the data shows that six of the seven
non-detects of this pollutant were measured between late July and early September. A
similar observation was made in the 2013 NMP report.
Concentrations of naphthalene appear higher during the second half of 2014, although
the quarterly average concentrations of naphthalene exhibit considerable variability,
as indicated by the confidence intervals. Concentrations of naphthalene measured at
SEWA range from 9.19 ng/m3 to 167 ng/m3. The maximum naphthalene
concentration was measured on November 19, 2014, the same day the highest
acetaldehyde and formaldehyde concentrations were measured and the second highest
benzene and 1,3-butadiene concentrations were measured.
23-11
-------
• Arsenic concentrations measured at SEWA during the second quarter of 2014 appear
lower than those measured during the other calendar quarters. A review of the data
shows that arsenic concentrations measured at SEWA range from 0.008 ng/m3 to
1.73 ng/m3. Six of the seven arsenic concentrations less than or equal to 0.1 ng/m3
measured at SEWA were measured during the second quarter (with one measured
during the first quarter and none measured during the third or fourth quarters).
Additionally, only one arsenic measurement greater than 1 ng/m3 was measured
during the second quarter of 2014, with the other eight spread out across the other
calendar quarters. The maximum arsenic concentrations measured during the first and
fourth quarters are roughly twice the maximum concentrations measured during the
second and third quarters of 2013. The second highest arsenic concentration was
measured at SEWA on November 19, 2014 (1.58 ng/m3).
• Concentrations of nickel measured at SEWA appear highly variable, with the
quarterly average for the third quarter nearly four times greater than the first quarter
average (and the other quarterly average concentrations falling in-between).
Concentrations of nickel measured at SEWA in 2014 range from 0.17 ng/m3 to
6.17 ng/m3. Three of the four nickel concentrations greater than 5 ng/m3 were
measured at SEWA in August, with the fourth measured on November 19, 2014
(5.32 ng/m3). None of the 17 nickel concentrations greater than 2 ng/m3 were
measured at SEWA before June.
Tables 4-9 through 4-12 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 4-9 for VOCs once; SEWA has the sixth highest annual
average concentration of carbon tetrachloride among sites sampling VOCs. SEWA is
the first NMP site outside of Calvert City, Kentucky to appear in Table 4-9 for carbon
tetrachloride. However, the annual average concentrations for most NMP sites span
less than 0.1 |ig/m3. A similar observation was made in the 2012 and 2013 NMP
reports.
• SEWA does not appear in Table 4-10 for carbonyl compounds. As indicated above,
SEWA has one of the lowest annual average acetaldehyde concentrations and the
lowest annual average concentration of formaldehyde among NMP sites sampling
these pollutants.
• SEWA does not appear in Table 4-11 for naphthalene. SEWA has the 15th highest
annual average concentration of naphthalene among NMP sites sampling PAHs.
• As shown in Table 4-12, SEWA has the third highest annual average concentration of
nickel among all NMP sites sampling metals (PMio), behind only ASKY-M and
BOMA. SEWA ranked second highest in the 2012 and 2013 NMP reports while
SEWA had the highest annual average nickel concentration in the 2010 and 2011
NMP reports.
23-12
-------
• SEWA also has the seventh highest annual average concentration of arsenic among
NMP sites sampling PMio metals.
23.4.2 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 23-4 for SEWA. Figures 23-4 through 23-12 overlay the site's minimum, annual average,
and maximum concentrations onto the program-level minimum, first quartile, median, average,
third quartile, and maximum concentrations for each pollutant, as described in Section 3.4.3.1,
and are discussed below.
Figure 23-4. Program vs. Site-Specific Average Acetaldehyde Concentration
SEWA
12 3
4 5 6
Concentration (ng/m3)
7
8 9
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site: Site Average
o
Site Concentration Range
Figure 23-4 presents the box plot for acetaldehyde for SEWA and shows the following:
• All but one of SEWA's acetaldehyde measurements are less than the program-level
third quartile (2.24 |ig/m3).
• SEWA's annual average acetaldehyde concentration is more than 1 |ig/m3 less than
the program-level average concentration for acetaldehyde (1.76 |ig/m3) and is also
less than the program-level first quartile (0.98 |ig/m3).
• This site has one of the lowest annual average concentrations of acetaldehyde among
NMP sites sampling carbonyl compounds, with only two sites having lower annual
average concentrations of acetaldehyde.
23-13
-------
Figure 23-5. Program vs. Site-Specific Average Arsenic (PMio) Concentration
6
Program Max Concentration = 10.1 ng/m3
?
i i i i i
0 1 2 3 4 5 6
Concentration {ng/m3)
Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 23-5 presents the box plot for arsenic for SEWA and shows the following:
• The program-level maximum arsenic concentration (10.1 ng/m3) is not shown directly
on the box plot in Figure 23-5 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 arsenic concentration measured at SEWA is considerably less than the
maximum concentration measured across the program.
• SEWA's annual average arsenic (PMio) concentration is very similar to the program-
level average concentration.
• There were no non-detects of arsenic measured at SEWA, although this is difficult to
discern in Figure 23-5.
Figure 23-6. Program vs. Site-Specific Average Benzene Concentration
Program Max Concentration = 12.4 |ig/m;
o
2
4
6
8
10
Concentration (ng/m3)
Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
23-14
-------
Figure 23-6 presents the box plot for benzene for SEWA and shows the following:
• The program-level maximum benzene concentration (12.4 |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.
• All benzene concentrations measured at SEWA in 2014 are less than 2 |ig/m3.
• The annual average benzene concentration for SEWA falls between the program-level
median concentration and program-level first quartile. Among other NMP sites,
SEWA has one of the lowest annual average concentrations of benzene (ranking
26th).
Figure 23-7. Program vs. Site-Specific Average 1,3-Butadiene Concentration
I
Program Max Concentration = 5.90 ng/m3
0
0.2
0.4 0.6
Concentration (ng/m3)
0.8
l
Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site: Site Average
o
Site Concentration Range
Figure 23-7 presents the box plot for 1,3-butadiene for SEWA and shows the following:
• Similar to benzene, the program-level maximum 1,3-butadiene concentration
(5.90 |ig/m3) is not shown directly on the box plot in Figure 23-7 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 |ig/m3.
• The annual average 1,3-butadiene concentration for SEWA is just greater than the
program-level median concentration and less than both the program-level average and
third quartile, which are very similar to each other.
Seven non-detects of 1,3-butadiene were measured at SEWA.
23-15
-------
Figure 23-8. Program vs. Site-Specific Average Carbon Tetrachloride Concentration
Program Max Concentration = 3.06 |ig/m3
0 0.5 1 1.5 2 2.5
Concentration (ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 23-8 presents the box plot for carbon tetrachloride 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. Note that the
program-level median and average concentrations are similar and plotted nearly on
top of each other.
• The range of carbon tetrachloride concentrations measured at SEWA spans
approximately 0.5 |ig/m3,
• The annual average concentration of carbon tetrachloride for SEWA is greater than
the program-level average concentration and just less than the program-level third
quartile, although less than 0.04 |ig/m3 separates these three values.
Figure 23-9. Program vs. Site-Specific Average 1,2-Dichloroethane Concentration
SEWA
-1
Program Max Concentration = 27.4 ng/m3
t 1 1 r
0 0.2 0.4 0.6 0.8 1
Concentration (ng/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range
23-16
-------
Figure 23-9 presents the box plot for 1,2-dichloroethane for SEWA and shows the
following:
• The scale of the box plot in Figure 23-9 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 (27.4 |ig/m3) is
considerably greater than the majority of measurements.
• All of the measurements of 1,2-dichloroethane measured at SEWA are less than the
program-level third quartile (0.101 |ig/m3). Note that the program-level average
concentration of 0.31 |ig/m3 is being driven by the measurements at the upper end of
the concentration range.
SEWA's annual average concentration of 1,2-dichloroethane is less than the program-
level first quartile.
• Seven non-detects of 1,2-dichloroethane were measured at SEWA.
Figure 23-10. Program vs. Site-Specific Average Formaldehyde Concentration
*
12 15
Concentration {[j.g/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 23-10 presents the box plot for formaldehyde for SEWA and shows the following:
• All formaldehyde concentrations measured at SEWA in 2014 are less than 2 |ig/m3.
The maximum formaldehyde concentration measured at SEWA falls between the
program-level first quartile and median concentrations.
• The annual average concentration for SEWA is less than the program-level first
quartile (1.40 |ig/m3) and is the lowest annual average among NMP sites sampling
formaldehyde, both for 2014 and in previous years.
23-17
-------
Figure 23-11. Program vs. Site-Specific Average Naphthalene Concentration
100
200
300
Concentration {ng/m3)
400
500
Program:
1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site:
Site Average
o
Site Concentration Range
Figure 23-11 presents the box plot for naphthalene for SEWA and shows the following:
• The maximum naphthalene concentration measured at SEWA is considerably less
than the maximum concentration measured across the program.
• The annual average concentration of naphthalene for SEWA is less than the program-
level average concentration and similar to the program-level median concentration.
Figure 23-12. Program vs. Site-Specific Average Nickel (PMio) Concentration
-
\J 1
0 2 4 6 8 10
Concentration {ng/m3)
Progra m: 1st Qua rti 1 e
2nd Quartile 3rd Quartile
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range
o
Figure 23-12 presents the box plot for nickel for SEWA and shows the following:
• Although the maximum nickel concentration measured at SEWA is less than the
maximum concentration measured across the program, it is among the higher
measurements across the program. SEWA is one of only five NMP sites at which
nickel concentrations greater than 6 ng/m3 were measured.
• SEWA's annual average concentration is greater than the program-level average
concentration and program-level third quartile. Recall from the previous section that
this site has the third highest annual average concentration of nickel among NMP
sites sampling PMio metals.
23-18
-------
23.4.3 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 3.4.3.2.
Sampling for PMio metals, VOCs, and carbonyl compounds under the NMP began in 2007 and
sampling for PAHs began in 2008. Thus, Figures 23-13 through 23-21 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.
Figure 23-13. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
SEWA
Maximum
Observations from Figure 23-13 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). Only one other acetaldehyde concentration greater
than 3 |ig/m3 has been measured at SEWA (September 2012).
23-19
-------
• The 1-year average concentrations have 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 1-year average acetaldehyde
concentration changed little between 2012 and 2014 and is at a minimum for 2014
compared to the other years of sampling (0.69 |ig/m3).
• 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). The median then
decreased each year between 2011 and 2014, reaching a minimum of 0.54 |ig/m3 for
the entire sampling period.
Figure 23-14. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at SEWA
o
2007 2008 2009
2010 2011
Year
t
2014
O 5th Percentile - Minimum
— Median
- Maximum o 95th Percentile
Observations from Figure 23-14 for arsenic (PMio) concentrations measured at SEWA
include the following:
• The maximum arsenic concentration was measured at SEWA on January 19, 2009
(2.69 ng/m3), although a similar concentration was also measured in 2007
(2.56 ng/m3). In total, 11 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, where it appears the minimum concentration is
zero. For each of these years, the minimum concentration of arsenic is around
23-20
-------
0.01 ng/m3. Seven of the nine arsenic concentrations less than or equal to 0.1 ng/m3
were measured in 2014.
• 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 by even less, from 0.50 ng/m3
(2011) to 0.63 ng/m3 (2013).
• All of the statistical parameters exhibit decreases between 2013 and 2014, with the
minimum, 5th percentile, 95th percentile, and maximum concentrations all at a
minimum since the onset of sampling in 2007.
Figure 23-15. Yearly Statistical Metrics for Benzene Concentrations Measured at SEWA
O 95th Percentile
Observations from Figure 23-15 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 was roughly half as high (2.55 |ig/m3,
measured in January 2011). Only five benzene concentrations greater than 2 |ig/m3
have been measured at SEWA.
• Overall, benzene concentrations have a slight decreasing trend at SEWA, although
this decrease is interrupted by the two years that the highest benzene concentrations
23-21
-------
were measured. If the maximum concentrations measured in 2009 and 2011 were
removed from the calculations, the 1-year average concentration would have a steady
decreasing trend for the entire period. The 1-year average concentration of benzene
has ranged from 0.50 |ig/m3 (2014) to 0.81 |ig/m3 (2009).
• The concentrations of benzene appear to have a seasonal trend at SEWA. Of the
68 benzene concentrations greater than 1 |ig/m3, 57 have been measured during the
colder months of the year, either during the first quarter (24) or fourth quarter (33) of
any given year.
Figure 23-16. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at SEWA
O 5th Percentile
O 95th Percentile
Observations from Figure 23-16 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 each year at SEWA 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.
Eleven percent of the measurements were non-detects for 2010, 17 percent were non-
23-22
-------
detects for 2011, 16 percent were non-detects for 2013, and 13 percent were non-
detects for 2014. The percentage of non-detects is 3 percent for each of the remaining
years.
• The 1-year average concentration has changed little over the years of sampling,
ranging from 0.06 |ig/m3 (2008) to 0.09 |ig/m3 (2011). Interestingly, the year with the
greatest number of non-detects (2011) also has the greatest number of measurements
greater than 0.2 |ig/m3 (seven).
• After fluctuating during the first few years of sampling, the 95th percentile was fairly
static between 2011 and 2013, then decreased by nearly half for 2014. The 95th
percentile is at a minimum for 2014, indicating that the majority of measurements are
less than 0.13 |ig/m3 in 2014.
Figure 23-17. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
SEWA
O 95th Percentile
Observations from Figure 23-17 for carbon tetrachloride concentrations measured at
SEWA include the following:
• 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. The maximum carbon
tetrachloride concentration (1.22 |ig/m3) has been measured twice at SEWA, once in
2008 and once in 2010.
23-23
-------
• 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.
• Between 2008 and 2011, a steady decreasing trend in the concentrations is shown,
with the 1-year average concentration for 2011 returning to 2007 levels.
• The range of measurements tightened for 2012 and is the smallest range of
measurements since the onset of sampling. 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 2011 to 43 in 2012,
the number of concentrations less than 0.6 |ig/m3 fell from 20 to seven during this
time frame.
• Despite the increase in the maximum concentration and the 95th percentile for 2013,
both the 1-year average and median concentrations exhibit slight decreases, although
the difference is not statistically significant. This is also true for 2014.
Figure 23-18. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured
at SEW A
pL-
pi-
pU
o
~
I
o
O...
2008
2007
2009
2010
Year
2011
2012
2013
2014
O 5th Percentile - Minimum - Med en — Maximum o 95th Percentile • ¦•¦^•¦¦Average
23-24
-------
Observations from Figure 23-18 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 there is an increasing
trend. For 2012, the percentage of measured detections is 93 percent, a considerable
increase from 26 percent in 2011. This percentage leveled off a bit for 2013 and 2014
(at 88 percent each).
• 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 2012,
2013, and 2014. This is because there were still several non-detects (or 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 23-19. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
SEWA
Maximum
Concentration for
2009 is 16.6 ng/m3
-r
2009
0-
2010 2011
Year
5th Percentile
— Minimum
- Med en - Maximum o 95th Percentile
23-25
-------
Observations from Figure 23-19 for formaldehyde concentrations measured at SEWA
include the following:
• The maximum formaldehyde concentration was measured at SEWA 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. Only nine concentrations greater than
2 |ig/m3 have been measured since the onset of carbonyl compound sampling at
SEWA, and all but one of these were measured prior to 2010.
• The 1-year average concentrations have an undulating pattern through 2012, with a
"down" year followed by an "up" year. Between 2007 and 2012, the 1-year average
formaldehyde concentrations have ranged from 0.53 |ig/m3 (2012) to 1.04 |ig/m3
(2009). The 1-year average formaldehyde concentration exhibits a very subtle
increase between 2012 to 2014, although the changes are not statistically significant.
The 1-year average concentrations calculated for 2012, 2013, and 2014 are the lowest
averages shown in Figure 23-19.
Figure 23-20. Yearly Statistical Metrics for Naphthalene Concentrations Measured at SEWA
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until March 2008.
23-26
-------
Observations from Figure 23-20 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 measurement greater than 250 ng/m3
measured at this site. Eight additional measurements greater than 200 ng/m3 have
been measured at SEWA and are spread across the years of sampling, except 2008.
• 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, from 19 in 2009 to seven for 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 this
year.
• Little change in the 1-year average concentration is shown between 2011 and 2013,
after which a significant decrease is shown for 2014. The fewest naphthalene
concentrations greater than 100 ng/m3 were measured in 2014 (one), with most years
having 10 or more (2010 is the exception at seven). In addition, the median
concentration for 2014 is less than 50 ng/m3 for the first time, indicating that more
than half of the measurements were less than 50 ng/m3.
23-27
-------
Figure 23-21. Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at SEWA
o
~
o
2010 2011
Year
T
2013
5th Percentile
— Minimum
- Median
— Maximum
O 95th Percentile
Observations from Figure 23-21 for nickel concentrations measured at SEWA include the
following:
• The two highest concentrations of nickel (14.3 ng/m3 and 11.8 ng/m3) were both
measured at SEWA in 2012, although concentrations greater than 10 ng/m3 were also
measured in 2009 (two) and 2010 (one).
• The 1-year average concentration exhibits an increase between 2007 and 2009, after
which a decrease in shown for 2010, with little change for 2011. An increase in the
1-year average concentration is shown for 2012, which is followed by a decrease for
2013 and little change for 2014. Confidence intervals calculated on the dataset
indicate that the changes shown are not statistically significant as the concentrations
measured are fairly variable. The median concentrations exhibit a similar pattern.
• The difference between the 1-year average and median concentrations is greater than
0.50 ng/m3 for all years (and greater than 1.0 ng/m3 for 2012). This also indicates that
there is considerable variability in the measurements of nickel.
• Despite this variability, all the 1-year average concentrations of nickel fall on either
side of 2 ng/m3. Exactly 1 ng/m3 separates the minimum 1-year average concentration
calculated (1.74 ng/m3, 2014) from the maximum 1-year average concentration
(2.74 ng/m3, 2012). If 2009 and 2012 are excluded, the difference among the 1-year
averages is less than 0.5 ng/m3.
23-28
-------
23.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at the Washington monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4
for definitions and explanations regarding the various toxicity factors, time frames, and
calculations associated with these risk-based screenings.
23.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Washington site 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 may want to shift their
air monitoring priorities. Refer to Section 3.4.3.3 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 23-5, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
23-29
-------
Table 23-5. Risk Approximations for the Washington Monitoring Site
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Seattle, Washington - SEWA
Acetaldehyde
0.0000022
0.009
61/61
0.69
±0.13
1.53
0.08
Benzene
0.0000078
0.03
60/60
0.50
±0.07
3.92
0.02
1.3 -Butadiene
0.00003
0.002
53/60
0.07
±0.01
2.08
0.03
Carbon Tetrachloride
0.000006
0.1
60/60
0.67
±0.02
4.04
0.01
1,2-Dichloroethane
0.000026
2.4
53/60
0.06
±0.01
1.60
<0.01
Formaldehyde
0.000013
0.0098
61/61
0.60
±0.10
7.85
0.06
Arsenic (PMi0)a
0.0043
0.000015
60/60
0.60
±0.11
2.58
0.04
Naphthalene1
0.000034
0.003
61/61
49.25
±7.46
1.67
0.02
Nickel (PMi,;,)a
0.00048
0.00009
60/60
1.74
±0.40
0.83
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.
Observations from Table 23-5 for SEWA include the following:
• The pollutants with the highest annual average concentrations for SEWA are
acetaldehyde, carbon tetrachloride, formaldehyde, and benzene.
• The pollutants with the highest cancer risk approximations are formaldehyde, carbon
tetrachloride, benzene, and arsenic. The cancer risk approximation for formaldehyde
for SEWA is the lowest among this pollutant's site-specific cancer risk
approximations and is one of only three 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), indicating that no adverse
noncancer health effects are expected from these individual pollutants.
23-30
-------
23.5.2 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 23-6 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
that have cancer toxicity factors. Table 23-6 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 23-6 provides the pollutants of interest with the highest cancer risk approximations (in-a-
million) for SEW A, as presented in Table 23-5. The emissions, toxicity-weighted emissions, and
cancer risk approximations are shown in descending order in Table 23-6. Table 23-7 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 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 23.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
23-31
-------
Table 23-6. 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)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Seattle, Washington (King County) - SEWA
Benzene
930.96
Formaldehyde
1.01E-02
Formaldehyde
7.85
Formaldehyde
776.28
Benzene
7.26E-03
Carbon Tetrachloride
4.04
Ethylbenzene
460.42
1,3-Butadiene
4.24E-03
Benzene
3.92
Acetaldehyde
442.08
Naphthalene
2.98E-03
Arsenic (PMio)
2.58
1.3 -Butadiene
141.43
POM, Group 2b
1.76E-03
1,3-Butadiene
2.08
T etrachloroethylene
95.67
POM, Group 2d
1.16E-03
Naphthalene
1.67
Naphthalene
87.72
Ethylbenzene
1.15E-03
1,2-Dichloroethane
1.60
POM, Group 2b
19.97
POM, Group 5a
1.11E-03
Acetaldehyde
1.53
POM, Group 2d
13.20
Acetaldehyde
9.73E-04
Nickel (PMio)
0.83
T richloroethylene
11.73
Nickel, PM
5.36E-04
-------
Table 23-7. 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)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Seattle, Washington (King County) - SEWA
Toluene
4,999.08
Acrolein
2,910,205.08
Acetaldehyde
0.08
Xylenes
1,895.75
Formaldehyde
79,212.57
Formaldehyde
0.06
Hexane
1,472.55
1,3-Butadiene
70,716.54
Arsenic (PMio)
0.04
Methanol
1,144.61
Cyanide Compounds, gas
63,595.60
1,3-Butadiene
0.03
Benzene
930.96
Acetaldehyde
49,120.41
Nickel (PMio)
0.02
Formaldehyde
776.28
Benzene
31,032.01
Benzene
0.02
Ethylbenzene
460.42
Naphthalene
29,239.94
Naphthalene
0.02
Ethylene glycol
455.61
Xylenes
18,957.50
Carbon Tetrachloride
0.01
Acetaldehyde
442.08
Lead, PM
16,900.94
1,2-Dichloroethane
<0.01
Methyl isobutyl ketone
205.29
Nickel, PM
12,405.52
-------
Observations from Table 23-6 for SEWA include the following:
• Benzene, formaldehyde, and ethylbenzene 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 1,3-butadiene.
• Eight of the highest emitted pollutants also have the highest toxicity-weighted
emissions for King County.
• Formaldehyde has the highest cancer risk approximation for SEWA. This pollutant
has the highest toxicity-weighted emissions and ranks second for quantity emitted.
Benzene, naphthalene, 1,3-butadiene, and acetaldehyde also appear on all three lists.
• Carbon tetrachloride and arsenic, which rank second and fourth, respectively, for
cancer risk approximations for SEWA, do not appear on either emissions-based list.
This is also true for 1,2-dichloroethane. Nickel, which ranks ninth among the
pollutants of interest for SEWA, has the 10th highest toxicity-weighted emissions for
King County, but is not among the highest emitted (of the pollutants with cancer
UREs).
• POM, Group 2b is the eighth highest emitted "pollutant" in King County and ranks
fifth for toxicity-weighted emissions. POM, Group 2b includes several PAHs sampled
for at SEWA including acenaphthene, fluorene, and perylene. A single concentration
of acenaphthene failed a screen, but this pollutant was not identified as a pollutant of
interest for SEWA.
• POM, Group 2d ranks ninth for total emissions and sixth for its toxicity-weighted
emissions. POM, Group 2d does not includes any PAHs sampled for at SEWA. POM,
Group 5a also has the eighth highest toxicity-weighted emissions for King County.
Benzo(a)pyrene is part of POM, Group 5a. A single concentration of benzo(a)pyrene
failed a screen, but this pollutant was not identified as a pollutant of interest for
SEWA.
Observations from Table 23-7 for SEWA include the following:
• Toluene, xylenes, and hexane are the highest emitted pollutants with noncancer RfCs
in King County. The quantity of the emissions of these pollutants are considerably
higher than the emissions for the pollutants topping the emissions-based list in
Table 23-6.
• Acrolein is the pollutant with the highest toxicity-weighted emissions (of the
pollutants with noncancer RfCs) for King County, followed by formaldehyde and
1,3-butadiene. 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.
23-34
-------
• 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 23-7.
• Naphthalene, 1,3-butadiene, and nickel are among SEWA's pollutants of interest that
also appear among those with the highest toxicity-weighted emissions, although none
of these appear among the highest emitted (of those with a noncancer RfC).
• Arsenic, carbon tetrachloride, and 1,2-dichloroethane are pollutants of interest for
SEWA that appear on neither emissions-based list.
23.6 Summary of the 2014 Monitoring Data for SEWA
Results from several of the data analyses described in this section include the following:
~~~ Thirteen pollutants failed at least one screen for SEWA.
~~~ Acetaldehyde has the highest annual average concentration for SEWA, although all
of the pollutants of interest for SEWA have annual average concentrations less than
1 ug nr.
~~~ The annual average concentration of nickel for SEWA is the third highest among
NMP sites sampling PMio metals. Conversely, the annual average concentration of
formaldehyde for SEWA is the lowest among NMP sites sampling carbonyl
compounds.
~~~ Concentrations of benzene exhibit an overall decreasing trend over the period
sampling period at SEWA. Concentrations of naphthalene decreased significantly for
2014. The number of non-detects of 1,2-dichloroethane has decreased considerably at
SEWA in recent years.
23-35
-------
24.0 Data Quality
This section discusses the data quality of the ambient air measurements that constitute the
2014 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 program. 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 designed to control and evaluate the various phases of the measurement process (sampling,
preparation, analysis, etc.) to ensure that the total measurement quality meets the overall program
data quality needs. In accordance with ERG's EPA-approved QAPP (ERG, 2013), the following
MQOs were assessed: completeness, precision, and accuracy (also called bias).
The quality assessments presented in this section show that the 2014 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 and ASTM D7614 for hexavalent
chromium measurement (which are both 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 10 percent. Audit samples show that ERG is meeting the accuracy
requirements of the NATTS TAD (EPA, 2009b). 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, 2013). The MQO of 85 percent completeness was met by all but
seven of 108 site-method combinations. 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 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 sampler 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 sampler 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 2014 sampling year, where possible, duplicate or collocated samples were
collected on at least 10 percent of the scheduled sample days, as outlined in the EPA-approved
QAPP. This provides a minimum of six pairs of either duplicate or collocated samples per site
and method. 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. 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, additional replicate samples 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 data dispersion compared to
the mean. CV can be calculated two ways. The first, which expresses the CV as a ratio of
the standard deviation and the mean, is used for a single variable. The second, which is
provided below, is ideal when comparing paired values, such as a primary concentration
and a duplicate concentration. 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.
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).
CVs were based on every pair of duplicate or collocated samples collected during the
program year. However, only measurements at or above the MDL were used in these
calculations. 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 2014 NMP method precision for VOCs, SNMOCs, carbonyl
compounds, PAHs, metals, and hexavalent chromium, presented as the average CV (expressed as
a percentage). CVs exceeding the 15 percent MQO are bolded in the table. Four of the six
analytical methods met the program MQO of 15 percent CV for precision. TO-13A/PAHs and
hexavalent chromium results did not meet the MQO of 15 percent, although they are just outside
CV = 100X
Where:
24-3
-------
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 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)
9.63
2,782
3,366
SNMOC
7.50
340
428
Carbonyl Compounds
(TO-11 A)
5.41
1,578
1,579
PAHs
(TO-13 A)
18.73
291
314
Metals Analysis
(Method IO-3.5/FEM)
12.98
1,760
2,168
Hexavalent Chromium
(ASTM D7614)
17.96
4
4
MQO
15.00 percent CV
Bold = CV greater than or equal to 15 percent
Tables 24-2 through 24-7 present method precision for VOCs, SNMOCs, 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, as
these may result in relatively large CVs.
24-4
-------
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 94.84 percent (dichloromethane for GLKY). The CV for
dichloromethane for GLKY is based on four pairs of samples greater than the MDL. In three of
the four 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 (17 pollutants) to 15 (methyl isobutyl ketone). Dichloromethane is the only other
pollutant 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 (1,1-dichloroethene) to 23.51 percent (dichloromethane). For 1,1-dichloroethene, 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 5.53 percent (DEMI) to 14.79 percent
(SEW A). None of the sites have a site-specific average CV greater than or equal tol5 percent.
Note that TVKY collected collocated samples more frequently than the 10 percent requirement.
The overall average method precision for VOCs is 9.63 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 only two of the sites shown in Table 24-2 (PXSS and TVKY);
the remainder collected duplicate VOC samples. 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 meets the MQO of 15 percent, with the variability associated with collocated samples
(9.48 percent) slightly less than the variability associated with duplicate samples (10.06 percent).
24-5
-------
Table 24-2. VOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant
Pollutant
BTUT
CHNJ
CSNJ
DEMI
ELNJ
GLKY
Acetylene
2.57
0.79
1.75
2.37
3.62
3.18
fcrt-Amyl Methyl Ether
--
--
--
--
--
--
Benzene
4.04
7.90
2.17
10.31
12.57
6.39
Bromochloro methane
~
--
--
~
~
--
Bromodichloromethane
~
~
~
--
~
~
Bromoform
~
~
~
~
--
~
Bromomethane
5.58
9.34
3.45
~
8.76
5.45
1.3 -Butadiene
8.92
6.92
4.54
2.75
5.26
15.80
Carbon Tetrachloride
4.16
9.07
1.92
2.42
9.72
22.30
Chlorobenzene
--
--
--
2.86
61.16
--
Chloroethane
24.34
53.64
33.10
5.11
0.00
~
Chloroform
6.72
4.75
3.65
27.63
2.37
3.02
Chloromethane
2.91
7.36
5.57
2.85
3.57
1.41
Chloroprene
--
--
--
--
--
--
Dibromochloro methane
~
~
~
--
~
~
1,2-Dibromoethane
~
~
~
~
--
~
«/-Dichlorobenzene
--
~
~
~
~
--
o-Dichlorobenzene
~
--
~
~
~
~
/?-Dichlorobcnzcnc
8.66
~
--
~
13.71
~
Dichlorodifluoromethane
3.85
1.05
1.23
2.22
2.65
0.36
1,1 -Dichloroethane
~
~
~
--
--
~
1,2-Dichloroethane
4.99
7.17
4.23
5.87
4.99
4.13
1,1 -Dichloroethene
--
--
~
~
~
--
67.Y-1.2-Dichlorocthvlcnc
~
~
--
~
~
~
trans-1,2-Dichloroethylene
~
~
~
--
~
~
Dichloromethane
59.22
6.94
7.82
9.86
7.64
94.84
1,2-Dichloropropane
--
~
~
~
--
--
cis-1,3 -Dichloropropene
~
--
~
~
~
~
trans-1,3 -Dichloropropene
~
~
--
~
~
~
Dichlorotetrafluoroethane
5.03
4.15
2.29
1.65
4.31
2.52
Ethyl Acrylate
~
~
~
--
--
~
Ethyl tert-Butyl Ether
10.67
~
~
~
5.04
--
Ethylbenzene
8.22
31.76
5.44
3.47
3.23
37.10
Hexachloro-1,3 -butadiene
--
--
--
~
~
~
Methyl Isobutyl Ketone
26.22
20.91
4.19
10.64
17.97
0.00
Methyl Metliacrylate
~
~
~
--
5.44
~
Methyl tert-Butyl Ether
14.14
18.43
27.30
~
3.33
"
n-Octane
6.48
14.04
3.91
6.38
5.96
35.76
— = 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
BTUT
CHNJ
CSNJ
DEMI
ELNJ
GLKY
Propylene
14.70
21.86
3.76
4.29
3.69
10.92
Styrene
9.46
6.72
38.25
6.02
11.73
20.20
1,1,2,2 -T etrachloroethane
--
--
--
--
--
--
T etrachloroethylene
4.77
3.30
2.65
8.12
3.83
Toluene
21.77
15.49
16.28
3.90
2.78
8.58
1,2,4 -Trichlorobenzene
--
--
--
--
--
--
1,1,1 -T richloroethane
11.47
--
--
--
--
--
1,1,2 -T richloroethane
--
--
--
--
--
--
T richloroethylene
11.59
--
--
--
0.00
--
T richlorofluoro methane
3.47
1.54
1.85
2.07
2.65
0.42
T richlorotrifluoroethane
3.46
1.10
2.88
1.88
2.79
1.05
1,2,4 -T rimethylbenzene
9.86
12.88
12.17
3.98
13.68
--
1,3,5 -Trimethylbenzene
10.03
6.73
12.22
4.30
2.09
--
Vinyl chloride
--
--
10.88
--
--
--
m,p-Xylene
27.07
41.57
3.22
4.01
4.22
16.98
o-Xylene
21.22
34.65
5.64
3.26
2.64
19.84
Average CV by Site
11.85
13.46
8.24
5.53
7.46
14.77
# of Pairs Collected by Site
6
6
5
6
5
4
— = 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
OCOK
PXSS
ROIL
Acetylene
4.98
4.34
4.17
2.91
1.18
5.82
fcrt-Amyl Methyl Ether
~
--
--
--
~
--
Benzene
4.91
8.35
4.27
10.95
2.16
3.13
Bromochloro methane
--
~
~
~
--
--
Bromodichloromethane
~
20.61
~
~
2.89
~
Bromoform
~
0.00
4.88
~
~
~
Bromomethane
22.26
10.73
4.79
10.88
5.24
0.00
1.3 -Butadiene
6.44
3.32
9.01
11.76
5.13
1.88
Carbon Tetrachloride
24.95
6.26
4.24
7.35
4.13
19.50
Chlorobenzene
--
--
0.00
--
--
--
Chloroethane
~
--
11.54
30.15
~
7.86
Chloroform
7.77
9.79
2.07
6.48
4.00
2.95
Chloromethane
3.61
5.12
4.67
4.37
2.08
4.08
Chloroprene
--
~
--
--
--
--
Dibromochloro methane
~
9.73
--
~
0.00
~
1,2-Dibromoethane
~
--
--
~
~
~
«/-Dichlorobenzene
~
~
~
--
~
~
o-Dichlorobenzene
~
~
~
~
"
~
/?-Dichlorobcnzcnc
--
~
~
~
8.65
--
Dichlorodifluoromethane
2.91
5.26
2.95
2.14
3.03
2.60
1,1 -Dichloroethane
~
--
--
~
~
~
1,2-Dichloroethane
7.14
10.14
4.66
4.80
6.39
3.07
1,1 -Dichloroethene
~
~
~
--
--
~
67.Y-1.2-Dichlorocthvlcnc
--
~
~
~
~
--
trans-1,2-Dichloroethylene
~
--
~
~
~
~
Dichloromethane
19.89
7.10
5.39
10.38
42.99
10.66
1,2-Dichloropropane
~
~
--
--
~
~
cis-1,3 -Dichloropropene
~
~
~
~
--
~
trans-1,3 -Dichloropropene
--
~
~
~
~
--
Dichlorotetrafluoroethane
4.00
5.56
3.01
2.25
5.57
3.48
Ethyl Acrylate
~
--
--
~
~
~
Ethyl tert-Butyl Ether
10.92
3.36
3.63
--
~
~
Ethylbenzene
5.78
7.08
9.09
7.95
1.53
4.53
Hexacliloro-1,3 -butadiene
--
~
~
~
--
--
Methyl Isobutyl Ketone
27.19
16.68
24.84
34.75
24.57
18.73
Methyl Methacrylate
18.20
--
--
~
8.32
~
Methyl tert-Butyl Ether
~
~
~
--
~
~
n-Octane
11.52
3.13
21.42
14.94
5.43
8.32
— = 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
OCOK
PXSS
ROIL
Propylene
8.71
7.51
7.61
21.60
9.95
3.16
Styrene
31.68
3.01
3.25
24.69
7.72
33.55
1,1,2,2 -T etrachloroethane
--
--
--
--
--
--
T etrachloroethylene
1.76
8.47
6.26
5.06
3.07
8.12
Toluene
16.86
5.05
5.12
9.21
1.14
3.87
1,2,4 -Trichlorobenzene
--
--
--
--
--
--
1,1,1 -T richloroethane
--
--
--
--
--
--
1,1,2 -T richloroethane
--
--
--
--
--
--
T richloroethylene
--
--
--
--
--
0.00
T richlorofluoro methane
4.09
6.62
2.91
2.38
2.22
1.94
T richlorotrifluoroethane
3.07
3.94
3.38
2.87
1.79
3.06
1,2,4 -T rimethylbenzene
8.00
8.84
14.24
6.19
2.62
12.94
1,3,5 -Trimethylbenzene
9.71
6.90
7.44
16.28
5.29
14.73
Vinyl chloride
--
--
--
--
--
--
m,p-Xylene
5.03
6.28
7.27
6.19
1.60
3.57
o-Xylene
6.06
7.51
8.03
6.67
2.07
3.75
Average CV by Site
10.67
7.17
6.79
10.53
6.10
7.13
# of Pairs Collected by Site
5
3
6
6
4
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.
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
S4MO
SEWA
SPIL
TMOK
TOOK
TROK
Acetylene
2.87
2.52
30.47
4.30
9.02
4.26
fcrt-Amyl Methyl Ether
--
~
--
--
--
--
Benzene
23.32
4.71
24.63
5.69
5.57
4.98
Bromochloro methane
~
--
--
~
~
~
Bromodichloromethane
6.43
~
~
--
~
~
Bromoform
~
~
~
~
--
~
Bromomethane
4.56
~
12.76
3.58
7.33
3.70
1.3 -Butadiene
10.86
6.82
3.60
9.56
9.07
8.77
Carbon Tetrachloride
15.30
2.72
21.30
7.97
7.72
6.83
Chlorobenzene
--
--
--
--
~
--
Chloroethane
18.69
~
22.61
4.88
--
~
Chloroform
4.73
22.48
6.47
6.31
7.24
6.55
Chloromethane
3.63
3.16
4.36
7.91
8.98
3.94
Chloroprene
--
--
--
~
~
--
Dibromochloro methane
~
~
~
--
~
~
1,2-Dibromoethane
~
~
~
~
--
~
«/-Dichlorobenzene
--
~
~
~
~
--
o-Dichlorobenzene
~
--
~
~
~
~
/?-Dichlorobcnzcnc
5.71
~
--
18.58
21.43
29.74
Dichlorodifluoromethane
0.79
2.95
2.41
3.99
8.83
2.76
1,1 -Dichloroethane
~
~
~
--
--
~
1,2-Dichloroethane
8.53
5.45
9.70
7.05
1.92
5.07
1,1 -Dichloroethene
--
--
~
~
~
--
67.Y-1.2-Dichlorocthvlcnc
~
~
--
~
~
~
trans-1,2-Dichloroethylene
~
~
~
5.66
~
6.15
Dichloromethane
18.19
61.92
4.43
19.83
15.57
28.79
1,2-Dichloropropane
--
~
~
--
--
--
cis-1,3 -Dichloropropene
~
--
~
~
~
~
trans-1,3 -Dichloropropene
~
~
--
~
~
~
Dichlorotetrafluoroethane
7.06
8.96
5.94
4.37
8.25
2.70
Ethyl Acrylate
~
~
~
--
--
~
Ethyl tert-Butyl Ether
6.07
~
~
~
~
--
Ethylbenzene
9.54
39.31
3.85
7.05
11.87
10.57
Hexachloro-1,3 -butadiene
--
--
--
~
~
~
Methyl Isobutyl Ketone
19.62
42.09
9.10
24.95
12.08
15.83
Methyl Metliacrylate
~
~
~
17.37
--
~
Methyl tert-Butyl Ether
--
~
~
--
~
--
n-Octane
18.23
8.67
4.15
7.55
3.81
4.98
— = 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
S4MO
SEWA
SPIL
TMOK
TOOK
TROK
Propylene
4.42
8.74
34.64
9.84
6.93
4.06
Styrene
12.62
6.17
3.21
21.99
24.44
10.87
1,1,2,2 -T etrachloroethane
--
--
--
--
--
--
T etrachloroethylene
7.68
3.17
37.36
6.41
7.51
2.59
Toluene
8.05
30.19
30.91
6.56
3.08
7.80
1,2,4 -Trichlorobenzene
--
--
--
--
--
--
1,1,1 -T richloroethane
--
--
--
--
--
--
1,1,2 -T richloroethane
--
--
--
--
--
--
T richloroethylene
--
--
7.06
--
--
--
T richlorofluoro methane
0.85
2.85
2.83
3.85
9.01
2.70
T richlorotrifluoroethane
2.14
4.26
3.60
4.05
6.66
1.62
1,2,4 -T rimethylbenzene
7.11
3.10
4.80
9.97
7.09
11.95
1,3,5 -Trimethylbenzene
19.36
4.16
5.05
15.29
6.26
11.67
Vinyl chloride
--
--
--
--
--
--
m,p-Xylene
7.95
35.47
2.28
6.07
15.04
11.20
o-Xylene
9.43
30.41
5.32
6.33
15.18
11.38
Average CV by Site
9.42
14.79
11.65
9.18
9.60
8.52
# of Pairs Collected by Site
6
7
5
5
5
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.
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
6.14
3.42
127
5.03
5.19
3.66
fcrt-Amyl Methyl Ether
~
~
0
--
--
--
Benzene
12.04
3.23
127
8.07
8.17
7.10
Bromochloro methane
~
--
0
--
--
--
Bromodichloromethane
~
~
5
9.98
13.52
2.89
Bromoform
~
~
2
2.44
2.44
~
Bromomethane
11.14
20.20
54
8.32
8.34
8.19
1.3 -Butadiene
9.25
3.85
110
7.18
7.17
7.19
Carbon Tetrachloride
6.51
3.74
127
9.41
9.86
5.32
Chlorobenzene
--
--
4
21.34
21.34
--
Chloroethane
11.61
~
31
18.63
19.26
11.61
Chloroform
14.66
5.04
117
7.73
7.56
9.33
Chloromethane
5.05
7.60
127
4.61
4.73
3.57
Chloroprene
--
--
0
--
--
--
Dibromochloro methane
~
~
3
4.87
9.73
0.00
1,2-Dibromoethane
~
~
0
--
--
~
«/-Dichlorobenzene
--
~
0
--
--
"
o-Dichlorobenzene
~
--
0
--
--
~
/?-Dichlorobcnzcnc
~
~
15
15.21
16.30
8.65
Dichlorodifluoromethane
2.89
3.40
127
2.91
2.91
2.96
1,1 -Dichloroethane
9.20
~
5
9.20
--
9.20
1,2-Dichloroethane
8.67
6.78
113
6.04
5.87
7.53
1,1 -Dichloroethene
0.00
--
1
0.00
--
0.00
6V.S-1,2-Dichlorocthy lcnc
14.14
~
1
14.14
--
14.14
trans-1,2-Dichloroethylene
7.31
1.86
5
5.25
4.56
7.31
Dichloromethane
30.98
7.81
124
23.51
22.02
36.99
1,2-Dichloropropane
--
~
0
--
--
--
cis-1,3 -Dichloropropene
--
--
0
--
--
--
trans-1,3 -Dichloropropene
~
~
0
--
--
~
Dichlorotetrafluoroethane
6.53
3.97
127
4.58
4.42
6.05
Ethyl Acrylate
~
~
0
--
--
~
Ethyl tort-Butyl Ether
--
~
19
6.61
6.61
--
Ethylbenzene
12.84
8.10
118
11.41
11.88
7.19
Hexacliloro-1,3 -butadiene
~
--
0
--
--
~
Methyl Isobutyl Ketone
39.20
36.60
84
21.31
20.13
31.88
Methyl Metliacrylate
~
~
4
12.33
13.67
8.32
Methyl tcrt-But\i Ether
--
~
12
15.80
15.80
--
n-Octane
21.74
4.63
117
10.55
10.22
13.59
— = 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
Propylene
9.31
22.80
127
10.92
11.07
9.63
Styrene
46.18
11.59
74
16.67
15.53
26.95
1,1,2,2 -T etrachloroethane
--
--
0
--
--
--
T etrachloroethylene
7.64
7.38
74
7.11
7.32
5.36
Toluene
36.25
11.42
127
12.22
11.49
18.70
1,2,4 -Trichlorobenzene
--
--
0
--
--
--
1,1,1 -T richloroethane
--
--
1
11.47
11.47
--
1,1,2 -T richloroethane
7.79
--
3
7.79
--
7.79
T richloroethylene
6.73
--
8
5.08
4.66
6.73
T richlorofluoro methane
21.56
3.34
127
3.96
3.08
11.89
T richlorotrifluoroethane
3.79
3.46
126
3.04
3.07
2.79
1,2,4 -T rimethylbenzene
14.50
10.99
95
9.21
9.28
8.56
1,3,5 -Trimethylbenzene
10.85
13.75
59
9.59
9.76
8.07
Vinyl chloride
7.07
--
17
8.97
10.88
7.07
m,p-Xylene
12.59
5.30
120
11.15
11.60
7.09
o-Xylene
14.26
8.22
118
11.09
11.42
8.17
Average CV by Site
13.39
8.74
2,782
9.63
10.06
9.48
# of Pairs Collected by Site
25
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.
BOLD ITALICS = EPA-designated NATTS Site
24-13
-------
24.2.2 SNMOC Method Precision
The SNMOC method precision for duplicate samples is presented in Table 24-3 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 SNMOCs listed. The duplicate sample precision results from duplicate
samples exhibit low- to mid-level variability among the pollutants and sites, ranging from a CV
of 0.62 percent (isobutane for NBIL) to 32.07 percent (methylcyclohexane for NBIL). The CVs
for 34 pollutants are less than 15 percent for both sites; conversely, there is only one pollutant
listed where the CV is greater than or equal to 15 percent for both sites: 2-methylhexane.
The pollutant-specific average CV, as shown in orange in Table 24-3, ranges from
0.92 percent («-dodecane) to 22.58 percent (2-methylhexane). The site-specific average CV, as
shown in green in Table 24-3, ranges from 7.21 percent (NBIL) to 7.78 percent (BTUT); these
are the only sites at which duplicate SNMOC samples were collected. No collocated SNMOC
samples were collected during the 2014 program year. The overall average method precision for
SNMOCs is 7.50 percent. Note that the results for TNMOC were not included in the precision
calculations.
Table 24-3. SNMOC Method Precision: Coefficient of Variation
Based on Duplicate Samples by Site and Pollutant
Pollutant
BTUT
NBIL
# of
pairs
Average
by
Pollutant
Acetylene
4.72
2.85
9
3.79
Benzene
10.70
6.77
9
8.74
1.3 -Butadiene
2.72
--
1
2.72
//-Butane
7.10
1.73
9
4.42
1-Butene
--
--
0
--
67.Y-2-Bute nc
4.72
--
3
4.72
;ra«.v-2-Butcne
20.00
--
3
20.00
Cyclohexane
8.32
4.98
8
6.65
Cyclopentane
4.48
23.67
5
14.07
Cyclopentene
--
--
0
--
w-Decane
6.68
6.73
5
6.71
— = 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
24-14
-------
Table 24-3. SNMOC Method Precision: Coefficient of Variation
Based on Duplicate Samples by Site and Pollutant (Continued)
Pollutant
BTUT
NBIL
# of
pairs
Average
by
Pollutant
1-Decene
—
—
0
--
w-Diethylbenzene
—
—
0
--
/j-Dicthvlbcnzcne
—
—
0
--
2,2-Dimethylbutane
7.66
2.13
4
4.89
2,3 -Dimethylbutane
3.44
5.63
8
4.54
2,3 -Dimethylpentane
3.38
4.64
9
4.01
2,4-Dimethylpentane
3.52
8.41
7
5.96
n-Dodecane
0.92
--
1
0.92
1-Dodecene
--
--
0
--
Ethane
1.87
2.63
9
2.25
2-Ethyl-l-butene
--
--
0
--
Ethylbenzene
4.25
2.70
6
3.47
Ethylene
7.11
3.68
9
5.39
OT-Ethyltoluene
2.11
4.72
4
3.41
o-Ethyltoluene
1.00
--
1
1.00
/?-Ethyltolucnc
14.51
12.31
3
13.41
//-Heptane
1.91
6.70
8
4.30
1-Heptene
--
--
0
--
w-Hexane
6.51
2.70
9
4.60
1-Hexene
--
~
0
--
67.Y-2-Hc\cnc
--
~
0
--
-------
Table 24-3. SNMOC Method Precision: Coefficient of Variation
Based on Duplicate Samples by Site and Pollutant (Continued)
Pollutant
BTUT
NBIL
# of
pairs
Average
by
Pollutant
3-Methylhexane
15.71
—
3
15.71
2-Methylpentane
13.31
13.72
9
13.51
3-Methylpentane
3.70
4.79
9
4.25
//-Nona ne
5.78
7.13
5
6.45
1-Nonene
--
--
0
--
//-Octane
4.03
0.92
6
2.47
1-Octene
--
--
0
--
n-Pentane
2.98
1.02
9
2.00
1-Pentene
29.72
5.93
4
17.82
c/.v-2-Pcntcnc
7.25
--
3
7.25
;ra/?.v-2-Pcntene
3.60
--
4
3.60
fl-Pincnc
15.40
11.69
7
13.55
/>-Pincnc
--
--
0
--
Propane
1.65
1.30
9
1.47
n-Propylbenzene
3.55
2.93
2
3.24
Propylene
11.67
8.49
9
10.08
Propyne
--
--
0
--
Styrene
--
~
0
--
Toluene
21.86
2.81
9
12.33
/?-T ridccanc
--
--
0
--
1-Tridecene
--
~
0
--
1,2,3 -Trimethylbenzene
4.61
3.21
2
3.91
1,2,4 -T rimethylbenzene
8.57
2.12
8
5.35
1,3,5 -Trimethylbenzene
16.30
~
2
16.30
2,2,3 -Trimethylpentane
2.61
--
1
2.61
2,2,4 -T rimethy lpentane
4.66
9.70
9
7.18
2,3,4 -T rimethy lpentane
14.26
24.55
8
19.40
n-Undecane
10.68
~
1
10.68
1-Undecene
--
~
0
--
/w-Xylene/p-Xylene
6.88
6.16
8
6.52
o-Xylene
5.59
3.34
8
4.46
SNMOC (Sum of Knowns)
3.34
1.62
9
2.48
Sum of Unknowns
7.39
23.38
9
15.39
Average CV by Site
7.78
7.21
340
7.50
# of Pairs Collected by Site
6
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.
BOLD ITALICS = EPA-designated NATTS Site
24-16
-------
24.2.3 Carbonyl Compound Method Precision
Table 24-4 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 0.00 percent
(hexaldehyde for GLKY and propionaldehyde for NBIL) to 32.86 percent (2-butanone for
SYFL). The number of sites for which a given pollutant has a CV greater than or equal to
15 percent varies from none (six pollutants) to two (acetone and tolualdehydes).
The pollutant-specific average CV, as shown in orange in Table 24-4, ranges from
2.85 percent (acetaldehyde) to 9.57 percent (tolualdehydes). The site-specific average CV, as
shown in green in Table 24-4, ranges from 2.13 percent (NBNJ) to 11.56 percent (SYFL). None
of the sites collecting duplicate or collocated carbonyl compound samples have a site-specific
average CV greater than or equal to 15 percent. The overall average method precision is
5.41 percent for carbonyl compounds.
Sites at which duplicate samples were collected are highlighted in blue in Table 24-4
while sites at which collocated samples were collected are highlighted in purple. Collocated
carbonyl compound samples were collected at only three of the sites shown in Table 24-4
(DEMI, INDEM, and PXSS); the remainder collected duplicate samples. The average CV for
sites that collected duplicate samples was calculated and is shown at the end of Table 24-4 in
blue while the average CV for sites collecting collocated samples is shown in purple. The
average CV for both precision types meets the MQO of 15 percent, with the variability
associated with collocated samples (8.77 percent) greater than the variability associated with
duplicate samples (4.95 percent).
24-17
-------
Table 24-4. Carbonyl Compound Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant
Pollutant
AZFL
BTUT
CHNJ
CSNJ
DEMI
ELNJ
GLKY
Acetaldehyde
2.15
1.98
1.05
1.10
11.99
1.04
1.09
Acetone
6.40
3.62
4.72
14.47
1.83
16.13
1.49
Benzaldehyde
4.64
3.65
5.59
8.37
22.45
4.04
7.96
2-Butanone
13.81
2.30
6.39
7.98
5.24
11.17
2.17
Butyraldehyde
7.72
2.35
2.02
3.90
18.87
2.66
0.51
Crotonaldehyde
4.18
3.60
3.09
4.36
3.42
5.44
2.09
2,5-Dimethylbenzaldehyde
--
--
--
--
--
--
--
Formaldehyde
10.34
2.00
2.14
1.71
8.37
1.22
0.61
Hexaldehyde
12.74
7.09
6.86
4.56
11.00
5.04
0.00
Isovaleraldehyde
--
--
--
--
--
--
--
Propionaldehyde
4.14
2.64
1.68
2.34
6.29
1.11
1.43
Tolualdehydes
9.57
8.03
6.15
8.81
29.03
6.59
11.16
Valeraldehyde
11.12
4.98
6.25
4.33
6.71
3.66
5.33
Average CV by Site
7.89
3.84
4.18
5.63
11.38
5.28
3.08
# of Pairs Collected by Site
5
6
6
6
7
6
5
— = 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
24-18
-------
Table 24-4. Carbonyl Compound Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
GPCO
INDEM
NBIL
NBNJ
OCOK
OR II
PXSS
Acetaldehyde
1.44
8.63
1.84
1.50
5.95
1.08
4.41
Acetone
2.16
9.94
1.95
1.09
4.55
9.87
5.64
Benzaldehyde
4.17
12.83
7.00
4.62
6.66
3.40
4.78
2-Butanone
1.90
6.94
0.86
1.28
7.63
9.99
3.21
Butyraldehyde
1.99
10.26
2.09
1.45
4.48
4.26
7.92
Crotonaldehyde
3.21
8.94
3.96
2.33
5.84
4.06
6.64
2,5 -Dimethylbenzaldehyde
--
--
--
--
--
--
--
Formaldehyde
1.60
6.27
1.13
1.26
4.01
1.78
4.34
Hexaldehyde
2.83
10.97
4.33
2.02
6.44
5.32
4.31
Isovaleraldehyde
--
--
--
--
--
Propionaldehyde
2.70
8.76
0.00
0.50
6.56
1.58
4.51
Tolualdehydes
9.34
6.74
11.79
6.25
8.36
5.91
8.23
Valeraldehyde
6.51
13.94
6.04
1.12
6.03
9.66
6.02
Average CV by Site
3.44
9.47
3.73
2.13
6.05
5.17
5.46
# of Pairs Collected by Site
6
11
3
2
6
6
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.
BOLD ITALICS = EPA-designated NATTS Site
24-19
-------
Table 24-4. Carbonyl Compound Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
ROIL
S4MO
SEWA
SKFL
SPIL
SYFL
TMOK
Acetaldehyde
1.75
1.15
5.18
3.07
3.59
1.56
1.45
Acetone
7.58
4.13
1.95
2.94
7.72
22.30
5.81
Benzaldehyde
6.23
2.25
9.50
5.16
3.38
10.31
4.66
2-Butanone
8.71
3.91
2.06
6.56
4.99
32.86
4.42
Butyraldehyde
6.74
4.01
7.81
3.94
3.60
8.47
3.21
Crotonaldehyde
7.12
2.45
3.78
4.18
4.33
7.75
3.70
2,5-Dimethylbenzaldehyde
--
--
--
--
--
--
--
Formaldehyde
3.55
1.61
3.71
3.38
3.75
2.60
1.08
Hexaldehyde
5.83
7.28
6.63
2.49
6.01
6.44
3.35
Isovaleraldehyde
--
--
--
--
--
--
--
Propionaldehyde
2.73
2.73
5.08
3.02
2.10
3.79
1.85
Tolualdehydes
7.75
9.15
8.77
10.25
7.94
21.88
7.70
Valeraldehyde
6.92
6.12
4.84
7.23
5.07
9.22
6.45
Average CV by Site
5.90
4.07
5.39
4.75
4.77
11.56
3.97
# of Pairs Collected by Site
6
6
6
4
5
6
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.
BOLD ITALICS = EPA-designated NATTS Site
24-20
-------
Table 24-4. Carbonyl Compound Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
TOOK
TROK
WPIN
YUOK
# of
pairs
Average
by
Pollutant
Average
for
Duplicate
Pairs
Average
for
Collocated
Pairs
Acetaldehyde
1.09
0.70
3.81
2.71
148
2.85
2.10
8.34
Acetone
1.41
1.26
3.26
11.12
148
6.13
6.18
5.80
Benzaldehyde
4.17
3.95
8.84
4.93
142
6.54
5.61
13.35
2-Butanone
6.53
1.18
3.09
9.56
139
6.59
6.79
5.13
Butyraldehyde
2.71
1.83
4.59
5.29
145
4.91
3.89
12.35
Crotonaldehyde
2.44
1.50
5.23
11.22
147
4.59
4.36
6.33
2,5 -Dimethylbenzaldehyde
--
--
--
--
--
--
--
--
Formaldehyde
2.41
0.90
4.18
1.56
148
3.02
2.57
6.33
Hexaldehyde
2.74
3.99
6.57
5.77
148
5.62
5.20
8.76
Isovaleraldehyde
--
--
--
--
--
--
--
--
Propionaldehyde
1.22
2.13
5.40
3.28
148
3.10
2.64
6.52
Tolualdehydes
9.42
6.07
6.52
7.81
122
9.57
8.87
14.67
Valeraldehyde
6.18
7.35
7.08
5.33
143
6.54
6.22
8.89
Average CV by Site
3.67
2.80
5.33
6.24
1,578
5.41
4.95
8.77
# of Pairs Collected by Site
6
6
10
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.
BOLD ITALICS = EPA-designated NATTS Site
24.2.4 PAH Method Precision
The method precision results for collocated PAH samples are shown in Table 24-5 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.
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 sites had collocated
samplers. Therefore, the method precision for PAHs for 2014 is based on data from three sites
(DEMI, RUCA, and SEW A) and a total of only 18 sample pairs. The results from collocated
samples exhibit low- to high-level variability, ranging from a CV of 2.30 percent (perylene for
DEMI) to 67.91 percent (indeno(l,2,3-cd)pyrene for RUCA). The overall average method
precision was 18.73 percent, which is greater than the MQO of 15 percent CV.
24-21
-------
The pollutant-specific average CV, as shown in orange in Table 24-5, ranges from
2.30 percent (perylene) to 36.18 percent (benzo(a)anthracene). The site-specific average CVs, as
shown in green in Table 24-5, vary across the sites, from 9.35 percent for DEMI, to
19.32 percent for SEW A, and 34.68 percent for RUCA. None of the CVs for DEMI are greater
than or equal to 15 percent while many PAHs are greater than or equal to 15 percent for RUCA
and SEW A, with the CVs for RUCA most often the highest of the three sites. A review of the
individual sample pairs for RUCA shows that higher CVs for this site is being driven by two
sample pairs. One pair in particular (the July 22, 2014 sample pair) has rather poor precision,
with 11 of the 22 pollutants with an individual pollutant-specific CV greater than or equal to
15 percent, and six of these approaching or greater than 100 percent. If this sample is removed
from the CV calculation for the program, the MQO of 15 percent would be met.
24-22
-------
Table 24-5. PAH Method Precision: Coefficient of Variation
Based on Collocated Samples by Site and Pollutant
Pollutant
DEMI
RUCA
SEWA
# of
Pairs
Average by
Pollutant
Acenaphthene
9.71
9.39
11.05
18
10.05
Acenaphthylene
10.05
41.91
21.35
7
24.44
Anthracene
7.58
28.87
11.85
14
16.10
Benzo(a)anthracene
13.62
60.94
33.99
15
36.18
Benzo(a)pyrene
7.61
--
46.52
8
27.07
Benzo(b)fluoranthene
11.52
60.43
21.64
16
31.20
Benzo(e)pyrene
10.76
57.25
27.25
15
31.75
Benzo(g,h,i)perylene
8.64
49.23
25.84
16
27.90
Benzo(k)fluoranthene
10.25
--
--
6
10.25
Chrysene
7.84
44.20
22.29
18
24.78
Coronene
8.61
31.69
19.45
14
19.92
Cyclopenta[cd]pyrene
14.31
--
--
1
14.31
Dibenz(a,h)anthracene
6.19
--
--
4
6.19
Fluoranthene
8.57
13.77
12.69
18
11.68
Fluorene
10.74
13.15
8.89
15
10.93
9-Fluorenone
10.19
14.18
10.41
18
11.60
Indeno( 1,2,3 -cd)pyrene
9.43
67.91
25.54
14
34.29
Naphthalene
11.15
12.83
8.43
18
10.80
Perylene
2.30
--
--
2
2.30
Phenanthrene
7.96
12.07
11.07
18
10.37
Pyrene
8.65
18.39
20.15
18
15.73
Retene
10.10
53.28
9.44
18
24.27
Average CV by Site
9.35
34.68
19.32
291
18.73
# of Pairs Collected by Site
6
6
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 in the final column of the table.
BOLD ITALICS = EPA-designated NATTS Site
24-23
-------
24.2.5 Metals Method Precision
The method precision for all collocated metals samples are presented 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 metals listed. All samples evaluated in this section are collocated samples.
The results from collocated samples exhibit low- to mid-level variability, ranging from a CV of
0 percent (beryllium and mercury for GLKY and cobalt for UNVT) to 45.43 percent (chromium
for BOMA). The number of sites for which a given pollutant has a CV greater than or equal to
15 percent varies from none (manganese) to five (cadmium), with several metals exceeding
15 percent CV for only one or two sites. 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 24-6, ranges from
6.22 percent (manganese) to 20.70 percent (nickel), with four of the 11 metals with an average
CV greater than 15 percent. The site-specific average CV, as shown in green in Table 24-6,
ranges from 6.72 percent (UNVT) to 20.94 percent (BOMA). Three sites (BOMA, GLKY and
S4MO) have site-specific average CVs greater than or equal to 15 percent. The overall average
method precision for metals is 12.98 percent.
24-24
-------
Table 24-6. Metals Method Precision: Coefficient of Variation
Based on Collocated Samples by Site and Pollutant
Pollutant
ASKY-M
BOMA
BTUT
GLKY
GPCO
Antimony
2.43
11.34
4.59
21.00
1.98
Arsenic
6.28
26.97
15.37
17.32
8.48
Beryllium
15.67
18.52
19.30
0.00
11.66
Cadmium
3.44
27.31
16.06
27.35
6.02
Chromium
10.74
45.43
--
0.51
--
Cobalt
7.39
18.14
5.78
19.73
6.58
Lead
3.78
5.86
1.06
23.53
2.94
Manganese
3.37
4.74
3.33
12.62
3.73
Mercury
25.49
27.21
28.28
0.00
--
Nickel
33.57
32.74
5.35
37.78
12.35
Selenium
5.07
12.04
17.26
12.54
12.34
Average CV by Site
10.66
20.94
11.64
15.67
7.34
# of Pairs Collected by Site
5
35
7
25
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
Table 24-6. Metals Method Precision: Coefficient of Variation
Based on Collocated Samples by Site and Pollutant (Continued)
Pollutant
S4MO
TOOK
UNVT
# of
pairs
Average by
Pollutant
Antimony
12.99
16.57
1.78
204
9.08
Arsenic
20.04
4.74
5.80
167
13.12
Beryllium
23.52
10.55
--
99
14.17
Cadmium
13.24
21.00
23.05
206
17.18
Chromium
--
7.10
--
54
15.94
Cobalt
14.47
7.14
0.00
199
9.91
Lead
12.16
8.59
4.32
206
7.78
Manganese
11.80
5.71
4.46
206
6.22
Mercury
12.96
16.46
--
86
18.40
Nickel
28.18
6.87
8.73
191
20.70
Selenium
13.49
3.56
5.66
142
10.24
Average CV by Site
16.28
9.84
6.72
1,760
12.98
# of Pairs Collected by Site
60
56
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 in the final column of the table.
BOLD ITALICS = EPA-designated NATTS Site
24-25
-------
24.2.6 Hexavalent Chromium Method Precision
Table 24-7 presents the method precision results from collocated hexavalent chromium
samples as the C V per site and the overall average CV for the method. All samples evaluated in
this section are collocated samples. Only two NATTS sites sampled hexavalent chromium in
2014, RIVA and S4MO, although sampling was discontinued in July 2014 at S4MO. The site-
specific CV ranges from 6.13 percent for RIVA to 29.80 percent for S4MO; note that these CVs
are based on two collocated pairs for each site. The overall average method precision for
hexavalent chromium is 17.96 percent, as shown in orange in Table 24-7, which is greater than
the MQO of 15 percent CV for method precision. Note, however, that the precision calculations
are based on only four collocated pairs.
Table 24-7. Hexavalent Chromium Method Precision: Coefficient of Variation
Based on Collocated Samples by Site
Pollutant
RIVA
S4MO
# of
pairs
Average by
Pollutant
Hexavalent Chromium
6.13
29.80
4
17.96
# of Pairs Collected by Site
2
2
4
Bold = CV greater than or equal to 15 percent
Orange shading indicates the average CV for this method.
BOLD ITALICS = EPA-designated NATTS Site
24.3 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 year. 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 to provide an indication of
analytical precision for monitoring sites unable to collect duplicate or collocated samples
(i.e., samplers "unequipped" to collect duplicate or collocated samples). Individual samples with
replicate analyses were also factored into the CV calculations for analytical precision. However,
only results at or above the MDL were used in these calculations, similar to the calculation of
method precision discussed in Section 24.2.
24-26
-------
Table 24-8 presents the 2014 NMP analytical precision for VOCs, SNMOCs, carbonyl
compounds, PAHs, metals, and hexavalent chromium, presented as average CV (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 9 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 24-8 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.
Table 24-8. 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)
6.09
6,923
8,284
SNMOCs
3.78
2,047
2,601
Carbonyl Compounds
(TO-11A)
2.59
3,550
3,552
PAHs
(TO-13A)
4.87
2,607
2,887
Metals Analysis
(Method IO-3.5/FEM)
6.51
4,276
5,226
Hexavalent Chromium
(ASTMD7641)
8.75
8
8
MQO
15.00 percent CV
Bold = CV greater than or equal to 15 percent
Tables 24-9 through 24-14 present analytical precision for VOCs, SNMOCs, 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 24-9 through 24-14, the number of
pairs in comparison to the respective tables listed for duplicate or collocated analyses in
Tables 24-2 through 24-7 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 replicates run on individual samples. This is also the reason the number of sites provided
in Tables 24-9 through 24-14 is higher than Tables 24-2 through 24-7. The replicate analyses of
duplicate, collocated, and individual samples indicate that the analytical precision level is within
the program MQOs.
24-27
-------
24.3.1 VOC Analytical Precision
Table 24-9 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. 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 34.69 percent (methyl tert-butyl ether for SPIL). The number of sites for which a
given pollutant has a CV greater than or equal to 15 percent varies from none (32 pollutants) to
four (styrene).
The pollutant-specific average CV, as shown in orange in Table 24-9, ranges from
1.98 percent (1,1,1-trichloroethane) to 16.02 percent (cis-1,2-dichloroethylene). The CV for
6'/.s -1,2-dichloroethylene is the only pollutant-specific CV greater than or equal to 15 percent and
is based on the replicate analysis of two samples collected at TVKY, the only site for which at
least one pair of measurements greater than the MDL were collected. The site-specific average
CV, as shown in green in Table 24-9, ranges from 3.12 percent (SPAZ) to 7.35 percent (CHNJ).
The overall average analytical precision is 6.09 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 24-9,
sites at which collocated samples were collected are highlighted in purple, and sites for which
replicates were run on individual samples are highlighted in brown. Collocated VOC samples
were collected at only two of the sites shown in Table 24-9 (PXSS, and TVKY); replicates were
run on individual VOC samples for seven sites, and the remainder of sites collected duplicate
VOC samples. The average CV for sites that collected duplicate samples was calculated and is
shown at the end of Table 24-9 in blue, the average CV for sites collecting collocated samples is
shown in purple, and the average CV for sites for which replicates were run on individual
samples is shown in brown. The average CV for all three precision types meets the MQO of
15 percent, with the CV ranging from 5.01 percent (replicates run on individual samples) to
6.53 percent (replicates run on collocated samples).
24-28
-------
Table 24-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant
Pollutant
ASKY
ATKY
BLKY
BTUT
CCKY
CHNJ
Acetylene
4.03
2.56
2.54
4.25
3.30
2.49
fcrt-Amyl Methyl Ether
--
--
--
~
--
~
Benzene
4.88
4.05
4.05
6.14
6.77
6.11
Bromochloro methane
--
~
~
--
~
--
Bromodichloromethane
--
--
~
~
~
~
Bromoform
--
~
~
~
--
~
Bromomethane
8.88
7.97
8.64
4.28
3.63
6.87
1.3 -Butadiene
11.47
6.92
5.79
6.08
8.71
9.04
Carbon Tetrachloride
2.98
2.02
2.31
5.05
3.31
4.27
Chlorobenzene
--
--
--
--
~
--
Chloroethane
12.12
9.38
7.10
6.33
--
5.88
Chloroform
3.85
7.32
6.76
7.97
8.63
6.32
Chloromethane
1.91
1.78
1.57
4.40
2.72
2.07
Chloroprene
--
~
--
--
~
--
Dibromochloro methane
~
--
~
~
~
~
1,2-Dibromoethane
~
~
~
~
--
~
«/-Dichlorobenzene
~
~
--
~
~
~
o-Dichlorobenzene
--
~
~
~
~
--
p-Dichlorobenzene
~
~
~
13.04
~
~
Dichlorodifluoromethane
1.94
1.42
1.34
4.79
2.85
2.25
1,1 -Dichloroethane
~
2.77
~
--
--
~
1,2-Dichloroethane
5.36
8.05
8.45
6.53
8.32
8.83
1,1 -Dichloroethene
--
0.00
--
~
~
--
cis-1,2 -Dichloroethy lene
~
--
~
--
~
~
trans-1,2-Dichloroethylene
~
--
~
~
~
~
Dichloromethane
3.07
2.24
2.01
9.54
4.94
7.21
1,2-Dichloropropane
~
~
--
~
--
~
cis-1,3 -Dichloropropene
--
~
~
~
~
--
trans-1,3 -Dichloropropene
~
~
~
--
~
~
Dichlorotetrafluoroethane
5.03
6.45
4.50
5.22
2.99
4.49
Ethyl Acrylate
~
--
~
~
--
~
Ethyl tert-Butyl Ether
~
~
--
4.18
~
~
Ethylbenzene
8.51
16.32
9.58
4.30
8.69
9.88
Hexachloro-1,3 -butadiene
--
~
~
--
~
--
Methyl Isobutyl Ketone
7.29
5.58
7.95
7.01
11.57
5.67
Methyl Methacrylate
~
6.73
~
~
--
~
Methyl tert-Butyl Ether
3.01
--
--
1.90
~
6.90
— = 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 individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-29
-------
Table 24-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
ASKY
ATKY
BLKY
BTUT
CCKY
CHNJ
//-Octane
9.89
14.66
11.63
5.33
9.21
9.36
Propylene
2.42
1.02
3.23
4.88
2.93
2.76
Styrene
10.68
13.30
17.72
6.53
10.47
23.55
1,1,2,2 -T etrachloroethane
--
--
--
--
--
--
T etrachloroethylene
9.43
5.98
--
6.41
--
3.55
Toluene
4.21
4.70
5.48
4.22
9.94
6.00
1,2,4 -Trichlorobenzene
--
--
--
--
--
--
1,1,1 -T richloroethane
--
0.00
--
3.97
--
--
1,1,2 -T richloroethane
--
8.00
--
--
--
--
T richloroethylene
--
0.00
--
3.78
--
--
T richlorofluoro methane
1.35
1.32
1.51
4.54
2.58
2.29
T richlorotrifluoroethane
1.77
2.73
2.45
4.78
2.52
2.26
1,2,4 -T rimethylbenzene
9.40
12.58
12.86
4.40
8.11
19.75
1,3,5 -Trimethylbenzene
10.46
12.15
15.71
3.17
9.87
15.48
Vinyl chloride
--
0.95
1.35
--
0.00
--
m,p-Xylene
7.07
14.66
10.99
3.91
7.71
7.92
o-Xylene
6.77
15.96
9.90
3.68
9.27
9.98
Average CV by Site
6.07
6.24
6.62
5.35
6.21
7.35
# of Pairs Collected by Site
8
7
7
12
4
14
— = 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 individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-30
-------
Table 24-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
CSNJ
DEMI
ELNJ
GLKY
GPCO
LAKY
Acetylene
1.48
1.95
3.31
5.53
1.85
1.28
fcrt-Amyl Methyl Ether
--
--
~
--
--
~
Benzene
3.41
3.32
4.21
4.90
3.35
2.37
Bromochloro methane
--
~
--
~
~
--
Bromodichloromethane
--
~
~
--
~
~
Bromoform
~
--
~
~
~
~
Bromomethane
7.58
0.00
1.33
4.08
3.19
6.67
1.3 -Butadiene
9.15
4.14
9.41
14.33
5.20
5.10
Carbon Tetrachloride
2.13
3.24
3.72
1.63
2.65
1.77
Chlorobenzene
--
10.22
3.02
--
--
--
Chloroethane
9.26
4.70
9.90
6.92
4.37
0.00
Chloroform
5.31
3.28
6.43
4.44
6.08
3.38
Chloromethane
1.31
1.52
3.05
1.39
1.49
1.47
Chloroprene
--
--
--
~
--
--
Dibromochloro methane
~
~
--
--
~
~
1,2-Dibromoethane
~
--
~
~
~
~
«/-Dichlorobenzene
~
~
~
~
--
~
o-Dichlorobenzene
~
~
--
~
~
~
p-Dichlorobenzene
--
~
7.73
~
~
--
Dichlorodifluoromethane
1.49
1.65
3.18
1.24
1.73
1.37
1,1 -Dichloroethane
~
--
~
--
~
~
1,2-Dichloroethane
6.84
3.75
5.21
4.15
4.74
5.19
1,1 -Dichloroethene
~
~
--
~
--
~
cis-1,2 -Dichloroethy lene
--
~
~
~
~
--
trans-1,2-Dichloroethylene
~
~
~
--
~
~
Dichloromethane
4.53
2.10
3.94
3.14
2.32
1.84
1,2-Dichloropropane
~
--
~
~
--
~
cis-1.3 -Dichloropropene
~
~
--
~
~
~
trans-1,3 -Dichloropropene
--
~
~
~
~
--
Dichlorotetrafluoroethane
2.85
2.45
6.78
3.01
3.68
3.05
Ethyl Acrylate
~
--
~
--
~
~
Ethyl tort-Butyl Ether
~
~
5.60
~
5.30
~
Ethylbenzene
6.97
3.65
6.54
12.45
5.71
6.02
Hexachloro-1,3 -butadiene
--
~
--
~
--
--
Methyl Isobutyl Ketone
6.05
6.73
8.87
6.28
6.66
6.73
Methyl Methacrylate
~
--
10.88
--
9.68
~
Methyl tert-Butyl Ether
3.12
~
7.79
~
--
~
— = 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 individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-31
-------
Table 24-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
CSNJ
DEMI
ELNJ
GLKY
GPCO
LAKY
//-Octane
5.35
4.69
7.87
16.57
7.08
8.23
Propylene
1.96
1.51
3.24
2.63
2.05
1.48
Styrene
5.78
4.43
15.53
10.59
3.99
8.02
1,1,2,2 -T etrachloroethane
--
--
--
--
--
--
T etrachloroethylene
5.01
4.45
4.42
--
2.68
7.44
Toluene
3.87
3.64
3.51
4.99
3.04
2.71
1,2,4 -Trichlorobenzene
--
--
--
--
--
--
1,1,1 -T richloroethane
--
--
--
--
--
--
1,1,2 -T richloroethane
--
--
--
--
--
--
T richloroethylene
--
--
6.25
--
--
3.45
T richlorofluoro methane
1.71
1.50
3.37
1.56
1.50
1.26
T richlorotrifluoroethane
3.21
1.34
3.50
2.03
1.86
2.16
1,2,4 -T rimethylbenzene
6.70
4.54
6.56
16.77
6.37
7.38
1,3,5 -Trimethylbenzene
12.31
4.31
12.04
14.89
11.89
7.00
Vinyl chloride
3.70
--
--
--
--
1.10
m,p-Xylene
6.63
3.90
5.26
10.12
4.55
4.97
o-Xylene
7.05
4.24
5.56
12.66
5.55
6.26
Average CV by Site
4.99
3.51
6.06
6.93
4.39
3.99
# of Pairs Collected by Site
10
12
10
8
10
7
— = 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 individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-32
-------
Table 24-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
LEKY
NBIL
NBNJ
OCOK
PXSS
ROIL
Acetylene
5.65
2.81
3.95
2.82
2.45
2.74
fcrt-Amyl Methyl Ether
--
--
--
~
--
~
Benzene
3.60
3.38
6.03
4.64
2.74
2.22
Bromochloro methane
--
~
~
--
~
--
Bromodichloromethane
--
5.06
~
~
8.16
~
Bromoform
~
7.51
7.14
~
--
~
Bromomethane
4.92
7.15
3.45
5.53
3.70
3.72
1.3 -Butadiene
12.35
11.10
11.65
10.92
7.59
6.88
Carbon Tetrachloride
2.90
2.90
4.59
3.52
2.31
3.54
Chlorobenzene
--
--
4.56
--
~
--
Chloroethane
10.28
0.00
5.27
7.63
--
12.51
Chloroform
5.65
2.71
4.24
5.88
3.29
6.13
Chloromethane
1.35
3.01
3.34
2.24
1.48
2.30
Chloroprene
--
--
--
--
~
--
Dibromochloro methane
~
5.32
--
~
8.84
~
1,2-Dibromoethane
~
--
~
~
--
~
«/-Dichlorobenzene
~
~
--
~
~
~
o-Dichlorobenzene
--
~
~
~
~
--
p-Dichlorobenzene
~
~
~
--
8.24
3.63
Dichlorodifluoromethane
1.25
3.04
3.40
2.08
1.67
2.28
1,1 -Dichloroethane
~
--
~
~
--
~
1,2-Dichloroethane
4.29
7.90
7.74
5.76
5.34
6.66
1,1 -Dichloroethene
--
~
--
~
~
--
cis-1,2 -Dichloroethy lene
~
~
~
--
~
~
trans-1,2-Dichloroethylene
~
--
~
~
~
~
Dichloromethane
1.76
3.34
3.21
2.99
1.85
3.41
1,2-Dichloropropane
~
~
--
~
--
~
cis-1.3 -Dichloropropene
--
~
~
~
~
--
trans-1,3 -Dichloropropene
~
~
~
--
~
~
Dichlorotetrafluoroethane
5.98
6.31
3.70
3.59
4.30
5.06
Ethyl Acrylate
~
--
~
~
--
~
Ethyl tort-Butyl Ether
~
5.05
8.73
~
~
~
Ethylbenzene
12.07
3.72
14.01
6.45
4.51
5.60
Hexachloro-1,3 -butadiene
--
~
--
--
~
--
Methyl Isobutyl Ketone
4.22
3.48
5.13
5.63
7.45
14.44
Methyl Methacrylate
7.71
--
~
~
7.92
~
Methyl tert-Butyl Ether
3.07
~
--
~
--
~
— = 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 individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-33
-------
Table 24-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
LEKY
NBIL
NBNJ
OCOK
PXSS
ROIL
//-Octane
15.62
7.43
12.67
10.47
6.01
3.55
Propylene
2.26
4.17
3.54
2.59
1.76
2.45
Styrene
5.54
4.64
11.71
12.15
11.30
12.71
1,1,2,2 -T etrachloroethane
--
--
--
--
--
--
T etrachloroethylene
5.75
4.78
8.59
4.33
4.37
5.85
Toluene
2.20
3.82
5.46
3.44
2.91
2.57
1,2,4 -Trichlorobenzene
--
--
--
--
--
--
1,1,1 -T richloroethane
--
--
--
--
--
--
1,1,2 -T richloroethane
--
--
--
--
--
--
T richloroethylene
--
--
--
--
--
2.89
T richlorofluoro methane
1.37
3.03
3.16
1.98
1.26
2.04
T richlorotrifluoroethane
2.58
3.11
3.03
3.53
1.77
3.16
1,2,4 -T rimethylbenzene
11.76
4.49
15.36
7.04
6.58
5.20
1,3,5 -Trimethylbenzene
6.75
3.93
14.31
9.67
8.78
8.14
Vinyl chloride
--
--
--
--
--
--
m,p-Xylene
10.15
3.43
11.56
5.48
4.40
3.17
o-Xylene
11.58
4.10
13.13
5.88
5.41
3.58
Average CV by Site
6.02
4.51
7.24
5.45
4.87
5.05
# of Pairs Collected by Site
8
7
12
12
8
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; and brown shading identifies sites for which replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-34
-------
Table 24-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
S4MO
SEWA
SPAZ
SPIL
TMOK
TOOK
TROK
Acetylene
1.53
3.29
1.67
5.35
3.70
7.96
3.84
tert-Amyl Methyl Ether
--
--
~
--
~
--
--
Benzene
2.63
4.96
3.09
4.64
4.67
5.94
3.82
Bromochloromethane
~
~
--
~
--
~
~
Bromodichloro methane
2.13
~
~
~
~
--
~
Bromofonn
--
~
~
--
~
~
~
Bromo methane
10.32
--
0.00
12.81
6.97
6.67
3.80
1.3 -Butadiene
10.61
4.60
2.43
4.94
8.19
9.94
6.67
Carbon Tetrachloride
8.95
3.55
2.08
3.95
4.71
7.69
4.44
Chlorobenzene
--
~
--
~
--
--
--
Chloroethane
7.37
8.32
3.75
4.93
4.56
~
2.48
Chloroform
3.42
4.23
3.31
7.13
6.21
6.80
6.69
Chloromethane
1.69
2.56
1.13
3.10
3.11
8.10
3.94
Chloroprene
~
--
--
--
--
~
--
Dibromochloromethane
--
~
~
~
~
--
~
1,2-Dibromoethane
~
~
~
--
~
~
~
«7-Dichlorobenzene
~
--
~
~
~
~
--
o-Dichlorobenzene
~
~
~
~
--
~
~
p-Dichlorobenzene
6.21
~
6.22
~
13.93
6.55
6.03
Dichlorodifluoro methane
1.29
2.55
1.12
2.62
3.05
7.93
3.70
1,1 -Dichloroethane
--
~
--
--
~
--
~
1,2-Dichloroethane
9.54
5.11
3.38
11.04
11.05
9.01
3.47
1,1 -Dichloroethene
~
--
~
~
--
~
--
cis-1,2-Dichloroethylene
~
~
--
~
~
~
~
trans-1,2-Dichloroethylene
--
~
2.57
~
18.66
--
2.22
Dichloromethane
2.81
2.45
2.48
3.09
3.39
4.66
4.09
1,2-Dichloropropane
~
--
~
--
~
~
--
cis-1,3 -Dichloropropene
~
~
~
~
--
~
~
trans-1,3 -Dichloropropene
~
~
--
~
~
~
~
Dichlorotetrafluoroethane
7.72
2.61
2.89
7.84
7.01
6.14
3.53
Ethyl Acrylate
--
~
~
--
~
--
~
Ethyl tort-Butyl Ether
12.71
--
~
3.14
~
~
4.29
Ethylbenzene
9.57
4.22
3.88
7.85
6.51
3.64
4.17
Hexachloro -1,3 -butadiene
~
~
--
~
--
~
--
Methyl Isobutyl Ketone
6.75
6.96
5.58
5.97
5.29
6.41
7.07
Methyl Methacrylate
--
~
~
--
12.86
--
~
Methyl tert-Butyl Ether
~
--
~
34.69
~
~
--
— = 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 individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-35
-------
Table 24-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
S4MO
SEWA
SPAZ
SPIL
TMOK
TOOK
TROK
//-Octane
10.87
6.22
3.84
7.69
6.20
6.59
6.68
Propylene
3.48
2.85
1.59
2.87
2.58
5.74
2.82
Styrene
5.95
9.00
4.66
8.68
17.26
6.33
5.89
1.1,2,2-Tetrachloroethane
--
--
--
--
--
--
--
T etrachloroethylene
11.33
4.97
4.57
5.13
6.48
5.97
6.35
Toluene
4.44
4.05
2.91
4.38
4.59
3.00
3.55
1,2,4-Trichlorobenzene
--
--
--
--
--
--
--
1,1,1 -T richloroethane
--
--
--
--
--
--
~
1.1,2-Trichloroethane
--
--
--
--
--
--
~
T richloroethylene
--
--
--
7.85
--
--
--
T richlorofluoro methane
1.12
2.51
1.08
2.80
2.92
8.36
3.81
T richlorotrifluoroethane
2.57
2.28
0.89
2.49
3.44
7.13
2.43
1,2,4-Trimethylbenzene
8.59
5.31
5.17
8.35
7.79
4.04
6.59
1,3,5 -T rimethy lbenzene
6.63
8.84
5.41
15.48
13.91
3.99
6.00
Vinyl chloride
--
--
--
--
--
--
--
m,p-Xylene
7.58
4.78
4.37
6.67
6.08
3.53
4.18
o-Xylene
9.52
4.71
4.18
7.57
6.03
3.21
4.35
Average CV by Site
6.33
4.62
3.12
7.25
7.18
6.21
4.53
# of Pairs Collected by Site
13
14
10
12
10
10
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; and brown shading identifies sites for which replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-36
-------
Table 24-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
TVKY
YUOK
# of
Pairs
Average
by
Pollutant
Average
for
Duplicate
Pairs
Average
for
Collocated
Pairs
Average
for
Unequipped
Acetylene
3.15
5.80
311
3.38
3.59
2.80
3.01
fcrt-Amyl Methyl Ether
~
--
0
--
--
--
--
Benzene
4.65
5.57
311
4.30
4.44
3.69
4.12
Bromochloro methane
--
~
0
--
--
--
--
Bromodichloromethane
~
--
11
5.12
3.60
8.16
~
Bromoform
10.10
~
5
8.25
7.32
10.10
~
Bromomethane
5.94
0.00
149
5.31
5.16
4.82
5.81
1.3 -Butadiene
6.36
11.79
275
8.20
8.59
6.97
7.54
Carbon Tetrachloride
3.30
5.17
311
3.65
4.20
2.80
2.48
Chlorobenzene
--
--
8
5.93
5.93
--
--
Chloroethane
5.90
3.02
99
6.33
6.09
5.90
7.11
Chloroform
5.95
9.26
289
5.58
5.70
4.62
5.56
Chloromethane
2.68
4.49
311
2.56
2.95
2.08
1.70
Chloroprene
--
~
0
--
--
--
--
Dibromochloro methane
~
--
7
7.08
5.32
8.84
~
1,2-Dibromoethane
~
~
0
--
--
~
~
«/-Dichlorobenzene
~
~
0
--
--
~
~
o-Dichlorobenzene
--
~
0
--
--
--
--
/?-Dichlorobcnzcnc
~
0.00
46
7.16
7.14
8.24
6.22
Dichlorodifluoromethane
2.89
4.22
311
2.53
2.92
2.28
1.61
1,1 -Dichloroethane
6.30
--
11
4.54
--
6.30
2.77
1,2-Dichloroethane
6.37
8.63
282
6.69
7.00
5.86
6.15
1,1 -Dichloroethene
11.40
~
3
5.70
--
11.40
0.00
6V.S-1,2-Dichlorocthy lcnc
16.02
~
2
16.02
--
16.02
--
trans-1,2-Dichloroethylene
6.63
9.70
13
7.96
10.20
6.63
2.57
Dichloromethane
5.15
6.59
305
3.63
4.04
3.50
2.62
1,2-Dichloropropane
--
--
0
--
--
--
~
cis-1,3 -Dichloropropene
--
~
0
--
--
--
--
trans-1,3 -Dicliloropropene
~
~
0
--
--
~
~
Diclilorotetrafluoroethane
7.51
7.61
310
4.90
4.98
5.91
4.41
Ethyl Acrylate
~
--
0
--
~
~
~
Ethyl tert-Butyl Ether
~
~
45
6.12
6.12
~
~
Ethylbenzene
11.07
11.09
291
7.67
7.02
7.79
9.29
Hexachloro-1,3 -butadiene
--
~
0
--
--
--
--
Methyl Isobutyl Ketone
8.09
4.28
223
6.78
6.59
7.77
6.99
Methyl Metliacrylate
~
--
9
9.30
11.14
7.92
7.22
Methyl tert-Butyl Ether
~
7.19
31
8.46
10.26
~
3.04
— = 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 individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-37
-------
Table 24-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
TVKY
YUOK
# of
Pairs
Average
by
Pollutant
Average
for
Duplicate
Pairs
Average
for
Collocated
Pairs
Average
for
Unequipped
//-Octane
12.66
7.99
285
8.68
7.92
9.33
10.44
Propylene
3.58
5.42
311
2.88
3.20
2.67
2.13
Styrene
10.10
9.79
204
9.86
9.69
10.70
10.06
1,1,2,2 -T etrachloroethane
--
--
0
--
--
--
--
T etrachloroethylene
5.51
5.69
167
5.79
5.65
4.94
6.64
Toluene
4.69
4.80
311
4.19
4.08
3.80
4.59
1,2,4 -Trichlorobenzene
--
--
0
--
--
--
--
1,1,1 -T richloroethane
--
--
3
1.98
3.97
--
0.00
1,1,2 -T richloroethane
9.75
--
8
8.88
--
9.75
8.00
T richloroethylene
5.19
--
20
4.20
5.19
5.19
1.72
T richlorofluoro methane
2.83
4.37
311
2.49
2.92
2.04
1.50
T richlorotrifluoroethane
3.27
5.78
310
2.89
3.22
2.52
2.16
1,2,4 -T rimethylbenzene
10.18
9.87
242
8.58
8.21
8.38
9.61
1,3,5 -Trimethylbenzene
11.82
6.02
146
9.59
9.50
10.30
9.62
Vinyl chloride
5.59
--
46
2.12
3.70
5.59
0.85
m,p-Xylene
10.54
9.53
296
6.78
6.02
7.47
8.56
o-Xylene
11.55
9.89
294
7.47
6.71
8.48
9.13
Average CV by Site
7.17
6.55
6,923
6.09
5.95
6.53
5.01
# of Pairs Collected by Site
50
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; and brown shading identifies sites for which replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-38
-------
24.3.2 SNMOC Analytical Precision
Table 24-10 presents analytical precision results from replicate analyses of duplicate 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. The CV ranges from 0 percent
(p-ethyltoluene for BMCO and 2,3,4-trimethylpentane for RFCO) to 15.44 percent (a-pinene for
PACO). a-Pinene is the only pollutant with a CV greater than or equal to 15 percent.
The pollutant-specific average CV, as shown in orange in Table 24-10, ranges from
0.71 percent (ethane) to 9.67 percent (p-diethylbenzene). None of the SNMOCs shown in
Table 24-10 have an average CV greater than or equal to 15 percent. The site-specific average
CV, as shown in green in Table 24-10, ranges from 3.14 percent (BRCO) to 3.90 percent
(RICO); analytical precision for all seven sites sampling SNMOCs falls between 3 percent and
4 percent. The overall average analytical precision is 3.78 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 24-10
while sites for which replicates were run on individual samples are highlighted in brown.
Collocated SNMOC samples were not collected at the NMP sites sampling SNMOC. Duplicate
SNMOC samples were collected at only BTUT and NBIL; replicates were run on individual
SNMOC samples collected at the five Garfield County, Colorado sites. The average CV for sites
that collected duplicate samples was calculated and is shown at the end of Table 24-10 in blue
while the average CV for sites for which replicates were run on individual samples is shown in
brown. The variability ranges from 3.63 percent (replicates run on individual samples) to
4.08 percent (replicates run on duplicate samples).
24-39
-------
Table 24-10. SNMOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant
Pollutant
BMCO
BRCO
BTUT
NBIL
PACO
Acetylene
3.87
3.26
1.89
1.46
4.55
Benzene
3.60
3.97
2.43
4.43
2.86
1.3 -Butadiene
--
--
5.25
--
--
//-Butane
0.69
0.38
0.96
1.74
0.71
1-Butene
--
--
--
--
~
67.Y-2-Bute nc
~
~
5.50
~
9.24
;ra«.v-2-Butcne
~
~
4.46
~
8.08
Cyclohexane
1.20
2.19
4.31
6.70
1.32
Cyclopentane
3.08
2.83
2.47
13.23
1.54
Cyclopentene
--
--
--
--
--
w-Decane
6.19
8.85
7.40
5.18
5.85
1-Decene
~
~
--
~
--
/w-Diethylbenzene
--
~
~
~
~
p-Diethylbenzene
~
6.33
13.01
--
~
2,2-Dimethylbutane
4.05
1.76
3.93
6.36
2.82
2,3 -Dimethylbutane
2.00
2.50
1.15
5.30
1.71
2,3 -Dimethylpentane
5.45
3.81
2.84
4.86
3.08
2,4-Dimethylpentane
8.64
6.51
3.89
4.85
5.66
n-Dodecane
--
--
7.76
--
--
1-Dodecene
~
--
--
~
~
Ethane
0.65
0.50
0.71
0.59
0.70
2-Ethyl-l-butene
~
~
--
~
--
Ethylbenzene
3.34
~
7.48
5.22
10.72
Ethylene
1.68
0.69
1.42
2.34
1.11
/w-Ethyltoluene
6.09
5.72
3.08
7.26
2.96
o-Ethyltoluene
--
--
4.39
--
--
/?-Ethyltolucnc
0.00
4.03
6.57
3.39
5.87
//-Heptane
2.52
3.37
2.65
7.59
2.11
1-Heptene
--
~
--
--
~
n-Hexane
1.68
2.01
2.66
2.33
2.00
1-Hexene
~
--
~
~
--
c/.v-2-Hc.\cnc
~
~
--
~
~
-------
Table 24-10. SNMOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
BMCO
BRCO
BTUT
NBIL
PACO
Isopropylbenzene
--
--
1.77
—
--
2 -Methyl -1 -butene
12.43
0.32
5.58
—
7.28
3 -Methyl -1 -butene
--
~
~
—
~
2 -Methyl -1 -pentene
--
~
--
—
~
4 -Methyl -1 -pentene
--
~
~
—
--
2 -Methyl -2 -butene
--
--
5.33
4.82
0.28
Methylcyclohexane
1.14
2.80
2.35
5.24
0.73
Methylcyclopentane
1.46
1.96
3.21
2.66
1.95
2-Methylheptane
2.96
8.11
5.79
--
5.60
3-Methylheptane
2.34
1.33
4.85
--
3.28
2-Methylhexane
1.69
2.74
1.65
3.69
2.06
3-Methylhexane
2.58
5.09
2.29
--
1.28
2-Methylpentane
0.90
1.63
2.34
1.73
0.70
3-Methylpentane
2.30
2.12
2.96
2.63
1.24
//-Nona ne
1.84
3.55
2.02
4.02
2.64
1-Nonene
--
2.54
5.84
--
3.08
//-Octane
2.18
2.24
3.17
5.54
1.56
1-Octene
5.35
9.44
4.49
--
3.55
n-Pentane
1.02
0.54
1.41
1.09
1.30
1-Pentene
2.86
1.86
3.08
1.36
3.12
cis-2 -Pentene
--
--
5.80
~
--
trans-2-Pentene
--
--
3.63
--
2.44
fl-Pincnc
4.06
5.91
6.10
6.04
15.44
/>-Pincnc
--
--
--
~
--
Propane
0.58
0.47
0.97
0.89
0.45
n-Propylbenzene
--
3.38
4.48
3.65
6.94
Propylene
7.47
4.26
2.66
2.51
3.93
Propyne
--
--
--
--
--
Styrene
6.67
2.70
--
~
5.45
Toluene
3.11
2.98
2.29
1.68
2.75
/?-T ridccanc
~
--
--
~
--
1-Tridecene
~
~
--
--
--
1,2,3 -Trimethylbenzene
0.54
~
3.94
~
7.36
1,2,4 -T rimethylbenzene
6.51
7.34
6.09
3.18
4.96
1,3,5 -Trimethylbenzene
7.44
~
9.35
~
5.43
2,2,3 -Trimethylpentane
--
--
4.38
~
2.48
— = 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 and brown shading identifies sites for which
replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-41
-------
Table 24-10. SNMOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
BMCO
BRCO
BTUT
NBIL
PACO
2,2,4 -T rimethy lpentane
--
--
3.48
5.69
--
2,3,4 -T rimethy lpentane
--
0.69
5.03
5.72
0.90
/7-Undecane
--
--
4.38
1.57
5.24
1-Undecene
--
--
--
--
--
«/-Xylene//?-Xylene
3.78
4.30
2.66
4.06
2.50
o-Xylene
4.25
5.08
4.13
6.37
3.14
SNMOC (Sum of Knowns)
0.56
0.46
2.36
0.69
1.17
Sum of Unknowns
2.51
1.77
2.77
3.25
0.78
Average CV by Site
3.17
3.14
3.82
3.86
3.48
# of Pairs Collected by Site
8
7
12
6
7
— = 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 and brown shading identifies sites for which
replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-42
-------
Table 24-10. SNMOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
RICO
RICO
# of
pairs
Average
by
Pollutant
Average
for
Duplicate
Pairs
Average
for
Unequipped
Acetylene
2.52
2.35
54
2.84
1.67
3.31
Benzene
2.85
1.81
54
3.14
3.43
3.02
1.3 -Butadiene
--
11.31
3
8.28
5.25
11.31
//-Butane
1.14
1.72
54
1.05
1.35
0.93
1-Butene
--
--
0
--
--
--
67.Y-2-Butcnc
4.07
6.51
20
6.33
5.50
6.61
;ra«.v-2-B lite nc
1.61
2.98
20
4.29
4.46
4.23
Cyclohexane
5.11
1.04
48
3.12
5.51
2.17
Cyclopentane
4.66
4.19
39
4.57
7.85
3.26
Cyclopentene
--
--
--
--
--
--
w-Decane
3.31
4.36
31
5.88
6.29
5.71
1-Decene
--
~
0
--
--
--
/w-Diethylbenzene
~
--
0
--
--
--
/j-Dicthvlbcnzcne
~
~
2
9.67
13.01
6.33
2,2 -Dimethy lbutane
2.20
4.99
34
3.73
5.14
3.17
2,3-Dimethylbutane
4.79
2.03
48
2.78
3.23
2.61
2,3 -Dimethy lpentane
1.67
2.61
49
3.47
3.85
3.32
2,4 -Dimethy lpentane
--
6.06
40
5.94
4.37
6.72
n-Dodecane
2.07
--
3
4.92
7.76
2.07
1-Dodecene
~
~
0
--
--
--
Ethane
0.49
1.35
54
0.71
0.65
0.74
2-Ethyl-l-butene
--
~
0
--
--
--
Ethylbenzene
5.67
8.49
31
6.82
6.35
7.05
Ethylene
1.22
2.63
54
1.58
1.88
1.47
/w-Ethyltoluene
2.97
4.39
39
4.64
5.17
4.43
o-Ethyltoluene
3.86
8.34
4
5.53
4.39
6.10
/?-Ethyltolucnc
3.25
4.58
28
3.96
4.98
3.55
//-Heptane
2.89
2.31
50
3.35
5.12
2.64
1-Heptene
--
--
0
--
--
--
n-Hexane
2.60
1.79
54
2.15
2.49
2.01
1-Hexene
--
--
0
--
--
--
c/.v-2-Hc\cnc
~
~
0
--
--
--
tr <7/75-2 -Hexene
~
--
0
--
--
~
Isobutane
0.84
1.77
54
1.15
1.91
0.84
Isobutylene
~
~
0
--
--
~
Isopentane
--
2.55
9
1.05
0.68
1.29
Isoprene
2.46
1.82
32
2.24
2.20
2.25
— = 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 and brown shading identifies sites for which replicates
were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-43
-------
Table 24-10. SNMOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
RICO
RICO
# of
pairs
Average
by
Pollutant
Average
for
Duplicate
Pairs
Average for
Unequipped
Isopropylbenzene
--
~
1
1.77
1.77
--
2-Methyl-1 -butene
7.42
4.17
23
6.20
5.58
6.33
3 -Methyl-1 -butene
~
~
0
--
--
--
2-Methyl-1 -pentene
~
--
0
--
--
--
4-Methyl-1 -pentene
~
~
0
--
--
--
2-Methyl-2-butene
4.08
5.50
19
4.00
5.07
3.28
Methylcyclohexane
2.93
2.00
50
2.46
3.80
1.92
Methylcyclopentane
5.31
1.52
54
2.58
2.93
2.44
2-Methylheptane
--
9.06
32
6.30
5.79
6.43
3-Methylheptane
~
2.76
30
2.91
4.85
2.43
2-Methylhexane
2.34
1.67
54
2.26
2.67
2.10
3-Methylhexane
3.58
4.74
24
3.26
2.29
3.46
2-Methylpentane
2.70
1.16
54
1.59
2.04
1.42
3-Methylpentane
5.23
2.68
54
2.74
2.80
2.71
w-Nonane
4.27
5.28
37
3.38
3.02
3.52
1-Nonene
--
3.78
4
3.81
5.84
3.14
//-Octane
2.57
2.07
45
2.76
4.36
2.12
1-Octene
--
2.26
15
5.02
4.49
5.15
/7-Pentane
0.74
1.18
54
1.04
1.25
0.96
1-Pentene
4.65
3.76
20
2.96
2.22
3.25
c7.v-2-Pcntcnc
--
--
5
5.80
5.80
--
;ra«.v-2-Pcntene
6.06
4.16
19
4.07
3.63
4.22
fl-Pincnc
2.06
--
28
6.60
6.07
6.87
/>-Pincnc
~
--
0
--
--
--
Propane
1.02
1.86
54
0.89
0.93
0.88
/7-Propylbenzene
3.43
6.01
15
4.65
4.06
4.94
Propylene
3.76
2.43
54
3.86
2.58
4.37
Propyne
--
--
0
--
--
--
Styrene
1.40
6.83
14
4.61
4.61
Toluene
1.09
2.09
54
2.28
1.98
2.40
«-T ridccanc
--
--
0
--
--
--
1-Tridecene
~
~
0
--
--
--
1,2,3 -T rimethy lbenzene
3.61
2.92
12
3.67
3.94
3.61
1,2,4-Trimethy lbenzene
4.87
4.49
42
5.35
4.64
5.63
1,3,5 -T rimethy lbenzene
5.17
8.93
15
7.26
9.35
6.74
2,2,3-Trimethylpentane
--
--
4
3.43
4.38
2.48
— = 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 and brown shading identifies sites for which replicates
were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-44
-------
Table 24-10. SNMOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
RICO
RICO
# of
pairs
Average
by
Pollutant
Average
for
Duplicate
Pairs
Average for
Unequipped
2,2,4-Trimethylpentane
3.64
4.57
21
4.35
4.59
4.11
2,3,4 -T rimethy lpentane
0.00
--
20
2.47
5.37
0.53
/7-Undecane
--
--
4
3.73
2.98
5.24
1-Undecene
--
--
0
--
--
--
«/-Xylene//?-Xylene
5.20
2.51
50
3.57
3.36
3.66
o-Xylene
7.52
2.83
50
4.76
5.25
4.56
SNMOC (Sum of Knowns)
0.99
8.96
54
2.17
1.52
2.43
Sum of Unknowns
2.33
6.52
54
2.85
3.01
2.78
Average CV by Site
3.20
3.90
2,047
3.78
4.08
3.63
# of Pairs Collected by Site
7
7
— = 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 and brown shading identifies sites for which replicates
were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-45
-------
24.3.3 Carbonyl Compound Analytical Precision
Table 24-11 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. The overall average CV was 2.59 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 14.37 percent (tolualdehydes for PACO),
indicating that every pollutant-site combination has a CV less than or equal to 15 percent.
The pollutant-specific average CV, as shown in orange in Table 24-11, ranges from
0.57 percent (acetone) to 4.60 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 24-11, ranges from 1.68 percent (NBNJ) to 4.08 percent (PACO), indicating that all of the
site-specific average CVs are also less than 5 percent.
Sites at which duplicate samples were collected are highlighted in blue in Table 24-11,
sites at which collocated samples were collected are highlighted in purple, and sites for which
replicates were run on individual samples are highlighted in brown. Collocated carbonyl
compound samples were collected at three of the sites shown in Table 24-11 (DEMI, INDEM,
and PXSS); replicates were run on individual samples for seven sites, and the remainder of sites
collected duplicate samples. The average CV for sites that collected duplicate samples was
calculated and is shown at the end of Table 24-11 in blue, the average CV for sites collecting
collocated samples is shown in purple, and the average CV for sites for which replicates were
run on individual samples is shown in brown. The average CV for all three precision types are
less than 3 percent, meeting the MQO of 15 percent, with the variability ranging from
2.49 percent (replicates run on duplicate samples) to 2.94 percent (replicates run on individual
samples).
24-46
-------
Table 24-11. Carbonyl Compound Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant
Pollutant
ASKY
AZFL
BMCO
BRCO
BTUT
CHNJ
CSNJ
DEMI
Acetaldehyde
0.62
1.44
0.44
0.38
0.38
0.62
0.93
0.46
Acetone
0.40
0.72
0.33
0.30
0.30
0.70
0.59
0.53
Benzaldehyde
4.69
2.97
2.44
9.94
3.37
5.22
3.33
3.70
2-Butanone
0.79
3.28
0.66
2.17
0.73
2.07
2.63
1.49
Butyraldehyde
2.37
2.60
1.45
5.54
1.12
2.62
1.59
1.79
Crotonaldehyde
1.82
3.34
2.38
1.15
2.90
2.40
1.90
2.55
2,5 -Dimethylbenzaldehyde
--
--
--
--
--
--
--
--
Formaldehyde
0.42
2.17
0.68
3.29
0.63
1.48
0.74
0.87
Hexaldehyde
7.70
4.75
2.70
0.00
4.01
5.42
2.43
4.11
Isovaleraldehyde
--
--
--
--
--
--
--
--
Propionaldehyde
1.61
2.03
0.00
2.08
1.29
1.95
1.87
1.39
Tolualdehydes
5.53
5.28
5.12
7.71
4.10
4.78
3.97
4.07
Valeraldehyde
4.33
4.34
3.70
3.55
3.01
4.59
3.10
3.34
Average CV by Site
2.75
2.99
1.81
3.28
1.99
2.90
2.10
2.21
# of Pairs Collected by Site
6
10
2
3
12
12
12
17
— = 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; brown shading identifies sites for which replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-47
-------
Table 24-11. Carbonyl Compound Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
ELNJ
GLKY
GPCO
INDEM
LEKY
NBIL
NBNJ
OCOK
Acetaldehyde
0.51
0.97
0.62
2.02
0.38
0.25
0.57
1.11
Acetone
0.57
0.40
0.38
0.90
0.53
0.50
0.00
1.58
Benzaldehyde
3.80
2.88
2.91
4.10
5.06
2.37
5.10
5.15
2-Butanone
2.94
2.78
1.72
2.54
1.96
1.89
0.71
1.22
Butyraldehyde
2.56
2.10
2.20
3.08
1.56
2.48
0.55
1.97
Crotonaldehyde
1.70
1.45
3.00
3.10
3.07
4.89
1.65
0.91
2,5 -Dimethylbenzaldehyde
--
--
--
--
--
--
--
--
Formaldehyde
0.41
0.85
0.69
1.45
0.48
0.49
0.43
1.36
Hexaldehyde
3.38
8.61
3.17
4.88
2.18
5.82
2.40
3.62
Isovaleraldehyde
--
--
--
--
--
--
--
--
Propionaldehyde
0.96
1.05
1.79
2.56
1.90
2.55
0.44
2.10
Tolualdehydes
5.07
3.47
4.67
5.32
2.91
3.27
5.70
4.84
Valeraldehyde
2.42
6.96
4.36
4.70
4.63
4.72
0.94
4.65
Average CV by Site
2.21
2.87
2.32
3.15
2.24
2.66
1.68
2.59
# of Pairs Collected by Site
12
10
12
22
6
7
4
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; brown shading identifies sites for which replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-48
-------
Table 24-11. Carbonyl Compound Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
OR II
PACO
PXSS
RICO
RICO
ROIL
S4MO
SEWA
Acetaldehyde
0.82
3.03
0.61
1.69
0.85
0.62
0.54
1.88
Acetone
0.95
0.22
0.35
0.36
0.73
0.52
0.40
0.60
Benzaldehyde
2.87
7.09
3.77
2.72
7.47
3.50
5.53
3.70
2-Butanone
2.48
1.99
2.01
1.74
1.22
2.39
0.89
2.27
Butyraldehyde
2.97
4.98
1.99
9.98
4.59
3.00
2.86
3.92
Crotonaldehyde
1.46
2.67
1.86
2.60
1.40
2.80
2.97
4.79
2,5 -Dimethylbenzaldehyde
--
--
--
--
--
--
--
--
Formaldehyde
0.96
2.05
0.53
1.07
3.63
0.76
0.97
1.87
Hexaldehyde
3.36
1.90
3.70
5.80
4.20
3.11
2.95
4.58
Isovaleraldehyde
--
--
--
--
--
--
--
--
Propionaldehyde
1.68
1.48
1.93
3.97
2.58
1.74
1.61
2.86
Tolualdehydes
4.61
14.37
4.60
4.75
0.00
4.55
4.64
2.66
Valeraldehyde
3.78
5.04
2.78
3.23
5.70
4.47
4.22
3.79
Average CV by Site
2.36
4.08
2.19
3.45
2.94
2.50
2.51
2.99
# of Pairs Collected by Site
12
3
12
3
3
12
12
15
— = 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; brown shading identifies sites for which replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-49
-------
Table 24-11. Carbonyl Compound Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
SKFL
SPIL
SYFL
TMOK
TOOK
TROK
WPIN
Acetaldehyde
1.46
0.73
0.81
0.41
1.04
0.56
1.47
Acetone
0.83
0.79
0.61
0.84
0.41
0.29
0.78
Benzaldehyde
4.15
2.96
5.29
3.65
5.30
2.18
4.14
2-Butanone
2.22
2.30
2.62
1.01
0.63
1.78
2.97
Butyraldehyde
1.88
2.12
3.23
1.59
3.10
3.77
4.19
Crotonaldehyde
1.86
2.24
1.42
2.30
2.01
1.28
3.79
2,5 -Dimethylbenzaldehyde
--
--
--
--
--
--
--
Formaldehyde
1.90
1.20
1.05
1.12
1.85
1.04
1.56
Hexaldehyde
4.43
4.49
4.40
3.51
4.33
2.54
3.91
Isovaleraldehyde
--
--
--
--
--
--
--
Propionaldehyde
2.99
1.80
2.82
1.55
1.57
2.04
3.01
Tolualdehydes
3.87
2.57
4.67
3.79
4.33
4.19
3.81
Valeraldehyde
3.75
3.36
3.97
3.82
4.09
4.47
4.41
Average CV by Site
2.67
2.23
2.81
2.15
2.61
2.20
3.09
# of Pairs Collected by Site
11
10
12
12
12
12
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; brown shading identifies sites for which replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-50
-------
Table 24-11. Carbonyl Compound Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
YUOK
# of
pairs
Average
by
Pollutant
Average
for
Duplicate
Pairs
Average
for
Collocated
Pairs
Average
for
Unequipped
Acetaldehyde
0.50
333
0.90
0.83
1.03
1.06
Acetone
0.82
333
0.57
0.62
0.60
0.41
Benzaldehyde
4.85
321
4.26
3.87
3.85
5.63
2-Butanone
2.43
313
1.89
2.00
2.01
1.50
Butyraldehyde
1.58
326
2.85
2.45
2.29
4.35
Crotonaldehyde
2.57
332
2.38
2.44
2.51
2.16
2,5 -Dimethylbenzaldehyde
--
--
--
--
--
--
Formaldehyde
0.71
333
1.21
1.10
0.95
1.66
Hexaldehyde
4.33
331
3.96
4.07
4.23
3.50
Isovaleraldehyde
--
--
--
--
--
--
Propionaldehyde
1.53
331
1.90
1.87
1.96
1.95
Tolualdehydes
4.07
277
4.60
4.22
4.66
5.77
Valeraldehyde
3.89
320
4.00
3.96
3.61
4.31
Average CV by Site
2.48
3,550
2.59
2.49
2.52
2.94
# of Pairs Collected by Site
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; brown shading identifies sites for which replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-51
-------
24.3.4 PAH Analytical Precision
Table 24-12 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. The CV ranges from
0.37 percent (coronene for BTUT) to 22.41 percent (perylene for BXNY). CVs for only four
pollutant-site combinations are greater than or equal to 15 percent.
The pollutant-specific average CV, as shown in orange in Table 24-12, ranges from
2.19 percent (phenanthrene) to 8.66 percent (perylene). The site-specific average CV, as shown
in green in Table 24-12, ranges from 3.25 percent (BTUT) to 8.61 percent (RIVA). The overall
average analytical precision CV is 4.87 percent.
Sites at which collocated PAH samples were collected are highlighted in blue in
Table 24-12 while sites for which replicates were run on individual samples are highlighted in
brown. Collocated PAH samples were collected at DEMI, RUCA, and SEW A; replicates were
run on individual PAH samples for the remaining sites. The average CV for sites that collected
collocated PAH samples was calculated and is shown at the end of Table 24-12 in blue while the
average CV for sites for which replicates were run on individual samples is shown in brown. The
variability ranges from 4.18 percent (replicates run on collocated samples) to 5.02 percent
(replicates run on individual samples).
24-52
-------
Table 24-12. PAH Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant
Pollutant
BOMA
BTUT
BXNY
CELA
DEMI
GLKY
GPCO
NBIL
Acenaphthene
3.13
4.38
2.57
4.39
2.45
1.78
1.61
4.72
Acenaphthylene
7.69
2.96
3.35
0.86
8.05
6.31
2.23
5.62
Anthracene
5.41
5.67
1.91
6.14
4.92
9.23
2.02
6.49
Benzo(a)anthracene
4.16
0.66
3.38
3.67
3.21
4.95
4.60
5.69
Benzo(a)pyrene
4.20
1.90
4.01
4.13
4.08
6.88
1.08
5.97
B enzo (b )fluoranthene
4.94
4.30
2.16
2.45
2.28
3.40
1.55
5.34
Benzo(e)pyrene
4.43
1.07
2.43
3.20
2.29
4.83
3.70
7.07
Benzo(g,hi)perylene
4.22
8.03
1.82
3.73
2.68
5.86
1.26
5.45
B enzo (k)fluoranthene
4.30
4.52
12.70
10.48
5.92
5.15
6.66
9.35
Chrysene
2.57
3.25
1.51
2.54
2.86
3.47
2.53
5.32
Coronene
3.22
0.37
3.70
4.79
5.04
10.39
6.87
8.37
Cyclopenta | cdlpyrene
--
--
4.84
3.54
9.79
--
9.56
--
Dibenz(a,h)anthracene
--
--
--
--
7.61
--
9.23
10.50
Fluoranthene
1.97
2.72
2.90
4.76
2.70
2.26
2.92
5.65
Fluorene
2.20
1.29
2.70
5.16
2.13
3.18
2.86
4.50
9-Fluorenone
2.64
3.94
3.38
5.36
2.00
3.12
4.19
5.83
Indeno( 1,2,3 -cd)pyrene
10.18
1.26
8.56
4.06
5.41
8.39
4.10
9.13
Naphthalene
4.06
4.04
1.50
3.69
3.59
2.56
3.97
5.97
Perylene
--
--
22.41
--
6.79
--
7.78
9.41
Phenanthrene
1.39
0.99
1.43
1.14
1.36
2.29
1.13
4.94
Pyrene
2.15
3.50
2.97
5.10
2.37
2.10
2.78
5.69
Retene
5.47
6.94
14.63
3.69
5.15
5.94
2.80
6.95
Average CV by Site
4.12
3.25
4.99
4.14
4.21
4.85
3.88
6.57
# of Pairs Collected by Site
7
7
7
6
12
11
7
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 collocated samples and brown shading identifies sites for which replicates
were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-53
-------
Table 24-12. PAH Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
PRRI
PXSS
RIVA
ROCH
RUCA
S4MO
SEWA
SJJCA
Acenaphthene
2.71
4.97
12.66
2.56
3.35
3.68
3.01
4.48
Acenaphthylene
3.05
3.97
5.05
3.35
2.63
3.20
2.92
5.03
Anthracene
4.59
3.65
17.09
10.59
4.73
9.38
3.25
3.68
Benzo(a)anthracene
5.20
2.31
2.21
2.00
2.97
2.84
2.80
2.43
Benzo(a)pyrene
4.69
6.70
2.41
3.50
0.46
7.50
6.41
3.54
B enzo (b )fluoranthene
1.86
5.95
10.51
3.21
5.78
3.85
2.51
8.29
Benzo(e)pyrene
3.06
6.61
1.53
3.58
3.56
2.75
5.29
3.45
Benzo(g,hi)perylene
3.44
4.43
2.62
5.33
6.14
3.67
3.36
4.25
B enzo (k)fluoranthene
8.17
4.30
6.43
8.47
8.29
7.10
5.82
6.78
Chrysene
1.36
3.12
12.52
1.40
3.88
1.80
3.81
2.44
Coronene
2.38
3.87
4.34
3.37
6.03
3.90
3.76
4.29
Cyclopenta | cdlpyrene
10.40
--
--
--
--
8.14
--
2.48
Dibenz(a,h)anthracene
1.61
--
5.41
--
0.70
6.31
--
10.02
Fluoranthene
2.23
2.39
11.17
2.80
2.10
4.27
3.34
6.08
Fluorene
2.07
1.65
13.36
2.97
2.66
3.36
2.44
3.61
9-Fluorenone
2.30
2.70
12.02
2.86
3.23
3.25
3.66
2.50
Indeno( 1,2,3 -cd)pyrene
3.76
4.68
6.36
8.33
8.81
5.46
5.63
2.85
Naphthalene
3.93
3.49
14.69
4.88
4.13
6.74
2.17
5.16
Perylene
4.23
13.06
3.00
--
6.94
5.59
--
7.44
Phenanthrene
0.89
2.19
12.59
0.69
1.55
2.21
1.09
1.49
Pyrene
2.09
3.04
11.63
3.49
2.66
3.90
3.45
6.95
Retene
5.09
6.83
13.13
6.10
3.41
7.18
3.84
6.42
Average CV by Site
3.60
4.50
8.61
4.18
4.00
4.82
3.61
4.71
# of Pairs Collected by Site
9
8
7
9
14
9
14
8
— = 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 replicates
were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-54
-------
Table 24-12. PAH Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
SKFL
UNVT
WADC
# of
Pairs
Average
by
Pollutant
Average
for
Collocated
Pairs
Average
for
Unequipped
Acenaphthene
1.79
9.87
1.71
155
3.99
2.93
4.19
Acenaphthylene
7.92
7.26
2.60
83
4.42
4.54
4.40
Anthracene
4.84
5.06
4.07
124
5.93
4.30
6.24
Benzo(a)anthracene
6.55
1.20
6.41
122
3.54
2.99
3.64
Benzo(a)pyrene
4.38
9.06
5.84
98
4.57
3.65
4.74
B enzo (b )fluoranthene
5.09
1.33
3.65
146
4.13
3.52
4.24
Benzo(e)pyrene
1.51
13.50
2.65
126
4.03
3.71
4.09
Benzo(g,hi)perylene
4.95
14.19
2.75
139
4.64
4.06
4.75
B enzo (k)fluoranthene
5.33
19.29
7.92
80
7.74
6.68
7.94
Chrysene
3.70
4.27
3.57
155
3.47
3.52
3.46
Coronene
7.19
4.68
4.70
106
4.80
4.94
4.78
Cyclopenta[cd]pyrene
--
--
--
16
6.97
9.79
6.49
Dibenz(a,h)anthracene
--
--
--
20
6.42
4.16
7.18
Fluoranthene
1.44
5.46
2.11
164
3.65
2.71
3.82
Fluorene
2.16
1.59
1.60
121
3.24
2.41
3.39
9-Fluorenone
2.40
3.91
1.80
164
3.74
2.96
3.89
Indeno( 1,2,3 -cd)pyrene
4.64
2.64
5.33
125
5.77
6.62
5.61
Naphthalene
5.18
2.41
1.28
164
4.39
3.30
4.60
Perylene
--
--
--
14
8.66
6.86
9.12
Phenanthrene
1.12
1.28
1.78
164
2.19
1.33
2.35
Pyrene
1.53
2.93
2.33
164
3.72
2.83
3.89
Retene
3.69
22.24
5.22
157
7.09
4.13
7.65
Average CV by Site
3.97
6.96
3.54
2,607
4.87
4.18
5.02
# of Pairs Collected by Site
8
7
8
— = 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 replicates
were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-55
-------
24.3.5 Metals Analytical Precision
Table 24-13 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. The CVs exhibit low- to
mid-level variability, ranging from 0 percent (for several sites and pollutants) to 33.33 percent
(beryllium for PAFL).
The pollutant-specific average CV, as shown in orange in Table 24-13, ranges from
1.67 percent (manganese) to 15.95 percent (beryllium). Beryllium is the only pollutant with a
pollutant-specific average CV greater than or equal to 15 percent. The site-specific average CV,
as shown in green in Table 24-13, ranges from 3.03 percent (LEKY) to 10.48 percent (BAKY);
all 21 sites sampling metals have site-specific average CVs less than 15 percent. The overall
average analytical precision CV is 6.51 percent.
Sites at which collocated metals samples were collected are highlighted in blue in
Table 24-13 while sites for which replicates were run on individual samples are highlighted in
brown. Collocated metals samples were collected at eight sites; replicates were run on individual
PAH samples at the remaining 13 sites. The average CV for sites that collected collocated metals
samples was calculated and is shown at the end of Table 24-13 in blue while the average CV for
sites for which replicates were run on individual samples is shown in brown. The variability
ranges from 5.95 percent (replicates run on individual samples) to 7.35 percent (replicates run on
collocated samples).
24-56
-------
Table 24-13. Metals Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant
Pollutant
ASKY-M
BAKY
BLKY
BOMA
BTUT
CCKY
Antimony
1.31
1.62
0.79
1.38
2.14
2.12
Arsenic
7.69
15.06
8.05
19.44
13.62
14.67
Beryllium
17.59
24.43
--
27.75
20.54
0.00
Cadmium
4.06
9.05
2.32
7.90
13.85
10.79
Chromium
1.56
--
--
0.80
--
--
Cobalt
1.36
6.51
0.00
5.22
4.45
0.00
Lead
0.85
1.64
1.30
1.37
1.66
0.85
Manganese
0.94
1.98
1.05
1.22
1.55
0.64
Mercury
13.42
26.24
--
21.04
11.03
0.00
Nickel
4.65
8.28
13.86
4.96
4.04
17.23
Selenium
3.45
9.98
1.30
8.91
8.88
8.06
Average CV by Site
5.17
10.48
3.58
9.09
8.17
5.44
# of Pairs Collected by Site
15
6
1
70
14
5
— = 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 replicates
were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-57
-------
Table 24-13. Metals Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
GLKY
GPCO
LEKY
NBIL
OCOK
PAFL
Antimony
1.89
1.44
1.71
4.17
1.87
1.64
Arsenic
15.14
11.35
5.68
4.80
5.45
3.80
Beryllium
24.59
15.79
--
15.18
10.34
33.33
Cadmium
13.40
7.88
5.25
18.15
4.28
17.93
Chromium
0.00
--
--
2.54
2.39
--
Cobalt
13.08
5.48
3.40
4.00
6.18
10.95
Lead
1.39
0.60
0.79
3.83
6.52
4.03
Manganese
1.00
0.97
0.81
4.82
5.16
3.63
Mercury
7.86
--
--
12.24
7.23
9.36
Nickel
16.03
5.24
4.18
0.92
4.68
0.65
Selenium
9.93
7.30
2.41
4.48
3.96
1.77
Average CV by Site
9.48
6.23
3.03
6.83
5.28
8.71
# of Pairs Collected by Site
55
24
6
7
8
4
— = 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 replicates
were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-58
-------
Table 24-13. Metals Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
PXSS
S4MO
SEWA
SJJCA
TMOK
TOOK
Antimony
1.22
3.89
1.50
0.89
1.19
1.37
Arsenic
14.66
13.17
12.51
21.86
1.42
1.95
Beryllium
17.41
19.93
--
--
6.80
7.30
Cadmium
1.70
6.04
8.96
16.41
23.26
2.69
Chromium
--
--
--
--
1.90
1.58
Cobalt
0.65
5.04
1.13
3.26
1.09
2.67
Lead
0.52
0.83
0.96
0.85
1.12
1.62
Manganese
0.78
0.86
0.78
1.10
2.37
1.88
Mercury
21.35
21.15
21.69
12.50
4.98
8.18
Nickel
3.27
5.52
1.48
8.95
2.67
1.94
Selenium
1.68
7.07
--
5.36
1.24
2.35
Average CV by Site
6.32
8.35
6.13
7.91
4.37
3.05
# of Pairs Collected by Site
6
121
6
6
3
113
— = 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 replicates
were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-59
-------
Table 24-13. Metals Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
TROK
UNVT
YUOK
# of
pairs
Average
by
Pollutant
Average
for
Collocated
Pairs
Average
for
Unequipped
Antimony
1.93
3.57
1.55
494
1.87
2.12
1.71
Arsenic
2.81
8.74
3.44
406
9.78
11.39
8.78
Beryllium
9.99
--
4.27
249
15.95
19.07
13.53
Cadmium
7.66
27.36
4.11
496
10.14
10.40
9.99
Chromium
3.01
--
2.33
128
1.79
0.98
2.43
Cobalt
1.75
11.59
2.03
479
4.28
6.11
3.15
Lead
3.64
1.60
1.90
496
1.80
1.24
2.15
Manganese
1.39
1.17
0.93
496
1.67
1.20
1.96
Mercury
10.87
--
12.95
226
13.06
13.78
12.67
Nickel
2.64
17.74
1.66
455
6.22
7.51
5.42
Selenium
2.43
8.86
1.07
351
5.02
7.09
3.64
Average CV by Site
4.37
10.08
3.29
4,276
6.51
7.35
5.95
# of Pairs Collected by Site
6
13
7
- = 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 replicates were
run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24.3.6 Hexavalent Chromium Analytical Precision
Table 24-14 presents analytical precision results from replicate analyses of collocated
samples as the CV per site and the overall average CV for hexavalent chromium. Recall from
Section 24.2.6 that two NATTS sites sampled hexavalent chromium in 2014, RIVA and S4MO,
although sampling was discontinued in July 2014 at S4MO. The site-specific CV for RIVA
(14.81 percent) is considerably higher than the CV for S4MO (2.69 percent), although both CVs
are less than 15 percent. The CVs for both sites are based on two sets of collocated samples and
their replicates, resulting in four pairs each. The overall average analytical precision of
hexavalent chromium is 8.75 percent, as shown in orange in Table 24-14.
24-60
-------
Table 24-14. Hexavalent Chromium Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site
Pollutant
RIVA
S4MO
# of pairs
Average
by
Pollutant
Hexavalent Chromium
14.81
2.69
8
8.75
# of Pairs Collected by Site
4
4
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 (or bias) 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 compares the measured concentrations to the reference concentrations 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, metals, and hexavalent chromium, which
are used to quantitatively measure analytical accuracy. Tables 24-15 through 24-19 present
ERG's results for PT audit samples analyzed in 2014. Note that the way in which the results of
the PT audit are reported changed mid-2014. Results for PT audit samples prepared prior to May
2014 are presented as percent recovery while the results from the second half of the year are
presented as percent difference. Percent recovery-based audit results are calculated as follows:
Percent of True (% Recovery) = —— x 100
^true
Where:
Xhb is the analytical result from the laboratory;
Xtrue is the true concentration of the audit sample.
This calculation results in percent recovery values near 100 percent. The program MQO is
± 25 percent recovery; thus, percent recovery values between 75 percent and 125 percent are
24-61
-------
acceptable. Beginning with May 2014 PT audit samples, the results are presented as percent
difference, which is calculated as follows:
Percent Difference = ———— x 100
y
true
Where:
Xiab is the analytical result from the laboratory;
Xtms is the true concentration of the audit sample.
This calculation results in values that appear much lower than percent recovery. The program
MQO of ± 25 percent still applies but percent difference values between 0 percent and
± 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.
The results of the 2014 PT audit samples show that few of the pollutants for which PT
audit samples were analyzed exceed the MQO for accuracy. Of the 67 results provided in
Tables 24-15 through Table 24-19, only four exceed the MQO for accuracy (three for VOCs and
one for metals). However, none failed multiple audits in 2014.
24-62
-------
Table 24-15. TO-15 NATTS PT Audit Samples
Pollutant
February 20141
August 20142
% Recovery
% Difference
Acrolein
127.1
5.2
Benzene
111.1
14.5
1,3-Butadiene
104.2
-1.8
Carbon Tetrachloride
130.9
21.6
Chloroform
129.7
7.2
1,2-Dibromoethane
106.3
-6.9
1,2-Dichloroethane
123.0
-3.8
Dichloro methane
113.5
13.8
1,2-Dichloropropane
112.9
2.3
cis-1.3 -Dichloropropene
107.9
-1.6
trans-1,3 -Dichloropropene
119.5
-17.7
1,1,2,2 -T etrachloroethane
119.1
13.0
T etrachloroethylene
112.3
0.1
T richloroethylene
112.9
-4.9
Vinyl chloride
116.8
12.5
1 The true value is based on the confirmation laboratory's mean.
2 The true value is based on the mean of participating NATTS laboratories.
Bold = Greater than ± 25 percent MQO
Table 24-16. TO-11A NATTS PT Audit Samples
Pollutant
February 20141
August 20142
% Recovery
% Difference
Acetaldehyde
NS
-1.0
Benzaldehyde
98.5
4.2
Formaldehyde
99.0
-1.5
Propionaldehyde
86.5
-4.7
NS = Not spiked onto PT audit sample provided to the laboratory
1 The true value is based on the confirmation laboratory's mean.
2 The true value is based on the mean of participating NATTS laboratories.
Bold = Greater than ± 25 percent MQO
24-63
-------
Table 24-17. TO-13A NATTS PT Audit Samples1
Pollutant
May 2014
December 2014
Acenaphthene
6.2
6.2
Anthracene
4.6
12.5
Benzo(a)pyrene
NS
15.4
Fluoranthene
8.2
NS
Fluorene
-5.1
12.4
Naphthalene
6.6
13.3
Phenanthrene
-5.2
4.3
Pyrene
1.9
9.6
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 24-18. Metals NATTS PT Audit Samples1
Pollutant
May 2014
December 2014
Antimony
-8.8
-29.1
Arsenic
4.6
17.0
Beryllium
NS
8.4
Cadmium
NS
0.1
Cobalt
3.5
3.1
Lead
-2.0
-0.5
Manganese
4.4
NS
Nickel
-19.5
0.1
Selenium
17.6
0.6
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 24-19. Hexavalent Chromium NATTS PT Audit Sample1
Pollutant
May 2014
Hexavalent Chromium
8.0
1 Audit result based on percent difference from mean of participating
NATTS laboratories.
Bold = Greater than ± 25 percent MQO
24-64
-------
ERG's use of the ICP/MS was approved in 2012 as a FEM for the sampling and analysis
of lead for adherence to the National Ambient Air Quality Standards (NAAQS) (EPA 2012a).
This approval requires additional quality assurance steps, including the analysis of quarterly
audit strips. Table 24-20 provides the results of the quarterly NAAQS audit results for lead for
ERG for 2014. All results are within the percent recovery target of ± 15 percent.
Table 24-20. Lead NAAQS Quarterly Audit Samples1'2
Pollutant
Filter #
Analysis #
March 2014
September 2014
December 2014
1
95.8
102.0
99.4
1
2
91.2
102.6
97.6
Lead
3
97.0
101.3
95.2
1
97.8
98.2
100.0
2
2
98.9
97.3
100.0
3
96.6
97.2
96.9
1 Audit result represents percent of nominal spike value.
2 A second quarter 2014 audit sample was not prepared by the confirmation laboratory.
Bold = Greater than ±15 percent recovery target
The accuracy of the 2014 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 2014 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 2014 monitoring data accurately represent ambient air quality.
24-65
-------
25.0 Results, Conclusions, and Recommendations
The following discussion 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 2014 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 51 monitoring sites are EPA-designated
NATTS sites. An additional 30 UATMP sites participated in the NMP in 2014. Data
from one special study site (ROIL) are also included in the report. The number of
NATTS sites whose data are included in the 2014 NMP report is less than in previous
years due to the removal of hexavalent chromium from the list of target pollutants for
which to monitoring is required.
• Total number of samples collected and analyzed. Over 7,800 valid samples were
collected at participating program sites and analyzed at the ERG laboratory, yielding
nearly 225,000 valid measurements of air toxics, including primary, duplicate,
collocated, and replicate results.
• Detects. Of the 198 pollutants for which statistical summaries are provided in
Tables 4-1 through 4-6, all but five were detected at least once over the course of the
2014 monitoring effort. The detection of a given pollutant is subject to the sensitivity
limitation 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 55 percent of the reported
measurements were greater than the associated MDLs. At the method level, this
percentage varies considerably, from 42 percent for SNMOCs to 83 percent for
carbonyl compounds. Quantification below the MDL is possible and an acceptable
analytical result; therefore, these results are incorporated into the data analyses. These
measurements account for 9 percent of concentrations. Non-detects account for the
remaining 36 percent of results.
25-1
-------
• 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". Thirty-four pollutants failed at least one
risk screening value; of those pollutants, 12 were identified as program-level
pollutants of interest.
• Seasonal Trends. Fewer pollutants exhibited identifiable seasonal trends in the
concentrations measured during the 2014 program year (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. Conversely, benzene and naphthalene
concentrations tended to be higher during the colder months of the year, particularly
during the first quarter, although higher benzene concentrations were also measured
during the third quarter of 2014.
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.
• Fifteen pollutants failed screens for PXSS, 11 of which contributed to 95 percent of
failed screens. PXSS failed the highest number of screens among all NMP sites. Six
pollutants failed screens for SPAZ, all of which contributed to 95 percent of failed
screens.
• Of the pollutants of interest for PXSS, formaldehyde has the highest annual average
concentration, followed by acetaldehyde and benzene. These are the only pollutants
of interest with annual average concentrations greater than 1 |ig/m3.
• Benzene has the highest annual average concentration for SPAZ, and is the only
pollutant with an annual average concentration greater than 1 |ig/m3.
• SPAZ and PXSS have the highest annual average concentrations of
/;-dichlorobenzene and ethylbenzene among NMP sites sampling this pollutant. These
two sites also have the second and third highest annual average concentrations of
1,3-butadiene.
• 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 site for
the site-specific pollutants of interest. Benzene and ethylbenzene concentrations
measured at both sites have decreased over recent years. The detection rate of
1,2-dichloroethane at both sites has been steadily increasing over the years.
25-2
-------
• Formaldehyde has the highest cancer risk approximation for PXSS and is the only
pollutant of interest with a cancer risk approximation greater than 10 in-a-million for
either site. Benzene has the highest cancer risk approximation for SPAZ. None of the
pollutants of interest for either site have a noncancer hazard approximation greater
than an HQ of 1.0.
• Benzene 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 screens for all three sites. A single concentration of
benzo(a)pyrene also failed a screen for CELA. Four other pollutants, in addition to
naphthalene and benzo(a)pyrene, failed screens for SJJCA.
• Naphthalene has the highest annual average concentration among the pollutants of
interest for CELA, RUCA, and SJJCA.
• 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. Concentrations of naphthalene
exhibit a decreasing trend at CELA in recent years while progressively higher
concentrations of nickel have been measured at SJJCA over the last several years.
• Of the pollutants of interest for each site, naphthalene has the highest cancer risk
approximation for all three California sites. The noncancer hazard approximations for
each pollutant of interest are considerably less than an HQ of 1.0 for all three sites.
• Formaldehyde is the highest emitted pollutant with a cancer toxicity factor in Los
Angeles and Riverside Counties, while benzene is the highest emitted pollutant with a
cancer toxicity factor in Santa Clara County. Formaldehyde has the highest cancer
toxicity-weighted emissions for all three 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
five UATMP sites located northeast of Grand Junction in Garfield County. The sites
are located in the towns of Battlement Mesa (BMCO), Silt (BRCO), Parachute
(PACO), Carbondale (RFCO), and Rifle (RICO).
• VOCs, carbonyl compounds, PAHs, and metals (PMio) were sampled for at GPCO.
SNMOCs and carbonyl compounds were sampled for at the Garfield County sites.
• Sixteen pollutants failed at least one screen for GPCO, 11 of which contributed to
95 percent of failed screens. Five pollutants failed screens for PACO and RICO,
while four pollutants failed screens for BRCO and three pollutants failed screens for
BMCO. Benzene, formaldehyde, and acetaldehyde were identified as pollutants of
interest for all five Garfield County sites as well as GPCO.
• Of the pollutants of interest for GPCO, formaldehyde has the highest annual average
concentration, followed by acetaldehyde and benzene.
• RICO is the only Garfield County site for which annual average concentrations could
be calculated for all of the site-specific pollutants of interest. Benzene has the highest
annual average concentration for RICO. This is also true for RFCO and PACO,
although annual average concentrations could not be calculated for the carbonyl
compounds for these sites. Formaldehyde has the highest annual average
concentration for BMCO, although annual average concentrations could not be
calculated for the SNMOCs for this site. Annual average concentrations could not be
calculated for any of the pollutants of interest for BRCO due to a series of invalid
samples during sampling.
• PACO has the highest annual average concentration of benzene among NMP sites
sampling this pollutant for the second year in a row. RICO has the third highest
annual average concentration of benzene. GPCO has the second highest annual
average concentration of acetaldehyde and the third highest annual average
concentration of ethylbenzene among NMP sites sampling these pollutants.
• Sampling for the site-specific pollutants of interest has occurred at GPCO, BRCO,
PACO, and RICO for at least 5 consecutive years; thus, a trends analysis was
conducted for the site-specific pollutants of interest. Notable trends include: Benzene
concentrations at GPCO have an overall decreasing trend across the years of sampling
while concentrations of naphthalene have decreased in recent years. Concentrations
of acetaldehyde and formaldehyde have a decreasing trend at RICO.
• Formaldehyde has the highest cancer risk approximation for GPCO (by an order of
magnitude) and is the fifth highest cancer risk approximation calculated across the
program for 2014. Formaldehyde has the highest cancer risk approximation for RICO
as well, which is the only other Colorado site for which cancer risk approximations
for all of the site-specific pollutants of interest could be calculated. All noncancer
hazard approximations are less than an HQ of 1.0 for all of the Colorado sites, where
noncancer hazard approximations could be calculated.
25-4
-------
• 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.
• While toluene is the highest emitted pollutant with a noncancer toxicity factor for
both Mesa and Garfield Counties, 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 accounted for more than 96 percent of failed screens for this site and
was the only pollutant identified as a pollutant of interest. Benzo(a)pyrene and
fluorene each failed a single screen.
• Naphthalene was detected in every valid PAH sample collected at WADC. The
annual average concentration of naphthalene for WADC is the ninth highest annual
average concentration among NMP sites sampling this pollutant.
• 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 approximation for naphthalene is 2.29 in-a-million. The noncancer
hazard approximation for naphthalene is 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.
• 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.
25-5
-------
• Acetaldehyde and formaldehyde failed screens for all four Florida sites sampling
carbonyl compounds. Naphthalene also failed screens for SKFL. Arsenic was the
only speciated metal to fail screens for PAFL.
• Formaldehyde has the highest annual average concentration for AZFL, SYFL, and
ORFL; annual average concentrations of acetaldehyde were just slightly less for each
site. Annual average concentrations could not be calculated for the carbonyl
compounds for SKFL. PAFL's annual average arsenic concentration ranks fourth
highest among NMP sites sampling metals (PMio).
• Sampling for the site-specific pollutants of interest has occurred at all of 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: Acetaldehyde concentrations have a decreasing trend over the last few
years at SKFL, while concentrations of acetaldehyde exhibit an increasing trend at
ORFL.
• Formaldehyde has the highest cancer risk approximation for AZFL, SYFL, and
ORFL, ranging from roughly 25 in-a-million to 30 in-a-million. The cancer risk
approximation for arsenic for PAFL is 3.50 in-a-million. Naphthalene's cancer risk
approximation for SKFL is 1.85 in-a-million. 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. Benzene has the highest cancer toxicity-
weighted emissions for Pinellas County; formaldehyde has the highest cancer
toxicity-weighted emissions for Hillsborough County; and hexavalent chromium has
the highest cancer toxicity-weighted emissions for Orange County.
• Toluene is the highest emitted pollutant with a noncancer toxicity factor in all three
Florida counties with NMP sites while 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.
• 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.
• Eighteen pollutants failed screens for NBIL; 13 pollutants failed screens for SPIL;
and 12 pollutants failed screens for ROIL. Among the site-specific pollutants of
interest, the three Illinois sites have seven pollutants in common: two carbonyl
compounds (acetaldehyde and formaldehyde) and five VOCs (benzene,
25-6
-------
1,3-butadiene, carbon tetrachloride, 1,2-dichloroethane, and hexachloro-1,3-
butadiene).
• Acetaldehyde has the highest annual average concentration for NBIL, while
formaldehyde has the highest annual average concentration for SPIL and ROIL.
• NBIL has the second highest annual average concentration of naphthalene among
NMP sites sampling PAHs. ROIL has the second highest annual average
concentration of benzene among NMP sites sampling VOCs. SPIL has the fourth
highest annual average concentration of acetaldehyde among sites sampling carbonyl
compounds.
• 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. Most notably, concentrations of acetaldehyde have
been increasing significantly at NBIL in recent years, while concentrations of
formaldehyde have been decreasing. In addition, the detection rate of
1,2-dichloroethane at both NBIL and SPIL has been increasing steadily over the last
few years of sampling.
• Formaldehyde has the highest cancer risk approximation for all three Illinois 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.
• Toluene is the highest emitted pollutant with a noncancer toxicity factor for both
counties, while 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.
• Formaldehyde and acetaldehyde failed screens for both INDEM and WPIN; all of the
measured detections of formaldehyde failed screens for both sites.
• Formaldehyde has the highest annual average concentration for both sites.
25-7
-------
• 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 acetaldehyde and formaldehyde
have decreased at WPIN over the last few years.
• 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. Coke oven emissions (PM) have the highest cancer toxicity-
weighted emissions for Lake County while formaldehyde has the highest cancer
toxicity-weighted emissions for Marion County.
• 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 a NATTS site. One monitoring site is located south of Evansville,
Indiana (BAKY). Five monitoring sites are located in or near the Calvert City area
(ATKY, BLKY, CCKY, LAKY, and TVKY). The final monitoring site is located in
Lexington, in north-central Kentucky (LEKY).
• All of the Kentucky monitoring sites sampled for VOCs except ASKY-M and
BAKY. PAHs, carbonyl compounds, and PMio metals were also sampled for at
GLKY. Carbonyl compounds were also sampled for at ASKY and LEKY and PMio
metals were also sampled for at ASKY-M, BAKY, CCKY, and LEKY. The CCKY
site was discontinued in October 2014 and the metals instrumentation moved to the
BLKY site.
• The number of pollutants failing screens for the Kentucky sites varies from two
(BAKY) to 11 (LEKY and TVKY). Most of the Kentucky sites had nine or more
pollutants fail screens.
• Of the pollutants of interest for each site, formaldehyde has the highest annual
average concentration for all three sites sampling carbonyl compounds (GLKY,
ASKY, and LEKY). Manganese has the highest annual average concentration for
ASKY-M, while arsenic has the highest annual average concentration for BAKY
(although arsenic was also the only pollutant of interest for this site). Carbon
tetrachloride has the highest annual average concentration for CCKY, while
1,2-dichloroethane has the highest annual average concentration for BLKY, LAKY,
and TVKY and vinyl chloride has the highest annual average concentration for
ATKY.
25-8
-------
• ASKY-M has the highest annual average concentrations of arsenic and nickel among
NMP sites sampling PMio metals.
• The Calvert City sites account for the five highest annual average concentrations of
both carbon tetrachloride and 1,2-dichloroethane, and account for three of the highest
annual average concentrations of 1,3-butadiene.
• Sampling for the site-specific pollutants of interest has occurred at GLKY for at least
5 consecutive years; thus, a trends analysis was conducted for the site-specific
pollutants of interest. Benzene concentrations measured at GLKY have a slight
decreasing trend over the years of sampling.
• Formaldehyde has the highest cancer risk approximations among the pollutants of
interest for the three Kentucky sites sampling carbonyl compounds while arsenic has
the highest cancer risk approximation among the two Kentucky sites sampling only
PMio metals. Among the Calvert City sites, 1,2-dichloroethane has the highest cancer
risk approximations. The cancer risk approximation for TVKY for 1,2-dichloroethane
is the highest cancer risk approximation calculated among the site-specific pollutants
of interest across the program. 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 Kentucky
counties with NMP sites, except Henderson County, where benzene ranks second to
formaldehyde. Coke oven emissions have the highest cancer toxicity-weighted
emissions for Boyd County; formaldehyde has the highest cancer toxicity-weighted
emissions for Carter, Henderson, Livingston, and Fayette Counties; 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.
Massachusetts.
• The Massachusetts monitoring site (BOMA) is a NATTS site located in Boston.
• Metals (PMio) and PAHs were sampled for at BOMA.
• Four pollutants failed screens for BOMA. Arsenic and naphthalene each accounted
for at least 40 percent of the site's failed screens.
• Of the pollutants of interest, naphthalene has the highest annual average
concentration.
25-9
-------
• BOMA has the second highest annual average concentration of nickel among NMP
sites sampling PMio metals.
• 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.
• Naphthalene and arsenic are the only pollutants of interest to have cancer risk
approximations greater than 1.0 in-a-million for BOMA. 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.
• Thirteen pollutants failed screens for DEMI, of which nine were identified as
pollutants of interest.
• Formaldehyde and acetaldehyde have the highest annual average concentrations for
DEMI. DEMI has the highest annual average concentration of naphthalene among
NMP sites sampling PAHs.
• 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
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. In addition, the detection rate of 1,2-dichloroethane at DEMI has been
increasing steadily over the last few years of sampling.
• Formaldehyde has the highest cancer risk approximation 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.
25-10
-------
Missouri.
• The NATTS site in Missouri (S4MO) is located in St. Louis.
• VOCs, carbonyl compounds, PAHs, metals (PMio), and hexavalent chromium were
sampled for at S4MO, although hexavalent chromium sampling was discontinued
after July 4, 2014.
• Twenty-three pollutants failed at least one screen for S4MO, 14 of which contributed
to 95 percent of failed screens. S4MO has the second 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 second highest annual average concentration of arsenic (PMio) and the
third highest annual average concentration of />dichlorobenzene 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 benzene have an overall decreasing trend at
S4MO, and concentrations of ethylbenzene and cadmium have decreased as well.
• Formaldehyde has the highest cancer risk approximation 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 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).
New Jersey.
• Three of the 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), and
North Brunswick (NBNJ). A fourth UATMP site (CSNJ) is located in the
Philadelphia-Camden-Wilmington CBS A.
• VOCs and carbonyl compounds were sampled for at all four New Jersey sites.
• Fourteen pollutants failed at least one screen for CSNJ; nine pollutants failed at least
one screen for CHNJ; and 12 pollutants failed at least one screen for both ELNJ and
NBNJ. The New Jersey sites have six pollutants of interest in common: acetaldehyde,
formaldehyde, benzene, carbon tetrachloride, 1,3-butadiene, and 1,2-dichloroethane.
25-11
-------
• Of the site-specific pollutants of interest, formaldehyde and acetaldehyde have the
highest annual average concentrations for CSNJ, CHNJ, and ELNJ. Carbon
tetrachloride had the highest annual average concentration for NBNJ, although annual
average concentrations for the carbonyl compounds could not be calculated. A
defective sampler resulted in the invalidation of samples collected between
May 5, 2014 and December 31, 2014.
• CSNJ and ELNJ rank second and third, respectively, for their annual average
concentrations of formaldehyde; ELNJ also has the third highest annual average
concentration of acetaldehyde among NMP sites sampling carbonyl compounds.
• Sampling for the site-specific pollutants of interest has occurred at three of the four
New Jersey sites 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 ELNJ, CHNJ, and NBNJ.
Benzene and ethylbenzene concentrations have decreased significantly at ELNJ since
sampling began. At CHNJ, concentrations of 1,3-butadiene have been increasing in
recent years. In addition, the detection rates of 1,2-dichloroethane and hexachloro-
1,3-butadience have been increasing steadily over the last few years of sampling at
CHNJ, ELNJ, and NBNJ.
• Formaldehyde has the highest cancer risk approximation for CSNJ, CHNJ, and ELNJ.
Benzene has the highest cancer risk approximation for NBNJ (where cancer risk
approximations could not be calculated for the 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 and formaldehyde are the highest emitted pollutants with cancer toxicity
factors in Camden, Union, Middlesex, and Morris Counties. These two pollutants also
have the highest toxicity-weighted emissions for each county, although the order
varied.
• 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 New York City (BXNY) and Rochester
(ROCH). Both are NATTS sites.
• PAHs were sampled for at both BXNY and ROCH.
• Six pollutants failed screens for BXNY and four pollutants failed screens for ROCH.
Naphthalene failed the majority of screens for both sites.
25-12
-------
• Naphthalene has the highest annual average concentration for BXNY and ROCH,
although the annual average concentration for BXNY is nearly twice the annual
average calculated for ROCH.
• BXNY has the third highest annual average concentration of naphthalene among
NMP sites sampling PAHs and is one of only four sites with an annual average
concentration greater than 100 ng/m3.
• Sampling for the site-specific pollutants of interest has occurred at ROCH for greater
than 5 consecutive years; thus, a trends analysis was conducted for each of the site-
specific pollutants of interest. The maximum concentrations for each of ROCH's
three pollutants of interest were measured in 2014.
• Naphthalene has the highest cancer risk approximation 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 these two sites are considerably less than an HQ of 1.0.
• Benzene is the highest emitted pollutant with a cancer toxicity factor for Bronx and
Monroe 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 both
Bronx and Monroe Counties while acrolein has the highest noncancer toxicity-
weighted emissions for both counties.
Oklahoma.
• There are five UATMP sites in Oklahoma: three are located in Tulsa (TOOK,
TMOK, and TROK) and two are located in or near Oklahoma City (OCOK and
YUOK).
• VOCs, carbonyls compounds, and metals (TSP) were sampled for at each of the
Oklahoma sites. The Oklahoma sites are the only NMP sites sampling TSP metals.
• Sixteen pollutants failed screens for TOOK; 15 failed screens for TMOK; 14 failed
screens for TROK and OCOK; and 13 failed screens for YUOK.
• Formaldehyde and acetaldehyde have the highest annual average concentrations for
each of the five Oklahoma sites.
• The three Tulsa sites have the fourth through sixth highest annual average
concentrations of />dichlorobenzene among NMP sites sampling this pollutant.
TOOK and TROK also have the fourth and fifth highest annual average
concentrations of ethylbenzene, respectively, with the annual average concentration
for TMOK ranking eighth. These sites also have some of the highest annual average
concentrations of hexachloro-l,3-butadiene among NMP sites sampling VOCs.
25-13
-------
• 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. Acetaldehyde, ethylbenzene, benzene, and
manganese concentrations have decreased at TOOK in recent years. Acetaldehyde,
benzene and ethylbenzene concentrations at TMOK have also decreased in recent
years. Acetaldehyde and formaldehyde concentrations at OCOK have also decreased.
Detection rates of 1,2-dichloroethane have increased at TOOK, TMOK, and OCOK
in recent years.
• Formaldehyde has the highest cancer risk approximations for each of the Oklahoma
monitoring sites. 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 and has the highest cancer toxicity-weighted emissions for both
counties. Formaldehyde is the highest emitted pollutant with a cancer toxicity factor
in Canadian County and has the highest cancer toxicity-weighted emissions for that
county.
• Toluene is the highest emitted pollutant with a noncancer toxicity factor in Oklahoma
and Tulsa Counties, while xylenes are the highest emitted pollutant with a noncancer
toxicity factor in Canadian County. Acrolein has the highest noncancer toxicity-
weighted emissions for all three counties.
Rhode Island.
• The Rhode Island monitoring site (PRRI) is located in Providence and is a NATTS
site.
• PAHs were sampled for at PRRI.
• Two pollutants failed screens for PRRI, although all but one of PRRI's failed screens
were attributable to naphthalene. As a result, naphthalene is PRRI's only pollutant of
interest.
• Naphthalene concentrations measured at PRRI span an order of magnitude, ranging
from 14.7 ng/m3 to 163 ng/m3.
• 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
in recent years.
• The cancer risk approximation for naphthalene for PRRI is 1.77 in-a-million. The
noncancer hazard approximation for this pollutant is considerably less than an HQ of
1.0.
25-14
-------
• 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.
• Nineteen pollutants failed screens for BTUT, 12 of which contributed to 95 percent of
this site's failed screens.
• Of the site-specific pollutants of interest, dichloromethane has the highest annual
average concentration for BTUT, which is consistent with previous years of
sampling. BTUT has the highest annual average concentrations of hexachloro-1,3-
butadiene, formaldehyde, and acetaldehyde among NMP sites sampling these
pollutants.
• 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 also have
also decreased in recent years. Concentrations of acetaldehyde and formaldehyde
exhibit decreases for 2014 following a significant increase in concentrations for 2013.
• The pollutant with the highest cancer risk approximation for BTUT is formaldehyde;
this is the second 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
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 for Davis County.
Vermont.
• The NATTS site in Vermont (UNVT) is located in Underhill, near the city of
Burlington.
• PAHs and metals (PMio) were sampled for at UNVT. However, the Vermont
Department of Environmental Conservation invalidated all of its nickel and total
chromium concentrations for the second half of 2014. This is due to a contamination
issue related to a new weighing and equilibration chamber at their laboratory
25-15
-------
• Arsenic (PMio) was the only pollutant to fail screens for UNVT.
• Arsenic concentrations measured at UNVT were less than 1.0 ng/m3, ranging from
0.003 ng/m3 to 0.83 ng/m3.
• UNVT has the lowest annual average concentration of arsenic (PMio) among NMP
sites sampling the pollutant.
• Sampling for site-specific pollutants of interest has occurred at UNVT for at least 5
consecutive years; thus, a trends analysis was conducted, where applicable. Changes
in annual average concentrations of arsenic for UNVT are not statistically significant.
• The cancer risk approximation for arsenic (PMio) for UNVT is 0.92 in-a-million. The
noncancer hazard approximation for this pollutant is considerably less than an HQ of
1.0.
• Benzene is the highest emitted pollutant with a cancer toxicity factor in Chittenden
County, while formaldehyde 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 is located near Richmond (RIVA).
• PAHs and hexavalent chromium were sampled for at RIVA. RIVA is the only site at
which hexavalent chromium was sampled for year-round.
• Three pollutants failed screens for RIVA, with concentrations of naphthalene
accounting for 96 percent of failed screens, and thus, is the only pollutant of interest
for this site.
• Naphthalene concentrations measured at RIVA range from 21.3 ng/m3 to 178 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.
• The cancer risk approximation for naphthalene at RIVA is 2.13 in-a-million, while
the noncancer hazard approximation is 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.
Toluene is the highest emitted pollutant with a noncancer toxicity factor in Henrico
County, while acrolein has the highest noncancer toxicity-weighted emissions.
25-16
-------
Washington.
• The NATTS site in Washington is located in Seattle (SEWA).
• VOCs, carbonyl compounds, PAHs, and metals (PMio) were sampled for at SEWA.
• Thirteen pollutants failed screens for SEWA, of which nine 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 and carbon tetrachloride have the
highest annual average concentrations for this site. The annual average concentration
of formaldehyde for SEWA is the lowest among NMP sites sampling this pollutant.
• SEWA has the third highest annual average concentration of nickel among NMP sites
sampling metals (PMio). This site had the second highest annual average nickel
concentration for 2012 and 2013.
• 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 a decreasing trend at SEWA.
Concentrations of naphthalene exhibit a significant decrease for 2014. In addition, the
detection rate of 1,2-dichloroethane at SEWA has been increasing steadily over the
last few years of sampling.
• Formaldehyde has the highest cancer risk approximation for SEWA, although it is the
lowest cancer risk approximation for formaldehyde among NMP sites. All of the
noncancer hazard approximations for the pollutants of interest for SEWA are less
than an HQ of 1.0.
• 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.
25-17
-------
25.1.3 Composite Site-level Results Summary
• Twenty-two pollutants were identified as site-specific pollutants of interest, based on
the risk-based screening process. Acetaldehyde, formaldehyde, and benzene were the
most common pollutants of interest among the monitoring sites. Benzene was
identified as a pollutant of interest for all 32 sites that sampled this pollutant (with
Method TO-15 or SNMOC). Acetaldehyde and formaldehyde were identified as
pollutants of interest for all 32 sites that sampled carbonyl compounds. Naphthalene
was identified as a pollutant of interest for 17 of the 19 sites that sampled PAHs (with
GLKY and UNVT as the exceptions). Arsenic was identified as a pollutant of interest
for 20 of the 21 sites that sampled metals (with BLKY, at which metals sampling did
not begin until late October, as the exception).
• 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 BTUT;
trichloroethylene is a pollutant of interest for only SPIL; and 1,1,2-trichloroethane is a
pollutant of interest for only TVKY.
• 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; as such, all but two of the participating NATTS sites
(S4MO and RIVA) stopped sampling this pollutant prior to 2014. Concentrations of
hexavalent chromium measured at S4MO and RIVA failed few screens in 2014 and
hexavalent chromium was not identified as a pollutant of interest for either site.
• 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 23 sites. Naphthalene had the next highest at 10 followed
by benzene with four.
• Five sites have cancer risk approximations greater than 50 in-a-million, four for
formaldehyde (BTUT, GPCO, CSNJ, ELNJ) and one for 1,2-dichloroethane (TVKY).
Formaldehyde tended to have the highest cancer risk approximation on a site-specific
basis. This is true for 27 NMP sites. The highest cancer risk approximation for
formaldehyde was calculated for BTUT (76.95 in-a-million). Yet, this is the second
highest the annual average-based cancer risk approximation. The cancer risk
approximation for 1,2-dichloroethane based on TVKY's annual average
concentration is 91.92 in-a-million. Benzene and 1,3-butadiene are the only other
pollutants for which a cancer risk approximation greater than 10 in-a-million was
calculated (one each).
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
1,3-Butadiene
Cadmium
Carbon Tetrachloride
p-Dichlorobenzene
1,2-Dichloroethane
Dichloromethane
Ethylbenzene
Fluoranthene
Fluorene
Formaldehyde
Hexachloro-l,3-butadiene
Manganese
Naphthalene
Nickel
1,1,2-T richloroethane
T richlo roethy lene
Vinyl chloride
AZ
PXSS
11
X
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
4
X
X
X
X
CO
BMCO
3
X
X
X
CO
BRCO
3
X
X
X
CO
GPCO
11
X
X
X
X
X
X
X
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
WADC
1
X
FL
AZFL
2
X
X
FL
ORFL
2
X
X
FL
PAFL
1
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
8
X
X
X
X
X
X
X
X
IL
SPIL
8
X
X
X
X
X
X
X
X
IN
INDEM
2
X
X
IN
WPIN
2
X
X
BOLD ITALICS = EPA-designated NATTS Site
-------
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
1,3-Butadiene
Cadmium
Carbon Tetrachloride
p-Dichlorobenzene
1,2-Dichloroethane
Dichloromethane
Ethylbenzene
Fluoranthene
Fluorene
Formaldehyde
Hexachloro-l,3-butadiene
Manganese
Naphthalene
Nickel
1,1,2-T richloroethane
T richlo roethy lene
Vinyl chloride
KY
ASKY
6
X
X
X
X
X
X
KY
ASKY-M
4
X
X
X
X
KY
ATKY
6
X
X
X
X
X
X
KY
BAKY
1
X
KY
BLKY
6
X
X
X
X
X
X
KY
CCKY
7
X
X
X
X
X
X
X
KY
GLKY
7
X
X
X
X
X
X
X
KY
LAKY
6
X
X
X
X
X
X
KY
LEKY
7
X
X
X
X
X
X
X
KY
TVKY
6
X
X
X
X
X
X
X
MA
BOMA
3
X
X
X
MI
DEMI
9
X
X
X
X
X
X
X
X
X
MO
S4MO
14
X
X
X
X
X
X
X
X
X
X
X
X
X
X
NJ
CHNJ
6
X
X
X
X
X
X
NJ
CSNJ
8
X
X
X
X
X
X
X
X
NJ
ELNJ
8
X
X
X
X
X
X
X
X
NJ
NBNJ
8
X
X
X
X
X
X
X
X
NY
BXNY
5
X
X
X
X
X
NY
ROCH
3
X
X
X
OK
OCOK
7
X
X
X
X
X
X
X
OK
TMOK
10
X
X
X
X
X
X
X
X
X
X
OK
TOOK
12
X
X
X
X
X
X
X
X
X
X
X
X
OK
TROK
10
X
X
X
X
X
X
X
X
X
X
BOLD ITALICS = EPA-designated NATTS Site
-------
Table 25-1. Summary of Site-Specific Pollutants of Interest (Continued)
State
Site
# of Pollutants of Interest
Acenaphthene
Acetaldehyde
Arsenic
Benzene
Benzo(a)pyrene
1,3-Butadiene
Cadmium
Carbon Tetrachloride
p-Dichlorobenzene
1,2-Dichloroethane
Dichloromethane
Ethylbenzene
Fluoranthene
Fluorene
Formaldehyde
Hexachloro-l,3-butadiene
Manganese
Naphthalene
Nickel
1,1,2-T richloroethane
T richlo roethy lene
Vinyl chloride
OK
YUOK
8
X
X
X
X
X
X
X
X
RI
PRRI
1
X
UT
BTUT
12
X
X
X
X
X
X
X
X
X
X
X
X
VA
RIVA
1
X
VT
UNVT
1
X
WA
SEWA
9
X
X
X
X
X
X
X
X
X
Total
287
5
32
20
32
2
30
2
27
6
27
1
14
2
5
32
19
2
17
6
1
1
5
BOLD ITALICS = EPA-designated NATTS Sites
-------
Carbon tetrachloride often had relatively high cancer risk approximations (based on
annual average concentrations) 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 four of the five 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 (with an HQ of 0.60) 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, 1,3-butadiene, and arsenic.
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, hexane, 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 acrolein emissions were
generally low when compared to other pollutants. Acrolein appears only twice among
the 10 highest emitted pollutants for counties with NMP sites (Garfield County,
Colorado and Canadian County, 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 rarely measured at detectable
levels (this pollutant has a 16 percent detection rate across the program). The five
Calvert City, Kentucky sites together account for more than 72 percent of the
measured detections of vinyl chloride for 2014 (which is a similar percentage as
2013). Individually, these sites have the highest number of measured detections of
vinyl chloride among NMP sites sampling VOCs. The Calvert City sites also account
25-22
-------
for the 124 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 (16 sites), a dramatic increase
in the number of measured detections is shown over the most recent years of
sampling, particularly for 2012, which was mostly sustained for 2013 and 2014. 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 75 percent and 100 percent for most NMP sites for 2014.
25.1.4 Data Quality Results Summary
Completeness, precision, and accuracy were assessed for the 2014 monitoring effort. The
quality assessments presented in this report show that the 2014 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.
Only seven out of 108 site- and method-specific datasets failed to comply with the MQO of
85 percent completeness while 30 datasets achieved 100 percent completeness.
Method (i.e., sampling and analytical) precision and analytical precision were determined
for the 2014 NMP monitoring efforts using CV calculations based on duplicate, collocated, and
replicate samples. Method precision for most analytical methods utilized during the 2014 NMP
was within the MQO of 15 percent CV (with the exceptions of TO-13A (PAHs) and hexavalent
chromium). 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
25-23
-------
the MQO for accuracy. Of the 37 pollutants analyzed for via audit samples, only four exceeded
the MQO of ± 25 percent recovery (and none failed multiple audits).
25.2 Conclusions
Conclusions extrapolated from the data analyses of the data generated from the 2014
NMP monitoring efforts are presented below.
• A large number 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, 1,2-dichloroethane, and
1,3-butadiene. Some of the lesser detected pollutants still fail relatively large numbers
of screens. For example, even though hexachloro-1,3-butadiene was detected
relatively infrequently, most of the measured detections failed screens. The MDL for
this pollutant is relatively high (0.29 |ig/m3) 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
hovered around 36 percent for the last four years. Risk screening values are often
updated from year-to-year, although the only changes for the 2014 report have to do
with how the PAHs are grouped into POM Groups and not the screening values
themselves.
• For those pollutants for which annual average concentrations could be calculated and
that have available cancer UREs, none of the cancer risk approximations were greater
than 100 in-a-million. In total, 31 site- and pollutant-specific cancer risk
approximations were greater than 10 in-a-million (24 for formaldehyde, five for
1,2-dichloroethane, and one each for benzene and 1,3-butadiene); and nearly
80 percent were greater than 1.0 in-a-million.
• For 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
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
25-24
-------
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 changes from year-to-year.
The number of sites participating in the 2014 NMP decreased considerably, 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.
Many of the data analyses utilized in this report require data from year-round (or
nearly year-round) sampling. Of the 108 site-method combinations, only three site-
method combinations did not cover the entire year: Sampling at the CCKY site was
discontinued in October 2014 and the metals instrumentation was moved to BLKY,
where sampling resumed. Hexavalent chromium sampling was discontinued at S4MO
in July 2014. Thus, the percentage of time-period averages and subsequent risk-based
analyses that could not be calculated decreased for 2014 compared to 2013. Fewer
data gaps allow for more complete results and inter-site comparisons.
Of the 51 monitoring sites participating in the 2014 NMP, none sampled for all six
available pollutant groups under the NMP through the national contract laboratory.
Three sites (S4MO, BTUT, and NBIL) sampled for five pollutant groups and another
four sites (GLKY, PXSS, GPCO, 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 2014 NMP report reflect the inclusion of data from
a number of source-oriented monitoring sites. Newer 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 2014 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 for 2014
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 2013 and 2014 NMP reports. The major difference between the 2014
report and other reports in recent years is the use of meteorological measurements
collected at the sites themselves (as opposed to NWS data).
25-25
-------
25.3 Recommendations
Based on the conclusions from the 2014 NMP, a number of recommendations for future
ambient air monitoring efforts are presented below.
• Continue participation in the National Monitoring Programs. Ongoing ambient air
monitoring at fixed locations can provide insight into long-term trends in air quality
and the potential for air pollution to cause adverse health effects among the general
population. Therefore, state and local agencies should be encouraged to either 1)
develop and implement their own ambient air monitoring programs based on proven,
consistent sampling and analysis methods and EPA technical and quality assurance
guidance, or 2) consider long-term participation in the NMP.
• Participate in the National Monitoring Programs year-round. Many of the analyses
presented in the 2014 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 a site is not sampling for a pollutant or pollutant
group identified as particularly hazardous for a given area, the agency responsible for
that site should consider sampling for those compounds.
• Strive to develop standard conventions for interpreting air monitoring data. The lack
of consistent approaches to present and summarize ambient air monitoring data
complicates direct comparisons between different studies. Thought should be given to
the feasibility of establishing standard approaches for analyzing and reporting air
monitoring data for programs with similar objectives.
• Continue to identify and implement improvements to the sampling and analytical
methods. In 2012, two analytical methods were accepted by governing bodies as
approved methods with which to analyze specific pollutants. ERG's hexavalent
chromium method was approved as an ASTM method and ERG's inorganic method
for both TSP and PMio was accepted as a FEM for lead (NAAQS). These approvals
were obtained after various method enhancements that improve the detection and
recovery of these pollutants. Further research is encouraged to identify other 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.
• Revise the pollutants targetedfor sampling based on lessons learned in the field, in
the laboratory, and/or from the annual report. In conjunction with method
improvements, the analytes targeted for monitoring should/needs to be reviewed and
revised periodically based on experience with the collection and analysis methods and
25-26
-------
based on the findings in the annual report. Pollutants initially targeted for ambient
monitoring may no longer be considered problematic based on monitoring results and
could be discontinued. The removal of hexavalent chromium from the target analyte
list for the NATTS program is an example of this. Other pollutants may prove
problematic from a sampling and/or analytical stand point and can be removed from
the target analyte list due to uncertainties associated with its analytical results. In
addition, studies may indicate that one analytical method is better than another at
providing accurate results for a given pollutant. All of these factors should be
considered when determining the pollutants for which to monitor.
Require consistency in sampling and analytical methods. The development of the
NATTS program has shown that there are inconsistencies in collection and analytical
methods that make data comparison difficult across agencies. Requiring agencies to
use specified and accepted measurement methods, consistent with the guidelines
presented in the NATTS TAD, is integral to the identification of trends and
measuring the effectiveness of regulation. Revisions to the NATTS TAD were
approved by EPA in 2016 and implementation is required by participating agencies
by October 31, 2017. It is expected that the revised document will enhance method
consistency.
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.
Consider more rigorous study of the effect of automobile emissions on ambient air
quality using multiple years of data. Because many NMP sites have generated years
of continuous data, a real opportunity exists to evaluate the importance and impact of
automobile emissions on ambient air quality. Suggested areas of study include
additional signature compound assessments and parking lot characterizations.
Develop and/or verify HAP and VOC emissions inventories. State/local/tribal
agencies should use the data collected from NMP sites to develop and validate
emissions inventories, or at the very least, identify and/or verify emissions sources of
concern. Ideally, state/local/tribal agencies would compare the ambient monitoring
results with an emissions inventory for source category completeness. The emissions
inventory could then be used to develop modeled concentrations useful to compare
against ambient monitoring data.
Promulgate ambient air standards for HAPs. Concentrations of many pollutants
sampled during the 2014 program year were greater than risk screening values
developed by various government agencies. One way to reduce the risk to human
health would be to develop standards similar to the NAAQS for pollutants that
frequently exceed published risk screening levels.
25-27
-------
• Incorporate/Update Risk in State Implementation Plans (SIPs). Use risk calculations
to design State Implementation Plans to implement policies that reduce the potential
for human health risk. This would be easier to enforce if ambient standards for certain
HAPs were developed (refer to above recommendation).
25-28
-------
26.0 References
ACE, 2016. 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: 7/18/2016.
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: 1/29/2016.
ASTM, 2013. ASTM, International. ASTMD6209-13 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: 1/29/2016.
AZ DOT, 2016. Arizona Department of Transportation. Transportation Data Management
System website. 2010 and 2011 AADTs.
http://adot.ms2soft.com/tcds/tsearch.asp?loc=Adot&mod= Date Last Accessed: 1/7/2016.
CA DOT, 2014. State of California Department of Transportation, Division of Traffic
Operations. 2014 Traffic Volumes on the California State Highway System. Sacramento,
CA. http://traffic-counts.dot.ca.gov/ Date Last Accessed: 1/7/2016.
Calvert City, 2016. Calvert City - Industry website, http://www.calvertcitv.com/industry.htm
Date Last Accessed: 7/18/2016.
Carbondale, 2016. Carbondale, Colorado. The Town website, http://www.carbondale.com/town
Date Last Accessed: 6/10/2016.
CCRPC, 2016. Chittenden County Regional Planning Commission. Chittenden County Traffic
Counts website. 2011 Data. http://www.ccmpo.us/data/index.php?count=ATR Date Last
Accessed: 1/11/2016.
Census Bureau, 2013a. U.S. Census Bureau. February 2013. List 1. Core Based Statistical Areas
(CBSAs) and Combined Statistical Areas (CSAs).
http://www.census.gov/population/metro/data/def.html Date Last Accessed: 1/11/2016.
Census Bureau, 2013b. U.S. Census Bureau. Metropolitan and Micropolitan Statistical Areas
website, http://www.census.gov/population/metro/ Date Last Accessed: 1/11/2016.
CO DOT, 2014. Colorado Department of Transportation. Online Transportation Information
System (OTIS), Traffic Data Explorer. 2014 data, http://dtdapps.coloradodot.info/otis/
Date Last Accessed: 1/7/2016.
DC DOT, 2014. District Department of Transportation. October 2014. 2013 Traffic Volumes
Map. Washington, D.C. http://ddot.dc.gov/page/traffic-volume-maps Date Last
Accessed: 1/8/2016.
26-1
-------
DC WSA, 2016. District of Columbia Water and Sewer Authority. Tunnel Overview website.
https://www.dcwater.com/workzones/proiects/first street tunnel/summary.cfm Date Last
Accessed: 10/27/2016.
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:
1/11/2016.
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: 1/26/2016.
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:
1/26/2016.
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: 1/26/2016.
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: 1/26/2016.
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.pdfDate Last Accessed:
1/26/2016.
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.pdfDate Last Accessed:
5/6/2016.
EPA, 2009a. U.S. EPA. April 2009. UATMP, NATTS, C SAT AM, PAMS, andNMOC Support,
Contract No. EP-D-09-048. pp. 1-2 through 1-3.
EPA, 2009b. U.S. EPA. April 1, 2009. Technical Assistance Document for the National Air
Toxics Trends Stations Program, Revision 2. Research Triangle Park, NC.
http://www.epa.gov/ttn/amtic/airtox.html Date Last Accessed: 1/11/2016.
26-2
-------
EPA, 2009c. 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: 5/6/2016.
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/ZvPURL.cgi?Dockev=P1009A7C.TXT Date Last Accessed: 3/7/2016.
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: 3/7/2016.
EPA, 201 la. 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: 1/19/2016.
EPA, 2012a. U.S. EPA. June 2012. Ambient Air Monitoring Reference and Equivalent Methods:
Designation of Three New Equivalent Methods. Federal Register Volume 77, Number
106, p. 32632. http://www.gpo.gov/fdsvs/pkg/FR-2012-06-01/html/2012-1335Q.htm Date
Last Accessed: 1/29/2016.
EPA, 2012b. U.S. EPA. September 2012. National Air Toxics Trends Stations (NATTS)
Network Assessment, Revised Draft. Prepared by Eastern Research Group, Inc.
EPA, 2014. U.S. EPA. September 2014. UATMP, PAMS, and NMOC Support, Contract No.
EP-D-14-030.
EPA, 2015a. U.S. EPA. March 4, 2015. 2011 National Emissions Inventory, Version 2.
http://www.epa.gov/ttnchiel/net/2011inventory.html Date Last Accessed: 3/7/2016.
EPA, 2015b. 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: 3/7/2016.
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: 5/13/2016.
EPA, 2015d. 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: 1/6/2017.
EPA, 2016a. U.S. EPA. Air Toxics - National Air Toxics Trends Stations website.
http://www3.epa.gov/ttnamtil/natts.html Date Last Accessed: 1/11/2016.
EPA, 2016b. U.S. EPA. Technology Transfer Network (TTN), Air Quality System (AQS)
website, http://www.epa.gov/aqs Date Last Accessed: 1/11/2016.
26-3
-------
EPA, 2016c. U.S. EPA. What are Hazardous Air Pollutants? website.
https://www.epa.gov/haps/what-are-hazardous-air-pollutants Date Last Accessed:
3/7/2016.
EPA, 2016d 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: 5/10/2016.
EPA, 2017. U.S. EPA. National Emissions Inventory (NEI) website, https://www.epa.gov/air-
emissions-inventories/national-emissions-inventory-nei Date Last Accessed: 2/9/17.
ERG, 2013. Eastern Research Group, Inc. 2013. Support for the EPA National Monitoring
Programs (UATMP, NATTS, C SAT AM, PAMS, and NMOC Support), Quality
Assurance Project Plan, Category 1. Contract No. EP-D-09-048. Morrisville, NC.
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: 1/6/2017.
FL DOT, 2014. Florida Department of Transportation. Florida Traffic Online Mapping
Application (2014). http://www2.dot.state.fl.us/FloridaTrafficOnline/viewer.html Date
Last Accessed: 1/8/2016.
FHWA, 2015. Federal Highway Administration. 2015. Highway Statistics 2015. Glossary
website, https://www.fhwa.dot.gov/policvinformation/statistics/2015/16 glossary.cfm
Date Last Accessed: 2/9/17.
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. p. 4-4. http://www.garfield-countv.com/air-qualitv/ Date Last
Accessed: 6/10/2016.
GCRBD, 2014. Garfield County Road and Bridge Department. Garfield County Road Traffic
Counts. 2014 Data, http://www.garfield-countv.com/road-bridge/countv-roads.aspx Date
Last Accessed: 1/6/2016.
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: 1/6/2017.
IL DOT, 2013/2014. Illinois Department of Transportation. Getting Around Illinois website.
Average Annual Daily Traffic map viewer. 2013 and 2014 data.
http://www.gettingaroundillinois.com/ Date Last Accessed: 1/8/2016.
IN DOT, 2011. Indiana Department of Transportation. Average Daily Traffic and Commercial
Vehicles Interactive map. 2011 Data, http://www.in.gov/indot/2469.htm Date Last
Accessed: 1/8/2016.
26-4
-------
KY, 2016. Kentucky State Parks. Grayson Lake History website.
http://parks.kv.gov/parks/recreationparks/gravson-lake/history.aspx Date Last Accessed:
7/18/2016.
KYTC, 2014. Kentucky Transportation Cabinet. Interactive Statewide Traffic Counts mapping
application. 2012-2014 Data. http://transportation.kv.gov/Planning/Pages/count-
maps.aspx Date Last Accessed: 1/8/2016.
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: 1/11/2016.
Lakes, 2011. Lakes Environmental. June 2011. WRPLOT View 7.0.0.
http://www.weblakes.com/products/wrplot/index.html Date Last Accessed: 11/10/2015.
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: 1/8/2016.
MI DOT, 2014. Michigan Department of Transportation. Traffic Monitoring Information System
(TMIS) website. Annual Average Daily Traffic Report (2014).
http://mdotnetpublic.state.mi.us/tmispublic/Search.aspx Date Last Accessed: 1/8/2016.
MO DOT, 2013. Missouri Department of Transportation, Transportation Planning. 2013 St.
Louis District Traffic Volume and Commercial Vehicle Count Map. Jefferson City, MO.
http://www.modot.org/safetv/trafficvolumemaps.htm Date Last Accessed: 1/8/2016.
NCDC, 2014. National Climatic Data Center. Quality Controlled Local Climatological Data
(QCLCD). 2014 Data, https://www.ncdc.noaa.gov/data-access/land-based-station-
data/land-based-datasets/qualitv-controlled-local-climatological-data-qclcd Date Last
Accessed: 1/6/2017.
NHMU, 2017. 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: 1/3/2017.
NJ DOT, 2014. New Jersey Department of Transportation. Roadway Information and Traffic
Monitoring System Program interactive Traffic Count Reports. 2010, 2012 Data.
http://www.state.ni.us/transportation/refdata/roadwav/traffic counts/ Date Last Accessed:
1/8/2016.
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:
1/8/2016.
26-5
-------
NOAA, 1999. NOAA. March 1998. Automatic Surface Observing System (ASOS) User's
Guide, http://www.nws.noaa.gov/asos/ Date Last Accessed: 10/31/2016.
NYS DOT, 2013. New York State Department of Transportation. Traffic Data Viewer mapping
application. 2013 Data.
https://www.dot.nv.gov/divisions/engineering/applications/traffic-data-viewer Date Last
Accessed: 1/8/2016.
OK DOT, 2014. Oklahoma Department of Transportation, Planning and Research Division. 2014
Annual Average Daily Traffic Oklahoma Highway System maps. Oklahoma City, OK.
http://www.okladot.state.ok.us/maps/aadt/index.htm Date Last Accessed: 1/11/2016.
Peltier, 2016. Peltier Technical Services, Inc. Peltier Tech Charts for Excel 3.0. Downloaded
March 8, 2016. http://peltiertech.com/Utilitv30/ Date Last Accessed: 3/22/2016.
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: 2/9/2017.
RI DOT, 2009. Rhode Island Department of Transportation. December 2009. State Highway
Map of Rhode Island, Traffic Flow Map 2009.
http://www.dot.ri.gov/about/maproom/index.php Date Last Accessed: 1/8/2016.
Seattle, 2016. Seattle Parks and Recreation. Jefferson Park website.
http://www.seattle.gov/parks/find/parks/iefferson-park Date Last Accessed: 12/16/2016.
UT DOT, 2013. Utah Department of Transportation. AADT Traffic Map Updated Using 2013
Data (via Google Earth kmz file).
http://www.udot.utah.gov/main/f?p=100:pg:0::::V.T:.528 Date Last Accessed: 1/8/2016.
UVM, 2015. The University of Vermont. Proctor Maple Research Center Mission website.
http://www.uvm.edu/~pmrc/?Page=mission.htm Date Last Accessed: 11/28/2016.
VA DOT, 2013. Virginia Department of Transportation, Traffic Engineering Division. Average
Daily Traffic Volumes with Vehicle Classification Data on Interstate, Arterial, and
Primary Routes 2013. http://www.virginiadot.org/info/2014 traffic data.asp Date Last
Accessed: 1/11/2016.
WS DOT, 2014. Washington State Department of Transportation. Traffic Volume Map
Application. 2014 Data.
http://www.wsdot.wa.gov/mapsdata/tools/trafficplanningtrends.htm Date Last Accessed:
1/11/2016.
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: 9/1/2016.
26-6
------- |