&EPA
Unfed SUBS
Enviroinwrilfll PiutmBmi
Agancy
2012 National Monitoring Programs Annual
Report (UATMP, NATTS, CSATAM)
September 2014
Final Report
-------
EPA-454/R-14-006a
September 2014
2012 National Monitoring Programs Annual Report (UATMP, NATTS, CSATAM)
By:
Eastern Research Group, Inc.
Morrisville, NC 27560
Prepared for:
Margaret Dougherty and David Shelow
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Contract No. EP-D-09-048
Delivery Orders 20, 25, 30, 31, 33, 36, 37, 40, 42, & 44
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Air Quality Assessment Division
Research Triangle Park, NC 27711
-------
2012 National Monitoring Programs
Annual Report
(UATMP, NATTS, and CSATAM)
Final Report
EPA Contract No. EP-D-09-048
Delivery Orders 20, 25, 30, 31, 33, 36, 37, 40, 42, & 44
Prepared for:
Margaret Dougherty 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
September 2014
-------
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 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 xx
List of Figures xxi
List of Tables xxxix
List of Acronyms xlix
Abstract li
1.0 Introduction 1-1
1.1 Background 1-1
1.2 The Report 1-2
2.0 The 2012 National Monitoring Programs Network 2-1
2.1 Monitoring Locations 2-1
2.2 Analytical Methods and Pollutants Targeted for Monitoring 2-13
2.2.1 VOC and SNMOC Concurrent Sampling and Analytical Methods ... 2-15
2.2.2 Carbonyl Compound Sampling and Analytical Method 2-19
2.2.3 PAH Sampling and Analytical Method 2-20
2.2.4 Metals Sampling and Analytical Method 2-21
2.2.5 Hexavalent Chromium Sampling and Analytical Method 2-22
2.3 Sample Collection Schedules 2-23
2.4 Completeness 2-30
3.0 Summary of the 2012 National Monitoring Programs Data Treatment and
Methods 3-1
3.1 Approach to Data Treatment 3-1
3.2 Human Health Risk and the Pollutants of Interest 3-4
3.3 Noncancer Risk-Based Screening Evaluation Using Minimum Risk Levels 3-7
3.4 Additional Program-Level Analyses of the 2012 National Monitoring
Programs Dataset 3-8
3.4.1 The Effect of Mobile Source Emissions on Spatial Variations 3-8
3.4.2 Variability Analyses 3-9
3.4.3 Greenhouse Gas Assessment 3-11
in
-------
TABLE OF CONTENTS (Continued)
Page
3.5 Additional Site-Specific Analyses 3-11
3.5.1 Site Characterization 3-11
3.5.2 Meteorological Analysis 3-12
3.5.2.1 Back Trajectory Analysis 3-13
3.5.2.2 Wind Rose Analysis 3-14
3.5.3 Preliminary Risk-Based Screening and Pollutants of Interest 3-15
3.5.3.1 Site-Specific Comparison to Program-level Average
Concentrations 3-15
3.5.3.2 Site Trends Analysis 3-16
3.5.3.3 Emission Tracer Analysis 3-17
3.5.3.4 Cancer Risk and Noncancer Hazard Approximations 3-17
3.5.3.5 Risk-Based Emissions Assessment 3-18
4.0 Summary of the 2012 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 4-15
4.2.1 Concentrations of the Pollutants of Interest 4-21
4.2.2 Risk-Based Screening Assessment Using MRLs 4-26
4.3 The Effect of Mobile Sources 4-29
4.3.1 Mobile Source Emissions 4-29
4.3.2 Hydrocarbon Concentrations 4-31
4.3.3 Motor Vehicle Ownership 4-32
4.3.4 Estimated Traffic Volume 4-33
4.3.5 Vehicle Miles Traveled 4-34
4.4 Variability Analysis 4-35
4.4.1 Coefficient of Variation and Inter-site Variability 4-3 5
4.4.2 Quarterly Variability Analysis 4-54
4.5 Greenhouse Gases 4-75
5.0 Sites in Arizona 5-1
5.1 Site Characterization 5-1
5.2 Meteorological Characterization 5-7
5.2.1 Climate Summary 5-7
iv
-------
TABLE OF CONTENTS (Continued)
Page
5.2.2 Meteorological Summary 5-7
5.2.3 Back Trajectory Analysis 5-9
5.2.4 Wind Rose Comparison 5-13
5.3 Pollutants of Interest 5-16
5.4 Concentrations 5-18
5.4.1 2012 Concentration Averages 5-18
5.4.2 Concentration Comparison 5-22
5.4.3 Concentration Trends 5-29
5.5 Additional Risk-Based Screening Evaluations 5-48
5.5.1 Risk-Based Screening Assessment Using MRLs 5-48
5.5.2 Cancer Risk and Noncancer Hazard Approximations 5-49
5.5.3 Risk-Based Emissions Assessment 5-51
5.6 Summary of the 2012 Monitoring Data for PXS Sand SPAZ 5-55
6.0 Sites in California 6-1
6.1 Site Characterization 6-1
6.2 Meteorological Characterization 6-12
6.2.1 Climate Summary 6-12
6.2.2 Meteorological Summary 6-13
6.2.3 Back Trajectory Analysis 6-15
6.2.4 Wind Rose Comparison 6-22
6.3 Pollutants of Interest 6-29
6.4 Concentrations 6-30
6.4.1 2012 Concentration Averages 6-31
6.4.2 Concentration Comparison 6-33
6.4.3 Concentration Trends 6-37
6.5 Additional Risk-Based Screening Evaluations 6-45
6.5.1 Risk-Based Screening Assessment Using MRLs 6-45
6.5.2 Cancer Risk and Noncancer Hazard Approximations 6-45
6.5.3 Risk-Based Emissions Assessment 6-47
6.6 Summary of the 2012 Monitoring Data for the California Monitoring Sites .... 6-53
v
-------
TABLE OF CONTENTS (Continued)
Page
7.0 Sites in Colorado 7-1
7.1 Site Characterization 7-1
7.2 Meteorological Characterization 7-15
7.2.1 Climate Summary 7-15
7.2.2 Meteorological Summary 7-15
7.2.3 Back Trajectory Analysis 7-18
7.2.4 Wind Rose Comparison 7-26
7.3 Pollutants of Interest 7-34
7.4 Concentrations 7-37
7.4.1 2012 Concentration Averages 7-37
7.4.2 Concentration Comparison 7-43
7.4.3 Concentration Trends 7-53
7.5 Additional Risk-Based Screening Evaluations 7-77
7.5.1 Risk-Based Screening Assessment Using MRLs 7-77
7.5.2 Cancer Risk and Noncancer Hazard Approximations 7-77
7.5.3 Risk-Based Emissions Assessment 7-80
7.6 Summary of the 2012 Monitoring Data for the Colorado Monitoring Sites 7-88
8.0 Site in the District of Columbia 8-1
8.1 Site Characterization 8-1
8.2 Meteorological Characterization 8-6
8.2.1 Climate Summary 8-6
8.2.2 Meteorological Summary 8-6
8.2.3 Back Trajectory Analysis 8-8
8.2.4 Wind Rose Comparison 8-10
8.3 Pollutants of Interest 8-12
8.4 Concentrations 8-12
8.4.1 2012 Concentration Averages 8-13
8.4.2 Concentration Comparison 8-14
8.4.3 Concentration Trends 8-15
VI
-------
TABLE OF CONTENTS (Continued)
Page
8.5 Additional Risk-Based Screening Evaluations 8-16
8.5.1 Risk-Based Screening Assessment Using MRLs 8-16
8.5.2 Cancer Risk and Noncancer Hazard Approximations 8-17
8.5.3 Risk-Based Emissions Assessment 8-18
8.6 Summary of the 2012 Monitoring Data for WADC 8-21
9.0 Sites in Florida 9-1
9.1 Site Characterization 9-1
9.2 Meteorological Characterization 9-13
9.2.1 Climate Summary 9-14
9.2.2 Meteorological Summary 9-14
9.2.3 Back Trajectory Analysis 9-17
9.2.4 Wind Rose Comparison 9-25
9.3 Pollutants of Interest 9-33
9.4 Concentrations 9-35
9.4.1 2012 Concentration Averages 9-35
9.4.2 Concentration Comparison 9-38
9.4.3 Concentration Trends 9-42
9.5 Additional Risk-Based Screening Evaluations 9-55
9.5.1 Risk-Based Screening Assessment Using MRLs 9-55
9.5.2 Cancer Risk and Noncancer Hazard Approximations 9-55
9.5.3 Risk-Based Emissions Assessment 9-57
9.6 Summary of the 2012 Monitoring Data for the Florida Monitoring Sites 9-65
10.0 Site in Georgia 10-1
10.1 Site Characterization 10-1
10.2 Meteorological Characterization 10-6
10.2.1 Climate Summary 10-6
10.2.2 Meteorological Summary 10-6
10.2.3 Back Trajectory Analysis 10-8
10.2.4 Wind Rose Comparison 10-10
10.3 Pollutants of Interest 10-12
vii
-------
TABLE OF CONTENTS (Continued)
Page
10.4 Concentrations 10-13
10.4.1 2012 Concentration Averages 10-13
10.4.2 Concentration Comparison 10-14
10.4.3 Concentration Trends 10-15
10.5 Additional Risk-Based Screening Evaluations 10-19
10.5.1 Risk-Based Screening Assessment Using MRLs 10-19
10.5.2 Cancer Risk and Noncancer Hazard Approximations 10-19
10.5.3 Risk-Based Emissions Assessment 10-20
10.6 Summary of the 2012 Monitoring Data for SDGA 10-24
11.0 Sites in Illinois 11-1
11.1 Site Characterization 11-1
11.2 Meteorological Characterization 11-10
11.2.1 Climate Summary 11-10
11.2.2 Meteorological Summary 11-11
11.2.3 Back Trajectory Analysis 11-13
11.2.4 Wind Rose Comparison 11-18
11.3 Pollutants of Interest 11-23
11.4 Concentrations 11-26
11.4.1 2012 Concentration Averages 11-27
11.4.2 Concentration Comparison 11-33
11.4.3 Concentration Trends 11-41
11.5 Additional Risk-Based Screening Evaluations 11-64
11.5.1 Risk-Based Screening Assessment Using MRLs 11-64
11.5.2 Cancer Risk and Noncancer Hazard Approximations 11 -64
11.5.3 Risk-Based Emissions Assessment 11-67
11.6 Summary of the 2012 Monitoring Data for NBIL, SPIL, and ROIL 11-73
12.0 Sites in Indiana 12-1
12.1 Site Characterization 12-1
12.2 Meteorological Characterization 12-8
12.2.1 Climate Summary 12-8
viii
-------
TABLE OF CONTENTS (Continued)
Page
12.2.2 Meteorological Summary 12-9
12.2.3 Back Trajectory Analysis 12-11
12.2.4 Wind Rose Comparison 12-14
12.3 Pollutants of Interest 12-18
12.4 Concentrations 12-19
12.4.1 2012 Concentration Averages 12-19
12.4.2 Concentration Comparison 12-21
12.4.3 Concentration Trends 12-22
12.5 Additional Risk-Based Screening Evaluations 12-28
12.5.1 Risk-Based Screening Assessment Using MRLs 12-28
12.5.2 Cancer Risk and Noncancer Hazard Approximations 12-28
12.5.3 Risk-Based Emissions Assessment 12-29
12.6 Summary of the 2012 Monitoring Data for INDEM and WPIN 12-33
13.0 Sites in Kentucky 13-1
13.1 Site Characterization 13-1
13.2 Meteorological Characterization 13-22
13.2.1 Climate Summary 13-22
13.2.2 Meteorological Summary 13-22
13.2.3 Back Trajectory Analysis 13-26
13.2.4 Wind Rose Comparison 13-38
13.3 Pollutants of Interest 13-52
13.4 Concentrations 13-58
13.4.1 2012 Concentration Averages 13-59
13.4.2 Concentration Comparison 13-69
13.4.3 Concentration Trends 13-76
13.5 Additional Risk-Based Screening Evaluations 13-76
13.5.1 Risk-Based Screening Assessment Using MRLs 13-76
13.5.2 Cancer Risk and Noncancer Hazard Approximations 13-76
13.5.3 Risk-Based Emissions Assessment 13-81
13.6 Summary of the 2012 Monitoring Data for the Kentucky Monitoring Sites... 13-96
IX
-------
TABLE OF CONTENTS (Continued)
Page
14.0 Site in Massachusetts 14-1
14.1 Site Characterization 14-1
14.2 Meteorological Characterization 14-6
14.2.1 Climate Summary 14-6
14.2.2 Meteorological Summary 14-6
14.2.3 Back Trajectory Analysis 14-8
14.2.4 Wind Rose Comparison 14-10
14.3 Pollutants of Interest 14-12
14.4 Concentrations 14-13
14.4.1 2012 Concentration Averages 14-14
14.4.2 Concentration Comparison 14-16
14.4.3 Concentration Trends 14-18
14.5 Additional Risk-Based Screening Evaluations 14-23
14.5.1 Risk-Based Screening Assessment Using MRLs 14-24
14.5.2 Cancer Risk and Noncancer Hazard Approximations 14-24
14.5.3 Risk-Based Emissions Assessment 14-25
14.6 Summary of the 2012 Monitoring Data for BOMA 14-29
15.0 Sites in Michigan 15-1
15.1 Site Characterization 15-1
15.2 Meteorological Characterization 15-9
15.2.1 Climate Summary 15-9
15.2.2 Meteorological Summary 15-9
15.2.3 Back Trajectory Analysis 15-11
15.2.4 Wind Rose Comparison 15-16
15.3 Pollutants of Interest 15-20
15.4 Concentrations 15-22
15.4.1 2012 Concentration Averages 15-23
15.4.2 Concentration Comparison 15-26
15.4.3 Concentration Trends 15-32
-------
TABLE OF CONTENTS (Continued)
Page
15.5 Additional Risk-Based Screening Evaluations 15-44
15.5.1 Risk-Based Screening Assessment Using MRLs 15-44
15.5.2 Cancer Risk and Noncancer Hazard Approximations 15-44
15.5.3 Risk-Based Emissions Assessment 15-46
15.6 Summary of the 2012 Monitoring Data for DEMI, RRMI, and SWMI 15-52
16.0 Site in Minnesota 16-1
16.1 Site Characterization 16-1
16.2 Meteorological Characterization 16-6
16.2.1 Climate Summary 16-6
16.2.2 Meteorological Summary 16-6
16.2.3 Back Trajectory Analysis 16-8
16.2.4 Wind Rose Comparison 16-10
16.3 Pollutants of Interest 16-12
16.4 Concentrations 16-13
16.4.1 2012 Concentration Averages 16-13
16.4.2 Concentration Comparison 16-15
16.4.3 Concentration Trends 16-16
16.5 Additional Risk-Based Screening Evaluations 16-16
16.5.1 Risk-Based Screening Assessment Using MRLs 16-16
16.5.2 Cancer Risk and Noncancer Hazard Approximations 16-17
16.5.3 Risk-Based Emissions Assessment 16-18
16.6 Summary of the 2012 Monitoring Data for STMN 16-21
17.0 Site in Missouri 17-1
17.1 Site Characterization 17-1
17.2 Meteorological Characterization 17-6
17.2.1 Climate Summary 17-6
17.2.2 Meteorological Summary 17-6
17.2.3 Back Trajectory Analysis 17-8
17.2.4 Wind Rose Comparison 17-10
17.3 Pollutants of Interest 17-13
xi
-------
TABLE OF CONTENTS (Continued)
Page
17.4 Concentrations 17-14
17.4.1 2012 Concentration Averages 17-15
17.4.2 Concentration Comparison 17-19
17.4.3 Concentration Trends 17-28
17.5 Additional Risk-Based Screening Evaluations 17-46
17.5.1 Risk-Based Screening Assessment Using MRLs 17-46
17.5.2 Cancer Risk and Noncancer Hazard Approximations 17-46
17.5.3 Risk-Based Emissions Assessment 17-48
17.6 Summary of the 2012 Monitoring Data for S4MO 17-52
18.0 Sites in New Jersey 18-1
18.1 Site Characterization 18-1
18.2 Meteorological Characterization 18-10
18.2.1 Climate Summary 18-10
18.2.2 Meteorological Summary 18-10
18.2.3 Back Trajectory Analysis 18-12
18.2.4 Wind Rose Comparison 18-16
18.3 Pollutants of Interest 18-21
18.4 Concentrations 18-23
18.4.1 2012 Concentration Averages 18-24
18.4.2 Concentration Comparison 18-28
18.4.3 Concentration Trends 18-36
18.5 Additional Risk-Based Screening Evaluations 18-63
18.5.1 Risk-Based Screening Assessment Using MRLs 18-64
18.5.2 Cancer Risk and Noncancer Hazard Approximations 18-64
18.5.3 Risk-Based Emissions Assessment 18-67
18.6 Summary of the 2012 Monitoring Data for the New Jersey Monitoring
Sites 18-73
19.0 Sites in New York 19-1
19.1 Site Characterization 19-1
xn
-------
TABLE OF CONTENTS (Continued)
Page
19.2 Meteorological Characterization 19-10
19.2.1 Climate Summary 19-10
19.2.2 Meteorological Summary 19-11
19.2.3 Back Trajectory Analysis 19-13
19.2.4 Wind Rose Comparison 19-18
19.3 Pollutants of Interest 19-23
19.4 Concentrations 19-25
19.4.1 2012 Concentration Averages 19-25
19.4.2 Concentration Comparison 19-28
19.4.3 Concentration Trends 19-30
19.5 Additional Risk-Based Screening Evaluations 19-30
19.5.1 Risk-Based Screening Assessment Using MRLs 19-30
19.5.2 Cancer Risk and Noncancer Hazard Approximations 19-31
19.5.3 Risk-Based Emissions Assessment 19-32
19.6 Summary of the 2012 Monitoring Data for BXNY, MONY, and ROCH 19-38
20.0 Sites in Oklahoma 20-1
20.1 Site Characterization 20-1
20.2 Meteorological Characterization 20-14
20.2.1 Climate Summary 20-14
20.2.2 Meteorological Summary 20-14
20.2.3 Back Trajectory Analysis 20-17
20.2.4 Wind Rose Comparison 20-24
20.3 Pollutants of Interest 20-31
20.4 Concentrations 20-35
20.4.1 2012 Concentration Averages 20-35
20.4.2 Concentration Comparison 20-43
20.4.3 Concentration Trends 20-56
20.5 Additional Risk-Based Screening Evaluations 20-69
20.5.1 Risk-Based Screening Assessment Using MRLs 20-70
20.5.2 Cancer Risk and Noncancer Hazard Approximations 20-70
20.5.3 Risk-Based Emissions Assessment 20-74
Xlll
-------
TABLE OF CONTENTS (Continued)
Page
20.6 Summary of the 2012 Monitoring Data for the Oklahoma Monitoring Sites.. 20-82
21.0 Site in Rhode Island 21-1
21.1 Site Characterization 21-1
21.2 Meteorological Characterization 21-6
21.2.1 Climate Summary 21-6
21.2.2 Meteorological Summary 21-6
21.2.3 Back Trajectory Analysis 21-8
21.2.4 Wind Rose Comparison 21-10
21.3 Pollutants of Interest 21-12
21.4 Concentrations 21-12
21.4.1 2012 Concentration Averages 21-13
21.4.2 Concentration Comparison 21-14
21.4.3 Concentration Trends 21-15
21.5 Additional Risk-Based Screening Evaluations 21-17
21.5.1 Risk-Based Screening Assessment Using MRLs 21-18
21.5.2 Cancer Risk and Noncancer Hazard Approximations 21-18
21.5.3 Risk-Based Emissions Assessment 21-19
21.6 Summary of the 2012 Monitoring Data for PRRI 21-23
22.0 Site in South Carolina 22-1
22.1 Site Characterization 22-1
22.2 Meteorological Characterization 22-5
22.2.1 Climate Summary 22-6
22.2.2 Meteorological Summary 22-6
22.2.3 Back Trajectory Analysis 22-8
22.2.4 Wind Rose Comparison 22-10
22.3 Pollutants of Interest 22-12
22.4 Concentrations 22-12
22.4.1 2012 Concentration Averages 22-13
22.4.2 Concentration Comparison 22-14
22.4.3 Concentration Trends 22-15
xiv
-------
TABLE OF CONTENTS (Continued)
Page
22.5 Additional Risk-Based Screening Evaluations 22-16
22.5.1 Risk-Based Screening Assessment Using MRLs 22-16
22.5.2 Cancer Risk and Noncancer Hazard Approximations 22-16
22.5.3 Risk-Based Emissions Assessment 22-17
22.6 Summary of the 2012 Monitoring Data for CHSC 22-21
23.0 Site in South Dakota 23-1
23.1 Site Characterization 23-1
23.2 Meteorological Characterization 23-6
23.2.1 Climate Summary 23-6
23.2.2 Meteorological Summary 23-6
23.2.3 Back Trajectory Analysis 23-8
23.2.4 Wind Rose Comparison 23-10
23.3 Pollutants of Interest 23-12
23.4 Concentrations 23-13
23.4.1 2012 Concentration Averages 23-13
23.4.2 Concentration Comparison 23-15
23.4.3 Concentration Trends 23-18
23.5 Additional Risk-Based Screening Evaluations 23-25
23.5.1 Risk-Based Screening Assessment Using MRLs 23-25
23.5.2 Cancer Risk and Noncancer Hazard Approximations 23-25
23.5.3 Risk-Based Emissions Assessment 23-26
23.6 Summary of the 2012 Monitoring Data for SSSD 23-30
24.0 Sites in Texas 24-1
24.1 Site Characterization 24-1
24.2 Meteorological Characterization 24-8
24.2.1 Climate Summary 24-8
24.2.2 Meteorological Summary 24-9
24.2.3 Back Trajectory Analysis 24-11
24.2.4 Wind Rose Comparison 24-14
24.3 Pollutants of Interest 24-17
xv
-------
TABLE OF CONTENTS (Continued)
Page
24.4 Concentrations 24-19
24.4.1 2012 Concentration Averages 24-19
24.4.2 Concentration Comparison 24-20
24.4.3 Concentration Trends 24-21
24.5 Additional Risk-Based Screening Evaluations 24-21
24.5.1 Risk-Based Screening Assessment Using MRLs 24-22
24.5.2 Cancer Risk and Noncancer Hazard Approximations 24-22
24.5.3 Risk-Based Emissions Assessment 24-23
24.6 Summary of the 2012 Monitoring Data for CAMS 35 and CAMS 85 24-27
25.0 Site in Utah 25-1
25.1 Site Characterization 25-1
25.2 Meteorological Characterization 25-6
25.2.1 Climate Summary 25-6
25.2.2 Meteorological Summary 25-6
25.2.3 Back Trajectory Analysis 25-8
25.2.4 Wind Rose Comparison 25-10
25.3 Pollutants of Interest 25-12
25.4 Concentrations 25-13
25.4.1 2012 Concentration Averages 25-13
25.4.2 Concentration Comparison 25-17
25.4.3 Concentration Trends 25-24
25.5 Additional Risk-Based Screening Evaluations 25-37
25.5.1 Risk-Based Screening Assessment Using MRLs 25-37
25.5.2 Cancer Risk and Noncancer Hazard Approximations 25-37
25.5.3 Risk-Based Emissions Assessment 25-39
25.6 Summary of the 2012 Monitoring Data for BTUT 25-43
26.0 Sites in Vermont 26-1
26.1 Site Characterization 26-1
26.2 Meteorological Characterization 26-10
26.2.1 Climate Summary 26-10
xvi
-------
TABLE OF CONTENTS (Continued)
Page
26.2.2 Meteorological Summary 26-11
26.2.3 Back Trajectory Analysis 26-11
26.2.4 Wind Rose Comparison 26-17
26.3 Pollutants of Interest 26-22
26.4 Concentrations 26-24
26.4.1 2012 Concentration Averages 26-24
26.4.2 Concentration Comparison 26-29
26.4.3 Concentration Trends 26-36
26.5 Additional Risk-Based Screening Evaluations 26-38
26.5.1 Risk-Based Screening Assessment Using MRLs 26-38
26.5.2 Cancer Risk and Noncancer Hazard Approximations 26-38
26.5.3 Risk-Based Emissions Assessment 26-41
26.6 Summary of the 2012 Monitoring Data for the Vermont Monitoring Sites .... 26-47
27.0 Site in Virginia 27-1
27.1 Site Characterization 27-1
27.2 Meteorological Characterization 27-6
27.2.1 Climate Summary 27-6
27.2.2 Meteorological Summary 27-6
27.2.3 Back Trajectory Analysis 27-8
27.2.4 Wind Rose Comparison 27-10
27.3 Pollutants of Interest 27-12
27.4 Concentrations 27-13
27.4.1 2012 Concentration Averages 27-13
27.4.2 Concentration Comparison 27-14
27.4.3 Concentration Trends 27-16
27.5 Additional Risk-Based Screening Evaluations 27-16
27.5.1 Risk-Based Screening Assessment Using MRLs 27-16
27.5.2 Cancer Risk and Noncancer Hazard Approximations 27-16
27.5.3 Risk-Based Emissions Assessment 27-17
27.6 Summary of the 2012 Monitoring Data for RIVA 27-21
xvn
-------
TABLE OF CONTENTS (Continued)
Page
28.0 Site in Washington 28-1
28.1 Site Characterization 28-1
28.2 Meteorological Characterization 28-6
28.2.1 Climate Summary 28-6
28.2.2 Meteorological Summary 28-6
28.2.3 Back Trajectory Analysis 28-8
28.2.4 Wind Rose Comparison 28-10
28.3 Pollutants of Interest 28-12
28.4 Concentrations 28-13
28.4.1 2012 Concentration Averages 28-13
28.4.2 Concentration Comparison 28-16
28.4.3 Concentration Trends 28-22
28.5 Additional Risk-Based Screening Evaluations 28-33
28.5.1 Risk-Based Screening Assessment Using MRLs 28-33
28.5.2 Cancer Risk and Noncancer Hazard Approximations 28-33
28.5.3 Risk-Based Emissions Assessment 28-35
28.6 Summary of the 2012 Monitoring Data for SEWA 28-39
29.0 Sites in Wisconsin 29-1
29.1 Site Characterization 29-1
29.2 Meteorological Characterization 29-9
29.2.1 Climate Summary 29-9
29.2.2 Meteorological Summary 29-9
29.2.3 Back Trajectory Analysis 29-11
29.2.4 Wind Rose Comparison 29-14
29.3 Pollutants of Interest 29-18
29.4 Concentrations 29-19
29.4.1 2012 Concentration Averages 29-20
29.4.2 Concentration Comparison 29-21
29.4.3 Concentration Trends 29-22
xvin
-------
TABLE OF CONTENTS (Continued)
Page
29.5 Additional Risk-Based Screening Evaluations 29-22
29.5.1 Risk-Based Screening Assessment Using MRLs 29-22
29.5.2 Cancer Risk and Noncancer Hazard Approximations 29-22
29.5.3 Risk-Based Emissions Assessment 29-23
29.6 Summary of the 2012 Monitoring Data for HOWI and MIWI 29-27
30.0 Data Quality 30-1
30.1 Completeness 30-1
30.2 Method Precision 30-2
30.2.1 VOC Method Precision 30-4
30.2.2 SNMOC Method Precision 30-12
30.2.3 Carbonyl Compound Method Precision 30-15
30.2.4 PAH Method Precision 30-17
30.2.5 Metals Method Precision 30-18
30.2.6 Hexavalent Chromium Method Precision 30-20
30.3 Analytical Precision 30-21
30.3.1 VOC Analytical Precision 30-22
30.3.2 SNMOC Analytical Precision 30-30
30.3.3 Carbonyl Compound Analytical Precision 30-33
30.3.4 PAH Analytical Precision 30-36
30.3.5 Metals Analytical Precision 30-39
30.3.6 Hexavalent Chromium Analytical Precision 30-41
30.4 Accuracy 30-43
31.0 Results, Conclusions, and Recommendations 31-1
31.1 Summary of Results 31-1
31.1.1 National-level Results Summary 31-1
31.1.2 State-level Results Summary 31-2
31.1.3 Composite Site-level Results Summary 31-22
31.1.4 Data Quality Results Summary 31-24
31.2 Conclusions 31-25
31.3 Recommendations 31-26
32.0 References 32-1
xix
-------
List of Appendices
Appendix A AQS Site Descriptions for the 2012 NATTS, UATMP, and CSATAM
Monitoring Sites
Appendix B Range of Method Detection Limits (MDLs)
Appendix C 2012 VOC Raw Data
Appendix D 2012 SNMOC Raw Data
Appendix E 2012 Carbonyl Compounds Raw Data
Appendix F 2012 PAH Raw Data
Appendix G 2012 Metals Raw Data
Appendix H 2012 Hexavalent Chromium Raw Data
Appendix I Summary of Invalidated 2012 Samples
Appendix J 2012 Summary Statistics for VOC Monitoring
Appendix K 2012 Summary Statistics for SNMOC Monitoring
Appendix L 2012 Summary Statistics for Carbonyl Compounds Monitoring
Appendix M 2012 Summary Statistics for PAH Monitoring
Appendix N 2012 Summary Statistics for Metals Monitoring
Appendix O 2012 Summary Statistics for Hexavalent Chromium Monitoring
Appendix P Glossary of Terms
Appendix Q Risk Factors Used Throughout the 2012 NMP Report
xx
-------
LIST OF FIGURES
Page
2-1 Locations of the 2012 National Monitoring Programs Monitoring Sites 2-3
4-la Coefficient of Variation Analysis of Acenaphthene Across 20 Sites 4-39
4-lb Inter-Site Variability for Acenaphthene 4-39
4-2a Coefficient of Variation Analysis of Acetaldehyde Across 28 Sites 4-40
4-2b Inter-Site Variability for Acetaldehyde 4-40
4-3a Coefficient of Variation Analysis of Arsenic Across 19 Sites 4-41
4-3b Inter-Site Variability for Arsenic 4-41
4-4a Coefficient of Variation Analysis of Benzene Across 22 Sites 4-42
4-4b Inter-Site Variability for Benzene 4-42
4-5a Coefficient of Variation Analysis of 1,3-Butadiene Across 22 Sites 4-43
4-5b Inter-Site Variability for 1,3-Butadiene 4-43
4-6a Coefficient of Variation Analysis of Carbon Tetrachloride Across 22 Sites 4-44
4-6b Inter-Site Variability for Carbon Tetrachloride 4-44
4-7a Coefficient of Variation Analysis ofp-Dichlorobenzene Across 22 Sites 4-45
4-7b Inter-Site Variability for/>-Dichlorobenzene 4-45
4-8a Coefficient of Variation Analysis of 1,2-Dichloroethane Across 22 Sites 4-46
4-8b Inter-Site Variability for 1,2-Dichloroethane 4-46
4-9a Coefficient of Variation Analysis of Ethylbenzene Across 22 Sites 4-47
4-9b Inter-Site Variability for Ethylbenzene 4-47
4-10a Coefficient of Variation Analysis of Fluorene Across 20 Sites 4-48
4-10b Inter-Site Variability for Fluorene 4-48
4-1 la Coefficient of Variation Analysis of Formaldehyde Across 28 Sites 4-49
4-llb Inter-Site Variability for Formaldehyde 4-49
4-12a Coefficient of Variation Analysis of Hexachloro-1,3-Butadiene Across 22 Sites 4-50
4-12b Inter-Site Variability for Hexachloro-l,3-butadiene 4-50
4-13a Coefficient of Variation Analysis of Manganese Across 19 Sites 4-51
4-13b Inter-Site Variability for Manganese 4-51
4-14a Coefficient of Variation Analysis of Naphthalene Across 20 Sites 4-52
4-14b Inter-Site Variability for Naphthalene 4-52
4-15a Coefficient of Variation Analysis of Nickel Across 19 Sites 4-53
4-15b Inter-Site Variability for Nickel 4-53
4-16 Comparison of Average Quarterly Acenaphthene Concentrations 4-57
4-17 Comparison of Average Quarterly Acetaldehyde Concentrations 4-58
4-18a Comparison of Average Quarterly Arsenic (PMi0) Concentrations 4-59
4-18b Comparison of Average Quarterly Arsenic (TSP) Concentrations 4-60
4-19 Comparison of Average Quarterly Benzene Concentrations 4-61
4-20 Comparison of Average Quarterly 1,3-Butadiene Concentrations 4-62
4-21 Comparison of Average Quarterly Carbon Tetrachloride Concentrations 4-63
4-22 Comparison of Average Quarterly />-Dichlorobenzene Concentrations 4-64
4-23 Comparison of Average Quarterly 1,2-Dichloroethane Concentrations 4-65
4-24 Comparison of Average Quarterly Ethylbenzene Concentrations 4-66
4-25 Comparison of Average Quarterly Fluorene Concentrations 4-67
4-26 Comparison of Average Quarterly Formaldehyde Concentrations 4-68
4-27 Comparison of Average Quarterly Hexachloro-l,3-butadiene Concentrations 4-69
xxi
-------
LIST OF FIGURES (Continued)
Page
4-28a Comparison of Average Quarterly Manganese (PMio) Concentrations 4-70
4-28b Comparison of Average Quarterly Manganese (TSP) Concentrations 4-71
4-29 Comparison of Average Quarterly Naphthalene Concentrations 4-72
4-30a Comparison of Average Quarterly Nickel (PMio) Concentrations 4-73
4-30b Comparison of Average Quarterly Nickel (TSP) Concentrations 4-74
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 Composite Back Trajectory Map for PXSS 5-10
5-5 Back Trajectory Cluster Map for PXSS 5-10
5-6 Composite Back Trajectory Map for SPAZ 5-11
5-7 Back Trajectory Cluster Map for SPAZ 5-11
5-8 Wind Roses for the Phoenix Sky Harbor International Airport Weather Station
near PXSS 5-14
5-9 Wind Roses for the Phoenix Sky Harbor International Airport Weather Station
near SPAZ 5-15
5-10 Program vs. Site-Specific Average Acetaldehyde Concentration 5-22
5-11 Program vs. Site-Specific Average Arsenic (PMio) Concentration 5-22
5-12 Program vs. Site-Specific Average Benzene Concentrations 5-23
5-13 Program vs. Site-Specific Average 1,3-Butadiene Concentrations 5-23
5-14 Program vs. Site-Specific Average Carbon Tetrachloride Concentration 5-24
5-15 Program vs. Site-Specific Average/>-Dichlorobenzene Concentration 5-24
5-16 Program vs. Site-Specific Average 1,2-Dichloroethane Concentration 5-25
5-17 Program vs. Site-Specific Average Ethylbenzene Concentration 5-25
5-18 Program vs. Site-Specific Average Formaldehyde Concentration 5-25
5-19 Program vs. Site-Specific Average Manganese (PMio) Concentration 5-26
5-20 Program vs. Site-Specific Average Naphthalene Concentration 5-26
5-21 Program vs. Site-Specific Average Nickel (PMio) Concentration 5-26
5-22 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at PXSS 5-30
5-23 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at PXSS 5-31
5-24 Yearly Statistical Metrics for Benzene Concentrations Measured at PXSS 5-32
5-25 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at PXSS 5-33
5-26 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
PXSS 5-34
5-27 Yearly Statistical Metrics for/»-Dichlorobenzene Concentrations Measured at
PXSS 5-35
5-28 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
PXSS 5-36
5-29 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at PXSS 5-37
5-30 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at PXSS 5-38
5-31 Yearly Statistical Metrics for Manganese (PMio) Concentrations Measured at
PXSS 5-39
5-32 Yearly Statistical Metrics for Naphthalene Concentrations Measured at PXSS 5-40
5-33 Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at PXSS 5-41
5-34 Yearly Statistical Metrics for Benzene Concentrations Measured at SPAZ 5-42
xxii
-------
LIST OF FIGURES (Continued)
Page
5-35 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at SPAZ 5-43
5-36 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
SPAZ 5-44
5-37 Yearly Statistical Metrics for/»-Dichlorobenzene Concentrations Measured at
SPAZ 5-45
5-38 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
SPAZ 5-46
5-39 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at SPAZ 5-47
6-1 Los Angeles, California (CELA) Monitoring Site 6-2
6-2 Long Beach, California (LBHCA) Monitoring Site 6-3
6-3 NEI Point Sources Located Within 10 Miles of CELA and LBHCA 6-4
6-4 Rubidoux, California (RUCA) Monitoring Site 6-5
6-5 NEI Point Sources Located Within 10 Miles of RUCA 6-6
6-6 San Jose, California (SJJCA) Monitoring Site 6-7
6-7 NEI Point Sources Located Within 10 Miles of SJJCA 6-8
6-8 Composite Back Trajectory Map for CELA 6-16
6-9 Back Trajectory Cluster Map for CELA 6-16
6-10 Composite Back Trajectory Map for LBHCA 6-17
6-11 Back Trajectory Cluster Map for LBHCA 6-17
6-12 Composite Back Trajectory Map for RUCA 6-18
6-13 Back Trajectory Cluster Map for RUCA 6-18
6-14 Composite Back Trajectory Map for SJJCA 6-19
6-15 Back Trajectory Cluster Map for SJJCA 6-19
6-16 Wind Roses for the Downtown Los Angeles/USC Campus Weather Station
near CELA 6-23
6-17 Wind Roses for the Long Beach/Daugherty Field Airport Weather Station
near LBHCA 6-24
6-18 Wind Roses for the Riverside Municipal Airport Weather Station near RUCA 6-25
6-19 Wind Roses for the San Jose International Airport Weather Station near SJJCA 6-26
6-20 Program vs. Site-Specific Average Arsenic (PMio) Concentration 6-34
6-21 Program vs. Site-Specific Average Fluorene Concentrations 6-34
6-22 Program vs. Site-Specific Average Manganese (PMio) Concentration 6-34
6-23 Program vs. Site-Specific Average Naphthalene Concentration 6-35
6-24 Program vs. Site-Specific Average Nickel (PMio) Concentrations 6-35
6-25 Yearly Statistical Metrics for Fluorene Concentrations Measured at CELA 6-38
6-26 Yearly Statistical Metrics for Naphthalene Concentrations Measured at CELA 6-39
6-27 Yearly Statistical Metrics for Naphthalene Concentrations Measured at RUCA 6-40
6-28 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at SJJCA 6-41
6-29 Yearly Statistical Metrics for Manganese (PMio) Concentrations Measured at
SJJCA 6-42
6-30 Yearly Statistical Metrics for Naphthalene Concentrations Measured at SJJCA 6-43
6-31 Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at SJJCA 6-44
7-1 Grand Junction, Colorado (GPCO) Monitoring Site 7-2
7-2 NEI Point Sources Located Within 10 Miles of GPCO 7-3
xxiii
-------
LIST OF FIGURES (Continued)
Page
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 Composite Back Trajectory Map for GPCO 7-20
7-11 Back Trajectory Cluster Map for GPCO 7-20
7-12 Composite Back Trajectory Map for BMCO 7-21
7-13 Back Trajectory Cluster Map for BMCO 7-21
7-14 Composite Back Trajectory Map for BRCO 7-22
7-15 Back Trajectory Cluster Map for BRCO 7-22
7-16 Composite Back Trajectory Map for PACO 7-23
7-17 Back Trajectory Cluster Map for PACO 7-23
7-18 Composite Back Trajectory Map for RICO 7-24
7-19 Back Trajectory Cluster Map for RICO 7-24
7-20 Composite Back Trajectory Map for RFCO 7-25
7-21 Wind Roses for the Walker Field Airport Weather Station near GPCO 7-27
7-22 Wind Roses for the Garfield County Regional Airport Weather Station near BMCO.. 7-28
7-23 Wind Roses for the Garfield County Regional Airport Weather Station near BRCO... 7-29
7-24 Wind Roses for the Garfield County Regional Airport Weather Station near PACO... 7-30
7-25 Wind Roses for the Garfield County Regional Airport Weather Station near RICO.... 7-31
7-26 Wind Roses for the Aspen-Pitkin County Airport Weather Station near RFCO 7-32
7-27 Program vs. Site-Specific Average Acenaphthene Concentration 7-43
7-28 Program vs. Site-Specific Average Acetaldehyde Concentration 7-44
7-29a Program vs. Site-Specific Average Benzene (Method TO-15) Concentrations 7-44
7-29b Program vs. Site-Specific Average Benzene (SNMOC) Concentrations 7-45
7-30a Program vs. Site-Specific Average 1,3-Butadiene (Method TO-15) Concentration 7-45
7-30b Program vs. Site-Specific Average 1,3-Butadiene (SNMOC) Concentration 7-46
7-31 Program vs. Site-Specific Average Carbon Tetrachloride Concentrations 7-46
7-32 Program vs. Site-Specific Average/>-Dichlorobenzene Concentration 7-46
7-33 Program vs. Site-Specific Average 1,2-Dichloroethane Concentration 7-47
7-34 Program vs. Site-Specific Average Dichloromethane Concentration 7-47
7-35 Program vs. Site-Specific Average Ethylbenzene Concentration 7-47
7-36 Program vs. Site-Specific Average Fluorene Concentration 7-48
7-37 Program vs. Site-Specific Average Formaldehyde Concentration 7-48
7-38 Program vs. Site-Specific Average Hexachloro-l,3-Butadiene Concentration 7-49
7-39 Program vs. Site-Specific Average Naphthalene Concentration 7-49
7-40 Yearly Statistical Metrics for Acenaphthene Concentrations Measured at GPCO 7-54
7-41 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at GPCO 7-55
7-42 Yearly Statistical Metrics for Benzene Concentrations Measured at GPCO 7-56
7-43 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at GPCO 7-57
7-44 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
GPCO 7-58
XXIV
-------
LIST OF FIGURES (Continued)
Page
7-45 Yearly Statistical Metrics for/»-Dichlorobenzene Concentrations Measured at
GPCO 7-59
7-46 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
GPCO 7-60
7-47 Yearly Statistical Metrics for Dichloromethane Concentrations Measured at GPCO... 7-61
7-48 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at GPCO 7-62
7-49 Yearly Statistical Metrics for Fluorene Concentrations Measured at GPCO 7-63
7-50 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at GPCO 7-64
7-51 Yearly Statistical Metrics for Hexachloro-1,3-Butadiene Concentrations Measured
at GPCO 7-65
7-52 Yearly Statistical Metrics for Naphthalene Concentrations Measured at GPCO 7-66
7-53 Yearly Statistical Metrics for Benzene Concentrations Measured at BRCO 7-67
7-54 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at BRCO 7-68
7-55 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at PACO 7-69
7-56 Yearly Statistical Metrics for Benzene Concentrations Measured at PACO 7-70
7-57 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at PACO 7-71
7-58 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at PACO 7-72
7-59 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at RICO 7-73
7-60 Yearly Statistical Metrics for Benzene Concentrations Measured at RICO 7-74
7-61 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at RICO 7-75
7-62 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at RICO 7-76
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 Composite Back Trajectory Map for WADC 8-9
8-4 Back Trajectory Cluster Map for WADC 8-9
8-5 Wind Roses for the Ronald Reagan Washington National Airport Weather Station
near WADC 8-11
8-6 Program vs. Site-Specific Average Naphthalene Concentration 8-14
8-7 Yearly Statistical Metrics for Naphthalene Concentrations Measured at WADC 8-15
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 Composite Back Trajectory Map for AZFL 9-18
9-10 Back Trajectory Cluster Map for AZFL 9-18
9-11 Composite Back Trajectory Map for SKFL 9-19
9-12 Back Trajectory Cluster Map for SKFL 9-19
9-13 Composite Back Trajectory Map for SYFL 9-20
9-14 Back Trajectory Cluster Map for SYFL 9-20
9-15 Composite Back Trajectory Map for ORFL 9-21
XXV
-------
LIST OF FIGURES (Continued)
Page
9-16 Back Trajectory Cluster Map for ORFL 9-21
9-17 Composite Back Trajectory Map for PAFL 9-22
9-18 Back Trajectory Cluster Map for PAFL 9-22
9-19 Wind Roses for the St. Petersburg/Whitted Airport Weather Station near AZFL 9-26
9-20 Wind Roses for the St. Petersburg/Clearwater International Airport Weather
Station near SKFL 9-27
9-21 Wind Roses for the Plant City Municipal Airport Weather Station near SYFL 9-28
9-22 Wind Roses for the Orlando Executive Airport Weather Station near ORFL 9-29
9-23 Wind Roses for the Orlando Executive Airport Weather Station near PAFL 9-30
9-24 Program vs. Site-Specific Average Acetaldehyde Concentrations 9-39
9-25 Program vs. Site-Specific Average Arsenic (PMio) Concentration 9-39
9-26 Program vs. Site-Specific Average Formaldehyde Concentrations 9-40
9-27 Program vs. Site-Specific Average Manganese (PMio) Concentration 9-40
9-28 Program vs. Site-Specific Average Naphthalene Concentrations 9-41
9-29 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at AZFL 9-43
9-30 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at AZFL 9-44
9-31 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SKFL 9-45
9-32 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at SKFL 9-46
9-33 Yearly Statistical Metrics for Naphthalene Concentrations Measured at SKFL 9-47
9-34 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SYFL 9-48
9-35 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at SYFL 9-49
9-36 Yearly Statistical Metrics for Naphthalene Concentrations Measured at SYFL 9-50
9-37 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at ORFL 9-51
9-38 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at ORFL 9-52
9-39 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at PAFL 9-53
9-40 Yearly Statistical Metrics for Manganese (PMio) Concentrations Measured at PAFL. 9-54
10-1 Decatur, Georgia (SDGA) Monitoring Site 10-2
10-2 NEI Point Sources Located Within 10 Miles of SDGA 10-3
10-3 Composite Back Trajectory Map for SDGA 10-8
10-4 Back Trajectory Cluster Map for SDGA 10-9
10-5 Wind Roses for the Hartsfield International Airport Weather Station near SDGA 10-11
10-6 Program vs. Site-Specific Average Hexavalent Chromium Concentration 10-15
10-7 Yearly Statistical Metrics for Benzo(a)pyrene Concentrations Measured at SDGA... 10-16
10-8 Yearly Statistical Metrics for Hexavalent Chromium Concentrations Measured
at SDGA 10-17
10-9 Yearly Statistical Metrics for Naphthalene Concentrations Measured at SDGA 10-18
11-1 Northbrook, Illinois (NBIL) Monitoring Site 11-2
11-2 Schiller Park, Illinois (SPIL) Monitoring Site 11-3
11-3 NEI Point Sources Located Within 10 Miles of NBIL and SPIL 11-4
11-4 Roxana, Illinois (ROIL) Monitoring Site 11-5
11-5 NEI Point Sources Located Within 10 Miles of ROIL 11-6
11-6 Composite Back Trajectory Map for NBIL 11-14
11-7 Back Trajectory Cluster Map for NBIL 11-14
11-8 Composite Back Trajectory Map for SPIL 11-15
xxvi
-------
LIST OF FIGURES (Continued)
Page
11-9 Back Trajectory Cluster Map for SPIL 11-15
11-10 Composite Back Trajectory Map for ROIL 11-16
11-11 Back Trajectory Cluster Map for ROIL 11-16
11-12 Wind Roses for the Palwaukee Municipal Airport Weather Station nearNBIL 11-20
11-13 Wind Roses for the O'Hare International Airport Weather Station near SPIL 11-21
11-14 Wind Roses for the Lambert/St. Louis International Airport Weather Station near
ROIL 11-22
11-15 Program vs. Site-Specific Average Acenaphthene Concentrations 11-33
11-16 Program vs. Site-Specific Average Acetaldehyde Concentrations 11-33
11-17 Program vs. Site-Specific Average Arsenic (PMio) Concentration 11-34
11-18 Program vs. Site-Specific Average Benzene Concentrations 11-34
11-19 Program vs. Site-Specific Average 1,3-Butadiene Concentrations 11-34
11 -20 Program vs. Site-Specific Average Carbon Tetrachloride Concentrations 11-35
11-21 Program vs. Site-Specific Average/>-Dichlorobenzene Concentrations 11-35
11-22 Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations 11-36
11-23 Program vs. Site-Specific Average Ethylbenzene Concentrations 11-36
11-24 Program vs. Site-Specific Average Fluoranthene Concentrations 11-36
11-25 Program vs. Site-Specific Average Fluorene Concentrations 11-37
11-26 Program vs. Site-Specific Average Formaldehyde Concentrations 11-37
11-27 Program vs. Site-Specific Average Manganese (PMio) Concentration 11-37
11 -28 Program vs. Site-Specific Average Naphthalene Concentration 11-38
11 -29 Program vs. Site-Specific Average Trichloroethylene Concentrations 11-38
11-30 Yearly Statistical Metrics for Acenaphthene Concentrations Measured at NBIL 11-42
11-31 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at NBIL 11-43
11-32 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at NBIL 11-44
11-33 Yearly Statistical Metrics for Benzene Concentrations Measured at NBIL 11-45
11-34 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at NBIL 11-46
11-35 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
NBIL 11-47
11-36 Yearly Statistical Metrics for/?-Dichlorobenzene Concentrations Measured at
NBIL 11-48
11-37 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
NBIL 11-49
11-38 Yearly Statistical Metrics for Fluoranthene Concentrations Measured at NBIL 11-50
11-39 Yearly Statistical Metrics for Fluorene Concentrations Measured at NBIL 11-51
11-40 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at NBIL 11-52
11-41 Yearly Statistical Metrics for Manganese (PMio) Concentrations Measured at
NBIL 11-53
11 -42 Yearly Statistical Metrics for Naphthalene Concentrations Measured at NBIL 11 -54
11-43 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SPIL 11-55
11 -44 Yearly Statistical Metrics for Benzene Concentrations Measured at SPIL 11-56
11-45 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at SPIL 11-57
11-46 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
SPIL 11-58
11-47 Yearly Statistical Metrics for/»-Dichlorobenzene Concentrations Measured at
SPIL 11-59
xxvii
-------
LIST OF FIGURES (Continued)
Page
11-48 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
SPIL 11-60
11 -49 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at SPIL 11-61
11-50 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at SPIL 11 -62
11-51 Yearly Statistical Metrics for Trichloroethylene Concentrations Measured at SPIL .. 11-63
12-1 Gary, Indiana (INDEM) Monitoring Site 12-2
12-2 NEI Point Sources Located Within 10 Miles of INDEM 12-3
12-3 Indianapolis, Indiana (WPIN) Monitoring Site 12-4
12-4 NEI Point Sources Located Within 10 Miles of WPIN 12-5
12-5 Composite Back Trajectory Map for INDEM 12-12
12-6 Back Trajectory Cluster Map for INDEM 12-12
12-7 Composite Back Trajectory Map for WPIN 12-13
12-8 Back Trajectory Cluster Map for WPIN 12-13
12-9 Wind Roses for the Lansing Municipal Airport Weather Station near INDEM 12-15
12-10 Wind Roses for the Indianapolis International Airport Weather Station near WPIN.. 12-16
12-11 Program vs. Site-Specific Average Acetaldehyde Concentration 12-21
12-12 Program vs. Site-Specific Average Formaldehyde Concentration 12-22
12-13 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at INDEM.... 12-23
12-14 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
INDEM 12-24
12-15 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at WPIN 12-26
12-16 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at WPIN 12-27
13-1 Ashland, Kentucky (ASKY) Monitoring Site 13-2
13-2 Ashland, Kentucky (ASKY-M) Monitoring Site 13-3
13-3 NEI Point Sources Located Within 10 Miles of ASKY and ASKY-M 13-4
13-4 Grayson, Kentucky (GLKY) Monitoring Site 13-5
13-5 NEI Point Sources Located Within 10 Miles of GLKY 13-6
13-6 Baskett, Kentucky (BAKY) Monitoring Site 13-7
13-7 NEI Point Sources Located Within 10 Miles of BAKY 13-8
13-8 Calvert City, Kentucky (ATKY) Monitoring Site 13-9
13-9 Smithland, Kentucky (BLKY) Monitoring Site 13-10
13-10 Calvert City, Kentucky (CCKY) Monitoring Site 13-11
13-11 Calvert City, Kentucky (LAKY) Monitoring Site 13-12
13-12 Calvert City, Kentucky (TVKY) Monitoring Site 13-13
13-13 NEI Point Sources Located Within 10 Miles of ATKY, BLKY, CCKY, LAKY,
and TVKY 13-14
13-14 Lexington, Kentucky (LEKY) Monitoring Site 13-15
13-15 NEI Point Sources Located Within 10 Miles of LEKY 13-16
13-16 Composite Back Trajectory Map for ASKY 13-27
13-17 Composite Back Trajectory Map for ASKY-M 13-27
13-18 Back Trajectory Cluster Map for ASKY-M 13-28
13-19 Composite Back Trajectory Map for GLKY 13-28
13-20 Back Trajectory Cluster Map for GLKY 13-29
13-21 Composite Back Trajectory Map for BAKY 13-29
xxviii
-------
LIST OF FIGURES (Continued)
Page
13-22 Composite Back Trajectory Map for BAKY 13-30
13-23 Composite Back Trajectory Map for ATKY 13-30
13-24 Composite Back Trajectory Map for BLKY 13-31
13-25 Composite Back Trajectory Map for CCKY 13-31
13-26 Back Trajectory Cluster Map for CCKY 13-32
13-27 Composite Back Trajectory Map for LAKY 13-32
13-28 Composite Back Trajectory Map for TVKY 13-33
13-29 Composite Back Trajectory Map for LEKY 13-33
13-30 Back Trajectory Cluster Map for LEKY 13-34
13-31 Wind Roses for the Tri-State/M. J. Ferguson Field Airport Weather Station near
ASKY 13-39
13-32 Wind Roses for the Tri-State/MJ. Ferguson Field Airport Weather Station near
ASKY-M 13-40
13-33 Wind Roses for the Tri-State/MJ. Ferguson Field Airport Weather Station near
GLKY 13-41
13-34 Wind Roses for the Evansville Regional Airport Weather Station near BAKY 13-43
13-35 Wind Roses for the Barkley Regional Airport Weather Station near ATKY 13-45
13-36 Wind Roses for the Barkley Regional Airport Weather Station near BLKY 13-46
13-37 Wind Roses for the Barkley Regional Airport Weather Station near CCKY 13-47
13-38 Wind Roses for the Barkley Regional Airport Weather Station near LAKY 13-48
13-39 Wind Roses for the Barkley Regional Airport Weather Station near TVKY 13-49
13-40 Wind Roses for the Blue Grass Airport Weather Station near LEKY 13-51
13-41 Program vs. Site-Specific Average Acetaldehyde Concentration 13-69
13-42 Program vs. Site-Specific Average Arsenic (PMio) Concentration 13-70
13-43 Program vs. Site-Specific Average Benzene Concentration 13-70
13-44 Program vs. Site-Specific Average 1,3-Butadiene Concentration 13-71
13-45 Program vs. Site-Specific Average Cadmium Concentration 13-71
13-46 Program vs. Site-Specific Average Carbon Tetrachloride Concentration 13-71
13-47 Program vs. Site-Specific Average 1,2-Dichloroethane Concentration 13-72
13-48 Program vs. Site-Specific Average Formaldehyde Concentration 13-72
13-49 Program vs. Site-Specific Average Lead (PMio) Concentration 13-72
13-50 Program vs. Site-Specific Average Manganese (PMio) Concentration 13-73
13-51 Program vs. Site-Specific Average Nickel (PMio) Concentration 13-73
14-1 Boston, Massachusetts (BOMA) Monitoring Site 14-2
14-2 NEI Point Sources Located Within 10 Miles of BOMA 14-3
14-3 Composite Back Trajectory Map for BOMA 14-9
14-4 Back Trajectory Cluster Map for BOMA 14-10
14-5 Wind Roses for the Logan International Airport Weather Station near BOMA 14-11
14-6 Program vs. Site-Specific Average Arsenic (PMio) Concentration 14-16
14-7 Program vs. Site-Specific Average Hexavalent Chromium Concentration 14-16
14-8 Program vs. Site-Specific Average Manganese (PMio) Concentration 14-17
14-9 Program vs. Site-Specific Average Naphthalene Concentration 14-17
14-10 Program vs. Site-Specific Average Nickel (PMio) Concentration 14-17
14-11 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at BOMA... 14-19
XXIX
-------
LIST OF FIGURES (Continued)
Page
14-12 Yearly Statistical Metrics for Hexavalent Chromium Concentrations Measured
atBOMA 14-20
14-13 Yearly Statistical Metrics for Manganese (PMio) Concentrations Measured
atBOMA 14-21
14-14 Yearly Statistical Metrics for Naphthalene Concentrations Measured at BOMA 14-22
14-15 Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at BOMA .... 14-23
15-1 Dearborn, Michigan (DEMI) Monitoring Site 15-2
15-2 River Rouge, Michigan (RRMI) Monitoring Site 15-3
15-3 Detroit, Michigan (SWMI) Monitoring Site 15-4
15-4 NEI Point Sources Located Within 10 Miles of DEMI, RRMI, and SWMI 15-5
15-5 Composite Back Trajectory Map for DEMI 15-12
15-6 Back Trajectory Cluster Map for DEMI 15-12
15-7 Composite Back Trajectory Map for RRMI 15-13
15-8 Back Trajectory Cluster Map for RRMI 15-13
15-9 Composite Back Trajectory Map for SWMI 15-14
15-10 Back Trajectory Cluster Map for SWMI 15-14
15-11 Wind Roses for the Detroit City Airport Weather Station near DEMI 15-17
15-12 Wind Roses for the Detroit City Airport Weather Station near RRMI 15-18
15-13 Wind Roses for the Detroit City Airport Weather Station near SWMI 15-19
15-14 Program vs. Site-Specific Average Acenaphthene Concentrations 15-27
15-15 Program vs. Site-Specific Average Acetaldehyde Concentrations 15-27
15-16 Program vs. Site-Specific Average Benzene Concentration 15-27
15-17 Program vs. Site-Specific Average 1,3-Butadiene Concentration 15-28
15-18 Program vs. Site-Specific Average Carbon Tetrachloride Concentration 15-28
15-19 Program vs. Site-Specific Average 1,2-Dichloroethane Concentration 15-28
15-20 Program vs. Site-Specific Average Ethylbenzene Concentration 15-29
15-21 Program vs. Site-Specific Average Fluorene Concentration 15-29
15-22 Program vs. Site-Specific Average Formaldehyde Concentration 15-29
15-23 Program vs. Site-Specific Average Hexavalent Chromium Concentrations 15-30
15-24 Program vs. Site-Specific Average Naphthalene Concentration 15-30
15-25 Yearly Statistical Metrics for Acenaphthene Concentrations Measured at DEMI 15-33
15-26 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at DEMI 15-34
15-27 Yearly Statistical Metrics for Benzene Concentrations Measured at DEMI 15-35
15-28 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at DEMI 15-36
15-29 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
DEMI 15-37
15-30 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
DEMI 15-38
15-31 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at DEMI 15-39
15-32 Yearly Statistical Metrics for Fluorene Concentrations Measured at DEMI 15-40
15-33 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at DEMI 15-41
15-34 Yearly Statistical Metrics for Hexavalent Chromium Concentrations Measured at
DEMI 15-42
15-35 Yearly Statistical Metrics for Naphthalene Concentrations Measured at DEMI 15-43
XXX
-------
LIST OF FIGURES (Continued)
Page
16-1 St. Cloud, Minnesota (STMN) Monitoring Site 16-2
16-2 NEI Point Sources Located Within 10 Miles of STMN 16-3
16-3 Composite Back Trajectory Map for STMN 16-9
16-4 Back Trajectory Cluster Map for STMN 16-9
16-5 Wind Roses for the St. Cloud Regional Airport Weather Station near STMN 16-11
16-6 Program vs. Site-Specific Average Hexavalent Chromium Concentration 16-15
17-1 St. Louis, Missouri (S4MO) Monitoring Site 17-2
17-2 NEI Point Sources Located Within 10 Miles of S4MO 17-3
17-3 Composite Back Trajectory Map for S4MO 17-9
17-4 Back Trajectory Cluster Map for S4MO 17-9
17-5 Wind Roses for the St. Louis Downtown Airport Weather Station near S4MO 17-12
17-6 Program vs. Site-Specific Average Acenaphthene Concentrations 17-19
17-7 Program vs. Site-Specific Average Acetaldehyde Concentrations 17-20
17-8 Program vs. Site-Specific Average Arsenic (PMio) Concentrations 17-20
17-9 Program vs. Site-Specific Average Benzene Concentrations 17-20
17-10 Program vs. Site-Specific Average 1,3-Butadiene Concentrations 17-21
17-11 Program vs. Site-Specific Average Cadmium (PMio) Concentrations 17-21
17-12 Program vs. Site-Specific Average Carbon Tetrachloride Concentrations 17-21
17-13 Program vs. Site-Specific Average/>-Dichlorobenzene Concentrations 17-22
17-14 Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations 17-22
17-15 Program vs. Site-Specific Average Ethylbenzene Concentrations 17-22
17-16 Program vs. Site-Specific Average Fluorene Concentrations 17-23
17-17 Program vs. Site-Specific Average Formaldehyde Concentrations 17-23
17-18 Program vs. Site-Specific Average Hexachloro-1,3-Butadiene Concentrations 17-23
17-19 Program vs. Site-Specific Average Lead (PMio) Concentrations 17-24
17-20 Program vs. Site-Specific Average Manganese (PMio) Concentrations 17-24
17-21 Program vs. Site-Specific Average Naphthalene Concentrations 17-24
17-22 Program vs. Site-Specific Average Nickel (PMio) Concentrations 17-25
17-23 Yearly Statistical Metrics for Acenaphthene Concentrations Measured at S4MO 17-29
17-24 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at S4MO 17-30
17-25 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at S4MO.... 17-31
17-26 Yearly Statistical Metrics for Benzene Concentrations Measured at S4MO 17-32
17-27 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at S4MO 17-33
17-28 Yearly Statistical Metrics for Cadmium (PMio) Concentrations Measured at
S4MO 17-34
17-29 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
S4MO 17-35
17-30 Yearly Statistical Metrics for/?-Dichlorobenzene Concentrations Measured at
S4MO 17-36
17-31 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
S4MO 17-37
17-32 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at S4MO 17-38
17-33 Yearly Statistical Metrics for Fluorene Concentrations Measured at S4MO 17-39
17-34 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at S4MO 17-40
xxxi
-------
LIST OF FIGURES (Continued)
Page
17-35 Yearly Statistical Metrics for Hexachloro-1,3-Butadiene Concentrations Measured
atS4MO 17-41
17-36 Yearly Statistical Metrics for Lead (PMio) Concentrations Measured at S4MO 17-42
17-37 Yearly Statistical Metrics for Manganese (PMio) Concentrations Measured at
S4MO 17-43
17-38 Yearly Statistical Metrics for Naphthalene Concentrations Measured at S4MO 17-44
17-39 Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at S4MO 17-45
18-1 Chester, New Jersey (CHNJ) Monitoring Site 18-2
18-2 NEI Point Sources Located Within 10 Miles of CHNJ 18-3
18-3 Elizabeth, New Jersey (ELNJ) Monitoring Site 18-4
18-4 North Brunswick, New Jersey (NBNJ) Monitoring Site 18-5
18-5 NEI Point Sources Located Within 10 Miles of ELNJ and NBNJ 18-6
18-6 Composite Back Trajectory Map for CHNJ 18-13
18-7 Back Trajectory Cluster Map for CHNJ 18-13
18-8 Composite Back Trajectory Map for ELNJ 18-14
18-9 Back Trajectory Cluster Map for ELNJ 18-14
18-10 Composite Back Trajectory Map for NBNJ 18-15
18-11 Back Trajectory Cluster Map for NBNJ 18-15
18-12 Wind Roses for the Summerville-Somerset Airport Weather Station near CHNJ 18-18
18-13 Wind Roses for the Newark International Airport Weather Station near ELNJ 18-19
18-14 Wind Roses for the Summerville-Somerset Airport Weather Station near NBNJ 18-20
18-15 Program vs. Site-Specific Average Acetaldehyde Concentrations 18-29
18-16 Program vs. Site-Specific Average Benzene Concentrations 18-29
18-17 Program vs. Site-Specific Average 1,3-Butadiene Concentrations 18-30
18-18 Program vs. Site-Specific Average Carbon Tetrachloride Concentrations 18-30
18-19 Program vs. Site-Specific Average/>-Dichlorobenzene Concentrations 18-31
18-20 Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations 18-31
18-21 Program vs. Site-Specific Average Ethylbenzene Concentrations 18-32
18-22 Program vs. Site-Specific Average Formaldehyde Concentrations 18-32
18-23 Program vs. Site-Specific Average Hexachloro-1,3-Butadiene Concentrations 18-33
18-24 Program vs. Site-Specific Average Propionaldehyde Concentrations 18-33
18-25 Program vs. Site-Specific Average 1,1,2,2-Tetrachloroethane Concentrations 18-33
18-26 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at CHNJ 18-37
18-27 Yearly Statistical Metrics for Benzene Concentrations Measured at CHNJ 18-38
18-28 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at CHNJ 18-39
18-29 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
CHNJ 18-40
18-30 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
CHNJ 18-41
18-31 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at CHNJ 18-42
18-32 Yearly Statistical Metrics for Hexachloro-1,3-Butadiene Concentrations Measured
at CHNJ 18-43
18-33 Yearly Statistical Metrics for 1,1,2,2-Tetrachloroethane Concentrations Measured
at CHNJ 18-44
18-34 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at ELNJ 18-45
xxxii
-------
LIST OF FIGURES (Continued)
Page
18-35 Yearly Statistical Metrics for Benzene Concentrations Measured at ELNJ 18-46
18-36 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at ELNJ 18-47
18-37 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
ELNJ 18-48
18-38 Yearly Statistical Metrics for/»-Dichlorobenzene Concentrations Measured at
ELNJ 18-49
18-39 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
ELNJ 18-50
18-40 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at ELNJ 18-51
18-41 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at ELNJ 18-52
18-42 Yearly Statistical Metrics for Propionaldehyde Concentrations Measured at ELNJ... 18-53
18-43 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at NBNJ 18-54
18-44 Yearly Statistical Metrics for Benzene Concentrations Measured at NBNJ 18-55
18-45 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at NBNJ 18-56
18-46 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
NBNJ 18-57
18-47 Yearly Statistical Metrics for/»-Dichlorobenzene Concentrations Measured at
NBNJ 18-59
18-48 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
NBNJ 18-60
18-49 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at NBNJ 18-61
18-50 Yearly Statistical Metrics for Hexachloro-l,3-Butadiene Concentrations Measured
at NBNJ 18-62
18-51 Yearly Statistical Metrics for 1,1,2,2-Tetrachloroethane Concentrations Measured
at NBNJ 18-63
19-1 New York City, New York (BXNY) Monitoring Site 19-2
19-2 New York City, New York (MONY) Monitoring Site 19-3
19-3 NEI Point Sources Located Within 10 Miles of BXNY and MONY 19-4
19-4 Rochester, New York (ROCH) Monitoring Site 19-5
19-5 NEI Point Sources Located Within 10 Miles of ROCH 19-6
19-6 Composite Back Trajectory Map for BXNY 19-14
19-7 Back Trajectory Cluster Map for BXNY 19-14
19-8 Composite Back Trajectory Map for MONY 19-15
19-9 Back Trajectory Cluster Map for MONY 19-15
19-10 Composite Back Trajectory Map for ROCH 19-16
19-11 Back Trajectory Cluster Map for ROCH 19-16
19-12 Wind Roses for the La Guardia Airport Weather Station near BXNY 19-19
19-13 Wind Roses for the La Guardia Airport Weather Station near MONY 19-20
19-14 Wind Roses for the Greater Rochester International Airport Weather Station near
ROCH 19-21
19-15 Program vs. Site-Specific Average Acenaphthene Concentrations 19-28
19-16 Program vs. Site-Specific Average Fluoranthene Concentrations 19-28
19-17 Program vs. Site-Specific Average Fluorene Concentrations 19-29
19-18 Program vs. Site-Specific Average Naphthalene Concentrations 19-29
xxxin
-------
LIST OF FIGURES (Continued)
Page
20-1 Tulsa, Oklahoma (TOOK) Monitoring Site 20-2
20-2 Tulsa, Oklahoma (TMOK) Monitoring Site 20-3
20-3 NEI Point Sources Located Within 10 Miles of TMOK and TOOK 20-4
20-4 Pryor Creek, Oklahoma (PROK) Monitoring Site 20-5
20-5 NEI Point Sources Located Within 10 Miles of PROK 20-6
20-6 Oklahoma City, Oklahoma (ADOK) Monitoring Site 20-7
20-7 Oklahoma City, Oklahoma (OCOK) Monitoring Site 20-8
20-8 NEI Point Sources Located Within 10 Miles of ADOK and OCOK 20-9
20-9 Composite Back Trajectory Map for TOOK 20-18
20-10 Back Trajectory Cluster Map for TOOK 20-18
20-11 Composite Back Trajectory Map for TMOK 20-19
20-12 Back Trajectory Cluster Map for TMOK 20-19
20-13 Composite Back Trajectory Map for PROK 20-20
20-14 Back Trajectory Cluster Map for PROK 20-20
20-15 Composite Back Trajectory Map for ADOK 20-21
20-16 Back Trajectory Cluster Map for ADOK 20-21
20-17 Composite Back Trajectory Map for OCOK 20-22
20-18 Back Trajectory Cluster Map for OCOK 20-22
20-19 Wind Roses for the Richard Lloyd Jones Jr. Airport Weather Station near TOOK.... 20-25
20-20 Wind Roses for the Tulsa International Airport Weather Station near TMOK 20-26
20-21 Wind Roses for the Claremore Regional Airport Weather Station near PROK 20-27
20-22 Wind Roses for the Tinker Air Force Base Airport Weather Station near ADOK 20-28
20-23 Wind Roses for the Wiley Post Airport Weather Station near OCOK 20-29
20-24 Program vs. Site-Specific Average Acetaldehyde Concentration 20-43
20-25 Program vs. Site-Specific Average Arsenic (TSP) Concentration 20-44
20-26 Program vs. Site-Specific Average Benzene Concentration 20-45
20-27 Program vs. Site-Specific Average 1,3-Butadiene Concentration 20-46
20-28 Program vs. Site-Specific Average Carbon Tetrachloride Concentration 20-47
20-29 Program vs. Site-Specific Average/>-Dichlorobenzene Concentration 20-48
20-30 Program vs. Site-Specific Average 1,2-Dichloroethane Concentration 20-49
20-31 Program vs. Site-Specific Average Ethylbenzene Concentration 20-50
20-32 Program vs. Site-Specific Average Formaldehyde Concentration 20-51
20-33 Program vs. Site-Specific Average Hexachloro-1,3-Butadiene Concentration 20-52
20-34 Program vs. Site-Specific Average Manganese (TSP) Concentration 20-52
20-35 Program vs. Site-Specific Average Nickel (TSP) Concentration 20-53
20-36 Program vs. Site-Specific Average Propionaldehyde Concentration 20-53
20-37 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at TOOK 20-57
20-38 Yearly Statistical Metrics for Arsenic (TSP) Concentrations Measured at TOOK 20-58
20-39 Yearly Statistical Metrics for Benzene Concentrations Measured at TOOK 20-59
20-40 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at TOOK 20-60
20-41 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
TOOK 20-61
20-42 Yearly Statistical Metrics for/?-Dichlorobenzene Concentrations Measured at
TOOK 20-62
20-43 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
TOOK 20-63
xxxiv
-------
LIST OF FIGURES (Continued)
Page
20-44 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at TOOK 20-64
20-45 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at TOOK 20-65
20-46 Yearly Statistical Metrics for Hexachloro-1,3-Butadiene Concentrations Measured
at TOOK 20-66
20-47 Yearly Statistical Metrics for Manganese (TSP) Concentrations Measured at
TOOK 20-67
20-48 Yearly Statistical Metrics for Nickel (TSP) Concentrations Measured at TOOK 20-68
20-49 Yearly Statistical Metrics for Propionaldehyde Concentrations Measured at TOOK. 20-69
21-1 Providence, Rhode Island (PRRI) Monitoring Site 21-2
21-2 NEI Point Sources Located Within 10 Miles of PRRI 21-3
21-3 Composite Back Trajectory Map for PRRI 21-8
21-4 Back Trajectory Cluster Map for PRRI 21-9
21-5 Wind Roses for the T.F. Green State Airport Weather Station near PRRI 21-11
21-6 Program vs. Site-Specific Average Hexavalent Chromium Concentration 21-14
21-7 Program vs. Site-Specific Average Naphthalene Concentration 21-15
21-8 Yearly Statistical Metrics for Hexavalent Chromium Concentrations Measured
at PRRI 21-16
21-9 Yearly Statistical Metrics for Naphthalene Concentrations Measured at PRRI 21-17
22-1 Chesterfield, South Carolina (CHSC) Monitoring Site 22-2
22-2 NEI Point Sources Located Within 10 Miles of CHSC 22-3
22-3 Composite Back Trajectory Map for CHSC 22-9
22-4 Back Trajectory Cluster Map for CHSC 22-9
22-5 Wind Roses for Monroe Airport Weather Station near CHSC 22-11
22-6 Program vs. Site-Specific Average Naphthalene Concentrations 22-14
22-7 Yearly Statistical Metrics for Naphthalene Concentrations Measured at CHSC 22-15
23-1 Sioux Falls, South Dakota (SSSD) Monitoring Site 23-2
23-2 NEI Point Sources Located Within 10 Miles of SSSD 23-3
23-3 Composite Back Trajectory Map for SSSD 23-9
23-4 Back Trajectory Cluster Map for SSSD 23-9
23-5 Wind Roses for the Joe Foss Field Airport Weather Station near SSSD 23-11
23-6 Program vs. Site-Specific Average Acetaldehyde Concentration 23-15
23-7 Program vs. Site-Specific Average Benzene Concentration 23-16
23-8 Program vs. Site-Specific Average 1,3-Butadiene Concentration 23-16
23-9 Program vs. Site-Specific Average Carbon Tetrachloride Concentration 23-16
23-10 Program vs. Site-Specific Average 1,2-Dichloroethane Concentration 23-17
23-11 Program vs. Site-Specific Average Formaldehyde Concentration 23-17
23-12 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SSSD 23-19
23-13 Yearly Statistical Metrics for Benzene Concentrations Measured at SSSD 23-20
23-14 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at SSSD 23-21
23-15 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
SSSD 23-22
23-16 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
SSSD 23-23
XXXV
-------
LIST OF FIGURES (Continued)
Page
23-17 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at SSSD 23-24
24-1 Deer Park, Texas (CAMS 3 5) Monitoring Site 24-2
24-2 NEI Point Sources Located Within 10 Miles of CAMS 35 24-3
24-3 Karnack, Texas (CAMS 85) Monitoring Site 24-4
24-4 NEI Point Sources Located Within 10 Miles of CAMS 85 24-5
24-5 Composite Back Trajectory Map for CAMS 35 24-12
24-6 Back Trajectory Cluster Map for CAMS 35 24-12
24-7 Composite Back Trajectory Map for CAMS 85 24-13
24-8 Back Trajectory Cluster Map for CAS 85 24-13
24-9 Wind Roses for the William P. Hobby Airport Weather Station near CAMS 35 24-15
24-10 Wind Roses for the Shreveport Regional Airport Weather Station near CAMS 85.... 24-16
24-11 Program vs. Site-Specific Average Hexavalent Chromium Concentration 24-21
25-1 Bountiful, Utah (BTUT) Monitoring Site 25-2
25-2 NEI Point Sources Located Within 10 Miles of BTUT 25-3
25-3 Composite Back Trajectory Map for BTUT 25-8
25-4 Back Trajectory Cluster Map for BTUT 25-9
25-5 Wind Roses for the Salt Lake City International Airport Weather Station near
BTUT 25-11
25-6 Program vs. Site-Specific Average Acetaldehyde Concentration 25-17
25-7 Program vs. Site-Specific Average Arsenic (PMio) Concentration 25-18
25-8 Program vs. Site-Specific Average Benzene Concentrations 25-18
25-9 Program vs. Site-Specific Average 1,3-Butadiene Concentration 25-18
25-10 Program vs. Site-Specific Average Carbon Tetrachloride Concentration 25-19
25-11 Program vs. Site-Specific Average 1,2-Dichloroethane Concentration 25-19
25-12 Program vs. Site-Specific Average Dichloromethane Concentration 25-19
25-13 Program vs. Site-Specific Average Ethylbenzene Concentration 25-20
25-14 Program vs. Site-Specific Average Formaldehyde Concentration 25-20
25-15 Program vs. Site-Specific Average Manganese (PMio) Concentration 25-20
25-16 Program vs. Site-Specific Average Naphthalene Concentration 25-21
25-17 Program vs. Site-Specific Average Nickel (PMio) Concentration 25-21
25-18 Program vs. Site-Specific Average Propionaldehyde Concentration 25-21
25-19 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at BTUT 25-25
25-20 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at BTUT.... 25-26
25-21 Yearly Statistical Metrics for Benzene Concentrations Measured at BTUT 25-27
25-22 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at BTUT 25-28
25-23 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
BTUT 25-29
25-24 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
BTUT 25-30
25-25 Yearly Statistical Metrics for Dichloromethane Concentrations Measured at
BTUT 25-31
25-26 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at BTUT 25-32
25-27 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at BTUT 25-33
xxxvi
-------
LIST OF FIGURES (Continued)
Page
25-28 Yearly Statistical Metrics for Manganese (PMio) Concentrations Measured at
BTUT 25-34
25-29 Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at BTUT 25-35
25-30 Yearly Statistical Metrics for Propionaldehyde Concentrations Measured at BTUT.. 25-36
26-1 Burlington, Vermont (BURVT) Monitoring Site 26-2
26-2 Underbill, Vermont (UNVT) Monitoring Site 26-3
26-3 NEI Point Sources Located Within 10 Miles of BURVT and UNVT 26-4
26-4 Rutland, Vermont (RUVT) Monitoring Site 26-5
26-5 NEI Point Sources Located Within 10 Miles of RUVT 26-6
26-6 Composite Back Trajectory Map for BURVT 26-13
26-7 Back Trajectory Cluster Map for BURVT 26-13
26-8 Composite Back Trajectory Map for RUVT 26-14
26-9 Back Trajectory Cluster Map for RUVT 26-14
26-10 Composite Back Trajectory Map for UNVT 26-15
26-11 Back Trajectory Cluster Map for UNVT 26-15
26-12 Wind Roses for the Burlington International Airport Weather Station near
BURVT 26-18
26-13 Wind Roses for the Rutland State Airport Weather Station near RUVT 26-19
26-14 Wind Roses for the Morrisville-Stowe State Airport Weather Station near UNVT.... 26-20
26-15 Program vs. Site-Specific Average Arsenic (PMio) Concentration 26-29
26-16 Program vs. Site-Specific Average Benzene Concentration 26-30
26-17 Program vs. Site-Specific Average 1,3-Butadiene Concentration 26-30
26-18 Program vs. Site-Specific Average Carbon Tetrachloride Concentration 26-31
26-19 Program vs. Site-Specific Average 1,2-Dibromoethane Concentration 26-31
26-20 Program vs. Site-Specific Average/>-Dichlorobenzene Concentration 26-32
26-21 Program vs. Site-Specific Average 1,2-Dichloroethane Concentration 26-32
26-22 Program vs. Site-Specific Average Ethylbenzene Concentration 26-33
26-23 Program vs. Site-Specific Average Hexachloro-l,2-Butadiene Concentration 26-33
26-24 Program vs. Site-Specific Average 1,1,2,2-Tetrachloroethane Concentration 26-33
26-25 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at UNVT ... 26-37
27-1 East Highland Park, Virginia (RIVA) Monitoring Site 27-2
27-2 NEI Point Sources Located Within 10 Miles of RIV A 27-3
27-3 Composite Back Trajectory Map for RIV A 27-9
27-4 Back Trajectory Cluster Map for RIV A 27-9
27-5 Wind Roses for the Richmond International Airport Weather Station near RIVA 27-11
27-6 Program vs. Site-Specific Average Fluorene Concentration 27-15
27-7 Program vs. Site-Specific Average Naphthalene Concentration 27-15
28-1 Seattle, Washington (SEWA) Monitoring Site 28-2
28-2 NEI Point Sources Located Within 10 Miles of SEWA 28-3
28-3 Composite Back Trajectory Map for SEWA 28-8
28-4 Back Trajectory Cluster Map for SEWA 28-9
28-5 Wind Roses for the Boeing Field/King County International Airport Weather
Station near SEWA 28-11
xxxvii
-------
LIST OF FIGURES (Continued)
Page
28-6 Program vs. Site-Specific Average Acetaldehyde Concentration 28-17
28-7 Program vs. Site-Specific Average Arsenic (PMio) Concentration 28-17
28-8 Program vs. Site-Specific Average Benzene Concentration 28-17
28-9 Program vs. Site-Specific Average 1,3-Butadiene Concentration 28-18
28-10 Program vs. Site-Specific Average Carbon Tetrachloride Concentration 28-18
28-11 Program vs. Site-Specific Average 1,2-Dichloroethane Concentration 28-18
28-12 Program vs. Site-Specific Average Formaldehyde Concentration 28-19
28-13 Program vs. Site-Specific Average Manganese (PMio) Concentration 28-19
28-14 Program vs. Site-Specific Average Naphthalene Concentration 28-19
28-15 Program vs. Site-Specific Average Nickel (PMio) Concentration 28-20
28-16 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SEWA 28-23
28-17 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at
SEWA 28-24
28-18 Yearly Statistical Metrics for Benzene Concentrations Measured at SEWA 28-25
28-19 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at SEWA 28-26
28-20 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
SEWA 28-27
28-21 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
SEWA 28-28
28-22 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at SEWA 28-29
28-23 Yearly Statistical Metrics for Manganese (PMio) Concentrations Measured
at SEWA 28-30
28-24 Yearly Statistical Metrics for Naphthalene Concentrations Measured at SEWA 28-31
28-25 Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at SEWA 28-32
29-1 Horicon, Wisconsin (HOWI) Monitoring Site 29-2
29-2 NEI Point Sources Located Within 10 Miles of HOWI 29-3
29-3 Milwaukee, Wisconsin (MIWI) Monitoring Site 29-4
29-4 NEI Point Sources Located Within 10 Miles of MIWI 29-5
29-5 Composite Back Trajectory Map for HOWI 29-12
29-6 Back Trajectory Cluster Map for HOWI 29-12
29-7 Composite Back Trajectory Map for MWIW 29-13
29-8 Back Trajectory Cluster Map for MIWI 29-13
29-9 Wind Roses for the Dodge County Airport Weather Station near HOWI 29-16
29-10 Wind Roses for the Lawrence J. Timmerman Airport Weather Station near MIWI... 29-17
29-11 Program vs. Site-Specific Average Hexavalent Chromium Concentration 29-21
xxxvin
-------
LIST OF TABLES
Page
1-1 Organization of the 2012 National Monitoring Programs Report 1-4
2-1 2012 National Monitoring Programs Sites and Past Program Participation 2-4
2-2 Site Characterizing Information for the 2012 National Monitoring Programs Sites 2-8
2-3 2012 VOC Method Detect!on Limits 2-17
2-4 2012 SNMOC Method Detect!on Limits 2-18
2-5 2012Carbonyl Compound Method Detection Limits 2-20
2-6 2012 PAH Method Detect!on Limits 2-21
2-7 2012 Metals Method Detection Limits 2-22
2-8 2012 Hexavalent Chromium Method Detect!on Limit 2-23
2-9 2012 Sampling Schedules and Completeness Rates 2-25
2-10 Method Completeness Rates for 2012 2-32
3-1 Overview and Lay out of Data Presented 3-1
3-2 NATTS MQO Core Analytes 3-7
3-3 POM Groups for PAHs 3-20
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 4-16
4-8 Site-Specific Risk-Based Screening Comparison 4-19
4-9 Annual Average Concentration Comparison of the VOC 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 Pollutants of Interest 4-23
4-12 Annual Average Concentration Comparison of the Metals Pollutants of Interest 4-24
4-13 Comparison of Maximum Concentrations vs. ATSDRMRLs 4-27
4-14 Summary of Mobile Source Information by Monitoring Site 4-30
4-15 Greenhouse Gases Measured by Method TO-15 4-76
5-1 Geographical Information for the Arizona Monitoring Sites 5-5
5-2 Population, Motor Vehicle, and Traffic Information for the Arizona Monitoring
Sites 5-6
5-3 Average Meteorological Conditions near the Arizona Monitoring Sites 5-8
5-4 Risk-Based Screening Results for the Arizona Monitoring Sites 5-16
5-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Arizona Monitoring Sites 5-19
5-6 Risk Approximations for the Arizona Monitoring Sites 5-50
5-7 Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Arizona Monitoring Sites 5-52
XXXIX
-------
LIST OF TABLES (Continued)
Page
5-8 Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Arizona Monitoring
Sites 5-53
6-1 Geographical Information for the California Monitoring Sites 6-9
6-2 Population, Motor Vehicle, and Traffic Information for the California Monitoring
Sites 6-12
6-3 Average Meteorological Conditions near the California Monitoring Sites 6-14
6-4 Risk-Based Screening Results for the California Monitoring Sites 6-29
6-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
California Monitoring Sites 6-31
6-6 Risk Approximations for the California Monitoring Sites 6-46
6-7 Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the California Monitoring Sites 6-48
6-8 Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the California Monitoring
Sites 6-50
7-1 Geographical Information for the Colorado Monitoring Sites 7-11
7-2 Population, Motor Vehicle, and Traffic Information for the Colorado Monitoring
Sites 7-14
7-3 Average Meteorological Conditions near the Colorado Monitoring Sites 7-16
7-4 Risk-Based Screening Results for the Colorado Monitoring Sites 7-35
7-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Colorado Monitoring Sites 7-38
7-6 Risk Approximations for the Colorado Monitoring Sites 7-78
7-7 Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Colorado Monitoring Sites 7-81
7-8 Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Colorado Monitoring
Sites 7-84
8-1 Geographical Information for the Washington, D.C. Monitoring Site 8-4
8-2 Population, Motor Vehicle, and Traffic Information for the Washington, D.C.
Monitoring Site 8-5
8-3 Average Meteorological Conditions near the Washington, D.C. Monitoring Site 8-7
8-4 Risk-Based Screening Results for the Washington, D.C. Monitoring Site 8-12
8-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Washington, D.C. Monitoring Site 8-13
8-6 Risk Approximations for the Washington, D.C. Monitoring Site 8-17
8-7 Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Washington, D.C. Monitoring Site 8-19
8-8 Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Washington, D.C.
Monitoring Site 8-20
xl
-------
LIST OF TABLES (Continued)
Page
9-1 Geographical Information for the Florida Monitoring Sites 9-10
9-2 Population, Motor Vehicle, and Traffic Information for the Florida Monitoring
Sites 9-13
9-3 Average Meteorological Conditions near the Florida Monitoring Sites 9-15
9-4 Risk-Based Screening Results for the Florida Monitoring Sites 9-34
9-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Florida Monitoring Sites 9-36
9-6 Risk Approximations for the Florida Monitoring Sites 9-56
9-7 Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Florida Monitoring Sites 9-58
9-8 Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Florida Monitoring
Sites 9-61
10-1 Geographical Information for the Georgia Monitoring Site 10-4
10-2 Population, Motor Vehicle, and Traffic Information for the Georgia Monitoring
Site 10-5
10-3 Average Meteorological Conditions near the Georgia Monitoring Site 10-7
10-4 Risk-Based Screening Results for the Georgia Monitoring Site 10-12
10-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Georgia Monitoring Site 10-14
10-6 Risk Approximations for the Georgia Monitoring Site 10-20
10-7 Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Georgia Monitoring Site 10-21
10-8 Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Georgia Monitoring
Site 10-22
11-1 Geographical Information for the Illinois Monitoring Sites 11-7
11-2 Population, Motor Vehicle, and Traffic Information for the Illinois Monitoring
Sites 11-9
11-3 Average Meteorological Conditions near the Illinois Monitoring Sites 11-12
11-4 Risk-Based Screening Results for the Illinois Monitoring Sites 11-24
11-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Illinois Monitoring Sites 11-28
11-6 Risk Approximations for the Illinois Monitoring Sites 11-65
11-7 Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Illinois Monitoring Sites 11 -68
11-8 Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Illinois Monitoring
Sites 11-70
12-1 Geographical Information for the Indiana Monitoring Sites 12-6
12-2 Population, Motor Vehicle, and Traffic Information for the Indiana Monitoring
Sites 12-8
12-3 Average Meteorological Conditions near the Indiana Monitoring Sites 12-10
xli
-------
LIST OF TABLES (Continued)
Page
12-4 Risk-Based Screening Results for the Indiana Monitoring Sites 12-18
12-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Indiana Monitoring Sites 12-20
12-6 Risk Approximations for the Indiana Monitoring Sites 12-29
12-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Indiana Monitoring Sites 12-31
12-8 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Indiana Monitoring
Sites 12-32
13-1 Geographical Information for the Kentucky Monitoring Sites 13-17
13-2 Population, Motor Vehicle, and Traffic Information for the Kentucky Monitoring
Sites 13-21
13-3 Average Meteorological Conditions near the Kentucky Monitoring Sites 13-23
13-4 Overview of Sampling Performed at the Kentucky Monitoring Sites 13-53
13-5 Risk-Based Screening Results for the Kentucky Monitoring Sites 13-53
13-6 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Kentucky Monitoring Sites 13-60
13-7 Risk Approximations for the Kentucky Monitoring Sites 13-77
13-8 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Kentucky Monitoring Sites 13-82
13-9 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Kentucky Monitoring
Sites 13-87
14-1 Geographical Information for the Massachusetts Monitoring Site 14-4
14-2 Population, Motor Vehicle, and Traffic Information for the Massachusetts
Monitoring Site 14-5
14-3 Average Meteorological Conditions near the Massachusetts Monitoring Site 14-7
14-4 Risk-Based Screening Results for the Massachusetts Monitoring Site 14-13
14-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Massachusetts Monitoring Site 14-14
14-6 Risk Approximations for the Massachusetts Monitoring Site 14-25
14-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Massachusetts Monitoring Site 14-26
14-8 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Massachusetts
Monitoring Site 14-27
15-1 Geographical Information for the Michigan Monitoring Sites 15-6
15-2 Population, Motor Vehicle, and Traffic Information for the Michigan Monitoring
Sites 15-8
15-3 Average Meteorological Conditions near the Michigan Monitoring Sites 15-10
15-4 Risk-Based Screening Results for the Michigan Monitoring Sites 15-21
15-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Michigan Monitoring Sites 15-24
xlii
-------
LIST OF TABLES (Continued)
Page
15-6 Risk Approximations for the Michigan Monitoring Sites 15-45
15-7 Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Michigan Monitoring Sites 15-47
15-8 Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Michigan Monitoring
Sites 15-49
16-1 Geographical Information for the Minnesota Monitoring Site 16-4
16-2 Population, Motor Vehicle, and Traffic Information for the Minnesota Monitoring
Site 16-5
16-3 Average Meteorological Conditions near the Minnesota Monitoring Site 16-7
16-4 Risk-Based Screening Results for the Minnesota Monitoring Site 16-13
16-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Minnesota Monitoring Site 16-14
16-6 Risk Approximations for the Minnesota Monitoring Site 16-17
16-7 Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Minnesota Monitoring Site 16-19
16-8 Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Minnesota Monitoring
Site 16-20
17-1 Geographical Information for the Missouri Monitoring Site 17-4
17-2 Population, Motor Vehicle, and Traffic Information for the Missouri Monitoring
Site 17-5
17-3 Average Meteorological Conditions near the Missouri Monitoring Site 17-7
17-4 Risk-Based Screening Results for the Missouri Monitoring Site 17-14
17-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Missouri Monitoring Site 17-16
17-6 Risk Approximations for the Missouri Monitoring Site 17-47
17-7 Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Missouri Monitoring Site 17-49
17-8 Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Missouri Monitoring
Site 17-50
18-1 Geographical Information for the New Jersey Monitoring Sites 18-7
18-2 Population, Motor Vehicle, and Traffic Information for the New Jersey Monitoring
Sites 18-9
18-3 Average Meteorological Conditions near the New Jersey Monitoring Sites 18-11
18-4 Risk-Based Screening Results for the New Jersey Monitoring Sites 18-22
18-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
New Jersey Monitoring Sites 18-25
18-6 Risk Approximations for the New Jersey Monitoring Sites 18-65
18-7 Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the New Jersey Monitoring Sites 18-68
xliii
-------
LIST OF TABLES (Continued)
Page
18-8 Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the New Jersey Monitoring
Sites 18-70
19-1 Geographical Information for the New York Monitoring Sites 19-7
19-2 Population, Motor Vehicle, and Traffic Information for the New York Monitoring
Sites 19-9
19-3 Average Meteorological Conditions near the New York Monitoring Sites 19-12
19-4 Risk-Based Screening Results for the New York Monitoring Sites 19-24
19-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
New York Monitoring Sites 19-26
19-6 Risk Approximations for the New York Monitoring Sites 19-32
19-7 Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the New York Monitoring Sites 19-33
19-8 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the New York Monitoring
Sites 19-35
20-1 Geographical Information for the Oklahoma Monitoring Sites 20-10
20-2 Population, Motor Vehicle, and Traffic Information for the Oklahoma Monitoring
Sites 20-13
20-3 Average Meteorological Conditions near the Oklahoma Monitoring Sites 20-15
20-4 Risk-Based Screening Results for the Oklahoma Monitoring Sites 20-31
20-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Oklahoma Monitoring Sites 20-36
20-6 Risk Approximations for the Oklahoma Monitoring Sites 20-71
20-7 Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Oklahoma Monitoring Sites 20-75
20-8 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Oklahoma Monitoring
Sites 20-78
21-1 Geographical Information for the Rhode Island Monitoring Site 21-4
21-2 Population, Motor Vehicle, and Traffic Information for the Rhode Island Monitoring
Site 21-5
21-3 Average Meteorological Conditions near the Rhode Island Monitoring Site 21-7
21-4 Risk-Based Screening Results for the Rhode Island Monitoring Site 21-12
21-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Rhode Island Monitoring Site 21-13
21-6 Risk Approximations for the Rhode Island Monitoring Site 21-18
21-7 Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Rhode Island Monitoring Site 21 -20
21-8 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Rhode Island
Monitoring Site 20-21
xliv
-------
LIST OF TABLES (Continued)
Page
22-1 Geographical Information for the South Carolina Monitoring Site 22-4
22-2 Population, Motor Vehicle, and Traffic Information for the South Carolina
Monitoring Site 22-5
22-3 Average Meteorological Conditions near the South Carolina Monitoring Site 22-7
22-4 Risk-Based Screening Results for the South Carolina Monitoring Site 22-12
22-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
South Carolina Monitoring Site 22-13
22-6 Risk Approximations for the South Carolina Monitoring Site 22-17
22-7 Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the South Carolina Monitoring Site 22-18
22-8 Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the South Carolina
Monitoring Site 22-19
23-1 Geographical Information for the South Dakota Monitoring Site 23-4
23-2 Population, Motor Vehicle, and Traffic Information for the South Dakota
Monitoring Site 23-5
23-3 Average Meteorological Conditions near the South Dakota Monitoring Site 23-7
23-4 Risk-Based Screening Results for the South Dakota Monitoring Site 23-12
23-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
South Dakota Monitoring Site 23-14
23-6 Risk Approximations for the South Dakota Monitoring Site 23-26
23-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the South Dakota Monitoring Site 23-28
23-8 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the South Dakota
Monitoring Site 23-29
24-1 Geographical Information for the Texas Monitoring Sites 24-6
24-2 Population, Motor Vehicle, and Traffic Information for the Texas Monitoring Sites... 24-8
24-3 Average Meteorological Conditions near the Texas Monitoring Sites 24-10
24-4 Risk-Based Screening Results for the Texas Monitoring Sites 24-18
24-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Texas Monitoring Sites 24-20
24-6 Risk Approximations for the Texas Monitoring Sites 24-22
24-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Texas Monitoring Sites 24-24
24-8 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Texas Monitoring
Sites 24-25
25-1 Geographical Information for the Utah Monitoring Site 25-4
25-2 Population, Motor Vehicle, and Traffic Information for the Utah Monitoring Site 25-5
25-3 Average Meteorological Conditions near the Utah Monitoring Site 25-7
25-4 Risk-Based Screening Results for the Utah Monitoring Site 25-12
xlv
-------
LIST OF TABLES (Continued)
Page
25-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Utah Monitoring Site 25-14
25-6 Risk Approximations for the Utah Monitoring Site 25-38
25-7 Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Utah Monitoring Site 25-40
25-8 Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Utah Monitoring
Site 25-41
26-1 Geographical Information for the Vermont Monitoring Sites 26-7
26-2 Population, Motor Vehicle, and Traffic Information for the Vermont
Monitoring Sites 26-9
26-3 Average Meteorological Conditions near the Vermont Monitoring Sites 26-12
26-4 Risk-Based Screening Results for the Vermont Monitoring Sites 26-22
26-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Vermont Monitoring Sites 26-25
26-6 Risk Approximations for the Vermont Monitoring Sites 26-39
26-7 Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Vermont Monitoring Sites 26-42
26-8 Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Vermont
Monitoring Sites 26-44
27-1 Geographical Information for the Virginia Monitoring Site 27-4
27-2 Population, Motor Vehicle, and Traffic Information for the Virginia Monitoring
Site 27-5
27-3 Average Meteorological Conditions near the Virginia Monitoring Site 27-7
27-4 Risk-Based Screening Results for the Virginia Monitoring Site 27-12
27-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Virginia Monitoring Site 27-14
27-6 Risk Approximations for the Virginia Monitoring Site 27-17
27-7 Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Virginia Monitoring Site 27-18
27-8 Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Virginia Monitoring
Site 27-19
28-1 Geographical Information for the Washington Monitoring Site 28-4
28-2 Population, Motor Vehicle, and Traffic Information for the Washington
Monitoring Site 28-5
28-3 Average Meteorological Conditions near the Washington Monitoring Site 28-7
28-4 Risk-Based Screening Results for the Washington Monitoring Site 28-12
28-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Washington Monitoring Site 28-14
28-6 Risk Approximations for the Washington Monitoring Site 28-34
xlvi
-------
LIST OF TABLES (Continued)
Page
28-7 Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Washington Monitoring Site 28-36
28-8 Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Washington
Monitoring Site 28-37
29-1 Geographical Information for the Wisconsin Monitoring Sites 29-6
29-2 Population, Motor Vehicle, and Traffic Information for the Wisconsin
Monitoring Sites 29-8
29-3 Average Meteorological Conditions near the Wisconsin Monitoring Sites 29-10
29-4 Risk-Based Screening Results for the Wisconsin Monitoring Sites 29-19
29-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Wisconsin Monitoring Sites 29-20
29-6 Risk Approximations for the Wisconsin Monitoring Sites 29-23
29-7 Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Wisconsin Monitoring Sites 29-24
29-8 Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Wisconsin
Monitoring Sites 29-25
30-1 Method Precision by Analytical Method 30-4
30-2 VOC Method Precision: Coefficient of Variation Based on Duplicate and
Collocated Samples by Site and Pollutant 30-5
30-3 SNMOC Method Precision: Coefficient of Variation Based on Duplicate and
Collocated Samples by Site and Pollutant 30-13
30-4 Carbonyl Compound Method Precision: Coefficient of Variation Based on
Duplicate and Collocated Samples by Site and Pollutant 30-15
30-5 PAH Method Precision: Coefficient of Variation Based on Collocated Samples
by Site and Pollutant 30-18
30-6 Metals Method Precision: Coefficient of Variation Based on Collocated Samples
by Site and Pollutant 30-19
30-7 Hexavalent Chromium Method Precision: Coefficient of Variation Based on
Collocated Samples by Site 30-20
30-8 Analytical Precision by Analytical Method 30-22
30-9 VOC Analytical Precision: Coefficient of Variation Based on Replicate Analyses
by Site and Pollutant 30-23
30-10 SNMOC Analytical Precision: Coefficient of Variation Based on Replicate
Analyses by Site and Pollutant 30-31
30-11 Carbonyl Compound Analytical Precision: Coefficient of Variation Based on
Replicate Analyses by Site and Pollutant 30-33
30-12 PAH Analytical Precision: Coefficient of Variation Based on Replicate Analyses
by Site and Pollutant 30-36
30-13 Metals Analytical Precision: Average Coefficient of Variation Based on Replicate
Analyses by Site and Pollutant 30-40
30-14 Hexavalent Chromium Analytical Precision: Coefficient of Variation Based on
Replicate Analyses by Site 30-42
xlvii
-------
LIST OF TABLES (Continued)
Page
30-15 VOC Internal PT Audit Samples-Percent Difference from True Value 30-44
30-16 Carbonyl Compound Internal PT Audit Samples-Percent Difference from
True Value 30-45
30-17 PAH NATTSPT Audit Samples-Percent Difference from Mean 30-46
30-18 Metals NATTS PT Audit Samples-Percent Difference from True Value 30-46
30-19 Hexavalent Chromium NATTS PT Audit Samples-Percent Difference from True
Value 30-46
30-20 Hexavalent Chromium Internal PT Audit Samples-Percent Difference from True
Value 30-46
30-21 NAAQS Lead PT Audit Samples-Percent Difference from True Value for Multiple
Concentrations 30-47
xlviii
-------
LIST OF ACRONYMS
AADT Annual Average Daily Traffic
AGL Above Ground Level
AQS Air Quality System
ASE Accelerated Solvent Extractor
ATSDR Agency for Toxic Substances and Disease Registry
CBSA Core-Based Statistical Area(s)
CFR Code of Federal Regulations
CNG Compressed Natural Gas
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
FHWA Federal Highway Administration
GC/MS-FID Gas Chromatography/Mass Spectrometry and Flame lonization Detection
GHG Greenhouse Gas(es)
GIS Geographical Information System
GMT Greenwich Mean Time
GWP Global Warming Potential
HAP Hazardous Air Pollutant(s)
HPLC High-Performance Liquid Chromatography
HQ Hazard Quotient
HYSPLIT Hybrid Single-Particle Lagrangian Integrated Trajectory
1C Ion Chromatography
ICP-MS Inductively Coupled Plasma/Mass Spectrometry
IPCC Intergovernmental Panel on Climate Change
kt Knots
mb Millibar
MDL Method Detection Limit
mg/m3 Milligrams per cubic meter
mL Milliliter
MQO Measurement Quality Objective(s)
MRL Minimal Risk Level
MSA Metropolitan or Micropolitan Statistical Area(s)
NAAQS National Ambient Air Quality Standard
NATA National Air Toxics Assessment
xlix
-------
LIST OF ACRONYMS (Continued)
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
NOAA National Oceanic and Atmospheric Administration
NOx Oxides of Nitrogen
NWS National Weather Service
PAH Poly cyclic Aromatic Hydrocarbon(s)
PAMS Photochemical Assessment Monitoring Stations
PM Particulate Matter
PMio Particulate Matter less than 10 microns
POM Poly cyclic Organic Matter
ppbC Parts per billion carbon
ppbv Parts per billion by volume
ppm Parts per million
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
TSV Total Spatial Variance
UATMP Urban Air Toxics Monitoring Program
l^g/m3 Micrograms per cubic meter
I^L Microliter
URE Unit Risk Estimate(s)
UTC Universal Time Coordinated
UV Ultraviolet
UV-VIS Ultraviolet Visible
VMT Vehicle Miles Traveled
VOC Volatile Organic Compound(s)
WBAN Weather Bureau/Army/Navy ID
-------
Abstract
This report presents the results and conclusions from the ambient air monitoring conducted
as part of the 2012 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 2012 NMP includes data from
samples collected at 64 monitoring sites that collected 24-hour air samples, typically on a l-in-6
or l-in-12 day schedule. Thirty sites sampled for 59 volatile organic compounds (VOCs); 37 sites
sampled for 15 carbonyl compounds; eight sites sampled for 80 speciated nonmethane organic
compounds (SNMOCs); 25 sites sampled for 22 poly cyclic aromatic hydrocarbons (PAHs); 19
sites sampled for 11 metals; and 25 sites sampled for hexavalent chromium. Over 233,000
ambient air concentrations were measured during the 2012 NMP. 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 significantly from city-to-city and from season-to-season.
The ambient air monitoring data collected during the 2012 NMP serve a wide range of
purposes. Not only do these data characterize the nature and extent of air pollution close to the 64
individual monitoring sites participating in these programs, but they also identify 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.
-------
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 impacts, 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). These programs have the
following program-specific objectives:
• The primary objective of the UATMP is to characterize the composition and
magnitude of air toxics pollution through ambient air monitoring.
• The primary 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.
• The primary objective of the CSATAM Program is to conduct local-scale
investigative ambient air toxics monitoring projects.
1.1 Background
EPA began the NMOC program in 1984. Monitoring for selected NMOCs was performed
during the morning hours of the summer ozone season. NMOC data were to be used to better
understand ozone formation and to develop ozone control strategies. The UATMP was initiated
by EPA in 1988 as an extension of the existing NMOC program 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
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
1-1
-------
over an extended period of time (EPA, 2009a). The generation of large quantities of high-quality
data over an extended period may allow concentration trends (i.e., any substantial increase or
decrease over a period of time) to 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, 2013a;
EPA 2009b). 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 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 (EPA, 2009b). Currently, 27 NATTS sites
are strategically placed across the country (EPA, 2013 a).
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 (EPA, 2009a). The objectives of the CSATAM Program include
identifying and profiling air toxics sources; developing and assessing 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. 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 contract 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
monitoring study in which their data are included in the report as well.
1-2
-------
In past reports, measurements from UATMP, NATTS, and CSATAM monitoring sites
have been presented together and referred to as "UATMP sites." In more recent reports, a
distinction is 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 64 monitoring sites around the country. The
64 sites included in this report are located in or near 38 urban or rural locations in 24 states and
the District of Columbia, including 35 metropolitan or micropolitan statistical areas (MS As).
This report provides both a qualitative overview of air toxics pollution at selected 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 64 different air sampling 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?
• Have pollutant concentrations decreased as a result of regulations (or increased
despite regulation)?
• Which pollutants contribute the greatest health risk on a short-term, intermediate-
term, and long-term basis?
1-3
-------
The data analyses contained in this report are applied to each participating UATMP,
NATTS, or CSATAM 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 2012 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, 2013b).
This report is organized into 32 sections and 17 appendices. While each state section is
designed to be a stand-alone section to allow those interested in a particular site or state to
understand the associated data analyses without having to read the entire report, it is
recommended that Sections 1 through 4 (Introduction, Monitoring Programs Network Overview,
Data Treatments and Methods, and Summary of Results) and Sections 29 and 30 (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 2012 National Monitoring Programs Report
Report
Section
Section Title
Overview of Contents
Introduction
This section serves as an introduction to the
background and scope of the NMP (specifically, the
UATMP, NATTS, and CSATAM Programs).
The 2012 National Monitoring
Programs Network
This section provides information on the 2012 NMP
monitoring effort:
• Monitoring locations
• Pollutants selected for monitoring
• Sampling and analytical methods
• Sampling schedules
• Completeness of the air monitoring programs.
Summary of the 2012 National
Monitoring Programs Data
Treatments and Methods
This section presents and discusses the data treatments
applied to the 2012 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.
1-4
-------
Table 1-1. Organization of the 2012 National Monitoring Programs Report (Continued)
Report
Section
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Section Title
Summary of the 2012 National
Monitoring Programs Data
Sites in Arizona
Sites in California
Sites in Colorado
Site in the District of Columbia
Sites in Florida
Site in Georgia
Sites in Illinois
Sites in Indiana
Sites in Kentucky
Site in Massachusetts
Sites in Michigan
Site in Minnesota
Site in Missouri
Sites in New Jersey
Overview of Contents
This section presents and discusses the results of the
data treatments from the 2012 NMP data.
Monitoring results for the sites in the Phoenix-Mesa-
Scottsdale, AZ MSA (PXSS and SPAZ)
Monitoring results for the sites in the Los Angeles-
Long Beach-Anaheim, CA MSA (CELA and
LBHCA), the Riverside-San Bernardino-Ontario, CA
MSA (RUCA), and the San Jose-Sunnyvale-Santa
Clara, CA MSA (SJJCA)
Monitoring results for the sites in the Grand Junction,
CO MSA (GPCO) and the Glenwood Springs, CO
MSA (BMCO, BRCO, PACO, RFCO, and RICO)
Monitoring results for the site in the Washington-
Arlington-Alexandria, DC-VA-MD-WV MSA
(WADC)
Monitoring results for the sites in the Orlando-
Kissimmee-Sanford, FL MSA (ORFL and PAFL) and
the Tampa-St. Petersburg-Clearwater, FL MSA
(AZFL, SKFL, and SYFL)
Monitoring results for the site in the Atlanta-Sandy
Springs-Roswell, GA MSA (SDGA)
Monitoring results for the sites in the Chicago-
Naperville-Elgin, IL-IN-WI MSA (NBIL and SPIL)
and the St. Louis, MO-IL MSA (ROIL)
Monitoring results for the sites in the Chicago-
Naperville-Elgin, IL-IN-WI MSA (INDEM) and the
Indianapolis-Carmel-Anderson, IN MSA (WPIN)
Monitoring results for the sites in the Huntington-
Ashland, WV-KY-OH MSA (ASKY and ASKY-M),
the Lexington-Fayette, KY MSA (LEKY), the
Evansville, IN-KY MSA (BAKY), the Paducah, KY-
IL MSA (BLKY) and the sites in Marshall County
(ATKY, CCKY, LAKY, and TVKY) and Carter
County (GLKY)
Monitoring results for the site in the Boston-
Cambridge-Newton, MA-NH MSA (BOMA)
Monitoring results for the sites in the Detroit-Warren-
Dearborn, MI MSA (DEMI, RRMI, and SWMI)
Monitoring results for the site in the St. Cloud, MN
MSA (STMN)
Monitoring results for the site in the St. Louis, MO-IL
MSA (S4MO)
Monitoring results for the sites in the New York-
Newark-Jersey City, NY-NJ-PA MSA (CHNJ, ELNJ,
and NBNJ)
1-5
-------
Table 1-1. Organization of the 2012 National Monitoring Programs Report (Continued)
Report
Section
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Section Title
Sites in New York
Sites in Oklahoma
Site in Rhode Island
Site in South Carolina
Site in South Dakota
Sites in Texas
Site in Utah
Sites in Vermont
Site in Virginia
Site in Washington
Sites in Wisconsin
Data Quality
Results, Conclusions, and
Recommendations
References
Overview of Contents
Monitoring results for the sites in the New York-
Newark-Jersey City, NY-NJ-PA MSA (BXNY and
MONY) and the Rochester, NY MSA (ROCH)
Monitoring results for the sites in the Tulsa, OK MSA
(TOOK and TMOK), the Oklahoma City, OK MSA
(ADOK and OCOK), and Pryor Creek, OK (PROK)
Monitoring results for the site in the Providence-
Warwick, RI-MA MSA (PRRI)
Monitoring results for the site in Chesterfield, SC
(CHSC)
Monitoring results for the site in the Sioux Falls, SD
MSA (SSSD)
Monitoring results for the sites in the Houston-The
Woodlands-Sugar Land, TX MSA (CAMS 35) and the
Marshall, TX MSA (CAMS 85)
Monitoring results for the site in the Ogden-Clearfield,
UT MSA (BTUT)
Monitoring results for the sites in the Burlington-South
Burlington, VT MSA (BURVT and UNVT) and the
Rutland, VT MSA (RUVT)
Monitoring results for the site in the Richmond, VA
MSA (RIVA)
Monitoring results for the site in the Seattle-Tacoma-
Bellevue, WA MSA (SEWA)
Monitoring results for the sites in the Beaver Dam, WI
MSA (HOWI) and the Milwaukee-Waukesha-West
Allis, WI MSA (MIWI)
This section defines and discusses the concepts of
precision and accuracy. Based on quantitative and
qualitative analyses, this section comments on the
precision and accuracy of the 2012 NMP ambient air
monitoring data.
This section summarizes the most significant findings
of the report and makes several recommendations for
future projects that involve ambient air monitoring.
This section lists the references cited throughout the
report.
1-6
-------
2.0 The 2012 National Monitoring Programs Network
Agencies operating UATMP, NATTS, or CSATAM sites may choose to have their
samples analyzed by EPA's contract laboratory, ERG, in Morrisville, NC. Data from 64
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 selected hydrocarbons, halogenated
hydrocarbons, and polar compounds from canister
samples (Speciated Nonmethane Organic Compounds
(SNMOCs) and/or Method TO-15), carbonyl
compounds from sorbent cartridge samples (Method
TO-11 A), poly cyclic aromatic hydrocarbons (PAHs)
from polyurethane foam (PUF) and XAD-2® resin
samples (Method TO-13 A), trace metals from filters
(Method IO-3.5), and hexavalent chromium from
sodium bicarbonate-coated filters (ASTM D7614).
Section 2.2 provides additional information regarding
each of the sampling methodologies used to collect and
analyze samples.
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.
The following sections review the monitoring locations, pollutants selected for
monitoring, sampling and analytical methods, collection schedules, and completeness of the
2012NMPdataset.
2.1 Monitoring Locations
For the NATTS network, monitor siting is based on the need to assess population
exposure and background-level concentrations. For the UATMP and CSATAM 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.
Among these programs, monitors were placed in urban areas near the centers of heavily
populated cities (e.g., Chicago, IL and Phoenix, AZ), while others were placed in moderately
populated rural areas (e.g., Horicon, WI and Chesterfield, SC).
2-1
-------
Figure 2-1 shows the locations of the 64 monitoring sites participating in the 2012
programs, which encompass 38 different urban and rural areas. Outlined in Figure 2-1 are the
associated core-based statistical areas (CBSA), 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
for the 64 monitoring sites. Forty-nine monitoring sites have been included in previous annual
reports. Fifteen monitoring sites are new to their respective programs for 2012; these sites are
highlighted in green in Table 2-1. One NATTS site (BXNY) was relocated to a different location
in 2010 (MONY) while construction was ongoing near the monitoring site. In June 2012, the
instrumentation was moved back to the original location at BXNY.
As Figure 2-1 and Table 2-1 show, the 2012 NMP sites are widely distributed across the
country. Detailed information about the monitoring sites is provided in Table 2-2 and
Appendix A. 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 location of the monitoring sites
vary significantly from site to site. 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 selected monitoring
results. For reference, each site's AQS site code is provided in Table 2-2.
2-2
-------
Figure 2-1. Locations of the 2012 National Monitoring Programs Monitoring Sites
San Jose, CA /Parachute, CO \ silt, CO
Battlement Mesa,
Grand Junction. CO
Underbill,
f \ t^
Rutland, VT
Rochester, NY LJ^Boston, MA
Burlington, VT
Horicon, Wl
, Dearborn, Ml
Milwaukee, WJ-3petroit., MUfF^ V^Chester, NJ
NorthbrookTlL^aiZ-River Rouge, Ml
Schiller Park, lr*L^-TGary, JN
Indianapolis, IN
^ Roxana.lL^ l^» J
Providence, Rl
New York, NY (2)
Elizabeth, NJ
North Brunswick
Township, NJ
Washington, DC
[Ashland; KY (2)
.n KYC°
-------
Table 2-1. 2012 National Monitoring Programs Sites and Past Program Participation
Monitoring Location
and Site
Ashland, KY (ASKY)
Ashland, KY (ASKY-M)
Baskett, KY (BAKY)
Battlement Mesa, CO (BMCO)
Boston, MA (BOMA)
Bountiful, UT (BTUT)
Burlington, VT (BURVT)
Calvert City, KY (ATKY)
Calvert City, KY (CCKY)
Calvert City, KY (LAKY)
Calvert City, KY (TVKY)
Carbondale, CO (RFCO)
Chester, NJ (CHNJ)
Chesterfield, SC (CHSC)
Dearborn, MI (DEMI)
Decatur, GA (SDGA)
Deer Park, TX (CAMS 35)
Program
UATMP
UATMP
UATMP
UATMP
NATTS
NATTS
UATMP
UATMP
UATMP
UATMP
UATMP
UATMP
UATMP
NATTS
NATTS
NATTS
NATTS
2002 and Earlier
2001, 2002
2001, 2002
2003
•/
2004
•/
2005
•/
•/
•/
2006
•/
•/
•/
2007
•/
•/
•/
2008
•/
•/
•/
2009
•/
•/
•/
•/
2010
•/
•/
•/
•/
2011
•/
•/
•/
•/
2012
•/
•/
•/
•/
•/
•/
•/
•/
to
Green shading indicates new site participating in the NMP.
BOLD ITALICS = EPA-designated NATTS site
*Special air toxics monitoring study.
-------
Table 2-1. 2012 National Monitoring Programs Sites and Past Program Participation (Continued)
Monitoring Location
and Site
Detroit, MI (SWMI)
East Highland Park, VA (RIVA)
Elizabeth, NJ (ELNJ)
Gary, IN (INDEM)
Grand Junction, CO (GPCO)
Grayson, KY (GLKY)
Horicon, WI (HOWI)
Indianapolis, IN (WPIN)
Karnack, TX (CAMS 85)
Lexington, KY (LEKY)
Long Beach, CA (LBHCA)
Los Angeles, CA (CELA)
Milwaukee, WI (MIWI)
New York, NY (BXNY)
New York, NY (MONY)
Northbrook, IL (NBIL)
North Brunswick, NJ (NBNJ)
Program
UATMP
NATTS
UATMP
UATMP
NATTS
NATTS
NATTS
UATMP
NATTS
UATMP
CSATAM
NATTS
UATMP
NATTS
NATTS
NATTS
UATMP
2002 and Earlier
2001,2002
1999-2002
2001, 2002
2003
•/
•/
2004
•/
•/
•/
2005
•/
•/
•/
2006
•/
•/
•/
•/
2007
•/
•/
•/
•/
•/
•/
2008
•/
•/
•/
•/
•/
2009
•/
•/
•/
•/
•/
2010
•/
•/
•/
•/
•/
•/
•/
2011
•/
•/
•/
•/
•/
•/
2012
•/
•/
•/
•/
•/
•/
•/
to
Green shading indicates new site participating in the NMP.
BOLD ITALICS = EPA-designaled NATTS site
*Special air toxics monitoring study.
-------
Table 2-1. 2012 National Monitoring Programs Sites and Past Program Participation (Continued)
Monitoring Location
and Site
Oklahoma City, OK (ADOK)
Oklahoma City, OK (OCOK)
Orlando, FL (PAFL)
Parachute, CO (PACO)
Phoenix, AZ (PXSS)
Phoenix, AZ (SPAZ)
Pinellas Park, FL (SKFL)
Providence, RI (PRRI)
Pryor Creek, OK (PROK)
Rifle, CO (RICO)
River Rouge, MI (RRMI)
Rochester, NY (ROCH)
Roxana, IL (ROIL)
Rubidoux, CA (RUCA)
Rutland, VT (RUVT)
San Jose, CA (SJJCA)
Schiller Park, IL (SPIL)
Program
UATMP
UATMP
UATMP
UATMP
NATTS
UATMP
NATTS
NATTS
UATMP
UATMP
UATMP
NATTS
UATMP*
NATTS
UATMP
NATTS
UATMP
2002 and Earlier
2001, 2002
2001
2001
1995-1999, 2002
2003
•/
2004
•/
•/
2005
•/
2006
•/
•/
•/
2007
•/
•/
•/
•/
2008
•/
•/
•/
•/
•/
•/
•/
•/
2009
•/
•/
•/
•/
•/
•/
•/
•/
2010
•/
•/
•/
•/
•/
•/
•/
•/
2011
•/
•/
•/
•/
•/
•/
•/
•/
2012
•/
•/
•/
•/
•/
•/
•/
•/
to
Green shading indicates new site participating in the NMP.
BOLD ITALICS = EPA-designated NATTS site
*Special air toxics monitoring study.
-------
Table 2-1. 2012 National Monitoring Programs Sites and Past Program Participation (Continued)
Monitoring Location
and Site
Seattle, WA (SEWA)
Silt, CO (BRCO)
Sioux Falls, SD(SSSD)
Smithland, KY (BLKY)
St. Cloud, MN (STMN)
St. Louis, MO (S4MO)
St. Petersburg, FL (AZFL)
Tulsa, OK (TMOK)
Tulsa, OK (TOOK)
Underbill, VT (UNVT)
Valrico, FL (SYFL)
Washington, D.C. (WADC)
Winter Park, FL (ORFL)
Program
NATTS
UATMP
UATMP
UATMP
UATMP
NATTS
UATMP
UATMP
UATMP
NATTS
NATTS
NATTS
UATMP
2002 and Earlier
2002
1991-1992,2001-
2002
2002
1990-1991
2003
•/
2004
•/
2005
•/
•/
•/
2006
•/
•/
•/
•/
2007
•/
•/
•/
•/
2008
•/
•/
•/
•/
•/
2009
•/
•/
•/
•/
•/
2010
•/
•/
•/
•/
•/
2011
•/
•/
•/
•/
•/
2012
•/
•/
•/
•/
•/
•/
to
Green shading indicates new site participating in the NMP.
BOLD ITALICS = EPA-designated NATTS site
*Special air toxics monitoring study.
-------
Table 2-2. Site Characterizing Information for the 2012 National Monitoring Programs Sites
Site
Code
ADOK
ASKY
ASKY-M
ATKY
AZFL
BAKY
BLKY
BMCO
BOMA
BRCO
BTUT
BURVT
BXNY
AQS
Code
40-109-0042
21-019-0017
21-019-0002
21-157-0016
12-103-0018
21-101-0014
21-139-0004
NA
25-025-0042
08-045-0009
49-011-0004
50-007-0014
36-005-0110
Location
Oklahoma City, OK
Ashland, KY
Ashland, KY
Calvert City, KY
St. Petersburg, FL
Baskett, KY
Smithland, KY
Battlement Mesa, CO
Boston, MA
Silt, CO
Bountiful, UT
Burlington, VT
New York, NY
Land Use
Commercial
Residential
Industrial
Industrial
Residential
Commercial
Agricultural
Residential
Commercial
Agricultural
Residential
Commercial
Residential
Location
Setting
Urban/City
Center
Suburban
Urban/City
Center
Suburban
Suburban
Rural
Rural
Rural
Urban/City
Center
Rural
Suburban
Urban/City
Center
Urban/City
Center
County-level
Population"
741,781
49,164
49,164
31,344
921,319
46,513
9,423
56,953
744,426
56,953
315,809
158,504
1,408,473
County-level
Vehicle
Registration,
# of Vehicles'1
(Year)
847,824
(2012)
39,227
(2012)
39,227
(2012)
30,297
(2012)
872,813
(2012)
38,518
(2012)
8,281
(2012)
74,508
(2011)
362,899
(2012)
74,508
(2011)
259,319
(2012)
169,767
(2012)
251,398
(2012)
Estimated
Daily Traffic,
AADTb
(Year)
34,100
(2011)
7,229
(2011)
12,842
(2012)
3,262
(2012)
38,500
(2012)
922
(2012)
2,280
(2010)
2,527
(2002)
27,654
(2010)
1,102
(2002)
129,145
(2011)
14,000
(2007)
99,201
(2011)
County-level
Stationary
Source HAP
Emissions0
(tpy)
3,898.13
381.85
381.85
1,200.40
4,200.72
515.54
59.49
2,896.50
998.92
2,896.50
1,896.78
775.57
5,267.58
County-level
Mobile Source
HAP
Emissions0
(tpy)
2,760.20
133.65
133.65
467.15
2,592.37
238.32
116.57
284.27
965.67
284.27
844.04
505.84
1,158.43
to
oo
BOLD ITALICS = EPA-designaled NATTS site
""Reference: Census Bureau, 2013c
blndividual references provided in each state section.
"Reference: 2011 NEI (EPA, 2013c)
dThe proportion of county-level population to the state-level population was applied to state-level vehicle registration figure and used as a surrogate when county-
level vehicle registration counts were not available.
eGPCO's hexavalent chromium monitor is at a separate, but adjacent, location; thus, this site has two AQS codes.
NA = Data not loaded into AQS per agency request
-------
Table 2-2. Site Characterizing Information for the 2012 National Monitoring Programs Sites (Continued)
Site
Code
CAMS 35
CAMS 85
CCKY
CELA
CHNJ
CHSC
DEMI
ELNJ
GLKY
GPCOe
HOW
INDEM
LAKY
AQS
Code
48-201-1039
48-203-0002
21-157-0018
06-037-1103
34-027-3001
45-025-0001
26-163-0033
34-039-0004
21-043-0500
08-077-0017
08-077-0018
55-027-0001
18-089-0022
21-157-0019
Location
Deer Park, TX
Karnack, TX
Calvert City, KY
Los Angeles, CA
Chester, NJ
Chesterfield, SC
Dearborn, MI
Elizabeth, NJ
Grayson, KY
Grand Junction, CO
Horicon, WI
Gary, IN
Calvert City, KY
Land Use
Residential
Agricultural
Residential
Residential
Agricultural
Forest
Industrial
Industrial
Residential
Commercial
Agricultural
Industrial
Residential
Location
Setting
Urban/City
Center
Rural
Suburban
Urban/City
Center
Rural
Rural
Suburban
Suburban
Rural
Urban/City
Center
Rural
Urban/City
Center
Suburban
County-level
Population"
4,253,700
67,450
31,344
9,962,789
497,999
46,103
1,792,365
543,976
27,348
147,848
88,415
493,618
31,344
County-level
Vehicle
Registration,
# of Vehicles'1
(Year)
3,252,420
(2012)
71,658
(2012)
30,297
(2012)
7,422,254
(2012)
445,710
(Ratio)d
41,259
(2012)
1,337,797
(2012)
485,449
(Ratio)d
25,391
(2012)
179,213
(2011)
96,912
(2012)
419,431
(2011)
30,297
(2012)
Estimated
Daily Traffic,
AADTb
(Year)
31,043
(2004)
1,250
(2011)
4,742
(2010)
229,000
(2012)
11,215
(2012)
550
(2012)
87,500
(2012)
250,000
(2006)
303
(2012)
11,000
(2011)
5,100
(2011)
34,754
(2011)
1,189
(2012)
County-level
Stationary
Source HAP
Emissions0
(tpy)
23,207.29
926.93
1,200.40
28,724.47
1,117.70
277.26
11,321.82
2,367.55
144.52
921.64
672.72
2,720.85
1,200.40
County-level
Mobile Source
HAP
Emissions0
(tpy)
6,300.34
256.41
467.15
13,337.05
1,229.12
153.23
4,336.32
958.12
116.39
472.53
404.31
1,355.11
467.15
to
BOLD ITALICS = EPA-designaled NATTS site
""Reference: Census Bureau, 2013c
blndividual references provided in each state section.
"Reference: 2011 NEI (EPA, 2013c)
dThe proportion of county-level population to the state-level population was applied to state-level vehicle registration figure and used as a surrogate when county-
level vehicle registration counts were not available.
eGPCO's hexavalent chromium monitor is at a separate, but adjacent, location; thus, this site has two AQS codes.
NA = Data not loaded into AQS per agency request
-------
Table 2-2. Site Characterizing Information for the 2012 National Monitoring Programs Sites (Continued)
Site
Code
LBHCA
LEKY
MIWI
MONY
NBIL
NBNJ
OCOK
ORFL
PACO
PAFL
PROK
PRRI
PXSS
AQS
Code
06-037-4002
21-067-0012
55-079-0026
36-005-0080
17-031-4201
34-023-0006
40-109-1037
12-095-2002
08-045-0005
12-095-1004
40-097-0187
44-007-0022
04-013-9997
Location
Long Beach, CA
Lexington, KY
Milwaukee, WI
New York, NY
Northbrook, IL
North Brunswick, NJ
Oklahoma City, OK
Winter Park, FL
Parachute, CO
Orlando, FL
Pryor Creek, OK
Providence, RI
Phoenix, AZ
Land Use
Residential
Residential
Commercial
Residential
Residential
Agricultural
Residential
Commercial
Residential
Commercial
Industrial
Residential
Residential
Location
Setting
Suburban
Suburban
Urban/City
Center
Urban/City
Center
Suburban
Rural
Suburban
Urban/City
Center
Urban/City
Center
Suburban
Suburban
Urban/City
Center
Urban/City
Center
County-level
Population"
9,962,789
305,489
955,205
1,408,473
5,231,351
823,041
741,781
1,202,234
56,953
1,202,234
41,168
628,323
3,942,169
County-level
Vehicle
Registration,
# of Vehicles'1
(Year)
7,422,254
(2012)
207,043
(2012)
632,914
(2012)
251,398
(2012)
2,092,085
(2012)
733,908
(Ratio)d
847,824
(2012)
1,073,682
(2012)
74,508
(2011)
1,073,682
(2012)
41,391
(2012)
548,763
(Ratio)d
3,761,859
(2012)
Estimated
Daily Traffic,
AADTb
(Year)
282,000
(2012)
10,083
(2012)
12,800
(2013)
91,213
(2011)
115,100
(2012)
110,653
(2009)
40,900
(2011)
35,000
(2012)
16,000
(2011)
49,500
(2012)
15,100
(2011)
136,800
(2009)
184,000
(2010)
County-level
Stationary
Source HAP
Emissions0
(tpy)
28,724.47
1,466.57
5,075.77
5,267.58
21,497.97
2,531.15
3,898.13
5,649.93
2,896.50
5,649.93
351.44
2,745.08
16,951.30
County-level
Mobile Source
HAP
Emissions0
(tpy)
13,337.05
925.83
1,840.40
1,158.43
8,212.63
1,499.95
2,760.20
3,886.50
284.27
3,886.50
186.75
1,103.44
9,549.40
BOLD ITALICS = EPA-designaled NATTS site
""Reference: Census Bureau, 2013c
blndividual references provided in each state section.
"Reference: 2011 NEI (EPA, 2013c)
dThe proportion of county-level population to the state-level population was applied to state-level vehicle registration figure and used as a surrogate when county-
level vehicle registration counts were not available.
eGPCO's hexavalent chromium monitor is at a separate, but adjacent, location; thus, this site has two AQS codes.
NA = Data not loaded into AQS per agency request
-------
Table 2-2. Site Characterizing Information for the 2012 National Monitoring Programs Sites (Continued)
Site
Code
RFCO
RICO
RIVA
ROCH
ROIL
RRMI
RUCA
RUVT
S4MO
SDGA
SEWA
SJJCA
SKFL
AQS
Code
08-045-0018
08-045-0007
51-087-0014
36-055-1007
17-119-9010
26-163-0005
06-065-8001
50-021-0002
29-510-0085
13-089-0002
53-033-0080
06-085-0005
12-103-0026
Location
Carbondale, CO
Rifle, CO
East Highland Park,
VA
Rochester, NY
Roxana, IL
River Rouge, MI
Rubidoux, CA
Rutland, VT
St. Louis, MO
Decatur, GA
Seattle, WA
San Jose, CA
Pinellas Park, FL
Land Use
Residential
Commercial
Residential
Residential
Industrial
Industrial
Residential
Commercial
Residential
Residential
Residential
Commercial
Residential
Location
Setting
Rural
Urban/City
Center
Suburban
Urban/City
Center
Suburban
Suburban
Suburban
Urban/City
Center
Urban/City
Center
Suburban
Urban/City
Center
Urban/City
Center
Suburban
County-level
Population"
56,953
56,953
314,932
747,813
267,883
1,792,365
2,268,783
60,869
1,318,610
707,089
2,007,440
1,837,504
921,319
County-level
Vehicle
Registration,
# of Vehicles'1
(Year)
74,508
(2011)
74,508
(2011)
354,419
(2012)
556,055
(2012)
286,043
(2012)
1,337,797
(2012)
1,724,787
(2012)
70,900
(2012)
1,112,866
(2012)
472,535
(2011)
1,403,968
(2012)
1,529,351
(2012)
872,813
(2012)
Estimated
Daily Traffic,
AADTb
(Year)
16,000
(2011)
17,000
(2011)
72,000
(2012)
88,348
(2011)
9,400
(2011)
97,300
(2012)
145,000
(2012)
6,700
(2012)
79,558
(2011)
141,980
(2012)
224,000
(2012)
106,000
(2012)
49,000
(2012)
County-level
Stationary
Source HAP
Emissions0
(tpy)
2,896.50
2,896.50
1,531.17
3,932.40
1,807.49
11,321.82
5,424.56
307.04
1,714.27
5,444.91
9,553.33
5,252.06
4,200.72
County-level
Mobile Source
HAP
Emissions0
(tpy)
284.27
284.27
764.23
1,726.37
692.36
4,336.32
2,951.80
261.25
966.57
1,597.34
6,638.48
3,316.86
2,592.37
BOLD ITALICS = EPA-designaled NATTS site
""Reference: Census Bureau, 2013c
blndividual references provided in each state section.
"Reference: 2011 NEI (EPA, 2013c)
dThe proportion of county-level population to the state-level population was applied to state-level vehicle registration figure and used as a surrogate when county-
level vehicle registration counts were not available.
eGPCO's hexavalent chromium monitor is at a separate, but adjacent, location; thus, this site has two AQS codes.
NA = Data not loaded into AQS per agency request
-------
Table 2-2. Site Characterizing Information for the 2012 National Monitoring Programs Sites (Continued)
Site
Code
SPAZ
SPIL
SSSD
STMN
SWMI
SYFL
TMOK
TOOK
TVKY
UNVT
WADC
WPIN
AQS
Code
04-013-4003
17-031-3103
46-099-0008
27-145-3053
26-163-0015
12-057-3002
40-143-1127
40-143-0235
21-157-0014
50-007-0007
11-001-0043
18-097-0078
Location
Phoenix, AZ
Schiller Park, IL
Sioux Falls, SD
St. Cloud, MN
Detroit, MI
Valrico, FL
Tulsa, OK
Tulsa, OK
Calvert City, KY
Underbill, VT
Washington, D.C.
Indianapolis, IN
Land Use
Residential
Mobile
Commercial
Industrial
Commercial
Residential
Residential
Industrial
Industrial
Forest
Commercial
Residential
Location
Setting
Urban/City
Center
Suburban
Urban/City
Center
Suburban
Urban/City
Center
Rural
Urban/City
Center
Urban/City
Center
Suburban
Rural
Urban/City
Center
Suburban
County-level
Population"
3,942,169
5,231,351
175,037
151,606
1,792,365
1,277,746
613,816
613,816
31,344
158,504
632,323
918,977
County-level
Vehicle
Registration,
# of Vehicles'1
(Year)
3,761,859
(2012)
2,092,085
(2012)
212,507
(2012)
218,196
(2012)
1,337,797
(2012)
1,143,207
(2012)
618,359
(2012)
618,359
(2012)
30,297
(2012)
169,767
(2012)
316,231
(2011)
820,767
(2011)
Estimated
Daily Traffic,
AADTb
(Year)
128,000
(2010)
191,700
(2011)
18,575
(2012)
24,100
(2009)
94,400
(2012)
10,400
(2012)
12,600
(2011)
63,000
(2011)
2,231
(2011)
1,100
(2011)
7,400
(2010)
143,970
(2011)
County-level
Stationary
Source HAP
Emissions0
(tpy)
16,951.30
21,497.97
1,187.98
2,112.70
11,321.82
5,928.69
3,514.68
3,514.68
1,200.40
775.57
2,377.90
4,871.79
County-level
Mobile Source
HAP
Emissions0
(tpy)
9,549.40
8,212.63
481.52
1,198.42
4,336.32
3,869.11
2,195.17
2,195.17
467.15
505.84
863.89
3,218.51
BOLD ITALICS = EPA-designated NATTS site
""Reference: Census Bureau, 2013c
blndividual references provided in each state section.
"Reference: 2011 NEI (EPA, 2013c)
dThe proportion of county-level population to the state-level population was applied to state-level vehicle registration figure and used as a surrogate when county-
level vehicle registration counts were not available.
eGPCO's hexavalent chromium monitor is at a separate, but adjacent, location; thus, this site has two AQS codes.
NA = Data not loaded into AQS per agency request
-------
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 people living within each monitoring site's respective county.
• The county-level number of motor vehicles registered in each site's respective
county, based on total vehicle registrations.
• 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 the 2011 National Emissions
Inventory (NEI).
This information is discussed in further detail in Section 4.3 and 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,
volatile organic compounds (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 using specific methods, as
listed below. The target pollutants varied from monitoring site to monitoring site.
• 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. This
method was often 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.
• Compendium MethodIO-3.5 was used to measure ambient air concentrations of
11 metals.
• ASTMMethodD7614 was used to measure ambient air concentrations of hexavalent
chromium.
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
2-13
-------
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, 2013d) 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 sample contamination 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 FAC MDL procedure involves using historical blank filter data to
calculate MDLs for each pollutant. For arsenic, the procedure described in 40 CFR was used to
calculate the MDL rather than the FAC procedure because this metal is not present at a high
enough level in the background on the filters.
2-14
-------
Tables 2-3 through 2-8 identify the specific target pollutants for each analytical method
and their corresponding MDLs, as determined for 2012. For the VOC and SNMOC analyses, the
experimentally-determined MDLs do not change within a given year unless the sample was
diluted. The 2012 VOC and SNMOC MDLs are presented in Tables 2-3 and 2-4, respectively.
For the rest of the analyses, 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 in accordance
with a combination of 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, this report may refer to this
method as the "concurrent SNMOC method" or "concurrent SNMOC analysis" because both
methods were often 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 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
2-15
-------
into the canister. Site operators recovered and returned the canisters, along with the Chain of
Custody 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 nonmethane organic compounds (TNMOC) concentration.
TNMOC is the sum of all hydrocarbon concentrations within the sample. Because isobutene and
1-butene elute from the GC column at the same time, the SNMOC analytical method reports only
the sum concentration for these two compounds, and not the separate concentration for each
compound. The same approach applies to w-xylene and/?-xylene for both the VOC and
concurrent SNMOC methods. 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. Note that
beginning in 2012, two VOCs (chloromethylbenzene and methyl ethyl ketone) were removed
from the VOC list. The MDL for every VOC is less than 0.075 parts per billion by volume
(ppbv). SNMOC detection limits are expressed in parts per billion Carbon (ppbC). All of the
SNMOC MDLs are less than 0.40 ppbC.
2-16
-------
Table 2-3. 2012 VOC Method Detection Limits
Pollutant
Acetonitrile
Acetylene
Acrolein
Acrylonitrile
fer/-Amyl Methyl Ether
Benzene
Bromochloromethane
Bromodichloromethane
Bromoform
Bromomethane
1,3 -Butadiene
Carbon Bisulfide
Carbon Tetrachloride
Chlorobenzene
Chloroethane
Chloroform
Chloromethane
Chloroprene
Dibromochloromethane
1 ,2-Dibromoethane
/w-Dichlorobenzene
o-Dichlorobenzene
£>-Dichlorobenzene
Dichlorodifluoromethane
1 , 1 -Dichloroethane
1 ,2-Dichloroethane
1 , 1 -Dichloroethene
c/'s-l,2-Dichloroethylene
trans- 1 ,2-Dichloroethylene
2012
MDL
(ppbv)
0.073
0.072
0.060
0.020
0.016
0.061
0.014
0.021
0.020
0.013
0.011
0.014
0.025
0.025
0.017
0.014
0.033
0.012
0.018
0.017
0.024
0.021
0.019
0.023
0.015
0.016
0.014
0.018
0.012
Pollutant
Dichloromethane
1 ,2-Dichloropropane
cis- 1 ,3 -Dichloropropene
trans- 1 ,3 -Dichloropropene
Dichlorotetrafluoroethane
Ethyl Aery late
Ethyl tert-Butyl Ether
Ethylbenzene
Hexachloro- 1 ,3 -Butadiene
Methyl Isobutyl Ketone
Methyl Methacrylate
Methyl fer/-Butyl Ether
w-Octane
Propylene
Styrene
1 , 1 ,2,2-Tetrachloroethane
Tetrachloroethylene
Toluene
1 ,2,4-Trichlorobenzene
1,1,1 -Trichloroethane
1,1,2-Trichloroethane
Trichloroethylene
Trichlorofluoromethane
Trichlorotrifluoroethane
1 ,2,4-Trimethylbenzene
1 , 3 ,5 -Trimethy Ibenzene
Vinyl Chloride
/w.^-Xylene1
o-Xylene
2012
MDL
(ppbv)
0.023
0.019
0.015
0.016
0.012
0.014
0.011
0.023
0.022
0.015
0.025
0.011
0.020
0.033
0.024
0.019
0.020
0.045
0.022
0.021
0.021
0.022
0.015
0.017
0.025
0.022
0.011
0.037
0.020
1 Because /w-xylene and^-xylene elute from the GC column at the same time, the
VOC analytical method reports the sum of/w-xylene and^-xylene concentrations
and not concentrations of the individual isomers.
2-17
-------
Table 2-4. 2012 SNMOC Method Detection Limits
Pollutant
Acetylene
Benzene
1,3 -Butadiene
w-Butane
c/s-2-Butene
/raws-2-Butene
Cyclohexane
Cyclopentane
Cyclopentene
w-Decane
1-Decene
/w-Diethylbenzene
£>-Diethylbenzene
2,2-Dimethylbutane
2,3 -Dimethy Ibutane
2,3 -Dimethylpentane
2,4-Dimethylpentane
w-Dodecane
1-Dodecene
Ethane
2-Ethyl-l-butene
Ethylbenzene
Ethylene
/w-Ethyltoluene
o-Ethyltoluene
£>-Ethyltoluene
w-Heptane
2012
MDL
(ppbC)1
0.151
0.192
0.199
0.198
0.199
0.145
0.180
0.149
0.260
0.155
0.215
0.215
0.172
0.197
0.241
0.225
0.174
0.383
0.383
0.102
0.342
0.115
0.063
0.122
0.135
0.187
0.151
Pollutant
1-Heptene
w-Hexane
1-Hexene
c/s-2-Hexene
/raws-2-Hexene
Isobutane
Isobutene/1 -Butene2
Isopentane
Isoprene
Isopropylbenzene
2-Methy 1-1 -Butene
3 -Methy 1-1 -Butene
2-Methyl- 1 -Pentene
4-Methyl-l-Pentene
2-Methyl-2-Butene
Methylcyclohexane
Methylcyclopentane
2-Methylheptane
3-Methylheptane
2-Methylhexane
3 -Methy Ihexane
2-Methylpentane
3-Methylpentane
w-Nonane
1-Nonene
w-Octane
1-Octene
2012
MDL
(ppbC)1
0.225
0.141
0.342
0.342
0.342
0.125
0.165
0.260
0.247
0.159
0.260
0.260
0.342
0.342
0.260
0.142
0.114
0.126
0.120
0.131
0.111
0.093
0.155
0.123
0.187
0.155
0.212
Pollutant
w-Pentane
1 -Pentene
c/s-2-Pentene
trans-1 -Pentene
a-Pinene
6-Pinene
Propane
w-Propylbenzene
Propylene
Propyne
Styrene
Toluene
w-Tridecane
1-Tridecene
1 ,2,3 -Trimethylbenzene
1 ,2,4-Trimethylbenzene
1 ,3 ,5-Trimethylbenzene
2,2,3 -Trimethylpentane
2,2,4-Trimethylpentane
2,3,4-Trimethylpentane
w-Undecane
1-Undecene
7w-Xylene/£>-Xylene2
o-Xylene
Sum of Knowns
Sum of Unknowns
TNMOC
2012
MDL
(ppbC)1
0.161
0.183
0.215
0.152
0.215
0.215
0.183
0.137
0.099
0.183
0.187
0.212
0.383
0.383
0.104
0.183
0.119
0.212
0.125
0.141
0.237
0.237
0.188
0.094
NA
NA
NA
1 Concentration in ppbC = concentration in ppbv * number of carbon atoms in the compound.
2 Because isobutene and 1-butene elute from the GC column at the same time, the SNMOC analytical method
reports the sum concentration for these two compounds and not concentrations of the individual compounds. For
the same reason, the /w-xylene and^-xylene concentrations are reported as a sum concentration.
NA = Not applicable
2-18
-------
2.2.2 Carbonyl Compound Sampling and Analytical Method
Following the specifications of 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 the cartridge 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 and returned
the cartridges, along with the Chain of Custody 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. 2-Butanone (methyl ethyl ketone) was added to the TO-11A analysis in
2012. Although the sensitivity varies from pollutant-to-pollutant and from site-to-site 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 0.011 ppbv.
2-19
-------
Table 2-5. 2012 Carbonyl Compound Method Detection Limits
Pollutant
Acetaldehyde
Acetone
Benzaldehyde
2-Butanone
Butyraldehyde
Crotonaldehyde
2,5-Dimethylbenzaldehyde
Formaldehyde
Hexaldehyde
Isovaleraldehyde
Propionaldehyde
Tolualdehydes1
Valeraldehyde
Minimum
MDL
(ppbv)
0.003
0.005
0.002
0.002
0.002
0.002
0.002
0.005
0.001
0.001
0.002
0.002
0.002
Maximum
MDL
(ppbv)
0.0212
0.0282
0.010
0.014
0.010
0.010
0.010
0.03 12
0.007
0.007
0.010
0.017
0.010
Average
MDL
(ppbv)
0.007
0.009
0.003
0.004
0.003
0.003
0.003
0.010
0.002
0.002
0.003
0.006
0.003
the analytical method reports only the sum concentration for these three isomers and
not the individual concentrations.
Indicates that sample dilution was required to perform analysis.
2.2.3 PAH Sampling and Analytical Method
PAH sampling and analysis was performed in accordance with 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 and
Chain of Custody 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 1 microliter (uL) 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 site-to-site due to the different volumes pulled through the samples, the
average detection limit for valid samples reported by the ERG laboratory range from
0.034 ng/m3 (acenaphthylene) to 0.199 ng/m3 (naphthalene).
2-20
-------
Table 2-6. 2012 PAH Method Detection Limits
Pollutant
Acenaphthene
Acenaphthylene
Anthracene
Benzo(a)anthracene
Benzo(a)pyrene
Benzo(b)fluoranthene
Benzo(e)pyrene
Benzo(g,h,i)perylene
Benzo(k)fluoranthene
Chrysene
Coronene
Cyclopenta[cd]pyrene
Dibenz(a,h)anthracene
Fluoranthene
Fluorene
9-Fluorenone
Indeno( 1 ,2,3 -cd)pyrene
Naphthalene
Perylene
Phenanthrene
Pyrene
Retene
Minimum
MDL
(ng/m3)
0.032
0.021
0.028
0.035
0.043
0.033
0.039
0.037
0.042
0.032
0.040
0.054
0.036
0.057
0.030
0.036
0.042
0.115
0.043
0.030
0.055
0.101
Maximum
MDL
(ng/m3)1
0.555
0.368
0.492
0.605
0.754
0.565
0.675
0.636
0.737
0.556
0.704
0.934
0.631
0.983
0.522
0.631
0.731
2.40
0.753
0.516
0.964
1.76
Average
MDL
(ng/m3)
0.051
0.034
0.046
0.056
0.070
0.052
0.063
0.059
0.068
0.052
0.065
0.087
0.059
0.091
0.048
0.059
0.068
0.197
0.070
0.048
0.090
0.163
Indicates that sample dilution was required to perform analysis.
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 provides 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 Chain of Custody forms and all associated documentation, to the ERG laboratory
for analysis.
Extraction and analysis for the determination of metals in or on particulate matter was
performed in accordance with EPA Compendium Method IO-3.5 and EPA FEM Method
"Standard Operating Procedure for the Determination of Lead in PMio (or TSP) by Inductively
2-21
-------
Coupled Plasma Mass Spectrometry (ICP-MS) with Hot Block Dilute Acid and Hydrogen
Peroxide Filter Extraction" (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 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 the metals samples. Due to the difference in
sample volume/filter collection media, there are two sets of MDLs listed in Table 2-7. Although
the sensitivity varies from pollutant-to-pollutant and from site-to-site due to the different
volumes pulled through the samples, the average MDL for valid samples ranges from
0.004 ng/m3 (beryllium) to 2.31 ng/m3 (chromium) for the quartz filters and from 0.010 ng/m3
(cadmium) to 24.45 ng/m3 (chromium) for the Teflon® filters.
Table 2-7. 2012 Metals Method Detection Limits
Pollutant
Minimum
MDL
(ng/m3)
Maximum
MDL
(ng/m3)
Average
MDL
(ng/m3)
8" X 10" Quartz Filters
Antimony
Arsenic
Beryllium
Cadmium
Chromium
Cobalt
Lead
Manganese
Mercury
Nickel
Selenium
0.008
0.036
0.003
0.004
1.58
0.017
0.066
0.081
0.005
0.253
0.010
0.028
0.5521
0.005
0.008
2.84
0.111
18.91
0.395
1.251
0.494
0.2191
0.023
0.058
0.004
0.006
2.31
0.090
0.135
0.320
0.009
0.402
0.023
Pollutant
Minimum
MDL
(ng/m3)
Maximum
MDL
(ng/m3)
Average
MDL
(ng/m3)
47mm Teflon® Filters
Antimony
Arsenic
Beryllium
Cadmium
Chromium
Cobalt
Lead
Manganese
Mercury
Nickel
Selenium
0.260
0.160
0.010
0.010
22.5
0.030
0.070
0.300
0.040
0.370
0.320
0.340
0.200
0.020
0.010
28.8
0.040
0.090
0.380
0.050
0.480
0.410
0.283
0.172
0.020
0.010
24.4
0.031
0.071
0.323
0.050
0.404
0.353
Indicates that sample dilution was required to perform analysis.
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 for hexavalent chromium analysis were collected by
passing ambient air through sodium bicarbonate impregnated acid-washed cellulose filters. ERG
prepared and distributed either filters secured in Teflon® cartridges or in petri dishes to the
monitoring sites prior to each scheduled sample collection event. Site operators connected the
2-22
-------
cartridges or installed the filters to the air sampling equipment. After a 24-hour sampling period,
site operators recovered the cartridges and Chain of Custody forms and returned them to the
ERG laboratory for analysis. Upon receipt at the laboratory, the filters were extracted using a
sodium bicarbonate solution. Ion chromatography (1C) analysis and ultraviolet-visible (UV-Vis)
detection of the extracts determined the amount of hexavalent chromium present in each sample.
Although the sensitivity varies from site-to-site due to the different volumes pulled
through the samples, Table 2-8 presents the range and average detection limit (0.0036 ng/m3) for
valid samples reported by the ERG laboratory across the program. Raw data for the hexavalent
chromium method are presented in Appendix H.
Table 2-8. 2012 Hexavalent Chromium Method Detection Limit
Pollutant
Hexavalent Chromium
Minimum
MDL
(ng/m3)
0.0028
Maximum
MDL1
(ng/m3)
0.0335
Average
MDL
(ng/m3)
0.0036
Indicates that sample dilution was required to perform analysis.
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 2012. The first sample date for each site is generally
at the beginning of January and sampling continued through the end of December, although there
were a few exceptions:
• The Oklahoma City, OK site (ADOK) began sampling TSP metals, carbonyl
compounds, and VOCs under the NMP in December 2011. As a result, data from the
five December 2011 samples collected at this site have been included with the 2012
data.
• The Milwaukee, WI (MIWI) and St. Cloud, MN (STMN) monitoring sites began
sampling hexavalent chromium under the NMP in February. Conversely, the Deer
Park, TX (CAMS 35) monitoring site discontinued hexavalent chromium sampling
under the NMP in February.
• Several Kentucky monitoring sites began sampling under the NMP between March
and July 2012. These sites sampled VOCs, carbonyl compounds, and/or PMio metals.
• The Carbondale, CO (RFCO) monitoring site began sampling SNMOC and carbonyl
compounds under the NMP in June.
• Sampling for PAHs at the Decatur, GA (SDGA) monitoring site under the NMP was
discontinued at the end of June.
2-23
-------
• The instrumentation at the New York, NY monitoring site (MONY) was relocated
back to its original NATTS location at PS 52 (BXNY) after the completion of
construction in the area. Monitoring at MONY stopped at the end of June after which
monitoring at BXNY began in July.
• The Vermont monitoring sites (BURVT, RUVT, and UNVT) began sending carbonyl
compound samples to ERG under the NMP in July.
• The Long Beach, CA (LBHCA) monitoring site began sampling PAHs under the
NMP in July.
• The Roxana, IL (ROIL) monitoring site began sampling VOCs and carbonyl
compounds in July.
• Monitoring at the Pry or Creek, OK (PROK) site was discontinued at the end of
October.
According to the NMP schedule, 24-hour integrated samples were collected at each
monitoring site on a l-in-6 day schedule and each 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:
• Prior to July 2012, the Garfield County, CO sites (BMCO, BRCO, PACO, RICO)
collected samples by initiating the samplers manually. Samples were generally
collected from mid-morning of one day to mid-morning of the next. However,
beginning in July 2012, timers were added to the samplers, allowing midnight-to-
midnight sampling. 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, which began sampling in June, was conducted
on a l-in-12 day schedule for both methods.
• The South Phoenix, AZ site (SPAZ) collected VOC samples on a l-in-12 day
schedule.
• The Orlando, FL site (PAFL) collected metals samples on a l-in-12 day schedule.
• The Detroit, MI site (SWMI) collected carbonyl compound samples on a l-in-12 day
schedule.
• The Burlington, VT and Rutland, VT sites (BURVT and RUVT) collected VOC
samples, and later carbonyl compound samples, on a l-in-12 day schedule.
2-24
-------
Table 2-9. 2012 Sampling Schedules and Completeness Rates
Site
ADOK2
ASKY
ASKY-M
ATKY
AZFL
BAKY
BLKY
BMCO
BOMA
BRCO
BTUT
BURVT3
BXNY
CAMS 35
CAMS 85
Monitoring Period1
First
Sample
12/5/11
7/14/12
3/4/12
7/14/12
1/4/12
3/4/12
7/14/12
1/4/12
1/4/12
1/4/12
1/4/12
1/4/12
7/2/12
1/4/12
1/4/12
Last
Sample
12/29/12
12/29/12
12/29/12
12/29/12
12/29/12
12/29/12
12/29/12
12/29/12
12/29/12
12/29/12
12/29/12
12/29/12
12/29/12
12/29/12
12/29/12
Carbonyl
Compounds
A
66
29
59
26
28
54
16
„
B
66
29
61
31
31
61
16
„
C
100
100
97
843
903
89
100
„
VOCs
A
66
29
29
26
„
„
56
31
B
66
29
29
29
„
„
61
31
„
C
100
100
100
90
„
„
92
100
„
Hexavalent
Chromium
A
61
59
„
31
61
59
B
„
61
„
61
„
31
61
61
C
100
97
100
100
97
Metals
A
64
50
50
61
57
„
B
65
51
51
61
60
C
98
98
98
10qq
9558
SNMOCs
A
__
56
„
„
B
61
61
61
„
„
C
87
95
92
„
„
PAHs
A
„
59
„
59
„
22
9
B
61
61
31
10
C
97
97
71
90
to
to
A = Number of valid samples collected.
B = Number of valid samples that should be collected in 2012 based on sample schedule and start/end date of sampling.
C = Completeness (%).
1 Begins with 1st sample collected and ends with last sample collected; date range presented may not be representative of each method-specific date range.
Includes five samples from December 2011.
3 Sampling schedule was a l-in-12 day schedule rather than a l-in-6 schedule.
BOLD ITALICS = EPA-designated NATTS site.
Shading indicates that completeness is below the MQO of 85%.
-------
Table 2-9. 2012 Sampling Schedules and Completeness Rates (Continued)
Site
CCKY
CELA
CHNJ
CHSC
DEMI
ELNJ
GLKY
GPCO
HOW
INDEM
LAKY
LBHCA
LEKY
MIWI
MONY
Monitoring Period1
First
Sample
3/4/12
1/4/12
1/4/12
1/4/12
1/4/12
1/4/12
1/4/12
1/4/12
1/4/12
1/4/12
7/14/12
7/14/12
3/4/12
2/27/12
1/4/12
Last
Sample
12/29/12
12/29/12
12/29/12
12/29/12
12/29/12
12/29/12
12/29/12
12/30/12
12/29/12
12/29/12
12/29/12
12/29/12
12/29/12
12/29/12
6/26/12
Carbonyl
Compounds
A
62
„
60
61
61
61
59
27
B
61
„
61
61
61
61
61
29
C
>100
„
98
100
100
100
97
93
VOCs
A
26
61
„
63
61
61
62
29
29
B
29
61
„
61
61
61
61
29
28
C
90
100
„
>100
100
100
>100
100
>100
Hexavalent
Chromium
A
50
62
61
61
61
„
52
30
B
61
61
61
61
61
„
52
30
C
82
>100
100
100
100
100
100
Metals
A
47
„
59
49
B
51
61
51
C
92
97
96
SNMOCs
A
„
„
B
„
__
„
C
__
„
„
PAHs
A
60
53
60
61
60
26
„
30
B
61
61
61
61
61
29
30
C
98
87
98
100
98
90
100
to
to
A = Number of valid samples collected.
B = Number of valid samples that should be collected in 2012 based on sample schedule and start/end date of sampling.
C = Completeness (%).
1 Begins with 1st sample collected and ends with last sample collected; date range presented may not be representative of each method-specific date range.
2 Includes five samples from December 2011.
Sampling schedule was a l-in-12 day schedule rather than a l-in-6 schedule.
BOLD ITALICS = EPA-designated NATTS site.
Shading indicates that completeness is below the MQO of 85%.
-------
Table 2-9. 2012 Sampling Schedules and Completeness Rates (Continued)
Site
NBIL
NBNJ
OCOK
ORFL
PACO
PAFL3
PROK
PRRI
PXSS
RFCO3
RICO
RIVA
ROCH
ROIL
RRMI
Monitoring Period1
First
Sample
1/4/12
1/4/12
1/4/12
1/4/12
1/4/12
1/10/12
1/4/12
1/1/12
1/4/12
6/8/12
1/4/12
1/4/12
1/4/12
6/8/12
1/4/12
Last
Sample
12/29/12
12/29/12
12/29/12
12/29/12
12/29/12
12/29/12
10/30/12
12/29/12
12/31/12
12/17/12
12/29/12
12/29/12
12/29/12
12/29/12
12/29/12
Carbonyl
Compounds
A
66
60
60
61
27
51
61
15
28
„
35
49
B
61
61
61
61
31
51
61
17
31
„
35
61
C
>100
98
98
100
873
100
100
88
903
„
100
80
VOCs
A
61
60
61
„
51
61
„
„
33
B
61
61
61
51
61
35
C
100
98
100
„
100
100
„
„
94
Hexavalent
Chromium
A
61
„
61
61
61
57
„
B
61
„
„
61
61
„
61
61
„
C
100
100
100
100
93
Metals
A
54
61
30
49
61
„
B
61
61
30
51
61
C
89
100
ioq5
96
100
17
60
SNMOCs
A
61
„
„
„
„
„
B
61
„
„
60
„
17
61
„
„
C
100
„
„
75
„
100
98
„
„
PAHs
A
57
„
„
60
59
„
56
58
„
B
61
61
61
61
61
C
93
98
97
92
95
to
to
A = Number of valid samples collected.
B = Number of valid samples that should be collected in 2012 based on sample schedule and_start/end date of.sampling^
C = Completeness (%).
1 Begins with 1st sample collected and ends with last sample collected; date range presented may not be representative of each method-specific date range.
Includes five samples from December 2011.
3 Sampling schedule was a l-in-12 day schedule rather than a l-in-6 schedule.
BOLD ITALICS = EPA-designated NATTS site.
Shading indicates that completeness is below the MQO of 85%.
-------
Table 2-9. 2012 Sampling Schedules and Completeness Rates (Continued)
Site
RUCA
RUVT3
S4MO
SDGA
SEWA
SJJCA
SKFL
SPAZ3
SPIL
SSSD
STMN
SWMI3
SYFL
TMOK
TOOK
Monitoring Period1
First
Sample
1/4/12
1/16/12
1/4/12
1/4/12
1/4/12
1/4/12
1/4/12
1/10/12
1/4/12
1/4/12
2/9/12
1/10/12
1/4/12
1/4/12
1/4/12
Last
Sample
12/29/12
12/29/12
12/29/12
12/29/12
12/29/12
12/29/12
12/29/12
12/23/12
12/29/12
12/29/12
12/29/12
12/23/12
12/29/12
12/29/12
12/29/12
Carbonyl
Compounds
A
16
61
60
59
61
58
30
60
59
61
B
16
61
61
61
61
61
30
61
61
61
C
100
100
98
97
100
95
100
98
97
100
VOCs
A
31
58
60
30
60
61
61
60
B
30
61
61
30
61
61
61
61
C
>100
95
98
100
98
100
100
98
Hexavalent
Chromium
A
60
54
60
60
„
54
62
B
61
61
61
61
„
55
„
61
C
98
89
98
98
98
>100
Metals
A
61
59
61
61
61
B
61
61
61
61
61
C
100
97
100
61
100
100
SNMOCs
A
„
B
61
„
C
__
100
„
PAHs
A
61
60
29
59
59
61
„
„
59
B
61
61
30
61
61
61
61
C
100
98
97
97
97
100
97
to
to
oo
A = Number of valid samples collected.
B = Number of valid samples that should be collected in 2012 based on sample schedule and start/end date of_sampling.
C = Completeness (%).
1 Begins with 1st sample collected and ends with last sample collected; date range presented may not be representative of each method-specific date range.
Includes five samples from December 2011.
3 Sampling schedule was a l-in-12 day schedule rather than a l-in-6 schedule.
BOLD ITALICS = EPA-designated NATTS site.
Shading indicates that completeness is below the MQO of 85%.
-------
Table 2-9. 2012 Sampling Schedules and Completeness Rates (Continued)
Site
TVKY
UNVT
WADC
WPIN
Monitoring Period1
First
Sample
7/14/12
1/4/12
1/4/12
1/4/12
Last
Sample
12/29/12
12/29/12
12/29/12
12/29/12
Carbonyl
Compounds
A
31
„
58
B
30
„
61
C
>100
„
95
VOCs
A
28
61
„
—
B
29
60
C
97
>100
„
—
Hexavalent
Chromium
A
61
61
B
61
61
—
C
100
100
Metals
A
61
„
B
61
C
100
SNMOCs
A
„
—
B
„
—
C
„
—
PAHs
A
58
61
—
B
61
61
C
95
100
A = Number of valid samples collected.
B = Number of valid samples that should be collected in 2012 based on sample schedule and.start/end date ofsampling.
C = Completeness (%).
1 Begins with 1st sample collected and ends with last sample collected; date range presented may not be representative of each method-specific date range.
2 Includes five samples from December 2011.
Sampling schedule was a l-in-12 day schedule rather than a l-in-6 schedule.
BOLD ITALICS = EPA-designated NATTS site.
to Shading indicates that completeness is below the MQO of 85%.
to
VO
-------
Table 2-9 shows the following:
• 30 sites collected VOC samples.
• 37 sites collected carbonyl compound samples.
• 8 sites collected SNMOC samples.
• 25 sites collected PAH samples.
• 19 sites collected metals samples.
• 25 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. A duplicate sample is a sample collected simultaneously
with a primary sample using the same sampling system (i.e., two separate samples through the
same sampling system at the same time). Collocated samples are samples collected
simultaneously using two independent collection systems at the same location at the same time.
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 for 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.
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
2-30
-------
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, 2012). The data in Table 2-9 show that five datasets from a total of 144
datasets from the 2012 NMP monitoring effort did not meet this MQO (shaded cells in
Table 2-9):
• Sampler issues at PACO resulted in an SNMOC completeness less than 85 percent.
Similarly, sampler issues at BMCO resulted in a carbonyl compound completeness
less than 85 percent.
• The PAH sampler at the BXNY monitoring site sustained damage during Hurricane
Sandy in late October 2012. The PAH sampler was back on-line by early December
2012.
• Intermittent sampler issues throughout much of 2012 resulted in a hexavalent
chromium completeness less than 85 percent for CHSC.
• A collection error at RRMI resulted in the invalidation of carbonyl compound
samples between May 15, 2012 and July 8, 2012.
Although the completeness for S4MO's VOCs is 95 percent, it should be noted that the
Missouri Department of Natural Resources discovered a sampler contamination issue and
invalidated all of its acrylonitrile results for this site through the end of October 2012. Similarly,
the Kentucky Department of Environmental Protection invalidated all of its acrylonitrile and
carbon disulfide data for GLKY through the end of September 2012. These issues are discussed
in more detail in the individual state sections.
Appendix I identifies samples that were invalidated and lists the reason for invalidation,
based on the applied AQS null code.
Table 2-10 presents method-specific completeness. Method-specific completeness was
greater than 90 percent for all six methods performed under the 2012 NMP and ranged from
92.8 percent for SNMOCs to 99.0 percent for VOCs.
2-31
-------
Table 2-10. Method Completeness Rates for 2012
Method
voc
SNMOC
Carbonyl Compounds
PAH
Metals Analysis
Hexavalent Chromium
#of
Valid
Samples
1,466
411
1,796
1,296
1,056
1,421
#of
Samples
Scheduled
1,481
443
1,846
1,350
1,081
1,449
Method
Completeness
(%)
99.0
92.8
97.3
96.0
97.7
98.1
Minimum
Site-Specific
Completeness
(%)
90
(BLKY)
75
(PACO)
80
(RRMI)
71
(BXNY)
89
(NBIL)
82
(CHSC)
Maximum
Site-Specific
Completeness
(%)
>100
(5 sites)
100
(3 sites)
>100
(3 sites)
100
(5 sites)
>100
(2 sites)
>100
(2 sites)
BOLD ITALICS = EPA-designated NATTS site.
2-32
-------
3.0 Summary of the 2012 National Monitoring Programs Data Treatment and Methods
This section summarizes the data treatment
and approaches used to evaluate the measurements
generated from samples collected during the 2012
NMP sampling year. These data were analyzed on
a program-wide basis as well as a site-specific
basis.
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 29.
A total of 233,600 valid air toxics concentrations (including non-detects, duplicate
analyses, replicate analyses, and analyses for collocated samples) were produced from 9,686
valid samples collected at 64 monitoring sites during the 2012 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
VOCs
SNMOCs
Carbonyl Compounds
PAHs
Metals
Hexavalent Chromium
Number
of Sites
30
8
37
25
19
25
A
Raw Data
C
D
E
F
G
H
ppendix
Statistical Summary
J
K
L
M
N
0
3.1 Approach to Data Treatment
This section examines the various statistical tools employed to characterize the data
collected during the 2012 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.
3-1
-------
Most monitoring sites collected duplicate or collocated samples on 10 percent of sample
days, as discussed in Section 2.3. At the laboratory, these duplicate or collocated samples were
then analyzed in replicate. Replicate measurements are repeated analyses performed on a
duplicate or collocated pair of samples. In the event duplicate or collocated collection events
were not possible at a given monitoring site, additional replicate samples were run on individual
samples to provide an indication of analytical precision. For each monitoring site with primary,
duplicate (or collocated), and replicate measurements, the results were averaged together for
each pollutant in order to calculate a single concentration per sample date and method. This is
referred to as the preprocessed daily measurement.
Concentrations of w,/>-xylene and o-xylene were summed together and are henceforth
referred to as "total xylenes," "xylenes (total)," 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-xy\ene and o-xylene species. This is also true of the Data Quality
section (Section 30).
For the 2012 NMP, where statistical parameters are calculated based on the preprocessed
daily measurements, zeros have been substituted for non-detect results. In past reports, the
substitution of zeros was applied only to risk-related analyses; however, beginning with the 2010
NMP report, the substitution of zeros was applied to all analyses. 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 associated reports, such as 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 (ug/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 units of measure associated with the
3-2
-------
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.
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 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 concentrations from January, February, and March; the
second quarter includes April, May, and June; the third quarter includes July, August, and
September; and the fourth quarter includes October, November, and December. A minimum of
75 percent of the total number of samples possible within a given quarter must be valid to have a
quarterly average presented. For sites sampling on a l-in-6 day sampling schedule, 12 samples
represents 75 percent; for sites sampling on a l-in-12 day schedule, six samples represents
75 percent. Sites that do not meet these minimum requirements 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 includes all measured detections and substituted zeros for non-detects
for a given calendar year (2012). 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, was 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. 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 to 1.50 |ig/m3 (EPA, 201 la).
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. In past NMP annual reports,
health risk-based calculations have been used to identify "pollutants of interest." For the 2012
NMP report, the pollutants of interest are also based on risk potential. 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, 201 Ib). Human
health risk can be 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, 2010). 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, 2013e).
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 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, 201 Ib). 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, 2010). The preliminary risk-based screening process provided in this report is an adaption
3-4
-------
of that approach and is a risk-based methodology for analysts and interested parties to identify
which pollutants may pose a risk in their area. Cancer UREs and noncancer RfCs are used as
screening values. Not all pollutants analyzed under the NMP have screening values; of the
pollutants sampled under the NMP, 71 pollutants have screening values in the guidance
document. 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 the risk screening value.
Concentrations that are greater than the risk screening value are described as "failing
the screen."
3. The number of failed screens was summed for each applicable pollutant.
4. The percent contribution of the number of failed screens to the total number of failed
screens program-wide was calculated for each applicable pollutant.
5. The pollutants contributing to the top 95 percent of the total failed screens were
identified as pollutants of interest.
In regards to Step 5 above, the actual cumulative contribution may exceed 95 percent in
order to include all pollutants contributing to the minimum 95 percent criteria (refer to
acenaphthene in Table 4-7 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 comes from 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, 2013f).
3-5
-------
Laboratory analysts have indicated that acetonitrile values may be artificially high (or
non-existent) due to site conditions and potential cross-contamination with concurrent sampling
of carbonyl compounds using Method TO-11A. The inclusion of acetonitrile in data analyses
must be determined on a site-specific basis by the agency responsible for the site. Thus,
acetonitrile results are excluded from certain program-wide and site-specific data analyses,
particularly those related to risk.
Laboratory analysts have indicated that acrylonitrile and carbon disulfide values may also
be artificially high due to potential contamination of the samplers using Method TO-15. The
inclusion of acrylonitrile and carbon disulfide 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. In past reports, these pollutants were considered pollutants of interest by default,
although this has changed for the 2012 report.
Acrolein was excluded from the preliminary risk-based screening process due to
questions about the consistency and reliability of the measurements (EPA, 2013g). Thus, the
results from sampling and analysis of this pollutant have been excluded from any risk-related
analyses presented in this report, similar to acetonitrile, acrylonitrile, and carbon disulfide.
-------
Table 3-2. NATTS MQO Core Analytes
Pollutant
Acrolein
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
Acetaldehyde
Formaldehyde
Naphthalene
Benzo(a)pyrene
Arsenic
Beryllium
Cadmium
Manganese
Lead
Nickel
Hexavalent chromium
Class/Method
VOCs/TO-15
Carbonyl Compounds/
TO-11A
PAHs/TO-13A
Metals/IO-3.5
Metals/ASTMD7614
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 NMP are not required to have their samples
analyzed by ERG or may measure analytes 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.
3.3 Noncancer Risk-Based Screening Evaluation Using Minimum Risk Levels
In addition to the preliminary risk-based screening process described above, a second
risk-based screening was conducted using the Agency for Toxic Substances and Disease Registry
(ATSDR) Minimal Risk Level (MRL) health risk benchmarks (ATSDR, 2013a). This screening
is simply informational and was not used to identify any additional pollutants of interest. An
MRL is a concentration of a hazardous substance that is "likely to be without appreciable risk of
adverse noncancer health effects over a specified duration of exposure," similar to EPA's RfCs
(ATSDR, 2013b). MRLs are intended to be used as screening tools, similar to the preliminary
risk-based screening process discussed above, although "exposure to a level above the MRL does
5-7
-------
not mean that adverse health effects will occur" (ATSDR, 2013b). ATSDR defines MRLs for
three durations of exposure: acute, intermediate, and chronic exposure. Acute risk results from
exposures of 1 to 14 days; intermediate risk results from exposures of 15 to 364 days; and
chronic risk results from exposures of 1 year or greater (ATSDR, 2013a). MRLs, as published by
ATSDR, are presented in parts per million (ppm) for gases and milligrams per cubic meter
(mg/m3) for particulates. The MRLs used in this report have been converted to |ig/m3, have one
significant figure, and are presented in Appendix Q.
For this risk-based screening evaluation, the preprocessed daily measurements were
compared to acute MRLs; quarterly averages were compared to intermediate MRLs; and annual
averages were compared to chronic MRLs. Section 4.2.2 presents the number of preprocessed
daily measurements, quarterly averages, and/or annual averages that are greater than their
respective MRL for each pollutant, summed to the program level. The number of site-specific
concentrations and/or time period averages that are greater than their respective MRLs is also
expanded upon in the individual state sections.
3.4 Additional Program-Level Analyses of the 2012 National Monitoring Programs
Dataset
This section summarizes additional analyses performed on the 2012 NMP dataset at the
program level. Additional program-level analyses include an examination of the potential effect
of motor vehicles and a review of how concentrations vary among the sites themselves and from
quarter-to-quarter. The results of these analyses are presented in Sections 4.3 and 4.4.
3.4.1 The Effect of Mobile Source Emissions on Spatial Variations
Mobile source emissions contribute significantly to air pollution. "Mobile sources" are
emitters of air pollutants that are capable of moving from place to place; mobile sources include
both on-road (i.e., passenger vehicles) and non-road emissions (i.e., lawnmowers). Pollutants
found in motor vehicle exhaust generally result from incomplete combustion of vehicle fuels.
Although modern vehicles and, more recently, vehicle fuels have been engineered to minimize
air emissions, all motor vehicles with internal combustion engines emit a wide range of
pollutants. The magnitude of these emissions primarily depends on the volume of traffic, while
the chemical profile of these emissions depends more on vehicle design and fuel formulation.
-------
This report uses a variety of parameters to quantify and relate motor vehicle emissions to
ambient air quality, which are discussed further in Section 4.3:
• Emissions data from the NEI
• Total hydrocarbon concentrations
• Motor vehicle ownership data
• Estimated daily traffic volume
• Vehicle miles traveled (VMT).
This report uses Pearson correlation coefficients to measure the degree of correlation
between two variables, such as the ones listed above. By definition, Pearson correlation
coefficients always lie between -1 and +1. Three qualification statements apply:
• A correlation coefficient of-1 indicates a perfectly "negative" relationship, indicating
that increases in the magnitude of one variable are associated with proportionate
decreases in the magnitude of the other variable, and vice versa.
• A correlation coefficient of+1 indicates a perfectly "positive" relationship, indicating
that the magnitudes of two variables both increase and both decrease proportionately.
• Data that are completely uncorrelated have Pearson correlation coefficients of 0.
Therefore, the sign (positive or negative) and magnitude of the Pearson correlation coefficient
indicate the direction and strength, respectively, of data correlations. In this report, correlation
coefficients greater than or equal to 0.50 and less than or equal to -0.50 are classified as strong,
while correlation coefficients less than 0.50 and greater than -0.50 are classified as weak.
The number of observations used in a calculation is an important factor to consider when
analyzing the correlations. A correlation using relatively few observations may skew the
correlation, making the degree of correlation appear higher (or lower) than it may actually be.
Thus, in this report, a minimum of five data points must be available to present a correlation.
3.4.2 Variability Analyses
Variability refers to the degree of difference among values in a dataset. Three types of
variability are analyzed for this report. The first type examines the coefficient of variation (CV)
for each of the program-level pollutants of interest across the program sites. The CV provides a
relative measure of variability by expressing the standard deviation to the magnitude of the
arithmetic mean for each of the program-level pollutants of interest, as identified in Section 4.2.
3-9
-------
It is particularly useful when comparing different sets of data because it is unitless (Pagano, P.
and Gauvreau, K., 2000). In this report, variability across data distributions for different sites and
different pollutants are compared. The CVs are shown in the form of scatter plots, where data
points represent the CV and a trend line is plotted to show linearity. In addition, the "R2" value is
also shown on each scatter plot. R2 is the coefficient of determination and is an indicator of how
dependant one variable is on the other. If R2 is equal to 1.0, the data exhibit perfect linearity; the
lower R2, the less dependent the variables are each other (Pagano, P. and Gauvreau, K., 2000).
Pollutants of interest whose data points are clustered together indicate uniformity in how the
concentrations are dispersed among the sites. This suggests that concentrations are affected by
typical and consistent sources (e.g., mobile sources). Data points that are not clustered suggest
the likelihood of a stationary source not typically found in most urban areas (e.g., coke
manufacturing facility). An example of a CV scatter plot is shown in Figure 4-la.
The second type of variability assessed in this report is inter-site variability and is paired
with the CV analysis in Section 4.4. 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 are discussed in Section 3.1 and sites that do not meet these requirements do
not have an annual average concentration presented. This assessment allows the reader to
visualize how concentrations varied across the sites for a particular pollutant of interest. In order
to further this analysis, the program-level average concentrations, as presented in Tables 4-1
through 4-6 in Section 4.1, are plotted against the site-specific annual averages. This allows the
reader to see how the site-specific annual averages compared to the program-level average for
each pollutant. An example of an inter-site variability bar graph is shown in Figure 4-lb. 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 third 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 averages are illustrated by bar graphs for each program-level pollutant
of interest. An example of a quarterly variability bar graph is shown in Figure 4-16. This analysis
3-10
-------
allows for a determination of a quarterly (or seasonal) correlation with the magnitude of
concentrations for a specific pollutant.
3.4.3 Greenhouse Gas Assessment
Currently, there is considerable discussion about climate change among atmospheric and
environmental scientists. Climate change refers to an extended period of change in
meteorological variables used to determine climate, such as temperature and precipitation.
Researchers are typically concerned with greenhouse gases (GHGs), which are those that cause
heat to be retained in the atmosphere (EPA, 2013h).
Agencies researching the effects of greenhouse gases tend to concentrate primarily on
tropospheric levels of these gases. The troposphere is the lowest level of the atmosphere, whose
height varies depending on season and latitude. This is also the layer in which weather
phenomenon occur (NOAA, 2013). A few VOCs measured with Method TO-15 are greenhouse
gases, although these measurements reflect the concentration at the surface, or in the breathing
zone, and do not represent the entire troposphere. Section 4.5 presents the 10 GHGs currently
measured with Method TO-15, their 100-year Global Warming Potential (GWP), and the average
concentration across the NMP program. GWP is a way to determine a pollutant's ability to retain
heat relative to carbon dioxide, which is one of the predominant anthropogenic GHGs in the
atmosphere; higher GWPs indicate a higher potential contribution to global warming (EPA,
20131). In the future, additional GHGs may be added to the NMP Method TO-15 target pollutant
list in order to assess their surface-level ambient concentrations.
3.5 Additional Site-Specific Analyses
In addition to many of 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 29).
3.5.1 Site Characterization
For each site participating in the 2012 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 population, vehicle
3-11
-------
registration, traffic data, and other characterizing information. For the 2012 NMP report, the
locations of point sources located near the monitoring sites were obtained from Version 1 of the
2011 NEI (EPA, 2013c). Sources for other site-characterizing data are provided in the individual
state sections.
3.5.2 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, an
overview of the general climatology is provided, based on the area where each site in located, to
give readers a general idea of what types of meteorological conditions likely affect the site. Next,
the average (or mean) for several meteorological parameters (such as temperature and relative
humidity) are provided. Two averages are presented for each parameter, one average for all days
in 2012 and one average for sample days only. These two averages allow for the determination
of how meteorological conditions on sample days varied from typical conditions experienced
throughout the year. These averages are 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) (NCDC, 2011 and 2012). Although some monitoring
sites have meteorological instruments on-site and report these data to AQS, NWS data were
chosen for this analysis for several reasons:
• Some sites do not have meteorological instruments on-site.
• Some sites collect meteorological data but do not report them to AQS; thus, they are
not readily available.
• There are differences among the sites in the meteorological parameters reported to
AQS.
Although there are limitations to using NWS data, the data used are standardized and quality-
assured per NWS protocol.
In addition to the climate summary and the statistical calculations performed on
meteorological observations collected near each monitoring site, the following sections describe
additional meteorological analyses that were performed for each monitoring site. These analyses
were performed to further characterize the meteorology at or near each monitoring site and to
determine if the meteorological conditions on days samples were collected were representative of
conditions typically experienced near each site.
3-12
-------
3.5.2.1 Back Trajectory Analysis
For all sites sampling under the NMP for 2012, a back trajectory analysis was conducted.
A back trajectory traces the origin of an air parcel in relation to the location where it is currently
being measured. The method of constructing a back trajectory uses the Lagrangian frame of
reference. In simplest terms, an air parcel can be traced back one hour to a new point of reference
based on the current measured wind speed and direction. At this new point of reference (that is
one hour prior to the current observation), the wind speed and direction are used again to
determine where the air was one hour before. Back trajectory calculations are also governed by
other meteorological parameters, such as pressure and temperature. Each time segment is
referred to as a "time step."
Gridded meteorological data and the model used for back trajectory analyses were
prepared and developed by the National Oceanic and Atmospheric Administration (NOAA)
using data from the NWS and other cooperative agencies. The model used is the Hybrid Single-
Particle Lagrangian Integrated Trajectory (HYSPLIT) model (Draxler, R.R. and Rolph, G.D.,
1997 and 1998; Draxler, R.R., 1999). Back trajectories were computed using the HYPLIT model
to represent four times for each sample day, one at OOZ, 06Z, 12Z, and 18Z. "Z" time is "Zulu
Time" and the same time as UTC (Universal Time Coordinated) or GMT (Greenwich Mean
Time), or the local time at the prime meridian (NOAA, 2013). Although back trajectories can be
modeled for extended periods of time, trajectories were constructed for durations of 24 hours to
match the 24-hour sampling duration. Trajectories are modeled with an initial height of
50 meters above ground level (AGL), and each sample day's back trajectories are plotted to
create a composite back trajectory map. A composite back trajectory map was constructed for
each monitoring site using Geographical Information System (GIS) software. The composite
back trajectory map can be used in the estimation of a 24-hour air shed domain for each site. An
air shed domain is the geographical area surrounding a site from which an air parcel may
typically travel within the 24-hour time frame. Information about the maximum and average
trajectory length may also be provided in reference to the composite back trajectory maps. Note
that the distances provided are straight-line distances, or the length from the site to end point, not
necessarily the length of the actual trajectory. Agencies can use the air shed domain to evaluate
regions where long-range transport may affect their monitoring site.
3-13
-------
In addition to the composite back trajectory map, the HYSPLIT model was used to
perform trajectory cluster analysis. This analysis is a grouping technique that allows the model to
create a subset of trajectories or "clusters" that represent back trajectories originating from
similar locations. For each monitoring site, data from each sample day's back trajectories were
used as input for the cluster analysis program. The model compares the end points between each
trajectory and calculates a spatial variance. Trajectories that are similar to each other have lower
spatial variances while trajectories that are dissimilar have larger spatial variances. The model
then provides the user with information about total spatial variance (TSV) among the trajectories,
which allows the user to determine how many clusters best represent a given group of
trajectories (Draxler, R.R., et. al., 2009). Similar to the composite map, once the cluster
trajectories for each site were computed, a cluster map was constructed for each monitoring site
using GIS software. Both the direction and the distance from the monitoring site are considered
in the clustering process. A minimum of 30 trajectories must be available for the model to run
the cluster analysis. Since four back trajectories were computed for each sample day, a minimum
of 30 sample days was the criteria used to perform the cluster analysis for this report. The cluster
analysis is useful for scientifically and quantitatively determining where air most often originates
for a given location.
3.5.2.2 Wind Rose Analysis
Wind roses were constructed for each site to help identify the predominant direction from
which the wind blows. 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 NWS surface wind data from the nearest weather station (with
sufficient data) into a wind rose software program, WRPLOT (Lakes, 2011). For each site, three
wind roses were constructed: first, historical data were used to construct a wind rose for up to
10 years prior to the current sampling year; second, 2012 data were used to construct a wind rose
presenting wind data for the entire calendar year; and lastly, a wind rose was constructed to
present wind data for sample days only. In addition to the wind roses, a map showing the
distance between the NWS station used and the monitoring site is presented. This allows for
topographical influences on the wind patterns to potentially be identified.
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
3-14
-------
determining whether high concentrations correlate with a specific wind direction. While the
composite back trajectory map shows where air parcels originated on a number of days, the wind
rose shows the frequency at which a given wind speed and direction are measured near the
monitoring site. Thus, the back trajectory analysis focuses on long range transport, while the
wind rose captures day-to-day fluctuations at the surface. Both are used to identify potential
meteorological influences on a monitoring site.
3.5.3 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 concentrations
• Trends Analysis
• Comparison to ATSDR MRLs, as described in Section 3.3, including the emission
tracer analysis described below
• The calculation of cancer risk and noncancer hazard approximations in relation to
cancer and noncancer health effects
• Risk-based emissions assessment.
3.5.3.1 Site-Specific Comparison to Program-level Average Concentrations
To better understand how an individual site's concentrations compare to the program-
level results, as presented in Tables 4-1 through 4-6 of Section 4.1, the site-specific and program-
level concentrations are presented together graphically for the site-specific pollutants of interest
indentified via the risk-based screening process. This analysis is an extension of the analysis
discussed in Section 3.4.2 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 Box
and Whisker Plot Utility (Peltier, 2012). 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 Tables 4-1 and 4-2 in Section 4.1.
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-
3-15
-------
level, the first, second (median), third, and fourth (maximum) quartiles are shown as colored
segments on a "bar" where the color changes indicate 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 5-10. 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 29, and are grouped by pollutant within each state section. This allows for both a "site
vs. program" comparison, and an inter-site comparison for sites within a given state.
3.5.3.2 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-eight
of the 64 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 29).
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.
• There must be a minimum of at least 5 years of consecutive data.
Five individual statistical metrics were used in this analysis and are presented as box and
whisker plots, an example of which can be seen in Figure 5-22. The statistical metrics shown
include the minimum and maximum concentration measured during each year (as shown by the
upper and lower value of the lines extending from the box); the 5th percentile, SOthpercentile (or
median), and 95th percentile (as shown by the y-values corresponding with the bottom, blue line,
or top of the box, respectively); and the average (or mean) concentration (as denoted by the
orange diamond). Each of the five metrics represents all measurements from that 1-year period.
3-16
-------
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 to be presented. For cases where sampling began
mid-year, a minimum of six months of sampling is required. In these cases, the 1-year average is
not provided but the range and quartiles are still presented.
Data used in this analysis were downloaded from EPA's AQS database (EPA, 2013b),
where 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. The 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.
3.5.3.3 Emission Tracer Analysis
The preprocessed daily measurements and time-period average concentrations for each
site-specific pollutant of interest were compared to the ATSDR MRL noncancer health risk
benchmarks in the same fashion described in Section 3.3. To further this analysis, pollution roses
were created for each of the site-specific pollutants of interest that have preprocessed daily
measurements greater than their respective ATSDR acute MRL health benchmark (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.
3.5.3.4 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 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
3-17
-------
Quotient (HQ), which is a unitless value. According to EPA, "If the HQ is calculated to be equal
to or less than 1.0, then no adverse health effects are expected as a result of exposure. If the HQ
is greater than 1.0, then adverse health effects are possible" (EPA, 201 Ib).
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, 2013f). Cancer URE and noncancer RfC toxicity factors
can be applied to the annual averages 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 29).
3.5.3.5 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 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
3-18
-------
concern (100 in-a-million) or potential noncancer effects above the level of concern
(e.g., HQ = 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.
The toxicity-weighted emissions approach consists of the following steps:
1. Obtain HAP emissions data for all anthropogenic sectors from the NEI. For point
sources, sum the process-level emissions to the county-level.
2. Apply the mass extraction speciation profiles to extract metal and cyanide mass. The
only exception is for two chromium species: chromium and chromium compounds.
3. Apply weight to the emissions derived from the steps above based on their toxicity.
The results of the toxicity-weighting process are unitless.
a. To apply weight based on cancer toxicity, multiply the emissions of each
pollutant by its cancer URE.
b. To apply weight based on noncancer toxicity, divide the emissions of each
pollutant by its noncancer RfC.
The PAHs measured using Method TO-13A are a sub-group of Poly cyclic 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 (EPA, 201 Ic). Thus, emissions data and toxicity-weighted emissions for 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. The POM groups are sub-grouped in
Table 3-3 because toxicity research has led to the refining of UREs for certain PAHs (EPA,
2013f). Note that naphthalene emissions are reported to the NEI individually; therefore,
naphthalene is not included in one of the POM Groups. Also note that four pollutants analyzed
by Method TO-13A and listed in Table 3-3 do not have assigned POM Groups.
3-19
-------
Table 3-3. POM Groups for PAHs
Pollutant
Acenaphthene
Acenaphthylene
Anthracene
Benzo(a)anthracene
Benzo(a)pyrene
Benzo(b)fluoranthene
Benzo(e)pyrene
Benzo(g,h,i)perylene
Benzo(k)fluoranthene
Chrysene
Coronene
Cyclopenta[cd]pyrene
Dibenz(a,h)anthracene
Fluoranthene
Fluorene
9-Fluorenone
Indeno( 1 ,2,3 -cd)pyrene
Naphthalene*
Perylene
Phenanthrene
Pyrene
Retene
POM Group
Group 2
Group 2
Group 2
POM
Subgroup
Group 2b
Group 2b
Group 2d
Group 6
Group 5
GroupSa
Group 6
Group 2
Group 2
Group 2b
Group 2b
Group 6
Group 7
NA
NA
Group 5
Group 2
Group 2
GroupSb
Group 2b
Group 2b
NA
Group 6
NA
Group 2
Group 2
Group 2
Group 2b
Group 2d
Group 2d
NA
* Naphthalene emissions are reported to the NEI individually;
therefore, naphthalene is not included in one of the POM Groups.
NA = no POM Group assigned.
3-20
-------
4.0 Summary of the 2012 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 the target pollutants were
detected out of the number of valid samples collected and analyzed. Approximately 53 percent of
the reported measurements (based on the preprocessed daily measurements) were above the
MDLs across the program. The following list provides the percentage of measurements that were
above the MDLs for each analytical method:
• 41.2 percent for VOCs
• 48.8 percent for SNMOCs
• 82.3 percent for carbonyl compounds
• 60.1 percent for PAHs
• 77.8 percent for metals
• 71.5 percent for hexavalent chromium samples.
4-1
-------
Some pollutants were detected in every sample collected while others were infrequently
detected or not detected at all. Similar to previous years' reports, acetaldehyde, formaldehyde,
and acetone had the greatest number of measured detections (1,796), using the preprocessed
daily measurements. These pollutants were reported in every valid carbonyl compound sample
collected (1,796). Eleven VOCs, including acetylene, benzene, carbon tetrachloride, and toluene,
were detected in every valid VOC sample collected (1,466). Ten pollutants, including acetylene,
ethylene, ethane, and propylene, were detected in every valid SNMOC sample collected (411).
Naphthalene, phenanthrene, fluoranthene, and pyrene were detected in every valid PAH sample
collected (1,296). Lead, manganese, and nickel were detected in every valid metal sample
collected (1,056). Hexavalent chromium was detected in 1,019 samples (out of 1,421 samples).
Although NBIL and BTUT have the greatest number of measured detections (6,980 for
NBIL and 6,708 for BTUT), they were also the only two sites that collected samples for all six
analytical methods/pollutant groups. However, the detection rates for these sites (63 percent and
65 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 the lowest percentage of concentrations greater than the MDLs (41.2 percent). A site
measuring only VOCs would be expected to have lower detection rates, such as SPAZ
(52.5 percent).
4-2
-------
Table 4-1. Statistical Summaries of the VOC Concentrations
Pollutant
Acetonitrile
Acetylene
Acrolein
Acrylonitrile3'4
tert-Amyl Methyl Ether
Benzene
Bromochloromethane
Bromodichloromethane
Bromoform
Bromomethane
1,3 -Butadiene
Carbon Bisulfide4
Carbon Tetrachloride
Chlorobenzene
Chloroethane
Chloroform
Chloromethane
Chloroprene
Dibromochloromethane
1 ,2-Dibromoethane
/w-Dichlorobenzene
o-Dichlorobenzene
£>-Dichlorobenzene
#of
Measured
Detections1
1,466
1,466
1,442
206
31
1,466
6
116
167
1,183
1,322
1,408
1,466
111
140
948
1,466
5
604
71
176
206
961
#
of Non-
Detects1
0
0
24
1,167
1,435
0
1,460
1,350
1,299
283
144
13
0
1,355
1,326
518
0
1,461
862
1,395
1,290
1,260
505
Minimum2
(ppbv)
0.033
0.091
0.054
0.018
0.003
0.034
0.008
0.006
0.004
0.007
0.006
0.004
0.018
0.004
0.010
0.010
0.288
0.007
0.001
0.004
0.003
0.002
0.002
Maximum
(ppbv)
437
17.2
10.8
2.47
0.016
1.79
0.012
4.10
0.088
0.141
1.85
16.2
0.781
0.291
0.341
9.37
1.66
0.021
1.42
0.017
0.641
0.107
0.228
Arithmetic
Mean
(ppbv)
10.7
0.892
0.521
0.051
O.001
0.281
0.001
0.010
0.001
0.011
0.049
0.669
0.110
0.002
0.005
0.049
0.570
0.001
0.006
O.001
0.002
0.001
0.011
Median
(ppbv)
0.361
0.579
0.366
0
0
0.216
0
0
0
0.011
0.031
0.021
0.109
0
0
0.019
0.555
0
0
0
0
0
0.007
Mode
(ppbv)
0.112
1.02
0
0
0
0.182
0
0
0
0
0
0.012
0.110
0
0
0
0.544
0
0
0
0
0
0
First
Quartile
(ppbv)
0.178
0.366
0.222
0
0
0.158
0
0
0
0.008
0.018
0.012
0.099
0
0
0
0.506
0
0
0
0
0
0
Third
Quartile
(ppbv)
1.89
0.956
0.601
0
0
0.323
0
0
0
0.013
0.064
0.197
0.118
0
0
0.032
0.610
0
0.006
0
0
0
0.014
Standard
Deviation
(ppbv)
39.8
1.27
0.699
0.179
0.001
0.213
0.001
0.152
0.004
0.010
0.075
1.66
0.030
0.017
0.023
0.350
0.107
0.001
0.057
0.002
0.019
0.006
0.016
2 Excludes zeros for non-detects
3 Because S4MO invalidated all acrylonitrile data through October 2012, the number of measured detections plus the number of non-detects does not equal the
total number of VOC samples collected.
4 Because GLKY invalidated all acrylonitrile and carbon disulfide data through September 24, 2012, the number of measured detections plus the number of non-
detects does not equal the total number of VOC samples collected.
NA = Not applicable for these parameters
-------
Table 4-1. Statistical Summaries of the VOC Concentrations (Continued)
Pollutant
Dichlorodifluoromethane
1 , 1 -Dichloroethane
1 ,2-Dichloroethane
1 , 1 -Dichloroethene
cis- 1 ,2-Dichloroethylene
trans- 1 ,2-Dichloroethylene
Dichloromethane
1 ,2-Dichloropropane
cis- 1 , 3 -Dichloropropene
trans- 1 ,3 -Dichloropropene
Dichlorotetrafluoroethane
Ethyl Aery late
Ethyl tort-Butyl Ether
Ethylbenzene
Hexachloro- 1 ,3 -butadiene
Methyl Isobutyl Ketone
Methyl Methacrylate
Methyl tert-Butyl Ether
w-Octane
Propylene
Styrene
1 , 1 ,2,2-Tetrachloroethane
Tetrachloroethylene
#of
Measured
Detections1
1,466
19
1,282
51
4
38
1,464
1
3
1
1,466
13
220
1,459
186
1,377
139
203
1,366
1,466
1,222
112
1,252
#
of Non-
Detects1
0
1,447
184
1,415
1,462
1,428
2
1,465
1,463
1,465
0
1,453
1,246
7
1,280
89
1,327
1,263
100
0
244
1,354
214
Minimum2
(ppbv)
0.266
0.004
0.009
0.005
0.036
0.007
0.034
0.012
0.015
0.016
0.008
0.004
0.004
0.004
0.002
0.006
0.002
0.005
0.008
0.091
0.006
0.004
0.004
Maximum
(ppbv)
1.17
0.092
4.21
0.109
0.063
0.214
214
0.012
0.020
0.016
0.055
0.071
0.273
0.834
0.019
1.17
0.213
0.089
0.663
43.7
9.14
0.026
0.792
Arithmetic
Mean
(ppbv)
0.502
O.001
0.037
O.001
O.001
0.001
0.727
0.001
0.001
O.001
0.018
O.001
0.011
0.081
0.001
0.038
0.003
0.005
0.057
0.694
0.069
0.001
0.022
Median
(ppbv)
0.498
0
0.019
0
0
0
0.116
0
0
0
0.016
0
0
0.056
0
0.031
0
0
0.041
0.348
0.029
0
0.013
Mode
(ppbv)
0.460
0
0
0
0
0
0.074
0
0
0
0.016
0
0
0.020
0
0
0
0
0
0.288
0
0
0
First
Quartile
(ppbv)
0.469
0
0.015
0
0
0
0.083
0
0
0
0.015
0
0
0.034
0
0.021
0
0
0.026
0.257
0.015
0
0.007
Third
Quartile
(ppbv)
0.529
0
0.023
0
0
0
0.182
0
0
0
0.018
0
0
0.096
0
0.046
0
0
0.070
0.566
0.045
0
0.024
Standard
Deviation
(ppbv)
0.053
0.004
0.180
0.003
0.003
0.006
7.22
0.001
0.001
O.001
0.005
0.003
0.033
0.082
0.002
0.041
0.012
0.013
0.059
2.07
0.347
0.002
0.039
2 Excludes zeros for non-detects
3 Because S4MO invalidated all acrylonitrile data through October 2012, the number of measured detections plus the number of non-detects does not equal the
total number of VOC samples collected.
4 Because GLKY invalidated all acrylonitrile and carbon disulfide data through September 24, 2012, the number of measured detections plus the number of non-
detects does not equal the total number of VOC samples collected.
NA = Not applicable for these parameters
-------
Table 4-1. Statistical Summaries of the VOC Concentrations (Continued)
Pollutant
Toluene
1 ,2,4-Trichlorobenzene
1,1,1 -Trichloroethane
1 , 1 ,2-Trichloroethane
Trichloroethylene
Trichlorofluoromethane
Trichlorotrifluoroethane
1 ,2,4-Trimethylbenzene
1,3,5 -Trimethy Ibenzene
Vinyl chloride
m,p-Xylene
o-Xylene
#of
Measured
Detections1
1,466
120
1,268
38
365
1,466
1,466
1,426
1,305
154
1,462
1,454
#
of Non-
Detects1
0
1,346
198
1,428
1,101
0
0
40
161
1,312
4
12
Minimum2
(ppbv)
0.017
0.003
0.004
0.006
0.005
0.140
0.060
0.006
0.004
0.004
0.007
0.004
Maximum
(ppbv)
5.70
0.062
0.066
0.074
3.25
1.04
0.177
1.05
0.663
3.83
3.42
0.981
Arithmetic
Mean
(ppbv)
0.596
0.001
0.008
O.001
0.009
0.270
0.085
0.076
0.029
0.013
0.218
0.088
Median
(ppbv)
0.362
0
0.008
0
0
0.263
0.083
0.054
0.023
0
0.137
0.059
Mode
(ppbv)
0.092
0
0.008
0
0
0.256
0.079
0
0
0
0.055
0.037
First
Quartile
(ppbv)
0.176
0
0.007
0
0
0.245
0.078
0.031
0.013
0
0.071
0.032
Third
Quartile
(ppbv)
0.737
0
0.010
0
0
0.282
0.090
0.093
0.034
0
0.258
0.107
Standard
Deviation
(ppbv)
0.683
0.005
0.005
0.004
0.093
0.055
0.010
0.080
0.032
0.135
0.270
0.096
Excludes zeros for non-detects
3 Because S4MO invalidated all acrylonitrile data through October 2012, the number of measured detections plus the number of non-detects does not equal the
total number of VOC samples collected.
4 Because GLKY invalidated all acrylonitrile and carbon disulfide data through September 24, 2012, the number of measured detections plus the number of non-
detects does not equal the total number of VOC samples collected.
NA = Not applicable for these parameters
-------
Table 4-2. Statistical Summaries of the SNMOC Concentrations
Pollutant
Acetylene
Benzene3
1,3 -Butadiene3
w-Butane
c/s-2-Butene
/ra«s-2-Butene3
Cyclohexane
Cyclopentane3
Cyclopentene3
w-Decane3
1-Decene
Tw-Diethylbenzene3
£>-Diethylbenzene3
2,2-Dimethylbutane
2,3 -Dimethylbutane
2,3 -Dimethylpentane
2,4-Dimethylpentane
w-Dodecane
1-Dodecene3
Ethane
2-Ethyl-l-butene
Ethylbenzene
Ethylene
/w-Ethyltoluene3
o-Ethyltoluene
#of
Measured
Detections1
411
368
277
409
277
267
402
349
36
356
6
158
77
363
330
405
365
310
89
411
1
376
411
330
206
#
of Non-
Detects1
0
43
134
2
134
144
9
62
375
55
405
253
334
48
81
6
46
101
322
0
410
35
0
81
205
Minimum2
(ppbC)
0.166
0.229
0.048
0.648
0.061
0.065
0.074
0.084
0.094
0.083
0.109
0.039
0.061
0.079
0.078
0.095
0.074
0.056
0.061
2.14
4.62
0.073
0.709
0.067
0.060
Maximum
(ppbC)
13.0
5.74
1.03
113
2.53
2.72
13.0
4.76
1.59
6.50
0.418
6.87
2.41
2.74
3.33
4.32
2.98
2.78
3.04
276
4.62
2.36
13.4
2.26
3.28
Arithmetic
Mean
(ppbC)
1.37
1.38
0.193
12.6
0.191
0.226
1.89
0.505
0.029
0.428
0.003
0.239
0.057
0.436
0.608
0.566
0.359
0.188
0.083
40.1
0.011
0.354
2.79
0.374
0.134
Median
(ppbC)
1.02
1.27
0.140
9.40
0.137
0.160
1.20
0.434
0
0.344
0
0
0
0.386
0.466
0.485
0.335
0.147
0
21.2
0
0.268
2.34
0.284
0.060
Mode
(ppbC)
1.31
0
0
14.8
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
13.8
0
0
2.00
0
0
First
Quartile
(ppbC)
0.659
0.711
0
4.02
0
0
0.274
0.238
0
0.181
0
0
0
0.200
0.148
0.262
0.154
0.066
0
7.99
0
0.145
1.72
0.133
0
Third
Quartile
(ppbC)
1.61
1.93
0.298
16.2
0.238
0.286
3.01
0.699
0
0.541
0
0.242
0
0.592
0.955
0.708
0.496
0.236
0
59.4
0
0.487
3.22
0.516
0.211
Standard
Deviation
(ppbC)
1.25
0.963
0.201
12.5
0.281
0.338
1.97
0.444
0.130
0.475
0.029
0.711
0.187
0.350
0.577
0.430
0.290
0.230
0.253
44.1
0.228
0.306
1.72
0.368
0.239
1 Out of 411 valid samples
2 Excludes zeros for non-detects
3 The number of non-detects includes those samples 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
p-Ethyltoluene
w-Heptane
1-Heptene3
w-Hexane
1-Hexene
c/s-2-Hexene
/ra«s-2-Hexene
Isobutane3
Isobutene/ 1 -Butene3
Isopentane3
Isoprene
Isopropylbenzene
2-Methyl- 1 -butene3
3 -Methyl- 1-butene3
2-Methyl- 1 -pentene
4-Methyl- 1 -pentene
2-Methyl-2 -butene3
Methylcyclohexane3
Methylcyclopentane
2-Methylheptane3
3-Methylheptane
2-Methylhexane3
3 -Methy Ihexane3
2-Methylpentane3
3-Methylpentane
#of
Measured
Detections1
301
408
189
411
131
8
24
410
20
285
291
95
244
25
8
27
253
375
408
289
318
408
334
411
411
#
of Non-
Detects1
110
3
222
0
280
403
387
1
391
126
120
316
167
386
403
384
158
36
3
122
93
3
77
0
0
Minimum2
(ppbC)
0.063
0.077
0.047
0.147
0.077
0.062
0.062
0.332
0.211
0.726
0.070
0.060
0.070
0.088
0.060
0.081
0.074
0.082
0.116
0.068
0.060
0.154
0.096
0.304
0.130
Maximum
(ppbC)
2.65
8.56
2.36
19.7
0.439
0.153
0.234
61.2
5.50
68.5
7.92
0.449
1.96
1.14
0.356
0.169
2.02
23.1
9.65
2.62
2.02
4.88
4.95
17.5
9.45
Arithmetic
Mean
(ppbC)
0.226
1.58
0.267
3.38
0.066
0.002
0.006
9.69
0.157
8.29
0.545
0.034
0.186
0.021
0.003
0.008
0.235
3.34
1.79
0.396
0.319
1.30
1.02
3.49
1.86
Median
(ppbC)
0.175
1.22
0
2.60
0
0
0
6.53
0
3.99
0.186
0
0.150
0
0
0
0.210
1.90
1.40
0.322
0.271
1.12
0.935
3.01
1.65
Mode
(ppbC)
0
2.59
0
11.7
0
0
0
13.8
0
0
0
0
0
0
0
0
0
0
1.74
0
0
1.19
0
1.36
3.10
First
Quartile
(ppbC)
0
0.301
0
0.805
0
0
0
1.83
0
0
0
0
0
0
0
0
0
0.286
0.475
0
0.095
0.593
0.300
1.15
0.643
Third
Quartile
(ppbC)
0.307
2.49
0.496
5.01
0.131
0
0
13.8
0
12.7
0.645
0
0.309
0
0
0
0.370
5.66
2.65
0.655
0.489
1.83
1.55
5.02
2.65
Standard
Deviation
(ppbC)
0.252
1.43
0.410
3.09
0.109
0.014
0.028
10.4
0.814
10.6
0.914
0.068
0.220
0.101
0.024
0.030
0.273
3.64
1.54
0.395
0.294
0.866
0.871
2.78
1.50
1 Out of 411 valid samples
2 Excludes zeros for non-detects
3 The number of non-detects includes those samples 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
w-Nonane
1-Nonene3
w-Octane3
1-Octene
w-Pentane
1-Pentene3
c/s-2-Pentene
/ra«s-2-Pentene
a-Pinene
(3-Pinene3
Propane
w-Propylbenzene
Propylene
Propyne
Styrene3
Toluene
w-Tridecane
1-Tridecene
1 ,2,3 -Trimethylbenzene
1 ,2,4-Trimethylbenzene
1 ,3 ,5-Trimethylbenzene
2,2,3 -Trimethylpentane
2,2,4-Trimethylpentane3
2,3 ,4-Trimethylpentane
w-Undecane
1-Undecene
#of
Measured
Detections1
377
153
404
227
411
361
187
299
116
151
411
218
411
1
91
411
71
5
202
397
228
126
257
313
324
29
#
of Non-
Detects1
34
258
7
184
0
50
224
112
295
260
0
193
0
410
320
0
340
406
209
14
183
285
154
98
87
382
Minimum2
(ppbC)
0.079
0.061
0.071
0.061
0.285
0.089
0.068
0.060
0.077
0.086
1.59
0.064
0.305
0.434
0.138
0.302
0.059
0.083
0.067
0.108
0.069
0.074
0.060
0.060
0.077
0.059
Maximum
(ppbC)
3.45
3.75
6.70
0.922
51.8
16.2
102
2.04
1.91
4.45
248
1.34
7.08
0.434
24.7
36.4
1.40
0.177
1.92
4.02
0.936
1.14
8.53
2.04
4.70
0.727
Arithmetic
Mean
(ppbC)
0.462
0.151
1.00
0.125
6.99
0.389
0.326
0.190
0.115
0.277
28.0
0.125
1.14
0.001
1.15
4.18
0.028
0.001
0.133
0.573
0.156
0.074
0.438
0.209
0.246
0.021
Median
(ppbC)
0.358
0
0.764
0.094
5.72
0.230
0
0.148
0
0
19.4
0.102
0.913
0
0
3.10
0
0
0
0.437
0.129
0
0.215
0.150
0.196
0
Mode
(ppbC)
0
0
0
0
11.2
0
0
0
0
0
28.9
0
1.15
0
0
1.44
0
0
0
0
0
0
0
0
0
0
First
Quartile
(ppbC)
0.170
0
0.280
0
2.22
0.145
0
0
0
0
6.82
0
0.640
0
0
1.42
0
0
0
0.278
0
0
0
0.083
0.107
0
Third
Quartile
(ppbC)
0.617
0.141
1.50
0.210
9.98
0.424
0.141
0.268
0.150
0.448
37.4
0.197
1.34
0
0
4.87
0
0
0.182
0.710
0.268
0.142
0.510
0.265
0.317
0
Standard
Deviation
(ppbC)
0.424
0.384
0.902
0.149
6.28
1.05
5.02
0.214
0.261
0.513
29.9
0.165
0.819
0.021
3.48
4.56
0.103
0.014
0.210
0.489
0.175
0.132
0.757
0.232
0.322
0.089
1 Out of 411 valid samples
2 Excludes zeros for non-detects
3 The number of non-detects includes those
NA = Not applicable for these parameters
samples where no value could be reported due to co-elution.
-------
Table 4-2. Statistical Summaries of the SNMOC Concentrations (Continued)
Pollutant
/w-Xylene/p-Xylene
o-Xylene
SNMOC (Sum of Knowns)
Sum of Unknowns
TNMOC
#of
Measured
Detections1
409
398
411
411
411
#
of Non-
Detects1
2
13
NA
NA
NA
Minimum2
(ppbC)
0.125
0.091
20.8
15.3
11.2
Maximum
(ppbC)
7.39
1.87
793
1,880
1,970
Arithmetic
Mean
(ppbC)
1.41
0.438
153
110
263
Median
(ppbC)
1.19
0.367
116
66.3
206
Mode
(ppbC)
1.99
0
184
116
160
First
Quartile
(ppbC)
0.570
0.217
52.0
42.9
127
Third
Quartile
(ppbC)
2.04
0.601
211
112
332
Standard
Deviation
(ppbC)
1.02
0.295
129
172
210
1 Out of 411 valid samples
2 Excludes zeros for non-detects
3 The number of non-detects includes those samples where no value could be reported due to co-elution.
NA = Not applicable for these parameters
-------
Table 4-3. Statistical Summaries of the Carbonyl Compound Concentrations
Pollutant
Acetaldehyde
Acetone
Benzaldehyde3
2-Butanone3
Butyraldehyde3
Crotonaldehyde3
2,5-Dimethylbenzaldehyde
Formaldehyde
Hexaldehyde3
Isovaleraldehyde
Propionaldehyde
Tolualdehydes3
Valeraldehyde3
#of
Measured
Detections1
1,796
1,796
1,740
1,734
1,780
1,755
o
J
1,796
1,761
9
1,790
1,538
1,719
#
of Non-
Detects1
0
0
56
57
16
41
1,793
0
35
1,787
6
258
77
Minimum2
(ppbv)
0.019
0.067
0.003
0.009
0.005
0.003
0.010
0.020
0.002
0.005
0.002
0.004
0.002
Maximum
(ppbv)
11.3
11.6
0.285
4.30
2.12
1.46
0.072
10.4
0.546
0.154
0.848
0.325
0.305
Arithmetic
Mean
(ppbv)
0.980
1.25
0.029
0.169
0.089
0.106
O.001
2.19
0.031
0.001
0.128
0.025
0.028
Median
(ppbv)
0.795
1.05
0.023
0.131
0.072
0.044
0
1.84
0.023
0
0.106
0.022
0.022
Mode
(ppbv)
1.03
1.16
0.014
0
0.041
0
0
1.10
0.011
0
0.065
0
0
First
Quartile
(ppbv)
0.539
0.665
0.014
0.079
0.046
0.022
0
1.19
0.013
0
0.070
0.013
0.013
Third
Quartile
(ppbv)
1.23
1.55
0.034
0.211
0.106
0.129
0
2.77
0.037
0
0.160
0.034
0.034
Standard
Deviation
(ppbv)
0.712
0.913
0.025
0.170
0.085
0.148
0.002
1.49
0.036
0.004
0.090
0.022
0.025
Out of 1,796 valid samples for all compounds except 2-butanone. The total for 2-butanone is 1,791 due to the five carbonyl compound samples from 201 1 included
with ADOK's data, when 2-butanone was not part of the analytes included with this method.
2 Excludes zeros for non-detects
The number of non-detects includes those samples where no value could be reported due to co-elution.
-------
Table 4-4. Statistical Summaries of the PAH Concentrations
Pollutant
Acenaphthene
Acenaphthylene
Anthracene
Benzo(a)anthracene
Benzo(a)pyrene
Benzo(b)fluoranthene
Benzo(e)pyrene
Benzo(g,h,i)perylene
B enzo (k)fluoranthene
Chrysene
Coronene
Cyclopenta[cd]pyrene
Dibenz(a,h)anthracene
Fluoranthene
Fluorene
9-Fluorenone
Indeno( 1 ,2,3 -cd)pyrene
Naphthalene
Perylene
Phenanthrene
Pyrene
Retene
#of
Measured
Detections1
1,277
655
1,020
950
824
1,179
1,067
1,079
673
1,253
664
194
242
1,296
1,280
1,295
966
1,296
361
1,296
1,296
1,230
#
of Non-
Detects1
19
641
276
346
472
117
229
217
623
43
632
1,102
1,054
0
16
1
330
0
935
0
0
66
Minimum2
(ng/m3)
0.081
0.024
0.028
0.011
0.017
0.019
0.016
0.019
0.010
0.019
0.017
0.014
0.017
0.078
0.338
0.102
0.016
2.61
0.017
0.432
0.046
0.025
Maximum
(ng/m3)
182
14.0
18.9
2.67
3.60
5.93
2.81
1.94
2.06
2.94
0.590
0.718
0.396
42.9
93.4
14.2
2.36
822
0.853
251
17.4
29.6
Arithmetic
Mean
(ng/m3)
5.00
0.567
0.396
0.094
0.086
0.214
0.109
0.113
0.053
0.218
0.038
0.017
0.009
2.34
5.16
1.54
0.097
86.4
0.017
10.1
1.30
0.366
Median
(ng/m3)
2.28
0.034
0.207
0.045
0.038
0.108
0.059
0.060
0.022
0.132
0.020
0
0
1.32
3.08
1.07
0.053
63.3
0
5.68
0.797
0.157
Mode
(ng/m3)
0
0
0
0
0
0
0
0
0
0
0
0
0
1.04
0
1.25
0
101
0
10.3
1.03
0
First
Quartile
(ng/m3)
1.12
0
0.070
0
0
0.050
0.030
0.031
0
0.072
0
0
0
0.745
1.80
0.623
0
35.3
0
3.11
0.454
0.090
Third
Quartile
(ng/m3)
4.67
0.538
0.459
0.100
0.093
0.249
0.131
0.133
0.063
0.259
0.051
0
0
2.51
5.35
1.89
0.115
114
0.023
10.4
1.51
0.332
Standard
Deviation
(ng/m3)
9.87
1.28
0.949
0.178
0.177
0.327
0.165
0.159
0.109
0.262
0.062
0.061
0.026
3.38
7.10
1.52
0.146
78.4
0.043
15.5
1.57
1.05
1 Out of 1,296 valid samples
2 Excludes zeros for non-detects
-------
Table 4-5. Statistical Summaries of the Metals Concentrations
Pollutant
Antimony (PM10)
Arsenic (PM10)
Beryllium (PM10)
Cadmium (PM10)
Chromium (PMi0)
Cobalt (PM10)
Lead (PM10)
Manganese (PM10)
Mercury (PM10)
Nickel (PM10)
Selenium (PM10)
Antimony (TSP)
Arsenic (TSP)
Beryllium (TSP)
Cadmium (TSP)
Chromium (TSP)
Cobalt (TSP)
Lead (TSP)
Manganese (TSP)
Mercury (TSP)
Nickel (TSP)
Selenium (TSP)
#of
Measured
Detections112
754
749
697
755
652
755
760
760
719
760
730
296
296
296
296
296
296
296
296
295
296
296
#
of Non-
Detects1'2
6
11
63
5
108
5
0
0
41
0
30
0
0
0
0
0
0
0
0
1
0
0
Minimum3
(ng/m3)
0.009
0.003
0.00003
0.0005
0.0002
0.0001
0.070
0.190
0.0001
0.070
0.0001
0.061
0.162
0.002
0.029
0.796
0.060
0.562
1.36
0.003
0.288
0.113
Maximu
m (ng/m3)
24.8
7.23
0.550
2.91
20.9
7.26
111
275
0.328
17.3
4.92
7.22
2.28
0.525
2.30
11.0
15.4
304
273
0.088
12.8
2.88
Arithmetic
Mean
(ng/m3)
1.44
0.751
0.014
0.170
2.52
0.202
4.52
10.6
0.017
1.26
0.610
0.712
0.679
0.032
0.195
2.65
0.939
6.54
23.6
0.016
1.42
0.766
Median
(ng/m3)
0.903
0.542
0.007
0.090
2.44
0.100
2.62
5.80
0.010
0.841
0.410
0.520
0.600
0.020
0.144
2.59
0.459
3.75
17.1
0.013
1.08
0.721
Mode
(ng/m3)
0.870
0
0.010
0.060
0
0.060
1.27
10.1
0.020
0.420
0
0.324
0.634
0.017
0.061
1.17
0.330
2.17
20.4
0.011
1.10
0.702
First
Quartile
(ng/m3)
0.510
0.340
0.003
0.057
0.440
0.055
1.53
2.96
0.007
0.460
0.204
0.323
0.413
0.012
0.090
1.65
0.254
2.23
11.4
0.010
0.773
0.434
Third
Quartile
(ng/m3)
1.52
0.879
0.020
0.150
3.70
0.190
4.42
11.2
0.020
1.42
0.840
0.885
0.827
0.032
0.235
3.15
0.926
6.06
27.9
0.019
1.66
0.975
Standard
Deviation
(ng/m3)
1.86
0.739
0.028
0.306
2.10
0.441
7.73
18.0
0.022
1.57
0.594
0.788
0.371
0.055
0.196
1.30
1.49
19.0
27.5
0.011
1.23
0.426
to
1 For PM10, out of 760 valid samples
2 For TSP, out of 296 valid samples
3 Excludes zeros for non-detects
-------
Table 4-6. Statistical Summary of the Hexavalent Chromium Concentrations
Pollutant
Hexavalent Chromium
#of
Measured
Detections1
1,019
#
of Non-
Detects1
402
Minimum2
(ng/m3)
0.0025
Maximum
(ng/m3)
8.51
Arithmetic
Mean
(ng/m3)
0.037
Median
(ng/m3)
0.016
Mode
(ng/m3)
0
First
Quartile
(ng/m3)
0
Third
Quartile
(ng/m3)
0.029
Standard
Deviation
(ng/m3)
0.260
1 Out of 1,421 valid samples
2 Excludes zeros for non-detects
-------
4.1.2 Concentration Range and Data Distribution
The concentrations measured during the 2012 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 acetonitrile, had a wide range of concentrations measured,
while other pollutants, such as dichlorotetrafluoroethane, did not, even though they were both
detected frequently. The pollutant for each method-specific pollutant group with the largest
range in measured concentrations is as follows:
• For VOCs, acetonitrile (0.033 ppbv to 437 ppbv)
• For SNMOCs, ethane (2.14 ppbC to 276 ppbC)
• For carbonyl compounds, acetone (0.067 ppbv to 11.6 ppbv)
• For PAHs, naphthalene (2.61 ng/m3 to 822 ng/m3)
• For metals in PMio, manganese (0.190 ng/m3 to 275 ng/m3)
• For metals in TSP, lead (0.562 ng/m3 to 304 ng/m3)
• For hexavalent chromium, 0.0025 ng/m3 to 8.51 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, median, mode, first and third quartiles, and standard deviation) for each of the
pollutants sampled during the 2012 NMP in respective pollutant group units. A multitude of
observations can be made from these tables. The pollutants with the three highest average
concentrations, by mass, for each pollutant group are provided below, with respective confidence
intervals (although the 95 percent confidence interval is not provided in the table).
The top three VOCs by average mass concentration, as presented in Table 4-1, are:
• Acetonitrile (10.7 ±2.04 ppbv)
• Acetylene (0.892 ± 0.065 ppbv)
• Dichloromethane (0.727 ± 0.370 ppbv).
4-14
-------
The top three SNMOCs by average mass concentration, as presented in Table 4-2, are:
• Ethane (40.1± 4.27 ppbC)
• Propane (28.0 ± 2.90 ppbC)
• w-Butane (12.6 ± 1.21 ppbC).
The top three carbonyl compounds by average mass concentration, as presented in
Table 4-3, are:
• Formaldehyde (2.19 ±0.069 ppbv)
• Acetone (1.25 ± 0.042 ppbv).
• Acetaldehyde (0.980 ± 0.033 ppbv)
The top three PAHs by average mass concentration, as presented in Tables 4-4, are:
• Naphthalene (86.4 ± 4.27 ng/m3)
• Phenanthrene (10.1 ± 0.845 ng/m3)
• Fluorene(5.16±0.387ng/m3).
The top three metals by average mass concentration for both PMio and TSP fractions, as
presented in Table 4-5, are;
• Manganese (PMio = 10.6 ± 1.28 ng/m3, TSP = 23.6 ± 3.15 ng/m3)
•
•
3 3
Lead (PMio = 4.52 ± 0.551 ng/m, TSP = 6.54 ± 2.18 ng/m)
3 3
Total chromium (PMio = 2.52 ± 0.150 ng/m, TSP = 2.65 ± 0.148 ng/m).
The average mass concentration of hexavalent chromium, as presented in Table 4-6, is
0.037 ±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.
4.2 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.
4-15
-------
Table 4-7. Results of the Program-Level Preliminary Risk-Based Screening Process
Pollutant
Formaldehyde
Acetaldehyde
Benzene
Carbon Tetrachloride
1,3 -Butadiene
1 ,2-Dichloroethane
Naphthalene
Arsenic
Manganese
Ethylbenzene
£>-Dichlorobenzene
Nickel
Fluorene
Hexachloro- 1 ,3 -butadiene
Acenaphthene
1 , 1 ,2,2-Tetrachloroethane
1 ,2-Dibromoethane
Hexavalent Chromium
Propionaldehyde
Cadmium
Lead
Vinyl chloride
Trichloroethylene
Fluoranthene
Benzo(a)pyrene
Dichloromethane
1,1,2-Trichloroethane
Xylenes
Chloroprene
Beryllium
Chloroform
Acenaphthylene
Antimony
Benzo(b)fluoranthene
Bromoform
Bromomethane
Cobalt
Tetrachloroethylene
Total
Screening
Value
(Hg/m3)
0.077
0.45
0.13
0.17
0.03
0.038
0.029
0.00023
0.005
0.4
0.091
0.0021
0.011
0.045
0.011
0.017
0.0017
0.000083
0.8
0.00056
0.015
0.11
0.2
0.011
0.00057
7.7
0.0625
10
0.0021
0.00042
9.8
0.011
0.02
0.0057
0.91
0.5
0.01
3.8
#of
Failed
Screens
,792
,711
,693
,464
,375
,280
,013
944
706
399
321
146
135
130
127
112
71
64
59
52
50
49
46
36
28
24
14
7
5
4
o
J
1
1
1
1
1
1
1
13,867
#of
Measured
Detections
1,796
1,796
1,695
1,466
1,473
1,282
1,296
1,045
1,056
1,675
961
1,056
1,280
186
1,277
112
71
1,019
1,790
1,051
1,056
154
365
1,296
824
1,464
38
1,696
5
993
948
655
1,050
1,179
167
1,183
1,051
1,252
38,759
%of
Failed
Screens
99.78
95.27
99.88
99.86
93.35
99.84
78.16
90.33
66.86
23.82
33.40
13.83
10.55
69.89
9.95
100.00
100.00
6.28
3.30
4.95
4.73
31.82
12.60
2.78
3.40
1.64
36.84
0.41
100.00
0.40
0.32
0.15
0.10
0.08
0.60
0.08
0.10
0.08
35.78
%of
Total
Failures
12.92
12.34
12.21
10.56
9.92
9.23
7.31
6.81
5.09
2.88
2.31
1.05
0.97
0.94
0.92
0.81
0.51
0.46
0.43
0.37
0.36
0.35
0.33
0.26
0.20
0.17
0.10
0.05
0.04
0.03
0.02
0.01
0.01
0.01
0.01
0.01
0.01
0.01
Cumulative
%
Contribution
12.92
25.26
37.47
48.03
57.94
67.17
74.48
81.29
86.38
89.26
91.57
92.62
93.60
94.53
95.45
96.26
96.77
97.23
97.66
98.03
98.39
98.75
99.08
99.34
99.54
99.71
99.81
99.86
99.90
99.93
99.95
99.96
99.96
99.97
99.98
99.99
99.99
100.00
4-16
-------
The results in Table 4-7 are listed in descending order by number of screens failed.
Table 4-7 shows that formaldehyde failed the greatest number of screens (1,792), although
acetaldehyde and benzene were not far behind (1,711 and 1,693, respectively). These three
pollutants were also among those with the greatest number of measured detections. Conversely,
seven pollutants listed in Table 4-7 failed only one screen each. The number of measured
detections for these seven pollutants varied significantly. Tetrachloroethylene was detected in
1,252 samples while bromoform was detected less frequently (167), both out of 1,466 valid
samples. Although three pollutants exhibited a failure rate of 100 percent
1,1,2,2-tetrachloroethane, 1,2-dibromoethane, and chloroprene), all of these were infrequently
detected (less than 10 percent). 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.
The program-level pollutants of interest, as indicated by the shading in Table 4-7, were
identified as follows:
• Acenaphthene • Ethylbenzene
• Acetaldehyde • Fluorene
• Arsenic • Formaldehyde
• Benzene • Hexachloro-1,3-butadiene
• 1,3-Butadiene • Manganese
• Carbon Tetrachloride • Naphthalene
• />-Dichlorobenzene • Nickel.
• 1,2-Dichloroethane
The pollutants of interest identified via the preliminary risk-based screening approach for
2012 is similar to the list of pollutants identified in previous years.
Of the 71 pollutants sampled for under the NMP that have corresponding screening
values, concentrations of 38 pollutants failed at least one screen (or roughly 54 percent of
pollutants). Of these, a total of 13,867 out of 38,759 concentrations (or nearly 36 percent) failed
screens. If all of the pollutants with screening values are considered (including those that did not
fail any screens), the percentage of concentrations failing screens is less (13,867 of 55,796, or
4-17
-------
25 percent). Note that this percentage excludes acrolein, acetonitrile, acrylonitrile, and carbon
disulfide measurements per the explanations provided in Section 3.2.
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.
As shown, S4MO has the largest number of failed screens (692), followed by PXSS (671) and
TOOK (605). 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 failed screens and the total number of screens conducted (based on applicable
measured detections) and is also provided in Table 4-8. Note that the results in this table also
exclude acrolein, acetonitrile, acrylonitrile, and carbon disulfide.
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,
five, or six pollutant groups tended to have a higher number of failed screens. Although WPIN,
RRMI, ORFL, AZFL, and INDEM have the highest failure rates (66 percent to 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 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 S4MO,
PXSS, NBIL, GLKY, BTUT, and SEWA. These sites each sampled five or six pollutant groups
and have a failure rate between 19 percent and 26 percent. For this reason, the number of
pollutant groups for which sampling was conducted is also presented in Table 4-8. Note that
measurements for two sites, HOWI and CAMS 85, did not fail any screens. Both of these sites
sampled only hexavalent chromium.
The following sections from this point forward focus primarily on those pollutants
designated as program-level pollutants of interest.
4-18
-------
Table 4-8. Site-Specific Risk-Based Screening Comparison
Site
S4MO
PXSS
TOOK
GPCO
NBIL
TMOK
DEMI
ADOK
SEWA
BTUT
OCOK
ELNJ
GLKY
PROK
SPIL
NBNJ
CHNJ
SSSD
UNVT
LEKY
ROIL
ASKY
BURVT
SKFL
RUVT
CCKY
RICO
SPAZ
ASKY-M
SYFL
BOMA
LAKY
TVKY
SJJCA
BRCO
ATKY
BMCO
ORFL
PACO
AZFL
#of
Failed
Screens
692
671
605
600
562
559
520
514
512
508
507
432
424
409
409
399
382
359
288
253
228
181
177
176
174
173
173
166
155
153
140
137
130
128
127
121
121
121
121
117
Total # of
Measured
Detections1
2,745
2,565
1,767
2,070
2,591
1,793
2,101
1,858
2,433
2,265
1,767
1,319
2,261
1,458
1,236
1,360
1,267
1,215
1,992
1,003
678
556
596
941
558
875
481
526
498
780
1,438
522
504
1,159
415
509
396
183
357
177
%of
Failed
Screens
25.21
26.16
34.24
28.99
21.69
31.18
24.75
27.66
21.04
22.43
28.69
32.75
18.75
28.05
33.09
29.34
30.15
29.55
14.46
25.22
33.63
32.55
29.70
18.70
31.18
19.77
35.97
31.56
31.12
19.62
9.74
26.25
25.79
11.04
30.60
23.77
30.56
66.12
33.89
66.10
#of
Pollutant
Groups
Analyzed
5
5
3
4
6
3
4
4
5
6
3
2
5
o
J
2
2
2
3
5
3
2
2
2
o
J
2
2
2
1
1
3
3
1
1
2
2
1
2
1
2
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)
Site
INDEM
WPIN
BLKY
RRMI
ROCH
BAKY
CELA
WADC
RIVA
PRRI
RUCA
SWMI
RFCO
MONY
BXNY
PAFL
SDGA
LBHCA
CAMS 35
MIWI
CHSC
STMN
CAMS 85
HOW
#of
Failed
Screens
117
117
102
98
91
83
74
63
59
58
58
58
49
46
38
37
31
26
13
12
7
6
0
0
Total # of
Measured
Detections1
177
174
385
147
757
496
669
673
612
811
647
90
150
464
312
300
342
250
168
41
409
39
47
35
%of
Failed
Screens
66.10
67.24
26.49
66.67
12.02
16.73
11.06
9.36
9.64
7.15
8.96
64.44
32.67
9.91
12.18
12.33
9.06
10.40
7.74
29.27
1.71
15.38
0
0
#of
Pollutant
Groups
Analyzed
1
1
1
1
2
1
1
2
2
2
1
1
2
2
2
1
2
1
2
1
2
1
1
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
4-20
-------
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 VOCs, 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 presented where there are at least three quarterly averages and where the site-specific
method completeness is at least 85 percent. The annual average concentrations for PAHs in
Table 4-11 and metals in Table 4-12 are reported in ng/m3 for ease of viewing, while annual
average concentrations in Tables 4-9 and 4-10, for VOCs and carbonyl compounds, respectively,
are reported in ug/m3. Note that not all sites sampled each pollutant; thus, the list of possible
sites presented in Tables 4-9 through 4-12 is limited to those sites sampling each pollutant. For
example, only five sites sampled TSP metals; thus, all five sites 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 Pollutants of Interest
Rank
1
2
3
4
5
6
7
8
9
10
Benzene
(jig/m )
TOOK
2.21 ±0.31
SPAZ
1.43 ±0.31
PXSS
1.28 ±0.21
GPCO
1.28 ±0.12
TMOK
1.25 ±0.16
BMCO
1.09 ±0.12
RUVT
1.05 ±0.20
ELNJ
1.04 ±0.14
BTUT
1.02 ±0.13
RICO
1.00 ±0.12
1,3-Butadiene
(Hg/m3)
SPAZ
0.26 ±0.07
PXSS
0.22 ±0.05
RICO
0.18 ±0.03
GPCO
0.18 ±0.03
ELNJ
0.14 ±0.02
SPIL
0.14 ±0.03
RUVT
0.13 ±0.04
TMOK
0.13 ±0.02
BTUT
0.12 ±0.03
DEMI
0.12 ±0.03
Carbon
Tetrachloride
(Hg/m3)
DEMI
0.71 ±0.02
NBIL
0.71 ±0.05
SEWA
0.70 ± 0.02
PROK
0.70 ±0.03
GLKY
0.69 ±0.03
PXSS
0.68 ±0.02
SPIL
0.68 ±0.03
TMOK
0.68 ±0.02
RUVT
0.68 ±0.04
S4MO
0.68 ±0.02
P-
Dichlorobenzene
(Hg/m3)
SPAZ
0.26 ± 0.06
PXSS
0.20 ±0.03
S4MO
0.18 ±0.06
ADOK
0.13 ±0.04
TOOK
0.09 ±0.01
PROK
0.09 ±0.02
TMOK
0.08 ±0.01
ELNJ
0.07 ±0.02
GPCO
0.07 ±0.01
OCOK
0.07 ±0.01
1,2-
Dichloro ethane
(Hg/m3)
S4MO
0.08 ±0.01
GPCO
0.08 ±0.01
NBNJ
0.08 ±0.01
BTUT
0.08 ±0.01
SPIL
0.08 ±0.01
BURVT
0.08 ±0.01
SPAZ
0.08 ±0.01
ELNJ
0.08 ±0.01
RUVT
0.07 ±0.01
TOOK
0.07 ±0.01
Ethylbenzene
(Hg/m3)
TOOK
0.91 ±0.17
SPAZ
0.84 ±0.18
PXSS
0.73 ±0.12
GPCO
0.70 ±0.11
TMOK
0.56 ±0.08
DEMI
0.53 ±0.14
ELNJ
0.41 ±0.05
BTUT
0.36 ±0.06
RUVT
0.36 ±0.05
S4MO
0.36 ±0.05
Hexachloro-1,3-
Butadiene
(Hg/m3)
NBNJ
0.02 ±0.01
S4MO
0.02 ±0.01
CHNJ
0.02 ±0.01
GPCO
0.02 ±0.01
BURVT
0.01 ±0.01
TOOK
0.01 ±0.01
UNVT
0.01 ±0.01
NBIL
0.01 ±0.01
ELNJ
0.01 ±0.01
TMOK
0.01 ±0.01
to
to
BOLD ITALICS = EPA-designated NATTS Site
-------
Table 4-10. Annual Average Concentration Comparison of the
Carbonyl Compound Pollutants of Interest
Rank
1
2
3
4
5
6
7
8
9
10
Acetaldehyde
(Ug/m3)
PXSS
2.90 ±0.30
GPCO
2.89 ±0.27
TOOK
2.78 ±0.43
SPIL
2.72 ±0.77
ELNJ
2.66 ±0.34
BTUT
2.54 ±0.35
OCOK
2.34 ±0.32
TMOK
2.33 ±0.32
WPIN
2.28 ±0.27
S4MO
1.86 ±0.22
Formaldehyde
(Ug/m3)
BTUT
4.44 ±0.75
WPIN
4.31 ±0.61
PXSS
3.96 ±0.27
ELNJ
3.89 ±0.47
TMOK
3.63 ±0.47
PROK
3. 58 ±0.65
OCOK
3.49 ±0.54
DEMI
3.45 ±0.44
TOOK
3.42 ±0.54
S4MO
3.26 ±0.52
BOLD ITALICS = EPA-designated NATTS Site
Table 4-11. Annual Average Concentration Comparison of the PAH Pollutants of Interest
Rank
1
2
o
6
4
5
6
7
8
9
10
Acenaphthene
(ng/m3)
GPCO
20.53 ±7.27
DEMI
12.60 ±4.41
ROCH
12.21 ±3 A3
NBIL
11. 51 ±3.62
S4MO
7.37 ±1.59
CELA
5.44 ±0.70
RIVA
3.76 ±0.75
SEWA
3.60 ±0.88
BOMA
3.04 ±1.23
WADC
3.03 ±0.49
Fluorene
(ng/m3)
GPCO
12.56 ±2.86
NBIL
12.31 ±4.18
DEMI
11.35 ±3. 53
ROCH
9.95 ±2.68
S4MO
8.32 ±1.67
CELA
7.67 ±1.07
RUCA
4.27 ±0.48
RIVA
4.16 ±0.65
WADC
4.03 ±0.53
BOMA
3.76 ±1.41
Naphthalene
(ng/m3)
GPCO
203.78 ±35.24
CELA
179.67 ±20.99
DEMI
141.70 ±23. 82
S4MO
110.45 ±19.71
WADC
104.38 ±19.17
PXSS
97.83 ± 19.46
RUCA
96.96 ±15.56
SKFL
96.91 ±21.04
RIVA
93. 95 ±12.47
NBIL
77.94 ±17.78
BOLD ITALICS = EPA-designated NATTS Site
4-23
-------
Table 4-12. Annual Average Concentration Comparison of the Metals Pollutants of Interest
Rank
1
2
3
4
5
6
7
8
9
10
Arsenic
(PM10)
(ng/m3)
ASKY-M
1.79 ±0.37
S4MO
1.09 ±0.25
PAFL
1.03 ±0.33
BAKY
0.93 ±0.20
LEKY
0.92 ±0.18
CCKY
0.86 ±0.28
NBIL
0.73 ±0.12
SEWA
0.68 ±0.11
PXSS
0.68 ±0.11
GLKY
0.59 ±0.10
Arsenic
(TSP)
(ng/m3)
TOOK
0.92 ±0.10
TMOK
0.77 ±0.11
PROK
0.63 ±0.10
OCOK
0.57 ±0.07
ADOK
0.49 ±0.05
Manganese
(PM10)
(ng/m3)
ASKY-M
34.09 ±10.54
PXSS
22.75 ±4.01
S4MO
22.66 ± 9.60
SEWA
9.80 ±2.88
NBIL
9.11 ±1.86
BTUT
7.97 ±1.24
BAKY
6.74 ±0.84
LEKY
6.69 ±0.96
CCKY
6.50 ±0.96
SJJCA
6.22 ±1.18
Manganese
(TSP)
(ng/m3)
TOOK
38.33 ±8.81
TMOK
26.22 ± 8.46
OCOK
21. 10 ±4.26
PROK
18.66 ±8.09
ADOK
13.09 ±2.635
Nickel
(PM10)
(ng/m3)
ASKY-M
2.94 ±0.90
SEWA
2.74 ±0.71
PXSS
2.04 ±0.34
S4MO
1.42 ±0.36
BOMA
1.41 ±0.29
BTUT
1.41±0.19
SJJCA
1.17±0.19
NBIL
1.04 ±0.16
PAFL
0.81±0.11
LEKY
0.62 ±0.21
Nickel
(TSP)
(ng/m3)
TOOK
2.42 ±0.49
TMOK
1.67 ±0.26
OCOK
1.10±0.16
PROK
0.99 ±0.21
ADOK
0.86 ±0.10
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 (4.44 ± 0.75 |ig/m3). As shown in
Table 4-10, WPIN also has an annual average concentration greater than 4 |ig/m3
(4.31 ± 0.61 |ig/m3) and all of the sites shown in Table 4-10 have annual average
concentrations greater than 3 |ig/m3. Formaldehyde and acetaldehyde together
account for 19 of the 20 annual average concentrations greater than 2.0 |ig/m3 in
Tables 4-9 through 4-12 (the one exception being for TOOK's annual average
concentration of benzene).
• The annual average concentrations of benzene are the only annual averages among
the VOCs shown greater than 1 |ig/m3. TOOK's annual average benzene
concentration (2.21 ± 0.31 |ig/m3) is significantly higher than the next highest annual
average benzene concentration (1.43 ± 0.31 |ig/m3 for SPAZ), but is considerably less
than its annual average for the 2011 NMP report (3.59 ± 0.98 |ig/m3). Across the
program, six of the 11 individual benzene measurements greater than 4 |ig/m3 were
measured at TOOK. The other Tulsa site (TMOK) ranks fifth for benzene. The two
Phoenix sites (SPAZ and PXSS) rank second and third, respectively, for benzene.
Three of the six Colorado sites also appear among the sites with the 10 highest annual
average benzene concentrations.
• Concentrations of some of the VOCs vary significantly while others do not. The
difference between the highest and 10th highest annual average concentration of
carbon tetrachloride is only 0.035 |ig/m3. The difference between the highest and 10th
4-24
-------
highest annual average concentrations of both 1,2-dichloroethane and hexachloro-1,3-
butadiene is even less, approximately 0.011 |ig/m3 for both pollutants. Conversely,
the difference between the highest and 10th highest annual average concentration of
benzene is 1.21 |ig/m3.
• The sites with the five highest annual averages concentrations of benzene are the
same sites with the five highest annual averages concentrations for ethylbenzene.
Altogether, eight of the 10 sites are the same in the benzene and ethylbenzene
columns.
• Although BTUT has the highest annual average concentration of formaldehyde
(4.44 ± 0.75 |ig/m3) shown in Table 4-10, the maximum concentration measured
across the program is shared between TOOK and SPIL (a concentration of
12.8 |ig/m3 was measured at each site). Of the five formaldehyde concentrations
greater than 12 |ig/m3 measured across the program, one was measured at BTUT, one
at TOOK, one at PROK, and two were measured at SPIL. However, SPIL does not
appear in Table 4-10 because its annual average concentration ranks 12th.
• While the three highest acetaldehyde concentrations across the program (ranging
from 8.74 |ig/m3 to 20.4 |ig/m3) were all measured at SPIL, its annual average
concentration ranked fourth among other sites sampling carbonyl compounds. The
variability in this site's acetaldehyde concentrations is indicated by its confidence
interval, which is nearly twice the confidence intervals shown for the other sites in
Table 4-10.
• Seven of the 10 sites shown in Table 4-10 for formaldehyde also appear among the
sites with the highest annual average concentrations of acetaldehyde.
• Table 4-11 shows that GPCO has the highest annual average concentration for each
of the program-wide PAH pollutants of interest. The annual average concentrations of
acenaphthene and naphthalene for GPCO are considerably higher than the next
highest annual averages and have relatively large confidence intervals associated with
them. GPCO has the four highest measurements of naphthalene program-wide
(ranging from 475 ng/m3 to 822 ng/m3). GPCO also has the only two concentrations
of acenaphthene greater than 100 ng/m3 measured across the program, as well as five
of the nine measurements greater than 50 ng/m3. GPCO's annual average
concentration of fluorene is relatively similar to the second highest annual average
concentration of this pollutant (calculated for NBIL).
• ASKY-M has the highest annual average concentration of each of the three program-
wide PMio metals pollutants of interest. All five Kentucky sites sampling PMio metals
appear in Table 4-11 for arsenic; four of the five Kentucky sites appear in Table 4-11
for manganese; and only two appear in Table 4-11 for nickel. S4MO, which has had
the highest concentration of arsenic and manganese in past reports, ranks second for
arsenic, third for manganese (behind PXSS), and fifth for nickel for 2012. S4MO and
ASKY-M each have one manganese concentration greater than 200 ng/m3 and
another greater than 100 ng/m3 (as does PXSS).
4-25
-------
• TOOK has the highest annual average concentration of each of the three program-
wide TSP metals pollutants of interest. Further, for the TSP metals, the Tulsa sites are
the two Oklahoma sites with the highest annual average concentrations.
• S4MO is on the top 10 list for 13 of the 19 program-level pollutants of interest; PXSS
is on the top 10 list for 11 of the 19 program-level pollutants of interest; and GPCO is
on the top 10 list for 10 of the 19 program-level pollutants of interest. NBIL, BTUT,
and ELNJ each appear in Tables 4-9 through 4-12 a total of 8 times. Conversely, 26
sites do not appear in Table 4-9 through 4-12 at all. Note, however, that some sites
did not meet the criteria for annual averages to be calculated.
4.2.2 Risk-Based Screening Assessment Using MRLs
Table 4-13 presents the pollutants analyzed under the NMP that have associated ATSDR
MRLs. Note that some pollutants do not have MRLs for one or more of the designated time
frames (acute, intermediate, or chronic). None of the preprocessed daily measurements are
greater than the associated acute MRL; none of the quarterly average concentrations, where they
could be calculated, are greater than the associated intermediate MRL; and none of the annual
average concentrations, where they could be calculated, are greater than the associated chronic
MRL. Thus, Table 4-13 also presents the maximum preprocessed daily measurement, quarterly
average concentration, and annual average concentration associated with each pollutant. This
allows the reader to see how close (or how far) from the MRL(s) some concentrations were. For
example, the acute MRL for benzene is 30 |ig/m3 and the maximum benzene concentration
measured in 2012 was 5.73 |ig/m3; the acute MRL for acetone is 60,000 |ig/m3 while the
maximum concentration measured was 27.6 |ig/m3.
4-26
-------
Table 4-13. Comparison of Maximum Concentrations vs. ATSDR MRLs
Pollutant
Acetone
Benzene
Bromomethane
Cadmium
Carbon Bisulfide
Carbon Tetrachloride
Chloroethane
Chloroform
Chloromethane
Cobalt
£>-Dichlorobenzene
1 ,2-Dichloroethane
1 , 1 -Dichloroethene
trans- 1 ,2-Dichloroethylene
cis/trans- 1 ,3 -dichloropropene2
Dichloromethane
1 ,2-Dichloropropane
Ethylbenzene
Formaldehyde
w-Hexane
Hexavalent Chromium
Manganese
Mercury
Methyl ter/-Butyl Ether
Naphthalene
Nickel
Styrene
ATSDR Acute
MRL1
(Hg/m3)
60,000
30
200
0.03
—
—
40,000
500
1,000
—
10,000
—
—
800
—
2,000
200
20,000
50
—
—
—
—
7,000
—
—
20,000
Maximum
Preprocessed Daily
Measurement
(Hg/m3)
27.60
5.73
0.55
0.003
—
—
0.90
45.80
3.43
—
1.37
—
—
0.85
—
745.00
0.06
3.63
12.80
—
—
—
—
0.32
—
—
39.00
ATSDR
Intermediate
MRL1
(Hg/m3)
30,000
20
200
—
—
200
—
200
400
—
1,000
—
80
800
40
1,000
30
9,000
40
—
0.3
—
—
3,000
—
0.2
-
Maximum
Quarterly Average
Concentration
(Hg/m3)
10.96
2.51
0.19
—
—
1.28
—
5.29
1.49
—
0.50
—
0.04
0.05
0.10
104.13
0.004
1.41
8.30
—
0.001
—
—
0.22
—
0.01
-
ATSDR
Chronic MRL1
(Hg/m3)
30,000
10
20
0.01
900
200
—
100
100
0.1
60
2,000
—
—
30
1,000
—
300
10
2,000
—
0.3
0.2
3,000
4
0.09
900
Maximum Annual
Average
Concentration
(Hg/m3)
7.88
2.21
0.11
0.001
11.56
0.71
—
2.48
1.32
0.002
0.26
0.08
—
—
0.02
40.23
—
0.91
4.44
3.34
—
0.04
0.00004
0.12
0.20
0.003
2.97
to
Reflects the use of one significant digit
2The MRL for 1,3-dichloropropene was
provided in Table 4-13.
for MRLs
applied to both isomers (cis-1,3 -dichloropropene and trans-1,3 -dichloropropene), with the maximum concentration for the pair
-------
Table 4-13. Comparison of Maximum Concentrations vs. ATSDRMRLs (Continued)
Pollutant
Tetrachloroethylene
Toluene
1,1,1 -Trichloroethane
Trichloroethylene
Vinyl Chloride
Xylenes
ATSDR Acute
MRL1
(Ug/m3)
1,000
4,000
10,000
—
1,000
9,000
Maximum
Preprocessed Daily
Measurement
(Ug/m3)
5.38
21.50
0.36
—
9.81
18.03
ATSDR
Intermediate
MRL1
(Ug/m3)
—
—
4,000
—
80
3,000
Maximum
Quarterly Average
Concentration
(Ug/m3)
—
—
0.08
—
1.05
5.63
ATSDR
Chronic MRL1
(Ug/m3)
300
300
—
2
~
200
Maximum Annual
Average
Concentration
(Ug/m3)
0.46
6.56
—
0.71
~
4.09
Reflects the use of one significant digit for MRLs
2The MRL for 1,3-dichloropropene was applied to both isomers (c/'s-l,3-dichloropropene and /ra«s-l,3-dichloropropene), with the maximum concentration for the pair
provided in Table 4-13.
to
oo
-------
The pollutant with the preprocessed daily measurement closest to the acute MRL is
dichloromethane (the acute MRL is 2000 |ig/m3 and the maximum dichloromethane measurement
is 745 |ig/m3). The pollutant with the quarterly average concentration closest to the intermediate
MRL is formaldehyde (the intermediate MRL is 40 |ig/m3 and the maximum quarterly average is
8.30 |ig/m3). The pollutant with the annual average concentration closest to the chronic MRL is
also formaldehyde (the chronic MRL is 10 |ig/m3 and the maximum annual average is 4.44 |ig/m3).
Because none of the preprocessed daily measurements are greater than associated acute
MRLs, the emission tracer analysis described in Section 3.5.3.3 was not performed.
4.3 The Effect of Mobile Sources
Ambient air is significantly affected by mobile sources, as discussed in Section 3.4.1.
Table 4-14 contains several parameters that are used to assess if mobile sources are affecting air
quality near the monitoring sites, including emissions data from the NEI, concentration data, and
site-characterizing data, such as vehicle ownership.
4.3.1 Mobile Source Emissions
Emissions from mobile sources contribute significantly to air pollution in the United States.
Mobile source emissions can be broken into two categories: on-road and non-road. On-road
emissions come from mobile sources such as automobiles, buses, and trucks that use roadways;
non-road emissions come from the remaining mobile sources such as locomotives, lawn mowers,
airplanes, and boats (EPA, 201 Ib). Table 4-14 contains county-level on-road and non-road HAP
emissions from the 2011 NEI.
Mobile source emissions tend to be highest in large urban areas and lowest in rural areas.
Estimated on-road county emissions were highest in Los Angeles County, CA (where CELA and
LBHCA are located), followed Maricopa County, AZ (where PXSS and SPAZ are located), and
Cook County, IL (where NBIL and SPIL are located). Estimated on-road emissions were lowest in
Livingston County, KY (BLKY), Chesterfield County, SC (CHSC), and Carter County, KY
(GLKY). Estimated non-road county emissions were also highest in Los Angeles County, CA;
Cook County, IL; and Maricopa County, AZ. Estimated non-road county emissions were lowest in
Carter County, KY; Boyd County, KY (where ASKY and ASKY-M are located); and Chesterfield
County, SC.
4-29
-------
Table 4-14. Summary of Mobile Source Information by Monitoring Site
Site
ADOK
ASKY
ASKY-M
ATKY
AZFL
BAKY
BLKY
BMCO
BOMA
BRCO
BTUT
BURVT
BXNY
CAMS 35
CAMS 85
CCKY
CELA
CHNJ
CHSC
DEMI
ELNJ
GLKY
GPCO
HOW
INDEM
LAKY
LBHCA
LEKY
MIWI
MONY
NBIL
NBNJ
OCOK
ORFL
PACO
PAFL
PROK
PRRI
PXSS
RFCO
RICO
County-level
Motor Vehicle
Registration1
(# of Vehicles)
847,824
39,227
39,227
30,297
872,813
38,518
8,281
74,508
362,899
74,508
259,319
169,767
251,398
3,252,420
71,658
30,297
7,422,254
445,710
41,259
1,337,797
485,449
25,391
179,213
96,912
419,431
30,297
7,422,254
207,043
632,914
251,398
2,092,085
733,908
847,824
1,073,682
74,508
1,073,682
41,391
548,763
3,761,859
74,508
74,508
Annual
Average Daily
Traffic1
(# of Vehicles)
34,100
7,229
12,842
3,262
38,500
922
2,280
2,527
27,654
1,102
129,145
14,000
99,201
31,043
1,250
4,742
229,000
11,215
550
87,500
250,000
303
11,000
5,100
34,754
1,189
282,000
10,083
12,800
91,213
115,100
110,653
40,900
35,000
16,000
49,500
15,100
136,800
184,000
16,000
17,000
County-level
Daily VMT1
27,411,171
1,281,000
1,281,000
1,292,000
21,387,550
1,417,000
398,000
1,902,077
10,890,178
1,902,077
6,866,779
4,032,329
8,178,210
57,020,660
2,405,125
1,292,000
214,458,140
14,844,444
1,228,145
40,951,779
12,264,174
1,080,000
2,009,730
2,626,054
16,226,000
1,292,000
214,458,140
7,545,000
17,532,434
8,178,210
86,217,829
20,644,392
27,411,171
34,099,958
1,902,077
34,099,958
1,662,076
NA
90,393,000
1,902,077
1,902,077
County-Level
On-road
Emissions2
(tpy)
2,075.41
110.93
110.93
116.37
1,716.20
137.79
36.29
209.80
621.13
209.80
586.54
315.99
917.00
4,639.61
158.35
116.37
9,326.27
697.15
96.86
3,354.28
636.49
104.91
325.68
227.28
831.30
116.37
9,326.27
591.61
1,365.75
917.00
4,729.93
976.00
2,075.41
2,663.78
209.80
2,663.78
126.46
798.94
6,467.29
209.80
209.80
County-Level
Non-road
Emissions2
(tpy)
684.79
22.72
22.72
350.78
876.17
100.53
80.28
74.47
344.54
74.47
257.49
189.85
241.43
1,660.72
98.06
350.78
4,010.78
531.97
56.37
982.04
321.63
11.48
146.85
177.03
523.81
350.78
4,010.78
334.21
474.65
241.43
3,482.70
523.95
684.79
1,222.71
74.47
1,222.71
60.29
304.50
3,082.11
74.47
74.47
Hydrocarbon
Average3
(ppbv)
1.46
2.86
NA
4.62
NA
NA
2.72
NA
NA
NA
3.33
2.18
NA
NA
NA
2.92
NA
2.74
NA
3.09
6.19
1.12
5.14
NA
NA
3.86
NA
1.90
NA
NA
2.13
2.99
2.01
NA
NA
NA
1.89
NA
4.70
NA
NA
Individual references provided in each state section.
Reference: EPA, 2013c
3This parameter is only available for monitoring sites sampling VOCs and is not limited by the annual average criteria.
BOLD ITALICS = EPA-designated NATTS Site
NA = VOC samples were not collected at this monitoring site.
4-30
-------
Table 4-14. Summary of Mobile Source Information by Monitoring Site (Continued)
Site
RIVA
ROCH
ROIL
RRMI
RUCA
RUVT
S4MO
SDGA
SEWA
SJJCA
SKFL
SPAZ
SPIL
SSSD
STMN
SWMI
SYFL
TMOK
TOOK
TVKY
UNVT
WADC
WPIN
County-level
Motor Vehicle
Registration1
(# of Vehicles)
354,419
556,055
286,043
1,337,797
1,724,787
70,900
1,112,866
472,535
1,403,968
1,529,351
872,813
3,761,859
2,092,085
212,507
218,196
1,337,797
1,143,207
618,359
618,359
30,297
169,767
316,231
820,767
Annual
Average Daily
Traffic1
(# of Vehicles)
72,000
88,348
9,400
97,300
145,000
6,700
79,558
141,980
224,000
106,000
49,000
128,000
191,700
18,575
24,100
94,400
10,400
12,600
63,000
2,231
1,100
7,400
143,970
County-level
Daily VMT1
8,232,198
15,980,952
7,867,318
40,951,779
55,717,760
1,745,205
23,994,911
20,113,000
23,044,858
41,250,490
21,387,550
90,393,000
86,217,829
3,778,321
4,983,115
40,951,779
34,061,637
20,402,564
20,402,564
1,292,000
4,032,329
9,775,000
32,005,000
County-Level
On-road
Emissions2
(tpy)
618.44
1,152.75
475.89
3,354.28
2,070.71
134.11
809.51
1,328.12
4,461.96
2,715.06
1,716.20
6,467.29
4,729.93
365.04
542.65
3,354.28
2,824.86
1,480.37
1,480.37
116.37
315.99
615.46
2,593.89
County-Level
Non-road
Emissions2
(tpy)
145.79
573.62
216.47
982.04
881.09
127.14
157.06
269.22
2,176.51
601.80
876.17
3,082.11
3,482.70
116.48
655.77
982.04
1,044.25
714.80
714.80
350.78
189.85
248.44
624.63
Hydrocarbon
Average3
(ppbv)
NA
NA
3.56
NA
NA
3.14
2.76
NA
2.01
NA
NA
5.38
2.76
1.62
NA
NA
NA
3.69
6.55
6.12
0.92
NA
NA
Individual references provided in each state section.
Reference: EPA, 2013c
3This parameter is only available for monitoring sites sampling VOCs
BOLD ITALICS = EPA-designaled NATTS Site
NA = VOC samples were not collected at this monitoring site.
and is not limited by the annual average criteria.
4.3.2 Hydrocarbon Concentrations
Hydrocarbons are organic compounds that contain only carbon and hydrogen.
Hydrocarbons are derived primarily from crude petroleum sources and are classified according to
their arrangement of atoms as alicyclic, aliphatic, and aromatic. Hydrocarbons are of prime
economic importance because they encompass the constituents of the major fossil fuels,
petroleum and natural gas, as well as plastics, waxes, and oils. Hydrocarbons in the atmosphere
originate from natural sources and from various anthropogenic sources, such as the combustion
of fuel and biomass, petroleum refining, petrochemical manufacturing, solvent use, and gas and
oil production and use. In urban air pollution, these components, along with oxides of nitrogen
(NOX) and sunlight, contribute to the formation of tropospheric ozone. Thus, the concentration of
4-31
-------
hydrocarbons in ambient air may act as an indicator of mobile source activity levels. Several
hydrocarbons are sampled with Method TO-15, including benzene, ethylbenzene, and toluene.
Table 4-14 presents the average of the sum of hydrocarbon concentrations for each site
sampling VOCs. Note that only sites sampling VOCs have data in this column. Table 4-14 shows
that TOOK, ELNJ, TVKY, and SPAZ have the highest hydrocarbon averages among the sites
monitoring VOCs. TOOK and ELNJ are located in highly populated urban areas and in relatively
close proximity to heavily traveled roadways. TOOK is located near Exit 3 A of 1-244 in Tulsa,
Oklahoma while ELNJ is location on Exit 13 A of the New Jersey Turnpike. SPAZ is located in a
highly urbanized area (Phoenix), but not near a major roadway. TVKY is located in a highly
industrialized area in a moderately populated area. The sites with the lowest hydrocarbon
averages are UNVT, GLKY, and ADOK. UNVT and GLKY are located in rural areas. ADOK is
located on the edge of an urbanized area just south of a major roadway. The average sum of
hydrocarbon concentrations can be compared to other indicators of mobile source activity, such
as the ones discussed below, to determine if correlations exist.
4.3.3 Motor Vehicle Ownership
Another indicator of motor vehicle activity near the monitoring sites is the total number
of vehicles owned by residents in the county where each monitoring site is located, which
includes passenger vehicles, trucks, and commercial vehicles, as well as vehicles that can be
regional in use such as boats or snowmobiles. Actual county-level vehicle registration data were
obtained from each applicable state or local agency, where possible. If data were not available,
vehicle registration data are available at the state-level (FJHWA, 2013a). The county proportion
of the state population was then applied to the state registration count.
The county-level motor vehicle ownership data and the average summed hydrocarbon
concentrations are presented in Table 4-14. As previously discussed, TOOK, ELNJ, TVKY, and
SPAZ have the highest average summed hydrocarbon concentrations, respectively, while UNVT,
GLKY, and ADOK have the lowest. Table 4-14 also shows that SPAZ, PXSS, NBIL, and SPIL
have the highest county-level vehicle ownership of the sites sampling VOCs, while the Kentucky
sites located in Livingston, Carter, and Marshall Counties have the lowest. The Pearson
correlation coefficient calculated between these two datasets is 0.19, which is considered a weak
correlation. CELA and LBHC A, which have the highest county-level vehicle ownership of all
4-32
-------
NMP sites, did not sample VOCs under the NMP; this is also true for many of the sites with
larger vehicle ownership counts.
The vehicle ownership at the county-level may not be completely indicative of the
ownership in a particular area. As an illustration, for a county with a large city in the middle of
its boundaries and less populated areas surrounding it, the total county-level ownership may be
more representative of areas inside the city limits than in the rural outskirts.
Other factors may affect the reliability of motor vehicle ownership data as an indicator of
ambient air monitoring data results:
• Estimates of higher vehicle ownership surrounding a monitoring site do not
necessarily imply increased motor vehicle use in the immediate vicinity of a
monitoring site. Conversely, sparsely populated regions often contain heavily traveled
roadways.
• Emissions sources in the area other than motor vehicles may significantly affect
levels of hydrocarbons in ambient air.
4.3.4 Estimated Traffic Volume
Traffic data for each of the participating monitoring sites were obtained from state and
local agencies, primarily departments of transportation. Most of the traffic counts in this report
reflect AADT, which is "the annual traffic count divided by the number of days in the year," and
incorporates both directions of traffic (FHWA, 2013b). AADT counts obtained were based on
data from 2002 to 2013, primarily 2011 forward. The updated traffic values are presented in
Table 4-14. The traffic data presented in Table 4-14 represent the most recently available data
applicable to the monitoring sites.
There are several limitations to obtaining the AADT near each monitoring site. AADT
statistics are developed for roadways, such as interstates, state highways, or local roadways,
which are managed by different municipalities or government agencies. AADT is not always
available for rural areas or for secondary roadways. For monitoring sites located near interstates,
the AADT for the interstate segment closest to the site was obtained. For other monitoring sites,
the highway or secondary road closest to the monitoring site was used. Only one AADT value
was obtained for each monitoring site. The intersection or roadway chosen for each monitoring
site is discussed in each individual state section (Sections 5 through 29).
4-33
-------
Table 4-14 shows that ELNJ, SEW A, and SPIL have the highest daily traffic volumes of
the sites sampling VOCs, while GLKY, UNVT, and LAKY have the lowest. For all monitoring
sites (not just those sampling VOCs), the highest daily traffic volume occurs near LBHCA,
ELNJ, CELA, and SEWA. LBHCA is near 1-405 east of the intersection with 1-710; ELNJ is
located near Exit 13A on 1-95; CELA is located in downtown Los Angeles; and SEWA is located
in Seattle near 1-5 south of its intersection with 1-9. ELNJ has the second highest traffic volume
and the second highest hydrocarbon average, but SEWA, which has the fourth highest traffic
volume, has the 23rd highest hydrocarbon average. Again, LBHCA and CELA did not measure
VOCs under the NMP. A Pearson correlation coefficient calculated between the average
summed hydrocarbon calculations and the traffic counts is 0.25, which is also considered a weak
correlation.
4.3.5 Vehicle Miles Traveled
Another approach to determine how mobile sources affect urban air quality is to review
VMT. VMT is "the mileage traveled by all vehicles on a road system over a period of time such
as a year" (FHWA, 2013b). Thus, VMT values tend to be large (in the millions). County-level
VMT was obtained for each of the participating monitoring sites from state organizations,
primarily departments of transportation. However, these data are not readily available for all
states. In addition, not all states provide this information on the same level. For example, many
states provide VMT for all public roads, while the state of Colorado provided this information
for state highways only. County-level VMT are presented in Table 4-14, where available.
Of the sites sampling VOCs, county-level VMT, where available, was highest for PXSS
and SPAZ, SPIL and NBIL, and DEMI (Wayne County, MI). The sites with the lowest county-
level VMT, where available, are BLKY, GLKY, and ASKY. A Pearson correlation coefficient
calculated between the average summed hydrocarbon concentrations and VMT, where available,
is 0.12, indicating little correlation between hydrocarbon concentrations and county-level VMT.
It is important to note that many of the sites with larger VMT did not measure VOCs under the
NMP (such as CELA, LBHCA, CAMS 35, RUCA, and SJJCA). In addition, county-level VMT
was not readily available for Rhode Island.
4-34
-------
4.4 Variability Analysis
This section presents the results of the three variability analyses described in
Section 3.4.2.
4.4.1 Coefficient of Variation and Inter-site Variability
The site-specific CVs and the inter-site comparison analyses are discussed together in this
section. Figures 4-la through 4-15a are graphical displays of site-specific CVs (standard
deviation vs. annual average concentration) for the program-level pollutants of interest.
Figures 4-lb through 4-15b are bar graphs depicting the site-specific annual averages overlain on
the program-level averages (indicated by the yellow shading), as presented in Section 4.1. For
each program-level pollutant of interest, the CV graph is shown first, followed by the inter-site
variability graph. The figures are aligned this way because they complement each other; the data
point with the highest annual average concentration and/or standard deviation in the CV graph is
easily identifiable in the inter-site variability graph. Further, the inter-site variability graphs
allow the reader to see how the individual site-specific annual averages feed into the program-
level averages (i.e., if a specific site(s) is driving the program average). In addition to the
standard deviations on the CV graphs, the confidence intervals provided on the inter-site
variability graphs are a further indication of the amount of variability contained within the site-
specific annual averages.
Several items to note about these figures: Some sites do not have annual averages
presented on the inter-site variability graphs because they did not meet the criteria specified in
Section 3.1. These same sites without annual averages on the inter-site variability graphs are not
represented by a data point on the corresponding CV graphs. 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. The annual averages for the
sites sampling only SNMOCs are not included in the graphs for benzene, 1,3-butadiene, or
ethylbenzene.
The CV figures show that few of the pollutants appear to exhibit the "clustering"
discussed in Section 3.4.2. Figure 4-6a for carbon tetrachloride exhibits clustering, or uniformity
in concentrations. 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
4-35
-------
banned at the Kyoto Protocol. 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 CVs shown in Figure 4-6a not only support the expected uniformity
(i.e., lack of variability) in "background" concentrations of carbon tetrachloride, but are also a
testament to the representativeness of the data generated under the NMP. Figure 4-6b supports
what is shown in Figure 4-6a. The inter-site variability is relatively low, with the annual average
concentrations of carbon tetrachloride ranging from 0.64 |ig/m3 for SSSD to 0.71 |ig/m3 for
DEMI. Further, the confidence intervals for all sites shown are less than ± 0.05 |ig/m3.
Figure 4-8a shows that 1,2-dichloroethane also exhibits clustering, and is supported in
Figure 4-8b by the relatively small differences in the annual averages and confidence intervals
shown for every site. The annual average concentrations of 1,2-dichloroethane ranged from
0.061 |ig/m3 for UNVT to 0.083 |ig/m3 for S4MO. Further, the confidence intervals for all sites
shown are less than ± 0.014 |ig/m3. However, the program-level average concentration
(approximately 0.15 |ig/m3), as indicated by the yellow shading, is roughly twice the site-specific
annual averages shown. This is because data for all sites are included in the program-level
averages, not just those with valid annual averages; thus, one or more sites without an annual
average shown in Figure 4-8b are driving the program-level average. A review of the data shows
that concentrations of 1,2-dichloroethane from the Calvert City, Kentucky sites are driving this
program-level average. These five sites account for the highest 56 measurements of this
pollutant, which range from 0.18 |ig/m3 to 17.1 |ig/m3. Annual averages for these sites could not
be calculated because they did not begin sampling until July 2012 under the NMP. However, the
average concentrations for 1,2-dichloroethane for these sites over the period of sampling is
provided in Appendix J.
Hexchloro-1,3-butadiene is another pollutant that exhibits clustering. Figure 4-12a shows
that the annual average concentrations have a very small range, ranging from 0 |ig/m3 (SEWA) to
0.02 |ig/m3 (NBNJ). Figure 4-12a also shows that the standard deviations tended to be higher
than the averages themselves, the exception being for the site that did not detect this pollutant
(SEWA). Hexchloro-1,3-butadiene was detected in fewer than 15 percent of samples collected,
resulting in a large number of zero substitutions. Thus, the standard deviations are relatively
large, as are the associated confidence intervals shown in Figure 4-12b. Even the site with the
4-36
-------
highest annual average concentration of hexachloro-1,3-butadiene (NBNJ) detected this pollutant
in fewer than one quarter of the samples collected.
The CVs for several of the program-level pollutants of interest follow a linear trend line.
Examples of pollutants whose annual average concentrations exhibit this trend include
acetaldehyde, benzene, fluorene, formaldehyde, and naphthalene. This means that as the annual
averages increase, so do the standard deviations, indicating increasing variability. This increased
variability is often a result of an increased range of individual measurements that are used to
calculate the annual average. This is supported by the corresponding inter-site variability graphs
for each pollutant. The site-specific annual averages that extend well above the program-level
average concentration for each pollutant tend to have a larger confidence interval associated with
them, indicating a wider range of measurements and the possible influence of outliers. The
annual averages considerably less than the program-level average concentration tend to have
much smaller confidence intervals. Figures 4-10a and 4-10b for fluorene and Figures 4-14a and
4-14b for naphthalene are good examples of this trend. The higher annual averages for sites such
as CELA, DEMI, and GPCO have large confidence intervals associated with them while sites
such as CHSC, GLKY, and UNVT have significantly lower annual averages as well as very
small confidence intervals. To illustrate this point, the range of measured detections of fluorene
for GPCO is 1.93 ng/m3 to 68.2 ng/m3 while the range of measurements for GLKY was 0 ng/m3
to 2.36 ng/m3.
Some of the pollutants' annual averages follow a linear pattern, but one of the annual
average concentrations is significantly higher than the annual average concentrations for the
other sites, one of the standard deviations is significantly higher than other sites, or both.
Examples of this include acetaldehyde, arsenic, benzene. Figures 4-4a and 4-4b show that the
annual average benzene concentration for TOOK is more than 85 percent higher than the next
highest annual average concentration of this pollutant. A review of TOOK's benzene data shows
that all but eight of TOOK's preprocessed daily measurements (out of 60) were greater than the
program-level average concentration of 0.90 |ig/m3. Thus, concentrations of benzene at TOOK
tend to run higher than at other sites. Although the annual average concentration of acetaldehyde
for SPIL does not stand out in Figure 4-2b, its standard deviation is more than twice the standard
deviations for the other monitoring sites. A review of this site's data shows that the three highest
concentrations of acetaldehyde across the program were measured at SPIL, ranging from
4-37
-------
8.74 |ig/m3 to 20.4 |ig/m3, yet more than 75 percent of the measurements from this site are less
than 3 |ig/m3. The confidence interval for SPIL's acetaldehyde annual average is reflecting the
influence of the higher concentrations. Figure 4-3 a shows that one of the annual average
concentrations of arsenic is considerably higher than the annual averages calculated for other
sites and has a relatively high standard deviation associated with it. Figure 4-3b shows that this
CV is based on arsenic data for ASKY-M. Although the maximum arsenic concentration
measured across the program was not measured at ASKY-M, this site has the greatest number of
arsenic concentrations greater than 2 ng/m3 (23, vs. the next highest site, LEKY at 6). These
account for nearly half of the samples collected at this site. However, arsenic concentrations
measured at this site range from 0.1 ng/m3 to 5.90 ng/m3, explaining the variability reflected in
the CV.
Figure 4-7a for/>-dichlorobenzene is an example where a relatively high annual average
and/or a relatively high confidence interval are affecting the graph. If the CVs for S4MO,
ADOK, PXSS, and SPAZ were removed from Figure 4-7a, this graph would exhibit easily
identifiable clustering. Figure 4-7b shows that the confidence intervals for S4MO, ADOK, and
SPAZ are relatively large, indicating that these annual averages may be influenced by outlier(s).
Collectively, these three sites account for all 13 measurements ofp-dichlorobenzene greater than
0.5 |ig/m3 measured across the program. Conversely, the confidence interval for PXSS is smaller
than the others, indicating that concentrations of this pollutant may run higher on a more regular
basis. Another consideration for SPAZ is the sampling frequency. VOC sampling at SPAZ
occurs on a l-in-12 day schedule, resulting in fewer overall samples and generally a higher
confidence interval. The calculation of the median concentration for all four datasets completes
the story. The median concentrations for ADOK and S4MO are 0.060 |ig/m3 and 0.096 |ig/m3,
respectively, which are roughly half their annual averages (0.125 |ig/m3 and 0.180 |ig/m3),
indicating that concentrations at the higher end of the range are driving the average
concentrations. Conversely, the median concentrations for PXSS and SPAZ are 0.160 |ig/m3 and
0.217 |ig/m3, respectively, which are more similar to their respective annual average
concentrations of 0.195 |ig/m3 and 0.258 |ig/m3, indicating that the/>-dichlorobenzene
concentrations measured at the two Phoenix sites tend to run higher on a regular basis.
4-38
-------
Figure 4-la. Coefficient of Variation Analysis of Acenaphthene Across 20 Sites
10 15 20
Annual Average Concentration (ng/m3)
Figure 4-lb. Inter-Site Variability for Acenaphthene
rii
rii
Monitoring Site
D Site-Specific Average
4-39
-------
Figure 4-2a. Coefficient of Variation Analysis of Acetaldehyde Across 28 Sites
i
i
s
3
y=0.5762x-0.0971
R2 = 0.588
0°
1.5 2
Annual Average Concentration {jig/m3]
Figure 4-2b. Inter-Site Variability for Acetaldehyde
Program Average
D Site-Specific Ave rage
4-40
-------
Figure 4-3a. Coefficient of Variation Analysis of Arsenic Across 19 Sites
y=0.8055x-0.037
R2 = 0.8693
OOO
o _-.---"
Q.-O"
o-
y=0.4269x+0.0397
R2 = 0.6909
0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8
Annual Average Concentration {ng/m3)
O PM10
O TSP
- Linear (PM10)
Linear(TSP)
Figure 4-3. Inter-Site Variability for Arsenic
ASKY-M BAKY BOMA BTUT CCKY GLKY LEKY NBIL PAFL PXSS S4MO SEWA SJJCA UNVT ADOK OCOK PROK TMOK TOOK
Monitoring Site
D Program PM10 Average
D Program TSP Average
Site-Specific Average
4-41
-------
Figure 4-4a. Coefficient of Variation Analysis of Benzene Across 22 Sites
i"
y=0.5695x-0.0394
R2 = 0.8537
1 1.5
Annual Average Concentration {jig/m3]
Figure 4-4b. Inter-Site Variability for Benzene
_ 2
"E
1
o
S
S 1.5
s
I
il
il
il
Monitoring Site
Program Average
D Site-Specific Average
4-42
-------
Figure 4-5a. Coefficient of Variation Analysis of 1,3-Butadiene Across 22 Sites
i
i
|c
&
y=0.658x +0.0146
R2 = 0.7246
0.1 0.15 0.2
Annual Average Concentration {jig/m3]
Figure 4-5b. Inter-Site Variability for 1,3-Butadiene
rfl
Monitoring Site
Program Average D Site-Specific Average
4-43
-------
Figure 4-6a. Coefficient of Variation Analysis of Carbon Tetrachloride Across 22 Sites
I
I
s
3
y=0.1841x-0.0229
R2 = 0.0242
0.4 0.6
Annual Average Concentration {jig/m3]
Figure 4-6b. Inter-Site Variability for Carbon Tetrachloride
3! 0.6
o
S
I
I
$0.4
s
**
hh
A
1
rli
-ti
Monitoring Site
Program Average D Site-Specific Average
4-44
-------
Figure 4-1 a. Coefficient of Variation Analysis of /7-Dichlorobenzene Across 22 Sites
— 0.25
i
o
0.15 0.2 0.25
Annual Average Concentration {jig/m3]
Figure 4-7b. Inter-Site Variability for/7-Dichlorobenzene
Monitoring Site
Program Average D Site-Specific Average
4-45
-------
Figure 4-8a. Coefficient of Variation Analysis of 1,2-Dichloroethane Across 22 Sites
E 0.06
I
0.01 0.02 0.03
0.04 0.05 0.06
Annual Average Concentration {jig/m3]
0.07 0.08 0.09
Figure 4-8b. Inter-Site Variability for 1,2-Dichloroethane
Monitoring Site
Program Average D Site-Specific Average
4-46
-------
Figure 4-9a. Coefficient of Variation Analysis of Ethylbenzene Across 22 Sites
r
I
.| 0.5
&
y=0.6572x-0.0034
R2 = 0.7854
0.1 0.2
0.4 0.5 0.6 0.7
Annual Average Concentration {jig/m3]
0.8 0.9
Figure 4-9b. Inter-Site Variability for Ethylbenzene
^E
o
IS
s
S
S
<0.4
0.2 -
i r
rr.
ft
i
±1 r
^
1
if
h
••
••
rh
p-
ft
ii
di
ft
r-
ft
T
rn
J-
—
Q
Monitoring Site
Program Average
D Site -Specific Average
4-47
-------
Figure 4-10a. Coefficient of Variation Analysis of Fluorene Across 20 Sites
t12
y=1.156x-1.5977
R2 = 0.891
8 10 12 14
Annual Average Concentration (ng/m3)
16 18
Figure 4-10b. Inter-Site Variability for Fluorene
_ 12
"E
5 10
s
I
I 8
n
n
rii
-in
rfi
il
O
/./
Monitoring Site
Program Average D Site-Specific Average
4-48
-------
Figure 4-lla. Coefficient of Variation Analysis of Formaldehyde Across 28 Sites
r
y=0.5415x-0.0411 _
R2= 0.7275 U
2 2.5 3
Annual Average Concentration {jig/m3]
Figure 4-1 Ib. Inter-Site Variability for Formaldehyde
Monitoring Site
Program Average
D Site-SpecificAverage
4-49
-------
Figure 4-12a. Coefficient of Variation Analysis of Hexachloro-l,3-Butadiene Across 22 Sites
I
I
s
3
y=1.6785x
R2 = 0.8377
0.02 0.03
Annual Average Concentration {jig/m3]
Figure 4-12b. Inter-Site Variability for Hexachloro-l,3-butadiene
1
i
Monitoring Site
Program Average D Site-Specific Average
4-50
-------
Figure 4-13a. Coefficient of Variation Analysis of Manganese Across 19 Sites
E
If
— 25
.e
I
3
•o 20
15 20 25 30
Annual Average Concentration (ng/m3)
O PM10
O TSP
- Linear (PM10)
Linear(TSP)
Figure 4-13b. Inter-Site Variability for Manganese
I
is
= 25-1-
ASKY-M BAKY BOMA BTUT CCKY GLKY LEKY NBIL PAFL PXSS S4MO SEWA SJJCA UNVT ADOK OCOK PROK TMOK TOOK
Monitoring Site
D Program PM10 Average
D Program TSP Average
Site-Specific Average
4-51
-------
Figure 4-14a. Coefficient of Variation Analysis of Naphthalene Across 20 Sites
-
s
3
^ 100
y=0.5931x
R' = 0.8645
o o
75 100 125 150
Annual Average Concentration (ng/m3)
Figure 4-14b. Inter-Site Variability for Naphthalene
rfl
il
0
n
t
rh
Program Average
Monitoring Site
D Site-Specific Average
4-52
-------
Figure 4-15a. Coefficient of Variation Analysis of Nickel Across 19 Sites
X 1.5
_n_
_ _, — x- y=0.8921x-0.3222
O ^5Q ° R2 = "^09
1.5 2 2.5
Annual Average Concentration (ng/m3)
O PM10
O TSP
- Linear (PM10)
Linear (TSP)
Figure 4-15b. Inter-Site Variability for Nickel
ASKY-M BAKY BOMA BTUT CCKY GLKY LEKY NBIL PAFL PXSS S4MO SEWA SJJCA UNVT ADOK OCOK PROK TMOK TOOK
Monitoring Site
D Program PM10 Average
D Program TSP Average
Site-Specific Average
4-53
-------
4.4.2 Quarterly Variability Analysis
Figures 4-16 through 4-30 provide a graphical display of the site-specific quarterly
average concentrations for each of the program-level pollutants of interest. Quarterly averages
are calculated based on the criteria specified in Section 3.1. If the pollutant of interest has a
corresponding ATSDR Intermediate MRL, as defined in Section 3.3, then this value is indicated
on the graph and is plotted where applicable. Note that the scales on the PMio and TSP graphs
are the same for a given speciated metal.
Data gaps, or missing quarterly averages, 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 thus have a quarterly average concentration of zero as
a result of the substitution of zeros for non-detects. One example of this is Figure 4-27 for
hexachloro-1,3-butadiene. This pollutant was infrequently detected (186 measured detections out
of 1,466 valid samples); of the 102 possible quarterly averages of this pollutant, 30 of them are
zero. Thus, relatively few quarterly averages appear in Figure 4-27. Further, most of the
remaining quarterly averages have relatively few measured detections and include many zero
substitutions for non-detects, resulting in relatively low quarterly averages. (Although this
pollutant was detected in less than 13 percent of VOC samples collected, its risk screening value
is relatively low; thus, 70 percent of the measured detections of this pollutant failed screens.)
Another reason for data gaps in the figures is due to the sampling duration of each site.
Some sites started late or ended early, which may result in a lack of quarterly averages. For
example, benzene is almost always detected in VOC samples, thus the gaps in Figure 4-19 are
primarily due to sampling duration. Many of the Kentucky sites started sampling VOCs in July
2012; thus, the first and second quarterly averages could not be calculated and therefore appear
as gaps in the figure.
In addition, the criteria 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). GPCO
experienced sampling issues for VOCs during the month of August which led to the invalidation
of several samples. As a result, there were fewer than 12 valid samples during the third quarter of
2012 and thus no third quarter benzene average could be calculated for GPCO in Figure 4-19.
4-54
-------
Some pollutants of interest, such as acetaldehyde, benzene, carbon tetrachloride,
ethylbenzene, formaldehyde, and naphthalene, were detected year-round. Comparing the
quarterly averages for sites with four valid quarterly averages in a year may reveal a temporal
trend for these pollutants. For example, formaldehyde averages tended to be highest during the
third quarter, as shown in Figure 4-26, with 27 of the 37 sites sampling formaldehyde exhibiting
the highest quarterly average for the period from July through September (although quarterly
averages could not be calculated for every quarter for every site). Thus, it appears that
formaldehyde concentrations tend to be highest during the summer months. Conversely,
1,3-butadiene averages tended to be higher during the fourth quarter of 2012, as shown in
Figure 4-20. Twenty-six of 35 sites have their highest quarterly 1,3-butadiene concentration for
the fourth quarter. However, several of the sites shown in Figure 4-20 did not begin sampling
until halfway through the year. Of the 20 sites with four quarterly 1,3-butadiene averages
presented in Figure 4-20, 16 have the fourth quarter average as the maximum quarterly average
concentration.
Other notable trends include benzene with higher concentrations in the first and fourth
quarters and acenaphthene, acetaldehyde, and fluorene with higher concentrations in the third
quarter. Arsenic tended to be highest during the second quarter for all five sites sampling TSP
metals, although a similar trend is not shown for the sites sampling PMio metals.
The quarterly average 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. This is evident in Figures 4-21, 4-22, 4-23, and 4-30a
for carbon tetrachloride, />-dichlorobenzene, 1,2-dichloroethane, and nickel, respectively, to
name a few. For example, Figure 4-23 shows that the quarterly averages of 1,2-dichloroethane
for the Calvert City, Kentucky sites (ATKY, BLKY, CCKY, LAKY, and TVKY) are
significantly higher than for other sites sampling VOCs, as most of the other bars are less than
the first gridline on the graph. Figure 4-22 shows that the fourth quarter average concentration of
/7-dichlorobenzene for SPAZ is significantly higher than this site's other quarterly averages as
well as most other sites' quarterly averages. Similarly, SEWA's third quarter average
concentration of nickel is more than twice this site's other quarterly averages and is the highest
quarterly average calculated for this pollutant.
4-55
-------
Figure 4-21 shows that the quarterly average concentrations of carbon tetrachloride that
are available for TVKY (this site did not begin sampling VOCs until July) are significantly
higher than all of the other sites sampling VOCs. These graphs may also reveal when there is
very little variability in the quarterly averages across other sites. Figure 4-21 for carbon
tetrachloride also shows that the quarterly averages of this pollutant did not vary significantly
across the sites, with the exception of TVKY.
Other notable trends are revealed in these graphs. For example, SPAZ and PXSS have
relatively high fourth quarter average concentrations for four of the VOC pollutants of interest
(benzene, ethylbenzene, 1,3-butadiene, and/?-dichlorobenzene), compared to their other
quarterly averages and most other NMP sites. While benzene tended to be highest during the first
and fourth quarters of 2012 for most sites, the second and third quarter averages were highest for
TOOK, the site with the highest annual average concentration of this pollutant, as shown in
Figure 4-19. For ASKY-M, the second quarter average concentrations of the PMio metal
pollutants of interest were considerably higher than the other quarterly averages for this site
(although sampling began in March so no first quarter average is available). In the case of
arsenic, both the second and third quarter average concentrations for ASKY-M are much higher
than the other sites' quarterly averages in Figure 4-18a.
These graphs also show that only six of the 16 program-level pollutants of interest have
ATSDR Intermediate MRLs. For the six that do, the quarterly average concentrations are
significantly less than their respective ATSDR Intermediate MRLs, generally by an order of
magnitude or more, which is also discussed in Section 4.2.2. In all six cases, the ATSDR
Intermediate MRL is considerably greater than the scale on the graph and is provided in a text
box rather than plotted in the figure.
4-56
-------
Figure 4-16. Comparison of Average Quarterly Acenaphthene Concentrations
Monitoring Site
llstQuarter
!2ndQuarter
SrdQuarter
4thQuarter
-------
Figure 4-17. Comparison of Average Quarterly Acetaldehyde Concentrations
J^.
I
oo
Monitoring Site
11st Quarter
12nd Quarter
3rd Quarter
14th Quarter
-------
Figure 4-18a. Comparison of Average Quarterly Arsenic (PMi0) Concentrations
0.2
ASKY-M BAKY BOMA BTUT CCKY GLKY
LEKY NBIL
Monitoring Site
PAFL PXSS S4MO SEWA SJJCA UNVT
llstQuarter
12nd Quarter
3rd Quarter
14th Quarter
-------
Figure 4-18b. Comparison of Average Quarterly Arsenic (TSP) Concentrations
J^.
I
o
ADOK
OCOK
PROK
Monitoring Site
TMOK
TOOK
llstOuarter
12nd Quarter
3rd Quarter
14th Quarter
-------
Figure 4-19. Comparison of Average Quarterly Benzene Concentrations
| ATSDRI intermediate MRL= 20 pig/m3
llstQuarter
Monitoring Site
12nd Quarter • 3rd Quarter
14th Quarter
-------
Figure 4-20. Comparison of Average Quarterly 1,3-Butadiene Concentrations
J^.
I
to
Monitoring Site
llstQuarter
12nd Quarter
3rd Quarter
14th Quarter
-------
Figure 4-21. Comparison of Average Quarterly Carbon Tetrachloride Concentrations
1.4
1.2
| ATSDR Intermediate MRL= 200 pig/m
...J
llstQuarter
Monitoring Site
12nd Quarter • 3rd Quarter • 4th Quarter
-------
Figure 4-22. Comparison of Average Quarterly /7-Dichlorobenzene Concentrations
0.6 TT^
i ATSDR Intermediate MRL= 1,000
llstQuarter
Monitoring Site
12nd Quarter • 3rd Quarter
14th Quarter
-------
Figure 4-23. Comparison of Average Quarterly 1,2-Dichloroethane Concentrations
j 3rd Quarter= 2.77 u.g/m3
4th Quarter = 1.91 u.g/m3
llstQuarter
Monitoring Site
12nd Quarter • 3rd Quarter
14th Quarter
-------
Figure 4-24. Comparison of Average Quarterly Ethylbenzene Concentrations
Oi
Oi
1.6
i ATSDR Intermediate MRL= 9,000 pig/m3 \
L...
Monitoring Site
llstQuarter
12nd Quarter
3rd Quarter
14th Quarter
-------
Figure 4-25. Comparison of Average Quarterly Fluorene Concentrations
Monitoring Site
llstQuarter
!2ndQuarter
SrdQuarter
4thQuarter
-------
Figure 4-26. Comparison of Average Quarterly Formaldehyde Concentrations
J^.
I
oo
=-5
• ATSDR Intermediate MRL= 40 pig/m3
Monitoring Site
11st Quarter
12nd Quarter
3rd Quarter
14th Quarter
-------
Figure 4-27. Comparison of Average Quarterly Hexachloro-l,3-butadiene Concentrations
0.045
0.04
llstQuarter
Monitoring Site
12nd Quarter • 3rd Quarter
14th Quarter
-------
Figure 4-28a. Comparison of Average Quarterly Manganese (PMi0) Concentrations
J^.
I
o
ASKY-M
BAKY
BOMA
BTUT
CCKY
GLKY
LEKY NBIL
Monitoring Site
PAFL
PXSS
S4MO
SEWA
SJJCA
UNVT
llstQuarter
12nd Quarter
3rd Quarter
14th Quarter
-------
Figure 4-28b. Comparison of Average Quarterly Manganese (TSP) Concentrations
ADOK
OCOK
PROK
Monitoring Site
TMOK
TOOK
llstOuarter
12nd Quarter
3rd Quarter
14th Quarter
-------
Figure 4-29. Comparison of Average Quarterly Naphthalene Concentrations
300
J^.
I
to
Monitoring Site
llstQuarter
1 2nd Quarter
3rd Quarter
4th Quarter
-------
Figure 4-30a. Comparison of Average Quarterly Nickel (PMi0) Concentrations
ATSDR Intermediate MRL= 200 ng/m3
ASKY-M BAKY BOMA BTUT CCKY GLKY
LEKY NBIL
Monitoring Site
PAFL PXSS S4MO SEWA SJJCA UNVT
11st Quarter
12nd Quarter
3rd Quarter
14th Quarter
-------
Figure 4-30b. Comparison of Average Quarterly Nickel (TSP) Concentrations
j ATSDR Intermediate MRL= 200 ng/m3
ADOK
OCOK
PROK
Monitoring Site
TMOK
TOOK
11st Quarter
12nd Quarter
3rd Quarter
14th Quarter
-------
4.5 Greenhouse Gases
Table 4-15 presents the program-level average concentrations for the 10 GHGs measured
using Method TO-15, in descending order by GWP. As shown, most of the GHGs were detected
in nearly every sample collected (a total 1,466 valid VOC samples). Chloroform, bromomethane,
and 1,1,1-trichloroethane were the only pollutants detected in less than 95 percent of VOC
samples collected, although even these were detected in greater than 50 percent of samples.
Dichlorodifluoromethane and dichlorotetrafluoroethane have the highest GWPs of the GHGs
measured by Method TO-15 (10,900 and 10,000 respectively), while bromomethane and
dichloromethane have the lowest GWPs (5 and 9, respectively).
Dichloromethane has the highest program-level average concentration among the GHGs
measured (2.53 ± 1.29 |ig/m3), although the program-level average concentration for
dichlorodifluoromethane is similar in magnitude (2.49 ± 0.01 jig/m3). The confidence interval for
dichloromethane indicates that this concentration is likely influenced by outliers, while the
confidence interval for dichlorodifluoromethane indicates little variability. A review of the data
shows that high concentrations for a few sites contributed to this dichloromethane average
concentration. The highest concentrations of this pollutant were measured at GPCO and ranged
from 124 |ig/m3 to 745 |ig/m3. An additional concentration of 153 |ig/m3 was also measured at
BTUT. However, the median concentration of this pollutant is less than 0.5 |ig/m3, indicating
that these high concentrations are the exception and not the rule. The median concentration for
dichlorodifluoromethane (2.47 |ig/m3) is very similar to its program average; the similarities in
these two calculations indicate little variability in the central tendency of this pollutant. Besides
dichloromethane and dichlorodifluoromethane, only two additional GHGs shown in Table 4-15
have program-level average concentrations greater than 1 |ig/m3: trichlorofluoromethane and
chloromethane.
4-75
-------
Table 4-15. Greenhouse Gases Measured by Method TO-15
Pollutant
Dichlorodifluoromethane
Dichlorotetrafluoroethane
Trichlorotrifluoroethane
Trichlorofluoromethane
Carbon Tetrachloride
1,1,1 -Trichloroethane
Chloroform
Chloromethane
Dichloromethane
Bromomethane
Global
Warming
Potential1
(100 yrs)
10,900
10,000
6,130
4,750
1,400
146
31
13
9
5
Total # of
Measured
Detections2
1,466
1,466
1,466
1,466
1,466
1,268
948
1,466
1,464
1,183
2012
Program
Average
(Hg/m3)
2.49
±0.01
0.12
±<0.01
0.65
±0.01
1.52
±0.02
0.69
±0.01
0.04
±<0.01
0.24
±0.09
1.18
±0.01
2.53
±1.29
0.04
±<0.01
:GWPs presented here are from the Intergovernmental Panel on Climate Change
(IPCC) Fourth Assessment Report (AR4) (IPCC, 2012).
2 Out of 1,466 valid samples
4-76
-------
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. 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 radii are still visible on the map, but have been grayed out in order to show emissions
sources just outside 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
to
W Maripos
-------
Figure 5-2. South Phoenix, Arizona (SPAZ) Monitoring Site
-------
Figure 5-3. NEI Point Sources Located Within 10 Miles of PXSS and SPAZ
112'25'0"W 112°20'CrW 112°15'0"W 112°10'0"W 112"5'0"W 112°0'0'W
112155'0"W 11230'0"W 111 °55'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
Legend
PXSS NATTS site
SPAZ UATMP site O 10 mile radius
County boundary
Source Category Group (No. of Facilities)
^f Aerospace/Aircraft Manufacturing (2)
"1" Airport/Airline/Airport Support Operations (36)
it Asphalt Production/Hot Mix Asphalt Plant (2)
X Battery Manufacturing (1)
B Bulk Terminals/Bulk Plants (3)
C Chemical Manufacturing (4)
6 Electrical Equipment Manufacturing (6)
* Electricity Generation via Combustion (4)
E Electroplating. Plating, Polishing, Anodizing, and Coloring (1) W
Foundries, Iron and Steel (1)
Metal Can, Box, and Other Metal Container Manufacturing (1)
Metal Coating, Engraving, and Allied Services to Manufacturers (1)
Metals Processing/Fabrication (4)
Mine/Quarry/Mineral Processing (1)
Miscellaneous Commercial/Industrial (5)
Plastic, Resin, or Rubber Products Plant (5)
Rail Yard/Rail Line Operations (3)
Woodwork, Furniture. Millwork & Wood Preserving (2)
5-4
-------
Table 5-1. Geographical Information for the Arizona Monitoring Sites
Site
Code
PXSS
SPAZ
AQS Code
04-013-9997
04-013-4003
Location
Phoenix
Phoenix
County
Maricopa
Maricopa
Micro- or
Metropolitan
Statistical Area
Phoenix-Mesa-
Scottsdale, AZ
MSA
Phoenix-Mesa-
Scottsdale, AZ
MSA
Latitude
and
Longitude
33.503833,
-112.095767
33.40316,
-112.07533
Land Use
Residential
Residential
Location
Setting
Urban/City
Center
Urban/City
Center
Additional Ambient Monitoring Information1
Haze, CO, SO2, NO, NO2, NOX, NOy, PAMS, O3,
Meteorological parameters, PM10, PM Coarse, PM25,
PM2 5 Speciation, IMPROVE Speciation.
CO, O3, Meteorological parameters, PM10,PM Coarse,
PM25.
:Data for additional pollutants are reported to AQS for these sites (EPA, 2013b); however, these data are not generated by ERG and are therefore not included in this report.
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 3/4 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/I-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.
Table 5-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Arizona monitoring sites. Table 5-2 includes both county-level
population and vehicle registration information. Table 5-2 also contains traffic volume
information for each site as well as the location for which the traffic volume was obtained.
Additionally, Table 5-2 presents the county-level daily VMT for Maricopa County.
Table 5-2. Population, Motor Vehicle, and Traffic Information for the Arizona
Monitoring Sites
Site
PXSS
SPAZ
Estimated
County
Population1
3,942,169
County-level
Vehicle
Registration2
3,761,859
Annual
Average Daily
Traffic3
184,000
128,000
Intersection
Used for
Traffic Data
1-17 b/w Exits 202 and 203
1-17 b/w Exits 195B and 196
County-
level Daily
VMT4
90,393,000
County-level vehicle registration reflects 2012 data (AZ DOT, 2011)
3AADT reflects 2010 data (AZ DOT, 2010)
4County-level VMT reflects 2012 data (AZ DOT, 2013)
BOLD ITALICS = EPA-designated NATTS Site
5-6
-------
Observations from Table 5-2 include the following:
• Maricopa County has the fourth highest county-level population and second highest
county-level vehicle registration compared to other counties with NMP sites.
• PXSS experiences a higher traffic volume compared to SPAZ, based on locations
along 1-17. The traffic volume near PXSS is the sixth highest compared to traffic
volumes near other NMP sites, with the traffic volume near SPAZ ranking 12th.
• The daily VMT for Maricopa County is the second highest compared to other
counties with NMP sites (where VMT data were available).
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 Climate Summary
Phoenix is located in the Salt River Valley, which is part of the Sonora Desert. The area
experiences mild winters and extremely hot and dry summers. Differences between the daytime
maximum temperature and overnight minimum temperature can be as high as 50°F. A summer
"monsoon" period brings precipitation to the area for part of the summer, while storm systems
originating over the Pacific Ocean bring rain in the winter and early spring. However, normal
monthly rainfall totals are generally less than one inch. Winds are generally light and out of the
east for much of the year (Wood, 2004; WRCC, 2013).
5.2.2 Meteorological Summary
Hourly meteorological data for 2012 were retrieved from NCDC for the weather station
closest to the Arizona monitoring sites (NCDC, 2012), as described in Section 3.5.2. The closest
weather station to both PXSS and SPAZ is located at Phoenix Sky Harbor International Airport
(WBAN 23183). Additional information about the Phoenix Sky Harbor weather station, such as
the distance between the sites and the weather station, is provided in Table 5-3. These data were
used to determine how meteorological conditions on sample days vary from conditions
experienced throughout the year.
5-7
-------
Table 5-3. Average Meteorological Conditions near the Arizona Monitoring Sites
Closest Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Phoenix, Arizona - PXSS
Phoenix Sky
Harbor Intl.
Airport
23183
(33.44, -111.99)
7.3
miles
146°
(SE)
Sample
Days
(69)
2012
88.6
±3.8
87.7
±1.6
77.4
±3.8
76.7
± 1.6
37.6
±3.2
36.8
±1.5
56.5
±2.5
56.0
±1.0
28.7
±3.3
28.3
±1.4
1011.7
±1.2
1011.4
±0.5
5.0
±0.5
5.2
±0.2
South Phoenix, Arizona - SPAZ
Phoenix Sky
Harbor Intl.
Airport
23183
(33.44, -111.99)
4.3
miles
77°
(ENE)
Sample
Days
(31)
2012
86.4
±6.0
87.7
±1.6
75.4
±5.9
76.7
±1.6
37.9
±5.0
36.8
±1.5
55.7
±3.9
56.0
±1.0
30.0
±4.2
28.3
±1.4
1011.8
±1.8
1011.4
±0.5
5.2
±0.8
5.2
±0.2
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
oo
-------
Table 5-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for all of 2012. Also included in Table 5-3 is the 95 percent
confidence interval for each parameter. As shown in Table 5-3, average meteorological
conditions on sample days were representative of average weather conditions experienced
throughout the year. Even though the observations are from the same weather station, roughly
two degrees separates the sample day averages for the maximum and average temperatures for
PXSS and SPAZ. This is primarily due to the sampling schedule. 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 number of sample days for each site is
provided in Table 5-3. Some of the hottest sampling days of 2012 for PXSS were days sampling
did not occur at SPAZ. The difference in the number of observations included in the calculations
is also reflected in the larger confidence intervals for SPAZ, as is the increased variability in the
observations themselves. These sites experienced the lowest relative humidity level and sea level
pressures among all NMP sites. Temperatures were also warmest near these sites.
5.2.3 Back Trajectory Analysis
Figure 5-4 is the composite back trajectory map for days on which samples were
collected at the PXSS monitoring site. Included in Figure 5-4 are four back trajectories per
sample day. Figure 5-5 is the corresponding cluster analysis. Similarly, Figures 5-6 and 5-7 are
the composite back trajectory map and corresponding cluster analysis for days on which samples
were collected at SPAZ. An in-depth description of these maps and how they were generated is
presented in Section 3.5.2.1. For the composite maps, each line represents the 24-hour trajectory
along which a parcel of air traveled toward the monitoring site on a given sample day and time,
based on an initial height of 50 meters AGL. For the cluster analyses, each line corresponds to a
trajectory representative of a given cluster of back trajectories. Each concentric circle around the
sites in Figures 5-4 through 5-7 represents 100 miles.
5-9
-------
Figure 5-4. Composite Back Trajectory Map for PXSS
Figure 5-5. Back Trajectory Cluster Map for PXSS
D
\ ^ \
100 200
^^1 "" \ '
y A
400 ^
^M Miles - - v |_ ^ - ' / ,
5-10
-------
Figure 5-6. Composite Back Trajectory Map for SPAZ
Figure 5-7. Back Trajectory Cluster Map for SPAZ
5-11
-------
Observations from Figures 5-4 and 5-5 for PXSS include the following:
• The 24-hour air shed domain for PXSS is among the smallest in size, based on
average back trajectory length, compared to other NMP sites. Only the Colorado
monitoring sites have smaller air shed domains than PXSS. The farthest away a back
trajectory originated from PXSS was off the coast of California and over the Channel
Islands, or just greater than 450 miles away. However, most back trajectories
(93 percent) originated less than 250 miles from PXSS and the average trajectory
length was approximately 141 miles.
• Back trajectories originated from a variety of directions at PXSS, although many back
trajectories originated from the southwest and west. Back trajectories also originated
from the north, northeast, and east of the site. Few back trajectories originated from
the northwest or south.
• The cluster analysis map supports the observations above regarding the direction of
trajectory origin as well as the observations about trajectory distances. Nearly
40 percent of back trajectories originated to the southwest and west of PXSS, over
southwest Arizona, southern California, and Baja California, Mexico. The short
cluster trajectory (33 percent) represents back trajectories originating from nearly all
directions, but generally over southwest and central Arizona. Another 12 percent of
back trajectories originated over the northern half of the state while 16 percent
originated to the northeast, east, and southeast of the site.
Observations from Figures 5-6 and 5-7 for SPAZ include the following:
• Samples were collected every 12 days at SPAZ, which is half the frequency of sample
collection at PXSS, as discussed in Section 5.2.2. As a result, fewer back trajectories
are shown in Figure 5-6 than Figure 5-4.
• The 24-hour air shed domain for SPAZ is similar in size to the air shed domain for
PXSS, based on average back trajectory length. The farthest away a back trajectory
originated from SPAZ was off the coast of California and over the Channel Islands,
or just greater than 450 miles away. However, most trajectories (91 percent)
originated less than 250 miles from SPAZ and the average trajectory length was
approximately 147 miles.
• The composite trajectory map for SPAZ has a trajectory distribution pattern similar to
PXSS. The cluster analysis maps are similar to each other directionally, although their
percentages differ. One cluster trajectory for SPAZ is short enough that it is covered
up by the star symbol; thus, the trajectory is presented in the inset map in Figure 5-7.
This shorter trajectory includes back trajectories of varying directions but generally
short distances.
5-12
-------
5.2.4 Wind Rose Comparison
Hourly surface wind data from the weather station at Phoenix Sky Harbor International
Airport were uploaded into a wind rose software program to produce customized wind roses, as
described in Section 3.5.2.2. 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 5-8 presents a map showing the distance between the weather station and PXSS,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 5-8 also presents three different wind roses for the
PXSS monitoring site. First, a historical wind rose representing 2002 to 2011 wind data is
presented, which shows the predominant surface wind speed and direction over an extended
period of time. Second, a wind rose representing wind observations for all of 2012 is presented.
Next, a wind rose representing wind observations for days on which samples were collected in
2012 is presented. These can be used to identify the predominant wind speed and direction for
2012 and to determine if wind observations on sample days were representative of conditions
experienced over the entire year and historically. Figure 5-9 presents the distance map and three
wind roses for SPAZ.
Observations from Figures 5-8 and 5-9 for the Arizona monitoring sites include the
following:
• The weather station at Phoenix Sky Harbor International Airport is the closest
weather station to both PXSS and SPAZ. The Phoenix Sky Harbor weather station is
located 7.3 miles southeast of PXSS and 4.3 miles east-northeast of SPAZ.
• Because the Phoenix Sky Harbor weather station is the closest weather station to both
sites, the historical and 2012 wind roses for PXSS are the same as those for SPAZ.
• The historical wind rose shows that easterly winds were the most commonly observed
winds near PXSS and SPAZ (accounting for approximately 20 percent of wind
observations), followed by westerly (12 percent) and east-southeasterly (9 percent)
winds. Winds from the northwest to north to northeast were infrequently observed, as
were winds from the south-southeast to south-southwest. Calm winds (< 2 knots)
account for 16 percent of the hourly wind measurements from 2002 to 2011.
• The 2012 wind patterns are similar to the historical wind patterns. Further, the sample
day wind patterns for each site resemble both the historical and 2012 wind patterns,
indicating that wind conditions on sample days were representative of those
experienced over the entire year and historically.
5-13
-------
Figure 5-8. Wind Roses for the Phoenix Sky Harbor International Airport Weather Station
near PXSS
Location of PXSS and Weather Station
2002-2011 Historical Wind Rose
Cshns: 15.92%
2012 Wind Rose
Sample Day Wind Rose
WIND SPEED
(Knots)
^| 17-21
^| 11-17
^| 7- 11
HI - •*
Calms: 13.37%
WWD SPEED
(Knots)
[ | >=22
^| 17 - 21
HI 11 - 17
^| 7- 11
5-14
-------
Figure 5-9. Wind Roses for the Phoenix Sky Harbor International Airport Weather Station
near SPAZ
Location of SPAZ and Weather Station
2002-2011 Historical Wind Rose
2012 Wind Rose
Sample Day Wind Rose
5-15
-------
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 5-4. 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 5-4. It is important to note which pollutants were sampled for at each site when reviewing
the results of this analysis. PXSS sampled for VOCs, carbonyl compounds, PAHs, metals
(PMio), and hexavalent chromium; SPAZ sampled for VOCs only.
Table 5-4. Risk-Based Screening Results for the Arizona Monitoring Sites
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Phoenix, Arizona - PXSS
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Formaldehyde
Manganese
Arsenic
Naphthalene
£>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Nickel
Hexavalent Chromium
Propionaldehyde
Benzo(a)pyrene
Hexachloro- 1 ,3 -butadiene
1 , 1 ,2,2-Tetrachloroethane
1 ,2-Dibromoethane
Benzo(b)fluoranthene
Cadmium
Dichloromethane
0.45
0.13
0.03
0.17
0.077
0.005
0.00023
0.029
0.091
0.038
0.4
0.0021
0.000083
0.8
0.00057
0.045
0.017
0.0017
0.0057
0.00056
7.7
Total
61
61
61
61
61
61
58
53
50
47
42
24
9
6
4
4
3
2
1
1
1
671
61
61
61
61
61
61
61
59
60
47
61
61
61
61
36
5
3
2
53
61
61
1,058
100.00
100.00
100.00
100.00
100.00
100.00
95.08
89.83
83.33
100.00
68.85
39.34
14.75
9.84
11.11
80.00
100.00
100.00
1.89
1.64
1.64
63.42
9.09
9.09
9.09
9.09
9.09
9.09
8.64
7.90
7.45
7.00
6.26
3.58
1.34
0.89
0.60
0.60
0.45
0.30
0.15
0.15
0.15
9.09
18.18
27.27
36.36
45.45
54.55
63.19
71.09
78.54
85.54
91.80
95.38
96.72
97.62
98.21
98.81
99.25
99.55
99.70
99.85
100.00
5-16
-------
Table 5-4. Risk-Based Screening Results for the Arizona Monitoring Sites (Continued)
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
South Phoenix, Arizona - SPAZ
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
Ethylbenzene
£>-Dichlorobenzene
Chloroprene
0.13
0.03
0.17
0.038
0.4
0.091
0.0021
Total
30
30
30
26
25
24
1
166
30
30
30
26
30
30
1
177
100.00
100.00
100.00
100.00
83.33
80.00
100.00
93.79
18.07
18.07
18.07
15.66
15.06
14.46
0.60
18.07
36.14
54.22
69.88
84.94
99.40
100.00
Observations from Table 5-4 include the following:
• The number of pollutants failing screens varied significantly between the two
monitoring sites; this is expected given the difference in pollutants measured at each
site.
• Twenty-one pollutants failed at least one screen for PXSS; 63 percent of
concentrations for these 21 pollutants were greater than their associated risk screening
value (or failed screens).
• Twelve pollutants contributed to 95 percent of failed screens for PXSS and therefore
were identified as pollutants of interest for PXSS. These 12 include two carbonyl
compounds, six VOCs, three PMio metals, and one PAH.
• PXSS failed the second highest number of screens (671) among all NMP sites, behind
only S4MO with 692 failed screens (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 26 percent. This is due primarily to the relatively high number of
pollutants sampled for at this site, as discussed in Section 4.2.
• Seven pollutants failed screens for SPAZ; approximately 94 percent of concentrations
for these seven 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;
for PXSS, the percentage of screens failed for each individual pollutant is more
varied.
• Six pollutants 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, 1,3-butadiene, carbon tetrachloride,
and 1,2-dichloroethane failed 100 percent of screens for each site. While other VOCs,
such chloroprene (for SPAZ) and 1,2-dibromoethane (for PXSS), also failed
100 percent of screens, they were detected infrequently.
5-17
-------
• Acetaldehyde, formaldehyde, and manganese also failed 100 percent of screens for
PXSS and were detected in all of the valid samples collected at this site.
5.4 Concentrations
This section presents various concentration averages used to characterize pollution 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. Additional site-specific statistical summaries for PXSS
and SPAZ are provided in Appendices J, L, M, N, and O.
5.4.1 2012 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 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 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
Arizona monitoring sites are presented in Table 5-5, where applicable. Note that concentrations
of the PAHs, metals, and hexavalent chromium 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
5-18
-------
average simply reflects "0" because only zeros substituted for non-detects were factored into the
quarterly average concentration.
Table 5-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Arizona Monitoring Sites
Pollutant
#of
Measured
Detections vs.
# 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)
Phoenix, Arizona - PXSS
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
£>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
Arsenic (PM10)a
Manganese (PM10)a
Naphthalene3
Nickel (PM10)a
61/61
61/61
61/61
61/61
60/61
47/61
61/61
61/61
61/61
61/61
59/59
61/61
2.93
±0.63
1.45
±0.41
0.25
±0.09
0.68
±0.03
0.18
±0.06
0.08
±0.02
0.72
±0.25
3.96
±0.53
0.93
±0.38
27.76
± 13.45
110.69
±27.32
2.39
±0.69
2.80
±0.52
0.99
±0.30
0.11
±0.03
0.69
±0.07
0.16
±0.03
0.09
±0.03
0.58
±0.15
3.84
±0.51
0.60
±0.12
25.58
±7.50
51.62
± 12.04
2.66
± 1.05
2.20
±0.45
0.76
±0.21
0.11
±0.03
0.67
±0.03
0.13
±0.04
0.05
±0.02
0.52
±0.16
3.74
±0.55
0.45
±0.09
16.75
±4.71
45.13
± 12.58
1.45
±0.34
3.72
±0.65
1.97
±0.49
0.42
±0.11
0.69
±0.04
0.31
±0.07
0.06
±0.03
1.11
±0.28
4.30
±0.62
0.75
±0.16
21.31
±4.16
190.04
±43.90
1.70
±0.31
2.90
±0.30
1.28
±0.21
0.22
±0.05
0.68
±0.02
0.20
±0.03
0.07
±0.01
0.73
±0.12
3.96
±0.27
0.68
±0.11
22.75
±4.01
97.83
± 19.46
2.04
±0.34
South Phoenix, Arizona - SPAZ
Benzene
1,3 -Butadiene
Carbon Tetrachloride
£>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
30/30
30/30
30/30
30/30
26/30
30/30
1.62
±0.74
0.27
±0.14
0.61
±0.14
0.17
±0.08
0.08
±0.03
0.85
±0.44
0.99
±0.27
0.13
±0.05
0.65
±0.07
0.17
±0.05
0.08
±0.03
0.62
±0.22
0.85
±0.31
0.13
±0.05
0.65
±0.05
0.22
±0.10
0.06
±0.02
0.56
±0.24
2.40
±0.50
0.53
±0.14
0.67
±0.02
0.50
±0.09
0.09
±0.03
1.41
±0.31
1.43
±0.30
0.26
±0.07
0.65
±0.04
0.26
±0.06
0.08
±0.01
0.84
±0.18
' Average concentrations provided for the pollutants below the blue line are presented in ng/m for ease of viewing.
5-19
-------
Observations for PXSS from Table 5-5 include the following:
• The pollutants of interest with the highest annual average concentrations are
formaldehyde (3.96 ± 0.27 |ig/m3) and acetaldehyde (2.90 ± 0.30 |ig/m3). Benzene is
the only other pollutant of interest with an annual average concentration greater than
1 |ig/m3 (1.28 ± 0.21 |ig/m3) for this site.
• The first and fourth quarter average concentrations for 1,3-butadiene are greater than
the second and third quarter average concentrations, supporting the seasonal tendency
discussed in Section 4.4.2, with higher quarterly averages for the quarters that include
colder months of the year. The quarterly averages for benzene exhibit a similar
tendency.
• The fourth quarter average concentrations of many of PXSS's pollutants of interest
(including acetaldehyde, formaldehyde, /?-dichlorobenzene, ethylbenzene, and
naphthalene) are higher than the other quarterly averages. A review of the data shows
that many of the highest concentrations of the VOCs were measured during the period
from October 30, 2012 through December 23, 2012. Higher measurements were also
collected on the first two sample days of 2012 (January 4th and January 10th). This is
particularly true for naphthalene. All but one of the nine concentrations of
naphthalene greater than 200 ng/m3 were measured at PXSS during the fourth quarter
of2012.
• Manganese is the pollutant with the highest annual average concentration
(22.75 ± 4.01 ng/m3) of the three PMi0 metals. The first and second quarter averages
are higher than the other quarterly averages and the first quarter average has a
relatively large confidence interval associated with it. The maximum concentration of
this pollutant (106 ng/m3) was measured at PXSS on January 22, 2012, is nearly twice
the next highest concentration of this pollutant measured at PXSS (62.2 ng/m3), and is
the fifth highest manganese concentration measured among NMP sites sampling PMio
metals. Three of the 10 highest manganese concentrations among NMP sites sampling
PMio metals were measured at PXSS. Figure 4-28a in Section 4.4.2 shows that PXSS
is one of the three NMP sites with the highest quarterly averages of manganese
(besides ASKY-M and S4MO).
• The first and second quarter averages of nickel are also higher than the other quarterly
averages while the second quarter average has a relatively large confidence interval
associated with it. The maximum nickel concentration (7.73 ng/m3) was measured at
PXSS on June 20, 2012 and ties for eighth highest among nickel concentrations
measured among NMP sites sampling PMio metals. The second highest nickel
concentration measured at PXSS (6.55 ng/m3) was also measured during the second
quarter and ranks 12th highest across the program. However, the third through
seventh highest concentrations of nickel were all measured in January and February.
• The quarterly averages of arsenic have a similar pattern as the VOCs and carbonyl
compounds in that they are higher during the first and fourth quarters of 2012. All but
one of the nine arsenic concentrations greater than 1 ng/m3 measured at PXSS were
measured in samples collected in January, February, November, or December.
5-20
-------
Observations for SPAZ from Table 5-5 include the following:
• The pollutant of interest with the highest annual average concentration for SPAZ is
benzene (1.43 ± 0.30 |ig/m3), which is the only pollutant of interest with an annual
average concentration greater than 1 |ig/m3.
• The fourth quarter average concentration of 1,3-butadiene for SPAZ is the highest
valid quarterly average of this pollutant among all NMP sites sampling this pollutant,
as shown in Figure 4-20 in Section 4.4.2. The maximum 1,3-butadiene concentration
measured at SPAZ (0.738 |ig/m3) was measured on November 29, 2012; further, the
top five 1,3-butadiene concentrations were all measured at this site in November and
December. In addition, the top 10 concentrations of 1,3-butadiene measured at SPAZ
(those greater than 0.25 |ig/m3) were all measured in January and February or
November and December, further supporting the seasonality observations in these
concentrations.
• A similar trend is shown for benzene. The top three concentrations of benzene were
all measured at SPAZ in November; further, six of the eight highest concentrations
(those greater than 2 |ig/m3) were measured at SPAZ during the fourth quarter of
2012 (with the other two measured during the first quarter). The three highest
concentrations of ethylbenzene and/>-dichlorobenzene were also measured at SPAZ
on the same days in November as benzene. Figures 4-22 and 4-24 for
/>-dichlorobenzene and ethylbenzene in Section 4.4.2 show that the maximum
quarterly average concentration for each pollutant across the program was calculated
for SPAZ for the fourth quarter. SPAZ's fourth quarter benzene concentration ranks
third highest among other NMP sites sampling this pollutant.
Tables 4-9 through 4-12 present the 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 16 times.
• SPAZ has the highest annual average concentration of 1,3-butadiene and
/>-dichlorobenzene, similar to 2011, among all NMP sites sampling VOCs. SPAZ also
has the second highest annual average concentration of benzene and ethylbenzene.
PXSS has the second highest annual average concentrations of 1,3-butadiene and
/>-dichlorobenzene and the third highest annual average concentrations of benzene
and ethylbenzene (behind SPAZ).
• PXSS has the highest annual average concentration of acetaldehyde and the third
highest annual average concentration of formaldehyde among NMP sites sampling
carbonyl compounds.
• The annual average concentration of naphthalene for PXSS ranks sixth among NMP
sites sampling PAHs.
5-21
-------
• PXSS appears in Table 4-12 for all three speciated metal pollutants of interest, with
its annual averages ranking second highest for manganese, third highest for nickel,
and ninth highest for arsenic, among NMP sites sampling PMi0 metals.
5.4.2 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 shaded in
gray in Table 5-4 for PXSS and SPAZ. Figures 5-10 through 5-21 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 3.5.3.1.
Figure 5-10. Program vs. Site-Specific Average Acetaldehyde Concentration
PXSS
•-
)
3691
Concentration {[og/m3
Program: 1st Qua
D
Site: SiteAve
0
2 15 18
rtile 2nd Quartile 3rd Quartile 4th Quartile Average
rage Site Concentration Range
2
Figure 5-11. Program vs. Site-Specific Average Arsenic (PMio) Concentration
PXSS
•I
3 4
Concentration {ng/m3
Program:
Site:
1st Qua rtile
D
Site Average
o
2nd Qua rtile 3rd Qua rtile
D D
Site Concentration Range
^^^^—
4th Quartile Average
D 1
5-22
-------
Figure 5-12. Program vs. Site-Specific Average Benzene Concentrations
SPAZ
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site:
Site Average Site Concentration Range
o
Figure 5-13. Program vs. Site-Specific Average 1,3-Butadiene Concentrations
PXSS
1 1 o ' '
1 ' , ,
] Program Max Concentration = 4.10 ug/m3
1
1
O.
1
] Program Max Concentration
- 4 10 ue/m3
0.25
0.5
0.75 1
Concentration {[og/m3;
1.25
1.5
1.75
Program:
Site:
IstQuartile
D
Site Average
O
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
5-23
-------
Figure 5-14. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations
SPAZ
2 3
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site:
Site Average Site Concentration Range
o
Figure 5-15. Program vs. Site-Specific Average/>-Dichlorobenzene Concentrations
PXSS
-f
1 1
0.2
0.4
0.6 0.8
Concentration {[og/m3)
1.2
Program:
Site:
IstQuartile
D
Site Average
O
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
1.4
5-24
-------
Figure 5-16. Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations
Program Max Concentration = 17.01 ug/m3
SPAZ
i™
Program Max Concentration = 17.01 ug/m3
0.4 0.5 0.6
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o
Figure 5-17. Program vs. Site-Specific Average Ethylbenzene Concentrations
PXSS
i
,
!
-
1 i i i i i
0.5
1.5 2
Concentration {[og/m3]
2.5
3.5
Program:
Site:
IstQuartile
D
Site Average
O
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
Figure 5-18. Program vs. Site-Specific Average Formaldehyde Concentration
PXSS
10
12
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o —
14
5-25
-------
Figure 5-19. Program vs. Site-Specific Average Manganese (PMi0) Concentration
PXSS
1. 0
p
Program Max Concentration = 275
ng/m3
30
60 90
Concentration (ng/m3)
120
Program:
Site:
IstQuartile
D
Site Average
0
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
150
Figure 5-20. Program vs. Site-Specific Average Naphthalene Concentration
•=£
400 500
Concentration (ng/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: SiteAverage Site Concentration Range
o
Figure 5-21. Program vs. Site-Specific Average Nickel (PMio) Concentration
I
8 10
Concentration (ng/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site:
SiteAverage Site Concentration Range
O
Observations from Figures 5-10 through 5-21 include the following:
• Figure 5-10 for acetaldehyde shows that PXSS's annual average concentration of
nearly 3 |ig/m3 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 highest annual average concentration among NMP sites sampling this
pollutant. The minimum concentration measured at PXSS is just less than the
program-level first quartile.
5-26
-------
• Figure 5-11 shows that the annual average arsenic (PMio) concentration for PXSS
is just less than the program-level average for arsenic (PMio) and ranked ninth
highest among the 14 NMP sites sampling PMio metals. Although the maximum
concentration of arsenic measured across the program was not measured at PXSS,
the maximum concentration measured at PXSS (2.76 ng/m3) is among the higher
arsenic measurements. There were no non-detects of arsenic measured at PXSS.
• Figure 5-12 for benzene shows both Arizona sites, as both SPAZ and PXSS
sampled VOCs. While neither Arizona site measured the maximum benzene
concentration measured across the program, both annual averages are greater than
the program-level average concentration. The annual average benzene
concentration for SPAZ is slightly higher than the annual average concentration
for PXSS, although the range of concentrations measured is greater for PXSS.
SPAZ and PXSS have the second and third highest annual average concentrations
of benzene, respectively, among NMP sites sampling this pollutant.
• Figure 5-13 for 1,3-butadiene also shows both sites. Note that the program-level
maximum concentration (4.10 |ig/m3) is not shown directly on the box plots
because the scale of the box plots would be too large to readily observe data
points at the lower end of the concentration range. Thus, the scale of the box plots
has been reduced to 2 |ig/m3. The annual average concentrations for both sites are
more than twice the program-level average concentration. Further, these two sites
have the highest annual average concentrations of this pollutant across the
program, as mentioned above, with the annual average concentration for SPAZ
slightly higher than the annual average concentration for PXSS. The minimum
concentrations measured at these two sites are greater than the program-level first
quartile.
• Figure 5-14 presents the box plots for carbon tetrachloride for both sites.
Figure 5-14 shows that the annual average concentration of carbon tetrachloride
for PXSS is nearly identical to the program-level average while the annual
average for SPAZ is just less than the program-level average concentration. The
range of concentrations measured at PXSS is slightly less than the range for
SPAZ, although the minimum concentration measured at SPAZ is less than that of
PXSS.
• Figure 5-15 presents the box plots for/»-dichlorobenzene for both sites. Note that
the program-level first quartile is zero and therefore not visible on the box plots.
Similar to 1,3-butadiene, SPAZ and PXSS have the highest annual average
concentrations of />-dichlorobenzene among NMP sites sampling VOCs. The
annual average for PXSS is three times the program-level average concentration
and the annual average for SPAZ is four times the program-level average.
Although the maximum concentrations measured at these sites are considerably
less than the program maximum concentration, several of the concentrations
measured at SPAZ are among the highest measured across the program. A single
non-detect of/?-dichlorobenzene was measured at PXSS while the minimum
concentration measured at SPAZ is equivalent to the program-level median
concentration.
5-27
-------
• Figure 5-16 presents the box plots for 1,2-dichloroethane for both sites. Note that
the program-level maximum concentration (17.01 |ig/m3) is not shown directly on
the box plots as the scale has been reduced to 1 |ig/m3 in order to allow for the
observation of data points at the lower end of the concentration range. The
program-level average concentration is greater than the program-level third
quartile for this pollutant and is greater than or similar to the maximum
concentration measured at most sites sampling 1,2-dichloroethane. This is
because the program-level average is being driven by the higher measurements
collected at a handful of monitoring sites. Figure 5-16 shows that the maximum
1,2-dichloroethane concentrations measured at the Arizona sites are two orders of
magnitude less than the maximum concentration measured across the program.
The annual averages for SPAZ and PXSS are similar to the median concentration
at the program level. The maximum concentration measured at PXSS is similar to
the program-level average concentration while the maximum concentration
measured at SPAZ is less than the program-level average concentration. Non-
detects of 1,2-dichloroethane were measured at both Arizona sites, although the
number is greater for PXSS (14) than SPAZ (4).
• Figure 5-17 presents that box plots for ethylbenzene for the Arizona monitoring
sites. While neither Arizona site measured the maximum ethylbenzene
concentration measured across the program, both annual averages are more than
twice the program-level average concentration, and both are greater than the
program-level the third quartile. The annual average ethylbenzene concentration
for SPAZ is slightly higher than the annual average concentration for PXSS,
although the maximum concentration measured at PXSS is slightly higher than
the maximum concentration measured at SPAZ. SPAZ and PXSS have the second
and third highest annual average concentrations of ethylbenzene, respectively,
among NMP sites sampling this pollutant. The minimum ethylbenzene
concentrations measured at PXSS and SPAZ are greater than the program-level
first quartile.
• Figure 5-18 is the box plot for formaldehyde. This figure shows that the annual
average concentration for PXSS is greater than both the program-level average
concentration and third quartile. Recall from the previous section that this site has
the third highest annual average concentration among NMP sites sampling
carbonyl compounds.
• Figure 5-19 is the box plot for manganese (PMio) for PXSS. Note that the
program-level maximum concentration (275 ng/m3) is not shown directly on the
box plot as the scale has been reduced to 150 ng/m3 in order to allow for the
observation of data points at the lower end of the concentration range. Note also
that the program-level average is just less than the program-level third quartile,
indicating that the measurements at the higher end of the concentration range are
driving the average. Figure 5-19 shows the annual average concentration of
manganese for PXSS (22.75 ng/m3) is more than twice the program-level average
concentration (10.58 ng/m3) and twice the program-level third quartile
(11.18 ng/m3). PXSS has the second highest annual average concentration of
manganese among NMP sites sampling PMio metals, as discussed above. While
5-28
-------
the maximum concentration measured at PXSS (106 ng/m3) is considerably less
than the program-level maximum concentration, this is the fifth highest
measurement of manganese measured among the NMP sites sampling PMio
metals. The minimum concentration measured at PXSS (5.07 ng/m3) is just less
than the program-level median concentration (5.80 ng/m3). There were no non-
detects of manganese measured among sites sampling PMio metals.
• Figure 5-20 is the box plot for naphthalene for PXSS. Figure 5-20 shows that the
annual average naphthalene concentration of just less than 100 ng/m3 is greater
than the program-level average concentration (86.37 ng/m3). The maximum
naphthalene concentration measured at PXSS (343 ng/m3) is considerably less
than the maximum concentration measured at the program level. There were no
non-detects of naphthalene measured at PXSS or among sites sampling PAHs.
• Figure 5-21 is the box plot for nickel (PMio) for PXSS. The program-level
average is just less than the program-level third quartile, indicating that the
measurements at the higher end of the concentration range are driving the
program average. Figure 5-21 shows the annual average concentration of nickel
for PXSS is greater than the program-level average concentration and the
program-level third quartile. The minimum concentration measured at PXSS is
greater than the program-level first quartile and is the highest minimum nickel
concentration among NMP sites sampling PMio metals. Recall from the previous
section that PXSS has the third highest annual average concentration of nickel.
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 3.5.3.2.
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 5-22 through 5-39 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 is not provided, although the range and quartiles are still presented.
5-29
-------
Figure 5-22. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at PXSS
O BthPercentile
— Minimum
— Maximum
0 95th Percentile
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 March 2011 was invalidated.
Observations from Figure 5-22 for acetaldehyde measurements collected 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 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.
• A distinct trend is hard to identify because few 1-year averages are shown. However,
the range of measurements has remained fairly static over the years. The median
concentrations have varied from 2.23 |ig/m3 (2011) to 3.24 |ig/m3 (2007).
5-30
-------
Figure 5-23. Yearly Statistical Metrics for Arsenic (PMi0) Concentrations Measured at PXSS
£ .„
centration (ng/
3 fc
r
o.o -
••
1 '
^ T T 1
I r— n
L r^n
1 1
A o- o o ..A
^ta ^™ ^M ^j£i |^^ ••
1 t~ — t— L~«~J — s— — •— *
2006 2007 2008 2009 2010 2011 2012
Year
O 5th Percentile — Minimum ~ Median — Maximum O 95th Percentile ...^>... Average
Observations from Figure 5-23 for arsenic measurements collected 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. The third highest concentration was
measured on January 10,2012(2.77 ng/m3).
• The 1-year average concentration increased from 2010 to 2011 after several years of a
slight decreasing trend, although the changes across the years of sampling are not
statistically significant. The 1-year averages range from 0.51 ng/m3 (2010) to
0.77 ng/m3 (2011).
• The median concentrations did not change between 2011 and 2012 (0.56 ng/m3).
However, the 95th percentile decreased by 0.5 ng/m3. For both years, the
95th percentile represents the fourth highest concentration measured at PXSS. While
the number of measurements between 1 ng/m3 and the 95th percentile decreased from
12 to five from 2011 to 2012, the number of measurements between the median and
1 ng/m3 increased from 14 to 21, resulting in an unchanged median (or 50th
percentile) concentration.
5-31
-------
Figure 5-24. Yearly Statistical Metrics for Benzene Concentrations Measured at PXSS
O BthPercentile
— Maximum
95thPercentile
A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 5-24 for benzene measurements collected 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 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). Only three additional measurements greater than 4 |ig/m3 have been
measured at this site (one each in 2007, 2009, and 2011).
• After an increase from 2008 to 2009, the 1-year average benzene concentration has a
decreasing trend, although the largest change is from 2009 to 2010. The median
concentration exhibits a similar trend.
• The median concentration increased significantly from 2008 to 2009 and is actually
greater than the 1-year average concentration for 2009. A review of the data shows
that the number of concentrations greater than 2 |ig/m3 increased from 15 to 24 from
2008 to 2009, representing more than 42 percent of the concentrations measured in
2009, as compared to 29 percent in 2008.
5-32
-------
Figure 5-25. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at PXSS
~ 0.6
£
o
2009 2010
Year
O BthPercentile
— Maximum
95thPercentile
A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 5-25 for 1,3-butadiene measurements collected at PXSS
include the following:
• The maximum 1,3-butadiene concentration (1.08 |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. All but one of the 76 concentrations greater
than 0.35 |ig/m3 were measured during the first or fourth quarters, supporting the
observations regarding the trend in the quarterly averages discussed in the previous
sections and Section 4.4.2.
• The 1-year average 1,3-butadiene concentrations exhibit little change over the periods
shown, ranging from 0.207 |ig/m3 (2010) to 0.230 |ig/m3 (both 2009 and 2011). The
median concentration exhibits a similar consistency in magnitude for the periods
where 1-year averages could be calculated.
• There have been eight 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, two
were measured in 2010, and one was measured in 2007. For 2011, the minimum and
5th percentile were both equal to zero. None of the non-detects of 1,3-butadiene were
measured in the first or fourth quarters of the year.
5-33
-------
Figure 5-26. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at PXSS
O 5th Percentile - Minimurr
— Maximum • 95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 5-26 for carbon tetrachloride measurements collected at PXSS
include the following:
• Six concentrations of carbon tetrachloride greater than 1.0 |ig/m3 have been measured
at PXSS since the onset of sampling in 2007. All of these were measured in 2008 and
2009.
• For 2007, 2010, and 2011, 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, which is more common (see
benzene or 1,3-butadiene as examples).
• The 1-year average exhibits a decreasing trend through 2011, after which an increase
is shown for 2012.
• The difference between the 1-year average and median concentrations is less than
0.025 |ig/m3 for each year (where both were calculated), with 2012 having the
smallest difference. This indicates decreasing variability in the central tendency of
this pollutant.
5-34
-------
Figure 5-27. Yearly Statistical Metrics for/7-Dichlorobenzene Concentrations Measured at PXSS
o..
O
....o
2009 2010
Year
O 5th Percentile
— Maximum O 95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 5-27 for/>-dichlorobenzene measurements collected at PXSS
include the following:
• The three highest concentrations of />-dichlorobenzene were all measured in
November 2007.
• 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 from 2008 through 2010. Prior to 2010, a single non-detect was
measured; for 2010, nine non-detects were measured. Each of the statistical
parameters increased for 2011, with the exception of the minimum and 5th percentile,
as six additional non-detects were measured in 2011. Only one non-detect was
measured in 2012.
• Although the range of measurements within which the majority of the concentrations
fall tightened up for 2012, little change is shown for the 1-year average or median
concentrations from 2011 to 2012.
5-35
-------
Figure 5-28. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at PXSS
,o
O 5th Percentile
— Maximum O 95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 5-28 for 1,2-dichloroethane measurements collected 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, and 47 in 2012.
• With the exception of 2012, the median concentration is zero for all years, indicating
that at least 50 percent of the measurements were non-detects.
• As the number of measured detections increase, so do each of the corresponding
statistical metrics shown in Figure 5-28.
• As the number of measured detections increased dramatically for 2012, the median
and 1-year average concentrations increased correspondingly. The median
concentration is actually greater than the 1-year average for 2012. This is because
there were still 14 non-detects (or zeros) factoring into the 1-year average
concentration for the year.
5-36
-------
Figure 5-29. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at PXSS
s
8 i.o
o....
r
-•4
r
O 5th Percentile
- Minimum
— Maximum
O 95th Percentile "-O-" Average
A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 5-29 for ethylbenzene measurements collected at PXSS include
the following:
• The maximum concentration of ethylbenzene measured at PXSS (2.16 |ig/m3) was
measured on January 1, 2009. The next four highest concentrations were all measured
in November 2011, including the only other concentration greater than 2 |ig/m3 that
has been measured at PXSS (2.01 jig/m3).
• Similar to 1,3-butadiene, the highest ethylbenzene concentrations were measured
during the first and fourth quarters of the years. All but one of the 30 highest
concentrations (those greater than 1.40 |ig/m3) were measured between October and
December or January and March of any given year.
• The median concentration has a decreasing trend through 2009, after which an
increasing trend is shown, reaching a maximum in 2011. The 1-year average
concentration follows a similar pattern. All of the statistical parameters shown
increased from 2010 to 2011. Nearly twice the number of measurements greater than
1 |ig/m3 (20) were measured in 2011 than the previous years (11 or less), accounting
for one-third of the total measurements for that year. The number of measurements
greater than 1 |ig/m3 for 2012 is down slightly (14) but still higher than years prior to
2011.
5-37
-------
Figure 5-30. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at PXSS
O 5th Percentile
— Minimum
— Maximum
0 95th Percentile
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 March 2011 was invalidated.
Observations from Figure 5-30 for formaldehyde measurements collected 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,
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 five highest formaldehyde concentrations (ranging from 6.28 |ig/m3 to
7.55 |ig/m3) were all measured in 2007. The next five highest concentrations were all
measured in either 2007 or 2011.
• 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 eight less than 2 |ig/m3 have been measured since 2007.
5-38
-------
Figure 5-31. Yearly Statistical Metrics for Manganese (PMi0) Concentrations Measured at PXSS
O 5th Percentile - Minimurr
Median — Maximum O 95th Percentile
Observations from Figure 5-31 for manganese measurements collected at PXSS include
the following:
• Four manganese concentrations greater than 100 ng/m3 have been measured at PXSS
since metals sampling began; three were measured in 2011 and the fourth was
measured in 2012. Of the 12 concentrations greater than 50 ng/m3, five were
measured during 2011, three in 2012, two in 2009, and one each in 2007 and 2008.
• The 1-year average concentration of manganese decreased significantly from 2009 to
2010 then increased significantly for 2011. The 1-year average concentration for 2011
is twice the 1-year average for 2010. Over the course of sampling, the measurements
from 2011 exhibit the most variability while the measurements from 2010 exhibit the
least.
• PXSS has the second highest annual average concentration of manganese for 2012.
Previous reports indicate that PXSS consistently has one of the highest annual
average concentrations of manganese among NMP sites sampling for PMi0 metals.
• Even though the maximum and 95th percentiles decreased from 2011 to 2012, the
median concentration increased for 2012. Although 2011 had a higher number of
measurements at the upper end of the scale, there were also more measurements at the
lower end of the scale for 2011 compared to 2012. For example, there were five
measurements less than 10 ng/m3 for 2012 compared to 14 for 2011.
5-39
-------
Figure 5-32. Yearly Statistical Metrics for Naphthalene Concentrations Measured at PXSS
I,
o
. o
2009 2010
Year
O 5th Percentile - Minimuir
Median — Maximum O 95th Percentile "-O-" Averagf
A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 5-32 for naphthalene measurements collected at PXSS include
the following:
• PXSS began sampling PAHs under the NMP in July 2007.
• The maximum naphthalene concentration was measured in December 2008. Although
this is the only measurement greater than 400 ng/m3 measured at PXSS, a similar
concentration was also measured twelve days later on January 1, 2009 (386 ng/m3).
The only other measurement greater than 300 ng/m3 was measured on December 23,
2012.
• Many of the statistical parameters were highest for 2009. The median, or midpoint,
for 2009 is 107 ng/m3. By comparison, the median concentrations for the other years
were less, ranging from 68.1 ng/m3 (2008) to 84.1 ng/m3 (2010).
• The difference between the 5th and 95th percentiles (the range of concentrations
where 90 percent of the measurements lie) has been increasing since 2010 and is
greatest for 2012. Thus, the range of concentrations within which the majority of
concentrations lie has an increasing trend.
5-40
-------
Figure 5-33. Yearly Statistical Metrics for Nickel (PMi0) Concentrations Measured at PXSS
~ 4.0
£
. O
o.
o
1
o
2009
Year
O 5th Percentile — Minimuir
Median — Maximum O 95th Percentile
Observations from Figure 5-33 for nickel measurements collected at PXSS include the
following:
• The maximum nickel concentration was measured at PXSS on June 20, 2012
(7.73 ng/m3). Four additional concentrations greater than 6 ng/m3 have been
measured at PXSS since metals sampling under the NMP began; two were measured
in 2008, one in 2009, and one in 2012.
• The 1-year average concentration of nickel exhibits a decreasing trend from 2007
through 2010, after which an increasing trend is shown. The increase from 2010 to
2011 is significant, representing a nearly 50 percent increase. The median
concentration exhibits a similar tendency between 2010 and 2012. The increase in the
median indicates that concentrations are running higher in these later years as the
median is less sensitive to outliers, or a few concentrations at the higher end of the
range, than the average concentration. PXSS has the third highest annual average
concentration of nickel for 2012 among NMP sites sampling PMio metals.
• The only two non-detects of nickel measured at PXSS were both measured in 2008.
5-41
-------
Figure 5-34. Yearly Statistical Metrics for Benzene Concentrations Measured at SPAZ
O BthPercentile
— Maximum
O 95thPercentile
Average
A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 5-34 for benzene measurements collected 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 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 concentration greater than 5 |ig/m3 measured at SPAZ.
Only five additional measurements greater than 4 |ig/m3 have been measured at this
site (one for each year of sampling except 2012).
• After several years of increasing, both the maximum and 95th percentile are at a
minimum for 2012. Although the 1-year average concentration is also down for 2012,
the median concentration actually increased. For 2011, the concentrations at the
higher end of the concentration range are driving the 1-year average concentration,
whereas there is less variability in the 2012 measurements.
• Forty-five of the 49 benzene concentrations greater than 2 |ig/m3 were measured
during the first or fourth quarters of the year.
5-42
-------
Figure 5-35. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at SPAZ
O 5th Percentile - Minimurr
— Maximum • 95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 5-35 for 1,3-butadiene measurements collected at SPAZ
include the following:
• The maximum 1,3-butadiene concentration (1.08 |ig/m3) was measured on
January 27, 2011. Thirty-seven of the 39 concentrations greater than 0.35 |ig/m3 were
measured during the first or fourth quarters of a 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 measurements collected were at the higher end of the concentration range. For
2012, the range of concentrations measured is smaller, as the maximum concentration
for 2012 is less than the 95th percentile for 2011. This is a pattern similar to that
exhibited by benzene in Figure 5-34.
• The 1-year average concentration decreased slightly from 2008 to 2009, then exhibits
a slight increasing trend through 2011, followed by a return to 2010 levels for 2012.
Confidence intervals calculated for the 1-year averages indicate that these changes are
not statistically significant. The median concentration exhibits a steeper decrease
from 2008 to 2009 and a steeper increase from 2009 to 2010. Little change is
exhibited by the median concentration between 2010 and 2012.
5-43
-------
Figure 5-36. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
SPAZ
O 5th Percentile - Minimurr
— Maximum • 95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 5-36 for carbon tetrachloride measurements collected 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 in 2007. 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 were
measured in 2009 and 2011.
• For the years 2009 and later, the box and whisker plots for this pollutant appear
"inverted," 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.
• The 1-year average exhibits a decreasing trend through 2011, after which a slight
increase is shown for 2012. However, these changes represent an overall change of
only 0.08 |ig/m3 and, based on the confidence intervals, are not statistically
significant. The median concentration exhibits little change between 2008 and 2010
then decreases substantially for 2011.
• The difference between the 1-year average and median concentrations is at a
minimum for 2012, indicating less variability in the central tendency than for other
years.
5-44
-------
Figure 5-37. Yearly Statistical Metrics for/7-Dichlorobenzene Concentrations Measured at SPAZ
O 5th Percentile — Minimum — Median
— Maximum
O 95th Percentile ...<>... Average
A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 5-37 for/>-dichlorobenzene measurements collected at SPAZ
include the following:
• The widest range ofp-dichlorobenzene concentrations is shown for 2008 (non-detect
to 0.90 |ig/m3), while the smallest range is shown for the following year (0.036 |ig/m3
to 0.51 |ig/m3).
• The 1-year average concentration decreased from 2008 to 2009, increased for 2010,
then decreased slightly for 2011 and 2012. However, confidence intervals calculated
for these averages indicate that the changes are not statistically significant.
• The median concentrations appear to exhibit larger fluctuations than the 1-year
average concentrations. Yet, the largest year-to-year difference is the change from
2009 to 2010 and represents a change of less than 0.12 |ig/m3.
5-45
-------
Figure 5-38. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at SPAZ
•2 0.15
o
O 5th Percentile
— Maximum O 95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 5-38 for 1,2-dichloroethane measurements collected 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, and 26 in 2012.
• The median concentration is zero for all years except 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 5-38.
• As the number of measured detections increased dramatically for 2012, the median
and 1-year average concentrations increased correspondingly. The median
concentration is greater than the 1-year average for 2012. This is because the four
non-detects (or zeros) factored into the 1-year average concentration are pulling the
average down (just like a maximum or outlier concentration can pull the average up).
5-46
-------
Figure 5-39. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at SPAZ
o
.. o
O 5th Percentile
— Maximum
O 95th Percentile "-O-" Averagf
A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 5-39 for ethylbenzene measurements collected 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 and3.0 |ig/m3 were measured in either 2007 or 2011.
• The median concentration is at a maximum for 2007, after which the median
decreases by half. Recall that 2007 includes only a 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. The 1-year average
concentration has a similar pattern, although no 1-year average is presented for 2007.
• The minimum concentration measured each year before 2010 is at or near zero (non-
detect); the minimum concentration in later years is an order of magnitude higher.
The 5th percentile for each of the later years is similar to the minimum concentrations
measured (less than 0.065 jig/m3).
5-47
-------
The following observations summarize some of the highlights from Figures 5-22 through
5-39 for PXSS and SPAZ:
• Several of the pollutants of interest for PXSS were highest on January 1, 2009 (or
measured their second highest concentration on this date). Some of the VOC
pollutants of interest for SPAZ were highest on January 27, 2011 (or measured their
second highest concentration on this date).
• The highest measurements of several of the VOCs, 1,3 -butadiene and ethylbenzene in
particular, were most often measured during the colder months of the year. This trend
is more prevalent at PXSS than SPAZ.
5.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk at
each Arizona monitoring site. Refer to Sections 3.3 and 3.5.3 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.
5.5.1 Risk-Based Screening Assessment Using MRLs
A risk-based screening was conducted by comparing the concentration data from the
Arizona monitoring sites to the ATSDR MRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 day to 14 days); intermediate (exposures of 15 days to 364 days); and chronic
(exposures of 1 year or greater). The preprocessed daily measurements of the pollutants of
interest were compared to the acute MRLs; the quarterly averages were compared to the
intermediate MRLs; and the annual averages were compared to the chronic MRLs.
As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2012.
5-48
-------
5.5.2 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 3.5.3.4 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 5-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Observations for PXSS from Table 5-6 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 (51.44 in-a-million), followed by benzene (10.01 in-a-million),
1,3-butadiene (6.66 in-a-million), and acetaldehyde (6.38 in-a-million).
• Formaldehyde's cancer risk approximation for PXSS is the third highest cancer risk
approximation among all of 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 health effects are expected from these
individual pollutants. The pollutant with the highest noncancer hazard approximation
for PXSS is manganese (0.45). This noncancer hazard approximation is the fourth
highest noncancer hazard approximation among all site-specific pollutants of interest.
5-49
-------
Table 5-6. Risk Approximations for the Arizona Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections vs.
# of Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
Phoenix, Arizona - PXSS
Acetaldehyde
Arsenic (PM10)a
Benzene
1,3 -Butadiene
Carbon Tetrachloride
£>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
Manganese (PM10)a
Naphthalene3
Nickel (PM10)a
0.0000022
0.0043
0.0000078
0.00003
0.000006
0.000011
0.000026
0.0000025
0.000013
0.000034
0.00048
0.009
0.000015
0.03
0.002
0.1
0.8
2.4
1
0.0098
0.00005
0.003
0.00009
61/61
61/61
61/61
61/61
61/61
60/61
47/61
61/61
61/61
61/61
59/59
61/61
2.90
±0.30
0.01
± 0.01
1.28
±0.21
0.22
±0.05
0.68
±0.02
0.20
±0.03
0.07
±0.01
0.73
±0.12
3.96
±0.27
0.02
±O.01
0.10
±0.02
O.01
±O.01
6.38
2.92
10.01
6.66
4.10
2.15
1.84
1.82
51.44
3.33
0.98
0.32
0.05
0.04
0.11
0.01
0.01
O.01
O.01
0.40
0.45
0.03
0.02
South Phoenix, Arizona - SPAZ
Benzene
1,3 -Butadiene
Carbon Tetrachloride
£>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
0.0000078
0.00003
0.000006
0.000011
0.000026
0.0000025
0.03
0.002
0.1
0.8
2.4
1
30/30
30/30
30/30
30/30
26/30
30/30
1.43
±0.30
0.26
±0.07
0.65
±0.04
0.26
±0.06
0.08
±0.01
0.84
±0.18
11.16
7.71
3.88
2.84
1.97
2.10
0.05
0.13
0.01
O.01
0.01
O.01
— = A Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 5-5.
Observations for SPAZ from Table 5-6 include the following:
• The pollutants with the highest annual average concentrations are benzene,
ethylbenzene, and carbon tetrachloride. Only benzene has an annual average
concentration greater than 1 |ig/m3.
5-50
-------
• Based on the annual averages and cancer UREs, benzene has the highest cancer risk
approximation for SPAZ (11.16 in-a-million), followed by 1,3-butadiene
(7.71 in-a-million), and carbon tetrachloride (3.88 in-a-million).
• None of the pollutants of interest for SPAZ have noncancer hazard approximations
greater than 1.0, indicating no adverse health effects are expected from these
individual pollutants. The pollutant with the highest noncancer hazard approximation
for SPAZ is 1,3-butadiene (0.13).
5.5.3 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 5-7 presents the 10 pollutants with the highest emissions from the 2011 NEI that have
cancer toxicity factors. Table 5-7 also presents the 10 pollutants with the highest toxicity-
weighted emissions, based on the weighting schema described in Section 3.5.3.5. Lastly,
Table 5-7 provides the 10 pollutants with the highest cancer risk approximations (in-a-million)
for each site, as presented in Table 5-6. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 5-7. Table 5-8 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 3.5.3.5. Similar to the cancer risk and
noncancer hazard approximations provided in Section 5.5.2, this analysis may help policy-
makers prioritize their air monitoring activities.
5-51
-------
Table 5-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Arizona Monitoring Sites
fj\
to
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Phoenix, Arizona (Maricopa County) - PXSS
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
Tetrachloroethylene
1,3 -Butadiene
Naphthalene
Dichloromethane
POM, Group 2b
POM, Group 2d
937.42
763.28
668.61
407.11
216.22
130.89
75.03
12.33
11.29
10.18
Formaldehyde
Benzene
1,3 -Butadiene
Naphthalene
Ethylbenzene
POM, Group 2b
POM, Group 2d
Acetaldehyde
Arsenic, PM
Hexavalent Chromium, PM
9.92E-03
7.31E-03
3.93E-03
2.55E-03
1.67E-03
9.94E-04
8.96E-04
8.96E-04
6.95E-04
5.02E-04
Formaldehyde
Benzene
1,3 -Butadiene
Acetaldehyde
Carbon Tetrachloride
Naphthalene
Arsenic
£>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
51.44
10.01
6.66
6.38
4.10
3.33
2.92
2.15
1.84
1.82
South Phoenix, Arizona (Maricopa County) - SPAZ
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
Tetrachloroethylene
1,3 -Butadiene
Naphthalene
Dichloromethane
POM, Group 2b
POM, Group 2d
937.42
763.28
668.61
407.11
216.22
130.89
75.03
12.33
11.29
10.18
Formaldehyde
Benzene
1,3 -Butadiene
Naphthalene
Ethylbenzene
POM, Group 2b
POM, Group 2d
Acetaldehyde
Arsenic, PM
Hexavalent Chromium, PM
9.92E-03
7.31E-03
3.93E-03
2.55E-03
1.67E-03
9.94E-04
8.96E-04
8.96E-04
6.95E-04
5.02E-04
Benzene
1,3 -Butadiene
Carbon Tetrachloride
£>-Dichlorobenzene
Ethylbenzene
1 ,2-Dichloroethane
11.16
7.71
3.88
2.84
2.10
1.97
-------
Table 5-8. 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
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Phoenix, Arizona (Marico]
Toluene
Ethylene glycol
Hexane
Xylenes
Methanol
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
Methyl isobutyl ketone
9,063.05
5,143.63
2,587.16
2,542.34
2,398.84
937.42
763.28
668.61
407.11
326.37
Acrolein
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Lead, PM
Benzene
Xylenes
Naphthalene
Ethylene glycol
Arsenic, PM
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer Hazard
Approximation
(HQ)
pa County) - PXSS
2,007,476.52
77,886.07
65,443.47
45,234.52
34,311.59
31,247.27
25,423.43
25,010.69
12,859.08
10,773.68
Manganese
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Arsenic
Benzene
Naphthalene
Nickel
Carbon Tetrachloride
Ethylbenzene
0.45
0.40
0.32
0.11
0.05
0.04
0.03
0.02
0.01
0.01
South Phoenix, Arizona (Maricopa County) - SPAZ
Toluene
Ethylene glycol
Hexane
Xylenes
Methanol
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
Methyl isobutyl ketone
9,063.05
5,143.63
2,587.16
2,542.34
2,398.84
937.42
763.28
668.61
407.11
326.37
Acrolein
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Lead, PM
Benzene
Xylenes
Naphthalene
Ethylene glycol
Arsenic, PM
2,007,476.52
77,886.07
65,443.47
45,234.52
34,311.59
31,247.27
25,423.43
25,010.69
12,859.08
10,773.68
1,3 -Butadiene
Benzene
Carbon Tetrachloride
Ethylbenzene
£>-Dichlorobenzene
1 ,2-Dichloroethane
0.13
0.05
0.01
0.01
0.01
0.01
-------
Observations from Table 5-7 include the following:
• Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
cancer UREs in Maricopa County.
• The pollutants with the highest toxi city-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 highest toxicity-weighted emissions 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.
• 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 ninth 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. Similarly, POM, Group 2d is the 10th
highest emitted "pollutant" and ranks seventh for toxicity-weighted emissions. POM,
Group 2d includes several PAHs sampled for at PXSS including anthracene,
phenanthrene, and pyrene. None of the PAHs included in POM, Groups 2b or 2d
were identified as pollutants of interest for PXSS (or failed any screens).
• 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.
Observations from Table 5-8 include the following:
• Toluene, ethylene glycol, 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.
• Five of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Maricopa County.
5-54
-------
• Acrolein has the highest toxi city-weighted emissions (by two orders of magnitude)
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.
• Manganese has the highest noncancer hazard approximation for PXSS (although
considerably less than an HQ of 1.0), followed by formaldehyde and acetaldehyde.
While all three of these pollutants appear among those with the highest toxi city-
weighted emissions, only formaldehyde and acetaldehyde appear among the highest
emitted.
• 1,3-Butadiene has the highest noncancer hazard approximation for SPAZ (0.13).
Although the noncancer hazard approximation for PXSS (0.11) is similar in
magnitude to that of SPAZ, it ranks fourth behind three pollutants for which SPAZ
does not sample. 1,3-Butadiene has the third highest toxicity-weighted emissions but
is not one of the highest emitted pollutants in Maricopa County (with a noncancer
RfC), as it ranks 12th.
5.6 Summary of the 2012 Monitoring Data for PXSS and SPAZ
Results from several of the data treatments described in this section include the
following:
»«» Twenty-one pollutants failed screens for PXSS; seven pollutants failed screens for
SPAZ.
*»* 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, including 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 1,3-butadiene andp-dichlorobenzene among NMP sites sampling VOCs. These
sites also rank second and third highest for benzene andethylbenzene. PXSS has the
highest annual average concentration of acetaldehyde among all NMP sites sampling
car bony I compounds. Among NMP sites sampling PMw metals, PXSS ranks second
for its annual average concentration of manganese.
»«» Concentrations of nickel have been increasing at PXSS over the last few years of
sampling. The detection rate of 1,2-dichloroethane has been increasing steadily at
both sites over the last few years of sampling.
5-55
-------
6.0 Sites in California
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at three NATTS sites and one CSATAM site 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.
The California monitoring sites are located in Los Angeles, Long Beach, Rubidoux, and
San Jose. Figure 6-1 and 6-2 are the composite satellite images retrieved from ArcGIS Explorer
showing the Los Angeles and Long Beach monitoring sites and their immediate surroundings.
Figure 6-3 identifies nearby point source emissions locations by source category for each site, as
reported in the 2011 NEI for point sources. Note that only sources within 10 miles of the sites are
included in the facility counts provided in Figure 6-3. 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 each 10-mile radius are still
visible on the map, but have been grayed out in order to show emissions sources just outside the
boundary. Figures 6-4 through 6-7 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
w/au. »*,H\ '
-------
Figure 6-2. Long Beach, California (LBHCA) Monitoring Site
-------
Figure 6-3. NEI Point Sources Located Within 10 Miles of CELA and LBHCA
118"40'0"W 118'35'0"W 118"30'0"W 118 25'0"W 118 20'0"W 118 15'0"W 118'10'CTW 118'5'0"W
Legend
lia°20'0"W 118C15'0"W 118U10'0"W 118J5'0"W 118"0'0"W 117°55'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
CELA NATTS site
LBHCA CSATAM site
10 mile radius
County boundary
Source Category Group (No. of Facilities)
Ij/l Aero space/ Ai re raft Manufacturing (14)
"1" Airport/Air! me/ Arr port Support Operations (88)
£ Asphalt Production/ Hot Mix Asphalt Plant (5)
0 Auto Body Shop/Pair ters/ Auto motive Stores (37)
5$ Automobile/Truck Manufacturing (7)
4J: Automotive/RV Dealership (3)
'X' Battery Manufacturing (3)
ft Building/Construction (4)
B Bulk Terminals/Bulk Plants (31)
C Chemical Manufacturing (35)
i Compressor Station (7)
XlCrematory -Animal/Human (1)
® Dry Cleaning (2)
G Electrical Equipment Manufacturing (12)
f Electricity Generation via Combustion (14)
E Electroplating, Plating. Polishing. Anoc^ing. and Goto;
F Food Processing/ Agriculture (20)
I Foundries. Iron and Steel (7)
A Foundries. Non-ferrojs (10)
• Gasoline/Diesel Service Station (3)
fS Glass Plant (1)
> Hotels/Motels/Lodging (2)
•3ft Industrial Machinery or Equipment Plant (7)
O Institution (school, hospital, prison, etc.) (41)
A Landfill (1)
^ Leather and Leather Products (1)
ft'1 Metal Can, Box. and Other Metal Container Manufacturing (6)
A Metal Coating, Engraving, and Allied Services to Manufacturers
(•) Metals Processing^brication (30)
A Military Base/National Security (1)
X Mine/Quarry/Mineral Processing (4)
? Miscellaneous Commercial/Industrial (81)
Municipal Waste Combustor (2)
Oil and/or Gas Production (48)
m
*
1113 '361 Q Paint and Coating Manufacturing (8)
Pesticide Manufacturing Plant (1)
\ Petroleum Products Manufacturing (2)
~A Petroleum Refinery (12)
d> Pharmaceutical Manufacturing (4)
R Plastic. Resin, or Rubber Products Plant (17)
V Port and Harbor Operations (3)
^ Printing. Coating & Dyeing of Fabrics (4)
P Printing; Publish ing/Pa per Product Manufacturing (36)
H Pulp and Paper Plant (2)
X Rail Yard/Rail Line Operations (5)
0 RestauranUCooking (1)
* A
«, Ship/Boat Manufacturing or Repair (2)
V Steel Mill (1)
TT Telecommunications/Radio (2)
^^ Testing Laboratories (1)
T Textile. Yarn, or Carpet Plant (8)
** Truck/Bus/Transportation Operations (1)
n Utilities/Pipeline Construction (2)
I Wastewater Treatment (8)
6 Water Treatment (3)
W \Abodwork, Furniture, Mil (work & Wood Preserving (25)
6-4
-------
Figure 6-4. Rubidoux, California (RUCA) Monitoring Site
-------
Figure 6-5. NEI Point Sources Located Within 10 Miles of RUCA
117°25'0-W 117320'0"W 117015'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
RUCA NATTS site
10 mile radius
County boundary
Source Category Group (No. of Facilities)
Aero space/A if era ft Manufacturing (1)
Airport/Airline/Airport Support Operations (12)
Animal Feediot or Farm (9)
Asphalt Production/Hot Mix Asphalt Plant (4)
Auto Body Shop/Painters/Automotive Stores (9)
Automobile/Truck Manufacturing (4)
Automotive/RV Dealership (2)
Brick. Structural Clay, or Clay Ceramics Plant (1)
Building/Construction (1)
Bulk Terminals/Bulk Plants (3)
Chemical Manufacturing (3)
Compressor Station (1)
Electrical Equipment Manufacturing {2}
Electricity Generation via Combustion (5)
Etnanol Biorefmenes (1)
Food Processing/Agriculture (7)
Foundries. Iron and Steel (1)
Gasoline/Diesel Service Station (1)
Industrial Machinery or Equipment Plant (5)
O Institution (school, hospital, prison, etc.) (9)
ft Landfill (5)
(•} Metal Can, Box. and Other Metal Container Manufacturing (4)
A Metal Coating. Engraving, andAllied Services to Manufacturers (1)
{•) Metals Processing/Fabrication (12)
A Military Base/National Security (1)
X Mine/Quarry/Mineral Processing (6)
I Miscellaneous Commercial/Industrial (18)
[M] Municipal Waste Combustor{1)
• Oil and/or Gas Production (2)
Q Paint and Coating Manufacturing (4)
R Plastic. Resm, or Rubber Products Plant (4)
7 Portland Cement Manufacturing (2)
P PrintingJPublishing/Paper Product Manufacturing (6)
X Rail Yard/Rail Line Operations (3)
V Steel Mill (2)
I Wastewater Treatment (3)
6 Water Treatment Facility (6)
W Woodwork, Furniture. Millwork S Wood Preserving (4)
6-6
-------
Figure 6-6. San Jose, California (SJJCA) Monitoring Site
-------
Figure 6-7. NEI Point Sources Located Within 10 Miles of SJJCA
Legend
12r55'0"W 121°50'0"W 12V45'Q"W 12 r 40*0" W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
•ft
SJJCA NATTS site O 10 mile radius
County boundary
Source Category Group (No. of Facilities)
ijl Aerospace/Aircraft Manufacturing (3)
^ Airport/Air line/Airport Support Operations (18)
4 Asphalt Production/Hot Mix Asphalt Plant (3)
0 Auto Body Shop/Pamters/Automotive Stores (219)
ft Automobile/Truck Manufacturing (3)
3b Automotive/RV Dealership (8)
LHJ Brick, Structural Clay, or Clay Ceramics Plant (1)
ft Building/Construction (13)
B Bulk Terminals;Bulk Plants (7)
C Chemical Manufacturing (12)
1 Compressor Station (2)
!><]Crematory - Animal/Human (5)
© Dry Cleaning (110)
6 Electrical Equipment Manufacturing (262)
f Electricity Generation via Combustion (60)
E Electroplating, Plating, Polishing. Anodizing, and Color
y Fertilizer Plant (1)
F Food Processing;Agriculture (20)
A Foundries, Non-ferrous (1)
If Gasoline/Diesel Service Station (7)
S Glass Plant (1)
> Hotels/Motels/Lodging (8)
•3ft" Industrial Machinery or Equipment Plant (22)
O Institution (school, hospital, prison, etc.) (138)
A Landfill (4)
« Leather and Leather Products (1)
(•) Metal Can. Box, and Other Metal Container Manufacturing (4)
A Metal Coating. Engraving, and Allied Services !o Manufacturers (27)
© Metals Processing/Fabrication (29)
A Military Base/National Security (3)
X Mine/Quarry/Mineral Processing (11)
£ Mineral WoolAAfoo! Fiberglass Manufacturing (1)
? Miscellaneous Commercial/Industrial (292)
ig(13) ^ oil and/or Gas Production (1)
[]
Paint and Coating Manufacturing (3)
3£ Petroleum Products Manufacturing (2)
^ Pharmaceutical Manufacturing (9)
R Plastic, Resin, or Rubber Products Plant (4)
¥ Portand Harbor Operations (1)
^ Printing. Coating & Dyeing of Fabrics (4)
P Printing/Publisriing/Paper Product Manufacturing (34)
ffl Pulp and Paper Plant (3)
X Rail Yard/Rail Line Operations (2)
TT Telecommunications/Radio (91)
^ Testing Laboratories (2)
T Textile, Yarn, or Carpet Plant (1)
M Tobacco Manufacturing (8)
rfV Truck/Bus/Transportation Operations (7)
CH Utilities/Pipeline Construction (1)
I Wbstewater Treatment (10)
A Water Treatment (30)
W Woodwork, Furniture. Millwork & Wood Preserving [36)
6-8
-------
Table 6-1. Geographical Information for the California Monitoring Sites
Site
Code
CELA
LBHCA
RUCA
SJJCA
AQS Code
06-037-1103
06-037-4002
06-065-8001
06-085-0005
Location
Los
Angeles
Long
Beach
Rubidoux
San Jose
County
Los
Angeles
Los
Angeles
Riverside
Santa
Clara
Micro- or
Metropolitan
Statistical Area
Los Angeles-Long
Beach-Anaheim,
CAMSA
Los Angeles-Long
Beach- Anaheim,
PA l\/T
-------
CELA is located on the rooftop of a two-story building northeast of downtown Los
Angeles, just southeast of Dodgers' Stadium. Figure 6-1 shows that 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 monitor.
The LBHCA monitoring site is located on the property of a church in Long Beach. The
surrounding area is considered residential and suburban, although commercial areas are also
located nearby and along Long Beach Blvd, as shown in Figure 6-2. Interstate-405 is located
approximately one-third of a mile from LBHCA and intersects with 1-710 just one mile west of
the site. This monitoring site is located approximately four miles north of the shores of Long
Beach as well as the Port of Long Beach, the second-busiest port in the U.S. (POLE, 2013).
Figure 6-3 shows that LBHCA is nearly 17 miles south of CELA. These sites are situated
among a high density of point sources. 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. Other source categories with a large number of emissions
sources within 10 miles of CELA and LBHCA are oil and gas production; institutions such as
school, hospitals, and/or prisons; auto body shops, painters, and automotive stores; printing,
publishing, and paper product manufacturing; electroplating, plating, polishing, anodizing, and
coloring; and chemical manufacturing. There is a cluster of emissions sources located just to the
west and southwest of CELA. There is also a second large cluster of sources to the south of the
site. 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. Several emissions sources are
located directly south of LBHCA, including several involved in oil and gas production.
RUCA is located just outside of Riverside, in a residential area of the suburban town of
Rubidoux. Figure 6-4 shows that RUCA is adjacent to a power substation west of a storage
facility 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,
6-10
-------
the football and baseball fields of which are prominent features in Figure 6-4. 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 45 miles west of CELA and 46 miles
northwest of LBHCA. Figure 6-5 shows that fewer emissions sources surround RUCA than
CELA and LBHCA. Most of the emissions sources are located to the northeast and northwest of
the site. 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
near RUCA include airport operations; metals processing; auto body shops, painters, and
automotive stores; animal feedlots or farms; and institutions such as school, hospitals, and/or
prisons.
SJJCA is located in central San Jose. Figure 6-6 shows that SJJCA is located in a
commercial area surrounded by residential areas. A railroad is shown just east of the monitoring
site, running north-south in Figure 6-6. 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 6-6. Figure 6-7
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
are electrical equipment manufacturing; auto body, paint, and automotive shops; institutions such
as school, hospitals, and/or prisons; dry cleaning; and telecommunications. Sources closest to
SJJCA include a food processing facility and several auto body shops.
Table 6-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the California monitoring sites. Table 6-2 includes both county-level
population and vehicle registration information. Table 6-2 also contains traffic volume
information for each site as well as the location for which the traffic volume was obtained.
Additionally, Table 6-2 presents the county-level daily VMT for Los Angeles, Riverside, and
Santa Clara Counties.
6-11
-------
Table 6-2. Population, Motor Vehicle, and Traffic Information for the California
Monitoring Sites
Site
CELA
LBHCA
RUCA
SJJCA
Estimated
County
Population1
9,962,789
2,268,783
1,837,504
County-level
Vehicle
Registration2
7,422,254
1,724,787
1,529,351
Annual
Average Daily
Traffic3
229,000
282,000
145,000
106,000
Intersection
Used for
Traffic Data
1-5 between Exits 136 and 137
1-405 between Exits 30 and 32
Mission Blvd between Rubidoux Blvd
& Valley Way
Guadalupe Pkwy (87) between
Julian St & W Taylor St
County-
level Daily
VMT4
214,458,140
55,717,760
41,250,490
Bounty-level population estimates reflect 2012 data (Census Bureau, 2013c)
2County-level vehicle registration reflects 2012 data (CA DMV, 2012)
3AADT reflects 2012 data (CA DOT, 2012a)
4County-level VMT reflects 2011 data (CA DOT, 2012b)
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 6-2 include the following:
• Los Angeles County (CELA and LBHCA) has the highest county-level population
and vehicle registration compared to all counties with NMP sites.
• Riverside and Santa Clara Counties are also in the top 10 for county-level population
and vehicle registration among counties with NMP sites.
• LBHCA experiences the highest annual average daily traffic among NMP sites, with
CELA's traffic ranking third. These two sites are located relatively close to major
freeways in the Los Angeles metro area. The traffic volume for RUCA also ranks
among the top 10. The traffic volume for SJJCA ranks 15th.
• Los Angeles County's daily VMT is the highest among all counties with NMP sites,
where VMT was available. This VMT is an order of magnitude higher than the next
highest county-level VMT (Maricopa County, AZ). The VMT for Riverside and
Santa Clara Counties are also in the top 10 for VMT among counties with NMP sites,
where VMT data were available, ranking fifth and sixth, respectively.
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.
6.2.1 Climate Summary
The climate of Los Angeles and the surrounding areas is generally mild. While the
proximity to the Pacific Ocean acts as a moderating influence on the Los Angeles area, the
elevation changes between the mountains and valleys allow the distance from the ocean to create
substantial differences in temperature, rainfall, and wind over a relatively short distance.
6-12
-------
Precipitation falls primarily in winter months, while summers tend to be dry. Westerly winds are
prevalent for much of the year. Stagnant wind conditions in the summer can result in air
pollution episodes, while breezy Santa Ana winds can create hot, dusty conditions. Fog and
cloudy conditions are more prevalent near the coast than farther inland (Wood, 2004; WRCC,
2013).
San Jose is located to the southeast of San Francisco, near the base of the San Francisco
Bay. The city is situated in the Santa Clara Valley, between the Santa Cruz Mountains to the
south and west and the Diablo Range to the east. San Jose experiences a Mediterranean climate,
with distinct wet-dry seasons. The period from November through March represents the wet
season, with cool but mild conditions prevailing. Little rainfall occurs the rest of the year and
conditions tend to be warm and sunny. San Jose is not outside the marine influences of the cold
ocean currents typically affecting the San Francisco area (Wood, 2004; NOAA, 1999).
6.2.2 Meteorological Summary
Hourly meteorological data for 2012 were retrieved from NCDC for the stations closest
to the California monitoring sites (NCDC, 2012), as described in Section 3.5.2. The weather
station nearest CELA is located at Downtown Los Angeles/USC Campus; the weather station
nearest LBHCA is located at Long Beach/Daugherty Field Airport; the nearest weather station to
RUCA is located at Riverside Municipal Airport; and the nearest station to SJJCA is located at
San Jose International Airport (WBANs 93134, 23129, 03171, and 23293, respectively).
Additional information about these weather stations, such as the distance between the sites and
the weather stations, is provided in Table 6-3. These data were used to determine how
meteorological conditions on sample days vary from conditions experienced throughout the year.
Table 6-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for all of 2012. Also included in Table 6-3 is the 95 percent
confidence interval for each parameter. As shown in Table 6-3, average meteorological
conditions on sample days near CELA, RUCA, and SJJCA were representative of average
weather conditions experienced throughout the year. The most significant difference in the table
for these sites is for average dew point temperature for SJJCA, but is still only 1°F different.
6-13
-------
Table 6-3. Average Meteorological Conditions near the California Monitoring Sites
Closest Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar
Wind
Speed
(kt)
Los Angeles, California - CELA
Downtown
L.A./USC Campus
Airport
93134
(34.03, -118.30)
4.6
miles
244°
(WSW)
Sample
Days
(65)
2012
73.9
±2.3
73.8
±0.9
63.9
±1.8
63.8
±0.8
49.7
±2.5
49.8
±1.0
56.4
±1.6
56.4
±0.7
64.4
±3.5
64.6
±1.5
1015.4
±0.9
1014.8
±0.4
1.2
±0.2
1.2
±0.1
Long Beach, California - LBHCA
Long
Beach/Daugherty
Field Airport
23129
(33.82, -118.15)
2.5
miles
124°
(ESE)
Sample
Days
(32)
2012
75.6
±3.8
74.0
±1.0
66.7
±2.9
64.2
±0.8
54.7
±2.9
51.1
±0.9
59.9
±2.3
57.1
±0.7
68.9
±4.5
66.4
± 1.3
1015.3
±1.3
1015.0
±0.4
3.8
±0.5
3.9
±0.2
Rubidoux, California - RUCA
Riverside Municipal
Airport
03171
(33.95, -117.44)
3.5
miles
214°
(SW)
Sample
Days
(63)
2012
80.3
±3.3
80.2
±1.3
66.5
±2.5
66.3
±1.1
44.6
±3.2
44.8
±1.3
55.1
±1.9
55.1
±0.8
52.6
±4.6
53.1
±1.8
1014.3
±1.0
1013.8
±0.4
3.6
±0.3
3.6
±0.1
San Jose, California - SJJCA
San Jose Intl.
Airport
23293
(37.36, -121.93)
1.7
miles
312°
(NW)
Sample
Days
(66)
2012
70.2
±2.3
70.0
±1.0
59.4
±1.7
58.9
±0.7
47.6
±1.8
46.6
±0.8
53.2
±1.5
52.5
±0.6
68.0
±2.4
67.4
±1.2
1016.9
±1.0
1016.7
±0.4
5.9
±0.6
5.7
±0.2
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
-------
The differences between the average meteorological conditions for 2012 and those
experienced on sample days near LBHCA are greater than the other sites. However, sampling at
LBHCA did not begin until July; therefore, the sample day averages for this site include only
data for the second half of 2012. However, the differences between the full-year averages and the
sample day averages are still relatively small, with the largest difference for dew point
temperature.
Table 6-3 shows that wind speeds near the southern California sites tend to be rather
light, particularly for CELA, which has the lowest average scalar wind speed among all NMP
sites. As expected, conditions tended to be cooler near SJJCA than near the other sites. For the
southern California sites, average temperatures tended to be slightly higher for RUCA, which is
farther inland than the other two sites.
6.2.3 Back Trajectory Analysis
Figure 6-8 is the composite back trajectory map for days on which samples were
collected at the CELA monitoring site in 2012. Included in Figure 6-8 are four back trajectories
per sample day. Figure 6-9 is the corresponding cluster analysis. Similarly, Figures 6-10 through
6-15 are the composite back trajectory maps for days on which samples were collected at
LBHCA, RUCA, and SJJCA, respectively, and the corresponding cluster analyses. An in-depth
description of these maps and how they were generated is presented in Section 3.5.2.1. For the
composite maps, each line represents the 24-hour trajectory along which a parcel of air traveled
toward the monitoring site on a given sample day and time, based on an initial height of 50
meters AGL. For the cluster analyses, each line corresponds to a trajectory representative of a
given cluster of back trajectories. Each concentric circle around the sites in Figures 6-8 through
6-15 represents 100 miles.
6-15
-------
Figure 6-8. Composite Back Trajectory Map for CELA
Figure 6-9. Back Trajectory Cluster Map for CELA
100 200
6-16
-------
Figure 6-10. Composite Back Trajectory Map for LBHCA
Figure 6-11. Back Trajectory Cluster Map for LBHCA
100 200
6-17
-------
Figure 6-12. Composite Back Trajectory Map for RUCA
Figure 6-13. Back Trajectory Cluster Map for RUCA
100 200
6-18
-------
Figure 6-14. Composite Back Trajectory Map for SJJCA
Figure 6-15. Back Trajectory Cluster Map for SJJCA
6-19
-------
Observations from Figures 6-8 and 6-9 for CELA include the following:
• The 24-hour air shed domain for CELA is among the smaller ones compared to other
NMP monitoring sites, based on the average length of back trajectories (174 miles).
Although the farthest away a back trajectory originated was off the northern
California coast, or nearly 500 miles away, most back trajectories (84 percent)
originated within 250 miles of CELA. Only three back trajectories originated greater
than 400 miles away.
• Back trajectories originated from a variety of directions at CELA. However, a large
number of back trajectories originated from the northwest over the Pacific Ocean and
along the California coastline. Another cluster originated from the east-northeast.
Fewer back trajectories originated from the east, southeast, south, or southwest.
• The cluster analysis shows that roughly three-quarters of back trajectories originated
from the northwest and/or offshore, although of varying distances. The shorter cluster
trajectory (25 percent) includes back trajectories originating to the northwest of Los
Angeles, south of Bakersfield and Santa Maria, as well as shorter back trajectories
originating just offshore. Another 23 percent of back trajectories originated offshore
west of San Luis Obispo and 28 percent originated towards Monterrey and San
Francisco and the adjacent offshore waters. The cluster trajectory originating over the
interior of California (16 percent) represents back trajectories originating over the
desert areas of southern California as well as southern portions of Nevada. The short
cluster trajectory (8 percent) originating due south of the Los Angeles area includes
back trajectories originating over the San Diego area as well as the offshore waters
between the two metro areas.
Observations from Figures 6-10 and 6-11 for LBHCA include the following:
• The composite back trajectory map for LBHCA is similar to the CELA map in back
trajectory distribution, although there are roughly half the back trajectories shown in
Figure 6-10, as this site did not begin sampling until July. The 24-hour air shed
domain for LBHCA is slightly smaller than CELA's, based on the average length of
back trajectories (160 miles). The farthest away a back trajectory originated was over
the Pacific Ocean, off the northern California coast, or just greater than 500 miles
away. However, most trajectories (89 percent) originated within 250 miles of
LBHCA and only three originated greater than 300 miles away.
• The cluster analysis for LHBCA is very similar to the cluster analysis for CELA in
trajectory distribution, although the percentages differ. While back trajectories
originating from a northwesterly direction account for more than 72 percent of the
back trajectories, back trajectories originating to the south account for a higher
percentage than those originating to the northeast compared to CELA.
6-20
-------
Observations from Figures 6-12 and 6-13 for RUCA include the following:
• The composite back trajectory map for RUCA is very similar to the one for CELA,
which is not surprising given their relatively close proximity to each other. The 24-
hour air shed domain for RUCA is smaller in size to CELA, based on the average
back trajectory length (147 miles). The farthest away a back trajectory originated was
off the northern California coast, nearly 500 miles away. However, nearly 95 percent
of back trajectories originated within 250 miles of RUCA and only one back
trajectory originated farther than 400 miles away.
• The cluster analysis for RUCA is similar to the cluster analysis for CELA in that
nearly 70 percent of back trajectories originated from the northwest of the site.
However, the cluster analysis splits these into two cluster trajectories rather than
three. The shorter cluster trajectory (40 percent) includes back trajectories originating
primarily to the west and northwest of the site, along the coastline and offshore
waters of the Santa Barbara Channel while the other cluster trajectory (28 percent)
represents those back trajectories originating farther up the coast as far north as the
San Francisco area. The cluster analysis splits the north and northeastward originating
back trajectories into two cluster trajectories. One cluster (15 percent) includes
relatively short back trajectories originating primarily to the north and northeast of the
site while the other cluster (9 percent) includes the longer back trajectories originating
near and beyond the California/Nevada border. The final cluster originating to the
south of RUCA includes relatively short back trajectories originating toward and
offshore of the San Diego area as well as those originating to the east and southeast
over the Mojave and Sonora Deserts.
Observations from Figures 6-14 and 6-15 for SJJCA include the following:
• Based on the average length of the back trajectories, the 24-hour air shed domain for
SJJCA is larger than the air shed domains for the other California sites. The average
length of back trajectories for SJJCA is 236 miles. The farthest away a back trajectory
originated was over northeast Oregon or greater than 500 miles away, although a
second back trajectory of similar distance also originated well over the Pacific Ocean,
southwest of the monitoring site. Only 56 percent of back trajectories originated
within 250 miles of SJJCA, while greater than 80 percent originated within 250 miles
of CELA, LBHCA, and RUCA. Eighteen back trajectories originated farther than
400 miles away from the site.
• Back trajectories originated from a variety of directions at SJJCA, seemingly more so
than for the other California sites. However, the composite map still shows a large
number of back trajectories originated from the north, northwest, and along the coast.
Fewer back trajectories originated from other directions.
• The cluster analysis shows that 25 percent of back trajectories originated to the north
of the site, along the northern California coastline, although these are split into two
cluster trajectories based on the length of the back trajectory. Another 27 percent of
back trajectory originated farther offshore. Nearly 25 percent of back trajectories
originated offshore and to the west of the site, although these tended to be shorter in
length than those originating from a more northwesterly or northerly direction.
6-21
-------
Fifteen percent of back trajectories originated over central California and west-central
Nevada. These too tended to be shorter in length (less than 200 miles long). Finally,
the last eight percent of back trajectories are represented by the cluster trajectory
originating to the south of SJJCA, and include back trajectories originating over the
Los Angeles area and both the adjacent Pacific waters as well as those farther
offshore.
6.2.4 Wind Rose Comparison
Hourly surface wind data from the weather stations at the Downtown Los Angeles/USC
Campus (for CELA), Long Beach/Daugherty Field Airport (for LBHCA), Riverside Municipal
Airport (for RUCA), and San Jose International Airport (for SJJCA) were uploaded into a wind
rose software program to produce customized wind roses, as described in Section 3.5.2.2. 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 6-16 presents a map showing the distance between the weather station and CELA,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 6-16 also presents three different wind roses for the
CELA monitoring site. First, a historical wind rose representing 2002 to 2011 wind data is
presented, which shows the predominant surface wind speed and direction over an extended
period of time. Second, a wind rose representing wind observations for all of 2012 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2012 is
presented. These can be used to identify the predominant wind speed and direction for 2012 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically. Figures 6-17 through 6-19 present the distance maps and
wind roses for LBHCA, RUCA, and SJJCA, respectively.
6-22
-------
Figure 6-16. Wind Roses for the Downtown Los Angeles/USC Campus Weather Station
near CELA
Location of CELA and Weather Station
2002-2011 Historical Wind Rose
pill' V ' ,
iHr;i^
2012 Wind Rose
Sample Day Wind Rose
6-23
-------
Figure 6-17. Wind Roses for the Long Beach/Daugherty Field Airport Weather Station
nearLBHCA
Location of LBHCA and Weather Station
2002-2011 Historical Wind Rose
)~
1 *""
if
;j
«SI i
a
2012 Wind Rose
Sample Day Wind Rose
6-24
-------
Figure 6-18. Wind Roses for the Riverside Municipal Airport Weather Station near RUCA
Locations of RUCA and Weather Station
2002-2011 Historical Wind Rose
Weathei
Slalioi
Calms: 30.91%
2012 Wind Rose
Sample Day Wind Rose
WIND SPEED
(Knots)
LH >=22-
Hi 17-21
^| 11 - 17
^1 7- 11
I I 4- 7
HI 2- 4
WIND SPEED
(Knots)
HI 11 - 17
CH 4-y1
6-25
-------
Figure 6-19. Wind Roses for the San Jose International Airport Weather Station near
SJJCA
Location of SJJCA and Weather Station
2002-2011 Historical Wind Rose
2012 Wind Rose
Sample Day Wind Rose
6-26
-------
Observations from Figure 6-16 for CELA include the following:
• The weather station at the Downtown Los Angeles/USC Campus is located
approximately 4.6 miles west-southwest of CELA.
• Historically, winds were generally light near this site, with calm winds (< 2 knots)
observed for 60 percent of the wind observations. For wind speeds greater than
2 knots, westerly winds were most common, followed by easterly and west-
southwesterly winds. Wind speeds greater than 11 knots were not measured at this
weather station.
• The 2012 full-year and sample day wind roses are similar to the historical wind rose
in that calm winds make up the majority of the observations and that westerly winds
were prominent. However, a higher percentage of calm winds were measured in 2012
while west-southwesterly winds were rarely observed. Yet, the wind patterns shown
on the full-year and sample day wind roses generally resemble the historical wind
patterns, indicating that conditions in 2012 and on sample days were representative of
those experienced historically.
Observations from Figure 6-17 for LBHCA include the following:
• The weather station at the Long Beach/Daugherty Field Airport is located
approximately 2.5 miles east-southeast of LBHCA.
• The historical wind rose shows that calm winds were observed for more than one-
third of the observations near LBHCA. Winds from the west-northwest and northwest
together account for approximately 20 percent of the wind observations while winds
from the south account for another 10 percent of observations. Winds from the
northeast quadrant were generally not observed near this site.
• The wind patterns on the 2012 full-year wind rose are very similar to the historical
wind patterns, indicating that conditions in 2012 were representative of those
experienced historically. The sample day wind rose has a higher percentage of west-
northwesterly and northwesterly winds and fewer winds from the south and south-
southwest. Recall however, that sampling at LBHCA began in July, and thus does not
include wind observations from the first half of the year. The wind patterns on the
sample day wind rose may be indicative of a seasonal wind pattern.
Observations from Figure 6-18 for RUCA include the following:
• The weather station at the Riverside Municipal Airport is located south of the Santa
Ana River and Wildlife Area, approximately 3.5 miles southwest of RUCA.
• Although calm winds were observed for approximately 31 percent of the wind
observations near RUCA, westerly and west-northwesterly winds were also
frequently observed, accounting for approximately 21 percent and 12 percent of wind
observations, respectively, based on the historical wind rose.
6-27
-------
• The 2012 wind rose exhibits a higher percentage of calm winds (38 percent)
compared to the historical wind rose. In addition, westerly winds make up almost the
same percentage of wind observations in 2012 as both westerly and west-
northwesterly winds on the historical wind rose, as west-northwesterly winds were
observed infrequently in 2012. As similar observation was noted in the 2011 NMP
report.
• 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 in 2012
were representative of those experienced over the entire year.
Observations from Figure 6-19 for SJJCA include the following:
• The weather station at the San Jose International Airport is located 1.7 miles
northwest of SJJCA, across 1-880, the Guadalupe Parkway, and the Guadalupe River.
• Between 2002 and 2011, approximately 45 percent of winds were from the west-
northwest to north. Another 18 percent of winds were from the southeast to south.
Winds from the northeastern and southwestern quadrants were rarely observed.
Approximately one-fifth of the winds were calm.
• The wind patterns on the full-year and sample day wind roses exhibit a shift in
primary wind direction, from west-northwest to north on the historical wind rose to
west to north-northwest on the 2012 wind roses. This shift is also shown in the
secondary wind directions, from southeast to south on the historical to east-southeast
to southeast on the 2012 wind rose. This shift was also shown on the 2009, 2010, and
2011 wind roses in the 2008-2009, 2010, and 2011 NMP reports.
• The wind patterns shown on the sample day wind rose generally 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.
6-28
-------
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 6-4. 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 6-4. It is important to note which pollutants were sampled for at each site when reviewing
the results of this analysis. All four California sites sampled PAHs; in addition, SJJCA also
sampled metals (PMio).
Table 6-4. Risk-Based Screening Results for the California Monitoring Sites
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Los Angeles, California - CELA
Naphthalene
Fluorene
Acenaphthene
0.029
0.011
0.011
Total
60
12
2
74
60
60
60
180
100.00
20.00
3.33
41.11
81.08
16.22
2.70
81.08
97.30
100.00
Long Beach, California - LBHCA
Naphthalene
Benzo(a)pyrene
0.029
0.00057
Total
25
1
26
26
13
39
96.15
7.69
66.67
96.15
3.85
96.15
100.00
Rubidoux, California - RUCA
Naphthalene
0.029
Total
58
58
61
61
95.08
95.08
100.00
100.00
San Jose, California - SJJCA
Arsenic (PM10)
Naphthalene
Manganese (PM10)
Nickel (PM10)
Benzo(a)pyrene
0.00023
0.029
0.005
0.0021
0.00057
Total
45
43
32
7
1
128
58
59
61
61
15
254
77.59
72.88
52.46
11.48
6.67
50.39
35.16
33.59
25.00
5.47
0.78
35.16
68.75
93.75
99.22
100.00
6-29
-------
Observations from Table 6-4 include the following:
• Naphthalene failed the majority of screens for all three California monitoring sites
where only PAHs were sampled. Naphthalene's site-specific contribution to the total
failed screens for these sites ranges from 81 percent (CELA) to 100 percent (RUCA).
• Fluorene and acenaphthene also failed screens for CELA; however, only naphthalene
and fluorene were identified as pollutants of interest for CELA.
• Benzo(a)pyrene failed a single screen for LBHCA. Since naphthalene accounts for
96 percent of failed screens for LBHCA, only naphthalene is a pollutant of interest
for this site.
• Naphthalene is the only pollutant to fail screens for RUCA and is therefore RUCA's
only pollutant of interest.
• SJJCA is the only site for which naphthalene does not account for the majority of
failed screens; arsenic failed two more screens than naphthalene. Together, these two
pollutants account for nearly 70 percent of SJJCA's total failed screens. Manganese,
nickel, and benzo(a)pyrene also failed screens for this site. Arsenic, naphthalene,
manganese, and nickel contributed to 95 percent of failed screens for SJJCA and were
therefore identified as pollutants of interest for this site.
6.4 Concentrations
This section presents various concentration averages used to characterize 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 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. Additional site-specific statistical summaries for the
California monitoring sites are provided in Appendices M and N.
6-30
-------
6.4.1 2012 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 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 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
California monitoring sites are presented in Table 6-5, 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 6-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest
for the California Monitoring Sites
Pollutant
#of
Measured
Detections
vs. # 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)
Los Angeles, California - CELA
Fluorene
Naphthalene
60/60
60/60
4.78
±1.02
147.31
± 40.04
7.63
±1.24
184.68
±44.58
13.12
±2.12
237.14
± 40.44
5.64
±1.35
155.03
± 36.73
7.67
±1.07
179.67
± 20.99
Long Beach, California - LBHCA
Naphthalene
26/26
NA
NA
52.95
±8.70
96.81
±31.44
NA
Rubidoux, California - RUCA
Naphthalene
61/61
109.91
± 46.06
74.53
±18.15
82.21
±21.13
119.68
± 32.49
96.96
± 15.56
San Jose, California - SJJCA |
Arsenic (PM10)
Manganese (PM10)
Naphthalene
Nickel (PM10)
58/61
61/61
59/59
61/61
0.54
±0.25
7.05
±3.15
101.47
±43.40
1.22
±0.52
0.42
±0.12
5.27
±1.78
41.90
±11.09
1.17
±0.33
0.35
±0.08
6.78
±1.32
32.26
±11.86
1.35
±0.26
0.27
±0.13
5.74
±3.21
100.84
±41.23
0.92
±0.43
0.39 1
±0.08
6.22
±1.18
69.73
± 16.96
1.17
±0.19
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
6-31
-------
Observations for the California monitoring sites from Table 6-5 include the following:
• Naphthalene was identified as a pollutant of interest for all four sites. Concentrations
of naphthalene were highest at CELA and lowest at SJJCA, based on the annual
averages. LBHCA does not have an annual average presented in Table 6-5 because
sampling did not begin at this site until July. However, summary statistics for
LBHCA covering the sampling period are provided in Appendix M.
• Concentrations of naphthalene for CELA were highest in the second and third
quarters of 2012, particularly the third quarter. However, the confidence intervals
calculated for these quarterly averages indicate a high level of variability is associated
with these measurements. For example, naphthalene concentrations measured at
CELA ranged from 41.2 ng/m3 to 369 ng/m3 with a median concentration of
168 ng/m3. CELA has the second highest number of naphthalene concentrations
greater than 300 ng/m3 (seven) among NMP sites sampling PAHs. Of these seven, all
but one was measured between June and August.
• Fluorene concentrations at CELA were also highest during the second and third
quarters of 2012, particularly the third quarter. Fluorene concentrations ranged from
2.06 ng/m3 to 19.3 ng/m3 with a median concentration of 6.74 ng/m3. Of the 15
concentrations greater than 10 ng/m3 measured at CELA, all but two were measured
in either the second or third quarter of 2012. Conversely, of the 13 concentrations less
than 4 ng/m3, all but one was measured during the first or fourth quarters of 2012.
This supports the observations in Section 4.4.2 regarding fluorene measurements
being higher in the warmer months of the year.
• Concentrations of naphthalene measured at LBHCA ranged from 27.7 ng/m3 to
245 ng/m3. Because this site began sampling in July, only third and fourth quarter
averages are presented in Table 6-5. The fourth quarter average concentration is
significantly higher than the third quarter average. All five concentrations greater than
100 ng/m3 measured at LBHCA were measured in October, November, or December.
Further, the measurements collected in the fourth quarter have more variability
associated with them, as indicated by the confidence intervals. Measurements
collected between July and September ranged from 27.7 ng/m3 to 76.7 ng/m3, with a
median concentration of 56.1 ng/m3; measurements collected between October and
December ranged from 39.3 ng/m3 to 245 ng/m3, with a median concentration of
81.6 ng/m3.
• Concentrations of naphthalene at RUCA also tended to be higher during the colder
months of the year. Not only are the first and fourth quarter averages higher than the
other quarterly averages, they also have more variability associated with them.
Concentrations measured during the first and fourth quarters range from 9.09 ng/m3
to 374 ng/m3 with a median concentration of 103 ng/m3. Concentrations measured
during the second and third quarters range from 32.8 ng/m3 to 166 ng/m3 with a
median concentration of 69.1 ng/m3.
• Naphthalene concentrations measured at SJJCA follow a similar pattern as those
measured at RUCA. The first and fourth quarter naphthalene averages are
significantly higher than the other quarterly averages, and they too have more
6-32
-------
variability associated with them. Concentrations measured during the first and fourth
quarters range from 23.5 ng/m3 to 294 ng/m3 with a median concentration of
66.8 ng/m3. Concentrations measured during the second and third quarters range from
13.2 ng/m3 to 101 ng/m3 with a median concentration of 30.2 ng/m3.
• Manganese has the highest annual average concentration of the PMio metal pollutants
of interest for SJJCA, followed by nickel and arsenic. Although the quarterly
averages of manganese are not significantly different from each other, the first and
fourth quarter average concentrations have a relatively high level of variability
associated with them, as indicated by the confidence intervals. Concentrations
measured in the first and fourth quarters span approximately 22 ng/m3 between the
minimum and maximum measurement in each quarter while the range is less than
10 ng/m3 for the second and third quarters.
Tables 4-9 through 4-12 present the 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 appear in Table 4-11 for PAHs a total of five times. CELA has the
second highest annual average concentration naphthalene among NMP sites sampling
PAHs (behind only GPCO); RUCA ranks seventh for naphthalene. CELA and RUCA
rank sixth and seventh for fluorene, respectively. CELA also ranks sixth for
acenaphthalene, although RUCA does not appear in Table 4-11 for this pollutant (it
ranks 13th). SJJCA does not appear in Table 4-11.
• SJJCA appears twice in Table 4-12 for PMio metals. SJJCA has the seventh highest
annual average concentration of nickel and 10th highest annual average concentration
of manganese among NMP sites sampling PMio metals.
6.4.2 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 shaded in
gray in Table 6-4 for CELA, RUCA, and SJJCA. Figures 6-20 through 6-24 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 3.5.3.1. Because annual averages could not be calculated for LBHCA, box plots were not
created for this site.
6-33
-------
Figure 6-20. Program vs. Site-Specific Average Arsenic (PMi0) Concentration
SJJCA
Concentration (ng/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o
Figure 6-21. Program vs. Site-Specific Average Fluorene Concentration
10
20
30
40 50 60
Concentration (ng/m3)
70
80
90
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: SiteAverage Site Concentration Range
O
100
Figure 6-22. Program vs. Site-Specific Average Manganese (PMi0) Concentration
SJJCA
lo
1
1
30
60 90
Concentration (ng/m3)
120
150
Program:
Site:
IstQuartile
D
SiteAverage
0
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
6-34
-------
Figure 6-23. Program vs. Site-Specific Average Naphthalene Concentrations
CELA
RUCA
SJJCA
•• 1
• Ic
r
• ol
°l
0 1
Figure
|l_
l~1
0
,
1 1 1 1 1 1 1
1 1 1 1 1 1 1
1
1
JO 200 300 400 500 600 700 800 900
Concentration (ng/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
D D D D 1
Site: SiteAverage Site Concentration Range
o —
6-24. Program vs. Site-Specific Average Nickel (PMio) Concentration
1 1 1
1 1 1
I 4 6 8 10 12 14 16 18
Concentration (ng/m3)
Program:
Site:
IstQuartile
D
SiteAverage
O
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
6-35
-------
Observations from Figures 6-20 through 6-24 include the following:
• Figure 6-20 shows that the annual average arsenic (PMio) concentration for
SJJCA is less than both the program-level average and median concentrations of
arsenic (PMio). The annual average concentration of arsenic for SJJCA
(0.39 ng/m3) is just greater than the program-level first quartile (0.34 ng/m3).
Three non-detects of arsenic were measured at SJJCA. SJJCA is one of only three
sites to measure non-detects of this pollutant (UNVT and BTUT are the others).
• Figure 6-21 for fluorene includes only CELA because this is the only site for
which fluorene is a pollutant of interest. Figure 6-21 shows that the annual
average concentration of fluorene for CELA is greater than both the program-
level average and third quartile. Although the maximum concentration measured
at CELA is significantly less than the maximum concentration measured across
the program, the minimum concentration measured at CELA is greater than the
program-level first quartile. There were no non-detects of fluorene measured at
CELA, although a few were measured at other NMP sites sampling PAHs.
• Figure 6-22 is the box plot for manganese (PMio) for SJJCA. Note that the
program-level maximum concentration (275 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 to 150 ng/m3. Figure 6-22 shows that the annual average concentration of
manganese (PMio) for SJJCA is less than the program-level average concentration
and just greater than the program-level median concentration. The maximum
manganese concentration measured at SJJCA is an order of magnitude less than
the maximum concentration measured across the program. The minimum
concentration measured at SJJCA is one of the lowest concentrations measured
among NMP sites sampling PMio metals (only five measurements are lower).
• Figure 6-23 for naphthalene shows all three sites with available annual averages.
The box plots make an inter-site comparison relatively easy; the annual average
concentration is highest for CELA, followed by RUCA, and lowest SJJCA. The
annual average naphthalene concentration for CELA is greater than the program-
level average concentration and third quartile; the annual average concentration
for RUCA is just greater than the program-level average concentration; and the
annual average concentration for SJJCA is less than the program-level average
concentration but just greater than the program-level median concentration.
Figure 6-23 also shows the range of concentrations measured at each site.
Although the maximum concentrations measured at CELA and RUCA are
similar, the minimum concentration measured at RUCA is less than the minimum
concentration measured at CELA; further, the minimum concentration measured
at CELA is greater than the program-level first quartile. There were no non-
detects of naphthalene measured at CELA, RUCA, SJJCA, or across the program.
6-36
-------
• Figure 6-24 is the box plot for nickel (PMio) for SJJCA. Figure 6-24 shows that
the annual average concentration of nickel for SJJCA is just less than the
program-level average concentration. The maximum nickel concentration
measured at SJJCA is considerably less than the maximum concentration
measured across the program. There were no non-detects of nickel measured at
SJJCA or across the program.
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 3.5.3.2.
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 6-25 through 6-31 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 is not provided,
although the range and quartiles are still presented. A trends analysis was not conducted for
LBHCA because this site has not sampled under the NMP for at least 5 consecutive years.
6-37
-------
Figure 6-25. Yearly Statistical Metrics for Fluorene Concentrations Measured at CELA
O 5th Percentile
— Maximum O 95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until April 2007.
Observations from Figure 6-25 for fluorene measurements collected at CELA include the
following:
• CELA began sampling PAHs under the NMP at the end of April 2007. Because a full
year's worth of data is not available, a 1-year average is not presented, although the
range of measurements is provided.
• The smallest range of measurements was collected in 2007, although the statistical
metrics do not represent a full year of sampling. This was also the only year a non-
detect was measured. The range of measurements, and thus the statistical parameters
shown, increase through 2009, when the maximum fluorene concentration was
measured. The maximum concentration for 2009 is the only measurement greater
than 25 ng/m3 measured at this site. The maximum, 95th percentile, 1-year average,
and median concentrations decrease from 2009 to 2010 and again for 2011.
Concentrations measured in 2011 exhibit the least amount of variability besides the
initial year of sampling.
• All of the statistical parameters shown in Figure 6-25 exhibit an increase from 2011
to 2012.
6-38
-------
Figure 6-26. Yearly Statistical Metrics for Naphthalene Concentrations Measured at CELA
~ 400
O
•
-2
O...
O
O 5th Percentile
— Maximum • 95th Percentile "-O-" Averagf
A 1-year average is not presented because sampling under the NMP did not begin until April 2007.
Observations from Figure 6-26 for naphthalene measurements collected at CELA include
the following:
• The statistical parameters shown for naphthalene in Figure 6-26 exhibit a similar
pattern as the statistical parameters for fluorene shown in Figure 6-25.
• The smallest range of measurements was again collected 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); further, 2007 is
the only year in which a concentration less than 19 ng/m3 was measured. The range of
naphthalene measurements, and thus the statistical parameters shown, increase
through 2009, when the maximum concentration was measured (736 ng/m3).
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 6-26 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.
6-39
-------
Figure 6-27. Yearly Statistical Metrics for Naphthalene Concentrations Measured at RUCA
I,
...o
O 5th Percentile - Minimurr
— Maximum • 95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until May 2007.
Observations from Figure 6-27 for naphthalene measurements collected 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 is not presented, although the range of
measurements is provided.
• The smallest range of measurements was collected in 2007, although the statistical
metrics do not represent a full year of sampling.
• The maximum naphthalene concentration was measured at RUCA in 2009. This
concentration (406 ng/m3) is the only one greater than 400 ng/m3 measured at RUCA.
The second highest naphthalene concentration (374 ng/m3) was measured in 2012.
• The 1-year average concentration has an increasing trend over most of the years of
sampling, although 2010 was down slightly. The range of concentrations measured at
RUCA reflects the relatively high level of variability of the measurements collected.
For some years, the maximum concentration is driving the average upward. In the
case of 2009, the maximum concentration is twice the 95th percentile. Even though
the majority of concentrations measured in 2012 fall within a tighter range of
measurements, the 1-year average concentration is still higher for 2012 than 2011,
due in part to the maximum concentration measured. However, the 20 percent
increase in the median concentration indicates that concentrations were higher overall
for 2012.
6-40
-------
Figure 6-28. Yearly Statistical Metrics for Arsenic (PMi0) Concentrations Measured at SJJCA
O 5th Percentile
— Maximum O 95th Percentile "-O-" Averagf
Observations from Figure 6-28 for arsenic measurements collected at SJJCA include the
following:
• The maximum concentration of arsenic was measured on the first day of sampling at
this site (January 1, 2008). The second highest concentration was measured at the end
of 2008 and was roughly half as high.
• The 1-year average arsenic concentration decreased from 2008 to 2009. Although this
mostly due to the high concentration measured in 2008, the 95th percentile, median
(50th percentile), and 5th percentile all decreased from 2008 to 2009, indicating that
the decrease is not only due to the difference in the maximum concentrations.
• After a slight increase from 2009 to 2010, the 1-year average arsenic concentration
has not changed significantly. Between 2010 and 2012, the 1-year average
concentration ranged from 0.37 ng/m3 to 0.39 ng/m3. Even though the maximum and
95th percentile exhibit increases for 2012, the 5th percentile decreased to zero,
indicating that additional non-detects were measured in 2012. Thus, the number of
concentrations on both the low- and high-end of the concentration range increased for
2012.
6-41
-------
Figure 6-29. Yearly Statistical Metrics for Manganese (PMi0) Concentrations Measured at SJJCA
~ 15.0
2010
Year
O 5th Percentile
— Maximum
95th Percentile
Observations from Figure 6-29 for manganese measurements collected at SJJCA include
the following:
• The maximum concentration of manganese was measured in 2011. The eight highest
concentrations of manganese were all measured at SJJCA in either 2011 or 2012.
• After a slight decreasing trend, the 1-year average manganese concentration increased
significantly from 2010 to 2011. The median concentration nearly doubled for this
timeframe. The 95th percentile for both 2011 and 2012 is greater than the maximum
concentration measured in previous years. The difference between the 5th and 95th
percentiles nearly doubled from 2010 to 2011, indicating that the magnitude of the
majority of the measurements is higher for these years compared to previous years.
6-42
-------
Figure 6-30. Yearly Statistical Metrics for Naphthalene Concentrations Measured at SJJCA
O 5th Percentile
— Maximum
O 95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until May 2008.
Observations from Figure 6-30 for naphthalene measurements collected 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 is not presented, although the range of
measurements is provided.
• The maximum concentration of naphthalene was measured at SJJCA in 2009
(496 ng/m3). A measurement of similar magnitude has not been measured a second
time at SJJCA.
• The median concentration has changed little over the years of sampling, ranging from
43.0 ng/m3 (2010) to 49.9 ng/m3 (2011). The 1-year average concentration exhibits
more variability, ranging from 63.4 ng/m3 (2010) to 81.0 ng/m3 (2009), although the
changes are not statistically significant.
• There is very little change among the minimum concentrations and 5th percentiles
across the years of sampling while there are significant fluctuations in the statistical
parameters at the higher end of the concentration range. For example, the 95th
percentile increased by 70 percent from 2010 to 2011.
6-43
-------
Figure 6-31. Yearly Statistical Metrics for Nickel (PMi0) Concentrations Measured at SJJCA
O 5th Percentile - Minimurr
Median — Maximum O 95th Percentile
Observations from Figure 6-31 for nickel measurements collected at SJJCA include the
following:
• The statistical parameters shown for nickel in Figure 6-31 exhibit a similar pattern as
the statistical parameters for manganese shown in Figure 6-29.
• The two maximum concentrations of nickel were both measured in 2012 and are the
only concentrations measured at SJJCA greater than 3 ng/m3. The nine highest
concentrations of nickel were all measured in either 2011 or 2012.
• After a significant decreasing trend between 2008 and 2010, the 1-year average nickel
concentration increased significantly from 2010 to 2011. This trend is reflected in the
median concentrations as well. The 95th percentile for 2011 is greater than the
maximum concentration measured in previous years.
• Even though the nine highest concentrations of nickel were measured in 2011 and
2012, the six lowest concentrations were also measured in these years. The minimum
concentration decreased by half between 2009 and 2012.
6-44
-------
6.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk at
each California monitoring site. Refer to Sections 3.3 and 3.5.3 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.
6.5.1 Risk-Based Screening Assessment Using MRLs
A risk-based screening was conducted by comparing the concentration data from the
California monitoring sites to the ATSDR MRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 day to 14 days); intermediate (exposures of 15 days to 364 days); and chronic
(exposures of 1 year or greater). The preprocessed daily measurements of the pollutants of
interest were compared to the acute MRLs; the quarterly averages were compared to the
intermediate MRLs; and the annual averages were compared to the chronic MRLs.
As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2012.
6.5.2 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 3.5.3.4 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 6-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
6-45
-------
Table 6-6. Risk Approximations for the California Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections vs.
# of Samples
Annual
Average
(ng/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
Los Angeles, California - CELA
Fluorene
Naphthalene
0.000088
0.000034
0.003
60/60
60/60
7.67
±1.07
179.67
±20.99
0.67
6.11
0.06 |
Long Beach, California - LBHCA |
Naphthalene
0.000034
0.003
26/26
NA
NA
NA |
Rubidoux, California - RUCA
Naphthalene
0.000034
0.003
61/61
96.96
± 15.56
O O f\
3.30
0.03
San Jose, California - SJJCA
Arsenic (PM10)
Manganese (PM10)
Naphthalene
Nickel (PM10)
0.0043
0.000034
0.00048
0.000015
0.00005
0.003
0.00009
58/61
61/61
59/59
61/61
0.39
±0.08
6.22
±1.18
69.73
± 16.96
1.17
±0.19
1.69
2.37
0.56
0.03
0.12
0.02
0.01
— = A Cancer URE or Noncancer RfC is not available.
NA = Not available due to the criteria for calculating an annual average.
Observations for the California sites from Table 6-6 include the following:
• Naphthalene has the highest 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 for CELA is more than double the annual
average for SJJCA and is significantly higher than the annual average for RUCA.
• Naphthalene also has the highest cancer risk approximation among the site-specific
pollutants of interest for the California monitoring sites. The cancer risk
approximations range from 2.37 in-a-million for SJJCA to 6.11 in-a-million for
CELA.
• Of the metals listed for SJJCA, manganese has the highest annual average
concentration; however, this pollutant has no cancer toxicity factor. Arsenic has the
highest cancer risk approximation among the metals in Table 6-6 (1.69 in-a-million).
Even though the annual average concentration of arsenic is two orders of magnitude
less than the annual average of naphthalene, the cancer risk approximations are not
much different. This is an indication of the relative toxicity of arsenic compared to
naphthalene.
• All of the noncancer hazard approximations for the pollutants of interest for the
California monitoring sites are less than 1.0, indicating that no adverse health effects
are expected from these individual pollutants.
6-46
-------
• Cancer risk and noncancer hazard approximations could not calculated for LBHCA
due to the July start date of sampling, as discussed in the previous sections.
6.5.3 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 6-7 presents the 10 pollutants with the highest emissions from the 2011 NEI that have
cancer toxicity factors. Table 6-7 also presents the 10 pollutants with the highest toxicity-
weighted emissions, based on the weighting schema described in Section 3.5.3.5. Lastly,
Table 6-7 provides the 10 pollutants with the highest cancer risk approximations (in-a-million)
for each site, as presented in Table 6-6. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 6-7. Table 6-8 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 those pollutants for which each respective site
sampled. As discussed in Section 6.3, each of the California monitoring sites sampled PAHs;
SJJCA also sampled metals (PMio). 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. Thus, LBHCA does not have cancer risk and noncancer hazard
approximations in Tables 6-7 and 6-8. A more in-depth discussion of this analysis is provided in
Section 3.5.3.5. Similar to the cancer risk and noncancer hazard approximations provided in
Section 6.5.2, this analysis may help policy-makers prioritize their air monitoring activities.
6-47
-------
Table 6-7. 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)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
Los Angeles, California (Los Angeles County) - CELA
Formaldehyde
Dichloromethane
Tetrachloroethylene
Benzene
Ethylbenzene
Acetaldehyde
£>-Dichlorobenzene
1,3 -Butadiene
POM, Group la
Naphthalene
2,039.76
1,707.53
1,424.90
1,381.37
849.87
795.99
339.36
292.06
252.09
131.79
Formaldehyde
POM, Group la
Benzene
1,3 -Butadiene
Naphthalene
Arsenic, PM
£>-Dichlorobenzene
Hexavalent Chromium, PM
Ethylbenzene
POM, Group 2b
2.65E-02
2.22E-02
1.08E-02
8.76E-03
4.48E-03
4.29E-03
3.73E-03
2.62E-03
2.12E-03
1.89E-03
Naphthalene 6.11
Fluorene 0.67
Long Beach, California (Los Angeles County) - LBHCA
Formaldehyde
Dichloromethane
Tetrachloroethylene
Benzene
Ethylbenzene
Acetaldehyde
£>-Dichlorobenzene
1,3 -Butadiene
POM, Group la
Naphthalene
2,039.76
1,707.53
1,424.90
1,381.37
849.87
795.99
339.36
292.06
252.09
131.79
Formaldehyde
POM, Group la
Benzene
1,3 -Butadiene
Naphthalene
Arsenic, PM
£>-Dichlorobenzene
Hexavalent Chromium, PM
Ethylbenzene
POM, Group 2b
2.65E-02
2.22E-02
1.08E-02
8.76E-03
4.48E-03
4.29E-03
3.73E-03
2.62E-03
2.12E-03
1.89E-03
oo
-------
Table 6-7. 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
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
Rubidoux, California (Riverside County) - RUCA
Formaldehyde
Benzene
Tetrachloroethylene
Acetaldehyde
Dichloromethane
Ethylbenzene
£>-Dichlorobenzene
1,3 -Butadiene
POM, Group la
Naphthalene
532.83
284.75
272.91
246.53
212.10
178.59
70.48
66.97
58.14
30.14
Formaldehyde
POM, Group la
Benzene
1,3 -Butadiene
Hexavalent Chromium, PM
Arsenic, PM
Naphthalene
£>-Dichlorobenzene
Acetaldehyde
Ethylbenzene
6.93E-03
5.12E-03
2.22E-03
2.01E-03
1.90E-03
1.03E-03
1.02E-03
7.75E-04
5.42E-04
4.46E-04
Naphthalene 3.30
San Jose, California (Santa Clara County) - SJJCA
Formaldehyde
Benzene
Ethylbenzene
Dichloromethane
Tetrachloroethylene
Acetaldehyde
1,3 -Butadiene
POM, Group la
£>-Dichlorobenzene
Naphthalene
363.98
302.63
218.82
191.74
153.82
151.80
62.80
62.72
60.37
31.71
POM, Group la
Formaldehyde
Benzene
1,3 -Butadiene
Hexavalent Chromium, PM
Naphthalene
Arsenic, PM
£>-Dichlorobenzene
POM, Group 2b
Ethylbenzene
5.52E-03
4.73E-03
2.36E-03
1.88E-03
1.59E-03
1.08E-03
1.04E-03
6.64E-04
5.58E-04
5.47E-04
Naphthalene 2.37
Arsenic 1.69
Nickel 0.56
VO
-------
Table 6-8. 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)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Noncancer Hazard
Approximation
Pollutant (HQ)
Los Angeles, California (Los Angeles County) - CELA
Toluene
1,1,1 -Trichloroethane
Ethylene glycol
Xylenes
Hexane
Formaldehyde
Dichloromethane
Tetrachloroethylene
Benzene
Methanol
8,302.59
6,903.35
4,337.04
4,120.59
3,927.94
2,039.76
1,707.53
1,424.90
1,381.37
1,338.87
Acrolein
Chlorine
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Arsenic, PM
Cyanide Compounds, PM
Cadmium, PM
Trichloroethylene
Benzene
5,981,887.03
290,023.06
208,138.95
146,028.22
88,443.41
66,543.97
63,440.92
62,581.30
60,450.02
46,045.79
Naphthalene 0.06
Long Beach, California (Los Angeles County) - LBHCA
Toluene
1,1,1 -Trichloroethane
Ethylene glycol
Xylenes
Hexane
Formaldehyde
Dichloromethane
Tetrachloroethylene
Benzene
Methanol
8,302.59
6,903.35
4,337.04
4,120.59
3,927.94
2,039.76
1,707.53
1,424.90
1,381.37
1,338.87
Acrolein
Chlorine
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Arsenic, PM
Cyanide Compounds, PM
Cadmium, PM
Trichloroethylene
Benzene
5,981,887.03
290,023.06
208,138.95
146,028.22
88,443.41
66,543.97
63,440.92
62,581.30
60,450.02
46,045.79
-------
Table 6-8. 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)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Noncancer Hazard
Approximation
Pollutant (HQ)
Rubidoux, California (Riverside County) - RUCA
Toluene
Xylenes
Hexane
Ethylene glycol
1,1,1 -Trichloroethane
Formaldehyde
Benzene
Tetrachloroethylene
Acetaldehyde
Methanol
1,799.61
1,020.69
958.45
835.09
617.83
532.83
284.75
272.91
246.53
218.81
Acrolein
Chlorine
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Arsenic, PM
Propionaldehyde
Bromomethane
Trichloroethylene
Lead, PM
1,281,660.39
98,782.17
54,370.16
33,482.53
27,392.75
15,991.32
14,957.55
13,246.82
12,385.28
11,114.81
Naphthalene 0.03
San Jose, California (Santa Clara County) - SJJCA
Toluene
1,1,1 -Trichloroethane
Xylenes
Hexane
Ethylene glycol
Formaldehyde
Benzene
Ethylbenzene
Methanol
Dichloromethane
1,704.32
1,289.63
979.05
892.98
826.56
363.98
302.63
218.82
216.15
191.74
Acrolein
Chlorine
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Arsenic, PM
Trichloroethylene
Naphthalene
Benzene
Xylenes
2,001,785.79
139,092.17
37,140.79
31,399.05
16,866.55
16,104.51
14,797.50
10,571.46
10,087.68
9,790.52
Manganese 0.12
Arsenic 0.03
Naphthalene 0.02
Nickel 0.01
-------
Observations from Table 6-7 include the following:
• Formaldehyde is the highest emitted pollutant with cancer UREs in all three
California counties. The quantity emitted is greater for Los Angeles County than
Riverside and Santa Clara Counties. Dichloromethane is the second highest emitted
pollutant in Los Angeles County but ranks fourth and fifth for Santa Clara and
Riverside Counties, respectively. Benzene is the second highest emitted pollutant in
Santa Clara and Riverside Counties but ranks fourth for Los Angeles County.
• Formaldehyde and POM, Group 1 are the pollutants with the highest toxicity-
weighted emissions (of the pollutants with cancer UREs) for Los Angeles and
Riverside Counties, while the order is reversed for Santa Clara County. Benzene
ranks third for all three counties.
• Seven of the highest emitted pollutants also have the highest toxi city-weighted
emissions for Los Angeles and Santa Clara Counties, while there are eight in common
for Riverside County. While dichloromethane and tetrachloroethylene are among the
highest emitted pollutants for each county, neither pollutant appears among those
with the highest toxicity-weighted emissions. Conversely, hexavalent chromium and
arsenic are among those with the highest toxicity-weighted emissions for all three
counties, but are not among the highest emitted pollutants.
• Naphthalene has the highest cancer risk approximation for all three sites for which
annual averages could be calculated. Naphthalene appears on both emissions-based
lists for all three counties.
• Arsenic, which has the second highest cancer risk approximation for SJJCA, has the
seventh highest toxicity-weighted emissions for Santa Clara County, but is not among
the highest emitted pollutants for this county (and ranks 20th). Nickel, the only other
pollutant of interest for SJJCA, does not appear on either emissions-based list.
Observations from Table 6-8 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 fifth for
Riverside County. Xylenes are the second highest emitted pollutant in Riverside
County but ranks third and fourth for Santa Clara and Los Angeles Counties,
respectively.
• Acrolein, chlorine, and formaldehyde are the pollutants with the highest toxicity-
weighted emissions (of the pollutants with noncancer RfCs) for all three counties.
While acrolein and chlorine rank highest for toxicity-weighted emissions for each
county, neither pollutant appears among the highest emitted. Conversely,
formaldehyde has the sixth highest emissions for each county.
6-52
-------
• Three of the highest emitted pollutants also have the highest toxi city -weighted
emissions for Santa Clara County, while only two of the highest emitted pollutants
also have the highest toxicity-weighted emissions for Los Angeles and Riverside
Counties.
• Naphthalene, the only pollutant for which a noncancer hazard approximation could be
calculated for CELA and RUCA, does not appear on either emissions based list in
Table 6-8. Naphthalene ranks eighth for toxicity-weighted emissions for Santa Clara
County but is not one of the highest emitted (of pollutants with noncancer RfCs).
• Manganese, which has the highest noncancer hazard approximation for SJJCA, does
not appear on either emissions-based list in Table 6-8. This is also true for nickel.
Arsenic ranks sixth for its toxicity-weighted emissions but is also not one of the
highest emitted pollutants in Santa Clara County.
6.6 Summary of the 2012 Monitoring Data for the California Monitoring Sites
Results from several of the data treatments described in this section include the
following:
»«» Naphthalene failed screens for all four California sites. Two additional PAH s failed
screens for CELA, one additional PAH failed screens for LBHCA, and only
naphthalene failed screens for RUCA. Two PAHs and three PMw metals failed
screens for SJJCA.
»«» Naphthalene had the highest annual average concentration among the site-specific
pollutants of interest for each of the California monitoring sites. CELA has the
second highest annual average concentration of naphthalene among NMP sites
sampling PAHs. Among the metals sampled at SJJCA, manganese had the highest
annual average concentration, which ranks tenth among other NMP sites sampling
metals.
Concentrations of naphthalene were higher during the first and fourth quarters (or
the colder months) of 2012 for RUCA and SJJCA; conversely, naphthalene
concentrations were higher during the second and third quarters (or warmer months)
for CELA.
Concentrations of naphthalene andfluorene increased at CELA from 2011 to 2012.
Significant increases in manganese and nickel concentrations at SJJCA occurred
between 2010 and 2011, with little change for 2012.
6-53
-------
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 other five
sites are located in Garfield County, between 35 miles and 76 miles northeast of Grand Junction,
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. Note that only sources within 10 miles of the site 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 radius are still visible on the map,
but have been grayed out in order to show emissions sources just outside 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
»t a • -Tfj i E •
VJJf ^\j|r, . , ,%
^ *" ' •^BSR™
-------
Figure 7-2. NEI Point Sources Located Within 10 Miles of GPCO
0 2.5 5 10 ,
I 1 1 1 1 1 I 1 1
Miles
Legend
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
GPCO NATTS site
O 10 mile radius
County boundary
Source Category Group (No. of Facilities)
"T Airport/Airline/Airport Support Operations (9) ^
'i Asphalt Production/Hot Mix Asphalt Plant (2) O
0 Auto Body Shop/Painters/Automotive Stores (4) •
•[-' Automotive/RV Dealership (1) A
— Brick, Structural Clay, or Clay Ceramics Plant (1) ©
B Bulk Terminals/Bulk Plants (4) X
C Chemical Manufacturing (3) ?
i Compressor Station (1) •
EX] Crematory - Animal/Human (4) R
(!) Dry Cleaning (3) X
6 Electrical Equipment Manufacturing (1) T
E Electroplating, Plating, Polishing, Anodizing, and Coloring (2) *
IT Gasoline/Diesel Service Station (43) W
Industrial Machinery or Equipment Plant (3)
Institution (school, hospital, prison, etc.) (4)
Landfill (1)
Metal Coating, Engraving, and Allied Services to Manufacturers (1)
Metals Processing/Fabrication (2)
Mine/Quarry/Mineral Processing (35)
Miscellaneous Commercial/Industrial (1)
Oil and/or Gas Production (2)
Plastic, Resin, or Rubber Products Plant (1)
Rail Yard/Rail Line Operations (2)
Textile, Yarn, or Carpet Plant (1)
Wastewater Treatment (1)
Woodwork, Furniture, Millwork & Wood Preserving (1)
7-3
-------
Figure 7-3. Battlement Mesa, Colorado (BMCO) Monitoring Site
-------
Figure 7-4. Silt, Colorado (BRCO) Monitoring Site
-------
Figure 7-5. Parachute, Colorado (PACO) Monitoring Site
-------
Figure 7-6. Rifle, Colorado (RICO) Monitoring Site
-------
Figure 7-7. NEI Point Sources Located Within 10 Miles of BMCO, BRCO, PACO, and
RICO
108°15'0"W 108'10'trW 108C5'0"W 108°0'0"W 107'55'0"W 107°50'0"W 107°45'0"W 107°40'0"W 107"35'0"W
Legend
l^
108*5'0"W lOS'O'tTW 107'55'0"W 107-50'0"W 107'45'0"W 107r40'0"W 107°35'0"W 107;30'0"W 107125'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
RICO UATMP site
BMCO UATMP site * BRCO UATMP site
O 10 mile radius
Source Category Group (No. of Facilities)
t Airport/Airline/Airport Support Operations (7)
£ Asphalt Production/Hot Mix Asphalt Plant (1 )
B Bulk Terminals/Bulk Plants (1 )
: Compressor Station (20)
<]> Dry Cleaning (1)
* Electricity Generation via Combustion (1)
© Gas Plant (2)
UATMP site
County boundary
r Gasoline/Diesel Service Station (19)
• Landfill (1)
Mine/Quarry/Mineral Processing (10)
? Miscellaneous Commercial/Industrial (1)
• Oil and/or Gas Production (1,056)
x Rail Yard/Rail Line Operations (1)
TT Telecommunications/Radio (1)
7-8
-------
Figure 7-8. Carbondale, Colorado (RFCO) Monitoring Site
-------
Figure 7-9. NEI Point Sources Located Within 10 Miles of RFCO
Legend
RFCO UATMP site
107°15'0"W 107'10'0"W 107°51011W
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
Source Category Group (No. of Facilities)
T Airport/Airline/Airport Support Operations (6)
f Building/Construction (1)
i Compressor Station (1)
IX Crematory - Animal/Human (1)
If Gasoline/Diesel Service Station (10)
o Institution (school, hospital, prison, etc.) (1)
x Mine/Quarry/Mineral Processing (3)
7-10
-------
Table 7-1. Geographical Information for the Colorado Monitoring Sites
Site
Code
GPCO
BMCO
BRCO
PACO
RICO
RFCO
AQS Code
08-077-0017
08-077-0018
None
08-045-0009
08-045-0005
08-045-0007
08-045-0018
Location
Grand
Junction
Battlement
Mesa
Silt
Parachute
Rifle
Carbondale
County
Mesa
Garfield
Garfield
Garfield
Garfield
Garfield
Micro- or
Metropolitan
Statistical Area
Grand Junction,
CO MSA
Glenwood Springs,
CO MSA
Glenwood Springs,
CO MSA
Glenwood Springs,
CO MSA
Glenwood Springs,
CO MSA
Glenwood Springs,
CO MSA
Latitude
and
Longitude
39.064289,
-108.56155
39.439989,
-108.029769
39.487755,
-107.659685
39.453654,
-108.053259
39.531813,
-107.782298
39.412278,
-107.230397
Land Use
Commercial
Residential
Agricultural
Residential
Commercial
Residential
Location
Setting
Urban/City
Center
Rural
Rural
Urban/City
Center
Urban/City
Center
Rural
Additional Ambient Monitoring Information1
Meteorological parameters, CO, PM10, PM10
Speciation, PM Coarse, PM2 5, and PM2 5 Speciation,
IMPROVE Speciation.
No AQS entry.
None.
PM10.
PM10.
PM10.
:Data for additional pollutants are reported to AQS for these sites (EPA, 2013b); however, these data are not generated by ERG and are therefore not included in this report.
BOLD ITALICS = EPA-designaled 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 hexavalent chromium 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 one
of the major east-west roads in Grand Junction (1-70 Business). A railroad runs east-west to the
south of the GPCO monitoring site, and merges with another railroad to the southwest of the site.
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 and 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.
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
cemetery is located to the south of the site and a church is located to the east.
The BRCO monitoring site is located on Bell/Melton Ranch, off Owens Drive,
approximately 4 miles south of the town of Silt. The site is both rural and agricultural in nature.
As shown in Figure 7-4, the closest major roadway is County Road 331, Dry Hollow Road.
PACO is located on the roof of the old Parachute High School building, which is
presently operating as a day care facility. This location is in the center of the town of Parachute,
as shown in Figure 7-5. The surrounding area is considered residential. Interstate-70 is less than
a quarter of a mile from the monitoring site.
7-12
-------
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 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 considered commercial.
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. One reason Garfield County is conducting air
monitoring is to characterize the effects these wells may have on the surrounding areas (GCPH,
2013).
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, 2014). 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 considered residential and rural. 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, two
mine/quarry/mineral processing facilities, and an airport within a few miles of this site.
7-13
-------
Table 7-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Colorado monitoring sites. Table 7-2 includes both county-level
population and vehicle registration information. Table 7-2 also contains traffic volume
information for each site as well as the location for which the traffic volume was obtained.
Additionally, Table 7-2 presents the county-level daily VMT for Mesa and Garfield Counties.
Note that the VMT presented is for state highways only, which differs from the VMT presented
in this table in most other state sections.
Table 7-2. Population, Motor Vehicle, and Traffic Information for the Colorado
Monitoring Sites
Site
GPCO
BMCO
BRCO
PACO
RFCO
RICO
Estimated
County
Population1
147,848
56,953
County-level
Vehicle
Registration2
179,213
74,508
Annual
Average Daily
Traffic3
11,000
2,527
1,102
16,000
16,000
17,000
Intersection
Used for
Traffic Data
PitkinAve, east of 7*81
S. Battlement Pkwy (CO Road 300)
Diy Hollow Rd (CO Road 331)
1-70 near exit 75
Route 133, south of 82
Route 13, Route 6 at 1-70
County-level
Daily VMT4
2,009,730
1,902,077
Bounty-level population estimates reflect 2012 data (Census Bureau, 2013c).
2County-level vehicle registration reflects 2011 data (CO DOR, 2012).
3AADT reflects 2002 data for BMCO and BRCO from Garfield County (GCRBD, 2002) and 2011 data for GPCO,
PACO, RFCO and RICO from the Colorado DOT (CO DOT, 2011).
4County-level VMT reflects 2012 data for state highways only (CO DOT, 2012).
BOLD ITALICS = EPA-designaled NATTS Site
Observations from Table 7-2 include the following:
• Mesa County's population and vehicle ownership are considerably higher than those
for Garfield County. However, both counties rank in the bottom-third compared to
other counties with NMP sites.
• 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.
• While the Mesa and Garfield County VMTs are fairly similar to each other, they are
also among the lowest for counties with NMP sites, where VMT data were available.
However, the county-level VMT available from the Colorado DOT is for state
highways only and is therefore biased low compared to other sites.
7-14
-------
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 Climate Summary
Grand Junction is located in a mountain valley on the west side of the Rockies. The
mountains surrounding the valley help protect the city from dramatic weather changes. The area
tends to be fairly dry, with annual precipitation amounts less than 10 inches. On average, one to
two snowfalls occur during each of the winter months, but tend to be short-lived in duration.
Winds tend to flow out of the east-southeast on average, due to the valley breeze effect (Wood,
2004). Valley breezes occur as the sun heats up the side of a mountain; the warm air rises,
creating a current that will move up the valley walls (Boubel, et al., 1994).
The towns of Battlement Mesa, Parachute, Rifle, and Silt are located to the northeast of
Grand Junction, across the county line and along the 1-70 corridor. These towns are located along
a river valley running north of the Grand Mesa. The town of Carbondale is farther east, in a river
valley in the southeast corner of Garfield County. Similar to Grand Junction, these towns are
shielded from drastic changes in weather by the surrounding terrain and tend to experience fairly
dry conditions for most of the year. Wind patterns in these towns are affected by the high
canyons, the Colorado River and its tributaries, and valley breezes (GCPH, 2013; WRCC, 2013).
7.2.2 Meteorological Summary
Hourly meteorological data for 2012 were retrieved from NCDC for the weather stations
closest to the Colorado monitoring sites (NCDC, 2012), as described in Section 3.5.2. The
weather station nearest GPCO is located at Walker Field Airport (WBAN 23066). The closest
weather station to four of the five Garfield County sites is located at Garfield County Regional
Airport (WBAN 03016) while the weather station closest to RFCO is located at Aspen-Pitkin
County Airport (WBAN 93073). Additional information about these weather stations, such as the
distance between the sites and the weather stations, is provided in Table 7-3. These data were
used to determine how meteorological conditions on sample days vary from conditions
experienced throughout the year.
7-15
-------
Table 7-3. Average Meteorological Conditions near the Colorado Monitoring Sites
Closest Weather
Station (WBAN
and Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
<°F)
Average
Dew Point
Temperature
<°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Grand Junction, Colorado - GPCO
Walker Field
Airport
23066
(39.13, -108.54)
4.9
miles
21°
(NNE)
Sample
Day
(75)
2012
65.9
±5.1
68.3
±2.2
52.9
±4.7
55.4
±2.0
25.5
±2.9
26.1
±1.3
40.1
±3.2
41.6
±1.3
43.7
±5.0
40.7
±2.1
1016.1
±1.8
1014.5
±0.8
5.8
±0.5
6.3
±0.3
Battlement Mesa, Colorado - BMCO
Garfield County
Regional Airport
03016
(39.53, -107.73)
16.7
miles
76°
(ENE)
Sample
Day
(59)
2012
65.5
±5.5
65.9
±2.2
50.2
±4.8
50.5
±1.9
27.2
±3.2
26.2
±1.3
39.2
±3.4
39.1
±1.4
50.0
±4.9
47.7
±1.9
1017.5
±1.9
1016.4
±0.8
4.1
±0.6
4.5
±0.3
Silt, Colorado - BRCO
Garfield County
Regional Airport
03016
(39.53, -107.73)
4.1
miles
320°
(NW)
Sample
Day
(61)
2012
66.0
±5.4
65.9
±2.2
50.7
±4.8
50.5
±1.9
27.8
±3.2
26.2
±1.3
39.7
±3.4
39.1
±1.4
50.0
±4.7
47.7
±1.9
1017.3
±1.8
1016.4
±0.8
4.2
±0.6
4.5
±0.3
Parachute, Colorado - PACO
Garfield County
Regional Airport
03016
(39.53, -107.73)
17.5
miles
81°
(E)
Sample
Day
(57)
2012
66.3
±5.7
65.9
±2.2
50.7
±5.1
50.5
±1.9
26.8
±3.3
26.2
±1.3
39.3
±3.6
39.1
±1.4
48.6
±4.9
47.7
±1.9
1017.1
±1.9
1016.4
±0.8
4.3
±0.6
4.5
±0.3
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
-------
Table 7-3. Average Meteorological Conditions near the Colorado Monitoring Sites (Continued)
Closest NWS
Station (WBAN
and Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
<°F)
Average
Dew Point
Temperature
<°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Rifle, Colorado - RICO
Garfield County
Regional Airport
03016
(39.53, -107.73)
3.2
miles
102°
(ESE)
Sample
Day
(61)
2012
65.8
±5.5
65.9
±2.2
50.6
±4.8
50.5
±1.9
27.4
±3.1
26.2
±1.3
39.5
±3.3
39.1
±1.4
49.7
±4.9
47.7
±1.9
1017.1
±1.9
1016.4
±0.8
4.3
±0.6
4.5
±0.3
Carbondale, Colorado - RFCO
Aspen-Pitkin
County Airport
93073
(39.23, -106.87)
22.1
miles
132°
(SE)
Sample
Day
(18)
2012
64.7
±9.4
58.3
±1.9
50.5
±8.0
43.7
±1.7
27.9
±6.6
22.7
±1.3
39.7
±6.0
34.3
±1.3
49.7
±8.8
50.8
±1.7
1015.5
±2.8
1014.9
±0.8
4.6
±0.5
5.1
±0.2
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages
-------
Table 7-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for all of 2012. Also included in Table 7-3 is the 95 percent
confidence interval for each parameter. As shown in Table 7-3, average meteorological
conditions on sample days near BMCO, BRCO, PACO, and RICO were representative of
average weather conditions experienced throughout the year. The parameter with the highest
difference between the full-year average and the sample day average for these sites is relative
humidity.
The differences between the sample day and full-year averages for the temperature,
relative humidity, and pressure parameters for GPCO are higher than most of the Garfield
County sites. A review of the data shows that there were 14 make-up days for GPCO, the
majority of which were collected in the cooler months of the year (10 of these were collected
between January and February or October through December). This explains why conditions on
sample days appear cooler than conditions experienced over the entire year.
For RFCO, the temperature parameters on sample days are considerably higher than
those shown for the entire year. RFCO did not begin sampling until June 2012, thereby missing
the coldest months of the year. RFCO also sampled on a l-in-12 day schedule, yielding roughly
half the number of collection events as the other sites; thus, the number of observations included
in each calculation for RFCO is less than the other sites. As a result, the confidence intervals
indicate a higher level of variability in the meteorological parameters for this site.
The lowest average dew point and wet bulb temperatures among NMP sites were
calculated for the Colorado monitoring sites. These sites also experienced some of the lowest
relative humidity levels among NMP sites.
7.2.3 Back Trajectory Analysis
Figure 7-10 is the composite back trajectory map for days on which samples were
collected at the GPCO monitoring site. Included in Figure 7-10 are four back trajectories per
sample day. Figure 7-11 is the corresponding cluster analysis. Similarly, Figures 7-12 through
7-20 are the composite back trajectory maps and corresponding cluster analyses for the Garfield
7-18
-------
County monitoring sites. An in-depth description of these maps and how they were generated is
presented in Section 3.5.2.1. For the composite maps, each line represents the 24-hour trajectory
along which a parcel of air traveled toward the monitoring site on a given sample day and time,
based on an initial height of 50 meters AGL. For the cluster analyses, each line corresponds to a
trajectory representative of a given cluster of back trajectories. Each concentric circle around the
sites in Figures 7-10 through 7-20 represents 100 miles.
Observations for GPCO from Figures 7-10 and 7-11 include the following:
• The 24-hour air shed domain for GPCO is the second smallest in size, based on
average back trajectory length (132 miles), compared to other NMP monitoring sites.
Only RFCO has a smaller average back trajectory length (131 miles). The farthest
away a back trajectory originated was near the western border of Idaho, or just less
than 500 miles away. However, most trajectories (90 percent) originated within
250 miles of GPCO.
• Back trajectories originated from a variety of directions at GPCO, although a majority
of the back trajectories had a westerly component. A large cluster of back trajectories
originated to the southwest of GPCO and a second cluster originated to the northwest
of the site.
• The cluster analysis shows that about one-third of back trajectories originated from
the southwest and west of GPCO. These are split into two cluster trajectories, one
representing shorter back trajectories originating over southeast Utah and the other
representing those originating over northern Arizona. Another 31 percent of back
trajectories originated within approximately 100 miles of GPCO and are represented
by the short cluster trajectory shown in the inset in the bottom-right side of the figure.
Seventeen percent of back trajectories originated from the northwest of GPCO. These
too are split into two cluster trajectories, one representing shorter back trajectories
originating over northern Utah and the other representing longer back trajectories
originating over Idaho and northeast Nevada. Back trajectories originating over the
southeast corner of Colorado account for 17 percent of back trajectories while back
trajectories originating over the northwest corner of Colorado or south-central
Wyoming account for 3 percent of back trajectories.
7-19
-------
Figure 7-10. Composite Back Trajectory Map for GPCO
Figure 7-11. Back Trajectory Cluster Map for GPCO
7-20
-------
Figure 7-12. Composite Back Trajectory Map for BMCO
Figure 7-13. Back Trajectory Cluster Map for BMCO
7-21
-------
Figure 7-14. Composite Back Trajectory Map for BRCO
Figure 7-15. Back Trajectory Cluster Map for BRCO
7-22
-------
Figure 7-16. Composite Back Trajectory Map for PACO
Figure 7-17. Back Trajectory Cluster Map for PACO
7-23
-------
Figure 7-18. Composite Back Trajectory Map for RICO
Figure 7-19. Back Trajectory Cluster Map for RICO
7-24
-------
Figure 7-20. Composite Back Trajectory Map for RFCO
Observations from Figures 7-12 through 7-20 for the Garfield County sites include the
following:
• The composite back trajectory maps for the Garfield County sites resemble the one
for GPCO. This is expected, given the sites' relatively close proximity to GPCO (and
to each other). Even the composite map for RFCO has a similar back trajectory
distribution as the other sites, even though the number of back trajectories in
Figure 7-20 for RFCO is less than half the back trajectories compared to the other
Garfield County sites. This is due to a combination of late start to sampling (June)
and a l-in-12 day sampling schedule.
• The 24-hour air shed domains for the Garfield County sites were similar in size to
GPCO, with the average trajectory length ranging from 130 miles (RFCO) to
139 miles (RICO). The longest back trajectories for these sites originated over Idaho.
The longest back trajectory for each site except RFCO represents the back trajectory
constructed for midday January 22, 2012.
• The cluster maps for the Garfield County sites resemble the cluster map for GPCO, in
that most of the back trajectories have a southwesterly or northwesterly component,
although the exact clusters constructed and the associated percentages vary. The
HYSPLIT model grouped the back trajectories for BMCO into five clusters but
grouped the back trajectories for RICO into just three clusters. However, common
elements of the cluster analyses include: 1) between 30 percent and 40 percent of
back trajectories originated with approximately 100 miles of the sites and are
represented by the short cluster trajectory originating towards the Colorado/Utah
7-25
-------
border, 2) approximately one-third of back trajectories originated to the southwest of
the sites, and 3) between 10 percent and 20 percent of back trajectories originated to
the northwest of the sites.
• Because RFCO has fewer than 30 sample days, a cluster analysis was not performed
for this site, as specified in Section 3.5.2.1.
7.2.4 Wind Rose Comparison
Hourly surface wind data from the weather stations at the Walker Field Airport (for
GPCO), Garfield County Regional Airport (for BMCO, BRCO, PACO, and RICO), and Pitkin-
Aspen County Airport (for RFCO) were uploaded into a wind rose software program to produce
customized wind roses, as described in Section 3.5.2.2. 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 7-21 presents a map showing the distance between the weather station and GPCO,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 7-21 also presents three different wind roses for the
GPCO monitoring site. First, a historical wind rose representing 2002 to 2011 wind data is
presented, which shows the predominant surface wind speed and direction over an extended
period of time. Second, a wind rose representing wind observations for all of 2012 is presented.
Next, a wind rose representing wind observations for days on which samples were collected in
2012 is presented. These can be used to identify the predominant wind speed and direction for
2012 and to determine if wind observations on sample days were representative of conditions
experienced over the entire year and historically. Figures 7-22 through 7-26 present the distance
maps and wind roses for the five Garfield County sites.
7-26
-------
Figure 7-21. Wind Roses for the Walker Field Airport Weather Station near GPCO
Location of GPCO and Weather Station
2002-2011 Historical Wind Rose
Calms: 16.49%
2012 Wind Rose
Sample Day Wind Rose
WIND SPEED
(Knots)
n »22
^| 11 - 17
n I."
HI 2- 4
Calms: 16.77%
Calms: 1939%
7-27
-------
Figure 7-22. Wind Roses for the Garfield County Regional Airport Weather Station near
BMCO
Location of BMCO and Weather Station
2002-2011 Historical Wind Rose
N
+
2012 Wind Rose
WIND SPEED
(Knots)
^| 17 - 21
^| 11 . 17
O «-7
HI -
Calms: 37.94%
Sample Day Wind Rose
Calms: 37.74%
7-28
-------
Figure 7-23. Wind Roses for the Garfield County Regional Airport Weather Station near
BRCO
Location of BRCO and Weather Station
2002-2011 Historical Wind Rose
2012 Wind Rose
WIND SPEED
(Knots)
^| 17 - 21
^| 11 . 17
O «-7
HI -
Calms: 37.94%
Sample Day Wind Rose
Calms: 3e.91%
7-29
-------
Figure 7-24. Wind Roses for the Garfield County Regional Airport Weather Station near
PACO
Location of PACO and Weather Station
2002-2011 Historical Wind Rose
2012 Wind Rose
Sample Day Wind Rose
Calms: 3650%
7-30
-------
Figure 7-25. Wind Roses for the Garfield County Regional Airport Weather Station near
RICO
Location of RICO and Weather Station
2002-2011 Historical Wind Rose
N
+
2012 Wind Rose
Sample Day Wind Rose
WIND SPEED
(Knots)
^| 17 - 21
^| 11 . 17
O «-7
HI -
Calms: 37.94%
7-31
-------
Figure 7-26. Wind Roses for the Aspen-Pitkin County Airport Weather Station near RFCO
Location of RFCO and Weather Station
2002-2011 Historical Wind Rose
Calms: 18.57%
2012 Wind Rose
Sample Day Wind Rose
7-32
-------
Observations from Figure 7-21 for GPCO include the following:
• The Walker Field Airport weather station is located approximately 5 miles north-
northeast of GPCO. Most of the city of Grand Junction lies between the site and the
airport. The airport property where the weather station is located is adjacent to where
the elevation begins to increase on the north side of the city.
• The historical wind rose shows that easterly, east-southeasterly, and southeasterly
winds were prevalent near GPCO. Winds from the west to northwest make up a
secondary wind grouping. Winds from the southwest quadrant and north-northeast to
northeast directions were rarely observed. Calm winds (< 2 knots) were observed for
approximately 15 percent of the hourly wind measurements.
• The 2012 wind rose exhibits similar wind patterns as the historical wind rose. Further,
the sample day wind patterns also resemble the historical and full-year wind patterns,
indicating that wind conditions on sample days were representative of those
experienced over the entire year and historically.
Observations from Figures 7-22 through 7-25 for BMCO, BRCO, PACO, and RICO
include the following:
• The weather station at Garfield County Regional Airport is the closest weather station
to four of the five monitoring sites in Garfield County. The weather station is located
east of Rifle, just south of 1-70. The distance from the weather station to the sites
varies from about 3 miles (RICO) to greater than 17 miles (PACO).
• The historical and 2012 wind roses for these Garfield County sites are identical to
each other because the wind observations come from the same weather station for all
four sites.
• The historical wind roses show that calm winds were prevalent near the monitoring
sites, representing one third of wind observations. Westerly and southerly winds were
also common. Winds from the northeast quadrant were rarely observed.
• Calm winds were observed for 38 percent of the wind observations in 2012. Fewer
southerly and south-southwesterly winds and more easterly winds were observed in
2012 near the Garfield County sites compared to the historical wind rose. A similar
observation was made in the 2011 NMP report.
• The sample day wind patterns for each site resemble the full-year wind patterns. This
resemblance indicates that conditions on sample days were representative of those
experienced over the entire year.
Observations from Figure 7-26 for RFCO include the following:
• The Aspen-Pitkin County Airport weather station is located approximately 22 miles
southeast of RFCO. The mountainous terrain surrounding the site and weather station
is visible in Figure 7-26.
7-33
-------
• The historical wind rose shows that winds from the south and south-southwest are
prevalent near RFCO, accounting for one third of the wind observations from this
weather station. Winds from the north-northwest and north make up another roughly
20 percent of wind observations, as do calm winds. Winds from due east and due
west were not observed.
• The 2012 wind rose exhibits similar wind patterns as the historical wind rose,
indicating that conditions in 2012 were similar to conditions experienced over the last
10 years.
• The sample day wind rose has a higher percentage of northerly winds and a lower
percentage of southerly winds than the historical and 2012 wind rose. The differences
in the sample day wind rose may be indicative of a seasonal pattern, as this wind rose
includes data from June through December only.
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 7-4. 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 7-4. It is important to note which pollutants were sampled for at each site when reviewing
the results of this analysis. GPCO sampled for VOCs, carbonyl compounds, PAHs, and
hexavalent chromium; the Garfield County sites sampled for SNMOCs and carbonyl compounds
only.
7-34
-------
Table 7-4. Risk-Based Screening Results for the Colorado Monitoring Sites
Pollutant
Screening
Value
(ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Grand Junction, Colorado - GPCO
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Acetaldehyde
Formaldehyde
Naphthalene
1,2-Dichloroethane
Ethylbenzene
Acenaphthene
Fluorene
£>-Dichlorobenzene
Dichloromethane
Hexachloro- 1 ,3 -butadiene
1 , 1 ,2,2-Tetrachloroethane
Benzo(a)pyrene
1 ,2-Dibromoethane
Trichloroethylene
Bromomethane
Xylenes
0.13
0.03
0.17
0.45
0.077
0.029
0.038
0.4
0.011
0.011
0.091
7.7
0.045
0.017
0.00057
0.0017
0.2
0.5
10
Total
62
62
62
61
61
60
56
49
32
27
17
15
10
7
6
6
5
1
1
600
62
62
62
61
61
60
56
62
60
60
54
62
10
7
40
6
25
57
62
929
100.00
100.00
100.00
100.00
100.00
100.00
100.00
79.03
53.33
45.00
31.48
24.19
100.00
100.00
15.00
100.00
20.00
1.75
1.61
64.59
10.33
10.33
10.33
10.17
10.17
10.00
9.33
8.17
5.33
4.50
2.83
2.50
1.67
1.17
1.00
1.00
0.83
0.17
0.17
10.33
20.67
31.00
41.17
51.33
61.33
70.67
78.83
84.17
88.67
91.50
94.00
95.67
96.83
97.83
98.83
99.67
99.83
100.00
Battlement Mesa, Colorado - BMCO
Benzene
1,3 -Butadiene
Formaldehyde
Acetaldehyde
Ethylbenzene
0.13
0.03
0.077
0.45
0.4
Total
53
28
24
15
1
121
53
28
26
26
53
186
100.00
100.00
92.31
57.69
1.89
65.05
43.80
23.14
19.83
12.40
0.83
43.80
66.94
86.78
99.17
100.00
Silt, Colorado - BRCO
Benzene
Formaldehyde
1,3 -Butadiene
Acetaldehyde
0.13
0.077
0.03
0.45
Total
57
28
22
20
127
57
28
23
28
136
100.00
100.00
95.65
71.43
93.38
44.88
22.05
17.32
15.75
44.88
66.93
84.25
100.00
Parachute, Colorado - PACO
Benzene
1,3 -Butadiene
Formaldehyde
Acetaldehyde
Ethylbenzene
0.13
0.03
0.077
0.45
0.4
Total
43
32
26
19
1
121
43
32
27
27
43
172
100.00
100.00
96.30
70.37
2.33
70.35
35.54
26.45
21.49
15.70
0.83
35.54
61.98
83.47
99.17
100.00
7-35
-------
Table 7-4. Risk-Based Screening Results for the Colorado Monitoring Sites (Continued)
Pollutant
Screening
Value
(ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Carbondale, Colorado - RFCO
Benzene
Formaldehyde
1,3 -Butadiene
Acetaldehyde
0.13
0.077
0.03
0.45
Total
16
15
12
6
49
16
15
12
15
58
100.00
100.00
100.00
40.00
84.48
32.65
30.61
24.49
12.24
32.65
63.27
87.76
100.00
Rifle, Colorado - RICO
Benzene
1,3 -Butadiene
Formaldehyde
Acetaldehyde
Ethylbenzene
0.13
0.03
0.077
0.45
0.4
Total
60
56
28
24
5
173
60
56
28
28
60
232
100.00
100.00
100.00
85.71
8.33
74.57
34.68
32.37
16.18
13.87
2.89
34.68
67.05
83.24
97.11
100.00
Observations from Table 7-4 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 each site.
• Nineteen pollutants failed at least one screen for GPCO; nearly 65 percent of the
concentrations for these 19 pollutants were greater than their associated risk screening
value (or failed screens).
• Thirteen pollutants contributed to 95 percent of failed screens for GPCO and
therefore were identified as pollutants of interest for GPCO. These 13 include two
carbonyl compounds, eight VOCs, and three PAHs.
• GPCO failed the fourth highest number of screens (600) among all NMP sites, behind
only S4MO, PXSS, and TOOK (refer to Table 4-8 of Section 4.2). However, the
failure rate for GPCO, when incorporating all pollutants with screening values, is
relatively low (less than 29 percent). This is due primarily to the relatively high
number of pollutants sampled for at this site, as discussed in Section 4.2.
• The number of pollutants failing screens for the Garfield County sites range from four
to five. Four pollutants (benzene, 1,3-butadiene, formaldehyde, and acetaldehyde)
failed screens for each Garfield County site. These four pollutants were identified as
pollutants of interest for all five sites. Ethylbenzene also failed screens for three of the
five Garfield County sites (BRCO and RFCO being the exceptions), but was not
identified as a pollutant of interest for any of them.
• Benzene is the only pollutant to fail 100 percent of screens for all six Colorado sites.
7-36
-------
• Note that carbonyl compounds were collected on a l-in-12 day sampling schedule at
BMCO, BRCO, PACO, and RICO, while SNMOCs were collected on a l-in-6 day
sampling schedule; thus, the number of carbonyl compound samples collected at
these sites were often less than half the number of SNMOC samples. Both carbonyl
compounds and SNMOCs were collected on a l-in-12 day sampling schedule at
RFCO, although sampling did not begin at RFCO until June.
7.4 Concentrations
This section presents various concentration averages used to characterize pollution 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.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Additional site-specific statistical summaries for the
Colorado monitoring site are provided in Appendices J through M and O.
7.4.1 2012 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 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 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
Colorado monitoring sites are presented in Table 7-5, where applicable. Note that concentrations
of the PAHs for GPCO are presented in ng/m3 for ease of viewing. Also note that if a pollutant
7-37
-------
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 7-5. Quarterly and Annual Average Concentrations of the Pollutants of
Interest for the Colorado Monitoring Sites
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(jig/m3)
2nd
Quarter
Average
(jig/m3)
3rd
Quarter
Average
(jig/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Ug/m3)
Grand Junction, Colorado - GPCO
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
£>-Dichlorobenzene
Dichloromethane
Ethylbenzene
Formaldehyde
Hexachloro- 1 ,3 -butadiene
Acenaphthene3
Fluorene3
Naphthalene3
61/61
62/62
62/62
62/62
56/62
54/62
62/62
62/62
61/61
10/62
60/60
60/60
60/60
2.96
±0.35
1.54
±0.24
0.18
±0.04
0.65
±0.06
0.09
±0.01
0.08
±0.01
0.68
±0.30
0.58
±0.11
2.63
±0.23
0.04
±0.03
28.18
±28.14
13.03
± 10.01
240.71
±113.20
3.75
±0.57
1.02
±0.22
0.10
±0.04
0.66
±0.03
0.09
±0.01
0.10
±0.03
0.51
±0.23
0.55
±0.13
2.49
±0.35
0.01
±0.02
26.54
±8.74
17.63
±5.01
201.08
± 47.87
2.39
±0.44
NA
NA
NA
NA
NA
NA
NA
3.15
±0.29
NA
18.42
±2.54
12.68
± 1.67
148.69
± 19.64
2.49
±0.55
1.40
±0.20
0.26
±0.05
0.68
±0.03
0.06
±0.02
0.04
±0.01
104.13
±75.41
0.77
±0.17
3.81
±0.78
0
8.32
±3.84
6.49
± 1.58
230.09
±79.34
2.89
±0.27
1.28
±0.12
0.18
±0.03
0.67
±0.02
0.08
±0.01
0.07
±0.01
40.23
± 28.78
0.70
±0.11
3.02
±0.25
0.02
±0.01
20.53
±7.27
12.56
±2.86
203.78
±35.24
Battlement Mesa, Colorado - BMCO
Acetaldehyde
Benzene
1,3 -Butadiene
Formaldehyde
26/26
53/53
28/53
26/26
0.62
±0.16
1.21
±0.29
0.05
±0.03
1.05
±0.16
NA
0.90
±0.27
0.01
±0.02
NA
0.45
±0.49
1.21
±0.19
0.13
±0.06
0.82
±0.71
0.44
±0.10
1.04
±0.17
0.12
±0.09
0.73
±0.12
NA
1.09
±0.12
0.08
±0.03
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.
7-38
-------
Table 7-5. Quarterly and Annual Average Concentrations of the Pollutants of
Interest for the Colorado Monitoring Sites (Continued)
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Hg/m3)
2nd
Quarter
Average
(Hg/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Hg/m3)
Silt, Colorado - BRCO
Acetaldehyde
Benzene
1,3 -Butadiene
Formaldehyde
28/28
57/58
23/58
28/28
0.49
±0.14
0.81
±0.16
0.04
±0.03
0.8
±0.14
0.77
±0.17
0.44
±0.11
0.01
±0.01
1.07
±0.18
0.74
±0.39
0.68
±0.12
0.09
±0.06
1.50
±0.73
0.49
±0.11
0.72
±0.22
0.12
±0.07
0.77
±0.20
0.61
±0.11
0.67
±0.08
0.06
±0.03
1.02
±0.21
Parachute, Colorado - PACO
Acetaldehyde
Benzene
1,3 -Butadiene
Formaldehyde
21121
43/45
32/45
21121
0.79
±0.37
1.43
±0.35
0.10
±0.04
1.32
±0.32
0.79
±0.30
1.10
±0.32
0.03
±0.02
1.31
±0.59
0.68
±0.47
NA
NA
1.22
±0.75
0.54
±0.20
NA
NA
1.03
±0.41
0.69
±0.15
NA
NA
1.20
±0.24
Carbondale, Colorado - RFCO
Acetaldehyde
Benzene
1,3 -Butadiene
Formaldehyde
15/15
16/17
12/17
15/15
NA
NA
NA
NA
NA
NA
NA
NA
0.72
±0.60
0.59
±0.24
0.11
±0.09
1.08
±0.73
0.47
±0.27
0.36
±0.09
0.18
±0.08
0.65
±0.25
NA
NA
NA
NA
Rifle, Colorado - RICO
Acetaldehyde
Benzene
1,3 -Butadiene
Formaldehyde
28/28
60/60
56/60
28/28
1.06
±0.35
1.18
±0.34
0.21
±0.07
1.42
±0.37
NA
0.77
±0.17
0.10
±0.02
NA
1.08
±0.49
0.93
±0.11
0.17
±0.05
1.55
±0.64
0.75
±0.32
1.08
±0.24
0.24
±0.06
1.06
±0.32
1.04
±0.19
1.00
±0.12
0.18
±0.03
1.39
±0.22
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 7-5 include the following:
• The pollutants with the highest annual average concentrations are dichloromethane
(40.23 ± 28.78 |ig/m3), formaldehyde (3.02 ± 0.25 |ig/m3), acetaldehyde (2.89 ± 0.27
|ig/m3), and benzene (1.28 ± 0.12 |ig/m3). These are also the only pollutants with
annual average concentrations greater than 1 |ig/m3.
7-39
-------
• The annual average concentration of dichloromethane for GPCO is significantly
higher than annual average concentrations for the other pollutants of interest and has
a relatively large confidence interval associated with it. A review of the quarterly
averages shows that concentrations measured during the fourth quarter of 2012 may
be driving the annual average. (Note that third quarter averages for the VOCs could
not be calculated due several invalidated samples during the month of August). A
review of the preprocessed daily measurements shows that the highest concentrations
of dichloromethane were measured at GPCO between late September and mid
November. Fifteen concentrations greater than 25 |ig/m3 were measured at GPCO
during this time frame and ranged from 29.5 |ig/m3to 745 |ig/m3. Measurements
collected at GPCO account for seven of the eight concentrations of dichloromethane
greater than 100 |ig/m3 and 15 of the 19 concentrations greater than 25 |ig/m3 among
all NMP sites sampling VOCs (with BTUT accounting for the other four).
• The fourth quarter formaldehyde concentration is higher than the other quarterly
averages and has a relatively large confidence interval associated with it. A review of
the data shows that the three highest concentrations of formaldehyde were measured
at GPCO during the last three scheduled sample days of December, ranging from
4.59 |ig/m3 on December 30th to 8.33 |ig/m3 on December 17th. The highest
formaldehyde concentrations measured at GPCO were collected during the second
half of 2012. Of the 25 concentrations greater than 3 |ig/m3 measured at GPCO, only
five were measured between January and June, with the other 20 measured between
July and December (four in July, three each in August and September, one in
October, five in November, and four in December).
• Of the PAH pollutants of interest, naphthalene has the highest annual average
concentration by an order of magnitude. Each of the PAHs in Table 7-5 has a large
confidence interval associated with its first quarter average concentration.
Naphthalene's fourth quarter average also has a relatively large confidence interval
associated with it. This indicates that outliers are likely influencing these calculations
and each pollutant's measurements are discussed in the bullets that follow.
• A review of the naphthalene data shows that the two highest concentrations of this
pollutant were measured on March 22nd and March 16th at GPCO (822 ng/m3 and
633 ng/m3, respectively). The third and fourth highest concentrations of this pollutant
were measured on November 17th and November 29th (525 ng/m3 and 475 ng/m3,
respectively). These are the four highest naphthalene concentrations measured among
all NMP sites sampling PAHs. GPCO has the highest number of naphthalene
measurements greater than 300 ng/m3 (nine) among all NMP sites. These nine
concentrations are split evenly among the first, second, and fourth quarters of 2012.
GPCO also had some of the highest measurements of naphthalene in 2011.
• The maximum concentration of fluorene was measured at GPCO on the same day as
the maximum concentration of naphthalene (68.2 ng/m3 on March 22nd) and is the
third highest fluorene concentration measured among NMP sites sampling PAHs.
Three of the five highest fluorene concentrations measured at GPCO were measured
in March and ranged from 30.1 ng/m3 to 68.2 ng/m3 (with the other two measured in
April). The next highest measurement collected during the first quarter is
7-40
-------
considerably less, with the remaining concentrations ranging from 1.93 ng/m3 to
7.33 ng/m3. Six of the seven lowest concentrations of fluorene were also measured
during the first quarter of 2012. This variability explains the large confidence interval
calculated for the first quarter of 2012.
• The confidence interval for the first quarter average concentration of acetnaphthene is
almost equivalent to the average itself. The two highest concentrations of
acenaphthene were also measured at GPCO on March 22nd and March 16th (182
ng/m3 and 101 ng/m3), with the third highest measured on March 28th (86.4 ng/m3).
Similar to fluorene, the next highest acenaphthalene concentration measured during
the first quarter is considerably less (14.1 ng/m3) and 12 of the 15 lowest
concentrations of acenaphthalene were measured during the first quarter. This
indicates that the three highest measurements are driving the first quarter average
acenaphthalene concentration. The two acenaphthene concentrations greater than 100
ng/m3 measured at GPCO are the highest concentrations of this pollutant measured
among all NMP sites sampling PAHs. Further, five of the nine acenaphthene
concentrations greater than 50 ng/m3 across the program were measured at GPCO
(with the others measured at DEMI and NBIL).
Observations for the Garfield County sites from Table 7-5 include the following:
• Acetaldehyde, benzene, 1,3-butadiene, and formaldehyde are pollutants of interest for
each Garfield County site.
• Because sampling at RFCO began in June, first quarter, second quarter, and annual
average concentrations could not be calculated. Issues with the sampler used to
collect SNMOC samples at PACO resulted in fewer than three quarterly averages and
a low method completeness; thus, annual averages could not be calculated for this site
for benzene or 1,3-butadiene. Sampler issues at BMCO also resulted in low carbonyl
compound completeness; thus, second quarter and annual average concentrations
could not be calculated for acetaldehyde and formaldehyde for this site. However,
Appendix K and Appendix L provide the pollutant-specific average concentrations
for all valid samples collected over the entire sample period for each site.
• Formaldehyde is the pollutant with the highest annual average concentration among
the pollutants of interest for each of the Garfield County sites (except BMCO, where
an annual average could not be calculated). However, the annual averages of
formaldehyde for these sites, where they could be calculated, are among the lowest
for NMP sites sampling carbonyl compounds, as shown in Figure 4-12b in Section 4.
A similar observation can be made for acetaldehyde.
• Concentrations of acetaldehyde and formaldehyde are highest at RICO, followed by
PACO and BRCO. However, the differences among the annual averages are not
statistically significant, with the exception of RICO's annual average of acetaldehyde.
RICO's annual average is about one-third higher than the other sites' annual
averages.
7-41
-------
• All of the confidence intervals associated with the third quarter average
concentrations of acetaldehyde and formaldehyde for the Garfield County sites are
relatively high compared to the other quarterly averages, particularly for BMCO. A
review of the data shows that the maximum acetaldehyde concentration measured at
all five sites was measured on July 1, 2012. Three of the five sites measured the
maximum formaldehyde concentration on this date too. For BRCO and PACO, the
July 1 formaldehyde concentration was the second highest measured. However, these
concentrations are generally low compared to measurements from other NMP sites.
For example, the maximum formaldehyde concentration measured at a Garfield
County site is 3.16 |ig/m3 (RFCO). Compared to other NMP sites sampling carbonyl
compounds, this measurement ranks 519th.
• Concentrations of benzene were highest at BMCO, followed by RICO and BRCO,
although BRCO's annual average is significantly less than the other two sites.
Concentrations of 1,3-butadiene were highest at RICO, followed by BMCO and
BRCO, although RICO's annual average concentration is significantly higher the
other two sites.
• Among the Garfield County sites, only BRCO has a quarterly average for all four
quarters for all four pollutants. The lack of quarterly averages across all sites and all
quarters makes a seasonal trend difficult to determine for these sites.
Tables 4-9 through 4-12 present the 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 GPCO appear in Tables 4-9 through 4-12 a total of
10 times.
• GPCO appears in Table 4-9 for six of the seven VOCs. Its highest ranking is second
for 1,2-dichloroethane. GPCO also ranks fourth for benzene, 1,3-butadiene,
ethylbenzene, and hexachloro-1,3-butadiene. RICO's annual average concentration
for 1,3-butadiene ranks third among NMP sites, just ahead of GPCO. BMCO's and
RICO's annual average benzene concentrations rank sixth and tenth among NMP
sites sampling this pollutant.
• GPCO's annual average acetaldehyde concentration ranks second highest among
NMP sites sampling carbonyl compounds, as shown in Table 4-10. GPCO's annual
average acetaldehyde concentration is between two and five times greater than the
annual averages calculated for the Garfield County sites. GPCO's formaldehyde
concentration does not appear in this table (it ranks 14th).
• GPCO has the highest annual concentration of naphthalene, acenaphthene, and
fluorene among all NMP sites sampling PAHs, as shown in Table 4-11. GPCO also
had the highest annual average concentration of naphthalene in the 2010 and 2011
NMP reports.
7-42
-------
7.4.2 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 each of the pollutants
shaded in gray in Table 7-4 for each site. Note that the box plots for benzene and 1,3-butadiene
were split into separate figures, one for measurements sampled with Method TO-15 (GPCO) and
one for measurements sampled with the SNMOC method (the Garfield County sites), where
annual averages could be calculated. Figures 7-27 through 7-39 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 3.5.3.1.
Figure 7-27. Program vs. Site-Specific Average Acenaphthene Concentration
Program Max Concentration = 182 ng/m3
40 50 60
Concentration (ng/m3)
Program: 1st Quartile 2nd Quartile 3rd Quartile
Site: SiteAverage Site Concentration Range
4th Quartile
Average
o
7-43
-------
Figure 7-28. Program vs. Site-Specific Average Acetaldehyde Concentrations
BRCO ^Q 1
1 1 1 1 1
GPCO 1 1 1
1 0
1 °
PACO >—O—l
RICO 1 Q 1 1
II II
0
Figure 7-2
GPCO
0
3 6 9 12 15 18 21
Concentration {[og/m3
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
• n n n I
Site: Site Average Site Concentration Range
o —
9a. Program vs. Site-Specific Average Benzene (Method TO-15) Concentration
_ ,
I 0
0
123456
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
n n n n I
Site: Site Average Site Concentration Range
o —
7-44
-------
Figure 7-29b. Program vs. Site-Specific Average Benzene (SNMOC) Concentrations
BMCO
RICO
1.5 2
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o —
Figure 7-30a. Program vs. Site-Specific Average 1,3-Butadiene (Method TO-15)
Concentration
Program Max Concentration = 4.10 ug/m3
0.75 1
Concentration {[og/m3)
Program:
Site:
IstQuartile
Site Average
O
2ndQuartile SrdQuartile
Site Concentration Range
4thQuartile Average
7-45
-------
Figure 7-30b. Program vs. Site-Specific Average 1,3-Butadiene (SNMOC) Concentrations
BRCO
0.3
Concentration {pg/m3]
Program
Site:
: IstQuartile
Site Average
o
2ndQuartile SrdQuartile 4thQuartile Av
Site Concentration Range
?rage
Figure 7-31. Program vs. Site-Specific Average Carbon Tetrachloride Concentration
2 3
Concentration {[og/m3)
Program:
Site:
IstQuartile
D
Site Average
0
2ndQuartile SrdQuartile
D D
Site Concentration Range
4thQuartile Average
D 1
Figure 7-32. Program vs. Site-Specific Aver age />-Dichlorobenzene Concentration
0.6 0.8
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site:
Site Average Site Concentration Range
O
7-46
-------
Figure 7-33. Program vs. Site-Specific Average 1,2-Dichloroethane Concentration
GPCO
Program Max Concentration = 17.01 ug/m3
0.1
0.2
0.3
0.4 0.5 0.6
Concentration {[og/m3)
0.7
0.8 0.9
Program:
Site:
IstQuartile
D
Site Average
o
2ndQuartile SrdQuartile
• a
Site Concentration Range
^^^^—
4thQuartile Average
D 1
Figure 7-34. Program vs. Site-Specific Average Dichloromethane Concentration
0
Figure 7-3
10 20 30 40
Concentration {[og/m3)
Prog
iram Max Concentra
tion = 745 jig/m3
50 60
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
D D D D 1
Site: Site Average Site Concentration Range
^
70
5. Program vs. Site-Specific Average Ethylbenzene Concentration
0
0.5 1
1.5
2 2.5
Concentration {[og/m3)
'
3
3.5
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
D D D D 1
Site: Site Average Site Concentration Range
o —
4
7-47
-------
Figure 7-36. Program vs. Site-Specific Average Fluorene Concentration
GPCO
0 10 20 30 40 50 60 70 80
Concentration (ng/m3)
90 100
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: SiteAverage Site Concentration Range
O
Figure 7-37. Program vs. Site-Specific Average Formaldehyde Concentrations
BRCO
GPCO
— H-
II
RICO
Concentration {[og/m3)
10
12
Program:
Site:
IstQuartile
n
SiteAverage
O
2ndQuartile SrdQuartile
• a
Site Concentration Range
^^^^—
4thQuartile Average
D 1
14
7-48
-------
Figure 7-38. Program vs. Site-Specific Average Hexachloro-l,3-Butadiene Concentration
GPCO
0.1 0.15
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site:
Site Average
Site Concentration Range
o
Figure 7-39. Program vs. Site-Specific Average Naphthalene Concentration
400 500
Concentration (ng/m3)
Program:
Site:
IstQuartile
Site Average
O
2ndQuartile SrdQuartile
Site Concentration Range
4thQuartile Average
Observations from Figures 7-27 through 7-39 include the following:
• Figure 7-27 is the box plot for acenaphthene for GPCO. The program-level
maximum concentration (182 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 to
100 ng/m3. The maximum concentration of acenaphthene across the program was
measured at GPCO, as discussed in the previous section, as was the second
highest measurement (101 ng/m3). Note how the first quartile, median, third
quartile, and program-level average concentration are all less than 5 ng/m3. This
provides an indication of just how high these GPCO measurements are compared
to the rest of the data. The annual average acenaphthalene concentration for
GPCO is more than four times the program-level average. The minimum
concentration measured at GPCO is greater than the program-level first quartile
but less than the program-level median.
• Figure 7-28 presents the acetaldehyde box plots for the four Colorado sites for
which annual averages could be calculated. The box plots show that GPCO has
the highest annual average acetaldehyde concentration among the Colorado sites.
The annual average for GPCO is greater than the program-level third quartile;
conversely, most of the annual average concentrations for the Garfield County
sites are less than the program-level first quartile (RICO is the exception; its
7-49
-------
annual average concentration is just greater than the program-level first quartile).
The minimum acetaldehyde concentration measured at GPCO is greater than the
annual average concentrations for all of the Garfield County sites while the
maximum acetaldehyde concentration for each Garfield County site is less than
the program-level average, with the exception of RICO. The maximum
acetaldehyde concentration measured at GPCO is significantly less than the
maximum concentration measured across the program.
• Figures 7-29a and 7-29b present the box plots for benzene. Figure 7-29a
compares to the benzene concentrations measured at GPCO to those measured
across the program for NMP sites sampling VOCs with Method TO-15;
Figure 7-29b presents the annual average benzene concentrations for the Garfield
County sites compared to the benzene concentrations measured across the
program for NMP sites sampling SNMOCs. The box plots are presented this way
to correspond with Tables 4-1 and 4-2 in Section 4.1, as discussed in
Section 3.5.3.1.
• Figure 7-29a shows that the annual average benzene concentration for GPCO is
greater than the program-level average concentration as well as the third quartile
for the program. The minimum benzene concentration measured at GPCO is just
less than the program-level first quartile. The maximum benzene concentration
measured at GPCO is less than half the maximum benzene concentration
measured across the program.
• Figure 7-29b includes a box plot for BMCO, BRCO, and RICO only because
annual averages could not be calculated for PACO and RFCO. The maximum
benzene concentration measured at RICO is the maximum concentration
measured among the eight sites sampling SNMOCs (3.06 jig/m3). Note that the
scale in Figure 7-29b is roughly half the scale for Figure 7-29a. Of the Garfield
County sites shown, BMCO has the highest annual average concentration of
benzene, followed by RICO then BRCO. The annual average concentration for
BMCO is greater than the program-level third quartile; the annual average for
RICO is just less than the program-level third quartile but greater than the
program-level average; and the annual average for BRCO is less than the
program-level average but similar to the program-level median concentration.
• Similar to the box plots for benzene, Figure 7-3Oa presents the 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 7-3Ob presents the 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.
• The program-level maximum concentration (4.10 |ig/m3) is not shown directly on
the box plot in Figure 7-3Oa as the scale has been reduced to 2 |ig/m3 in order to
allow for the observation of data points at the lower end of the concentration
range. GPCO's annual average 1,3-butadiene concentration is greater than the
program-level average concentration and program-level third quartile. The
7-50
-------
minimum 1,3-butadiene concentration measured at GPCO is the same as the
program-level first quartile. Even though the annual average concentration of 1,3-
butadiene for GPCO is among the higher annual averages for this pollutant, the
maximum concentration measured at GPCO (0.596 |ig/m3) is considerably less
than the maximum concentration measured across the program.
The program-level first quartile is zero, and thus, not shown in Figure 7-30b,
indicating that at least 25 percent of the 1,3-butadiene concentrations measured by
sites sampling SNMOCs were non-detects. The maximum 1,3-butadiene
concentration measured at RICO is the maximum concentration measured among
the eight NMP sites sampling SNMOCs (0.571 |ig/m3). Of the Garfield County
sites shown, RICO has the highest annual average concentration of 1,3-butadiene,
followed by BMCO then BRCO. The annual average concentration for BRCO is
less than the program-level median concentration; the annual average for BMCO
is similar to the program-level median concentration; and the annual average for
RICO is greater than the program-level third quartile.
Figure 7-31 is the box plot for carbon tetrachloride for GPCO and shows that the
range of measurements collected is rather small as the difference between the
minimum and maximum concentrations is 0.385 |ig/m3. This box plot also shows
that the annual average carbon tetrachloride concentration for GPCO is just less
than the program-level median and average concentrations.
The program-level first quartile for/>-dichlorobenzene is zero, and thus, not
shown in Figure 7-32, indicating that at least 25 percent of the/>-dichlorobenzene
concentrations measured were non-detects. Eight non-detects were reported for
GPCO. The annual average concentration of this pollutant for GPCO is just
greater than the program-level average and just less than the program-level third
quartile. The maximum />-dichlorobenzene concentration measured at GPCO is
significantly less than the maximum concentration measured across the program.
The program-level maximum concentration (17.01 |ig/m3) is not shown directly
on the box plot for 1,2-dichloroethane in Figure 7-33 as the scale has been
reduced to 1 |ig/m3 in order to allow for the observation of data points at the
lower end of the concentration range. All of GPCO's 1,2-dichloroethane
measurements are less than the program-level average concentration. This, as well
as the magnitude of the maximum concentration at the program-level, indicate
that there are potential outliers in the 1,2-dichloroethane dataset. The annual
average for GPCO is roughly half the program-level average concentration and
just greater than the program-level median.
The program-level maximum concentration (745 |ig/m3) is not shown directly on
the box plot for dichloromethane in Figure 7-34 as the scale has been reduced by
an order of magnitude (70 |ig/m3) in order to allow for the observation of data
points at the lower end of the concentration range. Seven of GPCO's
dichloromethane measurements are greater than the top of the scale in
Figure 7-34. GPCO's annual average concentration of dichloromethane
(40.23 ± 28.78 |ig/m3) is 16 times greater than the program-level average
7-51
-------
concentration. Eighteen individual concentrations measured at GPCO are greater
than the program-level average concentration.
• Figure 7-35 for ethylbenzene shows that GPCO's annual average concentration is
approximately twice the program-level average concentration. While the
maximum ethylbenzene concentration was not measured at this site, GPCO's
maximum ethylbenzene measurement is the second highest among sites sampling
VOCs. The minimum ethylbenzene concentration measured at GPCO is just less
than the program-level median concentration, indicating that nearly 50 percent of
the ethylbenzene concentrations across the program are less than GPCO's
minimum concentration.
• The program-level average concentration of fluorene is just less than the program-
level third quartile and thus, the two cannot be differentiated in Figure 7-36.
GPCO's annual average concentration is more than twice the program-level
concentration. While the maximum fluorene concentration across the program
was not measured at GPCO, GPCO's maximum concentration ranks third among
all fluorene measurements. The minimum fluorene concentration measured at
GPCO is greater than the program-level first quartile. Recall from the previous
section that GPCO has the highest annual average concentration of fluorene
among all NMP sites sampling PAHs.
• Figure 7-37 presents the box plots for formaldehyde. These box plots share some
of the same characteristics as the box plots for acetaldehyde. The box plots show
that GPCO has the highest annual average formaldehyde concentration among the
Colorado sites and is the only site for which the annual average concentration is
greater than the program-level average concentration. The minimum
formaldehyde concentration measured at GPCO is greater than the program-level
first quartile as well as the annual average concentrations for all of the Garfield
County sites shown. The maximum formaldehyde concentration for each Garfield
County site is less than the program-level third quartile.
• Figure 7-38 is the box plot for hexachloro-1,3-butadiene for GPCO. 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 (87 percent). Hexachloro-1,3-butadiene was detected
10 times at GPCO. The maximum concentration of hexachloro-1,3-butadiene
across the program was measured at GPCO (0.203 |ig/m3). The annual average
concentration of hexachloro-1,3-butadiene for GPCO (0.016 ± 0.010 |ig/m3) is
almost twice the program-level average concentration (0.009 jig/m3).
• Figure 7-39 is the box plot for naphthalene and shows that the maximum
concentration of naphthalene across the program was measured at GPCO. The
annual average naphthalene concentration for GPCO (203.78 ±35.24 ng/m3) is
more than twice the program-level average concentration and is greater than the
program-level third quartile. Recall from the previous section that GPCO has the
highest annual average naphthalene concentration among all sites sampling PAHs.
7-52
-------
The minimum concentration of naphthalene measured at GPCO (45.5 ng/m3) is
greater than the program-level first quartile (35.3 ng/m3).
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 3.5.3.2.
GPCO has sampled carbonyl compounds and VOCs under the NMP since 2004 and PAHs since
2008; BRCO, PACO, and RICO began sampling SNMOCs and carbonyl compounds under the
NMP in 2008. Thus, Figures 7-40 through 7-62 present the 1-year statistical metrics for each of
the pollutants of interest first for GPCO then for BRCO, PACO, and RICO. Note, however, that
the 1-year statistical metrics are not provided for the carbonyl compounds for BRCO. This is
because sampling was discontinued in October 2010 and did not begin again until September
2011. Thus, 5 consecutive years of data are not available for BRCO for acetaldehyde and
formaldehyde. 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 is not provided,
although the range and quartiles are still presented. BMCO began sampling SNMOCs and
carbonyl compounds under the NMP at the end of 2010 and RFCO is new for 2012; thus, the
trends analysis was not conducted for these sites.
7-53
-------
Figure 7-40. Yearly Statistical Metrics for Acenaphthene Concentrations Measured at GPCO
~ 100
2010
Year
O 5th Percentile
— Maximum
95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until April 2008.
Observations from Figure 7-40 for acenaphthene measurements collected at GPCO
include the following:
• Sampling for PAHs at GPCO began in April 2008. Because a full year's worth of data
is not available for 2008, a 1-year average is not presented, although the range of
measurements is provided.
• The three highest concentrations of acenaphthene were measured at GPCO in March
2012 and ranged from 86.4 ng/m3 to 182 ng/m3. Although the three highest
concentrations were all measured in March, concentrations measured in 2012 were
higher in general as nine of the 15 concentrations greater than 30 ng/m3 were
measured in 2012 while only one or two were measured in each of the remaining
years of sampling.
• Concentrations of acenaphthene decreased significantly from 2009 to 2010, based on
the 1-year averages, after which a steady increasing trend is shown. Even if the
highest concentrations measured in 2012 were removed from the dataset, the 1-year
average concentration for acenaphthene for 2012 would still represent more than a 50
percent increase from 2011.
7-54
-------
Figure 7-41. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at GPCO
Maximum
Concentration for
2004 is 93.0 u-g/m3.
O,
*
2008
Year
O 5th Percentile — Minimum — Median — Maximum
0 95th Percentile
Observations from Figure 7-41 for acetaldehyde measurements collected at GPCO
include the following:
• The maximum acetaldehyde concentration was measured at GPCO in 2004. The
maximum concentrations measured in subsequent time periods were significantly
lower. The two highest acetaldehyde concentrations (93.0 |ig/m3 and 54.9 |ig/m3)
were both measured in 2004 and the six highest acetaldehyde concentrations (those
greater than 6 |ig/m3) were all measured in 2004 and 2005.
• After the first two years of sampling, the 1-year average concentrations vary by less
than 1 |ig/m3 from year to year. The 1-year average has ranged from 2.00 |ig/m3
(2010) to 2.90 |ig/m3 (2009). The 1-year average and median concentrations are both
at a minimum for 2010, representing a statistically significant decrease from 2009.
The 1-year average concedntration increases from 2010 to 2011 and again for 2012,
back to 2009 levels.
• The 1-year average and median concentrations differ by less than 0.15 |ig/m3 for each
year after 2005, indicating relatively little variability in the central tendency of the
acetaldehyde concentrations measured over the period after 2005.
7-55
-------
Figure 7-42. Yearly Statistical Metrics for Benzene Concentrations Measured at GPCO
O BthPercentile
- Minimum
— Maximum
O 95thPercentile
Observations from Figure 7-42 for benzene measurements collected at GPCO include the
following:
• The maximum benzene concentration (10.6 |ig/m3) was measured on June 8, 2011.
Only three additional 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 from 2004 through 2007, based on
the 1-year averages. After a period of increasing 1-year averages through 2009, a
significant decrease is shown for 2010. Although the decreasing trend continued into
2011, the maximum concentration measured in 2011 results in a higher level of
variability, as indicated by the confidence intervals. The median concentrations
follow a similar pattern as the 1-year averages.
• Even though the range of benzene concentrations is at a minimum for 2012 and the 1-
year average decreased slightly, the median increased from 1.02 |ig/m3to 1.24 |ig/m3
from 2011 to 2012. While the maximum concentration is driving the 1-year average
for 2011, there are more concentrations at the upper end of the concentration range
for 2012, even if that range is more compact. There are also fewer concentrations at
the lower end of the concentration range for 2012; there is only one concentration less
than 0.5 |ig/m3 for 2012 (0.48 |ig/m3) while there are five for 2011.
7-56
-------
Figure 7-43. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at GPCO
~ 0.6
±
o
o
2008
Year
O BthPercentile
- Minimum
— Maximum
95thPercentile
Observations from Figure 7-43 for 1,3-butadiene measurements collected at GPCO
include the following:
• The maximum 1,3-butadiene concentration was measured on December 11, 2004 and
is the only 1,3-butadiene concentration greater than 1 |ig/m3 measured at GPCO. 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 varied by less than 0.065 |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 and median concentrations from 2011 to 2012
represent the largest year to year change (approximately 0.05 |ig/m3 for each). Not
only are the measurements at the upper end of the concentration range higher for
2012, there were also no non-detects reported for 2012, while there were seven
reported for 2011.
• The number of non-detects, and subsequently zeros substituted for non-detects, has
varied significantly across the period of sampling. The number of non-detects
decreased from approximately 30 percent in 2004 and 2005, to 8 percent in 2006, to
none in 2007, 2008, and 2009. The number of non-detects began to increase after
2009, up to 3 percent in 2010 and 12 percent in 2011, after which non-detects were
not reported for 2012.
7-57
-------
Figure 7-44. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured
at GPCO
o
o
f
2008
Year
O 5th Percentile - Minimurr
— Maximum • 95th Percentile
Observations from Figure 7-44 for carbon tetrachloride measurements collected at GPCO
include the following:
• Six concentrations of carbon tetrachloride greater than 1 |ig/m3 have been measured at
GPCO (one in 2006, four in 2008, and one in 2009). Conversely, 15 non-detects have
been measured (nine in 2004, five in 2005, and one in 2006).
• The year with the least variability is 2012, with the difference between the minimum
and maximum concentrations less than 0.40 |ig/m3 and the difference between the 5th
and 95th percentiles less than 0.26 |ig/m3. The year with the highest 1-year average
and median concentrations (0.67 |ig/m3 and 0.68 |ig/m3, respectively) is also 2012.
• 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 sampling range are pulling down the 1-year average.
• Three significant changes in the 1-year average concentrations are shown in
Figure 7-44. There is a significant increase 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. The increase in the 1-year average and median concentrations from
2011 to 2012 is greater than 0.1 |ig/m3 each. Although each of these changes is
statistically significant, the magnitude of the actual changes across the 1-year
averages is less than 0.2 |ig/m3.
7-58
-------
Figure 7-45. Yearly Statistical Metrics for/;-Dichlorobenzene Concentrations Measured at GPCO
O BthPercentile
- Minimum
— Maximum
O 95thPercentile
Observations from Figure 7-45 for/>-dichlorobenzene measurements collected at GPCO
include the following:
• There were no measured detections of />-dichlorobenzene during the first year of
VOC sampling at GPCO. After 2004, the percentage of non-detects decreased to
59 percent for 2005, 39 percent for 2006, and 8 percent for 2007. This corresponds to
a significant increase in the statistical parameters shown in Figure 7-45. However, the
5th percentile is still zero for all years of sampling, indicating the presence non-
detects each year.
• The maximum concentration ofp-dichlorobenzene was measured in 2006
(0.54 |ig/m3). In addition, eight of the 10 highest concentrations of this compound
were measured in 2006, with the other two measured in 2005 and 2007. This is
reflected in the statistical parameters shown for 2006.
• The 1-year average concentration increased from zero to 0.036 |ig/m3 from 2004 to
2005 and more than doubled for 2006. Nearly all of the statistical parameters
decreased from 2006 to 2007 with additional decreases for 2008. While the change in
the 1-year average from 2008 to 2009 is not significant, the decrease shown from
2009 to 2010 represents a 60 percent decrease. Even though the range of
measurements is similar between 2009 and 2010, the number of measured detections
decreased significantly in 2010, as indicated by the median concentration returning to
zero. Thus, the 1-year average is being pulled down by the number of zeros factored
into the calculation for 2010.
7-59
-------
• The number of measured detections increased for 2011 and again for 2012; in
addition, the magnitude of the measurements increased, resulting in an overall
increasing trend for the most recent years of sampling.
Figure 7-46. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations
Measured at GPCO
3 o.io
o
2006 2007 2008 2009
Year
5th Percentile
- Minimum
— Maximum
95th Percentile
••<>•" Average
Observations from Figure 7-46 for 1,2-dichloroethane measurements collected 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
except 2012, 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.
• As the number of measured detections increases, so do each of the corresponding
statistical metrics shown in Figure 7-46.
• As the number of measured detections increased dramatically for 2012, so do the 1-
year average and median concentrations. The median concentration is actually greater
than the 1-year average for 2012. This is because there were still six non-detects (or
zeros) factoring into the 1-year average concentration for the year. Excluding the
non-detects, the minimum concentration would be 0.04 |ig/m3, with a difference
between the minimum and maximum concentration measured for 2012 of less than
0.1 |ig/m3.
7-60
-------
Figure 7-47. Yearly Statistical Metrics for Dichloromethane Concentrations
Measured at GPCO
1
Concentrat
1 1
Maximum
Concentrationfor
2010 is 5,255 U£/m3.
2004 2005 2006 2007 2008
Year
• 5th Percentile - Minimum ~ Median
-, «S*- , i
2009
— Maximum •
<
— (
•
20
r '*•
10 2011
— i
J —
.O
^H 1
2012
95th Percentile ...^... Average
Observations from Figure 7-47 for dichloromethane measurements collected at GPCO
include the following:
• The maximum dichloromethane concentration measured at GPCO (5,256 |ig/m3) is
two orders of magnitude higher than the next highest concentration measured in 2010
(67.9 |ig/m3). This explains why the 1-year average concentration for 2010 is more
than five times greater than the 95th percentile for that year (the 1-year average is
being driven by the outlier).
• The second highest dichloromethane concentration measured at GPCO (745 |ig/m3)
was collected in 2012, as were all six additional measurements greater than
100 |ig/m3 collected at GPCO.
• Higher measurements of dichloromethane were not measured before 2008. The
1-year average dichloromethane concentration was less than 0.5 |ig/m3 for each year
through 2007, after which higher concentrations were measured more often.
However, 2012 is the only year for which concentrations greater than 5 |ig/m3
account for more than 10 percent of the measurements.
• Additional years of sampling are needed in order to determine if higher
dichloromethane measurements continue to be collected.
7-61
-------
Figure 7-48. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at GPCO
I 3.0
o.
••o-l
o
1
o
*
2008
Year
O 5th Percentile
- Minimum
— Maximum
O 95th Percentile
Observations from Figure 7-48 for ethylbenzene measurements collected 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 three of the 15 measurements greater than 2 |ig/m3 (but less than
3 |ig/m3) were measured during these two years.
• The 1-year average concentration increased slightly 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, a decrease
that continues through 2008.
• Although the maximum concentration measured increased 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 is shown from 2010 to 2011 and again for 2012. The median
concentration exhibits a slight increasing trend beginning with 2009 and continuing
through 2012.
7-62
-------
Figure 7-49. Yearly Statistical Metrics for Fluorene Concentrations Measured at GPCO
2010
Year
O 5th Percentile
— Maximum
95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until April 2008.
Observations from Figure 7-49 for fluorene measurements collected at GPCO include the
following:
• Because sampling for PAHs at GPCO began in April 2008, a 1-year average is not
presented for 2008, although the range of measurements is provided.
• The range of measurements collected at GPCO is between 15 ng/m3 and 17 ng/m3 for
each year of sampling until 2012. For 2012, the range of measurements is
significantly higher, with a maximum concentration nearly four times higher than
those measured in previous years.
• The 1-year average concentration decreased significantly from 2009 to 2010. A slight
increase from 2010 to 2011 is followed by a more significant increase for 2012. The
nine highest concentrations measured at GPCO were all collected in 2012 and ranged
from 19.9 ng/m3 to 68.2 ng/m3. Additional years of sampling are needed to determine
if this trend will continue.
7-63
-------
Figure 7-50. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at GPCO
oncentration (ng/m
D C
<
•
Maximum
Concentration for
2004 is 40.5 u.g/m3.
^n rln _L r-S-, JL
-^— — u— _^_ — w— A
- .~. - ^ -vs. rh i n
H-i ifii ^^ "^ -.... |—*-|
, -
L^ UJ ^ k*
• ^^ _._ 1 -o- L-QJ T~^
!_•_! --J--
^ T
^tr i
2004 2005 2006 2007 2008 2009 2010 2011 2012
Year
O 5th Percentile - Minimum ~ Median — Maximum O 95th Percentile ...^>... Average
Observations from Figure 7-50 for formaldehyde measurements collected 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). The
three highest concentrations of formaldehyde were measured on the same days in
2004 and 2005 as the three highest acetaldehyde concentrations.
• 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.
• The maximum concentration measured in 2012 is the highest formaldehyde
measurement collected since 2005. The 95th percentile for 2012 is greater than the
maximum concentration measured in 2011. The 1-year average calculated for 2012 is
slightly higher than the 1-year averages for the previous two years, although the
increase is not statistically significant.
7-64
-------
Figure 7-51. Yearly Statistical Metrics for Hexachloro-l,3-Butadiene Concentrations
Measured at GPCO
1
Concentrat
c
i-
_.-"
-*••
N
^. .~j
T
-O O
\ - *. J^...
2004 2005 2006 2007 2008 2009 2010 2011 2012
Year
• 5th Percentile
- Minimum ~ Median — Maximum O 95th Percentile ...^... Average
Observations from Figure 7-51 for hexachloro-l,3-butadiene measurements collected at
GPCO include the following:
• The number of measured detections for each year is very low, from zero measured
detections in 2004, 2008, and 2009 to 10 (or 17 percent) for 2005. This explains why
the minimum, 5th percentile, and median concentrations (and in some cases, the 1-
year averages) are all zero.
• The maximum hexachloro-1,3-butadiene concentration was measured during 2005
(0.26 |ig/m3), although nine additional measurements greater than 0.20 |ig/m3 have
been measured at GPCO across the years. Not only was the maximum concentration
measured in 2005, this was also the year with the greatest number of measured
detections. This explains the large increase in the 1-year average from 2004 to 2005.
• 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 very large confidence intervals.
• The number of measured detections for 2011 is approximately 13 percent, the highest
percentage since 2005. A similar number of measured detections (15 percent) were
collected in 2012. Additional years of sampling are needed to determine if this trend
continues.
7-65
-------
Figure 7-52. Yearly Statistical Metrics for Naphthalene Concentrations Measured at GPCO
o
2010
Year
O 5th Percentile
— Maximum
O 95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until April 2008.
Observations from Figure 7-52 for naphthalene measurements collected at GPCO include
the following:
• Because sampling for PAHs at GPCO began in April 2008, a 1-year naphthalene
average is not presented for 2008, although the range of measurements is provided.
• The maximum naphthalene concentration measured at GPCO was measured in 2012
(822 ng/m3). Concentrations of approximately 500 ng/m3 or more have been
measured in all years of sampling except 2010.
• Figure 7-52 resembles Figure 7-49 for fluorene. The 1-year average concentration
decreased significantly from 2009 to 2010. A slight increase from 2010 to 2011 is
followed by an additional increase for 2012. Five of the 11 concentrations greater
than 400 ng/m3 measured at GPCO were collected in 2012 and all of the statistical
parameters increased from 2011 to 2012. Additional years of sampling are needed to
determine if this trend continues.
7-66
-------
Figure 7-53. Yearly Statistical Metrics for Benzene Concentrations Measured at BRCO
• BthPercentile
— Minimum
— Maximum
95thPercentile
••<>•" Average
Observations from Figure 7-53 for benzene measurements collected at BRCO include the
following:
• BRCO began sampling benzene under the NMP in January 2008. The maximum
benzene concentration (13.66 |ig/m3) was measured on July 29, 2008 and is three
times higher than the next highest concentration (4.55 |ig/m3, measured on
January 7, 2009), although a similar concentration was also measured on
December 21, 2009 (4.49 |ig/m3).
• The statistical parameters for benzene exhibit a steady decreasing trend over the years
of sampling at BRCO. The 1-year average concentration has 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 has also decreased, from 1.05 |ig/m3 in 2008 to 0.65 |ig/m3
in 2012.
• The difference between the 1-year average and the median concentration has
decreased as well for each year, from a difference of 0.43 |ig/m3 for 2009 to
0.03 |ig/m3 for 2012. This indicates a decreasing variability in the central tendency of
the measurements.
7-67
-------
Figure 7-54. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
BRCO
Concentration (jig/m3)
2008
<
>
2009
O 5th Percentile — Minimum
Median
o
2010
Year
-
o
2011
Maximum
2012
O 95th Percentile ...^... Average
Observations from Figure 7-54 for 1,3-butadiene measurements collected at BRCO
include the following:
• Although the maximum 1,3-butadiene concentration (0.37 |ig/m3) was measured at
BRCO in 2010, the next 11 highest concentrations (those greater than 0.20 |ig/m3)
were all measured in 2012. Of the 32 concentrations greater than 0.05 |ig/m3, none
were measured in 2008, two were measured in 2009, three in 2010, six in 2011, and
21 in 2012.
• The median 1,3-butadiene concentration is zero for all five years of sampling. This
indicates that at least 50 percent of the measurements are zero (or non-detects). In
2008, only three measured detections were reported; for 2009 through 2011, there
were between six and seven measured detections each year; for 2012, 23 measured
detections (out of 58) were reported.
• The increase in the number of detections, particularly for 2012, is reflected in the
1-year average concentrations shown. The 1-year average increased nearly six-fold
from 2011 to 2012.
7-68
-------
Figure 7-55. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
PACO
£ "
1
§
1
I
3 1.0
2010
Year
O 5th Percentile - Minimuir
— Maximum • 95th Percentile
A 1-year average is not presented due to low method completeness in 2011.
Observations from Figure 7-55 for acetaldehyde measurements collected at PACO
include the following:
• PACO began sampling acetaldehyde under the NMP in February 2008. A 1-year
average is not presented for 2011 due to low method completeness.
• The maximum acetaldehyde concentration (2.04 |ig/m3) was measured on
January 13, 2009 and is the only acetaldehyde concentration greater than 2 |ig/m3
measured at this site.
• The 1-year averages shown have a decreasing trend, with the exception of 2011, the
only year for which a 1-year average is not presented. Nearly all of the statistical
parameters shown also have a decreasing trend. For 2011, the maximum, 95th
percentile, and 5th percentile all exhibit decreases, while the median concentration
increased. Even though the range of measurements is at a minimum for 2011, those
concentrations greater than 1 |ig/m3 represent a higher percentage of measurements
for 2011 compared to the previous year.
7-69
-------
Figure 7-56. Yearly Statistical Metrics for Benzene Concentrations Measured at PACO
5th Percentile
— Maximum
95th Percentile
A 1-year average is not presented due to low method completeness in 2012.
Observations from Figure 7-56 for benzene measurements collected at PACO include the
following:
• PACO began sampling SNMOCs under the NMP in January 2008. A 1-year average
is not presented for 2012 due to sampler issues resulting in low method completeness.
• The maximum benzene concentration (11.1 |ig/m3) was measured on October 15,
2008. The next highest measurement (10.1 |ig/m3) was measured three months later
on January 7, 2009. The third highest concentration was measured on the next sample
day in 2009 but was considerably less (7.52 |ig/m3). The 16 highest concentrations
were all measured in either 2008 or 2009.
• Even though the maximum concentration was measured in 2008, benzene
concentrations increased from 2008 to 2009, as indicated by the 1-year average, the
median, and the 95th percentile. However, concentrations of benzene exhibit a
significant decreasing trend between 2009 and 2010. The difference between the 5th
and 95th percentile decreased by half from 2009 to 2010. The decreasing trend
continued into 2011 and 2012, as no benzene concentrations greater than 3 |ig/m3
were measured in 2012. In addition, the maximum, 95th percentile, and median
concentrations are at a minimum for 2012.
• The difference between the 1-year average and median concentrations decreased
significantly from 2009 to 2010, a trend that continued into 2011. This trend indicates
decreasing variability in the central tendency of the measurements.
7-70
-------
Figure 7-57. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at PACO
O 5th Percentile
— Maximum
O 95th Percentile
A 1-year average is not presented due to low method completeness in 2012.
Observations from Figure 7-57 for 1,3-butadiene measurements collected 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 measurement greater than 1 |ig/m3 measured at
this site. The increase in the 1-year average 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, and a
decrease in the 1-year average concentration by almost half from 2008 to 2009.
• The second, third, fourth, and fifth highest 1,3-butadiene concentrations measured at
PACO were all measured in December 2010 and ranged from 0.39 |ig/m3to 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 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.
• With the exception of 2012, the number of non-detects measured at PACO has ranged
from 47 percent (2008) to 58 percent (2009 and 2011). This explains why the median
concentration is at or near zero for most years. For 2012, the number of non-detects is
less (29 percent) and explains why the median is greater than zero.
7-71
-------
• Nearly all of the statistical parameters decreased from 2010 to 2011 (except the
minimum and 5th percentile, which are both years for these years). Most (90 percent)
of the measurements for 2012 fall into the same range as 2011, as indicated by the 5th
and 95th percentiles. While the median increased as a result of fewer non-detects
reported in 2012, no conclusion can be made about the 1-year average. Additional
years of sampling are needed to determine if a viable trend in 1,3-butadiene
concentrations measured at PACO can be identified.
Figure 7-58. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at PACO
2010
Year
O 5th Percentile
- Minimum
— Maximum
95th Percentile
A 1-year average is not presented due to low method completeness in 2011.
Observations from Figure 7-58 for formaldehyde measurements collected 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 did not change between 2008 and 2009. The
decreases in the minimum and maximum concentrations for 2009 are countered by
the increase in the measurements at the higher end of the 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.
7-72
-------
• 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. This year has the greatest number of
measurements less than 1 |ig/m3 (nine). Note that the median concentration is greater
than the 1-year average for 2012. This indicates that the measurements at the lower
end of the concentration range are pulling down the 1-year average. A similar
observation can be made for 2009.
Figure 7-59. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at RICO
2010
Year
O 5th Percentile - Minimum ~ Median — Maximum
95th Percentile
A 1-year average is not presented due to low method completeness in 2010.
A 1-year average is not presented due to low method completeness in 2011.
Observations from Figure 7-59 for acetaldehyde measurements collected at RICO include
the following:
• RICO began sampling acetaldehyde under the NMP in February 2008. A 1-year
average is not presented for 2010 or 2011 due to low method completeness. However,
the range of measurements is provided for both years.
• The maximum acetaldehyde concentration (2.91 |ig/m3) was measured at RICO in
July 2008, although a similar concentration was also measured one month earlier.
• Because few 1-year average concentrations are shown, a distinct trend is hard to
identify. However, the measurements appear to have a decreasing trend, based on the
7-73
-------
decreases shown for nearly all of the other statistical parameters. Additional years of
sampling are needed to confirm if this trend is real particularly because the median
concentration does not exactly follow this trend.
Figure 7-60. Yearly Statistical Metrics for Benzene Concentrations Measured at RICO
5th Percentile - Minimurr
Median — Maximum
95th Percentile
Observations from Figure 7-60 for benzene measurements collected 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.
Seven of the nine benzene concentrations greater than 4 |ig/m3 were measured in
2009 (with the other two in 2008).
• The number of measurements greater than 2 |ig/m3 increased from 18 to 24 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
from 2008 to 2009 as well as the subsequent decreases in the years that follow. The
median concentration is less than 1 |ig/m3 for 2012, indicating that half of the
measurements are less than this concentration. The 1-year average concentration is
also less than 1 |ig/m3 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.
7-74
-------
Figure 7-61. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at RICO
~ 0.6
o
2010
Year
5th Percentile
— Maximum
95th Percentile
Observations from Figure 7-61 for 1,3-butadiene measurements collected 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/m3to 0.98 |ig/m3.
• 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 measurements
in 2010. Even though the 95th percentile more than doubled and the 1-year average
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.
• Although the range of concentrations measured decreased from 2010 to 2011, the
1-year average concentration decreases only slightly while the median concentration
increases. The 1-year average also decreases slightly for 2012 while the median
continues its increase. This is a result of a decreasing maximum concentration paired
with an increasing number of measurements at the mid- to upper-end of the
concentration range, as well as decreasing number of non-detects (and hence zeroes)
paired with an increasing number of measurements at the lower end of the
concentration range.
7-75
-------
Figure 7-62. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at RICO
2010 '
Year
O 5th Percentile — Minimuir
Median — Maximum O 95th Percentile "-O-" Average
A 1-year average is not presented due to low method completeness in 2010.
' A 1 -year average is not presented due to low method completeness in 2011.
Observations from Figure 7-62 for formaldehyde measurements collected at RICO
include the following:
• The maximum formaldehyde concentration (4.82 |ig/m3) was measured at RICO in
November 2008. The next highest concentration was measured in 2011 and is
considerably less (3.40 |ig/m3). Only four concentrations measured at RICO are
greater than 3 |ig/m3 with two measured in 2008 and one each in 2010 and 2011.
• The 1-year average concentrations, where they are presented, appear to have an
overall decreasing trend. However, additional years of sampling are needed to
confirm if this trend is real particularly because the median concentration does not
exactly follow this trend. The median increases from 2009 to 2010 even though the
majority of concentrations fall into a smaller concentration range, as indicated by the
difference between the 5th and 95th percentiles. The minimum concentration
measured for 2010 is greater than the 5th percentile for most of the years of sampling
and 2010 is the only year without a concentration less than 1 |ig/m3.
7-76
-------
7.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk at
each Colorado monitoring site. Refer to Sections 3.3 and 3.5.3 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.
7.5.1 Risk-Based Screening Assessment Using MRLs
A risk-based screening was conducted by comparing the concentration data from the
Colorado monitoring sites to the ATSDR MRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 day to 14 days); intermediate (exposures of 15 days to 364 days); and chronic
(exposures of 1 year or greater). The preprocessed daily measurements of the pollutants of
interest were compared to the acute MRLs; the quarterly averages were compared to the
intermediate MRLs; and the annual averages were compared to the chronic MRLs.
As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2012.
7.5.2 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 3.5.3.4 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 7-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
7-77
-------
Table 7-6. Risk Approximations for the Colorado Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
Grand Junction, Colorado - GPCO
Acenaphthene3
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
/>-Dichlorobenzene
1 ,2-Dichloroethane
Dichloromethane
Ethylbenzene
Fluorene3
Formaldehyde
Hexachloro- 1 , 3 -butadiene
Naphthalene3
0.000088
0.0000022
0.0000078
0.00003
0.000006
0.000011
0.000026
0.00000013
0.0000025
0.000088
0.000013
0.000022
0.000034
0.009
0.03
0.002
0.1
0.8
2.4
0.6
1
0.0098
0.09
0.003
60/60
61/61
62/62
62/62
62/62
54/62
56/62
62/62
62/62
60/60
61/61
10/62
60/60
0.02
±0.01
2.89
±0.27
1.28
±0.12
0.18
±0.03
0.67
±0.02
0.07
±0.01
0.08
±0.01
40.23
± 28.78
0.70
±0.11
0.01
±<0.01
3.02
±0.25
0.02
±0.01
0.20
±0.04
1.81
6.35
10.00
5.42
4.00
0.79
2.06
5.23
1.74
1.11
39.31
0.35
6.93
0.32
0.04
0.09
0.01
<0.01
0.01
0.07
0.01
0.31
O.01
0.07
Battlement Mesa, Colorado - BMCO
Acetaldehyde
Benzene
1,3 -Butadiene
Formaldehyde
0.0000022
0.0000078
0.00003
0.000013
0.009
0.03
0.002
0.0098
26/26
53/53
28/53
26/26
NA
1.09
±0.12
0.08
±0.03
NA
NA
8.50
2.28
NA
NA
0.04
0.04
NA
Silt, Colorado - BRCO
Acetaldehyde
Benzene
1,3 -Butadiene
Formaldehyde
0.0000022
0.0000078
0.00003
0.000013
0.009
0.03
0.002
0.0098
28/28
57/58
23/58
28/28
0.61
±0.11
0.67
±0.08
0.06
±0.03
1.02
±0.21
1.34
5.20
1.91
13.31
0.07
0.02
0.03
0.10
— = A Cancer URE or Noncancer RfC is not available.
NA = Not available due to the criteria for calculating an annual average.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 7-5.
7-78
-------
Table 7-6. Risk Approximations for the Colorado Monitoring Sites (Continued)
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
Parachute, Colorado - PACO
Acetaldehyde
Benzene
1,3 -Butadiene
Formaldehyde
0.0000022
0.0000078
0.00003
0.000013
0.009
0.03
0.002
0.0098
27/27
43/45
32/45
21121
0.69
±0.15
NA
NA
1.20
±0.24
1.52
NA
NA
15.64
0.08
NA
NA
0.12
Carbondale, Colorado - RFCO
Acetaldehyde
Benzene
1,3 -Butadiene
Formaldehyde
0.0000022
0.0000078
0.00003
0.000013
0.009
0.03
0.002
0.0098
15/15
16/17
12/17
15/15
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Rifle, Colorado - RICO
Acetaldehyde
Benzene
1,3 -Butadiene
Formaldehyde
0.0000022
0.0000078
0.00003
0.000013
0.009
0.03
0.002
0.0098
28/28
60/60
56/60
28/28
1.04
±0.19
1.00
±0.12
0.18
±0.03
1.39
±0.22
2.30
7.77
5.42
18.11
0.12
0.03
0.09
0.14
- = A Cancer URE or Noncancer RfC is not available.
NA = Not available due to the criteria for calculating an annual average.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 7-5.
Observations for GPCO from Table 7-6 include the following:
• Dichloromethane, formaldehyde, acetaldehyde, and benzene have the highest annual
average concentrations among GPCO's pollutants of interest.
• Formaldehyde has the highest cancer risk approximation (39.31 in-a-million) for this
site, followed by benzene (10.00 in-a-million), naphthalene (6.93 in-a-million), and
acetaldehyde (6.35 in-a-million).
• None of the pollutants of interest for GPCO have noncancer hazard approximations
greater than 1.0, indicating no adverse health effects are expected from these
individual pollutants. Acetaldehyde and formaldehyde have the highest noncancer
hazard approximations (0.32 and 0.31, respectively) among the pollutants of interest
for GPCO.
7-79
-------
Observations for the Garfield County sites from Table 7-6 include the following:
• Formaldehyde has the highest annual average concentration among the four pollutants
of interest for each Garfield County site, with the exception of BMCO. For BMCO,
benzene has the highest annual average concentration. Recall however, that annual
averages could not be calculated for the carbonyl compounds for BMCO.
• Formaldehyde also has the highest cancer risk approximation for each Garfield
County site, ranging from 13.31 in-a-million (BRCO) to 18.11 in-a-million (RICO),
where a cancer risk approximation could be calculate. All of these are less than half
the cancer risk approximation for formaldehyde for GPCO.
• For BMCO, benzene has the highest cancer risk approximation (8.50 in-a-million).
This is the highest cancer risk approximation for benzene among the Garfield County
sites, where annual averages are available.
• None of the noncancer hazard approximations calculated for the Garfield County sites
are greater than 1.0, indicating no adverse health effects are expected from these
individual pollutants. The highest noncancer hazard approximation was calculated for
formaldehyde for RICO (0.14).
• Annual averages, and therefore cancer risk and noncancer hazard approximations,
could not be calculated for RFCO. This is also true for benzene and 1,3-butadiene for
PACO and acetaldehyde and formaldehyde for BMCO.
7.5.3 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 7-7 presents the 10 pollutants with the highest emissions from the 2011 NEI that have
cancer toxicity factors. Table 7-7 also presents the 10 pollutants with the highest toxicity-
weighted emissions, based on the weighting schema described in Section 3.5.3.5. Lastly,
Table 7-7 provides the 10 pollutants with the highest cancer risk approximations (in-a-million)
for each site, as presented in Table 7-6. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 7-7. Table 7-8 presents similar
information, but is limited to those pollutants with noncancer toxicity factors.
7-80
-------
Table 7-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Colorado Monitoring Sites
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity- Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Grand Junction, Colorado (Mesa County) - GPCO
Benzene
Formaldehyde
Acetaldehyde
Ethylbenzene
1,3 -Butadiene
Naphthalene
Dichloromethane
Tetrachloroethylene
POM, Group 2b
POM, Group 2d
108.00
106.09
39.65
34.52
10.71
5.88
5.44
1.86
1.46
0.92
Formaldehyde
Benzene
1,3 -Butadiene
POM, Group 3
Naphthalene
POM, Group 2b
Acetaldehyde
Ethylbenzene
POM, Group 2d
POM, Group 5a
1.38E-03
8.42E-04
3.21E-04
3.04E-04
2.00E-04
1.28E-04
8.72E-05
8.63E-05
8.13E-05
6.03E-05
Formaldehyde
Benzene
Naphthalene
Acetaldehyde
1,3 -Butadiene
Dichloromethane
Carbon Tetrachloride
1 ,2-Dichloroethane
Acenaphthene
Ethylbenzene
39.31
10.00
6.93
6.35
5.42
5.23
4.00
2.06
1.81
1.74
Battlement Mesa, Colorado (Garfield County) - BMCO
Formaldehyde
Benzene
Acetaldehyde
Ethylbenzene
1,3 -Butadiene
Naphthalene
Tetrachloroethylene
POM, Group 2b
POM, Group 2d
Dichloromethane
385.39
381.10
94.18
42.90
9.16
4.05
1.01
0.67
0.51
0.25
Formaldehyde
Benzene
1,3 -Butadiene
Acetaldehyde
Naphthalene
Ethylbenzene
POM, Group 3
POM, Group 2b
POM, Group 2d
1,2-Dibromoethane
5.01E-03
2.97E-03
2.75E-04
2.07E-04
1.38E-04
1.07E-04
1.06E-04
5.94E-05
4.47E-05
2.63E-05
Benzene
1,3 -Butadiene
8.50
2.28
oo
-------
Table 7-7. 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
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations
Based on Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
Silt, Colorado (Garfield County) - BRCO
Formaldehyde
Benzene
Acetaldehyde
Ethylbenzene
1,3 -Butadiene
Naphthalene
Tetrachloroethylene
POM, Group 2b
POM, Group 2d
Dichloromethane
385.39
381.10
94.18
42.90
9.16
4.05
1.01
0.67
0.51
0.25
Formaldehyde
Benzene
1,3 -Butadiene
Acetaldehyde
Naphthalene
Ethylbenzene
POM, Group 3
POM, Group 2b
POM, Group 2d
1 ,2-Dibromoethane
5.01E-03
2.97E-03
2.75E-04
2.07E-04
1.38E-04
1.07E-04
1.06E-04
5.94E-05
4.47E-05
2.63E-05
Formaldehyde 13.31
Benzene 5.20
1,3 -Butadiene 1.91
Acetaldehyde 1.34
Parachute, Colorado (Garfield County) - PACO
Formaldehyde
Benzene
Acetaldehyde
Ethylbenzene
1,3 -Butadiene
Naphthalene
Tetrachloroethylene
POM, Group 2b
POM, Group 2d
Dichloromethane
385.39
381.10
94.18
42.90
9.16
4.05
1.01
0.67
0.51
0.25
Formaldehyde
Benzene
1,3 -Butadiene
Acetaldehyde
Naphthalene
Ethylbenzene
POM, Group 3
POM, Group 2b
POM, Group 2d
1 ,2-Dibromoethane
5.01E-03
2.97E-03
2.75E-04
2.07E-04
1.38E-04
1.07E-04
1.06E-04
5.94E-05
4.47E-05
2.63E-05
Formaldehyde 15.64
Acetaldehyde 1.52
oo
to
-------
Table 7-7. 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
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations
Based on Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
Carbondale, Colorado (Garfield County) - RFCO
Formaldehyde
Benzene
Acetaldehyde
Ethylbenzene
1,3 -Butadiene
Naphthalene
Tetrachloroethylene
POM, Group 2b
POM, Group 2d
Dichloromethane
385.39
381.10
94.18
42.90
9.16
4.05
1.01
0.67
0.51
0.25
Formaldehyde
Benzene
1,3 -Butadiene
Acetaldehyde
Naphthalene
Ethylbenzene
POM, Group 3
POM, Group 2b
POM, Group 2d
1 ,2-Dibromoethane
5.01E-03
2.97E-03
2.75E-04
2.07E-04
1.38E-04
1.07E-04
1.06E-04
5.94E-05
4.47E-05
2.63E-05
Rifle, Colorado (Garfield County) - RICO
Formaldehyde
Benzene
Acetaldehyde
Ethylbenzene
1,3 -Butadiene
Naphthalene
Tetrachloroethylene
POM, Group 2b
POM, Group 2d
Dichloromethane
385.39
381.10
94.18
42.90
9.16
4.05
1.01
0.67
0.51
0.25
Formaldehyde
Benzene
1,3 -Butadiene
Acetaldehyde
Naphthalene
Ethylbenzene
POM, Group 3
POM, Group 2b
POM, Group 2d
1 ,2-Dibromoethane
5.01E-03
2.97E-03
2.75E-04
2.07E-04
1.38E-04
1.07E-04
1.06E-04
5.94E-05
4.47E-05
2.63E-05
Formaldehyde 18.11
Benzene 7.77
1,3-Butadiene 5.42
Acetaldehyde 2.30
oo
-------
Table 7-8. 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
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Grand Junction, Colorado (Mesa County) - GPCO
Toluene
Xylenes
Ethylene glycol
Hexane
Benzene
Formaldehyde
Methanol
Acetaldehyde
Ethylbenzene
Styrene
407.44
181.20
180.58
112.50
108.00
106.09
102.15
39.65
34.52
12.68
Acrolein
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Benzene
Naphthalene
Xylenes
Antimony, PM
Lead, PM
Ethylene glycol
507,830.37
10,825.37
5,355.90
4,405.19
3,600.06
1,959.57
1,811.96
1,050.00
767.77
451.46
Acetaldehyde
Formaldehyde
1,3 -Butadiene
Naphthalene
Dichloromethane
Benzene
Carbon Tetrachloride
Ethylbenzene
Hexachloro- 1 ,3 -butadiene
£>-Dichlorobenzene
0.32
0.31
0.09
0.07
0.07
0.04
0.01
0.01
0.01
0.01
Battlement Mesa, Colorado (Garfield County) - BMCO
Toluene
Methanol
Xylenes
Formaldehyde
Benzene
Hexane
Acetaldehyde
Ethylene glycol
Acrolein
Ethylbenzene
760.37
623.54
550.01
385.39
381.10
147.32
94.18
69.40
68.36
42.90
Acrolein
Formaldehyde
Benzene
Acetaldehyde
Xylenes
1,3 -Butadiene
Naphthalene
Lead, PM
Arsenic, PM
Hexane
3,417,970.87
39,325.75
12,703.42
10,464.10
5,500.07
4,580.22
1,350.07
412.39
227.73
210.45
1,3 -Butadiene
Benzene
0.04
0.04
oo
-------
Table 7-8. 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)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Noncancer
Hazard
Approximation
Pollutant (HQ)
Silt, Colorado (Garfield County) - BRCO
Toluene
Methanol
Xylenes
Formaldehyde
Benzene
Hexane
Acetaldehyde
Ethylene glycol
Acrolein
Ethylbenzene
760.37
623.54
550.01
385.39
381.10
147.32
94.18
69.40
68.36
42.90
Acrolein
Formaldehyde
Benzene
Acetaldehyde
Xylenes
1,3 -Butadiene
Naphthalene
Lead, PM
Arsenic, PM
Hexane
3,417,970.87
39,325.75
12,703.42
10,464.10
5,500.07
4,580.22
1,350.07
412.39
227.73
210.45
Formaldehyde 0.10
Acetaldehyde 0.07
1,3-Butadiene 0.03
Benzene 0.02
Parachute, Colorado (Garfield County) - PACO
Toluene
Methanol
Xylenes
Formaldehyde
Benzene
Hexane
Acetaldehyde
Ethylene glycol
Acrolein
Ethylbenzene
760.37
623.54
550.01
385.39
381.10
147.32
94.18
69.40
68.36
42.90
Acrolein
Formaldehyde
Benzene
Acetaldehyde
Xylenes
1,3 -Butadiene
Naphthalene
Lead, PM
Arsenic, PM
Hexane
3,417,970.87
39,325.75
12,703.42
10,464.10
5,500.07
4,580.22
1,350.07
412.39
227.73
210.45
Formaldehyde 0.12
Acetaldehyde 0.08
oo
-------
Table 7-8. 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)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Noncancer
Hazard
Approximation
Pollutant (HQ)
Carbondale, Colorado (Garfield County) - RFCO
Toluene
Methanol
Xylenes
Formaldehyde
Benzene
Hexane
Acetaldehyde
Ethylene glycol
Acrolein
Ethylbenzene
760.37
623.54
550.01
385.39
381.10
147.32
94.18
69.40
68.36
42.90
Acrolein
Formaldehyde
Benzene
Acetaldehyde
Xylenes
1,3 -Butadiene
Naphthalene
Lead, PM
Arsenic, PM
Hexane
3,417,970.87
39,325.75
12,703.42
10,464.10
5,500.07
4,580.22
1,350.07
412.39
227.73
210.45
Rifle, Colorado (Garfield County) - RICO
Toluene
Methanol
Xylenes
Formaldehyde
Benzene
Hexane
Acetaldehyde
Ethylene glycol
Acrolein
Ethylbenzene
760.37
623.54
550.01
385.39
381.10
147.32
94.18
69.40
68.36
42.90
Acrolein
Formaldehyde
Benzene
Acetaldehyde
Xylenes
1,3 -Butadiene
Naphthalene
Lead, PM
Arsenic, PM
Hexane
3,417,970.87
39,325.75
12,703.42
10,464.10
5,500.07
4,580.22
1,350.07
412.39
227.73
210.45
Formaldehyde 0.14
Acetaldehyde 0.12
1,3-Butadiene 0.09
Benzene 0.03
oo
-------
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 3.5.3.5. Similar to the cancer risk and
noncancer hazard approximations provided in Section 7.5.2, this analysis may help policy-
makers prioritize their air monitoring activities.
Observations from Table 7-7 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 eight pollutants in common among the pollutants
with the highest toxicity-weighted emissions.
• Eight of the highest emitted pollutants also have the highest toxicity-weighted
emissions in Mesa County; the same eight pollutants have the highest emitted
pollutants and highest toxicity-weighted emissions for Garfield County.
• For GPCO, six of the 10 pollutants with the highest cancer risk approximations also
appear on both emissions-based lists for Mesa County. Dichloromethane has the sixth
highest cancer risk approximation and is the seventh highest emitted pollutant in
Mesa County, but does not appear among those with the highest toxicity-weighted
emissions (its ranks 27th). POM, Group 2b is the ninth highest emitted "pollutant" in
Mesa County and ranks sixth for toxicity-weighted emissions. POM, Group 2b
includes several PAHs sampled for at GPCO including acenaphthene, which has the
ninth highest cancer risk approximation for GPCO.
• The four pollutants of interest identified for each of the Garfield County sites appear
on both emissions-based lists in Table 7-7.
Observations from Table 7-8 include the following:
• Toluene is the highest emitted pollutant with a noncancer RfC in both Mesa and
Garfield Counties, although the emissions are higher in Garfield County. These two
counties have an additional eight pollutants in common on their lists of highest
emitted pollutants with noncancer RfCs.
7-87
-------
• 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. Although
acrolein has the highest toxicity-weighted emissions for every county with an NMP
site, rarely does it appear among the highest emitted pollutants. Garfield County is the
only county with an NMP site for which acrolein ranks among the highest emitted. A
similar observation was made in the 2011 NMP report.
• 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 with the highest
toxicity-weighted emissions.
• Formaldehyde, acetaldehyde, and benzene appear on all three lists for GPCO.
Additionally, 1,3-butadiene and naphthalene appear among the pollutants with the
highest noncancer hazard approximations and highest toxicity-weighted emissions,
but are not among the highest emitted pollutants with a noncancer RfC in Mesa
County. Ethylbenzene appears among the pollutants with the highest noncancer
hazard approximations and highest emissions, but is not among those with the highest
toxicity-weighted emissions.
• Formaldehyde and acetaldehyde appear on all three lists for the Garfield County sites
(except RFCO and BMCO, because noncancer hazard approximations could not be
calculated for these sites). This is also true for benzene, where a noncancer hazard
approximation could be calculated.
7.6 Summary of the 2012 Monitoring Data for the Colorado Monitoring Sites
Results from several of the data treatments described in this section include the
following:
»«» Nineteen pollutants failed screens for GPCO. The number of pollutants failing
screens for the Garfield County sites ranged from four to five.
»«» Dichloromethane has highest annual average concentration for GPCO, followed by
formaldehyde, acetaldehyde, and benzene. These were the only pollutants with annual
average concentrations greater than 1 jug/m , although the annual average
concentration for dichloromethane is an order of magnitude greater than the others.
Formaldehyde had the highest annual average concentration for each of the Garfield
County sites, except those for which an annual average could not be calculated.
*»* GPCO has the highest annual average concentrations of naphthalene, acenaphthene,
andfluorene among all NMP sites sampling PAHs.
7-S
-------
Benzene concentrations at GPCO have an overall decreasing trend across the years
of sampling, as do benzene concentrations measured at BRCO and, in more recent
years, RICO. In recent years, concentrations ofp-dichlorobenzene have an
increasing trend at GPCO. The range of concentrations of naphthalene, fluorene, and
acenaphthene measured at GPCO exhibit significant increases for 2012. In addition,
the detection rate of 1,2-dichloroethane at GPCO has been increasing steadily over
the last few years of sampling.
7-89
-------
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. 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 radius are still visible on the map, but have been grayed out in order to show
emissions sources just outside the boundary. Table 8-1 provides supplemental geographical
information such as land use, location setting, and locational coordinates.
3-1
-------
Figure 8-1. Washington, B.C. (WADC) Monitoring Site
oo
-------
Figure 8-2. NEI Point Sources Located Within 10 Miles of WADC
Legend
WADC NATTS site
77"510"W 77 WW 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.
Q 10 mile radius
County boundary
Source Category Group (No. of Facilities)
t Airport/Airline/Airport Support Operations (26) •< Mine/Quarry/Mineral Processing (1)
i Asphalt Production/Hot Mix Asphalt Plant (5) ? Miscellaneous Commercial/Industrial (6)
B Bulk Terminals/Bulk Plants (1) "Q Paint and Coating Manufacturing (1)
* Electricity Generation via Combustion (3) P Printing/Publishing/Paper Product Manufacturing (6)
> Hotels/Motels/Lodging (5) x Rail Yard/Rail Line Operations (1)
o Institution (school, hospital, prison, etc.) (19) » Water Treatment (2)
A Military Base/National Security (11)
8-3
-------
Table 8-1. Geographical Information for the Washington, D.C. Monitoring Site
Site
Code
WADC
AQS Code
11-001-0043
Location
Washington
County
District
Of
Columbia
Micro- or
Metropolitan
Statistical Area
Washington-
Arlington-
Alexandria, DC-
VA-MD-WVMSA
Latitude
and
Longitude
38.921847,
-77.013178
Land Use
Commercial
Location
Setting
Urban/City
Center
Additional Ambient Monitoring Information1
Arsenic, Lead, CO, VOCs, SO2, NOy, NO, NO2,
NOX, PAMS, Carbonyl compounds, O3,
Meteorological parameters, PM10, PM10
Speciation, Black carbon, PM Coarse, PM2 5, PM2 5
Speciation, IMPROVE Speciation.
BOLD ITALICS = EPA-designated NATTS Site
oo
-------
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 near
several heavily traveled roadways. The site is located in a commercial area, and is surrounded by
a hospital, a cemetery, and a university. As Figure 8-2 shows, WADC is surrounded by many
sources in the airport and airport support operations source category and the institution 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 televisions stations. The
institution source category includes hospital, schools, and prisons, etc. The closest sources to
WADC are a wastewater treatment facility, hospitals, and heliports at hospitals.
Table 8-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Washington D.C. monitoring site. Table 8-2 includes both county-
level population and vehicle registration information. Table 8-2 also contains traffic volume
information for WADC, as well as the location for which the traffic volume was obtained.
Additionally, Table 8-2 presents the daily VMT for the District of Columbia.
Table 8-2. Population, Motor Vehicle, and Traffic Information for the Washington, D.C.
Monitoring Site
Site
WADC
Estimated
County
Population1
632,323
County-level
Vehicle
Registration2
316,231
Annual
Average Daily
Traffic3
7,400
Intersection
Used for
Traffic Data
1st Street between W St. and V St.
County-
level Daily
VMT4
9,775,000
1 County-level population estimate reflects 2012 data (Census Bureau, 2013c)
2 County-level vehicle registration reflects 2011 data (FHWA, 2013a)
3 AADT reflects 2010 data (DC DOT, 2012a)
4 County-level VMT reflects 2011 data (DC DOT, 2012b)
BOLD ITALICS = EPA-designaled NATTS Site
Observations from Table 8-2 include the following:
• The District's population is in the middle of the range compared to other counties
with NMP sites. The District-level vehicle registration is also in the middle of the
range compared to other counties with NMP sites.
• The traffic volume experienced near WADC is in the bottom third compared to other
NMP monitoring sites. The traffic volume provided is for 1st Street, the closest
roadway east of the monitoring site, between W Street and V Street, three to four
blocks south of the site.
• The district-level VMT is in the middle-third compared to other county-level VMT,
where VMT is available.
8-5
-------
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 Climate Summary
Located on the Potomac River that divides Virginia and Maryland, the capital
experiences all four seasons, although its weather is somewhat variable. Summers are warm and
often humid, as southerly winds prevail. Summertime temperatures can be accentuated by the
urban heat island effect. Winters are typical of the Mid-Atlantic region, where cool, blustery air
masses are common followed by a fairly quick return to mild temperatures. Winds out of the
northwest are prevalent in the period from December to March. Precipitation is evenly
distributed across the seasons (Wood, 2004).
8.2.2 Meteorological Summary
Hourly meteorological data for 2012 were retrieved from NCDC for the station closest to
the Washington, D.C. monitoring site (NCDC, 2012), as described in Section 3.5.2. The closest
weather station to WADC is located at Ronald Reagan Washington National Airport (WBAN
13743). Additional information about the Reagan National Airport weather station, such as the
distance between the site and the weather station, is provided in Table 8-3. These data were used
to determine how meteorological conditions on sample days vary from conditions experienced
throughout the year.
Table 8-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for all of 2012. Also included in Table 8-3 is the 95 percent
confidence interval for each parameter. As shown in Table 8-3, average meteorological
conditions on sample days were representative of average weather conditions experienced
throughout the year near WADC.
-------
Table 8-3. Average Meteorological Conditions near the Washington, D.C. Monitoring Site
Closest Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Washington, D.C. - WADC
Ronald Reagan
Washington
National Airport
13743
(38.87, -77.03)
5.2
Miles
180°
(S)
Sample
Days
(70)
2012
69.9
±3.9
69.5
±1.7
61.4
±3.7
61.2
± 1.6
46.6
±4.1
46.4
±1.7
53.8
±3.4
53.5
± 1.5
61.2
±3.2
61.4
±1.4
1017.5
±1.6
1017.0
±0.7
7.0
±0.6
6.9
±0.3
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
oo
-------
8.2.3 Back Trajectory Analysis
Figure 8-3 is the composite back trajectory map for days on which samples were
collected at the WADC monitoring site. Included in Figure 8-3 are four back trajectories per
sample day. Figure 8-4 is the corresponding cluster analysis. An in-depth description of these
maps and how they were generated is presented in Section 3.5.2.1. For the composite map, each
line represents the 24-hour trajectory along which a parcel of air traveled toward the monitoring
site on a given sample day and time, based on an initial height of 50 meters AGL. For the cluster
analysis, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric circle around the site in Figures 8-3 and 8-4 represents 100 miles.
Observations from Figures 8-3 and 8-4 include the following:
• Back trajectories originated from a variety of directions at WADC. The longest back
trajectories originated from the northwest. Few back trajectories originated from the
east.
• The 24-hour air shed domain for WADC was comparable in size to many other NMP
monitoring sites. While the farthest away a back trajectory originated was towards
Lake Michigan, or just greater than 550 miles away, the average trajectory length was
203 miles and nearly 90 percent of back trajectories originated within 350 miles of
the site.
• The cluster analysis confirms that back trajectories originated from a variety of
directions of WADC. Back trajectories originating from the northwest account for
20 percent of the back trajectories, but are split into two cluster trajectories based on
back trajectory length. Eleven percent of these back trajectories originated over
western Pennsylvania, while nine percent originated over Lake Huron, Lake Erie,
Toronto, Canada, and western New York. Another 10 percent of back trajectories
originated over Michigan, Ohio, and Indiana. The cluster trajectory originating over
the Blue Ridge Mountains of Virginia (18 percent) represents back trajectories
originating over West Virginia, central and western Virginia, and the western half of
North Carolina. The short cluster trajectory originating just south of the monitoring
site represents the 15 percent of back trajectories originating less than 100 miles away
and over east-central Virginia. Another 18 percent originated to the south over
southeastern Virginia, eastern North Carolina, and the adjacent coastal waters. Ten
percent of back trajectories originated to the northeast to east of WADC, over New
Jersey, the Delmarva Peninsula or farther offshore. Finally, nine percent of back
trajectories originated to the north of WADC, over eastern Pennsylvania and New
York City and the surrounding urban areas.
-------
Figure 8-3. Composite Back Trajectory Map for WADC
Figure 8-4. Back Trajectory Cluster Map for WADC
8-9
-------
8.2.4 Wind Rose Comparison
Hourly surface wind data from the weather station at Ronald Reagan Washington
National Airport were uploaded into a wind rose software program to produce customized wind
roses, as described in Section 3.5.2.2. 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 8-5 presents a map showing the distance between the weather station and WADC,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 8-5 also presents three different wind roses for the
WADC monitoring site. First, a historical wind rose representing 2002 to 2011 wind data is
presented, which shows the predominant surface wind speed and direction over an extended
period of time. Second, a wind rose representing wind observations for all of 2012 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2012 is
presented. These can be used to identify the predominant wind speed and direction for 2012 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.
Observations from Figure 8-5 for WADC include the following:
• The weather station at Reagan National Airport is located approximately 5.2 miles to
the south of WADC. Between WADC and Washington National is the city of
Washington and the Potomac River.
• Historically, southerly to south-south westerly winds account for approximately
25 percent of wind observations near WADC, while northwesterly to northerly winds
account for another 25 percent of observations. Calm winds (< 2 knots) were
observed for less than 10 percent of the hourly measurements.
• The wind patterns on the full-year wind rose are similar to the wind patterns shown
on the historical wind rose. The sample day wind patterns also resemble those on the
historical wind rose, although there are a few differences. Northerly winds accounted
for fewer wind observations on sample days while north-northwesterly winds were
observed more often. Overall, though, the similarities in the three wind roses indicate
that wind patterns in 2012 were similar to what is expected climatologically near this
site.
8-10
-------
Figure 8-5. Wind Roses for the Ronald Reagan Washington National Airport Weather
Station near WADC
Location of WADC and Weather Station
2002-2011 Historical Wind Rose
/
T
an
•-•j ,„ *
Esili^-XiP
a ilt « ,*^ ...„...-
2012 Wind Rose
Sample Day Wind Rose
WIND SPEED
(Knots)
n >==
^| 17-21
^| 11-17
^| 7- 11
Calms: 10.52%
8-11
-------
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 8-4.
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 8-4. It is
important to note which pollutants were sampled for at each site when reviewing the results of
this analysis. WADC sampled for hexavalent chromium and PAHs.
Table 8-4. Risk-Based Screening Results for the Washington, D.C. Monitoring Site
Pollutant
Screening
Value
(Hg/m3)
# of Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
Washington, D.C - WADC
Naphthalene
Benzo(a)pyrene
Fluorene
0.029
0.00057
0.011
Total
61
1
1
63
61
31
61
153
100.00
3.23
1.64
41.18
96.83
1.59
1.59
96.83
98.41
100.00
Observations from Table 8-4 include the following:
• Three pollutants failed screens for WADC. While naphthalene failed 100 percent of
its 61 screens, benzo(a)pyrene and fluorene each failed a single screen.
• Naphthalene accounted for nearly 97 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 pollution 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 to illustrate how the site's
concentrations compare to the program-level averages, as presented in Section 4.1.
8-12
-------
• 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. Additional site-specific statistical summaries for WADC
are provided in Appendices M and O.
8.4.1 2012 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 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 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
WADC are presented in Table 8-5, 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 8-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest
for the Washington, D.C. Monitoring Site
Pollutant
#of
Measured
Detections
vs. # 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
61/61
NA
87.26
± 18.49
86.81
± 16.99
137.46
±64.58
104.38
± 19.17
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
8-13
-------
Observations for WADC from Table 8-5 include the following:
• Naphthalene was detected in every PAH sample collected at WADC. However,
sampler issues experienced in February and March resulted in the invalidation of
several samples and thus, no first quarter average was calculated. Many of these
samples were made up later in the year.
• The second and third quarter average concentrations of naphthalene are fairly similar
to each other in magnitude. The fourth quarter average is higher than the other
quarterly averages and has a relatively large confidence interval associated with it,
indicating that outliers may be present. Two naphthalene concentrations greater than
400 ng/m3 were measured at WADC, one in November (404 ng/m3) and one in
December (473 ng/m3); the next highest concentration measured during the fourth
quarter is considerably less (168 ng/m3). No other naphthalene concentration
measured at WADC was greater than 225 ng/m3.
• The maximum naphthalene concentration measured at WADC is the fifth highest
naphthalene concentration measured across NMP sites sampling PAHs. As shown in
Table 4-11, WADC has the fifth highest annual average concentration of naphthalene
and is one of only five NMP sites with annual average concentrations greater than
100 ng/m3.
8.4.2 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 the site-specific pollutants of
interest, where applicable. Thus, a box plot was created for naphthalene for WADC. Figure 8-6
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 3.5.3.1.
Figure 8-6. Program vs. Site-Specific Average Naphthalene Concentration
WADC
•j
D 100 200 300 400 500 600 700 800 9C
Concentration (ng/m3)
Program:
Site:
1st Quartile
D
Site Average
o
2nd Quartile 3rd Quartile
D D
Site Concentration Range
^^^^—
4th Quartile Average
D 1
8-14
-------
Observations from Figure 8-6 include the following:
• The annual average concentration of naphthalene for WADC is greater than the
program-level average concentration but less than the program-level third
quartile. The annual average concentration of naphthalene for WADC ranks fifth
compared to other NMP sites sampling PAHs. The maximum naphthalene
concentration measured at WADC is less than the program-level maximum
concentration, although it is among the higher measurements across the program.
The minimum concentration measured at WADC is similar to the program-level
first quartile.
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 3.5.3.2.
WADC has sampled PAHs under the NMP since mid-2008. Thus, Figure 8-7 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 is not provided, although the range and quartiles are still presented.
Figure 8-7. Yearly Statistical Metrics for Naphthalene Concentrations Measured at WADC
2010
Year
O 5th Percentile - Minimurr
— Maximum • 95th Percentile ....0... Average
A 1-year average is not presented because sampling under the NMP did not begin until late June 2008.
8-15
-------
Observations from Figure 8-7 for naphthalene measurements collected at WADC include
the following:
• WADC began sampling PAHs under the NMP in late June 2008. Because a full
year's worth of data is not available, a 1-year average is not presented for 2008,
although the range of measurements is provided.
• 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 all years of sampling
except 2008 (which included only half a year's worth of samples).
• The 1-year average concentration exhibits a slight decreasing trend between 2009 and
2011. However, confidence intervals calculated for these averages indicate that the
changes are not statistically significant.
• The difference between the 5th and 95th percentiles is at a minimum for 2012,
excluding 2008, indicating that the majority of concentrations measured are falling
into a tighter range of measurements. Although 2011 and 2012 have the same number
of measurements greater than 100 ng/m3(19), 2012 has none in the 225 ng/m3 to
400 ng/m3 range while 2011 has four in this concentration range. This explains why
the 95th percentile for 2011 is greater than the 95th percentile for 2012. Additionally,
2011 has a greater number of measurements at the lower end of the concentration
range than 2012 (almost twice as many measurements are less than 50 ng/m3 for 2011
compared to 2012). The number of concentrations in the 75 ng/m3 to 100 ng/m3 range
is higher in 2012 than in 2011. As a result, the median concentration is higher for
2012 than 2011.
8.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk at the
WADC monitoring site. Refer to Sections 3.3 and 3.5.3 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.
8.5.1 Risk-Based Screening Assessment Using MRLs
A risk-based screening was conducted by comparing the concentration data from the
Washington D.C. monitoring site to the ATSDR MRLs, where available. As described in
Section 3.3, MRLs are noncancer health risk benchmarks and are defined for three exposure
periods: acute (exposures of 1 day to 14 days); intermediate (exposures of 15 days to 364 days);
and chronic (exposures of 1 year or greater). The preprocessed daily measurements of the
pollutants of interest were compared to the acute MRLs; the quarterly averages were compared
to the intermediate MRLs; and the annual averages were compared to the chronic MRLs.
8-16
-------
As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2012.
8.5.2 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 3.5.3.4 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 8-6, where applicable. Cancer risk approximations are presented as
probabilities while the noncancer hazard approximations are ratios and thus, unitless values.
Table 8-6. Risk Approximations for the Washington, D.C. Monitoring Site
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections vs.
# of Samples
Annual
Average
(ng/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
Washington, D.C. - WADC
Naphthalene
0.000034
0.003
61/61
104.38
±19.17
3.55
0.03
Observations for WADC from Table 8-6 include the following:
• As discussed in Section 8.4.1, the annual average concentration of naphthalene for
WADC is among the higher annual average concentrations compared to other
NMP sites sampling this pollutant.
• The cancer risk approximation for naphthalene is greater than 1.0 in-a-million
(3.55 in-a-million). Its noncancer hazard approximation is significantly less than
1.0, indicating no adverse health effects are expected from this individual
pollutant.
8-17
-------
8.5.3 Risk-Based Emissions Assessment
In addition to the risk-based screenings discussed above, Tables 8-7 and 8-8 present an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 8-7 presents the 10 pollutants with the highest emissions from the 2011 NEI that have
cancer toxicity factors. Table 8-7 also presents the 10 pollutants with the highest toxicity-
weighted emissions, based on the weighting schema described in Section 3.5.3.5. Lastly,
Table 8-7 provides the cancer risk approximation (in-a-million) for the pollutant of interest for
WADC, as presented in Table 8-6. The emissions, toxicity-weighted emissions, and cancer risk
approximations are shown in descending order in Table 8-7. Table 8-8 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 3.5.3.5. Similar to the cancer risk and
noncancer hazard approximations provided in Section 8.5.2, this analysis may help policy-
makers prioritize their air monitoring activities.
8-18
-------
Table 8-7. 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)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
Washington, D.C. - WADC
Benzene
Formaldehyde
Acetaldehyde
Ethylbenzene
Tetrachloroethylene
1,3 -Butadiene
Naphthalene
POM, Group 2b
POM, Group 2d
Dichloromethane
119.25
108.89
61.97
58.43
25.40
19.26
11.14
2.50
1.76
0.81
Formaldehyde
Benzene
1,3 -Butadiene
POM, Group 3
Naphthalene
POM, Group 2b
Nickel, PM
POM, Group 2d
Ethylbenzene
Acetaldehyde
1.42E-03
9.30E-04
5.78E-04
4.99E-04
3.79E-04
2.20E-04
1.55E-04
1.55E-04
1.46E-04
1.36E-04
Naphthalene 3.55
oo
VO
-------
oo
to
o
Table 8-8. 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)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity- Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Noncancer
Hazard
Approximation
Pollutant (HQ)
Washington, B.C. - WADC
Toluene
Ethylene glycol
Methanol
Xylenes
Hexane
Benzene
Formaldehyde
Acetaldehyde
Ethylbenzene
Methyl isobutyl ketone
1,099.24
761.10
352.77
238.17
226.27
119.25
108.89
61.97
58.43
26.85
Acrolein
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Benzene
Naphthalene
Nickel, PM
Chlorine
Xylenes
Ethylene glycol
264,897.44
11,110.93
9,627.53
6,885.06
3,975.06
3,712.17
3,595.22
3,176.67
2,381.66
1,902.74
Naphthalene 0.03
-------
Observations from Table 8-7 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
toxi city-weighted emissions (of the pollutants with cancer UREs).
• Eight of the highest emitted pollutants also have the highest toxi city-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 fifth 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 includes several PAHs sampled for
at WADC but none of these failed any screens. POM, Group 3 does not include any
PAHs sampled for with Method TO-13.
Observations from Table 8-8 include the following:
• Toluene, ethylene glycol, 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.
• Five of the highest emitted pollutants in the District of Columbia also have the
highest toxicity-weighted emissions.
• Naphthalene has the sixth highest toxicity-weighted emissions but is not one of the 10
highest emitted pollutants (of the pollutants with noncancer RfCs).
• None of the other pollutants sampled for at WADC appear in Table 8-8.
8.6 Summary of the 2012 Monitoring Data for WADC
Results from several of the data treatments described in this section include the
following:
»«» Although 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 fifth among NMP
sites sampling this pollutant.
8-21
-------
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 different urban areas. Three sites (AZFL, SKFL,
and SYFL) are located in the Tampa-St. Petersburg-Clearwater, FL MSA. ORFL and PAFL are
located in the Orlando-Kissimmee-Sanford, FL MSA. Figures 9-1 and 9-2 are composite satellite
images retrieved from ArcGIS Explorer showing the St. Petersburg 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. 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 radii are still visible on the map, but have been grayed out in order
to show emissions sources just outside the boundary. 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
to
-------
Figure 9-2. Pinellas Park, Florida (SKFL) Monitoring Site
-------
Figure 9-3. NEI Point Sources Located Within 10 Miles of AZFL and SKFL
Legend
82°50'Q"W 82a45'0"W 82°40-0"W 82°35'Q"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
AZFL UATMP site
SKFL NATTS site
) 10 mile radius
I County boundary
Source Category Group (No. of Facilities)
•i" Aerospace/Aircraft Manufacturing (1)
T Airport/Airline/Airport Support Operations (9)
i Asphalt Production/Hot Mix Asphalt Plant (1)
B Bulk Terminals/Bulk Plants (1)
C Chemical Manufacturing (2)
9) Dry Cleaning (1)
e Electrical Equipment Manufacturing (4)
* Electricity Generation via Combustion (2)
F Food Processing/Agriculture (2)
•#• Industrial Machinery or Equipment Plant (3)
o Institution (school, hospital, prison, etc.) (1)
• Landfill (2)
A Metal Coating, Engraving, and Allied Services to Manufacturers (1)
© Metals Processing/Fabrication (6)
? Miscellaneous Commercial/Industrial (6)
[HI Municipal Waste Combustor (1)
E] Paint and Coating Manufacturing (2)
^— Pharmaceutical Manufacturing (1)
R Plastic, Resin, or Rubber Products Plant (4)
P Printing/Publishing/Paper Product Manufacturing (9)
4i. Ship/Boat Manufacturing or Repair (5)
1 Wastewater Treatment (2)
W Woodwork, Furniture, Millwork & Wood Preserving (1)
9-4
-------
Figure 9-4. Valrico, Florida (SYFL) Monitoring Site
-------
Figure 9-5. NEI Point Sources Located Within 10 Miles of SYFL
Legend
>< SYFL NATTS site
82'20'0"W 62-15'0-W 82'10'CTW 82 5'0'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
JCounty boundary
Source Category Group (No. of Facilities)
>j< Aerospace/Aircraft Manufacturing (1) A
"t Airport/Airline/Airport Support Operations (7) ©
* Asphalt Production/Hot Mix Asphalt Plant (2) x
0 Auto Body Shop/Painters/Automotive Stores (1} ?
8 Automobile/Truck Manufacturing (1) H
Brick, Structural Clay, or Clay Ceramics Plant (1) <
i Compressor Station (1) A
e Electrical Equipment Manufacturing (1) R
Fertilizer Plant (1) P
F Food Processing/Agriculture (3) X
• Landfill (1) W
Metal Coating, Engraving, and Allied Services to Manufacturers (1)
Metals Processing/Fabrication (5)
Mine/Quarry/Mineral Processing (1)
Miscellaneous Commercial/Industrial (3)
Municipal Waste Combustor (1)
Pesticide Manufacturing (1)
Petroleum Refinery (1)
Plastic, Resin, or Rubber Products Plant (1)
Printing/Publishing/Paper Product Manufacturing (2)
Rail Yard/Rail Line Operations (2)
Woodwork, Furniture, Millwork & Wood Preserving (1)
9-6
-------
VO
Figure 9-6. Winter Park, Florida (ORFL) Monitoring Site
TravillioivAve"
llas-Ave
-------
Figure 9-7. Orlando, Florida (PAFL) Monitoring Site
oo
-------
Figure 9-8. NEI Point Sources Located Within 10 Miles of ORFL and PAFL
Legend
>f ORFL UATMP site
81C25'0"W 81320'0"W 8V1510"W 8nt)'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
PAFL U ATM P site O 10 mile radius
County boundary
Source Category Group (No. of Facilities)
"1" Airport/Airline/Airport Support Operations (23)
i Asphalt Production/Hot Mix Asphalt Plant (5)
0 Auto Body Shop/Painters/Automotive Stores (1)
W Automobile/Truck Manufacturing (3)
B Bulk Terminals/Bulk Plants (1)
i Compressor Station (1)
8 Electrical Equipment Manufacturing (2)
f Electricity Generation via Combustion (1)
F Food Processing/Agriculture (5)
-#• Industrial Machinery or Equipment Plant (2)
O Institution (school, hospital, prison, etc.) (6)
• Landfill (1)
(i; Metal Can, Box, and Other Metal Container Manufacturing (1)
A Metal Coating, Engraving, and Allied Services to Manufacturers (2)
© Metals Processing/Fabrication (2)
x Mine/Quarry/Mineral Processing (1)
? Miscellaneous Commercial/Industrial (5)
rj Paint and Coating Manufacturing (4)
R Plastic, Resin, or Rubber Products Plant (1)
P Printing/Publishing/Paper Product Manufacturing (4)
X Rail Yard/Rail Line Operations (2)
ii. Ship/Boat Manufacturing or Repair (1)
9-9
-------
Table 9-1. Geographical Information for the Florida Monitoring Sites
VO
o
J
Site
Code
AZFL
SKFL
SYFL
ORFL
PAFL
AQS Code
12-103-0018
12-103-0026
12-057-3002
12-095-2002
12-095-1004
Location
St.
Petersburg
Pinellas
Park
Valrico
Winter
Park
Orlando
County
Pinellas
Pinellas
Hillsborough
Orange
Orange
Micro- or
Metropolitan
Statistical Area
Tampa-St.
Petersburg-
Clearwater, FL
Tampa-St.
Petersburg-
Clearwater, FL
Tampa-St.
Petersburg-
Clearwater, FL
Orlando-
Kissimmee-
Sanford, FL
Orlando-
Kissimmee-
Sanford, FL
Latitude
and
Longitude
27.785556,
-82.74
27.850348,
-82.714465
27.96565,
-82.2304
28.596389,
-81.3625
28.550833,
-81.345556
Land Use
Residential
Residential
Residential
Commercial
Commercial
Location
Setting
Suburban
Suburban
Rural
Urban/City
Center
Suburban
Additional Ambient Monitoring Information1
NO, NO2, NOX, VOCs, O3, Meteorological
parameters, PM10, PM10 Speciation, PM25.
VOCs, Meteorological parameters, PM10 Speciation,
Black carbon, PM25)PM25 Speciation, IMPROVE
Speciation.
CO, SO2, NOy, NO, NO2, NOX, VOCs, O3,
Meteorological parameters, PM10, PM10 Speciation,
PM2 5, PM2 5 Speciation, PM Coarse, IMPROVE
Speciation.
CO, SO2, NO, NO2, NOX, VOCs, O3, Meteorological
parameters, PM10, PM25.
PM10.
Data for additional pollutants are reported to AQS for these sites (EPA, 2013b); however, these data are not generated by ERG and are therefore not included in this report
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 St. North is a former
electronics manufacturer and is a permanently closed facility (EPA, 2014). Heavily traveled
roadways are located less than 1 mile from the monitoring site. AZFL is located just over 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 on the property of
Skyview Elementary School near 86th Avenue North. Figure 9-2 shows that SKFL is located in a
primarily residential area. However, a railroad intersects the Pinellas Park Ditch near a
construction company in the bottom left corner 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 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; and metals processing and fabrication 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 and an industrial machinery or equipment plant are also located within 2 miles of SKFL.
SYFL is located in Valrico, which is also part of the Tampa-St. Petersburg-Clearwater,
FL MSA, although it is on the eastern outskirts of the area. Unlike the other Florida sites, 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) is a tank that is part of the local water
treatment facility. This site serves as a background site, although the effect of increased
development in the area is likely being captured by the monitoring site. This site is the Tampa
NATTS site.
9-11
-------
Figure 9-5 shows that most of the emissions sources surrounding SYFL are greater than
5 miles away from the site. 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 is the water treatment facility pictured in Figure 9-4. However, this facility is not
shown in Figure 9-5 because they had no reportable air emissions in the 2011 NEI. Besides the
water treatment facility, a food processing facility is the next closest emissions source to SYFL.
ORFL is located in Winter Park, north of Orlando. Figure 9-6 shows that ORFL is
located near Lake Mendsen, east of Lake Killarney and south of Winter Park Village. This site
lies in a commercial area and serves as a population exposure monitor.
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 considered commercial 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 less than 2 miles to the west of the
monitoring site.
Figure 9-8 shows that ORFL is located a few miles north of PAFL. Most of the point
sources are located on the western side of the 10-mile radii. 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.
Table 9-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Florida monitoring sites. Table 9-2 includes both county-level
population and vehicle registration information. Table 9-2 also contains traffic volume
information for each site as well as the location for which the traffic volume was obtained.
Additionally, Table 9-2 presents the county-level daily VMT for Pinellas, Hillsborough, and
Orange Counties.
9-12
-------
Table 9-2. Population, Motor Vehicle, and Traffic Information for the Florida Monitoring
Sites
Site
AZFL
SKFL
SYFL
ORFL
PAFL
Estimated
County
Population1
921,319
1,277,746
1,202,234
County-level
Vehicle
Registration2
872,813
1,143,207
1,073,682
Annual
Average Daily
Traffic3
38,500
49,000
10,400
35,000
49,500
Intersection
Used for
Traffic Data
66th Street N, north of Route 19
Park Blvd, east of 66th Street N
E Dr. Martin Luther King Jr. Blvd, east
of Mclntosh Road
Orlando Avenue, north of Morse Drive
E Colonial Drive, between Primrose
Road & Bumbv Ave.
County-
level
Daily VMT4
21,387,550
34,061,637
34,099,958
Bounty-level population estimates reflect 2012 data (Census Bureau, 2013c)
2County-level vehicle registration reflects 2012 data (FL DHSMV, 2012)
3AADT reflects 2012 data (FL DOT, 2012a)
4County-level VMT reflects 2012 data (FL DOT, 2012b)
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 9-2 include the following:
• Hillsborough County, where SYFL is located, is the most populous of the Florida
counties with monitoring sites, although Orange County also has more than 1 million
people. Pinellas County ranks slightly lower in population as these counties rank
11th, 12th, and 14th in population compared to other counties with NMP sites.
• The vehicle registration counts for two of the three Florida counties are greater than
1 million, with Hillsborough County having the most and Pinellas County having the
least. The vehicle registration rankings for the Florida sites are very similar to the
county population rankings compared to other NMP sites.
• The traffic volume is lowest near SYFL and highest near PAFL, among the Florida
sites, although the traffic volume for SKFL is similar to the traffic volume near
PALF). Traffic volumes for four of the Florida monitoring sites are in the middle of
the range compared to other NMP sites, with traffic near SYFL in the bottom third
compared to other NMP sites.
• VMT is highest for Orange County and lowest for Pinellas County (among the
Florida sites), although the VMTs for Hillsborough County and Orange County are
similar. The Hillsborough, Orange, and Pinellas County VMTs ranked eighth, ninth,
and 14th highest among counties with NMP sites, respectively.
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.
9-13
-------
9.2.1 Climate Summary
The Tampa and Orlando areas experience very mild winters and warm, humid summers.
Temperatures below freezing are infrequent while temperatures greater than 90°F are common
from May to September. Precipitation tends to be concentrated during the summer months, as
afternoon thunderstorms occur almost daily. Semi-permanent high pressure offshore over the
Atlantic Ocean extends westward towards Florida in the winter, resulting in reduced
precipitation amounts. Land and sea breezes affect coastal locations and the proximity to the
Atlantic Ocean or Gulf of Mexico can have a marked affect on the local meteorological
conditions. Florida's orientation and location between the warm waters of the Gulf of Mexico,
the Atlantic Ocean, and Caribbean Sea make it susceptible to tropical systems. However,
Orlando's land-locked location generally makes it less vulnerable than the Tampa/St. Petersburg
area (Wood, 2004; FCC, 2014).
9.2.2 Meteorological Summary
Hourly meteorological data for 2012 were retrieved from NCDC for the weather stations
closest to the Florida monitoring sites (NCDC, 2012), as described in Section 3.5.2. The weather
station closest to the AZFL monitoring site is located at St. Petersburg/Whitted Airport (WBAN
92806); closest to SYFL is at Plant City Municipal Airport (WBAN 92824); closest to SKFL is
at St. Petersburg/Clearwater International Airport (WBAN 12873); and closest to both ORFL
and PAFL is at Orlando Executive Airport (WBAN 12841). Additional information about each
of these weather stations, such as the distance between the sites and the weather stations, is
provided in Table 9-3. These data were used to determine how meteorological conditions on
sample days vary from conditions experienced throughout the year.
9-14
-------
Table 9-3. Average Meteorological Conditions near the Florida Monitoring Sites
Closest Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar
Wind Speed
(kt)
St. Petersburg, Florida - AZFL
St. Petersburg/
Whirled Airport
92806
(27.77, -82.63)
6.9
miles
95°
(E)
Sample
Days
(61)
2012
81.0
± 1.9
80.8
±0.8
74.8
±2.1
74.7
±0.8
66.0
±2.6
66.2
±1.0
69.3
±2.2
69.3
±0.8
75.8
±2.6
76.2
±1.0
1017.0
± 1.2
1016.9
±0.4
7.6
±0.9
7.4
±0.3
Pinellas Park, Florida - SKFL
St Petersburg-
Clearwater Intl.
Airport
12873
(27.91, -82.69)
4.4
miles
13°
(NNE)
Sample
Days
(63)
2012
82.1
±1.9
81.9
±0.8
73.8
±2.0
73.8
±0.8
63.6
±2.6
63.8
±1.0
67.5
±2.1
67.6
±0.9
72.5
±2.4
72.9
±1.0
1017.7
±1.2
1017.4
±0.4
6.6
±0.8
6.6
±0.3
Valrico, Florida - SYFL
Plant City
Municipal Airport
92824
(28.00, -82.16)
4.6
miles
50°
(NE)
Sample
Days
(68)
2012
84.7
±1.8
84.4
±0.8
73.8
±2.1
73.7
±0.9
63.3
±2.8
63.5
±1.1
68.0
±2.3
68.1
±1.0
69.2
±2.3
69.6
±1.0
NA
NA
4.6
±0.6
4.4
±0.2
Winter Park, Florida - ORFL
Orlando Executive
Airport
12841
(28.55, -81.33)
3.9
miles
145°
(SE)
Sample
Days
(61)
2012
83.0
±2.1
82.6
±0.8
73.1
±2.2
73.0
±0.8
61.6
±2.9
61.9
±1.1
66.2
±2.3
66.3
±0.9
70.2
±2.6
70.9
±1.1
1018.1
±1.2
1017.9
±0.5
6.2
±0.7
6.0
±0.3
VO
1 Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
NA= Sea level pressure was not recorded at the Plant City Municipal Airport.
-------
Table 9-3. Average Meteorological Conditions near the Florida Monitoring Sites (Continued)
Closest Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar
Wind Speed
(kt)
Orlando, Florida - PAFL
Orlando Executive
Airport
12841
(28.55, -81.33)
0.8
miles
108°
(ESE)
Sample
Days
(30)
2012
84.0
±2.5
82.6
±0.8
74.1
±2.6
73.0
±0.8
63.3
±3.5
61.9
±1.1
67.5
±2.7
66.3
±0.9
71.9
±3.7
70.9
±1.1
1017.8
± 1.6
1017.9
±0.5
6.2
±0.9
6.0
±0.3
1 Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
NA= Sea level pressure was not recorded at the Plant City Municipal Airport.
-------
Table 9-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for all of 2012. Also included in Table 9-3 is the 95 percent
confidence interval for each parameter. As shown in Table 9-3, average meteorological
conditions on sample days in 2012 at the Florida monitoring sites were representative of average
weather conditions experienced throughout the entire year. The largest differences are shown for
PAFL. However, sampling at PAFL took place on a l-in-12 day schedule, yielding roughly half
the sample days as the other Florida monitoring sites and results in more variability in the sample
day averages.
The highest average dew point and wet bulb temperatures among NMP sites were
calculated for the Florida monitoring sites. AZFL and SKFL also experienced some of the
highest relative humidity levels among NMP sites.
9.2.3 Back Trajectory Analysis
Figure 9-9 is the composite back trajectory map for days on which samples were
collected at the AZFL monitoring site in 2012. Included in Figure 9-9 are four back trajectories
per sample day. Figure 9-10 is the corresponding cluster analysis. Similarly, Figures 9-11
through 9-18 are the composite back trajectory maps and corresponding cluster analyses for the
remaining Florida monitoring sites. An in-depth description of these maps and how they were
generated is presented in Section 3.5.2.1. For the composite maps, each line represents the
24-hour trajectory along which a parcel of air traveled toward the monitoring site on a given
sample day and time, based on an initial height of 50 meters AGL. For the cluster analyses, each
line corresponds to a trajectory representative of a given cluster of back trajectories. Each
concentric circle around the sites in Figures 9-9 through 9-18 represents 100 miles.
9-17
-------
Figure 9-9. Composite Back Trajectory Map for AZFL
Figure 9-10. Back Trajectory Cluster Map for AZFL
9-18
-------
Figure 9-11. Composite Back Trajectory Map for SKFL
V
/""/--./
Figure 9-12. Back Trajectory Cluster Map for SKFL
9-19
-------
Figure 9-13. Composite Back Trajectory Map for SYFL
Figure 9-14. Back Trajectory Cluster Map for SYFL
0 100 200
9-20
-------
Figure 9-15. Composite Back Trajectory Map for ORFL
Figure 9-16. Back Trajectory Cluster Map for ORFL
r >\ -'' /
9-21
-------
Figure 9-17. Composite Back Trajectory Map for PAFL
Figure 9-18. Back Trajectory Cluster Map for PAFL
100 200
9-22
-------
Observations from Figures 9-9 through 9-14 for the Tampa/St. Petersburg sites include
the following:
• The composite back trajectory maps for the Tampa/St. Petersburg sites are similar to
each other in trajectory distribution, which is not unexpected given their close
proximity to each other. Back trajectories originated from a variety of directions at
the Tampa/St. Petersburg sites.
• The 24-hour air shed domains for these sites were comparable in size to other NMP
sites, with the average trajectory length ranging from 227 miles for AZFL to 232
miles for SYFL. The farthest away a back trajectory originated was nearly 570 miles
away, originating over Tennessee, although back trajectories of similar length also
originated towards the Yucatan Peninsula of Mexico and eastward over the Atlantic
Ocean. However, most trajectories (roughly 86 percent for each site) originated
within 400 miles of the Tampa/St. Petersburg monitoring sites.
• The cluster maps for AZFL and SKFL are similar to each other in geographical
breakup and the percentages differ only slightly. The cluster maps show that
approximately one-quarter of back trajectories originated to the northwest, north, and
northeast of the sites, primarily over Alabama, Georgia, and the offshore waters of
Georgia and northeast Florida. Another one-quarter of back trajectories originated to
the northeast, east, and southeast of the sites, over the Atlantic Ocean and northern
Bahamas. Roughly 15 percent of back trajectories originated southward towards the
Straights of Florida, western Cuba, the Gulf of Mexico, and the Yucatan Peninsula.
Greater than one-third of the back trajectories are represented by the short cluster
trajectory originating just west of the Tampa/St. Petersburg area and over the Gulf of
Mexico. This cluster includes back trajectories of varying lengths originating over the
Gulf of Mexico as well as shorter trajectories originating from a variety of directions
around the sites but generally within 200 miles of the sites.
• The cluster map for SYFL has more cluster trajectories than the cluster maps for
AZFL and SKFL. The cluster analysis splits the northward-originating cluster
trajectory for AZFL and SKFL into two cluster trajectories for SYFL; one
representing back trajectories originating over Alabama and Georgia, the other
representing the back trajectories originating offshore. Similarly, the cluster analysis
splits the short cluster trajectory originating just offshore the Tampa/St. Petersburg
area for AZFL and SKFL into two back trajectories for SYFL; one representing the
short trajectories originating over central Florida or just south of the St. Petersburg
peninsula and one representing longer back trajectories originating farther westward
over the Gulf of Mexico. The cluster trajectory originating eastward over the Atlantic
Ocean and the cluster trajectory originating southward towards the Florida Keys are
similar to the cluster trajectories for AZFL and SKFL.
9-23
-------
Observations from Figures 9-15 through 9-18 for ORFL and PAFL include the
following:
• The composite back trajectory map for PAFL has fewer back trajectories compared to
the composite map for ORFL. This is because sampling at PAFL occurred on a 1-in-
12 day schedule, yielding approximately half the sample days as ORFL. The long
back trajectories originating over western Cuba are for the June 26, 2012 sample day;
samples were not collected on this day at PAFL; thus, these back trajectories are not
shown on the composite map for PAFL.
• The 24-hour air shed domain for ORFL is the largest in size compared to the other
Florida monitoring sites, with an average back trajectory length of 250 miles. The
longest back trajectory originated over central Tennessee, or approximately 580 miles
away, with a few additional back trajectories of similar length originating over and
south of western Cuba. However, greater than 90 percent of back trajectories
originated with 450 miles of ORFL.
• Nearly half of all back trajectories are represented by the short cluster originating to
the southwest of ORFL (45 percent), as shown on this site's cluster map. This cluster
includes back trajectories originating to the south of a diagonal line drawn across the
Panhandle of Florida, through ORFL, and extending across the Bahamas. The cluster
map groups the remaining back trajectories into three directions: those originating
northwestward over the Florida Panhandle, Georgia, and Alabama; those originating
northeastward off the Southeast Coast; and those originating eastward over the
Atlantic Ocean and northern Bahamas.
• The composite map for PAFL shows that the longest back trajectories originated over
Alabama and Tennessee or over the Atlantic Ocean, predominantly east of the
monitoring site. The back trajectories originating over northern Florida and southeast
Georgia, south Florida, or the Gulf of Mexico were generally of shorter length.
• The cluster map for PAFL has almost twice the number of cluster trajectories (7) than
the cluster map for ORFL (4). This can be attributed to the difference in the number
of sample days. One-third of back trajectories originated over south Florida and the
adjacent offshore waters. Nearly 40 percent of back trajectories originated over the
Atlantic Ocean, but are represented by three separate cluster trajectories. Ten percent
of back trajectories originated to the southwest of PAFL, over the waters south of the
St. Petersburg Peninsula. The cluster trajectory originating over the Florida/Georgia
border represents back trajectories originating over southeast Georgia and over north
Florida as well as those originating over the Panhandle of Florida and the adjacent
waters. Finally, the four back trajectories originating over Alabama and Tennessee
are grouped together in a single cluster trajectory and represent three percent of the
sample day back trajectories.
9-24
-------
9.2.4 Wind Rose Comparison
Hourly surface wind data from the weather stations nearest the Florida sites, as presented
in Section 9.2.2, were uploaded into a wind rose software program to produce customized wind
roses, as described in Section 3.5.2.2. 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 9-19 presents a map showing the distance between the weather station and AZFL,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 9-19 also presents three different wind roses for the
AZFL monitoring site. First, a historical wind rose representing 2002 to 2011 wind data is
presented, which shows the predominant surface wind speed and direction over an extended
period of time. Second, a wind rose representing wind observations for all of 2012 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2012 is
presented. These can be used to identify the predominant wind speed and direction for 2012 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically. Figures 9-20 through 9-23 present the three wind roses and
distance maps for SKFL, SYFL, ORFL, and PAFL, respectively.
9-25
-------
Figure 9-19. Wind Roses for the St. Petersburg/Whitted Airport Weather Station near
AZFL
Location of AZFL and Weather Station
2002-2011 Historical Wind Rose
2012 Wind Rose
Sample Day Wind Rose
WIND SPEED
(Knots)
^| 17 - 21
^| 11 . 17
O «-7
HI -
Calms: 9.37%
9-26
-------
Figure 9-20. Wind Roses for the St. Petersburg/Clearwater International Airport Weather
Station near SKFL
Location of SKFL and Weather Station
2002-2011 Historical Wind Rose
/ 5
2012 Wind Rose
Sample Day Wind Rose
Calms: 13.53%
9-27
-------
Figure 9-21. Wind Roses for the Plant City Municipal Airport Weather Station near SYFL
Location of SYFL and Weather Station
2008-2011 Historical Wind Rose
Calms: 24.88%
2012 Wind Rose
Sample Day Wind Rose
Calms: 23.-60%
9-28
-------
Figure 9-22. Wind Roses for the Orlando Executive Airport Weather Station near ORFL
Location of ORFL and Weather Station
2002-2011 Historical Wind Rose
Calms: 14.54%
2012 Wind Rose
Sample Day Wind Rose
Calms: 15.37%
9-29
-------
Figure 9-23. Wind Roses for the Orlando Executive Airport Weather Station near PAFL
Location of PAFL and Weather Station
2002-2011 Historical Wind Rose
ff
Oxford Si
PichlairSI
N5
(l< «
Calms: 14.54%
2012 Wind Rose
Sample Day Wind Rose
Calms: 15.37%
9-30
-------
Observations from Figure 9-19 for AZFL include the following:
• The weather station at St. Petersburg/Whitted Airport is located 6.9 miles east 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 historical wind rose shows that winds from the north, northeast quadrant, and
east were the most commonly observed wind directions near AZFL while winds from
the western quadrants were observed less frequently. Calm winds (< 2 knots)
accounted for less than 8 percent of the hourly wind measurements.
• The full-year wind rose shows that winds from the north, east-northeast, and east are
the predominant wind directions for 2012. While winds from the northwest quadrant
and north-northeast to northeast were observed less frequently than in previous years,
winds from the southeast and southwest quadrant were observed more often.
• The sample day wind patterns favor the full-year wind patterns, with east-
northeasterly and easterly winds observed the most. However, fewer northerly winds
were observed with a greater percentage of winds from the southeast and west-
southwest observed on sample days.
Observations from Figure 9-20 for SKFL include the following:
• The weather station at St. Petersburg/Clearwater International Airport is located
4.4 miles north-northeast of SKFL. The St. Petersburg/Clearwater Airport is located
on Old Tampa Bay while SKFL is farther inland.
• The historical wind rose shows that winds from a variety of directions were observed
near SKFL, although winds from the north, northeast quadrant, east, and
east-southeast were the most commonly observed wind directions. Calm winds
accounted for approximately 10 percent of the hourly wind measurements.
• The 2012 wind rose resembles the historical wind rose in that winds from the
northeast to east-southeast account for a majority of the wind observations. There is a
higher percentage of calm winds for 2012 (nearly 13 percent) while winds from the
north and north-northeast were observed less frequently.
• The predominance of winds from the northeast to east-southeast is even more evident
on the sample day wind rose. With the exception of north-northeast, none of the other
directions account for more than 5 percent of wind observations while calm winds
account for nearly 14 percent of observations on sample days.
9-31
-------
Observations from Figure 9-21 for SYFL include the following:
• The weather station at Plant City Municipal Airport is located 4.6 miles northeast of
SYFL. Note that this weather station has less historical data than the other stations.
This station did not begin operating until 2006 and data availability is intermittent
until mid-2007; thus, the historical wind rose includes data from the first full-year of
data (2008) through 2011.
• The historical wind rose shows that calm winds (< 2 knots) account for approximately
25 percent of the hourly wind measurements between 2008 and 2011. Winds from the
eastern quadrants were observed more often than the western quadrants, although
winds from all directions were observed near SYFL. Winds from due east account for
the highest percentage of winds near SYFL (10 percent).
• Both the full-year and sample day wind patterns are similar to the historical wind
patterns, indicating that conditions on sample days were representative of wind
conditions experienced throughout the year and historically.
Observations from Figures 9-22 and 9-23 for ORFL and PAFL include the following:
• The closest weather station to both ORFL and PAFL is the Orlando Executive
Airport. The weather station is located just less than 4 miles southeast of ORFL and
less than 1 mile east-southeast of PAFL, as PAFL is located on the edge of the
Orlando Executive Airport property. Thus, the historical and full-year wind roses for
these sites are identical.
• The historical wind roses show that winds from all directions were observed near
these sites, with easterly winds being observed the most, followed by winds from due
north and due south. Winds with an easterly component were observed more often
than winds with a westerly component. Calm winds were observed for less than
15 percent of the wind observations.
• The wind patterns shown on the full-year wind roses resemble the wind patterns on
the historical wind roses.
• The sample day wind rose for ORFL exhibits the same prominence of easterly,
northerly, and southerly winds, but winds from the entire northeast quadrant as well
as winds from the south-southwest account for a higher percentage of wind
observations than they do for the historical and full-year wind roses.
• The sample day wind rose for PAFL shares the easterly and southerly prominence of
the full-year wind rose; however, winds from the northwest to north and southwest to
west-southwest are reduced. The reductions in the wind observations from these
directions are seen in additional observations in winds from the northeast to east to
east-southeast as well as south-southwest. Note, however, that PAFL samples on a 1-
in-12 day sampling schedule, leading to roughly half the sample days included in the
sample day wind rose as ORFL.
9-32
-------
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 9-4.
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 9-4. It is
important to note which pollutants were sampled for at each site when reviewing the results of
this analysis. AZFL and ORFL sampled for carbonyl compounds only. SKFL and SYFL sampled
for hexavalent chromium and PAHs in addition to carbonyl compounds. PAFL sampled for only
metals.
Observations from Table 9-4 include the following:
• For AZFL and ORFL, the two sites sampling only carbonyl compounds, acetaldehyde
and formaldehyde were the only two pollutants to fail screens. For both sites,
formaldehyde failed one additional screen than acetaldehyde. Among the carbonyl
compounds, only acetaldehyde, formaldehyde, and propionaldehyde have risk
screening values. Propionaldehyde did not fail any screens for these two sites.
• Eight pollutants failed at least one screen for SKFL; 39 percent of concentrations for
these eight pollutants were greater than their associated risk screening value (or failed
screens). Three pollutants (acetaldehyde, formaldehyde, and naphthalene) contributed
to 95 percent of failed screens for SKFL and therefore were identified as pollutants of
interest for this site. Note that each of the remaining pollutants failed only one screen
each.
• Five pollutants failed at least one screen for SYFL; 58 percent of concentrations for
these five pollutants were greater than their associated risk screening value (or failed
screens). Similar to SKFL, three pollutants (acetaldehyde, formaldehyde, and
naphthalene) contributed to 95 percent of failed screens for SYFL and therefore were
identified as pollutants of interest for this site. Note that each of the remaining
pollutants failed only one screen each.
• Formaldehyde failed 100 percent of screens for all four sites sampling carbonyl
compounds.
9-33
-------
• Arsenic, manganese, and lead fail screens for PAFL, with arsenic contributing to
nearly 80 percent of the total failed screens. Arsenic and manganese contributed to
95 percent of failed screens for PAFL and therefore were identified as pollutants of
interest for this site.
Table 9-4. Risk-Based Screening Results for the Florida Monitoring Sites
Pollutant
Screening
Value
(Ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
St. Petersburg, Florida - AZFL
Formaldehyde
Acetaldehyde
0.077
0.45
Total
59
58
117
59
59
118
100.00
98.31
99.15
50.43
49.57
50.43
100.00
Pinellas Park, Florida - SKFL
Acetaldehyde
Formaldehyde
Naphthalene
Acenaphthene
Benzo(a)pyrene
Fluorene
Hexavalent Chromium
Propionaldehyde
0.45
0.077
0.029
0.011
0.00057
0.011
0.000083
0.8
Total
59
59
53
176
59
59
61
61
50
60
48
59
457
100.00
100.00
86.89
1.64
2.00
1.67
2.08
1.69
38.51
33.52
33.52
30.11
0.57
0.57
0.57
0.57
0.57
33.52
67.05
97.16
97.73
98.30
98.86
99.43
100.00
Valrico, Florida - SYFL
Acetaldehyde
Formaldehyde
Naphthalene
Benzo(a)pyrene
Propionaldehyde
0.45
0.077
0.029
0.00057
0.8
Total
60
60
31
1
1
153
60
60
59
23
60
262
100.00
100.00
52.54
4.35
1.67
58.40
39.22
39.22
20.26
0.65
0.65
39.22
78.43
98.69
99.35
100.00
Winter Park, Florida - ORFL
Formaldehyde
Acetaldehyde
0.077
0.45
Total
61
60
121
61
61
122
100.00
98.36
99.18
50.41
49.59
50.41
100.00
Orlando, Florida - PAFL
Arsenic (PM10)
Manganese (PM10)
Lead (PM10)
0.00023
0.005
0.015
Total
29
7
1
37
30
30
30
90
96.67
23.33
3.33
41.11
78.38
18.92
2.70
78.38
97.30
100.00
9-34
-------
9.4 Concentrations
This section presents various concentration averages used to characterize pollution 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. Additional site-specific statistical summaries for each of
the Florida monitoring sites are provided in Appendices L, M, N, and O.
9.4.1 2012 Concentration Averages
Quarterly and annual concentration averages were calculated for the pollutants of interest
for each Florida site, as described in Section 3.1. The quarterly average 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 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 Florida monitoring
sites are presented in Table 9-5, 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.
9-35
-------
Table 9-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest
for the Florida Monitoring Sites
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Ug/m3)
2nd
Quarter
Average
(Ug/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Ug/m3)
St. Petersburg, Florida - AZFL
Acetaldehyde
Formaldehyde
59/59
59/59
1.17
±0.16
2.27
±0.22
1.11
±0.24
2.05
±0.35
1.24
±0.27
1.79
±0.32
2.25
±0.89
1.45
±0.24
1.41
±0.24
1.90
±0.16
Pinellas Park, Florida - SKFL
Acetaldehyde
Formaldehyde
Naphthalene a
59/59
59/59
61/61
1.58
±0.22
2.58
±0.27
89.29
± 36.62
1.80
±0.53
3.13
±0.46
115.51
±60.39
1.04
±0.26
2.73
±1.25
71.88
± 19.66
1.23
±0.39
2.27
±0.44
112.64
±49.83
1.41
±0.19
2.69
±0.36
96.91
±21.04
Valrico, Florida - SYFL
Acetaldehyde
Formaldehyde
Naphthalene a
60/60
60/60
59/59
1.24
±0.34
1.77
±0.25
42.07
±13.19
1.66
±0.66
2.95
±0.99
38.46
±13.52
1.36
±0.24
2.38
±0.36
28.18
±6.63
1.55
±0.36
1.80
±0.26
39.08
± 14.00
1.45
±0.20
2.24
±0.29
36.75
±5.79
Winter Park, Florida - ORFL
Acetaldehyde
Formaldehyde
61/61
61/61
1.51
±0.55
1.74
±0.23
0.93
±0.19
2.45
±0.44
0.87
±0.14
2.26
±0.39
1.03
±0.30
1.75
±0.32
1.08
±0.17
2.05
±0.19
Orlando, Florida - PAFL
Arsenic (PM10)a
Manganese (PM10)a
30/30
30/30
1.10
±1.03
2.02
±0.60
0.61
±0.32
2.53
±1.77
0.86
±0.47
7.53
±2.97
1.49
±0.81
2.32
±0.97
1.02
±0.33
3.69
±1.20
1 Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m
for ease of viewing.
Observations from Table 9-5 include the following:
• The annual average concentration of formaldehyde is higher than the annual average
concentration of acetaldehyde, for the sites where these two pollutants were
measured.
• The annual average concentrations of formaldehyde range from 1.90 ± 0.16 |ig/m3
(AZFL) to 2.69 ± 0.36 |ig/m3 (SKFL). The annual average concentrations of
acetaldehyde varied less, ranging from 1.08 ± 0.17 |ig/m3 (ORFL) to
1.45 ± 0.20 |ig/m3 (SYFL).
9-36
-------
• The quarterly average concentrations of acetaldehyde and formaldehyde do not
appear to exhibit a seasonal trend of any type. However, a few of the quarterly
averages do stand out, as described in the bullets that follow.
• The fourth quarter acetaldehyde average concentration for AZFL is greater than the
other quarterly average concentrations and has a relatively large confidence interval
associated with it. Two concentrations greater than 5 |ig/m3 were measured at AZFL,
one in November (5.69 |ig/m3) and one in December (5.43 |ig/m3). The next highest
concentration was also measured during the fourth quarter but was roughly half as
high (2.49 |ig/m3). Five of the eight concentrations greater than 2 |ig/m3 were
measured at AZFL during the fourth quarter of 2012.
• The second quarter acetaldehyde average for SKFL is greater than the other quarterly
average concentrations and has a relatively large confidence interval associated with
it. The maximum acetaldehyde concentration was measured at SKFL in May
(5.02 |ig/m3). The next two highest concentrations were measured on the same days
in November and December as the maximum acetaldehyde concentrations measured
at AZFL but were roughly half as high (2.67 |ig/m3 and 2.65 |ig/m3).
• Although the second quarter formaldehyde average for SKFL is greater than the other
quarterly averages, the third quarter average has a large confidence interval associated
with it. The maximum formaldehyde concentration was measured at SKFL on July
20, 2013 (11.43 |ig/m3). This concentration is two and a half times higher than the
next two highest measurements, both of which were measured in May, and is among
the highest formaldehyde concentrations measured across the program. No other
formaldehyde measurements greater than 4 |ig/m3 were collected at this site.
• The second quarter formaldehyde average for SYFL is greater than the other quarterly
averages and has a relatively large confidence interval associated with it. The
maximum formaldehyde concentration was measured at SYFL on May 27, 2013
(9.08 |ig/m3). This concentration is more than twice the next highest measurement
(4.02 |ig/m3 collected on September 24, 2013). No other formaldehyde measurements
greater than 4 |ig/m3 were collected at this site. The highest and third highest
formaldehyde concentrations were collected at SYFL on the same days in May as the
second and third highest formaldehyde concentrations were collected at SKFL.
• Naphthalene was identified as a pollutant of interest for both SKFL and SYFL. The
annual average concentration of naphthalene for SKFL is more than twice the annual
average concentration for SYFL. A single measurement greater than 100 ng/m3 was
collected at SYFL while 19 measurements greater than 100 ng/m3 were measured at
SKFL, including seven greater than 200 ng/m3 and one greater than 400 ng/m3. The
maximum naphthalene concentration measured at SKFL (435 ng/m3) is among the
highest concentrations of naphthalene measured across the program.
• PAFL is the only Florida monitoring site that did not sample carbonyl compounds or
PAHs. The confidence interval for the first quarter average concentration of arsenic is
nearly equivalent to the average itself, indicating the potential for outliers. The
maximum arsenic concentration was measured at PAFL on January 22, 2012
9-37
-------
(3.86 ng/m3). The next highest measurement collected during this quarter was
significantly less (1.21 ng/m3). All other concentrations measured during this quarter
were less than 0.65 ng/m3.
• Not only is the third quarter average concentration of manganese significantly greater
than the other quarterly averages, it also has a relatively large confidence interval.
The five highest concentrations of manganese were all measured at PAFL during the
third quarter and ranged from 8.05 ng/m3 to 13.1 ng/m3. Manganese concentrations
measured at PAFL span an order of magnitude, ranging from 1.06 ng/m3 to
13.1 ng/m3, with a median concentration of 2.08 ng/m3.
Tables 4-9 through 4-12 present the 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 carbonyl compounds.
• SKFL has the eighth highest annual average concentration of naphthalene among
NMP sites sampling this pollutant, as shown in Table 4-11. Note that the confidence
interval associated with SKFL's annual average is among the larger confidence
intervals, indicating more variability associated with this site's measurements.
• The annual average concentration of arsenic for PAFL ranked third highest among
NMP sites sampling PMio metals. This site is one of only three sites with annual
average arsenic concentrations greater than 1 ng/m3. The confidence interval
associated with this annual average is also among the larger confidence intervals
shown, indicating a higher level of variability associated with this site's
measurements.
9.4.2 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 shaded in
gray in Table 9-4 for each of the Florida monitoring sites. Figures 9-24 through 9-28 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 3.5.3.1.
9-38
-------
Figure 9-24. Program vs. Site-Specific Average Acetaldehyde Concentrations
AZFL
ORFL
9 12
Concentration {[og/m3)
Program:
Site:
IstQuartile
Site Average
O
2ndQuartile SrdQuartile
Site Concentration Range
4thQuartile Average
Figure 9-25. Program vs. Site-Specific Average Arsenic (PMi0) Concentration
3 4
Concentration (ng/m3)
Program:
Site:
IstQuartile
D
Site Average
O
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^~
4thQuartile Average
D 1
9-39
-------
Figure 9-26. Program vs. Site-Specific Average Formaldehyde Concentrations
AZFL
ORFL
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o
Figure 9-27. Program vs. Site-Specific Average Manganese (PMi0) Concentration
E
Program Max Concentration = 275 ng/m3
60 90
Concentration (ng/m3)
Program:
Site:
IstQuartile
D
Site Average
O
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
9-40
-------
Figure 9-28. Program vs. Site-Specific Average Naphthalene Concentrations
SKFL
400 500
Concentration {ng/m3)
Program
Site:
: IstQuartile
Site Average
o
2ndQuartile SrdQuartile 4thQuartile Av
• n
Site Concentration Range
?rage
Observations from Figures 9-24 through 9-28 include the following:
• Figure 9-24 for acetaldehyde shows that the range of acetaldehyde measurements
collected at the Florida sites were not significantly different from each other. The annual
average acetaldehyde concentrations for each of the Florida sites are less than the
program-level average concentration. The maximum concentration measured at each site
is significantly less than the maximum concentration measured across the program.
• Figure 9-25 for arsenic shows that PAFL's annual average concentration is greater than
the program-level average concentration as well as the program-level third quartile. The
maximum arsenic concentration measured at PAFL is roughly half the maximum
concentration measured among sites sampling PMi0 metals. There were no non-detects of
arsenic measured at PAFL, although there were a few reported across the program.
• Figure 9-26 for formaldehyde shows there is more variability in the measurements of
formaldehyde among the Florida sites than there is for acetaldehyde. AZFL and ORFL
measured roughly the same range of measurements of formaldehyde and their annual
averages are both less than the program-level average concentration. Although the
maximum concentration of formaldehyde measured at SYFL is more than twice the
maximum concentrations measured at AZFL or ORFL, the annual average for SYFL is
just slightly greater than those calculated for AZFL or ORFL and roughly equivalent to
the program-level median concentration. The maximum formaldehyde concentration
measured at SKFL is one of the highest concentrations measured among NMP sites
sampling this pollutant. The annual average concentration for SKFL is the only annual
average among the Florida sites greater than the program-level average concentration
(but just barely).
• Figure 9-27 presents the box plot for manganese. Note that the program-level maximum
concentration (275 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
9-41
-------
concentration range. Thus, the scale has been reduced to 150 ng/m3. Figure 9-27 for
manganese shows that PAFL's annual average concentration is less than both the
program-level average and median concentrations, despite the relative variability in the
data set observed from the quarterly average concentrations discussed above. Compared
to other NMP sites sampling manganese, this site's annual average concentration ranks
13th (out of 14). The maximum manganese concentration measured at PAFL is
considerably less than the maximum concentration measured among NMP sites sampling
metals.
• Figure 9-28 presents the box plots for naphthalene. The range of measurements collected
at SKFL and SYFL are considerably different. The maximum concentration measured at
SYFL is roughly equivalent to the program-level third quartile while the maximum
concentration measured at SKFL is roughly four times higher. The annual average
concentration for SYFL is just greater than the program-level first quartile while the
annual average concentration for SKFL is greater than the program-level average
concentration.
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 3.5.3.2.
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 SYFL and PMio metals at
PAFL began in 2008. Thus, Figures 9-29 through 9-40 present the 1-year statistical metrics for
each of the pollutants of interest for each of the 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 is not provided, although the range and quartiles are still presented.
9-42
-------
Figure 9-29. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at AZFL
i
.... Average
Observations from Figure 9-29 for acetaldehyde measurements collected 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 a similar concentration was also measured in 2003 (8.00 |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, stayed elevated through 2004, then began to
decrease significantly, a trend that continued through 2008.
• The 1-year average and median began to increase again in 2009. This increase cannot
be attributed to an outlier here or there because the trend continued into 2010 and the
all statistical metrics exhibited this increase. The 95th percentile more than doubled
from 2008 to 2009, as did the 1-year average concentration. A significant decrease is
shown for 2011 and continues into 2012. Additional years of sampling are required to
determine if this decreasing trend continues or if another round of increasing will be
exhibited.
9-43
-------
Figure 9-30. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at AZFL
1
j, 8.0
§
u
4.0 -
-(
; f
o ^ ^ 1^1^
. S O. ... xv . '" 'Si >vJ
I — -^ ^ ^r z. *M f^T ^1
-sJLfJLfJLIJi ~LSJL2-ILQJ
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Year
O 5th Percentile — Minimum — Median — Maximum O 95th Percentile ...<>... Average
Observations from Figure 9-30 for formaldehyde measurements collected at AZFL
include the following:
• The maximum formaldehyde concentration was measured in 2001, after which the
highest concentration measured decreased by nearly half. The three highest
concentrations of formaldehyde (ranging from 9.30 |ig/m3 to 16.1 |ig/m3) were all
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. Very little change is shown for the last
2 years of sampling.
• The trends shown for formaldehyde in Figure 9-30 are almost the opposite of the
trends shown for acetaldehyde in Figure 9-29, particularly for the period between
2004 through 2008.
• The difference between the 5th and 95th percentiles, the range within which the
majority of the concentrations lie, is at a minimum for 2012.
9-44
-------
Figure 9-31. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SKFL
-r
T
2008 2009
Year
O 5th Percentile - Minimum
Median — Maximum O 95th Percentile
Observations from Figure 9-31 for acetaldehyde measurements collected at SKFL include
the following:
• Sampling for carbonyl compounds began at SKFL under the NMP in late July 2004.
Because this represents less than half of the sampling year, Figure 9-31 excludes data
from 2004.
• The maximum acetaldehyde concentration shown was measured in
2010 (10.3 |ig/m3). Although the second highest concentration was measured in 2011
(8.94 |ig/m3), the third, fourth, and fifth highest concentrations of acetaldehyde were
also 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 continues
into 2012.
9-45
-------
Figure 9-32. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at SKFL
oncentration (ng/m
D C
Maximum
Concentration for
2005 is 91.7 u.g/m3.
[ r 1
i-V-i j ^_i_, 1
1 "• _^_ ^^ 0 _£_ •
2005 2006 2007 2008 2009 2010 2011 2012
Year
O 5th Percentile - Minimum ~ Median — Maximum O 95th Percentile ...^>... Average
Observations from Figure 9-32 for formaldehyde measurements collected 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, but is considerably less (11.4 |ig/m3).
• For 2005, the 1-year average concentration is greater than the 95th percentile,
reflecting the effects that an outlier can have on statistical measurements. With the
exception of the maximum concentration measured in 2012, all other concentrations
measured at this site were less than 6 |ig/m3 for the years shown.
• The 1-year average and median concentrations exhibit a steady decreasing trend
through 2010. The range of measurements is at a minimum for 2010 and the 1-year
average and median concentration are nearly equivalent, reflecting little variability in
the central tendency of the measurements.
• The range of concentrations measurements increased significantly from 2010 to 2011,
with the range within which 90 percent of the concentrations fall more than doubling.
• All of the statistical parameters increased from 2011 to 2012, indicating that
concentrations of formaldehyde were higher overall at SKFL for 2012.
9-46
-------
Figure 9-33. Yearly Statistical Metrics for Naphthalene Concentrations Measured at SKFL
~ 250
2010
Year
O 5th Percentile - Minimuir
— Maximum • 95th Percentile "-O-" Averagf
A 1-year average is not presented because sampling under the NMP did not begin until March 2008.
Observations from Figure 9-33 for naphthalene measurements collected at SKFL include
the following:
• Sampling for PAHs began at SKFL under the NMP in March 2008. A 1-year average
is not presented for 2008 because a full year's worth of data is not available, although
the range of measurements is provided.
• The maximum naphthalene concentration was measured at SKFL in 2012
(435 ng/m3). Two additional measurements greater than 300 ng/m3 have been
measured at SKFL, one in 2008 and the other in 2010.
• The range within which the majority of naphthalene concentrations fall has changed
very little across the years of sampling, although there is an increase shown for 2012
as 2012 has the greatest number of measurements greater than 200 ng/m3 (seven).
• The 1-year average concentrations have varied from 82.22 ng/m3 (2011) to
96.91 ng/m3 (2012). Confidence intervals calculated for these averages indicate that
the changes over the years are not statistically significant.
9-47
-------
Figure 9-34. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SYFL
i
i
i
2006 2007 2008
Year
O BthPercentile
— Minimum
O 95th Percentile
Observations from Figure 9-34 for acetaldehyde measurements collected at SYFL include
the following:
• Carbonyl compounds have been measured at SYFL under the NMP since January
2004.
• The maximum acetaldehyde concentration was measured on January 18, 2007
(15.3 |ig/m3). The next highest concentration, also measured in 2007, is roughly half
as high (7.55 |ig/m3). Only one additional acetaldehyde measurement collected at
SYFL is greater than 7 |ig/m3 and was measured in 2008.
• 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
from the calculation, the 1-year average concentration for 2007 is still 50 percent
greater than 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 highest measurements
driving this 1-year average concentration.
• With the exception of 2007, the 1-year average concentrations have fluctuated
between 1.03 |ig/m3 (2011) and 1.60 |ig/m3 (2004). Confidence intervals calculated
for the 1-year averages indicate that the year-to-year changes for years 2009 through
2012 are statistically significant.
9-48
-------
Figure 9-35. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at SYFL
sr
entration (ngy
3
I
S r
T 1
T T
L, T ^ ^ ^ i ^k
0., •<>.._ .... -T ^ jjfe ^i mm ^ ^
_X_| _.-,_ — o— ! T~^ -Q- -a- -o-1
1-£-1 L-«-l >— Q-1 ^-^ ^"^ •"• ^^^ ^^
2004 2005 2006 2007 2008 2009 2010 2011 2012
Year
• 5th Percentile — Minimum ~ Median — Maximum O 95th Percentile ...^... Average
Observations from Figure 9-35 for formaldehyde measurements collected at SYFL
include the following:
• The maximum formaldehyde concentration was measured at SYFL in 2005
(32.5 |ig/m3) and was nearly twice the next highest concentration (17.1 |ig/m3,
measured in 2008), although several measurements of similar magnitude were also
measured in 2007. In all, eight formaldehyde concentrations greater than 10 |ig/m3
have been measured at SYFL, five in 2007 and one each in 2005, 2008, and 2010.
• Even though the maximum concentration was measured in 2005, the next 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. The outlier measured in 2005 is mostly reflected in
the confidence intervals calculated for this 1-year average concentration.
• Although the maximum concentration for 2007 is considerably less than the
maximum measured in 2005, the other statistical parameters exhibit significant
increases. In particular, the 95th percentile is four times higher and the 1-year average
doubled from 2006 to 2007. These statistical parameters indicate that the
measurements collected in 2007 were higher overall compared to other years. The
number of formaldehyde concentrations greater than 5 |ig/m3 is highest for 2007
(seven), while every other year of sampling has two or less.
9-49
-------
• The 1-year average formaldehyde concentration has fluctuated over the years, ranging
from 1.58 |ig/m3 (2006) to 3.19 |ig/m3 (2007), with little change in the last 2 years.
Figure 9-36. Yearly Statistical Metrics for Naphthalene Concentrations Measured at SYFL
'a 80
c
§
s
O
2010
Year
0 5th Percentile — Minimum ~ Median — Maximum • 95th Percentile
1A 1-year average is not presented because sampling under the NMP did not begin until April 2008.
Observations from Figure 9-36 for naphthalene measurements collected at SYFL include
the following:
• Sampling for PAHs began at SYFL under the NMP in April 2008. A 1-year average
is not presented for 2008 because a full year's worth of data is not available, although
the range of measurements is provided.
• The two highest naphthalene concentrations were both measured in 2011 (132 ng/m3
and 131 ng/m3), although measurements greater than 100 ng/m3 were also measured
2008, 2009, and 2012.
• The range within which the majority of naphthalene concentrations fall, as indicated
by the difference between the 5th and 95th percentile for each year, has changed very
little across the years of sampling. Although there is a slight increase shown for 2012,
both the median and 1-year average concentrations exhibit slight decreases for 2012.
This decrease is a result of a higher number of measurements at the lower end of the
concentration range.
9-50
-------
• The 1-year average concentrations have varied from 36.75 ng/m3 (2012) to
43.38 ng/m3 (2010), although confidence intervals calculated for these averages
indicate that the changes over the years are not statistically significant.
Figure 9-37. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at ORFL
I
T
2007 2008 2009
Year
5th Percentile - Minimum ~ Median — Maximum
95thPercentile —O—Averagf
A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
Observations from Figure 9-37 for acetaldehyde measurements collected at ORFL
include the following:
• Sampling for carbonyl compounds began at ORFL under the NMP in April 2003. A
1-year average 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). The
next three highest concentrations are the maximum concentrations shown for the
three years that follow.
• Between 2007 and 2011, the 1-year average concentrations have varied from
1.45 |ig/m3 (2010) to 1.85 |ig/m3 (2011). The 1-year average concentration is at a
minimum for 2012 (1.08 |ig/m3), which represents a significant decrease from 2011.
The median concentration decreased by almost half from 2011 to 2012. The number
of concentrations less than 1 |ig/m3 is one for 2011 but accounts for more than half of
the measurements for 2012.
9-51
-------
Figure 9-38. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at ORFL
^H
T
••O-
2007 2008
Year
O 5th Percentile
— Maximum O 95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
Observations from Figure 9-38 for formaldehyde measurements collected at ORFL
include the following:
• The maximum formaldehyde concentration was measured in 2007 (16.1 |ig/m3),
although concentrations greater than 10 |ig/m3 were also measured in 2005 and 2008.
• Even with the relatively high concentrations measured in the middle years of
sampling, the 1-year average concentrations exhibit a steady decreasing trend through
2011. The median concentrations have decreased as well, but exhibited an increase in
2009, followed by additional decreases.
• The range of formaldehyde concentrations is at a minimum for 2012, and the
maximum concentration for 2012 is the lowest maximum concentration shown for all
years of sampling. Despite this, both the 1-year average and median concentrations
increased slightly for 2012. Compared to 2011, concentrations measured in 2012 are
just higher overall. There are fewer measurements at the lower end of the
concentration range for 2012, as there were no measurements less than 1 |ig/m3
measured in 2012 (compared to four in 2011). In addition, the number of
measurements at the upper end of the concentration range for 2012 is higher, as the
number of measurements greater than 3 |ig/m3 is nearly double for 2012 than 2011.
9-52
-------
Figure 9-39. Yearly Statistical Metrics for Arsenic (PMi0) Concentrations Measured at PAFL
O 5th Percentile
— Maximum
95th Percentile
Observations from Figure 9-39 for arsenic measurements collected at PAFL include the
following:
• All four of the arsenic concentrations greater than 2 ng/m3 were measured in 2012,
and ranged from 2.08 ng/m3 to 3.86 ng/m3.
• The range of arsenic measurements collected is at a minimum for 2010, increases for
2011, then doubles for 2012. The range within which the majority of concentrations
fall, 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 has a slight decreasing trend through 2010. After a
slight increase for 2011, the 1-year average increases substantially from 2011 to 2012.
The median concentration exhibits a decreasing trend through 2011, even though the
range of measurements increases from 2010 to 2011.
• The difference between the 1-year average and median concentrations is at a
minimum for 2010. The increasing difference between these two statistical
parameters for 2011 and 2012 indicates an increasing level of variability within the
measurements. The number of measurements at the upper end of the concentration
range has been increasing at PAFL, as the number of measurements greater than 1
ng/m3 increased from two in 2010 to five in 2011 to nine in 2012. Conversely, the
number of concentrations at the lower end of the range has been decreasing, even
9-53
-------
though the minimum concentration for each year is relatively unchanged. Additional
years of sampling are needed to determine if this trend continues.
Figure 9-40. Yearly Statistical Metrics for Manganese (PMio) Concentrations Measured at PAFL
5th Percentile - Minimuir
Median — Maximum
95th Percentile —O—Averagf
Observations from Figure 9-40 for manganese measurements collected at PAFL include
the following:
• The maximum concentration of manganese was measured in 2010 (13.9 ng/m3),
although similar measurements were also collected in 2009 and 2012 (13.1 ng/m3 for
both years).
• With the exception of 2011, the 1-year average concentrations have an overall
increasing trend since the onset of sampling at PAFL. However, the variability in the
measurements, as indicated by confidence intervals calculated for each 1-year
average concentration, indicates that the changes are not statistically significant.
• Similar to arsenic, the increase in the 95th percentile of manganese from 2011 to
2012 is substantial. But the 1-year average concentration for 2012 is greater than the
95th percentile for 2011, so this is not surprising. Eight measurements collected in
2012 are greater than the maximum concentration measured in 2011. Even if the
maximum concentration was removed from the dataset for 2012, the increase in the 1-
year average from 2011 to 2012 would still be greater than 1 ng/m3.
9-54
-------
9.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk at
each Florida monitoring site. Refer to Sections 3.3 and 3.5.3 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.
9.5.1 Risk-Based Screening Assessment Using MRLs
A risk-based screening was conducted by comparing the concentration data from the
Florida monitoring sites to the ATSDR MRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 day to 14 days); intermediate (exposures of 15 days to 364 days); and chronic
(exposures of 1 year or greater). The preprocessed daily measurements of the pollutants of
interest were compared to the acute MRLs; the quarterly averages were compared to the
intermediate MRLs; and the annual averages were compared to the chronic MRLs.
As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2012.
9.5.2 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 3.5.3.4 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 9-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
9-55
-------
Table 9-6. Risk Approximations for the Florida Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections vs.
# of Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
St. Petersburg, Florida - AZFL
Acetaldehyde
Formaldehyde
0.0000022
0.000013
0.009
0.0098
59/59
59/59
1.41
±0.24
1.90
±0.16
3.10
24.76
0.16
0.19
Pinellas Park, Florida - SKFL
Acetaldehyde
Formaldehyde
Naphthalene a
0.0000022
0.000013
0.000034
0.009
0.0098
0.003
59/59
59/59
61/61
1.41
±0.19
2.69
±0.36
0.10
±0.02
3.11
35.03
3.30
0.16
0.27
0.03
Valrico, Florida - SYFL
Acetaldehyde
Formaldehyde
Naphthalene a
0.0000022
0.000013
0.000034
0.009
0.0098
0.003
60/60
60/60
59/59
1.45
±0.20
2.24
±0.29
0.04
±0.01
3.20
29.07
1.25
0.16
0.23
0.01
Winter Park, Florida - ORFL
Acetaldehyde
Formaldehyde
0.0000022
0.000013
0.009
0.0098
61/61
61/61
1.08
±0.17
2.05
±0.19
2.38
26.68
0.12
0.21
Orlando, Florida - PAFL
Arsenic (PM10)a
Manganese (PM10) a
0.0043
0.000015
0.00005
30/30
30/30
<0.01
±<0.01
0.01
±0.01
4.41
0.07
0.07
- = A Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 9-5.
Observations for the Florida sites from Table 9-6 include the following:
• Formaldehyde has the highest cancer risk approximations among the sites sampling
carbonyl compounds, ranging from 24.76 in-a-million (AZFL) to 35.03 in-a-million
(SKFL).
• The cancer risk approximations for acetaldehyde are an order of magnitude less than
the cancer risk approximations for formaldehyde, ranging from 2.38 in-a-million
(ORFL) to 3.20 in-a-million (SYFL).
• The cancer risk approximation for naphthalene for SKFL (3.30 in-a-million) is twice
the cancer risk approximation for naphthalene for SYFL (1.25 in-a-million), although
both less than a level of concern.
9-56
-------
• For PAFL, arsenic has a cancer risk approximation of 4.41 in-a-million. A cancer
URE is not available for manganese; thus, a cancer risk approximation could not be
calculated.
• All of the noncancer hazard approximations for the site-specific pollutants of interest
are less than 1.0, indicating that no adverse health effects are expected from these
individual pollutants. The highest noncancer hazard approximation was calculated for
formaldehyde (0.27), based on the annual average concentration for SKFL.
9.5.3 Risk-Based Emissions Assessment
In addition to the risk-based screenings discussed above, Tables 9-7 and 9-8 present an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 9-7 presents the 10 pollutants with the highest emissions from the 2011 NEI that have
cancer toxicity factors. Table 9-7 also presents the 10 pollutants with the highest toxicity-
weighted emissions, based on the weighting schema described in Section 3.5.3.5. Lastly,
Table 9-7 provides the 10 pollutants with the highest cancer risk approximations (in-a-million)
for each site, as presented in Table 9-6. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 9-7. Table 9-8 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 3.5.3.5. Similar to the cancer risk and
noncancer hazard approximations provided in Section 9.5.2, this analysis may help policy-
makers prioritize their air monitoring activities.
9-57
-------
Table 9-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Florida Monitoring Sites
oo
Top 10 Total Emissions for Pollutants
with Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
St. Petersburg, Florida (Pinellas County) - AZFL
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Naphthalene
POM, Group la
Dichloromethane
POM, Group 2b
POM, Group 2d
281.41
179.80
149.82
98.95
40.57
17.93
11.61
3.85
2.14
1.93
Benzene
Formaldehyde
1,3 -Butadiene
POM, Group la
Naphthalene
Ethylbenzene
Arsenic, PM
Acetaldehyde
Hexavalent Chromium, PM
POM, Group 2b
2.19E-03
1.95E-03
1.22E-03
1.02E-03
6.10E-04
4.49E-04
2.34E-04
2.18E-04
2.17E-04
1.88E-04
Formaldehyde 24.76
Acetaldehyde 3.10
Pinellas Park, Florida (Pinellas County) - SKFL
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Naphthalene
POM, Group la
Dichloromethane
POM, Group 2b
POM, Group 2d
281.41
179.80
149.82
98.95
40.57
17.93
11.61
3.85
2.14
1.93
Benzene
Formaldehyde
1,3 -Butadiene
POM, Group la
Naphthalene
Ethylbenzene
Arsenic, PM
Acetaldehyde
Hexavalent Chromium, PM
POM, Group 2b
2.19E-03
1.95E-03
1.22E-03
1.02E-03
6.10E-04
4.49E-04
2.34E-04
2.18E-04
2.17E-04
1.88E-04
Formaldehyde 35.03
Naphthalene 3.30
Acetaldehyde 3.11
-------
Table 9-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Florida Monitoring Sites (Continued)
VO
Top 10 Total Emissions for Pollutants
with Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
Valrico, Florida (Hillsborough County) - SYFL
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Naphthalene
POM, Group la
Methyl tert butyl ether
POM, Group 2b
POM, Group 2d
419.18
276.57
259.18
162.39
62.74
28.97
8.57
7.67
3.78
3.11
Formaldehyde
Benzene
1,3 -Butadiene
Cadmium, PM
Arsenic, PM
Nickel, PM
Naphthalene
POM, Group la
Ethylbenzene
Hexavalent Chromium, PM
3.37E-03
3.27E-03
1.88E-03
1.37E-03
1.20E-03
1.15E-03
9.85E-04
7.54E-04
6.91E-04
6.78E-04
Formaldehyde 29.07
Acetaldehyde 3.20
Naphthalene 1.25
Winter Park, Florida (Orange County) - ORFL
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
POM, Group la
POM, Group 2b
POM, Group 2d
Tetrachloroethylene
418.04
289.94
284.85
161.98
64.14
29.54
10.73
4.76
3.49
2.91
Hexavalent Chromium, PM
Formaldehyde
Benzene
1,3 -Butadiene
Naphthalene
POM, Group la
Ethylbenzene
POM, Group 2b
Acetaldehyde
Arsenic, PM
5.36E-03
3.77E-03
3.26E-03
1.92E-03
l.OOE-03
9.44E-04
7.12E-04
4.19E-04
3.56E-04
3.49E-04
Formaldehyde 26.68
Acetaldehyde 2.38
-------
VO
ON
O
Table 9-7. 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)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity- Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
Orlando, Florida (Orange County) - PAFL
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
POM, Group la
POM, Group 2b
POM, Group 2d
Tetrachloroethylene
418.04
289.94
284.85
161.98
64.14
29.54
10.73
4.76
3.49
2.91
Hexavalent Chromium, PM
Formaldehyde
Benzene
1,3 -Butadiene
Naphthalene
POM, Group la
Ethylbenzene
POM, Group 2b
Acetaldehyde
Arsenic, PM
5.36E-03
3.77E-03
3.26E-03
1.92E-03
l.OOE-03
9.44E-04
7.12E-04
4.19E-04
3.56E-04
3.49E-04
Arsenic 4.41
-------
Table 9-8. 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)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard
Approximations Based on Annual
Average Concentrations
(Site-Specific)
Noncancer Hazard
Approximation
Pollutant (HQ)
St. Petersburg, Florida (Pinellas County) - AZFL
Toluene
Ethylene glycol
Xylenes
Hexane
Methanol
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
Hydrochloric acid
2,255.54
1,129.96
744.68
740.44
533.81
281.41
179.80
149.82
98.95
87.44
Acrolein
1,3 -Butadiene
Formaldehyde
Acetaldehyde
Benzene
Xylenes
Naphthalene
Lead, PM
Hydrochloric acid
Arsenic, PM
376,906.97
20,283.73
15,287.30
10,994.91
9,380.21
7,446.78
5,976.63
4,943.69
4,371.98
3,633.59
Formaldehyde 0.19
Acetaldehyde 0.16
Pinellas Park, Florida (Pinellas County) - SKFL
Toluene
Ethylene glycol
Xylenes
Hexane
Methanol
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
Hydrochloric acid
2,255.54
1,129.96
744.68
740.44
533.81
281.41
179.80
149.82
98.95
87.44
Acrolein
1,3 -Butadiene
Formaldehyde
Acetaldehyde
Benzene
Xylenes
Naphthalene
Lead, PM
Hydrochloric acid
Arsenic, PM
376,906.97
20,283.73
15,287.30
10,994.91
9,380.21
7,446.78
5,976.63
4,943.69
4,371.98
3,633.59
Formaldehyde 0.27
Acetaldehyde 0.16
Naphthalene 0.03
VO
-------
Table 9-8. 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)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard
Approximations Based on Annual
Average Concentrations
(Site-Specific)
Noncancer Hazard
Approximation
Pollutant (HQ)
Valrico, Florida (Hillsborough County) - SYFL
Toluene
Ethylene glycol
Xylenes
Hexane
Methanol
Benzene
Hydrochloric acid
Ethylbenzene
Formaldehyde
Acetaldehyde
3,156.55
1,555.96
1,077.05
951.56
723.09
419.18
389.70
276.57
259.18
162.39
Acrolein
Cadmium, PM
1,3 -Butadiene
Nickel, PM
Formaldehyde
Hydrochloric acid
Arsenic, PM
Acetaldehyde
Benzene
Manganese, PM
743,682.57
76,142.15
31,371.77
26,715.11
26,447.31
19,484.80
18,554.76
18,043.57
13,972.74
13,932.27
Formaldehyde 0.23
Acetaldehyde 0.16
Naphthalene 0.01
Winter Park, Florida (Orange County) - ORFL
Toluene
Ethylene glycol
Xylenes
Hexane
Methanol
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
Hydrochloric acid
3,175.01
1,451.68
1,148.79
933.11
678.41
418.04
289.94
284.85
161.98
136.30
Acrolein
1,3 -Butadiene
Hexamethylene-l,6-diisocyanate, gas
Formaldehyde
Acetaldehyde
Benzene
Xylenes
Naphthalene
Hydrochloric acid
Arsenic, PM
835,285.94
32,071.75
30,043.31
29,586.10
17,997.24
13,934.55
11,487.90
9,845.69
6,814.94
5,407.85
Formaldehyde 0.21
Acetaldehyde 0.12
VO
ON
to
-------
Table 9-8. 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)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard
Approximations Based on Annual
Average Concentrations
(Site-Specific)
Pollutant
Orlando, Florida (Orange County) - PAFL
Toluene
Ethylene glycol
Xylenes
Hexane
Methanol
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
Hydrochloric acid
3,175.01
1,451.68
1,148.79
933.11
678.41
418.04
289.94
284.85
161.98
136.30
Acrolein
1,3 -Butadiene
Hexamethylene- 1 ,6-diisocyanate, gas
Formaldehyde
Acetaldehyde
Benzene
Xylenes
Naphthalene
Hydrochloric acid
Arsenic, PM
835,285.94
32,071.75
30,043.31
29,586.10
17,997.24
13,934.55
11,487.90
9,845.69
6,814.94
5,407.85
Manganese
Arsenic
Noncancer
Hazard
Approximation
(HQ)
0.07
0.07
-------
Observations from Table 9-7 include the following:
• Benzene, ethylbenzene, and formaldehyde 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 toxi city-weighted
emissions for Pinellas and Hillsborough Counties. 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 six of the highest emitted pollutants in
Hillsborough County also have the highest toxicity-weighted emissions.
• Formaldehyde, which has the highest cancer risk approximations for all sites
sampling carbonyl compounds, is one of the highest emitted pollutants in each county
and has one of the highest toxicity-weighted emissions for each county. This is also
true for acetaldehyde for Pinellas and Orange Counties, but acetaldehyde does not
appear among those pollutants with the highest toxicity-weighted emissions for
Hillsborough County (although it ranks 11th).
• Naphthalene, which is a pollutant of interest for both SFKL and SYFL, is one of the
highest emitted pollutants in both counties and has one of the highest toxicity-
weighted emissions for each county.
• Arsenic is the only pollutant with a cancer risk approximation for PAFL. Arsenic
ranks 10th for toxicity-weighted emissions for Orange County, but is not among the
highest emitted pollutants, indicating the relative toxi city of a low quantity of
emissions. Several metals appear among those with the highest toxicity-weighted
emissions for Hillsborough County, but metals were not sampled for under the NMP
at SYFL.
• POM, Groups la, 2b, and 2d are among the highest emitted "pollutants" in all three
counties and appear among the pollutants with the highest toxicity-weighted
emissions. POM, Group 2b includes several PAHs sampled for at SKFL and SYFL
including acenaphthene and fluorene, both of which failed screens for SKFL but were
not identified as site-specific pollutants of interest. POM, Group 2d also includes
several PAHs sampled for at SKFL and SYFL including phenanthrene and pyrene,
neither of which failed any screens for these sites. POM, Group la does not include
any PAHs sampled for with Method TO-13.
Observations from Table 9-8 include the following:
• Toluene, ethylene glycol, and xylenes are the highest emitted pollutants with
noncancer RfCs in all three Florida counties.
9-64
-------
• Acrolein has the highest toxi city-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.
• Five of the highest emitted pollutants in Pinellas and Orange Counties also have the
highest toxicity-weighted emissions. Four of the highest emitted pollutants in
Hillsborough County also have the highest toxicity-weighted emissions.
• Formaldehyde and acetaldehyde appear on both emissions-based lists for each
site/county. Naphthalene is among the pollutants with the highest toxicity-weighted
emissions for Pinellas and Orange Counties but is not among the highest emitted
(with a noncancer RfC) in any of the three counties.
• Several metals appear among those with the highest toxicity-weighted emissions for
Hillsborough County, but are not among the highest emitted. Metals were not
sampled for at SYFL under the NMP.
• Arsenic is the only metal that appears among the pollutants with the highest toxicity-
weighted emissions for Orange County (ranking 10th). There are no metals among
the highest emitted pollutants in Orange County.
9.6 Summary of the 2012 Monitoring Data for the Florida Monitoring Sites
Results from several of the data treatments described in this section include the
following:
»«» Acetaldehyde and'formaldehyde failedscreens for AZFL andORFL, where only
carbonyl compounds were sampled. Eight pollutants (three carbonyls, four PAHs,
and hexavalent chromium) failed screens for SKFL. Five pollutants (three carbonyls
and two PAHs) failed screens for SYFL. Arsenic, manganese, and lead failed screens
forPAFL.
»«» Formaldehyde had the highest annual average concentration for each of the Florida
sites where carbonyl compounds were sampled. The annual average concentration of
naphthalene for SKFL was more than twice the annual average concentration for
SYFL, the two sites where naphthalene was a pollutant of interest. Manganese had
the highest annual average concentration of the metals identified as pollutants of
interest for PAFL.
»«» Concentrations of formaldehyde have an overall decreasing trend at ORFL. A similar
trend in formaldehyde concentrations is shown at SKFL until recent years where an
increasing trend is shown. Concentrations of acetaldehyde decreased significantly
between 2010 and 2012 at AZFL and SKFL with a significant decrease also shown at
ORFL from 2011 and 2012. Conversely, acetaldehyde concentrations at SYFL
increased significantly from 2011 to 2012. Concentrations of naphthalene have not
changed significantly at SKFL or SYFL. Both arsenic and manganese exhibit
increases at PAFL from 2011 to 2012.
9-65
-------
10.0 Site in Georgia
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS site in Georgia, 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 SDGA 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 SDGA monitoring site is located in Decatur, Georgia, a suburb of Atlanta.
Figure 10-1 is a composite satellite image retrieved from ArcGIS Explorer showing the
monitoring site and its immediate surroundings. Figure 10-2 identifies nearby point source
emissions locations by source category, as reported in the 2011 NEI for point sources. Note that
only sources within 10 miles of the site are included in the facility counts provided in
Figure 10-2. 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 radius are still visible on the map, but have been grayed out in order
to show emissions sources just outside the boundary. Table 10-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates.
10-1
-------
Figure 10-1. Decatur, Georgia (SDGA) Monitoring Site
-------
Figure 10-2. NEI Point Sources Located Within 10 Miles of SDGA
Legend
84020'(rW 84°15'0"W 84"10'OnW
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
SDGA NATTS site
10 mile radius
County boundary
Source Category Group (No. of Facilities)
1T Airport/Airline/Airport Support Operations (21)
A Animal Feedlot or Farm (1)
3»t Battery Manufacturing (1)
C Chemical Manufacturing (2)
i Compressor Station (1)
F Food Processing/Agriculture (3)
W Glass Plant (1)
O Institution (school, hospital, prison, etc.) (1)
• Landfill (3)
j] Paint and Coating Manufacturing (2)
R Plastic, Resin, or Rubber Products Plant (3)
P Printing/Publishing/Paper Product Manufacturing (1}
X Rail Yard/Rail Line Operations (3)
10-3
-------
Table 10-1. Geographical Information for the Georgia Monitoring Site
Site
Code
SDGA
AQS Code
13-089-0002
Location
Decatur
County
DeKalb
Micro- or
Metropolitan
Statistical Area
Atlanta-Sandy
Springs-Roswell,
GA
Latitude
and
Longitude
33.68797,
-84.29048
Land Use
Residential
Location
Setting
Suburban
Additional Ambient Monitoring Information1
CO, SO2, NOy, NO, NO2, NOX, PAMS, Carbonyl
compounds, VOCs, O3, Meteorological parameters,
PMio, PM Coarse, PM10 Speciation, Black carbon,
PM2 5, and PM2 5 Speciation, Haze, IMPROVE
Speciation.
BOLD ITALICS = EPA-designated NATTS Site
-------
SDGA is located on the DeKalb County Schools Environmental Education property off
Wildcat Road and is the South DeKalb NATTS site. Residential subdivisions, a greenhouse and
horse barn, an athletic field, and a high school surround the monitoring site. A golf course backs
up against the school property on the south and east sides. Interstate-285 is located less than
1 mile north of the site, as shown in Figure 10-1. As Figure 10-2 shows, only one point source (a
food processing facility) is located in close proximity to SDGA. Additional sources are located
primarily on the west side of the 10-mile radius. 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, is the source category with the greatest number of emissions
sources within 10 miles of SDGA.
Table 10-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Georgia monitoring site. Table 10-2 includes both county-level
population and vehicle registration information. Table 10-2 also contains traffic volume
information for SDGA as well as the location for which the traffic volume was obtained.
Additionally, Table 10-2 presents the county-level daily VMT for DeKalb County.
Table 10-2. Population, Motor Vehicle, and Traffic Information for the Georgia Monitoring Site
Site
SDGA
Estimated
County
Population1
707,089
County-level
Vehicle
Registration2
472,535
Annual
Average
Daily Traffic3
141,980
Intersection
Used for
Traffic Data
1-285, north of Clifton Spring Rd
County-
level Daily
VMT4
20,113,000
1 County-level population estimate reflects 2012 data (Census Bureau, 2013c)
2County-level vehicle registration reflects 2011 data (GA DOR, 2011)
3AADT reflects 2012 data from the Georgia DOT (GA DOT, 2012a)
4County-level VMT reflects 2012 data (GA DOT, 2012b)
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 10-2 include the following:
• SDGA's county-level population and vehicle registration are in the middle of the
range compared to other counties with NMP sites.
• The traffic volume experienced near SDGA ranks ninth highest compared to other
NMP sites. The traffic estimate provided is for 1-285, north of Clifton Spring Road.
• The daily VMT for DeKalb County is in the middle third compared to other counties
with NMP sites (where VMT data were available).
10-5
-------
10.2 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
site in Georgia on sample days, as well as over the course of the year.
10.2.1 Climate Summary
Atlanta is the largest city in Georgia and is located at the base of the Blue Ridge
Mountains. The Gulf of Mexico to the south is the major moisture source for weather systems
that move across the region. Both topographical features, in addition to the Atlantic Ocean to the
east, exert moderating influences on the area's climate, tempering cold air outbreaks from the
north as well as summer heat waves. Summers are warm and humid while winters are relatively
mild, although snow is not uncommon. The semi-permanent Bermuda High Pressure offshore
over the Atlantic Ocean is a dominant weather feature affecting the Atlanta area, which pulls
warm, moist air into the region. Precipitation is ample throughout the year, although autumn is
the driest season. Westerly and northwesterly winds prevail throughout much of the year,
although east winds are more common in the late summer and fall (Wood, 2004; GSCO, 1998;
NCDC, 2014).
10.2.2 Meteorological Summary
Hourly meteorological data for 2012 were retrieved from NCDC for the weather station
closest to the Georgia monitoring site (NCDC, 2012), as described in Section 3.5.2. The closest
weather station to SDGA is located at W. B. Hartsfield/Atlanta International Airport (WBAN
13874). Additional information about the Hartsfi eld weather station, such as the distance
between the site and the weather station, is provided in Table 10-3. These data were used to
determine how meteorological conditions on sample days vary from conditions experienced
throughout the year.
10-6
-------
Table 10-3. Average Meteorological Conditions near the Georgia Monitoring Site
Closest Weather
Station (WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Decatur, Georgia - SDGA
W.B.
Hartsfield/ Atlanta
Intl. Airport
13874
(33.64, -84.43)
9.2
miles
237°
(WSW)
Sample
Days
(61)
2012
73.8
±3.3
74.4
±1.4
64.8
±3.3
65.0
±1.3
50.8
±3.9
51.0
±1.5
57.2
±3.2
57.3
±1.3
63.7
±3.6
63.7
±1.4
1018.1
±1.5
1017.6
±0.5
6.4
±0.7
6.3
±0.3
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
-------
Table 10-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for all of 2012. Also included in Table 10-3 is the 95 percent
confidence interval for each parameter. As shown in Table 10-3, average meteorological
conditions on sample days near SDGA were representative of average weather conditions
experienced throughout the year.
10.2.3 Back Trajectory Analysis
Figure 10-3 is the composite back trajectory map for days on which samples were
collected at the SDGA monitoring site. Included in Figure 10-3 are four back trajectories per
sample day. Figure 10-4 is the corresponding cluster analysis. An in-depth description of these
maps and how they were generated is presented in Section 3.5.2.1. For the composite map, each
line represents the 24-hour trajectory along which a parcel of air traveled toward the monitoring
site on a given sample day and time, based on an initial height of 50 meters AGL. For the cluster
analysis, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric circle around the site in Figures 10-3 and 10-4 represents 100 miles.
Figure 10-3. Composite Back Trajectory Map for SDGA
10-8
-------
Figure 10-4. Back Trajectory Cluster Map for SDGA
Observations from Figures 10-3 and 10-4 include the following:
• The composite back trajectory map for SDGA looks like a pinwheel, indicating that
back trajectories originated from a variety of directions around SDGA. Back
trajectories originating from the northwest and north tended to be longer than those
originating from other directions.
• Size-wise, the 24-hour air shed domain for SDGA is in the bottom-third compared to
other NMP monitoring sites. While the farthest away a back trajectory originated was
central Illinois, or greater than 450 miles away, the average back trajectory length is
165 miles. Three-quarters of back trajectories originated within 200 miles of SDGA
and greater than 90 percent of back trajectories originated within 300 miles of the
site. The four longest back trajectories originated over Illinois and represent a single
sample day (October 30, 2012).
• The cluster analysis shows that 26 percent of back trajectories originated to the west,
northwest, and north of SDGA and are generally less than 200 miles in length.
Another 25 percent of back trajectories originated to the south of SDGA over central
and southeast Georgia. The cluster trajectory originating over upstate South Carolina
represents both shorter back trajectories originating to the northeast and east of
SDGA over Georgia and South Carolina as well as longer trajectories originating over
the mountains of North Carolina, Tennessee, Virginia, and West Virginia. Fifteen
percent of back trajectories originated along Georgia's western border or the
southeast portion of Alabama. Twelve percent of back trajectories originated from
the northwest to north of SDGA, over Tennessee, Kentucky, Indiana, or Illinois.
10-9
-------
10.2.4 Wind Rose Comparison
Hourly surface wind data from the weather station at Hartsfield International Airport near
SDGA were uploaded into a wind rose software program to produce customized wind roses, as
described in Section 3.5.2.2. 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 10-5 presents a map showing the distance between the weather station and SDGA,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 10-5 also presents three different wind roses for the
SDGA monitoring site. First, a historical wind rose representing 2002 to 2011 wind data is
presented, which shows the predominant surface wind speed and direction over an extended
period of time. Second, a wind rose representing wind observations for all of 2012 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2012 is
presented. These can be used to identify the predominant wind speed and direction for 2012 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.
Observations from Figure 10-5 for SDGA include the following:
• The weather station at Hartsfield International Airport is the closest weather station to
SDGA and is located 9.2 miles west-southwest of SDGA.
• The historical wind rose shows that winds from the west to north-northwest account
for nearly 40 percent of wind observations. Easterly winds were also common. Winds
from the northeast quadrant were rarely observed. Calm winds (< 2 knots) were
observed for less than 10 percent of the hourly wind measurements.
• The wind patterns on the full-year wind rose are similar to those of the historical wind
rose. The reduced percentage of wind observations from the west to northwest and
east are accounted for in the increased percentage of calm winds.
• Although the predominant wind patterns on the sample day wind rose still resemble
those on the full-year wind rose, there are additional differences. Further decreases in
the percentage of wind observations from the west to north-northwest are shown but
are accompanied by increases in the percentage winds from the southeast quadrant.
10-10
-------
Figure 10-5. Wind Roses for the Hartsfield International Airport Weather Station near
SDGA
Location of SDGA and Weather Station
2002-2011 Historical Wind Rose
2012 Wind Rose
Sample Day Wind Rose
Calms: 10.79%
10-11
-------
10.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for the Georgia
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 10-4.
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 10-4. It is
important to note which pollutants were sampled for at each site when reviewing the results of
this analysis. SDGA sampled for PAHs and hexavalent chromium only, although the sampling of
PAHs was discontinued at SDGA at the end of June 2012.
Table 10-4. Risk-Based Screening Results for the Georgia Monitoring Site
Pollutant
Screening
Value
(Ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Decatur, Georgia - SDGA
Naphthalene
Benzo(a)pyrene
Hexavalent Chromium
0.029
0.00057
0.000083
Total
29
1
1
31
29
16
32
77
100.00
6.25
3.13
40.26
93.55
3.23
3.23
93.55
96.77
100.00
Observations from Table 10-4 for SDGA include the following:
• Three pollutants failed at least one screen for SDGA: naphthalene, benzo(a)pyrene,
and hexavalent chromium.
• Naphthalene failed 100 percent of its screens, accounting for 29 of the 31 total failed
screens (or roughly 94 percent); the other two pollutants failed only one screen each.
• Although naphthalene and benzo(a)pyrene together account for more than 95 percent
of the total failed screens for SDGA and are therefore identified as pollutants of
interest, hexavalent chromium failed the same number of screens as benzo(a)pyrene;
thus, hexavalent chromium was also added as pollutants of interest for SDGA, per the
procedure described in Section 3.2.
10-12
-------
10.4 Concentrations
This section presents various concentration averages used to characterize pollution levels
at the Georgia 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 the
monitoring site.
• Annual concentration averages are presented graphically for SDGA 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 the site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Additional site-specific statistical summaries for SDGA
are provided in Appendices M and O.
10.4.1 2012 Concentration Averages
Quarterly and annual concentration averages were calculated for the pollutants of interest
for SDGA, as described in Section 3.1. The quarterly average 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 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 Georgia monitoring site are
presented in Table 10-5, 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.
10-13
-------
Table 10-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Georgia Monitoring Site
Pollutant
#of
Measured
Detections
vs. # 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)
Decatur, Georgia - SDGA
Benzo(a)pyrene
Hexavalent Chromium
Naphthalene
16/29
32/54
29/29
0.08
±0.05
0.01
±0.01
92.21
± 18.02
0.05
±0.09
0.02
±0.01
105.82
± 27.28
NA
0.01
±0.01
NA
NA
0.01
±0.01
NA
NA
0.01
±<0.01
NA
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
Observations for SDGA from Table 10-5 include the following:
• Naphthalene was detected in every PAH sample collected at SDGA while
benzo(a)pyrene was detected in just greater than 50 percent of the samples collected.
The detection rate of benzo(a)pyrene was significantly higher in the first quarter than
the second. There were 12 measured detections and three non-detects for the first
quarter of 2012 while there were four measured detections and 10 non-detects for the
second quarter.
• Third and fourth quarter average concentrations could not be calculated for these two
pollutants because sampling was discontinued at the end of June 2012. As a result,
annual averages could not be calculated either.
• The second quarter average concentration of naphthalene is higher than the first
quarter average concentration, although not statistically so. The two highest
concentrations of naphthalene measured at SDGA were both measured in June
(180 ng/m3 and 183 ng/m3). Aside from these two measurements, the concentrations
measured during the first quarter are similar to those measured during the second
quarter.
• Hexavalent chromium was detected in nearly 60 percent of the samples collected at
SDGA and ranged from 0.0054 ng/m3 to 0.0954 ng/m3.
10.4.2 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. Figure 10-6 overlays the site's minimum, annual
average, and maximum hexavalent chromium concentrations onto the program-level minimum,
first quartile, median, average, third quartile, and maximum concentrations, as described in
10-14
-------
Section 3.5.3. Box plots were not created for the PAHs because annual average concentrations
could not be calculated due to the short sampling duration.
Figure 10-6. Program vs. Site-Specific Average Hexavalent Chromium Concentration
SDGA
t
Program Max Concentration = 8.51 ng/m3
i J
0.1
0.2 0.3
Concentration (ng/m3)
0.4
0.5
Program:
Site:
IstQuartile
D
Site Average
o
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
Observations from Figure 10-6 include the following:
• Figure 10-6 is the box plot for hexavalent chromium. Note that the program-level
maximum concentration (8.51 ng/m3) is not shown directly on the box plot
because the scale of the box plot would be too large to readily observe data points
at the lower end of the concentration range. Thus, the scale of the box plot has
been reduced to 0.5 ng/m3. In addition, the program-level first quartile is zero and
therefore not visible on the box plot.
• Figure 10-6 shows that the annual average concentration of hexavalent chromium
for SDGA is less than the program-level average concentration. SDGA's annual
average concentration is also less than the program-level median concentration.
The maximum concentration measured at SDGA is two orders of magnitude less
than the program-level maximum concentration.
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 3.5.3.2.
SDGA has sampled hexavalent chromium under the NMP since 2005 and PAHs since 2007.
Thus, Figures 10-7 through 10-9 present the 1-year statistical metrics for each of the pollutants of
interest for SDGA. The statistical metrics presented for assessing trends include the substitution
of zeros for non-detects. If sampling began (or ended) mid-year, a minimum of 6 months of
sampling is required for inclusion in the trends analysis; in these cases, a 1-year average is not
provided, although the range and quartiles are still presented.
10-15
-------
Figure 10-7. Yearly Statistical Metrics for Benzo(a)pyrene Concentrations Measured at SDGA
I
I0'6
O 5th Percentile
• 95th Percentile
1A 1-year average is not presented because sampling under the NMP did not begin until April 2007.
2 A 1-year average is not presented because sampling under the NMP was discontinued in June 2012.
Observations from Figure 10-7 for benzo(a)pyrene measurements collected at SDGA
include the following:
• Sampling under the NMP for PAHs began in April 2007 at SDGA. However, a
1-year average is not presented for 2007 because a full year's worth of data is not
available, although the range of measurements is provided. In addition, a 1-year
average is not provided for 2012 due to the discontinuation of sampling in June
2012.
• Only one benzo(a)pyrene concentration measured at SDGA is greater than
1 ng/m3, which was measured in 2010 (1.01 ng/m3). The next highest
concentration was measured in 2009 and was nearly half as high (0.597 ng/m3).
• The minimum, 5th percentile, and/or the median concentration has been equal to
zero for each year of sampling, indicating the presence of non-detects. The
number of non-detects was at a minimum in 2007, but ranged from 40 percent to
60 percent for the other years of sampling.
• The difference between the 5th and 95th percentiles is highest for 2007 then
decreased by more than half for 2008, after which an increasing trend is shown.
10-16
-------
Figure 10-8. Yearly Statistical Metrics for Hexavalent Chromium Concentrations
Measured at SDGA
0 5th Percentile — Minimum ~ Median — Maximum • 95th Percentile
" Average
A 1-year average is not presented because sampling under the NMP did not begin until February 2005.
21-Year averages are not presented because there was a break in sampling between Sept 2007 and May
2008.
Observations from Figure 10-8 for hexavalent chromium measurements collected at
SDGA include the following:
• Although hexavalent chromium sampling under the NMP began in 2005 at
SDGA, a 1-year average is not presented because a full year's worth of data is not
available, although the range of measurements is provided. In addition, there was
a break in sampling between September 2007 and May 2008 due to sampler
issues; as a result, a 1-year average is not provided for 2007 or 2008.
• The maximum concentration was measured in 2006 (0.300 ng/m3). Only four
additional concentrations greater than 0.1 ng/m3 have been measured at SDGA,
all of which were measured in either 2005 or 2006.
• The difference between the 5th and 95th percentiles exhibits little change over the
last several years of sampling, indicating that a majority of the measurements fall
within roughly the same range, at least since 2007.
• The median concentration decreased significantly between 2006 and 2009,
reaching a minimum of zero for 2009, which indicates that at least half of the
measurements were non-detects. Since 2009, the number of non-detects has
varied from 23 percent (2011) to 41 percent (2012). The increase in the measured
10-17
-------
detections from 2009 to 2010 explains the increase in the 1-year average, despite
the lower maximum concentration and 95th percentile.
Figure 10-9. Yearly Statistical Metrics for Naphthalene Concentrations Measured at SDGA
30
„
.c.
c
.2
1
u
T
—
LrJ
r
o
U
I
....•o--
1
— * —
..-.<>....
^^m
U
'"••...
— I
j —
0
-r
I
•
2007 l 2008 2009 2010 2011 2012 2
Year
• 5th Percentile — Min mum ~ Median — Maximum O 95th Percentile ...^... Average
2 A 1-year average is not presented because sampling under the NMP was discontinued in June 2012.
Observations from Figure 10-9 for naphthalene measurements collected at SDGA include
the following:
• Three naphthalene concentrations greater than 300 ng/m3 have been measured at
SDGA, two in 2010 (322 ng/m3 and 301 ng/m3) and one in 2008 (318 ng/m3). Ten
of the 18 concentrations greater than 250 ng/m3 were measured in 2010 (with two
measured in 2008, four in 2009, and two in 2011).
• The difference between the 5th and 95th percentiles increases significantly from
2007 to 2010, more than doubling over the 4-year period. This indicates that a
majority of the concentrations are falling into a wider range of measurements each
year. This range decreases for 2011 and again for 2012.
• The 1-year average concentration increases from 2008 through 2010, then
decreases for 2011. The median changes little from 2008 to 2009, then follows a
pattern similar to the 1-year average for 2010 and 2011.
10-18
-------
• Although the maximum and 95th percentile both decrease for 2012, the median
concentration exhibits an increase. This is due to the increase in magnitude for the
concentrations at the lower end of the concentration range. While 13
concentrations measured were less than 35 ng/m3 in 2011, there were none in
2012. The number of concentrations between 35 ng/m3 and 75 ng/m3 decreased by
half from 2011 (16) to 2012 (8). Recall, however, that 2012 includes only six
months of sampling.
10.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk at the
SDGA monitoring site. Refer to Sections 3.3 and 3.5.3 for definitions and explanations regarding
the various toxicity factors, time frames, and calculations associated with these risk-based
screenings.
10.5.1 Risk-Based Screening Assessment Using MRLs
A risk-based screening was conducted by comparing the concentration data from the
Georgia monitoring site to the ATSDRMRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 day to 14 days); intermediate (exposures of 15 days to 364 days); and chronic
(exposures of 1 year or greater). The preprocessed daily measurements of the pollutants of
interest were compared to the acute MRLs; the quarterly averages were compared to the
intermediate MRLs; and the annual averages were compared to the chronic MRLs.
As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2012.
10.5.2 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for SDGA 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 3.5.3.4 for an explanation of how cancer risk and noncancer hazard
approximations are calculated and what limitations are associated with them. Annual averages,
10-19
-------
cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard approximations are
presented in Table 10-6, where applicable. Cancer risk approximations are presented as
probabilities while the noncancer hazard approximations are ratios and thus, unitless values.
Table 10-6. Risk Approximations for the Georgia Monitoring Site
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
Decatur, Georgia - SDGA
Benzo(a)pyrene
Hexavalent Chromium
Naphthalene
0.00176
0.012
0.000034
0.0001
0.003
16/29
32/54
29/29
NA
0.01
±<0.01
NA
NA
0.16
NA
NA
<0.01
NA
NA = Not available due to the criteria for calculating an annual average.
- = A Cancer URE or Noncancer RfC is not available.
Observations for SDGA from Table 10-6 include the following:
• The cancer risk approximation for hexavalent chromium is 0.16 in-a-million,
considerably less than a level of concern.
• The noncancer hazard approximation for hexavalent chromium is significantly less
than 1.0, indicating that no adverse health effects are expected from this individual
pollutant.
• Cancer risk and noncancer hazard approximations could not be calculated for
benzo(a)pyrene or naphthalene because annual averages are not available.
10.5.3 Risk-Based Emissions Assessment
In addition to the risk-based screenings discussed above, Tables 10-7 and 10-8 present an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 10-7 presents the 10 pollutants with the highest emissions from the 2011 NEI that have
cancer toxicity factors. Table 10-7 also presents the 10 pollutants with the highest toxicity -
weighted emissions, based on the weighting schema described in Section 3.5.3.5. Lastly,
Table 10-7 provides the 10 pollutants with the highest cancer risk approximations (in-a-million)
for the site, as presented in Table 10-6. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 10-7. Table 10-8 presents similar
information, but is limited to those pollutants with noncancer toxicity factors.
10-20
-------
Table 10-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Georgia Monitoring Site
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
Decatur, Georgia (DeKalb County) - SDGA
Tetrachloroethylene
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Naphthalene
Trichloroethylene
Dichloromethane
POM, Group 2b
2,565.31
167.19
117.84
115.79
74.29
27.20
13.97
2.32
2.21
1.93
Formaldehyde
Benzene
1,3 -Butadiene
Tetrachloroethylene
Naphthalene
Ethylbenzene
POM, Group 2b
POM, Group 2d
Acetaldehyde
Arsenic, PM
1.51E-03
1.30E-03
8.16E-04
6.67E-04
4.75E-04
2.95E-04
1.70E-04
1.66E-04
1.63E-04
1.53E-04
Hexavalent Chromium 0.16
o
to
-------
Table 10-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Georgia Monitoring Site
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Noncancer
Hazard
Approximation
Pollutant (HQ)
Decatur, Georgia (DeKalb County) - SDGA
Tetrachloroethylene
Toluene
Ethylene glycol
Hexane
Xylenes
Methanol
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
2,565.31
1,551.83
851.66
472.63
436.95
395.07
167.19
117.84
115.79
74.29
Acrolein
Tetrachloroethylene
1,3 -Butadiene
Formaldehyde
Acetaldehyde
Benzene
Naphthalene
Xylenes
Lead, PM
Arsenic, PM
291,469.12
64,132.78
13,598.64
11,815.64
8,254.75
5,573.13
4,656.58
4,369.46
3,306.94
2,377.39
Hexavalent Chromium <0.01
to
to
-------
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 3.5.3.5. Similar to the cancer risk and
noncancer hazard approximations provided in Section 10.5.2, this analysis may help policy-
makers prioritize their air monitoring activities.
Observations from Table 10-7 include the following:
• Tetrachloroethylene, benzene, and ethylbenzene are the highest emitted pollutants
with cancer UREs in DeKalb County, although the tetrachloroethylene emissions are
significantly higher than the emissions of the other two pollutants.
• The pollutants with the highest toxicity-weighted emissions (of the pollutants with
cancer UREs) are formaldehyde, benzene, 1,3-butadiene, and tetrachloroethylene.
• Eight of the highest emitted pollutants also have the highest toxicity-weighted
emissions for DeKalb County.
• Naphthalene, one of the pollutants of interest for SDGA, has the fifth highest toxicity-
weighted emissions and seventh highest emissions for DeKalb County.
• Hexavalent chromium is not among the highest emitted pollutants in DeKalb County
nor is it among those with the highest toxicity-weighted emissions. Benzo(a)pyrene is
part of POM, Group 5a. POM, Group 5a does not appear on either emissions-based
list in Table 10-7.
• POM, Group 2b is the tenth highest emitted "pollutant" in DeKalb County and ranks
seventh for toxicity-weighted emissions. POM, Group 2b includes several PAHs
sampled for at SDGA including acenaphthene, benzo(e)pyrene, fluorene, and
perylene, although none of these pollutants failed screens for SDGA. POM, Group 2d
ranks eighth for toxicity-weighted emissions and includes three PAHs sampled for at
SDGA (anthracene, phenanthrene, and pyrene). None of these pollutants failed
screens either.
Observations from Table 10-8 include the following:
• Tetrachloroethylene is the highest emitted pollutant with a noncancer RfC in DeKalb
County, followed by toluene and ethylene glycol.
10-23
-------
• The pollutants with the highest toxi city-weighted emissions (of the pollutants with
noncancer RfCs) are acrolein, tetrachloroethylene, and 1,3-butadiene.
• Five of the highest emitted pollutants in DeKalb County also have the highest
toxicity-weighted emissions.
• While naphthalene is not one of the 10 highest emitted pollutants with a noncancer
RfC in DeKalb County, its toxicity-weighted emissions rank seventh. Hexavalent
chromium does not appear on either emissions-based list; nor does POM, Group 5a.
10.6 Summary of the 2012 Monitoring Data for SDGA
Results from several of the data treatments described in this section include the
following:
»«» Naphthalene, hexavalent chromium, and benzo(a)pyrene failed screens for SDGA,
although naphthalene accounted for the majority of failed screens.
»«» PAH sampling was discontinued at SDGA at the end of June 2012.
»«» Concentrations of hexavalent chromium have not changed significantly at SDGA over
the last few years of sampling.
10-24
-------
11.0 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.
11.1 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 MSA. Figures 11-1 and 11-2 are composite satellite images
retrieved from ArcGIS Explorer showing the Chicago monitoring sites and their immediate
surroundings. Figure 11-3 identifies the nearby point source emissions locations by source
category, as reported in the 2011 NEI for point sources, for NBIL and SPIL. Note that only
sources within 10 miles of the sites are included in the facility counts provided in Figure 11-3.
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 radii are still visible on the map, but have been grayed out in order
to show emissions sources just outside the boundary. Figures 11-4 and 11-5 are the composite
satellite image and facility map for ROIL, respectively. Table 11-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates for each
site.
11-1
-------
Figure 11-1. Northbrook, Illinois (NBIL) Monitoring Site
-------
Figure 11-2. Schiller Park, Illinois (SPIL) Monitoring Site
-------
Figure 11-3. NEI Point Sources Located Within 10 Miles of NBIL and SPIL
87"55'Q"vV 87'50'CTW 87'45'0"W 87°40'0"W 87"3510"W
Legend
88"0'Q"W 87|;55'0"W B7"50'0"W 87°45'0"W 87G40'0"W 87°35'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
NBIL NATTS site
SPIL UATMP site
10 mile radius
Source Category Group (No. of Facilities)
•f Airporf Airline/Airport Support Operations (31)
£ Asphalt Production/Hot Mix Asphalt Plant (9)
0 Auto Body Shop/Painters/Automotive Stores (1)
£3 Automobile/Truck Manufacturing <6)
££ AutomoBve/RV Dealership (1)
i " Brick, Structural Clay, or Clay Ceramics Plant (1)
f Building/Construction (A)
B Bulk Terminals/Bulk Plants (11)
C Chemical Manufacturing (35)
i Compressor Station (7)
)X] Crematory -Animal/Human [20]
0 Dry Cleaning (63)
6 Etectncal Equipment Manufacturing (39)
^ Electricity Generation via Combustion (7)
E Electroplating. Plating, Polishing, Anodizing, and Coloring (57)
F Food Processlng/Agncullure (45)
I Foundries, Iron and Steel (2)
A. Foundries, Non-ferrous (16)
H* Gasoline/Diesel Service Station (2)
$• Glass Plant (3)
> Hole Is/Mote Is/Lodging (2)
-^f- Industrial Machinery or Equipment Plant (31)
O Institution (school, hospital, prison, etc.) (47)
• Landfill (8)
(||j Metal Can, Box, and Other Metal Container Ma
A Metal Coating. Engraving, and Allied Services t
(•) Metals Processing/Fabrication (77)
X Mine'Quarry/Mmeral Processing (38)
^ Miscellaneous Commercial/Industrial (103)
• Oil and/or Gas Production (5)
h Pairrt and Coating Manufacturing (11)
lufacturing (2)
i Manufacturers (25)
County boundary
c i Pharmaceutical Manufacturing (7)
R Plastic. Resin, or Rubber Products Plant (34)
^ Printing, Coating & Dyeing of Fabrics (2)
P Printing/Publisuing/Paper Product Manufacturing (84)
H Pulp aid Paper Plant (1)
X Rail YardJRail Line Operations (6)
^ Steel Mill (3)
TT Telecommunications/Radio (21)
^ Testing Laboratories (1)
T Texiite. Yam, or Carpet Plant (1)
M Tobacco Manufactunng (1)
[CJ Utilities/Pipeline Constructor (2)
I Wastewater Treatment (6)
4 Water Treatment (4)
W Woodwork, Furniture. Mlllwork & Wood Preserving (11)
11-4
-------
Figure 11-4. Roxana, Illinois (ROIL) Monitoring Site
-------
Figure 11-5. NEI Point Sources Located Within 10 Miles of ROIL
QO'IS'CTW 9CMO'0"W 90"5'0"W 90'W'W 89"55'0"W
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 Q 10 mile radius
County boundary
Source Category Group (No. of Facilities)
>%t Aerospace/Aircraft Manufacturing (1)
^ AirponVAirliiWAIrport Support Operations (27)
£ Asphalt Production/Hot Mix Asphalt Plant (9)
Y Brewenes/Distilleries/Wineries (1)
_ .. BncK, Structural Clay, or Clay Ceramics Plant (1)
f BuildtngJConstruclion (1)
B Bulk Terminals; Bulk Plants (20)
C Chemical Manufacturing (28)
(3 Coke Battery (2)
1 Compressor Station (9)
JX] CrematOfy-Animal/Human (3)
(D Dry Clearing (4)
© Electncal Equipment Manufacturing (1)
f Electricity Generation via Combustion (7)
E Electroplating. Plating. Polishing. Anodizing, and Colonng (3)
=)j= Etnanol Biorefinenes (1)
Food Processing'Agriculture (10)
Foundries, Iron and Steel (2)
Foundries. Non-ferrous (2)
Gasoline/Diesel Service Station (2)
Industrial Machinery or Equipment Plant (4)
Institution (school, hospital, prison, etc.) (13)
Landfill (4)
Leather and Leather Products (1)
Metal Can, Box, and Other Metal Container Manufacturing (1)
Metal Coating. Engraving, and Allied Services to Manufacturers (4)
Metals Processing/Fabrication (12)
Military Base/National Security (i)
Mine/Quarry/Mmeral Processing (30)
Miscellaneous Commercial/Industnal (43)
Pesticide Manufacturing Plant (2)
Petroleum Products Manufacturing (2)
Petroleum Refinery (1)
Pharmaceutical Manufacturing (1)
Plastic, Resin, or Rubber Products Plant (3)
Port and Harbor Operations (8!
Printing/Publishing'Paper Product Manufacturing (7)
Pulp and Paper Plant (1)
Rail Yard/Rail Line Operations (15)
Railroad Engines/Parts Manufacturing (1)
Steel Mill (3)
Textile. Yarn, or Carpel Plant (1)
Truck/Bus/Transportation Operations (1)
Wastewater Treatment (8)
Water Treatment (1)
11-6
-------
Table 11-1. Geographical Information for the Illinois Monitoring Sites
Site
Code
NBIL
SPIL
ROIL
AQS Code
17-031-4201
17-031-3103
17-119-9010
Location
Northbrook
Schiller
Park
Roxana
County
Cook
County
Cook
County
Madison
County
Micro- or
Metropolitan
Statistical Area
Chicago-
Naperville-Elgin
IL-IN-WI MSA
Chicago-
Naperville-Elgin
IL-IN-WI MSA
St. Louis, MO-IL
MSA
Latitude
and
Longitude
42.139996,
-87.799227
41.965193,
-87.876265
38.848382,
-90.076413
Land Use
Residential
Mobile
Industrial
Location
Setting
Suburban
Suburban
Suburban
Additional Ambient Monitoring Information1
TSP, TSP Metals, CO, SO2, NO, NO2, NOX, NOy, O3,
Meteorological parameters, PM10, PM25, PM2 5
Speciation, IMPROVE Speciation.
TSP, TSP Metals, CO, NO, NO2, NOX, Meteorological
parameters, PM2 5.
None.
BOLD ITALICS = EPA-designated NATTS Site
-------
NBIL is located on the property of the Northbrook Water Filtration Station. Figure 11-1
shows that NBIL is located off State Highway 68 (Dundee Road), near Exit 30 on 1-94 (the
clover leaf of which is located on the lower right hand side of Figure 11-1). A railway runs
north-south in front of the water filtration station and intersects Dundee Readjust south of the
monitoring site. The surrounding area is classified as suburban and residential. Commercial,
residential, and forested areas are nearby, as well as 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 1/2 mile
from the site. The surrounding area is classified as suburban and mobile. Commercial and
residential areas are nearby and a railyard is located to the east of 1-294.
NBIL and SPIL are located within approximately 12 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 11-3. 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 11-3; 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 classified as industrial, a residential area is wedged between the industrial
properties, as Figure 11-4 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.
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 11-5. The Mississippi River, which is the
border between Missouri and Illinois, is just over a mile and a half west of the monitoring site.
11-8
-------
In addition to showing the ROIL monitoring site's location relative to the S4MO
monitoring site, Figure 11-5 also shows that there is a large cluster of emissions sources
surrounding and mostly to the south of ROIL. 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 the site. Other nearby sources include a rail yard, an
industrial machinery/equipment facility, and several chemical manufacturers.
Table 11-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Illinois monitoring sites. Table 11-2 includes both county-level
population and vehicle registration information. Table 11-2 also contains traffic volume
information for each site as well as the location for which the traffic volume was obtained.
Additionally, Table 11-2 presents the county-level daily VMT for Cook County and Madison
County.
Table 11-2. Population, Motor Vehicle, and Traffic Information for the Illinois Monitoring
Sites
Site
NBIL
SPIL
ROIL
Estimated
County
Population1
5,231,351
267,883
County-level
Vehicle
Registration2
2,092,085
286,043
Annual
Average Daily
Traffic3
115,100
191,700
9,400
Intersection
Used for
Traffic Data
1-94, north of intersection with
Dundee Rd
1-294 at Lawrence Ave
Route 1 1 1 at railroad tracks, where
Rt 1 1 1 becomes S. Central Ave
County-
level Daily
VMT4
86,217,829
7,867,318
Bounty-level population estimates reflect 2012 data (Census Bureau, 2013c).
2County-level vehicle registration reflects 2012 data (IL SOS, 2012).
3AADT reflects 2011 data for SPIL and ROIL and 2012 data for NBIL (IL DOT, 2011/2012).
4County-level VMT reflects 2012 data (IL DOT, 2012).
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 11-2 include the following:
• Cook County has the second highest county-level population (behind Los Angeles
County) and fourth highest county-level vehicle registration (behind Los Angeles
County, CA; Maricopa County, AZ; and Harris County, TX) compared to other
counties with NMP sites.
• Both the county-level population and vehicle registration for Madison County are an
order of magnitude less than Cook County and are in the middle of the range
compared to other counties with NMP sites.
11-9
-------
• SPIL experiences the highest traffic volume compared to the other sites in Illinois,
although both Chicago sites experience a significantly higher traffic volume than
ROIL. SPIL's traffic volume is the fifth highest among all NMP sites, behind
LBHCA, ELNJ, CELA, and SEWA. The traffic volume for NBIL is in the top third
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 interstates while the traffic
volume for ROIL is not.
• The Cook County daily VMT ranks third highest among counties with NMP sites,
behind only Los Angeles County, CA and Maricopa County, AZ. The daily VMT for
Madison County is an order of magnitude less than the VMT for Cook County,
ranking in the middle third among VMT for counties with NMP sites (where VMT
data were available).
11.2 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.
11.2.1 Climate Summary
Daily weather fluctuations are common for the Chicago area. The proximity of Chicago
to Lake Michigan offers moderating effects from the continental climate of the region. In the
winter, cold air masses may be moderated by their passage over the relatively warm Lake
Michigan while in the summer, afternoon lake breezes can cool the city when winds from the
south and southwest push temperatures upward. The lake also influences precipitation as the
origin of an air mass determines the amount and type of precipitation. The largest snowfalls tend
to occur when cold air masses flow southward over Lake Michigan, most of which does not
freeze in winter. Wind speeds average around 10 miles per hour, but can be greater due to winds
channeling between tall buildings downtown, giving the city its nickname, "The Windy City".
The urban heat island effect is another climatic feature of the Chicago area, as the highly
developed urban area absorbs and retains more heat than outlying areas (TL SCO, 2014; Wood,
2004).
Roxana is northeast of St. Louis and located just north of the confluence of the
Mississippi and Missouri Rivers, which acts as Illinois' western border. The area has a climate
that is continental in nature, with cold, dry winters; warm, somewhat wetter summers; and
significant seasonal variability. Warm, moist air flowing northward from the Gulf of Mexico
alternates with cold, dry air marching southward from Canada and the northern U.S., resulting in
11-10
-------
weather patterns that do not persist for very long. Precipitation tends to be higher in the summer
months than the winter months and severe weather in the form of thunderstorms, flooding, and
tornadoes have been known to occur within the region. Southerly winds prevail in the summer
while northwesterly winds are prevalent during the colder months of the year. (Wood, 2004;
MCC, 2014).
11.2.2 Meteorological Summary
Hourly meteorological data for 2012 were retrieved from NCDC for the weather stations
closest to the Illinois monitoring sites (NCDC, 2012), as described in Section 3.5.2. The closest
weather stations are located at Palwaukee Municipal Airport (near NBIL), O'Hare International
Airport (near SPIL), and Lambert-St. Louis International Airport (near ROIL), WBANs 04838,
94846, and 13994, respectively. Additional information about these weather stations, such as the
distance between the sites and the weather stations, is provided in Table 11-3. These data were
used to determine how meteorological conditions on sample days vary from conditions
experienced throughout the year.
Table 11-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for all of 2012. Also included in Table 11-3 is the 95 percent
confidence interval for each parameter. As shown in Table 11-3, average meteorological
conditions on sample days near NBIL and SPIL were representative of average weather
conditions experienced throughout the year. Conditions on sample days appear slightly warmer
than temperatures experienced throughout the year near ROIL. However, sampling at this site
did not begin until June, thereby missing the coldest months of the year. 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.
11-11
-------
Table 11-3. Average Meteorological Conditions near the Illinois Monitoring Sites
Closest Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Northbrook, Illinois - NBIL
Palwaukee
Municipal
Airport
04838
(42.12, -87.91)
5.3
miles
250°
(WSW)
Sample
Days
(71)
2012
62.4
±4.7
62.8
±2.0
54.1
±4.3
54.2
±1.9
41.0
±3.7
41.3
±1.7
47.4
±3.6
47.7
±1.6
64.3
±2.7
65.0
±1.2
1016.1
±1.6
1016.3
±0.7
6.9
±0.8
6.6
±0.3
Schiller Park, Illinois - SPIL
O'Hare
International
Airport
94846
(41.99, -87.91)
2.5
miles
301°
(WNW)
Sample
Days
(64)
2012
63.6
±5.1
63.3
±2.1
55.2
±4.6
54.9
±1.9
41.4
±3.9
41.5
±1.6
48.1
±3.8
48.1
±1.6
63.3
±3.2
64.2
±1.4
1015.2
±1.8
1015.7
±0.7
8.7
±0.8
8.4
±0.3
Roxana, Illinois - ROIL
Lambert/
St. Louis
International
Airport
13994
(38.75, -90.37)
16.5
miles
243°
(WSW)
Sample
Days
(38)
2012
72.6
±6.1
70.7
±2.0
63.6
±5.9
61.3
±1.9
47.5
±5.1
45.0
± 1.6
54.7
±4.8
52.5
±1.5
59.2
±3.9
58.7
± 1.3
1016.3
±1.9
1016.2
±0.6
6.7
±0.9
7.1
±0.3
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
-------
11.2.3 Back Trajectory Analysis
Figure 11-6 is the composite back trajectory map for days on which samples were
collected at the NBIL monitoring site. Included in Figure 11-6 are four back trajectories per
sample day. Figure 11-7 is the corresponding cluster analysis. Similarly, Figures 11-8 through
11-11 are the composite back trajectory maps for days on which samples were collected at SPIL
and ROIL and the corresponding cluster analyses. An in-depth description of these maps and
how they were generated is presented in Section 3.5.2.1. For the composite maps, each line
represents the 24-hour trajectory along which a parcel of air traveled toward the monitoring site
on a given sample day and time, based on an initial height of 50 meters AGL. For the cluster
analyses, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric circle around the sites in Figures 11-6 through 11-11 represents
100 miles.
11-13
-------
Figure 11-6. Composite Back Trajectory Map for NBIL
Figure 11-7. Back Trajectory Cluster Map for NBIL
11-14
-------
Figure 11-8. Composite Back Trajectory Map for SPIL
Figure 11-9. Back Trajectory Cluster Map for SPIL
11-15
-------
Figure 11-10. Composite Back Trajectory Map for ROIL
Figure 11-11. Back Trajectory Cluster Map for ROIL
11-16
-------
Observations from Figures 11-6 through 11-9 for NBIL and SPIL include the following:
• The composite back trajectory maps for NBIL and SPIL are similar to each other in
back trajectory distribution, which is expected given their proximity to each other.
• Back trajectories originated from a variety of directions at the sites, with the longest
trajectories originating from the northwest and north. The predominant direction of
back trajectory origin appears to be from the south, northwest, and north.
• The 24-hour air shed domains for NBIL and SPIL were among the largest in size
compared to other NMP sites. The longest back trajectories for each site were greater
than 750 miles in length and originated over North Dakota. These back trajectories
represent the November 23, 2012 sample day. However, the average back trajectory
length for these sites is approximately 280 miles and greater than 80 percent of back
trajectories originated within 400 miles of the sites.
• Back trajectories often originated to the north of NBIL (16 percent) and over Lake
Superior, the Upper Peninsula of Michigan, Lake Michigan, and northern Michigan.
Another 16 percent of back trajectories originated to the northwest of NBIL, over
Minnesota and the Dakotas. Nearly one-quarter of back trajectories originated to the
west and southwest of NBIL, over Iowa, Missouri, and western Illinois, although
these back trajectories tended to be shorter than those originating from the northwest
or north of the site. Fifteen percent of back trajectories originated to the south of the
site, over western Kentucky and Tennessee, northeastern Arkansas, or southeast
Missouri. The short cluster trajectory originating over northwest Indiana represents
back trajectories originating to the east, southeast, and south of the site and generally
less than 300 miles in length, but also shorter back trajectories originating along the
Wisconsin/Illinois border or over the southern half of Lake Michigan. The HYSPLIT
model is grouping these back trajectories together due to their relatively short length
rather than their directional similarities.
• The cluster analysis for SPIL has more cluster trajectories than the cluster analysis for
NBIL. The differences in grouping often result in a single cluster trajectory for one
site being split into two for another or vice versa. The cluster analysis shows that back
trajectories originated to the north of SPIL (13 percent); to the northwest of SPIL
(22 percent, although these are split into two cluster trajectories based on length); to
the west and southwest of SPIL (8 percent); to the south of SPIL (13 percent); over
the state of Illinois and generally less than 200 miles in length (20 percent); and over
Indiana and less than 300 miles in length (11 percent). The short cluster trajectory
originating over Lake Michigan and representing 14 percent of back trajectories
includes those trajectories originating less than 150 miles away and looping around
Lake Michigan towards the site.
Observations from Figures 11-10 and 11-11 for ROIL include the following:
• The composite back trajectory map for ROIL has fewer back trajectories than the
composite maps for the Chicago sites because sampling at this site did not begin until
June 2012.
11-17
-------
• Back trajectories originated from a variety of directions at the site, with the longest
trajectories originating from the northwest. Few back trajectories originated from due
east or west.
• One of the longest back trajectories computed among all NMP sites was generated for
ROIL. This back trajectory represents the November 23, 2012 sample day, originates
over western North Dakota, and is nearly 900 miles long. This is the same date for
which the longest back trajectories were generated for the Chicago sites. Yet, the
24-hour air shed domain for ROIL is similar in size to many other NMP sites. The
average back trajectory length for this site is 234 miles and nearly 84 percent of back
trajectories originated within 350 miles of the site.
• The cluster map for ROIL bears some resemblance to the cluster map for NBIL in the
geographical distribution of the clusters. Back trajectories originating to the
northwest, north, and northeast of ROIL account for nearly 25 percent of back
trajectories. Another 12 percent originated to the northwest of the site, but are split
into two cluster trajectories based on length. Eleven percent of back trajectories
originated to the southwest of ROIL, over Missouri and Arkansas. Nearly one quarter
of back trajectories originated to the south of the site, over western Kentucky and
Tennessee, northeastern Arkansas, or northern Mississippi. The short cluster
trajectory originating over south-central Illinois represents back trajectories
originating to the northeast and east over Indiana as well as shorter back trajectories
that spiraled around the southern half of Illinois on the way to ROIL.
11.2.4 Wind Rose Comparison
Hourly surface wind data from the weather stations at Palwaukee Municipal Airport (for
NBIL), O'Hare International Airport (for SPIL), and Lambert/St. Louis International Airport (for
ROIL) were uploaded into a wind rose software program to produce customized wind roses, as
described in Section 3.5.2.2. 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 11-12 presents a map showing the distance between the weather station and NBIL,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 11-12 also presents three different wind roses for the
NBIL monitoring site. First, a historical wind rose representing 2002 to 2011 wind data is
presented, which shows the predominant surface wind speed and direction over an extended
period of time. Second, a wind rose representing wind observations for all of 2012 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2012 is
presented. These can be used to identify the predominant wind speed and direction for 2012 and
to determine if wind observations on sample days were representative of conditions experienced
11-18
-------
over the entire year and historically. Figures 11-13 and 11-14 present the distance map and three
wind roses for SPIL and ROIL, respectively.
Observations from Figure 11-12 for NBIL include the following:
• The Palwaukee Municipal Airport weather station is located 5.3 miles west-southwest
ofNBIL.
• The historical wind rose shows that winds from a variety of directions were observed
near NBIL, although winds from the south, south-southwest, and west accounted for
one-quarter of wind observations. Winds from the east-southeast to south-southeast
were observed the least often. Calm winds (< 2 knots) were observed for
approximately 16 percent of the hourly measurements.
• The 2012 wind rose exhibits similar patterns in wind directions as the historical wind
rose, although a higher percentage of winds from the south and south-southwest and
fewer winds from the west were observed in 2012.
• The sample day wind patterns resemble the full-year wind patterns, with an even
higher percentage of winds from the south and south-southwest and even fewer winds
from the west.
Observations from Figure 11-13 for SPIL include the following:
• The O'Hare International Airport weather station is located 3.5 miles west-northwest
of SPIL. Most of the airport property lies between the weather station and the
monitoring site.
• The historical wind rose for SPIL shows that winds from a variety of directions were
observed, although winds from the south to southwest to west account for the highest
percentage of observations (nearly 40 percent). Winds from the southeast quadrant
were observed the least. Calm winds (< 2 knots) were observed for less than 8 percent
of the hourly measurements.
• The 2012 wind rose exhibits similar patterns in wind directions as the historical wind
rose, although winds from the predominant directions accounted for an even higher
percentage of the wind observations. The strongest winds were from the south and
south-southwest.
• The sample day wind pattern resemble those of the full-year wind rose, with the
winds from the south to southwest to west accounting for nearly 50 percent of the
wind observations.
11-19
-------
Figure 11-12. Wind Roses for the Palwaukee Municipal Airport Weather Station near
NBIL
Location of NBIL and Weather Station
2002-2011 Historical Wind Rose
2012 Wind Rose
Sample Day Wind Rose
Calms: 17.0B%
11-20
-------
Figure 11-13. Wind Roses for the O'Hare International Airport Weather Station near
SPIL
Location of SPIL and Weather Station
2002-2011 Historical Wind Rose
2012 Wind Rose
Sample Day Wind Rose
11-21
-------
Figure 11-14. Wind Roses for the Lambert/St. Louis International Airport Weather Station
near ROIL
Location of ROIL and Weather Station
2002-2011 Historical Wind Rose
2012 Wind Rose
Sample Day Wind Rose
11-22
-------
Observations from Figure 11-14 for ROIL include the following:
• The Lambert/St. Louis International Airport weather station is located 16.5 miles
west-southwest of ROIL. The airport lies to the northwest of the city of St. Louis and
south of the Missouri River.
• The historical wind rose for ROIL shows that winds from a variety of directions were
observed, with winds from the south observed the most. Winds from the west to
northwest were also common while winds from the northeast quadrant were observed
the least. Calm winds (< 2 knots) were observed for less than 12 percent of the hourly
measurements.
• The 2012 wind rose exhibits similar patterns in wind directions as the historical wind
rose, although winds from the south accounted for an even higher percentage of the
wind observations.
• The predominant wind direction on the sample day wind rose is still south, but the
similarities in the wind patterns are fewer. Winds from the east, east-southeast, south-
southeast, and west-northwest account for a higher percentage of wind observations
than they do on the full-year wind rose. However, this sample day wind rose includes
only seven months of the year as sampling at ROIL did not begin until June 2012.
11.3 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 11-4.
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 11-4. It is
important to note which pollutants were sampled for at each site when reviewing the results of
this analysis. NBIL sampled for VOCs, carbonyl compounds, SNMOCs, metals (PMio), PAHs,
and hexavalent chromium, while SPIL and ROIL sampled for VOCs and carbonyl compounds
only.
11-23
-------
Table 11-4. Risk-Based Screening Results for the Illinois Monitoring Sites
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Northbrook, Illinois - NBIL
Formaldehyde
Acetaldehyde
Benzene
Carbon Tetrachloride
1 ,2-Dichloroethane
Arsenic (PM10)
Naphthalene
1,3 -Butadiene
Manganese (PM10)
Fluorene
Acenaphthene
Fluoranthene
£>-Dichlorobenzene
Ethylbenzene
Hexachloro- 1 ,3 -butadiene
Trichloroethylene
Chloroform
Benzo(a)pyrene
Bromoform
Dichloromethane
Hexavalent Chromium
Nickel (PM10)
0.077
0.45
0.13
0.17
0.038
0.00023
0.029
0.03
0.005
0.011
0.011
0.011
0.091
0.4
0.045
0.2
9.8
0.00057
0.91
7.7
0.000083
0.0021
Total
66
65
61
61
53
47
44
42
31
22
19
14
12
8
5
4
3
1
1
1
1
1
562
66
66
61
61
53
54
57
45
54
57
57
57
31
61
6
25
61
55
13
61
44
54
1,099
100.00
98.48
100.00
100.00
100.00
87.04
77.19
93.33
57.41
38.60
33.33
24.56
38.71
13.11
83.33
16.00
4.92
1.82
7.69
1.64
2.27
1.85
51.14
11.74
11.57
10.85
10.85
9.43
8.36
7.83
7.47
5.52
3.91
3.38
2.49
2.14
1.42
0.89
0.71
0.53
0.18
0.18
0.18
0.18
0.18
11.74
23.31
34.16
45.02
54.45
62.81
70.64
78.11
83.63
87.54
90.93
93.42
95.55
96.98
97.86
98.58
99.11
99.29
99.47
99.64
99.82
100.00
Schiller Park, Illinois - SPIL
Acetaldehyde
Formaldehyde
Benzene
Carbon Tetrachloride
1,3 -Butadiene
1 ,2-Dichloroethane
Trichloroethylene
£>-Dichlorobenzene
Ethylbenzene
Hexachloro- 1 ,3 -butadiene
1, 1,2,2-Tetrachloroethane
1 ,2-Dibromoethane
Propionaldehyde
0.45
0.077
0.13
0.17
0.03
0.038
0.2
0.091
0.4
0.045
0.017
0.0017
0.8
Total
61
61
60
60
58
55
22
11
11
4
4
1
1
409
61
61
60
60
59
55
49
39
60
9
4
1
61
579
100.00
100.00
100.00
100.00
98.31
100.00
44.90
28.21
18.33
44.44
100.00
100.00
1.64
70.64
14.91
14.91
14.67
14.67
14.18
13.45
5.38
2.69
2.69
0.98
0.98
0.24
0.24
14.91
29.83
44.50
59.17
73.35
86.80
92.18
94.87
97.56
98.53
99.51
99.76
100.00
11-24
-------
Table 11-4. Risk-Based Screening Results for the Illinois Monitoring Sites (Continued)
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Roxana, Illinois - ROIL
Formaldehyde
Acetaldehyde
Benzene
Carbon Tetrachloride
1,3 -Butadiene
1 ,2-Dichloroethane
Ethylbenzene
£>-Dichlorobenzene
Hexachloro- 1 ,3 -butadiene
1, 1,2,2-Tetrachloroethane
1 ,2-Dibromoethane
0.077
0.45
0.13
0.17
0.03
0.038
0.4
0.091
0.045
0.017
0.0017
Total
35
34
33
33
30
29
19
5
4
4
2
228
35
35
33
33
31
29
33
22
5
4
2
262
100.00
97. 14
100.00
100.00
96.77
100.00
57.58
22.73
80.00
100.00
100.00
87.02
15.35
14.91
14.47
14.47
13.16
12.72
8.33
2.19
1.75
1.75
0.88
15.35
30.26
44.74
59.21
72.37
85.09
93.42
95.61
97.37
99.12
100.00
Observations from Table 11-4 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 the pollutants measured at
each site.
• Twenty-two pollutants failed at least one screen for NBIL; 51 percent of
concentrations for these 22 pollutants were greater than their associated risk screening
value (or failed screens).
• Thirteen pollutants contributed to 95 percent of failed screens for NBIL and therefore
were identified as pollutants of interest for this site. These 13 include two carbonyl
compounds, five VOCs, two PMio metals, and four PAHs.
• NBIL failed the fifth highest number of screens (562) among all NMP sites, as shown
in Table 4-8 of Section 4.2. However, the failure rate for NBIL, when incorporating
all pollutants with screening values, is relatively low, at 22 percent. This is due
primarily to the relatively high number of pollutants sampled for at this site, as
discussed in Section 4.2. NBIL is one of only two NMP sites sampling for all six
pollutant groups. 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 71 percent of
concentrations for these 13 pollutants were greater than their associated risk screening
value (or failed screens).
11-25
-------
• Nine pollutants contributed to 95 percent of failed screens for SPIL and therefore
were identified as pollutants of interest for this site. These nine include two carbonyl
compounds and seven VOCs.
• Eleven pollutants failed screens for ROIL; approximately 87 percent of
concentrations for these 11 pollutants were greater than their associated risk screening
value (or failed screens). Although this percentage is higher for ROIL than the
Chicago sites, nearly all of the measured detections for the pollutants listed for ROIL
failed screens while the percentage of screens failed for each individual pollutant is
more varied for the Chicago sites.
• 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, />-dichlorobenzene, and 1,2-dichloroethane). Of
these, benzene, carbon tetrachloride, 1,2-dichloroethane, and formaldehyde failed 100
percent of screens for each site.
11.4 Concentrations
This section presents various concentration averages used to characterize pollution 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. Additional site-specific statistical summaries for NBIL,
SPIL, and ROIL are provided in Appendices J through O.
11-26
-------
11.4.1 2012 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 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 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 Illinois monitoring
sites are presented in Table 11-5, 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.
11-27
-------
Table 11-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest
for the Illinois Monitoring Sites
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Hg/m3)
2nd
Quarter
Average
(Hg/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Hg/m3)
Northbrook, Illinois - NBIL
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
£>-Dichlorobenzene
1,2-Dichloroethane
Formaldehyde
Acenaphthene3
Arsenic (PM10)a
Fluoranthene3
Fluorene3
Manganese (PM10)a
Naphthalene3
66/66
61/61
45/61
61/61
31/61
53/61
66/66
57/57
54/54
57/57
57/57
54/54
57/57
1.82
±0.41
0.70
±0.18
0.05
±0.02
0.75
±0.18
0.03
±0.03
0.09
±0.02
1.73
±0.26
4.49
±3.76
0.66
±0.21
2.95
±2.00
4.89
±3.99
8.26
±3.55
62.56
± 29.90
1.89
±0.43
0.54
±0.14
0.04
±0.02
0.73
±0.04
0.05
±0.03
0.07
±0.02
2.78
±0.82
11.12
±5.35
0.67
±0.24
7.69
±3.46
13.09
±6.11
10.93
±5.65
63.79
± 22.73
1.72
±0.43
0.70
±0.18
0.08
±0.04
0.67
±0.04
0.07
±0.03
0.06
±0.02
3.52
±0.92
20.20
±8.52
0.85
±0.23
14.15
±5.88
23.97
±11.51
9.37
±2.91
98.01
±26.37
1.73
±0.38
0.64
±0.14
0.08
±0.04
0.69
±0.04
0.02
±0.02
0.06
±0.02
2.00
±0.24
NA
0.72
±0.33
NA
NA
7.92
±3.84
NA
1.78
±0.19
0.64
±0.08
0.06
±0.02
0.71
±0.04
0.04
±0.01
0.07
±0.01
2.49
±0.33
11.51
±3.62
0.73
±0.12
7.07
±2.27
12.31
±4.18
9.11
±1.86
77.94
± 17.77
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.
11-28
-------
Table 11-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest
for the Illinois Monitoring Sites (Continued)
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Hg/m3)
2nd
Quarter
Average
(Hg/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Hg/m3)
Schiller Park, Illinois - SPIL
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
£>-Dichlorobenzene
1,2-Dichloroethane
Ethylbenzene
Formaldehyde
Trichloroethylene
61/61
60/60
59/60
60/60
39/60
55/60
60/60
61/61
49/60
1.57
±0.33
1.10
±0.40
0.13
±0.03
0.63
±0.07
0.05
±0.03
0.09
±0.01
0.26
±0.09
1.94
±0.40
0.64
±0.66
1.44
±0.22
0.91
±0.33
0.10
±0.03
0.73
±0.04
0.06
±0.02
0.09
±0.02
0.28
±0.08
3.13
±0.73
0.56
±0.40
2.04
±0.62
0.84
±0.17
0.15
±0.03
0.65
±0.04
0.06
±0.03
0.07
±0.01
0.31
±0.09
4.64
± 1.51
0.36
±0.21
5.65
±2.41
0.96
±0.54
0.17
±0.12
0.71
±0.04
0.04
±0.04
0.07
±0.02
0.28
±0.19
2.66
± 1.36
1.20
±2.24
2.72
±0.77
0.95
±0.19
0.14
±0.03
0.68
±0.03
0.05
±0.02
0.08
±0.01
0.29
±0.06
3.09
±0.58
0.71
±0.60
Roxana, Illinois - ROIL
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
£>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
35/35
33/33
31/33
33/33
22/33
29/33
33/33
35/35
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
5.00
±0.67
1.77
±0.64
0.09
±0.03
0.66
±0.04
0.06
±0.02
0.08
±0.02
0.56
±0.15
1.13
±0.16
3.51
±0.78
1.90
±0.53
0.10
±0.04
0.67
±0.04
0.03
±0.02
0.09
±0.02
0.42
±0.13
0.91
±0.23
NA
NA
NA
NA
NA
NA
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 are presented in ng/m3 for ease of viewing.
11-29
-------
Observations for NBIL from Table 11-5 include the following:
• The pollutants with the highest annual average concentrations are formaldehyde
(2.49 ± 0.33 |ig/m3) and acetaldehyde (1.78 ± 0.19 jig/m3). The annual average
concentrations for the remaining pollutants of interest are less than 1 |ig/m3.
• The second and third quarter average formaldehyde concentrations are higher than the
first and fourth quarter averages and have relatively large confidence intervals
associated with them. The two highest formaldehyde concentrations measured at
NBIL were measured on back-to-back sample days in August (7.18 |ig/m3 and
6.52 jig/m3). The 10 highest concentrations were all measured between May and
September at this site, and ranged from 3.38 |ig/m3 to 7.18 |ig/m3.
• The first quarter average concentration of carbon tetrachloride is not significantly
different than the other quarterly averages but does have a relatively large confidence
interval associated with it. A review of the data shows that the measurement collected
on January 10, 2012 (1.88 |ig/m3) is more than twice the next highest concentration
measured at NBIL. This measurement is the fourth highest concentration of carbon
tetrachloride measured among all NMP sites sampling VOCs.
• Fourth quarter average concentrations could not be calculated for the PAHs because
fewer than 75 percent of samples were valid during this quarter.
• The second and third quarter average concentrations of acenaphthene, fluoranthene,
and fluorene were considerably higher than the first quarter averages, although all of
the quarterly averages shown exhibit a relatively large amount of variability based on
the confidence intervals.
• The maximum concentration of fluorene was measured on September 6, 2012
(93.4 ng/m3) and is nearly twice the next highest concentration, which was measured
on July 2, 2012 (55.0 ng/m3). Concentrations of fluorene measured at NBIL span two
orders of magnitude (0.934 ng/m3 to 93.4 ng/m3), with all but two of the 17
concentrations greater than 15 ng/m3 measured in samples collected between May
and September. Conversely, all 18 concentrations less than 3 ng/m3 were measured
between January and April or October and December. This seasonality in the
concentrations is also apparent in the fluoranthene measurements collected at NBIL
(and to some extent the acenaphthene measurements).
• Although the third quarter average concentration of naphthalene is greater than the
first and second quarterly averages, the maximum concentrations of naphthalene were
measured in December (359 ng/m3), November (233 ng/m3), and March (228 ng/m3).
Despite this, the median concentration for the third quarter is two to three times
greater than the median concentrations for the other quarters. The third quarter has the
greatest number of naphthalene concentrations greater than 100 ng/m3 (five) and the
least less than 40 ng/m3 (zero).
11-30
-------
• Concentrations of manganese also exhibit a considerable amount of variability,
particularly the second quarter. Although the two highest concentrations of
manganese were measured at NBIL in May (32. 1 ng/m3) and June (25.4 ng/m3), the
minimum manganese concentration was measured in April (1.87 ng/m3).
Observations for SPIL from Table 11-5 include the following:
• The pollutants with the highest annual average concentrations are formaldehyde
(3.09 ± 0.58 (ig/m3) and acetaldehyde (2.72 ± 0.77 |ig/m3). These are the only
pollutants with annual average concentrations greater than 1 |ig/m3.
• The fourth quarter average concentration of acetaldehyde 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 13 of the 15 concentrations greater than
3 |ig/m3 were measured at SPIL during the fourth quarter of 2012 and ranged from
3.19 |ig/m3 to 20.4 |ig/m3. The three highest concentrations of acetaldehyde measured
among NMP sites sampling carbonyl compounds were measured at SPIL
(8.74 |ig/m3, 1 1.8 |ig/m3, and 20.4 |ig/m3) between November and December.
• The third and fourth quarter average concentrations of formaldehyde have relatively
large confidence intervals associated with them. Six of the eight highest
concentrations of formaldehyde were measured at SPIL in July or August, including
the maximum concentration measured (12.8 |ig/m3). The second highest
concentration, however, was measured in December (12.3 |ig/m3). The next highest
concentration measured during the fourth quarter of 2012 was considerably less
(3.67
Several fourth quarter averages for the VOCs have relatively large confidence
intervals associated with them. The maximum concentration of benzene,
1,3-butadiene, ethylbenzene, and/>-dichlorobenzene were all measured on
November 17, 2012. The next highest concentration measured during the fourth
quarter for each of these pollutants was considerably less.
All four quarterly averages of trichloroethylene have relatively large confidence
intervals associated with them, particularly the fourth quarter average concentration.
This indicates that the concentrations of trichloroethylene are highly variable. A
review of the data shows that the maximum trichloroethylene concentration was
measured on November 17, 2012 (17.5 |ig/m3) and is the highest trichloroethylene
concentration measured among NMP sites sampling VOCs. Of the 10 concentrations
of trichloroethylene greater than 1 |ig/m3 across the program, eight of these were
measured at SPIL. The maximum concentration of trichloroethylene measured at
SPIL is nearly four times higher than the next highest concentration measured at this
site (4.57 |ig/m3). Trichloroethylene concentrations measured at this site range from
0.043 1 |ig/m3 to 17.5 |ig/m3, with a median concentration of 0. 127 |ig/m3 (including
1 1 non-detects). Similar observations were also made in the 201 1 NMP report.
11-31
-------
Observations for ROIL from Table 11-5 include the following:
• First and second quarter average concentrations are not provided for ROIL because
sampling did not begin until June 2012, and because at least three quarterly averages
are not available, annual averages could not be calculated for this site.
• The quarterly averages for acetaldehyde are significantly higher than the quarterly
averages for the other pollutants of interest for ROIL. A review of the data shows that
acetaldehyde measurements collected at ROIL span an order of magnitude, ranging
from 0.303 |ig/m3 to 7.47 |ig/m3. The three highest concentrations measured at ROIL
are among the 10 highest acetaldehyde concentrations measured among NMP sites
sampling carbonyl compounds. Interestingly, four of the five lowest acetaldehyde
concentrations were measured at ROIL on the first four sample days at ROIL (in
June).
• Conversely, the quarterly average concentrations of formaldehyde (that could be
calculated) are among the lowest for NMP sites sampling carbonyl compounds.
Tables 4-9 through 4-12 present the sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for NBIL and
SPIL from those tables include the following:
• NBIL and SPIL appear in Tables 4-9 through 4-12 a total of 12 times. Annual
averages could not be calculated for ROIL and therefore could not appear in these
tables.
• The annual average concentrations for both Chicago sites appear in Table 4-9 for
carbon tetrachloride, although there is little variability in the annual average
concentrations of this pollutant. SPIL also appears in Table 4-9 for 1,3-butadiene
(sixth) and 1,2-dichloroethane (fifth), while NBIL appears under hexachloro-1,3-
butadiene (eighth).
• SPIL has the fourth highest annual average concentration of acetaldehyde among
NMP sites sampling carbonyl compounds, as shown in Table 4-10. Note however,
that the confidence interval for SPIL is nearly twice the confidence intervals of the
other sites shown, indicating that the concentrations collected at SPIL have a higher
level of variability associated with them.
• NBIL has the second highest annual average concentration of fluorene and fourth
highest annual average concentration of acenaphthene among NMP sites sampling
PAHs, as shown in Table 4-11. The annual average concentration of naphthalene for
NBIL ranks 10th.
• As shown in Table 4-12, NBIL's annual average concentration of manganese ranks
fifth among NMP sites sampling PMi0 metals, with NBIL's annual average
concentrations of arsenic and nickel ranking seventh and eighth, respectively.
11-32
-------
11.4.2 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 shaded in
gray in Table 11-4 for NBIL and SPIL. Figures 11-15 through 11-29 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 3.5.3.1.
Site-specific box plots were not created for ROIL because annual averages could not be
calculated.
NBIL
Figure 11-15. Program vs. Site-Specific Average Acenaphthene Concentration
Program Max Concentration = 182 ng/m3
10
20
30
40 50 60
Concentration (ng/m3)
70
80
90
100
Program:
Site:
1st Quartile
D
Site Average
0
2nd Quartile 3rd Quartile
D D
Site Concentration Range
4th Quartile Average
D 1
Figure 11-16. Program vs. Site-Specific Average Acetaldehyde Concentrations
NBIL
9 12
Concentration {[og/m3)
Program: 1st Quartile
2nd Quartile
3rd Quartile
4th Quartile
Average
Site:
Site Average Site Concentration Range
o
11-33
-------
Figure 11-17. Program vs. Site-Specific Average Arsenic (PMi0) Concentration
NBIL
34
Concentration (ng/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o
Figure 11-18. Program vs. Site-Specific Average Benzene Concentrations
SPIL
Concentration {[ig/m3)
Program:
Site:
IstQuartile 2ndQuartile SrdQuartile
• D D
Site Average Site Concentration Range
o —
4thQuartile Average
D 1
Figure 11-19 Program vs. Site-Specific Average 1,3-Butadiene Concentrations
in-
Program Max Concentration = 4.10 ug/m3
I Pn
Program Max Concentration = 4.10 ug/m3
0.25
0.5
0.75 1 1.25
Concentration {[og/m3)
1.5
1.75
Program:
Site:
IstQuartile
D
Site Average
0
2ndQuartile SrdQuartile
D D
Site Concentration Range
4thQuartile Average
D 1
11-34
-------
Figure 11-20. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations
SPIL
2 3
Concentration {[og/m3)
Program:
Site:
IstQuartile
D
Site Average
o
2ndQuartile SrdQuartile
• •
Site Concentration Range
^^^^—
4thQuartile Average
D 1
Figure 11-21. Program vs. Site-Specific Average />-Dichlorobenzene Concentrations
SPIL
0.2
0.4
0.6 0.8
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile
Site:
SiteAverage Site Concentration Range
O
4thQuartile
1.2
Average
1.4
11-35
-------
Figure 11-22. Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations
Program Max Concentration = 17.01 ug/m3
—
Program Max Concentration = 17.01 ug/m3
0.1
0.2
0.3
0.4 0.5 0.6
Concentration {[og/m3)
0.7
0.8 0.9
Program:
Site:
IstQuartile
D
Site Average
0
2ndQuartile SrdQuartile
D D
Site Concentration Range
4thQuartile Average
D 1
Figure 11-23. Program vs. Site-Specific Average Ethylbenzene Concentration
SPIL
1.5 2 2.5
Concentration {[ig/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o
Figure 11-24. Program vs. Site-Specific Average Fluoranthene Concentration
10
15
20 25
Concentration {ng/m3]
30
35
40
Program:
Site:
IstQuartile
D
Site Average
O
2ndQuartile SrdQuartile
• D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
45
11-36
-------
Figure 11-25. Program vs. Site-Specific Average Fluorene Concentration
NBIL
L
0
1
10 20 30 40 50 60 70 80 90 1C
Concentration (ng/m3)
Program:
Site:
IstQuartile 2ndQuartile
Site Average Site
0
SrdQuartile 4thQuartile
n n
Average
1
i i i — i i i |
Concentration Range
^^^^—
Figure 11-26. Program vs. Site-Specific Average Formaldehyde Concentrations
Concentration {
Program:
Site:
IstQuartile
D
Site Average
o
2ndQuartile SrdQuartile
• •
Site Concentration Range
^^^^—
4thQuartile Average
D 1
Figure 11-27. Program vs. Site-Specific Average Manganese (PMio) Concentration
Program Max Concentration = 275 ng/m3
30
60 90
Concentration (ng/m3)
120
150
Program:
Site:
IstQuartile
D
Site Average
O
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
11-37
-------
Figure 11-28. Program vs. Site-Specific Average Naphthalene Concentration
NBIL
• ol
°l
1 1 1 1
III 1
D 100 200 300 400 500 600 700 800 9C
Concentration (ng/m3)
Program:
Site:
IstQuartile
D
Site Average
0
2ndQuartile SrdQuartile
D D
Site Concentration Range
4thQuartile Average
D 1
Figure 11-29. Program vs. Site-Specific Average Trichloroethylene Concentration
SPIL
8 10
Concentration (jig/m3)
12
14
16
18
Program:
Site:
IstQuartile
D
Site Average
o
2ndQuartile SrdQuartile
• a
Site Concentration Range
^^^^—
4thQuartile Average
D 1
Observations from Figures 11-15 through 11-29 include the following:
• Figure 11 -15 is the box plot for acenaphthene for NBIL. Note that the program-
level maximum concentration (182 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 to
100 ng/m3. Figure 11-15 shows thatNBIL's annual average acenaphthene
concentration is more than twice the program-level average concentration.
Although the maximum concentration of acenaphthene measured at NBIL was
less than the maximum concentration measured at the program-level, this
measurement is among the higher concentrations of this pollutant. 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. Although non-detects were measured across
the program, none were measured at NBIL.
• Figure 11-16 presents the acetaldehyde box plots for both Chicago sites. The box
plots show that the maximum acetaldehyde concentration across the program was
measured at SPIL; as discussed previously, the three highest concentrations of this
pollutant were all measured at SPIL. Thus, the annual average acetaldehyde
concentration for SPIL is greater than the annual average acetaldehyde
concentration for NBIL and the program-level average. NBIL's annual average is
11-38
-------
similar to the program-level average concentration. Even the minimum
acetaldehyde concentrations measured at these sites are significantly different.
The minimum acetaldehyde concentration measured atNBIL is among the lower
concentrations measured program-wide; the minimum acetaldehyde concentration
measured at SPIL is nearly seven times higher.
• Figure 11-17 is the box plot for arsenic, which was measured at NBIL but not at
SPIL. The box plot shows that the annual average concentration for NBIL is
similar to the program-level average concentration. The maximum concentration
measured at NBIL is considerably less than the maximum concentration measured
across the program. While a few non-detects of arsenic were measured among
sites sampling PMio metals, none were measured at NBIL.
• Figure 11-18 shows the box plots for benzene. The range of concentrations
measured at SPIL is more than twice the range of concentrations measured at
NBIL. Thus, SPIL's annual average benzene concentration is greater than NBIL's
annual average benzene concentration; in addition, the annual average benzene
concentration for SPIL is greater than the program-level average concentration
while NBIL's annual average is less than the program-level median concentration.
• Similar to the box plots for acenaphthene, the program-level maximum
1,3-butadiene concentration (4.10 |ig/m3) is not shown directly on the box plots as
the scale has been reduced to 2 |ig/m3 in Figure 11-19 to allow for the observation
of data points at the lower end of the concentration range. Figure 11-19 shows
that the Chicago sites' concentrations of 1,3-butadiene follow a similar patterns as
the sites' benzene concentrations. The range of concentrations measured is larger
for SPIL than for NBIL. While NBIL's annual average concentration of
1,3-butadiene is less than both the program-level average and median
concentrations, SPIL's annual average is greater than the program-level average
and just less than the program-level third quartile. A single non-detect of
1,3-butadiene was measured at SPIL while 16 non-detects were measured at
NBIL.
• Figure 11-20 presents the box plots for carbon tetrachloride. Even though the
range of measurements appears much larger for NBIL, this is a result of a single
"high" concentration measured atNBIL (1.88 |ig/m3). If this concentration was
excluded, the range of measurements for these two sites would be very similar.
Thus, the annual average concentrations of carbon tetrachloride for the Chicago
sites are very similar to each other and the program-level average concentration,
as discussed in the previous section.
• The first quartile for/?-dichlorobenzene is zero due to the number of non-detects
and, as a result, is not visible on the box plots in Figure 11-21. Although the
annual average concentration ofp-dichlorobenzene for NBIL is just slightly less
than the annual average for SPIL, both are less than the program-level average
concentration but greater than the program-level median concentration. However,
less than 0.022 |ig/m3 separates these four statistical parameters.
/>-Dichlorobenzene was detected in half of the samples collected atNBIL and
11-39
-------
65 percent of the samples collected at SPIL, which is similar to the percentage of
measured detections at the program-level.
The program-level maximum 1,2-dichloroethane concentration (17.01 |ig/m3) is
significantly higher than most of the concentrations measured at NMP sites
sampling VOCs. Therefore, the maximum concentration is not shown directly on
the box plots in Figure 11-22 as the scale has been reduced to 1 |ig/m3.
Figure 11-22 shows that the majority of the 1,2-dichloroethane measurements
collected at the Chicago sites are less than the program-level average
concentration of this pollutant. The wide disparity between most of the program-
level statistical parameters and the maximum concentration indicates that outliers
were measured at other NMP site(s).
Figure 11-23 shows that the annual average ethylbenzene concentration for NBIL
is slightly less than the program-level average concentration but greater than the
program-level median concentration. The maximum ethylbenzene concentration
measured at NBIL is less than program-level maximum ethylbenzene
concentration. There were no non-detects of ethylbenzene measured at NBIL.
Figures 11-24 and 11-25 present the box plots for fluoranthene and fluorene,
respectively, for NBIL. These box plots show that the maximum concentrations of
fluoranthene and fluorene at the program-level were measured at NBIL.
Concentrations of these pollutants measured at NBIL span two orders of
magnitude. For both PAHs, the annual average concentration for NBIL is more
than twice the program-level average concentrations.
Figure 11-26 presents the box plots for formaldehyde. The maximum
formaldehyde concentration measured across the program was measured at SPIL,
although an equivalent measurement was also collected at TOOK. SPIL's annual
average formaldehyde concentration is greater than the program-level average
concentration but less than the program-level third quartile. Even though the
maximum formaldehyde concentration was measured at SPIL, this site's annual
average concentration ranks 12th among other NMP sites. NBIL's annual average
formaldehyde concentration is less than the program-level average concentration
but greater than the program-level median. The difference between the minimum
formaldehyde concentrations measured at these two sites is similar to the
differences noted for acetaldehyde.
Figure 11-27 is the box plot for manganese (PMio). Note that the program-level
maximum concentration (275 ng/m3) is not shown directly on the box plot as the
scale has been reduced to 150 ng/m3 to allow for the observation of data points at
the lower end of the concentration range. Although the maximum concentration
measured at NBIL is considerably less than the maximum measured across the
program, the box plot shows that the annual average concentration for NBIL is
just less than the program-level average concentration.
11-40
-------
• Figure 11-28 is the box plot for naphthalene. The maximum concentration
naphthalene measured at NBIL (359 ng/m3) is less than half the maximum
concentration measured across the program (822 ng/m3). The annual average
concentration for NBIL is just less than the program-level average concentration
but greater than the program-level median.
• The first, second, and third quartiles for trichloroethylene are all zero in
Figure 11-29 due to the large number of non-detects; thus, only the fourth quartile
is visible. The maximum concentration of trichloroethylene across the program
was measured at SPIL. The annual average concentration for SPIL (0.71 |ig/m3) is
seven times greater than the next highest annual average concentration for this
pollutant (calculated for GPCO, 0.10 |ig/m3) and an order of magnitude higher
than the program-level average concentration (0.050 |ig/m3).
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 3.5.3.2.
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 11-30 through 11-52 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 is not provided, although the range and quartiles
are still presented.
11-41
-------
Figure 11-30. Yearly Statistical Metrics for Acenaphthene Concentrations Measured at NBIL
~ 50.0
O
2010
Year
O 5th Percentile - Minimuir
— Maximum • 95th Percentile "-O-" Averagf
A 1-year average is not presented because sampling under the NMP did not begin until June 2008.
Observations from Figure 11-30 for acenaphthene measurements collected at NBIL
include the following:
• Although PAH sampling under the NMP at NBIL began 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.
• Although the maximum acenaphthene concentration was measured at NBIL in 2008
(93.5 ng/m3), the second highest concentration measured that year was considerably
less (29.0 ng/m3). Although a concentration greater than 30 ng/m3 has been measured
each year since sampling began, nearly half of them were measured in 2011, with one
measured in 2008, two in 2009, three in 2010, 10 in 2011, and five in 2012.
• 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 after 2009.
• 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.
11-42
-------
Figure 11-31. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at NBIL
±
&I
,o
,o
2008 2009
Year
O 5th Percentile
• 95th Percentile
1A 1-year average is not presented because sampling under the NMP did not begin until March 2005.
Observations from Figure 11-31 for acetaldehyde measurements collected at NBIL
include the following:
• Carbonyl compound sampling at NBIL under the NMP 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 (4.22 |ig/m3) was measured in 2012,
although a similar concentration (4.12 |ig/m3) was measured in 2011. The highest
acetaldehyde concentrations were measured in the most recent years; of the 22
acetaldehyde concentrations greater than 2.5 |ig/m3 measured at NBIL, one was
measured in 2005, two in 2010, eight in 2011, and 11 were measured in 2012.
• After a decreasing trend through 2007, the 1-year average fluctuated between
0.69 |ig/m3 and 0.89 |ig/m3 between 2007 and 2009. After 2009, acetaldehyde
concentrations measured at NBIL increase significantly as all of the statistical metrics
exhibit an increase from 2009 to 2010 and again for 2011 and 2012 (although the
minimum concentration decreased for 2012). The 95th percentile for 2012 is greater
than the maximum concentration measured for most years of sampling. The 5th
percentile for 2012 is greater than the median and 1-year averages for some of the
earlier years of sampling.
• The increase in the 1-year average concentration of acetaldehyde between 2009 and
2012 represents a 159 percent increase.
11-43
-------
Figure 11-32. Yearly Statistical Metrics for Arsenic (PMi0) Concentrations Measured at NBIL
o
o
o
2008 2009
Year
O 5th Percentile — Minimurr
Median — Maximum O 95th Percentile
Observations from Figure 11-32 for arsenic (PMio) measurements collected at NBIL
include the following:
• Metals sampling at NBIL began in January 2005.
• The maximum arsenic concentration was measured on July 12, 2009, although a
similar concentration was also measured in 2010. Only four concentrations equal to
or 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 has fluctuated somewhat each year, the 1-year average concentrations exhibit
little significant change over the course of sampling. The 1-year average
concentration increased from 2005 to 2006, reached a maximum for 2007
(0.86 ng/m3), decreased slightly for 2008, after which the 1-year average
concentration has remained steady. Since 2008, the 1-year average concentrations
have ranged from 0.730 ng/m3 (2012) to 0.753 ng/m3 (2010).
• The minimum concentration for each year is greater than zero, indicating that there
were no non-detects of arsenic reported since the onset of metals sampling.
11-44
-------
Figure 11-33. Yearly Statistical Metrics for Benzene Concentrations Measured at NBIL
s "
8 2.0
o
o
Lrl
2007 2008
Year
O 5th Percentile
— Maximum
• 95th Percentile "-O-" Averagf
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 11-33 for benzene measurements collected 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 2004, 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 measurement greater than 4 |ig/m3 measured at NBIL. The three
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.
• The 1-year average concentration decreased significantly from 2005 to 2006, and
decreased slightly for 2007, then remained at the same level 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.
11-45
-------
Figure 11-34. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at NBIL
~ 0.8
Maximum
Concentration for
2011 is 2.68 ug/m3.
"OH
e
2007 2008
Year
O 5th Percentile
— Maximum
O 95th Percentile
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 11-34 for 1,3-butadiene measurements collected 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
concentrations greater than 1 |ig/m3 have been measured at NBIL, two in 2011 and
one in 2010. All other measurements of 1,3-butadiene 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, even the median concentration is zero. The number 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). Since 2010, the percentage of non-detects has hovered
around 25 percent.
• The 1-year average concentration changed little through 2009, after which an
increasing trend is shown through 2011, although 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. For example, for 2010, only three
measurements were greater than the 95th percentile; further, the maximum
11-46
-------
concentration was an order of magnitude higher than the 95th percentile for this year.
This is also true for 2011.
Figure 11-35. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
NBIL
S 1.0
5th Percentile
JL
T
T
2007 2008
Year
- Minimum
— Maximum
95th Percentile —O—Averagf
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 11-35 for carbon tetrachloride measurements collected at NBIL
include the following:
• The maximum concentration of carbon tetrachloride was measured in 2004
(4.81 jig/m3). No other measurements greater than 2.0 |ig/m3 have been measured at
NBIL and only one measurement greater than 1.5 |ig/m3 has been measured
(1.88 |ig/m3in2012).
• 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).
• After a slight decreasing trend between 2005 and 2007, the 1-year average increased
significantly for 2008. The 1-year average concentration exhibits a decreasing trend
after 2008 that continued through 2011, when the 1-year average reached a minimum
(0.64 |ig/m3). The median concentration exhibits a similar pattern.
11-47
-------
• Even though the difference between the minimum and maximum concentrations for
2012 is at the highest level since 2004, the difference between the 5th and 95th
percentiles is at a minimum. This indicates that the majority of concentrations are
falling within a relatively small range. That said, the increase shown for the 1-year
average and median is a result of an overall increase in the measurements rather than
the influence of the relatively high concentration measured in 2012. For example, the
number of concentrations greater than 0.70 |ig/m3 doubled from 201 1 to 2012 while
the number of measurements less than 0.50 |ig/m3 dropped from six to one for 2012.
Figure 11-36. Yearly Statistical Metrics for/7-Dichlorobenzene Concentrations Measured at
NBIL
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
O 5th Percentile - Minimum
Median — Maximum O 95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until April:
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 11-36 for/>-dichlorobenzene measurements collected at NBIL
include the following:
• The number of non-detects measured was greater than 95 percent for the first 2 years
of sampling. The number of non-detects decreased steadily through 2007, reaching a
minimum of 28 percent. After 2007, the percentage ranges from 39 percent (2009) to
73 percent (20 11).
• As the number of non-detects decreases through 2007, the range of concentrations
measured increases, resulting a dramatic increase in most of the statistical parameters
shown. All but two of the seven measurements greater than 1 |ig/m3 were measured in
2007, with the other two measured in 2004 and 2006.
11-48
-------
• The concentrations measured decreased significantly between 2007 and 2009. The
range of measurements increased by a factor of four between 2009 and 2010 then
returned to previous levels for 2011.
Figure 11-37. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at NBIL
o
2007 2008
Year
5th Percentile
- Minimum
— Maximum
95th Percentile —O—Averagf
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 11-37 for 1,2-dichloroethane measurements collected 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 for 2003 through 2008. The median concentration continued
to be zero for all years except 2012, 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 percentage of non-detects was at a minimum for 2012
(13 percent). As the number of measured detections increased, the 1-year average
concentrations exhibit significant increases.
11-49
-------
• For the first time, the median concentration is greater than zero for 2012 and is
greater than the 1-year average for 2012. This is because the eight non-detects (or
zeros) factored into the 1-year average concentration are pulling the average down
(just like a maximum or outlier concentration can pull the average up) and are not
contributing to the majority of measurements any longer.
Figure 11-38. Yearly Statistical Metrics for Fluoranthene Concentrations Measured at NBIL
B2
c
125'°
3 20.0
2010
Year
O 5th Percentile - Minimuir
— Maximum • 95th Percentile •••<>"• Averagf
A 1-year average is not presented because sampling under the NMP did not begin until June 2008.
Observations from Figure 11-38 for fluoranthene measurements collected at NBIL
include the following:
• The trends graph for fluoranthene resembles the trends graph for acenaphthene in that
the median concentration decreased significantly 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 (from nine to 27). Recall, however, that 2008 does not include a
full year's worth of sampling.
• Like acenaphthene, the 1-year average concentration of fluoranthene increases
between 2009 and 2011 and decreases slightly for 2012. 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.
• Although the maximum fluoranthene concentration was measured in 2012
(42.9 ng/m3), all of the other statistical parameters decreased at least slightly.
11-50
-------
Figure 11-39. Yearly Statistical Metrics for Fluorene Concentrations Measured at NBIL
~ 50.0
2010
Year
O 5th Percentile - Minimuir
— Maximum • 95th Percentile "-O-" Averagf
A 1-year average is not presented because sampling under the NMP did not begin until June 2008.
Observations from Figure 11-39 for fluorene measurements collected at NBIL include the
following:
• The statistical patterns for fluorene resemble the statistical patterns shown on the
trends graphs for acenaphthene and 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 2 ng/m3 more than doubled
from 2008 to 2009 (from six to 16). Recall, however, that 2008 does not include a full
year's worth of sampling.
• Like acenaphthene and fluoranthene, the 1-year average concentration of fluorene
increases between 2009 and 2011 and decreases slightly for 2012. 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 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.
11-51
-------
Figure 11-40. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at NBIL
2008 2009
Year
O 5th Percentile
— Minimum
— Maximum
0 95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until March 2005.
Observations from Figure 11-40 for formaldehyde measurements collected at NBIL
include the following:
• The maximum formaldehyde concentration was measured on January 5, 2006
(91.7 |ig/m3). However, the next five highest concentrations, ranging from 14.4 |ig/m3
to 53.5 |ig/m3, were all measured in 2010. The only other formaldehyde concentration
greater than 10 |ig/m3 was measured in 2011 (13.7 |ig/m3).
• The maximum concentration measured in 2006 is 20 times higher than the next
highest concentration measured that year (4.46 |ig/m3). The magnitude of this outlier
explains why the 1-year average concentration for 2006 is greater than the 95th
percentile.
• The 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 11-40, 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. This is because the number of concentrations in the middle of the concentration
11-52
-------
range increased. The number of measurements between 2 |ig/m3 and 4 |ig/m3 doubled
from 2011 to 2012.
Figure 11-41. Yearly Statistical Metrics for Manganese (PMio) Concentrations Measured at
NBIL
130-°
I
o
o
..o
2008 2009
Year
5th Percentile — Minimum ~ Median — Maximum
95th Percentile ...^... Average
Observations from Figure 11-41 for manganese (PMio) measurements collected at NBIL
include the following:
• The maximum manganese concentration was measured on August 26, 2005
(54.6 ng/m3). Concentrations in the 40 ng/m3 to 45 ng/m3 range have been measured
in 2005, 2008, and 2010.
• The 1-year average concentration decreased significantly from 2005 to 2006. The
1-year average increased from 2006 to 2007, then decreased between 2007 and 2009.
These changes, however, are statistically insignificant.
• After 2009, both the median and 1-year average concentrations increase steadily
through 2012. Even though the maximum and 95th percentile decreased for 2012, the
median and 1-year average increased. Although the minimum concentrations were
similar, there were fewer concentrations at the lower end of the concentration range
measured in 2012. The number of manganese measurements less than 3 ng/m3
decreased from 14 in 2011 to six in 2012. There were also more concentrations in the
mid- to upper-end of the concentration range. The number of measurements greater
than 20 ng/m3 increased from three in 2011 to five in 2012; the number of
measurements in the 10 ng/m3 to 20 ng/m3 range increased from nine in 2011 to 14 in
2012.
11-53
-------
Figure 11-42. Yearly Statistical Metrics for Naphthalene Concentrations Measured at NBIL
•••O
2010
Year
O 5th Percentile
— Maximum
O 95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until June 2008.
Observations from Figure 11-42 for naphthalene measurements collected at NBIL include
the following:
• The maximum naphthalene concentration was measured on September 23, 2010
(869 ng/m3). The next highest concentration measured in 2010 was considerably less
(363 ng/m3). All of the concentrations greater than 250 ng/m3 were measured in 2010
(four), 2011 (five), or 2012 (one).
• The 1-year average concentration of naphthalene increased between 2009 and 2010.
However, the large confidence interval calculated for 2010 indicates that the increase
is not statistically significant due to the relatively large amount of variability in the
2010 measurements. The range of naphthalene measurements for 2010 spans two
orders of magnitude, with a minimum concentration of 8.31 ng/m3 and a maximum
concentrations 869 ng/m3. The concentrations measured in 2011 also exhibit this type
of variability, ranging from 6.74 ng/m3 to 779 ng/m3.
• Most of the statistical parameters exhibit decreases for 2012. The maximum
concentration decreased by more than half from 2011 to 2012, and the 95th percentile
decrease by nearly 100 ng/m3. The decreases shown for the 1-year average and
median concentrations are less substantial.
11-54
-------
Figure 11-43. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SPIL
1
— 2— '
o-
^H
-r
2008 2009
Year
O 5th Percentile
— Maximum O 95th Percentile
A 1-year average is not presented because consistent sampling did not begin until March 2005.
Observations from Figure 11-43 for acetaldehyde measurements collected at SPIL
include the following:
• Although carbonyl compound sampling at SPIL began in early 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). Sixteen of the 18 concentrations of acetaldehyde
greater than 5 |ig/m3 were measured in 2011 (eight) or 2012 (eight), with the other
two measured in 2006.
• The 1-year average concentration decreased significantly from 2006 to 2007, then
held fairly steady through 2009. 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 41 in 2011.
• Although the maximum concentration increased from 2011 to 2012, most of the other
statistical parameters exhibit decreases.
11-55
-------
Figure 11-44. Yearly Statistical Metrics for Benzene Concentrations Measured at SPIL
•o..
±
T
2005 2006
2007 2008
Year
2011 2012
O 5th Percentile - Minimurr
— Maximum • 95th Percentile "-O-" Averagf
A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
Observations from Figure 11-44 for benzene measurements collected at SPIL include the
following:
• The only two concentrations of benzene greater than 5 |ig/m3 were both measured in
2005.
• The 1-year average benzene concentration has decreased over the years, reaching a
minimum of 0.68 |ig/m3 for 2009. The 1-year average concentration then increased
for 2010 (0.94 |ig/m3).
• Even though the maximum concentration increased for 2011 and again for 2012, the
majority of concentrations measured (as indicated by the 5th and 95th percentiles) fell
within roughly the same range. The 1-year average decreases just slightly for 2011
and returns to 2010 levels for 2012. The median concentration decreased slightly for
2011 then held steady for 2012.
11-56
-------
Figure 11-45. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at SPIL
O
2007 2008
Year
O 5th Percentile
- Minimum
— Maximum
O 95th Percentile ...<>... Average
A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
Observations from Figure 11-45 for 1,3-butadiene measurements at SPIL include the
following:
• The maximum concentration of 1,3-butadiene was measured at SPIL on
February 3, 2005 (1.29 |ig/m3) and is the only measurement greater than 1 |ig/m3. In
total, only six concentrations greater than 0.5 |ig/m3 have been measured at SPIL, one
in 2004, two in 2005, two in 2011, and one in 2012.
• The 1-year average concentrations of 1,3-butadiene decreased from 2006 through
2009. The increase from 2009 to 2010 is significant, representing a 67 percent
increase from 2009 levels. The median concentrations follow a similar pattern.
• The range of concentrations measured, as indicated by both the minimum and
maximum concentrations and the 5th and 95th percentiles, decreased after the initial
years of sampling through 2009, but increased significantly for 2010. Although the
maximum concentration increased for 2011, the range within which the majority of
concentrations fell stayed the same. Even though the maximum concentration
increased further for 2012, the range within which the majority of concentrations fall
decreased slightly.
• The detection rate for 1,3-butadiene has increased over time, ranging from
approximately 45 percent non-detects in 2004 to zero in 2008 and 2009, with one
non-detect each for 2010, 2011, and 2012.
11-57
-------
Figure 11-46. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
SPIL
i
i£i
rln
-------
Figure 11-47. Yearly Statistical Metrics for/7-Dichlorobenzene Concentrations Measured at SPIL
2007 2008
Year
O 5th Percentile
— Maximum
O 95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
Observations from Figure 11-47 for/>-dichlorobenzene measurements collected at SPIL
include the following:
• The median concentration for the first 3 years of sampling was zero, indicating that at
least half of the measurements are non-detects. Nearly 90 percent of the
measurements were non-detects for 2003 and 2004, after which the number of non-
detects began to decrease, reaching a minimum of 16 percent for 2007. After 2007,
the percentage of non-detects ranged from 27 percent (2009) to 55 percent (2010).
• The maximum concentration was measured at SPIL in 2008 (2.71 |ig/m3). Only two
additional concentrations greater than 1 |ig/m3 have been measured at SPIL, one in
2007 and one in 2009.
• The 1-year average concentration increased steadily through 2007, then decreased
steadily through 2010. An increase in the 1-year average concentration is shown for
2011 followed by another decrease. However, due to the wide range of concentrations
measured each year, the confidence intervals calculated are relatively large, indicating
a high level of variability in the measurements and that the changes are not
statistically significant.
• The difference between the 5th and 95th percentiles, or the range within which the
majority of concentrations fall, is at a minimum for 2012 (aside from 2004 when
95 percent of the measurements were non-detects).
11-59
-------
Figure 11-48. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at SPIL
I
FfrT
2007 2008
Year
O 5th Percentile
— Maximum
O 95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
Observations from Figure 11-48 for 1,2-dichloroethane measurements collected 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 number of non-detects was 95 percent or greater.
Thus, the minimum, 5th percentile, median, and in some cases, the 1-year average
concentrations were zero through 2009. The median concentration continued to be
zero for 2010 and 2011, indicating that at least half the measurements are non-detects.
However, 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.
• The maximum concentration of 1,2-dichloroethane was measured at SPIL in 2003
(0.75 |ig/m3). This is the only measured detection for 2003 as all other measurements
were non-detects. The next three highest concentrations (ranging from 0.12 |ig/m3 to
0.14 |ig/m3) were all measured in 2012, although similar concentrations were also
measured in 2009, 2010, and 2011.
• As the number of non-detects decreases and the number of measured detections
increase, the statistical parameters begin 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 the
sharp increase in the 1-year average concentration shown for 2012.
11-60
-------
Figure 11-49. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at SPIL
I
—
— <
•
L
t-
•
^
<
•
1 rh 1 .
, ,-i-, T *
--•••-.. r\ i
o- ^ -0 A
2007 2008
Year
O 5th Percentile
— Maximum
O 95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
Observations from Figure 11-49 for ethylbenzene measurements collected at SPIL
include the following:
• The maximum concentration of ethylbenzene was measured at SPIL in 2005
(2.39 |ig/m3), although a similar measurement was also collected in 2004. The five
highest concentrations of ethylbenzene were measured in 2004 or 2005.
• The 1-year average concentration has a steady decreasing trend between 2004 and
2009, although the largest decreases were between 2004 and 2006.
• The 1-year average increased significantly from 2009 to 2010, nearly doubling. The
range of measurements collected doubled from 2009 to 2010 as did the range within
which the majority of measurements fall.
• Little change in the measurements is shown after 2010.
11-61
-------
Figure 11-50. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at SPIL
o.,
—•—
2006
••^H • g&-
i
O 5th Percentile
— Minimum
— Maximum
0 95th Percentile
A 1-year average is not presented because consistent sampling did not begin until March 2005.
Observations from Figure 11-50 for formaldehyde measurements collected at SPIL
include the following:
• The maximum formaldehyde concentration (162 |ig/m3) was measured on
May 29, 2006 and is more than 10 times the maximum concentrations for any of the
other years shown in Figure 11-50 other than 2005. Of the 29 formaldehyde
concentrations greater than 15 |ig/m3, 12 were measured 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
2009. Although difficult to discern in Figure 11-50, an increasing trend in the 1-year
average concentration is shown between 2009 and 2011.
• Although the maximum concentration increased and the difference between the 5th
and 95th percentiles did not change, both the median and 1-year average
concentration exhibit slight decreases for 2012.
11-62
-------
Figure 11-51. Yearly Statistical Metrics for Trichloroethylene Concentrations Measured at SPIL
2007 2008
Year
O 5th Percentile
— Maximum
O 95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
Observations from Figure 11-51 for trichloroethylene measurements collected 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 14 percent
(2007) 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 measurement
(17.5 |ig/m3), which was measured in 2012.
• The concentrations of trichloroethylene exhibit considerable variability, as indicated
by confidence intervals calculated for the 1-year average concentrations, particularly
for 2012, where the maximum concentration is more than twice the maximum
concentrations for previous years (except 2003).
• The 1-year average concentrations have fluctuated between 0.34 |ig/m3 (2009) to
0.79 |ig/m3 (2010), with no distinct trend in the concentrations.
11-63
-------
11.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk at
each Illinois monitoring site. Refer to Sections 3.3 and 3.5.3 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.
11.5.1 Risk-Based Screening Assessment Using MRLs
A risk-based screening was conducted by comparing the concentration data from the
Illinois monitoring sites to the ATSDR MRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 day to 14 days); intermediate (exposures of 15 days to 364 days); and chronic
(exposures of 1 year or greater). The preprocessed daily measurements of the pollutants of
interest were compared to the acute MRLs; the quarterly averages were compared to the
intermediate MRLs; and the annual averages were compared to the chronic MRLs.
As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2012.
11.5.2 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 3.5.3.4 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 11-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
11-64
-------
Table 11-6. Risk Approximations for the Illinois Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections vs.
# of Samples
Annual
Average
(jig/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
Northbrook, Illinois - NBIL
Acenaphthene a
Acetaldehyde
Arsenic (PM10)a
Benzene
1,3 -Butadiene
Carbon Tetrachloride
/>-Dichlorobenzene
1 ,2-Dichloroethane
Fluoranthene a
Fluorene a
Formaldehyde
Manganese (PM10)a
Naphthalene a
0.000088
0.0000022
0.0043
0.0000078
0.00003
0.000006
0.000011
0.000026
0.000088
0.000088
0.000013
0.000034
0.009
0.000015
0.03
0.002
0.1
0.8
2.4
0.0098
0.00005
0.003
57/57
66/66
54/54
61/61
45/61
61/61
31/61
53/61
57/57
57/57
66/66
54/54
57/57
0.01
±0.01
1.78
±0.19
0.01
±0.01
0.64
±0.08
0.06
±0.02
0.71
±0.04
0.04
±0.01
0.07
±0.01
0.01
±O.01
0.01
±0.01
2.49
±0.33
0.01
±0.01
0.08
±0.02
1.01
3.92
3.14
5.01
1.87
4.27
0.47
1.81
0.62
1.08
32.31
2.65
0.20
0.05
0.02
0.03
0.01
O.01
0.01
0.25
0.18
0.03
Schiller Park, Illinois - SPIL
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
£>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
Trichloroethylene
0.0000022
0.0000078
0.00003
0.000006
0.000011
0.000026
0.0000025
0.000013
0.0000048
0.009
0.03
0.002
0.1
0.8
2.4
1
0.0098
0.002
61/61
60/60
59/60
60/60
39/60
55/60
60/60
61/61
49/60
2.72
±0.77
0.95
±0.19
0.14
±0.03
0.68
±0.03
0.05
±0.02
0.08
±0.01
0.29
±0.06
3.09
±0.58
0.71
±0.60
5.99
7.44
4.14
4.10
0.57
1.99
0.71
40.12
3.39
0.30
0.03
0.07
0.01
0.01
O.01
0.01
0.31
0.35
— = a Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 11-5.
NA = Not available due to the criteria for calculating an annual average.
11-65
-------
Table 11-6. Risk Approximations for the Illinois Monitoring Sites (Continued)
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections vs.
# of Samples
Annual
Average
(jig/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer Risk
Approximation
(HQ)
Roxana, Illinois - ROIL
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
£>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
0.0000022
0.0000078
0.00003
0.000006
0.000011
0.000026
0.0000025
0.000013
0.009
0.03
0.002
0.1
0.8
2.4
1
0.0098
35/35
33/33
31/33
33/33
22/33
29/33
33/33
35/35
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
— = a Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 11-5.
NA = Not available due to the criteria for calculating an annual average.
Observations for the Illinois sites from Table 11-6 include the following:
• Formaldehyde and acetaldehyde are the pollutants with the highest annual average
concentrations for both NBIL and SPIL, although the annual averages were higher for
SPIL.
• Formaldehyde has the highest cancer risk approximation for both sites (40.12 in-a-
million for SPIL and 32.31 in-a-million for NBIL). There were no other pollutants for
which a cancer risk approximation greater than 10 in-a-million was calculated.
• None of the pollutants of interest for NBIL or SPIL have noncancer hazard
approximations greater than 1.0, indicating that no adverse health effects are expected
from these individual pollutants. The pollutant with the highest noncancer hazard
approximation for NBIL is formaldehyde (0.25), although acetaldehyde has a
noncancer hazard approximation of similar magnitude (0.20). The pollutant with the
highest noncancer hazard approximation for SPIL is trichloroethylene (0.35),
although formaldehyde and acetaldehyde have noncancer hazard approximations of
similar magnitudes (0.31 and 0.30, respectively).
• Cancer risk and noncancer hazard approximations could not be calculated for ROIL
because annual average concentrations are not available.
11-66
-------
11.5.3 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 11-7 presents the 10 pollutants with the highest emissions from the 2011 NEI that have
cancer toxicity factors. Table 11-7 also presents the 10 pollutants with the highest toxicity -
weighted emissions, based on the weighting schema described in Section 3.5.3.5. Lastly,
Table 11-7 provides the 10 pollutants with the highest cancer risk approximations (in-a-million)
for each site, as presented in Table 11-6. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 11-7. Table 11-8 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 3.5.3.5. Similar to the cancer risk and
noncancer hazard approximations provided in Section 11.5.2, this analysis may help policy-
makers prioritize their air monitoring activities.
11-67
-------
Table 11-7. 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)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Northbrook, Illinois (Cook County) - NBIL
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
Tetrachloroethylene
1,3 -Butadiene
Trichloroethylene
Naphthalene
Dichloromethane
POM, Group 2b
961.69
821.78
565.60
463.56
257.15
155.11
99.56
90.61
35.41
15.48
Formaldehyde
Arsenic, PM
Benzene
Hexavalent Chromium, PM
1,3 -Butadiene
Naphthalene
Nickel, PM
Ethylbenzene
POM, Group 2b
Acetaldehyde
1.07E-02
8.12E-03
7.50E-03
5.42E-03
4.65E-03
3.08E-03
2.05E-03
1.41E-03
1.36E-03
1.02E-03
Formaldehyde
Benzene
Carbon Tetrachloride
Acetaldehyde
Arsenic
Naphthalene
1,3 -Butadiene
1 ,2-Dichloroethane
Fluorene
Acenaphthene
32.31
5.01
4.27
3.92
3.14
2.65
1.87
1.81
1.08
1.01
Schiller Park, Illinois (Cook County) - SPIL
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
Tetrachloroethylene
1,3 -Butadiene
Trichloroethylene
Naphthalene
Dichloromethane
POM, Group 2b
961.69
821.78
565.60
463.56
257.15
155.11
99.56
90.61
35.41
15.48
Formaldehyde
Arsenic, PM
Benzene
Hexavalent Chromium, PM
1,3 -Butadiene
Naphthalene
Nickel, PM
Ethylbenzene
POM, Group 2b
Acetaldehyde
1.07E-02
8.12E-03
7.50E-03
5.42E-03
4.65E-03
3.08E-03
2.05E-03
1.41E-03
1.36E-03
1.02E-03
Formaldehyde
Benzene
Acetaldehyde
1,3 -Butadiene
Carbon Tetrachloride
Trichloroethylene
1 ,2-Dichloroethane
Ethylbenzene
£>-Dichlorobenzene
40.12
7.44
5.99
4.14
4.10
3.39
1.99
0.71
0.57
oo
-------
Table 11-7. 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)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations
Based on Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
Roxana, Illinois (Madison County) - ROIL
Coke Oven Emissions, PM
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
Naphthalene
1,3 -Butadiene
Dichloromethane
POM, Group la
Tetrachloroethylene
137.23
115.44
114.84
50.92
48.47
13.98
13.21
12.11
6.59
3.60
Coke Oven Emissions, PM
Hexavalent Chromium, PM
Formaldehyde
Arsenic, PM
Benzene
POM, Group la
Naphthalene
1,3 -Butadiene
Nickel, PM
POM, Group 3
1.36E-01
8.20E-03
1.49E-03
1.03E-03
9.00E-04
5.80E-04
4.75E-04
3.96E-04
3.23E-04
2.95E-04
-------
Table 11-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Illinois Monitoring Sites
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Northbrook, Illinois (Cook County) - NBIL
Toluene
Ethylene glycol
Methanol
Hexane
Xylenes
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
Methyl isobutyl ketone
Toluene
Ethylene glycol
Methanol
Hexane
Xylenes
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
Methyl isobutyl ketone
9,526.78
6,437.35
3,042.16
2,625.12
2,404.38
961.69
821.78
565.60
463.56
342.65
Acrolein
Arsenic, PM
Cyanide Compounds, gas
Formaldehyde
1,3 -Butadiene
Manganese, PM
Cadmium, PM
Acetaldehyde
Trichloroethylene
Nickel, PM
3,182,872.70
125,965.03
86,973.50
83,855.56
77,557.35
70,630.83
52,253.81
51,506.84
49,780.32
47,566.89
Formaldehyde
Acetaldehyde
Manganese
Arsenic
1,3 -Butadiene
Naphthalene
Benzene
Carbon Tetrachloride
£>-Dichlorobenzene
1 ,2-Dichloroethane
0.25
0.20
0.18
0.05
0.03
0.03
0.02
0.01
<0.01
<0.01
Schiller Park, Illinois (Cook County) - SPIL
9,526.78
6,437.35
3,042.16
2,625.12
2,404.38
961.69
821.78
565.60
463.56
342.65
Acrolein
Arsenic, PM
Cyanide Compounds, gas
Formaldehyde
1,3 -Butadiene
Manganese, PM
Cadmium, PM
Acetaldehyde
Trichloroethylene
Nickel, PM
3,182,872.70
125,965.03
86,973.50
83,855.56
77,557.35
70,630.83
52,253.81
51,506.84
49,780.32
47,566.89
Trichloroethylene
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Benzene
Carbon Tetrachloride
Ethylbenzene
£>-Dichlorobenzene
1 ,2-Dichloroethane
0.35
0.31
0.30
0.07
0.03
0.01
<0.01
<0.01
<0.01
-------
Table 11-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Illinois Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Noncancer
Hazard
Approximation
Pollutant (HQ)
Roxana, Illinois (Madison County) - ROIL
Toluene
Ethylene glycol
Hexane
Xylenes
Methanol
Hydrochloric acid
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
600.92
331.04
190.00
187.92
178.11
128.58
115.44
114.84
50.92
48.47
Acrolein
Manganese, PM
Chlorine
Hexamethylene- 1 ,6-diisocyanate, gas
Arsenic, PM
Lead, PM
Formaldehyde
Cyanide Compounds, PM
Cyanide Compounds, gas
Nickel, PM
243,888.11
99,821.87
95,420.68
25,000.00
16,018.46
14,477.27
11,717.88
7,689.48
7,489.00
7,480.96
-------
Observations from Table 11-7 include the following:
• Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
cancer UREs in Cook County. Coke oven emissions is the highest emitted "pollutant"
with a cancer URE in Madison County, followed by benzene, formaldehyde, and
ethylbenzene.
• The pollutants with the highest toxi city-weighted emissions (of the pollutants with
cancer UREs) for Cook County are formaldehyde, arsenic, and benzene. Coke oven
emissions top Madison County's toxicity-weighted emissions, followed by
hexavalent chromium and formaldehyde.
• 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. Benzene, acetaldehyde, and 1,3-butadiene also
appear on all three list for both sites.
• Carbon tetrachloride, which has the third highest cancer risk approximation for NBIL
and sixth highest cancer risk approximation for SPIL, does not appear on either
emissions-based list.
• Trichloroethylene has the seventh highest cancer risk approximation for SPIL and is
the seventh highest emitted pollutant in Cook County, but does not appear among the
pollutants with the highest toxicity-weighted emissions (this pollutant ranks 14th).
• Several metals appear among the pollutants with the highest toxicity-weighted
emissions for Cook County, including arsenic, which has the fifth highest cancer risk
approximation for NBIL (SPIL did not sample metals). None of these metals appear
among the highest emitted pollutants for Cook County.
• POM, Group 2b ranks tenth for quantity emitted and ninth 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.
• NBIL is one of two NMP sites that sampled pollutants from all six methods. At least
one pollutant from each of the six methods appears among the pollutants with the
highest toxicity-weighted emissions.
• While seven of the 10 highest emitted pollutants in Madison County are sampled for
at ROIL, only three of the pollutants with the highest toxicity-weighted emissions are
sampled for at ROIL.
11-72
-------
Observations from Table 11-8 include the following:
• Toluene and ethylene glycol 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 two of the highest emitted pollutants also have the highest toxicity-weighted
emissions (formaldehyde, and acetaldehyde) for Cook County. The highest emitted
pollutants and the pollutants with the highest toxicity-weighted emissions for
Madison County have only one pollutant in common (formaldehyde).
• Formaldehyde and acetaldehyde have the highest noncancer hazard approximations
for NBIL (albeit less than an HQ of 1.0) and are the only pollutants that appear on
both emissions-based lists for Cook County.
• Trichloroethylene has the highest noncancer hazard approximation for SPIL; this
pollutant has the ninth highest toxicity-weighted emissions, but is not among the
highest emitted pollutants in Cook County (with a noncancer RfC).
• Several metals appear among the pollutants with the highest toxicity-weighted
emissions for Cook County, including manganese and arsenic, which have the third
and fourth highest noncancer hazard approximations for NBIL. (SPIL did not sample
metals). None of these metals appear among the highest emitted pollutants.
• While seven of the 10 highest emitted pollutants in Madison County (with noncancer
RfCs) are sampled for at ROIL, only two of the pollutants with the highest toxicity-
weighted emissions are sampled for at ROIL.
11.6 Summary of the 2012 Monitoring Data for NBIL, SPIL, and ROIL
Results from several of the data treatments described in this section include the
following:
»«» Twenty-two pollutants (two carbonyl compounds, 11 VOCs, five PAHs, three metals,
and hexavalent chromium) failed screens for NBIL; 13 pollutants (three carbonyl
compounds and 10 VOCs) failed screens for SPIL; and 11 pollutants (two carbonyl
compounds and nine VOCs) failed screens for ROIL.
»«» Formaldehyde had the highest annual average concentration among the pollutants of
interest for NBIL and SPIL. Although ROIL did not sample long enough for annual
averages to be calculated, the quarterly averages of acetaldehyde are significantly
higher than the quarterly averages for the remaining pollutants of interest.
11-73
-------
The maximum concentrations of several pollutants of interest across the program
were measured at the Chicago sites. The maximum concentrations ofacetaldehyde,
formaldehyde, and trichloroethylene program-wide were measured at SPIL. The
maximum concentrations offluorene andfluoranthene program-wide were measured
atNBIL.
Concentrations ofacetaldehyde and manganese have been increasing in recent years
atNBIL. Like many other NMP sites, a significant decrease in the number ofnon-
detects reported for 1,2-dichloroethane has occurred at both Chicago sites.
11-74
-------
12.0 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.
12.1 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 MSA, while a second site (WPIN) is located in the Indianapolis-Carmel-Anderson, IN MSA.
Figures 12-1 and 12-3 are composite satellite images retrieved from ArcGIS Explorer showing
the monitoring sites and their immediate surroundings. Figures 12-2 and 12-4 identify nearby
point source emissions locations by source category near INDEM and WPIN, respectively, as
reported in the 2011 NEI for point sources. Note that only sources within 10 miles of the sites are
included in the facility counts provided in Figures 12-2 and 12-4. 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
radius are still visible on each map, but have been grayed out in order to show emissions sources
just outside the boundary. Table 12-1 provides supplemental geographical information such as
land use, location setting, and locational coordinates.
12-1
-------
Figure 12-1. Gary, Indiana (INDEM) Monitoring Site
to
to
-------
Figure 12-2. NEI Point Sources Located Within 10 Miles of INDEM
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
Source Category Group (No. of Facilities)
"t" Airport/Airline/Airport Support Operations (14)
§ Asphalt Production/Hot Mix Asphalt Plant (1)
Brick. Structural Clay, or Clay Ceramics Plant (1)
B Bulk Terminals/Bulk Plants (5)
C Chemical Manufacturing (11)
S Coke Battery (2)
1 Compressor Station (6)
? Electricity Generation via Combustion (6)
E Electroplating, Plating, Polishing, Anodizing, and Coloring (2)
• Landfill (2)
• Metal Can. Box, and Other Metal Container Manufactunng (1)
A Metal Coating, Engraving, and Allied Services to Manufacturers (1)
© Metals Processing/Fabrication (7)
X Mine/Quarry/Mineral Processing (16)
? Miscellaneous Commercial/Industrial (20)
[] Paint and Coating Manufacturing (1)
* Petroleum Products Manufacturing (2)
a Petroleum Refinery (2)
R Plastic, Resin, or Rubber Products Plant (1)
P Printing/Publishing/Paper Product Manufacturing (1)
X Rail Yard/Rail Line Operations (6)
Steel Mill (10)
12-3
-------
Figure 12-3. Indianapolis, Indiana (WPIN) Monitoring Site
to
-------
Figure 12-4. NEI Point Sources Located Within 10 Miles of WPIN
Legend
86;10'0"W 86'5'0"W 86°0'0"W 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.
WPIN UATMP site O 10 mile radius
Source Cateogory Group (No. of Facilities)
County boundary
lj< Aerospace/Aircraft Manufacturing (2)
"f Airport/Airline/Airport Support Operations (26)
£ Asphalt Production/Hot Mix Asphalt Plant (1)
W Automobile/Truck Manufacturing (2)
B Bulk Terminals/Bulk Plants (3)
C Chemical Manufacturing (6)
NJ Coke Battery (1)
i Compressor Station (1)
8 Electrical Equipment Manufacturing (2)
f Electricity Generation via Combustion (3)
F Food Processing/Agriculture (2)
I Foundries, Iron and Steel (1)
O Institution (school, hospital, prison, etc.) (2}
• Landfill (1)
A Metal Coating. Engraving, and Allied Services to Manufacturers (2)
<•> Metals Processing/Fabrication (5)
? Miscellaneous Commercial/Industrial (4)
El Municipal Waste Combustor (1)
0 Paint and Coating Manufacturing (2)
cz> Pharmaceutical Manufacturing (2)
R Plastic. Resin, or Rubber Products Plant (3)
X Rail Yard/Rail Line Operations (1)
* Wastewater Treatment (1)
12-5
-------
Table 12-1. Geographical Information for the Indiana Monitoring Sites
Site
Code
INDEM
WPIN
AQS Code
18-089-0022
18-097-0078
Location
Gary
Indianapolis
County
Lake
Marion
Micro- or
Metropolitan
Statistical Area
Chicago-
Naperville-Elgin,
IL-IN-WI MSA
Indianapolis-
Carmel-Anderson,
IN MSA
Latitude
and
Longitude
41.606680,
-87.304729
39.811097,
-86.114469
Land Use
Industrial
Residential
Location
Setting
Urban/City
Center
Suburban
Additional Ambient Monitoring Information1
VOCs, SO2, NO, NO2, NOX, PAMS, O3,
Meteorological parameters, PM10, PM25, PM2.5
Speciation, IMPROVE Speciation.
TSP Metals, CO, VOCs, SNMOCs, SO2, NOy, NO,
O3, Meteorological parameters, PM10, PM25, PM25
Speciation, PM Coarse, IMPROVE Speciation.
Data for additional pollutants are reported to AQS for these sites (EPA, 2013b); however, these data are not generated by ERG and are therefore not included in this report
to
-------
INDEM is located in Gary, Indiana, a few miles east of the Indiana-Illinois border and
southeast of Chicago. Gary is located on the southernmost bank of Lake Michigan. The site is
located just north of 1-90, the edge of which can be seen in the bottom left portion of
Figure 12-1, and 1-65. Although INDEM resides on the Indiana Dunes National Lakeshore, the
surrounding area is highly industrialized, as shown in Figure 12-1, and several railroads
transverse the area. Figure 12-2 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 a mine/quarry
and another steel mill; a heliport at a police station and a hospital; and a mine/quarry.
WPIN is located in the parking lot of George Washington Park, near East 30th Street in
northeast Indianapolis. Figure 12-3 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 12-4 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, a heliport, and a fabricated metal products
facility. Each of these facilities is greater than 1 mile from WPIN but less than 2 miles out.
Table 12-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Indiana monitoring sites. Table 12-2 includes both county-level
population and vehicle registration information. Table 12-2 also contains traffic volume
information for each site as well as the location for which the traffic volume was obtained.
Additionally, Table 12-2 presents the county-level daily VMT for Marion and Lake Counties.
12-7
-------
Table 12-2. Population, Motor Vehicle, and Traffic Information for the Indiana Monitoring
Sites
Site
INDEM
WPIN
Estimated
County
Population1
493,618
918,977
County-level
Vehicle
Registration2
419,431
820,767
Annual
Average
Daily Traffic3
34,754
143,970
Intersection
Used for
Traffic Data
1-90, north of 1-65 interchange
1-70 between Exits 85 & 87
County-
level Daily
VMT4
16,226,000
32,005,000
Bounty-level population estimates reflect 2012 data (Census Bureau, 2013c)
2County-level vehicle registration reflects 2011 data (IN BMV, 2012)
3 AADT reflects 2011 data (IN DOT, 2011)
4County-level VMT reflects 2011 data (IN DOT, 2012)
Observations from Table 12-2 include the following:
• Marion County has almost twice the county-level population and vehicle registration
as Lake County.
• The county-level population for Marion County rounds out the top third among
county-level populations for other NMP sites, while the population for Lake County
is in the middle of the range. The county-level vehicle registrations mimic these
rankings.
• WPIN experiences a significantly higher traffic volume than INDEM. The traffic
estimate for WPIN is based on data from 1-70 between exits 85 and 87. Interstate-70
is just over 1 mile south of WPIN. Traffic data were not available for a location closer
to WPIN. The traffic volume near WPIN is the eighth highest among NMP sites.
• The traffic volume for INDEM is based on data from the 1-90 toll road. Traffic near
INDEM is in the middle of the range among traffic volumes for all NMP sites.
• The VMT for Marion County is almost twice the VMT for Lake County. The Marion
County VMT ranks 10th among counties with NMP sites, while the VMT for Lake
County is in the middle of the range (19th).
12.2 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.
12.2.1 Climate Summary
The city of Gary is located to the southeast of Chicago, at the southern-most tip of Lake
Michigan. Gary's proximity to Lake Michigan is an important factor controlling the weather of
the area. In the summer, warm temperatures can be suppressed, while cold winter temperatures
are often moderated. Winds that blow across Lake Michigan and over Gary in the winter can
12-8
-------
provide abundant amounts of lake-effect snow while lake breezes can bring relief from summer
heat (Wood, 2004; ISCO, 2002).
The city of Indianapolis is located in the center of Indiana, and experiences a temperate
continental climate and frequently changing weather patterns. Summers are warm and often
humid, as moist air flows northward out of the Gulf of Mexico. Winters are chilly with
occasional Arctic outbreaks. Precipitation is spread rather evenly throughout the year, with much
of the spring and summer precipitation resulting from showers and thunderstorms. Annual
snowfall totals average around 30 inches, with winters receiving less than 10 inches being
uncommon. The prevailing wind direction is southwesterly (Wood, 2004; ISCO, 2002).
12.2.2 Meteorological Summary
Hourly meteorological data for 2012 were retrieved NCDC for the weather stations
closest to the Indiana monitoring sites (NCDC, 2012), as described in Section 3.5.2. The two
closest weather stations are located at Lansing Municipal Airport (near INDEM) and Eagle
Creek Airpark (near WPIN), WBAN 04879 and 53842, respectively. Additional information
about these weather stations, such as the distance between the sites and the weather stations, is
provided in Table 12-3. These data were used to determine how meteorological conditions on
sample days vary from conditions experienced throughout the year.
Table 12-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for all of 2012. Also included in Table 12-3 is the 95 percent
confidence interval for each parameter. As shown in Table 12-3, average meteorological
conditions on sample days at WPIN and INDEM were representative of average weather
conditions experienced throughout the year near these locations. For both sites, the
meteorological parameter with the largest difference is relative humidity, although the
differences are not statistically significant.
12-9
-------
Table 12-3. Average Meteorological Conditions near the Indiana Monitoring Sites
Closest Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Gary, Indiana - INDEM |
Lansing Municipal
Airport
04879
(41.54, -87.52)
12.0
miles
242°
(WSW)
Sample
Days
(61)
2012
63.2
±5.1
62.5
±2.0
54.1
±4.5
53.5
±1.8
42.5
±3.9
42.6
±1.6
48.1
±3.8
47.9
± 1.6
68.6
±3.2
70.1
±1.3
NA
NA
7.3
±0.9
6.8
±0.4
Indianapolis, Indiana - WPIN
Eagle Creek
Airpark
53842
(39.83, -86.30)
9.1
miles
270°
(W)
Sample
Days
(63)
2012
64.4
±4.6
64.8
±2.0
55.4
±4.4
55.8
±1.8
42.1
±3.8
43.4
±1.6
48.6
±3.7
49.3
±1.6
64.6
±3.3
67.0
±1.3
1016.3
±1.5
1016.5
±0.6
5.8
±0.7
5.4
±0.3
to
o
Sample day averages are shaded in orange help differentiate the sample day averages from the full-year averages.
NA= Sea level pressure was not recorded at the Lansing Municipal Airport.
-------
12.2.3 Back Trajectory Analysis
Figure 12-5 is the composite back trajectory map for days on which samples were
collected at the INDEM monitoring site. Included in Figure 12-5 are four back trajectories per
sample day. Figure 12-6 is the corresponding cluster analysis. Similarly, Figure 12-7 is the
composite back trajectory map for days on which samples were collected at WPIN and
Figure 12-8 is the corresponding cluster analysis. An in-depth description of these maps and how
they were generated is presented in Section 3.5.2.1. For the composite maps, each line represents
the 24-hour trajectory along which a parcel of air traveled toward the monitoring site on a given
sample day and time, based on an initial height of 50 meters AGL. For the cluster analyses, each
line corresponds to a trajectory representative of a given cluster of back trajectories. Each
concentric circle around the sites in Figures 12-5 through 12-8 represents 100 miles.
Observations from Figures 12-5 and 12-6 for INDEM include the following:
• Back trajectories originated from a variety of directions at INDEM, with the south,
northwest, and north appearing to be the predominant directions of trajectory origin.
• The 24-hour air shed domain for INDEM was among the larger in size compared to
other NMP sites, with an average back trajectory length of 279 miles. The farthest
away a back trajectory originated was over central North Dakota, or nearly 800 miles
away. However, most trajectories (approximately 88 percent) originated within 450
miles of INDEM, with the longest trajectories originating from the northwest and
north.
• The cluster analysis shows that less than 20 percent of back trajectories originated to
the north of INDEM, primarily over eastern Wisconsin or Upper Peninsula of
Michigan, the Great Lakes, or northwest Michigan. Another 18 percent of back
trajectories originated to the northwest of INDEM. Back trajectories originating to the
west and southwest of INDEM, over Illinois, Iowa, and Missouri, account for more
than one-quarter of back trajectories.
• Back trajectories originating to the south of INDEM are split into two cluster
trajectories. Fifteen percent of back trajectories originated over western Kentucky and
Tennessee and are represented by the cluster trajectory originating south of Paducah,
Kentucky. This cluster includes back trajectories greater than 200 miles in length. The
short cluster trajectory originating over Indianapolis represents relatively short back
trajectories (generally less than 200 miles in length), originating to the east, southeast,
and south of INDEM, predominantly over western Ohio, northern Kentucky, and
Indiana.
12-11
-------
Figure 12-5. Composite Back Trajectory Map for INDEM
Figure 12-6. Back Trajectory Cluster Map for INDEM
12-12
-------
Figure 12-7. Composite Back Trajectory Map for WPIN
Figure 12-8. Back Trajectory Cluster Map for WPIN
12-13
-------
Observations from Figures 12-7 and 12-8 for WPIN include the following:
• The composite back trajectory map for WPIN resembles the composite back
trajectory map for INDEM in the geographic distribution of back trajectories.
• The 24-hour air shed domain for WPIN is similar in size to many other NMP
monitoring sites, with an average trajectory length of 252 miles. The farthest away a
back trajectory originated was over south-central North Dakota, or greater than
800 miles away, although most trajectories (nearly 90 percent) originated within
400 miles of WPIN. The longest back trajectories tended to originate from west,
northwest, and north.
• The cluster analysis for WPIN resembles the cluster analysis for INDEM. One major
difference is the additional cluster trajectory originating over the northwest corner of
Indiana and representing 17 percent of back trajectories. This cluster represents the
relatively short back trajectories originating over the northern half of Indiana and
Illinois. Common back trajectory origination includes from the north (10 percent),
northwest (17 percent), west and southwest (17 percent), and south (16 percent).
Similar to INDEM, the short cluster trajectory originating to the southeast of WPIN
represents the relatively short back trajectories (less than 250 miles in length)
originating from the northeast, east, southeast, and south of the monitoring site.
12.2.4 Wind Rose Comparison
Hourly surface wind data from the weather stations nearest the Indiana sites, as presented
in Section 12.2.2, were uploaded into a wind rose software program to produce customized wind
roses, as described in Section 3.5.2.2. 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 12-9 presents a map showing the distance between the weather station and
INDEM, which may be useful for identifying topographical influences that may affect the
meteorological patterns experienced at this location. Figure 12-9 also presents three different
wind roses for the INDEM monitoring site. First, a historical wind rose representing 2003 to
2011 wind data is presented, which shows the predominant surface wind speed and direction
over an extended period of time. Second, a wind rose representing wind observations for all of
2012 is presented. Next, a wind rose representing wind data for days on which samples were
collected in 2012 is presented. These can be used to identify the predominant wind speed and
direction for 2012 and to determine if wind observations on sample days were representative of
conditions experienced over the entire year and historically. Figure 12-10 presents the distance
map and three wind roses for WPIN.
12-14
-------
Figure 12-9. Wind Roses for the Lansing Municipal Airport Weather Station near INDEM
Location of INDEM and Weather Station
—
2003-2011 Historical Wind Rose
Calms: 19.5EW
2012 Wind Rose
Sample Day Wind Rose
Calms: 16.36%
12-15
-------
Figure 12-10. Wind Roses for the Indianapolis International Airport Weather Station near
WPIN
Location of WPIN and Weather Station
2002-2011 Historical Wind Rose
2012 Wind Rose
Sample Day Wind Rose
Calms: 1865%
12-16
-------
Observations from Figure 12-9 for INDEM include the following:
• The weather station at Lancing Municipal Airport is the closest weather station to
INDEM, although it is located 12 miles west-southwest of INDEM. The location of
the weather station is just east of the Illinois-Indiana state line and farther inland than
INDEM and thus, farther away from the influences of Lake Michigan than INDEM.
• The historical wind rose for INDEM shows that winds from the south to south-
southwest and west are the predominant wind directions over the 2003-2011 time
frame. Northerly to northeasterly winds off Lake Michigan accounted for less than
20 percent of the wind measurements, as did calm winds (< 2 knots). The strongest
winds were those from the south to southwest to west.
• The wind patterns shown on the 2012 wind rose generally resemble the wind patterns
shown on the historical wind rose. There were, however, slightly fewer calm winds
and a higher percentage of winds from the south to southwest.
• The differences in the wind patterns shown on the full-year wind rose continue on the
sample day wind rose. The calm rate is slightly lower and there is an even higher
percentage of winds from the south, south-southwest, and southwest.
Observations from Figure 12-10 for WPIN include the following:
• The weather station at Eagle Creek Airpark is the closest weather station to WPIN
and is located approximately 9 miles west of WPIN. Eagle Creek Airpark is located
on the southeast edge of the Eagle Creek Reservoir.
• Winds from the south, from the western quadrants, and from the north account for the
majority (nearly 60 percent) of wind observations from 2002 to 2011, while winds
from the eastern quadrants were observed for approximately one-quarter of the
observations. Calm winds were observed for 18 percent of observations. The
strongest winds tended to flow from the northwest.
• The wind patterns on the 2012 wind rose resemble the historical wind patterns,
although there were more southerly and south-southwesterly winds and fewer
southwesterly and west-southwesterly winds. The calm rate was also higher
(accounting for nearly 22 percent of observations).
• The sample day wind rose resembles the full-year wind rose but with fewer calm
winds and a higher percentage of winds from the south to southwest and west to
northwest.
12-17
-------
12.3 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 12-4. 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 12-4. It is important to note which pollutants were sampled for at each site when reviewing
the results of this analysis. INDEM and WPIN sampled for carbonyl compounds only.
Table 12-4. Risk-Based Screening Results for the Indiana Monitoring Sites
Pollutant
Screening
Value
(Ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Gary, Indiana - INDEM
Formaldehyde
Acetaldehyde
0.077
0.45
Total
59
58
117
59
59
118
100.00
98.31
99.15
50.43
49.57
50.43
100.00
Indianapolis, Indiana - WPIN
Acetaldehyde
Formaldehyde
Propionaldehyde
0.45
0.077
0.8
Total
58
58
1
117
58
58
58
174
100.00
100.00
1.72
67.24
49.57
49.57
0.85
49.57
99.15
100.00
Observations from Table 12-4 include the following:
• Formaldehyde, acetaldehyde, and propionaldehyde are the only carbonyl compounds
with risk screening values.
• Acetaldehyde and formaldehyde failed screens for INDEM. Acetaldehyde failed 58
out of 59 screens while formaldehyde failed 100 percent of screens for this site. Both
pollutants were identified as pollutants of interest for INDEM.
• All three carbonyl compounds with risk screening values failed screens for WPIN.
Acetaldehyde and formaldehyde each failed 100 percent of screens while
propionaldehyde failed only one screen. Acetaldehyde and formaldehyde were also
identified as pollutants of interest for WPIN.
12-18
-------
12.4 Concentrations
This section presents various concentration averages used to characterize pollution 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. Additional site-specific statistical summaries are provided
in Appendix L.
12.4.1 2012 Concentration Averages
Quarterly and annual concentration averages were calculated for the pollutants of interest
for each Indiana site, as described in Section 3.1. The quarterly average 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 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 Indiana monitoring
sites are presented in Table 12-5, 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.
12-19
-------
Table 12-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest
for the Indiana Monitoring Sites
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Ug/m3)
2nd
Quarter
Average
(Ug/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Ug/m3)
Gary, Indiana - INDEM
Acetaldehyde
Formaldehyde
59/59
59/59
1.06
±0.16
1.82
±0.33
1.25
±0.26
3.16
±0.87
1.44
±0.25
3.32
±0.60
1.04
±0.29
1.66
±0.29
1.20
±0.12
2.50
±0.33
Indianapolis, Indiana - WPIN |
Acetaldehyde
Formaldehyde
58/58
58/58
1.62
±0.57
3.07
± 1.26
2.39
±0.45
5.19
± 1.21
3.17
±0.48
5.91
±0.99
1.81
±0.32
2.73
±0.45
2.28
±0.27
4.31
±0.61
Observations for the Indiana sites from Table 12-5 include the following:
• For both sites, acetaldehyde and formaldehyde were detected in all of the carbonyl
compound samples collected.
• The annual average concentration of formaldehyde is greater than the annual average
concentration of acetaldehyde for INDEM. The same is true for WPIN. In both cases,
the acetaldehyde averages are almost half the formaldehyde average.
• The annual average concentrations of acetaldehyde and formaldehyde are higher at
WPIN than INDEM.
• The second and third quarter average concentrations of formaldehyde are
significantly higher than the first and fourth quarter averages for INDEM. This is also
true for acetaldehyde, although the differences are not statistically significant. The 13
highest formaldehyde concentrations were measured between May and September at
INDEM and ranged from 3.46 |ig/m3 to 6.83 |ig/m3; conversely, the 19 lowest
concentrations (those less than 1.70 |ig/m3) were measured between January and
April or October and December. This supports the trend identified in Section 4.4.2
where formaldehyde concentrations tended to be higher during the warmer months of
the year.
• With the exception of the fourth quarter, the quarterly averages of formaldehyde for
WPIN have rather large confidence intervals associated with them. A review of the
data shows that the two highest formaldehyde concentrations for WPIN were
measured on March 10, 2012 (10.7 |ig/m3) and May 27, 2012 (10.5 |ig/m3). Fifteen
additional concentrations of formaldehyde greater than 6 jig/m3 were measured at
WPIN between May and September. Conversely, the 19 lowest concentrations of
formaldehyde (those less than 3 |ig/m3) were measured between January and April or
October through December, again supporting the seasonal trend discussed in
Section 4.4.2.
12-20
-------
Tables 4-9 through 4-12 present the 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:
• WPIN's annual average concentration of formaldehyde is the second highest annual
average of this pollutant among NMP sites sampling carbonyl compounds, behind
only BTUT. The confidence interval for WPIN's annual average is among the largest
shown in Table 4-10, indicating a relatively high level of variability in this site's
measurements. Concentrations measured at this site range from 1.62 |ig/m3 to
10.7 |ig/m3, with a median concentration of 3.58 |ig/m3.
• 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.
12.4.2 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 shaded in
gray in Table 12-4 for INDEM and WPIN. Figures 12-11 and 12-12 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 3.5.3.1.
Figure 12-11. Program vs. Site-Specific Average Acetaldehyde Concentrations
INDEM
•H-
«•
9 12
Concentration {[og/m3)
Program: 1st Quartile 2nd Quartile 3rd Quartile 4th Quartile Average
Site:
Site Average Site Concentration Range
o
12-21
-------
Figure 12-12. Program vs. Site-Specific Average Formaldehyde Concentrations
INDEM
WPIN
10
12
14
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile 3rd Quartile 4thQuartile Average
Site: Site Average Site Concentration Range
o —
Observations from Figures 12-11 and 12-12 include the following:
• Figure 12-11 presents the box plots for both sites for acetaldehyde. The box plots
show that the annual average concentration for INDEM is less than both the
program-level average and median concentrations. The maximum concentration
of acetaldehyde measured at INDEM is similar to the annual average
concentration for WPIN. WPIN's annual average is greater than the program-
level average and third quartile. The minimum concentration measured at WPIN
is similar to the program-level first quartile.
• Figure 12-12 presents the box plots for formaldehyde for both sites. Although the
range of concentrations measured at each site is higher for formaldehyde than
acetaldehyde, these box plots share similarities with the acetaldehyde box plots.
The annual average concentration for INDEM is less than the program-level
average while the annual average for WPIN is greater than both the program-level
average and third quartile. Although the maximum formaldehyde concentration
measured at WPIN is not the maximum concentration measured across the
program, it is among the top 10. The minimum formaldehyde concentration
measured at WPIN is also greater than the program-level first quartile.
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 3.5.3.2.
INDEM and WPIN have sampled carbonyl compounds under the NMP since 2004 and 2007,
respectively. Thus, Figures 12-13 through 12-16 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
12-22
-------
minimum of 6 months of sampling is required for inclusion in the trends analysis; in these cases,
a 1-year average is not provided, although the range and quartiles are still presented.
Figure 12-13. Yearly Statistical Metrics for Acetaldehyde Concentrations
Measured at INDEM
T
T
2004 2005 2006 2007 2008 2009 2010 2011 2012
O 5th Percentile - Minimurr
Median — Maximum • 95th Percentile
A 1-year average is not presented due to a break in sampling between September 2005 and November
2005.
Observations from Figure 12-13 for acetaldehyde measurements collected at INDEM
include the following:
• Although carbonyl compound sampling under the NMP began in 2003, samples were
only collected for three months. Carbonyl compound sampling began in earnest at
INDEM at the beginning of 2004; thus, Figure 12-13 begins with 2004. However, a
1-year average is not presented for 2005 due to a break in sampling between
September 2005 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
12-23
-------
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 1.50 |ig/m3 increased from zero for
2007 to 15 for 2008 and the number of concentrations between 1.50 |ig/m3 and
2 |ig/m3 increased from three to six.
• 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 impact on the concentrations
measured, particularly with respect to formaldehyde, which is discussed in more
detail below.
• The statistical parameters shown for 2010, 2011, and 2012 are similar in magnitude to
each other (although the maximum concentration decreased for 2012). The 1-year
averages range from 1.20 |ig/m3 (2012) to 1.39 |ig/m3 (2010) over the period from
2009 to 2012.
Figure 12-14. Yearly Statistical Metrics for Formaldehyde Concentrations
Measured at INDEM
^J
2
2007 2008 2009
Year
5th Percentile - Minimurr
Median — Maximum • 95th Percentile ...^>... Average
1 A 1-year average is not presented due to a break in sampling between September 2005 and November
2005.
Observations from Figure 12-14 for formaldehyde measurements collected at INDEM
include the following:
12-24
-------
• Five formaldehyde concentrations greater than 400 |ig/m3 were measured in the
summer of 2008 (ranging from 414 |ig/m3to 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 12-14. 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 100 |ig/m3.
Further, 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.
• 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 the outliers is driving the 1-year average concentration upward, there were a larger
number of concentrations at the lower end of the concentration range as well. For
2008, 40 percent of measurements were less than 5 |ig/m3; for the years prior to 2008,
the number of measurements less than 5 |ig/m3 ranged from zero (2006) to three
(2005).
• All the statistical metrics decreased significantly for 2009 and the years that follow.
The 1-year average concentration ranged from 2.30 |ig/m3 (2011) to 2.58 |ig/m3
(2009). In contrast to the previous bullet, the number of measurements greater than
5 |ig/m3 ranged from one to four for each year between 2009 and 2012 (with the most
measured in 2012).
• 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 for 2009-2012
compared to previous years. 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 number of concentrations greater than
5 |ig/m3 before and after 2009, as discussed in the previous bullets.
12-25
-------
Figure 12-15. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at WPIN
~ 4.0
o
2009 2010
Year
O 5th Percentile - Minimurr
Median — Maximum O 95th Percentile
Observations from Figure 12-15 for acetaldehyde measurements collected 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 12-15 begins with 2007.
• The three highest acetaldehyde concentrations were measured at WPIN in 2010 and
ranged from 5.96 |ig/m3to 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 increased,
particularly the maximum (which doubled) and the 95th percentile (which increased
by 60 percent). The 1-year average has a slight decreasing trend again after 2010.
12-26
-------
Figure 12-16. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at WPIN
2009 2010
Year
O 5th Percentile - Minimurr
Median — Maximum O 95th Percentile
Observations from Figure 12-16 for formaldehyde measurements collected 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/m3to 10.7 |ig/m3.
• The 1-year average concentration has a decreasing trend through 2009, similar to
acetaldehyde, after which an increasing trend is shown through 2011. Although the
95th percentile increased for 2012 and the 1-year average did not change
significantly, the median concentration decreased. A review of the data for 2011 and
2012 shows that the number of concentrations in the 3 |ig/m3 to 4 |ig/m3 range
doubled from 2011 to 2012 (from seven to 15); in addition, the number of
concentrations in the 4 |ig/m3 to 6 |ig/m3 range decreased by nearly half (from 20 to
11). 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.
12-27
-------
12.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk at
each Indiana monitoring site. Refer to Sections 3.3 and 3.5.3 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.
12.5.1 Risk-Based Screening Assessment Using MRLs
A risk-based screening was conducted by comparing the concentration data from the
Indiana monitoring sites to the ATSDR MRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 day to 14 days); intermediate (exposures of 15 days to 364 days); and chronic
(exposures of 1 year or greater). The preprocessed daily measurements of the pollutants of
interest were compared to the acute MRLs; the quarterly averages were compared to the
intermediate MRLs; and the annual averages were compared to the chronic MRLs.
As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2012.
12.5.2 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 3.5.3.4 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.
12-28
-------
Table 12-6. Risk Approximations for the Indiana Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer Risk
Approximation
(HQ)
Gary, Indiana - INDEM
Acetaldehyde
Formaldehyde
0.0000022
0.000013
0.009
0.0098
59/59
59/59
1.20
±0.12
2.50
±0.33
2.65
32.55
0.13
0.26
Indianapolis, Indiana - WPIN
Acetaldehyde
Formaldehyde
0.0000022
0.000013
0.009
0.0098
58/58
58/58
2.28
±0.27
4.31
±0.61
5.01
56.06
0.25
0.44
Observations for the Indiana sites from Table 12-6 include the following:
• For both sites, the annual average concentration of formaldehyde is greater than the
annual average concentration of acetaldehyde. The annual averages for WPIN are
greater than the annual averages for INDEM.
• 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
approximations for WPIN are nearly twice the cancer risk approximations for
INDEM.
• The cancer risk approximation for formaldehyde for WPIN is the second highest
cancer risk approximation among all pollutants of interest program-wide.
• Neither pollutant of interest for INDEM or WPIN have noncancer hazard
approximations greater than 1.0, indicating that no adverse health effects are expected
from these individual pollutants. The noncancer hazard approximation for WPIN
ranks seventh highest among all pollutants of interest program-wide.
12.5.3 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 12-7 presents the 10 pollutants with the highest emissions from the 2011 NEI 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 3.5.3.5. 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
12-29
-------
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 3.5.3.5. Similar to the cancer risk and
noncancer hazard approximations provided in Section 12.5.2, this analysis may help policy-
makers prioritize their air monitoring activities.
Observations from Table 12-7 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
roughly twice as high in Marion County.
• Coke oven emissions, formaldehyde, and POM, Group Ib are the pollutants with the
highest toxicity-weighted emissions (of the pollutants with cancer UREs) for Lake
County. Formaldehyde, benzene, and hexavalent chromium are the pollutants with
the highest toxicity-weighted emissions for Marion County.
• Seven of the highest emitted pollutants in Lake County also have the highest toxicity-
weighted emissions; six 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.
• 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 toxic.
• 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.
12-30
-------
Table 12-7. 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)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
Gary, Indiana (Lake County) - INDEM
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
POM, Group Ib
Naphthalene
Coke Oven Emissions, PM
POM, Group 2d
POM, Group 2b
153.11
135.29
81.19
73.26
23.02
21.84
12.14
2.41
2.39
2.02
Coke Oven Emissions, PM
Formaldehyde
POM, Group Ib
Benzene
Hexavalent Chromium, PM
1,3 -Butadiene
Arsenic, PM
Naphthalene
Nickel, PM
POM, Group 2d
2.38E-03
1.76E-03
1.20E-03
1.19E-03
1.02E-03
6.91E-04
6.55E-04
4.13E-04
2.50E-04
2.10E-04
Formaldehyde 32.55
Acetaldehyde 2.65
Indianapolis, Indiana (Marion County) - WPIN
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
Tetrachloroethylene
POM, Group 2b
Propylene oxide
Trichloroethylene
364.14
306.72
235.70
177.27
56.98
32.73
16.40
5.64
4.72
4.71
Formaldehyde
Benzene
Hexavalent Chromium, PM
1,3 -Butadiene
Naphthalene
Arsenic, PM
Ethylbenzene
POM, Group 2b
Nickel, PM
POM, Group 3
3.99E-03
2.84E-03
2.41E-03
1.71E-03
1.11E-03
1.10E-03
5.89E-04
4.97E-04
4.60E-04
4.09E-04
Formaldehyde 56.06
Acetaldehyde 5.01
to
-------
Table 12-8. 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)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted
Emissions (County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Noncancer Hazard
Approximation
Pollutant (HQ)
Gary, Indiana (Lake County) - INDEM
Toluene
Ethylene glycol
Hexane
Xylenes
Methanol
Hydrochloric acid
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,162.14
610.79
421.01
390.84
327.88
323.74
153.11
135.29
81.19
73.26
Acrolein
Manganese, PM
Lead, PM
Hydrochloric acid
Formaldehyde
1,3 -Butadiene
Chlorine
Arsenic, PM
Acetaldehyde
Nickel, PM
412,327.83
134,984.66
52,706.82
16,187.23
13,805.18
11,511.05
11,183.33
10,154.19
8,139.87
5,794.59
Formaldehyde 0.26
Acetaldehyde 0.13
Indianapolis, Indiana (Marion County) - WPIN
Toluene
Ethylene glycol
Xylenes
Hexane
Methanol
Hydrochloric acid
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
2,485.16
1,135.51
889.12
741.29
532.81
466.75
364.14
306.72
235.70
177.27
Acrolein
Formaldehyde
1,3 -Butadiene
Hydrochloric acid
Acetaldehyde
Arsenic, PM
Lead, PM
Benzene
Naphthalene
Nickel, PM
1,142,806.73
31,298.11
28,488.29
23,337.36
19,697.08
17,076.36
13,691.58
12,138.08
10,908.60
10,648.54
Formaldehyde 0.44
Acetaldehyde 0.25
to
OJ
to
-------
Observations from Table 12-8 include the following:
• While toluene is the highest emitted pollutant with a noncancer RfC in both counties,
the toluene emissions in Marion County are more than twice that of Lake County.
Ethylene glycol is the second highest emitted pollutant in both counties, with a
similar pattern in the quantity emitted.
• Acrolein is the pollutant with the highest toxi city-weighted emissions (of the
pollutants with noncancer RfCs) for both counties. Manganese and lead rank second
and third for Lake County, while formaldehyde and 1,3-butadiene rank second and
third for Marion County.
• Only three of the highest emitted pollutants in Lake County also have the highest
toxicity-weighted emissions (formaldehyde, acetaldehyde, and hydrochloric acid).
Several metals (manganese, lead, nickel, and arsenic) are among the pollutants with
the highest toxicity-weighted emissions for Lake County, although none of these
appear among the highest emitted pollutants.
• Four of the highest emitted pollutants in Marion County also have the highest
toxicity-weighted emissions (formaldehyde, acetaldehyde, hydrochloric acid, and
benzene). Nickel, lead, and arsenic are also among the pollutants with the highest
toxicity-weighted emissions for Marion County, although none of these appear
among the highest emitted pollutants.
12.6 Summary of the 2012 Monitoring Data for INDEM and WPIN
Results from several of the data treatments described in this section include the
following:
»«» Two carbonyl compounds failed screens for INDEM and three failed screens for
WPIN.
»«» The annual average concentration of formaldehyde is greater than the annual
average concentration of acetaldehyde for both sites, with the annual averages for
WPIN nearly twice the annual averages for INDEM.
»«» Concentrations of formaldehyde and acetaldehyde exhibited a significant decreasing
trend at INDEM from 2008 to 2009, after which concentrations appear to be holding
steady. These changes may be explained by a sampler change out.
12-33
-------
13.0 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.
13.1 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. 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. Five monitoring sites are located in or
near the Calvert City area. 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 13-1
through 13-15. 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. 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 radius are still visible on the maps, but have been grayed out in order to show emissions
sources just outside the boundary. Table 13-1 provides supplemental geographical information
such as land use, location setting, and locational coordinates for each site.
13-1
-------
Figure 13-1. Ashland, Kentucky (ASKY) Monitoring Site
~^~
-------
Figure 13-2. Ashland, Kentucky (ASKY-M) Monitoring Site
-------
Figure 13-3. NEI Point Sources Located Within 10 Miles of ASKY and ASKY-M
Legend
82'45'O'W 82'40'0"W 82'35'0-W 82'30'0'W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
•ff
ASKY UATMP site
ASKY-M UATMP site O 10 mile radius
County boundary
Source Category Group (No. of Facilities)
t Airport/Airline/Airport Support Operations (6)
V Asphalt Production/Hot Mix Asphalt Plant (1)
B Bulk Terminals/Bulk Plants (1)
c Chemical Manufacturing (5)
S Coke Battery (1)
i Compressor Station (3)
* Electricity Generation via Combustion (3)
F Food Processing/Agriculture (2)
If Gasoline/Diesel Service Station (1)
• Landfill (2)
© Metals Processing/Fabrication (2)
Mine/Quarry/Mineral Processing (8)
Miscellaneous Commercial/Industrial (3)
Paint and Coating Manufacturing (1)
Pesticide Manufacturing Plant (1)
Petroleum Refinery (1)
Plastic, Resin, or Rubber Products Plant (2)
Port and Harbor Operations (2)
Rail Yard/Rail Line Operations (2)
Ship/Boat Manufacturing or Repair (1)
Steel Mill (2)
Testing Laboratories (1)
13-4
-------
Figure 13-4. Grayson, Kentucky (GLKY) Monitoring Site
-------
Figure 13-5. NEI Point Sources Located Within 10 Miles of GLKY
Legend
83"5'0-W 83"0'0"W 82"55'(rW 62'50'0'W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
GLKY NATTS site O 10 mile radius
County boundary
Source Category Group (No. of Facilities)
i Asphalt Production/Hot Mix Asphalt Plant (1)
Brick, Structural Clay, or Clay Ceramics Plant (2)
B Bulk Terminals/Bulk Plants (1)
F Food Processing/Agriculture (1)
o Institution (school, hospital, prison, etc.) (1)
x Mine/Quarry/Mineral Processing (2)
13-6
-------
Figure 13-6. Baskett, Kentucky (BAKY) Monitoring Site
-------
Figure 13-7. NEI Point Sources Located Within 10 Miles of BAKY
i7"20'0"W 87 15'0"W
87:35'0"W 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.
Legend
>{ BAKY UATMP site O 10 mile radius
County boundary
Source Category Group (No. of Facilities)
f Airport/Airline/Airport Support Operations (6)
i Asphalt Production/Hot Mix Asphalt Plant (1)
B Bulk Terminals/Bulk Plants (1)
c Chemical Manufacturing (2)
® Dry Cleaning (2)
* Electricity Generation via Combustion (4)
E Electroplating, Plating, Polishing, Anodizing, and Coloring (1)
F Food Processing/Agriculture (3)
© Metals Processing/Fabrication (5)
x Mine/Quarry/Mineral Processing (3)
? Miscellaneous Commercial/Industrial (3)
D Paint and Coating Manufacturing (1)
R Plastic, Resin, or Rubber Products Plant (1)
B Pulp and Paper Plant (2)
x Rail Yard/Rail Line Operations (1)
A. Ship/Boat Manufacturing or Repair (1)
® Testing Laboratories (1)
13-8
-------
Figure 13-8. Calvert City, Kentucky (ATKY) Monitoring Site
-------
Figure 13-9. Smithland, Kentucky (BLKY) Monitoring Site
-------
Figure 13-10. Calvert City, Kentucky (CCKY) Monitoring Site
-------
Figure 13-11. Calvert City, Kentucky (LAKY) Monitoring Site
to
-------
Figure 13-12. Calvert City, Kentucky (TVKY) Monitoring Site
-------
Figure 13-13. NEI Point Sources Located Within 10 Miles of ATKY, BLKY, CCKY,
LAKY, and TVKY
Legend
ATKY UATMP site
8°25'0"W 8ar20'0"W 88J15'0"W 88"1010"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
CCKY UATMP site
BLKY UATMP site yf LAKY UATMP site
Source Category Group (No. of Facilities)
TVKY UATMP site
O 10 mile radius
| County boundary
t Airport/Airline/Airport Support Operations (1) © Metals Processing/Fabrication (3)
'i Asphalt Production/Hot Mix Asphalt Plant (1) .< Mine/Quarry/Mineral Processing (7)
B Bulk Terminals/Bulk Plants (1)
C Chemical Manufacturing (10)
i Compressor Station (1)
t Electricity Generation via Combustion (1)
+ Industrial Machinery or Equipment Plant (1)
? Miscellaneous Commercial/Industrial (5)
R Plastic, Resin, or Rubber Products Plant (4)
^ Ship/Boat Manufacturing or Repair (4)
W Steel Mill (1)
13-14
-------
Figure 13-14. Lexington, Kentucky (LEKY) Monitoring Site
-------
Figure 13-15. NEI Point Sources Located Within 10 Miles of LEKY
8435'CrW W30'0"W 84'25'0"W 64 20'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
LEKY UATMP site
10 mile radius
County boundary
Source Category Group (No. of Facilities)
f 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)
H Automobile/Truck Manufacturing (4)
B Bulk Terminals/Bulk Plants (3)
XI Crematory -Animal/Human (1)
® Dry Cleaning (1)
6 Electrical Equipment Manufacturing (3)
f Electricity Generation via Combustion (1)
E Electroplating, Plating, Polishing, Anodizing, and Coloring (1)
F Food Processing/Agriculture (2)
$ Glass Plant (1)
+ Industrial Machinery or Equipment Plant (6)
O Institution (school, hospital, prison, etc.) (10)
• Landfill (2)
<•) Metals Processing/Fabrication (2)
A Military Base/National Security (1)
x Mine/Quarry/Mineral Processing (13)
? Miscellaneous Commercial/Industrial (4)
D Paint and Coating Manufacturing (1)
R Plastic, Resin, or Rubber Products Plant (3)
P Printing/Publishing/Paper Product Manufacturing (3)
TT Telecommunications/Radio (1)
M Tobacco Manufacturing (1)
W Woodwork, Furniture, Millwork & Wood Preserving (4)
13-16
-------
Table 13-1. Geographical Information for the Kentucky Monitoring Sites
Site Code
ASKY
ASKY-M
GLKY
BAKY
ATKY
BLKY
CCKY
LAKY
TVKY
LEKY
AQS Code
21-019-0017
21-019-0002
21-043-0500
21-101-0014
21-157-0016
21-139-0004
21-157-0018
21-157-0019
21-157-0014
21-067-0012
Location
Ashland
Ashland
Gray son
Baskett
Calvert
City
Smithland
Calvert
City
Calvert
City
Calvert
City
Lexington
County
Boyd
Boyd
Carter
Henderson
Marshall
Livingston
Marshall
Marshall
Marshall
Fayette
Micro- or
Metropolitan
Statistical Area
Huntington-
Ashland, WV-
KY-OH
Huntington-
Ashland, WV-
KY-OH
Not in an MSA
Evansville, IN-
KY
Not in an MSA
Paducah, KY-IL
Not in an MSA
Not in an MSA
Not in an MSA
Lexington-
Fayette, KY
Latitude
and
Longitude
38.45934,
-82.64041
38.476,
-82.63137
38.23887,
-82.9881
37.8712,
-87.46375
37.04176,
-88.35407
37.07151,
-88.33389
37.02702,
-88.34387
37.03718,
-88.33411
37.0452,
-88.33087
38.06503,
-84.49761
Land Use
Residential
Industrial
Residential
Commercial
Industrial
Agricultural
Residential
Residential
Industrial
Residential
Location
Setting
Suburban
Urban/City
Center
Rural
Rural
Suburban
Rural
Suburban
Suburban
Suburban
Suburban
Additional Ambient Monitoring Information1
SO2, NO, NO2, O3, Meteorological parameters,
PM2 5, PM2 5 Speciation, IMPROVE Speciation.
PM10.
O3, Meteorological parameters, PM10, PM25, and
PM25 Speciation, IMPROVE Speciation.
SO2, O3, Meteorological parameters, PM10, PM2 5.
None.
Meteorological parameters.
Meteorological parameters, PM10.
None.
None.
SO2, NO, NO2, O3, Meteorological parameters,
PM10, PM25, PM25 Speciation, IMPROVE
Speciation.
Data for additional pollutants are reported to AQS for these sites (EPA, 2013b); however, these data are not generated by ERG and are therefore not included in this report.
BOLD ITALICS = EPA-de signaled 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 MSA. The ASKY site is located behind the
county health department which is nestled in a residential area in the center of town, as shown in
Figure 13-1. The ASKY-M site is located on the roof of an oil company complex in the northern
part of Ashland, which is more industrial. The monitoring site is located less than one-quarter
mile from the Ohio River, and between the two lie a rail yard, a scrap yard, and other industries,
as shown in Figure 13-2. The ASKY-M monitoring site is located on Greenup Road (Route
60/23), a major thoroughfare through downtown Ashland.
Figure 13-3 shows that ASKY and ASKY-M are approximately 1.25 miles apart. 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 are involved in a variety of industries
including asphalt production, chemical manufacturing, food processing, metals
processing/fabrication, pesticide manufacturing, petroleum refining, and ship/boat
manufacturing. A cluster of emissions sources are located very close to ASKY-M, 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, a heliport at a hospital, and an
asphalt production plant. There are no emissions sources within 1 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 MSA. 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, 2014; ACE, 2014). The closest road to the monitoring
site is a service road feeding into Camp Grayson, as shown in Figure 13-4. This site serves as the
Grayson Lake NATTS site. Figure 13-5 shows that few point sources surround GLKY and that
most of them are on the outer periphery of the 10-mile radius 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 13-5. Sources within 10 miles of GLKY are involved in
13-18
-------
asphalt production, brick/structural clay/clay ceramics manufacturing, food processing, and
mining, among others.
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 railroad that runs through town. Open
fields surround the town, as shown in Figure 13-6, and there are no emissions sources within a
few miles of BAKY, as shown in Figure 13-7. 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 railroad that transverses the
area acting as a geographical boundary. The city is home to some 17 industrial plants, including
metal, steel, and chemical plants (Calvert City, 2014).
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 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 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 railroad. 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 13-8 through 13-12.
13-19
-------
Figure 13-13 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
located between ATKY 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, 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 hospital is located immediately to the south. Although the area is classified as
residential and suburban, most of the residences are located to the west of Newtown Pike (922).
A major electrical equipment and ink manufacturer is located to the northeast of the site, as
shown in Figure 13-14. LEKY is located just over a half-mile south of New Circle Road, a loop
encircling the city of Lexington.
Figure 13-15 shows that most of the emissions sources within 10 miles of LEKY are
within a few miles of the site. Emissions sources in the immediate vicinity of LEKY include a
food processing plant, the aforementioned electrical equipment manufacturing plant, and a
metals processing and fabrication facility.
Table 13-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Kentucky monitoring sites. Table 13-2 includes both county-level
population and vehicle registration information. Table 13-2 also contains traffic volume
information for each site as well as the location for which the traffic volume was obtained.
Additionally, Table 13-2 presents the county-level daily VMT for Boyd, Carter, Henderson,
Marshall, Livingston, and Fayette Counties.
13-20
-------
Table 13-2. Population, Motor Vehicle, and Traffic Information for the Kentucky
Monitoring Sites
Site
ASKY
ASKY-M
GLKY
BAKY
ATKY
CCKY
LAKY
TVKY
BLKY
LEKY
Estimated
County
Population1
49,164
27,348
46,513
31,344
9,423
305,489
County-level
Vehicle
Registration2
39,227
25,391
38,518
30,297
8,281
207,043
Annual
Average
Daily
Traffic3
7,229
12,842
303
922
3,262
4,742
1,189
2,231
2,280
10,083
Intersection Used for Traffic Data
29th Street between Newman St and Lynwood Ave
Greenup (23rd) between 16th St and 17th St
State Hwy 1496, south of Camp Webb Rd
Route 1078, north of Highway 60
Main St, south of Johnson Riley Road
Industrial Pkwy, south of E. 5th Ave
Route 282, east of Industrial Lane
Industrial Pkwy, east of Plant Cut-off Road
Route 93/453
W Loudon Ave, east of Newton Pike
County-
level Daily
VMT4
1,281,000
1,080,000
1,417,000
1,292,000
398,000
7,545,000
Bounty-level population estimates reflects 2012 data (Census Bureau, 2013c)
2County-level vehicle registration reflects 2012 data (KYTC, 2013 a)
3AADT reflects 2010, 2011, or 2012 data (KYTC, 2013b)
4County-level VMT reflects 2012 data (KYTC, 2013c)
BOLD ITALICS = EPA-designaled NATTS Site
Observations from Table 13-2 include the following:
• Fayette County (LEKY) is the most populous of the Kentucky counties with
monitoring sites (by an order of magnitude). Yet this county ranks 29th in population
compared to other counties with NMP sites. The remaining Kentucky counties are
among the least populated compared to other counties with NMP sites. Livingston
County (BLKY) is the least populated of all counties with NMP sites, followed by
Carter County (GLKY) as the second least populated, Marshall County (the Calvert
City sites) third, Henderson County (BAKY) sixth, and Boyd County (ASKY/ASKY-
M) seventh.
• The corresponding vehicle ownership data mimicked these rankings, with Fayette
County in the middle of the range compared to other counties with NMP sites and the
remaining Kentucky counties accounting for the bottom five county-level vehicle
counts.
• Traffic is highest near ASKY-M and LEKY 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.
• The daily VMT for Fayette County is significantly higher than the VMT for the other
Kentucky counties. The VMT for Fayette Count is in the middle of the range
compared to other counties with NMP sites (where VMT data were available), while
the other five Kentucky counties account for five of the six lowest county-level VMT.
13-21
-------
13.2 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.
13.2.1 Climate Summary
The monitoring sites in Kentucky are spread across four different regions across the state.
Elevation generally increases from west to east, with the famed Bluegrass Region in the north-
central portion of the state. The state of Kentucky experiences a continental climate, where
conditions tend to be slightly cooler and drier in the northeast portion of the state and warmer
and wetter in the southwest portion. Kentucky's mid-latitude location ensures an active weather
pattern, in a convergence zone between cooler air from the north and warm, moist air from the
south. The state enjoys all four seasons. Summers are persistently warm and humid; winters are
cloudy but not harsh; and spring and fall are considered pleasant. Precipitation is well distributed
throughout the year, although fall tends to be driest and spring wettest (NCDC, 2014).
13.2.2 Meteorological Summary
Hourly meteorological data for 2012 were retrieved from NCDC for the weather stations
closest to the Kentucky monitoring sites (NCDC, 2012), as described in Section 3.5.2. The
closest weather station to each site is as follows: For ASKY, ASKY-M, and GLKY, Tri-
State/MJ. Ferguson Field Airport (WBAN 03860); for BAKY, Evansville Regional Airport
(WBAN 93817); for BLKY, ATKY, CCKY, LAKY, and TVKY, Barkley Regional Airport
(WBAN 03816); and for LEKY, Blue Grass Airport (WBAN 93820). Additional information
about these weather stations, such as the distance between the sites and the weather stations, is
provided in Table 13-3. These data were used to determine how meteorological conditions on
sample days vary from conditions experienced throughout the year.
13-22
-------
Table 13-3. Average Meteorological Conditions near the Kentucky Monitoring Sites
Closest
Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Health Department, Ashland, Kentucky - ASKY
Tri-St/MJ.
Ferguson Field
Airport
03860
(38.38, -82.56)
7.8
miles
138°
(SE)
Sample
Days
(29)
2012
68.0
±6.6
68.0
±1.8
57.9
±5.9
57.5
±1.6
48.1
±6.1
46.2
±1.7
52.7
±5.6
51.6
±1.5
73.1
±3.8
69.7
±1.3
1016.7
±2.8
1017.0
±0.6
3.6
±0.8
4.0
±0.2
21st and Greenup, Ashland, Kentucky - ASKY-M
Tri-St/M.J.
Ferguson Field
Airport
03860
(38.38, -82.56)
8.6
miles
145°
(SE)
Sample
Days
(52)
2012
72.3
±4.6
68.0
±1.8
61.3
±4.1
57.5
± 1.6
49.4
±4.2
46.2
±1.7
54.8
±3.8
51.6
±1.5
69.0
±3.3
69.7
± 1.3
1016.3
±1.8
1017.0
±0.6
3.7
±0.6
4.0
±0.2
Grayson, Kentucky - GLKY
Tri-St/M.J.
Ferguson Field
Airport
03860
(38.38, -82.56)
23.8
miles
60°
(ENE)
Sample
Days
(69)
2012
66.8
±4.2
68.0
±1.8
56.0
±3.8
57.5
±1.6
44.2
±3.8
46.2
±1.7
50.0
±3.4
51.6
±1.5
68.6
±3.0
69.7
±1.3
1017.1
±1.6
1017.0
±0.6
4.0
±0.6
4.0
±0.2
Baskett, Kentucky - BAKY
Evansville
Regional
Airport
93817
(38.04, -87.52)
12.3
miles
341°
(NNW)
Sample
Days
(51)
2012
73.7
±4.7
70.4
±1.9
63.0
±4.3
60.0
±1.8
48.8
±4.0
46.9
±1.6
55.2
±3.7
52.9
±1.5
63.4
±2.9
65.5
±1.1
1016.8
±1.5
1017.0
±0.6
5.2
±0.7
5.3
±0.3
OJ
to
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
-------
Table 13-3. Average Meteorological Conditions near the Kentucky Monitoring Sites (Continued)
Closest
Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Atmos Energy, Calvert City, Kentucky - ATKY
Barkley
Regional
Airport
03816
(37.06, -88.77)
21.9
miles
268°
(W)
Sample
Days
(29)
2012
71.6
±6.9
71.8
±1.8
61.3
±6.3
61.2
±1.7
48.8
±5.8
48.2
±1.6
54.5
±5.5
54.1
±1.5
67.1
±3.8
66.0
±1.2
1016.6
±2.3
1016.5
±0.6
5.6
±1.1
5.6
±0.3
Smithland, Kentucky - BLKY
Barkley
Regional
Airport
03816
(37.06, -88.77)
23.0
miles
263°
(W)
Sample
Days
(29)
2012
71.6
±6.9
71.8
±1.8
61.3
±6.3
61.2
± 1.7
48.8
±5.8
48.2
±1.6
54.5
±5.5
54.1
±1.5
67.1
±3.8
66.0
± 1.2
1016.6
±2.3
1016.5
±0.6
5.6
± 1.1
5.6
±0.3
Calvert City Elementary, Calvert City, Kentucky - CCKY
Barkley
Regional
Airport
03816
(37.06, -88.77)
22.5
miles
270°
(W)
Sample
Days
(50)
2012
74.7
±4.7
71.8
±1.8
63.9
±4.3
61.2
±1.7
49.9
±4.0
48.2
±1.6
56.1
±3.7
54.1
±1.5
64.5
±3.2
66.0
±1.2
1016.2
±1.5
1016.5
±0.6
5.6
±0.7
5.6
±0.3
Lazy Daze, Calvert City, Kentucky - LAKY
Barkley
Regional
Airport
03816
(37.06, -88.77)
23.0
miles
269°
(W)
Sample
Days
(29)
2012
71.7
±6.9
71.8
±1.8
61.8
±6.1
61.2
±1.7
49.5
±5.6
48.2
±1.6
55.0
±5.3
54.1
±1.5
67.6
±4.0
66.0
±1.2
1016.0
±2.0
1016.5
±0.6
5.7
±1.0
5.6
±0.3
OJ
to
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
-------
Table 13-3. Average Meteorological Conditions near the Kentucky Monitoring Sites (Continued)
Closest
Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
TVA Substation, Calvert City, Kentucky - TVKY
Barkley
Regional
Airport
03816
(37.06, -88.77)
23.1
miles
267°
(W)
Sample
Days
(29)
2012
71.6
±6.9
71.8
±1.8
61.3
±6.3
61.2
±1.7
48.8
±5.8
48.2
±1.6
54.5
±5.5
54.1
±1.5
67.1
±3.8
66.0
±1.2
1016.6
±2.3
1016.5
±0.6
5.6
±1.1
5.6
±0.3
Lexington, Kentucky - LEKY
Blue Grass
Airport
93820
(38.04, -84.61)
5 8
miles
246°
(WSW)
Sample
Days
(53)
2012
70.9
±4.7
67.4
±1.8
60.9
±4.3
57.6
± 1.7
49.4
±4.0
47.1
±1.6
54.5
±3.8
52.0
±1.5
69.2
±3.3
71.0
± 1.3
1016.9
±1.6
1017.2
±0.6
6.2
±0.7
6.4
±0.3
OJ
to
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
-------
Table 13-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for all of 2012. Also included in Table 13-3 is the 95 percent
confidence interval for each parameter. GLKY is the only Kentucky site for which sampling was
conducted year-round. Table 13-3 shows that average meteorological conditions on sample days
near GLKY were generally representative of average weather conditions experienced throughout
the year. Although sample days appear slightly cooler and drier, this is probably due to the
majority of the make-up days occurring in the cooler months of the year (two in the first quarter,
one in the second quarter, and five in the fourth quarter).
The difference between the full-year and sample day temperature and moisture
parameters is wider for those Kentucky sites that began sampling under the NMP in March
(ASKY-M, BAKY, CCKY, and LEKY) than those sites that began sampling in July (ASKY,
ATKY, BLKY, LAKY, and TVKY). For those sites that began sampling in July, the largest
differences are seen in the moisture parameters. This is because the cooler months of the year,
which are not included in the sample day calculations, also tend to be the driest. For those sites
that began sampling in March, the largest differences are seen in the temperature parameters,
although the moisture parameters differ too. For these sites, the sample days appear warmer than
the full-year averages. This is because January and February, traditionally the coldest months of
the year, are not included in the sample day averages.
13.2.3 Back Trajectory Analysis
A composite back trajectory map representing days on which samples were collected is
presented for each site. Included on each composite map are four back trajectories per sample
day. Where sampling occurred for a long enough duration (30 sample days), per the criteria
described in Section 3.5.2.1, a corresponding cluster analysis is presented. Thus, Figures 13-16
through 13-30 are the composite back trajectory maps and corresponding cluster analyses for the
Kentucky monitoring sites. An in-depth description of these maps and how they were generated
is presented in Section 3.5.2.1. For the composite map, each line represents the 24-hour
trajectory along which a parcel of air traveled toward the monitoring site on a given sample day
and time, based on an initial height of 50 meters AGL. For the cluster analysis, each line
13-26
-------
corresponds to a trajectory representative of a given cluster of back trajectories. Each concentric
circle around the sites in Figures 13-16 and 13-30 represents 100 miles.
Figure 13-16. Composite Back Trajectory Map for ASKY
Figure 13-17. Composite Back Trajectory Map for ASKY-M
13-27
-------
Figure 13-18. Back Trajectory Cluster Map for ASKY-M
Figure 13-19. Composite Back Trajectory Map for GLKY
13-28
-------
Figure 13-20. Back Trajectory Cluster Map for GLKY
Figure 13-21. Composite Back Trajectory Map for BAKY
13-29
-------
Figure 13-22. Back Trajectory Cluster Map for BAKY
Figure 13-23. Composite Back Trajectory Map for ATKY
13-30
-------
Figure 13-24. Composite Back Trajectory Map for BLKY
Figure 13-25. Composite Back Trajectory Map for CCKY
13-31
-------
Figure 13-26. Back Trajectory Cluster Map for CCKY
Figure 13-27. Composite Back Trajectory Map for LAKY
13-32
-------
Figure 13-28. Composite Back Trajectory Map for TVKY
Figure 13-29. Composite Back Trajectory Map for LEKY
13-33
-------
Figure 13-30. Back Trajectory Cluster Map for LEKY
Observations from Figures 13-16 through 13-18 for ASKY and ASKY-M include the
following:
• The composite map for ASKY has fewer back trajectories than the composite map for
ASKY-M. This is because sampling at ASKY began in July while sampling at
ASKY-M began in March. Because there are fewer than 30 sample days for ASKY
(29), a cluster analysis is not presented for this site.
• The composite maps show that back trajectories originated from a variety of
directions at the Ashland sites, although fewer back trajectories originated from the
east of the sites. Back trajectories originating from the west, northwest, and north
tended to be the longest.
• The farthest away a back trajectory originated from the Ashland sites was over Lake
Superior, or greater than 650 miles away. A back trajectory of similar distance also
originated over central Iowa. The four long back trajectories originating over Lake
Superior are the four back trajectories representing October 30, 2012. The average
back trajectory length for ASKY (193 miles) is slightly less than the average
trajectory length for ASKY-M (202 miles). Recall, though, that the composite map
for ASKY includes four less months of sampling than the composite map for
ASKY-M.
• The four long back trajectories originating over Lake Superior are represented by
their own cluster trajectory in Figure 13-8, representing only 2 percent of back
trajectories. Nineteen percent of back trajectories originated to the west and northwest
13-34
-------
of ASKY-M, primarily over Indiana and Illinois. Nearly one third of back trajectories
originated to the southwest of the site, primarily over Tennessee and northern
Georgia. Eleven percent of back trajectories originated over the Appalachian
Mountains of Virginia, North Carolina, and Tennessee and generally less than 200
miles away. Thirteen percent of back trajectories originated to the northeast of
ASKY, over southeast Ohio, western Pennsylvania, and West Virginia. Only 7
percent of back trajectories originated to the north of ASKY-M, primarily over
Michigan, Lake Huron, Lake Erie, and southwest Ontario, Canada. The short cluster
trajectory originating over the Bluegrass Region of Kentucky represents those back
trajectories with a westerly component, originating over northeast Kentucky and
south-central Ohio, and generally less than 100 miles in length.
Observations from Figures 13-19 and 13-20 for GLKY include the following:
• The composite map for GLKY is similar in the geographic distribution of back
trajectories to the composite map for ASKY-M. This is not unexpected as these sites
are only 25 miles apart.
• The composite map shows that back trajectories originated from a variety of
directions at GLKY. Back trajectories with an easterly component appear to be
shorter than those originating from other directions.
• The farthest away a back trajectory originated from GLKY was over west-central
Iowa, or greater than 650 miles away. A back trajectory of similar distance also
originated over Lake Superior. The four long back trajectories originating over Lake
Superior and the Upper Peninsula of Michigan are the four back trajectories
representing October 30, 2012, similar to those shown for the Ashland sites. The
average back trajectory length for GLKY (205 miles) is just greater than the average
back trajectory length for ASKY-M (202 miles). More than 90 percent of back
trajectories were less than 400 miles in length.
• For GLKY, the four long back trajectories originating over Lake Superior are
represented by the same cluster trajectory representing other back trajectories
originating to the north of the site. Together, these represent 6 percent of back
trajectories. Nearly 40 percent of back trajectories originated to the west and
northwest of GLKY, but are split into two cluster trajectories. One (30 percent)
represents shorter back trajectories originating primarily over Indiana while the other
(9 percent) represents longer back trajectories originating primarily over Illinois.
One-quarter of back trajectories originated to the southwest of the site, primarily over
Tennessee and northern Georgia. The short cluster trajectory originating over the
Appalachian Plateau of Virginia represents the 18 percent of back trajectories
originating over the Appalachian Mountains of Virginia, North Carolina, and
Tennessee and southern West Virginia and generally within 200 miles of GLKY.
Back trajectories originating to the northeast of GLKY, over southeast Ohio, western
Pennsylvania, and West Virginia, account for 11 percent of back trajectories.
13-35
-------
Observations from Figures 13-21 and 13-22 for BAKY include the following:
• The composite map for BAKY shows that although most back trajectories originated
within 400 miles of the site, one back trajectory originated from greater than
850 miles away and another from nearly 800 miles away. These back trajectories
represent the November 23, 2012 sample day. The average back trajectory length for
BAKY is 223 miles, which is the highest average among the Kentucky sites.
• An imaginary line drawn east-west through the site on the composite map shows that
most back trajectories have either a northerly or a southerly component and that few
back trajectories originated from the east or west.
• Eighteen percent of back trajectories originated from the northwest of BAKY, with
the long cluster trajectory representing the two back trajectories originating over the
Dakotas and the shorter one representing back trajectories originating over Iowa,
Illinois, and Missouri. Nearly 30 percent of back trajectories originated to the
southwest of BAKY, primarily over western Tennessee or along the Mississippi
River. Another 22 percent of back trajectories originated to the southeast to south of
BAKY, over central Tennessee and northern Mississippi and Georgia. Nearly
10 percent of back trajectories originated to the north of BAKY but are split into two
cluster trajectories based on which side of Lake Michigan they originate. The short
cluster trajectory originating to the north of BAKY represents the 25 percent of back
trajectories originating less than 200 miles away from the site and with a northerly
component. These back trajectories originated over north-central Kentucky, the
southern half of Indiana, and southeast Illinois.
Observations from Figures 13-23 through 13-28 for the five sites in or near Calvert City
include the following:
• With the exception of CCKY, the composite maps for the Calvert City sites include
only half a year's worth of sample days due to the July start date. Thus, CCKY is the
only Calvert City site for which a cluster analysis could be performed.
• The composite maps for the Calvert City sites resemble each other, which is expected
given the relatively close proximity of these site to each other. These composite maps
also resemble the composite map for BAKY, which is located 75 miles northeast of
Calvert City. The composite maps show that most back trajectories originated within
400 miles of the sites and primarily to the northwest to northeast or southeast to
southwest of the sites.
• The average back trajectory length ranged from 201 miles to 207 miles for the four
sites that started sampling under the NMP in July; the average back trajectory length
for CCKY is 219 miles.
• Each composite map includes the two long back trajectories originating over South
Dakota, or greater than 800 miles away. These back trajectories represent the
November 23, 2012 sample day.
13-36
-------
• Nineteen percent of back trajectories originated to the northwest of CCKY, with the
long cluster trajectory representing the two back trajectories originating over South
Dakota and the shorter one representing back trajectories originating over Iowa,
Illinois, and Missouri. The short cluster trajectory originating to the northwest of
CCKY represents the 23 percent of back trajectories originating less than 200 miles
away from the site and over southern Illinois and southeast Missouri. Fifteen percent
of back trajectories originated to the southwest of CCKY, primarily along the
Mississippi River. Although 35 percent of back trajectories originated to the
southeast of CCKY, these are split into two cluster trajectories. One cluster trajectory
represents the longer back trajectories originating primarily over Alabama, and one
cluster trajectory represents the short back trajectories originating over central
Tennessee and northern Alabama. Ten percent of back trajectories originated to the
northeast of CCKY, but includes back trajectories of varying lengths.
Observations from Figures 13-29 and 13-30 for LEKY include the following:
• The composite map shows that back trajectories originated from a variety of
directions at LEKY, although few back trajectories originated from the east of the
site. The longest back trajectories tended to originate from the north, although the
longest back trajectory originated over South Dakota, or greater than 700 miles away.
This back trajectory also represents the November 23, 2012 sample day. The average
back trajectory length for LEKY is 202 miles.
• The cluster analysis for LEKY shows that 12 percent of back trajectories originated to
the north of the site, but these are split into two clusters, one representing the longer
back trajectories originating over the Upper Peninsula of Michigan, and one
representing the back trajectories east of Lake Michigan. Another 12 percent of back
trajectories originated to the northwest of LEKY, primarily over Illinois but as far
away as South Dakota. Twenty-five percent of back trajectories are represented by
the short cluster trajectory originating towards Louisville, Kentucky. These back
trajectories originated within 200 miles of LEKY and generally over southern Indiana
and central Kentucky. Nearly one-quarter of back trajectories originated to the
southwest of LEKY, over Tennessee and northern Georgia, Alabama, and
Mississippi. Sixteen percent of back trajectories originated over southeastern
Kentucky and northeastern Tennessee. The final 13 percent of back trajectories
originated to the north and northeast of LEKY, primarily over the southern half of
Ohio and West Virginia.
13-37
-------
13.2.4 Wind Rose Comparison
Hourly surface wind data from the weather stations nearest the Kentucky sites, as
presented in Section 13.2.2, were uploaded into a wind rose software program to produce
customized wind roses, as described in Section 3.5.2.2. 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 13-31 presents a map showing the distance between the Tri-State/MJ. Ferguson
Field Airport weather station and ASKY, which may be useful for identifying topographical
influences that may affect the meteorological patterns experienced at this location. Figure 13-31
also presents three different wind roses for the ASKY monitoring site. First, a historical wind
rose representing 2002 to 2011 wind data is presented, which shows the predominant surface
wind speed and direction over an extended period of time. Second, a wind rose representing
wind observations for all of 2012 is presented. Next, a wind rose representing wind data for days
on which samples were collected in 2012 is presented. These can be used to identify the
predominant wind speed and direction for 2012 and to determine if wind observations on sample
days were representative of conditions experienced over the entire year and historically. Figures
13-32 through 13-40 present the distance maps and wind roses for the remaining Kentucky
monitoring sites.
13-38
-------
Figure 13-31. Wind Roses for the Tri-State/M.J. Ferguson Field Airport Weather Station
nearASKY
Location of ASKY and Weather Station
2002-2012 Historical Wind Rose
2012 Wind Rose
Sample Day Wind Rose
13-39
-------
Figure 13-32. Wind Roses for the Tri-State/M.J. Ferguson Field Airport Weather Station
near ASKY-M
Location of ASKY-M and Weather Station
2002-2011 Historical Wind Rose
2012 Wind Rose
Sample Day Wind Rose
13-40
-------
Figure 13-33. Wind Roses for the Tri-State/M.J. Ferguson Field Airport Weather Station
nearGLKY
Location of GLKY and Weather Station
2002-2011 Historical Wind Rose
-_.
2012 Wind Rose
Sample Day Wind Rose
Calms: 3064%
13-41
-------
Observations from Figures 13-31 through 13-33 for ASKY, ASKY-M, and GLKY
include the following:
• The Tri-State/MJ. Ferguson Field weather station is the closest weather station to
both Ashland sites and GLKY. The weather station is located approximately 8 miles
southeast of the Ashland sites and nearly 24 miles to the east-northeast of GLKY.
This weather station is in West Virginia, south of the Ohio River and east of the Big
Sandy River.
• Because these three sites share the same weather station, the historical and full-year
wind roses are identical across the sites.
• The historical wind rose shows that winds from the south, southwest quadrant, and
west account for nearly half of the wind observations near these sites, particularly
those from south-southwest. Calm winds (< 2 knots) account for nearly 25 percent of
the hourly measurements.
• The wind patterns on the full-year wind rose are similar to those on the historical
wind rose, although calm winds accounted for a slightly higher percentage of the
wind observations in 2012 (30 percent). There were slightly fewer wind observations
from the southwest to west but additional wind observations from the south-
southwest.
• The sample day wind rose for ASKY resembles both the historical and full-year wind
roses, although there is a higher percentage of south-southwesterly winds as well as
calm winds (up nearly 10 percent from the historical wind rose). Recall that the
sample day wind rose includes only six months worth of sample days as ASKY did
not begin sampling under the NMP until July.
• The sample day wind rose for ASKY-M also resembles both the historical and full-
year wind roses, although there is a slightly higher percentage of southwesterly winds
and fewer southerly winds. The calm rate is also higher on the sample day wind rose.
Recall that ASKY-M began sampling under the NMP in March.
• The sample day wind rose for GLKY also resembles both the historical and full-year
wind roses.
• The sample day wind roses for all three sites have fewer winds on the top half of the
wind rose. The historical and full-year wind roses show that most of the directions
with a northerly component account for roughly 3 percent to 4 percent of
observations. These percentages are more variable for the sample day wind roses.
13-42
-------
Figure 13-34. Wind Roses for the Evansville Regional Airport Weather Station near BAKY
Location of BAKY and Weather Station
2002-2011 Historical Wind Rose
Calms: 23.73%
2012 Wind Rose
Sample Day Wind Rose
Calms: 24.65%
Calms: 24.34%
13-43
-------
Observations from Figure 13-34 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 historical wind rose shows that winds from a variety of directions are observed
near BAKY, although winds from the south and southwest quadrant are observed the
most and winds from the southeast quadrant are observed the least. Calm winds
account for just less than one-quarter of the observations.
• The full-year wind rose shows that winds from all directions were observed, with
winds from the south and south-south west accounting for the highest percentage of
winds greater than 2 knots. Calm winds account for approximately one-quarter of the
observations.
• The sample day wind rose for BAKY shares some similarities with the full-year and
historical wind roses, but exhibits some differences as well. Although southerly winds
are prevalent and calm winds still account for one-quarter of the observations, there is
a higher percentage of winds from the south-southeast and northwest, while fewer
south-southwesterly winds were observed. BAKY did not begin sampling until
March; thus, a full year's worth of wind observations may look different.
13-44
-------
Figure 13-35. Wind Roses for the Barkley Regional Airport Weather Station near ATKY
Location of ATKY and Weather Station
2002-2011 Historical Wind Rose
Calms: 25.35*
2012 Wind Rose
Sample Day Wind Rose
Calms: 19.92%
Calms: 21 12%
13-45
-------
Figure 13-36. Wind Roses for the Barkley Regional Airport Weather Station near BLKY
Location of BLKY and Weather Station
2002-2011 Historical Wind Rose
Calms: 25.35*
2012 Wind Rose
Sample Day Wind Rose
Calms: 19.92%
Calms: 21 12%
13-46
-------
Figure 13-37. Wind Roses for the Barkley Regional Airport Weather Station near CCKY
Location of CCKY and Weather Station
2002-2011 Historical Wind Rose
Calms: 25.35*
2012 Wind Rose
Sample Day Wind Rose
Calms: 19.92%
Calms: 20 OS%
13-47
-------
Figure 13-38. Wind Roses for the Barkley Regional Airport Weather Station near LAKY
Location of LAKY and Weather Station
2002-2011 Historical Wind Rose
Calms: 25.35*
2012 Wind Rose
Sample Day Wind Rose
Calms: 19.92%
Calms: 1997%
13-48
-------
Figure 13-39. Wind Roses for the Barkley Regional Airport Weather Station near TVKY
Location of TVKY and Weather Station
2002-2011 Historical Wind Rose
Calms: 25.35*
2012 Wind Rose
Sample Day Wind Rose
Calms: 19.92%
Calms: 21 12%
13-49
-------
Observations from Figures 13-35 through 13-39 for the Calvert City sites include the
following:
• The Barkley Regional Airport weather station is the closest weather station to all five
sites in and near Calvert City. The weather station is located between 20 miles and
25 miles west of the Calvert City monitoring sites and just west of the Paducah metro
area.
• The historical and full-year wind roses are identical across the sites because these five
sites share the same weather station.
• The historical 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 approximately 25 percent of the hourly measurements.
• The full-year wind rose resembles the historical wind rose, but has a higher
percentage of winds from the south to south-west and slightly fewer calm
observations (20 percent).
• The sample day wind roses for ATKY, BLKY, and TVKY are identical because these
three sites all began sampling in July and sampled on all the same days. Sampling at
LAKY also began in July but differs slightly from the other three sites because of
make-up days. The sample day wind roses for these sites show that southerly winds
were prevalent on sample days during the second half of 2012. In addition, the
percentage of winds from the south-southwest and southwest is less while the
percentage of winds from the south-southeast and northwest is higher. Calm winds
account for approximately 20 percent to 21 percent of the wind observations.
• Sampling at CCKY under the NMP began in March; thus, the sample day wind rose
reflects wind observations for an additional four months of sample days compared to
the other Calvert City sites. Yet, the differences between the sample day wind rose for
CCKY and the other Calvert City sites are not significant as southerly winds were
prevalent near CCKY and calm winds accounted for 20 percent of the observations.
13-50
-------
Figure 13-40. Wind Roses for the Blue Grass Airport Weather Station near LEKY
Location of LEKY and Weather Station
2002-2011 Historical Wind Rose
Calms: 12.64%
2012 Wind Rose
Sample Day Wind Rose
Calms: 1279%
Calms: 12_E-Q%
13-51
-------
Observations from Figure 13-40 for LEKY include the following:
• The Blue Grass Airport weather station is located approximately 6 miles west-
southwest of the LEKY monitoring site. As shown, the airport is located on the
western edge of the Lexington metro area.
• The historical 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 roughly 13 percent of observations. Winds from
other directions account for 5 percent of wind observations or less each. Calm winds
account for nearly 13 percent of the hourly measurements.
• The full-year wind rose resembles the historical wind rose, although the decrease in
the wind observations from the southwest to west-southwest is offset by the
additional wind observations from the south-southeast and south.
• Sampling at LEKY under the NMP began in March; thus, the sample day wind rose
reflects wind observations for 10 months of the year. The wind patterns on the sample
day wind rose for LEKY resemble the wind patterns on both the historical and full-
year wind roses. While southerly winds were still prevalent on sample days, an even
higher percentage of winds from the south-southeast were observed.
13.3 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 13-4 provides an overview of which analyses were performed at
each site and when each site began sampling under the NMP, as there are 10 monitoring sites in
Kentucky and their respective start dates are variable. The site-specific results of the risk-based
screening process are presented in Table 13-5, with the pollutants of interest shaded in gray.
13-52
-------
Table 13-4. Overview of Sampling Performed at the Kentucky Monitoring Sites
Site
ASKY
ASKY-M
GLKY
BAKY
ATKY
BLKY
CCKY
LAKY
TVKY
LEKY
VOCs
7/14/12
—
1/4/12
—
7/14/12
7/14/12
7/14/12
7/14/12
7/14/12
7/17/12
Carbonyl
Compounds
7/14/12
—
1/4/12
—
—
~
—
—
—
7/14/12
PAHs
~
—
1/4/12
—
—
~
—
—
—
-
PM10
Metals
~
3/4/12
1/4/12
3/4/12
—
~
3/4/12
—
—
3/4/12
Hexavalent
Chromium
~
—
1/4/12
—
—
~
—
—
—
-
~ = This pollutant group was not sampled for at this site.
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 13-4 include the following:
• GLKY sampled carbonyl compounds, VOCs, PAHs, PMio metals, and hexavalent
chromium throughout 2012.
• Those additional sites sampling PMio metals (ASKY-M, BAKY, CCKY, and
LEKY) began sampling under the NMP in March 2012.
• Those additional sites sampling VOCs (ASKY, ATKY, BLKY, CCKY, LAKY,
TVKY, and LEKY) began sampling under the NMP in July 2012.
• Those additional sites sampling carbonyl compounds (ASKY and LEKY) also
began sampling under the NMP in July 2012.
Table 13-5. Risk-Based Screening Results for the Kentucky Monitoring Sites
Pollutant
Screening
Value
(Ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Health Department, Ashland, Kentucky - ASKY
Acetaldehyde
Benzene
Carbon Tetrachloride
Formaldehyde
1,3 -Butadiene
1 ,2-Dichloroethane
Ethylbenzene
£>-Dichlorobenzene
Hexachloro- 1 ,3 -butadiene
1 ,2-Dibromoethane
1 , 1 ,2,2-Tetrachloroethane
0.45
0.13
0.17
0.077
0.03
0.038
0.4
0.091
0.045
0.0017
0.017
Total
29
29
29
29
26
24
9
2
2
1
1
181
29
29
29
29
28
24
29
21
o
J
1
1
223
100.00
100.00
100.00
100.00
92.86
100.00
31.03
9.52
66.67
100.00
100.00
81.17
16.02
16.02
16.02
16.02
14.36
13.26
4.97
1.10
1.10
0.55
0.55
16.02
32.04
48.07
64.09
78.45
91.71
96.69
97.79
98.90
99.45
100.00
13-53
-------
Table 13-5. Risk-Based Screening Results for the Kentucky Monitoring Sites (Continued)
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
21st and Greenup, Ashland, Kentucky - ASKY-M
Arsenic (PM10)
Manganese (PM10)
Nickel (PM10)
Cadmium (PM10)
Lead (PM10)
Antimony (PM10)
0.00023
0.005
0.0021
0.00056
0.015
0.02
Total
47
46
24
19
18
1
155
50
50
50
50
50
50
300
94.00
92.00
48.00
38.00
36.00
2.00
51.67
30.32
29.68
15.48
12.26
11.61
0.65
30.32
60.00
75.48
87.74
99.35
100.00
Grayson, Kentucky - GLKY
Benzene
Carbon Tetrachloride
Formaldehyde
1 ,2-Dichloroethane
Acetaldehyde
Arsenic (PM10)
1,3 -Butadiene
Manganese (PM10)
Naphthalene
1 , 1 ,2,2-Tetrachloroethane
Hexachloro- 1 ,3 -butadiene
1 ,2-Dibromoethane
Cadmium (PM10)
Ethylbenzene
0.13
0.17
0.077
0.038
0.45
0.00023
0.03
0.005
0.029
0.017
0.045
0.0017
0.00056
0.4
Total
61
61
61
56
52
51
51
13
6
5
o
J
2
1
1
424
61
61
61
56
61
59
59
59
61
5
4
2
55
61
665
100.00
100.00
100.00
100.00
85.25
86.44
86.44
22.03
9.84
100.00
75.00
100.00
1.82
1.64
63.76
14.39
14.39
14.39
13.21
12.26
12.03
12.03
3.07
1.42
1.18
0.71
0.47
0.24
0.24
14.39
28.77
43.16
56.37
68.63
80.66
92.69
95.75
97.17
98.35
99.06
99.53
99.76
100.00
Baskett, Kentucky - BAKY
Arsenic (PM10)
Manganese (PM10)
Cadmium (PM10)
0.00023
0.005
0.00056
Total
46
36
1
83
50
50
50
150
92.00
72.00
2.00
55.33
55.42
43.37
1.20
55.42
98.80
100.00
Atmos Energy, Calvert City, Kentucky - ATKY
Benzene
Carbon Tetrachloride
1 ,2-Dichloroethane
1,3 -Butadiene
Vinyl chloride
1 , 1 ,2-Trichloroethane
Hexachloro- 1 ,3 -butadiene
0.13
0.17
0.038
0.03
0.11
0.0625
0.045
Total
29
29
26
22
11
o
J
1
121
29
29
26
25
16
3
5
133
100.00
100.00
100.00
88.00
68.75
100.00
20.00
90.98
23.97
23.97
21.49
18.18
9.09
2.48
0.83
23.97
47.93
69.42
87.60
96.69
99.17
100.00
13-54
-------
Table 13-5. Risk-Based Screening Results for the Kentucky Monitoring Sites (Continued)
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Smithland, Kentucky - BLKY
Benzene
Carbon Tetrachloride
1 ,2-Dichloroethane
1,3 -Butadiene
Vinyl chloride
1,1,2-Trichloroethane
Hexachloro- 1 ,3 -butadiene
0.13
0.17
0.038
0.03
0.11
0.0625
0.045
Total
26
26
23
17
7
2
1
102
26
26
23
21
14
2
1
113
100.00
100.00
100.00
80.95
50.00
100.00
100.00
90.27
25.49
25.49
22.55
16.67
6.86
1.96
0.98
25.49
50.98
73.53
90.20
97.06
99.02
100.00
Calvert City Elementary, Calvert City, Kentucky - CCKY
Arsenic (PM10)
Manganese (PM10)
Carbon Tetrachloride
Benzene
1 ,2-Dichloroethane
1,3 -Butadiene
Vinyl chloride
1,1,2-Trichloroethane
Hexachloro- 1 ,3 -butadiene
0.00023
0.005
0.17
0.13
0.038
0.03
0.11
0.0625
0.045
Total
45
28
26
25
23
19
4
2
1
173
47
47
26
26
23
24
11
2
4
210
95.74
59.57
100.00
96.15
100.00
79.17
36.36
100.00
25.00
82.38
26.01
16.18
15.03
14.45
13.29
10.98
2.31
1.16
0.58
26.01
42.20
57.23
71.68
84.97
95.95
98.27
99.42
100.00
Lazy Daze, Calvert City, Kentucky - LAKY
Benzene
Carbon Tetrachloride
1 ,2-Dichloroethane
1,3 -Butadiene
Vinyl chloride
1, 1,2,2-Tetrachloroethane
1 ,2-Dibromoethane
Hexachloro- 1 ,3 -butadiene
1 , 1 ,2-Trichloroethane
/>-Dichlorobenzene
0.13
0.17
0.038
0.03
0.11
0.017
0.0017
0.045
0.0625
0.091
Total
29
29
27
24
12
6
o
J
o
J
o
J
1
137
29
29
27
28
19
6
3
7
5
12
165
100.00
100.00
100.00
85.71
63.16
100.00
100.00
42.86
60.00
8.33
83.03
21.17
21.17
19.71
17.52
8.76
4.38
2.19
2.19
2.19
0.73
21.17
42.34
62.04
79.56
88.32
92.70
94.89
97.08
99.27
100.00
13-55
-------
Table 13-5. Risk-Based Screening Results for the Kentucky Monitoring Sites (Continued)
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
TVA Substation, Calvert City, Kentucky - TVKY
Carbon Tetrachloride
Benzene
1 ,2-Dichloroethane
1,3 -Butadiene
Vinyl chloride
1, 1,2,2-Tetrachloroethane
1,1,2-Trichloroethane
Hexachloro- 1 ,3 -butadiene
1 ,2-Dibromoethane
£>-Dichlorobenzene
0.17
0.13
0.038
0.03
0.11
0.017
0.0625
0.045
0.0017
0.091
Total
28
27
27
22
14
4
4
2
1
1
130
28
28
27
24
19
4
6
5
1
9
151
100.00
96.43
100.00
91.67
73.68
100.00
66.67
40.00
100.00
11.11
86.09
21.54
20.77
20.77
16.92
10.77
3.08
3.08
1.54
0.77
0.77
21.54
42.31
63.08
80.00
90.77
93.85
96.92
98.46
99.23
100.00
Lexington, Kentucky - LEKY
Arsenic (PM10)
Manganese (PM10)
Benzene
Carbon Tetrachloride
Acetaldehyde
1,3 -Butadiene
Formaldehyde
1 ,2-Dichloroethane
£>-Dichlorobenzene
Ethylbenzene
1 ,2-Dibromoethane
1 , 1 ,2,2-Tetrachloroethane
Beryllium (PM10)
Hexachloro- 1 ,3 -butadiene
Nickel (PM10)
0.00023
0.005
0.13
0.17
0.45
0.03
0.077
0.038
0.091
0.4
0.0017
0.017
0.00042
0.045
0.0021
Total
45
32
29
29
27
27
27
24
o
3
o
3
2
2
1
1
1
253
49
49
29
29
27
27
27
24
14
29
2
2
48
3
49
408
91.84
65.31
100.00
100.00
100.00
100.00
100.00
100.00
21.43
10.34
100.00
100.00
2.08
33.33
2.04
62.01
17.79
12.65
11.46
11.46
10.67
10.67
10.67
9.49
1.19
1.19
0.79
0.79
0.40
0.40
0.40
17.79
30.43
41.90
53.36
64.03
74.70
85.38
94.86
96.05
97.23
98.02
98.81
99.21
99.60
100.00
Observations for the Ashland sites from Table 13-5 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 13-4. VOCs and carbonyl compounds were sampled for at ASKY while only
PMio metals were sampled for at ASKY-M.
• Eleven pollutants failed at least one screen for ASKY, with 81 percent of
concentrations for these 11 pollutants greater than their associated risk screening
value (or failed screens).
• Seven pollutants contributed to 95 percent of failed screens for ASKY and therefore
were identified as pollutants of interest. These seven include two carbonyl
compounds and five VOCs.
13-56
-------
• Six metals failed at least one screen for ASKY-M, with 52 percent of concentrations
for these six pollutants greater than their associated risk screening value (or failed
screens).
• Five metals contributed to 95 percent of failed screens for ASKY-M and therefore
were identified as pollutants of interest (arsenic, cadmium, lead, manganese, and
nickel).
Observations for GLKY from Table 13-5 include the following:
• GLKY sampled for all five pollutant groups shown in Table 13-4.
• Fourteen pollutants failed at least one screen for GLKY, with nearly 64 percent of
concentrations for these 14 pollutants greater than their associated risk screening
value (or failed screens).
• Eight 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 two metals.
Observations for BAKY from Table 13-5 include the following:
• Like ASKY-M, BAKY sampled for PMi0 metals only.
• Three pollutants failed at least one screen for BAKY, with 55 percent of
concentrations for these three pollutants greater than their associated risk screening
value (or failed screens). This site had the fewest pollutants fail screens among the
Kentucky monitoring sites.
• Arsenic and manganese contributed to 95 percent of failed screens for BAKY and
therefore were identified as pollutants of interest for this site.
Observations for the Calvert City sites from Table 13-5 include the following:
• VOCs were sampled for at all five Calvert City sites. PMio metals were also sampled
for at CCKY.
• The number of pollutants whose concentrations were greater than their associated risk
screening value varied from seven (ATKY and BLKY) to 10 (LAKY and TVKY).
• Five pollutants contributed to 95 percent of failed screens for ATKY and therefore
were identified as pollutants of interest for this site.
• Five pollutants contributed to 95 percent of failed screens for BLKY and therefore
were identified as pollutants of interest for this site.
• Six pollutants contributed to 95 percent of failed screens for CCKY and therefore
were identified as pollutants of interest for this site. The pollutants of interest for
13-57
-------
CCKY include two metals and four VOCs. Although arsenic and manganese failed
the greatest number of screens for CCKY, PMi0 metals sampling under the NMP
began 3 months before sampling for VOCs began.
• Nine pollutants contributed to 95 percent of failed screens for LAKY and therefore
were identified as pollutants of interest for this site. Although the pollutants through
hexachloro-1,3-butadiene together account for more than 95 percent of the total failed
screens for LAKY, 1,1,2-trichloroethane failed the same number of screens as
hexachloro-1,3-butadiene; thus, 1,1,2-trichloroethane was added as pollutants of
interest for LAKY, per the procedure described in Section 3.2.
• Seven pollutants contributed to 95 percent of failed screens for TVKY and therefore
were identified as pollutants of interest for this site.
• Benzene, carbon tetrachloride, 1,2-dichloroethane, and 1,3-butadiene were identified
as pollutants of interest for all five Calvert City sites.
Observations for LEKY from Table 13-5 include the following:
• Aside from GLKY, LEKY sampled for the most pollutant groups. Carbonyl
compounds, VOCs, and PMio metals were sampled for at LEKY.
• Fifteen pollutants failed at least one screen for LEKY, with 62 percent of
concentrations for these 15 pollutants greater than their associated risk screening
value (or failed screens).
• Ten pollutants contributed to 95 percent of failed screens for LEKY and therefore
were identified as pollutants of interest. These include two carbonyl compounds, six
VOCs, and two metals.
• Although arsenic and manganese failed the greatest number of screens for LEKY, it
should be noted that PMio metals sampling under the NMP began 3 months before
sampling for VOCs and carbonyl compounds began.
13.4 Concentrations
This section presents various concentration averages used to characterize pollution 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.
13-58
-------
• 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. Additional site-specific statistical summaries for the
Kentucky monitoring sites are provided in Appendices J, L, M, N, and O.
13.4.1 2012 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 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 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
Kentucky monitoring sites are presented in Table 13-6, where applicable. Many of the pollutants
of interest for the Kentucky sites do not have annual averages due to the relatively short
sampling duration (those sites/methods that began in July). However, pollutant-specific average
concentrations for all valid VOC and carbonyl compound samples collected over the entire
sample period are provided in Appendix J and Appendix L. 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.
13-59
-------
Table 13-6. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Kentucky Monitoring Sites
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Hg/m3)
2nd
Quarter
Average
(Hg/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Hg/m3)
Annual
Average
(Ug/m3)
Health Department, Ashland, Kentucky - ASKY
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
29/29
29/29
28/29
29/29
24/29
29/29
29/29
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
1.30
±0.20
0.76
±0.21
0.08
±0.03
0.64
±0.03
0.05
±0.01
0.32
±0.09
3.67
±0.80
1.44
±0.47
1.15
±0.30
0.15
±0.05
0.71
±0.03
0.07
±0.02
0.37
±0.11
1.82
±0.40
NA
NA
NA
NA
NA
NA
NA
21st and Greenup, Ashland, Kentucky - ASKY-M
Arsenic (PM10)a
Cadmium (PM10)a
Lead (PM10)a
Manganese (PM10)a
Nickel (PM10)a
50/50
50/50
50/50
50/50
50/50
NA
NA
NA
NA
NA
1.95
±0.56
0.69
±0.26
18.31
±7.06
46.19
± 15.74
3.36
±1.23
1.83
±0.81
0.41
±0.13
11.21
±5.11
22.93
±7.05
2.51
±1.55
1.39
±0.59
0.51
±0.41
8.43
±3.64
21.70
± 10.23
1.85
±0.88
1.79
±0.37
0.56
±0.16
14.35
±4.60
34.09
± 10.53
2.94
±0.90
Grayson, Kentucky - GLKY
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
Formaldehyde
Arsenic (PM10)a
Manganese (PM10)a
61/61
61/61
59/61
61/61
56/61
61/61
59/59
59/59
0.65
±0.14
0.55
±0.10
0.05
±0.01
0.66
±0.09
0.08
±0.02
0.80
±0.24
0.49
±0.15
2.57
±1.03
0.92
±0.14
0.41
±0.08
0.04
±0.02
0.71
±0.04
0.07
±0.01
2.27
±0.54
0.58
±0.23
5.56
±3.16
0.76
±0.18
0.34
±0.06
0.05
±0.03
0.65
±0.04
0.05
±0.01
2.44
±0.78
0.59
±0.24
3.23
±1.14
0.74
±0.26
0.57
±0.11
0.11
±0.04
0.73
±0.03
0.07
±0.01
1.00
±0.19
0.69
±0.24
3.48
±2.02
0.77
±0.09
0.46
±0.05
0.06
±0.01
0.69
±0.03
0.07
±0.01
1.64
±0.31
0.58
±0.10
3.71
±0.98
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.
13-60
-------
Table 13-6. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Kentucky Monitoring Sites (Continued)
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Hg/m3)
2nd
Quarter
Average
(Hg/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Hg/m3)
Annual
Average
(Ug/m3)
Baskett, Kentucky - BAKY
Arsenic (PM10)a
Manganese (PM10)a
50/50
50/50
NA
NA
0.76
±0.23
7.42
±1.48
1.10
±0.46
6.48
±1.32
1.00
±0.42
6.71
±1.91
0.93
±0.19
6.74
±0.84
Atmos Energy, Calvert City, Kentucky - ATKY
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
Vinyl chloride
29/29
25/29
29/29
26/29
16/29
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
0.58
±0.15
0.11
±0.10
0.70
±0.04
0.71
±0.84
0.43
±0.48
0.63
±0.15
0.10
±0.05
0.78
±0.09
0.48
±0.44
1.05
±1.35
NA
NA
NA
NA
NA
Smithland, Kentucky - BLKY
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
Vinyl chloride
26/26
21/26
26/26
23/26
14/26
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
0.50
±0.16
0.24
±0.39
0.70
±0.06
0.78
±1.11
0.11
±0.09
0.55
±0.15
0.38
±0.59
0.78
±0.09
0.89
±0.85
0.11
±0.11
NA
NA
NA
NA
NA
Calvert City Elementary, Calvert City, Kentucky - CCKY
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
Arsenic (PM10)a
Manganese (PM10)a
26/26
24/26
26/26
23/26
47/47
47/47
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
0.65
±0.19
7.20
±1.25
0.51
±0.16
0.06
±0.03
0.69
±0.04
0.46
±0.35
1.22
±0.86
6.85
±1.67
0.57
±0.10
0.09
±0.04
0.77
±0.04
0.17
±0.11
0.78
±0.24
6.35
±2.39
NA
NA
NA
NA
0.86
±0.28
6.50
±0.96
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.
13-61
-------
Table 13-6. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Kentucky Monitoring Sites (Continued)
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Hg/m3)
2nd
Quarter
Average
(Hg/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Hg/m3)
Annual
Average
(Ug/m3)
Lazy Daze, Calvert City, Kentucky - LAKY
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dibromoethane
1 ,2-Dichloroethane
Hexachloro- 1 , 3 -butadiene
1 , 1 ,2,2-Tetrachloroethane
1 , 1 ,2-Trichloroethane
Vinyl chloride
29/29
28/29
29/29
3/29
27/29
7/29
6/29
5/29
19/29
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
0.72
±0.27
0.09
±0.05
0.77
±0.07
0.01
±0.01
0.80
±0.51
0.03
±0.02
0.02
±0.01
0.02
±0.02
0.19
±0.13
0.61
±0.09
0.14
±0.09
0.76
±0.06
<0.01
±0.01
0.30
±0.23
0.01
±0.01
0.02
±0.02
0.03
±0.06
0.14
±0.12
NA
NA
NA
NA
NA
NA
NA
NA
NA
TVA Substation, Calvert City, Kentucky - TVKY
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
1 , 1 ,2,2-Tetrachloroethane
1,1,2-Trichloroethane
Vinyl chloride
28/28
24/28
28/28
27/28
4/28
6/28
19/28
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
1.15
±0.79
0.24
±0.23
1.14
±0.57
2.77
±2.75
0.01
±0.01
0.06
±0.08
0.87
±1.02
0.71
±0.32
0.17
±0.15
1.28
±0.60
1.91
±2.07
0.01
±0.02
0.01
±0.01
0.28
±0.26
NA
NA
NA
NA
NA
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 are presented in ng/m3 for ease of viewing.
13-62
-------
Table 13-6. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Kentucky Monitoring Sites (Continued)
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Hg/m3)
2nd
Quarter
Average
(Hg/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Hg/m3)
Annual
Average
(Ug/m3)
Lexington, Kentucky - LEKY
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
£>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
Arsenic (PM10)a
Manganese (PM10)a
21121
29/29
27/29
29/29
14/29
24/29
29/29
21121
49/49
49/49
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
0.67
±0.23
7.01
±1.69
NA
0.61
±0.11
0.10
±0.04
0.63
±0.03
0.06
±0.03
0.04
±0.02
0.28
±0.07
NA
0.86
±0.30
6.85
±1.64
1.19
±0.26
0.63
±0.13
0.12
±0.05
0.66
±0.03
0.01
±0.01
0.06
±0.01
0.22
±0.07
2.03
±0.44
1.25
±0.42
6.62
±2.33
NA
NA
NA
NA
NA
NA
NA
NA
0.92
±0.17
6.69
±0.96
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.
Observations for the Ashland sites from Table 13-6 include the following:
• First and second quarter and annual average concentrations could not be calculated
for the pollutants of interest for ASKY because sampling did not begin until July
2012. However, Appendix J and Appendix L provide the pollutant-specific average
concentrations for all valid VOC and carbonyl compound samples collected over the
entire sample period.
• With the exception of 1,3-butadiene and 1,2-dichloroethane, each of the pollutants of
interest for ASKY were detected in all the valid VOC samples collected.
• With the exception of formaldehyde, concentrations of the pollutants of interest were
higher in the fourth quarter than the third quarter, although in most cases, the
difference is not statistically significant.
• The third quarter formaldehyde concentration for ASKY is twice the fourth quarter
average. A review of the data shows that all eight measurements greater than
3.50 |ig/m3 were measured between July and September while four of the five
concentrations less than 1.50 |ig/m3 were measured in December (with the fifth
measured in October).
13-63
-------
• Sampling of PMio metals began in March at ASKY-M. Each of the five metal
pollutants of interest was detected in all of the samples collected. Thus, three
quarterly averages and an annual average are presented for ASKY-M.
• The pollutant of interest with the highest annual average concentration for ASKY-M
is manganese (34.09 ± 10.53 ng/m3), followed by lead (14.35 ± 4.60 ng/m3), and
nickel (2.94 ± 0.90 ng/m3).
• For each of the pollutants of interest for ASKY-M, the highest quarterly average was
calculated for the second quarter of 2012. However, with the exception of arsenic, the
maximum concentration for each of the pollutants of interest was measured on
March 22, 2012. This date is part of the first quarter, for which quarterly averages
could not be calculated.
• The second quarter manganese average is nearly twice the other quarterly averages,
although the confidence intervals for these averages indicate that there is a
considerable amount of variability associated with these averages. The manganese
concentrations range from 1.07 ng/m3 to 236 ng/m3, with a median concentration of
25.2 ng/m3. The maximum concentration measured at ASKY-M is the second highest
manganese concentration measured program-wide. Of the 18 concentrations greater
than 50 ng/m3 measured across the program (PMio only), 10 were measured at
ASKY-M (with three measured at PXSS and five measured at S4MO).
• Similarly, some of the highest concentrations of arsenic program-wide were measured
at ASKY-M. Of the 13 concentrations greater than 3 ng/m3 measured across the
program (PMio only), seven were measured at ASKY-M. Arsenic concentrations
measured at ASKY-M range from 0.10 ng/m3 to 5.90 ng/m3, with a median
concentration of 1.61 ng/m3. The maximum concentration measured at ASKY-M is
the second highest concentration measured program-wide.
• Some of the highest concentrations of lead measured program-wide were also
measured at ASKY-M. Of the 21 lead concentrations greater than 20 ng/m3 measured
across the program (PMio only), 12 were measured at ASKY-M (with one measured
at PAFL and eight measured at S4MO). Lead concentrations measured at ASKY-M
range from 1.27 ng/m3 to 100.1 ng/m3, with a median concentration of 9.38 ng/m3.
The maximum concentration measured at ASKY-M is again the second highest
concentration measured program-wide, with only S4MO and ASKY-M measuring
concentrations of lead greater than 100 ng/m3. However, the second highest
concentration measured at ASKY-M is considerably less (42.5 ng/m3).
• Concentrations of nickel measured at ASKY-M range from 0.14 ng/m3 to 17.3 ng/m3,
with a median concentration of 1.88 ng/m3. The maximum concentration measured at
ASKY-M is the highest nickel concentration measured program-wide. Of the four
3
10
nickel concentrations greater than 10 ng/m measured across the program (PM
only), two were measured at ASKY-M (with the other two measured at SEW A).
13-64
-------
• Table 4-12 presents the sites with the 10 highest annual average concentrations for
each of the metal program-level pollutants of interest. This table shows that the
highest annual averages for arsenic, manganese, and nickel calculated across the
program were all calculated for ASKY-M.
Observations for GLKY from Table 13-6 include the following:
• The only pollutant of interest with an annual average concentration greater than
1 |ig/m3 is formaldehyde (1.64 ± 0.31 |ig/m3). However, this is one of the lowest
annual averages of formaldehyde calculated for NMP sites sampling carbonyl
compounds.
• Concentrations of formaldehyde were considerably higher during the warmer months
of the year, based on the quarterly averages. All but one of the 15 measurements
greater than 2 |ig/m3 were measured during the second or third quarters, with the
three highest concentrations all measured in July. Conversely, all but one of the 16
concentrations less than 0.75 |ig/m3 were measured in the first or fourth quarters.
• Concentrations of acetaldehyde do not exhibit the same tendency as formaldehyde.
Concentrations of this pollutant were highest during the second quarter (although not
significantly so). The second quarter is the quarter during which the greatest number
of concentrations greater than 1 |ig/m3 were measured (eight), with three measured
during the third quarter (all in July), and two in the fourth quarter. Although the
maximum acetaldehyde concentration (2.28 |ig/m3) was measured at GLKY in
October, the next highest concentration measured during the fourth quarter was
roughly half as high (1.18 |ig/m3) and no other measurements greater than 1 |ig/m3
were measured that quarter. This explains the higher confidence interval calculated
for the fourth quarter.
• The second quarter average concentration of manganese is higher than the other
quarterly averages and has a relatively high confidence interval associated with it.
The maximum concentration of manganese was measured at GLKY on April 3, 2012
(24.4 ng/m3). The next highest concentration measured at GLKY was considerably
less (13.6 ng/m3) and the second highest concentration measured during the second
quarter was roughly half as high (13.0 ng/m3). No other manganese measurements
greater than 10 ng/m3 were measured at this site.
Observations for BAKY from Table 13-6 include the following:
• BAKY has only two pollutants of interest: arsenic and manganese.
• Concentrations of manganese are considerably higher than the concentrations of
arsenic measured at this site.
• Among NMP sites sampling PMio metals, BAKY has the fourth highest annual
average concentration of arsenic (0.93 ± 0.20 ng/m3). Arsenic concentrations
measured at BAKY range from 0.003 ng/m3 to 3.71 ng/m3. The maximum arsenic
13-65
-------
concentration measured at BAKY is one of the top 10 concentrations measured
among NMP sites sampling arsenic.
• BAKY also has the seventh highest annual average concentration of manganese
among NMP sites sampling PMi0 metals (6.74 ± 0.84 ng/m3). Manganese
concentrations measured at BAKY range from 0.99 ng/m3 to 13.5 ng/m3.
Observations for the Calvert City monitoring sites from Table 13-6 include the following:
• The only annual averages that could be calculated for the Calvert City sites are for the
metal pollutants of interest for CCKY. This is because CCKY is the only site
sampling PMi0 metals, for which sampling began in March. VOC sampling at all five
Calvert City sites began in July. However, Appendix J provides the pollutant-specific
average concentrations for all valid VOC samples collected over the entire sample
period for each site.
• CCKY has two metal pollutants of interest: arsenic and manganese. Concentrations of
manganese are considerably higher than the concentrations of arsenic measured at
this site.
• Among NMP sites sampling PMi0 metals, CCKY has the sixth highest annual
average concentration of arsenic (0.86 ± 0.28 ng/m3). Arsenic concentrations
measured at CCKY range from 0.15 ng/m3 to 5.86 ng/m3. The maximum arsenic
concentration measured at CCKY is the third highest concentration among NMP sites
sampling arsenic, behind only two measurements (from S4MO and ASKY-M).
• Among NMP sites sampling PMio metals, CCKY has the ninth highest annual
average concentration of manganese (6.50 ± 0.96 ng/m3). Manganese concentrations
measured at CCKY range from 1.55 ng/m3 to 17.9 ng/m3 with a median concentration
of 5.66 ng/m3.
• 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. Across the
program, this pollutant was detected in less than 12 percent of the total samples
collected. Together, the five Calvert City sites account for more than half (79) of the
154 measured detections of this pollutant. The Calvert City sites account for all 43
concentrations of vinyl chloride greater than 0.15 |ig/m3 measured across the
program. The maximum concentration of vinyl chloride was measured at ATKY
(9.81 |ig/m3), with 10 additional measurements greater than 1 |ig/m3 measured at
these sites (four at ATKY, one at CCKY, and five at TVKY). All of the quarterly
average concentrations of vinyl chloride for these sites, where they could be
calculated, have relatively large confidence intervals, indicating the relatively large
amount of variability associated with these measurements.
13-66
-------
• Another pollutant for which the highest concentrations program-wide were measured
at the Calvert City sites is 1,2-dichloroethane. The 56 highest concentrations of
1,2-dichloroethane across the program (those greater than 0.18 |ig/m3) were all
measured at the Calvert City sites. This includes all 29 measurements greater than
1 |ig/m3 and two greater than 10 |ig/m3. The four highest concentrations of
1,2-dichloroethane were measured at TVKY and ranged from 7.17 |ig/m3 to
17.1 |ig/m3. In many cases, the quarterly average concentrations of
1,2-dichloroethane for these sites have confidence intervals similar to or greater in
magnitude than the quarterly averages themselves, indicating the relatively large
amount of variability associated with these measurements. This is particularly true for
ATKY, BLKY, and TVKY.
• Some of the highest measurements of carbon tetrachloride were also measured at the
Calvert City sites, particularly TVKY. Of the 17 carbon tetrachloride concentrations
greater than 1 |ig/m3 measured across the program, 13 were measured at the Calvert
City sites (nine at TVKY, two at LAKY, and one each at BLKY and ATKY). The
quarterly average concentrations of carbon tetrachloride for TVKY are the two
highest quarterly averages of this pollutant among all NMP sites sampling VOC (and
the only ones greater than 1 |ig/m3). The fourth quarter averages of this pollutant for
BLKY and ATKY rank third and fourth, respectively. The third and fourth quarter
averages for LAKY and the fourth quarter average for CCKY are the only other
quarterly averages program-wide that are greater than 0.75 jig/m3 besides the second
quarter average for PROK.
• The highest quarterly average concentrations of benzene among the Calvert City sites
were calculated for TVKY and LAKY. The maximum benzene concentration was
measured at TVKY on July 20, 2012 (5.24 |ig/m3) and is the fourth highest benzene
concentration measured among NMP sites sampling for this pollutant. The next
highest concentration measured at TVKY is roughly half as high (2.83 |ig/m3). The
maximum benzene concentration measured at LAKY was also measured on
July 20, 2012 but was considerably less (2.04 |ig/m3).
• The confidence intervals for both quarterly averages of 1,3-butadiene for BLKY are
greater than the averages themselves, indicating that outliers may be affecting these
averages. The two highest 1,3-butadiene concentrations measured at BLKY are the
maximum concentrations measured across the program (4.10 |ig/m3 and 2.31 |ig/m3).
The next highest concentration measured at BLKY is an order of magnitude less
(0.344 |ig/m3). The third and fifth highest 1,3-butadiene measurements program-wide
were collected at TVKY. Thus, measurements from BLKY and TVKY account for
four of the five 1,3-butadiene measurements greater than 1 |ig/m3 across the program.
• The two highest 1,1,2,2-tetrachloroethylene concentrations measured program-wide
were both measured at LAKY and TVKY on October 12, 2012 (0.179 |ig/m3 and
0.169 |ig/m3, respectively). The next highest concentrations measured at these sites
were considerably less.
13-67
-------
• The Calvert City sites also measured some of the highest concentrations of
1,1,2-trichloroethane across the program, as these sites account for all 15
measurements greater than 0.06 |ig/m3. This pollutant was detected in only 38
samples across the program in 2012, and 18 of them were collected at the Calvert
City sites.
Observations for LEKY from Table 13-6 include the following:
• The only annual averages that could be calculated for LEKY are for the metal
pollutants of interest. This is because sampling of PMio metals began under the NMP
in March 2012 while sampling for VOCs and carbonyl compounds did not begin until
July. However, Appendix J and Appendix L provide the pollutant-specific average
concentrations for all valid VOC and carbonyl compound samples collected over the
entire sample period.
• The annual average concentration of arsenic for LEKY is the fifth highest annual
average concentration for this pollutant among NMP sites sampling PMio metals.
Arsenic concentrations measured at LEKY range from 0.03 ng/m3 to 2.35 ng/m3, with
a median concentration of 0.76 ng/m3. The fourth quarter average arsenic
concentration is greater than the other available quarterly averages for LEKY. Four of
the five highest concentrations of arsenic were measured at LEKY during the fourth
quarter of 2012.
• The annual average concentration of manganese for LEKY is the eighth highest
annual average for this pollutant among NMP sites sampling PMio metals.
Manganese concentrations measured at LEKY range from 0.66 ng/m3 to 16.6 ng/m3,
with a median concentration of 6.67 ng/m3. Although the fourth quarter average
manganese concentration is not the highest of the quarterly averages for LEKY, it has
a relatively large confidence interval, indicating a relatively large amount of
variability is associated with the measurements. Both the minimum and maximum
concentrations of manganese were measured at LEKY during this quarter.
• The third quarter average concentration ofp-dichlorobenzene for LEKY is
considerably higher than the fourth quarter average concentration. The two highest
/>-dichlorobenzene concentrations measured at LEKY (those greater than 0.1 |ig/m3)
were both measured during the third quarter of 2012. In addition, the detection rate of
/>-dichlorobenzene was considerably higher during the third quarter than the fourth
quarter. There were two non-detects of this pollutant reported for the third quarter
while there were 13 for the fourth quarter.
• The third quarter average concentrations did not vary significantly from the fourth
quarter average concentrations for the remaining pollutant of interest.
Additional observations for the Kentucky monitoring sites from Table 13-6 include:
• Some of the highest arsenic concentrations program-wide were measured at the
Kentucky sites. Of the 46 concentrations of arsenic greater than 2 ng/m3, 23 were
measured at ASKY-M, three were measured at BAKY, two at CCKY, and six at
13-68
-------
LEKY, with the Kentucky sites accounting for 34 of these measurements. Kentucky
sites account for half of the highest annual average concentrations of arsenic, as
shown in Table 4-12.
13.4.2 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 annual averages are available. Thus, box plots were created for the
pollutants of interest for GLKY and the PMio metals shaded in gray in Table 13-5 for ASKY-M,
BAKY, CCKY, and LEKY. Figures 13-41 through 13-51 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 3.5.3.1.
Figure 13-41. Program vs. Site-Specific Average Acetaldehyde Concentration
|±
9 12
Concentration {[og/m3)
Program:
Site:
1st Quartile
D
Site Average
o
2nd Quartile 3rd Quartile
D D
Site Concentration Range
^^^^—
4th Quartile Average
D 1
13-69
-------
Figure 13-42. Program vs. Site-Specific Average Arsenic (PMi0) Concentrations
BAKY
i±
CCKY
i±
GLKY
• 0
• °
1 1 1 1 1 1 1
Concentration {
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o
Figure 13-43. Program vs. Site-Specific Average Benzene Concentration
Concentration {[og/m3)
Program:
Site:
IstQuartile
D
Site Average
0
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
13-70
-------
Figure 13-44. Program vs. Site-Specific Average 1,3-Butadiene Concentration
GLKY
•H-
Program Max Concentration = 4.10 ug/m3
0.75 1 1.25
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o
Figure 13-45. Program vs. Site-Specific Average Cadmium Concentration
E
1.5
Concentration (ng/m3)
Program:
Site:
IstQuartile
D
Site Average
0
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
Figure 13-46. Program vs. Site-Specific Average Carbon Tetrachloride Concentration
GLKY
•4—
3
1
I 3 4
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile
D D D
Site: SiteAverage Site Concentration Range
o —
4thQuartile Average
5
13-71
-------
Figure 13-47. Program vs. Site-Specific Average 1,2-Dichloroethane Concentration
GLKY
Program Max Concentration = 17.01 ug/m3
0.4 0.5 0.6
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: SiteAverage Site Concentration Range
o
Figure 13-48. Program vs. Site-Specific Average Formaldehyde Concentration
—
10
12
Concentration {[og/m3;
Program:
Site:
IstQuartile
D
SiteAverage
O
2ndQuartile SrdQuartile
• a
Site Concentration Range
^^^^—
4thQuartile Average
D 1
14
Figure 13-49. Program vs. Site-Specific Average Lead (PMio) Concentration
ASKY-M
I
20
40
60
Concentration (ng/m3)
80
100
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: SiteAverage Site Concentration Range
O
120
13-72
-------
Figure 13-50. Program vs. Site-Specific Average Manganese (PMi0) Concentrations
Program Max Concentration = 275 ng/m3
BAKY
] Program Max Concentration = 275 ng/m3
] Program Max Concentration = 275 ng/m3
] Program Max Concentration = 275 ng/m3
LEKY
Program Max Concentration = 275 ng/m3
30
60 90
Concentration (ng/m3)
120
150
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: SiteAverage Site Concentration Range
o —
Figure 13-51. Program vs. Site-Specific Average Nickel (PMi0) Concentration
ASKY-M
10
0
3
2 4 6 8 10 12 14 16
Concentration (ng/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
• D D D 1
Site: SiteAverage Site Concentration Range
o —
1
13-73
-------
Observations from Figures 13-41 through 13-51 include the following:
• Figure 13-41 is the box plot for acetaldehyde for GLKY. The range of
acetaldehyde concentrations measured at GLKY is rather small, as the maximum
concentration measured at GLKY is just greater than the program-level third
quartile. The annual average acetaldehyde concentration for GLKY is less than
the program-level first quartile and ranks among the lowest annual averages of
this pollutant for NMP sites sampling carbonyl compounds.
• Figure 13-42 presents the box plots for the five Kentucky sites for which arsenic
is a pollutant of interest and an annual average concentration could be calculated.
The box plots show that the range of arsenic concentrations measured was
smallest for GLKY and largest for ASKY-M and CCKY. With the exception of
GLKY, all of the annual average concentrations of arsenic were greater than the
program-level average concentration. With the exception of CCKY, these annual
averages are also greater than the program-level third quartile. Note however, that
the maximum arsenic concentration across the program was not measured at any
of the Kentucky sites shown.
• Figure 13-43 is the box plot for benzene for GLKY. Similar to acetaldehyde, the
maximum concentration measured at GLKY is just greater than the program-level
third quartile and the annual average concentration for benzene is less than the
program-level first quartile.
• Figure 13-44 is the box plot for 1,3-butadiene for GLKY. Note that the program-
level maximum concentration (4.10 |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 2 |ig/m3. The annual average concentration for GLKY is similar
to the program-level median concentration but less than the program-level
average concentration. The maximum 1,3-butadiene concentration measured at
GLKY is an order of magnitude less than the maximum concentration measured
across the program. Note that the maximum concentration of 1,3-butadiene across
the program was measured at BLKY, as discussed in the previous section. Two
non-detects of 1,3-butadiene were measured at GLKY.
• Figure 13-45 is the box plot for cadmium for ASKY-M. Although cadmium was
sampled for by the other Kentucky sites sampling PMi0 metals, this is the only
site for which cadmium was identified as a pollutant of interest. Although the
maximum concentration across the program was not measured at ASKY-M, this
site does have one of the higher measurements. The annual average concentration
of cadmium for ASKY-M is more than three times greater than the program-level
average concentration. This site has the second highest annual average
concentration of cadmium, behind only S4MO.
• Figure 13-46 is the box plot for carbon tetrachloride for GLKY. This box plots
shows that the annual average concentration of this pollutant is nearly equivalent
to the program-level average concentration, both of which are similar in
magnitude to the program-level median concentration.
13-74
-------
• Figure 13-47 is the box plot for 1,2-dichloroethane for GLKY. Note that the
program-level maximum concentration (17.01 |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 entire range of 1,2-dichloroethane
concentrations measured at GLKY is less than the program-level average
concentration. The annual average concentration for GLKY is greater than the
program-level first quarter but less than the program level median concentration.
Note that the maximum concentration across the program was measured at
TVKY, as discussed in the previous section.
• Figure 13-48 is the box plot for formaldehyde for GLKY. The maximum
concentration measured at GLKY is roughly half the maximum concentration
measured across the program. The annual average formaldehyde concentration for
GLKY is less than both the program-level average and median concentrations and
just greater than the program-level first quartile.
• Figure 13-49 is the box plot for lead for ASKY-M. Although lead was sampled
for by the other Kentucky sites sampling PMio metals, ASKY-M is the only site
for which lead was identified as a pollutant of interest. Although the maximum
concentration across the program was not measured at ASKY-M, this site does
have one of the higher measurements. The annual average concentration of lead
for ASKY-M is more than three times greater than the program-level average
concentration and is the highest annual average concentration of lead calculated
among NMP sites sampling PMio metals.
• Figure 13-50 presents the box plots for the five Kentucky sites for which
manganese is a pollutant of interest and an annual average concentration could be
calculated. Note that the program-level maximum concentration (275 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 of the box plot has been reduced to 150 ng/m3. The box
plots show that the range of manganese concentrations measured was smallest for
BAKY and largest for ASKY-M. Although the maximum manganese
concentration across the program was not measured at ASKY-M, this site's
maximum concentration is greater than the scale in Figure 13-50. With the
exception of ASKY-M, all of the annual average concentrations of manganese are
less than the program-level average concentration. The annual average for ASKY-
M is more than three times greater than the program-level average concentration.
• Figure 13-51 is the box plot for nickel, and like lead and cadmium, is a pollutant
of interest for only ASKY-M. This box plot shows that the annual average
concentration of nickel for ASKY-M is more than two times greater than the
program-level average concentration. The maximum concentration measured
across the program was measured at ASKY-M and this site has the highest annual
average concentration of nickel calculated among NMP sites sampling
metals.
13-75
-------
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 3.5.3.2. The
only pollutant group for which GLKY has sampled under the NMP since at least 2008 is
hexavalent chromium and PAHs; however, hexavalent chromium did not fail any screens and
none of the PAHs that failed screens were identified as pollutants of interest for GLKY. Thus, a
trends analysis was not performed for this site.
13.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk at the
Kentucky monitoring sites. Refer to Sections 3.3 and 3.5.3 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.
13.5.1 Risk-Based Screening Assessment Using MRLs
A risk-based screening was conducted by comparing the concentration data from the
Kentucky monitoring sites to the ATSDR MRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 day to 14 days); intermediate (exposures of 15 days to 364 days); and chronic
(exposures of 1 year or greater). The preprocessed daily measurements of the pollutants of
interest were compared to the acute MRLs; the quarterly averages were compared to the
intermediate MRLs; and the annual averages were compared to the chronic MRLs.
As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2012.
13.5.2 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-
13-76
-------
monitoring priorities. Refer to Section 3.5.3.4 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 13-7, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Table 13-7. Risk Approximations for the Kentucky Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
Health Department, Ashland, Kentucky - ASKY
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
0.0000022
0.0000078
0.00003
0.000006
0.000026
0.0000025
0.000013
0.009
0.03
0.002
0.1
2.4
1
0.0098
29/29
29/29
28/29
29/29
24/29
29/29
29/29
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
21st and Greenup, Ashland, Kentucky - ASKY-M
Arsenic (PM10)a
Cadmium (PM10) a
Lead(PM10)a
Manganese (PM10) a
Nickel (PM10)a
0.0043
0.0018
0.00048
0.000015
0.00001
0.00015
0.00005
0.00009
50/50
50/50
50/50
50/50
50/50
O.01
±<0.01
0.01
±0.01
0.01
±0.01
0.03
±0.01
O.01
±0.01
7.70
1.01
1.41
0.12
0.06
0.10
0.68
0.03
NA = Not available due to the criteria for calculating an annual average.
- = A Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 13-6.
13-77
-------
Table 13-7. Risk Approximations for the Kentucky Monitoring Sites (Continued)
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
Grayson, Kentucky - GLKY
Acetaldehyde
Arsenic (PM10)a
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
Formaldehyde
Manganese (PM10) a
0.0000022
0.0043
0.0000078
0.00003
0.000006
0.000026
0.000013
0.009
0.000015
0.03
0.002
0.1
2.4
0.0098
0.00005
61/61
59/59
61/61
59/61
61/61
56/61
61/61
59/59
0.77
±0.09
0.01
±0.01
0.46
±0.05
0.06
±0.01
0.69
±0.03
0.07
±0.01
1.64
±0.31
O.01
±O.01
1.68
2.51
3.62
1.92
4.13
1.75
21.34
0.09
0.04
0.02
0.03
0.01
O.01
0.17
0.07
Baskett, Kentucky - BAKY
Arsenic (PM10)a
Manganese (PM10)a
0.0043
0.000015
0.00005
50/50
50/50
O.01
±O.01
0.01
±0.01
4.00
0.06
0.13
Atmos Energy, Calvert City, Kentucky - ATKY
Benzene
1,3 -Butadiene
1 ,2-Dichloroethane
Carbon Tetrachloride
Vinyl chloride
0.0000078
0.00003
0.000026
0.000006
0.0000088
0.03
0.002
2.4
0.1
0.1
29/29
25/29
26/29
29/29
16/29
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Smithland, Kentucky - BLKY
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
Vinyl chloride
0.0000078
0.00003
0.000006
0.000026
0.0000088
0.03
0.002
0.1
2.4
0.1
26/26
21/26
26/26
23/26
14/26
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA = Not available due to the criteria for calculating an annual average.
— = A Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 13-6.
13-78
-------
Table 13-7. Risk Approximations for the Kentucky Monitoring Sites (Continued)
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
Calvert City Elementary, Calvert City, Kentucky - CCKY
Arsenic (PM10)a
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
Manganese (PM10) a
0.0043
0.0000078
0.00003
0.000006
0.000026
0.000015
0.03
0.002
0.1
2.4
0.00005
47/47
26/26
24/26
26/26
23/26
47/47
<0.01
±<0.01
NA
NA
NA
NA
0.01
±<0.01
3.68
NA
NA
NA
NA
0.06
NA
NA
NA
NA
0.13
Lazy Daze, Calvert City, Kentucky - LAKY
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dibromoethane
1 ,2-Dichloroethane
Hexachloro- 1 , 3 -butadiene
1 , 1 ,2,2-Tetrachloroethane
1 , 1 ,2-Trichloroethane
Vinyl chloride
0.0000078
0.00003
0.000006
0.0006
0.000026
0.000022
0.000058
0.000016
0.0000088
0.03
0.002
0.1
0.009
2.4
0.09
_
0.4
0.1
29/29
28/29
29/29
3/29
27/29
7/29
6/29
5/29
19/29
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA = Not available due to the criteria for calculating an annual average.
- = A Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 13-6.
13-79
-------
Table 13-7. Risk Approximations for the Kentucky Monitoring Sites (Continued)
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
TVA Substation, Calvert City, Kentucky - TVKY
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
1 , 1 ,2,2-Tetrachloroethane
1,1,2-Trichloroethane
Vinyl chloride
0.0000078
0.00003
0.000006
0.000026
0.000058
0.000016
0.0000088
0.03
0.002
0.1
2.4
0.4
0.1
28/28
24/28
28/28
27/28
4/28
6/28
19/28
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Lexington, Kentucky - LEKY
Acetaldehyde
Arsenic (PM10)a
Benzene
1,3 -Butadiene
Carbon Tetrachloride
/>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
Manganese (PM10) a
0.0000022
0.0043
0.0000078
0.00003
0.000006
0.000011
0.000026
0.0000025
0.000013
0.009
0.000015
0.03
0.002
0.1
0.8
2.4
1
0.0098
0.00005
21121
49/49
29/29
27/29
29/29
14/29
24/29
29/29
21121
49/49
NA
<0.01
±<0.01
NA
NA
NA
NA
NA
NA
NA
0.01
±0.01
NA
3.94
NA
NA
NA
NA
NA
NA
NA
NA
0.06
NA
NA
NA
NA
NA
NA
NA
0.13
NA = Not available due to the criteria for calculating an annual average.
— = A Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 13-6.
Observations for the Kentucky monitoring sites from Table 13-7 include the following:
• Few annual averages, and thus, cancer risk and noncancer hazard approximations,
could be calculated for the Kentucky monitoring sites due to a relatively short
sampling duration.
13-80
-------
• The pollutants with the highest annual average concentrations for GLKY are
formaldehyde, acetaldehyde, and carbon tetrachloride. The pollutants with the highest
cancer risk approximations for GLKY are formaldehyde, carbon tetrachloride, and
benzene. All of the noncancer hazard approximations for the pollutants of interest for
GLKY are considerably less than an HQ of 1.0 (0.20 or less), indicating that no
adverse health effects are expected from these individual pollutants. The highest
noncancer hazard approximation was calculated for formaldehyde.
• The cancer risk approximations for arsenic range from 2.51 in-a-million (GLKY) to
7.70 in-a-million (ASKY-M). All of the noncancer hazard approximations for arsenic
are less than an HQ of 1.0 (0.12 or less), indicating that no adverse health effects are
expected from arsenic individually.
• A cancer risk factor is not available for manganese. All of the noncancer hazard
approximations for manganese are less than an HQ of 1.0 (0.68 or less), indicating
that no adverse health effects are expected from manganese individually. ASKY-M's
noncancer hazard approximation for manganese is the second highest noncancer
hazard approximation calculated for any program-wide pollutant of interest, behind
only TOOK's noncancer hazard approximation for manganese (0.77).
13.5.3 Risk-Based Emissions Assessment
In addition to the risk-based screenings discussed above, Tables 13-8 and 13-9 present an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 13-8 presents the 10 pollutants with the highest emissions from the 2011 NEI that have
cancer toxicity factors. Table 13-8 also presents the 10 pollutants with the highest toxicity-
weighted emissions, based on the weighting schema described in Section 3.5.3.5. Lastly,
Table 13-8 provides the 10 pollutants with the highest cancer risk approximations (in-a-million)
for each site, as presented in Table 13-7. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 13-8. Table 13-9 presents similar
information, but is limited to those pollutants with noncancer toxicity factors.
13-81
-------
Table 13-8. 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)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
Health Department, Ashland, Kentucky (Boyd County) - ASKY
Benzene
Formaldehyde
POM, Group la
Acetaldehyde
Ethylbenzene
Coke Oven Emissions, PM
1,3 -Butadiene
2,4-Dinitrotoluene
Naphthalene
Nickel, PM
55.37
20.84
14.58
11.02
10.16
7.32
2.89
2.20
2.16
1.40
Coke Oven Emissions, PM
POM, Group la
Hexavalent Chromium, PM
Nickel, PM
Benzene
Formaldehyde
2,4-Dinitrotoluene
1,3 -Butadiene
POM, Group 3
Naphthalene
7.25E-03
1.28E-03
9.89E-04
6.72E-04
4.32E-04
2.71E-04
1.96E-04
8.67E-05
7.99E-05
7.34E-05
21st and Greenup, Ashland, Kentucky (Boyd County) - ASKY-M
Benzene
Formaldehyde
POM, Group la
Acetaldehyde
Ethylbenzene
Coke Oven Emissions, PM
1,3 -Butadiene
2,4-Dinitrotoluene
Naphthalene
Nickel, PM
55.37
20.84
14.58
11.02
10.16
7.32
2.89
2.20
2.16
1.40
Coke Oven Emissions, PM
POM, Group la
Hexavalent Chromium, PM
Nickel, PM
Benzene
Formaldehyde
2,4-Dinitrotoluene
1,3 -Butadiene
POM, Group 3
Naphthalene
7.25E-03
1.28E-03
9.89E-04
6.72E-04
4.32E-04
2.71E-04
1.96E-04
8.67E-05
7.99E-05
7.34E-05
Arsenic 7.70
Nickel 1.41
Cadmium 1.01
oo
to
-------
Table 13-8. 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)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Grayson, Kentucky (Carter County) - GLKY
Formaldehyde
Benzene
Acetaldehyde
Ethylbenzene
Naphthalene
1,3 -Butadiene
POM, Group 2d
POM, Group 2b
POM, Group 6
Trichloroethylene
22.49
16.98
11.95
6.80
2.47
2.04
0.53
0.50
0.04
0.03
Formaldehyde
Benzene
Naphthalene
1,3 -Butadiene
POM, Group 2d
POM, Group 2b
POM, Group 3
Acetaldehyde
POM, Group 5a
Ethylbenzene
2.92E-04
1.32E-04
8.40E-05
6.13E-05
4.67E-05
4.41E-05
4.20E-05
2.63E-05
1.71E-05
1.70E-05
Formaldehyde
Carbon Tetrachloride
Benzene
Arsenic
1,3 -Butadiene
1 ,2-Dichloroethane
Acetaldehyde
21.34
4.13
3.62
2.51
1.92
1.75
1.68
Baskett, Kentucky (Henderson County) - BAKY
Formaldehyde
Benzene
Acetaldehyde
POM, Group la
Naphthalene
Ethylbenzene
1,3 -Butadiene
Tetrachloroethylene
POM, Group 2d
POM, Group 2b
50.13
40.75
25.97
20.37
16.70
15.75
6.55
4.75
4.19
2.77
POM, Group la
Formaldehyde
Naphthalene
POM, Group 2d
Benzene
Hexavalent Chromium, PM
Nickel, PM
POM, Group 2b
1,3 -Butadiene
POM, Group 3
1.79E-03
6.52E-04
5.68E-04
3.68E-04
3.18E-04
2.89E-04
2.70E-04
2.44E-04
1.96E-04
6.53E-05
Arsenic
4.00
OJ
oo
-------
Table 13-8. 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)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
Atmos Energy, Calvert City, Kentucky (Marshall County) - ATKY
Benzene
Ethylbenzene
Vinyl chloride
Acetaldehyde
Formaldehyde
1,2-Dichloroethane
1,3 -Butadiene
Naphthalene
POM, Group la
Carbon tetrachloride
61.70
37.84
30.93
26.73
23.00
9.25
7.11
2.78
2.66
2.32
Benzene
Formaldehyde
Vinyl chloride
Hexavalent Chromium, PM
1 ,2-Dichloroethane
POM, Group la
1,3 -Butadiene
Naphthalene
Ethylbenzene
Nickel, PM
4.81E-04
2.99E-04
2.72E-04
2.64E-04
2.41E-04
2.34E-04
2.13E-04
9.47E-05
9.46E-05
7.93E-05
Calvert City Elementary, Calvert City, Kentucky (Marshall County) - CCKY
Benzene
Ethylbenzene
Vinyl chloride
Acetaldehyde
Formaldehyde
1 ,2-Dichloroethane
1,3 -Butadiene
Naphthalene
POM, Group la
Carbon tetrachloride
61.70
37.84
30.93
26.73
23.00
9.25
7.11
2.78
2.66
2.32
Benzene
Formaldehyde
Vinyl chloride
Hexavalent Chromium, PM
1 ,2-Dichloroethane
POM, Group la
1,3 -Butadiene
Naphthalene
Ethylbenzene
Nickel, PM
4.81E-04
2.99E-04
2.72E-04
2.64E-04
2.41E-04
2.34E-04
2.13E-04
9.47E-05
9.46E-05
7.93E-05
Arsenic 3.68
OJ
oo
-------
Table 13-8. 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)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
Lazy Daze, Calvert City, Kentucky (Marshall County) - LAKY
Benzene
Ethylbenzene
Vinyl chloride
Acetaldehyde
Formaldehyde
1,2-Dichloroethane
1,3 -Butadiene
Naphthalene
POM, Group la
Carbon tetrachloride
61.70
37.84
30.93
26.73
23.00
9.25
7.11
2.78
2.66
2.32
Benzene
Formaldehyde
Vinyl chloride
Hexavalent Chromium, PM
1 ,2-Dichloroethane
POM, Group la
1,3 -Butadiene
Naphthalene
Ethylbenzene
Nickel, PM
4.81E-04
2.99E-04
2.72E-04
2.64E-04
2.41E-04
2.34E-04
2.13E-04
9.47E-05
9.46E-05
7.93E-05
TVA Substation, Calvert City, Kentucky (Marshall County) - TVKY
Benzene
Ethylbenzene
Vinyl chloride
Acetaldehyde
Formaldehyde
1 ,2-Dichloroethane
1,3 -Butadiene
Naphthalene
POM, Group la
Carbon tetrachloride
61.70
37.84
30.93
26.73
23.00
9.25
7.11
2.78
2.66
2.32
Benzene
Formaldehyde
Vinyl chloride
Hexavalent Chromium, PM
1 ,2-Dichloroethane
POM, Group la
1,3 -Butadiene
Naphthalene
Ethylbenzene
Nickel, PM
4.81E-04
2.99E-04
2.72E-04
2.64E-04
2.41E-04
2.34E-04
2.13E-04
9.47E-05
9.46E-05
7.93E-05
OJ
oo
-------
Table 13-8. 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)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
Smithland, Kentucky (Livingston County) - BLKY
Benzene
Formaldehyde
Acetaldehyde
Ethylbenzene
1,3 -Butadiene
Naphthalene
POM, Group 2b
POM, Group 2d
POM, Group 6
Nickel, PM
12.70
12.00
6.63
4.46
1.67
0.70
0.17
0.16
0.03
0.03
Formaldehyde
Benzene
1,3 -Butadiene
Naphthalene
POM, Group 3
POM, Group 2b
Acetaldehyde
POM, Group 2d
Nickel, PM
Ethylbenzene
1.56E-04
9.90E-05
5.02E-05
2.39E-05
1.68E-05
1.51E-05
1.46E-05
1.43E-05
1.26E-05
1.11E-05
Lexington, Kentucky (Fayette County) - LEKY
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
POM, Group 2b
Tetrachloroethylene
Trichloroethylene
POM, Group 2d
120.89
91.00
74.64
51.93
17.21
10.50
2.38
2.24
1.94
1.64
Formaldehyde
Benzene
1,3 -Butadiene
POM, Group 3
Naphthalene
POM, Group 2b
Ethylbenzene
Hexavalent Chromium, PM
POM, Group 2d
Arsenic, PM
1.18E-03
9.43E-04
5.16E-04
4.28E-04
3.57E-04
2.09E-04
1.87E-04
1.84E-04
1.44E-04
1.27E-04
Arsenic 3.94
OJ
oo
-------
Table 13-9. 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)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity Weight
Top 10 Noncancer Hazard
Approximations Based on Annual
Average Concentrations
(Site-Specific)
Noncancer
Hazard
Approximation
Pollutant (HQ)
Health Department, Ashland, Kentucky (Boyd County) - ASKY
Toluene
Ethylene glycol
Benzene
Hexane
Xylenes
Methanol
Hydrochloric acid
Formaldehyde
Acetaldehyde
Ethylbenzene
128.66
60.98
55.37
45.73
40.30
39.10
27.65
20.84
11.02
10.16
Manganese, PM
Acrolein
Chlorine
Nickel, PM
Lead, PM
Cadmium, PM
Formaldehyde
Benzene
1,3 -Butadiene
Hydrochloric acid
203,096.87
66,362.90
45,169.74
15,550.33
11,227.98
3,311.75
2,126.16
1,845.80
1,444.19
1,382.51
21s* and Greenup, Ashland, Kentucky (Boyd County) - ASKY-M
Toluene
Ethylene glycol
Benzene
Hexane
Xylenes
Methanol
Hydrochloric acid
Formaldehyde
Acetaldehyde
Ethylbenzene
128.66
60.98
55.37
45.73
40.30
39.10
27.65
20.84
11.02
10.16
Manganese, PM
Acrolein
Chlorine
Nickel, PM
Lead, PM
Cadmium, PM
Formaldehyde
Benzene
1,3 -Butadiene
Hydrochloric acid
203,096.87
66,362.90
45,169.74
15,550.33
11,227.98
3,311.75
2,126.16
1,845.80
1,444.19
1,382.51
Manganese 0.68
Arsenic 0.12
Lead 0.10
Cadmium 0.06
Nickel 0.03
OJ
oo
-------
Table 13-9. 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)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity Weight
Top 10 Noncancer Hazard
Approximations Based on Annual
Average Concentrations
(Site-Specific)
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Grayson, Kentucky (Carter County) - GLKY
Toluene
Ethylene glycol
Xylenes
Hexane
Formaldehyde
Benzene
Methanol
Acetaldehyde
Ethylbenzene
Naphthalene
75.43
34.12
26.97
23.33
22.49
16.98
15.68
11.95
6.80
2.47
Acrolein
Formaldehyde
Acetaldehyde
Cyanide Compounds, gas
1,3 -Butadiene
Naphthalene
Benzene
Xylenes
Propionaldehyde
Arsenic, PM
74,382.78
2,295.36
1,328.25
1,278.36
1,021.17
823.24
566.08
269.67
118.11
91.35
Formaldehyde
Acetaldehyde
Manganese
Arsenic
1,3 -Butadiene
Benzene
Carbon Tetrachloride
1 ,2-Dichloroethane
0.17
0.09
0.07
0.04
0.03
0.02
0.01
0.01
Baskett, Kentucky (Henderson County) - BAKY
Toluene
Carbonyl sulfide
Xylenes
Ethylene glycol
Hexane
Formaldehyde
Benzene
Methanol
Acetaldehyde
Naphthalene
161.88
128.78
77.22
56.92
54.77
50.13
40.75
28.37
25.97
16.70
Acrolein
Manganese, PM
Nickel, PM
Naphthalene
Formaldehyde
1,3 -Butadiene
Chlorine
Acetaldehyde
Cadmium, PM
Methylenediphenyl diisocyanate, 4,4'- (MDI), gas
65,506.85
43,233.90
6,258.97
5,566.87
5,115.54
3,274.37
3,245.93
2,885.41
2,792.02
2,483.57
Manganese
Arsenic
0.13
0.06
oo
oo
-------
Table 13-9. 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)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity Weight
Top 10 Noncancer Hazard
Approximations Based on Annual
Average Concentrations
(Site-Specific)
Noncancer
Hazard
Approximation
Pollutant (HQ)
Atmos Energy, Calvert City, Kentucky (Marshall County) - ATKY
Methanol
Toluene
Xylenes
Hydrochloric acid
Vinyl acetate
Benzene
Hexane
Ethylene glycol
Ethylbenzene
Chlorine
677.58
205.25
183.68
83.48
73.28
61.70
45.87
39.87
37.84
31.62
Chlorine
Acrolein
Manganese, PM
Hydrochloric acid
1,3 -Butadiene
Acetaldehyde
Acrylic acid
Formaldehyde
Benzene
Xylenes
210,803.93
77,112.59
5,023.60
4,173.99
3,557.33
2,970.22
2,916.21
2,346.51
2,056.78
1,836.76
Calvert City Elementary, Calvert City, Kentucky (Marshall County) - CCKY
Methanol
Toluene
Xylenes
Hydrochloric acid
Vinyl acetate
Benzene
Hexane
Ethylene glycol
Ethylbenzene
Chlorine
677.58
205.25
183.68
83.48
73.28
61.70
45.87
39.87
37.84
31.62
Chlorine
Acrolein
Manganese, PM
Hydrochloric acid
1,3 -Butadiene
Acetaldehyde
Acrylic acid
Formaldehyde
Benzene
Xylenes
210,803.93
77,112.59
5,023.60
4,173.99
3,557.33
2,970.22
2,916.21
2,346.51
2,056.78
1,836.76
Manganese 0.13
Arsenic 0.06
oo
VO
-------
Table 13-9. 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)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity Weight
Top 10 Noncancer Hazard
Approximations Based on Annual
Average Concentrations
(Site-Specific)
Noncancer
Hazard
Approximation
Pollutant (HQ)
Lazy Daze, Calvert City, Kentucky (Marshall County) - LAKY
Methanol
Toluene
Xylenes
Hydrochloric acid
Vinyl acetate
Benzene
Hexane
Ethylene glycol
Ethylbenzene
Chlorine
677.58
205.25
183.68
83.48
73.28
61.70
45.87
39.87
37.84
31.62
Chlorine
Acrolein
Manganese, PM
Hydrochloric acid
1,3 -Butadiene
Acetaldehyde
Acrylic acid
Formaldehyde
Benzene
Xylenes
210,803.93
77,112.59
5,023.60
4,173.99
3,557.33
2,970.22
2,916.21
2,346.51
2,056.78
1,836.76
TVA Substation, Calvert City, Kentucky (Marshall County) - TVKY
Methanol
Toluene
Xylenes
Hydrochloric acid
Vinyl acetate
Benzene
Hexane
Ethylene glycol
Ethylbenzene
Chlorine
677.58
205.25
183.68
83.48
73.28
61.70
45.87
39.87
37.84
31.62
Chlorine
Acrolein
Manganese, PM
Hydrochloric acid
1,3 -Butadiene
Acetaldehyde
Acrylic acid
Formaldehyde
Benzene
Xylenes
210,803.93
77,112.59
5,023.60
4,173.99
3,557.33
2,970.22
2,916.21
2,346.51
2,056.78
1,836.76
VO
o
-------
Table 13-9. 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)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity Weight
Top 10 Noncancer Hazard
Approximations Based on Annual
Average Concentrations
(Site-Specific)
Noncancer
Hazard
Approximation
Pollutant (HQ)
Smithland, Kentucky (Livingston County) - BLKY
Toluene
Xylenes
Benzene
Formaldehyde
Ethylene glycol
Hexane
Acetaldehyde
Methanol
Ethylbenzene
1,3 -Butadiene
48.42
35.09
12.70
12.00
11.72
11.09
6.63
5.38
4.46
1.67
Acrolein
Formaldehyde
Manganese, PM
1,3 -Butadiene
Acetaldehyde
Cyanide Compounds, gas
Benzene
Xylenes
Nickel, PM
Naphthalene
21,504.21
1,224.76
1,211.46
836.35
736.97
527.46
423.20
350.93
291.64
234.01
Lexington, Kentucky (Fayette County) - LEKY
Toluene
Ethylene glycol
Xylenes
Hexane
Methanol
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
Methyl isobutyl ketone
777.20
364.42
288.28
237.91
176.71
120.89
91.00
74.64
51.93
29.90
Acrolein
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Benzene
Naphthalene
Xylenes
Hexamethylene- 1 ,6-diisocyanate, gas
Arsenic, PM
Methylenediphenyl diisocyanate, 4,4'- (MDI), gas
261,778.93
9,286.02
8,604.26
5,769.93
4,029.83
3,500.50
2,882.83
2,051.30
1,974.60
1,757.48
Manganese 0.13
Arsenic 0.06
-------
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 3.5.3.5. Similar to the cancer risk and
noncancer hazard approximations provided in Section 13.5.2, this analysis may help policy-
makers prioritize their air monitoring activities.
Observations from Table 13-8 include the following:
• Among the Kentucky counties with monitoring sites, emissions (for pollutants with
cancer UREs) are highest in Fayette County (LEKY) and least in Livingston County
(BLKY).
• Benzene, formaldehyde, and POM, Group la are the highest emitted pollutants with
cancer UREs in Boyd County, where the Ashland sites are located. Coke oven
emissions, POM Group la, and hexavalent chromium are the pollutants with the
highest toxicity-weighted emissions (of the pollutants with cancer UREs) for Boyd
County. Eight of the highest emitted pollutants also have the highest toxi city-
weighted emissions for Boyd County.
• Cancer risk approximations could be calculated for arsenic, nickel, and cadmium for
ASKY-M. Although arsenic has the highest cancer risk approximation for ASKY-M,
this pollutant appears on neither emissions-based list (arsenic ranks 26th for total
emissions and 17th for toxi city-weighted emissions). This is also true for cadmium.
Conversely, nickel appears on both emissions based lists, ranking 10th for total
emissions and fourth for toxicity-weighted emissions.
• Formaldehyde, benzene, and acetaldehyde are the highest emitted pollutants with
cancer UREs in Carter County, where GLKY is located. Formaldehyde, benzene, and
naphthalene are the pollutants with the highest toxicity-weighted emissions (of the
pollutants with cancer UREs) for this county. Eight of the highest emitted pollutants
also have the highest toxicity-weighted emissions for Carter County.
• Formaldehyde has the highest cancer risk approximation for GLKY, and ranks first
on all three lists in Table 13-8. 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, 3, and 5a). Many
13-92
-------
of the PAHs sampled using Method TO-13 are part of POM, Groups 2b, 2d, 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. POM, Group la,
formaldehyde, and naphthalene 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 for which a cancer risk
approximation could be calculated. Arsenic appears on neither emissions-based list
for Henderson County (arsenic ranks 23th for total emissions and 16th for toxicity-
weighted emissions). Several POM Groups appear on the emissions-based lists for
Henderson County, but PAHs were not sampled at BAKY.
• Benzene, ethylbenzene, and vinyl chloride are the highest emitted pollutants with
cancer UREs in Marshall County, where four of the five Calvert City sites are
located. Marshall County is the only county with NMP sites for which vinyl chloride
appears among the highest emitted pollutants. The quantity of vinyl chloride emitted
in Marshall County (31 tpy) is the highest emissions for this pollutant among NMP
counties and is twice the quantity of the next highest emissions (16 tpy in Harris
County, Texas). This is also true for carbon tetrachloride. There are only three
counties with NMP sites that have carbon tetrachloride emissions greater than 1 tpy,
Marshall County, Kentucky (2.32 tpy), Harris County, Texas (1.25 tpy), and Harrison
County, Texas (1.06 tpy). Marshall County is also the only county with NMP sites for
which 1,2-dichloroethane appears among the highest emitted pollutants. The quantity
of 1,2-dichloroethane emitted in Marshall County (9.25 tpy) is the second highest
emissions for this pollutant among NMP sites, behind only Harris County, Texas
(16 tpy).
• Benzene, formaldehyde, and vinyl chloride are the pollutants with the highest
toxicity-weighted emissions (of the pollutants with cancer UREs) for Marshall
County. Marshall County is the only county for which vinyl chloride and
1,2-dichloroethane appear among the pollutants with the highest toxicity-weighted
emissions. Eight of the highest emitted pollutants also have the highest toxicity-
weighted emissions for Marshall County.
• Arsenic is the only pollutant of interest for CCKY for which a cancer risk
approximation could be calculated. Arsenic appears on neither emissions-based list
for Marshall County (arsenic ranks 27th for total emissions and 16th for toxicity-
weighted emissions).
• Benzene, formaldehyde, and acetaldehyde are the highest emitted pollutants with
cancer UREs in Livingston County, where BLKY is located. Benzene, formaldehyde,
and 1,3-butadiene 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 Livingston County.
13-93
-------
Several POM Groups appear on the emissions-based lists for Livingston County, but
PAHs were not sampled at BLKY.
• 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. Seven of the highest emitted
pollutants also have the highest toxicity-weighted emissions for Fayette County.
Several POM Groups appear on the emissions-based lists for Fayette County, but
PAHs were not sampled at LEKY.
• Arsenic is the only pollutant of interest for LEKY for which a cancer risk
approximation could be calculated. Arsenic has the 10th highest toxicity-weighted
emissions but does not appear among the highest emitted for Fayette County (arsenic
ranks 23rd for total emissions).
Observations from Table 13-9 include the following:
• Among the Kentucky counties with monitoring sites, emissions (for pollutants with
noncancer RfCs) are highest in Fayette County (LEKY) and least in Livingston
County (BLKY).
• Toluene, ethylene glycol, and benzene are the highest emitted pollutants with
noncancer RfCs in Boyd County. Manganese, acrolein, and chlorine are the pollutants
with the highest toxicity-weighted emissions (of the pollutants with noncancer RfCs)
for Boyd County. Three of the highest emitted pollutants also have the highest
toxicity-weighted emissions for Boyd County.
• Nonancer hazard approximations could be calculated for all five metal pollutants of
interest for ASKY-M. Manganese, which has the highest nonancer hazard
approximation, also has the highest toxicity-weighted emissions. Nickel, lead, and
cadmium are also among the pollutants with the highest toxicity-weighted emissions.
None of the metal pollutants of interest for ASKY-M are among the highest emitted
in Boyd County.
• Toluene, ethylene glycol, and xylenes are the highest emitted pollutants with
noncancer RfCs in Carter County. Acrolein, formaldehyde, and acetaldehyde are the
pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) 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. Five of the highest emitted
pollutants also have the highest toxicity-weighted emissions for Carter County.
• Formaldehyde and acetaldehyde have the highest nonancer hazard approximations for
GLKY and appear on both emissions-based lists. Benzene also appears on all three
lists. Arsenic and 1,3-butadiene are among the pollutants with the highest toxicity-
13-94
-------
weighted emissions but are not among the highest emitted in Carter County.
Manganese, which has the third highest nonancer hazard approximation, appears on
neither emissions-based list. This is also true for carbon tetrachloride and
1,2-dichloroethane.
• Toluene, carbonyl sulfide, and xylenes are the highest emitted pollutants with
noncancer RfCs in Henderson County. Acrolein, manganese, and nickel are the
pollutants with the highest toxi city-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.
• Manganese, which has the highest nonancer hazard approximation for BAKY, has the
second highest toxicity-weighted emissions but is not among the highest emitted.
Arsenic, the only other pollutant of interest for BAKY, appears on neither emissions-
based list.
• Methanol, toluene, and xylenes are the highest emitted pollutants with noncancer
RfCs in Marshall County. Chlorine, acrolein, and manganese 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. Four of the highest emitted
pollutants also have the highest toxicity-weighted emissions for Marshall County.
• Manganese, which has the highest nonancer hazard approximation for CCKY, has the
third highest toxicity-weighted emissions but is not among the highest emitted.
Arsenic, the only other pollutant of interest for CCKY for which a noncancer hazard
approximation could be calculated, appears on neither emissions-based list.
• Toluene, xylenes, and benzene are the highest emitted pollutants with noncancer
RfCs in Livingston County. Acrolein, formaldehyde, and manganese 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.
• Toluene, ethylene glycol, and xylenes 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.
• Manganese, which has the highest nonancer hazard approximation for LEKY, appears
on neither emissions-based list for Fayette County. Arsenic, the only other pollutant
of interest for LEKY for which a noncancer hazard approximation could be
calculated, has the ninth highest toxicity-weighted emissions but does not appear
among the highest emitted.
13-95
-------
13.6 Summary of the 2012 Monitoring Data for the Kentucky Monitoring Sites
Results from several of the data treatments described in this section include the
following:
<• Four monitoring sites (ASKY-M, BAKY, CCKY, andLEKY) began sampling PM10
metals under the NMP in March 2012. Seven monitoring sites began sampling VOCs
in July. Two monitoring sites (ASKY and LEKY) also began sampling car bony I
compounds in July. GLKY sampled VOCs, PAHs, car bony I compounds, PM10 metals
and hexavalent chromium year-round.
»«» The number of pollutants failing screens for the Kentucky sites varies from three
(BAKY) to 15 (LEKY).
»«» Because the start dates for sampling were staggered, annual average concentrations
could only be calculated for GLKY and those sites sampling PMw metals.
*»* ASKY-M had the highest annual average concentrations of arsenic, manganese, and
nickel among NMP sites sampling PM10 metals. Four additional Kentucky sites were
among the sites with the highest annual average concentrations of arsenic; three
additional Kentucky sites were among those with the highest annual average
concentrations of manganese; LEKY was also among the sites with the highest annual
average concentrations of nickel.
»«» Some of the highest concentrations of VOCs were measured at the Calvert City sites,
particularly vinyl chloride, carbon tetrachloride, and 1,2-dichloroethane.
13-96
-------
14.0 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.
14.1 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 14-1 is a 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. 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 radius
are still visible on the map, but have been grayed out in order to show emissions sources just
outside the boundary. Table 14-1 provides supplemental geographical information such as land
use, location setting, and locational coordinates.
14-1
-------
Figure 14-1. Boston, Massachusetts (BOMA) Monitoring Site
-------
Figure 14-2. NEI Point Sources Located Within 10 Miles of BOMA
Middlesex \
\ County \
\ --€
« - - *
71-25'O-W 71'20'0-W
Leqend
BOMA NATTS site
71"15'0"W 71'10'0'W 71'5'CTW 7111010I1W
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
Source Category Group (No. of Facilities)
•J< Aerospace/Aircraft Manufacturing (1)
*t" Airport/Airline/Airport Support Operations (18)
i Asphalt Production/Hot Mix Asphalt Plant (2)
W Automobile/Truck Manufacturing (1)
B Bulk Terminals/Bulk Plants (9)
C Chemical Manufacturing (3)
CX] Crematory - Animal/Human (1)
ft) Dry Cleaning (2)
6 Electrical Equipment Manufacturing (1)
f Electricity Generation via Combustion (8)
E Electroplating, Plating, Polishing. Anodizing, and Coloring (4)
F Food Processing/Agriculture (3)
I Foundries, Iron and Steel (1)
> Hotels/Motels/Uodging (1)
-^ Industrial Machinery or Equipment Plant (1)
O Institution (school, hospital, prison, etc.) (43)
A Metal Coating, Engraving, and Allied Services to Manufacturers (1)
County boundary
Metals Processing/Fabrication (2)
Military Base/National Security (1)
Mine/Quarry/Mineral Processing (2)
Miscellaneous Commercial/Industrial (38)
Municipal Waste Combustor (1)
Oil and/or Gas Production (2)
Paint and Coating Manufacturing (1)
Pharmaceutical Manufacturing (1)
Plastic, Resin, or Rubber Products Plant (1)
Printing/Publishing/Paper Product Manufactunng (1)
Rail Yard/Rail Line Operations (1)
Ship/Boat Manufacturing or Repair (1)
Telecommunications/Radio (2)
Testing Laboratories (1)
Truck/Bus/Transportation Operations (3)
Wastewater Treatment (3)
Water Treatment (1)
14-3
-------
Table 14-1. Geographical Information for the Massachusetts Monitoring Site
Site
Code
BOMA
AQS Code
25-025-0042
Location
Boston
County
Suffolk
Micro- or
Metropolitan
Statistical Area
Boston-
Cambridge-
Newton, MA-NH
AyTO A
IVloA
Latitude
and
Longitude
42.3295,
-71.0826
Land Use
Commercial
Location
Setting
Urban/City
Center
Additional Ambient Monitoring Information1
CO, VOCs, SO2, NO, NO2, NOX, NOy,
PAMS/NMOCs, Carbonyl compounds, O3,
Meteorological parameters, PM10, Black carbon, PM
coarse, PM2.5, PM25 Speciation, IMPROVE
Speciation.
BOLD ITALICS = EPA-designated NATTS Site
-------
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 14-1. The monitoring site is 1.25 miles south of 1-90 and
1 mile west of 1-93. The original purpose for the location of this site was to measure population
exposure to a city bus terminal located across the street from the monitoring site. In recent years,
the buses servicing the area were converted to compressed natural gas (CNG). As Figure 14-2
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 institution 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; electricity generating units (via
combustion); and bulk terminals and bulk plants.
Table 14-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Massachusetts monitoring site. Table 14-2 includes both county-
level population and vehicle registration information. Table 14-2 also contains traffic volume
information for BOMA as well as the location for which the traffic volume was obtained.
Additionally, Table 14-2 presents the county-level daily VMT for Suffolk County.
Table 14-2. Population, Motor Vehicle, and Traffic Information for the Massachusetts
Monitoring Site
Site
BOMA
Estimated
County
Population1
744,426
County-level
Vehicle
Registration2
362,899
Annual
Average
Daily Traffic3
27,654
Intersection
Used for
Traffic Data
Melnea Cass Blvd, near Shawmut Ave
County-
level Daily
VMT4
10,890,178
Bounty-level population estimate reflects 2012 data (Census Bureau, 2013c)
2County-level vehicle registration reflects 2012 data (MA RMV, 2013)
3AADT reflects 2010 data (MA DOT, 2010)
4County-level VMT reflects 2012 data (MA DOT, 2013)
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 14-2 include the following:
• The Suffolk County population is in the middle of the range, ranking 18th among
other counties with NMP sites.
• The Suffolk County vehicle registration is also in the middle of the range, ranking
24th among other counties with NMP sites.
14-5
-------
• The traffic volume experienced near BOMA is in the middle of the range compared to
other NMP sites. The traffic estimate provided is for Melnea Cass Boulevard near
Shawmut Avenue.
• The daily VMT for Suffolk County is also in the middle of the range compared to
other counties with NMP sites (where VMT data were available). The VMT for
Suffolk County ranks 23rd.
14.2 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.
14.2.1 Climate Summary
Boston's New England location ensures that the city experiences a fairly active weather
pattern. Storm systems frequently track across the region, bringing ample precipitation to the
area. The proximity to the Atlantic Ocean helps moderate temperatures, both in the summer and
the winter, while at the same time allowing winds to gust higher than they would farther inland.
Winds generally flow from the northwest in the winter and southwest in the summer. Coastal
storm systems called "Nor'easters," strong low pressure systems that produce heavy rain or snow
and winds up to hurricane strength along the Mid-Atlantic and northeast coastal states, often
produce the heaviest snowfalls for the area. This coastal location may also be affected by tropical
systems, approximately one every 5 years on average (Wood, 2004; NCDC, 2014).
14.2.2 Meteorological Summary
Hourly meteorological data for 2012 were retrieved from NCDC for the weather station
closest to the Massachusetts monitoring site (NCDC, 2012), as described in Section 3.5.2. The
closest weather station to BOMA is located at Logan International Airport (WBAN 14739).
Additional information about the Logan Airport weather station, such as the distance between the
site and the weather station, is provided in Table 14-3. These data were used to determine how
meteorological conditions on sample days vary from conditions experienced throughout the year.
14-6
-------
Table 14-3. Average Meteorological Conditions near the Massachusetts Monitoring Site
Closest Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Boston, Massachusetts - BOMA
Logan
International
Airport
14739
(42.36, -71.01)
4.1
Miles
42°
(NE)
Sample
Day
(74)
2012
58.8
±3.9
60.8
±1.7
52.2
±3.7
53.9
±1.6
40.2
±4.1
41.5
±1.8
46.8
±3.4
48.1
±1.5
66.3
±3.7
65.8
±1.6
1015.5
±1.8
1015.3
±0.8
8.6
±0.7
8.8
±0.3
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
-------
Table 14-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for all of 2012. Also included in Table 14-3 is the 95 percent
confidence interval for each parameter. As shown in Table 14-3, temperatures on sample days at
BOMA appear slightly cooler than temperatures experienced throughout 2012, although the
differences are not significant. This is due to the number of make-up sample days. The majority
of make-up samples were collected during the colder months of the year, specifically in January
or February or between October and December.
14.2.3 Back Trajectory Analysis
Figure 14-3 is the composite back trajectory map for days on which samples were
collected at the BOMA monitoring site. Included in Figure 14-3 are four back trajectories per
sample day. Figure 14-4 is the corresponding cluster analysis. An in-depth description of these
maps and how they were generated is presented in Section 3.5.2.1. For the composite map, each
line represents the 24-hour trajectory along which a parcel of air traveled toward the monitoring
site on a given sample day and time, based on an initial height of 50 meters AGL. For the cluster
analysis, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric circle around the site in Figures 14-3 and 14-4 represents 100 miles.
Observations from Figures 14-3 and 14-4 include the following:
• The composite back trajectory map shows that back trajectories originated from a
variety of directions at BOMA. The predominant direction of back trajectory origin is
from the northwest quadrant, with the longest trajectories originating offshore.
• The size of the 24-hour air shed domain for BOMA is in the upper end of the range
compared to other NMP sites. The farthest away a back trajectory originated was
approximately 800 miles, well offshore of the southeast coast. This back trajectory
and the others originating over eastern North Carolina that appear to spiral in towards
the site are those associated with the October 30, 2012 sample day, as Hurricane
Sandy moved onshore. The average back trajectory length was 255 miles with the
majority of back trajectories (86 percent) originating within 400 miles of the
monitoring site.
14-8
-------
• More than half of back trajectories originated to the west, northwest, and north of
BOMA, but are split into four cluster trajectories in Figure 14-4. One-quarter of back
trajectories originated over Maine, Vermont, New Hampshire, northeast New York,
and along the U.S./Canada border; another 10 percent originated over southeast
Ontario and Quebec, Canada; 5 percent of these back trajectories reach as far as Lake
Huron and central Ontario; and the 13 percent of back trajectories originating to the
west include shorter trajectories originating over New York and Pennsylvania, as well
as longer ones originating as far west as Indiana. One-third of back trajectories
originated to the southwest of the site, but are split into two back trajectories: those
relatively short in length (< 200 miles) and those originating toward the Delmarva
Peninsula, Virginia, and North Carolina. Six percent of back trajectories originated to
the south of the site over the Atlantic Ocean and another 9 percent originated to the
east of BOMA over the Gulf of Maine.
Figure 14-3. Composite Back Trajectory Map for BOMA
14-9
-------
Figure 14-4. Back Trajectory Cluster Map for BOMA
i"--, f ?i\
14.2.4 Wind Rose Comparison
Hourly surface wind data from the weather station at Logan International Airport near
BOMA were uploaded into a wind rose software program to produce customized wind roses, as
described in Section 3.5.2.2. 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 14-5 presents a map showing the distance between the weather station and
BOMA, which may be useful for identifying topographical influences that may affect the
meteorological patterns experienced at this location. Figure 14-5 also presents three different
wind roses for the BOMA monitoring site. First, a historical wind rose representing 2002 to 2011
wind data is presented, which shows the predominant surface wind speed and direction over an
extended period of time. Second, a wind rose representing wind observations for all of 2012 is
presented. Next, a wind rose representing wind data for days on which samples were collected in
2012 is presented. These can be used to identify the predominant wind speed and direction for
2012 and to determine if wind observations on sample days were representative of conditions
experienced over the entire year and historically.
14-10
-------
Figure 14-5. Wind Roses for the Logan International Airport Weather Station near BOMA
Location of BOMA and Weather Station
2002-2011 Historical Wind Rose
Calms: 3.26%
2012 Wind Rose
Sample Day Wind Rose
Calms: 4.50%
14-11
-------
Observations from Figure 14-5 for BOMA include the following:
• The Logan International Airport weather station is located approximately 4 miles
northeast of BOMA. Note that the airport is located on a peninsula in Boston Harbor
with downtown Boston to the west, Chelsea to the north, and Winthrop to the east,
while the BOMA monitoring site is located west of South Boston and farther inland
(less than 2 miles from the nearest coastline).
• The historical wind rose shows that calm winds (< 2 knots) account for only 3 percent
of wind observations. Winds with a westerly component (south-southwest to north-
northwest) make up the majority (nearly 60 percent) of winds greater than 2 knots.
• The wind patterns shown on the 2012 wind rose resemble the historical wind patterns,
indicating that wind conditions during 2012 were typical of conditions experienced
historically near BOMA.
• The sample day wind patterns resemble the full-year and historical wind patterns,
indicating that wind conditions on sample days were representative of those
experienced over the entire year and historically.
14.3 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 14-4.
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 14-4. It is
important to note which pollutants were sampled for at each site when reviewing the results of
this analysis. BOMA sampled for PMio metals, PAHs, and hexavalent chromium.
14-12
-------
Table 14-4. Risk-Based Screening Results for the Massachusetts Monitoring Site
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Boston, Massachusetts - BOMA
Arsenic (PM10)
Naphthalene
Manganese (PM10)
Hexavalent Chromium
Nickel (PM10)
Acenaphthene
Cadmium (PM10)
Fluoranthene
Fluorene
0.00023
0.029
0.005
0.000083
0.0021
0.011
0.00056
0.011
0.011
Total
57
54
14
6
5
1
1
1
1
140
61
59
61
50
61
59
61
59
59
530
93.44
91.53
22.95
12.00
8.20
1.69
1.64
1.69
1.69
26.42
40.71
38.57
10.00
4.29
3.57
0.71
0.71
0.71
0.71
40.71
79.29
89.29
93.57
97.14
97.86
98.57
99.29
100.00
Observations from Table 14-4 include the following:
• Nine pollutants failed at least one screen for BOMA; 26 percent of concentrations for
these nine pollutants were greater than their associated risk screening value (or failed
screens).
• Five pollutants contributed to 95 percent of failed screens for BOMA and therefore
were identified as pollutants of interest for this site. These include three PMi0 metals,
one PAH (naphthalene), and hexavalent chromium.
• Arsenic and naphthalene each account for roughly 40 percent of the total failed
screens for BOMA.
14.4 Concentrations
This section presents various concentration averages used to characterize pollution 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 the
monitoring site.
• Annual concentration averages are presented graphically 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 the site.
14-13
-------
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Additional site-specific statistical summaries for BOMA
are provided in Appendices M through O.
14.4.1 2012 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 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 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 BOMA are presented in Table 14-5,
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 14-5. Quarterly and Annual Average Concentrations of the Pollutants of
Interest for the Massachusetts Monitoring Site
Pollutant
#of
Measured
Detections
vs. # 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)
Boston, Massachusetts - BOMA
Arsenic (PM10)
Hexavalent Chromium
Manganese (PM10)
Naphthalene
Nickel (PM10)
61/61
50/61
61/61
59/59
61/61
0.56
±0.18
0.01
±0.01
3.73
±0.94
56.23
± 12.80
1.60
±0.38
0.43
±0.14
0.06
±0.04
3.63
±0.76
45.63
±9.63
1.65
±1.02
0.51
±0.07
0.03
±0.01
4.33
±0.97
65.47
±11.59
1.19
±0.35
0.54
±0.17
0.04
±0.04
4.52
±1.49
84.48
±25.40
1.21
±0.41
0.51
±0.07
0.03
±0.02
4.06
±0.51
63.55
±9.09
1.41
±0.29
14-14
-------
Observations for BOMA from Table 14-5 include the following:
• Naphthalene is the pollutant with the highest annual average concentration
(63.55 ± 9.09 ng/m3). The annual average concentrations for the remaining pollutants
of interest are at least an order of magnitude lower. Of the PMio metals, manganese is
the pollutant with the highest annual average concentration (4.06 ± 0.51 ng/m3).
• The fourth quarter concentration of naphthalene is higher than the other quarterly
averages and has a higher confidence interval associated with it than the others. A
review of the data shows that the maximum concentration of naphthalene was
measured at BOMA on November 17, 2012 (235 ng/m3). The second and third
highest measurements are lower but were also measured during the fourth quarter
(152 ng/m3 and 130 ng/m3). Of the 13 concentrations greater than 75 ng/m3 measured
at BOMA, nine were measured during the fourth quarter.
• The concentrations of manganese measured at BOMA span an order of magnitude,
ranging from 1.44 ng/m3 to 10.8 ng/m3. The fourth quarter average manganese
concentration is higher than the other quarterly averages and has a larger confidence
interval than the others. A review of the data shows that the maximum concentration
of manganese was measured at BOMA on November 29, 2012. Of the 10 highest
manganese concentrations measured at BOMA, six were measured during the fourth
quarter.
• The second quarter average concentration of nickel has a relatively large confidence
interval associated with it. The maximum nickel concentration was measured at
BOMA on May 21, 2012 (8.43 ng/m3). The next highest concentration measured
during the second quarter is much less (2.10 ng/m3). Nickel concentrations measured
during the second quarter range from 0.599 ng/m3 to 8.43 ng/m3, with a median
concentration of 1.23 ng/m3. The maximum nickel concentration measured at BOMA
is the sixth highest nickel concentrations measured across the program (PMio only).
Further, BOMA has the fifth highest annual average concentration among NMP sites
sampling PMio metals, as shown in Table 4-12.
• The confidence intervals calculated for the quarterly average concentrations of
hexavalent chromium for BOMA indicate that the concentrations are highly variable.
Measured detections of hexavalent chromium span two orders of magnitude, ranging
from 0.0033 ng/m3 to 0.314 ng/m3, although several non-detects were also reported.
The median concentration for BOMA is 0.018 ng/m3. Note that all but one of the 11
non-detects were reported for the colder months of the year (January through March
or November and December).
14-15
-------
14.4.2 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 shaded in
gray in Table 14-4 for BOMA. Figures 14-6 through 14-10 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 3.5.3.
Figure 14-6. Program vs. Site-Specific Average Arsenic (PMi0) Concentration
BOMA
. o
1°
1 1
312345678
Concentration (ng/m3)
Program:
Site:
1st Quartile
D
Site Average
0
2nd Quartile 3rd Quartile
D D
Site Concentration Range
^^^^—
4th Quartile Average
D 1
Figure 14-7. Program vs. Site-Specific Average Hexavalent Chromium Concentration
BOMA
Program Max Concentration = 8.51 ng/m3
0.1
0.2 0.3
Concentration (ng/m3)
0.4
Program:
Site:
1st Quartile
D
Site Average
o
2nd Quartile 3rd Quartile
D D
Site Concentration Range
4th Quartile Average
D 1
0.5
14-16
-------
Figure 14-8. Program vs. Site-Specific Average Manganese (PMi0) Concentration
BO MA
1-
3
—I Program Max Concentration = 275 ng/m3
30 60 90 120
Concentration (ng/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
n n n n I
Site: Site Average Site Concentration Range
o —
IE
Figure 14-9. Program vs. Site-Specific Average Naphthalene Concentration
I-
3
0
100 200 300
Program: IstQuartile ;
Site: Site Average !
o
400 500 600 700 800
Concentration (ng/m3)
>nd Qua rti 1 e 3rd Qua rti 1 e 4th Qua rti 1 e
E D D
>ite Concentration Range
Average
9C
Figure 14-10. Program vs. Site-Specific Average Nickel (PMio) Concentration
8 10
Concentration {ng/m3)
Program: IstQuartile 2ndQuartile SrdQuartile
4thQuartile
Average
Site:
Site Average
Site Concentration Range
O
Observations from Figures 14-6 through 14-10 include the following:
• Figure 14-6 is the box plot for arsenic and shows that BOMA's annual average
arsenic (PMio) concentration is less than the program-level average concentration
but similar to the program-level median concentration. The maximum
concentration measured at BOMA is considerably less than the maximum
concentration measured at the program level. There were no non-detects of
arsenic measured at BOMA.
14-17
-------
• Figure 14-7 is the box plot for hexavalent chromium. Note that the program-level
maximum concentration (8.51 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 to
0.5 ng/m3. In addition, the first quartile for hexavalent chromium is zero and thus,
not visible in Figure 14-7. This box plot shows that the annual average
concentration of hexavalent chromium for BOMA is just less than the program-
level average concentration. The maximum concentration measured at BOMA is
significantly less than the maximum concentration measured at the program level.
As discussed in the previous section, several non-detects of hexavalent chromium
were measured at BOMA.
• Figure 14-8 is the box plot for manganese (PMio). Similar to hexavalent
chromium, the program-level maximum concentration (275 ng/m3) is not shown
directly on the box plot in order to allow for the observation of data points at the
lower end of the concentration range. Thus, the scale has been reduced to
150 ng/m3. Figure 14-8 shows that the range of manganese concentrations
measured at BOMA is relatively small compared to the range of manganese
concentrations measured across the program. The maximum manganese
concentration measured at BOMA is similar to the program-level average
concentration. The annual average manganese concentration for BOMA falls
between the program-level first quarter and median concentration.
• Figure 14-9 is the box plot for naphthalene and shows that the annual average
naphthalene concentration for BOMA is less than the program-level average and
similar to the program-level median concentration. The maximum concentration
measured at BOMA is considerably less than the maximum concentration
measured at the program level.
• Figure 14-10 is the box plot for nickel (PMio). This box plot shows that BOMA's
annual average concentration of nickel is just greater than the program-level
average and is similar to the program-level third quartile. The minimum nickel
concentration measured at BOMA is greater than the program-level first quartile.
The maximum nickel concentration measured at BOMA is among the higher
nickel concentrations measured across the program.
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 3.5.3.2.
BOMA has sampled PMio metals under the NMP since 2003; hexavalent chromium since 2005;
and PAHs since 2008. Thus, Figures 14-11 through 14-15 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
14-18
-------
of 6 months of sampling is required for inclusion in the trends analysis; in these cases, a 1-year
average is not provided, although the range and quartiles are still presented.
Figure 14-11. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at
BOMA
5th Percent!le
— Minimum
— Maximum
O 95thPercentile
••<>•" Average
A 1-year average is not presented because there were breaks in sampling during portions of 2004.
Observations from Figure 14-11 for arsenic measurements collected 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 is not presented, although the range of measurements is provided.
• The maximum arsenic concentration shown was measured on July 5, 2008. The next
highest concentration measured is approximately half as high and was measured on
July 4, 2006.
• The 1-year average concentrations of arsenic at BOMA have fluctuated over the
years, ranging from 0.36 ng/m3 (2010) to 0.61 ng/m3 (2008). For 2008, the maximum
concentration (5.45 ng/m3) 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
14-19
-------
concentration for 2008 was removed from the dataset, the 1-year average for 2008
would decrease from 0.61 ng/m3 to 0.53 ng/m3, making the changes in the 1-year
averages between 2007 and 2009 more subtle.
• The 1-year average and median concentrations increased from 2010 to 2011 and
again for 2012. Additional years of sampling are needed to determine if this trend
continues.
Figure 14-12. Yearly Statistical Metrics for Hexavalent Chromium Concentrations Measured
at BOMA
I
~ 0.30
..o
2008 2009
Year
O 5th Percentile — Minimuir
Median — Maximurr
95th Percent!le
Observations from Figure 14-12 for hexavalent chromium measurements collected at
BOMA include the following:
• The maximum hexavalent chromium concentration was measured in 2008
(0.525 ng/m3). Less than 10 percent of hexavalent chromium concentrations measured
at BOMA are greater than 0.1 ng/m3. At least one concentration greater than
0.1 ng/m3 has been measured in each year since the onset of sampling, with 2005
having the most (eight) and 2011 having the least (one).
• The range of measurements has varied each year, as indicated by both the maximum
concentration and the 95th percentile. The 95th percentile for 2008 is greater than the
maximum concentrations for 2010 and 2011.
• The 1-year average concentration decreased significantly from 2006 to 2007, then
increased for 2008. A decreasing trend is also shown between 2008 and 2010,
14-20
-------
followed by a slight increasing trend. However, there is a considerable amount of
variability within the measurements collected each year, as indicated by the
confidence intervals calculated for the 1-year averages, particularly for 2008.
• The minimum and 5th percentile are both zero for each year of sampling, indicating
the presence of non-detects. The percentage of non-detects has varied between
11 percent (2006) to 43 percent (2009).
Figure 14-13. Yearly Statistical Metrics for Manganese (PMi0) Concentrations Measured at
BOMA
I
~ 8.0
o
2008
Year
O 5th Percentile — Minimurr
Median — Maximuir
O 95th Percentile
Average
A 1-year average is not presented because there were breaks in sampling during portions of 2004.
Observations from Figure 14-13 for manganese measurements collected at BOMA
include the following:
• The maximum manganese concentration shown was measured on November 11, 2004
(13.5 ng/m3). Six additional manganese concentrations measured at BOMA are
greater than 10 ng/m3, and were measured in 2004, 2005, 2008, 2010, and 2012.
• A steady decrease in the upper range of concentrations measured, as indicated by the
maximum and 95th percentile, is shown between 2004 and 2007. Both the median
and 1-year average concentrations of manganese exhibit a decreasing trend as well.
With the exception of 2008, when all of the statistical metrics except the median
exhibit increases, the 1-year average concentrations of manganese changed relatively
little between 2007 and 2010. The 1-year average concentration ranged from
3.16 ng/m3 (2009) to 3.57 ng/m3 (2008) between 2007 and 2010.
14-21
-------
• Although an increasing trend is shown for 2011 and 2012, additional sampling is
required to determine if this trend continues.
Figure 14-14. Yearly Statistical Metrics for Naphthalene Concentrations Measured at BOMA
-
2010
Year
5th Percent!le
— Minimum
0 95thPercentile •••«••• Aver;
A 1-year average is not presented because sampling under the NMP did not begin until May 2008.
Observations from Figure 14-14 for naphthalene measurements collected 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 is not presented, although the
range of measurements is provided.
• The maximum naphthalene concentration was measured on the very first sample day
(May 6, 2008), although a similar measurement was collected in 2012. Only two
additional concentrations greater than 200 ng/m3 have been measured at BOMA
(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.
• The median concentration decreased significantly from 2008 to 2009, from
84.0 ng/m3 to 56.3 ng/m3. Little change is shown after 2008, with the median varying
by only 5 ng/m3 between 2009 and 2012. Similarly, the 1-year average varies by only
14-22
-------
10 ng/m3 for the years shown, ranging from 60.3 ng/m3 for 2011 to 70.3 ng/m3 for
2009.
Figure 14-15. Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at BOMA
r°
§
s
S 8.0
o
2008
Year
5th Percent!le
— Minimum — Median
— Maximum
95thPercentile
• Average
A 1-year average is not presented because there were breaks in sampling during portions of 2004.
Observations from Figure 14-15 for nickel measurements collected at BOMA include the
following:
• The maximum concentration was measured at BOMA in 2004 (17.2 ng/m3). All but
one of the nickel concentrations greater than 7.50 ng/m3 were measured in 2004 or
2005 (with the other was measured in 2012).
• A steady decreasing trend in the nickel measurements collected at BOMA is shown
through 2010. Little change is shown between 2010 and 2011 with the exception of
the maximum concentration. Even with the higher concentrations measured in 2012,
the 1-year average concentration did not change significantly from the previous year
(from 1.38 ng/m3 for 2011 to 1.41 ng/m3 for 2012).
14.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk at the
BOMA monitoring site. Refer to Sections 3.3 and 3.5.3 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.
14-23
-------
14.5.1 Risk-Based Screening Assessment Using MRLs
A risk-based screening was conducted by comparing the concentration data from the
Massachusetts monitoring site to the ATSDR MRLs, where available. As described in
Section 3.3, MRLs are noncancer health risk benchmarks and are defined for three exposure
periods: acute (exposures of 1 day to 14 days); intermediate (exposures of 15 days to 364 days);
and chronic (exposures of 1 year or greater). The preprocessed daily measurements of the
pollutants of interest were compared to the acute MRLs; the quarterly averages were compared
to the intermediate MRLs; and the annual averages were compared to the chronic MRLs.
As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2012.
14.5.2 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 3.5.3.4 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 14-6, 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 14-6 include the following:
• Naphthalene has the highest annual average concentration for BOMA. Manganese
and nickel also have annual average concentrations greater than 1.0 ng/m3.
• 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 (2.16 in-a-million for
naphthalene and 2.20 in-a-million for arsenic). This speaks to the relative toxicity of
one pollutant compared to the other.
14-24
-------
• 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 0.1. This indicates that no adverse health effects are
expected due to these individual pollutants.
Table 14-6. Risk Approximations for the Massachusetts Monitoring Site
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
Boston, Massachusetts - BOMA
Arsenic (PM10)
Hexavalent Chromium
Manganese (PM10)
Naphthalene
Nickel (PM10)
0.0043
0.012
_
0.000034
0.00048
0.000015
0.0001
0.00005
0.003
0.00009
61/61
50/61
61/61
59/59
61/61
0.51
±0.07
0.03
±0.02
4.06
±0.51
63.55
±9.09
1.41
±0.29
2.20
0.41
_
2.16
0.68
0.03
<0.01
0.08
0.02
0.02
— = A Cancer URE or Noncancer RfC is not available.
14.5.3 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 14-7 presents the 10 pollutants with the highest emissions from the 2011 NEI that have
cancer toxicity factors. Table 14-7 also presents the 10 pollutants with the highest toxicity-
weighted emissions, based on the weighting schema described in Section 3.5.3.5. Lastly,
Table 14-7 provides the 10 pollutants with the highest cancer risk approximations (in-a-million)
for BOMA, as presented in Table 14-6. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 14-7. Table 14-8 presents similar
information, but is limited to those pollutants with noncancer toxicity factors.
14-25
-------
Table 14-7. 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)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Boston, Massachusetts (Suffolk County) - BOMA
Formaldehyde
Benzene
Acetaldehyde
Ethylbenzene
1,3 -Butadiene
Tetrachloroethylene
Naphthalene
POM, Group 2b
POM, Group 2d
Propylene oxide
158.03
146.66
72.50
68.14
25.92
14.49
13.28
3.67
1.81
0.98
Formaldehyde
Benzene
1,3 -Butadiene
POM, Group 3
Naphthalene
Nickel, PM
POM, Group 2b
Arsenic, PM
Hexavalent Chromium, PM
Ethylbenzene
2.05E-03
1.14E-03
7.78E-04
5.81E-04
4.51E-04
4.47E-04
3.23E-04
2.53E-04
2.17E-04
1.70E-04
Arsenic
Naphthalene
Nickel
Hexavalent Chromium
2.20
2.16
0.68
0.41
-^
to
-------
Table 14-8. 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)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer Hazard
Approximation
(HQ)
Boston, Massachusetts (Suffolk County) - BOMA
Toluene
Hexane
Xylenes
Formaldehyde
Benzene
Acetaldehyde
Ethylbenzene
Methyl isobutyl ketone
1,3 -Butadiene
Glycol ethers, gas
534.20
403.19
289.61
158.03
146.66
72.50
68.14
56.12
25.92
16.16
Acrolein
Formaldehyde
1,3 -Butadiene
Nickel, PM
Acetaldehyde
Benzene
Naphthalene
Arsenic, PM
Cadmium, PM
Xylenes
515,526.17
16,125.78
12,959.09
10,345.15
8,055.15
4,888.74
4,425.71
3,914.83
2,970.00
2,896.12
Manganese
Arsenic
Naphthalene
Nickel
Hexavalent Chromium
0.08
0.03
0.02
0.02
<0.01
-^
to
-------
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 3.5.3.5. Similar to the cancer risk and
noncancer hazard approximations provided in Section 14.5.2, this analysis may help policy-
makers prioritize their air monitoring activities.
Observations from Table 14-7 include the following:
• Formaldehyde, benzene, and acetaldehyde are the highest emitted pollutants with
cancer UREs in Suffolk County.
• The pollutants with the highest toxicity-weighted emissions (of the pollutants with
cancer UREs) are formaldehyde, benzene, and 1,3-butadiene.
• Six of the highest emitted pollutants also have the highest toxicity-weighted
emissions.
• Four out of five of BOMA' s pollutants of interest have cancer UREs and thus, have
cancer risk approximations presented in Table 14-7. All four of these pollutants are
among those with the highest toxicity-weighted emissions. Conversely, only one of
them (naphthalene) appears among the highest emitted.
• POM, Group 2b ranks eighth for quantity emitted and seventh for toxicity-weighted
emissions. POM, Group 2b includes several PAHs sampled for at BOMA including
acenaphthene, fluoranthene, and fluorene. Although all three of these pollutants failed
at least one screen for BOMA, none of them were identified as pollutants of interest
for BOMA.
Observations from Table 14-8 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, formaldehyde, and 1,3-butadiene.
• Five of the highest emitted pollutants also have the highest toxicity-weighted
emissions.
14-28
-------
• Manganese, which has the highest (albeit low) noncancer hazard approximation for
BOMA, appears on neither emissions-based list.
• Although arsenic, naphthalene, and nickel are among the pollutants with the highest
toxi city-weighted emissions, none of these appear among the highest emitted
pollutants.
14.6 Summary of the 2012 Monitoring Data for BOMA
Results from several of the data treatments described in this section include the
following:
»«» Nine pollutants failed screens for BOMA, with arsenic and naphthalene accounting
for a majority of the failed screens.
»«» Naphthalene had the highest annual average concentration among the pollutants of
interest for BOMA.
»«» Even though concentrations of nickel have a decreasing trend over the years of
sampling, BOMA has the fifth highest annual average concentration for 2012 among
NMP sites sampling PM 10 metals.
14-29
-------
15.0 Sites in Michigan
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS and UATMP sites 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.
15.1 Site Characterization
This section characterizes the 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.
The DEMI, RRMI, and SWMI monitoring sites are located in the Detroit-Warren-
Dearborn, MI MSA. Figures 15-1 through 15-3 are the composite satellite images retrieved from
ArcGIS Explorer showing the monitoring sites and their immediate surroundings. Figure 15-4
identifies nearby point source emissions locations by source category, as reported in the 2011
NEI for point sources. Note that only sources within 10 miles of the sites are included in the
facility counts provided in Figure 15-4. 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 radii are still visible on the map,
but have been grayed out in order to show emissions sources just outside the boundary.
Table 15-1 provides supplemental geographical information such as land use, location setting,
and locational coordinates.
15-1
-------
Figure 15-1. Dearborn, Michigan (DEMI) Monitoring Site
-------
Figure 15-2. River Rouge, Michigan (RRMI) Monitoring Site
-------
Figure 15-3. Detroit, Michigan (SWMI) Monitoring Site
-------
Figure 15-4. NEI Point Sources Located Within 10 Miles of DEMI, RRMI, and SWMI
83"15'0"W 83'10'0'W
_«—-
I Wayne -_-
j County
7 I
>
- —f
83"25'0"W 83°20'0"W 83°15'0°W 83010'ETW 83r5-0"W 83'0'0"W
I OfionH Note: Due to facility density and collocation, the total facilities
•ir
DEMI NATTS site
O 10 mile radius
displayed may not represent all facilities within the area of interest.
RRMI UATMP site J^ SWMI UATMP site
County boundary
Source Category Group (No. of Facilities)
•^ Airport/Airlme^Airport Support Operations (10)
£ Asphalt Production/Hot Mix Asphalt Plant (3)
0 Auto Body Shop'Pa inters/Automotive Stores (1)
iji} Auto mobile fl>uck Manufactunng (7)
B Bulk Terminals/Bulk Plants (10)
C Chemical Manufacturing (5)
i Compressor Station (2)
Q Electrical Equipment Manufacturing (1)
f Electricity Generation via Combustion (8)
E Electroplating. Plating, Polishing, Anodizing, and Coloring (4)
F Food Processing/Agriculture (2)
I Foundries, Iron and Steel (1)
-)^f- Industnal Machinery or Equipment Plant (3)
O Institution (school, hospital, prison, etc.) (9)
ft Landfill (1)
|'j|1 Metal Can, Box. and Other Metal Container Manufacturing (1)
A
0
X
Q
Metal Coating, Engraving, and Allied Services to Manufacturers (4)
Metals Processing/Fabrication (4)
Mme/Quarry/Mineral Processing (10)
Miscellaneous Commercial/Industrial (18)
Municipal Waste Combustor (1 )
Paint and Coating Manufacturing (2)
Petroleum Products Manufactunng (1)
Petroleum Refinery (1)
Pharmaceutical Manufacturing (1)
Plastic, Resin, or Rubber Products Plant (4)
Printing/ Publish ing/Pa per Product Manufacturing (1 )
Rait Yard/Rail Line Operations (6)
Steel Mill (3)
Testing Laboratories (4)
Wastewater Treatment (1)
15-5
-------
Table 15-1. Geographical Information for the Michigan Monitoring Sites
Site
Code
DEMI
RRMI
SWMI
AQS Code
26-163-0033
26-163-0005
26-163-0015
Location
Dearborn
River
Rouge
Detroit
County
Wayne
Wayne
Wayne
Micro- or
Metropolitan
Statistical Area
Detroit- Warren-
Dearborn, MI
MSA
Detroit- Warren-
Dearborn, MI
MSA
Detroit- Warren-
Dearborn, MI
MSA
Latitude
and
Longitude
42.306666,
-83.148889
42.267222,
-83.132222
42.302778,
-83.106667
Land Use
Industrial
Industrial
Commercial
Location
Setting
Suburban
Suburban
Urban/City
Center
Additional Ambient Monitoring Information1
TSP Metals, Meteorological parameters, PM10, PM10
Speciation, PM2 5, PM2.5 Speciation, and IMPROVE
Speciation.
TSP Metals, Meteorological parameters, PM10, PM10
Manganese.
Soil Index, SO2, TSP Metals, VOCs, Meteorological
parameters, PM10, PM10 Manganese, PM25, PM2 5
Speciation, IMPROVE Speciation.
BOLD ITALICS = EPA-designated NATTS Site
-------
DEMI is located in the parking lot of Salinas 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 15-1 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).
RRMI is located at John Bilak Park in River Rouge, a southwestern suburb of Detroit,
less than 1 mile from the Detroit River and the U.S./Canadian border. The surrounding area is of
mixed usage, with residential properties surrounded by highly industrial ones. A freight yard is
located to the west of the site past Haltiner Street while the Port of Detroit is located just to the
east and southeast, just beyond the bottom right-hand side of Figure 15-2. This site is also
downwind of a steel manufacturing facility.
SWMI is located on the property of Southwestern High School in the city of Detroit. The
high school's track can be seen just west of the site marker in Figure 15-3. Interstate-75 runs
northeast-southwest less than 0.3 miles north of SWMI. The surrounding area is considered
commercial, although the site lies approximately 1 mile north of Zug Island, a small, highly
industrialized area where the Rouge River empties into the Detroit River. This site is also less
than 1 mile west of the Detroit River and U.S./Canadian border.
Figure 15-4 shows that DEMI, RRMI, and SWMI are located within a few miles of each
other. Numerous point sources surround these sites. A cluster of sources is located just west of
DEMI. Another cluster of sources is located just north of RRMI. The source categories with the
most point sources within 10 miles of these sites 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 15-4, the closest source to DEMI is involved in
automobile/truck manufacturing; the closest source to SWMI is involved in electricity generation
via combustion; and the closest source to RRMI is involved in asphalt/hot mix asphalt
production.
15-7
-------
Table 15-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Michigan monitoring sites. Table 15-2 includes both county-level
population and vehicle registration information. Table 15-2 also contains traffic volume
information for each site as well as the location for which the traffic volume was obtained.
Additionally, Table 15-2 presents the county-level daily VMT for Wayne County.
Table 15-2. Population, Motor Vehicle, and Traffic Information for the Michigan
Monitoring Sites
Site
DEMI
RRMI
SWMI
Estimated
County
Population1
1,792,365
County-level
Vehicle
Registration2
1,337,797
Annual
Average
Daily
Traffic3
87,500
97,300
94,400
Intersection
Used for
Traffic Data
1-94 between Ford Road and Rotunda
Drive
1-75 between Outer Drive & M-85
1-75 between Springwells Street and
Livernois Avenue
County-
level Daily
VMT4
40,951,779
Bounty-level population estimate reflects 2012 data (Census Bureau, 2013c)
2County-level vehicle registration reflects 2012 data (MDS, 2013)
3AADT reflects 2012 data (MI DOT, 2012)
4County-level VMT reflects 2011 data (MI DOT, 2013)
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 15-2 include the following:
• Wayne County's population and vehicle registration both rank eighth highest among
counties with NMP sites.
• The traffic volumes near the Michigan sites are similar to each other and rank 17th,
18th, and 21st among NMP sites. Traffic for DEMI is provided for 1-94, between
Ford Road and Rotunda Drive; traffic data for RRMI is for 1-75 between Outer Drive
and South Fort Street/M-85; and traffic data for SWMI is for 1-75 between
Springwells Street and Livernois Avenue.
• The Wayne County daily VMT is the seventh highest VMT compared to other
counties with NMP sites (where VMT data were available).
15-8
-------
15.2 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
sites in Michigan on sample days, as well as over the course of the year.
15.2.1 Climate Summary
Detroit is located in a region of active weather. Winters tend to be cold and wet, with
snowfall averages between 35 inches and 40 inches per year. Summers are generally mild,
although temperatures exceeding 90°F are not uncommon. Precipitation is fairly well distributed
throughout the year, with summer precipitation coming primarily in the form of showers and
thunderstorms. The urbanization of the area and Lake St. Clair to the east are major influences on
the city's weather. The lake tends to keep the Detroit area warmer in the winter and cooler in the
summer than more inland areas. The urban heat island also keeps the city warmer than outlying
areas. Winds are often breezy and flow from the southwest on average (Wood, 2004; MSU,
2014).
15.2.2 Meteorological Summary
Hourly meteorological data for 2012 were retrieved from NCDC for the weather station
closest to the Michigan monitoring sites (NCDC, 2012), as described in Section 3.5.2. The
closest weather station to all three sites is located at Detroit City Airport (WBAN 14822).
Additional information about this weather station, such as the distance between the sites and the
weather station, is provided in Table 15-3. These data were used to determine how
meteorological conditions on sample days vary from conditions experienced throughout the year.
15-9
-------
Table 15-3. Average Meteorological Conditions near the Michigan Monitoring Sites
Closest Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average 1
Scalar
Wind
Speed
(kt)
Dearborn, Michigan - DEMI
Detroit City Airport
14822
(42.41, -83.01)
9.7
miles
35°
(NE)
Sample
Day
(72)
2012
61.6
±4.4
61.6
±2.0
53.2
±4.0
53.6
±1.8
41.1
±3.5
41.4
± 1.6
47.1
±3.4
47.4
±1.6
66.6
±2.8
66.4
± 1.3
1016.1
±1.6
1016.1
±0.7
6.8
±0.7
6.5
±0.3 |
River Rouge, Michigan - RRMI |
Detroit City Airport
14822
(42.41, -83.01)
11.4
miles
23°
(NNE)
Sample
Day
(60)
2012
62.4
±4.9
61.6
±2.0
53.8
±4.5
53.6
±1.8
40.9
±3.9
41.4
±1.6
47.3
±3.8
47.4
±1.6
64.8
±3.0
66.4
±1.3
1016.2
±1.8
1016.1
±0.7
6.9
±0.8
6.5
±0.3
Detroit, Michigan - SWMI |
Detroit City Airport
14822
(42.41, -83.01)
8.7 miles
24°
(NNE)
Sample
Day
(30)
2012
61.9
±6.6
61.6
±2.0
52.9
±6.2
53.6
±1.8
40.7
±5.6
41.4
±1.6
46.8
±5.4
47.4
±1.6
65.8
±3.9
66.4
±1.3
1018.2
±2.7
1016.1
±0.7
5.8
±0.9
6.5
±0.3
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
-------
Table 15-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for all of 2012. Also included in Table 15-3 is the 95 percent
confidence interval for each parameter. Average meteorological conditions on sample days near
the sites were generally representative of average weather conditions experienced throughout the
year. Note that the number of sample days for SWMI is roughly half the number for DEMI and
RRMI. This is because SWMI sampled on a l-in-12 day schedule compared to DEMI and
RRMI, which sampled on a l-in-6 day schedule. The biggest difference in Table 15-3 in the
meteorological parameters is for sea level pressure for SWMI. This is because SWMI did not
sample on some of the days with the lowest sea level pressures. For example, SWMI did not
sample on October 30, 2012, the day that Hurricane Sandy came ashore over New Jersey.
15.2.3 Back Trajectory Analysis
Figure 15-5 is the composite back trajectory map for days on which samples were
collected at the DEMI monitoring site. Included in Figure 15-5 are four back trajectories per
sample day. Figure 15-6 is the corresponding cluster analysis. An in-depth description of these
maps and how they were generated is presented in Section 3.5.2.1. For the composite map, each
line represents the 24-hour trajectory along which a parcel of air traveled toward the monitoring
site on a given sample day and time, based on an initial height of 50 meters AGL. For the cluster
analysis, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric circle around the site in Figures 15-5 and 15-6 represents 100 miles.
Figures 15-7 throughl5-10 are the composite back trajectory maps and corresponding cluster
analyses for RRMI and SWMI.
15-11
-------
Figure 15-5. Composite Back Trajectory Map for DEMI
Figure 15-6. Back Trajectory Cluster Map for DEMI
15-12
-------
Figure 15-7. Composite Back Trajectory Map for RRMI
Figure 15-8. Back Trajectory Cluster Map for RRMI
15-13
-------
Figure 15-9. Composite Back Trajectory Map for SWMI
Figure 15-10. Back Trajectory Cluster Map for SWMI
15-14
-------
Observations from Figures 15-5 through 15-10 for the Michigan sites include the
following:
• The composite back trajectory maps for DEMI and RRMI are similar to each other in
trajectory distribution. This is expected given the close proximity to each other and
the similarities in sample days. The composite map for SWMI has roughly half the
back trajectories because this site sampled on a l-in-12 day sampling schedule rather
than a l-in-6 day schedule.
• Back trajectories originated from a variety of directions near the monitoring sites.
Back trajectories originating to the east of the sites tended to be shorter in length than
back trajectories from other directions.
• The 24-hour air shed domain for DEMI was similar in size to RRMI. The farthest
away a back trajectory originated from these sites was over North Dakota, or
750 miles away. The average back trajectory lengths for these two sites were just less
than 270 miles. Approximately 88 percent of trajectories originated within 450 miles
of the sites.
• For SWMI, the air shed domain was slightly smaller than those for DEMI and RRMI,
with an average back trajectory length of 251 miles, with greater than 90 percent of
back trajectories originating within 450 miles of the site. The longest back trajectories
were approximately 590 miles in length (one over southern Minnesota and one over
central Ontario, Canada).
• The cluster analysis for DEMI shows that nearly 20 percent of back trajectories
originated from a direction with a northerly component, over Michigan and Lake
Huron, and are generally less than 300 miles in length. Another 7 percent of back
trajectories originated to the northwest to north but were longer in length. Twenty-one
percent of back trajectories originated to the southwest, west, and northwest of
DEMI. Another 6 percent of back trajectories originated farther west over Iowa and
Minnesota. Twenty-two percent of back trajectories originated to the south of the site.
The short cluster trajectory originating over Lake Erie represents back trajectories
originating over the eastern half of Ohio and Lake Erie as well as shorter back
trajectories (generally less than 150 miles) originating from other directions.
• The cluster analysis for RRMI is similar to the cluster analysis for DEMI. The
primary difference is how the HYSPLIT model grouped the back trajectories with a
northerly component. For RRMI, they are split into three cluster trajectories rather
than two.
• The cluster analysis for SWMI shows that one-third of back trajectories originated to
the northwest of the site but are split into two cluster trajectories based on length.
Another 20 percent originated to the southwest of SWMI over Illinois and Indiana.
Eleven percent of back trajectories originated to the south of the site, mostly over
Ohio and Kentucky. Nearly one third of back trajectories are represented by the short
cluster trajectory originating over Lake Erie. This cluster trajectory includes back
trajectories originating to the northeast, east, and southeast of SWMI as well as a few
15-15
-------
shorter back trajectories spiraling in toward SWMI from other directions. The final
5 percent of back trajectories originated over southeast Ontario, Canada.
15.2.4 Wind Rose Comparison
Hourly surface wind data from the weather station at the Detroit City Airport were
uploaded into a wind rose software program to produce customized wind roses, as described in
Section 3.5.2.2. 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 15-11 presents a map showing the distance between the weather station and
DEMI, which may be useful for identifying topographical influences that may affect the
meteorological patterns experienced at this location. Figure 15-11 also presents three different
wind roses for the DEMI monitoring site. First, a historical wind rose representing 2002 to 2011
wind data is presented, which shows the predominant surface wind speed and direction over an
extended period of time. Second, a wind rose representing wind observations for all of 2012 is
presented. Next, a wind rose representing wind data for days on which samples were collected in
2012 is presented. These can be used to identify the predominant wind speed and direction for
2012 and to determine if wind observations on sample days were representative of conditions
experienced over the entire year and historically. Figures 15-12 and 15-13 present the distance
maps and wind roses for RRMI and SWMI.
15-16
-------
Figure 15-11. Wind Roses for the Detroit City Airport Weather Station near DEMI
Location of DEMI and Weather Station
2002-2011 Historical Wind Rose
2012 Wind Rose
Sample Day Wind Rose
Calms: 1408%
Calms: 12BH6
15-17
-------
Figure 15-12. Wind Roses for the Detroit City Airport Weather Station near RRMI
Location of RRMI and Weather Station
2002-2011 Historical Wind Rose
2012 Wind Rose
Sample Day Wind Rose
Calms: 1408%
15-18
-------
Figure 15-13. Wind Roses for the Detroit City Airport Weather Station near SWMI
Location of SWMI and Weather Station
2002-2011 Historical Wind Rose
.« ,."**""
>
Wlmlsw
< % Vv %
k)'\
•••' V* y*~ M "
/ \ V v
2012 Wind Rose
Sample Day Wind Rose
Calms: 14.t3-E%
Calms: 16.27%
15-19
-------
Observations from Figures 15-11 through 15-13 include the following:
• The weather station at Detroit City Airport is the closest weather station to all three
monitoring sites. This weather station is located to the northeast of all three sites and
ranges from 8.7 miles (SWMI) to 11.4 miles (RRMI) away from the sites. Most of the
city of Detroit lies between the weather station and the monitoring sites.
• Because the Detroit City Airport weather station is the closest weather station to all
three sites, the historical and 2012 wind roses for DEMI are the same as those for
RRMI and SWMI.
• The historical wind roses show that winds from a variety of directions were observed
near these sites, although winds from the southwest to west were the most frequently
observed while winds from the northeast and southeast quadrants were observed the
least. Calm winds (< 2 knots) were observed for roughly 10 percent of the hourly
measurements.
• The wind patterns on the 2012 wind roses resemble the historical wind patterns,
although there were slightly fewer south-southwesterly to westerly winds and slightly
more winds from due north and due south. The percentage of calm winds is also
higher (14 percent).
• The sample day wind rose for DEMI generally resembles the full-year wind rose,
although there was a higher percentage of winds from the southwest to west and
slightly fewer winds from the south on sample days.
• The sample day wind patterns for RRMI resemble the full-year wind patterns,
although there was an even higher percentage of winds from the southwest to west as
well as south on sample days near RRMI.
• The wind patterns on the sample day wind rose for SWMI differ from those on the
sample day wind roses for DEMI and RRMI. The calm rate (16 percent) is higher for
SWMI and winds from the north account for the highest percentage of wind
observations. Winds from the northwest, north-northwest, and north-northeast
account for nearly as many wind observations as winds from the southwest to west.
Recall that the sample day wind rose for SWMI has half the wind observations
compared to the sample day wind roses for DEMI and RRMI due to the sampling
frequency.
15.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each
Michigan 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
15-20
-------
"failed the screen." The site-specific results of this risk-based screening process are presented in
Table 15-4. 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 15-4. It is important to note which pollutants were sampled for at each site when reviewing
the results of this analysis. All three Michigan monitoring sites sampled for carbonyl
compounds; in addition, DEMI sampled for VOCs, PAHs, and hexavalent chromium.
Table 15-4. Risk-Based Screening Results for the Michigan Monitoring Sites
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Dearborn, Michigan - DEMI
Benzene
Carbon Tetrachloride
1,3 -Butadiene
Acetaldehyde
Formaldehyde
Naphthalene
1 ,2-Dichloroethane
Ethylbenzene
Fluorene
Acenaphthene
Hexavalent Chromium
Fluoranthene
£>-Dichlorobenzene
1 ,2-Dibromoethane
Hexachloro- 1 ,3 -butadiene
1 , 1 ,2,2-Tetrachloroethane
Benzo(a)pyrene
Chloroprene
Tetrachloroethylene
Trichloroethylene
Vinyl chloride
Xylenes
0.13
0.17
0.03
0.45
0.077
0.029
0.038
0.4
0.011
0.011
0.000083
0.011
0.091
0.0017
0.045
0.017
0.00057
0.0021
3.8
0.2
0.11
10
Total
63
63
61
60
60
59
53
26
20
19
11
8
5
2
2
2
1
1
1
1
1
1
520
63
63
61
60
60
60
53
62
60
60
58
60
26
2
3
2
58
1
61
16
16
63
968
100.00
100.00
100.00
100.00
100.00
98.33
100.00
41.94
33.33
31.67
18.97
13.33
19.23
100.00
66.67
100.00
1.72
100.00
1.64
6.25
6.25
1.59
53.72
12.12
12.12
11.73
11.54
11.54
11.35
10.19
5.00
3.85
3.65
2.12
1.54
0.96
0.38
0.38
0.38
0.19
0.19
0.19
0.19
0.19
0.19
12.12
24.23
35.96
47.50
59.04
70.38
80.58
85.58
89.42
93.08
95.19
96.73
97.69
98.08
98.46
98.85
99.04
99.23
99.42
99.62
99.81
100.00
River Rouge, Michigan - RRMI
Formaldehyde
Acetaldehyde
0.077
0.45
Total
49
49
98
49
49
98
100
100
100.00
50.00
50.00
50.00
100.00
Detroit, Michigan - SWMI
Acetaldehyde
Formaldehyde
0.45
0.077
Total
29
29
58
30
30
60
96.67
96.67
96.67
50.00
50.00
50.00
100.00
15-21
-------
Observations from Table 15-4 for DEMI include the following:
• Twenty-two pollutants failed at least one screen for DEMI; nearly 54 percent of
concentrations for these 22 pollutants were greater than their associated risk screening
value (or failed screens).
• Eleven pollutants contributed to 95 percent of failed screens for DEMI and therefore
were identified as pollutants of interest for DEMI. These 11 include two carbonyl
compounds, five VOCs, three PAHs, and hexavalent chromium.
• Benzene, carbon tetrachloride, 1,3-butadiene, acetaldehyde, and formaldehyde each
failed 100 percent of screens, with each contributing to roughly 12 percent to the total
number of failed screens; thus, these five pollutants together account for nearly
60 percent of the total failed screens.
Observations from Table 15-4 for RRMI and SWMI include the following:
• Acetaldehyde, formaldehyde, and propionaldehyde are the only carbonyl compounds
with risk screening values.
• Acetaldehyde and formaldehyde each failed 100 percent of screens for RRMI.
• Acetaldehyde and formaldehyde also failed screens for SWMI. These pollutants
contributed equally to the total number of failed screens.
• Acetaldehyde and formaldehyde were identified as pollutants of interest for both
sites.
15.4 Concentrations
This section presents various concentration averages used to characterize pollution levels
at the Michigan 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 data analysis is performed where the data meet the applicable criteria specified in
the appropriate sections discussed below. Additional site-specific statistical summaries for
DEMI, RRMI, and SWMI are provided in Appendices J, L, M, and O.
15-22
-------
15.4.1 2012 Concentration Averages
Quarterly and annual concentration averages were calculated for the pollutants of interest
for the Michigan sites, as described in Section 3.1. The quarterly average 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 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
Michigan monitoring sites are presented in Table 15-5, where applicable. Note that
concentrations of the PAHs and hexavalent chromium are presented in ng/m3 for ease of
viewing. Also note that if a pollutant was not detected in a given calendar quarter, the quarterly
average simply reflects "0" because only zeros substituted for non-detects were factored into the
quarterly average concentration.
15-23
-------
Table 15-5. Quarterly and Annual Average Concentrations of the Pollutants of
Interest for the Michigan Monitoring Sites
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Hg/m3)
2nd
Quarter
Average
(Hg/m3)
3rd
Quarter
Average
(Hg/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Ug/m3)
Dearborn, Michigan - DEMI
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
Acenaphthene3
Fluorene3
Hexavalent Chromium3
Naphthalene3
60/60
63/63
61/63
63/63
53/63
62/63
60/60
60/60
60/60
58/62
60/60
1.59
±0.33
1.02
±0.32
0.08
±0.03
0.67
±0.07
0.08
±0.02
0.46
±0.21
2.15
±0.48
5.23
±2.29
5.68
±2.22
0.04
±0.01
135.38
±51.33
1.75
±0.30
0.84
±0.31
0.08
±0.03
0.74
±0.03
0.08
±0.01
0.55
±0.21
4.42
±0.93
25.03
± 14.73
20.44
±11.46
0.05
±0.02
155.61
± 48.26
1.98
±0.29
0.89
±0.20
0.11
±0.03
0.68
±0.03
0.05
±0.01
0.56
±0.21
4.65
±0.79
16.74
±5.90
15.88
±5.38
0.06
±0.03
165.71
±41.92
1.68
±0.49
0.93
±0.41
0.18
±0.08
0.74
±0.03
0.05
±0.02
0.54
±0.43
2.41
±0.45
3.40
±1.13
3.40
±0.92
0.04
±0.02
110.08
±55.09
1.75
±0.17
0.92
±0.15
0.11
±0.03
0.71
±0.02
0.07
±0.01
0.53
±0.14
3.45
±0.44
12.60
±4.40
11.35
±3.53
0.05
±0.01
141.70
±23.82
River Rouge, Michigan - RRMI
Acetaldehyde
Formaldehyde
49/49
49/49
1.29
±0.27
2.97
±0.46
NA
NA
2.05
±0.25
4.90
±0.66
1.63
±0.54
4.01
±0.91
NA
NA
Detroit, Michigan - SWMI
Acetaldehyde
Formaldehyde
30/30
30/30
1.35
±0.30
2.08
±0.41
1.63
±0.31
3.32
±0.79
2.22
±0.76
4.91
± 1.06
1.21
±0.59
1.86
±0.92
1.62
±0.28
3.11
±0.59
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 DEMI from Table 15-5 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.
The second and third quarter average concentrations of formaldehyde are roughly
twice the other quarterly averages, supporting the seasonal trend discussed in
Section 4.4.2. A review of the data shows that the maximum concentration of
formaldehyde was measured on June 20, 2012 (8.39 |ig/m3) and that the 12 highest
concentrations measured at DEMI were measured during the second and third
15-24
-------
quarters of 2012. Conversely, all but one of the 11 formaldehyde concentrations less
than 2 |ig/m3 were measured during the first or fourth quarters of 2012.
• Although the first quarter average concentration of benzene is higher than the other
quarterly averages, the fourth quarter average concentration has the largest
confidence interval. A review of the data shows that the maximum concentration of
benzene was measured on November 17, 2012 (4.00 |ig/m3). This measurement is
among the highest benzene measurements among NMP sites sampling this pollutant.
The next highest concentrations of benzene were measured at DEMI in April
(2.81 |ig/m3) and March (2.46 |ig/m3) and are the only concentrations greater than
2 |ig/m3 measured at this site.
• The maximum 1,3-butadiene concentration measured at DEMI was also measured on
November 17, 2012 (0.703 |ig/m3) and is more than twice the next highest
concentration (0.266 |ig/m3, also measured in November). The five highest
concentrations of 1,3-butadiene (ranging from 0.213 |ig/m3to 0.703 |ig/m3) were
measured at DEMI during the fourth quarter of 2012. Yet, the minimum measured
detection was also measured during the fourth quarter of 2012 (0.031 |ig/m3). This
explains why the fourth quarter average is higher than the other quarterly averages
and also has a relatively large confidence interval associated with it.
• The fourth quarter average concentration of ethylbenzene has a relatively large
confidence interval associated with it. A review of the data shows that the maximum
concentration of this pollutant was measured on October 6, 2012 (3.63 |ig/m3) and is
nearly twice the next highest concentration (1.81 |ig/m3, measured in June). The
maximum ethylbenzene concentration measured at DEMI is also the highest
ethylbenzene concentration measured among NMP sites sampling this pollutant. Only
two sites (DEMI and GPCO) have ethylbenzene measurements greater than 3 |ig/m3.
• The second and third quarter average concentrations of acenaphthene and fluorene are
significantly higher than the other quarterly averages and have relatively large
confidence intervals associated with them. The maximum concentrations of these
pollutants were measured on the same day, June 20, 2012. The highest concentrations
of these pollutants were measured in June, July, and August, generally on the same
days, although the order varied. A similar observation was made in the 2011 NMP
report.
• The quarterly average concentrations of naphthalene all have relatively large
confidence intervals associated with them, indicating that the measurements are
highly variable. The maximum concentration of naphthalene (455 ng/m3) was
measured on November 17, 2012, the same day as the maximum benzene and
1,3-butadiene concentrations. This was the sixth highest concentration of naphthalene
measured across the program. Four concentrations of naphthalene greater than
300 ng/m3 were measured at DEMI, one in the first quarter, one in the second quarter
and two in the third quarter. At least one concentration greater than 200 ng/m3 was
measured each quarter of 2012: three during the first quarter, four during the second,
five during the third, and one (the maximum) during the fourth.
15-25
-------
Observations for RRMI and SWMI from Table 15-5 include the following:
• A collection error at RRMI resulted in the invalidation of carbonyl compound
samples between May 15, 2012 and July 8, 2012. As a result, second quarter and
annual average concentrations could not be calculated. However, Appendix L
provides the pollutant-specific average concentrations for all valid samples collected
over the entire sample period for this site.
• Concentrations of formaldehyde measured at RRMI range from 1.93 |ig/m3 to
7.70 |ig/m3; acetaldehyde concentrations range from 0.54 |ig/m3 to 4.42 |ig/m3'
• The maximum formaldehyde concentration was measured at SWMI on the same day
as the maximum acetaldehyde concentration (September 18, 2012). Concentrations of
formaldehyde for SWMI are highly variable, ranging from 0.0246 |ig/m3 to
7.70 |ig/m3; acetaldehyde concentrations exhibit similar variability, ranging from
0.0343 |ig/m3 to 4.35 |ig/m3'
Tables 4-9 through 4-12 present the sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for the
Michigan sites from those tables include the following:
• DEMI appears in Tables 4-9 through 4-12 a total of seven times.
• DEMI has the highest annual average concentration of carbon tetrachloride, as shown
in Table 4-9. However, the difference between highest and 10th highest annual
average concentration of this pollutant is only 0.03 |ig/m3. DEMI also has the sixth
highest annual average concentration of ethylbenzene and 10th highest annual
average concentration of 1,3-butadiene among NMP sites sampling these pollutants.
• The annual average concentration of formaldehyde for DEMI ranks eighth highest
among sites sampling carbonyl compounds.
• The annual average concentration of acenaphthene for DEMI is the second highest
among NMP sites sampling PAHs. DEMFs annual average concentrations of
fluorene and naphthalene each rank third among sites sampling PAHs, as shown in
Table 4-11.
15.4.2 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 shaded in
gray in Table 15-4. Figures 15-14 through 15-24 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 3.5.3.1.
15-26
-------
Figure 15-14. Program vs. Site-Specific Average Acenaphthene Concentration
DEMI
Program Max Concentration = 182 ng/m3
10
20
30
40 50 60
Concentration (ng/m3)
70
80
90
Program:
Site:
IstQuartile
D
Site Average
0
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
100
Figure 15-15. Program vs. Site-Specific Average Acetaldehyde Concentrations
DEMI
9 12
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o
Figure 15-16. Program vs. Site-Specific Average Benzene Concentration
2 3
Concentration {
Program:
Site:
IstQuartile
D
Site Average
O
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
15-27
-------
Figure 15-17. Program vs. Site-Specific Average 1,3-Butadiene Concentration
DEMI
i Program Max Concentration = 4.10 ug/m3
0.75 1
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site:
Site Average Site Concentration Range
o
Figure 15-18. Program vs. Site-Specific Average Carbon Tetrachloride Concentration
2 3
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o —
Figure 15-19. Program vs. Site-Specific Average 1,2-Dichloroethane Concentration
DEMI
] Program Max Concentration = 17.01 ug/n
0.2
0.3
0.4 0.5 0.6
Concentration {[og/m3)
0.7
0.8 0.9
Program:
Site:
IstQuartile
D
Site Average
O
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
15-28
-------
Figure 15-20. Program vs. Site-Specific Average Ethylbenzene Concentration
DEMI
•-
1.5 2 2.5
Concentration {[og/m3
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o
Figure 15-21. Program vs. Site-Specific Average Fluorene Concentration
10
20
30
40 50 60
Concentration (ng/m3)
70
80
90
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: SiteAverage Site Concentration Range
O
100
Figure 15-22. Program vs. Site-Specific Average Formaldehyde Concentrations
SWMI
6 8
Concentration {[og/m3
10
12
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: SiteAverage Site Concentration Range
o —
14
15-29
-------
Figure 15-23. Program vs. Site-Specific Average Hexavalent Chromium Concentration
DEMI
1
Program Max Concentration = 8.51 ng/m3
0.2 0.3
Concentration (ng/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site:
Site Average Site Concentration Range
o
Figure 15-24. Program vs. Site-Specific Average Naphthalene Concentration
100
200
300
400 500
Concentration (ng/m3)
600
700
800
900
Program:
Site:
IstQuartile
D
Site Average
O
2ndQuartile SrdQuartile
• D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
Observations from Figures 15-14 through 15-24 include the following:
• Figure 15-14 is the box plot for acenaphthene for DEMI. Note that the program-
level maximum concentration (182 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 to
100 ng/m3. This box plot shows that although the maximum acenaphthene
concentration measured at DEMI is not the maximum concentration across the
program, it is among the higher measurements. The maximum acenaphthene
concentration measured at DEMI is the third highest acenaphthene measurement
program-wide. The annual average acenaphthene concentration for DEMI (12.60
ng/m3) is more than twice the program-level average concentration (5.00 ng/m3).
• Figure 15-15 includes the box plots for acetaldehyde for DEMI and SWMI. The
box plots show that the maximum acetaldehyde concentrations measured at DEMI
and SWMI are significantly less than the program-level maximum concentration.
The annual average concentration for DEMI is roughly the same as the program-
level average concentration while the annual average for SWMI is just less than
the program-level average concentration (but greater than the program-level
median). Although no non-detects of acetaldehyde were measured at the
Michigan sites or across the program, the minimum concentration of acetaldehyde
15-30
-------
measured at SWMI is the lowest concentration program-wide (although a
concentration of the same magnitude was also measured at BMCO).
Figure 15-16 is the box plot for benzene, which shows that DEMFs annual
average benzene concentration is similar to the program-level average
concentration (0.90 |ig/m3). The maximum concentration of benzene measured at
DEMI is less than the maximum concentration measured at the program level, but
still among the higher measurements program-wide.
Figure 15-17 is the box plot for 1,3-butadiene. Similar to the acenaphthene box
plot, the program-level maximum concentration (4.10 |ig/m3) is not shown
directly on the box plot as the scale has been reduced to 2 |ig/m3 to allow for the
observation of data points at the lower end of the concentration range.
Figure 15-17 shows that the annual average concentration of 1,3-butadiene for
DEMI is also similar to the program-level average concentration. The maximum
1,3-butadiene concentration measured at DEMI is considerably less than the
maximum concentration measured across the program. Two non-detects of
1,3-butadiene were measured at DEMI.
Figure 15-18 is the box plot for carbon tetrachloride. The annual average
concentration of carbon tetrachloride for DEMI is similar to the program-level
average concentration (0.69 |ig/m3). The range of measurements collected at
DEMI is relatively small, spanning approximately 0.5 |ig/m3. The maximum
concentration measured at DEMI is significantly less than the maximum
concentration measured across the program.
Figure 15-19 is the box plot for 1,2-dichloroethane. Note that the program-level
maximum concentration (17.01 |ig/m3) is not shown directly on the box plot as
the scale has been reduced to 1 |ig/m3 in order to allow for the observation of data
points at the lower end of the concentration range. The entire range of
measurements collected at DEMI is less than the program-level average
concentration. This is because the program-level average is being driven by the
higher measurements collected at a few monitoring sites. The annual average for
DEMI is less than the median concentration measured at the program level. Ten
non-detects of 1,2-dichloroethane were measured at DEMI.
Figure 15-20 is the box plot for ethylbenzene. Even though the maximum
ethylbenzene concentration program-wide was measured at DEMI, this site does
not have the highest annual average concentration of this pollutant. The annual
average ethylbenzene concentration for DEMI is greater than the program-level
median, average, and third quartile.
The box plot for fluorene presented in Figure 15-21 is similar to the box plot for
acenaphthene in that the maximum fluorene concentration measured at DEMI is
among highest measurements of fluorene program-wide. The annual average
concentration for DEMI is more than twice the program-level average
concentration of fluorene.
15-31
-------
• Figure 15-22 includes the box plots for formaldehyde for DEMI and SWMI. The
box plots show that the annual average concentration for DEMI is slightly greater
than the annual average concentration for SWMI. The annual average for DEMI
is similar to the program-level third quartile while the annual average for SWMI
less than the program-level third quartile, although both are greater than the
program-level average concentration. Although no non-detects of formaldehyde
were measured at the Michigan sites or across the program, the minimum
concentration of formaldehyde measured at SWMI is the minimum concentration
program-wide.
• Figure 15-23 is the box plot for hexavalent chromium. Note that the program-
level maximum concentration (8.51 ng/m3) is not shown directly on the box plot
as the scale has been reduced to 0.5 ng/m3 in order to allow for the observation of
data points at the lower end of the concentration range. In addition, the program-
level first quartile is zero and therefore not visible on the box plot. The box plot
shows that annual average concentration for DEMI is greater than the program-
level average concentration, even though the maximum concentration measured
across the program is substantially higher than the maximum concentration
measured at DEMI.
• Figure 15-24 is the box plot for naphthalene. The annual average concentration of
naphthalene for DEMI is greater than the program-level average concentration.
The maximum naphthalene concentration measured at DEMI is less than the
maximum concentration measured across program, although it was the sixth
highest naphthalene concentration measured across the program.
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 3.5.3.2.
DEMI has sampled VOCs and carbonyl compounds under the NMP since 2003, hexavalent
chromium since 2005, and PAHs since 2008. Thus, Figures 15-25 through 15-35 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 is not provided, although the range and quartiles are still presented.
Although RRMI and SWMI have sampled under the NMP previously, they have not sampled
continuously for 5 consecutive years; thus, a trends analysis was not performed for these sites.
15-32
-------
Figure 15-25. Yearly Statistical Metrics for Acenaphthene Concentrations Measured at DEMI
1
j'
§
g
2010
Year
O 5th Percentile
— Minimum
— Maximum
0 95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until April 2008.
Observations from Figure 15-25 for acenaphthene measurements collected 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 is not presented, although the
range of measurements is provided.
• The maximum acenaphthene concentration measured at DEMI was measured in
August 2010 (175 ng/m3). All five of the concentrations greater than 100 ng/m3 were
measured in either July or August; further, all 30 measurements greater than 25 ng/m3
were measured during the second or third quarters of the year (the warmer months of
the year).
• The range of concentrations measured decreased from 2008 to 2009 as the maximum
concentration for 2009 is less than the 95th percentile for 2008.
• Nearly all of the statistical metrics increased from 2009 to 2010, including the
median, which is influenced less by a few concentrations at the upper end of the
concentration range than the 1-year average, such as the two concentrations greater
than 100 ng/m3 that were measured in 2010; the next highest concentration was
considerably less (55.1 ng/m3).
15-33
-------
• Although the 95th percentile increased from 2010 to 2011, the 1-year average
exhibits a slight decrease, which continues into 2012. However, confidence intervals
calculated for these averages indicate that the measurements are highly variable.
Figure 15-26. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at DEMI
2003 2004 2005 2006 2007 2008 2009 2010 2011
5th Percentile — Minimum — Median — Maximurr
95th Percentile — «••
A 1-year average is not presented because data from March 2007 to March 2008 was invalidated.
Observations from Figure 15-26 for acetaldehyde measurements collected 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 from
March 13, 2007 through 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 in 2004 (7.84 |ig/m3). Of
the six concentrations greater than 5 |ig/m3 measured at DEMI, three were measured
in 2004, two were measured in 2005, and one was measured in 2006.
• The 1-year average concentration exhibits a decreasing trend after 2004 that
continues through 2006. The median concentration, which is available for 2008,
15-34
-------
changed little from 2006 to 2008, but decreased slightly for 2009. Both the 1-year
average and median concentrations exhibit an increasing trend after 2009 that levels
off for 2012, although these changes are not statistically significant.
Figure 15-27. Yearly Statistical Metrics for Benzene Concentrations Measured at DEMI
~ 4.0
T V
t^
2007 2008
Year
O 5th Percentile
— Maximum
95th Percentile
A 1-year average is not presented due to low completeness for 2003.
Observations from Figure 15-27 for benzene measurements collected at DEMI include
the following:
• VOCs have been sampled continuously at DEMI under the NMP since 2003.
However, the l-in-12 day schedule in 2003 combined with a number of invalids
resulted in low completeness; as a result, a 1-year average is not presented for 2003.
• The three highest benzene concentrations were all measured in 2004 and ranged 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
over the period shown, reaching a minimum in 2009. Between 2009 and 2012, the
1-year average fluctuated between 0.81 |ig/m3 (2009) and 0.94 |ig/m3 (2010).
• The difference between the 1-year average and median concentrations has decreased
over the years, indicating less variability in the measurements. Between 2009 and
2011, less than 0.1 |ig/m3 separates these two statistical parameters.
15-35
-------
• The minimum, 5th percentile and median concentration decreased from 2011 to 2012
while the 1-year average, 95th percentile, and maximum increased. Thus, the
measurements at the lower end of the concentration range decreased while the
measurements at the upper end of the concentration range increased.
Figure 15-28. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at DEMI
o
2007 2008
Year
5th Percentile
- Minimum
— Maximum
95th Percentile
••<>•" Average
A 1-year average is not presented due to low completeness for 2003.
Observations from Figure 15-28 for 1,3-butadiene measurements collected at DEMI
include the following:
• The maximum 1,3-butadiene concentration was measured on October 18, 2004. This
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 this site in 2004; in addition, the
maximum 95th percentile was calculated for 2004. This indicates there is a high level
of variability within the 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 down to 2 percent in 2008, then fluctuated between 2 percent and
7 percent for the years that follow.
15-36
-------
• Even as the number of non-detects decreased (and thus, the number of zeros factored
into the calculated decrease), the 1-year average concentration decreased between
2006 and 2009. An increasing trend is shown for the years after 2009.
Figure 15-29. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured
at DEMI
5th Percentile
1
T
g
2007 2008
Year
- Minimum
— Maximum
95th Percentile
••<>•" Average
A 1-year average is not presented due to low completeness for 2003.
Observations from Figure 15-29 for carbon tetrachloride measurements collected at
DEMI include the following:
• In 2003, the measured detections ranged from 0.25 |ig/m3 to 0.76 |ig/m3, plus two
non-detects. This is the only year of sampling for which half the measurements were
less than 0.5 |ig/m3.
• The range of concentrations in 2004 doubled from 2003 levels. The number of
measurements greater than 1 |ig/m3 increased from none in 2003 to 12 for 2004.
• The 1-year average decreased by more than 1 |ig/m3 from 2004 to 2005, as the range
of measurements decreased substantially, with little change in the 1-year average
from 2005 to 2007.
• With the exception of the 5th percentile, all of the statistical metrics increased
significantly for 2008, with the 1-year average and median concentrations for 2008
similar to the 95th percentile for 2007.
15-37
-------
• A steady decreasing trend is shown between 2008 and 2011. Between these years, the
majority of concentrations are falling within a tighter concentration range. 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.
Figure 15-30. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured
at DEMI
S 0.2
2007 2008
Year
5th Percentile - Minimum ~ Median — Maximum
95th Percentile
A 1-year average is not presented due to low completeness for 2003.
Observations from Figure 15-30 for 1,2-dichloroethane measurements collected at DEMI
include the following:
• There were no measured detections of 1,2-dichloroethane in 2003, 2004, 2007, or
2008; there was one measured detection in 2005, three in 2006, four in 2009, 12 in
2010, 11 in 2011, and 53 in 2012. With the exception of 2012, the median
concentration is zero for all years, indicating that at least half of the measurements are
non-detects.
• As the number of measured detections increase, so do each of the corresponding
statistical metrics shown in Figure 15-30.
• As the number of measured detections increased dramatically for 2012, the 1-year
average and median concentrations increased correspondingly. The median
concentration is actually greater than the annual average for 2012. This is because
there were still 10 non-detects (or zeros) factoring into the 1-year average
concentration for the year.
15-38
-------
• The maximum 1,2-dichloroethane concentration measured at DEMI was measured on
July 16, 2006 (3.45 |ig/m3). The next highest concentration was also measured in
2006 but was considerably less (0.16 jig/m3). A similar measurement was also
collected in 2005. All of the 10 remaining concentrations greater than 0.1 |ig/m3 were
measured in 2011 or 2012.
Figure 15-31. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at DEMI
L-j-l
L^J
2003 2004 2005 2006 2007 2008 2009 2010
2011 2012
5th Percentile - Minimum ~ Median — Maximum
95th Percentile
A 1-year average is not presented due to low completeness for 2003.
Observations from Figure 15-31 for ethylbenzene measurements collected at DEMI
include the following:
• The maximum ethylbenzene concentration was measured in September 2004
(4.35 |ig/m3). The next highest concentration was measured in 2012 (3.63 |ig/m3).
The only other ethylbenzene measurement greater than 3 |ig/m3 was measured in
2011. Only 10 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.
• All of the statistical parameters exhibit increases for 2010, with most continuing this
increase for 2011.
• The maximum concentration measured exhibits a steady increasing trend between
2008 and 2012.
15-39
-------
• For 2012, the magnitude of the measurements at the lower end of the concentration
range decreased (including one non-detect) while the measurements at the upper end
of the concentration range increased. However, 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, resulting in the slight decreases shown in the central tendency
statistics.
Figure 15-32. Yearly Statistical Metrics for Fluorene Concentrations Measured at DEMI
16C
14C
12C
IOC
•a
1 8C
1
6C
4C
2C
C
C
P
)
P
2008 1
9 5th Percentile
I
2009
vlinimum
(
<
•
C
T
"
5 1
> o o
" ^ ^
2010 2011 2012
Year
Median — Maximum O 95th Percentile ...^... Average
Observations from Figure 15-32 for fluorene measurements collected at DEMI include
the following:
• The maximum fluorene concentration was measured at DEMI in August 2010; only
two other measurements greater than 100 ng/m3 have been measured at DEMI (one in
August 2008 and one in August 2010). All eight concentrations greater than 50 ng/m3
have been measured in June, July, or August and all 32 concentrations greater than
20 ng/m3 were measured at DEMI during the second or third quarters of the year (the
warmer months of the year), similar to acenaphthene and fluoranthene.
• The median concentrations have varied less than 2 ng/m3 over the years, ranging from
4.92 ng/m3 (2011) to 6.82 ng/m3 (2010). The 1-year average concentrations exhibit
more variability, although little change is shown from 2011 to 2012.
• All of the statistical metrics increased (at least slightly) from 2009 to 2010. The
1-year average is being driven by the two highest concentrations measured in 2010
15-40
-------
(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 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 six during this
period. There were seven concentrations greater than 25 ng/m3 measured in 2012,
even though the 95th percentile exhibits a slight decrease.
Figure 15-33. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at DEMI
WrJ,
2007 2008
Year
O 5th Percentile — Minimum — Median — Maximum
0 95th Percentile
A 1-year average is not presented because data from March 2007 to March 2008 was invalidated.
Observations from Figure 15-33 for formaldehyde measurements collected 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 is not presented for 2008, although
the range of measurements is provided.
• The maximum formaldehyde concentration shown was measured in 2005
(33.1 |ig/m3). The next four highest concentrations measured at DEMI were also
measured in 2005 and ranged from 13.3 |ig/m3 to 20.9 |ig/m3. The only other
formaldehyde concentrations greater than 10 |ig/m3 were measured in 2004.
15-41
-------
• The decrease in the 1-year average concentration shown between 2005 and 2006 is
significant (from 5.35 |ig/m3to 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 measurements collected 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 range, there were nine concentrations
greater than 4 |ig/m3 in 2011 compared to 21 in 2012.
Figure 15-34. Yearly Statistical Metrics for Hexavalent Chromium Concentrations Measured
at DEMI
I
I °'3°
o
o
5th Percentile
- Minimum
— Maximum
95th Percentile
Observations from Figure 15-34 for hexavalent chromium measurements collected at
DEMI include the following:
• The minimum concentrations and 5th percentiles for several years are both zero,
indicating the presence of non-detects. The percentage of non-detects has ranged from
27 percent (2009) to less than 2 percent (2007 and 2011).
• The maximum hexavalent chromium concentration was measured in 2006. The two
highest hexavalent chromium concentrations for this site were measured on
July 4, 2006 (0.496 ng/m3) and on July 5, 2008 (0.392 ng/m3), although a similar
concentration was also measured on January 1, 2009 (0.372 ng/m3).
15-42
-------
• Although a decrease in the 1-year average concentration is shown from 2006
(0.068 ng/m3) to 2007 (0.042 ng/m3), the confidence intervals calculated are relatively
large as a result of the highest concentrations and indicate that these changes are not
statistically significant. The 1-year average concentration changed little after 2006,
ranging from 0.036 ng/m3 (2009) to 0.048 ng/m3 (2012). The median concentration
exhibits a little more variability, ranging from 0.019 ng/m3 (2009) to 0.041 ng/m3
(2011).
Figure 15-35. Yearly Statistical Metrics for Naphthalene Concentrations Measured at DEMI
r
§
s
3 200
o
2010
Year
O 5th Percentile
- Minimum
— Maximum
95th Percentile
••<>•" Average
A 1-year average is not presented because sampling under the NMP did not begin until April 2008.
Observations from Figure 15-35 for naphthalene measurements collected 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 since the onset of sampling).
• The median concentrations exhibit a slight increasing trend through 2011, as do the
1-year average concentrations, although the confidence intervals calculated are
relatively large as a result of the wide range of concentrations measured and indicate
that these changes are not statistically significant. The range of concentrations
measured each year at DEMI spans more than 350 ng/m3 each year.
• The difference between the 1-year average and median concentrations exhibits an
increase for each year, reaching a maximum for 2012.
15-43
-------
15.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk at the
Michigan monitoring sites. Refer to Sections 3.3 and 3.5.3 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.
15.5.1 Risk-Based Screening Assessment Using MRLs
A risk-based screening was conducted by comparing the concentration data from the
Michigan monitoring sites to the ATSDR MRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 day to 14 days); intermediate (exposures of 15 days to 364 days); and chronic
(exposures of 1 year or greater). The preprocessed daily measurements of the pollutants of
interest were compared to the acute MRLs; the quarterly averages were compared to the
intermediate MRLs; and the annual averages were compared to the chronic MRLs.
As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2012.
15.5.2 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Michigan 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 3.5.3.4 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 15-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
15-44
-------
Table 15-6. Risk Approximations for the Michigan Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
Dearborn, Michigan - DEMI
Acenaphthene3
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
Ethylbenzene
Fluorene3
Formaldehyde
Hexavalent Chromium3
Naphthalene3
0.000088
0.0000022
0.0000078
0.00003
0.000006
0.000026
0.0000025
0.000088
0.000013
0.012
0.000034
_
0.009
0.03
0.002
0.1
2.4
1
0.0098
0.0001
0.003
River Rou
Acetaldehyde
Formaldehyde
0.0000022
0.000013
0.009
0.0098
60/60
60/60
63/63
61/63
63/63
53/63
62/63
60/60
60/60
58/62
60/60
0.01
±0.01
1.75
±0.17
0.92
±0.15
0.11
±0.03
0.71
±0.02
0.07
±0.01
0.53
±0.14
0.01
±<0.01
3.45
±0.44
<0.01
±<0.01
0.14
±0.02
1.11
3.86
7.15
3.44
4.27
1.72
1.33
1.00
44.80
0.58
4.82
_
0.19
0.03
0.06
0.01
<0.01
0.01
0.35
O.01
0.05
»e, Michigan - RRMI
49/49
49/49
NA
NA
NA
NA
NA
NA
Detroit, Michigan - SWMI
Acetaldehyde
Formaldehyde
0.0000022
0.000013
0.009
0.0098
30/30
30/30
1.62
±0.28
3.11
±0.59
3.57
40.48
0.18
0.32
— = A Cancer URE or Noncancer RfC is not available.
3 For the annual average concentration of this pollutant in ng/m3, refer to Table 15-5.
Observations from Table 15-6 include the following:
• Formaldehyde has the highest annual average concentration for DEMI and SWMI.
This pollutant also has the highest cancer risk approximation for these sites, ranging
from 40.48 in-a-million for SWMI to 44.80 in-a-million for DEMI.
• The range of cancer risk approximations for acetaldehyde was even tighter, ranging
from 3.57 in-a-million for SWMI to 3.86 in-a-million for DEMI.
• Aside from formaldehyde and acetaldehyde, the pollutants with the highest cancer
risk approximations for DEMI were benzene, naphthalene, and carbon tetrachloride.
15-45
-------
• None of the pollutants of interest for DEMI or SWMI have noncancer hazard
approximations greater than 1.0, indicating that no adverse health effects are expected
from these individual pollutants. The pollutant with the highest noncancer hazard
approximation for DEMI and SWMI is formaldehyde (which ranged from 0.32 to
0.35).
15.5.3 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 15-7 presents the 10 pollutants with the highest emissions from the 2011 NEI that have
cancer toxicity factors. Table 15-7 also presents the 10 pollutants with the highest toxicity -
weighted emissions, based on the weighting schema described in Section 3.5.3.5. Lastly,
Table 15-7 provides the 10 pollutants with the highest cancer risk approximations (in-a-million)
for each site, as presented in Table 15-6. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 15-7. Table 15-8 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 3.5.3.5. Similar to the cancer risk and
noncancer hazard approximations provided in Section 15.5.2, this analysis may help policy-
makers prioritize their air monitoring activities.
15-46
-------
Table 15-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Michigan Monitoring Sites
Top 10 Total Emissions for Pollutants
with Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Dearborn, Michigan (Wayne County) - DEMI
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
Tetrachloroethylene
Trichloroethylene
Dichloromethane
Coke Oven Emissions, PM
516.86
422.71
350.07
251.15
81.46
46.49
30.63
17.05
10.96
8.71
Coke Oven Emissions, PM
Formaldehyde
Benzene
POM, Group 5a
Hexavalent Chromium, PM
1,3 -Butadiene
Arsenic, PM
Nickel, PM
Naphthalene
Ethylbenzene
8.62E-03
5.50E-03
4.03E-03
3.00E-03
2.72E-03
2.44E-03
2.15E-03
1.59E-03
1.58E-03
8.75E-04
Formaldehyde
Benzene
Naphthalene
Carbon Tetrachloride
Acetaldehyde
1,3 -Butadiene
1 ,2-Dichloroethane
Ethylbenzene
Acenaphthene
Fluorene
44.80
7.15
4.82
4.27
3.86
3.44
1.72
1.33
1.11
1.00
River Rouge, Michigan (Wayne County) - RRMI
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
Tetrachloroethylene
Trichloroethylene
Dichloromethane
Coke Oven Emissions, PM
516.86
422.71
350.07
251.15
81.46
46.49
30.63
17.05
10.96
8.71
Coke Oven Emissions, PM
Formaldehyde
Benzene
POM, Group 5a
Hexavalent Chromium, PM
1,3 -Butadiene
Arsenic, PM
Nickel, PM
Naphthalene
Ethylbenzene
8.62E-03
5.50E-03
4.03E-03
3.00E-03
2.72E-03
2.44E-03
2.15E-03
1.59E-03
1.58E-03
8.75E-04
-------
Table 15-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Michigan Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants
with Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
Detroit, Michigan (Wayne County) - SWMI
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
Tetrachloroethylene
Trichloroethylene
Dichloromethane
Coke Oven Emissions, PM
516.86
422.71
350.07
251.15
81.46
46.49
30.63
17.05
10.96
8.71
Coke Oven Emissions, PM
Formaldehyde
Benzene
POM, Group 5a
Hexavalent Chromium, PM
1,3 -Butadiene
Arsenic, PM
Nickel, PM
Naphthalene
Ethylbenzene
8.62E-03
5.50E-03
4.03E-03
3.00E-03
2.72E-03
2.44E-03
2.15E-03
1.59E-03
1.58E-03
8.75E-04
Formaldehyde 40.48
Acetaldehyde 3.57
oo
-------
Table 15-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Michigan Monitoring Sites
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer Hazard
Approximation
(HQ)
Dearborn, Michigan (Wayne County) - DEMI
Toluene
Hydrochloric acid
Ethylene glycol
Xylenes
Hexane
Methanol
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
4,113.77
3,022.42
2,244.10
1,299.49
1,288.67
1,113.11
516.86
422.71
350.07
251.15
Acrolein
Hydrochloric acid
Manganese, PM
Formaldehyde
1,3 -Butadiene
Nickel, PM
Arsenic, PM
Acetaldehyde
Benzene
Naphthalene
1,292,303.40
151,120.80
127,044.59
43,133.35
40,728.90
36,772.89
33,262.46
27,905.70
17,228.71
15,496.90
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Naphthalene
Benzene
Carbon Tetrachloride
Ethylbenzene
Hexavalent Chromium
1 ,2-Dichloroethane
0.35
0.19
0.06
0.05
0.03
0.01
<0.01
<0.01
<0.01
River Rouge, Michigan (Wayne County) - RRMI
Toluene
Hydrochloric acid
Ethylene glycol
Xylenes
Hexane
Methanol
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
4,113.77
3,022.42
2,244.10
1,299.49
1,288.67
1,113.11
516.86
422.71
350.07
251.15
Acrolein
Hydrochloric acid
Manganese, PM
Formaldehyde
1,3 -Butadiene
Nickel, PM
Arsenic, PM
Acetaldehyde
Benzene
Naphthalene
1,292,303.40
151,120.80
127,044.59
43,133.35
40,728.90
36,772.89
33,262.46
27,905.70
17,228.71
15,496.90
-------
Table 15-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Michigan Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Noncancer Hazard
Approximation
Pollutant (HQ)
Detroit, Michigan (Wayne County) - SWMI
Toluene
Hydrochloric acid
Ethylene glycol
Xylenes
Hexane
Methanol
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
4,113.77
3,022.42
2,244.10
1,299.49
1,288.67
1,113.11
516.86
422.71
350.07
251.15
Acrolein
Hydrochloric acid
Manganese, PM
Formaldehyde
1,3 -Butadiene
Nickel, PM
Arsenic, PM
Acetaldehyde
Benzene
Naphthalene
1,292,303.40
151,120.80
127,044.59
43,133.35
40,728.90
36,772.89
33,262.46
27,905.70
17,228.71
15,496.90
Formaldehyde 0.32
Acetaldehyde 0.18
-------
Observations from Table 15-7 include the following:
• Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
cancer UREs in Wayne County.
• The pollutants with the highest toxi city-weighted emissions (of the pollutants with
cancer UREs) for Wayne County are coke oven emissions, formaldehyde, and
benzene.
• Six of the highest emitted pollutants in Wayne County also have the highest toxicity-
weighted emissions.
• Formaldehyde has the highest cancer risk approximation for DEMI and SWMI. This
pollutant also appears on both emissions-based lists. Acetaldehyde is one of the
highest emitted pollutants but does not appear among those with the highest toxicity-
weighted emissions.
• In addition to formaldehyde, benzene, naphthalene, ethylbenzene, and 1,3-butadiene
are among the pollutants with the highest cancer risk approximations for DEMI and
appear on both emissions-based lists. Hexavalent chromium has the fifth highest
toxicity-weighted emissions but does not appear among the highest emitted
pollutants. Carbon tetrachloride does not appear on either emissions-based list.
Observations from Table 15-8 include the following:
• Toluene, hydrochloric acid, and ethylene glycol are the highest emitted pollutants
with noncancer RfCs in Wayne County. The quantity of emissions for highest ranking
the pollutants in Table 15-8 is an order of magnitude higher than the quantity of
emissions for the highest ranking pollutants in Table 15-7.
• The pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) for Wayne County are acrolein, hydrochloric acid, and manganese.
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 and acetaldehyde have the highest noncancer hazard approximations
for both DEMI and SWMI, although none of the pollutants of interest have associated
noncancer hazard approximations greater than 1.0. Formaldehyde emissions rank
eighth highest for Wayne County while the toxicity-weighted emissions (of the
pollutants with noncancer RfCs) rank fourth.
15-51
-------
• Acetaldehyde also appears on both emissions-based lists for DEMI and SWMI.
Acetaldehyde ranks tenth for quantity emitted and eighth for its toxicity-weighted
emissions.
• Benzene is the only other pollutant that appears on all three lists for DEMI.
15.6 Summary of the 2012 Monitoring Data for DEMI, RRMI, and SWMI
Results from several of the data treatments described in this section include the
following:
»«» Twenty-two pollutants failed screens for DEMI. Acetaldehyde and formaldehyde both
failed screens for RRMI and SWMI.
*»* Of the site-specific pollutants of interest, formaldehyde had the highest annual
average concentration for DEMI and SWMI. Annual average concentrations could
not be calculated for RRMI.
»«» DEMI has the highest annual average concentration of carbon tetrachloride among
NMP sites sampling VOCs. DEMI also has some of the highest annual average
concentrations ofacenaphthene,fluorene, and naphthalene among NMP sites
sampling PAHs. The highest concentration of ethylbenzene program-wide was
measured at DEMI.
»«» Concentrations of acenaphthene andfluorene measured at DEMI tended to be
highest during the summer months. A significant decrease in benzene concentrations
occurred at DEMI for many years, although concentrations have leveled off in recent
years. The detection rate of 1,2-dichloroethane has been increasing steadily at DEMI
over the last few years of sampling.
15-52
-------
16.0 Site in Minnesota
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the UATMP site in Minnesota, 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 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 STMN site is located in St. Cloud, Minnesota. Figure 16-1 is a composite satellite
image retrieved from ArcGIS Explorer showing the 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. 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 radius
are still visible on the map, but have been grayed out in order to show emissions sources just
outside the boundary. Table 16-1 provides supplemental geographical information such as land
use, location setting, and locational coordinates.
16-1
-------
Figure 16-1. St. Cloud, Minnesota (STMN) Monitoring Site
to
-------
Figure 16-2. NEI Point Sources Located Within 10 Miles of STMN
94:30'0"W 94=25'0"W
Legend
STMN UATMP site
WIS'O'W 94;10'0"W 9435'0"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
Source Category Group (No. of Facilities)
•*• Airport/Airline/Airport Support Operations (3)
* Asphalt Production/Hot Mix Asphalt Plant (1)
B Automobile/Truck Manufacturing (2)
c Chemical Manufacturing (1)
X Crematory-Animal/Human (3)
® Dry Cleaning (3)
* Electricity Generation via Combustion (1)
F Food Processing/Agriculture (1)
i Foundries, Iron and Steel (1)
•#• Industrial Machinery or Equipment Plant (3)
o Institution (school, hospital, prison, etc.) (9)
Landfill (1)
© Metals Processing/Fabrication (3)
A Military Base/National Security (1)
Mine/Quarry/Mineral Processing (4)
Miscellaneous Commercial/Industrial (2)
P Printing/Publishing/Paper Product Manufacturing (3)
H Pulp and Paper Plant (1)
n Telecommunications/Radio (1)
T Textile, Yarn, or Carpet Plant (1)
» Wastewater Treatment (6)
W Woodwork, Furniture, Millwork & Wood Preserving (12)
16-3
-------
Table 16-1. Geographical Information for the Minnesota Monitoring Site
Site
Code
STMN
AQS Code
27-145-3053
Location
St. Cloud
County
Stearns
Micro- or
Metropolitan
Statistical Area
St. Cloud, MN
MSA
Latitude
and
Longitude
45.564637,
-94.226345
Land Use
Industrial
Location
Setting
Suburban
Additional Ambient Monitoring Information1
TSP, TSP Metals.
Data for additional pollutants are reported to AQS for STMN (EPA, 2013b); however, these data are not generated by ERG and are therefore not included in this report.
-------
The STMN monitoring site is located on the property of Grede Foundries, St. Cloud, Inc.,
on the west side of St. Cloud, Minnesota, just north of the Waite Park town limits. Monitoring at
this site is source-oriented and part of a special assessment initiated based on elevated total
chromium levels (MPCA, 2013). An apartment complex and mobile home park are separated
from additional industrial properties, including a stainless steel tank manufacturing facility, by
54th Avenue North just west of the site. Farther west, the Sauk River runs northeast-southwest
through the area and is adjacent to additional residential properties to the north and northwest of
the site. A railway runs east-west to the south of the site with commercial properties immediately
adjacent.
Figure 16-2 shows that the monitoring site is located in close proximity to many
emissions sources. The source categories with the greatest number of emissions sources near
STMN include woodworking, institutions (which include schools, prisons, and hospitals),
wastewater treatment, and mine/quarry/mineral processing. The sources located to the east and
along the county boundary are located near the banks of the Mississippi River. The STMN site is
located in a highly industrial area, which includes a major hospital to the northeast, a metals
processing and fabrication facility, and foundry, iron, and steel facility, and an industrial
machinery/equipment plant. Additional facilities are located to the southwest of STMN.
Table 16-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Minnesota monitoring site. Table 16-2 includes both county-level
population and vehicle registration information. Table 16-2 also contains traffic volume
information for STMN as well as the location for which the traffic volume was obtained.
Additionally, Table 16-2 presents the county-level daily VMT for Stearns County.
Table 16-2. Population, Motor Vehicle, and Traffic Information for the Minnesota
Monitoring Site
Site
STMN
Estimated
County
Population1
151,606
County-level
Vehicle
Registration2
218,196
Annual
Average
Daily Traffic3
24,100
Intersection
Used for
Traffic Data
8th Street N, east of Anderson Ave
County-
level Daily
VMT4
4,983,115
Bounty-level population estimate reflects 2012 data (Census Bureau, 2013c)
2County-level vehicle registration reflects 2012 data (MN DPS, 2013)
3AADT reflects 2009 data (MN DOT, 2009)
4County-level VMT reflects 2012 data (MN DOT, 2013)
16-5
-------
Observations from Table 16-2 include the following:
• The Stearns County population is in the bottom-third compared to other counties with
NMP sites. The county-level vehicle registration has a similar ranking compared to
other counties with NMP sites.
• The traffic volume near STMN is in the middle of the range compared to other NMP
sites. The traffic estimate provided is for 8th Street North (Veterans Drive), east of
Anderson Avenue.
• The daily VMT for Stearns County is nearly 5 million miles and ranks 30th compared
to other counties with NMP sites (where VMT data were available).
16.2 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
site in Minnesota on sample days, as well as over the course of the year.
16.2.1 Climate Summary
The city of St. Cloud is located roughly in the center of the state of Minnesota. The area
experiences a continental climate, with summers characterized by warm days and cool nights and
winters that are long and cold. Annual precipitation is around 30 inches with more than half of
the precipitation concentrated between May and September and in the form of thunderstorms.
Nearly 50 inches of snow falls on average during the winter months. A northwest wind is
predominant in St. Cloud most of the year, although a southerly wind occurs during the summer
months (NCDC, 2014; MCWG, 2013).
16.2.2 Meteorological Summary
Hourly meteorological data for 2012 were retrieved from NCDC for the weather station
closest to the Minnesota monitoring site (NCDC, 2012), as described in Section 3.5.2. The
closest weather station is located at St. Cloud Regional Airport (WBAN 14926). Additional
information about the St. Cloud Regional Airport weather station, such as the distance between
the site and the weather station, is provided in Table 16-3. These data were used to determine
how meteorological conditions on sample days vary from conditions experienced throughout the
year.
16-6
-------
Table 16-3. Average Meteorological Conditions near the Minnesota Monitoring Site
Closest
Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
St. Cloud, Minnesota - STMN
St. Cloud
Regional
Airport
14926
(45.54, -94.05)
8.1
miles
99°
(ESE)
Sample
Days
(55)
2012
60.1
±6.0
57.6
+ 2.3
49.8
±5.5
47.6
+ 2.1
38.0
±5.1
36.0
+ 1.9
44.1
±4.9
42.1
+ 1.9
68.0
±3.5
68.1
+ 1.3
1015.3
±1.9
1015.6
±0.7
6.5
±0.8
6.6
+ 0.3
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
-------
Table 16-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for all of 2012. Also included in Table 16-3 is the 95 percent
confidence interval for each parameter. As shown in Table 16-3, average meteorological
conditions on sample days appear warmer than conditions experienced throughout 2012,
although the differences are not statistically significant. Sampling at STMN under the NMP did
not begin until February 2012, thereby missing the coldest month of the year.
16.2.3 Back Trajectory Analysis
Figure 16-3 is the composite back trajectory map for days on which samples were
collected at the STMN monitoring site. Included in Figure 16-3 are four back trajectories per
sample day. Figure 16-4 is the corresponding cluster analysis. An in-depth description of these
maps and how they were generated is presented in Section 3.5.2.1. For the composite map, each
line represents the 24-hour trajectory along which a parcel of air traveled toward the monitoring
site on a given sample day and time, based on an initial height of 50 meters AGL. For the cluster
analysis, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric circle around the site in Figures 16-3 and 16-4 represents 100 miles.
16-8
-------
Figure 16-3. Composite Back Trajectory Map for STMN
Figure 16-4. Back Trajectory Cluster Map for STMN
16-9
-------
Observations from Figures 16-3 and 16-4 for STMN include the following:
• Back trajectories originated from a variety of directions at STMN, although many of
the back trajectories originate from the northwest. The longest back trajectories
originated to the northwest of STMN.
• The 24-hour air shed domain for STMN is similar in size to other NMP monitoring
sites. The farthest away a back trajectory originated was Alberta, Canada, or greater
than 800 miles away. However, the average back trajectory length was 245 miles and
most back trajectories (90 percent) originated within 400 miles of the site.
• The cluster analysis shows that nearly 50 percent of the back trajectories originated
from the northwest of STMN, although the HYSPLIT model split them into two
clusters based length (43 percent originated over North Dakota, Manitoba, Canada,
and the northwest portion of Minnesota; the remaining 5 percent originated farther
west, over Montana or Saskatchewan and Alberta, Canada. Seventeen percent of back
trajectories originated to the southeast, south, or southwest of STMN, primarily over
Iowa, Nebraska, and South Dakota. The short cluster trajectory originating toward
Minneapolis represents relatively short back trajectories (generally less than
200 miles) originating from all quadrants except the northwest. These back
trajectories originated over northeast Iowa, the western half of Wisconsin, and
Minnesota.
16.2.4 Wind Rose Comparison
Hourly surface wind data from the weather station at St. Cloud Regional Airport near
STMN were uploaded into a wind rose software program to produce customized wind roses, as
described in Section 3.5.2.2. 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 16-5 presents a map showing the distance between the weather station and STMN,
which may be useful for identifying topographical influences that can affect the meteorological
patterns experienced at this location. Figure 16-5 also presents three different wind roses for the
STMN monitoring site. First, a historical wind rose representing 2002 to 2011 wind data is
presented, which shows the predominant surface wind speed and direction over an extended
period of time. Second, a wind rose representing wind observations for all of 2012 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2012 is
presented. These can be used to identify the predominant wind speed and direction for 2012 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.
16-10
-------
Figure 16-5. Wind Roses for the St. Cloud Regional Airport Weather Station near STMN
Location of STMN and Weather Station
2002-2011 Historical Wind Rose
Calms: 1620%
2012 Wind Rose
Sample Day Wind Rose
Calms: 15.75%
16-11
-------
Observations from Figure 16-5 for STMN include the following:
• The St. Cloud Regional Airport weather station is located approximately 8 miles east
of STMN. Most of the city of St. Cloud and the Mississippi River lie between the site
and the weather station. The area surrounding the airport is more rural in nature than
the more urbanized area surrounding STMN.
• The historical wind rose shows that winds from the northwest quadrant (including
west and north) and southeast quadrant (including east and south) were observed
more frequently than winds from the northeast or southwest. Winds from these
quadrants account for approximately one-third of observations. The strongest wind
speeds were most often associated with westerly to northwesterly winds. Calm winds
(<2 knots) were observed for 15 percent of the hourly measurements.
• The wind patterns shown on the 2012 wind rose resemble the historical wind patterns.
• The sample day wind rose exhibits some of the characteristics of the other wind roses,
with winds from the northwest and southeast quadrants accounting for the majority of
observations, but the individual direction percentages are more variable. There were
more wind observations from the east-northeast and south-southwest on sample days
and fewer wind observations from the east-southeast and west-northwest.
• The percentage of calm winds, however, is similar across all three wind roses.
16.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for STMN 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 16-4. 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 16-4. It is important to note which
pollutants were sampled for at each site when reviewing the results of this analysis. STMN
sampled only hexavalent chromium.
16-12
-------
Table 16-4. Risk-Based Screening Results for the Minnesota Monitoring Site
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
St. Cloud, Minnesota - STMN
Hexavalent Chromium
0.000083
Total
6
6
39
39
15.38
15.38
100.00
100.00
Table 16-4 presents the results of the preliminary risk-based screening process for
STMN. Observations from Table 16-4 include the following:
• Hexavalent chromium was detected in 39 of the 54 valid samples collected at STMN.
• Hexavalent chromium failed six screens for STMN, which represents a roughly
15 percent failure rate.
16.4 Concentrations
This section presents various concentration averages used to characterize pollution levels
at the Minnesota 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 the 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 the site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Additional site-specific statistical summaries for STMN
are provided in Appendix O.
16.4.1 2012 Concentration Averages
Quarterly and annual concentration averages were calculated for the pollutants of interest
for the Minnesota site, as described in Section 3.1. The quarterly average 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
16-13
-------
number of samples possible within a given quarter for a quarterly average to be calculated. An
annual average 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
STMN are presented in Table 16-5, 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 16-5. Quarterly and Annual Average Concentrations of the Pollutants of
Interest for the Minnesota Monitoring Site
Pollutant
#of
Measured
Detections
vs. # 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)
St. Cloud, Minnesota - STMN
Hexavalent Chromium
39/54
NA
0.757
±1.177
0.026
±0.011
0.022
±0.017
0.283
±0.337
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
Observations for STMN from Table 16-5 include the following:
• Measured detections of hexavalent chromium span three orders of magnitude, ranging
from 0.0044 ng/m3 to 8.51 ng/m3. This dataset also includes 15 non-detects.
• The maximum concentration measured at STMN is the single highest hexavalent
chromium concentration measured under the NMP since this method was added to the
program in 2005. This measurement was collected on May 9, 2012. Two additional
measurements greater than 1 ng/m3 were measured at STMN, one on
February 27, 2012 (3.07 ng/m3) and one on May 15, 2012 (2.15 ng/m3). These too are
among the highest measurements of hexavalent chromium collected program-wide. In
total, five hexavalent chromium measurements greater than 1 ng/m3 have been
collected under the NMP between 2005 and 2012.
• The fourth highest hexavalent chromium concentration measured at STMN is an
order of magnitude less than the others (0.331 ng/m3). Only three measurements
collected at STMN fall between 0.1 ng/m3 and 1 ng/m3.
• The second quarter average concentration is significantly higher than the other
quarter quarterly averages and the associated confidence interval is greater than the
average itself. The second quarter data set includes the two May concentrations
discussed above, two of the three measurements between 0.1 ng/m3 and 1 ng/m3, as
16-14
-------
well as four non-detects (zeros). This explains the variability associated with the
second quarter average concentration as well as the annual average.
• Because sampling did not begin until February 9, 2012, a first quarter average
concentration could not be calculated.
16.4.2 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, a hexavalent chromium box plot was created for
STMN. Figure 16-6 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 3.5.3.1.
Figure 16-6. Program vs. Site-Specific Average Hexavalent Chromium Concentration
i
] Program Max Concentration = 8.51 ng/m
0.2 0.3
Concentration (ng/m3)
Program:
Site:
1st Quartile
D
Site Average
0
2nd Quartile 3rd Quartile
D D
Site Concentration Range
4th Quartile Average
D 1
Observations from Figure 16-6 include the following:
• The program-level maximum concentration (8.51 ng/m3) is not shown directly on
the box plot in Figure 16-6 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 0.5 ng/m3. In addition, the program-level
first quartile is zero and therefore not visible on the box plot.
• The maximum hexavalent chromium concentration measured at STMN is the
maximum concentration measured across the program.
• The annual average for STMN is greater than the program-level first, second, and
third quartiles and is an order of magnitude greater than the program-level
average concentration.
16-15
-------
• The annual average concentration of hexavalent chromium for STMN is the
highest annual average concentration of this pollutant calculated among all NMP
sites sampling hexavalent chromium. STMN is one of only two sites with an
annual average concentration greater than 0.1 ng/m3.
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 3.5.3.2.
Because sampling under the NMP did not begin until February 2012 at STMN, a trends analysis
was not conducted for this site.
16.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk at the
Minnesota monitoring site. Refer to Sections 3.3 and 3.5.3 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.
16.5.1 Risk-Based Screening Assessment Using MRLs
A risk-based screening was conducted by comparing the concentration data from STMN
to the ATSDRMRLs, where available. As described in Section 3.3, MRLs are noncancer health
risk benchmarks and are defined for three exposure periods: acute (exposures of 1 day to
14 days); intermediate (exposures of 15 days to 364 days); and chronic (exposures of 1 year or
greater). The preprocessed daily measurements of the pollutants of interest were compared to the
acute MRLs; the quarterly averages were compared to the intermediate MRLs; and the annual
average was compared to the chronic MRLs.
As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2012.
Note that hexavalent chromium has an intermediate MRL (0.3 |ig/m3) only.
16-16
-------
16.5.2 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Minnesota 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 pollutant 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 3.5.3.4 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 16-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Table 16-6. Risk Approximations for the Minnesota Monitoring Site
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
St. Cloud, Minnesota - STMN
Hexavalent Chromium
0.012
0.0001
39/54
0.28
±0.34
3.40
0.01
Observations for STMN from Table 16-6 include the following:
• The annual average concentration of hexavalent chromium for STMN is the highest
annual average concentration of this pollutant program-wide.
• The cancer risk approximation for STMN for hexavalent chromium is 3.40 in-a-
million, the highest cancer risk approximation calculated for this pollutant across the
program.
• The noncancer hazard approximation for hexavalent chromium is less than 0.01,
indicating that no adverse noncancer health effects are expected from this individual
pollutant.
16-17
-------
16.5.3 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 16-7 presents the 10 pollutants with the highest emissions from the 2011 NEI that have
cancer toxicity factors. Table 16-7 also presents the 10 pollutants with the highest toxicity -
weighted emissions, based on the weighting schema described in Section 3.5.3.5. Lastly,
Table 16-7 provides the pollutants with the highest cancer risk approximations (in-a-million) for
the site, as presented in Table 16-6. The emissions, toxicity-weighted emissions, and cancer risk
approximations are shown in descending order in Table 16-7. Table 16-8 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 3.5.3.5. Similar to the cancer risk and
noncancer hazard approximations provided in Section 16.5.2, this analysis may help policy-
makers prioritize their air monitoring activities.
16-18
-------
Table 16-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Minnesota Monitoring Site
Top 10 Total Emissions for Pollutants with Cancer
UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations
Based on Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
St. Cloud, Minnesota (Stearns County) - STMN
Bis(2-ethylhexyl)phthalate (DEHP), gas1
Formaldehyde
Benzene
Acetaldehyde
Ethylbenzene
1,3 -Butadiene
Naphthalene
1 ,3 -Dichloropropene
Tetrachloroethylene
Dichloromethane
699.98
197.54
179.92
109.50
61.53
28.02
16.18
12.08
4.76
4.43
Formaldehyde
Bis(2-ethylhexyl)phthalate (DEHP), gas1
Benzene
1,3 -Butadiene
Naphthalene
Acetaldehyde
POM, Group 2b
POM, Group 5a
POM, Group 2d
POM, Group 3
2.57E-03
1.68E-03
1.40E-03
8.41E-04
5.50E-04
2.41E-04
2.38E-04
2.18E-04
1.79E-04
1.66E-04
Hexavalent Chromium 3.40
1 EPA has reviewed the reported emissions of this pollutant and has revised these emissions in version 2 of the 2011 NEI; however, version 1, which is the version
cited in this report, is the only version of the 2011 NEI publically available at the time of publication.
-------
Table 16-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Minnesota Monitoring Site
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard
Approximations Based on Annual Average
Concentrations
(Site-Specific)
Noncancer
Hazard
Approximation
Pollutant (HQ)
St. Cloud, Minnesota (Stearns County) - STMN
Toluene
Bis(2-ethylhexyl)phthalate (DEHP), gas1
Xylenes
Hexane
Ethylene glycol
Formaldehyde
Benzene
Acetaldehyde
Methanol
Ethylbenzene
796.08
699.98
304.31
206.81
199.63
197.54
179.92
109.50
93.71
61.53
Acrolein
Bis(2-ethylhexyl)phthalate (DEHP), gas1
Manganese, PM
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Lead, PM
Benzene
Naphthalene
Chlorine
390,174.17
69,998.10
21,950.40
20,156.75
14,010.08
12,166.99
6,189.57
5,997.44
5,391.85
5,240.71
Hexavalent Chromium O.01
to
o
1 EPA has reviewed the reported emissions of this pollutant and has revised these emissions in version 2 of the 2011 NEI; however, version 1, which is the version
cited in this report, is the only version of the 2011 NEI publically available at the time of publication.
-------
Observations from Table 16-7 include the following:
• Bis(2-ethylhexyl)phthalate (DEHP) gas, formaldehyde, and benzene are the highest
emitted pollutants with cancer UREs in Stearns County. The emissions of
bis(2-ethylhexyl)phthalate (DEHP) gas in Stearns County are two orders of
magnitude higher than the emissions of this pollutant for any other county with an
NMP site.
• Formaldehyde, bis(2-ethylhexyl)phthalate (DEHP) gas, and benzene are the
pollutants with the highest toxi city-weighted emissions (of the pollutants with cancer
UREs) for Stearns County.
• Six of the highest emitted pollutants in Stearns County also have the highest toxi city-
weighted emissions.
• Hexavalent chromium, which is the only pollutant sampled for at STMN, is not
among the highest emitted pollutants or those with the highest toxicity-weighted
emissions. Hexavalent chromium ranks 29th for total emissions and 12th for toxicity-
weighted emissions.
• Naphthalene and several POM Groups rank among Stearns County's highest toxi city-
weighted emissions. PAHs were not sampled for at STMN.
Observations from Table 16-8 include the following:
• Toluene, bis(2-ethylhexyl)phthalate (DEHP) gas, and xylenes are the highest emitted
pollutants with noncancer RfCs in Stearns County.
• The pollutants with the highest toxi city-weighted emissions (of the pollutants with
noncancer RfCs) are acrolein, bis(2-ethylhexyl)phthalate (DEHP) gas, and
manganese.
• Four of the highest emitted pollutants in Stearns County also have the highest
toxicity-weighted emissions.
• Again, hexavalent chromium does not appear among the pollutants with the highest
emissions or toxicity-weighted emissions. This pollutant's emissions rank 58th and its
toxicity-weighted emissions rank 31st (among the pollutants with noncancer RfCs).
16.6 Summary of the 2012 Monitoring Data for STMN
Results from several of the data treatments described in this section include the
following:
»«» Hexavalent chromium was the only pollutant sampled for at STMN.
»«» Concentrations of hexavalent chromium measured at STMN range from 0.0044 ng/m
to 8.51 ng/m3.
16-21
-------
The maximum concentration measured at STMN is the single highest hexavalent
chromium concentration measured under the NMP since this method was added to
the program in 2005.
16-22
-------
17.0 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.
17.1 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 MSA. Figure 17-1 is a
composite satellite image retrieved from ArcGIS Explorer showing the 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. Note that only sources within
10 miles of the site 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 radius are still visible on the map, but have been grayed out in order to show emissions
sources just outside the boundary. Table 17-1 provides supplemental geographical information
such as land use, location setting, and locational coordinates.
17-1
-------
Figure 17-1. St. Louis, Missouri (S4MO) Monitoring Site
to
-------
Figure 17-2. NEI Point Sources Located Within 10 Miles of S4MO
90'15'0"W 90"10'0"W 90"5'0"W WO'O'W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
S4MO NATTS site O 10 mile radius
County boundary
Source Category Group (No. of Facilities)
t
i
Ajrport/Airline/Airport Support Operations (22)
Asphalt Production/Hot Mix Asphall Plant (6)
y Breweries/Distilleries/Wineries (1)
§ Building/Construction (1)
B Bulk Terminals/Bulk Plants (9)
C Chemical Manufacturing (25)
^[ Coke Battery (2)
i Compressor Station (2)
[X] Crematory - Animal/Human (1)
££) Dry Cleaning (3)
G Electrical Equipment Manufacturing (1 ]
^ Electricity Generation via Combustion (5)
E Electroplating. Plating, Polishing. Anodizing, ana Coloring (3)
=)f= Ethanol Biorefinerles (1)
F Food p roc ess ingf Agriculture (8)
I Foundries, Iron and Steel (2)
A Foundries, Non-ferrous (1)
|f Gasoline/Diesel Service Station (2)
-j^- Industrial Machinery or Equipment Plant (3)
O institution (school, hospital, prison, etc.) (8)
• Landfill (3)
Q Leather and Leather Products (1)
• Metal Can, Box. and Other Melal Container Manufacturing (1)
A Metal Coating. Engraving, and Allied Services to Manufacturers
© Metals Processing/Fabrication (9)
^ Military Base/National Secunty (1)
X Mine/Quarry/Mineral Processing (20)
"> Miscellaneous Commercial/tndjstnal (34)
Q Pain! and Coating Manufacturing (6)
< Pesticide Manufacturing Plant (1)
$$• Petroleum Products Manufacturing (1)
d> Pharmaceutical Manufacturing (1)
R Plastic, Restn. or Rubber Products Plant (3)
y Port and Harbor Operations (6)
P Printing/Publishing/Paper Product Manufacturing (7)
X Rail Yard/Rail Line Operations (14)
J} Railroad Engines/Parts Manufacturing (1)
^T Steel Mill (2)
T Textile. Yam. orCarpet Plant (1)
rfV True fc/Bus/T ran spoliation Operations (1)
I Wastewater Treatment (7)
A Water Treatment (1)
17-3
-------
Table 17-1. Geographical Information for the Missouri Monitoring Site
Site
Code
S4MO
AQS Code
29-510-0085
Location
St. Louis
County
St. Louis
City
Micro- or
Metropolitan
Statistical Area
St. Louis, MO-IL
MSA
Latitude
and
Longitude
38.656449,
-90.198548
Land Use
Residential
Location
Setting
Urban/City
Center
Additional Ambient Monitoring Information1
TSP Lead, CO, SO2, NOy, NO, O3, Meteorological
parameters, PM10, PM Coarse, Black carbon, PM25,
PM2 5 Specialion, IMPROVE Specialion.
Data for additional pollutants are reported to AQS for this site (EPA, 2013b); however, these data are not generated by ERG and are therefore not included in this report.
BOL D ITALICS = EPA-de signaled NATTS Site
-------
S4MO is located in central St. Louis. Figure 17-1 shows that the S4MO monitoring site is
located less than 1/4 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 just on the other side of 1-70. Figure 17-2
shows that a large number of point sources are located within 10 miles of S4MO, particularly
east of the Missouri/Illinois border. The source categories with the greatest number of point
sources surrounding S4MO include mines, quarries, and mineral processing facilities; 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; 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.
Table 17-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Missouri monitoring site. Table 17-2 includes both county-level
population and vehicle registration information. Table 17-2 also contains traffic volume
information for S4MO as well as the location for which the traffic volume was obtained.
Additionally, Table 17-2 presents the county-level daily VMT for S4MO. Note that because the
state of Missouri provides data within the city of St. Louis separately from St. Louis County,
Table 17-2 includes the combination of the city and county data for county-level statistics in
order to compare these statistics with other sites' county-level data.
Table 17-2. Population, Motor Vehicle, and Traffic Information for the Missouri
Monitoring Site
Site
S4MO
Estimated
County
Population1
1,318,610
County-level
Vehicle
Registration2
1,112,866
Annual
Average
Daily Traffic3
79,558
Intersection
Used for
Traffic Data
1-70 near Exit 249
County-
level Daily
VMT4
23,994,911
Bounty-level population estimate reflects county and city data for 2012 (Census Bureau, 2013c)
2Vehicle registration reflects county and city data for 2012 (MO DOR, 2013)
3AADT reflects 2011 data (MO DOT, 2012)
4VMT reflects county and city data for 2012 (MO DOT, 2013)
BOLD ITALICS = EPA-designated NATTS Site
17-5
-------
Observations from Table 17-2 include the following:
• S4MO's county-level population and vehicle registration both rank 10th compared to
other counties with NMP sites.
• The traffic volume experienced near S4MO is in the middle of the range compared to
other NMP sites. The traffic estimate provided is for 1-70 near Exit 249.
• The VMT for S4MO ranks 12th among counties with NMP sites (where VMT data
were available).
17.2 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.
17.2.1 Climate Summary
The city of St. Louis is located along the Mississippi River, which acts as Missouri's
eastern border. St. Louis has a climate that is continental in nature, with cold, dry winters; warm,
muggy summers; and significant seasonal variability. Warm, moist air flowing northward from
the Gulf of Mexico alternates with cold, dry air marching southward from Canada and the
northern U.S., resulting in weather patterns that do not persist for very long. Southerly winds
prevail during the warmer months of the year, while west-northwesterly winds prevail the rest of
the year. Thunderstorms are common, particularly in the spring, summer, and fall, and annual
snowfall totals average around 20 inches. The city of St. Louis experiences the urban heat island
effect, retaining more heat within the city than outlying areas (Wood, 2004 and MCC, 2014).
17.2.2 Meteorological Summary
Hourly meteorological data for 2012 were retrieved from NCDC for the station closest to
the Missouri monitoring site (NCDC, 2012), as described in Section 3.5.2. The closest weather
station to S4MO is located at St. Louis Downtown Airport (WBAN 03960). Additional
information about this weather station, such as the distance between the site and the weather
station, is provided in Table 17-3. These data were used to determine how meteorological
conditions on sample days vary from conditions experienced throughout the year.
17-6
-------
Table 17-3. Average Meteorological Conditions near the Missouri Monitoring Site
Closest
Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
From Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
St. Louis, Missouri - S4MO
St. Louis
Downtown
Airport
03960
(38.57, -90.16)
6.2
miles
157°
(SSE)
Sample
Days
(65)
2012
68.4
±4.5
70.2
±2.0
58.1
±4.1
59.4
±1.8
44.8
±3.9
45.9
±1.6
51.2
±3.6
52.2
±1.5
64.8
±3.0
64.6
±1.2
1016.8
±1.5
1016.4
±0.6
5.5
±0.7
5.5
±0.3
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
-------
Table 17-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for all of 2012. Also included in Table 17-3 is the 95 percent
confidence interval for each parameter. Although average meteorological conditions on sample
days are not statistically different than the average meteorological conditions experienced
throughout 2012, the temperatures do appear slightly cooler on sample days, as shown in
Table 17-3. This is likely the result of the inclusion of dates for make-up samples, which were
collected during the colder months of the year (one in February and three in December 2012).
17.2.3 Back Trajectory Analysis
Figure 17-3 is the composite back trajectory map for days on which samples were
collected at the S4MO monitoring site. Included in Figure 17-3 are four back trajectories per
sample day. Figure 17-4 is the corresponding cluster analysis. An in-depth description of these
maps and how they were generated is presented in Section 3.5.2.1. For the composite map, each
line represents the 24-hour trajectory along which a parcel of air traveled toward the monitoring
site on a given sample day and time, based on an initial height of 50 meters AGL. For the cluster
analysis, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric circle around the site in Figures 17-3 and 17-4 represents 100 miles.
17-8
-------
Figure 17-3. Composite Back Trajectory Map for S4MO
Figure 17-4. Back Trajectory Cluster Map for S4MO
17-9
-------
Observations from Figures 17-3 and 17-4 for S4MO include the following:
• Back trajectories originated from a variety of directions at S4MO, although few back
trajectories originated from due east or west. The longest back trajectories originated
from the northwest.
• The farthest away a back trajectory originated was nearly 900 miles away from
S4MO, over western North Dakota. This was the third longest back trajectory
constructed for the 2012 NMP report (the other two were constructed for SSSD).
However, the 24-hour air shed domain for S4MO is similar in size to other NMP
sites, as the average back trajectory length was 251 miles and most back trajectories
(88 percent) originated within 400 miles of the monitoring site.
• The cluster analysis shows that nearly 50 percent back trajectories originated to the
northwest and north of S4MO, although the model split these into three different
clusters. Twenty-five percent originated less than 200 miles away from the site, over
the northern half Missouri and central Illinois. Another 18 percent of back trajectories
originated farther away, primarily over Iowa. The longest back trajectories originated
to the northwest over the Dakotas and Nebraska and account for another 4 percent of
back trajectories.
• Nine percent of back trajectories originated to the southwest and west of S4MO,
primarily over the western half of Arkansas, but includes the two longer back
trajectories that originated over Kansas and the panhandle of Oklahoma. Nearly one
third of back trajectories originated to the southeast and south of the site, over eastern
Arkansas, Mississippi, Alabama, Tennessee, and Kentucky. The cluster trajectory
originating to the northeast of S4MO (13 percent) represents relatively short back
trajectories originating over the southernmost portion of Indiana and Illinois, longer
back trajectories originating to the northeast of S4MO as far away as Lake Michigan,
and all back trajectories in between.
17.2.4 Wind Rose Comparison
Hourly surface wind data from the weather station at St. Louis Downtown Airport near
S4MO were uploaded into a wind rose software program to produce customized wind roses, as
described in Section 3.5.2.2. 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 a map showing the distance between the weather station and S4MO,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 17-5 also presents three different wind roses for the
S4MO monitoring site. First, a historical wind rose representing 2002 to 2011 wind data is
presented, which shows the predominant surface wind speed and direction over an extended
period of time. Second, a wind rose representing wind observations for all of 2012 is presented.
17-10
-------
Next, a wind rose representing wind data for days on which samples were collected in 2012 is
presented. These can be used to identify the predominant wind speed and direction for 2012 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.
Observations from Figure 17-5 for S4MO include the following:
• The St. Louis Downtown Airport weather station is located approximately 6 miles
south-southeast of S4MO. The weather station location is across the Mississippi River
and state border in Illinois.
• The historical wind rose shows that winds from the southeast, south-southeast, and
south were frequently observed near S4MO. Winds from these directions account for
approximately 28 percent of observations. Calm winds (<2 knots) were observed for
approximately 19 percent of the hourly wind measurements. Winds from the west to
northwest to north account for the majority of the remaining wind observations. The
strongest winds were from the west to northwest.
• The wind patterns shown on the 2012 wind rose generally resemble those shown on
the historical wind rose, although there were fewer southeasterly winds and more
south-southeasterly and southerly winds in 2012. The percentage of calm winds was
also higher (23 percent) in 2012.
• The primary wind directions on the sample day wind rose resemble the primary wind
directions on the historical and full-year wind roses, while the percentages for the
secondary wind directions are more variable.
17-11
-------
Figure 17-5. Wind Roses for the St. Louis Downtown Airport Weather Station near S4MO
Location of S4MO and Weather Station
2002-2011 Historical Wind Rose
2012 Wind Rose
Sample Day Wind Rose
Calms: 23.04%
Calms; 2213%
17-12
-------
17.3 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 17-4.
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 17-4. It is
important to note which pollutants were sampled for at each site when reviewing the results of
this analysis. S4MO sampled for VOCs, PAHs, carbonyl compounds, metals (PMi0), and
hexavalent chromium.
Observations from Table 17-4 include the following:
• Twenty-four pollutants failed at least one screen for S4MO; 56 percent of
concentrations for these 24 pollutants were greater than their associated risk screening
value (or failed screens).
• Seventeen pollutants contributed to 95 percent of failed screens for S4MO and
therefore were identified as pollutants of interest for this site. These 17 include two
carbonyl compounds, seven VOCs, five PMi0 metals, and three PAHs.
• S4MO failed the highest number of screens (692) among all NMP sites (refer to
Table 4-8 of Section 4.2). However, the failure rate for S4MO, when incorporating all
pollutants with screening values, is approximately 25 percent. This is due primarily to
the relatively high number of pollutants sampled for at this site, as discussed in
Section 4.2.
• Acetaldehyde, formaldehyde, benzene, carbon tetrachloride, and 1,2-dichloroethane
failed 100 percent of screens for S4MO and were detected in all or most of the
samples collected. 1,2-Dibromoethane and 1,1,2,2-tetrachloroethane also failed 100
percent of screens but were detected less frequently.
17-13
-------
Table 17-4. Risk-Based Screening Results for the Missouri Monitoring Site
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
St. Louis, Missouri - S4MO
Acetaldehyde
Formaldehyde
Arsenic (PM10)
Benzene
Carbon Tetrachloride
1,3 -Butadiene
Naphthalene
1 ,2-Dichloroethane
Manganese (PM10)
/>-Dichlorobenzene
Cadmium (PM10)
Ethylbenzene
Lead (PM10)
Fluorene
Acenaphthene
Hexachloro- 1 ,3 -butadiene
Nickel (PM10)
1 ,2-Dibromoethane
Hexavalent Chromium
1 , 1 ,2,2-Tetrachloroethane
Benzo(a)pyrene
Fluoranthene
Trichloroethylene
Propionaldehyde
0.45
0.077
0.00023
0.13
0.17
0.03
0.029
0.038
0.005
0.091
0.00056
0.4
0.015
0.011
0.011
0.045
0.0021
0.0017
0.000083
0.017
0.00057
0.011
0.2
0.8
Total
61
61
60
58
58
57
57
56
55
30
20
17
17
16
15
11
10
8
7
7
4
3
3
1
692
61
61
61
58
58
58
60
56
61
52
61
58
61
60
60
14
61
8
55
7
60
60
29
61
1,241
100.00
100.00
98.36
100.00
100.00
98.28
95.00
100.00
90.16
57.69
32.79
29.31
27.87
26.67
25.00
78.57
16.39
100.00
12.73
100.00
6.67
5.00
10.34
1.64
55.76
8.82
8.82
8.67
8.38
8.38
8.24
8.24
8.09
7.95
4.34
2.89
2.46
2.46
2.31
2.17
.59
.45
.16
.01
.01
0.58
0.43
0.43
0.14
8.82
17.63
26.30
34.68
43.06
51.30
59.54
67.63
75.58
79.91
82.80
85.26
87.72
90.03
92.20
93.79
95.23
96.39
97.40
98.41
98.99
99.42
99.86
100.00
17.4 Concentrations
This section presents various concentration averages used to characterize pollution 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
S4MO.
• Annual average concentrations are presented graphically 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 the site.
17-14
-------
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Additional site-specific statistical summaries for S4MO
are provided in Appendices J, L, M, N, and O.
17.4.1 2012 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 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 includes all measured detections and substituted 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 S4MO
are presented in Table 17-5, 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.
17-15
-------
Table 17-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Missouri Monitoring Site
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Ug/m3)
2nd
Quarter
Average
(Ug/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Ug/m3)
St. Louis, Missouri - S4MO
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
£>-Dichlorobenzene
1,2-Dichloroethane
Ethylbenzene
Formaldehyde
Hexachloro- 1 ,3 -butadiene
Acenaphthene3
Arsenic (PM10)a
Cadmium (PM10)a
Fluorene3
Lead (PM10)a
Manganese (PM10)a
Naphthalene3
Nickel (PM10)a
61/61
58/58
58/58
58/58
52/58
56/58
58/58
61/61
14/58
60/60
61/61
61/61
60/60
61/61
61/61
60/60
61/61
1.30
±0.28
0.85
±0.14
0.11
±0.03
0.64
±0.07
0.13
±0.08
0.10
±0.01
0.35
±0.10
1.84
±0.37
0.04
±0.03
4.27
±2.01
0.99
±0.37
0.38
±0.17
4.82
± 1.74
12.89
±8.01
18.94
±10.91
112.42
±42.88
1.31
±0.49
1.88
±0.41
0.64
±0.13
0.09
±0.04
0.72
±0.05
0.20
±0.14
0.09
±0.01
0.38
±0.13
3.71
±0.94
0.02
±0.02
9.11
±3.07
1.01
±0.33
0.37
±0.20
10.58
±2.99
11.44
±7.45
19.63
±7.24
103.15
±38.87
1.46
±0.59
2.57
±0.44
0.87
±0.40
0.11
±0.04
0.67
±0.03
0.21
±0.17
0.06
±0.01
0.33
±0.06
5.17
±1.20
0.02
±0.02
12.23
±3.71
1.02
±0.24
0.97
±0.45
13.86
±4.10
12.59
±4.17
21.64
± 13.25
116.58
±30.37
1.50
±0.46
1.62
±0.35
0.84
±0.21
0.13
±0.05
0.68
±0.03
0.18
±0.11
0.08
±0.01
0.37
±0.13
2.17
±0.39
0
3.87
±1.99
1.34
±0.90
0.53
±0.33
4.01
±1.30
16.45
± 14.32
30.51
±36.51
109.63
±52.51
1.40
±1.25
1.86
±0.21
0.80
±0.12
0.11
±0.02
0.68
±0.02
0.18
±0.06
0.08
±0.01
0.35
±0.05
3.26
±0.52
0.02
±0.01
7.37
±1.59
1.09
±0.25
0.57
±0.16
8.32
±1.67
13.33
±4.31
22.66
±9.60
110.45
± 19.71
1.42
±0.36
' Average concentrations provided for the pollutants below the blue line are presented in ng/m for ease of viewing.
17-16
-------
Observations for S4MO from Table 17-5 include the following:
• The pollutants with the highest annual average concentrations are formaldehyde
(3.26 ± 0.52 |ig/m3) and acetaldehyde (1.86 ± 0.21 |ig/m3). These are the only
pollutants of interest with annual averages greater than 1 |ig/m3.
• The second and third quarter average concentrations of formaldehyde are
significantly higher than the first and fourth quarter averages and have larger
confidence intervals. Concentrations of formaldehyde measured at S4MO ranged
from 1.01 |ig/m3to 11.8 |ig/m3. The 20 highest concentrations were measured
between May and September, or the warmest months of the year. Conversely, the four
lowest measurements of formaldehyde were all measured in January and February
and all but three of the 22 measurements less than 2 |ig/m3 were measured during the
first or fourth quarters of 2012. A similar observation can be made for acetaldehyde
but the seasonality is less pronounced.
• The confidence intervals associated with the quarterly averages ofp-dichlorobenzene
are relatively large compared to the averages themselves. A review of the data shows
that concentrations of />-dichlorobenzene span two orders of magnitude, ranging from
0.0241 |ig/m3to 1.37 |ig/m3, as well as six non-detects. The maximum
/>-dichlorobenzene concentration was measured in September and is the only
/>-dichlorobenzene concentration greater than 1 |ig/m3 measured across the program.
Six of the 13 />-dichlorobenzene measurements greater than 0.5 |ig/m3 were measured
at S4MO, with the others being measured at ADOK (3) and SPAZ (4).
• The fourth quarter average concentration of hexachlor-1,3-butadiene is zero,
indicating that there were no measured detections this quarter. This pollutant was
detected in less than one-quarter of the samples collected. Half of the measured
detections were measured during the first quarter of 2012, with three measured during
the second quarter and four during the third quarter.
• Manganese has the highest annual average concentration (22.66 ± 9.60 ng/m3) among
the PMio metals measured at S4MO. The confidence intervals associated with the
quarterly averages for manganese are relatively large, indicating that there is a high
level of variability in the measurements. This is particularly true for the fourth
quarter, where the confidence interval is greater than the average itself, indicating the
likely presence of outliers. Concentrations of manganese measured at S4MO range
from 2.09 ng/m3 to 275 ng/m3, with a median concentration of 12.70 ng/m3. The
maximum concentration of manganese was measured at S4MO on October 12, 2012
and is the maximum concentration measured across the program, although similar
measurements were also collected at the two Tulsa, Oklahoma sites (which are
sampling TSP metals rather than PMio metals). The next highest concentration
measured at S4MO is still greater than 100 ng/m3 and was measured in August. Three
additional measurements greater than 50 ng/m3 were measured in January, March,
and December.
• The fourth quarter averages of lead, arsenic, and nickel also reflect a high level of
variability, based on the associated confidence intervals. Although the maximum
concentration of each of these metals was measured during the fourth quarter, they
17-17
-------
were not measured on the same day. The maximum lead concentration (111 ng/m3)
was measured at S4MO on December 23, 2012 and is more than twice the next
highest concentration measured at this site. This is the second highest lead
concentration measured across the program. The maximum arsenic concentration was
measured on November 17, 2012 (7.23 ng/m3) and is two and a half times greater
than the next highest concentration measured at this site. This too is the highest
arsenic concentration measured across the program. The maximum nickel
concentration (9.74 ng/m3) was measured at S4MO on December 5, 2012 and is more
than twice the next highest concentration measured at this site. While not the highest
nickel concentration measured across the program, it does rank in the top five.
• Naphthalene has the highest annual average concentration among the PAHs measured
at S4MO. The confidence intervals calculated for the quarterly averages of
naphthalene indicate that there is a high level of variability in the measurements.
Concentrations of naphthalene measured at S4MO range from 21.3 ng/m3 to
360 ng/m3 with a median concentration of 80.5 ng/m3.
• Concentrations of acenaphthene and fluorene appear to be highest during the warmer
months of the year, particularly the third quarter of 2012. The averages for these
quarters have relatively large confidence intervals associated with them. A review of
the data shows that the maximum concentration of each pollutant was measured on
July 2, 2012 (27.9 ng/m3 and 31.3 ng/m3, respectively). The July 2nd sample day is
also the same sample day the highest concentrations were measured in 2011. Of the
concentrations of each pollutant greater than 10 ng/m3, the majority were measured
during the third quarter, followed by the second quarter. The nine highest
concentrations of these pollutants were measured on the same nine sample days,
although the exact order varies. For fluorene, eight of the nine measurements greater
than 15 ng/m3 were measured between June and August, while all 10 of the
measurements less than 2.50 ng/m3 were measured in either the first or fourth quarter
of the year. Similarly, eight of the nine highest acenaphthene measurements were
measured at S4MO between June and August, while all 12 measurements less than
2.0 ng/m3 were measured in either the first or fourth quarter of the year.
• At the beginning of 2013, the Missouri Department of Natural Resources discovered
that a sampler contamination issue resulted in artificially elevated concentrations of
acrylonitrile from September 2010 through October 2012. Thus, the acrylonitrile
results from this time period were invalidated, which includes all of the results for
2012 through October 30, 2012. All acrylonitrile measurements for November and
December were non-detects.
Tables 4-9 through 4-12 present the 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 13 times, the most of any NMP
site.
17-18
-------
• S4MO has the highest annual average concentration of 1,2-dichloroethane, the second
highest annual average concentration of hexachloro-l,3-butadiene, and the third
highest annual average concentration ofp-dichlorobenzene. This site also has the 10th
highest annual average concentration of carbon tetrachloride and ethylbenzene.
• S4MO has the 10th highest annual average concentration of both acetaldehyde and
formaldehyde among NMP sites sampling carbonyl compounds.
• S4MO's annual average concentration of naphthalene ranks fourth highest among
NMP sites sampling PAHs, while this site's annual average concentrations of
acenaphthene and fluorene each rank fifth among NMP sites sampling PAHs.
• S4MO has the second highest annual average concentration of arsenic, second only to
ASKY-M, among NMP sites sampling PMio metals (and those sampling TSP metals).
This site's annual average concentration of manganese ranked third among NMP sites
sampling PMio metals, while its annual average concentration of nickel ranked fourth.
17.4.2 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 shaded in
gray in Table 17-4 for S4MO. Figures 17-6 through 17-22 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 3.5.3.1.
Figure 17-6. Program vs. Site-Specific Average Acenaphthene Concentration
S4MO
I Program Max Concentration = 182 ng/m3 j
10 20 30 40 50 60 70 80 90 100
Concentration (ng/m3)
Program: 1st Quartile 2nd Quartile 3rd Quartile 4th Quartile Average
• D D D I
Site: Site Average Site Concentration Range
17-19
-------
Figure 17-7. Program vs. Site-Specific Average Acetaldehyde Concentration
S4MO
9 12
Concentration {[og/m3)
15
18
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o —
21
Figure 17-8. Program vs. Site-Specific Average Arsenic (PMio) Concentration
45
Concentration (ng/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site:
Site Average Site Concentration Range
O
S4MO
Figure 17-9. Program vs. Site-Specific Average Benzene Concentration
Concentration {[og/m3;
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o —
17-20
-------
Figure 17-10. Program vs. Site-Specific Average 1,3-Butadiene Concentration
S4MO
Program Max Concentration = 4.10 |jg/n
0.75 1
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site:
Site Average Site Concentration Range
o
Figure 17-11. Program vs. Site-Specific Average Cadmium (PMio) Concentration
E
1.5 2
Concentration (ng/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: SiteAverage Site Concentration Range
O
Figure 17-12. Program vs. Site-Specific Average Carbon Tetrachloride Concentration
2 3
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: SiteAverage Site Concentration Range
o —
17-21
-------
Figure 17-13. Program vs. Site-Specific Average /7-Dichlorobenzene Concentration
S4MO
,
3
°
0.2 0.4 0.6 0
Concentration {[og/m3
8 1 1.2
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o —
1.
Figure 17-14. Program vs. Site-Specific Average 1,2-Dichloroethane Concentration
•H
r.
Program Max Concentration = 17.01 ug/m3
0.4 0.5 0.6
Concentration {[og/m3)
0.8 0.9
Program:
Site:
IstQuartile
D
Site Average
0
2ndQuartile SrdQuartile
• •
Site Concentration Range
^^^^—
4thQuartile Average
D 1
Figure 17-15. Program vs. Site-Specific Average Ethylbenzene Concentration
S4MO
•-
X
!
^^^^^^^
3 0.5 1 1.5 2 2.5 3 3.5 4
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile
D D D
4thQuartile
Average
Site: Site Average Site Concentration Range
o —
17-22
-------
17-16. Program vs. Site-Specific Average Fluorene Concentration
S4MO
10
20
30
40 50 60
Concentration (ng/m3)
70
80
90
Program:
Site:
IstQuartile
Site Average
0
2ndQuartile SrdQuartile
Site Concentration Range
^^^^—
4thQuartile Average
100
Figure 17-17. Program vs. Site-Specific Average Formaldehyde Concentration
S4MO
10
Concentration {[og/m3;
12
Program:
Site:
IstQuartile
D
Site Average
o
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
14
Figure 17-18. Program vs. Site-Specific Average Hexachloro-l,3-Butadiene Concentration
S4MO
0.1 0.15
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
O
17-23
-------
Figure 17-19. Program vs. Site-Specific Average Lead (PMi0) Concentration
S4MO
I
20
40 60 80
Concentration (ng/m3)
100
120
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: SiteAverage Site Concentration Range
o —
Figure 17-20. Program vs. Site-Specific Average Manganese (PMio) Concentration
E
Program Max Concentration = 275 ng/m3
60 90
Concentration (ng/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: SiteAverage Site Concentration Range
O
Figure 17-21. Program vs. Site-Specific Average Naphthalene Concentration
S4MO
100 200
300
400 500
Concentration (ng/m3)
600 700 800 900
Program:
Site:
IstQuartile
D
SiteAverage
O
2ndQuartile SrdQuartile
• D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
17-24
-------
Figure 17-22. Program vs. Site-Specific Average Nickel (PMi0) Concentration
S4MO
8 10
Concentration {ng/m3)
12
14
16
18
Program:
Site:
IstQuartile
D
Site Average
o
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^—
4th Qua rti 1 e Avera ge
D 1
Observations from Figures 17-6 through 17-22 include the following:
• Figure 17-6 is the box plot for acenaphthene. Note that the program-level
maximum concentration (182 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 to
100 ng/m3. Figure 17-6 shows that the maximum acenaphthene concentration
measured at S4MO is considerably less than the maximum concentration
measured at the program-level. Yet, the annual average concentration of
acenaphthene for S4MO is greater than the program-level average concentration.
• Figure 17-7 shows that the annual average acetaldehyde concentration for S4MO
is just greater than the program-level average concentration. The maximum
acetaldehyde concentration measured at S4MO is considerably less than the
maximum concentration measured across the program. The minimum
concentration measured at S4MO is among the higher minimum concentrations
among NMP sites sampling this pollutant.
• Figure 17-8 shows that the maximum arsenic (PMio) concentration measured at
S4MO is the maximum concentration measured across the program. S4MO's
annual average arsenic (PMio) concentration is greater than the program-level
average and third quartile. Recall from the previous section that this site has the
second highest annual average arsenic concentration among NMP sites sampling
metals.
• Figure 17-9 is the box plot for benzene and shows that the annual average
benzene concentration for S4MO is just less than the program-level average
concentration but greater than the program-level median concentration. The
maximum benzene concentration measured at S4MO is less than the maximum
concentration measured at the program level. There were no non-detects of
benzene measured at S4MO or across the program.
17-25
-------
• Similar to the acenaphthene graph, the program-level maximum 1,3-butadiene
concentration (4.10 |ig/m3) is not shown directly on the box plot as the scale has
been reduced to 2 |ig/m3 to allow for the observation of data points at the lower
end of the concentration range. Figure 17-10 for 1,3-butadiene shows that the
maximum concentration measured at S4MO is considerably less than the
maximum concentration measured across the program. The annual average
1,3-butadiene concentration for S4MO is roughly equivalent to the program-level
average concentration. While there were non-detects of 1,3-butadiene measured
across the program, there were none measured at S4MO.
• Figure 17-11 shows that the maximum cadmium (PMio) concentration measured
at S4MO is the maximum concentration measured across the program. Of the 25
concentrations greater than 1 ng/m3 measured across the program, 12 were
measured at S4MO. S4MO's annual average cadmium concentration is more than
three times higher than the program-level average concentration. The minimum
concentration measured at S4MO is just less than the program-level first quartile.
• Figure 17-12 for carbon tetrachloride shows that the range of measurements
collected at S4MO is relatively small compared to those measured at the program-
level. The annual average concentration for S4MO is similar to the program-level
average concentration, which is also similar to the program-level median
concentration (less than 0.015 |ig/m3 separates these three parameters).
• Figure 17-13 is the box plot for/>-dichloromethane. Note that the first quartile is
zero and therefore not visible on the graph. This box plot shows that the
maximum /7-dichloromethane concentration across the program was measured at
S4MO. The annual average concentration of this pollutant for S4MO is nearly
three times greater the program-level average concentration. Even though the
maximum/?-dichlorobenzene concentration was measured at S4MO, this site does
not have the highest annual average concentration among sites sampling this
pollutant (although it does rank third highest).
• Figure 17-14 is the box plot for 1,2-dichloroethane. Note that the program-level
maximum concentration (17.01 |ig/m3) is not shown directly on the box plot as
the scale has been reduced to 1 |ig/m3 to allow for the observation of data points
at the lower end of the concentration range. Figure 17-14 shows that all of the
1,2-dichloromethane measurements collected at S4MO are less than the program-
level average concentration. The program-level average concentration is greater
than the program third quartile for this pollutant and is greater than or similar to
the maximum concentration measured at most sites sampling 1,2-dichloroethane.
This is because the program-level average is being driven by the higher
measurements collected at a handful of monitoring sites. The annual average
concentration for S4MO is just greater than the median concentration at the
program level. Recall from the previous section that S4MO has the highest annual
average concentration among NMP sites sampling this pollutant.
17-26
-------
• Figure 17-15 is the box plot for ethylbenzene and shows that the annual average
concentration for S4MO is similar to the program-level average concentration.
The maximum concentration of ethylbenzene measured at S4MO is considerably
less than the maximum concentration measured across the program. There were
no non-detects of ethylbenzene measured at S4MO.
• Figure 17-16 is the box plot for fluorene. The box plot shows that the majority of
the fluorene measurements program-wide are within a relatively small
concentration range as indicated by the first, second (median), and third quartiles,
which are relatively close together. Seventy-five percent of the fluorene
measurements program-wide are less than 5.35 ng/m3. But, the maximum
concentration measured across the program is significantly higher (93.4 ng/m3).
The annual average concentration of fluorene for S4MO is greater than the
program-level average, although the maximum fluorene concentration measured
at S4MO is considerably less than the program-level maximum concentration.
• Figure 17-17 for formaldehyde shows that the annual average concentration for
S4MO is greater than the program-level average concentration but just less than
the program-level third quartile. The maximum formaldehyde concentration
measured at S4MO is less than the maximum concentration measured across the
program but ranks sixth highest among sites sampling formaldehyde. The
minimum concentration measured at S4MO is greater than 1 |ig/m3.
• Figure 17-18 is the box plot for hexachloro-1,3-butadiene. Note that the first,
second, and third quartiles for this pollutant are zero and thus, not visible on the
box plot. The box plot shows that the annual average concentration of hexachloro-
1,3-butadiene for S4MO is greater than the program-level average concentration.
The maximum concentration measured at S4MO is one of the higher
measurements across the program and S4MO is in a three-way tie for the most
measured detections of this pollutant (14). Recall from the previous section that
S4MO has the second highest annual average concentration among sites sampling
this pollutant.
• Figure 17-19 shows that the majority of the lead measurements program-wide are
within a relatively small concentration range as indicated by the first, second
(median), and third quartiles, which are relatively close together. The annual
average lead (PMio) concentration for S4MO is nearly three times the program-
level average concentration. This site has the second highest annual average lead
concentration (behind ASKY-M) among sites sampling metals. In addition, the
maximum lead concentration measured at S4MO is the maximum concentration
measured across the program. The minimum lead concentration measured at
S4MO is greater than the program-level first quartile.
17-27
-------
• Figure 17-20 is the box plot for manganese (PMio). Note that the program-level
maximum concentration (275 ng/m3) is not shown directly on the box plot as the
scale has been reduced to 150 ng/m3 to allow for the observation of data points at
the lower end of the concentration range. Figure 17-20 shows that S4MO's annual
average manganese (PMio) concentration is roughly twice the program-level
average concentration and is also greater than the program-level third quartile.
Recall from the previous section that the maximum concentration of manganese
measured at S4MO is the maximum concentration measured across the program;
in addition, this site has the third highest annual average manganese concentration
among sites sampling PMio metals.
• Figure 17-21 is the box plot for naphthalene and shows that the annual average
naphthalene concentration for S4MO is greater than the program-level average
concentration and just less than the program-level third quartile. The maximum
naphthalene concentration measured at S4MO is considerably less than the
program-level maximum concentration.
• Figure 17-22 is the box plot for nickel. The maximum nickel concentration
measured at S4MO is among the higher nickel concentrations measured across the
program. S4MO's annual average nickel concentration is just greater than the
program-level average concentration and similar to the program-level third
quartile.
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 3.5.3.2.
S4MO has sampled VOCs and carbonyl compounds under the NMP since 2002, PMio metals
since 2003, and PAHs since 2008. Thus, Figures 17-23 through 17-39 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 is not provided, although the range and quartiles are still presented.
17-28
-------
Figure 17-23. Yearly Statistical Metrics for Acenaphthene Concentrations Measured at S4MO
O
2010
Year
O 5th Percentile
— Maximum
O 95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until April 2008.
Observations from Figure 17-23 for acenaphthene measurements collected 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 for 2008 is not presented, although the
range of measurements is provided.
• Two measurements greater than 30 ng/m3 were measured at S4MO in September
2008. Another measurement greater than 30 ng/m3 was also measured in July 2011.
• All of the statistical parameters shown exhibit decreases from 2008 to 2009. Although
the range of concentrations measured increased from 2009 to 2010, but the median
concentration decreased slightly, a trend that continued into 2011.
• 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 concentration range increased while the number of measurements at
the lower end of the concentration decreased. The number of concentrations greater
than 10 ng/m3 increased from 12 to 17 from 2011 to 2012 and the number of
concentrations between 5 ng/m3 and 10 ng/m3 increased from eight to 15. Conversely,
the number of concentrations between 1 ng/m3 and 5 ng/m3 decreased from 31 to 25
and the number of concentrations less than 1 ng/m3 decreased from 10 to three from
2011 to 2012.
17-29
-------
Figure 17-24. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at S4MO
„
c
.2
2
,1,0
| — i
T
n 1 n _L
II *
•
^ - ~t>~ £L JU, -"^ I
'• *£• ^ _jj_ ''•••UJ ^ ' ^ 1 o 1
• ^ e t^5 ~ " UJ HP
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Year
O 5th Percent le - M nimum ~ Median — Maximum O 95th Percentile ...^>... Average
Observations from Figure 17-24 for acetaldehyde measurements collected 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 in Figure 17-24, 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 2010 to
2011 and again for 2012. The range of measurements is at a minimum for 2012; the
difference between the 5th and 95th percentiles, or the range within which 90 percent
of the measurements fall, is also at a minimum for 2012.
17-30
-------
Figure 17-25. Yearly Statistical Metrics for Arsenic (PMi0) Concentrations Measured at S4MO
2007 2008
Year
O 5th Percentile
— Maximum
O 95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until July 2003.
Observations from Figure 17-25 for arsenic measurements collected 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 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). Only five additional arsenic concentrations greater than 10 ng/m3 have
been measured at S4MO (three in 2005 and one each in 2003, 2007, 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.
• Many of the statistical parameters are at a minimum for 2011. The difference between
the 5th and 95th percentiles is at a minimum for 2011, as is the difference between
the median and 1-year average concentrations (less than 0.12 ng/m3 separates these
two parameters for 2011).
• Many of the statistical parameters exhibit increases for 2012, although difficult to
discern in Figure 17-25. The maximum concentration nearly doubled and the number
of measurements greater than 2 ng/m3 increased from one in 2011 to five in 2012.
17-31
-------
Figure 17-26. Yearly Statistical Metrics for Benzene Concentrations Measured at S4MO
£
C
.2
2
1
1 — i
<
•
i — |
•
^
_
1 1 I
r^i r IT
n^ r- i— i rt ^
• rn ^
-o i rn
^^ m^ Q .^..
L^-1 '-?-' _^ , Ej^ Up-1
I .LjJiLjJ- ii_^j
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Year
O 5th Percentile - Minimum ~ Median — Maximum O 95th Percentile ...^>... Average
Observations from Figure 17-26 for benzene measurements collected 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.
• All four benzene concentrations greater than 5 |ig/m3 were measured in 2003.
• The 1-year average concentrations exhibit a steady decreasing trend through 2007,
representing a roughly 1 |ig/m3 decrease, although the most significant changes
occurred in the early years of sampling. In the years following 2007, the 1-year
average concentration has varied between 0.80 |ig/m3 (2011) and 1.03 |ig/m3 (2010).
• The range of benzene measurements is smallest for 2011, with a difference of
approximately 1 |ig/m3 between the minimum and maximum concentration measured.
• 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 did not change and the median decreased. Even though the maximum
concentration measured doubled from 2011 to 2012, it's the concentrations at the
lower end of the concentration range most affecting the calculations. The number of
concentrations less than 0.5 |ig/m3 increased from two to 11 from 2011 to 2012.
17-32
-------
Figure 17-27. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at S4MO
2007 2008
Year
O 5th Percentile
— Maximum
O 95th Percentile
Observations from Figure 17-27 for 1,3-butadiene measurements collected 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 ug/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 66 percent in 2004
to a minimum of zero percent in 2010 and 2012.
• Between 2004 and 2008, the 1-year average concentration changed very little,
ranging from 0.078 ug/m3 (2005) to 0.095 ug/m3 (2006). Greater fluctuations are
shown in the years that follow. Years with a higher number of non-detects, as
indicated by the minimum and 5th percentile, such as 2009 and 2011, alternate with
years without any non-detects and concentrations that are higher in magnitude, as
indicated by the 95th percentile and maximum concentration.
• Even with the variable range of measurements, the median concentration shown for
the period from 2010 to 2012 varies by less than 0.01 ug/m3.
17-33
-------
Figure 17-28. Yearly Statistical Metrics for Cadmium (PMi0) Concentrations Measured at
S4MO
§
I
I
3 i.o
o
•o-
••o-f
•Q-
±
I
2007 2008
Year
• 5th Percentile
— Maximum • 95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until July 2003.
Observations from Figure 17-28 for cadmium measurements collected at S4MO include
the following:
• The maximum concentration for most years of sampling is less than 3 ng/m3; the 12
measurements greater than 3 ng/m3 were measured at S4MO in 2004 (three), 2005
(two), 2008 (two), and 2009 (five).
• A steady decreasing trend is shown in the 1-year average and median concentrations
between 2004 and 2006. Even though the 1-year average exhibits an increasing trend
between 2006 and 2009, the median concentration does not, indicating that
concentrations at the upper end of the concentration range are driving the 1-year
average, particularly for 2009. The difference between the 1-year average and median
concentrations is at a maximum for 2009, indicating an increasing level of variability
in the measurements. The range of concentrations measured decreased significantly
from 2009 to 2010.
• A slight increasing trend in the maximum and 95th percentile is shown from 2010 to
2011 and 2012, yet the median cadmium concentration is at a minimum for 2012.
This is a result of the increasing number of concentrations at the lower end of the
concentration range. The minimum concentration measured in 2012 decreased by half
and the number of concentrations less than 0.25 ng/m3 increased from 10 in both
2010 and 2011 to 24 in 2012.
17-34
-------
Figure 17-29. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
S4MO
I
o
o*
T
2003 2004 2005 2006
2007 2008
Year
2009 2010
2011 2012
O 5th Percentile - Minimuir
— Maximum • 95th Percentile
Observations from Figure 17-29 for carbon tetrachloride measurements collected at
S4MO include the following:
• Twenty of the 21 non-detects of carbon tetrachloride were measured in 2003, 2004, or
2005, with the final non-detect measured in 2007.
• A steady increasing trend in the 1-year average in shown through 2006. Although the
maximum concentration decreased substantially from 2006 to 2007, the change in the
1-year average is slight and the median concentration did not change at all. In fact,
the median concentration is the same for 2005, 2006, and 2007.
• All of the statistical parameters exhibit increases from 2007 to 2008. Both the median
and 1-year average concentrations have a decreasing trend from 2008 through 2010,
after which these parameters begin increasing again.
• Although there appears to be significant variability in these measurements, the
changes shown in the 1-year averages vary by less than 0.25 ug/m3.
• The box and whisker plots for this pollutant appear "inverted," with the minimum
concentration extending farther away from the majority of measurements than the
maximum concentration (which is more common, see benzene as an example) for the
period between 2007 and 2011.
17-35
-------
• The range of concentrations measured is at a minimum for 2012. The difference
between the 5th and 95th percentiles is also at a minimum for 2012, indicating that
the majority of measurements are falling within a tighter range.
Figure 17-30. Yearly Statistical Metrics for/7-Dichlorobenzene Concentrations Measured at
S4MO
O 5th Percentile
— Maximum
95th Percentile
Observations from Figure 17-30 for/>-dichlorobenzene measurements collected 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 number of non-detects was at a maximum in 2003 (90 percent), after which the
percentage decreased, reaching a minimum of 5 percent for 2009. The percentage of
non-detects has varied between 10 percent and 20 percent since 2009.
• 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/>-dichlorobenzene concentration
was measured. The difference between the median and 1-year average concentration
is also an indicator of this. During this period, the 1-year average was at least three
times greater than the median.
17-36
-------
• The concentrations measured decreased considerably from 2008 to 2009 then
increased again in 2010. The increase in most of the statistical parameters for 2010,
particularly the 95th percentile, indicates that concentrations measured were higher in
general that year. The number of measurements greater than 0.75 ug/m3 increased
from two in 2009 to eight for 2010.
• Although the maximum concentration measured in 2011 is similar to the maximum
concentration measured in 2010, the 95th percentile and 1-year average decreased
while the median concentration increased. This is because there were fewer
measurements at the upper end of the concentration range and a greater number of
non-detects from the previous year.
• The concentrations measured in 2012 exhibit less variability than the previous two
years. The difference between the 1-year average and median concentration is at a
minimum over the 3-year period.
Figure 17-31. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
S4MO
¥
1
3
••
rn ^
T
1
fy L
,o
fc ^ o
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Year
O 5th Percentile - Minimum ~ Median — Maximum O 95th Percentile ...^>... Average
Observations from Figure 17-31 for 1,2-dichloroethane measurements collected at S4MO
include the following:
• With the exception of 2012, 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
17-37
-------
detection in 2005, two in 2006 and 2008, five in 2009, 10 in 2010, 18 in 2011, and 56
in 2012.
• As the number of measured detections increased in the later years of sampling, each
of the corresponding statistical metrics shown in Figure 17-31 also increased.
• As the number of measured detections increased dramatically for 2012, the median
and 1-year average concentrations increased correspondingly. The median
concentration is actually greater than the 1-year average for 2012, although the
difference is less than 0.005 ug/m3. The majority of measurements fall within a
relatively small range (roughly 0.06 ug/m3) as indicated by the 5th and 95th
percentiles.
Figure 17-32. Yearly Statistical Metrics for Ethylbenzne Concentrations Measured at S4MO
oncentration (ug/m3
D C
<
~
.
T
2. .<>... r-t-i ph
LjJ •• " :£. j^ rjt-] ••" ^m ^ rOJ
L-|-J L-t-l L»J LQJ 't^j L-g-l ^~^ T~^
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Year
O 5th Percentile - Minimum ~ Median — Maximum O 95th Percentile ...^>... Average
Observations from Figure 17-32 for ethylbenzene measurements collected at S4MO
include the following:
• The maximum concentration of ethylbenzene was measured at S4MO in 2003. In
fact, the eight highest concentrations (those greater than 2.50 u,g/m3) were all
measured in 2003. Nearly half of the ethylbenzene concentrations greater than
1 u,g/m3 were measured in 2003 (21 out of 44).
• The 1-year average concentration has a steady decreasing trend through 2009, when
nearly all of the statistical parameters were at a minimum. The maximum
17-38
-------
concentration for 2009 (0.44 ug/m3) is less than the 1-year average for some of the
earlier years of sampling.
• Nearly all of the statistical parameters exhibit considerable increases from 2009 to
2010, with the median and 1-year average concentrations doubling, the 95th
percentile tripling, and the maximum increasing by a factor of five.
• The range of measurements collected in 2011 and 2012 decreased by more than half
compared to 2010. A decrease in the 1-year average is shown from 2010 to 2011 and
again for 2012, although the changes are not statistically significant.
Figure 17-33. Yearly Statistical Metrics for Fluorene Concentrations Measured at S4MO
I20'0
2010
Year
5th Percentile - Minimum ~ Median — Maximum
95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until April 2008.
Observations from Figure 17-33 for fluorene measurements collected at S4MO include
the following:
• The box and whisker plots for fluorene measurements resemble the plots for
acenaphthalene presented in Figure 17-23.
• 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).
• All of the statistical parameters shown exhibit decreases from 2008 to 2009. From
2009 to 2010, the range of concentrations measured increased but the median
17-39
-------
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 1 ng/m3 decreased from 11 to three from 2011 to 2012.
Figure 17-34. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at S4MO
45
40
35
30
|25
§
1
1 20
3
15
10
5.
0.
o 4 —
t\
.
2003
•
— i
T
±
±r i-i-i
ft 1 T
JM 1 ^
~ *• A * A g ^ ^ i
2004 2005 2006 2007 2008 2009 2010 2011 2012
Year
5th Percent le - M nimum ~ Median — Maximum O 95th Percentile ...^>... Average
Observations from Figure 17-34 for formaldehyde measurements collected at S4MO
include the following:
• The maximum formaldehyde concentration (43.8 ug/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 ug/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 2003 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 ug/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.
17-40
-------
• Most of the statistical parameters exhibit considerably increases from 2010 to 2011.
There were 11 concentrations of formaldehyde measured in 2011 that were greater
than the maximum concentration measured in 2010.
• Most of the statistical parameters exhibit decreases from 2011 to 2012.
Figure 17-35. Yearly Statistical Metrics for Hexachloro-l,3-Butadiene Concentrations
Measured at S4MO
c
_o
u
— '
>—
•*••
2003 2004 2005
1
O 5th Percentile - Minimum
<
•
20
>-, r^ I
r-O-i
>yv yw
06 2007 2008 2009 2010 2011 2012
Year
Median — Maximum O 95th Percentile •••^•••Average
Observations from Figure 17-35 for hexachloro-l,3-butadiene measurements collected at
S4MO include the following:
• The median concentration for 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 number 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. Additional years of sampling are needed to
determine if the number of measured detections will continue to increase.
17-41
-------
Figure 17-36. Yearly Statistical Metrics for Lead (PMi0) Concentrations Measured at S4MO
£
1
C
.0 ,-f
* 6C
1
20 -
o -\ —
— '
•
-4
>—
•
1-
2003
O
i '
— i
<*
•
— 1_
i—
•
1-
2004
5th Percent le
1
-^
2005
-^
1>—
>
•
^-s-1
2006
Minimum
-H
3—
iifa
2007
Year
Median
"*••
,
2008
— i
>—
•^jj
— •-
2009
Maximum
—
•
:>—
-t-
2010
|
^
-£-
2011
O 95th Percentile
— <
>—
•C
%
-e-
2012
O-" Average
A 1-year average is not presented because sampling under the NMP did not begin until July 2003.
Observations from Figure 17-36 for lead measurements collected at S4MO include the
following:
• The maximum lead concentration was measured at S4MO in 2012 (111 ng/m3). This
is the only measurement greater than 100 ng/m3 measured at S4MO.
• The 95th percentile for 2012 is greater than the 95th percentiles for all other years as
well as the maximum concentration for some years. Even though the maximum, 95th
percentile, and 1-year average concentration increased from to 2011 to 2012, the
median concentration actually decreased. Concentrations less than 7 ng/m3 account
for more than half of the concentrations measured in 2012, up from 31 percent in
2011.
• The 1-year average concentration of lead at S4MO has fluctuated over the years and
exhibits no real trend. The 1-year averages have ranged from 9.94 ng/m3 (2009) to
14.46 ng/m3 (2006). The confidence intervals calculated for these averages are
relatively large and indicate a considerable amount of variability in the
measurements. This site has the second highest annual average concentration of lead
for 2012 and has had the highest annual average for the last several years compared to
other NMP sites sampling PMio metals under the NMP.
17-42
-------
Figure 17-37. Yearly Statistical Metrics for Manganese (PMi0) Concentrations Measured at
S4MO
-
2007 2008
Year
• 5th Percentile
— Maximum • 95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until July 2003.
Observations from Figure 17-37 for manganese measurements collected at S4MO include
the following:
• The maximum manganese concentration was measured on November 26, 2008
(734 ng/m3) and is nearly twice the next highest concentration (395 ng/m3, measured
in 2011). A similar concentration was also measured in 2004 (387 ng/m3). The
maximum manganese concentration measured in 2012 ranks fourth highest since
metals sampling began at S4MO.
• Seven manganese concentrations greater than 100 ng/m3 have been measured at
S4MO since 2003, with only 2012 having more than one. For each of these years, the
second highest concentration of manganese was at least half as high as the highest
concentration. For example, for 2010, the two highest concentrations are 200 ng/m3
and 84.5 ng/m3.
• The 1-year average concentration of manganese has ranged from 8.08 ng/m3 (2009)
to 22.66 ng/m3 (2012). The median concentration, which is influenced less by
outliers, has varied less, ranging from 6.82 ng/m3 (2009) to 13.15 ng/m3 (2003). The
median concentration actually has a decreasing trend from 2006 to 2009, despite the
outlier measured in 2008 (the two highest concentrations measured in 2008 were
734 ng/m3 and 31.2 ng/m3).
17-43
-------
Figure 17-38. Yearly Statistical Metrics for Naphthalene Concentrations Measured at S4MO
~ 400
O
•
-2-
2010
Year
O 5th Percentile - Minimuir
— Maximum • 95th Percentile "-O-" Averagf
A 1-year average is not presented because sampling under the NMP did not begin until April 2008.
Observations from Figure 17-38 for naphthalene measurements collected at S4MO
include the following:
• Naphthalene concentrations measured at S4MO exhibit considerable variability,
ranging from 18 ng/m3 (2011) to 784 ng/m3 (2010).
• The 1-year average concentration has ranged from 83.82 ng/m3 (2011) to 135 ng/m3
(2010). The median varies less, ranging from 72.20 ng/m3 (2011) to 89.85 ng/m3
(2010).
• The years when rather high concentrations were measured alternate with years when
the maximum concentration is considerably less, resulting in the 1-year average (and
median) concentrations having an undulating pattern.
17-44
-------
Figure 17-39. Yearly Statistical Metrics for Nickel (PMi0) Concentrations Measured at S4MO
8 4.0
r^
o
V
t&l
t
2007 2008
Year
O 5th Percentile
— Maximum
O 95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until July 2003.
Observations from Figure 17-39 for nickel measurements collected at S4MO include the
following:
• The two highest nickel concentration were measured in 2009 (9.82 ng/m3) and 2012
(9.74 ng/m3). No other concentrations greater than 7 ng/m3 have been measured at
S4MO.
• The 1-year average concentration has ranged from 1.04 ng/m3 (2010) to 1.45 ng/m3
(2007). The slight decreasing trend shown between 2004 and 2010 was interrupted by
the increase shown for 2007. This year has the highest minimum concentration, the
second fewest measurements less than 1 ng/m3, and the fourth highest concentration
measured at S4MO.
• The 1-year average, 95th percentile, and maximum concentrations exhibit an
increasing trend between 2010 and 2012. However, the wide range of concentrations
measured results in relatively large confidence intervals that indicate that the change
is not statistically significant.
17-45
-------
17.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk at the
S4MO monitoring site. Refer to Sections 3.3 and 3.5.3 for definitions and explanations regarding
the various toxicity factors, time frames, and calculations associated with these risk-based
screenings.
17.5.1 Risk-Based Screening Assessment Using MRLs
A risk-based screening was conducted by comparing the concentration data from the
Missouri monitoring site to the ATSDRMRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 day to 14 days); intermediate (exposures of 15 days to 364 days); and chronic
(exposures of 1 year or greater). The preprocessed daily measurements of the pollutants of
interest were compared to the acute MRLs; the quarterly averages were compared to the
intermediate MRLs; and the annual averages were compared to the chronic MRLs.
As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2012.
17.5.2 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 3.5.3.4 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 17-6, where applicable. Cancer risk approximations are presented as
probabilities while the noncancer hazard approximations are ratios and thus, unitless values.
17-46
-------
Table 17-6. Risk Approximations for the Missouri Monitoring Site
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
St. Louis, Missouri - S4MO
Acenaphthene3
Acetaldehyde
Arsenic (PM10)a
Benzene
1,3 -Butadiene
Cadmium (PM10)a
Carbon Tetrachloride
/>-Dichlorobenzene
1,2-Dichloroethane
Ethylbenzene
Fluorene3
Formaldehyde
Hexachloro- 1 ,3 -butadiene
Lead (PM10)a
Manganese (PM10)a
Naphthalene3
Nickel (PM10)a
0.000088
0.0000022
0.0043
0.0000078
0.00003
0.0018
0.000006
0.000011
0.000026
0.0000025
0.000088
0.000013
0.000022
0.000034
0.00048
0.009
0.000015
0.03
0.002
0.00001
0.1
0.8
2.4
1
0.0098
0.09
0.00015
0.00005
0.003
0.00009
60/60
61/61
61/61
58/58
58/58
61/61
58/58
52/58
56/58
58/58
60/60
61/61
14/58
61/61
61/61
60/60
61/61
0.01
±<0.01
1.86
±0.21
<0.01
±<0.01
0.80
±0.12
0.11
±0.02
<0.01
±<0.01
0.68
±0.02
0.18
±0.06
0.08
±0.01
0.35
±0.05
0.01
±0.01
3.26
±0.52
0.02
±0.01
0.01
±<0.01
0.02
±0.01
0.11
±0.02
0.01
±0.01
0.65
4.08
4.67
6.25
3.24
1.03
4.06
1.98
2.17
0.89
0.73
42.33
0.42
3.76
0.68
0.21
0.07
0.03
0.05
0.06
0.01
O.01
0.01
O.01
0.33
0.01
0.09
0.45
0.04
0.02
- = A Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 17-5.
Observations for S4MO from Table 17-6 include the following:
• The pollutants with the highest annual average concentrations for S4MO are
formaldehyde, acetaldehyde, and benzene.
• Formaldehyde has the highest cancer risk approximations for S4MO (42.33 in-a-
million), The cancer risk approximation for formaldehyde is among the higher cancer
risk approximations calculated among the site-specific pollutants of interest across the
program.
17-47
-------
• Benzene has the highest cancer risk approximation for S4MO among the VOCs
(6.25 in-a-million); arsenic has the highest cancer risk approximation for S4MO
among the metals (4.67 in-a-million); and naphthalene has the highest cancer risk
approximation for S4MO among the PAHs (3.76 in-a-million).
• None of the pollutants of interest for S4MO have noncancer hazard approximations
greater than 1.0, indicating that no adverse health effects are expected from these
individual pollutants. The pollutant with the highest noncancer hazard approximation
is manganese (0.45), which is the fifth highest noncancer hazard approximation
calculated for a site-specific pollutant interest among NMP sites.
17.5.3 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 17-7 presents the 10 pollutants with the highest emissions from the 2011 NEI that have
cancer toxicity factors. Table 17-7 also presents the 10 pollutants with the highest toxicity -
weighted emissions, based on the weighting schema described in Section 3.5.3.5. Lastly,
Table 17-7 provides the 10 pollutants with the highest cancer risk approximations (in-a-million)
for S4MO, as presented in Table 17-6. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 17-7. Table 17-8 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 3.5.3.5. Similar to the cancer risk and
noncancer hazard approximations provided in Section 17.5.2, this analysis may help policy-
makers prioritize their air monitoring activities.
17-48
-------
Table 17-7. 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)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
St. Louis, Missouri (St. Louis City) - S4MO
Formaldehyde
Acetaldehyde
Benzene
Ethylbenzene
Naphthalene
1,3 -Butadiene
Trichloroethylene
POM, Group 2d
POM, Group 2b
Dichloromethane
283.51
125.56
114.51
54.99
29.11
19.55
15.45
5.73
5.10
3.65
Formaldehyde
Naphthalene
Benzene
Hexavalent Chromium, PM
1,3 -Butadiene
POM, Group 2d
POM, Group 3
POM, Group 2b
Acetaldehyde
Arsenic, PM
3.69E-03
9.90E-04
8.93E-04
8.16E-04
5.86E-04
5.04E-04
5.02E-04
4.49E-04
2.76E-04
2.48E-04
Formaldehyde
Benzene
Arsenic
Acetaldehyde
Carbon Tetrachloride
Naphthalene
1,3 -Butadiene
1 ,2-Dichloroethane
£>-Dichlorobenzene
Cadmium
42.33
6.25
4.67
4.08
4.06
3.76
3.24
2.17
1.98
1.03
VO
-------
Table 17-8. 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)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer Hazard
Approximation
(HQ)
St. Louis, Missouri (St. Louis City) - S4MO
Toluene
Ethylene glycol
Formaldehyde
Hexane
Xylenes
Methanol
Acetaldehyde
Benzene
Hydrochloric acid
Ethylbenzene
703.90
393.90
283.51
232.58
221.77
208.08
125.56
114.51
70.78
54.99
Acrolein
Formaldehyde
Manganese, PM
Acetaldehyde
1,3 -Butadiene
Naphthalene
Trichloroethylene
Arsenic, PM
Benzene
Hydrochloric acid
888,399.09
28,929.21
18,321.76
13,951.64
9,774.06
9,702.37
7,726.86
3,850.67
3,816.95
3,539.11
Manganese
Formaldehyde
Acetaldehyde
Lead
Arsenic
Cadmium
1,3 -Butadiene
Naphthalene
Benzene
Nickel
0.45
0.33
0.21
0.09
0.07
0.06
0.05
0.04
0.03
0.02
-------
Observations from Table 17-7 include the following:
• Formaldehyde, acetaldehyde, and benzene are the highest emitted pollutants with
cancer UREs in the city of St. Louis.
• The pollutants with the highest toxi city-weighted emissions (of the pollutants with
cancer UREs) are formaldehyde, naphthalene, and benzene.
• Seven of the highest emitted pollutants also have the highest toxi city-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 third highest cancer risk approximation for S4MO. While arsenic is
not one of the highest emitted pollutants, it ranks 10th for its toxicity-weighted
emissions. Carbon tetrachloride has the fifth highest cancer risk approximation for
S4MO but appears on neither emissions-based list.
• POM, Group 2b is the ninth highest emitted "pollutant" in St. Louis and ranks eighth
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 17-8 include the following:
• Toluene, ethylene glycol, and formaldehyde 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 manganese. Although acrolein was
sampled for at S4MO, this pollutant was excluded from the pollutants of interest
designation, and thus subsequent risk-based screening evaluations, due to questions
about the consistency and reliability of the measurements, as discussed in Section 3.2.
• Four of the highest emitted pollutants in the city of St. Louis also have the highest
toxicity-weighted emissions.
• Manganese, the pollutant with highest noncancer hazard approximation, has the third
highest toxicity-weighted emissions but is not one of the highest emitted (it ranks
30th). Arsenic, naphthalene, and 1,3-butadiene are also among the pollutants with the
highest toxicity-weighted emissions, but are not among the highest emitted.
• Formaldehyde and acetaldehyde are the pollutants with the second and third highest
noncancer hazard approximations for S4MO, respectively; these two pollutants of
interest appear on both emissions-based lists. Benzene also appears on all three lists.
17-51
-------
17.6 Summary of the 2012 Monitoring Data for S4MO
Results from several of the data treatments described in this section include the
following:
»«» Twenty-four pollutants failed screens for S4MO. S4MO failed the highest number of
screens among allNMP 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 jug/m3.
»«» S4MO has the highest annual average concentration of 1,2-dichloroethane, the
second highest annual average concentration ofhexachloro-l,3-butadiene, and the
third highest annual average concentration ofp-dichlorobenzene. S4MO also has the
second highest annual average concentration of arsenic and the third highest annual
average concentration of manganese.
»«» Concentrations of acetaldehyde measured at S4MO since 2010 have a decreasing
trend. In addition, the detection rate of 1,2-dichloroethane has been increasing
steadily at S4MO over the last few years of sampling.
17-52
-------
18.0 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.
18.1 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.
The New Jersey sites are all located within the New York-Newark-Jersey City,
NY-NJ-PA MSA. Figure 18-1 is a composite satellite image retrieved from ArcGIS Explorer
showing the CHNJ monitoring site and its immediate surroundings. Figure 18-2 identifies nearby
point source emissions locations by source category, as reported in the 2011 NEI for point
sources. Note that only sources within 10 miles of the site are included in the facility counts
provided in Figure 18-2. 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 radius are still visible on the map, but have been
grayed out in order to show emissions sources just outside the boundary. Figures 18-3 through
18-5 are the composite satellite maps and emissions source map for ELNJ and NBNJ. Table 18-1
provides supplemental geographical information such as land use, location setting, and locational
coordinates.
18-1
-------
Figure 18-1. Chester, New Jersey (CHNJ) Monitoring Site
oo
I
to
-------
Figure 18-2. NEI Point Sources Located Within 10 Miles of CHNJ
Legend
74L45'0"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
Source Category Group (No. of Facilities)
•i" Aerospace/Aircraft Manufacturing (1)
T Airport/Airline/Airport Support Operations (12)
'A Asphalt Production/Hot Mix Asphalt Plant (1)
c Chemical Manufacturing (3)
e Electrical Equipment Manufacturing (1)
F Food Processing/Agriculture (1)
? Miscellaneous Commercial/Industrial (2)
W Woodwork, Furniture, Millwork & Wood Preserving (1)
18-3
-------
Figure 18-3. Elizabeth, New Jersey (ELNJ) Monitoring Site
oo
-------
Figure 18-4. North Brunswick, New Jersey (NBNJ) Monitoring Site
oo
-------
Figure 18-5. NEI Point Sources Located Within 10 Miles of ELNJ and NBNJ
74'35'0-W 74°30'0"W 74'25'0"W 74°20'()-W 74'15'0-W 74-10'0'W 74'5'0"W 74 WW 73'55'0"W z
r \
* Somerset ,
\ County
74545'0"W 74 40'0"W
74"35'0"W 74J3010"W
74*25'0"W 74C2D'0"W 7
-------
Table 18-1. Geographical Information for the New Jersey Monitoring Sites
Site
Code
CHNJ
ELNJ
NBNJ
AQS Code
34-027-3001
34-039-0004
34-023-0006
Location
Chester
Elizabeth
North
Brunswick
County
Morris
Union
Middlesex
Micro- or Metropolitan
Statistical Area
New York-Newark-
Jersey City, NY-NJ-PA
MSA
New York-Newark-
Jersey City, NY-NJ-PA
MSA
New York-Newark-
Jersey City, NY-NJ-PA
MSA
Latitude
and
Longitude
40.787628,
-74.676301
40.64144,
-74.208365
40.472825,
-74.422403
Land Use
Agricultural
Industrial
Agricultural
Location
Setting
Rural
Suburban
Rural
Additional Ambient Monitoring
Information1
SO2, NO, NO2, O3, Meteorological
parameters, PM2 5, PM2 5 Speciation,
IMPROVE Speciation.
CO, SO2, NO2, NOX, Meteorological
parameters, PM2 5, PM2 5 Speciation,
IMPROVE Speciation.
Meteorological parameters, PM2 5, PM2 5
Speciation, IMPROVE Speciation.
:Data for additional pollutants are reported to AQS for these sites (EPA, 2013b); however, these data are not generated by ERG and are therefore not included in this
report.
oo
-------
CHNJ is located in northern New Jersey, in the town of Chester, west of the New York
City metropolitan area. Figure 18-1 shows that CHNJ is located in an open area near Building 1
of the Department of Public Works off Route 513. The surrounding area is rural and agricultural
with a rolling topography, but surrounded by small neighborhoods. Although the location is
considered part of the New York City MSA, the site's location is outside most of the urbanized
areas. Figure 18-2 shows that few sources are within a few miles of CFINJ. The source category
with the greatest number of emissions sources surrounding CFINJ 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 sources closest to CFINJ include a
privately owned heliport and a wood work, furniture, mill work, and wood preserving facility.
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 18-3 shows, the monitoring site is located
just off Exit 13 of the New Jersey Turnpike (1-95), near the toll plaza. Interstate-278 intersects
the Turnpike here as well. The surrounding area is highly industrialized, with an oil refinery
located just southwest of the site. Additional industry is located to the southwest, west, and east,
while residential neighborhoods are located to the northwest and north of the site.
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 18-4. County Road 617 (Ryders Lane) and US-1 intersect just west of the site and 1-95
runs northeast-southwest about 1 mile east of the site, part of which can be seen in the lower
right hand corner of Figure 18-4.
Figure 18-5 shows that the outer portions of the 10-mile radii for ELNJ and NBNJ
intersect and that 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
18-8
-------
manufacturing, bulk terminals/bulk plant, and petroleum refining source categories. The
emissions sources in closest proximity to the NBNJ monitoring site are involved in airport and
airport support operations and pharmaceutical manufacturing.
Table 18-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the New Jersey monitoring sites. Table 18-2 includes a county-level
population for each site. County-level vehicle registration data for Union, Morris, and Middlesex
Counties were not available from the State of New Jersey. Thus, state-level vehicle registration,
which was obtained from the Federal Highway Administration (FFEWA), was allocated to the
county level using the county-level proportion of the state population from the U.S. Census
Bureau. Table 18-2 also contains traffic volume information for each site as well as the location
for which the traffic volume was obtained. Additionally, Table 18-2 presents the county-level
daily VMT for Middlesex, Morris, and Union Counties.
Table 18-2. Population, Motor Vehicle, and Traffic Information for the New Jersey
Monitoring Sites
Site
CHNJ
ELNJ
NBNJ
Estimated
County
Population1
497,999
543,976
823,041
County-level
Vehicle
Registration2
445,710
485,449
733,908
Annual
Average
Daily Traffic3
11,215
250,000
110,653
Intersection
Used for
Traffic Data
Route 510, east of Fox Chase Rd
1-95 between Exits 13 & 13 A
US-1, west of Route 617
County-
level Daily
VMT4
14,844,444
12,264,174
20,644,392
Bounty-level population estimates reflect 2012 data (Census Bureau, 2013c)
2County-level vehicle registration reflects ratios based on 2011 state-level vehicle registration data from the FHWA
and the 2011 county-level proportion of the state population data (FHWA, 2013a and Census Bureau, 2012)
3AADT for ELNJ reflects 2006 data from NJ Department of Treasury and 2009 data for NBNJ and 2012 data for
CHNJ from the New Jersey DOT (NJ DOTr, 2008 and NJ DOT, 2013)
4County-level VMT reflects 2011 data (NJ DOT, 2011)
Observations from Table 18-2 include the following:
• Middlesex County, where NBNJ is located, has the highest county-level population
for the New Jersey sites while Morris County, where CFINJ is located, has the least.
Compared to NMP monitoring sites in other locations, the county-level populations
are in the middle of the range, ranking 16th, 24th, and 25th.
• The estimated county-level vehicle registration is also highest for NBNJ and least for
CFINJ. The county-level registration estimates for the sites have similar rankings as
the county-level populations among NMP sites.
18-9
-------
• ELNJ and NBNJ experience a significantly higher average traffic volume than CHNJ.
Traffic data for ELNJ are provided for 1-95, between Exit 13 and 13 A; this is the
second highest traffic volume among all NMP sites. Traffic data for CHNJ are
provided for Route 510, east of Fox Chase Road; traffic data for NBNJ are provided
for US-1, west of State Road 617 (Ryders Lane).
• The daily VMT is highest for Middlesex County (NBNJ) and lowest for Union
County (ELNJ). However, VMT for the New Jersey counties are in the middle of the
range compared to other counties with NMP sites (where VMT data were available).
18.2 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.
18.2.1 Climate Summary
Frontal systems push across the state of New Jersey regularly, producing variable
weather conditions. The state's proximity to the Atlantic Ocean has a moderating effect on
temperatures. Summers along the coast tend to be cooler than areas farther inland, while winters
tend to be warmer. Large urban areas within the state experience the urban heat island effect, in
which urban areas retain more heat than outlying areas. New Jersey's Mid-Atlantic location also
allows for ample annual precipitation and relatively high humidity. Greater than 3 inches of
precipitation can be expected each month in the northeastern portion of the state. A
southwesterly wind is most common in the summer and a northwesterly wind is typical in the
winter. Winds from the west and northwest result in air masses that dry out, stabilize, and warm
as they move eastward from higher elevations to sea level (Wood, 2004; Rutgers, 2014).
18.2.2 Meteorological Summary
Hourly meteorological data for 2012 were retrieved from NCDC for the weather stations
closest to the New Jersey monitoring sites (NCDC, 2012), as described in Section 3.5.2. The
closest weather stations are located at Somerville-Somerset Airport (near CHNJ and NBNJ) and
Newark International Airport (near ELNJ), WBAN 54785 and 14734, respectively. Additional
information about these weather stations, such as the distance between the sites and the weather
stations, is provided in Table 18-3. These data were used to determine how meteorological
conditions on sample days vary from conditions experienced throughout the year.
18-10
-------
Table 18-3. Average Meteorological Conditions near the New Jersey Monitoring Sites
Closest Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Chester, New Jersey - CHNJ
Somerville, New
Jersey/Somerset
Airport
54785
(40.62, -74.67)
11.3
miles
165°
(SSE)
Sample
Days
(65)
2012
64.2
±4.3
65.2
±1.8
53.7
±4.0
54.6
±1.6
43.2
±4.4
43.4
±1.8
48.6
±3.8
49.3
±1.6
71.4
±3.3
70.1
±1.5
1017.4
±1.8
1015.9
±0.8
2.5
±0.4
2.9
±0.2
Elizabeth, New Jersey - ELNJ
Newark International
Airport
14734
(40.68, -74.17)
3.4
miles
20°
(NNE)
Sample
Days
(65)
2012
63.7
±4.2
65.4
±1.8
56.4
±4.0
57.8
±1.7
42.2
±4.4
43.1
±1.8
49.6
±3.6
50.7
±1.5
62.8
±4.2
61.8
±1.7
1016.6
±2.1
1015.9
±0.8
7.9
±0.7
7.8
±0.4
North Brunswick, New Jersey - NBNJ
Somerville, New
Jersey/Somerset
Airport
54785
(40.62, -74.67)
16.1
miles
297°
(WNW)
Sample
Days
(67)
2012
64.5
±4.3
65.2
±1.8
54.1
±4.0
54.6
±1.6
43.7
±4.3
43.4
±1.8
49.0
±3.8
49.3
±1.6
71.9
±3.5
70.1
±1.5
1017.2
±1.8
1015.9
±0.8
2.6
±0.4
2.9
±0.2
oo
1 Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
-------
Table 18-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for all of 2012. Also included in Table 18-3 is the 95 percent
confidence interval for each parameter. As shown in Table 18-3, average meteorological
conditions on sample days were generally representative of average weather conditions
experienced throughout the year near CHNJ, ELNJ, and NBNJ. The largest difference between a
sample day and a full-year average is for relative humidity at NBNJ, although the difference is
not statistically significant.
18.2.3 Back Trajectory Analysis
Figure 18-6 is the composite back trajectory map for days on which samples were
collected at the CHNJ monitoring site. Included in Figure 18-6 are four back trajectories per
sample day. Figure 18-7 is the corresponding cluster analysis. Similarly, Figures 18-8 through
18-11 are the composite back trajectory maps and corresponding cluster analyses for ELNJ and
NBNJ. An in-depth description of these maps and how they were generated is presented in
Section 3.5.2.1. For the composite maps, each line represents the 24-hour trajectory along which
a parcel of air traveled toward the monitoring site on a given sample day and time, based on an
initial height of 50 meters AGL. For the cluster analyses, each line corresponds to a trajectory
representative of a given cluster of back trajectories. Each concentric circle around the sites in
Figures 18-6 through 18-11 represents 100 miles.
18-12
-------
Figure 18-6. Composite Back Trajectory Map for CHNJ
f. VX
Figure 18-7. Back Trajectory Cluster Map for CHNJ
,
18-13
-------
Figure 18-8. Composite Back Trajectory Map for ELNJ
Figure 18-9. Back Trajectory Cluster Map for ELNJ
18-14
-------
Figure 18-10. Composite Back Trajectory Map for NBNJ
Figure 18-11. Back Trajectory Cluster Map for NBNJ
18-15
-------
Observations from Figures 18-6 through 18-11 include the following:
• Due to their relatively close proximity to each other and the standardization of sample
days, the back trajectories shown on each composite back trajectory map for the New
Jersey sites are similar to each other. The composite map for NBNJ is on a slightly
smaller scale because the relatively long back trajectory originating off the North
Carolina/South Carolina coast corresponds to a sample day in which a sample was not
collected at NBNJ.
• Back trajectories originated from a variety of directions at the sites. In general, the
longest back trajectories originated from west or northwest of the monitoring sites.
• The 24-hour air shed domains for the New Jersey sites were similar in size to each
other. Back trajectories greater than or approaching 600 miles in length were
constructed for each site. These back trajectories originated over Ontario, Canada,
Michigan, and Indiana. The average back trajectory length for these sites ranged from
223 miles (CHNJ) to 236 miles (ELNJ).
• The cluster trajectories for the New Jersey sites are similar to each other in
geographical distribution, although the percentages vary. Each of the cluster maps has
a long cluster trajectory originating over Lake Erie; another longer cluster trajectory
originating over southeast Ontario; a relatively short cluster trajectory originating
over central or southeast New York; a cluster trajectory originating to the south of
Nantucket; and a cluster trajectory originating off the Delmarva Peninsula. The
cluster trajectory originating over northern Virginia and the eastern panhandle of
West Virginia for CHNJ and NBNJ is split into two cluster trajectories for ELNJ.
18.2.4 Wind Rose Comparison
Hourly surface wind data from the weather stations nearest the New Jersey sites, as
presented in Section 18.2.2, were uploaded into a wind rose software program to produce
customized wind roses, as described in Section 3.5.2.2. 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 18-12 presents a map showing the distance between the weather station and
CHNJ, which may be useful for identifying topographical influences that may affect the
meteorological patterns experienced at this location. Figure 18-12 also presents three different
wind roses for the CHNJ monitoring site. First, a historical wind rose representing 2002 to 2011
wind data is presented, which shows the predominant surface wind speed and direction over an
extended period of time. Second, a wind rose representing wind observations for all of 2012 is
presented. Next, a wind rose representing wind data for days on which samples were collected in
18-16
-------
2012 is presented. These can be used to identify the predominant wind speed and direction for
2012 and to determine if wind observations on sample days were representative of conditions
experienced over the entire year and historically. Figures 18-13 and 18-14 present the distance
maps and wind roses for ELNJ and NBNJ, respectively.
Observations from Figures 18-12 and 18-14 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-southeast of CFINJ and 16.1 miles west-northwest of NBNJ.
• The historical and full-year wind roses for CFINJ are identical to the historical and
full-year wind roses for NBNJ because the data are from the same weather station.
• The historical wind roses for these sites show that calm winds (< 2 knots) accounted
for greater than 40 percent of observations. For wind speeds greater than 2 knots,
northerly winds were observed most frequently, accounting for 9 percent of the
observations, while winds from the southwest quadrant were rarely observed.
• Calm winds account 50 percent of the wind observations throughout 2012. Winds
from the west-northwest to north account for another one-quarter of wind
observations throughout 2012.
• Wind patterns on the sample day wind roses share many of the characteristics of the
full year wind roses, even if at first glance it does not appear that way. Calm winds
were still prevalent on sample days, accounting for more than half of wind
observations. Although northwesterly and north-northwesterly winds still account for
the majority of wind observations on sample days (for winds greater than 2 knots),
the percentages are more varied. There were fewer observations from the west-
northwest and north on sample days and a higher percentage of winds from the
northeast.
• While the 2012 wind roses do exhibit the same prevalence for calm winds as the
historical wind rose, they do not exhibit the same northerly predominance for wind
speeds greater than 2 knots. Instead, there was an increase in wind observations from
the northwest quadrant. A similar observation was made for 2009, 2010, and 2011 in
previous NMP reports.
18-17
-------
Figure 18-12. Wind Roses for the Somerville-Somerset Airport Weather Station near
CHNJ
Location of CHNJ and Weather Station
2002-2011 Historical Wind Rose
2012 Wind Rose
Sample Day Wind Rose
18-18
-------
Figure 18-13. Wind Roses for the Newark International Airport Weather Station near
ELNJ
Location of ELNJ and Weather Station
2002-2011 Historical Wind Rose
/
2012 Wind Rose
Sample Day Wind Rose
Calms: 73B%
18-19
-------
Figure 18-14. Wind Roses for the Somerville-Somerset Airport Weather Station near
NBNJ
Location of NBNJ and Weather Station
2002-2011 Historical Wind Rose
2012 Wind Rose
Sample Day Wind Rose
Calms: 52.13%
18-20
-------
Observations from Figure 18-13 for ELNJ include the following:
• The Newark International Airport weather station is located 3.4 miles north-northeast
of ELNJ.
• The historical wind rose shows that winds from a variety of directions were observed
near ELNJ, although winds from the east-northeast to southeast were observed
infrequently. Calm winds account for 6 percent of observations. The strongest winds
were associated with westerly to northwesterly winds.
• The wind patterns shown on the 2012 wind rose are similar to the historical wind
patterns, indicating that wind conditions observed throughout 2012 were similar to
those observed historically. This is also true for the sample day wind rose.
18.3 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 18-4. 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 18-4. It is important to note which pollutants were sampled for at each site when reviewing
the results of this analysis. All three New Jersey sites sampled for VOCs and carbonyl
compounds.
18-21
-------
Table 18-4. Risk-Based Screening Results for the New Jersey Monitoring Sites
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Chester, New Jersey - CHNJ
Acetaldehyde
Formaldehyde
Benzene
Carbon Tetrachloride
1 ,2-Dichloroethane
1,3 -Butadiene
Hexachloro- 1 , 3 -butadiene
1 . 1 .2.2-Tetrachloroethane
1 ,2-Dibromoethane
£>-Dichlorobenzene
Chloroprene
Ethylbenzene
Propionaldehyde
0.45
0.077
0.13
0.17
0.038
0.03
0.045
0.017
0.0017
0.091
0.0021
0.4
0.8
Total
62
62
61
61
58
43
12
9
7
4
1
1
1
382
62
62
61
61
58
51
14
9
7
36
1
61
62
545
100.00
100.00
100.00
100.00
100.00
84.31
85.71
100.00
100.00
11.11
100.00
1.64
1.61
70.09
16.23
16.23
15.97
15.97
15.18
11.26
3.14
2.36
1.83
1.05
0.26
0.26
0.26
16.23
32.46
48.43
64.40
79.58
90.84
93.98
96.34
98.17
99.21
99.48
99.74
100.00
Elizabeth, New Jersey - ELNJ
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Formaldehyde
1 ,2-Dichloroethane
Ethylbenzene
/>-Dichlorobenzene
Propionaldehyde
Hexachloro- 1 , 3 -butadiene
1 , 1 ,2,2-Tetrachloroethane
1 ,2-Dibromoethane
Chloroprene
Trichloroethylene
0.45
0.13
0.03
0.17
0.077
0.038
0.4
0.091
0.8
0.045
0.017
0.0017
0.0021
0.2
Total
61
61
61
61
61
55
27
19
11
6
5
2
1
1
432
61
61
61
61
61
55
60
51
61
7
5
2
1
30
577
100.00
100.00
100.00
100.00
100.00
100.00
45.00
37.25
18.03
85.71
100.00
100.00
100.00
3.33
74.87
14.12
14.12
14.12
14.12
14.12
12.73
6.25
4.40
2.55
1.39
1.16
0.46
0.23
0.23
14.12
28.24
42.36
56.48
70.60
83.33
89.58
93.98
96.53
97.92
99.07
99.54
99.77
100.00
North Brunswick, New Jersey - NBNJ
Benzene
Carbon Tetrachloride
Formaldehyde
Acetaldehyde
1,3 -Butadiene
1 ,2-Dichloroethane
Hexachloro- 1 , 3 -butadiene
1 , 1 ,2,2-Tetrachloroethane
/>-Dichlorobenzene
1 ,2-Dibromoethane
Ethylbenzene
Trichloroethylene
0.13
0.17
0.077
0.45
0.03
0.038
0.045
0.017
0.091
0.0017
0.4
0.2
Total
60
60
60
59
58
58
12
10
9
7
5
1
399
60
60
60
60
60
58
14
10
41
7
60
25
515
100.00
100.00
100.00
98.33
96.67
100.00
85.71
100.00
21.95
100.00
8.33
4.00
77.48
15.04
15.04
15.04
14.79
14.54
14.54
3.01
2.51
2.26
1.75
1.25
0.25
15.04
30.08
45.11
59.90
74.44
88.97
91.98
94.49
96.74
98.50
99.75
100.00
18-22
-------
Observations from Table 18-4 include the following:
• Thirteen pollutants failed at least one screen for CHNJ; 70 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 CHNJ and therefore
were identified as pollutants of interest for this site. These eight include two carbonyl
compounds and six VOCs.
• Fourteen pollutants failed at least one screen for ELNJ, with nearly 75 percent of
concentrations for these 14 pollutants greater than their associated risk screening
value.
• Nine pollutants contributed to 95 percent of failed screens for ELNJ and therefore
were identified as pollutants of interest for this site. These nine include three carbonyl
compounds and six VOCs.
• Twelve pollutants failed at least one screen for NBNJ, with 77 percent of
concentrations for these 12 pollutants greater than their associated risk screening
value.
• Nine pollutants contributed to 95 percent of failed screens for NBNJ and therefore
were identified as pollutants of interest for this site. These nine include two carbonyl
compounds and seven VOCs.
• CJrtNJ, ELNJ, and NBNJ have six pollutants of interest in common: acetaldehyde,
formaldehyde, benzene, carbon tetrachloride, 1,3-butadiene, and 1,2-dichloroethane.
Of these, benzene, carbon tetrachloride, formaldehyde, and 1,2-dichloroethane failed
100 percent of screens for each site.
18.4 Concentrations
This section presents various concentration averages used to characterize pollution 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.
• 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.
18-23
-------
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Additional site-specific statistical summaries for the three
New Jersey monitoring sites are provided in Appendices J and L.
18.4.1 2012 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 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 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
New Jersey monitoring sites are presented in Table 18-5, 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.
18-24
-------
Table 18-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the New Jersey Monitoring Sites
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Ug/m3)
2nd
Quarter
Average
(Ug/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Ug/m3)
Chester, New Jersey - CHNJ
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
Formaldehyde
1, 1,2,2-Tetrachloroethane
Hexachloro- 1 ,3 -butadiene
62/62
61/61
51/61
61/61
58/61
62/62
9/61
14/61
1.61
±0.59
0.61
±0.08
0.03
±0.01
0.61
±0.09
0.08
±0.01
2.23
±0.39
0.01
±0.01
0.02
±0.02
1.45
±0.28
0.67
±0.11
0.02
±0.01
0.71
±0.05
0.09
±0.01
2.55
±0.83
0.01
±0.02
0.02
±0.02
1.32
±0.24
0.60
±0.24
0.05
±0.02
0.66
±0.03
0.06
±0.01
3.44
±0.80
0.01
±0.01
0.02
±0.02
1.67
±0.33
0.67
±0.09
0.06
±0.02
0.70
±0.04
0.06
±0.01
1.46
±0.31
0.01
±0.01
0.01
±0.01
1.51
±0.18
0.64
±0.07
0.04
±0.01
0.67
±0.03
0.07
±0.01
2.46
±0.35
0.01
±0.01
0.02
±0.01
Elizabeth, New Jersey - ELNJ
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
£>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
Propionaldehyde
61/61
61/61
61/61
61/61
51/61
55/61
60/61
61/61
61/61
2.31
±0.46
1.02
±0.15
0.14
±0.03
0.57
±0.10
0.06
±0.02
0.09
±0.02
0.34
±0.07
2.98
±0.46
0.35
±0.09
2.20
±0.62
0.91
±0.33
0.08
±0.02
0.74
±0.03
0.07
±0.03
0.09
±0.01
0.34
±0.08
3.50
±1.01
0.48
±0.14
3.36
±0.69
0.97
±0.15
0.14
±0.04
0.68
±0.03
0.09
±0.03
0.07
±0.01
0.46
±0.07
5.51
±1.08
0.74
±0.16
2.71
±0.89
1.28
±0.44
0.20
±0.05
0.70
±0.02
0.07
±0.05
0.06
±0.02
0.50
±0.17
3.47
±0.63
0.49
±0.17
2.66
±0.34
1.04
±0.14
0.14
±0.02
0.67
±0.03
0.07
±0.02
0.08
±0.01
0.41
±0.05
3.89
±0.47
0.52
±0.08
18-25
-------
Table 18-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the New Jersey Monitoring Sites (Continued)
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Ug/m3)
2nd
Quarter
Average
(Ug/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Ug/m3)
North Brunswick, New Jersey - NBNJ
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
£>-Dichlorobenzene
1 ,2-Dichloroethane
Formaldehyde
Hexachloro- 1 ,3 -butadiene
1 , 1 ,2,2-Tetrachloroethane
60/60
60/60
60/60
60/60
41/60
58/60
60/60
14/60
10/60
1.97
±0.43
0.76
±0.11
0.06
±0.02
0.62
±0.07
0.03
±0.02
0.08
±0.01
1.53
±0.27
0.02
±0.02
0.01
±0.01
1.20
±0.33
0.94
±0.49
0.07
±0.01
0.70
±0.06
0.05
±0.03
0.08
±0.02
1.89
±0.73
0.02
±0.03
0.02
±0.02
1.37
±0.22
0.68
±0.07
0.09
±0.02
0.67
±0.04
0.07
±0.02
0.08
±0.01
2.64
±0.62
0.03
±0.03
0.01
±0.01
1.10
±0.19
1.06
±0.15
0.14
±0.03
0.70
±0.03
0.06
±0.03
0.07
±0.01
1.26
±0.15
<0.01
±0.01
0
1.41
±0.17
0.86
±0.12
0.09
±0.01
0.67
±0.03
0.05
±0.01
0.08
±0.01
1.82
±0.26
0.02
±0.01
0.01
±0.01
Observations for CHNJ from Table 18-5 include the following:
• The pollutants of interest with the highest annual average concentrations are
formaldehyde and acetaldehyde. These are the two pollutants with annual average
concentrations greater than 1 |ig/m3.
• Concentrations of formaldehyde were lowest during the fourth quarter of 2012. Three
of the four concentrations less than 1 |ig/m3 were measured in December. Further,
seven of the 13 concentrations less than 1.5 |ig/m3 were measured at CHNJ during the
fourth quarter of the year (with none measured during the first quarter and three each
measured in the second and third quarters). Conversely, eight of the 10 measurements
greater than 4 |ig/m3 were measured in July or August, and are reflected in the higher
third quarter average.
• The quarterly average concentrations of benzene are fairly similar to each other, but
the confidence interval for the third quarter average is two to three times higher than
the other confidence intervals. A review of the data shows that the two highest
benzene concentrations were both measured in July. Benzene concentrations
measured during the third quarter range from 0.371 |ig/m3 to 2.35 |ig/m3. This range
is more than three times greater than the range of concentrations measured in each of
the other quarters.
18-26
-------
• The quarterly average concentrations of acetaldehyde are also similar in magnitude to
each other, but the confidence interval for the first quarter average is nearly twice the
other confidence intervals. A review of the data shows that the maximum
acetaldehyde concentration was measured on March 28, 2012 (5.38 |ig/m3) and is
more than twice the next highest concentration measured at CHNJ. This is explains
the higher variability in the first quarter measurements reflected by the confidence
intervals.
Observations for ELNJ from Table 18-5 include the following:
• The pollutants of interest with the highest annual average concentrations are
formaldehyde, acetaldehyde, and benzene. These are the only pollutants with annual
average concentrations greater than 1 |ig/m3.
• The concentrations of the carbonyl compound pollutants of interest for ELNJ appear
higher during the warmer months of the year, as illustrated by the third quarter
average concentrations, particularly for formaldehyde. However, the differences are
not statistically significant.
• Concentrations of several of the VOCs appear highest during the fourth quarter.
Although not significantly different, the fourth quarter averages of benzene,
ethylbenzene, and 1,3-butadiene are higher than the other quarterly averages and have
relatively large confidence intervals. A review of the data shows that the maximum
concentration of each of these pollutants was measured on the same sample day,
November 23, 2012. "Higher" concentrations of these pollutants were also measured
on samples collected on November 11, 2012 and October 18, 2012 for each of these
pollutants. The maximum concentrations of acetaldehyde and propionaldehyde were
also collected on November 11, 2012 at ELNJ.
Observations for NBNJ from Table 18-5 include the following:
• The pollutants of interest with the highest annual average concentrations are
formaldehyde and acetaldehyde. These are the only pollutants with annual average
concentrations greater than 1 |ig/m3.
• Concentrations of formaldehyde appear higher during the warmer months of the year,
although the differences among the quarterly averages are not statistically significant.
A review of the data shows that the two highest concentrations of formaldehyde
(those greater than 5 |ig/m3) were measured in June and July; further, 16 of the 18
formaldehyde concentrations greater than 2 |ig/m3 were measured during the second
and third quarters of the year.
• Even though the highest quarterly average concentration of benzene was calculated
for the fourth quarter, the confidence interval for the second quarter average
concentration is considerably higher than the confidence intervals for the other
quarterly averages. This indicates a relatively high level of variability associated with
the second quarter measurements. A review of the data shows that the maximum
concentration of benzene was measured on April 3, 2012 (4.00 |ig/m3). The next
18-27
-------
seven highest concentrations were all measured during the fourth quarter and were all
3
less than 2 |ig/m . Aside from the maximum concentration, the concentrations
3
measured during the second quarter of 2012 ranged from 0.545 |ig/m to 0.938 |ig/m .
Thus, it is the maximum concentration that is driving the large confidence interval for
the second quarter average.
• The fourth quarter average 1,3-butadiene concentration is higher than the other
quarterly averages (although not significantly so). A review of the data shows that the
three highest concentrations were all measured during the fourth quarter of 2012;
further, of the 18 concentrations greater than 0.1 |ig/m3, 12 were measured during the
fourth quarter (with two in the first quarter, one in the second quarter, and three in the
third quarter).
Tables 4-9 through 4-12 present the 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 nine times (CHNJ,
once; ELNJ, six times; and NBNJ, twice). At least one New Jersey site appears
among the rankings for each of the program-level pollutants of interest except carbon
tetrachloride. All three New Jersey sites appear for hexachloro-1,3-butadiene, with
NBNJ ranking first for this pollutant.
• ELNJ appears in Table 4-10 for both carbonyl compounds. ELNJ has the fourth
highest annual average concentration of formaldehyde and the fifth highest annual
average concentration of acetaldehyde among NMP sites sampling carbonyl
compounds.
18.4.2 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 shaded in
gray in Table 18-4 for each of the New Jersey sites. Figures 18-15 through 18-25 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 3.5.3.1.
18-28
-------
Figure 18-15. Program vs. Site-Specific Average Acetaldehyde Concentrations
CHNJ
ELNJ
^
9 12
Concentration {[og/m3)
Program:
Site:
IstQuartile
•
Site Average
0
2ndQuartile SrdQuartile
• •
Site Concentration Range
4thQuartile Average
D 1
Figure 18-16. Program vs. Site-Specific Average Benzene Concentrations
CHNJ
ELNJ
NBNJ
Concentration {[og/m3)
Program:
Site:
IstQuartile
D
Site Average
0
2ndQuartile SrdQuartile
• •
Site Concentration Range
4thQuartile Average
D 1
18-29
-------
Figure 18-17. Program vs. Site-Specific Average 1,3-Butadiene Concentrations
ELNJ
NBNJ
IHh •
] Program Max Concentration = 4.10 ug/m3
1 lo
1 1°
j Program Max Concentration
/ 3 '
«H
| Program Max Concentration
= 4. 10 ug/m3
0.75 1 1.25
Concentration {[og/m3)
Program:
Site:
IstQuartile
D
Site Average
0
2ndQuartile SrdQuartile
D D
Site Concentration Range
4thQuartile Average
D 1
Figure 18-18. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations
CHNJ
ELNJ
2 3
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o
18-30
-------
Figure 18-19. Program vs. Site-Specific Average /7-Dichlorobenzene Concentrations
NBNJ
0.6 0.8
Concentration {[og/m3)
Program:
Site:
IstQuartile
D
Site Average
0
2ndQuartile SrdQuartile
D D
Site Concentration Range
4thQuartile Average
D 1
Figure 18-20. Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations
CHNJ
Program Max Concentration = 17.01 ug/m3
ELNJ
r;
Program Max Concentration = 17.01 ug/m3
Program Max Concentration = 17.01 ug/m3
0.4 0.5 0.6
Concentration {[og/m3)
Program:
Site:
IstQuartile
Site Average
o
2ndQuartile SrdQuartile
Site Concentration Range
4thQuartile Average
18-31
-------
Figure 18-21. Program vs. Site-Specific Average Ethylbenzene Concentration
ELNJ
Concentration {pg/m3]
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o
Figure 18-22. Program vs. Site-Specific Average Formaldehyde Concentrations
CHNJ
Concentration {[og/m3
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
O
18-32
-------
Figure 18-23. Program vs. Site-Specific Average Hexachloro-l,3-Butadiene Concentrations
0.1 0.15
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o —
Figure 18-24. Program vs. Site-Specific Average Propionaldehyde Concentration
1 1.5
Concentration {[og/m3)
Program
Site:
: IstQuartile
Site Average
0
2ndQuartile SrdQuartile 4thQuartile Av
Site Concentration Range
?rage
Figure 18-25. Program vs. Site-Specific Average l,l?2,2-Tetrachloroethane Concentrations
CHNJ
0.06 0.09 0.12
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: SiteAverage Site Concentration Range
O
18-33
-------
Observations from Figures 18-15 through 18-25 include the following:
• Figure 18-15 for acetaldehyde shows all three sites. The range of acetaldehyde
concentrations measured is largest at ELNJ and smallest at NBNJ. The annual
average concentration for ELNJ is greater than the program-level average
concentration as well as the program-level third quartile. The annual averages for
CFINJ and NBNJ are less than the program-level average concentration and
similar to the program-level median concentration. The minimum concentration
measured at NBNJ is considerably less than the minimum concentration for the
other two New Jersey sites. The minimum concentration measured at ELNJ is
similar to the program-level first quartile.
• Figure 18-16 presents the box plots for benzene. Among the New Jersey sites, the
smallest range of benzene measurements was measured at CFINJ. The annual
average benzene concentration for CFINJ is less than the program-level median
concentration. The annual average concentration for ELNJ is greater than the
program-level average concentration and similar to the program-level third
quartile. NBNJ's annual average benzene concentration is just less the program-
level average concentration, even though the maximum benzene concentration
among the New Jersey sites was measured at NBNJ. The minimum benzene
concentration measured at NBNJ is similar to the program-level first quartile.
• Figure 18-17 presents the box plots for 1,3-butadiene. The program-level
maximum concentration (4.10 |ig/m3) is not shown directly on the box plots as the
scale has been reduced to 2 |ig/m3 to allow for the observation of data points at the
lower end of the concentration range. Among the New Jersey sites, the smallest
range of 1,3-butadiene concentrations was measured at CFINJ while the largest
range was measured at ELNJ. However, the maximum concentration measured at
all three sites is considerably less than the program-level maximum concentration.
The annual average 1,3-butadiene concentration for CFINJ is greater than the
program-level first quartile but less than the program-level median concentration.
The annual average concentration for NBNJ is just less than the program-level
average concentration but greater than the median concentration. ELNJ's annual
average concentration is greater than the program-level average concentration and
similar to the program-level third quartile. Ten non-detects were measured at
CHNJ while none were measured at ELNJ or NBNJ.
• Figure 18-18 presents the box plots for carbon tetrachloride. The range of
measurements collected at CHNJ and ELNJ are similar to each other, while the
minimum concentration measured at NBNJ is higher than the other two sites. The
annual average concentrations are the same for each of the New Jersey sites
(0.67 |ig/m3) and are just less than both the program-level average and median
concentrations. Note that the program-level median and average concentrations
are very similar to each other (less than 0.005 |ig/m3 separates these two
parameters).
18-34
-------
• Figure 18-19 presents the box plots for/>-dichlorobenzene for ELNJ and NBNJ,
the two sites for which this is a pollutant of interest. Note that the program-level
first quartile is zero and therefore not visible on the box plot. The minimum
concentrations for both ELNJ and NBNJ are zero as non-detects were measured at
these sites. The range ofp-dichlorobenzene concentrations is slightly greater for
ELNJ than NBNJ. The annual average concentration for NBNJ is just less than
the program-level average concentration while the annual average concentration
for ELNJ is just greater than the program-level average concentration (roughly
0.02 |ig/m3 separates these two annual averages).
• Figure 18-20 presents the box plots for 1,2-dichloroethane for all three sites.
Similar to 1,3-butadiene, the program-level maximum concentration
(17.01 |ig/m3) is not shown directly on the box plots as the scale has been reduced
to 1 |ig/m3 to allow for the observation of data points at the lower end of the
concentration range. The program-level average concentration is greater than the
program-level third quartile for this pollutant and is greater than or similar to the
maximum concentration measured at most sites sampling 1,2-dichloroethane. This
is because the program-level average is being driven by the higher measurements
collected at a few monitoring sites. Figure 18-20 shows that the maximum
1,2-dichloroethane concentrations measured at the New Jersey sites are two
orders of magnitude less than the average concentration across the program. The
annual averages for each site are similar to the median concentration measured at
the program level. A few non-detects of 1,2-dichloroethane were measured at
each New Jersey site.
• Figure 18-21 presents the box plot for ethylbenzene for ELNJ, the only site for
which this is a pollutant of interest. The annual average concentration for ELNJ is
just greater than the program-level average and similar to the program-level third
quartile. A single non-detect of ethylbenzene was measured at ELNJ.
• Figure 18-22 presents the box plots for formaldehyde for all three sites. The
annual average concentration of formaldehyde is greatest for ELNJ and lowest for
NBNJ. The annual average for ELNJ is greater than both the program-level
average concentration and third quartile. The annual average concentration for
CFINJ is less than the program-level average but greater than the program-level
median. The annual average concentration for NBNJ is less than both the
program-level average and median concentrations. Similar to many of the other
pollutants of interest for the New Jersey sites, the minimum concentration for
ELNJ is just less than the program-level first quartile.
18-35
-------
• Figure 18-23 presents the box plots for hexchloro-1,3-butadiene for CHNJ and
NBNJ (this pollutant was not a pollutant of interest for ELNJ). 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 annual average hexchloro-1,3-butadiene concentrations for both sites
are greater than the program-level average concentration. There were 14
measured detections collected at each of these sites. The range of hexachloro-1,3-
butadiene measurements collected at CHNJ is fairly similar in magnitude to the
range of measurements collected at NBNJ.
• Figure 18-24 presents the box plot for propionaldehyde for ELNJ. The minimum
concentration measured at ELNJ is greater than the program-level first quartile.
The annual average concentration for ELNJ is greater than both the program-level
average and third quartile. ELNJ has the second highest annual average
concentration of propionaldehyde among NMP sites sampling carbonyl
compounds (second only to BTUT).
• Figure 18-25 presents the box plots for 1,1,2,2-tetrachloroethane for CHNJ and
NBNJ. Note that the first, second, and third quartiles for 1,1,2,2-tetrachloroethane
are zero at the program-level and therefore not visible on the box plots due to the
large number of non-detects. The annual average 1,1,2,2-tetrachloroethane
concentrations for both sites are greater than the program-level average
concentration. There were nine measured detections collected at CFINJ and 10
collected at NBNJ.
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 3.5.3.2.
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 CFINJ and NBNJ since 2001.
Thus, Figures 18-26 through 18-51 present the 1-year statistical metrics for each of the pollutants
of interest first for CFINJ, 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 is not provided, although the range and quartiles are still presented.
18-36
-------
Figure 18-26. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at CHNJ
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
O 5th Percentile
— Minimum
— Maximum
95th Percentile
..<>... Average
A 1 -year average is not presented because sampling under the NMP did not begin until May 2001.
Observations from Figure 18-26 for acetaldehyde measurements collected 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 is not
presented, although the range of measurements is provided.
• The maximum acetaldehyde concentration was measured on June 2, 2004
(29.1 |ig/m3). The second highest concentration was also measured in 2004
(11.5 |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, after which the median and average concentrations leveled out through
2010. Note that the high concentrations measured in 2004 and 2005 result in
confidence intervals that are relatively large.
• The maximum and 95th percentile increased from 2009 to 2010 and again in 2011.
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.
18-37
-------
Figure 18-27. Yearly Statistical Metrics for Benzene Concentrations Measured at CHNJ
ation (Mg/m
Concent
o
0.5
—i
t-
-P-
-i
<
'-1
>-
X.
f-'
I
T
-C
i-
O
I
n i T
n n
_ o-
IT
2001 1 2002 2003 2004 2005 2 2006 2007
Year
O 5th Percentile — Min mum — Median —
X>
u
2008
Vlaximum
-i
t-
u
2009
1
0
y
2010
-C
>-
o-
T
2011
O 95thPercenti e —O
"°~
O
-t-1
2012
•• Average
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 18-27 for benzene measurements collected 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 is
not presented, although the range of measurements is provided. In addition, a 1-year
average for 2005 is not provided due to completeness less than 85 percent.
• The maximum benzene concentration measured at CHNJ was measured in 2012
(2.35 |ig/m3), although a similar concentration was also measured in 2008. 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, three in 2009, and one
each in 2011 and 2012).
• The 1-year average and median concentrations exhibit a decreasing trend through
2007, although no 1-year average is shown for 2001 or 2005. Even though an increase
in the 1-year average concentration is shown from 2007 to 2008, this increase is being
driven not by the three measurements greater than 1 |ig/m3 but by the measurements
in the middle of the concentration range. This evident from the increase shown in the
median concentration. The number of concentrations between 0.5 |ig/m3 and
0.75 |ig/m3 doubled from 10 to 20 in 2008.
18-38
-------
• Even though the 1-year average and median concentrations exhibit decreases from
2008 to 2009, the 95th percentile is at a maximum for 2009. This is also true for the
difference between the 5th and 95th percentiles, or the range within which the
majority of concentrations fall. Conversely, the difference between the 5th and 95th
percentiles is at a minimum for 2010.
• An increase in the 1-year average, median, 95th percentile, and maximum
concentration is shown from 2010 to 2011 and again for 2012. Additional years of
sampling are needed to determine if this trend continues.
Figure 18-28. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at CHNJ
I °'30
s
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Year
5th Percentile — Minimum — Median — Maximuir
95th Percentile
A 1 -year average is not presented because sampling under the NMP did not begin until May 2001.
A 1-year average is not presented due to low completeness in 2005.
Observations from Figure 18-28 for 1,3-butadiene measurements collected at CHNJ
include the following:
• The maximum 1,3-butadiene concentration was measured in 2003 and is nearly twice
the next highest concentration, which was measured in 2008. 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,
18-39
-------
when the percentage of non-detects reached a minimum of 17 percent. After 2008, an
increasing number of non-detects was reported. After 2010, the number of non-
detects began decreasing again, dropping to 18 percent for 2012.
• The 1-year average and median concentrations have a decreasing trend from 2008
through 2010 and then an increasing trend through 2012. These changes correspond
with the changes in the number of non-detects/measured detections discussed above.
Figure 18-29. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
CHNJ
•1 0.8
§
s
u
0.4
••
-o-
_
T
f T
i
2001
-o-
O
r
o
r-(
O
T
~°~
-(
<
H
>-
H
I
H
o-
..-
i
-
y-
' '
T
o-...
"" ""--. ...•O
T UJ
1
1 2002 2003 2004 2005 2 2006 2007 2008 2009 2010 2011 2012
Year
O 5th Percentile — Min mum — Median —
Maximum
0 95th Percentile ...^>... Average
1A 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 18-29 for carbon tetrachloride measurements collected at
CFDSTJ include the following:
• The range of carbon tetrachloride measurements increased significantly from 2001 to
2002. This is predominantly due to a few non-detects that were measured between
2002 and 2005. After 2005, only one non-detect was reported.
• 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
for 2008. There were 14 measurements in 2008 that were greater than the maximum
concentration measured in 2007. The number of measurements greater than 0.6 |ig/m3
doubled from 2007 to 2008.
18-40
-------
• The minimum concentration measured in 2009 increased by an order of magnitude.
Although the 1-year average increased slightly from 2008 to 2009, the median
concentration decreased slightly. The decreasing trend in the median concentration
continues through 2011. The 1-year average exhibits a similar trend.
• Although the minimum concentration decreased from 2011 to 2012, the 5th percentile
is at a maximum for 2012.
Figure 18-30. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
CHNJ
£
C
Concentrat
3 C
i k
0.00
Maximum
Concentration for
2008 is 1.27 u.g/m3.
1 JL <
I
.<>*
***'v^\i| |
2001 1 2002 2003 2004 2005 2 2006 2007 2008 2009 2010 2011 2012
Year
O 5th Percentile — Minimum — Median — Maximum 0 95th Percentile ...^... Average
2 A 1-year average is not presented due to low completeness in 2005.
Observations from Figure 18-30 for 1,2-dichloroethane measurements collected 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 number 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 concentration shown for the
later years of sampling.
• 2012 is the first year that the 5th percentile and median concentration are greater than
zero. Aside from the three non-detects, the range of measurements collected in 2012
is relatively small, ranging from 0.0527 |ig/m3to 0.121 |ig/m3. The 1-year average
18-41
-------
and median concentrations calculated for 2012 are less than 0.001 |ig/m3 apart,
indicating little variability associated with the measurements collected in 2012.
Figure 18-31. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at CHNJ
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
O 5th Percentile — Minimum — Median — Maximurr
0 95th Percentile
A 1 -year average is not presented because sampling under the NMP did not begin until May 2001.
Observations from Figure 18-31 for formaldehyde measurements collected 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 and
almost four times the maximum concentrations shown for other years.
• A decreasing trend in the 1-year average and median formaldehyde concentrations is
shown though 2006, after which the 1-year average and median concentrations
leveled out through 2009. Less than 0.5 |ig/m3 separates the 1-year averages
calculated for the period between 2006 and 2009.
• The 1-year and median concentrations decreased significantly for 2010, when the
1-year average concentration reached a minimum. This is due primarily to the
measurements at the lower end of the concentration range. The number of
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. The 95th percentile for 2011 is greater than
18-42
-------
the maximum concentration for 2010. Although the range of measurements decreased
for 2012, little change is shown in the 1-year average concentration.
Figure 18-32. Yearly Statistical Metrics for Hexachloro-l,3-Butadiene Concentrations
Measured at CHNJ
•|o.20
.9
«
1
u
-i
>-
20011 2002 2003 2004 2005
_
111 11
T
0
2 2006 2007 2008 2009 2010 2011 2012
Year
• 5th Percentile — Minimum — Median — Maximum 0 95th Percentile ...^>... Average
1A 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 18-32 for hexachloro-l,3-butadiene measurements collected at
CHNJ include the following:
• There were no measured detections of hexachloro-1,3-butadiene measured during the
first 4 years of sampling.
• The number of measured detections increased to seven for 2005, representing
14 percent of measurements, then decreased each year through 2009, when again no
measured detections were measured. The number of measured detections increased to
one for 2010, then four for 2011, then up to 12 for 2012, or nearly 20 percent, the
maximum number of measured detections since sampling began.
18-43
-------
Figure 18-33. Yearly Statistical Metrics for 1,1,2,2-Tetrachloroethane Concentrations
Measured at CHNJ
oncentration (ng/m
D C
3 5
0.00
, 1
>
£"
2001 1 2002 2003 2004 2005 2 2006 2007 2008 2009 2010 2011 2012
Year
0 5th Percentile — Minimum — Median — Maximum O 95th Percentile ...^... Average
2 A 1-year average is not presented due to low completeness in 2005.
Observations from Figure 18-33 for 1,1,2,2-tetrachloroethane measurements collected at
CHNJ include the following:
• Between 2001 and 2011, a total of seven measured detections of 1,1,2,2-
tetrachloroethane were measured at CHNJ (two in 2005, one each in 2009 and 2010,
and three in 2011). The number of measured detections for 2012 is greater than all the
previous years combined (9).
• Additional years of sampling are needed to determine if the number of measured
detections continues to increase.
18-44
-------
Figure 18-34. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at ELNJ
o
T
i
T
T
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
O 5th Percentile - Minimurr
— Maximum O 95th Percentile
A 1-year average is not presented due to low completeness in 2000.
Observations from Figure 18-34 for acetaldehyde measurements collected 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 is not presented for 2000, although the range of
measurements is provided.
• The maximum acetaldehyde concentration was measured in 2007, although a
concentration of similar magnitude was also measured in 2005. In total, 22
concentrations greater than 10 |ig/m3 have been measured at ELNJ, all of which were
measured prior to 2008.
• The range of measurements 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
1-year average concentration is shown from 2007 to 2008, where the 1-year average
decreased by more than half. The range of measurements collected in 2008 is more
similar to the range shown before 2003.
• Although an increasing trend is also shown between 2008 and 2011, the 1-year
averages are roughly half the magnitude of those shown before 2008.
18-45
-------
All of the statistical parameters exhibit decreases from 2011 to 2012.
Figure 18-35. Yearly Statistical Metrics for Benzene Concentrations Measured at ELNJ
I
§
1
s
3
,o
2000 l 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
O 5th Percentile - Min mum
— Maximum O 95th Percenti e
A 1-year average is not presented due to low completeness in 2000.
Observations from Figure 18-35 for benzene measurements collected at ELNJ include the
following:
• The maximum benzene concentration (34.3 |ig/m3) was measured in 2008 and is
more than four times higher than the next highest concentration (measured in 2009).
The third highest concentration was also measured in 2009. In all, only five benzene
concentrations greater than 5 |ig/m3 have been measured at ELNJ.
• A 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 a negligible 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. That the median increased by less than
0.02 |ig/m3 between 2007 and 2008 further indicates that this outlier is the primary
driver pushing the 1-year average concentration upward. However, the minimum
concentration doubled from 2007 to 2008, indicating that the outlier may not be the
only factor.
18-46
-------
• Even though two of the three highest concentrations were measured 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.
• Figure 18-35 shows that benzene concentrations measured in 2010, 2011, and 2012
were fairly constant. The difference in the 1-year average concentrations for these
years is less than 0.02 |ig/m3.
Figure 18-36. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at ELNJ
2000 2001
5th Percentile
— Maximum
95th Percentile
A 1-year average is not presented due to low completeness in 2000.
Observations from Figure 18-36 for 1,3-butadiene measurements collected at ELNJ
include the following:
• The maximum concentration of 1,3-butadiene was measured 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.
• 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, even the
median is zero, indicating that at least half of the measurements were non-detects.
Between 2000 and 2006, the percentage has ranged from 5 percent (2006) to 57
percent (2004). After 2006, only two non-detects have been measured (both in 2011).
18-47
-------
• Figure 18-36 shows a decreasing trend in the 1-year average concentration through
2004, after which the 1-year average concentration remain fairly constant. 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 2012, the
1-year average concentration has ranged from 0.12 |ig/m3 (2010) to 0.16 |ig/m3 (2006
and 2009).
Figure 18-37. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
ELNJ
£
I
^
I
o
.0
T
2000 2001 2002 2003
2006 2007
Year
5th Percentile - Minimum ~ Median — Maximum
95th Percentile "-O"'Averagf
A 1-year average is not presented due to low completeness in 2000.
Observations from Figure 18-37 for carbon tetrachloride measurements collected 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 no more than
0.1 |ig/m3 during the period from 2001 to 2007, even though the range of
measurements varies. All of the statistical parameters exhibit an increase in
magnitude from 2007 to 2008. 2008 is the first year that the 1-year average
concentration is greater than 0.6 |ig/m3; all of the 1-year averages between 2008 and
2012 are greater than 0.6 |ig/m3.
18-48
-------
• The difference between the 5th percentile and 95th percentile, or the range within
which the majority of measurements fall, has been decreasing since 2005 and reaches
a minimum for 2012. Less than 0.35 |ig/m3 separates these parameters for 2012.
Figure 18-38. Yearly Statistical Metrics for/7-Dichlorobenzene Concentrations Measured at
ELNJ
E°'6
i
i
•43
c
s
8 0.4
T
I
O
20001 2001
o
r-C
•-••'"
y--"
->
r
-
-i
j-
o
T
i— i
H
o
5-I
UJ
2002 2003 2004 2005 2006 2007 2008
-<
•
r-O-
I 1
™ ^ "• o
-p-l *. ^ ^
2009 2010 2011 2012
Year
• 5th Percentile
— Minimum — Median
— Maximum 0
95th Percentile ...<>... Average
1A 1-year average is not presented due to low completeness in 2000.
Observations from Figure 18-38 for/>-dichlorobenzene measurements collected at ELNJ
include the following:
• At least one non-detect has been measured at this site each year since the onset of
sampling. For many years, this number is higher as indicated by the 5th percentile
and/or median concentration. The percentage of non-detects has ranged from
5 percent (2006 and 2009) to 95 percent (2004).
• Most of the increases or decreases shown in the 1-year average concentrations
correspond with a substantial increase or decrease in the number of non-detects for a
particular year. This is not true is for 2008 and 2009. For both years, the number of
non-detects is three, representing 5 percent of the measurements. The change in the
1-year average (and median) is a result in the concentrations measured, both at the
upper and lower end of the concentration range. The number of measurements greater
than 0.3 |ig/m3 is three times higher for 2008 than 2009; the number of measurements
less than 0.1 |ig/m3 increased from 22 in 2008 to 35 in 2009.
18-49
-------
• For 2010, the number of non-detects increased to 25 percent of the measurements (up
from 5 percent for 2009), yet only a small decrease in the 1-year average is shown
(compared to other years) and the median itself did not change.
• The data collected between 2010 and 2012 seem to exhibit less variability than the
previous years. The 1-year average concentration for each year between 2010 and
2012 is less than 0.1 |ig/m3 and non-detects account for between 15 percent and
25 percent of the measurements collected.
Figure 18-39. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
ELNJ
Concentration (jig/m3)
00000000
o -
^^ , -^ ^y ^^ ,^^r ^
a ^
rL
O
.'•
r
n
2000 1 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Year
O 5th Percentile - Minimum ~ Median — Maximum • 95th Percentile
H
>-|
/
.#"'
2011 2012
•••<>•" Average
A 1-year average is not presented due to low completeness in 2000.
Observations from Figure 18-39 for 1,2-dichloroethane measurements collected 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 no measured detections were 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 measurements collected in 2012 is relatively small,
ranging from 0.061 |ig/m3to 0.144 |ig/m3. The 1-year average and median
18-50
-------
concentrations calculated for 2012 are approximately 0.0015 |ig/m3 apart, indicating
relatively little variability associated with the measurements collected in 2012.
Figure 18-40. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at
ELNJ
20001 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
5th Percentile - Minimum ~ Median — Maximum
95th Percentile
A 1-year average is not presented due to low completeness in 2000.
Observations from Figure 18-40 for ethylbenzene measurements collected at ELNJ
include the following:
• There is a steady decreasing trend in the 1-year average and median concentrations
between 2001 and 2007.
• A significant increase in these parameters is shown for 2008. The median
concentration for 2008 is greater than the 95th percentile for 2007. The number of
measurements greater than 1 |ig/m3 increased from one in 2007 to 16 in 2008.
• The measurements collected in 2009 more resemble those collected in 2007 than
2008, with the exception of the maximum concentration measured.
• The smallest range of ethylbenzene measurements was collected in 2010, with all
measurements spanning less than 1 |ig/m3.
• Between 2009 and 2012, the majority of concentrations fell within a fairly similar
range and the 1-year average concentration did not change significantly, ranging from
0.41 |ig/m3 (2012) to 0.51 |ig/m3 (2011).
18-51
-------
Figure 18-41. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
ELNJ
I
i
I
T
o
T
i
T
O 5th Percentile - Minimurr
Median — Maximum • 95th Percentile
A 1-year average is not presented due to low completeness in 2000.
2000 l 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Observations from Figure 18-41 for formaldehyde measurements collected at ELNJ
include the following:
• The maximum formaldehyde concentration was measured at ELNJ in 2010, although
a similar measurement was also collected in 2000. A total of 12 concentrations
greater than 10 |ig/m3 have been measured at ELNJ.
• After a decreasing trend through 2002, there was a significant increase in
formaldehyde concentrations from 2002 to 2003, as shown by the median, which
more than doubled, and the 1-year average concentration, which increased by roughly
60 percent. The number of formaldehyde concentrations greater than 4 |ig/m3 nearly
tripled from 2002 to 2003 (from 9 to 25) while the number of measurements less than
2 |ig/m3 decreased by half (from 29 to 15).
• Between 2004 and 2007, there was relatively little change in the 1-year average
formaldehyde concentrations, which ranged from 4.52 |ig/m3 (2006) to 4.70 |ig/m3
(2005) during this time.
• Similar to acetaldehyde, the 1-year average concentration of formaldehyde decreased
significantly between 2007 and 2008, after which an increasing trend is shown
through 2010. While the trends graph for acetaldehyde shows a continued increase for
2011 followed by a decrease for 2012, formaldehyde concentrations decrease for
2011 then increase slightly for 2012.
18-52
-------
Figure 18-42. Yearly Statistical Metrics for Propionaldehyde Concentrations Measured at
ELNJ
SM
§
1
c
S 15
3
1.0
1
T
2000
I
0
2001
2002
• 5th Percentile
-<
^
1-
j
1 I 1 iLri
i i ^L rn rh
r°n r°n
1 pL
... 3. o .0; °" -
TT UjJ l-t-l LjJ Lf-l
i i *? i ^ i i i i i i i i
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Year
- M nimum ~ Median — Maximum O 95th Percenti e ...^... Average
Observations from Figure 18-42 for propionaldehyde measurements collected at ELNJ
include the following:
• A decrease in the concentrations is shown between 2000 and 2002, as the maximum
concentration measured in 2002 is the same as the median concentration calculated
for 2000 and is less than the 95th percentile for 2001.
• The maximum propionaldehyde concentration was measured at ELNJ in 2003
(3.48 |ig/m3). The next two highest concentrations were also measured in 2003 and
together these three measurements are the only concentrations greater than 2 |ig/m3
measured at ELNJ since the onset of sampling. In addition, there were 12
measurements collected in 2003 greater in magnitude than the maximum
concentration measured in 2002. This explains the significant increase in the 1-year
average concentration.
• A steady increasing trend in the 1-year average concentration of propionaldehyde is
shown between 2002 and 2010 (with the exception of 2003 to 2004). The 1-year
average concentration more than tripled over the period (from 0.16 |ig/m3 in 2002 to
0.54 |ig/m3 in 2010), after which some fluctuation is shown.
18-53
-------
Figure 18-43. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at NBNJ
25.0
20
i15
c
.9
S
0 -
0 -
8 10.0
5.
0.
0 -
o -
Maximum
Concentration for
2004 is 111 U£/m3.
T
u
-
2001
O
s
1
2002
j
o
2003
5th Percentile
/
;" |
-o-
^m
T
-i
1—
O.
u
1
T T
n
o T r*^
u •••§•-• H y 9 8 9
2004 2005 2006 2007 2008 2009 2010 2011 2012
- Min mum
Year
Median — Maximum O 95th Percentile ...^>... Average
A 1 -year average is not presented because sampling under the NMP did not begin until May 2001.
Observations from Figure 18-43 for acetaldehyde measurements collected 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 is not
presented, although the range of measurements is provided.
• The maximum acetaldehyde concentration was measured in 2004. This concentration
(111 |ig/m3) 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 29 concentrations greater than 8 |ig/m3, 28 were measured at NBNJ in 2004 or
2005 (the one other was measured in 2008). 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, 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 from 2005 through 2007, as
do all of the other statistical parameters. Between 2008 and 2011, the 1-year average
concentration fluctuates between 2 |ig/m3 and 3 |ig/m3. The 1-year average, as well as
most of the other statistical parameters, is at a minimum for 2012 (1.41 |ig/m3).
18-54
-------
Figure 18-44. Yearly Statistical Metrics for Benzene Concentrations Measured at NBNJ
±
o
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
O 5th Percentile - Minimurr
Median — Maximum O 95th Percentile
A 1 -year average is not presented because sampling under the NMP did not begin until May 2001.
Observations from Figure 18-44 for benzene measurements collected at NBNJ include
the following:
• Sampling for VOCs 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 is not presented,
although the range of measurements is provided.
• The maximum benzene concentration was measured in 2012 (4.00 |ig/m3); aside from
this measurement, only three 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,
when both the median and 1-year average concentrations were at a minimum.
• Between 2008 and 2011, the 1-year average concentration is fairly static, ranging
from 0.65 |ig/m3 (2010) to 0.70 |ig/m3 (2011), even though there is some fluctuation
in the range of concentrations measured.
• The 1-year average benzene concentration increased 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 measurements in 2011 that are less than the
minimum concentration measured in 2012 (0.49 |ig/m3). In addition, the number of
18-55
-------
measurements in the mid- to upper-end of the concentration increased substantially
for 2012. While the number of measurements between 0.5 |ig/m3 and 0.75 |ig/m3 was
the same for both years, the number of benzene measurements between 0.75 |ig/m3
'and 1 |ig/m3 increased from two in 2011 to 20 in 2012.
Figure 18-45. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at NBNJ
O
..o
^
2006 2007
Year
O 5th Percentile
— Maximum O 95th Percentile
A 1 -year average is not presented because sampling under the NMP did not begin until May 2001.
Observations from Figure 18-45 for 1,3-butadiene measurements collected at NBNJ
include the following:
• The maximum 1,3-butadiene concentration was measured in 2005 and is the only
measurement greater than 0.40 |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. The median concentration increased from 2004 to 2005, indicating that the
number of non-detects decreased, although 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. The number of non-
detects increased considerably for 2011, from 4 percent in 2010 to 29 percent for
2011, an increase that is evident from the return of the 5th percentile to zero for 2011.
There were no non-detects measured in 2012, as indicating by the minimum
concentration, which is greater than zero for the first time.
18-56
-------
• The 1-year average concentration of 1,3-butadiene decreased significantly from 2003
to 2004. This is likely a result of the change in the number of non-detects as well as a
reduction in the range of measurements. The number of non-detects increased from
69 percent to 93 percent from 2003 to 2004. Thus, many zeros were substituted into
this average. The increase in the 1-year average concentration shown from 2004 to
2005 results from a combination of fewer non-detects and a larger range of
measurements. The number of non-detects decreased to 47 percent for 2005.
• The 1-year average concentration exhibits little change between 2005 and 2011,
ranging from 0.046 |ig/m3 (2009) to 0.057 |ig/m3 (2008).
• The 1-year average concentration increases significantly from 2011 to 2012.
Increases are also exhibited by each of the other statistical parameters. This is
partially due to the decrease in non-detects (and thus, zeroes substituted for non-
detects in the calculations) from 29 percent to 0 percent from 2011 to 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 increased from eight in
2011 to 18 in 2012.
Figure 18-46. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
NBNJ
1
Concentrat
c
T
I
2001
—i
r°n T
)—
0
, T T H i r*i r
o
n
o
r"!
o
r*i £ ...^
..-••" •• * 0' p^ '^^
T T
T T U T
1 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Year
• 5th Percentile - M nimum ~ Median — Maximum O 95th Percenti e ...^... Average
18-57
-------
Observations from Figure 18-46 for carbon tetrachloride measurements collected at
NBNJ include the following:
• The range of carbon tetrachloride measurements collected 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 measurements collected each year.
• The 1-year average concentration changed little between 2002 and 2005, ranging
from 0.49 |ig/m3 to 0.60 |ig/m3. A slight increase in the 1-year average concentration
is shown from 2005 to 2006, a result of higher concentrations at both the lower and
upper end of the concentration range. Note that between 2004 and 2007, the median
concentration varied by 0.003 |ig/m3.
• All of the statistical parameters exhibit increases for 2008. The minimum
concentration increased six-fold from 2007 to 2008. In addition, there were 19
measurements collected in 2008 that were greater than the maximum concentration
for 2007.
• A decreasing trend in the measurements is shown after 2008 and continues through
2010. Even though the maximum concentrations continue to decrease for 2011 and
2012 and the difference between the 5th percentile and 95th percentile decreases each
year, the 1-year average and median concentrations exhibit an increasing trend for the
final 2 years shown.
• The box and whisker plots for this pollutant appear "inverted" for the second half of
the sampling period, with the minimum concentration extending farther away from
the majority of the measurements rather than the maximum (see benzene or
1,3-butadiene as examples). This is a common feature of the trends graphs for carbon
tetrachloride across many NMP sites.
18-58
-------
Figure 18-47. Yearly Statistical Metrics for/7-Dichlorobenzene Concentrations Measured at
NBNJ
Concentration (jig/m3)
1
2001
0
r <
] L
1 2002 2003 2004
SthPercentile
— Minimum
-\
m
20
Maximum
Concentration for
2005 s 21.9 u.g/m3.
*.
M T T
TTfTj A ,*, n r1!
'••^ <>. ....
... ^^ Kp
5 2006 2007 2008 2009 2010 2011 2012
Year
Med an - Maximum 0 95th Percentile ...^>... Average
Observations from Figure 18-47 for/>-dichlorobenzene measurements collected at NBNJ
include the following:
• The maximum />-dichlorobenzene concentration measured at NBNJ was measured on
July 27, 2005 (21.9 |ig/m3) and is significantly greater than all other concentrations of
this pollutant for this site. All other />-dichlorobenzene concentrations measured at
NBNJ are less than 0.60 |ig/m3.
• The median concentration ofp-dichlorobenzene is zero for the first 5 years of
sampling, indicating that at least half of the measurements were non-detects (or zeros
substituted for non-detects). There were no measured detections of this pollutant
measured at NBNJ in 2004. Between 2001 and 2005, the number of non-detects
ranged from 68 percent (2005) to 100 percent (2004).
• The number of non-detects decreased significantly after 2005. Between 2006 and
2009, the number of non-detects varied between 13 percent (2009) and 17 percent
(2007). The number of non-detects increased to 40 percent for 2010, after which it
hovered around 30 percent for the last 2 years shown in Figure 18-47.
• The 1-year average concentration is close to zero for the first few years of sampling
due to the low number of measured detections. For 2005, the 1-year average
increased significantly, but is driven solely by the outlying concentration measured
18-59
-------
that year. If the maximum concentration for 2005 is removed from the data set, the
1-year average concentration decreases by an order of magnitude.
• Between 2006 and 2012, the 1-year average concentration ofp-dichlorobenzene
ranges from 0.048 |ig/m3 (2010) and 0.098 |ig/m3 (2008). The maximum
concentration measured in 2008 is considerably higher than the maximum
concentrations measured for the other years between 2006 and 2012. Even the 95th
percentile for 2008 is greater than the maximum concentrations for each year during
this time frame. 2008 has the greatest number ofp-dichlorobenzene measurements
greater than 0.1 |ig/m3 (18).
Figure 18-48. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
NBNJ
•;O
1
o
1
1
O
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
O 5th Percentile - Minimum " Median — Maximum • 95th Percentile ...^>... Average
A 1 -year average is not presented because sampling under the NMP did not begin until May 2001.
Observations from Figure 18-48 for 1,2-dichloroethane measurements collected at NBNJ
include the following:
• There were no measured detections of 1,2-dichloroethane between 200land 2004.
Between one and four measured detections were measured between 2005 and 2007,
after which no measured detections were measured in 2008. After 2008, the number
of measured detections increased significantly, from three in 2009, to 11 for 2010, 18
in 2011, and 58 for 2012. This increase in the number of measured detections is very
similar to what was exhibited by the measurements collected at ELNJ. This also
explains the significant increase in the 1-year average concentrations shown for the
later years of sampling.
18-60
-------
• 2012 is the first year that the median concentration is greater than zero. Aside from
the two non-detects, the range of measurements collected in 2012 is relatively small,
ranging from 0.053 |ig/m3to 0.140 |ig/m3. The 1-year average and median
concentrations calculated for 2012 are less than 0.001 |ig/m3 apart, indicating
relatively little variability associated with the measurements collected in 2012.
Figure 18-49. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
NBNJ
Maximum
Concentration for
2004 is 96.1 ug/m3.
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
5th Percentile — Minimum ~ Median - Maximum
95th Percentile
>... Average
A 1 -year average is not presented because sampling under the NMP did not begin until May 2001.
Observations from Figure 18-49 for formaldehyde measurements collected 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 the
next highest concentration (27.7 |ig/m3, measured in 2011). Concentrations greater
than 20 |ig/m3 have been measured during five of the 12 years shown.
• 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 measurement
collected in 2004. If the maximum concentration was excluded from the calculations
for 2004, the statistical parameters shown for 2004 would fall between those of 2003
and 2005, exhibiting lesser increases.
18-61
-------
• After 2005, concentrations of formaldehyde decreased steadily, reaching a minimum
in 2008. This year also has the smallest range of formaldehyde measurements,
although a similar range was 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. That said, the median concentrations are nearly the same for 2011 and
2012 and vary by less than 0.20 |ig/m3 between 2010 and 2012.
Figure 18-50. Yearly Statistical Metrics for Hexachloro-l,3-Butadiene Concentrations
Measured at NBNJ
„
c
o
1
J0.15
_••"
—i
t-
2001 * 2002 2003 2004 2005
_
I | T
i— O-i
'"' •<>...... o o
2006 2007 2008 2009 2010 2011 2012
Year
• 5th Percentile - Minimum ~ Median — Maximum O 95th Percentile ...^>... Average
1A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
Observations from Figure 18-50 for hexachloro-l,3-butadiene measurements collected 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 for 2007 through 2009. A single measured
detection was reported for 2010. The number of measured detections increased to
18-62
-------
eight for 2011 and to 11 for 2012, representing 18 percent of the measurements, the
maximum number of measured detections since sampling began at NBNJ.
Figure 18-51. Yearly Statistical Metrics for 1,1,2,2-Tetrachloroethane Concentrations
Measured at NBNJ
0.1
0.1
0.0
•3
centration (\i
o
b
S
0.0
0.0
0.0
fc» -4
fcr 12
-(
i—
2001 1 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Year
• 5th Percentile — Minimum ~ Median — Maximum 0 95th Percentile ...^>... Average
A 1 -year average is not presented because sampling under the NMP did not begin until May 2001
Observations from Figure 18-51 for 1,1,2,2-tetrachloroethane measurements collected at
NBNJ include the following:
• Between 2001 and 2011, a total of six measured detections of 1,1,2,2-
tetrachloroethane were measured at NBNJ (one each in 2005, 2006, and 2010, and
three in 2011). The number of measured detections for 2012 is greater than all the
previous years combined (9).
• Additional sampling is needed to determine if the increase in the number of measured
detections continues.
18.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk at
each New Jersey monitoring site. Refer to Sections 3.3 and 3.5.3 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.
18-63
-------
18.5.1 Risk-Based Screening Assessment Using MRLs
A risk-based screening was conducted by comparing the concentration data from the New
Jersey monitoring sites to the ATSDR MRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 day to 14 days); intermediate (exposures of 15 days to 364 days); and chronic
(exposures of 1 year or greater). The preprocessed daily measurements of the pollutants of
interest were compared to the acute MRLs; the quarterly averages were compared to the
intermediate MRLs; and the annual averages were compared to the chronic MRLs.
As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2012.
18.5.2 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 3.5.3.4 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 18-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
18-64
-------
Table 18-6. Risk Approximations for the New Jersey Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
Chester, New Jersey - CHNJ
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
Formaldehyde
Hexachloro- 1 ,3 -butadiene
1 , 1 ,2,2-Tetrachloroethane
0.0000022
0.0000078
0.00003
0.000006
0.000026
0.000013
0.000022
0.000058
0.009
0.03
0.002
0.1
2.4
0.0098
0.09
62/62
61/61
51/61
61/61
58/61
62/62
14/61
9/61
1.51
±0.18
0.64
±0.07
0.04
±0.01
0.67
±0.03
0.07
±0.01
2.46
±0.35
0.02
±0.01
0.01
±0.01
3.31
4.97
1.27
4.03
1.87
31.92
0.40
0.46
0.17
0.02
0.02
0.01
<0.01
0.25
0.01
Elizabeth, New Jersey - ELNJ
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
/>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
Propionaldehyde
0.0000022
0.0000078
0.00003
0.000006
0.000011
0.000026
0.0000025
0.000013
0.009
0.03
0.002
0.1
0.8
2.4
1
0.0098
0.008
61/61
61/61
61/61
61/61
51/61
55/61
60/61
61/61
61/61
2.66
±0.34
1.04
±0.14
0.14
±0.02
0.67
±0.03
0.07
±0.02
0.08
±0.01
0.41
±0.05
3.89
±0.47
0.52
±0.08
5.85
8.12
4.18
4.03
0.81
1.97
1.02
50.57
0.30
0.03
0.07
0.01
0.01
O.01
0.01
0.40
0.06
— = A Cancer URE or Noncancer RfC is not available.
18-65
-------
Table 18-6. Risk Approximations for the New Jersey Monitoring Sites (Continued)
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
North Brunswick, New Jersey - NBNJ
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
£>-Dichlorobenzene
1 ,2-Dichloroethane
Formaldehyde
Hexachloro- 1 ,3 -butadiene
1 , 1 ,2,2-Tetrachloroethane
0.0000022
0.0000078
0.00003
0.000006
0.000011
0.000026
0.000013
0.000022
0.000058
0.009
0.03
0.002
0.1
0.8
2.4
0.0098
0.09
60/60
60/60
60/60
60/60
41/60
58/60
60/60
14/60
10/60
1.41
±0.17
0.86
±0.12
0.09
±0.01
0.67
±0.03
0.05
±0.01
0.08
±0.01
1.82
±0.26
0.02
±0.01
0.01
±0.01
3.10
6.74
2.76
4.03
0.58
2.04
23.68
0.43
0.57
0.16
0.03
0.05
0.01
<0.01
0.01
0.19
<0.01
— = A Cancer URE or Noncancer RfC is not available.
Observations from Table 18-6 include the following:
• For CHNJ, the pollutants with the highest annual averages are formaldehyde,
acetaldehyde, and carbon tetrachloride. Formaldehyde has the highest cancer risk
approximation for this site (31.92 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. None of
the pollutants of interest for CFDSTJ have noncancer hazard approximations greater
than 1.0, indicating that no adverse health effects are expected from these individual
pollutants. Formaldehyde is the pollutant with the highest noncancer hazard
approximation (0.25).
• For ELNJ, the pollutants with the highest annual averages 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
greater than the cancer risk approximation for acetaldehyde. ELNJ's cancer risk
approximation for formaldehyde (50.57 in-a-million) is the highest cancer risk
approximation among the pollutants of interest for the New Jersey sites and the fourth
highest among all NMP sites. None of the pollutants of interest for ELNJ have
noncancer hazard approximations greater than 1.0, indicating that no adverse health
effects are expected from these individual pollutants. Formaldehyde is the pollutant
with the highest noncancer hazard approximation (0.40).
18-66
-------
• For NBNJ, the pollutants with the highest annual averages are formaldehyde,
acetaldehyde, and benzene. Formaldehyde has the highest cancer risk approximation
for NBNJ, followed by benzene and carbon tetrachloride. None of the pollutants of
interest for NBNJ have noncancer hazard approximations greater than 1.0, indicating
that no adverse health effects are expected from these individual pollutants.
Formaldehyde is the pollutant with the highest noncancer hazard approximation for
NBNJ (0.19), although the noncancer hazard approximation for acetaldehyde is
similar (0.16).
18.5.3 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 18-7 presents the 10 pollutants with the highest emissions from the 2011 NEI that have
cancer toxicity factors. Table 18-7 also presents the 10 pollutants with the highest toxicity -
weighted emissions, based on the weighting schema described in Section 3.5.3.5. Lastly,
Table 18-7 provides the 10 pollutants with the highest cancer risk approximations (in-a-million)
for each site, as presented in Table 18-6. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 18-7. Table 18-8 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 3.5.3.5. Similar to the cancer risk and
noncancer hazard approximations provided in Section 18.5.2, this analysis may help policy-
makers prioritize their air monitoring activities.
18-67
-------
Table 18-7. 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)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Chester, New Jersey (Morris County) - CHNJ
Tetrachloroethylene
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
Dichloromethane
POM, Group 2b
Trichloroethylene
285.46
155.47
92.49
89.56
55.40
24.57
10.24
5.27
1.95
1.54
Benzene
Formaldehyde
1,3 -Butadiene
Naphthalene
POM, Group 3
Ethylbenzene
Nickel, PM
POM, Group 2b
Hexavalent Chromium, PM
Arsenic, PM
1.21E-03
1.20E-03
7.37E-04
3.48E-04
2.84E-04
2.24E-04
2.03E-04
1.72E-04
1.52E-04
1.30E-04
Formaldehyde
Benzene
Carbon Tetrachloride
Acetaldehyde
1 ,2-Dichloroethane
1,3 -Butadiene
1 , 1 ,2,2-Tetrachloroethane
Hexachloro- 1 , 3 -butadiene
31.92
4.97
4.03
3.31
1.87
1.27
0.46
0.40
Elizabeth, New Jersey (Union County) - ELNJ
Tetrachloroethylene
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
Dichloromethane
Trichloroethylene
POM, Group 2b
1,432.00
134.93
101.13
76.77
58.90
20.46
11.54
2.95
1.77
1.76
Formaldehyde
Benzene
1,3 -Butadiene
Naphthalene
Tetrachloroethylene
POM, Group 3
Hexavalent Chromium, PM
Ethylbenzene
POM, Group 2b
Acetaldehyde
1.31E-03
1.05E-03
6.14E-04
3.92E-04
3.72E-04
2.71E-04
2.65E-04
1.92E-04
1.55E-04
1.30E-04
Formaldehyde
Benzene
Acetaldehyde
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
Ethylbenzene
£>-Dichlorobenzene
50.57
8.12
5.85
4.18
4.03
1.97
1.02
0.81
oo
ON
oo
-------
Table 18-7. 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)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
North Brunswick, New Jersey (Middlesex County) - NBNJ
Tetrachloroethylene
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
Trichloroethylene
Dichloromethane
POM, Group 2b
1,352.40
205.10
142.95
115.00
79.40
30.98
16.17
3.19
2.77
2.64
Formaldehyde
Benzene
1,3 -Butadiene
Naphthalene
Hydrazine
POM, Group 3
Tetrachloroethylene
Ethylbenzene
POM, Group 2b
Hexavalent Chromium, PM
1.86E-03
1.60E-03
9.29E-04
5.50E-04
4.38E-04
4.08E-04
3.52E-04
2.87E-04
2.32E-04
2.12E-04
Formaldehyde
Benzene
Carbon Tetrachloride
Acetaldehyde
1,3 -Butadiene
1 ,2-Dichloroethane
£>-Dichlorobenzene
1 , 1 ,2,2-Tetrachloroethane
Hexachloro- 1 ,3 -butadiene
23.68
6.74
4.03
3.10
2.76
2.04
0.58
0.57
0.43
oo
-------
Table 18-8. 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)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer Hazard
Approximation
(HQ)
Chester, New Jersey (Morris County) - CHNJ
Toluene
Xylenes
Hexane
Tetrachloroethylene
Benzene
Formaldehyde
Ethylbenzene
Ethylene glycol
Acetaldehyde
Methyl isobutyl ketone
628.36
355.29
319.95
285.46
155.47
92.49
89.56
81.35
55.40
44.30
Acrolein
1,3 -Butadiene
Formaldehyde
Tetrachloroethylene
Acetaldehyde
Benzene
Nickel, PM
Xylenes
Naphthalene
Lead, PM
210,793.55
12,287.45
9,437.51
7,136.42
6,155.06
5,182.23
4,696.47
3,552.87
3,412.38
2,413.62
Formaldehyde
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Hexachloro- 1 ,3 -butadiene
1 ,2-Dichloroethane
0.25
0.17
0.02
0.02
0.01
0.01
0.01
Elizabeth, New Jersey (Union County) - ELNJ
Tetrachloroethylene
Toluene
Hexane
Xylenes
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
Ethylene glycol
Methyl isobutyl ketone
1,432.00
541.43
355.77
289.97
134.93
101.13
76.77
58.90
45.15
44.91
Acrolein
Cyanide Compounds, PM
Tetrachloroethylene
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Manganese, PM
Benzene
Chlorine
Naphthalene
279,707.82
37,500.01
35,799.91
10,319.26
10,228.03
6,544.54
4,514.27
4,497.65
4,370.00
3,848.04
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Propionaldehyde
Benzene
Carbon Tetrachloride
Ethylbenzene
£>-Dichlorobenzene
1 ,2-Dichloroethane
0.40
0.30
0.07
0.06
0.03
0.01
0.01
0.01
0.01
oo
o
-------
Table 18-8. 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)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer Hazard
Approximation
(HQ)
North Brunswick, New Jersey (Middlesex County) - NBNJ
Tetrachloroethylene
Toluene
Hexane
Xylenes
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
Methyl isobutyl ketone
Ethylene glycol
1,352.40
786.36
506.75
448.83
205.10
142.95
115.00
79.40
58.71
35.26
Acrolein
Tetrachloroethylene
1,3 -Butadiene
Formaldehyde
Acetaldehyde
Benzene
Manganese, PM
Naphthalene
Lead, PM
Titanium tetrachloride
367,039.29
33,810.00
15,490.94
14,586.38
8,822.64
6,836.50
5,831.26
5,389.62
5,101.39
4,535.00
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Benzene
Carbon Tetrachloride
Hexachloro- 1 , 3 -butadiene
£>-Dichlorobenzene
1 ,2-Dichloroethane
0.19
0.16
0.05
0.03
0.01
0.01
0.01
0.01
oo
-------
Observations from Table 18-7 include the following:
• Tetrachloroethylene is the highest emitted pollutant with a cancer URE in all three
New Jersey counties, followed by benzene, formaldehyde, and ethylbenzene. The
tetrachloroethylene emissions in Union and Middlesex Counties are an order of
magnitude greater than the emissions in Morris County.
• Benzene, formaldehyde, 1,3-butadiene, and naphthalene are the pollutants with the
highest toxi city-weighted emissions (of the pollutants with cancer UREs) for all three
New Jersey counties, although not necessarily in that order.
• Six of the 10 highest emitted pollutants in Morris County also have the highest
toxi city-weighted emissions. Eight of the highest emitted pollutants in Union County
also have the highest toxicity-weighted emissions. Seven of the highest emitted
pollutants in Middlesex County also have the highest toxicity-weighted emissions.
• Formaldehyde, benzene, and 1,3-butadiene are among the pollutants with the highest
cancer risk approximations for CFINJ and also appear on both emissions-based lists.
Acetaldehyde is also among the pollutants with the highest cancer risk
approximations for CFINJ; this pollutant also appears among the highest emitted
pollutants in Morris County but does not appear among those with the highest
toxicity-weighted emissions.
• Five of the 10 pollutants of interest listed in Table 18-7 for ELNJ also appear on both
emissions-based lists: formaldehyde, benzene, acetaldehyde, ethylbenzene, and
1,3-butadiene.
• Formaldehyde, benzene, and 1,3-butadiene are among those with the highest cancer
risk approximations for NBNJ, and each of these appear on both emissions-based lists
for Middlesex County. Acetaldehyde has one of the higher cancer risk
approximations for NBNJ; although acetaldehyde is one of the highest emitted
pollutants in Middlesex County, it is not among those with the highest toxicity-
weighted emissions.
• Carbon tetrachloride is among the pollutants with the highest cancer risk
approximations for each of the three New Jersey counties, ranking third, fifth, and
third for CFINJ, ELNJ, and NBNJ, respectively. This pollutant's total emissions rank
greater than 20th for each county and its toxicity-weighted emissions rank greater
than 30th for each county; thus, carbon tetrachloride does not appear on either
emissions-based list in Table 18-7.
Observations from Table 18-8 include the following:
• Tetrachloroethylene is also the highest emitted pollutant with a noncancer RfC in
Union and Middlesex Counties, but ranks fourth in Morris County. Toluene, xylenes,
and hexane are the highest emitted pollutants with noncancer RfCs in Morris County.
These pollutants are among the highest emitted (behind tetrachloroethylene) in Union
and Middlesex Counties.
18-72
-------
• Acrolein is the pollutant with the highest toxi city-weighted emissions (of the
pollutants with noncancer RfCs) for all three counties but is not among the highest
emitted pollutants for any of the New Jersey counties (acrolein ranks between 16th
and 17th for these counties). Although acrolein was sampled for at all three sites, this
pollutant was excluded from the pollutant of interest designation, and thus subsequent
risk-based screening evaluations, due to questions about the consistency and
reliability of the measurements, as discussed in Section 3.2.
• Formaldehyde, benzene, acetaldehyde, and tetrachloroethylene appear on both
emissions-based lists for Morris, Union, and Middlesex Counties. Xylenes also
appear on both lists for Morris County.
• Formaldehyde, acetaldehyde, and benzene are among the pollutants with the highest
noncancer hazard approximations for CHNJ, ELNJ, and NBNJ (although all were less
than an HQ of 1.0). These pollutants also appear on both emissions-based lists for all
three New Jersey counties. 1,3-Butadiene ranks among the pollutants with the highest
noncancer hazard approximations and is among the pollutants with the highest
toxicity-weighted emissions but is not among the pollutants with the highest
emissions (of the pollutants with noncancer RfCs).
18.6 Summary of the 2012 Monitoring Data for the New Jersey Monitoring Sites
Results from several of the data treatments described in this section include the
following:
»«» Thirteen pollutants failed at least one screen for CHNJ; 14 failed screens for ELNJ;
12 failed screens for NBNJ.
»«» Formaldehyde and acetaldehyde had the highest annual average concentrations for
each of the New Jersey sites.
»«» NBNJ has the highest annual average concentration ofhexachloro-l,3-butadiene
among allNMP sites sampling VOCs, with the annual average for CHNJ ranking
third and ELNJ ranking ninth.
»«» ELNJ is the longest running NMP site still participating in the program.
Concentrations ofpropionaldehyde measured at ELNJ have a fairly steady increasing
trend. In addition, the detection rate of 1,2-dichloroethane at all three New Jersey
sites has been increasing steadily over the last few years of sampling.
18-73
-------
19.0 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.
19.1 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.
Two New York monitoring sites are located in New York City (BXNY and MONY) and
one is located in Rochester (ROCH). Figures 19-1 and 19-2 are composite satellite images
retrieved from ArcGIS Explorer showing the New York City monitoring sites and their
immediate surroundings. Figure 19-3 identifies nearby point source emissions locations by
source category, as reported in the 2011 NEI for point sources. Note that only sources within
10 miles of the sites are included in the facility counts provided in Figure 19-3. 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 radius are still visible on the map, but have been grayed out in order to show emissions
sources just outside the boundary. Figures 19-4 and 19-5 are the composite satellite image and
point emissions sources map for ROCH. Table 19-1 provides supplemental geographical
information such as land use, location setting, and locational coordinates.
19-1
-------
Figure 19-1. New York City, New York (BXNY) Monitoring Site
to
-------
Figure 19-2. New York City, New York (MONY) Monitoring Site
-------
Figure 19-3. NEI Point Sources Located Within 10 Miles of BXNY and MONY
Legend
BXNY NATTS site
74"0'0"W 73"55'0"W 73°50rO"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
MONY NATTS site
10 mile radius
County boundary
Source Category Group (No. of Facilities)
"1" Airport/Airline/Airport Support Operations (26)
B Bulk Terminals/Bulk Plants (7)
C Chemical Manufacturing (5)
i Compressor Station (1)
f Electricity Generation via Combustion (17)
E Electroplating, Plating, Polishing, Anodizing, and Coloring (1)
F Food Processing/Agriculture (5)
A Foundries, Non-ferrous (1)
> Hotels/Motels/Lodging (1)
O Institution (school, hospital, prison, etc.) (27)
A Metal Coating, Engraving, and Allied Services to Manufacturers (1)
© Metals Processing/Fabrication (4)
x Mine/Quarry/Mineral Processing (1)
? Miscellaneous Commercial/Industrial (34)
• Oil and/or Gas Production (3)
"Q Paint and Coating Manufacturing (6)
a Pharmaceutical Manufacturing (2)
R Plastic, Resin, or Rubber Products Plant (4)
P Printing/Publishing/Paper Product Manufacturing (16)
«» Truck/Bus/Transportation Operations (2)
' Wastewater Treatment (5)
19-4
-------
Figure 19-4. Rochester, New York (ROCH) Monitoring Site
-------
Figure 19-5. NEI Point Sources Located Within 10 Miles of ROCH
Legend
77°40'Q"W 77°35'0"W 77C30'0"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.
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 (2)
c Chemical Manufacturing (4) ? Miscellaneous Commercial/Industrial (3)
e Electrical Equipment Manufacturing (2) => Pharmaceutical Manufacturing (1)
1< Glass Plant (1) R Plastic, Resin, or Rubber Products Plant (1)
-#- Industrial Machinery or Equipment Plant (1) P Printing/Publishing/Paper Product Manufacturing (3)
19-6
-------
Table 19-1. Geographical Information for the New York Monitoring Sites
Site
Code
BXNY
MONY
ROCH
AQS Code
36-005-0110
36-005-0080
36-055-1007
Location
New York
New York
Rochester
County
Bronx
Bronx
Monroe
Micro- or
Metropolitan
Statistical Area
New York-Newark-
Jersey City, NY-
NJ-PAMSA
New York-Newark-
Jersey City, NY-
NJ-PA MSA
Rochester, NY
MSA
Latitude
and
Longitude
40.81618,
-73.902
40.83606,
-73.92009
43.14618,
-77.54817
Land Use
Residential
Residential
Residential
Location
Setting
Urban/City
Center
Urban/City
Center
Urban/City
Center
Additional Ambient Monitoring Information1
Haze, SO2, NO, NO2, NOX, VOCs, Carbonyl
compounds, O3, Meteorological Parameters, PM
coarse, Black Carbon, PM10, PM10 Metals, PM2 5,
PM2 5 Speciation, IMPROVE Speciation.
Carbonyl Compounds, VOCs, Meteorological
Parameters, Black carbon, PM10, PM10 Speciation,
PM25.
CO, SO2, NO, NOy, VOCs, Carbonyl compounds,
O3, Meteorological parameters, Black Carbon, PM10,
PM10 Speciation, PM25, PM25 Speciation, IMPROVE
Speciation.
Data for additional pollutants are reported to AQS for these sites (EPA, 2013b); however, these data are not generated by ERG and are therefore not included in this report
BOLD ITALICS = EPA-designated NATTS Site
VO
-------
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 19-1. 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 19-1. BXNY is less than one-half mile from the East River.
The MONY site is located at the Morrisania Neighborhood Family Care center. This site
is considered the Bronx (#2) NATTS site and is a relocation of the BXNY site. MONY is located
less than three-quarters of a mile south of 1-95, one-half mile east of 1-87 and the Harlem River,
which separates the island of Manhattan from the Bronx. Part of the Harlem River can be seen in
the upper left-hand corner of Figure 19-2. The Hudson River is just a few blocks farther west.
The area surrounding MONY is primarily residential, although commercial areas are located
along Jerome Avenue and East 167th Street.
The BXNY site is 1.65 miles southeast of the MONY site. Figure 19-3 shows the
numerous point sources that are located within 10 miles of the sites. The majority of the
emissions sources are located to the south and west of the sites. The source categories with the
greatest number of emissions sources surrounding these sites 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 involved in oil and gas production.
Two point sources are located within 1 mile of MONY: one is a hospital/medical school and the
other falls into the miscellaneous commercial/industrial source category.
In June 2010, the monitoring instruments at BXNY were moved to MONY due to roofing
construction. At the end of June 2012, the instrumentation was returned to the BXNY site and
sampling resumed at this location in July 2012. Thus, this report includes the final 6 months of
sampling at MONY and the initial 6 months of sampling after the relocation back to BXNY.
19-8
-------
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 19-4, a railroad 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 19-5 shows, the relatively few point sources within 10 miles of ROCH
are located primarily on the west side of the 10-mile radius. 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 electrical equipment manufacturers nearby, to name a few.
The closest source to ROCH is an electrical equipment manufacturer
Table 19-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the New York monitoring sites. Table 19-2 includes both county-level
population and vehicle registration information. Table 19-2 also contains traffic volume
information for each site as well as the location for which the traffic volume was obtained.
Additionally, Table 19-2 presents the county-level daily VMT for Bronx and Monroe Counties.
Table 19-2. Population, Motor Vehicle, and Traffic Information for the New York
Monitoring Sites
Site
BXNY
MONY
ROCH
Estimated
County
Population1
1,408,473
747,813
County-level
Vehicle
Registration2
251,398
556,055
Annual
Average
Daily Traffic3
99,201
91,213
88,348
Intersection
Used for
Traffic Data
1-278 between 1-87 and 1-895
1-87 between Bronx Expressway &
Macombs Bridge
1-490 at 1-590
County-
level Daily
VMT4
8,178,210
15,980,952
Bounty-level population estimates reflect 2012 data (Census Bureau, 2013c)
2County-level vehicle registration reflects 2012 data (NYS DMV, 2012)
3AADT reflects 2011 data (NYS DOT, 2011)
4 County-level VMT reflects 2011 data (NYS DOT, 2013)
BOLD ITALICS = EPA-designaled NATTS Site
Observations from Table 19-2 include the following:
• Bronx County has the ninth highest county-level population among counties with
NMP sites. The population of Rochester County is less than the Bronx County
population and ranks 17th among NMP sites.
19-9
-------
• County-level vehicle ownership for Bronx County ranks 29th among counties with
NMP sites, which is in the middle of the range among NMP sites. The county-level
vehicle registration for Rochester County is more than twice the vehicle registration
for Bronx County and ranks 18th compared to other NMP sites.
• Although the population for Bronx County is twice the population for Rochester
County, the vehicle registration for Bronx County is roughly half the vehicle
registration for Rochester County The difference in county-level population and
vehicle registration ranking for Bronx County may be explained by mass
transportation systems.
• Of the New York monitoring sites, traffic is highest near BXNY, which ranks 16th
among NMP sites. The traffic volumes near MONY and ROCH are not that different
from each other and, compared to other NMP sites, the traffic volumes near MONY
and ROCH rank 19th and 20th, respectively. The traffic data for BXNY is for 1-278
between 1-87 and 1-895; the traffic data for MONY are provided for 1-87 between the
Bronx Expressway and Macombs Bridge; and the traffic data for ROCH are provided
for 1-490 at 1-590.
• County-level daily VMT for Monroe County is nearly twice the VMT for Bronx
County. These VMT are in the middle of the range compared to other counties with
NMP sites (where VMT data are available).
19.2 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.
19.2.1 Climate Summary
Weather conditions are somewhat variable in New York City as frontal systems
frequently affect the area. Precipitation is spread fairly evenly throughout the year, with
thunderstorms in the summer and fall and more significant rain or snow events in the winter and
spring. Wintertime snow accumulations generally range from 3 inches to 10 inches. The
proximity to the Atlantic Ocean offers a moderating influence from cold outbreaks as well as the
summertime heat. The urban heat island effect tends to keep the city warmer than outlying areas.
Both influences result in a relatively small diurnal range of temperatures. In addition, air sinking
down from the mountains to the west can help drive temperatures higher during warm spells.
Northwesterly winds prevail during the winter months while southwesterly winds are common
during the warmer months (Wood, 2004; Cornell, 2014).
19-10
-------
Rochester is located in western New York and borders Lake Ontario's south side.
Elevation increases significantly from the shore to the southern-most parts of the city, rising over
800 feet. The lake acts as a moderating influence on the city's temperatures, both in the summer
and the winter, as Lake Ontario does not freeze most winters. It also plays a major factor in the
city's precipitation patterns. Lake effect snow enhances the area's snowfall totals, although
snowfall rates tend to be higher near Lake Ontario and points east rather than farther inland. The
average winter sees 90 inches of snowfall in the city. Spring and summer tend to be sunny due to
the stabilizing effect of the lake, while cloudy conditions are prevalent in the fall and winter.
Prevailing winds are from the southwest year-round (Bair 1992; Wood, 2004).
19.2.2 Meteorological Summary
Hourly meteorological data for 2012 were retrieved from NCDC for the stations closest
to the New York monitoring sites (NCDC, 2012), as described in Section 3.5.2. The closest
weather station to both MONY and BXNY is located at La Guardia Airport, WBAN 14732. The
closest weather station to ROCH is located at Greater Rochester International Airport, WBAN
14768. Additional information about these weather stations, such as the distance between the
sites and the weather stations, is provided in Table 19-3. These data were used to determine how
meteorological conditions on sample days vary from conditions experienced throughout the year.
Table 19-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for the entire year. Also included in Table 19-3 is the 95 percent
confidence interval for each parameter. As expected, Table 19-3 shows that conditions were
cooler in western New York than in New York City. Temperatures near BXNY appear warmer
on sample days than they were the rest of the year. Recall, though, that sampling at this site
began in July 2012, thereby missing the coldest months of the year. The reverse is true for
MONY. Temperatures near MONY appear cooler on sample days than they were the rest of the
year; this site completed sampling June 2012, thereby missing the warmest months of the year.
Average meteorological conditions on sample days near ROCH were representative of average
weather conditions experienced throughout the year.
19-11
-------
Table 19-3. Average Meteorological Conditions near the New York Monitoring Sites
Closest Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
New York City, New York - BXNY
La Guardia
Airport
14732
(40.78, -73.88)
2.8
miles
143°
(SE)
Sample
Days
(34)
2012
67.7
±5.5
64.4
+ 1.7
62.2
±5.1
58.1
+ 1.6
49.5
±4.9
43.3
+ 1.8
55.3
±4.4
50.9
+ 1.5
65.7
±5.0
61.0
+ 1.6
1015.6
±2.7
1015.8
+ 0.8
8.5
±1.1
8.9
+ 0.4
New York City, New York - MONY
La Guardia
Airport
14732
(40.78, -73.88)
4.4
miles
139°
(SE)
Sample
Days
(30)
2012
59.2
±6.1
64.4
+ 1.7
52.6
±5.8
58.1
+ 1.6
36.7
±6.9
43.3
+ 1.8
45.7
±5.5
50.9
+ 1.5
58.2
±6.5
61.0
+ 1.6
1016.8
±3.0
1015.8
+ 0.8
9.1
± 1.2
8.9
+ 0.4
Rochester, New York - ROCH
Greater
Rochester Intl.
Airport
14768
(43.12, -77.68)
6.4
miles
240°
(WSW)
Sample
Days
(61)
2012
61.5
±4.7
60.3
+ 1.9
52.5
±4.3
51.9
+ 1.7
41.5
±4.0
41.3
+ 1.6
47.1
±3.8
46.6
+ 1.5
69.0
±2.7
70.1
+ 1.1
1015.5
±2.0
1015.6
+ 0.7
7.5
±0.8
7.2
+ 0.3
VO
to
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
-------
19.2.3 Back Trajectory Analysis
Figure 19-6 is the composite back trajectory map for days on which samples were
collected at the BXNY monitoring site. Included in Figure 19-5 are four back trajectories per
sample day. Figure 19-7 is the corresponding cluster analysis. Similarly, Figures 19-8 through
19-11 are the composite back trajectory maps and corresponding cluster analyses for days on
which samples were collected at MONY and ROCH. An in-depth description of these maps and
how they were generated is presented in Section 3.5.2.1. For the composite maps, each line
represents the 24-hour trajectory along which a parcel of air traveled toward the monitoring site
on a given sample day and time, based on an initial height of 50 meters AGL. For the cluster
analyses, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric circle around the sites in Figures 19-6 through 19-11 represents
100 miles.
Observations from Figures 19-6 and 19-7 for BXNY include the following:
• The back trajectory maps for BXNY include sample days between July and
December 2012 only to match the sampling period for this site.
• Back trajectories originated from a variety of directions at BXNY, with the longest
back trajectories originating to the northwest of BXNY.
• The 24-hour air shed domain for BXNY is similar in size to other NMP sites.
Although the farthest away a back trajectory originated was over Lake Michigan, or
less than 650 miles away, the average back trajectory length was 208 miles and nearly
90 percent of trajectories originated within 350 miles of the site. Recall, however that
this map includes only 6 months of sample days and that the map may look different
with a full year's worth of data.
• The cluster analysis shows that one-quarter back trajectories originated to the west
and northwest of BXNY, but are split into two cluster trajectories based on length.
The short cluster trajectory also originating to the northwest of BXNY and accounting
for 29 percent of back trajectories represents relatively short back trajectories
(generally less than 150 miles) originating from a variety of directions. Eight percent
of back trajectories originated to the north of BXNY over upstate New York and
Vermont. Twelve percent originated to the east of the site over the offshore waters of
the New England states. Another 15 percent of back trajectories originated to the
south of BXNY, primarily over the offshore waters of New Jersey and the Delmarva
Peninsula. The final 12 percent of back trajectories originated to the southwest of
BXNY, over Virginia, Maryland, and Pennsylvania. This cluster also includes three
of the four back trajectories that spiral around from the Appalachian Mountains,
along the NC/VA border, and offshore towards the monitoring site. These back
trajectories represent October 30, 2012, when Hurricane Sandy came ashore.
19-13
-------
Figure 19-6. Composite Back Trajectory Map for BXNY
Figure 19-7. Back Trajectory Cluster Map for BXNY
19-14
-------
Figure 19-8. Composite Back Trajectory Map for MONY
Figure 19-9. Back Trajectory Cluster Map for MONY
19-15
-------
Figure 19-10. Composite Back Trajectory Map for ROCH
Figure 19-11. Back Trajectory Cluster Map for ROCH
19-16
-------
Observations from Figures 19-8 and 19-9 for MONY include the following:
• The back trajectory maps for MONY include sample days between only January and
June 2012 to match the sampling period for this site.
• Relatively few back trajectories originated to the north and northeast of the MONY
site. The longest back trajectories originated from the west and northwest of MONY.
• The 24-hour air shed domain for MONY is larger in size to BXNY and many NMP
sites, with an average back trajectory length of 269 miles. The farthest away a back
trajectory originated was over central Indiana or greater than 600 miles away,
although 85 percent of trajectories originated within 400 miles of the site. Recall,
however, that this map includes only 6 months of sample days and that the map may
look different with a full year's worth of data.
• The cluster analysis shows that nearly one-quarter of back trajectories originated to
the northwest of MONY, over New York and southeast Ontario, Canada. Another
13 percent of back trajectories originated to the west of MONY, over Lake Huron and
southward to Ohio. The relatively short cluster trajectory (28 percent) represents back
trajectories originating to the south, southwest, and west of the site but generally less
than 200 miles away. Another 17 percent of back trajectories originated over the Mid-
Atlantic states and their offshore waters. The cluster trajectory originating toward
Cape Cod, Massachusetts represents all of the back trajectories originating to the
northeast, east, and southeast of MONY.
Observations from Figures 19-10 and 19-11 for ROCH include the following:
• Back trajectories originated from a variety of directions at ROCH, although relatively
few originated from the northeast and east of ROCH.
• The farthest away a back trajectory originated from ROCH was over New Brunswick,
Canada, or greater than 600 miles away. This back trajectory is also associated with
Hurricane Sandy's path. However, the average back trajectory length was 257 miles
and 93 percent of back trajectories originated within 450 miles of the site.
• The cluster analysis shows that nearly 70 percent of back trajectories originated from
a direction with a westerly component. These back trajectories are represented by
four cluster trajectories in Figure 19-11: 1) back trajectories originating over
southeast Ontario, Canada (23 percent); 2) back trajectories originating over
Michigan and Lake Michigan (16 percent); 3) longer trajectories originating towards
Illinois, Indiana, and western Ohio (8 percent); and 4) snorter back trajectories
originating from over western Pennsylvania and eastern Ohio (22 percent).
• Six percent of back trajectories originated to the northeast of ROCH; this cluster
includes the two long back trajectories originating farther than 500 miles away and
those back trajectories wedged between them. The very short cluster trajectory
originating just east of ROCH and then curving back towards the site represents those
back trajectories originating within 100 miles of the site and from a variety of
directions. This cluster trajectory also represents a few longer back trajectories
19-17
-------
originating along the NY-PA-NJ borders. The final 12 percent of back trajectories
originated to the south-southeast to south-southwest of the site and includes the two
curvy back trajectories originating over south-central Virginia associated with
Hurricane Sandy.
19.2.4 Wind Rose Comparison
Hourly surface wind data from the weather stations at La Guardia Airport (for BXNY and
MONY) and Greater Rochester International Airport (for ROCH) were uploaded into a wind
rose software program to produce customized wind roses, as described in Section 3.5.2.2. 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 19-12 presents a map showing the distance between the weather station and
BXNY, which may be useful for identifying topographical influences that may affect the
meteorological patterns experienced at this location. Figure 19-12 also presents three different
wind roses for the BXNY monitoring site. First, a historical wind rose representing 2002 to 2011
wind data is presented, which shows the predominant surface wind speed and direction over an
extended period of time. Second, a wind rose representing wind observations for all of 2012 is
presented. Next, a wind rose representing wind data for days on which samples were collected in
2012 is presented. These can be used to identify the predominant wind speed and direction for
2012 and to determine if wind observations on sample days were representative of conditions
experienced over the entire year and historically. Figures 19-13 and 19-14 present the distance
maps and wind roses for MONY and ROCH.
19-18
-------
Figure 19-12. Wind Roses for the La Guardia Airport Weather Station near BXNY
Location of BXNY and Weather Station
2002-2011 Historical Wind Rose
2012 Wind Rose
Sample Day Wind Rose
Calms: 4&1W
19-19
-------
Figure 19-13. Wind Roses for the La Guardia Airport Weather Station near MONY
Location of MONY and Weather Station
2002-2011 Historical Wind Rose
2012 Wind Rose
Sample Day Wind Rose
Calms: 4.72%
19-20
-------
Figure 19-14. Wind Roses for the Greater Rochester International Airport Weather Station
near ROCH
Location of ROCH and Weather Station
2002-2011 Historical Wind Rose
V «..~ / ,-.,. ' s /
! I ' - // X
Calms: 9.19%
2012 Wind Rose
Sample Day Wind Rose
Calms: 10 11%
19-21
-------
Observations from Figures 19-12 and 19-3 for BXNY and MONY include the following:
• The weather station at La Guardia Airport is located 2.8 miles southeast of BXNY
and 4.4 miles southeast of MONY. The East River and Rikers Island separate the sites
and the weather station.
• Because the La Guardia Airport weather station is the closest weather station to both
sites, the historical and 2012 wind roses for BXNY are the same as those for MONY.
• The historical wind rose shows that winds from a variety of directions are observed
near BXNY and MONY, although rarely from the southeast quadrant. Winds from
the west to northwest to north account for nearly 40 percent of the wind observations.
Winds from the northeast and east-northeast account for another 17 percent of
observations while winds from the south account for nearly 12 percent. Calm winds
(<2 knots) were observed for less than 5 percent of the hourly measurements near
BXNY and MONY.
• The full-year wind rose for 2012 shares many similarities with the historical wind
rose, such as the prominence of winds from the northwest and the lack of winds from
the southeast quadrant. There are some differences, though. For example, winds from
the northeast account for a higher percentage than winds from the east-northeast,
whereas the percentages are more similar historically.
• For BXNY, the sample day wind patterns resemble the wind patterns on the other
wind roses, particularly the full-year wind rose, although there are fewer strong winds
associated with winds from the northwest quadrant and more strong winds associated
with southerly winds.
• For MONY, westerly to northwesterly winds account for nearly 40 percent of the
wind observations, with southerly winds accounting for another 12 percent of
observations. While this is a common attribute of the historical and full-year wind
roses, the sample day wind rose lacks winds from the north-northwest to north to
northeast.
• The differences between the sample day wind rose for MONY and the wind rose for
BXNY likely results from seasonal differences in the wind observations experienced
near the sites. Recall that sampling at MONY was discontinued June, the
instrumentation moved, and sampling restarted at BXNY in July.
Observations from Figure 19-14 for ROCH include the following:
• The Greater Rochester International Airport weather station is located 6.4 miles west-
southwest of ROCH, with much of the southern half of the city of Rochester between
them.
• The historical wind rose shows that winds from the south-southwest to west were
frequently observed, accounting for nearly 50 percent of the wind observations. Calm
winds were observed for less than 10 percent of the hourly measurements near
19-22
-------
ROCH, while the strongest winds were most frequently observed with west-
southwesterly and westerly winds.
• The wind patterns shown on the 2012 wind rose are similar to the historical wind
patterns for ROCH, although the percentage of calm winds was slightly higher
(nearly 12 percent).
• The sample day wind patterns are similar to those shown on the full-year wind rose,
although the percentages differ somewhat.
19.3 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 19-4. 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 19-4. It is important to note which pollutants were sampled for at each site when reviewing
the results of this analysis. All three New York sites sampled for hexavalent chromium and
PAHs.
19-23
-------
Table 19-4. Risk-Based Screening Results for the New York Monitoring Sites
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
PS 52, New York City, New York - BXNY
Naphthalene
Fluorene
Acenaphthene
Acenaphthylene
Fluoranthene
0.029
0.011
0.011
0.011
0.011
Total
22
8
6
1
1
38
22
22
22
17
22
105
100.00
36.36
27.27
5.88
4.55
36.19
57.89
21.05
15.79
2.63
2.63
57.89
78.95
94.74
97.37
100.00
Morrisania, New York City, New York - MONY
Naphthalene
Acenaphthene
Fluorene
Benzo(a)pyrene
Fluoranthene
0.029
0.011
0.011
0.00057
0.011
Total
30
6
6
o
J
1
46
30
30
30
30
30
150
100.00
20.00
20.00
10.00
3.33
30.67
65.22
13.04
13.04
6.52
2.17
65.22
78.26
91.30
97.83
100.00
Rochester, New York - ROCH
Naphthalene
Acenaphthene
Fluorene
Fluoranthene
0.029
0.011
0.011
0.011
Total
44
22
17
8
91
58
58
58
58
232
75.86
37.93
29.31
13.79
39.22
48.35
24.18
18.68
8.79
48.35
72.53
91.21
100.00
Observations from Table 19-4 include the following:
• Five pollutants failed screens for BXNY; 36 percent of concentrations for these five
pollutants were greater than their associated risk screening value (or failed screens).
All five of these pollutants were identified as pollutants of interest for BXNY.
Although the first four pollutants together account for more than 95 percent of the
total failed screens for BXNY, fluoranthene failed the same number of screens as
acenaphthylene; thus, fluoranthene was also added as a pollutant of interest for
BXNY, per the procedure described in Section 3.2.
• Five pollutants also failed screens for MONY; 31 percent of concentrations for these
five pollutants were greater than their associated risk screening value (or failed
screens). Four of these five pollutants contributed to 95 percent of failed screens for
MONY and therefore were identified as pollutants of interest.
• 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).
All four of these pollutants contributed to 95 percent of failed screens; therefore, all
four were identified as pollutants of interest for this site.
• Although hexavalent chromium was sampled for at each of these sites, this pollutant
did not fail any screens.
19-24
-------
• For all three sites, naphthalene, acenaphthene, and fluorene were identified as
pollutants of interest. Naphthalene failed the majority of screens for each site,
accounting for between 48 percent (ROCH) and 65 percent (MONY) of failed
screens.
• The number of samples collected at ROCH is about twice as many as those collected
at MONY or BXNY. Recall that sampling occurred year-round at ROCH, while
6 months of sampling occurred at BXNY and MONY due to sampler relocation.
19.4 Concentrations
This section presents various concentration averages used to characterize pollution 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. Additional site-specific statistical summaries for BXNY,
MONY, and ROCH are provided in Appendices M and O.
19.4.1 2012 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 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 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
19-25
-------
New York monitoring sites are presented in Table 19-5, 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 19-5. Quarterly and Annual Average Concentrations of the Pollutants of
Interest for the New York Monitoring Sites
Pollutant
#of
Measured
Detections
vs. # 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)
PS 52, New York City, New York - BXNY
Acenaphthene
Acenaphthylene
Fluoranthene
Fluorene
Naphthalene
22/22
17/22
22/22
22/22
22/22
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
11.03
±2.50
0.30
±0.17
7.25
±2.08
12.43
±3.21
124.50
± 16.50
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Morrisania, New York City, New York - MONY
Acenaphthene
Benzo(a)pyrene
Fluorene
Naphthalene
30/30
30/30
30/30
30/30
5.51
±2.42
0.35
±0.10
6.31
±1.96
133.47
±30.63
11.22
±4.31
0.17
±0.12
10.85
±3.78
124.30
± 24.46
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Rochester, New York - ROCH
Acenaphthene
Fluoranthene
Fluorene
Naphthalene
58/58
58/58
58/58
58/58
3.43
±1.61
2.74
±1.17
3.43
±1.34
44.41
±9.37
16.47
±7.75
8.39
±4.07
13.80
±6.35
65.41
±21.35
22.86
±7.02
9.41
±2.65
18.01
±5.43
74.80
±20.11
5.27
±3.79
1.72
±0.46
3.70
±1.87
60.06
± 27.74
12.27
±3.43
5.68
±1.50
9.95
±2.68
61.48
± 10.13
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
Observations from Table 19-5 include the following:
• Sampling at BXNY began in July, which explains why there are no first or second
quarter averages. In addition, damage to the PAH sampler sustained during the
landfall of Hurricane Sandy resulted in no samples collected during the month of
November. Thus, there are no fourth quarter averages either. Annual averages were
not calculated for BXNY because there are fewer than three quarterly averages
available. However, Appendix M and Appendix O provide the pollutant-specific
19-26
-------
average concentrations for all valid samples collected over the entire sample period
for this site.
• Four of the five pollutants of interest for BXNY were detected in all of the valid
samples collected at this site. The third quarter average concentration of naphthalene
is an order of magnitude greater than the quarterly average concentrations for the
other pollutants of interest.
• Sampling at MONY ended in June, which explains why there are no third or fourth
quarter averages. Because there are fewer than three quarterly averages available for
MONY, annual averages were not calculated. However, the pollutant-specific average
concentrations for all valid samples collected over the entire sample period for this
site are provided in Appendix M and Appendix O.
• The available quarterly average concentrations of naphthalene for MONY are an
order of magnitude greater than the quarterly average concentrations for the other
pollutants of interest.
• The second quarter averages for acenaphthene and fluorene for MONY are greater
than the corresponding first quarter averages of these pollutants. The reverse is true
for benzo(a)pyrene. However, the differences between the first and second quarter
averages are not statistically significant and additional quarterly averages would be
needed to determine if there is a seasonal trend in the measurements.
• Naphthalene has the highest quarterly averages (and annual average) among the
pollutants of interest for ROCH. The quarterly averages of naphthalene for ROCH are
roughly half other quarterly averages calculated for BXNY and MONY (where
available). In fact, the annual average concentration of naphthalene for ROCH is
among the lower annual averages, ranking 15th out of 20 NMP sites (where annual
averages could be calculated).
• Quarterly averages of acenaphthene, fluoranthene, and fluorene were considerably
higher during the second and third quarters of the year. This supports the seasonal
trends discussed in Section 4.4.2.
Tables 4-9 through 4-12 present the sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for ROCH from
those tables include the following:
• ROCH has the third highest annual average concentration of acenaphthene and the
fourth highest annual average concentration of fluorene among NMP sites sampling
PAHs.
• ROCH does not appear in Table 4-11 for naphthalene. As discussed in the previous
section, the annual average concentration of naphthalene for ROCH ranks 15th
compared to other NMP sites sampling PAHs.
19-27
-------
19.4.2 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. Because ROCH is the only site for which annual
averages could be calculated, box plots were created for the four pollutants of interest for ROCH.
Figures 19-15 through 19-18 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 3.5.3.
Figure 19-15. Program vs. Site-Specific Average Acenaphthene Concentration
ROCH
| Pro
L
gram Max Concentration = 1
82 ng/m3
40 50 60
Concentration {ng/m3)
Program:
Site:
1st Quartile
D
Site Average
0
2nd Quartile 3rd Quartile
• D
Site Concentration Range
4th Quartile Average
D 1
Figure 19-16. Program vs. Site-Specific Average Fluoranthene Concentration
20 25
Concentration {ng/m3)
Program:
Site:
1st Quartile
Site Average
o
2nd Quartile 3rd Quartile
Site Concentration Range
4th Quartile Average
19-28
-------
Figure 19-17. Program vs. Site-Specific Average Fluorene Concentration
ROCH
-o-
10
20
30
40 50 60
Concentration (ng/m3)
70
80
90
100
Program:
Site:
IstQuartile
Site Average
0
2ndQuartile SrdQuartile
Site Concentration Range
^^^^—
4thQuartile Average
Figure 19-18. Program vs. Site-Specific Average Naphthalene Concentration
ROCH
3 100 200 300 400 500 600 700 800 9C
Concentration {ng/m3)
Program:
Site:
IstQuartile
D
Site Average
o
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
Observations from Figures 19-15 through 19-18 include the following:
• Figure 19-15 presents the box plot for acenaphthene. Note that the program-level
maximum concentration (182 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 to
100 ng/m3. The box plot shows that the annual average concentration for ROCH
is nearly two and half times greater the program-level average concentration. The
maximum concentration of acenaphthene for ROCH is considerably less than the
maximum concentration measured across the program, although the maximum
concentration measured at ROCH is among the higher measurements.
• Figure 19-16 presents the box plot for fluoranthene. Similar to acenaphthene, the
annual average concentration for ROCH is nearly two and half times greater the
program-level average concentration. Although the maximum concentration of
fluoranthene measured at ROCH is less than the maximum concentration
measured across the program, it is the fourth highest concentration measured
among NMP sites sampling PAHs. This site has the second highest annual
average concentration of fluoranthene among NMP sites sampling PAHs (behind
NBIL).
19-29
-------
• Figure 19-17 presents the box plot for fluorene. The annual average concentration
for ROCH is just less than twice the program-level average concentration. The
maximum concentration of fluorene measured at ROCH is considerably less than
the maximum concentration measured across the program, although the maximum
concentration measured at ROCH is among the higher measurements.
• Figure 19-18 presents the box plot for naphthalene. In contrast to the box plots for
the other pollutants of interest for ROCH, Figure 19-18 shows that the annual
average naphthalene concentration is less than the program-level average
concentration and is similar to the program-level median. The maximum
naphthalene concentration measured at ROCH is considerably less than the
program-level maximum concentration.
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 3.5.3.2. A
trends analysis was not performed for BXNY or MONY due to the sampler relocation. Although
sampling for PAHs at ROCH began in July 2008, a trends analysis was not performed for
ROCH. This is because a collection error was discovered at the site, resulting in the invalidation
of nearly one and one-half years' worth of samples. As a result, there is not 5 consecutive years
of data available for the ROCH monitoring site.
19.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk at
each New York monitoring site. Refer to Sections 3.3 and 3.5.3 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.
19.5.1 Risk-Based Screening Assessment Using MRLs
A risk-based screening was conducted by comparing the concentration data from the New
York monitoring sites to the ATSDR MRLs, where available. As described in Section 3.3, MRLs
are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 day to 14 days); intermediate (exposures of 15 days to 364 days); and chronic
(exposures of 1 year or greater). The preprocessed daily measurements of the pollutants of
interest were compared to the acute MRLs; the quarterly averages were compared to the
intermediate MRLs; and the annual averages were compared to the chronic MRLs.
19-30
-------
As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2012.
19.5.2 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 3.5.3.4 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 19-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Observations for the New York sites from Table 19-6 include the following:
• ROCH is the only site for which annual averages could be calculated. Naphthalene
has the highest annual average concentration among the pollutants of interest for
ROCH.
• Naphthalene also has the highest cancer risk approximation for ROCH
(2.09 in-a-million). Acenaphthene also has a cancer risk approximation greater than
1 in-a-million (1.08 in-a-million).
• Only naphthalene has a noncancer RfC. The noncancer hazard approximation for
naphthalene is 0.02, considerably less than 1.0, indicating that no adverse health
effects are expected from this individual pollutant.
19-31
-------
Table 19-6. Risk Approximations for the New York Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
PS 52, New York City, New York - BXNY
Acenaphthene
Acenaphthylene
Fluoranthene
Fluorene
Naphthalene
0.000088
0.000088
0.000088
0.000088
0.000034
0.003
22/22
17/22
22/22
22/22
22/22
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Morrisania, New York City, New York - MONY
Acenaphthene
Benzo(a)pyrene
Fluorene
Naphthalene
0.000088
0.00176
0.000088
0.000034
_
0.003
30/30
30/30
30/30
30/30
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Rochester, New York - ROCH
Acenaphthene
Fluoranthene
Fluorene
Naphthalene
0.000088
0.000088
0.000088
0.000034
_
0.003
58/58
58/58
58/58
58/58
12.27
±3.43
5.68
±1.50
9.95
±2.68
61.48
±10.13
1.08
0.50
0.88
2.09
_
0.02
— = A Cancer URE or Noncancer RfC is not available.
NA = Not available due to the criteria for calculating an annual average.
19.5.3 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 19-7 presents the 10 pollutants with the highest emissions from the 2011 NEI that have
cancer toxicity factors. Table 19-7 also presents the 10 pollutants with the highest toxicity -
weighted emissions, based on the weighting schema described in Section 3.5.3.5. Lastly,
Table 19-7 provides the 10 pollutants with the highest cancer risk approximations (in-a-million)
for each site, as presented in Table 19-6. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 19-7. Table 19-8 presents similar
information, but is limited to those pollutants with noncancer toxicity factors.
19-32
-------
Table 19-7. 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)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
PS 52, New York City, New York (Bronx County) - BXNY
Benzene
Ethylbenzene
Formaldehyde
Tetrachloroethylene
Acetaldehyde
1,3 -Butadiene
Naphthalene
POM, Group 2d
POM, Group 2b
POM, Group la
154.89
112.57
104.34
81.66
61.40
19.36
11.42
1.92
1.85
1.60
Formaldehyde
Benzene
1,3 -Butadiene
Naphthalene
Ethylbenzene
Arsenic, PM
POM, Group 2d
POM, Group 2b
Hexavalent Chromium, PM
POM, Group la
1.36E-03
1.21E-03
5.81E-04
3.88E-04
2.81E-04
2.64E-04
1.69E-04
1.63E-04
1.57E-04
1.41E-04
New York City, New York (Bronx County) - MONY
Benzene
Ethylbenzene
Formaldehyde
Tetrachloroethylene
Acetaldehyde
1,3 -Butadiene
Naphthalene
POM, Group 2d
POM, Group 2b
POM, Group la
154.89
112.57
104.34
81.66
61.40
19.36
11.42
1.92
1.85
1.60
Formaldehyde
Benzene
1,3 -Butadiene
Naphthalene
Ethylbenzene
Arsenic, PM
POM, Group 2d
POM, Group 2b
Hexavalent Chromium, PM
POM, Group la
1.36E-03
1.21E-03
5.81E-04
3.88E-04
2.81E-04
2.64E-04
1.69E-04
1.63E-04
1.57E-04
1.41E-04
VO
-------
Table 19-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the New York Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Rochester, New York (Monroe County) - ROCH
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
Dichloromethane
1,3 -Butadiene
Tetrachloroethylene
Naphthalene
POM, Group la
Trichloroethylene
263.68
198.08
146.98
107.77
46.10
40.49
24.16
23.67
9.25
6.40
Formaldehyde
Benzene
1,3 -Butadiene
POM, Group 3
POM, Group la
Naphthalene
POM, Group 2b
Arsenic, PM
Ethylbenzene
POM, Group 2d
2.58E-03
2.06E-03
1.21E-03
1.19E-03
8.14E-04
8.05E-04
5.27E-04
4.74E-04
3.67E-04
3.64E-04
Naphthalene
Acenaphthene
Fluorene
Fluoranthene
2.09
1.08
0.88
0.50
VO
-------
Table 19-8. 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)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations Based
on Annual Average Concentrations
(Site-Specific)
Noncancer
Hazard
Approximation
Pollutant (HQ)
PS 52, New York City, New York (Bronx County) - BXNY
Toluene
Ethylene glycol
Methanol
Hexane
Xylenes
Benzene
Ethylbenzene
Formaldehyde
Tetrachloroethylene
Methyl isobutyl ketone
2,284.17
1,704.89
793.11
507.63
368.69
154.89
112.57
104.34
81.66
63.81
Acrolein
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Benzene
Ethylene glycol
Cadmium, PM
Arsenic, PM
Naphthalene
Xylenes
251,938.53
10,647.01
9,680.02
6,822.71
5,162.93
4,262.23
4,115.92
4,095.60
3,805.84
3,686.91
Morrisania, New York City, New York (Bronx County) - MONY
Toluene
Ethylene glycol
Methanol
Hexane
Xylenes
Benzene
Ethylbenzene
Formaldehyde
Tetrachloroethylene
Methyl isobutyl ketone
2,284.17
1,704.89
793.11
507.63
368.69
154.89
112.57
104.34
81.66
63.81
Acrolein
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Benzene
Ethylene glycol
Cadmium, PM
Arsenic, PM
Naphthalene
Xylenes
251,938.53
10,647.01
9,680.02
6,822.71
5,162.93
4,262.23
4,115.92
4,095.60
3,805.84
3,686.91
VO
-------
Table 19-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the New York Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations Based
on Annual Average Concentrations
(Site-Specific)
Noncancer
Hazard
Approximation
Pollutant (HQ)
Rochester, New York (Monroe County) - ROCH
Toluene
Ethylene glycol
Xylenes
Methanol
Hexane
Benzene
Hydrochloric acid
Formaldehyde
Ethylbenzene
Acetaldehyde
1,721.15
917.86
529.64
510.18
504.94
263.68
257.86
198.08
146.98
107.77
Acrolein
1,3 -Butadiene
Formaldehyde
Hydrochloric acid
Acetaldehyde
Cadmium, PM
Benzene
Naphthalene
Arsenic, PM
Nickel, PM
528,479.46
20,242.88
20,212.66
12,893.11
11,974.58
9,105.49
8,789.41
7,889.94
7,350.08
6,416.61
Naphthalene 0.02
VO
-------
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 3.5.3.5. Similar to the cancer risk and
noncancer hazard approximations provided in Section 19.5.2, this analysis may help policy-
makers prioritize their air monitoring activities.
Observations from Table 19-7 include the following:
• Benzene, ethylbenzene, and formaldehyde are the highest emitted pollutants with
cancer UREs in Bronx and Monroe Counties (although not necessarily in that order).
• 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.
• Eight 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 all three sites and has the highest
concentrations measured at 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 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 and MONY, including acenaphthene, fluoranthene, and
fluorene. POM, Group 2d also appears on both emissions-based lists for Bronx
County and includes anthracene, phenanthrene, and pyrene. None of these pollutants
failed screens for BXNY or MONY. POM, Group la also appears on both emissions-
based lists for Bronx County but does not include any PAHs sampled for under
Method TO-13A.
• POM Group la appears on both emissions-based lists for Monroe County while
POM, Groups 2b and 2d are among the pollutants with the highest toxicity-weighted
emissions for Monroe County. Three of the four pollutants of interest for ROCH are
part of POM, Group 2b. POM, Group 3 is among the pollutants with the highest
toxicity-weighted emissions for this county.
19-37
-------
Observations from Table 19-8 include the following:
• Toluene and ethylene glycol are the highest emitted pollutants with noncancer RfCs
in both Bronx and Monroe Counties, although the emissions are higher for Bronx
County than Monroe County.
• 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.
• Four of the highest emitted pollutants in Bronx County are also among the pollutants
with the highest toxicity-weighted emissions; four of the highest emitted pollutants in
Monroe County are also among the pollutants with the highest toxicity-weighted
emissions.
• Naphthalene is 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. None of the other pollutants of interest for the three New
York sites have noncancer RfCs.
• Several metals appear among the pollutants with the highest toxicity weighted
emissions for Monroe County although none of these appear among the highest
emitted. Metals were not sampled at ROCH under the NMP.
19.6 Summary of the 2012 Monitoring Data for BXNY, MONY, and ROCH
Results from several of the data treatments described in this section include the
following:
»«» The instrumentation at the MONY monitoring site was relocated to the BXNY
monitoring site at the end of June 2012. This relocation returns the instruments to the
original NATTS location that was discontinued due to ongoing roofing construction.
»«» Five pollutants failed screens for BXNY, five pollutants failed screens for MONY, and
four pollutants failed screens for ROCH. Naphthalene, acenaphthene, andfluorene
were identified as pollutants of interest for each New York monitoring site.
»«» Naphthalene had the highest annual average concentration among the pollutants of
interest for ROCH. Concentrations of acenaphthene, fluoranthene, andfluorene were
highest at ROCH during the warmer months of the year.
*»* ROCH had the second, third, and fourth highest annual average concentrations of
fluoranthene, acenapthalene, andfluorene, respectively, among NMP sites sampling
PAHs.
19-38
-------
20.0 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.
20.1 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.
Two Oklahoma sites (TOOK and TMOK) are located in the Tulsa, Oklahoma MSA.
Another site, PROK, is located east of the Tulsa area in Pry or Creek, Oklahoma. There are also
two sites in the Oklahoma City, Oklahoma MSA (ADOK and OCOK).
Figures 20-1 and 20-2 are composite satellite images retrieved from ArcGIS Explorer
showing the Tulsa monitoring sites and their immediate surroundings. Figure 20-3 identifies
nearby point source emissions locations by source category, as reported in the 2011 NEI for
point sources. Note that only sources within 10 miles of the sites are included in the facility
counts provided in Figure 20-3. 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 radii are still visible on the map, but have been
grayed out in order to show emissions sources just outside the boundary. Figures 20-4 through
20-8 are the composite satellite maps and emissions source maps for the Pry or Creek and
Oklahoma City sites. Table 20-1 provides supplemental geographical information such as land
use, location setting, and locational coordinates.
20-1
-------
Figure 20-1. Tulsa, Oklahoma (TOOK) Monitoring Site
to
o
to
-------
Figure 20-2. Tulsa, Oklahoma (TMOK) Monitoring Site
to
o
-------
Figure 20-3. NEI Point Sources Located Within 10 Miles of TMOK and TOOK
Legend
96°5'0"W 96"0'0"W 95'J55'0"W 95"50'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
TMOK UATMP site
TOOK UATMP site O 10 mile radius
Source Category Group (No. of Facilities)
lj< Aerospace/Aircraft Manufacturing (4)
1T Airport/Airline/Airport Support Operations (14)
^ Automobile/Truck Manufacturing (1)
Z3 Brick, Structural Clay, or Clay Ceramics Plant (1)
B Bulk Terminals/Bulk Plants (1)
C Chemical Manufacturing (2)
i Compressor Station (2)
^ Electricity Generation via Combustion (2)
$ Glass Plant (2)
•^ Industrial Machinery or Equipment Plant (3)
O Institution (school, hospital, prison, etc.) (1)
• Landfill (2)
lip) Metal Can, Box, and Other Metal Container Manufacturing (2)
A Metal Coating. Engraving, and Allied Services to Manufacturers (2)
(•} Metals Processing/Fabrication (2)
X Mine/Quarry/Mineral Processing (1)
? Miscellaneous Commercial/1ndustnal (2)
[M] Municipal Waste Cornbustor (1)
• Oil and/or Gas Production (1)
County boundary
j] Pami and Coating Manufacturing (2)
•{$r Petroleum Products Manufacturing (1)
A Petroleum Refinery (2)
R Plastic. Resin, or Rubber Products Plant (4)
7 Portland Cement Manufacturing (1)
^ Printing, Coatng & Dyeing of Fabrics (1)
X Rail Yard/Rail Line Operations (1)
Q Railroad Engines/Parts Manufacturing (1)
V Steel Mill (1)
20-4
-------
Figure 20-4. Pryor Creek, Oklahoma (PROK) Monitoring Site
to
o
-------
Figure 20-5. NEI Point Sources Located Within 10 Miles of PROK
Legend
95'"20'0"W 95"15'0"W 95"10'Q"W 95"5'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
PROK UATMP site O 10 mile radius | | County boundary
Source Category Group (No. of Facilities)
T Airport/Airline/Airport Support Operations (5)
c Chemical Manufacturing (3)
* Electricity Generation via Combustion (2)
Fertilizer Plant (1)
F Food Processing/Agriculture (1)
i Foundries, Iron and Steel (1)
7 Portland Cement Manufacturing (1)
20-6
-------
Figure 20-6. Oklahoma City, Oklahoma (ADOK) Monitoring Site
to
o
-------
Figure 20-7. Oklahoma City, Oklahoma (OCOK) Monitoring Site
to
o
oo
• *5#i' -
-------
Figure 20-8. NEI Point Sources Located Within 10 Miles of ADOK and OCOK
97'50'0"W 97°45'0'W 97'40'frW 97r'35'0"W S7"30'0"W 97'25'0"W 97'20'0"W 97 15'0"W 97 10'CTW
97'50'0-W 97'45'0"W 97"40'0"W
Legend
97°35'0"W WMHTW 97'25'CTW 97'120'0"W 97a15'0"W 97°10'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
ADOK UATMP site >^ OCOK UATMP site
Source Category Group (No. of Facilities)
t Airport/Airline/Airport Support Operations (20)
~ Brick, Structural Clay, or Clay Ceramics Plant (1)
B Bulk Terminals/Bulk Plants (1)
C Chemical Manufacturing (1)
i Compressor Station (6)
F Food Processing/Agriculture (2)
o Institution (school, hospital, prison, etc.) (1)
• Landfill (3)
® Metals Processing/Fabrication (1)
G 10 mile radius | County boundary
Military Base/National Security (1)
Miscellaneous Commercial/Industrial (2)
Oil and/or Gas Production (5)
Paint and Coating Manufacturing (1)
Pesticide Manufacturing Plant (1)
Plastic, Resin, or Rubber Products Plant (1)
Printing/Publishing/Paper Product Manufacturing (1)
Railroad Engines/Parts Manufacturing (1)
20-9
-------
Table 20-1. Geographical Information for the Oklahoma Monitoring Sites
Site
Code
TOOK
TMOK
PROK
ADOK
OCOK
AQS Code
40-143-0235
40-143-1127
40-097-0187
40-109-0042
40-109-1037
Location
Tulsa
Tulsa
Pryor
Creek
Oklahoma
City
Oklahoma
City
County
Tulsa
Tulsa
Mayes
Oklahoma
Oklahoma
Micro- or
Metropolitan
Statistical Area
Tulsa, OK MSA
Tulsa, OK MSA
Not in an MSA
Oklahoma City,
OK MSA
Oklahoma City,
OK MSA
Latitude and
Longitude
36.126945,
-95.998941
36.204902,
-95.976537
36.292941,
-95.303409
35.3803163,
-97.4057199
35.614131,
-97.475083
Land Use
Industrial
Residential
Industrial
Commercial
Residential
Location
Setting
Urban/City
Center
Urban/City
Center
Suburban
Urban/City
Center
Suburban
Additional Ambient Monitoring Information1
SO2, H2S, and Meteorological parameters.
CO, SO2, NOy, NO, NO2, NOX, O3, Meteorological
parameters, PM10, PM Coarse, PM2 5, and PM25
Speciation, IMPROVE Speciation.
None.
None.
CO, SO2, NO, NO2, NOX, O3, Meteorological
parameters, PM coarse, PM10, PM2 5, and PM2 5
Speciation, IMPROVE Speciation.
Data for additional pollutants are reported to AQS for these sites (EPA, 2013b); however, these data are not generated by ERG and are therefore not included in this report
to
o
-------
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. The monitoring site is positioned
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. As shown in Figure 20-1, an oil refinery is located just south of W 25th Street S.
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, the edge of which can be seen on left-hand side of
Figure 20-1.
TMOK is located in north Tulsa on the property of Fire Station Number 24. As shown in
Figure 20-2, 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.
Figure 20-3 shows that the Tulsa sites are located approximately 5 miles apart, with
TMOK to the north and TOOK to the 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 close to TOOK include two petroleum refineries; a rail yard; a
municipal waste combustor; a compressor station; a metal coating, engraving, and allied services
to manufacturers facility; and a facility generating electricity via combustion.
PROK is located on the eastern edge of the town of Pry or Creek, on the property of Pry or
Creek High School. Residential areas are located to the northwest, west, and south of the site,
while agricultural areas are located to the east, as shown in Figure 20-4. The monitoring site is
located due north (and downwind) of an industrial park located a few miles to the south.
Figure 20-5 shows that there are relatively few emissions sources surrounding PROK and that
the airport source category has the greatest number of emissions sources near the site. An aircraft
operations facility is located one-quarter mile north of PROK but is located under the site symbol
Figure 20-5. A chemical manufacturer and a Portland cement plant are located south and east
-------
of PROK. The aforementioned industrial park is represented in Figure 20-5 by the six facilities
oriented north-south to the south-southeast of PROK. Two chemical manufacturers, a fertilizer
plant, a food processing/agricultural facility, a foundry, and a facility generating electricity via
combustion are located at this industrial park.
The instrumentation at the Midwest City, Oklahoma monitoring site was relocated from
north of Tinker Air Force Base to a location to the south of Tinker Air Force Base in December
2011. The new monitoring site (ADOK) is located on the property of the Oklahoma City Police
Department firing range, approximately one-half mile south of 1-240. The area is considered
commercial although the immediate area surrounding ADOK is open, with a residential
subdivision located farther west, as shown in Figure 20-6. This site lies just northwest of Stanley
Draper Lake and is surrounded by grasslands, with little activity or traffic. The monitoring site
was relocated to this location to capture any influence from Tinker Air Force Base and to collect
background data (OK DEQ, 2013).
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 20-7.
Figure 20-8 shows that ADOK and OCOK are approximately 13 miles apart and that
most of the point sources located within 10 miles of them are located between the sites in the
center of Oklahoma City (northwest of ADOK and south of OCOK). The source category with
the greatest number of sources surrounding these two sites is the airport source category. The
point sources closest to ADOK is a printing and publishing facility, although the southern-most
edge of Tinker Air Force Base lies just on the other side of 1-240; the source closest to OCOK is
involved in brink, structural clay, or clay ceramics.
Table 20-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Oklahoma monitoring sites. Table 20-2 includes both county-level
population and vehicle registration information. Table 20-2 also contains traffic volume
information for each site as well as the location for which the traffic volume was obtained.
20-12
-------
Additionally, Table 20-2 presents the county-level daily VMT for Tulsa, Mayes, and Oklahoma
Counties.
Table 20-2. Population, Motor Vehicle, and Traffic Information for the Oklahoma
Monitoring Sites
Site
TOOK
TMOK
PROK
ADOK
OCOK
Estimated
County
Population1
613,816
41,168
741,781
County-level
Vehicle
Registration2
618,359
41,391
847,824
Annual
Average Daily
Traffic3
63,000
12,600
15,100
34,100
40,900
Intersection
Used for
Traffic Data
1-244 at Southwest Blvd
Near intersection of E 36th St N. & N.
Peoria Ave
Highway 69, south of Route 20
1-240 between 1-35 and 1-40
Route 77, north of toll road
County-
level Daily
VMT4
20,402,564
1,662,076
27,411,171
Bounty-level population estimates reflect 2012 data (Census Bureau, 2013c)
2County-level vehicle registration reflects 2012 data (OKTC, 2012)
3AADT reflects 20 1 1 data (OK DOT, 20 1 1)
4County-level VMT reflects 2012 data (OK DOT, 2013)
Observations from Table 20-2 include the following:
• The Mayes County population is significantly less than the populations for Tulsa and
Oklahoma Counties. Compared to other NMP monitoring sites, the Tulsa and
Oklahoma County populations are in the middle of the range, while Mayes County's
population is on the low end.
• The Mayes County vehicle registration is also significantly less than vehicle
registrations for Tulsa and Oklahoma Counties. These observations are expected
given the relatively rural nature of the area surrounding PROK compared to the urban
locations of the Tulsa and Oklahoma City sites. Compared to other NMP sites, the
Oklahoma County vehicle ownership is in the top third while the vehicle ownership
for Tulsa County is in the middle third.
• The traffic volume passing the TMOK site is the lowest among the Oklahoma
monitoring sites and is similar to the traffic passing the PROK site, while the traffic
passing by TOOK is the highest of the five sites. The traffic data for all five
Oklahoma sites are in the middle third compared to other NMP sites. The following
list provides the roadways or intersections from which the traffic data were obtained:
TOOK - 1-244, near Southwest Boulevard; TMOK - intersection of East 36th Street
North and North Peoria Avenue; PROK - Highway 69, south of Graham Avenue
(Route 20); ADOK - 1-240, between 1-35 and 1-40; and OCOK - Route 77 north of
the toll road.
• County-level VMT is greatest for Oklahoma County and ranks 11th compared to
other NMP sites. VMT is the least for Mayes County and is among the lower VMTs
compared to other NMP sites.
20-13
-------
20.2 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.
20.2.1 Climate Summary
Tulsa is located in northeast Oklahoma, just southeast of the Osage Indian Reservation,
and along the Arkansas River. Pry or Creek is also in northeast Oklahoma, approximately
30 miles east of Tulsa. Oklahoma City is located in the center of the state. These areas are
characterized by a continental climate, with long, warm summers and relatively mild winters.
Precipitation is generally concentrated in the spring and summer months, with maximum
precipitation occurring in May, June, and September, although precipitation amounts generally
decrease across the state from east to west. Spring and summer precipitation usually results from
showers and thunderstorms, while fall and winter precipitation accompanies frontal systems.
Annual snowfall in these areas is less than 10 inches per year. A southerly wind prevails for
much of the year. Oklahoma is part of "Tornado Alley," where severe thunderstorms capable of
producing strong winds, hail, and tornadoes occur more frequently than other areas around the
country; tornadoes are more prevalent here than any other region in the U.S. (Wood, 2004;
NCDC, 2014; NOAA, 2014a).
20.2.2 Meteorological Summary
Hourly meteorological data for 2012 were retrieved from NCDC for the weather stations
closest to the Oklahoma monitoring sites (NCDC, 2011 and 2012), as described in Section 3.5.2.
The closest weather stations to the Tulsa sites are located at Richard Lloyd Jones Jr. Airport
(near TOOK) and Tulsa International Airport (near TMOK), WBAN 53908 and 13968,
respectively. The closest weather station to the Pry or Creek site is located at Claremore Regional
Airport, WBAN 53940. The two closest weather stations to the Oklahoma City sites are located
at Tinker Air Force Base Airport (near ADOK) and Wiley Post Airport (near OCOK), WBAN
13919 and 03954, respectively. Additional information about these weather stations, such as the
distance between the sites and the weather stations, is provided in Table 20-3. These data were
used to determine how meteorological conditions on sample days vary from conditions
experienced throughout the year.
20-14
-------
Table 20-3. Average Meteorological Conditions near the Oklahoma Monitoring Sites
Closest
Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Tulsa, Oklahoma - TOOK
Richard Lloyd
Jones Jr.
Airport
53908
(36.04, -95.98)
6.1
miles
173°
(S)
Sample
Days
(65)
2012
78.0
±4.1
76.5
+ 1.8
65.8
±4.2
64.5
+ 1.8
48.4
±4.0
47.3
+ 1.7
56.1
±3.5
55.1
+ 1.5
57.9
±3.0
58.3
+ 1.3
1015.8
±1.5
1016.3
±0.6
5.4
±0.8
5.4
+ 0.3
Tulsa, Oklahoma - TMOK
Tulsa
International
Airport
13968
(36.20, -95.89)
4.7
miles
95°
(E)
Sample
Days
(61)
2012
75.9
±4.3
75.7
+ 1.8
64.9
±4.2
65.0
+ 1.8
48.6
±4.1
48.3
+ 1.6
55.9
±3.6
55.7
+ 1.5
59.5
±3.3
58.9
+ 1.4
1014.9
±1.6
1015.1
+ 0.7
7.8
±1.0
8.0
+ 0.4
Pryor Creek, Oklahoma - PROK
Claremore
Regional
Airport
53940
(36.29, -95.47)
9.3
miles
270°
(W)
Sample
Days
(51)
2012
77.0
±4.4
73.5
+ 1.8
65.6
±4.1
61.8
+ 1.7
51.7
±3.8
47.9
+ 1.6
57.7
±3.4
54.1
+ 1.5
65.0
±3.5
64.6
+ 1.3
NA
NA
7.0
±1.2
7.0
+ 0.3
Oklahoma City, Oklahoma - ADOK
Tinker
AFB/Airport
13919
(35.42, -97.39)
2.8
miles
26°
(NNE)
Sample
Days
(69)
December
2011 &
2012
74.0
±4.4
73.2
+ 1.8
63.0
±4.1
62.4
+ 1.7
47.6
±3.9
47.4
+ 1.6
54.5
±3.5
54.2
+ 1.5
62.1
±3.8
62.4
+ 1.6
1016.1
±1.6
1015.8
±0.7
9.2
±0.9
9.6
+ 0.4
to
o
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
NA = Sea level pressure was not recorded at the Claremore Regional Airport.
-------
Table 20-3. Average Meteorological Conditions near the Oklahoma Monitoring Sites (Continued)
Closest
Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Oklahoma City, Oklahoma - OCOK
Wiley Post
Airport
03954
(35.53, -97.65)
10.7
miles
240°
(WSW)
Sample
Days
(64)
2012
75.8
±4.2
74.8
+ 1.8
64.3
±4.3
64.0
+ 1.8
47.6
±4.0
47.2
+ 1.7
55.1
±3.5
54.8
+ 1.5
59.0
±3.5
58.6
+ 1.5
1014.8
±1.6
1015.0
±0.7
9.7
±1.1
10.2
+ 0.5
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
NA = Sea level pressure was not recorded at the Claremore Regional Airport.
to
o
-------
Table 20-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for all of 2012. Also included in Table 20-3 is the 95 percent
confidence interval for each parameter. As shown in Table 20-3, average meteorological
conditions on sample days were representative of average weather conditions experienced
throughout the year for most of the Oklahoma monitoring sites. The differences are greatest for
PROK, where sample days appear warmer than conditions experienced throughout the year, but
the difference is not statistically significant. Sampling was discontinued at PROK at the end of
October 2012, thereby missing two of the cooler months of the year. Note that sampling at
ADOK began in December 2011 and data from the five samples collected that month are
included in this report. Thus, the meteorological averages provided in Table 20-3 include
meteorological observations from December 2011.
20.2.3 Back Trajectory Analysis
Figure 20-9 is the composite back trajectory map for days on which samples were
collected at the TOOK monitoring site. Included in Figure 20-9 are four back trajectories per
sample day. Figure 20-10 is the corresponding cluster analysis. Similarly, Figures 20-11 through
20-18 are the composite back trajectory maps for days on which samples were collected at the
remaining Oklahoma sites and the corresponding cluster analyses. An in-depth description of
these maps and how they were generated is presented in Section 3.5.2.1. For the composite maps,
each line represents the 24-hour trajectory along which a parcel of air traveled toward the
monitoring site on a given sample day and time, based on an initial height of 50 meters AGL. For
the cluster analyses, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric circle around the sites in Figures 20-9 through 20-18 represents
100 miles.
20-17
-------
Figure 20-9. Composite Back Trajectory Map for TOOK
Figure 20-10. Back Trajectory Cluster Map for TOOK
20-18
-------
Figure 20-11. Composite Back Trajectory Map for TMOK
Figure 20-12. Back Trajectory Cluster Map for TMOK
20-19
-------
Figure 20-13. Composite Back Trajectory Map for PROK
Figure 20-14. Back Trajectory Cluster Map for PROK
20-20
-------
Figure 20-15. Composite Back Trajectory Map for ADOK
Figure 20-16. Back Trajectory Cluster Map for ADOK
20-21
-------
Figure 20-17. Composite Back Trajectory Map for OCOK
Figure 20-18. Back Trajectory Cluster Map for OCOK
20-22
-------
Observations from Figures 20-9 through 20-18 include the following:
• The back trajectory maps for the Tulsa sites, the Pry or Creek site, and the Oklahoma
City sites are similar to each other in back trajectory distribution. This is somewhat
expected, given their relatively close proximity to each other and the similarity in
sample days, although not all sites sampled on the exact same days over the period.
• The air shed domain for the OCOK site was among the largest in size compared to
other NMP sites, based on an average back trajectory length of 281 miles. The
farthest away a back trajectory originated was over western Utah, or approximately
825 miles away. The air shed domains for the other Oklahoma sites are in the top
third compared to other NMP sites, based on the average back trajectory length. The
farthest away a back trajectory originated was greater than 800 miles away for all five
sites. The average back trajectory length for the Oklahoma sites ranged from
250 miles (PROK) to 281 miles (TOOK).
• Each of the sites exhibits a tendency for back trajectories to originate from the south-
southeast to south-south west of the sites and from the northwest to northeast of the
sites, with the longest back trajectories generally originating from the northwest.
• For the Tulsa sites, nearly 40 percent of back trajectories originated from the
southeast to southwest, generally over the eastern half of Texas, although these are
split into two cluster trajectories. Roughly one-quarter of back trajectories originated
to the east of the sites, primarily over Arkansas. The short cluster trajectory
originating to the north of the sites includes those back trajectories with a northern
component and that originated primarily over eastern Kansas and along the
Kansas/Missouri border. The remaining back trajectories originated from the west to
northwest to north of the sites, but of varying lengths.
• The cluster analysis for PROK groups together the relatively short back trajectories
originating from the north over eastern Kansas and along Kansas/Missouri border
with the relatively short back trajectories originating over the southern half of
Missouri and Arkansas. These back trajectories together account for greater than
40 percent of back trajectories. Those back trajectories originating to the southeast,
south, and southwest account for another 42 percent of back trajectories. Back
trajectories originating to the north account for 9 percent of back trajectories while
those originating to the west and northwest account for another 6 percent. Recall that
sampling was discontinued at PROK at the end of October 2012; thus, this site has
fewer sample days included in its back trajectory maps.
• The cluster analysis maps for the Oklahoma City sites are similar to each in cluster
distribution patterns, although the percentages vary. Greater than 50 percent of back
trajectories originated to the south of the sites, but are split into three different cluster
trajectories based on back trajectory length and the location in Texas the back
trajectory originated. Approximately 20 percent of back trajectories originated over
the eastern half of Oklahoma and southeast Kansas. This cluster trajectory for OCOK
includes a few of the longer trajectories originating over northeast Kansas, which
explains why this cluster trajectory is longer for OCOK than ADOK. Twenty-five
percent of back trajectories originated from the west, northwest, and north of the sites.
20-23
-------
These back trajectories are represented by two cluster trajectories for ADOK and
three for OCOK.
20.2.4 Wind Rose Comparison
Hourly surface wind data from the weather stations at Richard Lloyd Jones Junior Airport
(for TOOK), Tulsa International Airport (for TMOK), Claremore Regional Airport (for PROK),
Tinker Air Force Base (for ADOK), and Wiley Post Airport (for OCOK) were uploaded into a
wind rose software program to produce customized wind roses, as described in Section 3.5.2.2.
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 20-19 presents a map showing the distance between the weather station and
TOOK, which may be useful for identifying topographical influences that may affect the
meteorological patterns experienced at this location. Figure 20-19 also presents three different
wind roses for the TOOK monitoring site. First, a historical wind rose representing 2002 to 2011
wind data is presented, which shows the predominant surface wind speed and direction over an
extended period of time. Second, a wind rose representing wind observations for all of 2012 is
presented. Next, a wind rose representing wind observations for days on which samples were
collected in 2012 is presented. These can be used to identify the predominant wind speed and
direction for 2012 and to determine if wind observations on sample days were representative of
conditions experienced over the entire year and historically. Figures 20-20 through 20-23 present
the distance maps and wind roses for the remaining Oklahoma sites.
20-24
-------
Figure 20-19. Wind Roses for the Richard Lloyd Jones Jr. Airport Weather Station near
TOOK
Location of TOOK and Weather Station
2002-2011 Historical Wind Rose
Calms: 24.37%
2012 Wind Rose
Sample Day Wind Rose
WIND SPEED
(Knots)
n >==
^| 17-21
^| 11-17
^| 7- 11
n 4.7
HI - -
Calms: 27.52%
WIND SPEED
(Knots)
[Z] -2=
^| 17-21
Hf 11-17
^| 7- 11
n 4-7
^| i- -
Calms: 26.0SW
20-25
-------
Figure 20-20. Wind Roses for the Tulsa International Airport Weather Station near
TMOK
Location of TMOK and Weather Station
2002-2011 Historical Wind Rose
WIND SPEED
(Knots)
LJ = 22
^| 17-21
^| 11-17
^| 7- 11
Calms: 8.45%
2012 Wind Rose
Sample Day Wind Rose
WIND SPEED
(Knots)
n «22
^| 17-21
^| 11-17
^| 7- 11
n «-7
HI - -
Calms: 11 02%
WIND SPEED
(Knots)
CH -2=
^| 17-21
Hf 11-17
^| 7- 11
n 4-7
^| i- -
Calms: 11.61%
20-26
-------
Figure 20-21. Wind Roses for the Claremore Regional Airport Weather Station near
PROK
Location of PROK and Weather Station
2003-2011 Historical Wind Rose
25%
"*v, 20%.
15%
1 0%
WIND SPEED
(Knots)
LJ = 22
^| 17-21
^| 11-17
^| 7- 11
Calms: 16.34%
2012 Wind Rose
Sample Day Wind Rose
25%
"vs^ 20%
15%
WHO SPEED
(Knots)
CH -22
^| 17-21
Hf 11-17
^| 7- 11
n 4-7
^| i- -
Cdms: 15.45%
20-27
-------
Figure 20-22. Wind Roses for the Tinker Air Force Base Airport Weather Station near
ADOK
Location of ADOK and Weather Station
2006-2011 Historical Wind Rose
2012 Wind Rose
Sample Day Wind Rose
35%
"',s 28%
21%
14%
WIND SPEED
(Knots)
[Z] -2=
^| 17-21
Hf 11-17
^| 7- 11
n 4.7
^| i- -
Cdms: 5.11%
20-28
-------
Figure 20-23. Wind Roses for the Wiley Post Airport Weather Station near OCOK
Location of OCOK and Weather Station
2002-2011 Historical Wind Rose
2012 Wind Rose
Sample Day Wind Rose
35%
"S, 28%
21%
14%
WIN D S PE ED
(Knots)
n -22
^| 17-21
^| 11 - 17
^| 7- 11
Calms: 7.01%
20-29
-------
Observations from Figures 20-19 through 20-23 include the following:
• The distance maps show that the distances between the sites and the weather stations
varies from 2.8 miles between Tinker Air Force Base and ADOK to 10.7 miles
between OCOK and the Wiley Post Airport.
• Even though the historical data are from five different weather stations, the wind
patterns shown on the wind roses for the Oklahoma sites are similar to each other.
Each of the historical wind roses shows that southerly winds prevailed near each
Oklahoma monitoring site, accounting for roughly 20 percent to 30 percent of
observations among the historical time periods. The historical wind roses varied in
the percentage of calm winds (<2 knots) observed, ranging from as little as 3 percent
at the Tinker Air Force Base (ADOK) to as high as 24 percent at the Richard Lloyd
Jones Jr. Airport (TOOK). Calms winds, winds from the south-southeast to south-
southwest, and winds from the north-northwest to north-northeast account for the
majority of wind observations at each site while winds from the west or east were
rarely observed near each site.
• For TOOK, the 2012 wind patterns are similar to the historical wind patterns, as are
the sample day wind patterns, although a slightly higher percentage of calm winds
were observed in 2012 and on sample days. These similarities indicate that conditions
on sample days were representative of those experienced over the entire year and
historically.
• For TMOK, the 2012 wind patterns are similar to the historical wind patterns, as are
the sample day wind patterns, although a slightly higher percentage of calm winds
were observed in 2012 and on sample days. These similarities indicate that conditions
on sample days were representative of those experienced over the entire year and
historically.
• For PROK, the historical wind rose includes 9 years of data, starting with 2003. The
2012 wind patterns resemble the historical wind patterns. The sample day wind rose
for PROK is similar to the historical and full-year wind roses, indicating that
conditions on sample days were representative of conditions experienced throughout
the year and historically.
• For ADOK, the historical wind rose includes 6 years of data, starting with 2006. The
2012 wind patterns resemble the historical wind patterns, although there were slightly
more southerly wind observations in 2012. This is also true of calm winds. The
sample day wind patterns resemble the historical and the full-year wind patterns,
although there is an even higher percentage of calm winds on sample days.
• For OCOK, the wind patterns shown on the 2012 wind rose resemble the historical
wind patterns, but with a slightly higher percentage of calm winds. The sample day
wind rose for OCOK is similar to both the historical and full-year wind roses, but also
exhibits a higher percentage of calm winds.
20-30
-------
20.3 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 20-4. 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 20-4. It is important to note which pollutants were sampled for at each site when reviewing
the results of this analysis. The five Oklahoma sites sampled for VOCs, carbonyl compounds,
and metals (TSP).
Table 20-4. Risk-Based Screening Results for the Oklahoma Monitoring Sites
Pollutant
Screening
Value
(Ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Tulsa, Oklahoma - TOOK
Acetaldehyde
Arsenic (TSP)
Formaldehyde
Manganese (TSP)
Benzene
Carbon Tetrachloride
1,3 -Butadiene
Ethylbenzene
1,2-Dichloroethane
£>-Dichlorobenzene
Nickel (TSP)
Hexachloro- 1 ,3 -butadiene
Propionaldehyde
1 , 1 ,2,2-Tetrachloroethane
Xylenes
Cadmium (TSP)
Lead (TSP)
1 ,2-Dibromoethane
Beryllium (TSP)
Cobalt (TSP)
Trichloroethylene
0.45
0.00023
0.077
0.005
0.13
0.17
0.03
0.4
0.038
0.091
0.0021
0.045
0.8
0.017
10
0.00056
0.015
0.0017
0.00042
0.01
0.2
Total
61
61
61
61
60
60
58
51
41
29
24
7
7
5
5
4
4
3
1
1
1
605
61
61
61
61
60
60
60
60
41
55
61
10
61
5
60
61
61
3
61
61
11
1,035
100.00
100.00
100.00
100.00
100.00
100.00
96.67
85.00
100.00
52.73
39.34
70.00
11.48
100.00
8.33
6.56
6.56
100.00
1.64
1.64
9.09
58.45
10.08
10.08
10.08
10.08
9.92
9.92
9.59
8.43
6.78
4.79
3.97
1.16
1.16
0.83
0.83
0.66
0.66
0.50
0.17
0.17
0.17
10.08
20.17
30.25
40.33
50.25
60.17
69.75
78.18
84.96
89.75
93.72
94.88
96.03
96.86
97.69
98.35
99.01
99.50
99.67
99.83
100.00
20-31
-------
Table 20-4. Risk-Based Screening Results for the Oklahoma Monitoring Sites (Continued)
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Tulsa, Oklahoma - TMOK
Benzene
Carbon Tetrachloride
Manganese (TSP)
Acetaldehyde
Arsenic (TSP)
Formaldehyde
1,3 -Butadiene
1,2-Dichloroethane
Ethylbenzene
£>-Dichlorobenzene
Nickel (TSP)
Propionaldehyde
Hexachloro -1,3 -butadiene
1 , 1 ,2,2-Tetrachloroethane
1 ,2-Dibromoethane
Lead (TSP)
Beryllium (TSP)
Cadmium (TSP)
Trichloroethylene
0.13
0.17
0.005
0.45
0.00023
0.077
0.03
0.038
0.4
0.091
0.0021
0.8
0.045
0.017
0.0017
0.015
0.00042
0.00056
0.2
Total
61
61
61
59
59
59
57
42
40
20
15
7
6
4
3
2
1
1
1
559
61
61
61
59
61
59
60
42
61
54
61
59
9
4
3
61
61
61
19
917
100.00
100.00
100.00
100.00
96.72
100.00
95.00
100.00
65.57
37.04
24.59
11.86
66.67
100.00
100.00
3.28
1.64
1.64
5.26
60.96
10.91
10.91
10.91
10.55
10.55
10.55
10.20
7.51
7.16
3.58
2.68
1.25
1.07
0.72
0.54
0.36
0.18
0.18
0.18
10.91
21.82
32.74
43.29
53.85
64.40
74.60
82.11
89.27
92.84
95.53
96.78
97.85
98.57
99.11
99.46
99.64
99.82
100.00
Pryor Creek, Oklahoma - PROK
Acetaldehyde
Benzene
Carbon Tetrachloride
Formaldehyde
Arsenic (TSP)
1 ,2-Dichloroethane
Manganese (TSP)
1,3 -Butadiene
/>-Dichlorobenzene
Hexachloro- 1 ,3 -butadiene
1 , 1 ,2,2-Tetrachloroethane
Cadmium (TSP)
1 ,2-Dibromoethane
Ethylbenzene
Nickel (TSP)
Beryllium (TSP)
Propionaldehyde
Trichloroethylene
0.45
0.13
0.17
0.077
0.00023
0.038
0.005
0.03
0.091
0.045
0.017
0.00056
0.0017
0.4
0.0021
0.00042
0.8
0.2
Total
51
51
51
51
47
46
45
28
21
4
3
2
2
2
2
1
1
1
409
51
51
51
51
49
47
49
39
45
6
3
49
2
51
49
49
51
3
696
100.00
100.00
100.00
100.00
95.92
97.87
91.84
71.79
46.67
66.67
100.00
4.08
100.00
3.92
4.08
2.04
1.96
33.33
58.76
12.47
12.47
12.47
12.47
11.49
11.25
11.00
6.85
5.13
0.98
0.73
0.49
0.49
0.49
0.49
0.24
0.24
0.24
12.47
24.94
37.41
49.88
61.37
72.62
83.62
90.46
95.60
96.58
97.31
97.80
98.29
98.78
99.27
99.51
99.76
100.00
20-32
-------
Table 20-4. Risk-Based Screening Results for the Oklahoma Monitoring Sites (Continued)
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Oklahoma City, Oklahoma - ADOK
Acetaldehyde
Benzene
Formaldehyde
Carbon Tetrachloride
Arsenic (TSP)
Manganese (TSP)
1 ,2-Dichloroethane
1,3 -Butadiene
£>-Dichlorobenzene
Lead (TSP)
1 , 1 ,2,2-Tetrachloroethane
Ethylbenzene
Propionaldehyde
1 ,2-Dibromoethane
Hexachloro- 1 ,3 -butadiene
Cadmium (TSP)
Nickel (TSP)
0.45
0.13
0.077
0.17
0.00023
0.005
0.038
0.03
0.091
0.015
0.017
0.4
0.8
0.0017
0.045
0.00056
0.0021
Total
66
66
66
65
59
58
54
32
24
6
6
3
3
2
2
1
1
514
66
66
66
66
64
64
54
44
57
64
6
66
66
2
5
64
64
884
100.00
100.00
100.00
98.48
92.19
90.63
100.00
72.73
42.11
9.38
100.00
4.55
4.55
100.00
40.00
1.56
1.56
58.14
12.84
12.84
12.84
12.65
11.48
11.28
10.51
6.23
4.67
1.17
1.17
0.58
0.58
0.39
0.39
0.19
0.19
12.84
25.68
38.52
51.17
62.65
73.93
84.44
90.66
95.33
96.50
97.67
98.25
98.83
99.22
99.61
99.81
100.00
Oklahoma City, Oklahoma - OCOK
Benzene
Carbon Tetrachloride
Acetaldehyde
Arsenic (TSP)
Formaldehyde
Manganese (TSP)
1 ,2-Dichloroethane
1,3 -Butadiene
£>-Dichlorobenzene
Ethylbenzene
Propionaldehyde
Hexachloro- 1 ,3 -butadiene
1 , 1 ,2,2-Tetrachloroethane
Trichloroethylene
Nickel (TSP)
1 ,2-Dibromoethane
Cadmium (TSP)
0.13
0.17
0.45
0.00023
0.077
0.005
0.038
0.03
0.091
0.4
0.8
0.045
0.017
0.2
0.0021
0.0017
0.00056
Total
61
61
60
60
60
58
52
47
16
7
7
4
4
4
3
2
1
507
61
61
60
61
60
61
52
56
54
61
60
8
4
7
61
2
61
790
100.00
100.00
100.00
98.36
100.00
95.08
100.00
83.93
29.63
11.48
11.67
50.00
100.00
57.14
4.92
100.00
1.64
64.18
12.03
12.03
11.83
11.83
11.83
11.44
10.26
9.27
3.16
1.38
1.38
0.79
0.79
0.79
0.59
0.39
0.20
12.03
24.06
35.90
47.73
59.57
71.01
81.26
90.53
93.69
95.07
96.45
97.24
98.03
98.82
99.41
99.80
100.00
20-33
-------
Observations from Table 20-4 include the following:
• Twenty-one pollutants failed at least one screen for TOOK; 58 percent of
concentrations for these 21 pollutants were greater than their associated risk screening
value (or failed screens).
• Thirteen pollutants contributed to 95 percent of failed screens for TOOK and
therefore were identified as pollutants of interest for this site. These 13 include three
carbonyl compounds, seven VOCs, and three TSP metals.
• Nineteen pollutants failed at least one screen for TMOK; 61 percent of concentrations
for these 19 pollutants were greater than their associated risk screening value (or
failed screens).
• Eleven pollutants contributed to 95 percent of failed screens for TMOK and therefore
were identified as pollutants of interest for this site. These 11 include two carbonyl
compounds, six VOCs, and three TSP metals.
• Eighteen pollutants failed at least one screen for PROK; 59 percent of concentrations
for these 18 pollutants were greater than their associated risk screening value (or
failed screens).
• Nine pollutants contributed to 95 percent of failed screens for PROK and therefore
were identified as pollutants of interest for this site. These nine include two carbonyl
compounds, five VOCs, and two TSP metals.
• Seventeen pollutants failed at least one screen for ADOK; 58 percent of
concentrations for these 17 pollutants were greater than their associated risk screening
value (or failed screens).
• Nine pollutants contributed to 95 percent of failed screens for ADOK and therefore
were identified as pollutants of interest for this site. These nine include two carbonyl
compounds, five VOCs, and two TSP metals.
• Seventeen pollutants failed at least one screen for OCOK; 64 percent of
concentrations for these 17 pollutants were greater than their associated risk screening
value (or failed screens).
• Eleven pollutants contributed to 95 percent of failed screens for OCOK and therefore
were identified as pollutants of interest for this site. These 11 include three carbonyl
compounds, six VOCs, and two TSP metals.
• The number of pollutants identified as pollutants of interest range from nine to 13
among the Oklahoma sites. These sites have nine pollutants of interest in common:
acetaldehyde, arsenic, benzene, 1,3-butadiene, carbon tetrachloride,
/>-dichlorobenzene, 1,2-dichloroethane, formaldehyde, and manganese.
20-34
-------
• TOOK failed the third highest number of screens among all NMP sites, with the other
Oklahoma sites ranking sixth (TMOK), eighth (ADOK), 11th (OCOK), and 14th
(PROK), as shown in Table 4-8.
20.4 Concentrations
This section presents various concentration averages used to characterize pollution 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. Additional site-specific statistical summaries for the five
Oklahoma sites are provided in Appendices J, L, and N.
20.4.1 2012 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 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 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
Oklahoma monitoring sites are presented in Table 20-5, 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"
20-35
-------
because only zeros substituted for non-detects were factored into the quarterly average
concentration.
Table 20-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Oklahoma Monitoring Sites
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(jig/m3)
2nd
Quarter
Average
(jig/m3)
3rd
Quarter
Average
(jig/m3)
4th
Quarter
Average
(Hg/m3)
Annual
Average
(jig/m3)
Tulsa, Oklahoma - TOOK
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
£>-Dichlorobenzene
1,2-Dichloroethane
Ethylbenzene
Formaldehyde
Hexachloro- 1 ,3 -butadiene
Propionaldehyde
Arsenic (TSP)a
Manganese (TSP)a
Nickel (TSP)a
61/61
60/60
60/60
60/60
55/60
41/60
60/60
61/61
10/60
61/61
61/61
61/61
61/61
1.71
±0.28
2.05
±0.55
0.10
±0.03
0.64
±0.05
0.12
±0.02
0.06
±0.03
0.93
±0.42
1.85
±0.30
0.01
±0.01
0.29
±0.06
0.88
±0.19
33.02
±11.60
2.35
±0.83
3.25
±0.96
2.48
±0.69
0.07
±0.02
0.72
±0.04
0.11
±0.03
0.11
±0.01
1.01
±0.31
3.96
±1.42
0.02
±0.02
0.59
±0.22
1.08
±0.21
37.59
±8.17
2.34
±0.57
3.96
±0.97
2.51
±0.76
0.08
±0.03
0.63
±0.03
0.08
±0.02
0.05
±0.02
1.06
±0.38
5.11
±1.04
0.02
±0.02
0.69
±0.15
0.82
±0.19
32.49
±7.97
1.86
±0.38
2.11
±0.54
1.80
±0.48
0.14
±0.04
0.64
±0.05
0.07
±0.02
0.06
±0.04
0.66
±0.21
2.66
±0.47
0
0.42
±0.13
0.93
±0.23
50.62
±33.47
3.18
±1.73
2.78
±0.42
2.21
±0.31
0.10
±0.02
0.66
±0.02
0.09
±0.01
0.07
±0.01
0.91
±0.17
3.42
±0.54
0.01
±0.01
0.50
±0.08
0.92
±0.10
38.33
±8.81
2.42
±0.49
a Average concentrations provided for the pollutants below the blue line are presented in ng/m for ease of viewing.
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
20-36
-------
Table 20-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Oklahoma Monitoring Sites (Continued)
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Ug/m3)
2nd
Quarter
Average
(Ug/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Ug/m3)
Tulsa, Oklahoma - TMOK
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
£>-Dichlorobenzene
1,2-Dichloroethane
Ethylbenzene
Formaldehyde
Arsenic (TSP)a
Manganese (TSP)a
Nickel (TSP)a
59/59
61/61
60/61
61/61
54/61
42/61
61/61
59/59
61/61
61/61
61/61
1.52
±0.23
1.23
±0.25
0.11
±0.04
0.66
±0.06
0.07
±0.02
0.04
±0.02
0.52
±0.17
2.56
±0.39
0.72
±0.18
21.11
±9.34
1.57
±0.45
2.57
±0.53
1.47
±0.44
0.10
±0.03
0.74
±0.03
0.11
±0.03
0.11
±0.02
0.64
±0.16
4.17
± 1.04
0.99
±0.28
25.96
±5.17
1.85
±0.36
3.41
±0.84
1.06
±0.18
0.10
±0.03
0.65
±0.03
0.07
±0.02
0.05
±0.02
0.52
±0.12
4.99
±1.09
0.61
±0.16
19.90
±5.11
1.27
±0.34
1.77
±0.40
1.27
±0.43
0.18
±0.07
0.68
±0.04
0.08
±0.03
0.07
±0.03
0.55
±0.22
2.71
±0.49
0.78
±0.24
38.34
±33.71
2.02
±0.86
2.33
±0.32
1.25
±0.16
0.12
±0.02
0.68
±0.02
0.08
±0.01
0.06
±0.01
0.56
±0.08
3.63
±0.47
0.77
±0.11
26.22
±8.46
1.67
±0.26
Pryor Creek, Oklahoma - PROK
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
£>-Dichlorobenzene
1 ,2-Dichloroethane
Formaldehyde
Arsenic (TSP)a
Manganese (TSP)a
51/51
51/51
39/51
51/51
45/51
47/51
51/51
49/49
49/49
1.21
±0.19
0.62
±0.10
0.04
±0.02
0.67
±0.08
0.12
±0.03
0.09
±0.01
1.86
±0.31
0.58
±0.19
12.10
±6.96
1.64
±0.31
0.77
±0.36
0.03
±0.01
0.77
±0.07
0.14
±0.05
0.08
±0.02
4.27
±1.41
0.70
±0.12
15.16
±2.83
1.94
±0.41
0.50
±0.06
0.05
±0.03
0.65
±0.03
0.04
±0.01
0.05
±0.01
5.03
±1.10
0.58
±0.27
16.54
±3.67
NA
NA
NA
NA
NA
NA
NA
NA
NA
1.56
±0.19
0.61
±0.11
0.05
±0.01
0.69
±0.03
0.09
±0.02
0.07
±0.01
3.58
±0.65
0.63
±0.10
18.66
±8.09
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.
20-37
-------
Table 20-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Oklahoma Monitoring Sites (Continued)
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Ug/m3)
2nd
Quarter
Average
(Ug/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Ug/m3)
Oklahoma City, Oklahoma - ADOK
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
£>-Dichlorobenzene
1,2-Dichloroethane
Formaldehyde
Arsenic (TSP)a
Manganese (TSP)a
66/66
66/66
44/66
66/66
57/66
54/66
66/66
64/64
64/64
1.16
±0.16
0.65
±0.08
0.04
±0.02
0.62
±0.07
0.04
±0.01
0.07
±0.02
1.64
±0.25
0.48
±0.10
11.40
±7.40
1.98
±0.46
0.72
±0.26
0.02
±0.01
0.73
±0.04
0.05
±0.01
0.09
±0.01
3.81
±1.13
0.56
±0.10
14.28
±5.17
2.84
±0.58
0.57
±0.22
0.06
±0.03
0.67
±0.03
0.28
±0.13
0.05
±0.02
4.82
±0.92
0.49
±0.13
14.26
±3.19
1.39
±0.30
0.58
±0.10
0.05
±0.04
0.68
±0.05
0.16
±0.03
0.05
±0.02
2.06
±0.34
0.46
±0.10
12.80
±3.43
1.81
±0.24
0.63
±0.08
0.04
±0.01
0.67
±0.03
0.13
±0.04
0.06
±0.01
3.00
±0.46
0.49
±0.05
13.08
±2.62
Oklahoma City, Oklahoma - OCOK
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
£>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
Propionaldehyde
Arsenic (TSP)a
Manganese (TSP)a
60/60
61/61
56/61
61/61
54/61
52/61
61/61
60/60
60/60
61/61
61/61
1.47
±0.22
0.71
±0.08
0.05
±0.01
0.61
±0.06
0.04
±0.01
0.08
±0.01
0.20
±0.03
1.81
±0.28
0.29
±0.05
0.51
±0.17
15.45
±9.33
2.55
±0.52
1.01
±0.43
0.12
±0.14
0.73
±0.04
0.05
±0.02
0.09
±0.01
0.48
±0.37
4.16
± 1.15
0.57
±0.11
0.67
±0.14
16.09
±4.28
3.59
±0.70
0.61
±0.09
0.07
±0.03
0.65
±0.03
0.09
±0.02
0.06
±0.01
0.30
±0.04
5.46
±1.07
0.71
±0.13
0.58
±0.15
29.68
±11.40
1.67
±0.30
0.79
±0.16
0.08
±0.05
0.65
±0.05
0.08
±0.02
0.05
±0.02
0.26
±0.07
2.44
±0.43
0.33
±0.07
0.52
±0.10
22.63
±6.71
2.34
±0.32
0.78
±0.12
0.08
±0.04
0.66
±0.03
0.07
±0.01
0.07
±0.01
0.31
±0.09
3.49
±0.54
0.48
±0.06
0.57
±0.07
21.10
±4.26
a Average concentrations provided for the pollutants below the blue line are presented in ng/m for ease of viewing.
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
20-38
-------
Observations for all five Oklahoma sites from Table 20-5 include the following:
• Formaldehyde has the highest annual average concentration for each site, followed by
acetaldehyde. With the exception of the Tulsa sites, these were the only two
pollutants of interest with annual average concentrations greater than 1 |ig/m3 for
each site. For TOOK and TMOK, benzene also has an annual average concentration
greater than 1 |ig/m3. However, the annual average for TOOK (2.21 ± 0.31 |ig/m3) is
greater than the annual average for TMOK (1.25 ±0.16 |ig/m3). This observation was
also made in the 2011 report, but the difference has decreased for 2012.
• Annual average concentrations of formaldehyde range from 3.00 ± 0.46 |ig/m3 for
ADOK to 3.63 ± 0.47 |ig/m3 for TMOK. With the exception of ADOK, the annual
averages of formaldehyde span less than 0.25 |ig/m3. The annual average
concentration of acetaldehyde are more variable and ranged from 1.56 ± 0.19 |ig/m3
for PROK to 2.78 ± 0.42 |ig/m3 for TOOK.
• Concentrations of the carbonyl compounds, formaldehyde in particular, tended to be
highest in the summer months and lowest in the winter months. However, the
confidence intervals associated with some of these averages indicate that the
differences are not statistically significant for all locations. Three of the five
Oklahoma sites measured their maximum formaldehyde concentration on
June 26, 2012. The June 26th sample was invalid for OCOK and a make-up sample
was collected on June 28, 2012. The maximum formaldehyde concentration for
OCOK was measured in the June 28th sample. While the maximum formaldehyde
concentration was measured at TMOK on August 1, 2012 (10.1 |ig/m3), a similar
concentration was also measured on June 26, 2012 (10.0 |ig/m3). The formaldehyde
concentrations measured at the Oklahoma sites on or near June 26, 2012 account for
some of the highest concentrations measured program-wide.
• The annual average concentration of manganese is higher than the annual average
concentrations of the other TSP metals for each site. The annual average manganese
concentrations range from 13.08 ± 2.62 ng/m3 for ADOK to 38.33 ± 8.81 ng/m3 for
TOOK. Manganese concentrations greater than 70 ng/m3 were measured at all five
Oklahoma sites and concentrations greater than 200 ng/m3 were measured at TOOK,
TMOK, and PROK.
Observations for TOOK from Table 20-5 include the following:
• Although the third quarter average formaldehyde concentration is the highest among
the quarterly averages for TOOK, the confidence interval is larger for the second
quarter average. As discussed above, the maximum formaldehyde concentration was
measured on June 26, 2012 (12.8 |ig/m3). This is the highest concentration of
formaldehyde among all NMP sites across the program, although a concentration of
the same magnitude was also measured at SPIL. The maximum acetaldehyde
concentration was also measured at TOOK on June 26, 2012 and is the fourth highest
acetaldehyde concentration measured program-wide.
20-39
-------
• Of the 32 benzene concentrations greater than 3 |ig/m3 measured across the program,
13 were measured at TOOK. The site with the next highest number of benzene
concentrations greater than 3 |ig/m3 is ROIL, with five. Concentrations of benzene
measured at TOOK range from 0.46 |ig/m3to 5.73 |ig/m3.
• Of the 10 ethylbenzene concentrations greater than 2 |ig/m3 measured across the
program, five were measured at TOOK. At least one of these was measured in each
quarter except the fourth quarter, explaining the differences in the quarterly averages
and the associated confidence intervals shown in Table 20-5 for TOOK.
Concentrations of ethylbenzene measured at TOOK range from 0.14 |ig/m3to
2.89 |ig/m3.
• The fourth quarter average concentration of manganese is greater than the other
quarterly averages and has a relatively large confidence interval associated with it.
This is also true for nickel. A review of the data shows that the maximum
concentration of each of the metal pollutants of interest was measured at TOOK on
October 18, 2012. The maximum manganese concentration (273 ng/m3) is the only
manganese concentration measured at TOOK greater than 100 ng/m3. This
concentration is the second highest manganese concentration measured at an NMP
site sampling metals (behind only S4MO, 275 ng/m3). The maximum manganese
concentrations measured at TMOK and PROK were also measured on this sample
day (266 ng/m3 and 204 ng/m3, respectively). These maximum concentrations
correlate to the day a large dust storm affected the Tulsa area.
Observations for TMOK from Table 20-5 include the following:
• The fourth quarter average concentration of 1,3-butadiene is greater than the other
quarterly averages and has a larger confidence interval associated with it. A review of
the data shows that the two highest concentrations of 1,3-butaidene were both
measured during the fourth quarter (0.50 |ig/m3 on November 17, 2012 and
0.34 |ig/m3 on October 30, 2012). Further, eight of the 11 concentrations of
1,3-butadiene greater than 0.20 |ig/m3 were measured at TMOK during the fourth
quarter of 2012.
• The fourth quarter average concentration of manganese is greater than the other
quarterly averages and has a large confidence interval of nearly the same magnitude
associated with it. A review of the data shows that the maximum concentration of this
pollutant was measured at TMOK on October 18, 2012. The maximum manganese
concentration (266 ng/m3) is the only manganese concentration measured at TMOK
greater than 100 ng/m3. This concentration is the third highest manganese
concentration measured at an NMP site sampling metals. This is the same day the
maximum manganese concentrations were measured at TOOK and PROK. These
relatively high concentrations correlate to a dust storm affecting the Tulsa area. The
maximum manganese concentrations measured at TMOK and TOOK were also
measured on the same day in 2011 (October 6, 2011) and was discussed in the 2011
NMP report.
20-40
-------
• Nickel is also highest during the fourth quarter and has a relatively large confidence
interval associated with it. The maximum nickel concentration (7.01 ng/m3) was also
measured at TMOK on October 18th, and is nearly twice the next highest nickel
concentration measured at TMOK (3.78 ng/m3, also measured during the fourth
quarter).
Observations for PROK from Table 20-5 include the following:
• The formaldehyde concentration measured at PROK on June 26, 2012 (12.4 |ig/m3) is
the fourth highest formaldehyde concentration measured program-wide.
• The maximum benzene concentration was measured at PROK on April 3, 2012
(3.12 |ig/m3). The second highest benzene concentration (1.05 |ig/m3) was measured
at PROK on June 26, 2012, the same day the maximum formaldehyde concentration
was measured at PROK (and other Oklahoma sites).
• The maximum manganese concentration (204 ng/m3) was measured at PROK on
October 18, 2012, which is the same day the maximum manganese concentrations
were measured at TOOK and TMOK, and is the fifth highest manganese
concentration measured program-wide. The next highest manganese concentration
measured at PROK is considerably less (57.6 ng/m3) and no other measurements
greater than 30 ng/m3 were measured at PROK.
• Because sampling at PROK was discontinued at the end of October 2012, there are no
fourth quarter averages for this site in Table 20-5.
Observations for ADOK from Table 20-5 include the following:.
• The third and fourth quarter average concentrations of />-dichlorobenzene are
considerably higher than the other quarterly averages, particularly the third quarter. A
review of the data shows that the maximum concentration of this pollutant was
measured on August 13, 2012 (0.81 |ig/m3). The eight highest concentrations of
/>-dichlorobenzene were all measured at ADOK in August or September and ranged
from 0.283 |ig/m3 to 0.807 |ig/m3. Further, all but one of the 22 concentrations greater
than 0.1 |ig/m3 measured at ADOK were measured between August and December.
Observations for OCOK from Table 20-5 include the following:
• The second quarter average concentrations of 1,3-butadiene and ethylbenzene are
higher than the other quarterly averages and have relatively large confidence
intervals, particularly 1,3-butadiene, for which the confidence interval is greater than
the average itself. This is also true for benzene. The maximum concentrations of
1,3-butadiene and ethylbenzene were measured at OCOK on June 14, 2012 (and the
second highest concentration of benzene was also measured on this date). The
ethylbenzene concentration for this date (2.93 |ig/m3) is more than five times higher
than the next highest concentration measured at OCOK and is the third highest
ethylbenzene concentration measured across the program. The 1,3-butadiene
concentration for this date (1.09 |ig/m3) is nearly four times higher than the next
highest concentration measured at OCOK and the fourth highest 1,3-butadiene
20-41
-------
concentration measured across the program. Although the benzene concentration for
this date is the second highest benzene concentration measured at OCOK
(2.82 |ig/m3), it is just less than the maximum concentration measured at this site
(2.97 |ig/m3), which was also measured during the second quarter.
• The quarterly averages of propionaldehyde for OCOK exhibit the same trend as
formaldehyde and acetaldehyde in that the quarterly averages for the warmer months
are greater than the quarterly averages for the cooler months.
Tables 4-9 through 4-12 present the 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 37 times. However,
because they are the only sites sampling TSP metals, all five sites appear for each
metal, accounting for 15 of the appearances.
• TOOK has the highest annual average of concentrations of benzene and ethylbenzene
among all NMP sites sampling these pollutants. The annual average concentrations
for TMOK rank fifth for both pollutants. Similar findings for benzene were observed
in the 2010 and 2011 NMP reports.
• An annual average concentration for at least one Oklahoma site ranked among the
highest annual average concentrations for all of the VOC pollutants of interest
provided in Table 4-9.
• Four of the five Oklahoma sites appear in Table 4-10 for their annual average
concentrations of formaldehyde, ranking between fifth and ninth. Only ADOK does
not appear in this table for formaldehyde (it ranked 14th).
• TOOK has the third highest annual average concentration of acetaldehyde among
NMP sites sampling carbonyl compounds. OCOK and TMOK rank seventh and
eighth, respectively.
• TOOK has the highest annual average concentration of the three TSP metals shown in
Table 4-12, followed by TMOK. TOOK has the highest annual average manganese
concentration among all NMP sites sampling manganese.
20-42
-------
20.4.2 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 shaded in
gray in Table 20-4. Figures 20-24 through 20-36 overlay the sites' minimum, annual average,
and maximum concentrations onto the program-level minimum, first quartile, median, average,
third quartile, and maximum concentrations, as described in Section 3.5.3.1.
Figure 20-24. Program vs. Site-Specific Average Acetaldehyde Concentrations
ADOK
OCOK
lo ' '
1
it:
9 12
Concentration {[og/m3)
Program: IstQuartile
2ndQuartile
SrdQuartile
4thQuartile
Average
Site:
SiteAverage Site Concentration Range
o
20-43
-------
Figure 20-25. Program vs. Site-Specific Average Arsenic (TSP) Concentrations
ADOK
OCOK
TOOK
1 1.5
Concentration (ng/m3)
Site:
Site Average Site Concentration Range
o
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile
Average
20-44
-------
Figure 20-26. Program vs. Site-Specific Average Benzene Concentrations
OCOK
PROK
TMOK
TOOK
Concentration {[og/m3
Program:
Site:
IstQuartile
D
Site Average
o
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
20-45
-------
Figure 20-27. Program vs. Site-Specific Average 1,3-Butadiene Concentrations
Program Max Concentration = 4.10 |Jg/m3
OCOK
Program Max Concentration = 4.10 |Jg/m3
PROK
l^r-
Program Max Concentration = 4.10 |Jg/m3
TMOK
Program Max Concentration = 4.10 |Jg/m3
TOOK
K-
Program Max Concentration = 4.10 |Jg/m3
0.75 1 1.25
Concentration {(og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o
20-46
-------
Figure 20-28. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations
• «h
OCOK
PROK
TMOK
TOOK
2 3
Concentration {[ug/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o
20-47
-------
Figure 20-29. Program vs. Site-Specific Average /7-Dichlorobenzene Concentrations
OCOK
PROK
TMOK
l^^^^^_^^^_l
0.6 0.8
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: SiteAverage Site Concentration Range
o
20-48
-------
Figure 20-30. Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations
Program Max Concentration = 17.01 |Jg/m3
OCOK
Program Max Concentration = 17.01 |Jg/m3
PROK
Program Max Concentration = 17.01 |Jg/m3
TMOK
Program Max Concentration = 17.01 |Jg/m3
r.
Program Max Concentration = 17.01 |Jg/m
0.1 0.2 0.3 0.4 0.5 0.6
Concentration {[og/m3)
0.7 0.8
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: SiteAverage Site Concentration Range
o
20-49
-------
Figure 20-31. Program vs. Site-Specific Average Ethylbenzene Concentrations
TMOK
TOOK
0.5
1.5 2 2.5
Concentration {[og/m3)
3.5
Program:
Site:
IstQuartile
D
Site Average
o
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
20-50
-------
Figure 20-32. Program vs. Site-Specific Average Formaldehyde Concentrations
ADOK
OCOK
Concentration {[og/m3]
Program:
Site:
IstQuartile
Site Average
o
2ndQuartile SrdQuartile
Site Concentration Range
4thQuartile Average
20-51
-------
Figure 20-33. Program vs. Site-Specific Average Hexachloro-l,3-Butadiene Concentration
TOOK
0.1 0.15
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o
Figure 20-34. Program vs. Site-Specific Average Manganese (TSP) Concentrations
PROK
TOOK
50
100
150
Concentration (ng/m3)
200
O
250
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: SiteAverage Site Concentration Range
300
20-52
-------
Figure 20-35. Program vs. Site-Specific Average Nickel (TSP) Concentrations
TOOK
10
12
14
Concentration {
Program: IstQuartile 2nd Quartile SrdQuartile 4th Quartile Average
Site:
Site Average Site Concentration Range
o
Figure 20-36. Program vs. Site-Specific Average Propionaldehyde Concentrations
TOOK
0.5
1 1.5
Concentration {[og/m3)
2.5
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site:
SiteAverage Site Concentration Range
O
Observations from Figures 20-24 through 20-36 include the following:
• Figure 20-24 shows that the range of acetaldehyde concentrations is largest for
TOOK and smallest for PROK. The annual average acetaldehyde concentrations
for TOOK, TMOK, and OCOK are greater than the program-level average for
acetaldehyde. The annual average for TOOK is also greater than the program-
level third quartile. The annual average acetaldehyde concentration for ADOK is
similar to the program-level average while the annual average for PROK is just
less than the program-level average. The minimum acetaldehyde concentrations
measured at TOOK, TMOK, and OCOK are just less than the program-level first
quartile.
20-53
-------
• Because the Oklahoma sites are the only sites sampling TSP metals, Figure 20-25
compares the individual Oklahoma site data against the combined Oklahoma data.
Figure 20-25 shows that the annual average arsenic (TSP) concentration is
greatest for TOOK and least for ADOK. This figure also shows that the range of
measurements of arsenic is largest for PROK, where the maximum arsenic (TSP)
concentration was measured, although a similar concentration was also measured
at TMOK. The minimum arsenic concentration measured among the five sites
sampling TSP metals was measured at TMOK.
• Figure 20-26 presents the box plots for benzene. The maximum benzene
concentration measured across the program was measured at TOOK. The annual
average concentration of benzene for TOOK and TMOK are greater than the
program-level average while the annual average concentrations of benzene for
PROK, ADOK, and OCOK are less than the program-level average. The annual
average benzene concentration for TOOK is more than twice the program-level
average concentration and three to four times greater than the annual average
concentrations of the other Oklahoma sites except TMOK.
• Figure 20-27 presents the box plots for 1,3-butadiene. The program-level
maximum concentration (4.10 |ig/m3) is not shown directly on the box plots as the
scale has been reduced to 2 |ig/m3 to allow for the observation of data points at
the lower end of the concentration range. Among the Oklahoma sites, the
maximum concentration of 1,3-buadiene was measured at OCOK (1.09 |ig/m3);
this measurement is more than twice the next highest concentration (0.50 |ig/m3)
measured at one of these five sites. The annual average 1,3-butadiene
concentrations for four of the five Oklahoma sites OCOK are less than the
program-level average concentration, while the annual average for TMOK is just
greater than the program-level average. At least one non-detect of 1,3-butadiene
was measured at the Oklahoma sites, with the exception of TOOK.
• Figure 20-28 presents the box plots for carbon tetrachloride. Although the range
of carbon tetrachloride measurements varies by site, each of the annual average
concentrations of carbon tetrachloride is similar to the program-level average
concentration. The maximum carbon tetrachloride concentrations measured at the
Oklahoma sites are considerably less than the maximum concentration measured
at the program level.
• Figure 20-29 presents the box plots for/?-dichlorobenzene. Note that the program-
level first quartile is zero and therefore not visible on the box plots in
Figure 20-29. The range of measurements collected at ADOK is considerably
larger than those measured at the other Oklahoma sites. Even so, the annual
average concentrations ofp-dichlorobenzene for all of these sites are greater than
the program-level average concentration (although the difference for OCOK is
minimal). The annual average concentration for ADOK is twice the program-level
average. Several non-detects were measured at each of the Oklahoma sites.
20-54
-------
• Figure 20-30 presents the box plots for 1,2-dichloroethane for all five sites. Note
that the program-level maximum concentration (17.01 |ig/m3) is not shown
directly on the box plots as the scale has been reduced to 1 |ig/m3 in order to allow
for the observation of data points at the lower end of the concentration range.
Figure 20-30 for 1,2-dichloroethane shows that nearly the entire range of
1,2-dichloroethane measurements collected at the Oklahoma sites was less than
the program-level average concentration. This is because the program-level
average is being driven by the higher measurements collected at a few monitoring
sites. The maximum 1,2-dichloroethane concentrations measured at each
Oklahoma site is at least two orders of magnitude less than the maximum
concentration measured across the program. The annual average concentrations
for the Oklahoma sites are less than or similar to the median concentration for the
program.
• Figure 20-31 for ethylbenzene presents the concentration data for only three of the
five sites because these are the only sites for which this pollutant is a pollutant of
interest. The range of concentrations measured at TMOK is roughly half the range
of concentrations measured at OCOK and TOOK. Even though the range of
measurements shown in Figure 20-31 is roughly the same for OCOK and TOOK,
the annual average concentration for TOOK is twice that of OCOK. This is
because the maximum concentration measured at OCOK (2.93 |ig/m3) is so much
higher than the next highest measurement, as discussed in the previous section.
Aside from the maximum concentration, all ethylbenzene measurements collected
at OCOK are less than 0.60 |ig/m3. Conversely, nearly 70 percent of the
measurements from TOOK are greater than 0.60 |ig/m3. The annual average
ethylbenzene concentration for TMOK lies between the annual averages for
TOOK and OCOK. The annual averages for TOOK and TMOK are both greater
than the program-level average concentration and third quartile.
• Figure 20-32 shows that the annual average formaldehyde concentration for each
Oklahoma site is greater than the program-level average concentration and that
four of the five are greater than the program-level third quartile. The annual
average concentrations of formaldehyde did not vary significantly among the
Oklahoma sites. The maximum concentration measured across the program was
measured at TOOK, although a similar concentration was also measured at
PROK. The minimum concentration of formaldehyde measured at TMOK is
similar to the program-level first quartile.
• Figure 20-33 presents the hexachloro-1,3-butadiene concentration data for only
TOOK because TOOK is the only site for which hexachloro-l,3-butadiene is a
pollutant of interest. Note that the first, second, and third quartiles are zero due to
the large number of non-detects for this pollutant. The annual average
concentration of hexachloro-1,3-butadiene for TOOK is just greater than the
program-level average concentration. Of the 60 measurements collected at
TOOK, 50 were non-detects, or roughly 17 percent, which is slightly higher than
the percentage across the program (13 percent).
20-55
-------
• Figure 20-34 compares the manganese data for each individual Oklahoma site
against the combined Oklahoma data. The range of measurements collected at
each site increases in the same order as the sites are presented in Figure 20-34.
The minimum manganese (TSP) concentration was measured at ADOK while the
maximum manganese (TSP) concentration was measured at TOOK (although a
similar measurement was also collected at TMOK). The annual average
manganese concentration was greatest for TOOK and least for ADOK, among the
Oklahoma sites.
• Figure 20-35 presents the nickel concentration data for two of the five Oklahoma
sites because the Tulsa sites are the only ones for which nickel is a pollutant of
interest. The range of nickel measurements collected was greater at TOOK than
TMOK. The maximum nickel (TSP) concentration was measured at TOOK; this
measurement ranked fourth among all nickel (TSP and PMi0) measurements
collected across the program.
• Propionaldehyde is a pollutant of interest for OCOK and TOOK; thus, box plots
for propionaldehyde for these sites are presented in Figure 20-36. The maximum
propionaldehyde concentration measured across the program was measured at
TOOK. Even though the range of measurements was larger for TOOK than
OCOK, the annual average concentrations for these two sites are similar to each
other. Both annual averages are greater than the program-level average
concentration as well as the third quartile. The minimum propionaldehyde
concentration measured at OCOK is just greater than the program-level first
quartile.
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 3.5.3.2.
TOOK has sampled TSP metals, carbonyl compounds, and VOCs under the NMP since 2006;
thus, Figures 20-37 through 20-49 present the 1-year statistical metrics for each of the pollutants
of interest for TOOK. 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 is not
provided, although the range and quartiles are still presented. Although PROK has technically
sampled since 2008, sampling did not begin until late October 2008. Because this is less than
6 months of sampling, 2008 would not be included. This would result in fewer than 5 years of
data on the graph; thus, trends graphs were not created for PROK.
20-56
-------
Figure 20-37. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at TOOK
8 4.0
...o
2009
Year
O 5th Percentile
— Maximum
O 95th Percentile "-O-" Averagf
A 1-year average is not presented because issues at the onset of sampling resulted in low completeness.
Observations from Figure 20-37 for acetaldehyde measurements collected 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 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 12
highest concentrations were all measured in 2011 or 2012. Of the 30 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, one was measured in each
year between 2007 and 2009, and two were measured in 2006.
• The statistical metrics exhibit an increasing trend between 2009 and 2011. The 95th
percentile for 2011 and 2012 are greater than the maximum concentrations measured
prior to 2011.
• Little change is shown in the acetaldehyde measurements from 2011 to 2012.
20-57
-------
Figure 20-38. Yearly Statistical Metrics for Arsenic (TSP) Concentrations Measured at TOOK
1
c
0 -, n
Concentrat
4
•
' — !
T
± rn rU ^
> ,.
, - - ' f
2— ' 1 — £ — 1 1 — g — 1 II1
2007 2008 2009 2010 2011 2012
Year
O 5th Percentile - Minimum ~ Median — Maximum O 95th Percentile ...^>... Average
Observations from Figure 20-38 for arsenic (TSP) measurements collected 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 20-38 excludes data from 2006 per the criteria specified in
Section 3.5.3.2.
• The two highest concentrations of arsenic were measured at TOOK in September
2007. These are the only two concentrations greater than 4 ng/m3 measured at TOOK.
• The 1-year average and median concentrations exhibit a decreasing trend between
2007 and 2010, although the difference in relatively small between 2009 and 2010.
Although the range of measurements decreased slightly for 2011, the 1-year average
and median concentrations increased for 2011, an increase that continues into 2012.
20-58
-------
Figure 20-39. Yearly Statistical Metrics for Benzene Concentrations Measured at TOOK
Concentration (jig/m3)
l-i l-i NJ NJ
2006 '
1
>
JJjj
1
^
I
rn
^
j
— i
•
T
O
1 — u — • w z » ' — a — •
2007
O 5th Percentile
Minimum
2008
Median
2009
Year
2010
— Maximum
2011 2012
O 95th Percentile —O— Average
A 1-year average is not presented because issues at the onset of sampling resulted in low completeness.
Observations from Figure 20-39 for benzene measurements collected 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 is not presented for 2006, although the range of measurements is
provided.
• The maximum concentration of benzene was measured in 2011 (23.8 |ig/m3). The
four highest benzene concentrations measured at TOOK were measured in 2011 and
are greater than 10 |ig/m3. 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
substantial decrease from 2008 to 2009, most of the statistical parameters increased
for 2010, and again for 2011. All of the statistical parameters decreased for 2012,
particularly the maximum concentration and 95th percentile.
20-59
-------
Figure 20-40. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at TOOK
o.
o
..o
2009
Year
O 5th Percentile
— Maximum O 95th Percentile
A 1-year average is not presented because issues at the onset of sampling resulted in low completeness.
Observations from Figure 20-40 for 1,3-butadiene measurements collected 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).
• After an initial decrease from 2007 to 2008 and little change in 2009, the 1-year
average concentration began to increase, with the greatest increase occurring from
2010 to 2011. With the exception of the minimum and 5th percentile, all of the
statistical metrics increased between 2009 and 2011. Even though the maximum and
95th percentile decreased, additional increases are shown for the 1-year average and
median concentrations for 2012.
• The minimum concentration for most years is zero, indicating the presence of non-
detects. For 2006, 2010, and 2011, 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).
20-60
-------
Figure 20-41. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
TOOK
• 5th Percentile
- Minimum
— Maximum
• 95th Percentile
A 1-year average is not presented because issues at the onset of sampling resulted in low completeness.
Observations from Figure 20-41 for carbon tetrachloride measurements collected at TOOK
include the following:
• Similar to other compounds, the maximum concentration of carbon tetrachloride was
measured in 2011 (1.64 |ig/m3). With the exception of 2011, the range of carbon
tetrachloride measurements spans 1 |ig/m3 or less. The range of measurements is at a
minimum for 2012, when the difference between the minimum and maximum
concentration is less than 0.45 |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 averages
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.
• For each year shown, the median concentration is very similar to the 1-year average
concentration. The difference between these two parameters is greatest for 2009, yet
only 0.016 |ig/m3 separates them. This indicates that there is relatively little
variability in the central tendency of the measurements.
20-61
-------
Figure 20-42. Yearly Statistical Metrics for/7-Dichlorobenzene Concentrations Measured at
TOOK
Concentration (ng/m3)
pppppi->i->i->
I
2006 '
1
^fcii
' — a — '
2007
• 5th Percentile
Minimum
...<>....
2008
— (
; —
~
r^
2009
Year
Median
1
....<>...
2010
— Maximum
— i
i-<
> —
>....
^H
2011
• 95th Percentile .—O»
t^m
2012
Averag
A 1-yearavera
ge is not presented because issues
at the onset of sampling resulted in low completeness.
Observations from Figure 20-42 for/>-dichlorobenzene measurements collected at TOOK
include the following:
• The maximum concentration ofp-dichlorobenzene was measured in 2008
(1.33 |ig/m3) and is the only measurement greater than 0.70 |ig/m3 measured at
TOOK.
• There were no non-detects of />-dichlorobenzene measured at TOOK in 2006 or 2007.
After 2007, at least two non-detects were measured each year. For 2008 and 2010
through 2012, the minimum and 5th percentile are both zero, indicating the presence
of additional non-detects. For 2010 through 2012, six non-detects were measured
each year.
• The 1-year average concentration fluctuated between 0.12 |ig/m3 and 0.16 |ig/m3
between 2007 and 2011. The 1-year average decreased significantly from 2011 to
2012 (0.09 |ig/m3). 2012 is the first year that a/>-dichlorobenzene concentration
greater than 0.20 |ig/m3 was not measured. By comparison, 15 concentrations greater
than 0.20 |ig/m3 were measured in 2011 and at least eight concentrations greater than
0.20 |ig/m3 were measured every other year of sampling. Additional years of
sampling are needed to determine if this trend continues.
20-62
-------
Figure 20-43. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
TOOK
• 5th Percentile
— Maximum
• 95th Percentile
A 1-year average is not presented because issues at the onset of sampling resulted in low completeness.
Observations from Figure 20-43 for 1,2-dichloroethane measurements collected at TOOK
include the following:
• 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.
• The median concentration for all years through 2011 is zero, indicating that at least
half of the measurements were non-detects. The number of non-detects decreased to
22 for 2012, accounting for 37 percent of the valid samples collected.
• The 1-year average concentration for 2012 is less than the median concentration,
which is a little unusual. The 1-year average 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 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 pull the average up).
20-63
-------
Figure 20-44. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at TOOK
o
^m
-2-
...o
2009
Year
O 5th Percentile
— Maximum
O 95th Percentile
A 1-year average is not presented because issues at the onset of sampling resulted in low completeness.
Observations from Figure 20-44 for ethylbenzene measurements collected 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. Even the median increased, although slightly. Even if the two
highest concentrations measured in 2008 were excluded from the dataset, the 1-year
average would still exhibit a slight increase.
• Most of the statistical parameters are at a minimum for 2009. The 1-year average and
median concentrations decreased by half from 2008. There were no ethylbenzene
concentrations greater than 1 |ig/m3 measured at TOOK in 2009 while at least seven
concentrations greater than 1 |ig/m3 were measured during the other years of
sampling.
• After 2009, concentrations of ethylbenzene measured at TOOK exhibit a significant
increasing trend. Although several of the pollutants of interest for TOOK increased
slightly for 2012, the increase shown for ethylbenzene is the most significant.
20-64
-------
Figure 20-45. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at TOOK
.o
...o
o
2009
Year
O 5th Percentile
— Maximum O 95th Percentile
A 1-year average is not presented because issues at the onset of sampling resulted in low completeness.
Observations from Figure 20-45 for formaldehyde measurements collected at TOOK
include the following:
• The maximum concentration of formaldehyde (12.80 |ig/m3) was measured at TOOK
on June 26, 2012. Only one other measurement greater than 10 |ig/m3 has been
measured at TOOK (10.1 |ig/m3 measured in 2011).
• All but two of the 71 formaldehyde measurements greater than 5 |ig/m3 were
measured during the period between May and September, regardless of year.
• Similar to acetaldehyde, an increasing trend in the 1-year average concentration is
shown for formaldehyde between 2009 and 2011. The 1-year average increased by
1 |ig/m3 over this period.
• Even though the maximum formaldehyde concentration was measured in 2012, all of
the other statistical parameters exhibit slight decreases for this year. This is because
there were fewer concentrations at the upper end of the concentration range for 2012.
The number of formaldehyde measurements greater than 6 |ig/m3 decreased from 10
in 2011 to five in 2012. In addition, there were more concentrations at the lower end
of the concentration range for 2012. The number of formaldehyde measurements less
than 1.5 |ig/m3 increased from three in 2011 to eight in 2012.
20-65
-------
Figure 20-46. Yearly Statistical Metrics for Hexachloro-l,3-Butadiene Concentrations
Measured at TOOK
1
§
1
1
0.00
M
20
••
<•
rn
T
1^ ^^ ^^ o '
06 ' 2007 2008 2009 2010 2011 2012
Year
• 5th Percentile - Minimum ~ Median — Maximum O 95th Percentile ...^... Average
Observations from Figure 20-46 for hexachloro-l,3-butadiene measurements collected 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 few 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.
20-66
-------
Figure 20-47. Yearly Statistical Metrics for Manganese (TSP) Concentrations Measured at
TOOK
O 5th Percentile - Minimurr
— Maximum • 95th Percentile
Observations from Figure 20-47 for manganese (TSP) measurements collected at TOOK
include the following:
• The maximum concentration of manganese was measured in 2012 (273 ng/m3), on
the day of the dust storm. Measurements greater than 100 ng/m3 were also measured
in 2007 (131 ng/m3) and 2011 (104 ng/m3).
• A steady 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 five 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 25 ng/m3
decreased from 26 in 2011 to 19 in 2012).
20-67
-------
Figure 20-48. Yearly Statistical Metrics for Nickel (TSP) Concentrations Measured at TOOK
14
12
10
I8'
c
.9
S
I6'
4.
2.
0.
1
1
2007
O 5th Percentile
pL
^
'— s^
2008
- Min mum
^\
^3
2009
I
^H
2010
Year
rh
P^
i
2011
'
b —
O
^^~
—
2012
Median — Maximum O 95th Percentile ...^>... Average
Observations from Figure 20-48 for nickel (TSP) measurements collected at TOOK
include the following:
• The trends graph for nickel resembles the trends graph for manganese in several
ways.
• The maximum concentration of nickel was measured at TOOK on the same day as the
maximum concentration of manganese (October 18, 2012, the day of the dust storm).
The next two highest concentrations of nickel were also measured in 2012.
Collectively, these three measurements are the only nickel concentrations greater than
6 ng/m3 measured at TOOK since the onset of sampling.
• 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.
• The median concentration for 2011 is very similar to the median concentration for
2012. For 2011, 75 percent of measurements lie between 1 ng/m3 and 3 ng/m3. For
2012, this number is 79 percent. The higher concentrations measured in 2012 (there
are seven concentrations from 2012 that are greater than the maximum concentration
measured in 2011) are balanced by the lower concentrations measured in 2011 (there
are 11 measurements less than 1 ng/m3 in 2011 compared to two in 2012), resulting in
similar median concentrations.
20-68
-------
Figure 20-49. Yearly Statistical Metrics for Propionaldehyde Concentrations Measured at
TOOK
£
1
§
3 1.0
2009
Year
• 5th Percentile
— Maximum
• 95th Percentile
A 1-year average is not presented because issues at the onset of sampling resulted in low completeness.
Observations from Figure 20-49 for propionaldehyde measurements collected at TOOK
include the following:
• The maximum concentration of propionaldehyde (2.02 |ig/m3) was measured at
TOOK on the same day as the maximum formaldehyde concentration
(June 26, 2012). At least one measurement greater than 1.0 |ig/m3 has been measured
at TOOK each year of sampling, except 2009.
• Similar to acetaldehyde and formaldehyde, an increasing trend in the 1-year average
concentrations of propionaldehyde is shown between 2009 and 2011.
• With the exception of the maximum concentration, little change is shown for the
statistical parameters for 2012.
20.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk at
each Oklahoma monitoring site. Refer to Sections 3.3 and 3.5.3 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.
20-69
-------
20.5.1 Risk-Based Screening Assessment Using MRLs
A risk-based screening was conducted by comparing the concentration data from the
Oklahoma monitoring sites to the ATSDR MRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 day to 14 days); intermediate (exposures of 15 days to 364 days); and chronic
(exposures of 1 year or greater). The preprocessed daily measurements of the pollutants of
interest were compared to the acute MRLs; the quarterly averages were compared to the
intermediate MRLs; and the annual averages were compared to the chronic MRLs.
As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2012.
20.5.2 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 3.5.3.4 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 20-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
20-70
-------
Table 20-6. Risk Approximations for the Oklahoma Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections vs.
# of Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
Tulsa, Oklahoma - TOOK
Acetaldehyde
Arsenic (TSP)a
Benzene
1,3 -Butadiene
Carbon Tetrachloride
£>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
Hexachloro- 1 ,3 -butadiene
Manganese (TSP)a
Nickel (TSP)a
Propionaldehyde
0.0000022
0.0043
0.0000078
0.00003
0.000006
0.000011
0.000026
0.0000025
0.000013
0.000022
0.00048
0.009
0.000015
0.03
0.002
0.1
0.8
2.4
1
0.0098
0.09
0.00005
0.00009
0.008
61/61
61/61
60/60
60/60
60/60
55/60
41/60
60/60
61/61
10/60
61/61
61/61
61/61
2.78
±0.42
0.01
±0.01
2.21
±0.31
0.10
±0.02
0.66
±0.02
0.09
±0.01
0.07
±0.01
0.91
±0.17
3.42
±0.54
0.01
±0.01
0.04
±0.01
O.01
±O.01
0.50
±0.08
6.11
3.97
17.24
2.88
3.95
1.03
1.88
2.29
44.48
0.28
1.16
0.31
0.06
0.07
0.05
0.01
0.01
O.01
O.01
0.35
O.01
0.77
0.03
0.06
— = A Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 20-5.
20-71
-------
Table 20-6. Risk Approximations for the Oklahoma Monitoring Sites (Continued)
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections vs.
# of Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
Tulsa, Oklahoma - TMOK
Acetaldehyde
Arsenic (TSP)a
Benzene
1,3 -Butadiene
Carbon Tetrachloride
£>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
Manganese (TSP)a
Nickel (TSP)a
0.0000022
0.0043
0.0000078
0.00003
0.000006
0.000011
0.000026
0.0000025
0.000013
0.00048
0.009
0.000015
0.03
0.002
0.1
0.8
2.4
1
0.0098
0.00005
0.00009
59/59
61/61
61/61
60/61
61/61
54/61
42/61
61/61
59/59
61/61
61/61
2.33
±0.32
0.01
±0.01
1.25
±0.16
0.12
±0.02
0.68
±0.02
0.08
±0.01
0.06
±0.01
0.56
±0.08
3.63
±0.47
0.03
±0.01
0.01
±0.01
5.13
3.32
9.78
3.75
4.08
0.90
1.67
1.39
47.13
0.80
0.26
0.05
0.04
0.06
0.01
0.01
O.01
O.01
0.37
0.52
0.02
Pryor Creek, Oklahoma - PROK
Acetaldehyde
Arsenic (TSP)a
Benzene
1,3 -Butadiene
Carbon Tetrachloride
£>-Dichlorobenzene
1 ,2-Dichloroethane
Formaldehyde
Manganese (TSP)a
0.0000022
0.0043
0.0000078
0.00003
0.000006
0.000011
0.000026
0.000013
0.009
0.000015
0.03
0.002
0.1
0.8
2.4
0.0098
0.00005
51/51
49/49
51/51
39/51
51/51
45/51
47/51
51/51
49/49
1.56
±0.19
O.01
±0.01
0.61
±0.11
0.05
±0.01
0.69
±0.03
0.09
±0.02
0.07
±0.01
3.58
±0.65
0.02
±0.01
3.44
2.73
4.78
1.48
4.17
1.02
1.84
46.53
0.17
0.04
0.02
0.02
0.01
0.01
O.01
0.37
0.37
- = A Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 20-5.
20-72
-------
Table 20-6. Risk Approximations for the Oklahoma Monitoring Sites (Continued)
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections vs.
# of Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
Oklahoma City, Oklahoma - ADOK
Acetaldehyde
Arsenic (TSP)a
Benzene
1,3 -Butadiene
Carbon Tetrachloride
£>-Dichlorobenzene
1 ,2-Dichloroethane
Formaldehyde
Manganese (TSP)a
0.0000022
0.0043
0.0000078
0.00003
0.000006
0.000011
0.000026
0.000013
0.009
0.000015
0.03
0.002
0.1
0.8
2.4
0.0098
0.00005
66/66
64/64
66/66
44/66
66/66
57/66
54/66
66/66
64/64
1.81
±0.24
0.01
±0.01
0.63
±0.08
0.04
±0.01
0.67
±0.03
0.13
±0.04
0.06
±0.01
3.00
±0.46
0.01
±0.01
3.98
2.12
4.93
1.29
4.03
1.38
1.66
38.96
0.20
0.03
0.02
0.02
0.01
0.01
O.01
0.31
0.26
Oklahoma City, Oklahoma - OCOK
Acetaldehyde
Arsenic (TSP)a
Benzene
1,3 -Butadiene
Carbon Tetrachloride
/>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
Manganese (TSP)a
Propionaldehyde
0.0000022
0.0043
0.0000078
0.00003
0.000006
0.000011
0.000026
0.0000025
0.000013
0.009
0.000015
0.03
0.002
0.1
0.8
2.4
1
0.0098
0.00005
0.008
60/60
61/61
61/61
56/61
61/61
54/61
52/61
61/61
60/60
61/61
60/60
2.34
±0.32
O.01
±0.01
0.78
±0.12
0.08
±0.04
0.66
±0.03
0.07
±0.01
0.07
±0.01
0.31
±0.09
3.49
±0.54
0.02
±0.01
0.48
±0.06
5.14
2.46
6.07
2.30
3.96
0.74
1.80
0.78
45.35
0.26
0.04
0.03
0.04
0.01
0.01
O.01
0.01
0.36
0.42
0.06
- = A Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 20-5.
20-73
-------
Observations from Table 20-6 include the following:
• Formaldehyde and acetaldehyde have the highest annual average concentrations for
each site. Among the TSP metals, the annual average concentration of manganese is
the highest for each site.
• Formaldehyde and benzene have the highest cancer risk approximations among the
pollutants of interest for the Oklahoma monitoring sites. Formaldehyde cancer risk
approximations range from 38.96 in-a-million for ADOK to 47.13 in-a-million for
TMOK. The cancer risk approximations for formaldehyde for TMOK, PROK, and
OCOK rank fifth, sixth, and seventh highest among all cancer risk approximations
program-wide. Benzene cancer risk approximations range from 4.78 in-a-million for
PROK to 17.24 in-a-million for TOOK. The benzene cancer risk approximation for
TOOK is the highest cancer risk approximation calculated for benzene program-wide.
• Among the metals, arsenic has the highest cancer risk approximations for all of the
Oklahoma monitoring sites, ranging from 2.12 in-a-million for ADOK to
3.97 in-a-million for TOOK. Note that manganese do not have a cancer URE.
• None of the pollutants of interest have noncancer hazard approximations greater than
1.0, indicating that no adverse health effects are expected from these individual
pollutants.
• Among the noncancer hazard approximations for the Oklahoma sites, formaldehyde
and manganese have the highest noncancer hazard approximations for each site
(albeit less than 1.0). The noncancer hazard approximation for manganese for TOOK
(0.77) is the highest noncancer hazard approximations calculated across the program.
20.5.3 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 20-7 presents the 10 pollutants with the highest emissions from the 2011 NEI that have
cancer toxicity factors. Table 20-7 also presents the 10 pollutants with the highest toxicity -
weighted emissions, based on the weighting schema described in Section 3.5.3.5. Lastly,
Table 20-7 provides the 10 pollutants with the highest cancer risk approximations (in-a-million)
for each site, as presented in Table 20-6. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 20-7. Table 20-8 presents similar
information, but is limited to those pollutants with noncancer toxicity factors.
20-74
-------
Table 20-7. 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)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Tulsa, Oklahoma (Tulsa County) - TOOK
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
POM, Group la
1,3 -Butadiene
Tetrachloroethylene
Naphthalene
Trichloroethylene
Dichloromethane
303.37
202.14
166.00
84.86
63.48
36.70
30.44
26.24
16.89
8.60
POM, Group la
Benzene
Formaldehyde
1,3 -Butadiene
Hexavalent Chromium, PM
Naphthalene
Ethylbenzene
Nickel, PM
POM, Group 2b
Arsenic, PM
5.59E-03
2.37E-03
2.16E-03
1.10E-03
9.06E-04
8.92E-04
5.05E-04
2.74E-04
2.33E-04
1.98E-04
Formaldehyde
Benzene
Acetaldehyde
Arsenic
Carbon Tetrachloride
1,3 -Butadiene
Ethylbenzene
1 ,2-Dichloroethane
Nickel
£>-Dichlorobenzene
44.48
17.24
6.11
3.97
3.95
2.88
2.29
1.88
1.16
1.03
Tulsa, Oklahoma (Tulsa County) - TMOK
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
POM, Group la
1,3 -Butadiene
Tetrachloroethylene
Naphthalene
Trichloroethylene
Dichloromethane
303.37
202.14
166.00
84.86
63.48
36.70
30.44
26.24
16.89
8.60
POM, Group la
Benzene
Formaldehyde
1,3 -Butadiene
Hexavalent Chromium, PM
Naphthalene
Ethylbenzene
Nickel, PM
POM, Group 2b
Arsenic, PM
5.59E-03
2.37E-03
2.16E-03
1.10E-03
9.06E-04
8.92E-04
5.05E-04
2.74E-04
2.33E-04
1.98E-04
Formaldehyde
Benzene
Acetaldehyde
Carbon Tetrachloride
1,3 -Butadiene
Arsenic
1 ,2-Dichloroethane
Ethylbenzene
£>-Dichlorobenzene
Nickel
47.13
9.78
5.13
4.08
3.75
3.32
1.67
1.39
0.90
0.80
to
o
-------
Table 20-7. 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)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Pryor Creek, Oklahoma (Mayes County) - PROK
Benzene
Formaldehyde
Acetaldehyde
Ethylbenzene
POM, Group la
1,3 -Butadiene
Naphthalene
Chloromethylbenzene
Nickel, PM
Aniline
30.82
20.34
13.15
12.72
3.38
3.07
2.13
1.55
1.17
0.75
Hexavalent Chromium, PM
Arsenic, PM
Nickel, PM
POM, Group la
Formaldehyde
Benzene
Beryllium, PM
1,3 -Butadiene
Chloromethylbenzene
Naphthalene
2.94E-03
2.31E-03
5.63E-04
2.97E-04
2.64E-04
2.40E-04
1.14E-04
9.20E-05
7.58E-05
7.23E-05
Formaldehyde
Benzene
Carbon Tetrachloride
Acetaldehyde
Arsenic
1 ,2-Dichloroethane
1,3 -Butadiene
£>-Dichlorobenzene
46.53
4.78
4.17
3.44
2.73
1.84
1.48
1.02
Oklahoma City, Oklahoma (Oklahoma County) - ADOK
Benzene
Ethylbenzene
Formaldehyde
POM, Group la
Acetaldehyde
1,3 -Butadiene
Naphthalene
Dichloromethane
Tetrachloroethylene
POM, Group 2b
330.65
224.06
218.33
114.89
114.57
42.92
22.29
14.78
8.04
3.24
POM, Group la
Formaldehyde
Benzene
1,3 -Butadiene
Naphthalene
Ethylbenzene
Hexavalent Chromium, PM
POM, Group 2b
POM, Group 2d
Acetaldehyde
1.01E-02
2.84E-03
2.58E-03
1.29E-03
7.58E-04
5.60E-04
3.91E-04
2.85E-04
2.56E-04
2.52E-04
Formaldehyde
Benzene
Carbon Tetrachloride
Acetaldehyde
Arsenic
1 ,2-Dichloroethane
£>-Dichlorobenzene
1,3 -Butadiene
38.96
4.93
4.03
3.98
2.12
1.66
1.38
1.29
to
o
-------
Table 20-7. 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)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Oklahoma City, Oklahoma (Oklahoma County) - OCOK
Benzene
Ethylbenzene
Formaldehyde
POM, Group la
Acetaldehyde
1,3 -Butadiene
Naphthalene
Dichloromethane
Tetrachloroethylene
POM, Group 2b
330.65
224.06
218.33
114.89
114.57
42.92
22.29
14.78
8.04
3.24
POM, Group la
Formaldehyde
Benzene
1,3 -Butadiene
Naphthalene
Ethylbenzene
Hexavalent Chromium, PM
POM, Group 2b
POM, Group 2d
Acetaldehyde
1.01E-02
2.84E-03
2.58E-03
1.29E-03
7.58E-04
5.60E-04
3.91E-04
2.85E-04
2.56E-04
2.52E-04
Formaldehyde
Benzene
Acetaldehyde
Carbon Tetrachloride
Arsenic
1,3 -Butadiene
1 ,2-Dichloroethane
Ethylbenzene
£>-Dichlorobenzene
45.35
6.07
5.14
3.96
2.46
2.30
1.80
0.78
0.74
to
o
-------
Table 20-8. 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)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Tulsa, Oklahoma (Tulsa County) - TOOK
Toluene
Ethylene glycol
Xylenes
Hexane
Methanol
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
Methyl isobutyl ketone
1,823.14
742.90
742.04
689.70
360.74
303.37
202.14
166.00
84.86
78.48
Acrolein
1,3 -Butadiene
Formaldehyde
Manganese, PM
Benzene
Acetaldehyde
Naphthalene
Trichloroethylene
Xylenes
Nickel, PM
443,042.58
18,352.00
16,939.11
12,843.28
10,112.23
9,429.01
8,747.25
8,445.87
7,420.42
6,347.63
Manganese
Formaldehyde
Acetaldehyde
Benzene
Propionaldehyde
Arsenic
1,3 -Butadiene
Nickel
Carbon Tetrachloride
Ethylbenzene
0.77
0.35
0.31
0.07
0.06
0.06
0.05
0.03
0.01
<0.01
Tulsa, Oklahoma (Tulsa County) - TMOK
Toluene
Ethylene glycol
Xylenes
Hexane
Methanol
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
Methyl isobutyl ketone
1,823.14
742.90
742.04
689.70
360.74
303.37
202.14
166.00
84.86
78.48
Acrolein
1,3 -Butadiene
Formaldehyde
Manganese, PM
Benzene
Acetaldehyde
Naphthalene
Trichloroethylene
Xylenes
Nickel, PM
443,042.58
18,352.00
16,939.11
12,843.28
10,112.23
9,429.01
8,747.25
8,445.87
7,420.42
6,347.63
Manganese
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Arsenic
Benzene
Nickel
Carbon Tetrachloride
Ethylbenzene
£>-Dichlorobenzene
0.52
0.37
0.26
0.06
0.05
0.04
0.02
0.01
<0.01
<0.01
to
o
oo
-------
Table 20-8. 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)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Pryor Creek, Oklahoma (Mayes County) - PROK
Toluene
Hydrochloric acid
Xylenes
Ethylene glycol
Hexane
Benzene
Methanol
Formaldehyde
Acetaldehyde
Ethylbenzene
128.09
75.26
56.41
50.78
41.02
30.82
24.93
20.34
13.15
12.72
Acrolein
Chlorine
Arsenic, PM
Manganese, PM
Nickel, PM
Cyanide Compounds, PM
Lead, PM
Hydrochloric acid
Cadmium, PM
Hexavalent Chromium, PM
83,815.30
56,686.67
35,830.19
15,273.86
13,034.79
6,904.00
6,528.90
3,763.00
2,946.28
2,447.13
Manganese
Formaldehyde
Acetaldehyde
Arsenic
1,3 -Butadiene
Benzene
Carbon Tetrachloride
£>-Dichlorobenzene
1 ,2-Dichloroethane
0.37
0.37
0.17
0.04
0.02
0.02
0.01
<0.01
<0.01
Oklahoma City, Oklahoma (Oklahoma County) - ADOK
Toluene
Ethylene glycol
Xylenes
Hexane
Methanol
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
Methyl isobutyl ketone
2,213.83
943.84
886.13
734.44
445.10
330.65
224.06
218.33
114.57
71.17
Acrolein
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Benzene
Xylenes
Naphthalene
Arsenic, PM
Ethylene glycol
Lead, PM
668,119.67
22,278.35
21,459.32
12,730.19
11,021.63
8,861.28
7,429.80
2,849.44
2,359.59
2,138.17
Formaldehyde
Manganese
Acetaldehyde
Arsenic
1,3 -Butadiene
Benzene
Carbon Tetrachloride
£>-Dichlorobenzene
1 ,2-Dichloroethane
0.31
0.26
0.20
0.03
0.02
0.02
0.01
<0.01
<0.01
to
o
VO
-------
Table 20-8. 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)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Oklahoma City, Oklahoma (Oklahoma County) - OCOK
Toluene
Ethylene glycol
Xylenes
Hexane
Methanol
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
Methyl isobutyl ketone
2,213.83
943.84
886.13
734.44
445.10
330.65
224.06
218.33
114.57
71.17
Acrolein
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Benzene
Xylenes
Naphthalene
Arsenic, PM
Ethylene glycol
Lead, PM
668,119.67
22,278.35
21,459.32
12,730.19
11,021.63
8,861.28
7,429.80
2,849.44
2,359.59
2,138.17
Manganese
Formaldehyde
Acetaldehyde
Propionaldehyde
1,3 -Butadiene
Arsenic
Benzene
Carbon Tetrachloride
Ethylbenzene
/>-Dichlorobenzene
0.42
0.36
0.26
0.06
0.04
0.04
0.03
0.01
<0.01
<0.01
to
o
oo
o
-------
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 3.5.3.5. Similar to the cancer risk and
noncancer hazard approximations provided in Section 20.5.2, this analysis may help policy-
makers prioritize their air monitoring activities.
Observations from Table 20-7 include the following:
• Benzene is the highest emitted pollutant with a cancer URE in Mayes, Oklahoma, and
Tulsa Counties. The quantity of benzene emissions in Mayes County is an order of
magnitude lower than the quantity of emissions for Oklahoma and Tulsa Counties.
• The pollutant with the highest toxicity-weighted emissions (of the pollutants with
cancer UREs) for Oklahoma and Tulsa Counties is POM Group la, followed by
benzene, formaldehyde, and 1,3-butadiene (although not necessarily in that order).
POM, Group la includes all unspeciated POM. The pollutants with the highest
toxicity-weighted emissions for Mayes County are hexavalent chromium, arsenic, and
nickel.
• Six of the highest emitted pollutants in Tulsa County also have the highest toxicity-
weighted emissions. Seven of the highest emitted pollutants in Mayes County also
have the highest toxicity-weighted emissions. Eight of the highest emitted pollutants
in Oklahoma County also have the highest toxicity-weighted emissions. POM, Group
la, benzene, formaldehyde, naphthalene, and 1,3-butadiene appear on both emissions-
based lists for all three counties.
• Hexavalent chromium has the highest toxicity-weighted emissions for Mayes County
and is also listed for Tulsa and Oklahoma Counties, yet it is not among the highest
emitted pollutants for any of these counties (ranking 15th, 23rd, and 28th,
respectively). This indicates that lower emissions can translate to higher risk levels.
Hexavalent chromium was not sampled for at the Oklahoma monitoring sites.
• Formaldehyde and benzene have the highest cancer risk approximations among the
Oklahoma sites' pollutants of interest. These pollutants also appear on both
emissions-based lists for all five sites. Conversely, carbon tetrachloride, whose cancer
risk approximation is in the top five for each site, appears on neither emissions-based
list.
20-81
-------
Observations from Table 20-8 include the following:
• Toluene is the highest emitted pollutant with a noncancer RfC in all three counties,
although the quantity emitted is significantly higher in Tulsa and Oklahoma Counties
than in Mayes County. Xylenes and ethylene glycol are also among the highest
emitted pollutants in all three counties. Hydrochloric acid is also one of the highest
emitted pollutants in Mayes County.
• Acrolein is the pollutant with the highest toxi city-weighted emissions (of the
pollutants with noncancer RfCs) for all three counties. Yet, this pollutant is not
among the highest emitted pollutants for any of the three counties. This indicates that
lower emissions can translate to higher risk levels. 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 Tulsa County also have the highest toxi city-
weighted emissions; five of the highest emitted pollutants in Oklahoma County also
have the highest toxi city-weighted emissions. Only one of the highest emitted
pollutants in Mayes County also has one of the highest toxicity-weighted emissions
(hydrochloric acid). Note that although toluene is the highest emitted pollutant in all
three counties, this pollutant does not appear among those with the highest toxicity-
weighted emissions.
• Six of the 10 pollutants with the highest noncancer toxicity-weighted emissions in
Mayes County are metals. None of these appear among the highest emitted, though.
• Formaldehyde and manganese have the highest noncancer hazard approximations
among the Oklahoma sites. Formaldehyde appears on both emissions-based lists for
Tulsa and Oklahoma Counties but ranks 12th for toxicity-weighted emissions for
Mayes County and therefore does not appear in Table 20-8 in that column.
Manganese appears among the pollutants with the highest toxicity-weighted
emissions for Tulsa and Mayes Counties but ranks 17th for toxicity-weighted
emissions for Oklahoma County. There are no metals listed among the highest
emitted pollutants for any of the three counties.
• Note that for the metals, the emissions-based lists are PMio while the Oklahoma sites
sampled TSP metals.
20.6 Summary of the 2012 Monitoring Data for the Oklahoma Monitoring Sites
Results from several of the data treatments described in this section include the
following:
»«» Twenty-one pollutants failed at least one screen for TOOK; 19 pollutants failed
screens for TMOK; 18 pollutants failed screens for PROK; 17 pollutants failed
screens for ADOK; and 17 pollutants failed screens for OCOK.
20-82
-------
»«» Formaldehyde and acetaldehyde had the highest annual average concentration for
each site. Among the TSP metals, the annual average concentration of manganese
was the highest for each site.
»«» The maximum manganese concentrations measured at TOOK, TMOK, andPROK
correlate to the day a large dust storm affected the area.
»«» TOOK had the highest annual average of concentration of benzene and ethylbenzene
among all NMP sites sampling this pollutant.
»«» Concentrations of ethylbenzene, manganese, and nickel exhibit increasing trends at
TOOK. In addition, the detection rate of 1,2-dichloroethane has been increasing at
TOOK over the last few years of sampling, particularly for 2012.
20-83
-------
21.0 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.
21.1 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 21-1 is a composite
satellite image retrieved from ArcGIS Explorer showing the 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. 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 radius
are still visible on the map, but have been grayed out in order to show emissions sources just
outside the boundary. Table 21-1 provides supplemental geographical information such as land
use, location setting, and locational coordinates.
21-1
-------
Figure 21-1. Providence, Rhode Island (PRRI) Monitoring Site
to
to
-------
Figure 21-2. NEI Point Sources Located Within 10 Miles of PRRI
Legend
71 'SO'ETW 71 D25'
-------
Table 21-1. Geographical Information for the Rhode Island Monitoring Site
Site
Code
PRRI
AQS Code
44-007-0022
Location
Providence
County
Providence
Micro- or
Metropolitan
Statistical Area
Providence-
Warwick, RI-MA
MSA
Latitude
and
Longitude
41.807776,
-71.415105
Land Use
Residential
Location
Setting
Urban/City
Center
Additional Ambient Monitoring Information1
PAMS, VOCs, Carbonyl Compounds, Meteorological
parameters, PM10, PM10 Speciation, Black Carbon,
PM25, and PM25 Metals, TSP Germanium.
1 Data for additional pollutants are reported to AQS for PRRI (EPA, 2013b); however, these data are not generated by ERG and are therefore not included in this report.
BOLD ITALICS = EPA-designaled NATTS Site
to
I
-------
Figure 21-1 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 River leads into
Providence Harbor a few tenths of a mile farther to the east, just on the other side of 1-95.
Figure 21-2 shows that a large number of point sources are located within 10 miles of PRRI. 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; plastic, resin, or
rubber products plants; and facilities generating electricity via combustion. Sources within one-
half mile of PRRI include several hospitals and a heliport at a hospital.
Table 21-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Rhode Island monitoring site. Table 21-2 includes the county-
level population for the site. County-level vehicle registration data for Providence County were
not available from the State of Rhode Island. Thus, state-level vehicle registration, which was
obtained from the Federal Highway Administration, was allocated to the county level using the
county-level proportion of the state population from the U.S. Census Bureau. Table 21-2 also
contains traffic volume information for PRRI as well as the location for which the traffic volume
was obtained. County-level VMT data were not readily available for Providence County.
Table 21-2. Population, Motor Vehicle, and Traffic Information for the Rhode Island
Monitoring Site
Site
PRRI
Estimated
County
Population1
628,323
County-level
Vehicle
Registration2
548,763
Annual
Average Daily
Traffic3
136,800
Intersection
Used for
Traffic Data
1-95 near 1-195
County-
level Daily
VMT4
NA
Bounty-level population estimate reflects 2012 data (Census Bureau, 2013c)
2County-level vehicle registration is based on 2011 state-level vehicle registration data from the FHWA and the
2011 county-level proportion of the state population data (FHWA, 2013aand Census Bureau, 2012)
3AADT reflects 2009 data (RI DOT, 2009)
4County-level VMT was not available for this site.
BOLD ITALICS = EPA-designaled NATTS Site
Observations from Table 21-2 include the following:
• Providence County's population is in the middle of the range compared to other
counties with NMP sites.
21-5
-------
• The estimated county-level vehicle registration is also in the middle of the range
compared to other counties with NMP sites.
• The traffic volume experienced near PRRI is the 10th highest compared to traffic
volume near other NMP monitoring sites. The traffic estimate provided is for 1-95
near the 1-195 interchange.
21.2 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.
21.2.1 Climate Summary
Providence is a coastal city on the Narragansett Bay, which opens to the Rhode Island
Sound and the Atlantic Ocean. The city's proximity to the Sound and the Atlantic Ocean temper
cold air outbreaks, and breezes off the ocean moderate summertime heat. On average, southerly
and southwesterly winds in the summer become northwesterly in the winter. Weather is fairly
variable in Providence as storm systems frequently affect the New England region. Precipitation
occurs in Providence about one day in every three and is distributed fairly evenly throughout the
year. Thunderstorms are common between May and August, while coastal storms tend to
produce the greatest amounts of rain and snow. Thirty inches of snow is typical in winter (Wood,
2004; CoCoRaHS, 2011).
21.2.2 Meteorological Summary
Hourly meteorological data for 2012 were retrieved from NCDC for the weather station
closest to the Rhode Island monitoring sites (NCDC, 2012), as described in Section 3.5.2. The
closest weather station is located at Theodore F. Green State Airport (WBAN 14765). Additional
information about the T.F. Green Airport weather station, such as the distance between the site
and the weather station, is provided in Table 21-3. These data were used to determine how
meteorological conditions on sample days vary from conditions experienced throughout the year.
21-6
-------
Table 21-3. Average Meteorological Conditions near the Rhode Island Monitoring Site
Closest
Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Providence, Rhode Island - PRRI
Theodore F.
Green State
Airport
14765
(41.72, -71.43)
6.0
miles
173°
(S)
Sample
Days
(65)
2012
62.2
±4.2
62.1
+ 1.7
54.2
±4.0
53.8
+ 1.6
42.9
±4.4
42.2
+ 1.8
48.9
±3.7
48.4
+ 1.5
68.7
±3.6
68.0
+ 1.6
1015.9
±2.0
1015.7
+ 0.8
6.5
±0.6
6.8
+ 0.3
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
to
-------
Table 21-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for all of 2012. Also included in Table 21-3 is the 95 percent
confidence interval for each parameter. As shown in Table 21-3, average meteorological
conditions on sample days are representative of average weather conditions experienced
throughout the year near PRRI.
21.2.3 Back Trajectory Analysis
Figure 21-3 is the composite back trajectory map for days on which samples were
collected at the PRRI monitoring site. Included in Figure 21-3 are four back trajectories per
sample day. Figure 21-4 is the corresponding cluster analysis. An in-depth description of these
maps and how they were generated is presented in Section 3.5.2.1. For the composite map, each
line represents the 24-hour trajectory along which a parcel of air traveled toward the monitoring
site on a given sample day and time, based on an initial height of 50 meters AGL. For the cluster
analysis, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric circle around the site in Figures 21-3 and 21-4 represents 100 miles.
Figure 21-3. Composite Back Trajectory Map for PRRI
21-8
-------
Figure 21-4. Back Trajectory Cluster Map for PRRI
/
Observations from Figures 21-3 and 21-4 for PRRI include the following:
• Back trajectories originated from a variety of directions at PRRI.
• The airshed domain for PRRI was similar in size to other NMP sites, based on the
average back trajectory length. The average back trajectory length was 246 miles,
although the farthest away a back trajectory originated was over south-central
Ontario, Canada, or nearly 700 miles away. Yet, nearly 92 percent of back trajectories
originated within 450 miles of the site.
• The cluster analysis shows that 36 percent of back trajectories originated to the
northwest and north of PRRI, from the northern half of New York eastward toward
Maine. Another 20 percent originated to the west and southwest of the site, over the
southern half of New York, Pennsylvania, New Jersey and their offshore waters.
Back trajectories originating from the west and northwest but farther away are
represented by the longer cluster trajectory originating near Lake Huron (7 percent).
Thirteen percent of back trajectories originated over the offshore waters of the Mid-
Atlantic states. The back trajectories originating over North Carolina and Virginia and
curving eastward and then northward toward PRRI are included in this cluster
trajectory. These back trajectories represent the October 30, 2012 sample day, the day
after Hurricane Sandy made landfall in New Jersey. Nearly one-quarter of back
trajectories originated to the east of PRRI, over the Gulf of Maine and southward over
the Atlantic Ocean.
21-9
-------
21.2.4 Wind Rose Comparison
Hourly surface wind data from the weather station at T.F. Green Airport near PRRI were
uploaded into a wind rose software program to produce customized wind roses, as described in
Section 3.5.2.2. 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 21-5 presents a map showing the distance between the weather station and PRRI,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 21-5 also presents three different wind roses for the
PRRI monitoring site. First, a historical wind rose representing 2002 to 2011 wind data is
presented, which shows the predominant surface wind speed and direction over an extended
period of time. Second, a wind rose representing wind observations for all of 2012 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2012 is
presented. These can be used to identify the predominant wind speed and direction for 2012 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.
Observations from Figure 21-5 for PRRI include the following:
• The weather station at T.F. Green Airport is located 6 miles south of PRRI.
• The historical wind rose shows that while westerly winds were observed the most
(approximately 11 percent of observations), winds from the western quadrants, due
north, and due south are common near PRRI. Calm winds (< 2 knots) account for less
than nine percent of the hourly measurements.
• The wind patterns shown on the 2012 wind rose are similar to the historical wind
patterns, with winds from the western quadrants, due north, and due south prevalent
near PRRI. The calm rate for 2012 is 12 percent, which is slightly higher than the
calm rate for the historical wind rose.
• The wind patterns shown on the sample day wind rose continue the prevalence of
winds from the western quadrants and due south, but the number of observations from
the north is reduced. The sample day calm rate is nearly 14 percent compared to
12 percent for 2012 and 9 percent for the historical wind rose. Although still
accounting for relatively few observations, the number of observations from the
north-northeast and northeast on sample days is fewer (less than 3 percent each) than
the percentages shown on the full-year and historical wind rose while the number of
observations from the south-southeast is greater.
21-10
-------
Figure 21-5. Wind Roses for the T.F. Green State Airport Weather Station near PRRI
Location of PRRI and Weather Station
2002-2011 Historical Wind Rose
2012 Wind Rose
Sample Day Wind Rose
Calms: 1206%
Calms: 13.91%
21-11
-------
21.3 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 21-4. 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 21-4. 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 at PRRI.
Table 21-4. Risk-Based Screening Results for the Rhode Island Monitoring Site
Pollutant
Screening
Value
(Ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Providence, Rhode Island - PRRI
Naphthalene
Hexavalent Chromium
Benzo(a)pyrene
0.029
0.000083
0.00057
Total
55
2
1
58
60
41
56
157
91.67
4.88
1.79
36.94
94.83
3.45
1.72
94.83
98.28
100.00
Observations from Table 21-4 include the following:
• Three pollutants failed at least one screen for PRRI; 37 percent of concentrations for
these three pollutants were greater than their associated risk screening value (or failed
screens).
• Concentration of naphthalene failed 55 of the 58 total screens, accounting for just less
than 95 percent of all failed screens for PRRI.
• Naphthalene and hexavalent chromium contributed to 95 percent of failed screens for
PRRI and therefore were identified as pollutants of interest.
21.4 Concentrations
This section presents various concentration averages used to characterize pollution 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:
21-12
-------
• Time period-based concentration averages (quarterly and annual) are provided for the
site.
• Annual concentration averages are presented graphically 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 the site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Additional site-specific statistical summaries for PRRI are
provided in Appendices M and O.
21.4.1 2012 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 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 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 PRRI
are presented in Table 21-5, 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 21-5. Quarterly and Annual Average Concentrations of the Pollutants of
Interest for the Rhode Island Monitoring Site
Pollutant
#of
Measured
Detections vs.
# 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
Hexavalent Chromium
Naphthalene
41/61
60/60
0.006
± 0.004
74.35
±16.14
0.024
± 0.027
43.86
±11.91
0.023
±0.010
73.48
± 16.56
0.010
± 0.007
114.66
±39.59
0.016
± 0.007
76.41
± 12.42
21-13
-------
Observations for PRRI from Table 21-5 include the following:
• Hexavalent chromium was detected in 67 percent of the valid samples collected at
PRRI while naphthalene was detected in 100 percent of samples collected.
• The quarterly and annual average concentrations of naphthalene are significantly
higher than the quarterly and annual average concentrations of hexavalent chromium.
• The confidence interval for the second quarter average concentration of hexavalent
chromium is greater than the average itself, indicating potential outliers. The
maximum hexavalent chromium concentration was measured at PRRI on
May 9, 2012 (0.207 ng/m3). There were no other hexavalent chromium concentrations
greater than 0.1 ng/m3 measured at this site. The next highest concentration measured
during the second quarter of 2012 is an order of magnitude less (0.0269 ng/m3),
explaining the large confidence interval calculated for this quarter.
• Concentrations of naphthalene measured at PRRI span an order of magnitude, ranging
from 21.3 ng/m3 to 212 ng/m3. All four naphthalene concentrations greater than
200 ng/m3 were measured in November and December. All five naphthalene
concentrations less than 30 ng/m3 were measured during the second quarter. This
explains the variability exhibited by the quarterly averages of this pollutant.
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 shaded in
gray in Table 21-4 for PRRI. Figures 21-6 and 21-7 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 3.5.3.1.
Figure 21-6. Program vs. Site-Specific Average Hexavalent Chromium Concentration
.
r
. . ,, if on / 3 '
1
0.2 0.3
Concentration {ng/m3)
Program:
Site:
1st Quartile
D
Site Average
0
2nd Quartile 3rd Quartile
D D
Site Concentration Range
^^^^—
4th Quartile Average
D 1
21-14
-------
Figure 21-7. Program vs. Site-Specific Average Naphthalene Concentration
PRRI
1-4
100
200
300
400 500
Concentration (ng/m3)
600
700
800
900
Program:
Site:
IstQuartile
D
Site Average
0
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
Observations from Figures 21-6 and 21-7 include the following:
• Figure 21-6 is the box plot for hexavalent chromium. Note that the program-level
maximum concentration (8.51 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 to
0.5 ng/m3. In addition, the program-level first quartile is zero and therefore not
visible on the box plot. The annual average concentration of hexavalent chromium
for PRRI is less than the program-level average concentration and similar to
program-level median concentration. The maximum concentration measured at
PRRI is considerably less than the program-level maximum concentration. There
were 20 non-detects of hexavalent chromium measured at PRRI.
• Figure 21-7 is the box plot for naphthalene. The annual average naphthalene
concentration for PRRI is just less than the program-level average concentration.
The maximum naphthalene concentration measured at PRRI is considerably less
than the maximum concentration measured at the program-level. There were no
non-detects of naphthalene measured at PRRI (or across the program).
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 3.5.3.2.
PRRI has sampled hexavalent chromium under the NMP since 2005 and PAHs since 2008. Thus,
Figure 21-8 and 21-9 present the 1-year statistical metrics for each of the pollutants of interest
for PRRI. The statistical metrics presented for assessing trends include the substitution of zeros
for non-detects. If sampling began mid-year, a minimum of 6 months of sampling is required for
inclusion in the trends analysis; in these cases, a 1-year average is not provided, although the
range and quartiles are still presented.
21-15
-------
Figure 21-8. Yearly Statistical Metrics for Hexavalent Chromium Concentrations
Measured at PRRI
I
§
1
5 o.io
• 5th Percentile
- Minimum
— Maximum
• 95th Percentile ...^... Average
Observations from Figure 21-8 for hexavalent chromium measurements collected at
PRRI include the following:
• The maximum hexavalent chromium concentration was measured on May 9, 2012
(0.207 ng/m3), although similar concentrations were also measured in 2006 and 2007.
No other measurements greater than 0.15 ng/m3 have been measured at PRRI and
only eight concentrations greater than 0.1 ng/m3 have been measured since sampling
began in 2005.
• The 1-year average concentration of hexavalent chromium has fluctuated over the
years of sampling, with the 1-year average at a maximum for 2006 (0.027 ng/m3) and
a minimum for 2009 (0.007 ng/m3).
• For each year shown, the minimum and 5th percentile are zero, indicating the
presence of non-detects. The number of non-detects reported has varied by year, from
as low as 9 percent in 2011 to as high as 65 percent in 2009. This explains why the
median concentration is also zero for 2009.
21-16
-------
Figure 21-9. Yearly Statistical Metrics for Naphthalene Concentrations Measured at PRRI
I200
o
2010
Year
O 5th Percentile
— Maximum O 95th Percentile "-O-" Averagf
1A 1-year average is not presented because sampling under the NMP did not begin until July 2008.
Observations from Figure 21-9 for naphthalene measurements collected 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 is not presented for 2008, although the range
of measurements is provided.
• The maximum naphthalene concentration was measured in 2011 (301 ng/m3). Seven
of the 10 naphthalene concentrations greater than 200 ng/m3 were measured in
November of any given year.
• Although the maximum concentration measured each year varies, the 1-year average
concentration of naphthalene exhibits little variability, ranging from 71.39 ng/m3 for
2010 to 77.73 ng/m3 for 2009. This is also true for the median concentration, which
varies by less than 6 ng/m3 across the years of sampling.
21.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk at the
PRRI monitoring site. Refer to Sections 3.3 and 3.5.3 for definitions and explanations regarding
the various toxicity factors, time frames, and calculations associated with these risk-based
screenings.
21-17
-------
21.5.1 Risk-Based Screening Assessment Using MRLs
A risk-based screening was conducted by comparing the concentration data from the
Rhode Island monitoring site to the ATSDR MRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 day to 14 days); intermediate (exposures of 15 day to 364 days); and chronic
(exposures of 1 year or greater). The preprocessed daily measurements of the pollutants of
interest were compared to the acute MRLs; the quarterly averages were compared to the
intermediate MRLs; and the annual averages were compared to the chronic MRLs.
As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2012.
21.5.2 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 3.5.3.4 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 21-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Table 21-6. Risk Approximations for the Rhode Island Monitoring Site
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections vs.
# of Samples
Annual
Average
(ng/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
Providence, Rhode Island - PRRI
Hexavalent Chromium
Naphthalene
0.012
0.000034
0.0001
0.003
41/61
60/60
0.02
±0.01
76.41
± 12.42
0.19
2.60
<0.01
0.03
21-18
-------
Observations for PRRI from Table 21-6 include the following:
• Both pollutants of interest for PRRI have a cancer URE and a noncancer RfC.
• The cancer risk approximation for naphthalene (2.60 in-a-million) is greater than the
cancer risk approximation for hexavalent chromium (0.19 in-a-million).
• The noncancer hazard approximations for naphthalene and hexavalent chromium are
negligible (0.03 in-a-million and <0.01 in-a-million, respectively), indicating that no
adverse health effects are expected from these individual pollutants.
21.5.3 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 21-7 presents the 10 pollutants with the highest emissions from the 2011 NEI that have
cancer toxicity factors. Table 21-7 also presents the 10 pollutants with the highest toxicity -
weighted emissions, based on the weighting schema described in Section 3.5.3.5. Lastly,
Table 21-7 provides the 10 pollutants with the highest cancer risk approximations (in-a-million)
for PRRI, as presented in Table 21-6. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 21-7. Table 21-8 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 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 3.5.3.5. Similar to the cancer risk and
noncancer hazard approximations provided in Section 21.5.2, this analysis may help policy-
makers prioritize their air monitoring activities.
21-19
-------
Table 21-7. 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)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
Providence, Rhode Island (Providence County) - PRRI
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Tetrachloroethylene
Naphthalene
Trichloroethylene
POM, Group la
Dichloromethane
171.52
140.46
80.45
68.23
26.97
17.76
15.23
9.08
9.00
4.57
Formaldehyde
Benzene
POM, Group 3
1,3 -Butadiene
POM, Group la
Naphthalene
POM, Group 2b
POM, Group 2d
POM, Group 5a
Ethylbenzene
1.83E-03
1.34E-03
1.03E-03
8.09E-04
7.92E-04
5.18E-04
3.78E-04
2.28E-04
2.27E-04
2.01E-04
Naphthalene 2.60
Hexavalent Chromium 0.19
to
o
-------
Table 21-8. 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)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Noncancer Hazard
Approximation
Pollutant (HQ)
Providence, Rhode Island (Providence County) - PRRI
Toluene
Ethylene glycol
Methanol
Xylenes
Hexane
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
Methyl isobutyl ketone
1,262.41
776.88
386.62
338.43
311.14
171.52
140.46
80.45
68.23
41.59
Acrolein
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Benzene
Naphthalene
Trichloroethylene
Xylenes
Arsenic, PM
Ethylene glycol
296,350.76
14,333.09
13,484.18
7,581.54
5,717.41
5,075.86
4,539.58
3,384.29
2,555.13
1,942.20
Naphthalene 0.03
Hexavalent Chromium <0 . 0 1
to
to
-------
Observations from Table 21-7 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 toxi city-weighted emissions (of the
pollutants with cancer UREs), followed by benzene and POM, Group 3.
• Six of the highest emitted pollutants also have the highest toxi city-weighted
emissions for Providence County.
• Naphthalene, which has the highest cancer risk approximation among the pollutants
of interest for PRRI, has the seventh highest emissions and the sixth highest toxi city-
weighted emissions. Conversely, hexavalent chromium appears on neither emissions-
based list, ranking 31st for quantity emitted and 13th for its toxi city-weighted
emissions.
• Several POM Groups appear among the pollutants with the highest toxicity-weighted
emissions for Providence County. POM, Group 2b and 2d rank seventh and eighth for
their toxicity-weighted emissions, respectively. POM, Groups 2b and 2d include
several PAHs sampled for at PRRI, although none of these pollutants failed screens.
• POM, Group 5a ranks ninth 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 21-8 include the following:
• Toluene, ethylene glycol, 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, formaldehyde, and 1,3-butadiene.
• Five 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. Hexavalent chromium does not appear on either
emissions-based list. These are the only two pollutants of interest for PRRI.
21-22
-------
21.6 Summary of the 2012 Monitoring Data for PRRI
Results from several of the data treatments described in this section include the
following:
»«» Three pollutants failed at least one screen for PRRI, with naphthalene accounting for
the majority of failed screens.
*»* Of the site-specific pollutants of the interest, naphthalene had the highest annual
average concentration for PRRI.
»«» The maximum concentration ofhexavalent chromium since the onset of sampling at
PRRI was measured in 2012.
21-23
-------
22.0 Site in South Carolina
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS site in South Carolina, 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 South Carolina 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.
CHSC is located in central Chesterfield County, South Carolina. 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. 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 radius are still visible on the map, but have been grayed out in order to show emissions
sources just outside the boundary. Table 22-1 provides supplemental geographical information
such as land use, location setting, and locational coordinates.
22-1
-------
Figure 22-1. Chesterfield, South Carolina (CHSC) Monitoring Site
to
to
to
-------
Figure 22-2. NEI Point Sources Located Within 10 Miles of CHSC
Legend
80°20'0"W 80"15'0"W 80"10'(TW 80'5'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
CHSC NATTS site O 10 mile radius | [ County boundary
Source Category Group (No. of Facilities)
t Airport/Airline/Airport Support Operations (1)
•*• Industrial Machinery or Equipment Plant (1)
22-3
-------
Table 22-1. Geographical Information for the South Carolina Monitoring Site
Site
Code
CHSC
AQS Code
45-025-0001
Location
Not in a
city
County
Chesterfield
Micro- or
Metropolitan
Statistical Area
Not in an MSA
Latitude
and
Longitude
34.615367,
-80.198787
Land Use
Forest
Location
Setting
Rural
Additional Ambient Monitoring Information1
VOCs, Carbonyl Compounds, Hexachlorobutadiene,
O3, Meteorological parameters, PM10, PM10
Speciation, PM2 5, and PM2 5 Speciation, Black
Carbon, IMPROVE Speciation.
BOLD ITALICS = EPA-designated NATTS Site
to
to
-------
CHSC is located about 14 miles south of the North Carolina/South Carolina border,
between the towns of McBee and Chesterfield. The monitoring site is located near the Ruby fire
tower and, as Figure 22-1 shows, is located just off State Highway 145. The surrounding area is
rural in nature and is part of the Carolina Sandhills Wildlife Refuge. Figure 22-2 shows that few
point sources are located within 10 miles of CHSC, the closest of which is the Wild Irish Rose
Airport.
Table 22-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the South Carolina monitoring site. Table 22-2 includes both county-
level population and vehicle registration information. Table 22-2 also contains traffic volume
information for CHSC as well as the location for which the traffic volume was obtained.
Additionally, Table 22-2 presents the daily VMT for Chesterfield County.
Table 22-2. Population, Motor Vehicle, and Traffic Information for the South Carolina
Monitoring Site
Site
CHSC
Estimated
County
Population1
46,103
County-level
Vehicle
Registration2
41,259
Annual
Average Daily
Traffic3
550
Intersection
Used for
Traffic Data
Hwy 145 between US-1 and 109
County-
level Daily
VMT4
1,228,145
Bounty-level population estimate reflects 2012 data (Census Bureau, 2013c)
2County-level vehicle registration reflects 2012 data (SC DMV, 2012)
3AADT reflects 2012 data (SC DOT, 2012)
4County-level VMT reflects 2012 data (SC DOT, 2013)
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 22-2 include the following:
• Chesterfield County's population is among the lowest compared to other counties
with NMP sites. A similar ranking was found for the county-level vehicle ownership.
• The traffic volume experienced near CHSC is the second lowest compared to other
NMP monitoring sites. The traffic estimate provided is for State Highway 145
between State Highway 109 and US-1.
• The daily VMT for Chesterfield County is the third lowest VMT compared to other
counties with NMP sites (where VMT data were available).
22.2 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
site in South Carolina on sample days, as well as over the course of the year.
22-5
-------
22.2.1 Climate Summary
The town of Chesterfield is located just south of the North Carolina/South Carolina
border, about 35 miles northwest of the city of Florence. Although the area experiences all four
seasons, South Carolina's southeastern location ensures mild winters and long, hot summers.
Summers are dominated by the Bermuda high pressure system over the Atlantic Ocean, which
allows southwesterly winds to prevail, bringing in warm, moist air out of the Gulf of Mexico.
During winter, winds out of the southwest shift northeasterly after frontal systems move across
the area. The mountains to the northwest help shield the area from cold air outbreaks. Greater
than 2 inches of precipitation can be expected any given month, with the maximum typically
occurring in July (greater than 5 inches). Chesterfield County leads the state in the average
number of sleet and freezing rain events per year (Bair, 1992; SC SCO, 2014).
22.2.2 Meteorological Summary
Hourly meteorological data for 2012 were retrieved from NCDC for the weather station
closest to the South Carolina monitoring site (NCDC, 2012), as described in Section 3.5.2. The
closest weather station with adequate data is located at the Monroe Airport in Monroe, North
Carolina (WBAN 53872). Additional information about the Monroe Airport weather station,
such as the distance between the site and the weather station, is provided in Table 22-3. These
data were used to determine how meteorological conditions on sample days vary from conditions
experienced throughout the year.
Table 22-3 presents average temperature (average maximum and average daily),
moisture (average dew point temperature, average wet bulb temperature, and average relative
humidity), pressure (average sea level pressure), and wind (average scalar wind speed)
information for days samples were collected and for all of 2012. Also included in Table 22-3 is
the 95 percent confidence interval for each parameter. As shown in Table 22-3, average
meteorological conditions experienced on sample days were representative of average weather
conditions experienced throughout the year near CHSC.
22-6
-------
Table 22-3. Average Meteorological Conditions near the South Carolina Monitoring Site
Closest
Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Chesterfield, South Carolina - CHSC
Monroe Airport
53872
(35.02, -80.62)
35 8
miles
311°
(NW)
Sample
Days
(62)
2012
72.0
±3.6
72.4
+ 1.5
61.8
±3.5
62.1
+ 1.4
50.8
±4.0
51.4
+ 1.6
55.9
±3.4
56.4
+ 1.4
70.5
±3.2
71.3
+ 1.4
1018.9
±1.8
1018.2
+ 0.6
4.4
±0.6
4.5
+ 0.3
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
to
to
-------
22.2.3 Back Trajectory Analysis
Figure 22-3 is the composite back trajectory map for days on which samples were
collected at the CHSC monitoring site. Included in Figure 22-3 are four back trajectories per
sample day. Figure 22-4 is the corresponding cluster analysis. An in-depth description of these
maps and how they were generated is presented in Section 3.5.2.1. For the composite map, each
line represents the 24-hour trajectory along which a parcel of air traveled toward the monitoring
site on a given sample day and time, based on an initial height of 50 meters AGL. For the cluster
analysis, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric circle around the site in Figures 22-3 and 22-4 represents 100 miles.
Observations from Figures 22-3 and 22-4 for CHSC include the following:
• Back trajectories originated from a variety of directions at CHSC, with the longest
back trajectories originating to the northwest of the site.
• The 24-hour air shed domain for CHSC was among the smaller in size compared to
other NMP monitoring sites. Several of the longest back trajectories originated over
Indiana, or greater than 500 miles away. However, the average back trajectory length
was 183 miles and 87 percent of back trajectories originated within 300 miles of the
site.
• The cluster analysis shows that 13 percent of back trajectories originated from the
northwest of CHSC, over the Ohio Valley region. Nineteen percent of back
trajectories originated from the west and southwest of CHSC, primarily over
northwest South Carolina and Georgia. Another 24 percent of back trajectories
originated to the east, southeast, and south of CHSC, along the coasts and adjacent
waters of South Carolina, Georgia, and Florida. Twenty-one percent of back
trajectories originated to the north of the site, over the Mid-Atlantic states. Another
23 percent of back trajectories are represented by the cluster trajectory that is covered
up by the star symbol; thus, the cluster trajectory is presented in the inset map in
Figure 22-4. This short trajectory represents back trajectories originating from
varying directions but generally less than 100 miles from CHSC (plus a few
originating near the western and northern border of North Carolina).
22-8
-------
Figure 22-3. Composite Back Trajectory Map for CHSC
100 200
Figure 22-4. Back Trajectory Cluster Map for CHSC
22-9
-------
22.2.4 Wind Rose Comparison
Hourly surface wind data from the weather station at Monroe Airport near CHSC were
uploaded into a wind rose software program to produce customized wind roses, as described in
Section 3.5.2.2. 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 22-5 presents a map showing the distance between the weather station and CHSC,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 22-5 also presents three different wind roses for the
CHSC monitoring site. First, a historical wind rose representing 2002 to 2011 wind data is
presented, which shows the predominant surface wind speed and direction over an extended
period of time. Second, a wind rose representing wind observations for all of 2012 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2012 is
presented. These can be used to identify the predominant wind speed and direction for 2012 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.
Observations from Figure 22-5 for CHSC include the following:
• The Monroe Airport weather station is located across the North Carolina/South
Carolina border, approximately 36 miles northwest of CHSC.
• The historical wind rose for CHSC shows that calm winds (< 2 knots) account for
23 percent of the hourly measurements. Winds from the south-southwest to west
account for approximately one-third of observations, just slightly more than winds
from the north to east-northeast. Winds from the southeast quadrant are generally not
observed.
• The wind patterns shown on the 2012 wind rose for CHSC are similar to the historical
wind patterns, although there were slightly more calm observations and fewer winds
observations from the northeast quadrant. This indicates that wind conditions in 2012
were similar to what is expected climatologically near this site.
• The sample day wind patterns for 2012 also resemble the historical and full-year wind
patterns. However, the calm rate for sample days is approaching 30 percent and the
number of northerly observations is reduced to 3 percent.
22-10
-------
Figure 22-5. Wind Roses for the Monroe Airport Weather Station near CHSC
Location of CHSC and Weather Station
2002-2011 Historical Wind Rose
2012 Wind Rose
Sample Day Wind Rose
Calms: 2828%
Calms: 2933%
22-11
-------
22.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for CHSC 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 22-4. 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 22-4. It is important to note which
pollutants were sampled for at each site when reviewing the results of this analysis; CHSC
sampled hexavalent chromium and PAHs.
Table 22-4. Risk-Based Screening Results for the South Carolina Monitoring Site
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Chesterfield, South Carolina - CHSC
Naphthalene
0.029
Total
7
7
53
53
13.21
13.21
100.00
100.00
Observations from Table 22-4 include the following:
• Naphthalene was the only pollutant to fail screens for CHSC. This pollutant was
detected in all 53 valid samples collected at CHSC and failed seven screens, or
approximately 13 percent of screens.
• This site has the fourth lowest number of failed screens (7) among all NMP sites.
22.4 Concentrations
This section presents various concentration averages used to characterize pollution levels
at the South Carolina 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 the
site.
• Annual concentration averages are presented graphically 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 the site.
22-12
-------
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Additional site-specific statistical summaries for CHSC
are provided in Appendices M and O.
22.4.1 2012 Concentration Averages
Quarterly and annual concentration averages were calculated for the pollutants of interest
for the South Carolina site, as described in Section 3.1. The quarterly average 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 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
CHSC are presented in Table 22-5, 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 22-5. Quarterly and Annual Average Concentrations of the Pollutants of
Interest for the South Carolina Monitoring Site
Pollutant
#of
Measured
Detections
vs. # 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)
Chesterfield, South Carolina - CHSC
Naphthalene
53/53
24.41
±7.55
11.31
±1.88
9.50
±1.52
22.11
±7.63
17.26
±3.25
Observations for CHSC from Table 22-5 include the following:
• Naphthalene concentrations measured at CHSC span an order of magnitude, ranging
from 5.61 ng/m3 to 58.3 ng/m3, with a median concentration of 12.8 ng/m3.
• The annual average concentration of naphthalene is 17.26 ± 3.25 ng/m3. This is the
third lowest annual average concentration of naphthalene among NMP sites sampling
PAHs.
22-13
-------
• The first and fourth quarter average concentrations of naphthalene are significantly
higher than the other quarterly averages and have relatively large confidence intervals
associated with them. The maximum naphthalene concentration was measured on
November 11, 2012 (58.5 ng/m3) although a similar measurement (53.1 ng/m3) was
also measured in March. The 17 highest concentrations of naphthalene were
measured at CHSC between January and March or October and December 2012.
22.4.2 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, a box plot was created for the pollutants shaded in
gray in Table 22-4 for CHSC. Figure 22-6 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 of naphthalene, as described in
Section 3.5.3.1.
Figure 22-6. Program vs. Site-Specific Average Naphthalene Concentration
D
1
3 100 200 300 400 500 600 700 800 9C
Concentration (ng/m3)
Program:
Site:
1st Quartile
D
Site Average
o
2nd Quartile 3rd Quartile
D D
Site Concentration Range
^^^^—
4th Quartile Average
D 1
Observations from Figure 22-6 include the following:
• Figure 22-6 is the box plot for naphthalene. The annual average concentration of
naphthalene for CHSC is less than the program-level first quartile. The maximum
naphthalene concentration measured at CHSC is less than the program-level
average concentration as well as the program-level median concentration. There
were no non-detects of naphthalene measured at CHSC or across the program.
The range of naphthalene measurements collected at CHSC is among the smallest
measured at an NMP site sampling PAHs.
22-14
-------
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 3.5.3.2.
CHSC has sampled PAHs under the NMP since 2008. Thus, Figure 22-7 presents the 1-year
statistical metrics for the pollutant of interest for CHSC. 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 is not provided, although the range and quartiles are still presented.
Figure 22-7. Yearly Statistical Metrics for Naphthalene Concentrations Measured at CHSC
I200
2010
Year
5th Percentile - Minimurr
— Maximum
95th Percent!le
A 1-year average is not presented because sampling under the NMP did not begin until March 2008.
Observations from Figure 22-7 for naphthalene measurements collected at CHSC include
the following:
• CHSC began sampling PAHs under the NMP in March 2008. Because a full year's
worth of data is not available, a 1-year average is not presented for 2008, although the
range of measurements is provided.
• The maximum concentration of naphthalene was measured on May 1, 2009
(323 ng/m3). This is the only concentration of naphthalene greater than 200 ng/m3
measured at CHSC since the onset of PAH sampling. Only two measurements greater
22-15
-------
than 100 ng/m3 have been measured (one each in 2010 and 2011) and no other
concentrations greater than 60 ng/m3 have been measured at this site.
• The majority of naphthalene concentrations measured at CHSC fall within a relatively
small range, as indicated by the 5th and 95th percentiles.
• The 1-year average concentration of naphthalene has changed relatively little of the
years of sampling. The 1-year average has ranged from 16.42 ng/m3 (2011) to
21.71 ng/m3 (2009).
22.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk at the
CHSC monitoring site. Refer to Sections 3.3 and 3.5.3 for definitions and explanations regarding
the various toxicity factors, time frames, and calculations associated with these risk-based
screenings.
22.5.1 Risk-Based Screening Assessment Using MRLs
A risk-based screening was conducted by comparing the concentration data from the
South Carolina monitoring site to the ATSDR MRLs, where available. As described in
Section 3.3, MRLs are noncancer health risk benchmarks and are defined for three exposure
periods: acute (exposures of 1 day to 14 days); intermediate (exposures of 15 days to 364 days);
and chronic (exposures of 1 year or greater). The preprocessed daily measurements of the
pollutants of interest were compared to the acute MRLs; the quarterly averages were compared
to the intermediate MRLs; and the annual averages were compared to the chronic MRLs.
As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2012.
22.5.2 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the South Carolina 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 3.5.3.4 for an explanation of how cancer risk and
22-16
-------
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 22-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Table 22-6. Risk Approximations for the South Carolina Monitoring Site
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections vs.
# of Samples
Annual
Average
(ng/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
Chesterfield, South Carolina - CHSC
Naphthalene
0.000034
0.003
53/53
17.26
±3.25
0.59
0.01
Observations for CHSC from Table 22-6 include the following:
• Naphthalene has both a cancer URE and a noncancer RfC.
• The cancer risk approximation for naphthalene is less than 1 in-a-million
(0.59 in-a-million).
• The noncancer hazard approximation for naphthalene is very low (0.01), indicating
that no adverse health effects are expected from this individual pollutant.
22.5.3 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 22-7 presents the 10 pollutants with the highest emissions from the 2011 NEI that have
cancer toxicity factors. Table 22-7 also presents the 10 pollutants with the highest toxicity -
weighted emissions, based on the weighting schema described in Section 3.5.3.5. Lastly,
Table 22-7 provides the 10 pollutants with the highest cancer risk approximations (in-a-million)
for the site, as presented in Table 22-6. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 22-7. Table 22-8 presents similar
information, but is limited to those pollutants with noncancer toxicity factors.
22-17
-------
Table 22-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the South Carolina Monitoring Site
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
Chesterfield, South Carolina (Chesterfield County) - CHSC
Benzene
Formaldehyde
Acetaldehyde
Ethylbenzene
1,3 -Butadiene
Naphthalene
Trichloroethylene
POM, Group 2b
POM, Group la
POM, Group 2d
22.71
21.01
11.34
10.96
3.32
1.20
0.30
0.30
0.27
0.26
Formaldehyde
Benzene
1,3 -Butadiene
Naphthalene
Arsenic, PM
Ethylbenzene
POM, Group 2b
Acetaldehyde
POM, Group la
POM, Group 5a
2.73E-04
1.77E-04
9.95E-05
4.07E-05
3.20E-05
2.74E-05
2.64E-05
2.49E-05
2.38E-05
2.36E-05
Naphthalene 0.59
to
to
oo
-------
Table 22-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the South Carolina Monitoring Site
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Noncancer Hazard
Approximation
Pollutant (HQ)
Chesterfield, South Carolina (Chesterfield County) - CHSC
Toluene
Ethylene glycol
Xylenes
Hexane
Methanol
Benzene
Formaldehyde
Acetaldehyde
Ethylbenzene
Methyl isobutyl ketone
135.33
64.96
48.09
42.54
30.61
22.71
21.01
11.34
10.96
4.44
Acrolein
Formaldehyde
Cyanide Compounds, gas
1,3 -Butadiene
Acetaldehyde
Benzene
Lead, PM
Arsenic, PM
Xylenes
Cadmium, PM
30,334.22
2,143.93
2,002.83
1,658.43
1,259.78
756.90
568.43
495.51
480.89
419.93
Naphthalene 0.01
to
to
-------
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 3.5.3.5. Similar to the cancer risk and
noncancer hazard approximations provided in Section 22.5.2, this analysis may help policy-
makers prioritize their air monitoring activities.
Observations from Table 22-7 include the following:
• Benzene, formaldehyde, and acetaldehyde are the highest emitted pollutants with
cancer UREs in Chesterfield County.
• Formaldehyde, benzene, and 1,3-butadiene are the pollutants with the highest
toxicity-weighted emissions (of the pollutants with cancer UREs) for Chesterfield
County.
• Eight of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Chesterfield County.
• Naphthalene, the only pollutant of interest for CHSC, appears on both emissions-
based lists, with the sixth highest emissions and the fourth highest toxicity-weighted
emissions for Chesterfield County.
• Several POM Groups appear among the pollutants with the highest emissions and
toxicity-weighted emissions. POM, Group 2b appears on both emissions-based lists
and includes several PAHs sampled for at CHSC including acenaphthylene,
fluoranthene, and perylene. POM, Group 2d, which includes phenanthrene and
pyrene, ranks tenth for quantity emitted but is not among those with the highest
toxicity-weighted emissions. POM, Group 5a, which includes benzo(a)pyrene, ranks
tenth for toxicity weighted emissions but is not among the highest emitted. None of
the pollutants sampled for at CHSC and included in POM, Groups 2b, 2d, or 5a failed
screens for CHSC. POM, Group la, which appears on both emissions-based lists,
does not include any PAHs sampled for at CHSC.
Observations from Table 22-8 include the following:
• Toluene, ethylene glycol, and xylenes are the highest emitted pollutants with
noncancer RfCs in Chesterfield County.
• The pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) are acrolein, formaldehyde, and cyanide compounds (gaseous).
22-20
-------
• Four of the highest emitted pollutants in Chesterfield County also have the highest
toxicity-weighted emissions.
• Naphthalene does not appear on either emissions-based list in Table 22-8, ranking
16th for quantity emitted and 12th for its toxicity-weighted emissions.
22.6 Summary of the 2012 Monitoring Data for CHSC
Results from several of the data treatments described in this section include the
following:
»«» Naphthalene was the only pollutant to fail screens for CHSC. This site has the fourth
lowest number of failed screens (7) among allNMP sites.
»«» Concentrations of naphthalene measured during the colder months of the year were
greater than those measured during the warmer months of the year.
»«» Concentrations of naphthalene have changed little since the onset of PAH sampling
at CHSC.
22-21
-------
23.0 Site in South Dakota
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the UATMP site in South Dakota, 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 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 South Dakota monitoring site is located in Sioux Falls, South Dakota (SSSD).
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. 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 radius are still visible on the map, but have been grayed out in order to show
emissions sources just outside the boundary. Table 23-1 provides supplemental geographical
information such as land use, location setting, and locational coordinates.
23-1
-------
Figure 23-1. Sioux Falls, South Dakota (SSSD) Monitoring Site
to
OJ
to
-------
Figure 23-2. NEI Point Sources Located Within 10 Miles of SSSD
97'0'FW 60'55'0-W
Legend
6MCFO-W ee-35'O'W 96'3CTO-VV
Note: Due to facility density and collocation the total facilities
displayed may not represent all facilities within the area of interest
SSSD UATMP site 10 mile radius | | County boundary
Source Category Group (No. of Facilities)
•f AirportfAirline/Airport Support Operations (7)
. Compressor Station (2)
Food Processing/Agriculture (1)
Industrial Machinery or Equipment Plant (2)
Institution (school, hospital, prison, etc.) (1)
<•> Metals Processing/Fabrication (1)
R Plastic, Resin, or Rubber Products Plant (1)
' Wastewater Treatment (1)
W Woodwork. Furniture, Millwork 8 Wood Preserving (2)
F
•#•
O
23-3
-------
Table 23-1. Geographical Information for the South Dakota Monitoring Site
Site
Code
SSSD
AQS Code
46-099-0008
Location
Sioux Falls
County
Minnehaha
Micro- or
Metropolitan
Statistical
Area
Sioux Falls, SD
MSA
Latitude and
Longitude
43.54792,
-96.700769
Land Use
Commercial
Location
Setting
Urban/City
Center
Additional Ambient Monitoring Information1
CO, SO2, NO, NO2, NOX, NOy, O3, Meteorological
parameters, PM10, PM coarse, PM25, PM25
Speciation, IMPROVE Speciation.
Data for additional pollutants are reported to AQS for SSSD (EPA, 2013b); however, these data are not generated by ERG and are therefore not included in this report.
to
-k
-------
SSSD is located on the east side of Sioux Falls, in southeast South Dakota. The
monitoring site is located at the South Dakota School for the Deaf. The surrounding area is
mixed usage, with both commercial and residential areas surrounding the site. SSSD is less than
one-half mile from the intersection of Highway 42 (East 10th Street) and 1-229, as shown in
Figure 23-1. As Figure 23-2 shows, relatively few emissions sources are located within 10 miles
of SSSD. The source category with the greatest number of point sources shown in Figure 23-2 is
the airport and airport support operations 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 sources closest to SSSD are a hospital heliport and a food
processing facility.
Table 23-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the South Dakota monitoring site. Table 23-2 includes both county-
level population and vehicle registration information. Table 23-2 also contains traffic volume
information for SSSD site as well as the location for which the traffic volume was obtained.
Additionally, Table 23-2 presents the county-level daily VMT for Minnehaha County.
Table 23-2. Population, Motor Vehicle, and Traffic Information for the South Dakota
Monitoring Site
Site
SSSD
Estimated
County
Population1
175,037
County-level
Vehicle
Registration2
212,507
Annual
Average Daily
Traffic3
18,575
Intersection
Used for
Traffic Data
E 10th St, east of N. Mable Ave
County-
level Daily
VMT4
3,778,321
Bounty-level population estimate reflects 2012 data (Census Bureau, 2013c)
2County-level vehicle registration reflects 2012 data (SD DOR, 2012)
3AADT reflects 2012 data (SD DOT, 2012)
4County-level VMT reflects 2012 data (SD DOT, 2013)
Observations from Table 23-2 include the following:
• The county-level population for SSSD ranks in the bottom third compared to other
counties with NMP sites. The county-level vehicle registration for SSSD is similarly
ranked compared to other counties with NMP sites.
• The traffic volume for SSSD is in the middle of the range compared to other NMP
sites. Traffic data for SSSD are provided for East 10th Street, east of N. Mable
Avenue.
• The daily VMT for Minnehaha County is in the bottom third among counties with
NMP sites (where VMT was available).
23-5
-------
23.2 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
site in South Dakota on sample days, as well as over the course of the year.
23.2.1 Climate Summary
The Sioux Falls area has a continental climate, with cold winters, warm summers, and
often drastic day-to-day variations. Precipitation varies throughout the year, with the spring and
summer seasons receiving more than half of the annual rainfall, primarily in the form of
thunderstorms. On average, a south wind blows in the summer and fall and a northwest wind
blows in the winter and early spring. Flooding is often a concern in the area during springtime
when snow begins to melt, although a flood control system, including levees and a diversion
channel, was constructed to reduce the flood threat within the city limits. (Wood, 2004).
23.2.2 Meteorological Summary
Hourly meteorological data for 2012 were retrieved from NCDC for the weather station
closest to the South Dakota monitoring site (NCDC, 2012), as described in Section 3.5.2. The
closest weather station is located at Joe Foss Field Airport near SSSD, WBAN 14944. Additional
information about this weather station, such as the distance between the site and the weather
station, is provided in Table 23-3. These data were used to determine how meteorological
conditions on sample days vary from conditions experienced throughout the year.
Table 23-3 presents average temperature (average maximum and average daily),
moisture (average dew point temperature, average wet bulb temperature, and average relative
humidity), pressure (average sea level pressure), and wind (average scalar wind speed)
information for days samples were collected and for the entire year. Also included in Table 23-3
is the 95 percent confidence interval for each parameter. As shown in Table 23-3, average
meteorological conditions on sample days near SSSD were representative of average weather
conditions experienced throughout the year.
23-6
-------
Table 23-3. Average Meteorological Conditions near the South Dakota Monitoring Site
Closest
Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
| Sioux Falls, South Dakota - SSSD
Joe Foss Field
Airport
14944
(43.58, -96.75)
3 2
310°
(JN V\ )
Sample
Days
(62)
2012
61.3
±5.7
62.1
+ 2.4
50.3
±5.2
50.9
+ 2.2
37.4
±4.4
37.4
+ 1.9
43.8
±4.4
44.1
+ 1.9
65.8
±3.5
64.5
+ 1.4
1014.8
±1.8
1014.8
+ 0.8
8.6
±1.1
8.3
+ 0.4
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
to
-------
23.2.3 Back Trajectory Analysis
Figure 23-3 is the composite back trajectory map for days on which samples were
collected at the SSSD monitoring site. Included in Figure 23-3 are four back trajectories per
sample day. Figure 23-4 is the corresponding cluster analysis. An in-depth description of these
maps and how they were generated is presented in Section 3.5.2.1. For the composite map, each
line represents the 24-hour trajectory along which a parcel of air traveled toward the monitoring
site on a given sample day and time, based on an initial height of 50 meters AGL. For the cluster
analysis, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric circle around the site in Figures 23-3 and 23-4 represents 100 miles.
Observations from Figures 23-3 and 23-4 for SSSD include the following:
• Back trajectories originated from a variety of directions at the SSSD site. The longest
back trajectories originated from the northwest.
• The 24-hour air shed domain for SSSD is among the largest air sheds compared to
other NMP monitoring sites. The farthest away a back trajectory originated was over
British Columbia, Canada or greater than 950 miles away, although additional back
trajectories also originated over Alberta and Saskatchewan. SSSD is the only site with
back trajectories greater than 900 miles in length, although the average back
trajectory length was nearly 286 miles.
• The cluster analysis shows that back trajectories originating from the north, northwest
quadrant, and west account for more than 40 percent of back trajectories, but are split
into three clusters based on length. Another 17 percent of back trajectories originated
to the south of SSSD, over Nebraska, Kansas, and Oklahoma. Ten percent of back
trajectories originated to the southeast of SSSD, primarily over Iowa and Missouri.
The shorter cluster trajectory (32 percent) originating to the north of SSSD represents
shorter trajectories (< 200 miles in length) originating from a variety of directions,
although primarily along and east of the South Dakota border.
23-8
-------
Figure 23-3. Composite Back Trajectory Map for SSSD
Figure 23-4. Back Trajectory Cluster Map for SSSD
•
23-9
-------
23.2.4 Wind Rose Comparison
Hourly surface wind data from the weather station at Joe Foss Field Airport were
uploaded into a wind rose software program to produce customized wind roses, as described in
Section 3.5.2.2. 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 23-5 presents a map showing the distance between the weather station and SSSD,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 23-5 also presents three different wind roses for the
SSSD monitoring site. First, a historical wind rose representing 2002 to 2011 wind data is
presented, which shows the predominant surface wind speed and direction over an extended
period of time. Second, a wind rose representing wind observations for all of 2012 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2012 is
presented. These can be used to identify the predominant wind speed and direction for 2012 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.
Observations from Figure 23-5 for SSSD include the following:
• The Joe Foss Field Airport weather station is located approximately 3.2 miles
northwest of SSSD.
• The historical wind rose shows that winds from a variety of directions were observed
near SSSD, although winds from the south were observed the most (13 percent), and
southwesterly and west-southwesterly winds observed the least (less than 2 percent).
Calm winds were observed for approximately 12 percent of the observations. The
strongest winds tend to be from the south or the northwest quadrant.
• The 2012 wind patterns are very similar to the historical wind patterns, although calm
winds account for greater than 14 percent of the observations. This indicates that
wind conditions in 2012 near SSSD are similar to historical wind conditions.
• The sample day wind rose also resembles the full-year wind rose, but does exhibit
some differences. Southerly winds were still prominent, but there is a higher
percentage of wind observations from the north, northwest, west, and south-
southwest. There were also fewer observations from the southeast quadrant.
23-10
-------
Figure 23-5. Wind Roses for the Joe Foss Field Airport Weather Station near SSSD
Location of SSSD and Weather Station
2002-2011 Historical Wind Rose
- ..IF
"Z?J
" '!l-Tl'il A/+
2012 Wind Rose
Sample Day Wind Rose
23-11
-------
23.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for the South
Dakota 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 23-4.
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 23-4. It is
important to note which pollutants were sampled for at each site when reviewing the results of
this analysis. SSSD sampled for VOCs, SNMOCs, and carbonyl compounds.
Table 23-4. Risk-Based Screening Results for the South Dakota Monitoring Site
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Sioux Falls, South Dakota - SSSD
Benzene
Carbon Tetrachloride
Acetaldehyde
Formaldehyde
1,3 -Butadiene
1 ,2-Dichloroethane
Ethylbenzene
Hexachloro- 1 ,3 -butadiene
1 , 1 ,2,2-Tetrachloroethane
£>-Dichlorobenzene
Propionaldehyde
0.13
0.17
0.45
0.077
0.03
0.038
0.4
0.045
0.017
0.091
0.8
Total
61
60
58
58
54
53
4
4
4
2
1
359
61
61
58
58
58
53
61
4
4
25
58
501
100.00
98.36
100.00
100.00
93.10
100.00
6.56
100.00
100.00
8.00
1.72
71.66
16.99
16.71
16.16
16.16
15.04
14.76
1.11
1.11
1.11
0.56
0.28
16.99
33.70
49.86
66.02
81.06
95.82
96.94
98.05
99.16
99.72
100.00
Observations from Table 23-4 include the following:
• Eleven pollutants failed at least one screen for SSSD; nearly 72 percent of
concentrations for these 11 pollutants were greater than their associated risk screening
value (or failed screens).
• Many of the pollutants listed in Table 23-4 failed 100 percent of screens. However,
the detection rate of these pollutants varied. For example, benzene was detected in all
61 sampled collected at SSSD and failed all screens. Hexachloro-1,3-butadiene also
failed 100 percent of screens but was detected only four times.
23-12
-------
• Six pollutants contributed to 95 percent of failed screens for SSSD and therefore were
identified as pollutants of interest. These six include two carbonyl compounds and
four VOCs.
• 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 SSSD sampled both VOCs (TO-15) and SNMOCs, the TO-15
results were used for the 12 pollutants these methods have in common.
23.4 Concentrations
This section presents various concentration averages used to characterize pollution levels
at the South Dakota 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 the
site.
• Annual concentration averages are presented graphically 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 the site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Additional site-specific statistical summaries for SSSD are
provided in Appendices J through L.
23.4.1 2012 Concentration Averages
Quarterly and annual concentration averages were calculated for the pollutants of interest
for the South Dakota site, as described in Section 3.1. The quarterly average 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 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
South Dakota monitoring site are presented in Table 23-5, where applicable. Note that if a
23-13
-------
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 23-5. Quarterly and Annual Average Concentrations of the Pollutants of
Interest for the South Dakota Monitoring Site
Pollutant
#of
Measured
Detections
vs. # 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)
Sioux Falls, South Dakota - SSSD
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
Formaldehyde
58/58
61/61
58/61
61/61
53/61
58/58
1.37
±0.32
0.72
±0.12
0.06
±0.02
0.58
±0.08
0.08
±0.01
1.60
±0.27
1.39
±0.24
0.52
±0.07
0.04
±0.01
0.69
±0.06
0.07
±0.02
1.85
±0.43
2.45
±0.70
0.68
±0.26
0.12
±0.04
0.67
±0.04
0.05
±0.02
2.74
±0.41
1.57
±0.30
0.60
±0.09
0.11
±0.04
0.62
±0.09
0.06
±0.02
1.30
±0.19
1.69
±0.23
0.63
±0.08
0.08
±0.02
0.64
±0.03
0.07
±0.01
1.86
±0.21
Observations for SSSD from Table 23-5 include the following:
• The pollutants with the highest annual average concentrations are formaldehyde
(1.86 ± 0.21 |ig/m3) and acetaldehyde (1.69 ± 0.23 |ig/m3). These are the only two
pollutants of interest with an annual average greater than 1.0 |ig/m3.
• The third quarter average formaldehyde concentration is significantly higher than the
other quarterly averages. A review of the data shows that formaldehyde
concentrations measured at SSSD range from 0.68 |ig/m3to 3.73 |ig/m3. The
maximum concentration of formaldehyde was measured on September 30, 2012. Of
the eight concentrations greater than 3 |ig/m3 measured at SSSD, six were measured
during the third quarter, the maximum in September and the other five on each of the
sample days in July (the other two were measured in June). Conversely, none of the
seven concentrations of formaldehyde less than 1 |ig/m3 were measured during the
third quarter of 2012 (two were measured during the first quarter, two were measured
during the second, and three were measured during the fourth).
• The third quarterly average concentration of acetaldehyde for SSSD is also the
highest quarterly average and has a relatively large confidence interval associated
with it. A review of the data shows that acetaldehyde concentrations measured at
SSSD range from 0.68 |ig/m3to 6.73 |ig/m3. The maximum acetaldehyde
concentration was measured on the same day at SSSD as the maximum formaldehyde
concentration (September 30, 2012). The second highest acetaldehyde concentration
23-14
-------
measured at SSSD is roughly half as high (3.25 |ig/m3) and was measured in July.
There is more variability in the dates of the higher acetaldehyde concentrations than
there is with formaldehyde. But, similar to formaldehyde, none of 12 concentrations
of acetaldehyde less than 1 |ig/m3 were measured during the third quarter of 2012
(five were measured during the first quarter, four were measured during the second,
and three were measured during the fourth).
• The maximum concentration of benzene was also measured on September 30, 2012.
The maximum benzene concentration is 2.49 |ig/m3 and is the only concentration
greater than 2 |ig/m3 measured at SSSD. The next highest concentration measured
during the third quarter is considerably less (0.95 |ig/m3). This explains the relatively
large confidence interval shown for the third quarter. Only two benzene
concentrations greater than 1 |ig/m3 were measured at SSSD and were both measured
during the first quarter of 2012.
• The third and fourth quarter average concentrations of 1,3-butadiene are greater than
the first and second quarter averages. A review of the data shows that 1,3-butadiene
concentrations range from 0.024 |ig/m3 to 0.306 |ig/m3, with three non-detects also
measured. The maximum 1,3-butadiene concentration measured at SSSD was also
measured on September 30th. All six concentrations of 1,3-butadiene greater than
0.2 |ig/m3 were measured during September and October. Further, all but two of the
14 concentrations greater than 0.1 jig/m3 were measured at SSSD between August
and October 2012.
23.4.2 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 shaded in
gray in Table 23-4 for SSSD. Figures 23-6 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, as described in Section 3.5.3.1.
Figure 23-6. Program vs. Site-Specific Average Acetaldehyde Concentration
9 12
Concentration {[og/m3)
Program: 1st Quartile 2nd Quartile 3rd Quartile 4th Quartile Average
Site: Site Average Site Concentration Range
o —
23-15
-------
Figure 23-7. Program vs. Site-Specific Average Benzene Concentration
SSSD
Concentration {[og/m3]
Program:
Site:
IstQuartile
Site Average
o
2ndQuartile SrdQuartile
Site Concentration Range
4thQuartile Average
Figure 23-8. Program vs. Site-Specific Average 1,3-Butadiene Concentration
I Program Max Concentration = 4.10 ug/m3
0.75 1 1.25
Concentration {[og/m3)
Program:
Site:
IstQuartile
D
Site Average
0
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
Figure 23-9. Program vs. Site-Specific Average Carbon Tetrachloride Concentration
SSSD
2 3
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o —
23-16
-------
Figure 23-10. Program vs. Site-Specific Average 1,2-Dichloroethane Concentration
SSSD
Program Max Concentration = 17.01 ug/m3
0.4 0.5 0.6
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site:
Site Average Site Concentration Range
o
Figure 23-11. Program vs. Site-Specific Average Formaldehyde Concentration
10
12
14
Concentration {[og/m3]
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o —
Observations from Figures 23-6 through 23-11 include the following:
• Figure 23-6 shows that the annual average acetaldehyde concentration for SSSD
is similar to the program-level average concentration. The maximum
concentration measured at SSSD is considerably less than the maximum
concentration measured across the program.
• Figure 23-7 shows that the annual average benzene concentration for SSSD is less
than the program-level average and median concentrations. The annual average
concentration of benzene for SSSD is among the lowest annual average
concentrations among NMP sites sampling benzene. The maximum benzene
concentration measured at SSSD is less than the maximum concentration
measured across the program.
• The program-level maximum 1,3-butadiene concentration (4.10 |ig/m3) is not
shown directly on the box plot in Figure 23-8 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 2 |ig/m3. This figure
shows that the annual average 1,3-butadiene for SSSD falls between the program-
level average and median concentrations. The maximum 1,3-butadiene
concentration measured at this site is considerably less than the maximum
23-17
-------
concentration measured across the program. A few non-detects of 1,3-butadiene
were measured at this site.
• Figure 23-9 shows that the program-level average and median concentrations of
carbon tetrachloride are about the same. The annual average concentration for
SSSD is just less than these statistical parameters. The minimum carbon
tetrachloride concentration measured across the program was measured at SSSD
• Figure 23-10 presents the box plot for 1,2-dichloroethane. Similar to
1,3-butadiene, the program-level maximum concentration (17.01 |ig/m3) is not
shown directly on the box plot as the scale has been reduced to 1 |ig/m3 in order to
allow for the observation of data points at the lower end of the concentration
range. The program-level average concentration is greater than the program third
quartile for this pollutant and is greater than or similar to the maximum
concentration measured at most sites sampling 1,2-dichloroethane. This is
because the program-level average concentration is being driven by the higher
measurements collected at a few monitoring sites. Figure 23-1 1 shows that the
maximum 1,2-dichloroethane concentration measured at SSSD is two orders of
magnitude less than the maximum concentration measured across the program
and less than the program-level average concentration. The annual average
concentration for SSSD is just greater than the program -level first quartile. Eight
non-detects of 1,2-dichloroethane were measured at SSSD.
• Figure 23-1 1 shows that the annual average formaldehyde concentration for SSSD
is less than both the program-level average and median concentrations. The
maximum formaldehyde concentration measured at this site is considerably less
than the maximum concentration measured across the program. There were no
non-detects of formaldehyde measured at SSSD.
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 3.5.3.2.
Sampling at SSSD began in 2008 after a re-location from a previous sampling site. Thus,
Figures 23-12 through 23-17 present the 1-year statistical metrics for each of the pollutants of
interest for SSSD. 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 is not provided,
although the range and quartiles are still presented.
23-18
-------
Figure 23-12. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SSSD
12
10
8.
E
1 6-
1
4.
2.
0.
O
4
•
M
2008
1
••
J 1
o
— f — ' .
I 2 ! 2— ' ' S !
2009 2010 2011 2012
Year
5th Percentile - Minimum ~ Median — Maximum O 95th Percentile ...^>... Average
Observations from Figure 23-12 for acetaldehyde measurements collected at SSSD
include the following:
• SSSD began sampling carbonyl compounds under the NMP in 2008.
• The maximum acetaldehyde concentration (11.0 |ig/m3) was measured on
February 6, 2009, although a similar measurement (10.6 |ig/m3) was also measured in
January 2008. For both years, the second highest concentration measured was
considerably less than the maximum concentration (5.22 |ig/m3 for 2008 and
6.57|ig/m3for2009).
• Nearly all of the statistical parameters increased from 2008 to 2009, with the 95th
percentile exhibiting the largest increase. The number of concentrations greater than
4 |ig/m3 increased from three in 2008 to eight in 2009. The 1-year average
concentration increased from 2.00 |ig/m3 to 2.38 |ig/m3 from 2008 to 2009, although
confidence intervals indicate that the change is not statistically significant.
• A steady decreasing trend in the acetaldehyde concentrations measured at SSSD is
then shown through 2011, with little change shown for 2012. The range within which
the majority of measurements fall, as indicated by the 5th and 95th percentiles,
decreased by more than half from 2009 to 2011.
• Throughout the period of sampling, the median concentration exhibited little change,
ranging from 1.49 |ig/m3 (2011) to 1.79 |ig/m3 (2009).
23-19
-------
Figure 23-13. Yearly Statistical Metrics for Benzene Concentrations Measured at SSSD
O 5th Percentile - Minimurr
— Maximum O 95th Percentile
Observations from Figure 23-13 for benzene measurements collected at SSSD include the
following:
• SSSD also began sampling VOCs and SNMOCs under the NMP in 2008. Recall that
if both VOCs and SNMOCs are sampled concurrently at a site, the TO-15 results are
used for the 12 pollutants these methods have in common. Benzene is one of those 12
pollutants; thus, the results provided here are from TO-15.
• The maximum benzene concentration (2.49 |ig/m3) was measured at SSSD on
September 30, 2012. Only one other benzene concentration greater than 2 |ig/m3 has
been measured at SSSD (2.37 |ig/m3 on March 25, 2008).
• With the exception of the 95th percentile, nearly all of the statistical parameters
exhibit decreases from 2008 to 2009, which is the opposite of what is shown for
acetaldehyde for the same time frame. However, both pollutants exhibit an increase in
the range within which the majority of concentrations fall for 2009.
• The increases shown in the 1-year average and median concentrations for 2010 are
partly a result of higher concentrations on the lower end of the concentration range.
There were 14 concentrations in 2009 that are less than the minimum concentration
measured in 2010. In addition, the number of concentrations between 0.5 |ig/m3 and
1 |ig/m3 increased by nearly 70 percent. These two factors resulted in the increases
shown for 2010 as well as a tightening of the range within which a majority of the
concentrations fell.
23-20
-------
All of the statistical parameters exhibit a decrease from 2010 to 2011.
• Even though the maximum concentration was measured in 2012, the difference
between the 5th and 95th percentiles is at a minimum, as is the 1-year average
concentration. Even so, the 1-year average concentration of benzene has changed
relatively little over the years of sampling, ranging from 0.63 |ig/m3 (2012) to
0.74 |ig/m3 (2010).
Figure 23-14. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at SSSD
I
I
o
2010
Year
5th Percentile - Minimum " Median — Maximum
95th Percentile
Observations from Figure 23-14 for 1,3-butadiene measurements collected at SSSD
include the following:
• The maximum 1,3-butadiene concentration (0.31 |ig/m3) was measured at SSSD on
September 30, 2012, the same day as the maximum benzene concentration. Of the 14
1,3-butadiene concentration greater than 0.15 |ig/m3 measured at SSSD, 10 were
measured in 2012.
• Nearly all of the statistical parameters exhibit a decrease from 2008 to 2009, with the
exception of the minimum concentration. For both years, two non-detects were
measured.
• The number of non-detects increased from 2009 to 2010, as indicated by the decrease
in the 5th percentile. The number of non-detects increased from two for 2009 to 13
for 2010. Even so, the 1-year average and in particular, the median concentration
exhibit increases. This is because the number of concentrations in the mid- to upper-
23-21
-------
end of the concentration range increased while those on the lower end of the range
(but still detected) decreased. The number of 1,3-butadiene concentrations between
0.04 |ig/m3 and 0.08 |ig/m3 tripled from 2009 to 2010, increasing from nine to 27.
Conversely, the number of measurements less than 0.04 |ig/m3 decreased from 44 in
2009 to 16 in 2010.
• While relatively little change in the concentrations of 1,3-butadiene is shown from
2010 to 2011, concentrations increased significantly for 2012. The maximum, 95th
percentile, and 1-year average concentrations nearly doubled from 2011 to 2012. The
median concentration also exhibits an increase. The number of non-detects also
decreased significantly from 2011 to 2012, down to three from 17.
Figure 23-15. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
SSSD
2010
Year
O 5th Percentile - Minimum " Median — Maximum • 95th Percentile "-O-" Averagf
Observations from Figure 23-15 for carbon tetrachloride measurements collected at
SSSD include the following:
• Eleven concentrations of carbon tetrachloride greater than 1.0 |ig/m3 have been
measured at SSSD since the onset of sampling in 2008. All of these were measured in
2008 and 2009.
• The box and whisker plots for this pollutant appear "inverted," with the minimum
concentration extending farther away from the majority of the measurements than the
maximum (see benzene or 1,3-butadiene as examples). For 2010 and 2011, the central
tendency statistics are closer to the 95th percentile than the 5th percentile, with the
median concentration greater than the 1-year average concentration, both of which are
23-22
-------
a little unusual. The median concentration is the midpoint of a dataset. The difference
between the maximum concentration and median for 2010 is 0.23 |ig/m3; the
difference between the median and the minimum concentration is 0.57 |ig/m3. Thus, a
greater number of concentrations are clustered around the upper end of the
concentration range, while the concentrations on the lower end of the concentration
range are more spread apart. Because the 1-year average concentration is influenced
more by outlying concentrations, the 1-year average is being pulled downward by the
concentrations at the lower end of the range. The same is true for 2011.
The concentrations measured in 2012 exhibit less variability, as indicated by the
difference between the 5th and 95th percentiles, which decreased by almost half from
2011 to 2012.
• Even though the range of measurements across the years of sampling vary by more
than 1 |ig/m3, the median concentration for each year varied by 0.1 |ig/m3 and the
1-year average concentration varied by less than 0.2 |ig/m3.
Figure 23-16. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
SSSD
2010
Year
5th Percentile - Minimurr
Median — Maximum
95th Percentile
23-23
-------
Observations from Figure 23-16 for 1,2-dichloroethane measurements collected at SSSD
include the following:
• There was only one measured detection of 1,2-dichloromethane in 2008 and only two
in 2009; as a result, nearly all of the statistical metrics are equal to or just greater than
zero.
• The number of measured detections increased to nine in 2010 and to 17 in 2011. This
explains the significant increases shown in the 95th percentiles and 1-year average
concentrations. However, the median concentration is still zero because more than
half of the measurements are still non-detects.
• For 2012, measured detections account for nearly 87 percent of the measurements. As
a result, only the minimum and 5th percentile are zero for 2012 and both the median
and 1-year average concentration exhibit significant increases.
Figure 23-17. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at SSSD
I
2010
Year
O 5th Percentile - Minimurr
Median — Maximum • 95th Percentile ...^... Average
Observations from Figure 23-17 for formaldehyde measurements collected at SSSD
include the following:
• The maximum formaldehyde concentration (5.38 |ig/m3) was measured on
May 1, 2011, although concentrations greater than 5 |ig/m3 were also measured in
2008 and 2009.
23-24
-------
• While the maximum and 95th percentile did not change from 2008 to 2009, the
remaining statistical parameters decreased, particularly the 1-year average and
median concentrations. The number of concentrations greater than 3 |ig/m3 decreased
from 25 in 2008 to seven in 2009 while the number of concentrations less than
2 |ig/m3 increased from 10 in 2008 to 29 in 2009.
• The significant decrease in formaldehyde concentrations shown from 2008 to 2009 is
followed by a slight decreasing trend through 2012.
23.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk at the
South Dakota monitoring site. Refer to Sections 3.3 and 3.5.3 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.
23.5.1 Risk-Based Screening Assessment Using MRLs
A risk-based screening was conducted by comparing the concentration data from the
South Dakota monitoring site to the ATSDR MRLs, where available. As described in
Section 3.3, MRLs are noncancer health risk benchmarks and are defined for three exposure
periods: acute (exposures of 1 day to 14 days); intermediate (exposures of 15 days to 364 days);
and chronic (exposures of 1 year or greater). The preprocessed daily measurements of the
pollutants of interest were compared to the acute MRLs; the quarterly averages were compared
to the intermediate MRLs; and the annual averages were compared to the chronic MRLs.
As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2012.
23.5.2 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the South Dakota 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 3.5.3.4 for an explanation of how cancer risk and
23-25
-------
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 23-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Table 23-6. Risk Approximations for the South Dakota Monitoring Site
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(Ug/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
Sioux Falls, South Dakota - SSSD
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
Formaldehyde
0.0000022
0.0000078
0.00003
0.000006
0.000026
0.000013
0.009
0.03
0.002
0.1
2.4
0.0098
58/58
61/61
58/61
61/61
53/61
58/58
1.69
±0.23
0.63
±0.08
0.08
±0.02
0.64
±0.03
0.07
±0.01
1.86
±0.21
3.71
4.91
2.41
3.85
1.71
24.15
0.19
0.02
0.04
0.01
0.01
0.19
Observations from Table 23-6 for SSSD include the following:
• The pollutants with the highest annual average concentrations for SSSD are
formaldehyde and acetaldehyde.
• Formaldehyde has the highest cancer risk approximation (24.15 in-a-million) among
this site's pollutants of interest, followed by benzene (4.91 in-a-million) and carbon
tetrachloride (3.85 in-a-million).
• Acetaldehyde and formaldehyde have the highest noncancer hazard approximations
among SSSD's pollutants of interest, both with an HQ of 0.19. Because none of the
noncancer hazard approximations were greater than 1.0, no adverse health effects are
expected from these individual pollutants.
23.5.3 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 23-7 presents the 10 pollutants with the highest emissions from the 2011 NEI that have
cancer toxicity factors. Table 23-7 also presents the 10 pollutants with the highest toxicity-
23-26
-------
weighted emissions, based on the weighting schema described in Section 3.5.3.5. Lastly,
Table 23-7 provides the 10 pollutants with the highest cancer risk approximations (in-a-million)
for SSSD, as presented in Table 23-6. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 23-7. Table 23-8 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 3.5.3.5. Similar to the cancer risk and
noncancer hazard approximations provided in Section 23.5.2, this analysis may help policy-
makers prioritize their air monitoring activities.
Observations from Table 23-7 include the following:
• Formaldehyde, benzene, and ethylbenzene are the highest emitted pollutants with
cancer UREs in Minnehaha County.
• Formaldehyde, benzene, and POM, Group la are the pollutants with the highest
toxicity-weighted emissions (of the pollutants with cancer UREs) for Minnehaha
County.
• Eight of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Minnehaha County.
• Formaldehyde and benzene top both emissions-based lists and have the highest
cancer risk approximations for SSSD. Acetaldehyde and 1,3-butadiene also appear on
all three lists. Conversely, carbon tetrachloride and 1,2-dichloroethane appear on
neither emissions-based list but are among the pollutants of interest for SSSD.
• Naphthalene and several POM Groups appear among the highest emitted pollutants in
Minnehaha County and are among those with the highest toxicity-weighted
emissions. PAHs were not sampled for at SSSD.
23-27
-------
Table 23-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the South Dakota Monitoring Site
Top 10 Total Emissions for Pollutants
with Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Sioux Falls, South Dakota (Minnehaha County) - SSSD
Formaldehyde
Benzene
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
POM, Group la
POM, Group 2b
Trichloroethylene
POM, Group 2d
94.68
72.38
54.26
48.96
13.91
5.05
4.87
1.21
1.14
0.98
Formaldehyde
Benzene
POM, Group la
1,3 -Butadiene
Naphthalene
POM, Group 3
Ethylbenzene
Hexavalent Chromium, PM
Acetaldehyde
POM, Group 2b
1.23E-03
5.65E-04
4.28E-04
4.17E-04
1.72E-04
1.43E-04
1.36E-04
1.21E-04
1.08E-04
1.06E-04
Formaldehyde
Benzene
Carbon Tetrachloride
Acetaldehyde
1,3 -Butadiene
1 ,2-Dichloroethane
24.15
4.91
3.85
3.71
2.41
1.71
to
to
oo
-------
Table 23-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the South Dakota Monitoring Site
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer Hazard
Approximation
(HQ)
Sioux Falls, South Dakota (Minnehaha County) - SSSD
Toluene
Xylenes
Ethylene glycol
Hexane
Methanol
Formaldehyde
Benzene
Ethylbenzene
Acetaldehyde
Styrene
506.57
258.52
210.29
148.02
106.78
94.68
72.38
54.26
48.96
22.93
Acrolein
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Xylenes
Benzene
Naphthalene
Hydrochloric acid
Arsenic, PM
Lead, PM
124,884.21
9,661.40
6,957.09
5,440.09
2,585.24
2,412.62
1,684.45
973.59
958.91
944.44
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Benzene
Carbon Tetrachloride
1 ,2-Dichloroethane
0.19
0.19
0.04
0.02
0.01
0.01
to
to
VO
-------
Observations from Table 23-8 include the following:
• Toluene, xylenes, and ethylene glycol are the highest emitted pollutants with
noncancer RfCs in Minnehaha County.
• Acrolein is the pollutant with the highest toxi city-weighted emissions (of the
pollutants with noncancer RfCs), followed by formaldehyde and 1,3-butadiene.
Although acrolein was sampled for at SSSD, 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 one of the highest emitted
pollutants in Minnehaha County.
• Four of the highest emitted pollutants also have the highest toxi city-weighted
emissions for Minnehaha County.
• Formaldehyde and acetaldehyde, which have the highest noncancer hazard
approximations for SSSD, appear on both emissions-based lists, as does benzene.
1,3-Butadiene appears among those with the highest toxicity-weighted emissions but
is not one of the highest emitted (with a noncancer RfC) in Minnehaha County.
Carbon tetrachloride and 1,2-dichloroethane again appear on neither emissions-based
list.
23.6 Summary of the 2012 Monitoring Data for SSSD
Results from several of the data treatments described in this section include the
following:
»«» Eleven pollutants failed at least one screen for SSSD.
»«» Formaldehyde and acetaldehyde are the only pollutants of interest for which the
annual average concentrations were greater than 1 jug/m3.
»«» The maximum concentrations of several of SSSD's pollutants of interest were
measured at SSSD on September 30, 2012.
»«» Concentrations of 1,3-butadiene increased significantly from 2011 to 2012.
Conversely, formaldehyde concentrations measured at SSSD exhibit a steady
decreasing trend across the years, although the most significant decreases were
realized during the early years of sampling. In addition, the detection rate of
1,2-dichloroethane has been increasing steadily at SSSD over the years of sampling.
23-30
-------
24.0 Sites in Texas
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS sites in Texas, 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.
24.1 Site Characterization
This section characterizes the CAMS 35 and CAMS 85 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 CAMS 35 monitoring site is located in the Houston-The Woodlands-Sugarland,
Texas MSA and CAMS 85 is part of the Marshall, Texas MSA. Figure 24-1 is a composite
satellite image retrieved from ArcGIS Explorer showing the monitoring site and its immediate
surroundings. Figure 24-2 identifies nearby point source emissions locations by source category
for the site, as reported in the 2011 NEI for point sources. Note that only sources within 10 miles
of the site are included in the facility counts provided in Figure 24-2. 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
radius are still visible on the map, but have been grayed out in order to show emissions sources
just outside the boundary. Figures 24-3 and 24-4 are the composite satellite image and point
emissions sources map for CAMS 85. Table 24-1 provides supplemental geographical
information such as land use, location setting, and locational coordinates.
24-1
-------
Figure 24-1. Deer Park, Texas (CAMS 35) Monitoring Site
to
-^
to
-------
Figure 24-2. NEI Point Sources Located Within 10 Miles of CAMS 35
" "7 / ^]'*-;«
• + *?-<£ :/--• --^--^.. 3
J A **>. HB _
Legend
5"10'0"W 95'5'0"W 95'0'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
CAMS 35 NATTS site O 10 mile radius
Source Category Group (No. of Facilities)
County boundaries
1" Airport/Airline/Airport Support Operations (24) A
$ Asphalt Production/Hot Mix Asphalt Plant (1) ®
B Bulk Terminals/Bulk Plants (18) <•>
C Chemical Manufacturing (84) X
i Compressor Station (9} ?
6 Electrical Equipment Manufacturing (1) •
$ Electricity Generation via Combustion (6) $$
E Electroplating, Plating. Polishing, Anodizing, and Coloring (1) 3
Fertilizer Plant (1) R
F Food Processing/Agriculture (1) ^
© Gas Plant (1) H
{& Glass Plant (1) X
-^- Industrial Machinery or Equipment Plant (2) £.
O Institution (school, hospital, prison, etc.) (1) *
Landfill (2)
Metal Can, Box, and Other Metal Container Manufacturing (2)
Metals Processing/Fabrication (2)
Mine/Quarry/Mineral Processing (1)
Miscellaneous Commercial/Industrial (13)
Oil and/or Gas Production (7)
Petroleum Products Manufacturing (1)
Petroleum Refinery (5)
Plastic, Resin, or Rubber Products Plant (10)
Port and Harbor Operations (3)
Pulp and Paper Plant (1)
Rail Yard/Rail Line Operations (4)
Ship/Boat Manufacturing or Repair (4)
Wastewater Treatment (2)
24-3
-------
Figure 24-3. Karnack, Texas (CAMS 85) Monitoring Site
to
-------
Figure 24-4. NEI Point Sources Located Within 10 Miles of CAMS 85
94II1510"W 94"10'0"W 941J5'0"W 941J0'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
Legend
•fr
CAMS 85 NATTS site O 10 mile radius
County boundaries
Source Category Group (No. of Facilities)
T Airport/Airline/Airport Support Operations (3)
24-5
-------
Table 24-1. Geographical Information for the Texas Monitoring Sites
Site Code
CAMS 35
CAMS 85
AQS Code
48-201-1039
48-203-0002
Location
Deer Park
Karnack
County
Harris
Harrison
Micro- or
Metropolitan
Statistical Area
Houston-The
Woodlands-Sugar
Land, TX MSA
Marshall, TX
MSA
Latitude
and
Longitude
29.670025,
-95.128508
32.668987,
-94.167457
Land Use
Residential
Agricultural
Location
Setting
Urban/City
Center
Rural
Additional Ambient Monitoring Information1
Haze, TSP Lead, CO, SO2, NOy, NO, NO2, NOX,
PAMS/SNMOCs, VOCs, Carbonyl compounds, O3,
Meteorological parameters, PM10, PM Coarse, PM10
Speciation, PM2 5, and PM2 5 Speciation, Black
Carbon, IMPROVE Speciation, SVOCs.
SVOCs, NO2, NO, NOX, PAMS/SNMOCs, Carbonyl
Compounds, VOCs, O3, Meteorological parameters,
PMio, PM10 Speciation, PM25, PM2 5 Speciation,
IMPROVE Speciation.
:Data for additional pollutants are reported to AQS for these sites (EPA, 2013b); however, these data are not generated by ERG and are therefore not included in this report.
BOLD ITALICS = EPA-designated NATTS Site
to
-------
The CAMS 35 monitoring site is located in Deer Park, southeast of Houston, in east
Texas. This site serves as the Houston NATTS Site. The site is located at Brown Memorial Park,
in a primarily residential area, as shown in Figure 24-1. Major thoroughfares are near the site,
including Beltway 8 (1.6 miles to the west) and Highway 225 (2.8 miles to the north). Galveston
Bay is located to the east and southeast of the site and the Houston Ship Channel, which runs
from the Bay westward towards downtown Houston, is located to the north on the other side of
Highway 225. The east side of Houston has significant industry, including several major oil
refineries. As Figure 24-2 shows, a large number of emissions sources are located roughly along
a line that runs east to west just north of the site (or along the Houston Ship Channel). A second
cluster of emissions sources is located to the southeast of the monitoring site. The source
category with the greatest number of sources (84) surrounding CAMS 35 is chemical
manufacturing. Other source categories with a number of sources around CAMS 35 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; plastic, resin, or rubber products plants; compressor stations; and oil and gas
production. The point source located closest to the CAMS 35 monitoring site is a heliport at San
Jacinto College's Central Campus in Pasadena
The CAMS 85 NATTS site is located in Karnack, in northeast Texas. The monitoring site
is about 10 miles northeast of Marshall, Texas and about 7 miles west of the Texas-Louisiana
border. This site is located on the property of the Longhorn Army Ammunition Plant near the
intersection of FM Road 134 and Spur Road 449 (Taylor Avenue), as shown in Figure 24-3. The
surrounding area is rural and agricultural. As Figure 24-4 shows, there are few point sources
within 10 miles of CAMS 85 and these sources all fall into a single source category: the airport
source category. The closest source to CAMS 85 is the Fly-N-Fish Lodge Airport near Caddo
Lake.
Table 24-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Texas monitoring sites. Table 24-2 includes both county-level
population and vehicle registration information. Table 24-2 also contains traffic volume
information for each site as well as the location for which the traffic volume was obtained.
Additionally, Table 24-2 presents the county-level daily VMT for Harris and Harrison Counties.
24-7
-------
Table 24-2. Population, Motor Vehicle, and Traffic Information for the Texas
Monitoring Sites
Site
CAMS 35
CAMS 85
Estimated
County
Population1
4,253,700
67,450
County-level
Vehicle
Registration2
3,252,420
71,658
Annual
Average Daily
Traffic3
31,043
1,250
Intersection
Used for
Traffic Data
Spencer Hwy, between Red Bluff Rd
and Underwood Rd
FM Rd 134 at intersection with 449
County-level
Daily VMT4
57,020,660
2,405,125
Bounty-level population estimates reflect 2012 data (Census Bureau, 2013c)
2County-level vehicle registration reflects 2012 data (TX DMV, 2013)
3AADT reflects 2011 data for CAMS 85 and 2004 data for CAMS 35 (TX DOT, 2011 and HCPID, 2013)
4County-level VMT reflects 2012 data (TX DOT, 2012)
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 24-2 include the following:
• The population and vehicle ownership counts are significantly higher for CAMS 35
than CAMS 85. This is not surprising given the rural nature of the area surrounding
the CAMS 85 site and the large urban area encompassed within Harris County.
• Compared to other counties with NMP monitoring sites, Harris County is third
highest for both county-level population and county-level vehicle ownership.
Conversely, Harrison County is among the lowest for both county-level population
and vehicle ownership.
• The traffic volume passing CAMS 35 is substantially higher than the traffic volume
passing CAMS 85. The traffic volume for CAMS 35 is in the middle of the range
compared to other NMP sites while the traffic volume near CAMS 85 is among the
lower traffic volumes for NMP sites. Traffic data for CAMS 35 are provided for
Spencer Highway between Red Bluff Road and Underwood Road; the traffic data for
CAMS 85 are provided for FM Road 134 at the intersection with Spur Road 449.
• Like the other mobile source activity indicators, county-level daily VMT is
considerably higher for Harris County than Harrison County. Harris County ranks
fourth compared to other counties with NMP sites for VMT, while Harrison County
ranks in the bottom third.
24.2 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
sites in Texas on sample days, as well as over the course of the year.
24.2.1 Climate Summary
The eastern third of Texas is characterized by a subtropical humid climate, with the
climate becoming more continental in nature farther north and west. The proximity to the Gulf of
Mexico acts as a moderating influence as temperatures soar in the summer or dip in the winter.
24-8
-------
Areas closer to the coast, such as Houston, remain slightly cooler in the summer than
neighboring areas to the north. The reverse is also true, as coastal areas are warmer in the winter
than areas farther inland, although East Texas winters are relatively mild. The onshore flow from
the Gulf of Mexico also allows humidity levels to remain high in East Texas, particularly near
the coast. The winds flow out of the Gulf of Mexico a majority of the year, with the winter
months being the exception, as frontal systems allow colder air to filter in from the north.
Abundant rainfall is also typical of the region, again due in part to the nearness to the Gulf of
Mexico. Greater than 45 inches of precipitation can be expected annually. Severe weather is
most common in spring, particularly in May, and tropical systems can be a threat to the state
during the summer and fall. Snowfall is rare in East Texas but ice storms are more common in
northeast Texas than in other parts of the state (Wood, 2004; TAMU, 2014).
24.2.2 Meteorological Summary
Hourly meteorological data for 2012 were retrieved from NCDC for the weather stations
closest to the Texas monitoring sites (NCDC, 2012), as described in Section 3.5.2. The closest
weather station to CAMS 35 is located at William P. Hobby Airport, WBAN 12918; the closest
weather station to CAMS 85 is located at Shreveport Regional Airport, WBAN 13957.
Additional information about the Hobby Airport and Shreveport Regional Airport weather
stations, such as the distance between the sites and the weather stations, is provided in Table 24-
3. These data were used to determine how meteorological conditions on sample days vary from
conditions experienced throughout the year.
Table 24-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for the entire year. Also included in Table 24-3 is the 95 percent
confidence interval for each parameter. As shown in Table 24-3, average meteorological
conditions on sample days were representative of average weather conditions experienced
throughout the year near both sites.
24-9
-------
Table 24-3. Average Meteorological Conditions near the Texas Monitoring Sites
Closest
Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Deer Park, Texas - CAMS 35
William P.
Hobby Airport
12918
(29.65, -95.28)
9.1
miles
256°
(WSW)
Sample
Days
(61)
2012
80.3
±2.8
80.7
+ 1.1
72.1
±2.8
71.9
+ 1.1
61.9
±3.2
61.2
+ 1.3
65.9
±2.7
65.5
+ 1.1
72.9
±2.8
71.8
+ 1.2
1016.7
± 1.2
1016.7
+ 0.5
6.3
±0.6
6.1
+ 0.3 1
Karnack, Texas - CAMS 85 |
Shreveport
Regional
Airport
13957
(32.45, -93.82)
24.4
miles
127°
(SE)
Sample
Days
(61)
2012
79.3
±3.4
79.0
+ 1.4
68.4
±3.3
68.3
+ 1.4
56.7
±3.5
56.5
+ 1.5
61.5
±3.0
61.4
+ 1.3
69.3
±2.8
69.4
+ 1.2
1016.2
± 1.3
1016.2
+ 0.6
6.2
±0.7
5.8
+ 0.3
to
o
^ Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
-------
24.2.3 Back Trajectory Analysis
Figure 24-5 is the composite back trajectory map for days on which samples were
collected at the CAMS 35 monitoring site. Included in Figure 24-5 are four back trajectories per
sample day. Figure 24-6 is the corresponding cluster analysis. Similarly, Figures 24-7 and 24-8
are the composite back trajectory map and corresponding cluster analysis for days on which
samples were collected at CAMS 85. An in-depth description of these maps and how they were
generated is presented in Section 3.5.2.1. For the composite maps, each line represents the
24-hour trajectory along which a parcel of air traveled toward the monitoring site on a given
sample day and time, based on an initial height of 50 meters AGL. For the cluster analyses, each
line corresponds to a trajectory representative of a given cluster of back trajectories. Each
concentric circle around the sites in Figures 24-5 through 24-8 represents 100 miles.
Observations from Figures 24-5 and 24-6 for CAMS 35 include the following:
• Back trajectories originated from a variety of directions at the CAMS 35 monitoring
site, although the majority of trajectories originated over the Gulf of Mexico or to the
north of the site and rarely to the west of the site.
• The 24-hour air shed domain for CAMS 35 is similar in size to many other NMP
monitoring sites. Although the farthest away a back trajectory originated was over the
Gulf of Mexico, or nearly 600 miles away, the average back trajectory length was
252 miles. Approximately 85 percent of back trajectories originated within 400 miles
of the site.
• The cluster analysis shows that greater than 50 percent of back trajectories originated
over the Gulf of Mexico, although the position over the Gulf and the trajectory length
varies. Another common trajectory origin is from the northeast to east (13 percent),
over Louisiana and Mississippi. Another 7 percent of back trajectories originated over
Oklahoma and north-central Texas. One cluster trajectory for CAMS 35 is short
enough that it is covered up by the star symbol; thus, the cluster trajectory is
presented in the inset map in Figure 24-6. This cluster trajectory includes back
trajectories of varying directions but generally short distances (less than 200 miles in
length). The back trajectories are frequently curved in nature, looping around
Southeast Texas or offshore before arriving at the monitoring site. Most of these back
trajectories are obscured in Figure 24-5 by the density of trajectory pathways nearest
the site.
24-11
-------
Figure 24-5. Composite Back Trajectory Map for CAMS 35
Figure 24-6. Back Trajectory Cluster Map for CAMS 35
24-12
-------
Figure 24-7. Composite Back Trajectory Map for CAMS 85
Figure 24-8. Back Trajectory Cluster Map for CAMS 85
24-13
-------
Observations from Figures 24-7 and 24-8 for CAMS 85 include the following:
• Back trajectories originated from a variety of directions at the CAMS 85 monitoring
site, although back trajectories originating to the east and west are rare.
• The 24-hour air shed domain for CAMS 85 is slightly smaller in size compared to
CAMS 35. The average back trajectory length is 219 miles and most trajectories
(83 percent) originated less than 300 miles from CAMS 85. The farthest away a back
trajectory originated was nearly 600 miles away, over Nebraska.
• The cluster analysis for CAMS 85 shows that greater than one-third of back
trajectories originated to the south of the site, but are split into two different cluster
trajectories. Another 20 percent of back trajectories originated over East Texas and
Louisiana. Nine percent of back trajectories originated over eastern Arkansas and
along the Mississippi River. An additional 28 percent of back trajectories originated
to the north of the site, as indicated by the short cluster (20 percent) representing
relatively short back trajectories originating over the Ark-La-Tex region and eastern
Oklahoma, and the longer cluster (8 percent) originating over the central Plains.
24.2.4 Wind Rose Comparison
Hourly surface wind data from the weather stations at Hobby Airport near CAMS 35 and
Shreveport Regional Airport near CAMS 85 were uploaded into a wind rose software program to
produce customized wind roses, as described in Section 3.5.2.2. 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 24-9 presents a map showing the distance between the weather station and
CAMS 35, which may be useful for identifying topographical influences that may affect the
meteorological patterns experienced at this location. Figure 24-9 also presents three different
wind roses for the CAMS 35 monitoring site. First, a historical wind rose representing 2002 to
2011 wind data is presented, which shows the predominant surface wind speed and direction
over an extended period of time. Second, a wind rose representing wind observations for all of
2012 is presented. Next, a wind rose representing wind data for days on which samples were
collected in 2012 is presented. These can be used to identify the predominant wind speed and
direction for 2012 and to determine if wind observations on sample days were representative of
conditions experienced over the entire year and historically. Figure 24-10 presents the distance
map and three wind roses for CAMS 85.
24-14
-------
Figure 24-9. Wind Roses for the William P. Hobby Airport Weather Station near CAMS 35
Location of CAMS 35 and Weather Station
2002-2011 Historical Wind Rose
2012 Wind Rose
Sample Day Wind Rose
Calms: 17.04%
Calms: 1578%
24-15
-------
Figure 24-10. Wind Roses for the Shreveport Regional Airport Weather Station near
CAMS 85
Location of CAMS 85 and Weather Station
2002-2011 Historical Wind Rose
Cshns: 16.03%
2012 Wind Rose
Sample Day Wind Rose
24-16
-------
Observations from Figure 24-9 for CAMS 35 include the following:
• The Hobby Airport weather station is located approximately 9 miles west-southwest
of CAMS 35.
• The historical wind rose shows that winds from the southeast quadrant, including
both easterly and southerly winds, prevailed near the CAMS 35 site. Northerly winds
were also observed fairly often. Calm winds (<2 knots) were observed for
approximately 14 percent of the wind measurements.
• The wind patterns on the wind rose for 2012 resemble the historical wind patterns.
However, the percentage of calm winds was slightly higher for 2012 (17 percent).
• The wind patterns shown on the sample day wind rose resemble the wind patterns
shown on both the full-year and historical wind roses, indicating that conditions
experienced near CAMS 35 on sample days are representative of those experienced
throughout the year and over time.
Observations from Figure 24-10 for CAMS 85 include the following:
• The Shreveport Regional Airport weather station is located across the Texas-
Louisiana border, approximately 24 miles southeast of CAMS 85.
• The wind patterns on the historical wind rose for CAMS 85 bear some resemblance to
those on the historical wind rose for CAMS 35. The historical wind rose shows that
winds from the southeast to south account for approximately 30 percent of the wind
observations near the CAMS 85 site. Northerly winds were also observed fairly often.
Calm winds were observed for approximately 16 percent of the wind measurements.
• The wind patterns shown on the 2012 wind rose are similar to the historical wind
patterns, although the number of southerly and south-south westerly winds increased
for 2012. The calm rate also increased for 2012 (19 percent).
• The sample day wind patterns resemble the full-year wind patterns, although the calm
rate is slightly less (16 percent) while the number of southeasterly wind observations
increased.
24.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each Texas
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 24-4.
24-17
-------
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 24-4. It is
important to note which pollutants were sampled for at each site when reviewing the results of
this analysis. Although CAMS 35 sampled for hexavalent chromium and PAHs, sampling for
PAHs was discontinued at the end of February 2012. CAMS 85 sampled for hexavalent
chromium only.
Table 24-4. Risk-Based Screening Results for the Texas Monitoring Sites
Pollutant
Screening
Value
(Ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Deer Park, Texas - CAMS 35
Naphthalene
Hexavalent Chromium
0.029
0.000083
Total
8
5
13
9
60
69
88.89
8.33
18.84
61.54
38.46
61.54
100.00
Karnack, Texas - CAMS 85
Hexavalent Chromium
0.000083
Total
0
0
47
47
0.00
0.00
0.00
0.00
Observations from Table 24-4 include the following:
• Nine valid PAH samples were collected at CAMS 35 before sampling was
discontinued.
• Naphthalene was the only PAH to fail screens for CAMS 35. Of the nine valid
samples collected, naphthalene failed screens for eight of them (89 percent).
• Hexavalent chromium was detected in 60 of the 61 valid samples collected at
CAMS 35. This pollutant failed five screens (8 percent).
• Naphthalene accounted for roughly 62 percent of the failed screens for CAMS 35,
with hexavalent chromium accounting for the other 38 percent. Thus, both pollutants
were identified as pollutants of interest for this site.
• Hexavalent chromium is the only pollutant sampled for at CAMS 85. This pollutant
did not fail any screens during the 2012 monitoring effort. This was also true for
2011.
• Because CAMS 85 does not have any pollutants of interest, this site is excluded from
the sections that follow, with the exception of the emissions section (Section 24.5.3).
24-18
-------
24.4 Concentrations
This section presents various concentration averages used to characterize pollution levels
at the CAMS 35 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
CAMS 35.
• Annual concentration averages are presented graphically 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 the site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Additional site-specific statistical summaries for
CAMS 35 are provided in Appendices M and O. A site-specific statistical summary is also
provided for CAMS 85 in Appendix O.
24.4.1 2012 Concentration Averages
Quarterly and annual concentration averages were calculated for the pollutants of interest
for CAMS 35, as described in Section 3.1. The quarterly average 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
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 CAMS 35
monitoring site are presented in Table 24-5, 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.
24-19
-------
Table 24-5. Quarterly and Annual Average Concentrations of the Pollutants of
Interest for the Texas Monitoring Sites
Pollutant
#of
Measured
Detections
vs. # 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)
Deer Park, Texas - CAMS 35
Hexavalent Chromium
Naphthalene
60/61
9/9
0.037
± 0.007
NA
0.040
±0.012
NA
0.053
± 0.009
NA
0.058
± 0.027
NA
0.047
± 0.008
NA
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
Note: There are no pollutants of interest for CAMS 85.
Observations from Table 24-5 include the following:
• Concentrations of hexavalent chromium measured at CAMS 35 range from
0.0044 ng/m3 to 0.195 ng/m3, including a single non-detect. The median
concentration for the dataset is 0.042 ng/m3.
• Concentrations of hexavalent chromium appear lower during the first half of the year
and higher during the second half of the year, as indicated by the quarterly averages.
In addition, the fourth quarter average has a relatively large confidence interval
associated with it. A review of the data shows that the three highest concentrations of
this pollutant (those greater than 0.1 ng/m3) were measured between October and
December at CAMS 35. Further, 16 of the 22 measurements greater than 0.05 ng/m3
were measured at CAMS 35 between July and December (with only one measured
during the first quarter and five during the second).
• Compared to other NMP sites sampling hexavalent chromium, the annual average
concentration for CAMS 35 is among the higher annual averages, ranking fifth
among the NMP sites sampling hexavalent chromium.
• Concentrations of naphthalene measured at CAMS 35 range from 24.0 ng/m3 to
181 ng/m3.
• Because sampling for PAHs was discontinued in February 2012, no quarterly or
annual averages could be calculated.
24.4.2 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, a box plot was created for hexavalent chromium
for CAMS 35. Figure 24-11 overlays the site's minimum, annual average, and maximum
24-20
-------
concentrations onto the program-level minimum, first quartile, median, average, third quartile,
and maximum concentrations, as described in Section 3.5.3.1.
Figure 24-11. Program vs. Site-Specific Average Hexavalent Chromium Concentration
CAMS 35
• 1°
)
Program:
Site:
^^^ ^^^
0.1 0.2 0.3
Concentration (ng/m3)
.
0.4 0
1st Quartile 2nd Quartile 3rd Quartile 4th Quartile Average
• n n n I
Site Average Site Concentration Range
o —
Observations from Figure 24-11 include the following:
• Figure 24-11 is the box plot for hexavalent chromium for CAMS 35. Note that the
program-level maximum concentration (8.51 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 to 0.5 ng/m3. In addition, the program-level first quartile is zero and
therefore not visible on the box plot. Figure 24-11 shows that the annual average
hexavalent chromium concentration for CAMS 35 is just greater than the
program-level average concentration. The maximum hexavalent chromium
concentration measured at CAMS 35 is significantly less than the maximum
concentration measured at the program-level.
24.4.3 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 3.5.3.2.
Although CAMS 35 has sampled PAHs continuously since 2008 under the NMP, sampling was
discontinued at this site in February 2012. Hexavalent chromium sampling under the NMP did
not begin until 2010 and therefore does not meet the criteria specified above. As a result, a trends
analysis was not conducted.
24.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk at
CAMS 35. Refer to Sections 3.3 and 3.5.3 for definitions and explanations regarding the various
toxicity factors, time frames, and calculations associated with these risk-based screenings.
24-21
-------
24.5.1 Risk-Based Screening Assessment Using MRLs
A risk-based screening was conducted by comparing the concentration data from the
Texas monitoring sites to the ATSDRMRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 day to 14 days); intermediate (exposures of 15 days to 364 days); and chronic
(exposures of 1 year or greater). The preprocessed daily measurements of the pollutants of
interest were compared to the acute MRLs; the quarterly averages were compared to the
intermediate MRLs; and the annual averages were compared to the chronic MRLs.
As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2012.
24.5.2 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for CAMS 35 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 3.5.3.4 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 24-6, where applicable. Cancer risk approximations are presented as
probabilities while the noncancer hazard approximations are ratios and thus, unitless values.
Table 24-6. Risk Approximations for the Texas Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections vs.
# of Samples
Annual
Average
(ng/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer Risk
Approximation
(HQ)
Deer Park, Texas - CAMS 35
Hexavalent Chromium
Naphthalene
0.012
0.000034
0.0001
0.003
60/61
9/9
0.047
± 0.008
NA
0.56
NA
0.01
NA
NA = Not available due to the criteria for calculating an annual average.
Note: There are no pollutants of interest for CAMS 85.
24-22
-------
Observations from Table 24-6 include the following:
• The cancer risk approximation for hexavalent chromium for CAMS 35 is
0.56 in-a-million, which is less than a level of concern.
• The noncancer hazard approximation for hexavalent chromium for CAMS 35 is
considerably less than 1.0, indicating that no adverse health effects are expected from
this individual pollutant.
24.5.3 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 24-7 presents the 10 pollutants with the highest emissions from the 2011 NEI that have
cancer toxicity factors. Table 24-7 also presents the 10 pollutants with the highest toxicity-
weighted emissions, based on the weighting schema described in Section 3.5.3.5. Lastly,
Table 24-7 provides the 10 pollutants with the highest cancer risk approximations (in-a-million)
for each site, as presented in Table 24-6. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 24-7. Table 24-8 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 3.5.3.5. Similar to the cancer risk and
noncancer hazard approximations provided in Section 24.5.2, this analysis may help policy-
makers prioritize their air monitoring activities.
24-23
-------
Table 24-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Texas Monitoring Sites
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
Deer Park, Texas (Harris County) - CAMS 35
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Methyl tert butyl ether
Naphthalene
Propylene oxide
Dichloromethane
Trichloroethylene
1,003.70
609.51
600.25
358.08
332.11
109.34
90.96
59.07
49.15
27.58
1,3 -Butadiene
Benzidine, gas
Formaldehyde
Benzene
Hexavalent Chromium, PM
Naphthalene
Ethylene oxide
Nickel, PM
Acrylonitrile
Arsenic, PM
9.96E-03
8.83E-03
7.92E-03
7.83E-03
3.84E-03
3.09E-03
2.23E-03
1.98E-03
1.79E-03
1.66E-03
Hexavalent Chromium 0.56
Karnack, Texas (Harrison County) - CAMS 85
Formaldehyde
Benzene
Acetaldehyde
Ethylbenzene
Naphthalene
1,3 -Butadiene
Ethylene oxide
Dichloromethane
Tetrachloroethylene
Chloromethylbenzene
90.96
57.39
45.99
28.97
13.32
9.78
6.72
2.56
1.71
1.37
Formaldehyde
Ethylene oxide
Naphthalene
Benzene
Nickel, PM
1,3 -Butadiene
Hexavalent Chromium, PM
Arsenic, PM
Acetaldehyde
Ethylbenzene
1.18E-03
5.91E-04
4.53E-04
4.48E-04
3.50E-04
2.93E-04
1.98E-04
1.39E-04
1.01E-04
7.24E-05
to
-^
to
-------
Table 24-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Texas Monitoring Sites
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Noncancer Hazard
Approximation
Pollutant (HQ)
Deer Park, Texas (Harris County) - CAMS 35
Toluene
Ethylene glycol
Hexane
Methanol
Xylenes
Benzene
Formaldehyde
Methyl isobutyl ketone
Ethylbenzene
Acetaldehyde
8,992.02
5,139.15
3,952.83
2,806.38
2,282.67
1,003.70
609.51
609.01
600.25
358.08
Acrolein
1,3 -Butadiene
Chlorine
Titanium tetrachloride
Manganese, PM
Formaldehyde
Nickel, PM
Acetaldehyde
Cadmium, PM
Benzene
1,939,768.88
166,053.00
120,895.98
77,090.00
71,991.08
62,195.29
45,806.59
39,786.75
37,360.45
33,456.71
Hexavalent Chromium O.01
Karnack, Texas (Harrison County) - CAMS 85
Toluene
Ethylene glycol
Xylenes
Hexane
Formaldehyde
Benzene
Chloromethane
Methanol
Acetaldehyde
Glycol ethers, gas
237.26
159.89
109.93
98.06
90.96
57.39
48.40
46.62
45.99
30.73
Acrolein
Hexamethylene- 1 ,6-diisocyanate, gas
Manganese, PM
Chlorine
Formaldehyde
Cyanide Compounds, PM
Nickel, PM
Maleic anhydride
Acetaldehyde
1,3 -Butadiene
444,426.16
48,091.54
24,602.50
22,528.15
9,281.93
9,151.68
8,103.67
7,969.71
5,109.47
4,887.61
to
-^
to
-------
Observations from Table 24-7 include the following:
• Because Table 24-7 includes emissions data from the NEI, which is independent of
the sampling results at a specific site, data for Harrison County, where CAMS 85 is
located, is included.
• Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
cancer UREs in Harris County. Formaldehyde, benzene, and acetaldehyde are the
highest emitted pollutants with cancer UREs in Harrison County. The magnitude of
the emissions is significantly higher in Harris County than Harrison County.
• The pollutants with the highest toxi city-weighted emissions (of the pollutants with
cancer UREs) for Harris County are 1,3-butadiene, benzidine (gaseous), and
formaldehyde. The pollutants with the highest toxicity-weighted emissions for
Harrison County are formaldehyde, ethylene oxide, and naphthalene.
• Four of the highest emitted pollutants in Harris County also have the highest toxicity-
weighted emissions (1,3-butadiene, formaldehyde, benzene, and naphthalene).
• Formaldehyde tops both emissions-based lists for Harrison County. Another five of
the highest emitted pollutants in Harrison County also are among those with the
highest toxicity-weighted emissions.
• Naphthalene is the only pollutant of interest for CAMS 35 that appears on both
emissions-based lists for Harris County. The total emissions of naphthalene for Harris
County rank seventh while its toxicity-weighted emissions rank sixth.
• Although hexavalent chromium, the only pollutant for which a cancer risk
approximation could be calculated for CAMS 35, ranks fifth for its toxicity-weighted
emissions, this pollutant is not one of the highest emitted in Harris County (its
emissions rank 34th).
• Hexavalent chromium is the only pollutant sampled for at CAMS 85 under the NMP.
This pollutant has the seventh highest toxicity-weighted emissions for Harrison
County, but is not among the 10 highest emitted (its emissions rank 27th).
Observations from Table 24-8 include the following:
• Table 24-8 includes emissions data for Harrison County, similar to Table 24-7.
• Toluene and ethylene glycol are the highest emitted pollutants with noncancer RfCs
in both Harris and Harrison Counties. The magnitude of the emissions is significantly
higher for Harris County than Harrison County.
• The pollutant with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) for both counties is acrolein.
24-26
-------
• Three of the highest emitted pollutants also have the highest toxi city-weighted
emissions for Harris County (formaldehyde, acetaldehyde, and benzene) while only
two of the highest emitted pollutants (formaldehyde and acetaldehyde) also have the
highest toxicity-weighted emissions for Harrison County.
• Neither naphthalene nor hexavalent chromium appear on either emissions-based list
for Harris County, although naphthalene ranks 11th for its toxicity-weighted
emissions (of the pollutants with noncancer RfCs).
• Hexavalent chromium appears on neither emissions-based list for Harrison County
(ranking 58th for quantity emitted and 28th for its toxicity-weighted emissions).
24.6 Summary of the 2012 Monitoring Data for CAMS 35 and CAMS 85
Results from several of the data treatments described in this section include the
following:
»«» Naphthalene and hexavalent chromium failed at least one screen for CAMS 35, with
naphthalene accounting for 62 percent of the total failed screens, even though
sampling was discontinued in February.
»«» Hexavalent chromium, the only pollutant sampled for at CAMS 85, did not fail any
screens.
»«» The highest concentrations of hexavalent chromium were measured at CAMS 35
during the fourth quarter of 2012.
24-27
-------
25.0 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.
25.1 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 25-1 is a
composite satellite image retrieved from ArcGIS Explorer showing the monitoring site and its
immediate surroundings. Figure 25-2 identifies nearby point source emissions locations by
source category, as reported in the 2011 NEI for point sources. Note that only sources within
10 miles of the site are included in the facility counts provided in Figure 25-2. 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 radius are still visible on the map, but have been grayed out in order to show emissions
sources just outside the boundary. Table 25-1 provides supplemental geographical information
such as land use, location setting, and locational coordinates.
25-1
-------
Figure 25-1. Bountiful, Utah (BTUT) Monitoring Site
to
-------
Figure 25-2. NEI Point Sources Located Within 10 Miles of BTUT
Legend
111 °55'D-W 111c 50'0"W 111045'0"W 111 °4Q'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
BTUT NATTS site
O 10 mile radius
County boundary
Source Category Group (No. of Facilities)
•f Airport/Airline/Airport Support Operations (8) • Landfill (1)
tt Asphalt Production/Hot Mix Asphalt Plant (2) © Metals Processing/Fabrication (2)
B Bulk Terminals/Bulk Plants (1) ? Miscellaneous Commercial/Industrial (1)
* Electricity Generation via Combustion (2) 1) Paint and Coating Manufacturing (1)
+ Industrial Machinery or Equipment Plant (1) a Petroleum Refinery (5)
o Institution (school, hospital, prison, etc.) (1) x Rail Yard/Rail Line Operations (2)
25-3
-------
Table 25-1. Geographical Information for the Utah Monitoring Site
Site
Code
BTUT
AQS Code
49-011-0004
Location
Bountiful
County
Davis
Micro- or
Metropolitan
Statistical Area
Ogden-Clearfield,
UTMSA
Latitude
and
Longitude
40.902967,
-111.884467
Land Use
Residential
Location
Setting
Suburban
Additional Ambient Monitoring Information1
SO2, NO, NO2, NOX, O3, Meteorological parameters,
PM10, PM25, PM25 Speciation, Black Carbon,
IMPROVE Speciation.
BOLD ITALICS = EPA-designated NATTS Site
to
-k
-------
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 25-1 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 25-2 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, which includes airports and related operations as well as small
runways and heliports, such as those associated with hospitals or television stations, and
petroleum refineries. Point sources within 2 miles of BTUT include a metals
processing/fabrication facility, a facility generating electricity via combustion, a petroleum
refinery, and a painting and coatings manufacturer.
Table 25-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Utah monitoring site. Table 25-2 includes both county-level
population and vehicle registration information. Table 25-2 also contains traffic volume
information for BTUT as well as the location for which the traffic volume was obtained.
Additionally, Table 25-2 presents the county-level daily VMT for Davis County.
Table 25-2. Population, Motor Vehicle, and Traffic Information for the Utah Monitoring
Site
Site
BTUT
Estimated
County
Population1
315,809
County-level
Vehicle
Registration2
259,319
Annual
Average
Daily Traffic3
129,145
Intersection
Used for
Traffic Data
1-15, north of Hwy-89 junction
County-
level
Daily
VMT4
6,866,779
Bounty-level population estimate reflects 2012 data (Census Bureau, 2013c)
2County-level vehicle registration reflects 2012 data (UT TC, 2012)
3AADT reflects 2011 data (UT DOT, 2011)
4County-level VMT reflects 2011 data (UT DOT, 2012)
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 25-2 include the following:
• Davis County's population is in the middle of the range compared to other counties
with NMP sites. The county-level vehicle registration ranking is similar to the
population ranking.
25-5
-------
• The traffic volume experienced near BTUT is 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.
• The daily VMT for Davis County is in the middle of the range compared to other
counties with NMP sites (where VMT was available).
25.2 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.
25.2.1 Climate Summary
The Salt Lake City area's climate can be described as semi-arid and continental with
considerable seasonal variations. Summers are hot and dry while winters are cold and snow is
common. The area is generally dry, though, and sunshine prevails across the area during much of
the year. Most months average less than 2 inches of precipitation, with spring as the wettest
season. Precipitation that does fall can be enhanced over the eastern parts of the valley as storm
systems move up the side of the Wasatch Mountains, located to the east. Smaller mountain
ranges to the southwest and south protect the valley from winter storm systems moving in from
the southwest. The Great Salt Lake has a moderating influence on the area's temperature, as the
lake never freezes, and can enhance precipitation from storm systems that move over the lake.
Moderate winds flow out of the southeast on average, although there is a valley breeze/lake
breeze system that affects the area. High pressure systems that occasionally settle over the area
can result in stagnation episodes (Wood 2004; WRCC, 2013).
25.2.2 Meteorological Summary
Hourly meteorological data for 2012 were retrieved from NCDC for the weather station
closest to the Utah monitoring site (NCDC, 2012), as described in Section 3.5.2. The closest
weather station to BTUT is located at Salt Lake City International Airport (WBAN 24127).
Additional information about the Salt Lake City International Airport weather station, such as the
distance between the site and the weather station, is provided in Table 25-3. These data were
used to determine how meteorological conditions on sample days vary from conditions
experienced throughout the year.
25-6
-------
Table 25-3. Average Meteorological Conditions near the Utah Monitoring Site
Closest Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
<°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
<°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Bountiful, Utah - BTUT
Salt Lake City
International
24127
(40.79, -111.97)
9.6
miles
216°
(SW)
Sample
Days
(71)
2012
65.9
±4.5
67.1
+ 2.1
55.5
±4.2
56.5
+ 1.9
32.1
±2.2
32.7
+ 1.0
43.8
±2.7
44.4
+ 1.2
47.8
±4.3
47.7
+ 1.9
1015.5
± 1.8
1014.6
+ 0.8
7.1
±0.7
7.2
+ 0.3
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
to
-------
Table 25-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for all of 2012. Also included in Table 25-3 is the 95 percent
confidence interval for each parameter. As shown in Table 25-3, average meteorological
conditions on sample days near BTUT were representative of average weather conditions
experienced throughout the year.
25.2.3 Back Trajectory Analysis
Figure 25-3 is the composite back trajectory map for days on which samples were
collected at the BTUT monitoring site. Included in Figure 25-3 are four back trajectories per
sample day. Figure 25-4 is the corresponding cluster analysis. An in-depth description of these
maps and how they were generated is presented in Section 3.5.2.1. For the composite map, each
line represents the 24-hour trajectory along which a parcel of air traveled toward the monitoring
site on a given sample day and time, based on an initial height of 50 meters AGL. For the cluster
analysis, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric circle around the site in Figures 25-3 and 25-4 represents 100 miles.
Figure 25-3. Composite Back Trajectory Map for BTUT
25-8
-------
Figure 25-4. Back Trajectory Cluster Map for BTUT
Observations from Figures 25-3 and 25-4 include the following:
• Back trajectories originated from a variety of directions at BTUT. Back trajectories
often originated from the northwest and south of the site. Back trajectories originating
from a direction with a westerly component tended to be longer than those originating
from a direction with an easterly component.
• Similar to other sites located in the inter-mountain west, the 24-hour air shed domain
for BTUT is smaller in size than many other NMP sites. The farthest away a back
trajectory originated was over the south-central Oregon, or nearly 450 miles away.
However, the average back trajectory length was 162 miles and nearly 90 percent of
back trajectories originated within 300 miles of the site.
• The cluster analysis shows that 30 percent of back trajectories are represented by the
short cluster trajectory originating just south of the site. This cluster represents back
trajectories originating primarily to the south and within roughly 150 miles of BTUT
and those looping over the northern half of Utah. Nearly one-quarter of back
trajectories originated to the west and northwest of BTUT, although of varying
lengths, as indicated by the shorter cluster trajectory (17 percent), which represents
back trajectories originating over the southern half of Idaho and the northeast corner
of Nevada, and the longer cluster trajectory (6 percent), which represents longer back
trajectories originating over northern Nevada and southeast Oregon. Five percent of
back trajectories originated to southwest of BTUT, but are varied in length. Twenty
percent of back trajectories originated to the south of BTUT, over the southern half of
Utah and northern half of Arizona. Another 9 percent of back trajectories originated
25-9
-------
to the southeast of BTUT. The final 13 percent of back trajectories originated to the
east and northeast of the site, and include short back trajectories originating over the
Wasatch Mountains as well as longer back trajectories originating over central
Wyoming.
25.2.4 Wind Rose Comparison
Hourly surface wind data from the weather station at Salt Lake City International Airport
near BTUT were uploaded into a wind rose software program to produce customized wind roses,
as described in Section 3.5.2.2. 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 25-5 presents a map showing the distance between the weather station and BTUT,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 25-5 also presents three different wind roses for the
BTUT monitoring site. First, a historical wind rose representing 2002 to 2011 wind data is
presented, which shows the predominant surface wind speed and direction over an extended
period of time. Second, a wind rose representing wind observations for all of 2012 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2012 is
presented. These can be used to identify the predominant wind speed and direction for 2012 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.
Observations from Figure 25-5 for BTUT include the following:
• The Salt Lake City International Airport weather station is located 9.6 miles
southwest of BTUT.
• The historical wind rose shows that southeasterly, south-southeasterly, and southerly
winds were prevalent near BTUT, accounting for more than 40 percent of the wind
observations. Winds from the north-northwest and north were also common. Winds
from the northeast and southwest quadrants were rarely observed. Calm winds
(<2 knots) were observed for approximately 12 percent of the hourly measurements.
The strongest wind speeds were observed with south-southeasterly and southerly
winds.
• The wind patterns shown on the 2012 wind rose are similar to the historical wind
patterns, indicating that wind conditions in 2012 were similar to wind conditions
experienced historically near BTUT. This is also true for the sample day wind rose.
25-10
-------
Figure 25-5. Wind Roses for the Salt Lake City International Airport Weather Station near
BTUT
Location of BTUT and Weather Station
2002-2011 Historical Wind Rose
2012 Wind Rose
Sample Day Wind Rose
25-11
-------
25.3 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 25-4.
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 25-4. It is
important to note which pollutants each site sampled for when reviewing the results of this
analysis. BTUT sampled for VOCs, carbonyl compounds, SNMOCs, PAHs, metals (PMi0), and
hexavalent chromium and is one of only two sites sampling the entire suite of pollutants under
the NMP (NBIL is the other).
Table 25-4. Risk-Based Screening Results for the Utah Monitoring Site
Pollutant
Screening
Value
(Ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Bountiful, Utah - BTUT
Benzene
Carbon Tetrachloride
Acetaldehyde
Formaldehyde
1,3 -Butadiene
1 ,2-Dichloroethane
Arsenic (PM10)
Naphthalene
Manganese (PM10)
Ethylbenzene
Propionaldehyde
Dichloromethane
Nickel (PM10)
£>-Dichlorobenzene
1, 1,2,2-Tetrachloroethane
Hexachloro- 1 ,3 -butadiene
1 ,2-Dibromoethane
Benzo(a)pyrene
Chloroprene
Lead (PM10)
Trichloroethylene
0.13
0.17
0.45
0.077
0.03
0.038
0.00023
0.029
0.005
0.4
0.8
7.7
0.0021
0.091
0.017
0.045
0.0017
0.00057
0.0021
0.015
0.2
Total
56
56
54
54
52
49
45
41
37
21
10
7
7
5
5
3
2
1
1
1
1
508
56
56
54
54
55
49
55
59
57
56
54
56
57
24
5
o
6
2
16
1
57
12
838
100.00
100.00
100.00
100.00
94.55
100.00
81.82
69.49
64.91
37.50
18.52
12.50
12.28
20.83
100.00
100.00
100.00
6.25
100.00
1.75
8.33
60.62
11.02
11.02
10.63
10.63
10.24
9.65
8.86
8.07
7.28
4.13
1.97
1.38
1.38
0.98
0.98
0.59
0.39
0.20
0.20
0.20
0.20
11.02
22.05
32.68
43.31
53.54
63.19
72.05
80.12
87.40
91.54
93.50
94.88
96.26
97.24
98.23
98.82
99.21
99.41
99.61
99.80
100.00
25-12
-------
Observations from Table 25-4 include the following:
• Twenty-one pollutants failed at least one screen for BTUT; nearly 61 percent of
concentrations for these 21 pollutants were greater than their associated risk screening
value (or failed screens).
• Thirteen pollutants contributed to 95 percent of failed screens for BTUT and therefore
were identified as pollutants of interest for this site. These 13 include three carbonyl
compounds, six VOCs, three PMi0 metals, and one PAH.
• Acetaldehyde, benzene, carbon tetrachloride, and formaldehyde 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 much 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.
25.4 Concentrations
This section presents various concentration averages used to characterize pollution 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 the
monitoring site.
• Annual concentration averages are presented graphically for BTUT 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 the site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Additional site-specific statistical summaries for BTUT
are provided in Appendix J through Appendix O.
25.4.1 2012 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 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
25-13
-------
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 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 25-5, where applicable. Note that concentrations of the PAHs and PMi0
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 25-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Utah Monitoring Site
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Hg/m3)
2nd
Quarter
Average
(Ug/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Ug/m3)
Bountiful, Utah - BTUT
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
Dichloromethane
Ethylbenzene
Formaldehyde
Propionaldehyde
Arsenic (PM10)a
Manganese (PM10)a
Naphthalene3
Nickel (PM10)a
54/54
56/56
55/56
56/56
49/56
56/56
56/56
54/54
54/54
55/57
57/57
59/59
57/57
1.71
±0.46
1.10
±0.38
0.11
±0.05
0.62
±0.05
0.09
±0.01
1.22
±0.68
0.36
±0.17
2.26
±0.34
0.34
±0.09
0.47
±0.30
5.77
±2.01
51.84
±21.35
1.39
±0.52
1.91
±0.33
0.84
±0.19
0.06
±0.01
0.68
±0.07
0.09
±0.01
0.57
±0.15
0.32
±0.08
2.90
±0.39
0.47
±0.09
0.33
±0.12
9.09
±2.73
34.29
± 10.66
1.53
±0.43
3.05
±0.78
0.99
±0.25
0.10
±0.03
0.68
±0.06
0.06
±0.02
10.79
±9.56
0.36
±0.06
5.65
±1.46
0.65
±0.13
0.51
±0.11
11.14
±2.70
48.54
±8.71
1.53
±0.36
NA
1.14
±0.24
0.21
±0.06
0.67
±0.03
0.06
±0.03
20.85
± 26.90
0.42
±0.11
NA
NA
0.60
±0.21
6.01
± 1.84
64.56
±20.13
1.20
±0.25
2.54
±0.35
1.02
±0.13
0.12
±0.02
0.66
±0.03
0.08
±0.01
7.82
±6.53
0.36
±0.06
4.44
±0.75
0.55
±0.07
0.48
±0.10
7.97
± 1.24
49.56
±8.02
1.41
±0.19
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
25-14
-------
Observations for BTUT from Table 25-5 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 in
2012, but is considerably less than the annual average for 2011. Th annual average
concentration for 2012 has a very large confidence interval associated it, indicating
the likely presence of outliers. A review of the quarterly averages shows that the third
and fourth quarter average concentrations are significantly higher than the other two
quarterly averages and that their confidence intervals are nearly as high (third quarter)
or higher (fourth quarter) than the averages themselves. Concentrations of
dichloromethane measured at BTUT in 2012 range from 0.244 |ig/m3to 153 |ig/m3.
The maximum concentration of this pollutant was measured on December 5, 2012
and is the only one greater than 100 |ig/m3. However, two additional concentrations
are greater than 50 |ig/m3 were also measured at BTUT. Of the 22 dichloromethane
concentrations greater than 10 |ig/m3 measured across the program, six were
measured at BTUT. However, the median concentration of dichloromethane for
BTUT is 0.80 |ig/m3, as over half of the measurements are less than 1 |ig/m3.
• There are no fourth quarter average concentrations for the carbonyl compounds
because sampler issues during this quarter resulted in fewer valid samples than the
75 percent criteria. However, the maximum formaldehyde concentration measured at
BTUT was measured on December 5, 2012 (12.6 |ig/m3), the same day as the
maximum dichloromethane measurement. This formaldehyde concentration is the
third highest formaldehyde concentration measured across the program.
• The third quarter average formaldehyde concentration is significantly greater than the
first and second quarter averages. This is also true for acetaldehyde. Although no
fourth quarter average is provided, a review of the data shows that this trend likely
carries into the fourth quarter. The 18 highest formaldehyde concentrations measured
at BTUT (those greater than or equal to 4.5 |ig/m3) were all measured in the third and
fourth quarters of 2012. Conversely, all but one of the 23 formaldehyde
concentrations less than 3 |ig/m3 were measured during the first two quarters of 2012.
For acetaldehyde, all but one of the 11 concentrations greater than 4 |ig/m3 were
measured during the third and fourth quarters while all 14 concentrations less than
1.6 |ig/m3 were measured during the first and second quarters. The difference among
the quarterly averages is less significant for propionaldehyde, but the data shows a
similar pattern with the higher concentrations measured during the second half of the
year and lower concentrations measured during the first half of the year.
• Although the first quarter averages for benzene and ethylbenzene are slightly less
than the fourth quarter averages, the first quarter averages have higher confidence
intervals. A review of the data shows that the maximum concentrations of each of
these pollutants were measured at BTUT on the first and second sample days of the
year (January 4, 2012 and January 10, 2012). For benzene, these represent two of
three concentrations greater than 2 |ig/m3 measured at BTUT. For ethylbenzene, the
January 4th measurement (1.29 |ig/m3) is the only concentration greater than 1 |ig/m3
25-15
-------
measured at BTUT; the January 10th measurement (0.89 |ig/m3) is the only other
concentrations greater than 0.70 |ig/m3 measured at BTUT.
• Concentrations of 1,3-butadiene appear highest during the fourth quarter of 2012. A
review of the data shows that eight of the 11 concentrations greater than 0.2 |ig/m3
were measured at BTUT during the fourth quarter. Of the three measured outside the
fourth quarter, one was measured on January 4th and one was measured on
January 10th, the same days the highest benzene and ethylbenzene concentrations
were measured. The 1,3-butadiene concentrations for these days are the second and
fourth highest 1,3-butadiene concentrations, respectively, measured at BTUT.
• Concentrations of naphthalene appear highest during the colder months of the year,
although the confidence intervals shown for the first and fourth quarter averages
indicate that there is a considerable amount of variability associated with these
measurements. Concentrations of naphthalene measured at BTUT range from
5.50 ng/m3to 142 ng/m3. The maximum concentration of naphthalene was also
measured on January 4, 2012. The five concentrations of naphthalene greater than
100 ng/m3 were measured in January, November, or December. Of the 16
concentrations greater than 60 ng/m3 measured at BTUT, six were measured during
the first quarter and six were measured during the fourth quarter (with one measured
during the second and three measured during the third).
• Of the PMio metals identified as pollutants of interest for BTUT, manganese has the
highest annual average concentration. Manganese concentrations appear to be higher
during the warmer months of the year, although the confidence intervals calculated
for the quarterly averages indicate considerable variability in the individual
measurements. The maximum manganese concentration was measured at BTUT on
July 20, 2012 (24.78 ng/m3). The 10 highest manganese concentrations measured at
BTUT were all measured between May and September while 10 of the 12 lowest
manganese concentrations were measured during the first or fourth quarters of the
year.
• The first quarter average arsenic concentration has a relatively large confidence
interval associated with it. A review of the data shows that the two highest arsenic
concentrations measured at BTUT were measured on March 10, 2012 (1.87 ng/m3)
and January 10, 2012 (1.54 ng/m3). The next highest concentration measured during
the first quarter is one-third as high (0.55 ng/m3). In addition, the two lowest
measured detections of arsenic were also measured during the first quarter, explaining
the relatively high level of variability shown for this quarter.
Tables 4-9 through 4-12 present the 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 eight times for the program-level
pollutants of interest.
25-16
-------
• BTUT is listed for several of the program-level VOC pollutants of interest shown in
Table 4-9. BTUT ranks highest for 1,2-dichloroethane, ranking fourth among other
NMP sites sampling this pollutant.
• For the second 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. BTUT also has the sixth highest annual average concentration of
acetaldehyde.
• BTUT does not appear in Table 4-11 for PAHs. This site's annual average
concentrations of the PAHs are among the lower averages for sites sampling PAHs.
• BTUT ranks sixth for both manganese and nickel, as shown in Table 4-12. BTUT
does not appear in Table 4-12 for arsenic.
25.4.2 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 shaded in
gray in Table 25-4 for BTUT. Figures 25-6 through 25-18 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 3.5.3.1.
Figure 25-6. Program vs. Site-Specific Average Acetaldehyde Concentration
BTUT
9 12
Concentration {[og/m3)
15
18
21
Program:
Site:
1st Quartile
D
Site Average
o
2nd Quartile 3rd Quartile
D D
Site Concentration Range
^^^^—
4th Quartile Average
D 1
25-17
-------
Figure 25-7. Program vs. Site-Specific Average Arsenic (PMi0) Concentration
BTUT
B
, 1
3 1
1 1 1
2345678
Concentration (ng/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile
D D D D
Site:
Site Average Site Concentration Range
o —
Average
Figure 25-8. Program vs. Site-Specific Average Benzene Concentration
BTUT
Concentration {[og/m3;
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o —
Figure 25-9. Program vs. Site-Specific Average 1,3-Butadiene Concentration
Program Max Concentration = 4.10 ug/m3
0.75 1 1.25
Concentration (pg/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
O
25-18
-------
Figure 25-10. Program vs. Site-Specific Average Carbon Tetrachloride Concentration
BTUT
2 3
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o
Figure 25-11. Program vs. Site-Specific Average 1,2-Dichloroethane Concentration
-
r.
Program Max Concentration = 17.01 ug/m3
0.4 0.5 0.6
Concentration {[og/m3)
0.8 0.9
Program:
Site:
IstQuartile
D
Site Average
0
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
Figure 25-12. Program vs. Site-Specific Average Dichloromethane Concentration
BTUT
1
Prog
ram Max Concentration = 745 jig/m3
10
20
30 40
Concentration {[og/m3)
50
60
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o —
70
25-19
-------
Figure 25-13. Program vs. Site-Specific Average Ethylbenzene Concentration
BTUT
fit
1.5 2 2.5
Concentration (pg/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o
Figure 25-14. Program vs. Site-Specific Average Formaldehyde Concentration
10
12
Concentration {[og/m3;
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o —
14
Figure 25-15. Program vs. Site-Specific Average Manganese (PMi0) Concentration
H
Program Max Concentration = 275 ng/m3
60 90
Concentration (ng/m3)
Program:
Site:
IstQuartile
D
Site Average
O
2ndQuartile SrdQuartile
D D
Site Concentration Range
4thQuartile Average
D 1
25-20
-------
Figure 25-16. Program vs. Site-Specific Average Naphthalene Concentration
BTUT
irt
100
200
300
400 500
Concentration (ng/m3)
600
700
800
900
Program:
Site:
IstQuartile
D
Site Average
0
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
Figure 25-17. Program vs. Site-Specific Average Nickel (PMio) Concentration
8 10
Concentration (ng/m3)
Program:
Site:
IstQuartile
Site Average
o
2ndQuartile SrdQuartile
D D
Site Concentration Range
4thQuartile Average
Figure 25-18. Program vs. Site-Specific Average Propionaldehyde Concentration
BTUT
1 1.5
Concentration {[og/m3)
Program:
Site:
IstQuartile
D
Site Average
0
2ndQuartile SrdQuartile
D D
Site Concentration Range
4thQuartile Average
D 1
25-21
-------
Observations from Figures 25-6 through 25-18 include the following:
• Figure 25-6 shows that the annual average acetaldehyde concentration for BTUT
is greater than the program-level average concentration as well as the program-
level third quartile. The maximum acetaldehyde concentration measured at BTUT
is considerably less than the maximum acetaldehyde concentration measured at
the program-level. There were no non-detects of acetaldehyde measured at BTUT
or across the program.
• Figure 25-7 shows that BTUT's annual average arsenic concentration is less than
the program-level average as well as the program-level median concentration.
Concentrations of arsenic measured at BTUT range from non-detect to less than
2 ng/m3.
• Figure 25-8 shows that the annual average benzene concentration for BTUT is
greater than the program-level average concentration and similar to the program-
level third quartile, although the maximum concentration measured at BTUT is
roughly half the maximum concentration measured at the program level. There
were no non-detects of benzene measured at BTUT or across the program.
• Figure 25-9 the box plot for 1,3-butadiene. Note that the program-level maximum
1,3-butadiene concentration (4.10 |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 has been reduced to
2 |ig/m3. Figure 25-9 shows that the annual average concentration for BTUT is
just greater than the program-level average concentration. The maximum
concentration of benzene measured at BTUT is an order of magnitude less than
the maximum concentration measured across the program. There was a single
non-detect of 1,3-butadiene measured at BTUT.
• Figure 25-10 shows that the annual average concentration of carbon tetrachloride
for BTUT is just less than the program-level average concentration and the
program-level median concentration. The range of carbon tetrachloride
concentrations for BTUT spans roughly 0.5 |ig/m3, with the maximum
concentration less than 1 |ig/m3. There were no non-detects of carbon
tetrachloride measured at BTUT or across the program.
• Figure 25-11 is the box plot for 1,2-dichloroethane. Note that the program-level
maximum concentration (17.01 |ig/m3) is not shown directly on the box plot as
the scale has been reduced to 1 |ig/m3 in order to allow for the observation of data
points at the lower end of the concentration range. The program-level average
concentration is greater than the program third quartile for this pollutant and is
greater than or similar to the maximum concentration measured at most sites
sampling 1,2-dichloroethane. This is because the program-level average is being
driven by the higher measurements collected at a few monitoring sites.
Figure 25-11 shows that the maximum 1,2-dichloroethane concentration
measured at BTUT is two orders of magnitude less than the maximum
concentration measured across the program. The annual average for BTUT is
similar to the median concentration at the program level. The maximum
25-22
-------
1,2-dichloroethane concentration measured at BTUT is less than the program-
level average concentration. Seven non-detects of 1,2-dichloroethane were
measured at BTUT.
• Similar to other pollutants, the program-level maximum concentration of
dichloromethane (745 |ig/m3) is not shown directly on the box plot in
Figure 25-12 to allow for the observation of data points at the lower end of the
concentration range; thus, as the scale has been reduced to 70 |ig/m3, although
reducing the scale by an order of magnitude still does not allow for the first three
quartiles to be readily viewed. This is a result of a few measurements at the upper
end of the concentrations range driving the data. Two dichloromethane
concentrations measured at BTUT are greater than the top of the scale in
Figure 25-12. The maximum concentration measured at BTUT (153 |ig/m3) is the
sixth highest concentration program-wide, but is still considerably less than the
maximum dichloromethane concentration measured across the program. BTUT's
annual average dichloromethane concentration is roughly three times greater than
the program-level average concentration. BTUT has the second highest annual
average concentration of dichloromethane among sites sampling this pollutant,
behind only GPCO.
• Figure 25-13 shows that the annual average ethylbenzene concentration for BTUT
is similar to the program-level average concentration. The maximum
ethylbenzene concentration measured at BTUT is less than the maximum
concentration measured across the program. There were no non-detects of
ethylbenzene measured at BTUT.
• Figure 25-14 shows that the range of formaldehyde concentrations measured at
BTUT is large and that the maximum concentration measured at BTUT is just less
than the maximum formaldehyde concentration measured across the program. The
annual average formaldehyde concentration for BTUT is greater than both the
program-level average and third quartile. As discussed in the previous section,
BTUT has the highest annual average formaldehyde concentration among NMP
sites sampling carbonyl compounds.
• Figure 25-15 is the box plot for manganese (PMio). The program-level maximum
manganese concentration (275 ng/m3) is not shown directly on the box plot as the
scale has been reduced to 150 ng/m3 in order to allow for the observation of data
points at the lower end of the concentration range. Figure 25-15 shows that the
annual average concentration of manganese (PMio) for BTUT is less than the
program-level average concentration. The maximum concentration measured at
BTUT is an order of magnitude less than the program-level maximum
concentration. There were no non-detects of manganese measured at BTUT or
across the program.
• Figure 25-16 is the box plot for naphthalene, which shows that the annual average
naphthalene concentration for BTUT is less than both the program-level average
and median concentrations. The annual average concentration of naphthalene for
BTUT ranks 16th among the 20 sites for which annual averages could be
25-23
-------
calculated. The maximum naphthalene concentration measured at BTUT is
considerably less than the program-level maximum concentration. There were no
non-detects of naphthalene measured at BTUT or across the program.
• Figure 25-17 is the box plot for nickel (PMio). The maximum concentration of
nickel measured at BTUT is considerably less than the program-level maximum
concentration. The annual average concentration of nickel for BTUT is greater
than the program-level average concentration and similar to the program-level
third quartile. The minimum concentration of nickel measured at BTUT is just
less than the program-level first quartile.
• Figure 25-18 shows that the annual average concentration of propionaldehyde for
BTUT is greater than the program-level average concentration and third quartile.
This site has the highest annual average concentration of this pollutant among
NMP sites sampling carbonyl compounds. The maximum propionaldehyde
concentration was not measured at BTUT. There minimum concentration of this
pollutant measured at BTUT is just less than the program-level first quartile.
25.4.3 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 3.5.3.2.
BTUT has sampled carbonyl compounds, VOCs, metals, and SNMOCs under the NMP since
2003. Thus, Figures 25-19 through 25-30 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 is not
provided, although the range and quartiles are still presented. Because sampling for PAHs did
not begin in earnest at BTUT until late 2008, a trend analysis was not performed for naphthalene.
25-24
-------
Figure 25-19. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at BTUT
r^
2007 2008 2009
Year
• BthPercentile
— Minimum
— Maximum
95thPercentile
Observations from Figure 25-19 for acetaldehyde measurements collected 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 25-19 excludes
data from 2003.
• The maximum acetaldehyde concentration was measured in 2004 (32.7 |ig/m3). The
next highest concentrations of acetaldehyde were measured at BTUT in 2008
(20.0 |ig/m3) and 2007 (15.3 |ig/m3). No acetaldehyde concentrations greater than
8 |ig/m3 have been measured at BTUT since 2008.
• The 1-year average concentration exhibits a steady decreasing trend beginning with
2006 and continuing through 2009, after which the 1-year average concentration
changes little, ranging from 1.97 |ig/m3 (2009) to 2.54 |ig/m3 (2012).
• The range within which the majority of concentrations fall, as indicated by the
difference between the 5th and 95th percentiles, decreased steadily through 2008,
where it reached a minimum. This range then increased for 2009, an increasing trend
that continued through 2011, after which a slight decrease is shown for 2012. This is
due to a slight increase in the concentrations at the lower end of the concentration
range for 2012.
25-25
-------
Figure 25-20. Yearly Statistical Metrics for Arsenic (PM10) Concentrations Measured at BTUT
f
§
S 10l°
1
0.0 -
Maximum
Concentration for
2004 is 33.0 ng/m3. u
••
O.
^H
, ,11 1 T
J)-, "^ -^ I
••vv ° .<> O r~«-|
• IUK •• i^ O O -A
^~ , ~"~ , '-t-1 , '-•-1 , l=i=l , !_•_! , =•= , 1-5-1 , ^p ,
2004 2005 2006 2007 2008 2009 2010 2011 2012
Year
O 5th Percentile — Minimum — Median — Maximum O 95th Percentile ...<>... Average
Observations from Figure 25-20 for arsenic measurements collected at BTUT include the
following:
• Sampling for PMi0 metals under the NMP began at BTUT in late July 2003. Because
this represents less than half of the sampling year, Figure 25-20 excludes data from
2003.
• The maximum arsenic concentration was also measured at BTUT in 2004
(33.0 ng/m3) and is nearly twice the next highest concentration (16.8 ng/m3), also
measured in 2004. The three highest measurements of arsenic were all measured at
BTUT in 2004; further, eight of the 11 highest concentrations of arsenic (those
greater than 5 ng/m3) were measured in 2004.
• Of the 20 highest arsenic concentrations measured at BTUT, 12 were measured
during the first quarter of the calendar year and eight were 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) to 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 set, the
25-26
-------
1-year average concentrations for this period would all be less than 1 ng/m3. The
maximum concentrations for 2007 and 2009 were both measured in January.
• A slight decreasing trend is shown in the arsenic concentrations measured between
2009 and 2012. The 1-year average concentration is at a minimum for 2012. The
maximum arsenic concentration measured in 2012 is less than 2 ng/m3, the only year
for which this is true.
Figure 25-21. Yearly Statistical Metrics for Benzene Concentrations Measured at BTUT
5th Percent!le
95th Percent!le
....0... Average
Observations from Figures 25-21 for benzene measurements collected 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 25-21 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.
• Concentrations of benzene appear to be higher during the colder months of the year,
as all but one of the 40 highest concentrations (those greater than 2.75 |ig/m3) were
measured during the first or fourth 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 is then shown through 2009,
25-27
-------
after which another decreasing trend follows. The 1-year average benzene
concentrations for each year fall between 1 |ig/m3 and 2 |ig/m3, with the 1-year
average concentration at a minimum for 2012 (1.02 |ig/m3).
• Although the 1-year average concentration increased for 2009, the median
concentration decreased. The difference between these two parameters is highest for
2009, a reflection of increased variability in the measurements. The 1-year average is
being driven by a few higher concentrations measured in 2009, as discussed above.
Figure 25-22. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at BTUT
I ™
£
O
2008
Year
• 5th Percentile
— Minimum
— Maximum
95thPercentile
Observations from Figure 25-22 for 1,3-butadiene measurements collected 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 measurement was also collected 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 the 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
concentration 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
25-28
-------
2011, explaining why the 5th percentile returned to zero. For 2012, there was a single
non-detect of this pollutant.
The 1-year average concentration increased from 0.061 |ig/m3 for 2004 to
0.104 |ig/m3 for 2005. This increase is likely due to the decrease in non-detects (and
thus zeros substituted for them) as well as the higher concentrations measured in
2005, as discussed above. Between 2005 and 2012, the 1-year average concentration
has changed little, ranging from 0.099 |ig/m3 (2008, 2011) to 0.118 |ig/m3 (2012).
Figure 25-23. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
BTUT
~ 0.6
o
T
T
2008
Year
• 5th Percentile
95th Percentile
Observations from Figures 25-23 for carbon tetrachloride measurements collected 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 jig/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 of concentrations measured
decreased substantially for 2007, which is reflected in the dip in the 1-year average
concentration. A slight decreasing trend in the carbon tetrachloride measurements is
shown between 2008 and 2010, after which an increasing trend is shown through the
end of the sampling period.
25-29
-------
• Although the overall range within which most of the concentrations fall is decreasing,
the central tendency of the measurements has increased since the onset of sampling.
Figure 25-24. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
BTUT
I
.O
2008
Year
5th Percentile — Minimuir
Median — Maximurr
95th Percentile
Observations from Figures 25-24 for 1,2-dichloroethane measurements collected at
BTUT include the following:
• For the first several years of sampling, all of the statistical parameters shown were
zero. 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, and 47 in 2012. This explains the increases shown in the 1-year average
concentrations for 2010, 2011, and 2012.
• 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. The
median concentration is actually greater than the annual average concentration for
2012. This is because there were still seven non-detects (or zeros) factoring into the
average concentration for the year.
25-30
-------
Figure 25-25. Yearly Statistical Metrics for Dichloromethane Concentrations Measured at
BTUT
Concentration (jig/m3)
Maximum
Concentrationfor
2010 is 2,430 u.g/m3.
2004 2005 2006 2007 2008
Year
O 5th Percentile — Minimum — Median
-
Maxirr
fr
2009
um
2010
O 95th Percentile
i
* ±
2011 2012
••••O-" Average
Observations from Figures 25-25 for dichloromethane measurements collected at BTUT
include the following:
• Prior to 2008, the maximum concentration of dichloromethane measured at BTUT
was 1.64 |ig/m3 (in 2005). However, due to the scale on the graph, none of the
statistical parameters for the early years are visible.
• Beginning in 2008, "higher" concentrations of dichloromethane began to be measured
at BTUT. In 2008, concentrations of 33 |ig/m3 and 203 |ig/m3 were measured, both in
November. In 2009, four concentrations greater than 100 |ig/m3 and five
concentrations between 20 |ig/m3 and 80 |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, there was only one concentration greater
than 1,000 |ig/m3 measured, although two more greater than 500 |ig/m3 were also
measured. The maximum concentration for 2012 (152 |ig/m3) is considerably less, but
still greater than 100 |ig/m3.
• There does not appear to be a pattern in the time of year that these higher
measurements are collected. Of the 20 concentrations measured at BTUT greater than
100 |ig/m3, at least one has been measured in each month of the year except March,
April, and May. There is a 3-way tie for month with the greatest number of these
higher measurements: January, September, and December.
25-31
-------
• Even with these measurements, the median concentration for each year is less than
4 |ig/m3 and is less than 1 |ig/m3 for most years of sampling.
Figure 25-26. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at BTUT
I
•On
2007 2008 2009
Year
5th Percent!le
95thPercentile —O—Average
Observations from Figures 25-26 for ethylbenzene measurements collected at BTUT
include the following:
• The maximum concentration of ethylbenzene measured at BTUT was measured in
2006 (4.87 |ig/m3), although concentrations greater than 4 |ig/m3 were also measured
in 2004 and 2010. Only one additional concentration greater than 2 |ig/m3 has been
measured at BTUT (3.89 |ig/m3 in 2011).
• A steady decreasing trend in the 1-year average concentration is shown from 2004
through 2007, representing just less than a 50 percent decrease (from 0.70 |ig/m3 for
2004 to 0.39 |ig/m3 for 2007). However, most of the change is realized between 2004
and 2006.
• Between 2007 and 2009, little change is shown, 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, it is this concentration driving most of the
increases shown, as removing the maximum concentration 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.
25-32
-------
• The range of ethylbenzene concentrations measured in 2012 is the smallest among the
years of sampling. The 1-year average concentration is also at a minimum for 2012.
Figure 25-27. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at BTUT
1
° 250
re
1
5.0 -
— <
— <
— i
20
0 _
I r^ }
h rl i i i h i
T 1 +
n LJ D y fg y y g y
04 2005 2006 2007 2008 2009 2010 2011 2012
Year
thPercentile — Minimum — Median — Maximum O 95th Percentile '"O"- Average
Observations from Figure 25-27 for formaldehyde measurements collected 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 in 2004,
2005, 2006, 2007, and 2011.
• Although the maximum concentration decreased significantly from 2004 to 2005, the
other statistical metrics exhibit increases. 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). As an illustration, there were 11
concentrations greater than 5 |ig/m3 measured in 2004 compared to 31 in 2005.
• After 2005, the 1-year average concentration began to decrease, reaching a minimum
for 2008. After 2008, a steady increasing trend is shown in the 1-year average
formaldehyde concentrations, as well as most other statistical parameters. This trend,
however, levels out for 2012.
25-33
-------
• Although little change is shown in the 1-year average between 2011 and 2012, the
range of concentrations measured is smaller for 2012 and the median actually exhibits
an increase. The decrease in the concentrations in the upper end of the range from
2011 to 2012 are balanced out by a higher number of measurements in 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. In
addition, there are six concentrations measured in 2011 that are less than the
minimum concentration measured in 2012; thus, the concentrations at the lower end
of the concentration range increased for 2012.
Figure 25-28. Yearly Statistical Metrics for Manganese (PMio) Concentrations Measured at
BTUT
i
§
s
o
2007 2008 2009
Year
O BthPercentile
— Minimum
— Maximum
95th Percentile
. Average
Observations from Figure 25-28 for manganese measurements collected at BTUT include
the following:
• The maximum manganese concentration (40.4 ng/m3) was measured in 2004,
although the next highest concentration, measured in 2007, is similar in magnitude
(36.0 ng/m3). The third (28.3 ng/m3) and the fourth (27.2 ng/m3) highest
concentrations were also measured in 2007.
• The 1-year average concentration decreased from 2004 to 2005, after which an
increase shown through 2007, although these changes are not statistically significant.
However, a significant decrease in manganese concentrations is shown between 2007
25-34
-------
and 2010, which is followed by an increase for 2011. The median concentration
follows a similar trend.
• The 1-year average manganese concentration changed very little from 2011 to 2012,
while the median concentration increased considerably (from 6.48 ng/m3 to
8.11 ng/m3). The number of manganese concentrations in the mid- to upper-end of the
concentration range (between 8 ng/m3 and 15 ng/m3) increased from 19 to 27 from
2011 to 2012. At the same time, the number of manganese concentrations at the lower
end of the concentration range (less than 2 ng/m3) decreased from eight to three from
2011 to 2012.
Figure 25-29. Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at BTUT
Concentration (ng/m3)
T
0.0 - -
rh
2004
— <
> —
O
2005
• BthPercentile
Nfc
2006
— Minimum
2007
Median
^ T
r-6-,
••«.-... IT rjh
2008 2009 2010 2011 2012
Year
— Maximum O 95th Percentile ...^>... Average
Observations from Figure 25-29 for nickel measurements collected 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 also measured in 2008. The fifth highest
concentration measured was half as high (less than 10 ng/m3) and was also measured
in 2005.
• All 24 non-detects of nickel were measured in 2009.
• The range of nickel concentrations measured each year is highly variable.
Concentrations measured over a given year have spanned a little as 2.5 ng/m3 (2010)
25-35
-------
or up to nearly 30 ng/m3 (2005). This variability is reflected in the undulating pattern
shown in the central tendency statistics. The 1-year average concentrations have
ranged from 0.75 ng/m3 (2009) to 4.05 ng/m3 (2005).
Figure 25-30. Yearly Statistical Metrics for Propionaldehyde Concentrations Measured at
BTUT
I-
o
"-r1
^r
2007 2008 2009
Year
5th Percent!le
— Minimum
— Maximum
O 95thPercentile
••<>•" Average
Observations from Figure 25-30 for propionaldehyde measurements collected at BTUT
include the following:
• The maximum propionaldehyde concentration (3.38 |ig/m3) was measured on the
same day as the maximum acetaldehyde and formaldehyde concentrations
(August 31, 2004), although a similar concentration was also measured in 2007. No
other propionaldehyde concentrations greater than 2.5 |ig/m3 have been measured at
BTUT.
• Even though the maximum concentration decreased considerably from 2004 to 2005,
the other statistical metrics exhibit increases (similar to the formaldehyde
concentrations). The median increased four-fold from 2004 to 2005, indicating that
concentrations ran higher in 2005 than 2004 (as opposed to being driven by an
outlier, as in 2004). The number of concentrations greater than 1 |ig/m3 tripled from
2004 to 2005 and the number of concentrations between 0.5 |ig/m3 and 1 |ig/m3
quadrupled during this period.
• After 2005, the 1-year average concentration began to decrease, reaching a minimum
for 2009, where the entire set of measurements span less than 1 |ig/m3. The
25-36
-------
propionaldehyde concentrations increase significantly from 2009 to 2010, with an
undulating pattern in the 1-year average concentrations developing afterward.
25.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk at the
BTUT monitoring site. Refer to Sections 3.3 and 3.5.3 for definitions and explanations regarding
the various toxicity factors, time frames, and calculations associated with these risk-based
screenings.
25.5.1 Risk-Based Screening Assessment Using MRLs
A risk-based screening was conducted by comparing the concentration data from the
Utah monitoring site to the ATSDRMRLs, where available. As described in Section 3.3, MRLs
are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 day to 14 days); intermediate (exposures of 15 days to 364 days); and chronic
(exposures of 1 year or greater). The preprocessed daily measurements of the pollutants of
interest were compared to the acute MRLs; the quarterly averages were compared to the
intermediate MRLs; and the annual averages were compared to the chronic MRLs.
As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites are greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2012.
25.5.2 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 3.5.3.4 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 25-6, where applicable. Cancer risk approximations are presented as
probabilities while the noncancer hazard approximations are ratios and thus, unitless values.
25-37
-------
Table 25-6. Risk Approximations for the Utah Monitoring Site
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
Bountiful, Utah - BTUT
Acetaldehyde
Arsenic (PM10)a
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
Dichloromethane
Ethylbenzene
Formaldehyde
Manganese (PM10)a
Naphthalene3
Nickel (PM10)a
Propionaldehyde
0.0000022
0.0043
0.0000078
0.00003
0.000006
0.000026
0.00000013
0.0000025
0.000013
0.000034
0.00048
0.009
0.000015
0.03
0.002
0.1
2.4
0.6
1
0.0098
0.00005
0.003
0.00009
0.008
54/54
55/57
56/56
55/56
56/56
49/56
56/56
56/56
54/54
57/57
59/59
57/57
54/54
2.54
±0.35
0.01
±0.01
1.02
±0.13
0.12
±0.02
0.66
±0.03
0.08
±0.01
7.82
±6.53
0.36
±0.06
4.44
±0.75
0.01
±0.01
0.05
±0.01
O.01
±O.01
0.55
±0.07
5.58
2.05
7.92
3.53
3.97
2.00
1.02
0.91
57.67
1.68
0.67
0.28
0.03
0.03
0.06
0.01
O.01
0.01
O.01
0.45
0.16
0.02
0.02
0.07
— = A Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 25 -5.
Observations for BTUT from Table 25-6 include the following:
• The pollutants with the highest annual average concentrations are dichloromethane,
formaldehyde, acetaldehyde, and benzene, as discussed in Section 25.4.1.
• The pollutants with the highest cancer risk approximations are formaldehyde,
benzene, acetaldehyde, and carbon tetrachloride. The cancer risk approximation for
formaldehyde for BTUT (57.67 in-a-million) is the highest cancer risk approximation
calculated across the program.
• There were no pollutants of interest with noncancer hazard approximations greater
than 1.0, indicating that no adverse health effects are expected from these individual
pollutants. The highest noncancer hazard approximation was calculated for
formaldehyde (0.45), which is the sixth highest noncancer hazard approximation
calculated among the site-specific pollutants of interest with noncancer toxicity
factors. (Note that the five highest noncancer hazard approximations are all for
manganese.)
25-38
-------
• Dichloromethane's relatively high annual average concentration does not translate
into high risk approximations. This is an indication of the toxicity potential of
dichloromethane.
25.5.3 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 25-7 presents the 10 pollutants with the highest emissions from the 2011 NEI that have
cancer toxicity factors. Table 25-7 also presents the 10 pollutants with the highest toxicity -
weighted emissions, based on the weighting schema described in Section 3.5.3.5. Lastly,
Table 25-7 provides the 10 pollutants with the highest cancer risk approximations (in-a-million)
for BTUT, as presented in Table 25-6. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 25-7. Table 25-8 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 3.5.3.5. Similar to the cancer risk and
noncancer hazard approximations provided in Section 25.5.2, this analysis may help policy-
makers prioritize their air monitoring activities.
25-39
-------
Table 25-7. 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)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Bountiful, Utah (Davis County) - BTUT
Benzene
Formaldehyde
Ethylbenzene
Dichloromethane
Acetaldehyde
1,3 -Butadiene
Naphthalene
Methyl tert butyl ether
POM, Group 2b
POM, Group 2d
111.98
68.81
65.63
46.45
40.38
14.29
8.58
4.85
1.84
1.37
Formaldehyde
Benzene
Hexavalent Chromium, PM
1,3 -Butadiene
POM, Group 3
Naphthalene
Ethylbenzene
POM, Group 2b
POM, Group 2d
Acetaldehyde
8.95E-04
8.73E-04
6.56E-04
4.29E-04
3.20E-04
2.92E-04
1.64E-04
1.62E-04
1.21E-04
8.88E-05
Formaldehyde
Benzene
Acetaldehyde
Carbon Tetrachloride
1,3 -Butadiene
Arsenic
1 ,2-Dichloroethane
Naphthalene
Dichloromethane
Ethylbenzene
57.67
7.92
5.58
3.97
3.53
2.05
2.00
1.68
1.02
0.91
to
-------
Table 25-8. 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)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Bountiful, Utah (Davis County) - BTUT
Toluene
Ethylene glycol
Hexane
Xylenes
Methanol
Benzene
Formaldehyde
Ethylbenzene
Methyl isobutyl ketone
Dichloromethane
881.72
438.24
373.79
282.06
205.65
111.98
68.81
65.63
51.38
46.45
Acrolein
1,3 -Butadiene
Formaldehyde
Acetaldehyde
Benzene
Naphthalene
Xylenes
Ethylene glycol
Lead, PM
Arsenic, PM
179,411.58
7,147.14
7,021.22
4,486.93
3,732.63
2,860.46
2,820.62
1,095.59
980.24
707.34
Formaldehyde
Acetaldehyde
Manganese
Propionaldehyde
1,3 -Butadiene
Benzene
Arsenic
Naphthalene
Nickel
Dichloromethane
0.45
0.28
0.16
0.07
0.06
0.03
0.03
0.02
0.02
0.01
to
-k
-------
Observations from Table 25-7 include the following:
• Benzene, formaldehyde, ethylbenzene, and dichloromethane are the highest emitted
pollutants with cancer UREs in Davis County.
• The pollutants with the highest toxi city-weighted emissions (of the pollutants with
cancer UREs) are formaldehyde, benzene, hexavalent chromium, and 1,3-butadiene.
• Eight of the highest emitted pollutants also have the highest toxi city-weighted
emissions in Davis County.
• Formaldehyde and benzene, which have the highest and second highest cancer risk
approximations for BTUT, appear at or near the top of both emissions-based lists.
Acetaldehyde, 1,3-butadiene, naphthalene, and ethylbenzene also appear on all three
lists in Table 25-7. Dichloromethane, which has the highest annual average
concentration and the ninth highest cancer risk approximation for BTUT, ranks fourth
for emissions in Davis County but does not have one of the highest toxi city-weighted
emissions (it ranks 16th). Carbon tetrachloride, which has the fourth highest cancer
risk approximation for BTUT, appears on neither emissions-based list.
• POM, Group 2b is the ninth highest emitted "pollutant" in Davis County and ranks
eighth 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 were identified as pollutants of interest for
BTUT.
• POM, Group 2d is the tenth highest emitted "pollutant" in Davis County and ranks
ninth for toxicity-weighted emissions. POM, Group 2d also includes several PAHs
sampled for at BTUT including phenanthrene, anthracene, and pyrene. None of the
PAHs included in POM, Group 2b were identified as pollutants of interest for BTUT.
Observations from Table 25-8 include the following:
• Toluene, ethylene glycol, and hexane 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.
• Although less than 1.0, formaldehyde, acetaldehyde, and manganese have the highest
noncancer hazard approximations for BTUT. Formaldehyde and acetaldehyde rank
third and fourth (respectively) for toxicity-weighted emissions and formaldehyde
25-42
-------
ranks seventh for total emissions (acetaldehyde does not appear for total emissions
because it ranks 11th). Manganese does not appear on either emissions-based list in
Table 25-8.
25.6 Summary of the 2012 Monitoring Data for BTUT
Results from several of the data treatments described in this section include the
following:
»«» Twenty-one 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 second year in a row, BTUT has the highest annual average formaldehyde
concentration among NMP sites sampling this pollutant.
»«» Concentrations of benzene have an overall decreasing trend at BTUT. The 1-year
average concentration for 2012 is the lowest 1-year average concentration of
benzene calculated since the onset of sampling at BTUT. In addition, the detection
rate of 1,2-dichloroethane has been increasing steadily at BTUT over the last few
years of sampling.
25-43
-------
26.0 Sites in Vermont
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the UATMP and NATTS sites 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.
26.1 Site Characterization
This section characterizes the Vermont 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 Vermont NATTS site (UNVT) and one of the UATMP sites (BURVT) are located in
northwest Vermont in the Burlington-South Burlington, VT MSA. The third site (RUVT) is
located farther south in Rutland, Vermont. Figures 26-1 and 26-2 are the composite satellite
images retrieved from ArcGIS Explorer showing the Burlington monitoring sites and their
immediate surroundings. Figure 26-3 identifies nearby point source emissions locations by
source category, as reported in the 2011 NEI for point sources. Note that only sources within
10 miles of the sites are included in the facility counts provided in Figure 26-3. 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 radii are still visible on the map, but have been grayed out in order to show emissions
sources just outside the boundary. Figures 26-4 and 26-5 are the composite satellite image and
emissions sources map for the Rutland site. Table 26-1 provides supplemental geographical
information such as land use, location setting, and locational coordinates.
26-1
-------
Figure 26-1. Burlington, Vermont (BURVT) Monitoring Site
to
-------
Figure 26-2. Underbill, Vermont (UNVT) Monitoring Site
to
I
OJ
Source: NASA, N
V 2 0 Q a M i c r a 5 a
-------
Figure 26-3. NEI Point Sources Located Within 10 Miles of BURVT and UNVT
73115'0"W 73'10'CTW 73"5'0"W 73rWW 72°55'0"W 72°50'0"W 72"4510"W 72"40'0"W 72 35'0"W
73°30'0"W 73°2S'Q"\N 73°20'0"W 73°15'0"W 73°10'0"W 73°510"W 73°0'0"W 72"55'0"W 72050'0"W
Note: Due to facility density and collocation, the total facilities
. . displayed may not represent all facilities within the area of interest.
Legend
BURVT UATMP site "^ UNVT NATTS site O 10 mile radius | | County boundary
Source Category Group (No. of Facilities)
* Aerospace/Aircraft Manufacturing (1) o Institution (school, hospital, prison, etc.) (1)
t: Airport/Airline/Airport Support Operations (14) <•> Metals Processing/Fabrication (3)
B Bulk Terminals/Bulk Plants (1) A Military Base/National Security (1)
e Electrical Equipment Manufacturing (2) ? Miscellaneous Commercial/Industrial (1)
* Electricity Generation via Combustion (2) P Printing/Publishing/Paper Product Manufacturing (3)
-#- Industrial Machinery or Equipment Plant (1) ® Testing Laboratories (1)
26-4
-------
Figure 26-4. Rutland, Vermont (RUVT) Monitoring Site
to
I
(^
-------
Figure 26-5. NEI Point Sources Located Within 10 Miles of RUVT
72U55'0"W 72 50'0"W
73115'0'1W 73°101011W
Legend
73"5'0-W 73'WW 72°551 Metals Processing/Fabrication (1)
B Pulp and Paper Plant (1)
w Woodwork, Furniture, Millwork & Wood Preserving (2)
26-6
-------
Table 26-1. Geographical Information for the Vermont Monitoring Sites
Site
Code
BURVT
UNVT
RUVT
AQS Code
50-007-0014
50-007-0007
50-021-0002
Location
Burlington
Underbill
Rutland
County
Chittenden
Chittenden
Rutland
Micro- or
Metropolitan
Statistical Area
Burlington-South
Burlington, VT
MSA
Burlington-South
Burlington, VT
MSA
Rutland, VT MSA
Latitude
and
Longitude
44.4762,
-73.2106
44.52839,
-72.86884
43.608056,
-72.982778
Land Use
Commercial
Forest
Commercial
Location
Setting
Urban/City
Center
Rural
Urban/City
Center
Additional Ambient Monitoring Information1
CO, NO, NO2, NOX, Meteorological parameters,
PM25.
Haze, Sulfate, CO, SO2, NO, NOy, O3,
Meteorological parameters, PM10, PM Coarse, PM25,
PM2 5 Speciation, IMPROVE Speciation.
CO, SO2, NO, NO2, NOX, Meteorological parameters,
PM25.
BOLD ITALICS = EPA-designated NATTS Site
to
-------
BURVT is located in a municipal parking lot in downtown Burlington near the
intersection of Main Street and South Winooski Avenue. This location is less than 1 mile east of
Burlington Bay on Lake Champlain. The areas to the west of the site are primarily commercial
while the areas to the east are primarily residential, as shown in Figure 26-1. Route 2 (Main
Street) and Route 7 (South Willard Street) intersect two blocks east of the monitoring site and
1-89 runs north-south just over 1 mile east of the site. Between the two roadways and the
interstate lies the University of Vermont.
The UNVT monitoring site is located on the Proctor Maple Research Farm in Underhill,
Vermont, which is east of the Burlington area. Mount Mansfield, the highest peak in Vermont,
lies to the east in Underhill State Park, less than 3 miles away. The Underhill Artillery Range is
located a few miles to the south. Figure 26-2 shows that the area surrounding the site is rural in
nature and heavily forested. This site is intended to serve as a background site for the region for
trends assessment, standards compliance, and long-range transport assessment.
UNVT and BURVT are located approximately 16 miles apart, as shown in Figure 26-3.
Most of the emissions sources are located between these two sites, although closer to BURVT.
The source category with the greatest number of emissions sources surrounding these 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 sources closest to
BURVT are a medical school/hospital, a heliport at the medical school, and two facilities
generating electricity via combustion. The sources closest to UNVT are private airports.
The RUVT monitoring site is located in Rutland, in central Vermont. The city of Rutland
is in a valley between the Green Mountains to the east and Taconic Mountains to the west. The
monitoring site is located in the courthouse parking lot in downtown Rutland, just north of West
Street. Commercial areas are located to the east and south, while residential areas are located to
the north and west, as shown in Figure 26-4. A railway parallels Route 4 coming into Rutland
from the west, crosses under Route 4, then meanders around a shopping plaza just south of
Route 4. The intersection of Route 4-Business (West Street) and Route 7 is approximately one-
half mile east of the site. Figure 26-5 shows that relatively few point sources are located within
10 miles of RUVT. Most of the emissions sources near RUVT are located along Route 7 (Main
Street), just south of the monitoring site. The source categories with the greatest number of
26-8
-------
sources within 10 miles of the site include airport operations (6) and aerospace/aircraft
manufacturing (3). The source closest to RUVT is an aerospace/aircraft manufacturer.
Table 26-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Vermont monitoring sites. Table 26-2 includes both county-level
population and vehicle registration information. Table 26-2 also contains traffic volume
information for each site as well as the location for which the traffic volume was obtained.
Additionally, Table 26-2 presents the county-level daily VMT for Chittenden and Rutland
Counties.
Table 26-2. Population, Motor Vehicle, and Traffic Information for the Vermont
Monitoring Sites
Site
BURVT
UNVT
RUVT
Estimated
County
Population1
158,504
60,869
County-level
Vehicle
Registration2
169,767
70,900
Annual
Average Daily
Traffic3
14,000
1,100
6,700
Intersection
Used for
Traffic Data
Main St. between S. Union St. and
S. Willard St.
Pleasant Valley Rd, north of Harvey Rd
Bus US-4 between Grove St &
West St./Merchants Row
County-
level Daily
VMT4
4,032,329
1,745,205
Bounty-level population estimates reflect 2012 data (Census Bureau, 2013c)
2County-level vehicle registration reflects 2012 data (VT DMV, 2012)
3AADT reflects 2007 for BURVT and 2011 data for UNVT (CCRPC, 2013) and 2012 data for RUVT (VTrans, 2013a)
4County-level VMT reflects 2011 data (Vtrans, 2013b)
BOLD ITALICS = EPA-designaled NATTS Site
Observations from Table 26-2 include the following:
• The population for Chittenden County is more than twice the population for Rutland
County. The populations for both counties are in the bottom third compared to other
counties with NMP sites.
• A similar pattern is shown for the rankings of the vehicle ownership data for both
counties, although the number of vehicles registered in each county is higher than the
population counts.
• The traffic volume is highest near BURVT and lowest near UNVT among the
Vermont sites. The traffic estimate near BURVT is in the middle of the range
compared to other NMP sites while the traffic volumes for RUVT and UNVT are in
the bottom third compared to other NMP sites. The traffic estimate for BURVT is
provided for Main Street between South Union Street and South Willard Street; for
UNVT, the data is for Pleasant Valley Road, north of Harvey Road; and for RUVT,
the data is for US-4 Business between Merchants Row and Grove Street.
26-9
-------
• Even though the county-level daily VMT for Chittenden County is more than twice
the VMT for Rutland County, both VMTs are in the bottom third compared to other
counties with NMP sites (where VMT data were available).
26.2 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
sites in Vermont on sample days, as well as over the course of the year.
26.2.1 Climate Summary
The city of Burlington resides just to the east of Lake Champlain in northwest Vermont.
Lake Champlain has a moderating affect on the city, keeping the city slightly warmer in winter
than it could be given its New England location. The town of Underhill is located to the east of
Burlington but still within the Burlington MSA. The city of Rutland is located 60 miles south of
the Burlington area. Rutland resides within the same climatic division of Vermont as Burlington,
but misses the moderating influences of Lake Champlain. The state of Vermont is affected by
most storm systems that track across the country, producing variable weather and often cloudy
skies. Summers in Vermont are pleasant, with warm days and cool nights, escaping much of the
heat and humidity most of the East Coast experiences. Winters are warmer in the Champlain
Valley region than in other portions of the state but snow is common state-wide. The highest
precipitation amounts are generally received during the summer months while greater than
15 inches of snow can be expected each month during the winter. Average annual winds parallel
the valleys, generally from the south ahead of advancing weather systems, or from the north
behind these systems. These storm systems tend to be moderated somewhat due to the
Adirondacks to the west and Green Mountains to the east (Wood, 2004; NCDC, 2014; NOAA,
2014b).
26-10
-------
26.2.2 Meteorological Summary
Hourly meteorological data for 2012 were retrieved from NCDC for the weather stations
closest to the Vermont monitoring sites (NCDC, 2012), as described in Section 3.5.2. The closest
weather station to BURVT is located at Burlington International Airport; nearest RUVT is
Rutland State Airport; and nearest UNVT is Morrisville-Stowe State Airport (WBANs 14742,
94737, and 54771, respectively). Additional information about these weather stations, such as the
distance between the sites and the weather stations, is provided in Table 26-3. These data were
used to determine how meteorological conditions on sample days vary from conditions
experienced throughout the year.
Table 26-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for all of 2012. Also included in Table 26-3 is the 95 percent
confidence interval for each parameter. As shown in Table 26-3, meteorological conditions on
sample days were representative of weather conditions experienced throughout the year at these
sites. The greatest difference shown is for sea level pressure at BURVT, although the difference
is not statistically significant. Note that the number of sample days included in the sample day
average for UNVT is twice the number of sample days for BURVT and RUVT. This is because
sampling at UNVT occurs on a l-in-6 day schedule, while sampling at BURVT and RUVT
occurs on a l-in-12 day schedule.
26.2.3 Back Trajectory Analysis
Figure 26-6 is the composite back trajectory map for days on which samples were
collected at the BURVT monitoring site. Included in Figure 26-6 are four back trajectories per
sample day. Figure 26-7 is the corresponding cluster analysis. Similarly, Figures 26-8 through
26-11 are the composite back trajectory maps and corresponding cluster analyses for RUVT and
UNVT. An in-depth description of these maps and how they were generated is presented in
Section 3.5.2.1. For the composite maps, each line represents the 24-hour trajectory along which
a parcel of air traveled toward the monitoring site on a given sample day and time, based on an
initial height of 50 meters AGL. For the cluster analyses, each line corresponds to a trajectory
representative of a given cluster of back trajectories. Each concentric circle around the sites in
Figures 26-6 through 26-11 represents 100 miles.
26-11
-------
Table 26-3. Average Meteorological Conditions near the Vermont Monitoring Sites
Closest Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar
Wind
Speed
(kt)
Burlington, Vermont - BURVT
Burlington Intl.
Airport
14742
(44.48, -73.16)
2.9
miles
87°
(E)
Sample
Days
(31)
2012
59.3
±7.1
58.7
±2.1
50.8
±6.8
50.0
±1.9
39.5
±6.4
38.1
±1.8
45.3
±6.0
44.4
±1.7
67.9
±3.9
66.7
±1.1
1013.5
±2.7
1015.1
±0.8
6.7
±1.0
6.1
±0.3
Rutland, Vermont - RUVT
Rutland State Airport
94737
(43.53, -72.95)
5.4
miles
149°
(SSE)
Sample
Days
(31)
2012
58.5
±6.7
57.3
±2.0
49.0
±6.3
48.4
±1.8
37.2
±5.6
36.1
± 1.6
43.3
±5.4
42.6
±1.5
66.9
±4.6
65.6
± 1.3
NA
NA
6.6
±0.9
6.0
±0.3
Underbill, Vermont - UNVT
Morrisville-Stowe
State Airport
54771
(44.53, -72.61)
11.8
miles
73°
(ENE)
Sample
Days
(64)
2012
55.8
±5.1
56.8
+ 2.1
45.3
±4.8
46.4
+ 1.9
36.2
±4.6
36.8
+ 1.8
41.1
±4.4
42.0
+ 1.7
73.6
±2.5
72.4
+ 1.1
1016.0
±2.0
1015.8
+ 0.8
2.7
±0.5
2.9
+ 0.2
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
NA = Sea level pressure was not recorded at the Rutland State Airport.
to
-------
Figure 26-6. Composite Back Trajectory Map for BURVT
Figure 26-7. Back Trajectory Cluster Map for BURVT
,
26-13
-------
Figure 26-8. Composite Back Trajectory Map for RUVT
Figure 26-9. Back Trajectory Cluster Map for RUVT
26-14
-------
Figure 26-10. Composite Back Trajectory Map for UNVT
Figure 26-11. Back Trajectory Cluster Map for UNVT
26-15
-------
Observations from Figures 26-6 through 26-11 for the Vermont monitoring sites include
the following:
• Even though there are roughly half as many back trajectories on the composite maps
for BURVT and RUVT as there are for UNVT (due to the sampling schedules), the
composite back trajectory maps for the Vermont sites exhibit similarities.
• An imaginary line drawn roughly northwest to southeast through the site on each
composite map shows that the majority of back trajectories originated on the
southwestern side of that line. Few back trajectories originated from the northeast and
east of the sites.
• For each site, the farthest away a back trajectory originated was near Chicago, or
nearly 750 miles away. Back trajectories greater than 600 miles also originated well
offshore over the Atlantic Ocean. However, back trajectories of these lengths were
the exception rather than the norm, as nearly 90 percent of back trajectories originated
within 350 miles of each site. The average back trajectory length for both UNVT and
BURVT is 223 miles, while the average back length for RUVT is 218 miles.
• The cluster analysis for BURVT shows that greater than 50 percent of back
trajectories originated to the west of the site but are grouped into three clusters based
on distance and exact direction: 1) those originating over and just north of Lake
Huron and Georgian Bay, 2) those originating over New York and Lake Ontario, and
3) those originating farther west, primarily over Lake Erie, Michigan, and Lake
Michigan. Another one-third of back trajectories originated to the south of BURVT.
The cluster trajectory originating to the north of BURVT represents those back
trajectories originating to the north, northeast, and east of the site and generally less
than 200 miles in length. The long cluster trajectory originating well off-shore
represents the four long back trajectories originating over the Atlantic Ocean. These
are associated with Hurricane Sandy's landfall and subsequent inland motion on the
October 30, 2012 sample day.
• The cluster analysis for RUVT is similar directionally to the cluster analysis for
BURVT, although the percentages vary. Nearly 60 percent of back trajectories
originated to the west of the site, and like the cluster analysis for BURVT, are divided
into three clusters based on length and exact direction. Nearly 30 percent of back
trajectories originated to the south of the site. The two back trajectories originating
off the North Carolina coast are associated with Hurricane Sandy's landfall. The
cluster trajectory originating to the north of RUVT represents those back trajectories
originating to the north, northeast, and east of the site and generally less than
200 miles in length. The long cluster trajectory originating well off-shore represents
the two long back trajectories originating over the Atlantic Ocean. These are also
associated with Hurricane Sandy's inland motion.
• The cluster analysis for UNVT is similar to the cluster analyses for BURVT and
RUVT in the geographic distribution of back trajectories. Fifty percent of back
trajectories originated from a direction with a westerly component and are grouped
into three cluster trajectories: 1) those originating from the northwest, 2) those
26-16
-------
originating from the west, and 3) those originating from the southwest along Lake
Ontario, Lake Huron, and Michigan. Another 30 percent of back trajectories
originated to the south of UNVT. The cluster trajectory originating to the north of
UNVT represents those back trajectories originating primarily over south-central
Quebec, Canada. The final 5 percent of back trajectories originated to the east of the
site, although the length of these back trajectories varied.
26.2.4 Wind Rose Comparison
Hourly surface wind data from the weather stations at Burlington International Airport
(for BURVT), Rutland State Airport (for RUVT), and Morrisville-Stowe State Airport (for
UNVT) were uploaded into a wind rose software program to produce customized wind roses, as
described in Section 3.5.2.2. 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 26-12 presents a map showing the distance between the weather station and
BURVT, which may be useful for identifying topographical influences that may affect the
meteorological patterns experienced at this location. Figure 26-12 also presents three different
wind roses for the BURVT monitoring site. First, a historical wind rose representing 2002 to
2011 wind data is presented, which shows the predominant surface wind speed and direction
over an extended period of time. Second, a wind rose representing wind observations for all of
2012 is presented. Next, a wind rose representing wind data for days on which samples were
collected in 2012 is presented. These can be used to identify the predominant wind speed and
direction in 2012 and to determine if wind observations on sample days were representative of
conditions experienced over the entire year and historically. Figures 26-13 and 26-14 present the
three wind roses and distance maps for the RUVT and UNVT monitoring sites, respectively.
26-17
-------
Figure 26-12. Wind Roses for the Burlington International Airport Weather Station
near BURVT
Location of BURVT and Weather Station
2002-2011 Historical Wind Rose
2012 Wind Rose
Sample Day Wind Rose
40%
'X% 32%
24%
16%
40%
X% 32%
24%
16%
VflND SPEED
(Knots)
CH -22
^| 17-21
^| 11 - 17
n I"
^1 2- A
Calms: 21.10%
V/IND SPEED
[Knots)
CH -22
^| 17-21
^| 11-17
^| 7- 11
n 4-7
^1 2- 4
Calms: 18.01%
26-18
-------
Figure 26-13. Wind Roses for the Rutland State Airport Weather Station near RUVT
Location of RUVT and Weather Station
2003-2011 Historical Wind Rose
Calms: 17.26%
2012 Wind Rose
Sample Day Wind Rose
Calms: 13.73%
26-19
-------
Figure 26-14. Wind Roses for the Morrisville-Stowe State Airport Weather Station near
UNVT
Location of UNVT and Weather Station
2002-2011 Historical Wind Rose
\SR
2012 Wind Rose
Sample Day Wind Rose
26-20
-------
Observations from Figure 26-12 for BURVT include the following:
• The Burlington International Airport weather station is located approximately 3 miles
east of BURVT.
• The historical wind rose shows that southerly winds are prevalent near BURVT,
accounting for nearly 22 percent of the hourly measurements. Calm winds (< 2 knots)
account for another 20 percent of measurements. Winds from the northwest quadrant,
including north, account for another one-quarter of the wind observations. Winds
from the eastern quadrants are rarely observed.
• The wind patterns shown on the 2012 wind rose are similar to the historical wind
patterns, indicating that wind conditions observed during 2012 were similar to those
observed over the previous 10 years.
• The sample day wind rose shows that southerly winds prevailed on sample days, but
accounted for a higher percentage of observations (35 percent). In addition, fewer
winds from the north and northwest quadrant were observed while a higher
percentage of south-southeasterly winds was observed.
Observations from Figure 26-13 for RUVT include the following:
• The Rutland State Airport weather station is located 5.4 miles south-southeast of
RUVT.
• The historical wind rose shows that east-southeasterly and southeasterly winds were
prevalent near RUVT, as these directions account for more than one-quarter of the
hourly measurements. Winds from the southwest and northwest quadrants were also
observed while winds from the northeast quadrant were generally not observed. Calm
winds were observed for 17 percent of the hourly measurements.
• The wind patterns shown on the 2012 wind rose are similar to the historical wind
patterns, although a slightly higher percentage of winds from the southeast and
slightly fewer east-southeasterly winds were observed in 2012.
• The sample day wind rose exhibits similar wind patterns as the historical and full-
year wind roses, but with higher percentages of east-southeasterly and southeasterly
winds (together accounting for more than one-third of wind observations). This
corresponds with fewer calm observations (less than 14 percent).
Observations from Figure 26-14 for UNVT include the following:
• The Morrisville-Stowe Airport weather station is located approximately 12 miles east
of UNVT. Between the site and the weather station lie the Green Mountains.
• The historical wind rose shows that calm winds were prevalent near UNVT, as calm
winds were observed for nearly 45 percent of the hourly measurements. Winds from
the northwest to north account for 20 percent of the wind observations greater than
26-21
-------
2 knots. Winds from the south to south-southwest account for another 15 percent of
observations.
• The wind patterns shown on the 2012 wind rose are similar to the historical wind
patterns, although calm winds account for nearly 50 percent of the observations.
• The sample day wind rose shows that wind conditions on sample days were similar to
those experienced throughout 2012, although number of observations from the north-
northwest is less while the number of calms is up to nearly 52 percent.
26.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each
Vermont 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 26-4. 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 26-4. It is important to note which pollutants were sampled for at each site when reviewing
the results of this analysis. BURVT and RUVT sampled for year-round for VOCs, while UNVT
sampled for hexavalent chromium, PAHs, and metals (PMio) in addition to VOCs. All three sites
began sampling carbonyl compounds under the NMP in July 2012.
Table 26-4. Risk-Based Screening Results for the Vermont Monitoring Sites
Pollutant
Screening
Value
(Ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Burlington, Vermont - BURVT
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
Acetaldehyde
Formaldehyde
£>-Dichlorobenzene
Hexachloro- 1 ,3 -butadiene
1 ,2-Dibromoethane
1 , 1 ,2,2-Tetrachloroethane
Ethylbenzene
0.13
0.03
0.17
0.038
0.45
0.077
0.091
0.045
0.0017
0.017
0.4
Total
31
31
31
29
16
16
7
5
4
4
3
177
31
31
31
29
16
16
28
6
4
4
31
227
100.00
100.00
100.00
100.00
100.00
100.00
25.00
83.33
100.00
100.00
9.68
77.97
17.51
17.51
17.51
16.38
9.04
9.04
3.95
2.82
2.26
2.26
1.69
17.51
35.03
52.54
68.93
77.97
87.01
90.96
93.79
96.05
98.31
100.00
26-22
-------
Table 26-4. Risk-Based Screening Results for the Vermont Monitoring Sites (Continued)
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Rutland, Vermont - RUVT
Benzene
Carbon Tetrachloride
1,3 -Butadiene
1 ,2-Dichloroethane
Acetaldehyde
Formaldehyde
Ethylbenzene
£>-Dichlorobenzene
Hexachloro- 1 ,3 -butadiene
1 ,2-Dibromoethane
1 , 1 ,2,2-Tetrachloroethane
Trichloroethylene
0.13
0.17
0.03
0.038
0.45
0.077
0.4
0.091
0.045
0.0017
0.017
0.2
Total
31
31
30
29
16
16
9
4
3
2
2
1
174
31
31
30
29
16
16
31
26
5
2
2
4
223
100.00
100.00
100.00
100.00
100.00
100.00
29.03
15.38
60.00
100.00
100.00
25.00
78.03
17.82
17.82
17.24
16.67
9.20
9.20
5.17
2.30
1.72
1.15
1.15
0.57
17.82
35.63
52.87
69.54
78.74
87.93
93.10
95.40
97.13
98.28
99.43
100.00
Underbill Vermont - UNVT
Benzene
Carbon Tetrachloride
1 ,2-Dichloroethane
Formaldehyde
Arsenic (PM10)
Acetaldehyde
1,3 -Butadiene
Hexachloro- 1 ,3 -butadiene
1 , 1 ,2,2-Tetrachloroethane
1 ,2-Dibromoethane
Naphthalene
Manganese (PM10)
Ethylbenzene
0.13
0.17
0.038
0.077
0.00023
0.45
0.03
0.045
0.017
0.0017
0.029
0.005
0.4
Total
61
61
52
31
27
19
10
8
6
5
5
2
1
288
61
61
53
31
55
31
16
10
6
5
58
61
56
504
100.00
100.00
98.11
100.00
49.09
61.29
62.50
80.00
100.00
100.00
8.62
3.28
1.79
57.14
21.18
21.18
18.06
10.76
9.38
6.60
3.47
2.78
2.08
1.74
1.74
0.69
0.35
21.18
42.36
60.42
71.18
80.56
87.15
90.63
93.40
95.49
97.22
98.96
99.65
100.00
Observations from Table 26-4 include the following:
• Eleven pollutants failed at least one screen for BURVT; 78 percent of concentrations
for these 11 pollutants were greater than their associated risk screening value (or
failed screens).
• Ten pollutants contributed to 95 percent of failed screens for BURVT and therefore
were identified as pollutants of interest for this site. These 10 include two carbonyl
compounds and eight VOCs. Although the first nine pollutants listed account for
more than 95 percent of the total failed screens for BURVT, 1,1,2,2-tetrachloroethane
failed the same number of screens as 1,2-dibromoethane (4); thus,
1,1,2,2-tetrachloroethane was also added as a pollutant of interest for BURVT, per the
procedure described in Section 3.2.
26-23
-------
• Twelve pollutants failed at least one screen for RUVT; 78 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 RUVT and therefore
were identified as pollutants of interest for this site. These eight include two carbonyl
compounds and six VOCs.
• Thirteen pollutants failed at least one screen for UNVT; 57 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 UNVT and therefore
were identified as pollutants of interest for this site. These nine include two carbonyl
compounds, six VOCs, and one PMio metal.
• BURVT and RUVT have seven pollutants of interest in common. Even though three
additional pollutant groups were sampled for at UNVT, the Vermont sites have six
pollutants of interest in common (two carbonyl compounds and four VOCs).
26.4 Concentrations
This section presents various concentration averages used to characterize pollution levels
at the Vermont 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. Additional site-specific statistical summaries for BURVT,
RUVT, and UNVT are provided in Appendices J, L, M, N, and O.
26.4.1 2012 Concentration Averages
Quarterly and annual concentration averages were calculated for the pollutants of interest
for each Vermont site, as described in Section 3.1. The quarterly average of a particular
pollutant is simply the average concentration of the preprocessed daily measurements over a
26-24
-------
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 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
Vermont monitoring sites are presented in Table 26-5, where applicable. Note that
concentrations of arsenic for UNVT 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 26-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Vermont Monitoring Sites
Pollutant
#of
Measured
Detections vs.
# of Samples
Burlin
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dibromoethane
£>-Dichlorobenzene
1 ,2-Dichloroethane
Formaldehyde
Hexachloro- 1 ,3 -butadiene
1 , 1 ,2,2-Tetrachloroethane
16/16
31/31
31/31
31/31
4/31
28/31
29/31
16/16
6/31
4/31
1st
Quarter
Average
(jig/m3)
2nd
Quarter
Average
(jig/m3)
3rd
Quarter
Average
(jig/m3)
4th
Quarter
Average
(jig/m3)
Annual
Average
(jig/m3)
?ton, Vermont - BURVT
NA
0.97
±0.18
0.09
±0.03
0.60
±0.10
0.01
±0.01
0.06
±0.02
0.09
±0.03
NA
0.02
±0.02
0.01
±0.01
NA
0.74
±0.09
0.07
±0.01
0.73
±0.04
0.01
±0.02
0.08
±0.03
0.09
±0.01
NA
0.02
±0.04
0.01
±0.02
1.43
±0.36
0.65
±0.09
0.09
±0.03
0.65
±0.03
0.01
±0.02
0.06
±0.03
0.06
±0.02
3.41
±0.94
0.01
±0.02
0.01
±0.01
1.19
±0.35
0.78
±0.12
0.07
±0.02
0.71
±0.05
0
0.04
±0.02
0.07
±0.01
2.38
±0.79
0
0
NA
0.78
±0.07
0.08
±0.01
0.67
±0.03
0.01
±0.01
0.06
±0.01
0.08
±0.01
NA
0.01
±0.01
0.01
±0.01
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
a Average concentrations provided for the pollutant below the blue line are presented in ng/m3 for ease of viewing.
26-25
-------
Table 26-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Vermont Monitoring Sites (Continued)
Pollutant
#of
Measured
Detections vs.
# of Samples
1st
Quarter
Average
(Hg/m3)
2nd
Quarter
Average
(Hg/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Hg/m3)
Annual
Average
(Ug/m3)
Rutland, Vermont - RUVT
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
£>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
16/16
31/31
30/31
31/31
26/31
29/31
31/31
16/16
NA
1.33
±0.34
0.17
±0.07
0.60
±0.13
0.06
±0.02
0.08
±0.02
0.28
±0.07
NA
NA
0.75
±0.17
0.06
±0.02
0.75
±0.05
0.09
±0.03
0.09
±0.01
0.35
±0.12
NA
1.21
±0.32
0.76
±0.14
0.09
±0.04
0.68
±0.04
0.06
±0.02
0.06
±0.02
0.43
±0.11
2.56
±0.91
1.56
±0.67
1.33
±0.61
0.19
±0.10
0.70
±0.03
0.03
±0.02
0.06
±0.02
0.38
±0.11
2.57
±0.81
NA
1.05
±0.20
0.13
±0.04
0.68
±0.04
0.06
±0.01
0.07
±0.01
0.36
±0.05
NA
Underbill, Vermont - UNVT
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
Formaldehyde
Hexachloro- 1 ,3 -butadiene
1 , 1 ,2,2-Tetrachloroethane
Arsenic (PM10)a
31/31
61/61
16/61
61/61
53/61
31/31
10/61
6/61
55/61
NA
0.49
±0.08
0.01
±0.01
0.61
±0.05
0.08
±0.01
NA
0.01
±0.01
0.01
±0.01
0.21
±0.08
NA
0.33
±0.07
0.01
±0.01
0.70
±0.06
0.08
±0.01
NA
0.02
±0.02
0.01
±0.01
0.26
±0.09
0.56
±0.11
0.29
±0.04
0.03
±0.02
0.66
±0.03
0.04
±0.01
1.48
±0.40
0.01
±0.01
O.01
±0.01
0.24
±0.08
0.47
±0.13
0.42
±0.07
0.03
±0.04
0.71
±0.04
0.05
±0.02
0.68
±0.20
0
0
0.27
±0.15
NA
0.38
±0.04
0.02
±0.01
0.67
±0.02
0.06
±0.01
NA
0.01
±0.01
0.01
±0.01
0.25
±0.05
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
a Average concentrations provided for the pollutant below the blue line are presented in ng/m3 for ease of viewing.
Observations for BURVT and RUVT from Table 26-5 include the following:
• BURVT and RUVT sampled VOCs on a l-in-12 day schedule. Carbonyl compounds
were also sampled at these sites on a l-in-12 day schedule beginning in July 2012. As
a result of this start date, first quarter, second quarter, and annual average
concentrations are not available for the carbonyl compound pollutants of interest.
26-26
-------
• Benzene is the pollutant with the highest annual average concentration for BURVT,
followed by carbon tetrachloride. All of the annual average concentrations for the
pollutants of interest for BURVT are less than 1 |ig/m3, where they could be
calculated.
• Concentrations of benzene measured at BURVT range from 0.505 |ig/m3 to
1.41 |ig/m3, with a median benzene concentration of 0.75 |ig/m3. Three of the four
concentrations greater than 1 |ig/m3 were measured during the first quarter of 2012,
and all but two of the eight first quarter measurements are greater than the median
concentration. This explains why the first quarter average is higher than the other
quarterly averages. The difference, however, is not statistically significant. A similar
observation was made in the 2011 NMP report.
• A few of the VOCs listed for BURVT were detected relatively few times. Three of
these pollutants were not detected at all during the fourth quarter, resulting in a fourth
quarter average concentration of zero.
• The quarterly average concentrations of formaldehyde are at least twice the quarterly
averages of acetaldehyde. Concentrations of formaldehyde measured at BURVT
range from 1.33 |ig/m3 to 5.89 |ig/m3; concentrations of acetaldehyde range from
0.62|ig/m3to2.31 |ig/m3.
• Benzene, carbon tetrachloride, and ethylbenzene have the highest annual average
concentrations for RUVT. Only benzene has an annual average concentration greater
than 1 |ig/m3 (1.05 ± 0.20 |ig/m3). This is the highest annual average concentration
among the Vermont sites' pollutants of interest, where they could be calculated.
• The first and fourth quarter average concentrations of benzene are greater than the
other quarterly averages and have relatively large confidence intervals associated with
them. Concentrations of benzene measured at RUVT range from 0.54 |ig/m3 to
2.64 |ig/m3, with a median concentration of 0.90 |ig/m3. Of the 14 benzene
concentrations greater than 1 |ig/m3 measured at RUVT, seven were measured in the
first quarter and four were measured in the fourth quarter. The three concentrations
greater than 2 |ig/m3 were measured in January, November, and December.
• Concentrations of 1,3-butadiene also appear higher during the first and fourth quarters
of 2012. The three highest concentrations of 1,3-butadiene (those greater than
0.3 |ig/m3) were measured at RUVT on the same days as the three highest
concentrations of benzene (although not necessarily in same order). Concentrations of
1,3-butadiene measured at RUVT span an order of magnitude, ranging from
0.038 |ig/m3to 0.364 |ig/m3, including one non-detect.
• Concentrations of formaldehyde measured at RUVT range from 1.37 |ig/m3 to
4.96 g/m3; concentrations of acetaldehyde range from 0.57 |ig/m3to 2.71 |ig/m3.
26-27
-------
Observations for UNVT from Table 26-5 include the following:
• UNVT sampled VOCs, PAHs, PMio metals, and hexavalent chromium on a l-in-6
day schedule. Carbonyl compound sampling on the same schedule was added in July
2012.
• All of the annual average concentrations for the pollutants of interest for UNVT are
less than 1 |ig/m3, where they could be calculated.
• Carbon tetrachloride has the highest annual average concentration for UNVT
(0.67 ± 0.02 |ig/m3). The annual average concentrations of this pollutant are similar
across the three Vermont sites, differing by less than 0.01 |ig/m3.
• Benzene has the second highest annual average concentration of the pollutants of
interest for UNVT (0.38 ± 0.04 |ig/m3). However, this is the lowest annual average
concentration among all NMP sites sampling benzene. Similar to the other Vermont
sites, concentrations of benzene appear higher during the colder months of the year.
Of the 12 measurements greater than 0.5 |ig/m3, all but one was measured during the
first or fourth quarters of the year. Conversely, the nine lowest concentrations were
all measured during the second and third quarters.
• The first and second quarter average concentrations of 1,3-butadiene are at least an
order of magnitude less than the third and fourth quarter average concentrations. A
review of the data shows that this pollutant was detected only once during the first
quarter and twice during the second. Thus, the remaining the 13 measurements are
spread across the third and fourth quarters of the year, but with the majority measured
during the third quarter.
• A few of the VOCs listed for UNVT were detected relatively few times. Two
pollutants were not detected at all during the fourth quarter, resulting in fourth quarter
average concentrations of zero.
• Concentrations of formaldehyde measured at UNVT range from 0.22 |ig/m3 to
3.74 g/m3; concentrations of acetaldehyde range from 0.20 |ig/m3to 1.21 |ig/m3.
• Arsenic was detected in most of the metals samples collected at UNVT. In addition to
six non-detects, concentrations of arsenic range from 0.01 ng/m3to 0.90 ng/m3.
Among NMP sites sampling arsenic, UNVT has the lowest annual average
concentration of this pollutant (0.25 ± 0.05 ng/m3).
Tables 4-9 through 4-12 present the sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for the Vermont
monitoring sites from those tables include the following:
26-28
-------
• BURVT appears twice in Table 4-9 for VOCs. BURVT has the fifth highest annual
average concentration of hexachloro-1,3-butadiene and the sixth highest annual
average concentration of 1,2-dichloroethane among NMP sites sampling VOCs.
• RUVT appears in Table 4-9 five times, but ranks no higher than seventh for any of
the pollutants for which is appears.
• UNVT appears in Table 4-9 only once; UNVT has the seventh highest annual average
concentration of hexachloro-1,3-butadiene. UNVT does not appear in any of the other
tables.
26.4.2 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 shaded in
gray in Table 26-4 for BURVT, RUVT and UNVT. Figures 26-15 through 26-24 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 3.5.3.1.
Figure 26-15. Program vs. Site-Specific Average Arsenic (PMio) Concentration
4 5
Concentration (ng/m3)
Program: 1st Quartile 2nd Quartile 3rd Quartile 4th Quartile Average
Site: Site Average Site Concentration Range
o —
26-29
-------
Figure 26-16. Program vs. Site-Specific Average Benzene Concentrations
RUVT
UNVT
Concentration {[og/m3)
Program
Site:
: IstQuartile
Site Average
o
2ndQuartile SrdQuartile 4thQuartile Av
Site Concentration Range
?rage
Figure 26-17. Program vs. Site-Specific Average 1,3-Butadiene Concentrations
Program Max Concentration = 4.10 ug/m3
RUVT
Program Max Concentration = 4.10 ug/m3
UNVT
E
Program Max Concentration = 4.10 ug/m
0.25
0.5
0.75 1
Concentration {
1.25
1.5
1.75
Program:
Site:
IstQuartile
D
Site Average
O
2ndQuartile SrdQuartile
• •
Site Concentration Range
^^^^~
4thQuartile Average
D 1
26-30
-------
Figure 26-18. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations
BURVT
RUVT
UNVT
2 3
Concentration {[ig/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o —
Figure 26-19. Program vs. Site-Specific Average 1,2-Dibromoethane Concentration
BURVT
0.02
0.04
0.06 0.08
Concentration (jig/m3)
0.1
0.12
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o —
0.14
26-31
-------
Figure 26-20. Program vs. Site-Specific Average /7-Dichlorobenzene Concentrations
0.2
0.4
0.6 0.8
Concentration {[og/m3)
1.2
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o —
1.4
Figure 26-21. Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations
BURVT
r;
Program Max Concentration = 17.01 ug/m3
RUVT
Program Max Concentration = 17.01 ug/m3
Program Max Concentration = 17.01 ug/m3
0.4 0.5 0.6
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o —
26-32
-------
Figure 26-22. Program vs. Site-Specific Average Ethylbenzene Concentration
RUVT
i
1.5 2 2.5
Concentration (pg/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o
Figure 26-23. Program vs. Site-Specific Average Hexachloro-l,3-Butadiene Concentrations
UNVT
0.1 0.15
Concentration {[ug/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: SiteAverage Site Concentration Range
O
Figure 26-24. Program vs. Site-Specific Average 1,1^2,2-Tetrachloroethane Concentrations
0.09
Concentration {
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: SiteAverage Site Concentration Range
O
26-33
-------
Observations from Figures 26-15 through 26-24 include the following:
• Figure 26-15 is the box plot for arsenic. UNVT is the only Vermont site that
sampled PMio metals. UNVT's annual average arsenic (PMio) concentration is
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.
The maximum arsenic concentration measured at UNVT is just greater than the
program-level third quartile. A few non-detects of arsenic were measured at
UNVT.
• Figure 26-16 for benzene shows all three Vermont sites. The annual average
concentration of benzene is highest for RUVT and lowest for UNVT. The annual
average concentration for RUVT is the only one greater than the program-level
average concentration. The annual average for BURVT is less than the program-
level average but greater than the program-level median concentration. UNVT's
annual average benzene concentration is less than the program-level average,
median, and first quartile concentrations. The minimum benzene concentration
measured at BURVT and RUVT are greater than the annual average for UNVT
and the program-level first quartile.
• Figure 26-17 for 1,3-butadiene also shows all three sites. Note that the program-
level maximum concentration (4.10 |ig/m3) is not shown directly on the box plots
because the scale of the box plots would be too large to readily observe data
points at the lower end of the concentration range. Thus, the scale of the box plots
has been reduced to 2 |ig/m3. The box plots for 1,3-butadiene are similar to the
box plots for benzene: The annual average concentration for RUVT is the only
one greater than the program-level average concentration; the annual average for
BURVT is less than the program-level average but greater than the program-level
median concentration; and the annual average concentration for UNVT is less
than the program-level average, median, and first quartile concentrations. The
maximum 1,3-butadiene concentration measured at each site is at least an order of
magnitude less than the maximum concentration measured across the program.
One non-detect was measured at RUVT. Nearly 75 percent of the 1,3-butadiene
measurements were non-detects for UNVT. The minimum concentration of
1,3-butadiene measured at BURVT is the same as the program-level first quartile.
• Figure 26-18 presents the box plots for carbon tetrachloride for all three sites. The
range of measurements collected at the Vermont sites are very similar to each
other. The annual average concentration for each site is similar to the program-
level average concentration of carbon tetrachloride. The maximum concentrations
measured at these sites are significantly less than the maximum concentration
measured across the program.
• Figure 26-19 presents the box plot for 1,2-dibromoethane for BURVT. This
pollutant is not a pollutant of interest for RUVT or UNVT. The first, second, and
third quartiles are not visible on the box plot because they are all zero due to the
large number of non-detects of this pollutant. This pollutant was detected four
26-34
-------
times at BURVT. BURVT is one of only two NMP sites for which
1,2-dibromoethane is a pollutant of interest.
• Figure 26-20 is the box plot for/>-dichlorobenzene for BURVT and RUVT. This
pollutant is not a pollutant of interest for UNVT. The range ofp-dichlorobenzene
measurements is similar between these two sites. The annual average
concentrations for BURVT and RUVT were similar to each other and to the
program-level average concentration. The maximum concentration measured at
each of these two sites is considerably less than the maximum concentration
measured across the program.
• Figure 26-21 presents the box plots for 1,2-dichloroethane for all three sites. Note
that the program-level maximum concentration (17.01 |ig/m3) is not shown
directly on the box plot as the scale has been reduced to 1 |ig/m3 in order to allow
for the observation of data points at the lower end of the concentration range. The
program-level average concentration is greater than the program-level third
quartile for this pollutant and is greater than or similar to the maximum
concentration measured at most sites sampling 1,2-dichloroethane. This is
because the program-level average is being driven by the higher measurements
collected at a few monitoring sites. Figure 26-21 shows that the maximum
1,2-dichloroethane concentrations measured at the Vermont sites are two orders
of magnitude less than the maximum concentration measured across the program.
The maximum concentrations measured at the Vermont sites are also less than the
program-level average concentration. The annual averages for BURVT and
RUVT are similar to the median concentration at the program level, while the
annual average for UNVT is similar to the program-level first quartile. At least
two non-detects of 1,2-dichloroethane were measured at each Vermont site.
• Figure 26-22 is the box plot for ethylbenzene for RUVT, the only Vermont site
for which ethylbenzene is a pollutant of interest. The range of ethylbenzene
concentrations measured at RUVT is relatively small. The annual average
concentration for RUVT is similar to the program-level average concentration.
The maximum ethylbenzene concentration measured at RUVT is considerably
less than the maximum concentration measured across the program. The
minimum ethylbenzene concentration measured at RUVT is greater than the
program-level first quartile.
• Figure 26-23 presents the box plots for hexachloro-1,3-butadiene for BURVT and
UNVT. This pollutant is not a pollutant of interest for RUVT. The first, second,
and third quartiles are not visible on the box plots because they are all zero due to
the large number of non-detects of this pollutant. This pollutant was detected in
fewer than 20 percent of the samples collected at BURVT and UNVT in 2012.
The range of measurements is greater for BURVT than UNVT. The annual
average concentration for BURVT is slightly greater than the program-level
average while the annual average concentration for UNVT is similar to the
program-level average concentration.
26-35
-------
• Figure 26-24 presents the box plots for 1,1,2,2-tetrachloroehtane for BURVT and
UNVT. This pollutant is not a pollutant of interest for RUVT. The first, second,
and third quartiles are not visible on the box plots because they are all zero due to
the large number of non-detects of this pollutant. This pollutant was detected in
relatively few of the samples collected at BURVT and UNVT in 2012. The range
of measurements for BURVT is similar to the range of measurements for UNVT.
The annual average concentration for BURVT is just slightly greater than the
program-level average while the annual average concentration for UNVT is
similar to the program-level average concentration.
26.4.3 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 3.5.3.2.
UNVT has sampled PMio metals under the NMP since 2008. Thus, Figure 26-25 presents the
annual statistical metrics for arsenic for UNVT, respectively. Sampling under the NMP did not
begin at BURVT or RUVT until 2009; thus, a trends analysis was not performed for these sites.
The statistical metrics presented for assessing trends include the substitution of zeros for non-
detects. If sampling began mid-year, a minimum of 6 months of sampling is required for
inclusion in the trends analysis; in these cases, a 1-year average is not provided, although the
range and quartiles are still presented.
26-36
-------
Figure 26-25. Yearly Statistical Metrics for Arsenic (PMi0) Concentrations
Measured at UNVT
2010
Year
O 5th Percentile
— Minimum
0 95thPercentile
Observations from Figure 26-25 for arsenic measurements collected at UNVT include the
following:
• The maximum arsenic concentration was measured at UNVT on November 11, 2012
(0.90 ng/m3).
• With the exception of the 95th percentile, a slight decreasing trend is shown for all of
the statistical metrics between 2008 and 2010. The 1-year average concentration
during this time decreased slightly from 0.25 ng/m3 to 0.21 ng/m3. The minimum in
2008 was 0.05 ng/m3, which decreased to 0.02 ng/m3 for 2009, and the first non-
detects were measured in 2010 (three).
• Most of the statistical metrics exhibit increases for 2011, particularly the 95th
percentile, which is roughly equivalent to the previous year's maximum
concentration. Only the minimum concentration stayed the same for 2011 (zero), as
three non-detects were measured in 2011.
• Although the maximum concentration exhibits further increases for 2012, most of the
other statistical parameters either decreased slightly (5th percentile, 1-year average,
and 95th percentile) or did not change (minimum and median concentration). The
number of non-detects doubled for 2012. That is why the minimum concentration and
5th percentile are both zero for 2012.
26-37
-------
26.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk at the
Vermont monitoring sites. Refer to Sections 3.3 and 3.5.3 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.
26.5.1 Risk-Based Screening Assessment Using MRLs
A risk-based screening was conducted by comparing the concentration data from the
Vermont monitoring sites to the ATSDR MRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 day to 14 days); intermediate (exposures of 15 days to 364 days); and chronic
(exposures of 1 year or greater). The preprocessed daily measurements of the pollutants of
interest were compared to the acute MRLs; the quarterly averages were compared to the
intermediate MRLs; and the annual averages were compared to the chronic MRLs.
As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2012.
26.5.2 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Vermont 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 3.5.3.4 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 26-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
26-38
-------
Table 26-6. Risk Approximations for the Vermont Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
Burlington, Vermont - BURVT
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dibromoethane
£>-Dichlorobenzene
1 ,2-Dichloroethane
Formaldehyde
Hexachloro- 1 , 3 -butadiene
1 , 1 ,2,2-Tetrachloroethane
0.0000022
0.0000078
0.00003
0.000006
0.0006
0.000011
0.000026
0.000013
0.000022
0.000058
0.009
0.03
0.002
0.1
0.009
0.8
2.4
0.0098
0.09
16/16
31/31
31/31
31/31
4/31
28/31
29/31
16/16
6/31
4/31
NA
0.78
±0.07
0.08
±0.01
0.67
±0.03
0.01
±0.01
0.06
±0.01
0.08
±0.01
NA
0.01
±0.01
0.01
±0.01
NA
6.12
2.44
4.03
4.06
0.67
1.98
NA
0.28
0.35
NA
0.03
0.04
0.01
<0.01
0.01
0.01
NA
0.01
Rutland, Vermont - RUVT
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
£>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
0.0000022
0.0000078
0.00003
0.000006
0.000011
0.000026
0.0000025
0.000013
0.009
0.03
0.002
0.1
0.8
2.4
1
0.0098
16/16
31/31
30/31
31/31
26/31
29/31
31/31
16/16
NA
1.05
±0.20
0.13
±0.04
0.68
±0.04
0.06
±0.01
0.07
±0.01
0.36
±0.05
NA
NA
8.20
3.88
4.07
0.67
1.91
0.90
NA
NA
0.04
0.06
0.01
0.01
O.01
0.01
NA
26-39
-------
Table 26-6. Risk Approximations for the Vermont Monitoring Sites (Continued)
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
Underbill, Vermont - UNVT
Acetaldehyde
Arsenic (PM10)a
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
Formaldehyde
Hexachloro- 1 , 3 -butadiene
1 , 1 ,2,2-Tetrachloroethane
0.0000022
0.0043
0.0000078
0.00003
0.000006
0.000026
0.000013
0.000022
0.000058
0.009
0.000015
0.03
0.002
0.1
2.4
0.0098
0.09
31/31
55/61
61/61
16/61
61/61
53/61
31/31
10/61
6/61
NA
0.01
±0.01
0.38
±0.04
0.02
±0.01
0.67
±0.02
0.06
±0.01
NA
0.01
±0.01
0.01
±0.01
NA
1.06
2.98
0.54
4.02
1.58
NA
0.21
0.29
NA
0.02
0.01
0.01
0.01
O.01
NA
O.01
— = a Cancer URE or Noncancer RfC is not available.
NA = Not available due to the criteria for calculating an annual average.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 26-5.
Observations from Table 26-6 include the following:
• For BURVT, benzene and carbon tetrachloride have the highest annual average
concentrations. Benzene and 1,2-dibromoethane have the highest cancer risk
approximations for BURVT (6.12 in-a-million and 4.06 in-a-million, respectively),
with carbon tetrachloride ranking third (4.03 in-a-million).
• Benzene and carbon tetrachloride have the highest annual average concentrations for
RUVT. These pollutants also have the highest cancer risk approximations for RUVT
(8.20 in-a-million and 4.07 in-a-million, respectively).
• Carbon tetrachloride and benzene have the highest annual average concentrations for
UNVT. These two pollutants also have the highest cancer risk approximations for
UNVT (4.02 in-a-million and 2.98 in-a-million, respectively).
• The noncancer hazard approximations for the pollutants of interest for all three
Vermont sites are all considerably less than 1.0, indicating that no adverse health
effects are expected from these individual pollutants.
• Annual averages could not be calculated for the carbonyl compound pollutants of
interest due to the short sampling duration; as a result, cancer and noncancer hazard
approximations could not be calculated.
26-40
-------
26.5.3 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 26-7 presents the 10 pollutants with the highest emissions from the 2011 NEI that have
cancer toxicity factors. Table 26-7 also presents the 10 pollutants with the highest toxicity -
weighted emissions, based on the weighting schema described in Section 3.5.3.5. Lastly,
Table 26-7 provides the 10 pollutants with the highest cancer risk approximations (in-a-million)
for each site, as presented in Table 26-6. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 26-7. Table 26-8 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 3.5.3.5. Similar to the cancer risk and
noncancer hazard approximations provided in Section 26.5.2, this analysis may help policy-
makers prioritize their air monitoring activities.
26-41
-------
Table 26-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Vermont Monitoring Sites
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Burlington, Vermont (Chittenden County) - BURVT
Benzene
Formaldehyde
Acetaldehyde
Ethylbenzene
1,3 -Butadiene
Naphthalene
Dichloromethane
Tetrachloroethylene
POM, Group 2b
POM, Group 2d
87.16
55.98
32.76
32.08
11.30
6.46
2.54
2.21
1.63
0.96
Formaldehyde
Benzene
1,3 -Butadiene
Hexavalent Chromium, PM
Arsenic, PM
POM, Group 3
Naphthalene
POM, Group 2b
POM, Group 5a
Nickel, PM
7.28E-04
6.80E-04
3.39E-04
3.22E-04
2.71E-04
2.50E-04
2.20E-04
1.44E-04
1.02E-04
9.56E-05
Benzene
1 ,2-Dibromoethane
Carbon Tetrachloride
1,3 -Butadiene
1 ,2-Dichloroethane
£>-Dichlorobenzene
1 , 1 ,2,2-Tetrachloroethane
Hexachloro- 1 ,3 -butadiene
6.12
4.06
4.03
2.44
1.98
0.67
0.35
0.28
Underbill, Vermont (Chittenden County) - UNVT
Benzene
Formaldehyde
Acetaldehyde
Ethylbenzene
1,3 -Butadiene
Naphthalene
Dichloromethane
Tetrachloroethylene
POM, Group 2b
POM, Group 2d
87.16
55.98
32.76
32.08
11.30
6.46
2.54
2.21
1.63
0.96
Formaldehyde
Benzene
1,3 -Butadiene
Hexavalent Chromium, PM
Arsenic, PM
POM, Group 3
Naphthalene
POM, Group 2b
POM, Group 5a
Nickel, PM
7.28E-04
6.80E-04
3.39E-04
3.22E-04
2.71E-04
2.50E-04
2.20E-04
1.44E-04
1.02E-04
9.56E-05
Carbon Tetrachloride
Benzene
1 ,2-Dichloroethane
Arsenic
1,3 -Butadiene
1 , 1 ,2,2-Tetrachloroethane
Hexachloro- 1 ,3 -butadiene
4.02
2.98
1.58
1.06
0.54
0.29
0.21
to
ON
-k
to
-------
Table 26-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Vermont Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Rutland, Vermont (Rutland County) - RUVT
Benzene
Formaldehyde
Acetaldehyde
Ethylbenzene
1,3 -Butadiene
Naphthalene
POM, Group 2b
POM, Group 2d
Tetrachloroethylene
Trichloroethylene
49.23
25.76
18.18
16.11
5.57
3.49
0.82
0.47
0.38
0.30
Benzene
Formaldehyde
1,3 -Butadiene
POM, Group 3
Naphthalene
POM, Group 2b
POM, Group 5a
Arsenic, PM
POM, Group 2d
Ethylbenzene
3.84E-04
3.35E-04
1.67E-04
1.38E-04
1.19E-04
7.21E-05
6.43E-05
4.26E-05
4.12E-05
4.03E-05
Benzene
Carbon Tetrachloride
1,3 -Butadiene
1 ,2-Dichloroethane
Ethylbenzene
£>-Dichlorobenzene
8.20
4.07
3.88
1.91
0.90
0.67
to
-------
Table 26-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Vermont Monitoring Sites
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations (Site-
Specific)
Pollutant
Noncancer Hazard
Approximation
(HQ)
Burlington, Vermont (Chittenden County) - BURVT
Toluene
Ethylene glycol
Xylenes
Hexane
Methanol
Benzene
Formaldehyde
Hydrochloric acid
Acetaldehyde
Ethylbenzene
385.94
192.75
135.31
100.90
90.73
87.16
55.98
35.41
32.76
32.08
Acrolein
Manganese, PM
Chlorine
Formaldehyde
1,3 -Butadiene
Arsenic, PM
Acetaldehyde
Benzene
Nickel, PM
Cadmium, PM
508,436.14
59,493.59
12,098.33
5,712.53
5,649.93
4,200.83
3,639.92
2,905.47
2,212.99
2,178.46
1,3 -Butadiene
Benzene
Carbon Tetrachloride
1 ,2-Dibromoethane
Hexachloro- 1 , 3 -butadiene
£>-Dichlorobenzene
1 ,2-Dichloroethane
0.04
0.03
0.01
<0.01
<0.01
<0.01
<0.01
Underbill, Vermont (Chittenden County) - UNVT
Toluene
Ethylene glycol
Xylenes
Hexane
Methanol
Benzene
Formaldehyde
Hydrochloric acid
Acetaldehyde
Ethylbenzene
385.94
192.75
135.31
100.90
90.73
87.16
55.98
35.41
32.76
32.08
Acrolein
Manganese, PM
Chlorine
Formaldehyde
1,3 -Butadiene
Arsenic, PM
Acetaldehyde
Benzene
Nickel, PM
Cadmium, PM
508,436.14
59,493.59
12,098.33
5,712.53
5,649.93
4,200.83
3,639.92
2,905.47
2,212.99
2,178.46
Arsenic
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Hexachloro- 1 , 3 -butadiene
1 ,2-Dichloroethane
0.02
0.01
0.01
0.01
<0.01
<0.01
to
-------
Table 26-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Vermont Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer Hazard
Approximation
(HQ)
Rutland, Vermont (Rutland County) - RUVT
Toluene
Ethylene glycol
Xylenes
Benzene
Hexane
Methanol
Formaldehyde
Acetaldehyde
Ethylbenzene
1,3 -Butadiene
183.66
76.00
64.79
49.23
40.83
35.39
25.76
18.18
16.11
5.57
Acrolein
1,3 -Butadiene
Formaldehyde
Acetaldehyde
Benzene
Naphthalene
Arsenic, PM
Xylenes
Lead, PM
Nickel, PM
76,839.04
2,784.50
2,628.93
2,020.27
1,641.04
1,162.51
659.72
647.86
572.30
521.39
1,3 -Butadiene
Benzene
Carbon Tetrachloride
Ethylbenzene
£>-Dichlorobenzene
1 ,2-Dichloroethane
0.06
0.04
0.01
<0.01
<0.01
<0.01
to
-------
Observations from Table 26-7 include the following:
• Benzene, formaldehyde, and acetaldehyde are the highest emitted pollutants with
cancer UREs in both Chittenden and Rutland Counties, although the emissions in
Chittenden County were nearly twice those in Rutland County.
• Formaldehyde, benzene, and 1.3-butadiene are the pollutants with the highest
toxi city-weighted emissions (of the pollutants with cancer UREs) for both counties,
although not necessarily in that order.
• Five of the highest emitted pollutants also have the highest toxi city-weighted
emissions for Chittenden County while seven of the highest emitted pollutants also
have the highest toxicity-weighted emissions for Rutland County.
• Benzene is at or near the top of all three lists for both counties (for the emissions) and
for all three sites (for the cancer risk approximations). The cancer risk approximation
for carbon tetrachloride is among the highest for all three sites, but this pollutant
appears on neither emissions-based list for either county. Formaldehyde is also at or
near the top of the emissions-based lists for both counties, although a full-year's
worth of sampling is needed to determine how the concentrations of formaldehyde
rank among each site's pollutants of interest. 1,3-Butadiene is another pollutant for
which a cancer risk approximation could be calculated for all sites and that appears
near the top of both emissions-based lists.
• Arsenic has the fourth highest cancer risk approximation for UNVT and ranks fifth
for its toxicity-weighted emissions, but is not one of the highest emitted in Chittenden
County.
• Naphthalene ranks seventh for its toxicity-weighted emissions and ranks sixth for its
total emissions for Chittenden County. Naphthalene failed screens for UNVT but was
not identified as a pollutant of interest for this site.
• Several POM Groups appear on the emissions-based lists for Chittenden and Rutland
Counties. Several of the PAHs sampled for at UNVT are included in various POM
Groups. Benzo(a)pyrene is part of POM, Group 5a; POM, Group 2b includes
acenaphthylene, fluoranthene, and perylene; and POM, Group 2d includes
anthracene, phenanthrene, and pyrene. None of the pollutants sampled for at UNVT
and included in these POM groups failed screens.
Observations from Table 26-8 include the following:
• Toluene, ethylene glycol, and xylenes are the highest emitted pollutants with
noncancer RfCs in Chittenden and Rutland Counties, although the emissions in
Chittenden County were more than twice those in Rutland County.
• Acrolein is the pollutant with the highest toxicity-weighted emissions (of the
pollutants with noncancer RfCs) for both Chittenden and Rutland Counties. Although
acrolein was sampled for at all three sites, this pollutant was excluded from the
26-46
-------
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 for Chittenden County also have the highest
toxi city-weighted emissions while five of the highest emitted pollutants for Rutland
County also have the highest toxicity-weighted emissions.
• Although very low, 1,3-butadiene and benzene have the highest noncancer hazard
approximations for BURVT and RUVT. Benzene appears on both emissions-based
lists for both counties. Although 1,3-butadiene also appears on both emissions-based
lists for Rutland County, this pollutant ranks fifth for toxicity-weighted emissions in
Chittenden County but is not among the highest emitted.
• Although very low, arsenic has the highest noncancer hazard approximation for
UNVT. While this pollutant ranks sixth among the toxicity-weighted emissions for
Chittenden County, it is not among the highest emitted. Four of the metals sampled
for at UNVT appear among the pollutants with the highest toxicity-weighted
emissions but are not among the highest emitted.
26.6 Summary of the 2012 Monitoring Data for the Vermont Monitoring Sites
Results from several of the data treatments described in this section include the
following:
»«» A total of 11 pollutants failed screens for B UR VT; 12 pollutants failed screens for
RUVT; and 13 pollutants failed screens for UNVT.
»«» Among the site-specific pollutants of interest, only the annual average benzene
concentration for RUVT was greater than 1 jug/m3.
»«» The annual average concentrations for several of UNVT's pollutants of interest were
the lowest annual averages among NMP sites sampling those pollutants.
26-47
-------
27.0 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.
27.1 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 27-1 is a composite satellite image retrieved from ArcGIS Explorer
showing the monitoring site and its immediate surroundings. Figure 27-2 identifies nearby point
source emissions locations by source category, as reported in the 2011 NEI for point sources.
Note that only sources within 10 miles of the site are included in the facility counts provided in
Figure 27-2. 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 radius are still visible on the map, but have been grayed out in order
to show emissions sources just outside the boundary. Table 27-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates.
27-1
-------
Figure 27-1. East Highland Park, Virginia (RIVA) Monitoring Site
to
-------
Figure 27-2. NEI Point Sources Located Within 10 Miles of RIVA
Legend
77°30'0"W 77°25'
-------
Table 27-1. Geographical Information for the Virginia Monitoring Site
Site
Code
RIVA
AQS Code
51-087-0014
Location
East
Highland
Park
County
Henrico
Micro- or
Metropolitan
Statistical Area
Richmond, VA
MSA
Latitude
and
Longitude
37.55652,
-77.40027
Land Use
Residential
Location
Setting
Suburban
Additional Ambient Monitoring Information1
TSP Metals, CO, SO2, NOy, NO, NO2, NOX, VOCs,
Carbonyl compounds, O3, Meteorological parameters,
PM10, PM10 Metals, PM Coarse, PM25, PM2 5
Speciation, IMPROVE Speciation.
BOLD ITALICS = EPA-designated NATTS Site
to
-k
-------
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 less than one-half mile west of the site, as shown in Figure 27-1. 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 27-2 shows,
RIVA is located near several point sources, most of which are located to the southwest and south
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.
Table 27-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Virginia monitoring site. Table 27-2 includes both county-level
population and vehicle registration information. Table 27-2 also contains traffic volume
information for RIVA as well as the location for which the traffic volume was obtained.
Additionally, Table 27-2 presents the county-level daily VMT for Henrico County.
Table 27-2. Population, Motor Vehicle, and Traffic Information for the Virginia
Monitoring Site
Site
RIVA
Estimated
County
Population1
314,932
County-level
Vehicle
Registration2
354,419
Annual
Average Daily
Traffic3
72,000
Intersection
Used for
Traffic Data
1-64 at Mechanicsville Turnpike
County-level
Daily VMT4
8,232,198
Bounty-level population estimate reflects 2012 data (Census Bureau, 2013c).
2County-level vehicle registration reflects 2012 data (Henrico County, 2013).
3AADT reflects 2012 data (VA DOT, 2012).
4County-level VMT reflects 2012 data (VA DOT, 2013).
BOLD ITALICS = EPA-designaled NATTS Site
Observations from Table 27-2 include the following:
• RIVA's county-level population is in the middle third of the range compared to other
counties with NMP sites, as is its county-level vehicle ownership.
27-5
-------
• The traffic volume experienced near RIVA is also in the middle of the range
compared to other NMP monitoring sites. The traffic volume provided is for 1-64 at
US-360 (Mechanicsville Turnpike).
• The daily VMT for Henrico County is also in the middle of the range compared to
other counties with NMP sites (where VMT data are available).
27.2 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.
27.2.1 Climate Summary
The city of Richmond is located in east-central Virginia, east of the Blue Ridge
Mountains and west of the Chesapeake Bay and Atlantic Ocean. The James River flows through
the west, center, and south parts of town. Richmond has a modified continental climate. Winters
tend to be mild, as the mountains act as a barrier to cold air and the proximity to the Atlantic
Ocean prevents temperatures from plummeting too low. Summers are warm and humid, also due
to these influences. Precipitation is well distributed throughout the year, with 3 inches to 4 inches
typical during most months of the year. A northerly wind is common during the winter months
while southerly winds prevail during the warmest months of the year (Wood, 2004).
27.2.2 Meteorological Summary
Hourly meteorological data for 2012 were retrieved from NCDC for the weather station
closest to the Virginia monitoring site (NCDC, 2012), as described in Section 3.5.2. The closest
weather station is located at Richmond International Airport (WBAN 13740). Additional
information about the Richmond International Airport weather station, such as the distance
between the site and the weather station, is provided in Table 27-3. These data were used to
determine how meteorological conditions on sample days vary from conditions experienced
throughout the year.
27-6
-------
Table 27-3. Average Meteorological Conditions near the Virginia Monitoring Site
Closest Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
<°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
<°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
East Highland Park, Virginia - RIVA
Richmond
International Airport
13740
(37.51, -77.32)
5.5
miles
119°
(ESE)
Sample
Days
(67)
2012
71.6
±4.0
71.0
+ 1.6
61.0
±3.8
60.9
+ 1.5
48.7
±4.3
48.7
+ 1.7
54.6
±3.6
54.5
+ 1.5
67.5
±3.4
68.0
+ 1.4
1017.6
±1.8
1017.4
+ 0.7
5.7
±0.5
5.9
+ 0.3
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
to
-------
Table 27-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for all of 2012. Also included in Table 27-3 is the 95 percent
confidence interval for each parameter. As shown in Table 27-3, average meteorological
conditions on sample days were representative of average weather conditions experienced
throughout the year.
27.2.3 Back Trajectory Analysis
Figure 27-3 is the composite back trajectory map for days on which samples were
collected at the RIVA monitoring site. Included in Figure 27-3 are four back trajectories per
sample day. Figure 27-4 is the corresponding cluster analysis. An in-depth description of these
maps and how they were generated is presented in Section 3.5.2.1. For the composite map, each
line represents the 24-hour trajectory along which a parcel of air traveled toward the monitoring
site on a given sample day and time, based on an initial height of 50 meters AGL. For the cluster
analysis, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric circle around the site in Figures 27-3 and 27-4 represents 100 miles.
Observations from Figures 27-3 and Figure 27-4 for RIVA include the following:
• Back trajectories originated from a variety of directions near RIVA, although a
greater number of them originated from a direction with a westerly component.
• The 24-hour air shed domain for RIVA was similar in size to many other NMP
monitoring sites. The farthest away a back trajectory originated was over Lake
Michigan, or approximately 650 miles away. However, the average back trajectory
length is 214 miles and most back trajectories (89 percent) originated within
350 miles of the site.
• The cluster analysis shows that 10 percent of back trajectories originated to the
northeast and east of RIVA. Eighteen percent of back trajectories originated to the
north and northwest of the site, primarily over Pennsylvania, West Virginia, Mayland,
and Northern Virginia. Another 17 percent also originated to the northwest of the site
but farther away, over Michigan, Indiana, and Ohio. Nearly 30 percent of back
trajectories originated from the west, southwest, and south of the site, over western
Virginia and the Carolinas. The relatively short cluster trajectory originating over
southeast Virginia (26 percent) represents short back trajectories originating from a
variety of directions as well as back trajectories originating from the east, southeast,
and south of the site, primarily over eastern North Carolina but also over the offshore
waters of Virginia, North Carolina, and South Carolina.
27-8
-------
Figure 27-3. Composite Back Trajectory Map for RIVA
Figure 27-4. Back Trajectory Cluster Map for RIVA
27-9
-------
27.2.4 Wind Rose Comparison
Hourly surface wind data from the weather station at Richmond International Airport
near RIVA were uploaded into a wind rose software program to produce customized wind roses,
as described in Section 3.5.2.2. 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 27-5 presents a map showing the distance between the weather station and RIVA,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 27-5 also presents three different wind roses for the
RIVA monitoring site. First, a historical wind rose representing 2002 to 2011 wind data is
presented, which shows the predominant surface wind speed and direction over an extended
period of time. Second, a wind rose representing wind observations for all of 2012 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2012 is
presented. These can be used to identify the predominant wind speed and direction for 2012 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.
Observations from Figure 27-5 for RIVA include the following:
• The Richmond International Airport weather station is located 5.5 miles
east-southeast of RIVA.
• The historical wind rose shows that the most commonly observed wind direction is
north, although winds from the north-northeast, south, south-southwest, and
southwest were also frequently observed. Winds from the southeast and northwest
quadrants were observed less frequently. Calm winds (< 2 knots) were observed for
approximately 15 percent of the hourly wind measurements.
• The 2012 wind rose resembles the historical wind rose in some ways but exhibits
differences as well. Northerly, southerly and south-south westerly winds were still
prominent but accounted for a higher percentage of observations in 2012 while fewer
southwesterly to westerly and northeasterly winds were observed. Calm winds were
observed slightly more often in 2012.
• Southerly winds account for the greatest number of wind observations on sample days
near RIVA (approximately 13 percent), followed by south-southwesterly winds
(roughly 11 percent), both of which are greater than the number of northerly wind
observations (10 percent). The calm rate on sample days is nearly 18 percent.
27-10
-------
Figure 27-5. Wind Roses for the Richmond International Airport Weather Station near
RIVA
Location of RIVA and Weather Station
2002-2011 Historical Wind Rose
2012 Wind Rose
Sample Day Wind Rose
27-11
-------
27.3 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 27-4.
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 27-4. It is
important to note which pollutants were sampled for at each site when reviewing the results of
this analysis. RIVA sampled for PAHs and hexavalent chromium.
Table 27-4. Risk-Based Screening Results for the Virginia Monitoring Site
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
East Highland Park, Virginia - RIVA
Naphthalene
Fluorene
Acenaphthene
Benzo(a)pyrene
0.029
0.011
0.011
0.00057
Total
55
2
1
1
59
56
55
56
33
200
98.21
3.64
1.79
3.03
29.50
93.22
3.39
1.69
1.69
93.22
96.61
98.31
100.00
Observations from Table 27-4 include the following:
• Although four PAHs failed screens for RIVA, naphthalene contributed to 93 percent
of the total failed screens, while the other pollutants accounted for only one or two
failed screens each.
• Naphthalene and fluorene contributed to 95 percent of failed screens for RIVA and
therefore were identified as pollutants of interest for this site.
• Naphthalene failed greater than 98 percent of its screens, with 55 of 56 measured
detections of naphthalene failing screens. Conversely, only four percent of fluorene
concentrations failed screens.
27-12
-------
27.4 Concentrations
This section presents various concentration averages used to characterize pollution 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 the
site.
• Annual concentration averages are presented graphically for the 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 the site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Additional site-specific statistical summaries for RIVA are
provided in Appendices M and O.
27.4.1 2012 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 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 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 Virginia monitoring site are
presented in Table 27-5, 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.
27-13
-------
Table 27-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Virginia Monitoring Site
Pollutant
#of
Measured
Detections
vs. # 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
Fluorene
Naphthalene
55/56
56/56
3.22
±0.79
100.00
±20.81
5.14
±1.52
87.51
± 28.20
5.44
±1.62
73.84
± 14.44
2.84
±0.63
114.50
±33.71
4.16
±0.65
93.95
± 12.47
Observations for RIVA from Table 27-5 include the following:
• The quarterly and annual average concentrations of naphthalene are significantly
higher than the annual average concentrations of fluorene.
• Concentrations of naphthalene appear higher during the first and fourth quarters of
the year, although the confidence intervals indicate that naphthalene concentrations
measured at RIVA are fairly variable. Concentrations of naphthalene measured at
RIVA range from 23.9 ng/m3 to 268 ng/m3. Of the 21 concentrations greater than
100 ng/m3, five were measured during the first quarter, five were measured during the
second quarter, one was measured during the third quarter, and nine were measured
during the fourth quarter. Four of the five naphthalene concentrations greater than
100 ng/m3 and measured during the second quarter were measured in June.
• Concentrations of fluorene measured at RIVA span an order of magnitude, ranging
from 1.17 ng/m3 to 11.3 ng/m3. The quarterly averages of fluorene appear higher in
the warmer months of the year and lower in the colder months of the year, although
the differences are not statistically significant. Of the nine fluorene concentrations
greater than 6 ng/m3, eight were measured between June and August.
27.4.2 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 two pollutants
shaded in gray in Table 27-4 for RIVA. Figures 27-6 and 27-7 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 3.5.3.1.
27-14
-------
Figure 27-6. Program vs. Site-Specific Average Fluorene Concentration
RIVA
10 20 30 40 50 60
Concentration (ng/m3)
70
80
90 100
Program:
Site:
IstQuartile
Site Average
0
2ndQuartile SrdQuartile
Site Concentration Range
^^^^—
4thQuartile Average
Figure 27-7. Program vs. Site-Specific Average Naphthalene Concentration
RIVA
ij
1
|
3 100 200 300 400 500 600 700 800 9C
Concentration {ng/m3)
Program:
Site:
IstQuartile
D
Site Average
o
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
Observations from Figures 27-6 and 27-7 include the following:
• Figure 27-6 is the box plot for fluorene. This box plot shows that the annual
average concentration for RIVA is about halfway between the program-level
median and average concentration. Figure 27-6 also shows that the maximum
fluorene concentration measured at RIVA is considerably less than the maximum
concentration measured across the program. A single non-detect of fluorene was
measured at RIVA.
• Figure 27-7 is the box plot for naphthalene and shows that the annual average
concentration of naphthalene for RIVA is just greater than the program-level
average concentration. The maximum naphthalene concentration measured at
RIVA is considerably less than the program-level maximum concentration. There
were no non-detects of naphthalene measured at RIVA or across the program.
27-15
-------
27.4.3 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 3.5.3.2.
RIVA did not begin sampling PAHs under the NMP until October 2008. Because a minimum of
6 months of sampling is required for inclusion in the trends analysis and 2008 does not meet this
criterion, the trends analysis was not conducted.
27.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk at the
RIVA monitoring site. Refer to Sections 3.3 and 3.5.3 for definitions and explanations regarding
the various toxicity factors, time frames, and calculations associated with these risk-based
screenings.
27.5.1 Risk-Based Screening Assessment Using MRLs
A risk-based screening was conducted by comparing the concentration data from the
Virginia monitoring site to the ATSDR MRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 day to 14 days); intermediate (exposures of 15 days to 364 days); and chronic
(exposures of 1 year or greater). The preprocessed daily measurements of the pollutants of
interest were compared to the acute MRLs; the quarterly averages were compared to the
intermediate MRLs; and the annual averages were compared to the chronic MRLs.
As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2012.
27.5.2 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 3.5.3.4 for an explanation of how cancer risk and noncancer hazard
27-16
-------
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 27-6, where applicable. Cancer risk approximations are presented as
probabilities while the noncancer hazard approximations are ratios and thus, unitless values.
Table 27-6. Risk Approximations for the Virginia Monitoring Site
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections vs.
# of Samples
Annual
Average
(ng/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
East Highland Park, Virginia - RIVA
Fluorene
Naphthalene
0.000088
0.000034
0.003
55/56
56/56
4.16
±0.65
93.95
± 12.47
0.37
3.19
0.03
— = A Cancer URE or Noncancer RfC is not available.
Observations for RIVA from Table 27-6 include the following:
• The annual average concentration of naphthalene is greater than the annual average
concentration of fluorene.
• The cancer risk approximation for naphthalene is 3.19 in-a-million. The cancer risk
approximation for fluorene is less than 1.0 in-a-million.
• Only naphthalene has a noncancer toxicity factor. The noncancer hazard
approximation for naphthalene is considerably less than 1.0 (0.03), indicating that no
adverse health effects are expected from this individual pollutant.
27.5.3 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 27-7 presents the 10 pollutants with the highest emissions from the 2011 NEI that have
cancer toxicity factors. Table 27-7 also presents the 10 pollutants with the highest toxicity -
weighted emissions, based on the weighting schema described in Section 3.5.3.5. Lastly,
Table 27-7 provides the 10 pollutants with the highest cancer risk approximations (in-a-million)
for RIVA, as presented in Table 27-6. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 27-7. Table 27-8 presents similar
information, but is limited to those pollutants with noncancer toxicity factors.
27-17
-------
Table 27-7. 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)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
East Highland Park, Virginia (Henrico County) - RIVA
Benzene
Formaldehyde
Acetaldehyde
Ethylbenzene
1,3 -Butadiene
Tetrachloroethylene
Naphthalene
POM, Group 2b
POM, Group 2d
Trichloroethylene
105.08
100.16
55.95
51.83
18.95
17.17
10.37
2.59
1.60
0.85
Formaldehyde
Benzene
1,3 -Butadiene
POM, Group 3
Naphthalene
POM, Group 2b
POM, Group 2d
Ethylbenzene
Acetaldehyde
POM, Group 5a
1.30E-03
8.20E-04
5.68E-04
4.57E-04
3.52E-04
2.28E-04
1.41E-04
1.30E-04
1.23E-04
8.27E-05
Naphthalene 3.19
Fluorene 0.37
to
oo
-------
Table 27-8. 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)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations (Site-
Specific)
Noncancer Hazard
Approximation
Pollutant (HQ)
East Highland Park, Virginia (Henrico County) - RIVA
Toluene
Ethylene glycol
Xylenes
Hexane
Methanol
Benzene
Formaldehyde
Acetaldehyde
Ethylbenzene
Methyl isobutyl ketone
833.14
383.38
205.40
197.92
181.20
105.08
100.16
55.95
51.83
24.42
Acrolein
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Benzene
Naphthalene
Xylenes
Arsenic, PM
Ethylene glycol
Lead, PM
303,722.52
10,220.27
9,474.82
6,216.90
3,502.59
3,455.66
2,054.03
1,052.35
958.44
807.68
Naphthalene 0.03
to
-------
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 3.5.3.5. Similar to the cancer risk and
noncancer hazard approximations provided in Section 27.5.2, this analysis may help policy-
makers prioritize their air monitoring activities.
Observations from Table 27-7 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, one of two pollutants of interest for RIVA, has the seventh highest
emissions and the fifth highest toxicity-weighted emissions for Henrico County.
• POM, Group 2b is the eighth highest emitted "pollutant" in Henrico County and ranks
sixth for toxicity-weighted emissions. POM, Group 2b includes several PAHs
sampled for at RIVA, including fluorene. Acenaphthene, which failed one screen for
RIVA, is also part of this group.
• Several other POM Groups also appear in Table 27-7, particularly for toxicity-
weighted emissions. POM, Group 2d appears on both emissions-based lists for
Henrico County and includes anthracene, phenanthrene, and pyrene. POM, Groups 3
and 5a are also listed among those with the highest toxicity-weighted emission. POM,
Group 5a includes benzo(a)pyrene, which failed a single screen for RIVA but is not a
pollutant of interest. None of the PAHs sampled for at RIVA are included in POM,
Group 3.
Observations from Table 27-8 include the following:
• Toluene, ethylene glycol, and xylenes are the highest emitted pollutants with
noncancer RfCs in Henrico County.
• The pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) are acrolein, formaldehyde, and 1,3-butadiene.
27-20
-------
• Five 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.
27.6 Summary of the 2012 Monitoring Data for RIVA
Results from several of the data treatments described in this section include the
following:
»«» Four PAHs failed screens for RIVA, with naphthalene accounting for greater than
90 percent of the total failed screens. Hexavalent chromium did not fail any screens.
»«» Naphthalene andfluorene were identified as pollutants of interest for RIVA. The
annual average concentration of naphthalene was significantly higher than the
annual average concentration offluorene.
27-21
-------
28.0 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.
28.1 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 NATTS site in Washington is located in Seattle. Figure 28-1 is a composite satellite
image retrieved from ArcGIS Explorer showing the monitoring site and its immediate
surroundings. Figure 28-2 identifies nearby point source emissions locations by source category,
as reported in the 2011 NEI for point sources. Note that only sources within 10 miles of the site
are included in the facility counts provided in Figure 28-2. 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 radius
are still visible on the map, but have been grayed out in order to show emissions sources just
outside the boundary. Table 28-1 provides supplemental geographical information such as land
use, location setting, and locational coordinates.
28-1
-------
Figure 28-1. Seattle, Washington (SEWA) Monitoring Site
to
oo
to
-------
Figure 28-2. NEI Point Sources Located Within 10 Miles of SEWA
122*40'0"W 122°35'0"W 122°30'0"W 122°25'0"W 122°20'0"W 122°15'0"W 122°10'0"W
122;30'0"W 122C25'0"W 122=20'0"W
Legend
122a15'0"W 122°10'0"W 122°5'Q"W 122°0'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
County boundary
SEWA NATTS site Q 10 mile radius
Source Category Group (No. of Facilities)
•& Aerospace/Aircraft Manufacturing (2) A Metal Coating, Engraving, and Allied Services to Manufacturers (1)
f Airport/Airline/Airport Support Operations (27) © Metals Processing/Fabrication (4)
8 Automobile/Truck Manufacturing (2) ? Miscellaneous Commercial/Industrial (1)
c= Brick, Structural Clay, or Clay Ceramics Plant (2) "g Paint and Coating Manufacturing (1)
B Bulk Terminals/Bulk Plants (2) 7 Portland Cement Manufacturing (1)
6 Electrical Equipment Manufacturing (1) X Rail Yard/Rail Line Operations (2)
F Food Processing/Agriculture (2) ^ Ship/Boat Manufacturing or Repair Facility (1)
*} Glass Plant (1) W Steel Mill (1)
O Institution (school, hospital, prison, etc.) (1) * Wastewater Treatment (1)
28-3
-------
Table 28-1. Geographical Information for the Washington Monitoring Site
Site
Code
SEWA
AQS Code
53-033-0080
Location
Seattle
County
King
Micro- or
Metropolitan
Statistical Area
Seattle-Tacoma-
Bellevue, WA
MSA
Latitude
and
Longitude
47.568236,
-122.308628
Land Use
Residential
Location
Setting
Urban/City
Center
Additional Ambient Monitoring Information1
Haze, CO, SO2, NOy, NO, O3, Meteorological
parameters, PM Coarse, PM10, Black Carbon, PM2 5,
PM2 5 Speciation, IMPROVE Speciation.
BOLD ITALICS = EPA-designated NATTS Site
to
oo
-------
The SEWA monitoring site is located in Seattle, at the southeast corner of the Beacon
Hill Reservoir. The reservoir and the Jefferson Park Golf Course to the east are separated by
Beacon Avenue. A middle school and a hospital can be seen to the south of the site in the
bottom-most portion of Figure 28-1. 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 and intersects with 1-90 a couple of miles to the north of the site. The area to the west of
1-5 is 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
point sources located within 1 mile of SEWA are a metals processing and fabrication facility and
a food processing facility, as shown in Figure 28-2.
Table 28-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Washington monitoring site. Table 28-2 includes both county-
level population and vehicle registration information. Table 28-2 also contains traffic volume
information for SEWA as well as the location for which the traffic volume was obtained.
Additionally, Table 28-2 presents the county-level daily VMT for King County.
Table 28-2. Population, Motor Vehicle, and Traffic Information for the Washington
Monitoring Site
Site
SEWA
Estimated
County
Population1
2,007,440
County-level
Vehicle
Registration2
1,403,968
Annual
Average Daily
Traffic3
224,000
Intersection
Used for
Traffic Data
1-5, south of the Columbian Way
exit/Spokane St. Viaduct
County-
level Daily
VMT4
23,044,858
Bounty-level population estimate reflects 2012 data (Census Bureau, 2013c)
2County-level vehicle registration reflects 2012 data (WA DOL, 2012)
3AADT reflects 2012 data (WA DOT, 2012a)
4County-level VMT reflects 2012 data (WA DOT, 2012b)
BOLD ITALICS = EPA-designaled NATTS Site
Observations from Table 28-2 include the following:
• King County has the sixth highest county-level population among counties with NMP
sites.
• King County has the seventh highest county-level vehicle registration among counties
with NMP sites.
28-5
-------
• The traffic volume experienced near SEWA is the fourth highest compared to other
NMP monitoring sites. The traffic estimate provided is for 1-5 south of the Columbian
Way exit/Spokane Street Viaduct.
• The daily VMT for King County is in the top third compared to other counties with
NMP sites (where VMT data were available).
28.2 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.
28.2.1 Climate Summary
The city of Seattle is located between Puget Sound and Lake Washington. The entire
urban area is situated between the Olympic Mountains to the west and the Cascades to the east.
The area experiences a mild climate as the mountains moderate storm systems that move into the
Pacific Northwest and both the mountains and the sound shield the city from temperature
extremes. Although the city is known for its cloudy, rainy conditions, actual precipitation totals
tend to be comparable or less than many locations east of the Rocky Mountains. The majority of
precipitation falls during the winter months, with monthly totals greater than 5 inches common
between November and January while less than 2 inches is typical during the summer. Normal
annual snowfall amounts are around 10 inches. Prevailing winds in the Seattle area are out of the
south to southwest for much of the year (Wood, 2004; WRCC 2013).
28.2.2 Meteorological Summary
Hourly meteorological data for 2012 were retrieved from NCDC for the weather station
closest to the Washington monitoring site (NCDC, 2012), as described in Section 3.5.2. The
closest weather station to SEWA is located at Boeing Field/King County International Airport
(WBAN 24234). Additional information about this weather station, such as the distance between
the site and the weather station, is provided in Table 28-3. These data were used to determine
how meteorological conditions on sample days vary from conditions experienced throughout the
year.
28-6
-------
Table 28-3. Average Meteorological Conditions near the Washington Monitoring Site
Closest
Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Seattle, Washington - SEWA
Boeing Field/
King County
Intl Airport
94.9^4.
(47.53, -122.30)
2.6
miles
1CQO
(S)
Days
(68)
2012
59.8
±2.7
59.1
+ 1.2
53.0
±2.2
52.4
+ 1.0
43.2
±1.9
43.1
+ 0.8
48.2
±1.8
47.8
+ 0.8
72.0
±2.7
72.9
+ 1.1
1015.9
±1.8
1016.2
+ 0.7
4.6
±0.6
4.5
+ 0.2
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
to
oo
-------
Table 28-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for all of 2012. Also included in Table 28-3 is the 95 percent
confidence interval for each parameter. As shown in Table 28-3, average meteorological
conditions on sample days were representative of average weather conditions experienced
throughout the year.
28.2.3 Back Trajectory Analysis
Figure 28-3 is the composite back trajectory map for days on which samples were
collected at the SEWA monitoring site. Included in Figure 28-3 are four back trajectories per
sample day. Figure 28-4 is the corresponding cluster analysis. An in-depth description of these
maps and how they were generated is presented in Section 3.5.2.1. For the composite map, each
line represents the 24-hour trajectory along which a parcel of air traveled toward the monitoring
site on a given sample day and time, based on an initial height of 50 meters AGL. For the cluster
analysis, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric circle around the site in Figures 28-3 and 28-4 represents 100 miles.
Figure 28-3. Composite Back Trajectory Map for SEWA
28-8
-------
Figure 28-4. Back Trajectory Cluster Map for SEW A
Observations from Figures 28-3 and 28-4 for SEWA include the following:
• Back trajectories originated from a variety of directions from SEWA, although less
frequently from the northeast quadrant. The longest back trajectories originated
offshore.
• The 24-hour air shed domain for SEWA is smaller in size compared to many other
NMP sites. Although the longest trajectory originated 800 miles away over the
Pacific Ocean, the average back trajectory length was less than 200 miles long and
nearly 85 percent of trajectories originated within 300 miles of the site.
• The cluster analysis shows that 37 percent of back trajectories are represented by the
short cluster trajectory originating over the Puget Sound (and presented in the insert
map in Figure 28-4). This cluster trajectory includes back trajectories originating
from nearly any direction and generally less than 100 miles from the monitoring site.
Twenty-one percent of back trajectories originated over northwest Washington,
Vancouver Island, and the adjacent waters. Three percent of back trajectories
originated well offshore and over the Pacific Ocean. Nearly one-quarter of back
trajectories originated over southwest Washington and the western half of Oregon.
Another 14 percent of back trajectories originated primarily over southeast
Washington and northeast Oregon.
28-9
-------
28.2.4 Wind Rose Comparison
Hourly surface wind data from the weather station at Boeing Field/King County
International Airport were uploaded into a wind rose software program to produce customized
wind roses, as described in Section 3.5.2.2. 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 28-5 presents a map showing the distance between the weather station and SEW A,
which may be useful for identifying topographical influences that can affect the meteorological
patterns experienced at this location. Figure 28-5 also presents three different wind roses for the
SEWA monitoring site. First, a historical wind rose representing 2002 to 2011 wind data is
presented, which shows the predominant surface wind speed and direction over an extended
period of time. Second, a wind rose representing wind data for all of 2012 is presented. Next, a
wind rose representing wind data for days on which samples were collected in 2012 is presented.
These can be used to identify the predominant wind speed and direction for 2012 and to
determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.
Observations from Figure 28-5 for SEWA include the following:
• The Boeing Field/King County Airport weather station is located 2.7 miles south of
SEWA.
• The historical wind rose shows that southeasterly, south-southeasterly, and southerly
winds were frequently observed, accounting for nearly 40 percent of observations.
Calm winds (< 2 knots) accounted for 24 percent of wind observations near SEWA.
• The wind patterns shown on the 2012 wind rose are similar to the historical wind
patterns, although the percentage of calm winds is slightly higher (nearly 28 percent)
and the percentage of south-southeasterly winds is slightly lower in 2012.
• The wind patterns shown on the sample day wind rose resemble the 2012 wind
patterns, indicating that conditions on sample days were representative of those
experienced over the entire year (and historically).
28-10
-------
Figure 28-5. Wind Roses for the Boeing Field/King County International Airport Weather
Station near SEWA
Location of SEWA and Weather Station
2002-2011 Historical Wind Rose
Weather
Station
2012 Wind Rose
Sample Day Wind Rose
Calms: 27.50%
28-11
-------
28.3 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 28-4. 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 28-4. It is important to note which
pollutants were sampled for at the site when reviewing the results of this analysis. SEWA
sampled for PMi0 metals, VOCs, PAHs, carbonyl compounds, and hexavalent chromium.
Table 28-4. Risk-Based Screening Results for the Washington Monitoring Site
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Seattle, Washington - SEWA
Benzene
Carbon Tetrachloride
Formaldehyde
1 ,2-Dichloroethane
Arsenic (PM10)
1,3 -Butadiene
Naphthalene
Acetaldehyde
Manganese (PM10)
Nickel (PM10)
Ethylbenzene
Acenaphthene
Hexavalent Chromium
Fluorene
Lead (PM10)
0.13
0.17
0.077
0.038
0.00023
0.03
0.029
0.45
0.005
0.0021
0.4
0.011
0.000083
0.011
0.015
Total
60
60
60
57
56
55
53
43
29
22
8
3
3
2
1
512
60
60
60
57
59
58
59
60
59
59
60
58
53
59
59
880
100.00
100.00
100.00
100.00
94.92
94.83
89.83
71.67
49.15
37.29
13.33
5.17
5.66
3.39
1.69
58.18
11.72
11.72
11.72
11.13
10.94
10.74
10.35
8.40
5.66
4.30
1.56
0.59
0.59
0.39
0.20
11.72
23.44
35.16
46.29
57.23
67.97
78.32
86.72
92.38
96.68
98.24
98.83
99.41
99.80
100.00
Observations from Table 28-4 for SEWA include the following:
• Fifteen pollutants failed at least one screen for SEWA; 58 percent of concentrations
for these 15 pollutants were greater than their associated risk screening value (or
failed screens).
• Ten pollutants contributed to 95 percent of failed screens for SEWA and therefore
were identified as pollutants of interest for the site. These 10 include two carbonyl
compounds, four VOCs, three PMi0 metals, and one PAH.
28-12
-------
• Benzene, carbon tetrachloride, and formaldehyde were detected in every valid sample
collected at SEWA and failed 100 percent of screens. 1,2-Dichloroethane also failed
100 percent of screens for SEWA, but was not detected in every sample collected.
28.4 Concentrations
This section presents various concentration averages used to characterize pollution 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 the
site.
• Annual concentration averages are presented graphically for the 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 the site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Additional site-specific statistical summaries for SEWA
are provided in Appendices J, L, M, N, and O.
28.4.1 2012 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 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 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 Washington monitoring site are
presented in Table 28-5, 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.
28-13
-------
Table 28-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Washington Monitoring Site
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Ug/m3)
2nd
Quarter
Average
(Ug/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Ug/m3)
Seattle, Washington - SEWA
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
Formaldehyde
Arsenic (PM10)a
Manganese (PM10)a
Naphthalene3
Nickel (PM10)a
60/60
60/60
58/60
60/60
57/60
60/60
59/59
59/59
59/59
59/59
0.63
±0.19
0.75
±0.18
0.09
±0.03
0.68
±0.06
0.08
±0.01
0.47
±0.23
0.65
±0.30
8.20
±4.14
72.97
± 24.56
2.22
±1.31
0.52
±0.10
0.45
±0.07
0.05
±0.01
0.71
±0.02
0.08
±0.01
0.37
±0.09
0.61
±0.19
9.95
±5.99
52.01
± 14.50
2.29
±0.91
1.15
±0.30
0.53
±0.16
0.09
±0.04
0.68
±0.03
0.05
±0.01
0.83
±0.27
0.78
±0.20
14.13
±6.76
93.82
±28.34
5.14
±1.84
0.62
±0.14
0.64
±0.10
0.10
±0.03
0.73
±0.04
0.07
±0.0.1
0.44
±0.10
0.67
±0.27
6.43
±5.87
61.39
±13.01
1.08
±0.30
0.74
±0.12
0.59
±0.07
0.08
±0.02
0.70
±0.02
0.07
±0.01
0.53
±0.10
0.68
±0.11
9.80
±2.88
70.87
± 10.90
2.74
±0.71
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
Observations from Table 28-5 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 (0.74 ± 0.12 |ig/m3), carbon tetrachloride (0.70 ± 0.02 |ig/m3), benzene
(0.59 ± 0.07 (ig/m3), and formaldehyde (0.53 ± 0.10 |ig/m3).
• Even though acetaldehyde has the highest annual average concentration among
SEWA's pollutants of interest, this annual average is one of the lowest among other
NMP sites sampling carbonyl compounds. SEWA's annual average concentration of
formaldehyde is the lowest among all NMP sites. No other NMP site has an annual
average concentration of formaldehyde less than 1 |ig/m3. Similar observations were
made in previous NMP reports.
• The third quarter average acetaldehyde concentration is significantly higher than the
other quarterly average concentrations and has a larger confidence interval. Of the 10
concentrations of acetaldehyde greater than 1 |ig/m3 measured at SEWA, seven were
collected in the third quarter of 2012, including the maximum concentration of
3.12 |ig/m3, which was measured on September 18, 2012.
28-14
-------
• The maximum formaldehyde concentration was also measured on September 18th
(2.67 |ig/m3). The second highest formaldehyde concentration was measured in
January (1.76 |ig/m3). The only other formaldehyde concentration greater than
1 |ig/m3 was measured in August. The remaining 60 concentrations are less than
1 |ig/m3 and have a median concentration of 0.41 |ig/m3. This explains the large
confidence intervals associated with the first and third quarter averages of
formaldehyde.
• The maximum benzene concentration was also measured at SEWA on
September 18th (1.47 |ig/m3). Six measurements of benzene greater than 1 |ig/m3
were measured at SEWA (one in January, two in February, two in September, and
one in October).
• Of the metal pollutants of interest for SEWA, manganese has the highest annual
average concentration (9.80 ± 2.88 ng/m3), followed by nickel (2.74 ± 0.71 ng/m3)
and arsenic (0.68 ± 0.11 ng/m3).
• The third quarter average concentration of manganese is higher than the other
quarterly averages, although all of the quarterly averages have relatively large
confidence intervals. This indicates a relatively high level of variability in the
measurements. Concentrations of manganese range from 0.767 ng/m3 to 45.0 ng/m3,
with a median concentration of 4.91 ng/m3. The maximum concentration of
manganese was measured on October 18, 2012, although concentrations of similar
magnitude were also measured on September 8th and September 18th (the same day
the maximum concentrations of acetaldehyde, formaldehyde, and benzene were
measured).
• The highest concentration of arsenic was also measured on October 18, 2012
(2.02 ng/m3). This is the only arsenic measurement greater than 2 ng/m3 measured at
SEWA. Ten additional arsenic concentrations greater than 1 ng/m3 were measured at
SEWA and are spread across the calendar quarters (two each in the first and second
quarter, four in the third quarter, and three in the fourth quarter).
• The third quarter average concentration of nickel is more than twice the other
quarterly averages and has a larger confidence intervals associated with it (although
the first and second quarterly averages also have relatively large confidence
intervals). A review of the data shows that concentrations of nickel range from
0.495 ng/m3 to 14.3 ng/m3. The maximum concentration of nickel was measured on
September 18, 2012, the same day as several of SEWA's other pollutants of interest
and is the second highest nickel concentration measured among NMP sites sampling
this pollutant. Of the 25 nickel concentrations greater than 5 ng/m3 measured across
the program, eight were measured at SEWA (which is the highest for any single site).
Of the 20 concentrations greater than 3 ng/m3 measured at SEWA, three were
measured during the first quarter of 2012, five in the second quarter, 12 in the third
quarter, and none in the fourth quarter. This explains why the fourth quarter average
concentration is less than the other averages.
28-15
-------
• The third quarter average concentration of naphthalene is greater than the other
quarterly averages and has a large confidence interval associated with it. A review of
the data shows that naphthalene concentrations measured at SEWA range from
17.1 g/m3 to 234 ng/m3, with a median concentration of 61.8 ng/m3. The maximum
concentration of naphthalene was also measured on September 18, 2012. Two
additional concentrations greater than 150 ng/m3 were measured in August and
September.
Tables 4-9 through 4-12 present the 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 third highest annual
average concentration of carbon tetrachloride among sites sampling VOCs. Note,
however, that concentrations of carbon tetrachloride in Table 4-9 span only
0.03 |ig/m3.
• SEWA does not appear in Table 4-10 for carbonyl compounds or Table 4-11 for
PAHs.
• As shown in Table 4-12, SEWA has the second highest annual average concentration
of nickel among all sites sampling metals (PMW and TSP), behind only ASKY-M.
SEWA had the highest annual average nickel concentration for the 2010 and 2011
NMP reports.
• SEWA also has the fourth highest concentrations of manganese and ranks eighth
highest for arsenic.
28.4.2 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 shaded in
gray in Table 28-4 for SEWA. Figures 28-6 through 28-15 overlay the site's minimum, annual
average, and maximum concentrations onto the program-level minimum, first quartile, median,
average, third quartile, and maximum concentrations for each pollutant, as described in
Section 3.5.3.1.
28-16
-------
Figure 28-6. Program vs. Site-Specific Average Acetaldehyde Concentration
SEWA
9 12
Concentration {[og/m3)
15
18
Program:
Site:
IstQuartile
D
Site Average
0
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
21
Figure 28-7. Program vs. Site-Specific Average Arsenic (PMio) Concentration
4 5
Concentration (ng/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o —
Figure 28-8. Program vs. Site-Specific Average Benzene Concentration
Concentration {[og/m3)
Program:
Site:
IstQuartile
Site Average
O
2ndQuartile SrdQuartile
• n
i — i
Site Concentration Range
4thQuartile Average
28-17
-------
Figure 28-9. Program vs. Site-Specific Average 1,3-Butadiene Concentration
SEWA
fc
Program Max Concentration = 4.10 ug/m3
0.75 1 1.25
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o
Figure 28-10. Program vs. Site-Specific Average Carbon Tetrachloride Concentration
SEWA
2 3
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
O
Figure 28-11. Program vs. Site-Specific Average 1,2-Dichloroethane Concentration
-t-
1
] Program Max Concentration = 17.01 ug/m3
0.1
0.2
0.3
0.4 0.5 0.6
Concentration {[og/m3)
0.7
0.8 0.9
Program:
Site:
IstQuartile
•
Site Average
0
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^~
4thQuartile Average
D 1
28-18
-------
Figure 28-12. Program vs. Site-Specific Average Formaldehyde Concentration
SEWA
10
12
Concentration {[og/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
o —
14
Figure 28-13. Program vs. Site-Specific Average Manganese (PMio) Concentration
e
Program Max Concentration = 275 ng/m3
60 90
Concentration (ng/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
Site: Site Average Site Concentration Range
O
Figure 28-14. Program vs. Site-Specific Average Naphthalene Concentration
400 500
Concentration (ng/m3)
Program:
Site:
IstQuartile
Site Average
O
2ndQuartile SrdQuartile
Site Concentration Range
4thQuartile Average
28-19
-------
Figure 28-15. Program vs. Site-Specific Average Nickel (PMi0) Concentration
SEWA
8 10
Concentration (ng/m3)
12
14
16
18
Program:
Site:
IstQuartile
D
Site Average
o
2ndQuartile SrdQuartile
D D
Site Concentration Range
^^^^—
4thQuartile Average
D 1
Observations from Figures 28-6 through 28-15 include the following:
• Figure 28-6 shows that SEWA's annual average acetaldehyde concentration is
considerably less than the program-level average concentration for acetaldehyde
and is actually less than the program-level first quartile (25th percentile). This site
has the third lowest annual average concentration of acetaldehyde among NMP
sites sampling carbonyl compounds.
• Figure 28-7 shows that SEWA's annual average arsenic (PMio) concentration is
just less than the program-level average concentration of arsenic (PMio). The
maximum arsenic concentration measured at SEWA is considerably less than the
maximum concentration measured across the program. There were no non-detects
of arsenic measured at SEWA, although there were a few measured across the
program.
• Figure 28-8 shows that the annual average benzene concentration for SEWA is
less than the program-level average concentration as well as the program-level
median concentration. SEWA's annual average benzene concentration is the third
lowest annual average among sites sampling benzene. The maximum benzene
concentration measured at SEWA is considerably less than the maximum benzene
concentration measured across the program.
• Figure 28-9 is the box plot for 1,3-butadiene. Note that the program-level
maximum concentration (4.10 |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 2 |ig/m3. This figure shows that the annual average 1,3-butadiene
concentration for SEWA is less than the program-level average concentration but
greater than the program-level median concentration, although the difference
between the average and median concentrations is less than 0.04 |ig/m3. Figure
28-9 also shows that the maximum 1,3-butadiene concentration measured at
SEWA is considerably less than the maximum concentration measured across the
program. Two non-detects of 1,3-butadiene were measured at SEWA.
28-20
-------
• Figure 28-10 for carbon tetrachloride shows that the range of concentrations
measured at SEWA for this pollutant is relatively small. The annual average
concentration of carbon tetrachloride for SEWA is similar to the program-level
average and median concentrations (less than 0.012 |ig/m3 separates these three
values).
• Figure 28-11 is the box plot for 1,2-dichloroethane. Note that the program-level
maximum concentration (17.01 |ig/m3) is not shown directly on the box plot as
the scale has been reduced to 1 |ig/m3 in order to allow for the observation of data
points at the lower end of the concentration range. The program-level average
concentration is greater than the program third quartile for this pollutant and is
greater than or similar to the maximum concentration measured at most sites
sampling 1,2-dichloroethane. This is because the program-level average is being
driven by the higher measurements collected at a few monitoring sites.
Figure 28-11 shows that the maximum 1,2-dichloroethane concentration
measured at SEWA is similar to the program-level third quartile. The annual
average for SEWA is just less than the program-level median concentration but
greater than the first quartile. Three non-detects of 1,2-dichloroethane were
measured at SEWA.
• Figure 28-12 shows that SEWA's annual average formaldehyde concentration is
less than the program-level first quartile, similar to acetaldehyde. The entire range
of formaldehyde concentrations measured at SEWA is less than the program-level
average concentration. As previously discussed, SEWA has the lowest annual
average concentration of formaldehyde among NMP sites sampling carbonyl
compounds.
• Figure 28-13 is the box plot for manganese. The program-level maximum
concentration (275 ng/m3) is not shown directly on the box plot as the scale has
been reduced to 150 |ig/m3 in order to allow for the observation of data points at
the lower end of the concentration range. This figure shows that the annual
average concentration of manganese (PMio) for SEWA is just less than the
program-level average concentration. The maximum manganese concentration
measured at SEWA is considerably less than the maximum concentration
measured across the program. There were no non-detects of manganese measured
at SEWA.
• Figure 28-14 shows that the annual average concentration of naphthalene for
SEWA is less than the program-level average concentration. The maximum
naphthalene concentration measured at SEWA is considerably less than the
program-level maximum concentration. There were no non-detects of naphthalene
measured at SEWA or across the program.
• Figure 28-15 is the box plot for nickel. Although the maximum nickel
concentration measured at SEWA is not the maximum concentration measured
across the program, it is the second highest concentration program-wide. The
minimum concentration of nickel measured at SEWA is greater than the program-
level first quartile. SEWA's annual average concentration is the second highest
28-21
-------
among NMP sites sampling this pollutant and is more than twice the program-
level average concentration.
28.4.3 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 3.5.3.2.
Sampling for PMio metals, VOCs, and carbonyl compounds under the NMP began in 2007 and
sampling for PAHs began in 2008. Thus, Figures 28-16 through 28-25 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 is not provided, although the range and quartiles are still presented.
28-22
-------
Figure 28-16. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
SEWA
oncentration (Mg/m
i a
3 C
<
•
1 !
rh 1
a i — j> — i
2007 2008 2009 2010 2011 2012
Year
• 5th Percentile - Min mum ~ Median — Maximum O 95th Percentile ...^... Average
Observations from Figure 28-16 for acetaldehyde measurements collected 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 was considerably less (3.36 |ig/m3,
measured in September 2009). Only one other acetaldehyde concentration greater
than 3 |ig/m3 has been measured at SEWA and is the maximum concentration for
2012 measured on September 18th (3.12 |ig/m3).
• Even though the third highest acetaldehyde concentration was measured in 2012, the
1-year average acetaldehyde concentration is at a minimum for 2012 as compared to
the other years of sampling. However, the range is rather small, with the 1-year
average concentrations ranging from 0.74 |ig/m3 (2012) to 0.98 |ig/m3 (2009).
Confidence intervals calculated indicate that the 1-year average concentrations are not
statistically different.
• 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 from 2011 to 2012 (0.68 |ig/m3). These changes, though, are also relatively
small.
28-23
-------
Figure 28-17. Yearly Statistical Metrics for Arsenic (PMi0) Concentrations Measured at SEWA
O 5th Percentile - Minimurr
Median — Maximum O 95th Percentile
Observations from Figure 28-17 for arsenic (PMio) measurements collected 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 g/m3).
• The 1-year average concentration fluctuated only slightly between 2007 and 2009,
ranging from 0.69 ng/m3 (2008) to 0.76 |ig/m3 (2007). Although a decrease is shown
from 2009 to 2010, confidence intervals indicate that the change is not statistically
significant. Nearly all of the statistical parameters for 2011 returned to levels similar
to 2010. Little change in the 1-year average concentration is shown for 2012.
• There have been no non-detects of arsenic measured since the onset of sampling,
including in 2008, where it appears the minimum concentration is zero. For 2008, the
minimum is 0.011 ng/m3.
28-24
-------
Figure 28-18. Yearly Statistical Metrics for Benzene Concentrations Measured at SEWA
6.0
5.0
4.0
E
|3.0
1
2.0
1.0
0.0
I
1
^^
2007
O 5th Percentile
r^
— f —
2008
- Minimum
.....<
T
i rn i
>
^^•^^ ' ^\^^ . jjj^^_
I ^^ - 1
2009 2010 2011 2012
Year
Median — Maximum O 95th Percentile "-O-" Average
Observations from Figure 28-18 for benzene measurements collected 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 (2.48 |ig/m3,
measured in January 2011). Only five benzene concentrations greater than 2 |ig/m3
have been measured at SEWA.
• The 1-year average concentration of benzene ranges from 0.59 |ig/m3 (2012) to
0.81 |ig/m3 (2009). If the maximum concentration measured in 2009 was removed
from the calculation, the 1-year average concentration for 2009 would fall in line with
the others and the averages would exhibit a steady decreasing trend through 2010,
albeit very slight.
• The median concentration decreased from 2010 to 2011 because the number of
concentrations less than 0.4 |ig/m3 nearly doubled. However, the 1-year average
concentration increased because it is being driven by the higher concentrations
measured in 2011 (there are five concentrations measured in 2011 greater than the
maximum concentration measured in 2010).
• All of the statistical metrics exhibit slight decreases for 2012.
28-25
-------
Figure 28-19. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at SEWA
1.0
0.8
I"
|
1
1
3 0.4
0.2
0.0
rh
2007
T
• 5th Percentile
rn
i
2008
— Minimum
y
m ^L
~"n j^
n
f • Q I
2009 2010 2011 2012
Year
Median — Maximum 0 95th Percentile ...^>... Average
Observations from Figure 28-19 for 1,3-butadiene measurements collected 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 roughly half as high
(0.46 |ig/m3) and was measured on the same day in January 2011 as the second
highest benzene concentration.
• At least one non-detect 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 and 2011, both the minimum and 5th percentile are zero, indicating that the
number of non-detects has increased. Ten percent of the measurements were non-
detects for 2010, which increased to 15 percent for 2011. The number of non-detects
decreased to 3 percent for 2012.
• The 1-year average concentration has changed little over the course of sampling,
ranging from 0.064 |ig/m3 (2008) to 0.089 |ig/m3 (2011). Interestingly, the year with
the greatest number of non-detects (or zeros) also has the greatest number of
measurements greater than 0.2 |ig/m3 (seven).
• Little change is shown in the 1-year average and median concentration from 2011 to
2012. The decrease in the maximum concentration is balanced by the decrease in the
number of non-detects.
28-26
-------
Figure 28-20. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
SEWA
2009 2010
Year
O 5th Percentile - Minimurr
Median — Maximum • 95th Percentile ...^... Average
Observations from Figure 28-20 for carbon tetrachloride measurements collected at
SEWA include the following:
• Eighteen concentrations of carbon tetrachloride greater than 1.0 |ig/m3 have been
measured since the onset of sampling in 2007. All but one of these were measured in
2008 and 2009. The maximum carbon tetrachloride concentration (1.22 |ig/m3) has
been measured twice at SEWA, once in 2008 and once in 2010.
• All of the statistical metrics increased from 2007 to 2008, particularly the 1-year
average concentration. Between 2008 and 2011, a steady decreasing trend in the
concentrations is shown.
• The range of measurements compressed somewhat for 2012 and is the smallest range
of measurements since the onset of sampling. Yet, both the 1-year average and
median concentrations exhibit increases.
• The confidence intervals calculated for each year are very small, indicating that most
the concentrations fall within a relatively small range, particularly for 2012. The
difference between the median and 1-year average concentration is less than
0.03 |ig/m3 for each year, with one year having no difference. This indicates little
variability in the central tendency of this pollutant.
28-27
-------
Figure 28-21. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured
at SEWA
§
1
I
3 0.06
.. o
.0
O 5th Percentile - Minimum
Median — Maximum O 95th Percentile
Observations from Figure 28-21 for 1,2-dichloroethane measurements collected 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. By 2012, the percentage of measured detections is at 93 percent, a significant
increase from 26 percent in 2011.
• As the number of measured detections increased, particularly for 2012, the median
and 1-year average concentrations increased correspondingly. The median
concentration is actually greater than the 1-year average for 2012. This is because
there were still 14 non-detects (or zeros) factoring into the 1-year average
concentration for 2012, while the range of measured detections is rather small
(0.040 |ig/m3 to 0.095 |ig/m3).
28-28
-------
Figure 28-22. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
SEWA
Concentration (jig/m3)
<.
v...,
2007
T
• SthPercentile
0
•to
2008
— Min mum
Maximum
Concentration for
2009 is 16.6 u.g/mA
T
1 T
^" "t^ "^ ^^
2 1 Q 1 » 1 o 1
2009 2010 2011 2012
Year
Median — Maximum 0 95th Percentile ...^>... Average
Observations from Figure 28-22 for formaldehyde measurements collected 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. The fourth highest concentration is the
September 18, 2012 measurement (2.67 |ig/m3). A total of nine concentrations greater
than 2 |ig/m3 has been measured since the onset of carbonyl compound sampling at
SEWA.
• The 1-year average concentrations have an undulating pattern across the period of
sampling, with a "down" year followed by an "up" year. The 1-year average
formaldehyde concentration has ranged from 0.53 |ig/m3 (2012) to 1.04 |ig/m3
(2009).
• The level of variability in the measurements decreased significantly from 2009 to
2010. The difference between the 1-year average and median concentrations is less
than 0.1 |ig/m3 for 2010, 2011, and 2012. Further, the difference between the 5th and
95th percentiles is at a minimum for 2012. Thus, the majority of measurements fell
into a smaller range in 2012.
28-29
-------
Figure 28-23. Yearly Statistical Metrics for Manganese (PMi0) Concentrations Measured at
SEWA
Is
c
§
1
10.0
1
<
f ill
I 0
1 * 1
>• <;>.....
— o — 1 1 — o — 1 1 • 1 ^^^H ^»n 1 — a —
2007 2008 2009 2010 2011 2012
Year
• 5th Percentile - Minimum ~ Median — Max mum O 95th Percentile ...^... Average
Observations from Figure 28-23 for manganese (PMio) measurements collected at SEWA
include the following:
• The three highest manganese concentrations measured at SEWA were all measured in
2007 and are the only three measurements greater than 50 ng/m3 measured at this site,
although the maximum concentrations measured for several years are just less than
50 ng/m3.
• A steady decreasing trend in the 1-year average manganese concentration is shown
through 2010. The 95th percentiles also exhibit this decrease. The maximum and
median concentrations exhibit this trend for most years but not throughout the entire
4-year period.
• Most of the statistical metrics increased from 2010 to 2011. Although the 95th
percentile more than doubled and the 1-year average increased by 40 percent, the
median concentration increased just slightly and the minimum concentration
decreased. Additional increases are shown for most of the statistical parameters for
2012.
28-30
-------
Figure 28-24. Yearly Statistical Metrics for Naphthalene Concentrations Measured at SEWA
O 5th Percentile - Minimurr
— Maximum • 95th Percentile
A 1-year average is not presented because sampling under the NMP did not begin until March 2008.
Observations from Figure 28-24 for naphthalene measurements collected 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 is not presented for 2008, although the
range of measurements is provided.
• The maximum naphthalene concentration measured at SEWA was measured in 2011
(308 ng/m3). This is the only measurement greater than 250 ng/m3 measured at this
site. Seven additional measurements greater than 200 ng/m3 have been measured at
SEWA and are spread across the years of sampling, except 2008.
• The 1-year average concentrations of naphthalene have an undulating pattern across
the period of sampling, ranging from 61.44 ng/m3 (2010) to 78.67 ng/m3 (2009).
Little change in the 1-year average concentration is shown from 2011 to 2012.
28-31
-------
Figure 28-25. Yearly Statistical Metrics for Nickel (PM10) Concentrations Measured at SEWA
O
^m
-t-
O
^
-f-
o..
...o
O 5th Percentile - Minimurr
— Maximum O 95th Percentile
Observations from Figure 28-25 for nickel measurements collected at SEWA include the
following:
• The maximum concentration of nickel was measured at SEWA on
September 18, 2012. The largest range of measurements was collected during 2012.
Further, the range within which the majority of concentrations fall (as determined by
the 5th and 95th percentiles) is largest for 2012.
• The maximum and 1-year average concentrations exhibit an increasing trend between
2007 and 2009, after which a decrease in shown for 2010. Although the maximum
concentration decreased for 2011, the 95th percentile increased while little change is
shown for the 1-year average and median concentrations. All of the statistical metrics
exhibit increases for 2012. However, confidence intervals calculated on the dataset
indicate that the changes shown are not statistically significant.
• The difference between the 1-year average and median concentrations is greater than
0.65 ng/m3 for all years and greater than 1.0 ng/m3 for 2012. This indicates that there
is considerable variability in the measurements of nickel.
28-32
-------
28.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk at the
Washington monitoring site. Refer to Sections 3.3 and 3.5.3 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.
28.5.1 Risk-Based Screening Assessment Using MRLs
A risk-based screening was conducted by comparing the concentration data from the
Washington monitoring site to the ATSDR MRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 day to 14 days); intermediate (exposures of 15 days to 364 days); and chronic
(exposures of 1 year or greater). The preprocessed daily measurements of the pollutants of
interest were compared to the acute MRLs; the quarterly averages were compared to the
intermediate MRLs; and the annual averages were compared to the chronic MRLs.
As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2012.
28.5.2 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 3.5.3.4 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 28-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
28-33
-------
Table 28-6. Risk Approximations for the Washington Monitoring Site
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer Risk
Approximation
(HQ)
Seattle, Washington - SEWA
Acetaldehyde
Arsenic (PM10)a
Benzene
1,3 -Butadiene
Carbon Tetrachloride
1 ,2-Dichloroethane
Formaldehyde
Manganese (PM10) a
Naphthalene a
Nickel (PM10)a
0.0000022
0.0043
0.0000078
0.00003
0.000006
0.000026
0.000013
0.000034
0.00048
0.009
0.000015
0.03
0.002
0.1
2.4
0.0098
0.00005
0.003
0.00009
60/60
59/59
60/60
58/60
60/60
57/60
60/60
59/59
59/59
59/59
0.74
±0.12
0.01
±O.01
0.59
±0.07
0.08
±0.02
0.70
±0.02
0.07
±O.01
0.53
±0.10
0.01
± O.01
0.07
±0.01
O.01
±O.01
1.63
2.92
4.58
2.55
4.19
1.70
6.91
2.41
1.31
0.08
0.05
0.02
0.04
0.01
O.01
0.05
0.20
0.02
0.03
— = A Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 28-5.
Observations from Table 28-6 for SEWA include the following:
• The pollutants with the highest annual average concentrations for SEWA are
acetaldehyde, carbon tetrachloride, benzene, and formaldehyde.
• The pollutants with the highest cancer risk approximations are formaldehyde,
benzene, carbon tetrachloride, and arsenic. Although the pollutant with the highest
cancer risk approximation is formaldehyde, its cancer risk approximation is the
lowest among NMP sites sampling carbonyl compounds.
• The noncancer hazard approximations for SEWA are all less than 1.0, with the
highest calculated for manganese (0.20), indicating that no adverse health effects are
expected from these individual pollutants.
28-34
-------
28.5.3 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 28-7 presents the 10 pollutants with the highest emissions from the 2011 NEI that have
cancer toxicity factors. Table 28-7 also presents the 10 pollutants with the highest toxicity -
weighted emissions, based on the weighting schema described in Section 3.5.3.5. Lastly,
Table 28-7 provides the 10 pollutants with the highest cancer risk approximations (in-a-million)
for the site, as presented in Table 28-6. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 28-7. Table 28-8 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 3.5.3.5. Similar to the cancer risk and
noncancer hazard approximations provided in Section 28.5.2, this analysis may help policy-
makers prioritize their air monitoring activities.
28-35
-------
Table 28-7. 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)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Seattle, Washington (King County) - SEWA
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Tetrachloroethylene
Naphthalene
POM, Group 2b
POM, Group 2d
Trichloroethylene
900.20
750.12
465.05
421.81
138.68
95.67
87.42
16.86
11.80
11.73
Formaldehyde
Benzene
Hexavalent Chromium, PM
1,3 -Butadiene
POM, Group 3
Naphthalene
POM, Group 2b
Ethylbenzene
POM, Group 2d
Acetaldehyde
9.75E-03
7.02E-03
4.47E-03
4.16E-03
3.86E-03
2.97E-03
1.48E-03
1.16E-03
1.04E-03
9.28E-04
Formaldehyde
Benzene
Carbon Tetrachloride
Arsenic
1,3 -Butadiene
Naphthalene
1,2-Dichloroethane
Acetaldehyde
Nickel
6.91
4.58
4.19
2.92
2.55
2.41
1.70
1.63
1.31
to
oo
-------
Table 28-8. 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)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer Hazard
Approximation
(HQ)
Seattle, Washington (King County) - SEWA
Toluene
Ethylene glycol
Xylenes
Hexane
Methanol
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
Methyl isobutyl ketone
5,086.22
2,460.80
1,920.76
1,505.63
1,144.61
900.20
750.12
465.05
421.81
205.29
Acrolein
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Benzene
Naphthalene
Xylenes
Lead, PM
Nickel, PM
Manganese, PM
2,427,912.38
76,542.48
69,341.11
46,868.07
30,006.57
29,138.59
19,207.63
16,885.21
12,603.60
7,804.03
Manganese
Acetaldehyde
Formaldehyde
Arsenic
1,3 -Butadiene
Nickel
Naphthalene
Benzene
Carbon Tetrachloride
1 ,2-Dichloroethane
0.20
0.08
0.05
0.05
0.04
0.03
0.02
0.02
0.01
0.01
to
oo
-------
Observations from Table 28-7 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 toxi city-weighted emissions (of the pollutants with
cancer UREs) for King County are formaldehyde, benzene, and hexavalent
chromium.
• Eight of the highest emitted pollutants also have the highest toxi city-weighted
emissions for King County.
• Formaldehyde and benzene have the highest cancer risk approximations for SEWA.
These two pollutants top both emissions-based lists as well. Naphthalene,
1,3-butadiene, and acetaldehyde also appear on all three lists.
• Carbon tetrachloride, arsenic, and nickel, which rank third, fourth, and ninth,
respectively, for cancer risk approximations for SEWA, do not appear on either
emissions-based list.
• POM, Group 2b is the eighth highest emitted "pollutant" in King County and ranks
seventh for toxicity-weighted emissions. POM, Group 2b includes several PAHs
sampled for at SEWA including acenaphthene, fluorene, and perylene. POM, Group
2d ranks ninth for total emissions and its toxicity-weighted emissions. POM, Group
2d includes several PAHs sampled for at SEWA including anthracene, phenanthrene,
and pyrene. None of the PAHs included in POM, Groups 2b or 2d were identified as
pollutants of interest for SEWA, although fluorene and acenapthalene each failed a
few screens.
Observations from Table 28-8 for SEWA include the following:
• Toluene, ethylene glycol, and xylenes are the highest emitted pollutants with
noncancer RfCs in King County. The emissions of the pollutants with noncancer
RfCs are considerably higher than the emissions for the pollutants listed in
Table 28-7.
• 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.
• 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 28-8.
28-38
-------
• Manganese, which has the highest noncancer hazard approximation for SEW A, albeit
low, does not appear among the highest emitted pollutants but ranks 10th for its
toxicity-weighted emissions. Nickel, naphthalene, and 1,3-butadiene also appear
among those with the highest toxicity-weighted emissions but are not among the
highest emitted in King County (of those with a noncancer RfC).
28.6 Summary of the 2012 Monitoring Data for SEWA
Results from several of the data treatments described in this section include the
following:
*»* Fifteen pollutants failed at least one screen for SEWA.
»«» Acetaldehyde had the highest annual average concentration for SEWA, although all
of the pollutants of interest for SEWA had annual average concentrations less than
1 jug/m3.
»«» The annual average concentration of nickel for SEWA is the second highest among
NMP sites sampling metals. Conversely, the annual average concentration of
formaldehyde is the lowest among NMP sites sampling carbonyl compounds.
»«» Concentrations of carbon tetrachloride and manganese exhibited decreasing trends
over much of the sampling period, although these trends did not continue into the
later years of sampling.
28-39
-------
29.0 Sites in Wisconsin
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS and UATMP sites in Wisconsin, 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.
29.1 Site Characterization
This section characterizes the 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 HOWI site is located in Horicon, Wisconsin and is the relocated Mayville NATTS
site. The MIWI site is located in Milwaukee. Figure 29-1 is the composite satellite image
retrieved from ArcGIS Explorer showing the monitoring site and its immediate surroundings.
Figure 29-2 identifies nearby point source emissions locations by source category, as reported in
the 2011 NEI for point sources. Note that only sources within 10 miles of the site are included in
the facility counts provided in Figure 29-2. 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 radius are still visible on the map,
but have been grayed out in order to show emissions sources just outside the boundary.
Figures 29-3 and 29-4 are the composite satellite image and point emissions sources map for
MIWI. Table 29-1 provides supplemental geographical information such as land use, location
setting, and locational coordinates.
29-1
-------
Figure 29-1. Horicon, Wisconsin (HOWI) Monitoring Site
to
VO
-------
Figure 29-2. NEI Point Sources Located Within 10 Miles of HOWI
'45'0"W 88 40'0"W 88'J35'0"W 88"3010"W
Legend
88°40'0"W 88'35'0"W 88r30'
-------
Figure 29-3. Milwaukee, Wisconsin (MIWI) Monitoring Site
to
VO
-------
Figure 29-4. NEI Point Sources Located Within 10 Miles of MIWI
88 10'0"W 88°5'0"W
87"45'0"W 87L-40'0"W
830'0-W 87'55'0-W 87'50'0"W 87 45'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
MIWIUATMPsite
O 10 mile radius
County boundary
Source Category Group (No. of Facilities)
•f Airport/Airline/Airport Support Operations (8)
« Asphalt Production/Hot Mix Asphalt Plant (2)
B Automobile/Truck Manufacturing (2)
X Battery Manufacturing (1)
T Breweries/Distilleries/Wineries (1)
C Chemical Manufacturing (8)
® Dry Cleaning (3)
6 Electrical Equipment Manufacturing (2)
f Electricity Generation via Combustion (1)
E Electroplating, Plating, Polishing, Anodizing, and Coloring (5)
F Food Processing/Agriculture (3)
I Foundries, Iron and Steel (6)
A Foundries, Non-ferrous (1)
-#- Industrial Machinery or Equipment Plant (9)
O Institution (school, hospital, prison, etc.) (6)
9 Leather and Leather Products (3)
!j Metal Can, Box, and Other Metal Container Manufacturing (3)
<•} Metals Processing/Fabrication (11)
x Mine/Quarry/Mineral Processing (2)
? Miscellaneous Commercial/Industrial (7)
fj Paint and Coating Manufacturing (1)
R Plastic, Resin, or Rubber Products Plant (4)
P Printing/Publishing/Paper Product Manufacturing (11)
B Pulp and Paper Plant (1)
X Rail Yard/Rail Line Operations (3)
' Wastewater Treatment (1)
» Water Treatment (2)
W Woodwork, Furniture, Millwork & Wood Preserving (3)
29-5
-------
Table 29-1. Geographical Information for the Wisconsin Monitoring Sites
Site
Code
HOW
MIWI
AQS Code
55-027-0001
55-079-0026
Location
Horicon
Milwaukee
County
Dodge
Milwaukee
Micro- or
Metropolitan
Statistical Area
Beaver Dam, WI
MSA
Milwaukee -
Waukesha-West
Allis, WI MSA
Latitude
and
Longitude
43.466111,
-88.621111
43.061258,
-87.913520
Land Use
Agricultural
Commercial
Location
Setting
Rural
Urban/City
Center
Additional Ambient Monitoring Information1
SVOCs, PCBs, CO, SO2, NOy, NO, VOCs,
Carbonyl compounds, O3, Meteorological
parameters, PM10, PM10 Metals, PM Coarse, PM25,
PM2 5 Speciation, IMPROVE Speciation.
SNMOCs, SO2, NOy, NO, NO2, NOX, Carbonyl
compounds, O3, Meteorological parameters, PM10,
PM Coarse, PM25, PM25 Speciation, IMPROVE
Speciation.
BOLD ITALICS = EPA-designated NATTS Site
to
VO
-------
The HOWI monitoring site is located just north of the town of Horicon, in southeast
Wisconsin, within the boundaries of the Horicon Marsh Wildlife Area. HOWI is located roughly
in the center of a triangle formed by Milwaukee (32 miles to the southeast), Madison (37 miles to
the southwest), and Fond Du Lac (20 miles to the northeast). The surrounding area is rural and
agricultural in nature, although a residential subdivision is located just south of the site. The
HOWI monitoring site serves as a rural background site. However, the area is affected by nearby
urban areas, and thus, could show the effects on the wildlife sanctuary. State Highway 28, which
can be seen on the lower right-hand side of Figure 29-1, is the closest major roadway.
Figure 29-2 shows that a couple of point sources are located just south and west of HOWI, in the
town of Horicon. The closest point source to HOWI is an industrial machinery or equipment
plant. The source categories with the most emissions sources within 10 miles of HOWI are metal
processing/fabrication 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; and industrial machinery or equipment plants.
The city of Milwaukee is located in southeast Wisconsin on the western shores of Lake
Michigan. The MIWI monitoring site is located in the parking lot behind the Wisconsin
Department of Natural Resources headquarters building. The site is located in a commercial area
surrounded by residential areas, as shown in Figure 29-3. Interstate-43 runs north-south less than
one-half mile west of the site. The Milwaukee River runs roughly north-south about one-half of a
mile east of the site with the Milwaukee Bay and Lake Michigan approximately 2 miles farther
east. Figure 29-4 shows this proximity to Lake Michigan as well as the numerous point sources
within 10 miles of MIWI. A cluster of point sources is located to the east of the site as well as to
the south. The source categories with the most emissions sources within 10 miles of MIWI are
metals processing/fabrication; printing, publishing, and paper product manufacturing; industrial
machinery or equipment; chemical manufacturing; and airport and airport support operations.
Within 1.5 miles of MIWI are electroplating, plating, polishing, anodizing, and coloring facilities
to the south and a pulp and paper plant, a leather and leather products facility, and a chemical
manufacturing facility to the east.
29-7
-------
Table 29-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Wisconsin monitoring sites. Table 29-2 includes both county-level
population and vehicle registration information. Table 29-2 also contains traffic volume
information for HOWI and MIWI as well as the location for which each traffic volume was
obtained. Additionally, Table 29-2 presents the county-level daily VMT for Dodge County and
Milwaukee County.
Table 29-2. Population, Motor Vehicle, and Traffic Information for the Wisconsin
Monitoring Sites
Site
HOWI
MIWI
Estimated
County
Population1
88,415
955,205
County-level
Vehicle
Registration2
96,912
632,914
Annual
Average
Daily Traffic3
5,100
12,800
Intersection
Used for
Traffic Data
Route 28 (Clason St), north of
Route 33
N. Martin Luther King Jr. Drive,
north of W. North Ave.
County-
level Daily
VMT4
2,626,054
17,532,434
Bounty-level population estimate reflects 2012 data (Census Bureau, 2013c)
2County-level vehicle registration reflects 2012 data (WI DOT, 2012a)
3AADT reflects 2011 data for HOWI and 2013 data for MIWI (WI DOT, 2011 and 2013)
4County-level VMT reflects 2011 data (WI DOT, 2012b)
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 29-2 include the following:
• Dodge County's population is an order of magnitude less than the population for
Milwaukee County and in the bottom-third compared to other counties with NMP
sites. This is not unexpected given the rural nature of the area. Conversely,
Milwaukee County's population is in the top third compared to other counties with
NMP sites.
• The county-level vehicle registration for HOWI is considerably less than the vehicle
registration for MIWI, ranking similarly to the ranking for population among other
counties with NMP sites. The county-level vehicle registration for MIWI is not as
high as its ranking for population compared to other NMP sites, putting it in the
middle third of the range.
• The traffic volume near MIWI is more than twice the traffic volume near HOWI. The
traffic volume near HOWI is also on the low end compared to other NMP sites while
the traffic near MIWI falls in the middle of the range. The traffic estimate provided
for HOWI is for State Road 28 near State Road 33 on the east side of Horicon. The
traffic estimate for MIWI is for N. Martin Luther King Jr. Drive, north of W. North
Avenue.
• The daily VMT for Milwaukee County is considerably higher than the VMT for
Dodge County. VMTs for these sites rank 18th and 33rd, respectively, compared to
VMTs for other counties with NMP sites (and where VMT data were available).
29-8
-------
29.2 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
sites in Wisconsin on sample days, as well as over the course of the year.
29.2.1 Climate Summary
HOWI and MIWI are both located in southeast Wisconsin. The city of Milwaukee is
located along the western shores of Lake Michigan, while the town of Horicon is located about
40 miles west of Lake Michigan, between the towns of West Bend and Beaver Dam. The climate
in this part of the state is continental in nature, with an active weather pattern, as storm systems
frequently move eastward across the region. Lake Michigan has a significant influence on the
area, although the town of Horicon is far enough inland to limit some of the moderating
influences of the lake. Precipitation falls predominantly in the spring and summer months, with
thunderstorms most common in the summer. Summers tend to be mild, although southerly winds
out of the Gulf of Mexico can occasionally advect warm, humid air into the area while easterly
winds off Lake Michigan have a cooling effect on the Milwaukee area. Winters are cold and
snowfall is common, with an annual average snowfall around 50 inches near Milwaukee. Lake
Michigan can moderate cold air masses moving in from the north and may induce lake-effect
snow events. Lake effect snows can occur with winds with a northeasterly and easterly
component, although lake effect snows are often reduced farther inland. The number of days per
season with at least 1 inch snow cover on the ground can range from less than 20 days to greater
than 100 days (Wood, 2004; WI SCO, 2013a and 2013b).
29.2.2 Meteorological Summary
Hourly meteorological data for 2012 were retrieved from NCDC for the weather stations
closest to the Wisconsin monitoring sites (NCDC, 2012), as described in Section 3.5.2. The
closest weather stations are located at Dodge County Airport near HOWI and Lawrence J.
Timmerman Airport near MIWI (WBANs 04898 and 94869, respectively). Additional
information about these weather stations, such as the distance between each site and the weather
station, is provided in Table 29-3. These data were used to determine how meteorological
conditions on sample days vary from conditions experienced throughout the year.
29-9
-------
Table 29-3. Average Meteorological Conditions near the Wisconsin Monitoring Sites
Closest
Weather
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(°F)
Average
Temperature
(°F)
Average
Dew Point
Temperature
(°F)
Average
Wet Bulb
Temperature
(°F)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Horicon, Wisconsin - HOWI
Dodge County
Airport
04898
(43.43, -88.70)
4 7
230°
(C\\f\
(a\\ )
Sample
Days
(63)
2012
60.3
±5.2
58.5
+ 2.1
50.7
±4.6
49.6
+ 1.9
36.5
±3.8
34.4
+ 1.6
43.9
±3.8
43.3
+ 1.6
63.9
±4.1
65.6
+ 1.6
NA
NA
7.4
±0.9
6.9
+ 0.3
Milwaukee, Wisconsin - MIWI
Lawrence J
Timmerman
Airport
94869
(43.11, -88.03)
6 5
miles
295°
(WNW)
Sample
Days
(52)
2012
63.6
±5.2
59.2
+ 2.1
54.6
±4.7
50.5
+ 1.9
40.1
±4.2
37.5
+ 1.6
47.3
±4.0
44.2
+ 1.6
62.7
±3.8
65.6
+ 1.5
NA
NA
6.9
±1.1
6.8
+ 0.3
to
VO
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
NA = Sea level pressure was not recorded at either airport.
-------
Table 29-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for all of 2012. Also included in Table 29-3 is the 95 percent
confidence interval for each parameter. As shown in Table 29-3, average meteorological
conditions on sample days near HOWI appear slightly warmer than average weather conditions
experienced throughout 2012, although the differences are not statistically significant. The
differences are a little larger for MIWI. However, sampling at MIWI did not begin until the end
of February, thereby missing some of the coldest days of the year.
29.2.3 Back Trajectory Analysis
Figure 29-5 is the composite back trajectory map for days on which samples were
collected at the HOWI monitoring site. Included in Figure 29-5 are four back trajectories per
sample day. Figure 29-6 is the corresponding cluster analysis. Similarly, Figures 29-7 and 29-8
are the composite back trajectory map and corresponding cluster analysis for days on which
samples were collected at MIWI. An in-depth description of these maps and how they were
generated is presented in Section 3.5.2.1. For the composite maps, each line represents the
24-hour trajectory along which a parcel of air traveled toward the monitoring site on a given
sample day and time, based on an initial height of 50 meters AGL. For the cluster analyses, each
line corresponds to a trajectory representative of a given cluster of back trajectories. Each
concentric circle around the sites in Figures 29-5 through 29-8 represents 100 miles.
29-11
-------
Figure 29-5. Composite Back Trajectory Map for HOWI
Figure 29-6. Back Trajectory Cluster Map for HOWI
29-12
-------
Figure 29-7. Composite Back Trajectory Map for MIWI
Figure 29-8. Back Trajectory Cluster Map for MIWI
29-13
-------
Observations from Figures 29-5 through 29-8 for HOWI and MIWI include the
following:
• The composite back trajectory maps for HOWI and MIWI resemble each other,
although there are fewer individual back trajectories on the composite map for MIWI.
Back trajectories originated from a variety of directions at these sites, although fewer
back trajectories originate from a direction with an easterly component. Back
trajectories from the northwest quadrant tended to be the longest while those with an
easterly component tended to be the shortest.
• The 24-hour air shed domains for the Wisconsin sites are among the largest in size
compared to other NMP sites. Both of the sites have back trajectories greater than 700
miles in length. MIWI has the highest average back trajectory length among all NMP
sites (290 miles) while HOWFs average back trajectory length ranked third (286
miles). These two sites have the second and third highest number of back trajectories
greater than 500 miles in length.
• The cluster analyses for these two sites have many similarities. The primary
difference is how the model groups the shorter back trajectories with a northerly
component. For HOWI, the shorter back trajectories originating to the north and
northeast are represented by the short cluster trajectory originating over Green Bay;
the shorter back trajectories originating over northwest Wisconsin and Minnesota are
grouped with those shorter back trajectories originating over Iowa and are represented
by the westward originating cluster trajectory. For MIWI, the shorter back trajectories
originating from the northwest, north, and north-northeast are grouped together and
are represented by the cluster trajectory originating in the center of Wisconsin. The
shorter back trajectories originating over Michigan are included with those
originating from the east and southeast.
• Both cluster analyses show that the longest back trajectories originated to the
northwest over the Northern Plains, Minnesota, and Manitoba and Ontario, Canada.
Back trajectories of varying lengths also originated to the south of the sites, over
Illinois and Missouri. Shorter, westward-originating back trajectories were also
common. Back trajectories originating from a direction with an easterly component
account for less than 25 percent of back trajectories.
29.2.4 Wind Rose Comparison
Hourly surface wind data from the weather stations at Dodge County Airport near HOWI
and Lawrence J. Timmerman Airport near MIWI were uploaded into a wind rose software
program to produce customized wind roses, as described in Section 3.5.2.2. 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.
29-14
-------
Figure 29-9 presents a map showing the distance between the weather station and HOWI,
which may be useful for identifying topographical influences that can affect the meteorological
patterns experienced at this location. Figure 29-9 also presents three different wind roses for the
HOWI monitoring site. First, a historical wind rose representing 2003 to 2011 wind data is
presented, which shows the predominant surface wind speed and direction over an extended
period of time. Second, a wind rose representing wind observations for all of 2012 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2012 is
presented. These can be used to identify the predominant wind speed and direction for 2012 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically. Figure 29-10 presents the distance map and three wind roses
for MIWI.
Observations from Figure 29-9 for HOWI include the following:
• The Dodge County Airport weather station is located less than 5 miles southwest of
HOWI.
• The historical wind rose shows that winds from a variety of directions were observed
near HOWI. Winds from the south, southwest quadrant, and west account for
one-third of wind observations. The strongest wind speeds were associated with
southerly to west-southwesterly winds. Calm winds (<2 knots) were observed for
approximately 14 percent of the hourly measurements.
• The wind patterns shown on the 2012 wind rose resemble the historical wind patterns,
although winds from the south and south-southwest were observed more frequently as
were winds from the southeast and south-southeast. The percentage of calm winds
was less than 12 percent in 2012.
• The sample day wind rose shows that winds from the southeast and southwest
quadrants were observed even more frequently on sample days and that a higher
percentage of strong (> 22 knots) winds were observed with winds from the south-
southeast to south-southwest. Calm winds accounted for even fewer observations on
sample days (roughly 10 percent).
29-15
-------
Figure 29-9. Wind Roses for the Dodge County Airport Weather Station near HOWI
Location of HOWI and Weather Station
2003-2011 Historical Wind Rose
N
+
Calms: 1435%
2012 Wind Rose
Sample Day Wind Rose
Calms: 11.59%
29-16
-------
Figure 29-10. Wind Roses for the Lawrence J. Timmerman Airport Weather Station near
MIWI
Location of MIWI and Weather Station
2006-2011 Historical Wind Rose
\
vj » X^Mj-r-
s i "'"":.—
\ ""- : • —-
Calms: 16.77%
2012 Wind Rose
Sample Day Wind Rose
Calms: 1683%
29-17
-------
Observations from Figure 29-10 for MIWI include the following:
• The Timmerman Airport weather station is located 6.5 miles west-northwest of
MIWI. Note that the airport location is considerably farther inland than the
monitoring site location.
• The historical wind rose shows that winds from a variety of directions were observed
near MIWI, although westerly winds account for the greatest number of observations
greater than 2 knots. Winds with a westerly component were observed more
frequently than winds with an easterly component. Calm winds (<2 knots) were
observed for approximately 17 percent of the hourly measurements.
• The wind patterns shown on the 2012 wind rose resemble the historical wind patterns,
although winds from the south and south-southwest were observed more frequently.
• The sample day wind rose does not show the prominence of the westerly wind.
Instead, southwesterly winds account for the highest percentage of wind observations.
However, winds from the south-southwest, west, and west-northwest each account for
approximately 7 percent of wind observations on sample days. Calm winds accounted
for a greater percentage of wind observations on sample days (roughly 19 percent).
Some of the differences between the full-year and sample day wind rose may be
attributable to the shortened sampling duration (sampling at MIWI did not begin until
the end of February).
29.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each
Wisconsin 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 29-4. 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 29-4. It is important to note which pollutants were sampled for at each site when reviewing
the results of this analysis. Only hexavalent chromium was sampled for at these two sites.
29-18
-------
Table 29-4. Risk-Based Screening Results for the Wisconsin Monitoring Sites
Pollutant
Screening
Value
(ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Horicon, Wisconsin - HOWI
Hexavalent Chromium
0.000083
Total
0
0
35
35
0.00
0.00
0.00
0.00
Milwaukee, Wisconsin - MIWI
Hexavalent Chromium
0.000083
Total
12
12
41
41
29.27
29.27
100.00
100.00
Observations from Table 29-4 include the following:
• Hexavalent chromium was detected in 35 of the 61 valid samples collected at HOWI.
• Hexavalent chromium did not fail any screens during the 2012 monitoring effort at
HOWI. This was also true for 2011.
• Because HOWI does not have any pollutants of interest, this site is excluded from the
sections that follow, with the exception of the emissions section (Section 29.5.3).
• Hexavalent chromium was detected in 41 of the 52 valid samples collected at MIWI.
• Hexavalent chromium failed 12 screens for MIWI (or nearly 30 percent); thus,
hexavalent chromium is a pollutant of interest for MIWI.
29.4 Concentrations
This section presents various concentration averages used to characterize pollution levels
at the MIWI 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
MIWI.
• Annual concentration averages are presented graphically for the 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. Additional site-specific statistical summaries for MIWI are
provided in Appendix O. A site-specific statistical summary for HOWI is also provided in
Appendix O.
29-19
-------
29.4.1 2012 Concentration Averages
Quarterly and annual concentration averages were calculated for the pollutants of interest
for MIWI, as described in Section 3.1. The quarterly average 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 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 MIWI are presented in Table 29-5,
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 29-5. Quarterly and Annual Average Concentrations of the Pollutants of
Interest for the Wisconsin Monitoring Sites
Pollutant
#of
Measured
Detections
vs. # 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)
Milwaukee, Wisconsin - MIWI
Hexavalent Chromium
41/52
NA
0.115
±0.128
0.273
±0.308
0.062
± 0.064
0.166
±0.111
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
Note: There are no pollutants of interest for HOWL
Observations for MIWI from Table 29-5 include the following:
• Concentrations of hexavalent chromium span three orders of magnitude, ranging from
0.0045 ng/m3 to 2.30 ng/m3 (as well as 11 non-detects). Three of the six hexavalent
chromium concentrations greater than 1 ng/m3 measured across the program were
measured at MIWI (while the other three were measured at STMN).
• The maximum hexavalent chromium concentration for MIWI was measured on
August 13, 2012. Although this is the third highest hexavalent chromium
concentration measured across the program in 2012, it is also one of the highest
concentrations of this pollutant measured across all years of hexavalent chromium
sampling.
29-20
-------
• A first quarter average concentration could not be calculated because sampling at
MIWI did not begin until February 27, 2012.
• For each available quarterly average, the confidence interval is larger than the
average itself, indicating that the measurements factoring into each average are highly
variable.
• Of the 11 non-detects, seven were measured during the fourth quarter, with the other
four measured during the second quarter.
• MIWI has the second highest annual average concentration of hexavalent chromium
among all NMP sites sampling this pollutant. MIWI (and STMN) are the only two
NMP sites with annual averages greater than 0.1 ng/m3.
29.4.2 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, a box plot was created for hexavalent chromium
for MIWI. Figure 29-11 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 3.5.3.1.
Figure 29-11. Program vs. Site-Specific Average Hexavalent Chromium Concentration
MIWI
i
Program Max Concentration = 8.51 ng/m3
0.2 0.3
Concentration (ng/m3)
Program:
Site:
1st Quartile
Site Average
o
2nd Quartile 3rd Quartile
Site Concentration Range
4th Quartile Average
Observations from Figure 29-11 include the following:
• The program-level maximum concentration (8.51 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 to 0.5 ng/m3. In addition, the program-level first quartile is zero
and therefore not visible on the box plot. Figure 29-11 shows that the annual
average hexavalent chromium concentration for MIWI is more than four times
greater than the program-level average concentration. The maximum hexavalent
29-21
-------
chromium concentration measured at MIWI is greater than the scale in
Figure 29-11 but is not the maximum concentration measured at the program-
level, as previously discussed.
29.4.3 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 3.5.3.2.
MIWI is a new site under the NMP for 2012 and therefore does not meet the criteria specified
above. As a result, a trends analysis was not conducted.
29.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk at
MIWI. Refer to Sections 3.3 and 3.5.3 for definitions and explanations regarding the various
toxicity factors, time frames, and calculations associated with these risk-based screenings.
29.5.1 Risk-Based Screening Assessment Using MRLs
A risk-based screening was conducted by comparing the concentration data from the
Wisconsin monitoring sites to the ATSDR MRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 day to 14 days); intermediate (exposures of 15 days to 364 days); and chronic
(exposures of 1 year or greater). The preprocessed daily measurements of the pollutants of
interest were compared to the acute MRLs; the quarterly averages were compared to the
intermediate MRLs; and the annual averages were compared to the chronic MRLs.
As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2012.
29.5.2 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for MIWI 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 pollutant of interest. Although the use of these approximations is
limited, they may help identify where policy-makers want to shift their air-monitoring priorities.
29-22
-------
Refer to Section 3.5.3.4 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 29-6, where applicable. Cancer risk approximations are presented as
probabilities while the noncancer hazard approximations are ratios and thus, unitless values.
Table 29-6. Risk Approximations for the Wisconsin Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Risk
Approximation
(HQ)
Milwaukee, Wisconsin - MIWI
Hexavalent Chromium
0.012
0.0001
41/52
0.166
±0.111
1.99
<0.01
Observations for MIWI from Table 29-6 include the following:
• The cancer risk approximation for hexavalent chromium is 1.99 in-a-million, one of
only two cancer risk approximations for this pollutant greater than 1 in-a-million
program-wide.
• The noncancer hazard approximation for hexavalent chromium is less than 0.01,
indicating that no adverse health effects are expected from this pollutant.
29.5.3 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 29-7 presents the 10 pollutants with the highest emissions from the 2011 NEI that have
cancer toxicity factors. Table 29-7 also presents the 10 pollutants with the highest toxicity -
weighted emissions, based on the weighting schema described in Section 3.5.3.5. Lastly,
Table 29-7 provides the 10 pollutants with the highest cancer risk approximations (in-a-million)
for each site, as presented in Table 29-6. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 29-7. Table 29-8 presents similar
information, but is limited to those pollutants with noncancer toxicity factors.
29-23
-------
Table 29-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Wisconsin Monitoring Sites
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
Horicon, Wisconsin (Dodge County) - HOWI
Benzene
Formaldehyde
Acetaldehyde
Ethylbenzene
1,3 -Butadiene
Naphthalene
POM, Group 2b
Trichloroethylene
POM, Group 2d
POM, Group 6
80.96
48.25
33.00
22.67
8.03
5.84
1.27
0.85
0.81
0.17
Benzene
Formaldehyde
POM, Group 3
1,3 -Butadiene
Naphthalene
POM, Group 2b
POM, Group 5a
Acetaldehyde
POM, Group 2d
Hexavalent Chromium, PM
6.31E-04
6.27E-04
2.56E-04
2.41E-04
1.99E-04
1.12E-04
1.09E-04
7.26E-05
7.15E-05
6.00E-05
Milwaukee, Wisconsin (Milwaukee County) - MIWI
Benzene
Formaldehyde
POM, Group la
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
Dichloromethane
POM, Group 2b
POM, Group 2d
229.78
183.31
179.42
145.70
111.59
37.48
20.83
14.76
3.92
2.97
POM, Group la
Hexavalent Chromium, PM
Formaldehyde
Benzene
1,3 -Butadiene
Nickel, PM
Naphthalene
Arsenic, PM
POM, Group 3
Ethylbenzene
1.58E-02
3.19E-03
2.38E-03
1.79E-03
1.12E-03
1.08E-03
7.08E-04
4.78E-04
4.27E-04
3.64E-04
Hexavalent Chromium 1.99
to
VO
to
-------
Table 29-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Wisconsin Monitoring Sites
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations Based
on Annual Average Concentrations
(Site-Specific)
Noncancer Hazard
Approximation
Pollutant (HQ)
Horicon, Wisconsin (Dodge County) - HOWI
Toluene
Ethylene glycol
Xylenes
Benzene
Hexane
Methanol
Formaldehyde
Acetaldehyde
Ethylbenzene
Hydrochloric acid
423.43
113.15
109.42
80.96
72.68
50.62
48.25
33.00
22.67
18.25
Acrolein
Manganese, PM
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Benzene
Cyanide Compounds, gas
Naphthalene
Lead, PM
Xylenes
131,802.77
5,388.88
4,923.22
4,012.73
3,667.03
2,698.61
2,508.26
1,947.34
1,151.06
1,094.24
Milwaukee, Wisconsin (Milwaukee County) - MIWI
Toluene
Ethylene glycol
Methanol
Xylenes
Hexane
Hydrochloric acid
Benzene
Formaldehyde
Ethylbenzene
Hydrofluoric acid
2,087.97
1,316.47
644.01
581.41
578.74
451.78
229.78
183.31
145.70
127.65
Acrolein
Manganese, PM
Nickel, PM
Hydrochloric acid
1,3 -Butadiene
Formaldehyde
Acetaldehyde
Hydrofluoric acid
Benzene
Arsenic, PM
563,909.98
63,774.13
24,907.34
22,589.06
18,738.94
18,705.01
12,398.97
9,117.92
7,659.40
7,409.16
Hexavalent Chromium O.01
to
VO
to
-------
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 3.5.3.5. Similar to the cancer risk and
noncancer hazard approximations provided in Section 29.5.2, this analysis may help policy-
makers prioritize their air monitoring activities.
Observations from Table 29-7 include the following:
• Because Table 29-7 includes emissions data from the NEI, which is independent of
the sampling results at a specific site, data for Dodge County, where HOWI is
located, is included.
• Benzene and formaldehyde are the highest emitted pollutants with cancer UREs in
both Dodge and Milwaukee County, although the emissions are higher in Milwaukee
County.
• Benzene is the pollutant with the highest toxicity-weighted emissions (of the
pollutants with cancer UREs), followed by formaldehyde and POM, Group 3 for
Dodge County. POM, Group la, hexavalent chromium, and formaldehyde are the
pollutants with the highest toxicity-weighted emissions for Milwaukee County.
• Seven of the highest emitted pollutants in Dodge County also have the highest
toxicity-weighted emissions. Six of the highest emitted pollutants in Milwaukee
County also have the highest toxicity-weighted emissions.
• Hexavalent chromium, which is the only pollutant sampled for at MIWI, has the
second highest toxicity-weighted emissions for Milwaukee County, but is not among
the highest emitted. Hexavalent chromium emissions in Milwaukee County rank 17th.
• Several POM Groups rank among Milwaukee County's highest total emissions and
toxicity-weighted emissions. PAHs were not sampled at this site.
Observations from Table 29-8 include the following:
• Toluene and ethylene glycol are the highest emitted pollutants with noncancer RfCs
in both counties, although the emissions are considerably higher for Milwaukee
County.
• The pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) for both counties are acrolein and manganese.
29-26
-------
• Four of the highest emitted pollutants in Dodge County also have the highest toxicity-
weighted emissions. The same is true for Milwaukee County, although the actual
pollutants differ.
• Hexavalent chromium does not appear among the pollutants with the highest
emissions or toxicity-weighted emissions for either county (among pollutants with
noncancer RfCs).
29.6 Summary of the 2012 Monitoring Data for HOWI and MIWI
Results from several of the data treatments described in this section include the
following:
»«» Hexavalent chromium was the only pollutant sampled for at HO WI and did not fail
any screens.
»«» Hexavalent chromium was also the only pollutant sampled for at MIWI; hexavalent
chromium failed 30 percent of screens for this site.
»«» Concentrations of hexavalent chromium measured at MIWI ranged from
0.0045 ng/m to 2.30 ng/m , including several of the highest concentrations program-
wide. The annual average concentration of hexavalent chromium for MIWI ranked
second highest compared to other NMP sites sampling this pollutant.
29-27
-------
30.0 Data Quality
This section discusses the data quality of the ambient air measurements that constitute the
2012 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 Measurement Quality Objectives (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, 2012), the following MQOs were assessed: completeness,
precision, and accuracy (also called bias).
The quality assessments presented in this section show that the 2012 monitoring data are
of a known and high quality, consistent with the intended data use. The overall method-specific
completeness was greater than 90 percent for each method. The method precision for collocated
and duplicate analyses met the precision MQO of 15 percent CV for all methods. The analytical
precision for replicate analyses also met the precision MQO of 15 percent CV. 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.
30.1 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, 2012). The MQO of 85 percent completeness was met by all but five
of 144 site-method combinations. Completeness statistics are presented and discussed more
thoroughly in Section 2.4.
30-1
-------
30.2 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 using the same sampling system (i.e., two
separate samples through the same sampling system at the same time). 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 using two
independent collection systems at the same location at the same time.
Both approaches provide valuable, but different, assessments of method precision:
• Analysis of duplicate samples provides information on the potential for variability (or
precision) expected from a single collection system (intra-system assessment).
• Analysis of collocated samples provides information on the potential for variability
(or precision) expected between different collection systems (inter-system
assessment).
During the 2012 sampling year, duplicate and 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. These duplicate or collocated
samples were then analyzed in replicate. Replicate measurements are repeated analyses
performed on a duplicate or collocated pair of samples and are discussed in greater detail in
Section 30.3. In the event 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.
30-2
-------
Method precision was 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 vs. 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-r) l2
2n
i
Where:
p = the primary result from a duplicate or collocated pair;
r = the secondary result from a duplicate or collocated pair;
n = the number of valid data pairs (the 2 adjusts for the fact that there are two
values with error).
Coefficients of variation were based on every pair of duplicate or collocated samples
collected during the program year. However, only results 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. This is a change in procedure compared to NMP reports prior to
2010, where comparison to the MDL was not considered and 1/2 MDL was substituted for non-
detects. 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.
30-3
-------
Table 30-1 presents the 2012 NMP method precision for VOCs, SNMOCs, carbonyl
compounds, PAHs, metals, and hexavalent chromium, presented as the average CV (expressed as
a percentage). Each analytical method met the program MQO of 15 percent CV for precision.
This table also includes the number of pairs that were included in the calculation of the method
precision. The total number of pairs for each method is also included in Table 30-1 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 30-1. Method Precision by Analytical Method
Method/Pollutant
Group
voc
(TO- 15)
SNMOC
Carbonyl Compounds
(TO- 11 A)
PAH
(TO-13)
Metals Analysis
(Method IO-3.5/FEM)
Hexavalent Chromium
(ASTMD7614)
MQO
Average
Coefficient of
Variation
(%)
9.28
8.84
9.14
14.72
9.12
13.77
Number of
Pairs Included
in the
Calculation
3,147
857
1,740
351
1,733
102
Total Number
of Pairs Without
the > MDL
exclusion
4,046
1,127
1,761
494
2,157
109
15.00 percent CV
Tables 30-2 through 30-7 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 C V calculations. C Vs 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.
30.2.1 VOC Method Precision
Table 30-2 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 all VOCs listed. The individual method precision results exhibit low-
to high-level variability, where the CV ranged from 0 percent for a few pollutants for several
30-4
-------
sites to 56.50 percent (methyl isobutyl ketone for ROIL, although a similar CV was also
calculated for GLKY and this pollutant). The pollutant-specific average CV ranged from 2.43
percent (£ram'-l,2-dichloroethylene) to 27.46 percent (methyl isobutyl ketone). The site-specific
average CV ranged from 5.08 percent (GPCO) to 14.40 percent (PROK). None of the sites had a
site-specific average CV greater than 15 percent. The overall average method precision for
VOCs was 9.28 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.
Table 30-2. VOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant
Pollutant
Acetylene
fer/-Amyl Methyl Ether
Benzene
Bromochloromethane
Bromodichloromethane
Bromoform
Bromomethane
1,3 -Butadiene
Carbon Tetrachloride
Chlorobenzene
Chloroethane
Chloroform
Chloromethane
Chloroprene
Dibromochloromethane
1 ,2-Dibromoethane
/w-Dichlorobenzene
o-Dichlorobenzene
£>-Dichlorobenzene
Dichlorodifluoromethane
1 , 1 -Dichloroethane
1 ,2-Dichloroethane
1 , 1 -Dichloroethene
cis- 1 ,2-Dichloroethylene
trans- 1 ,2-Dichloroethylene
Dichloromethane
ADOK
9.81
NA
16.40
NA
NA
NA
5.66
8.87
8.88
NA
NA
13.29
9.96
NA
NA
NA
NA
NA
NA
7.36
NA
6.46
NA
NA
NA
17.81
BTUT
2.40
NA
9.02
NA
NA
NA
NA
7.75
10.87
NA
NA
9.33
4.18
NA
NA
NA
NA
NA
NA
4.04
NA
8.01
NA
NA
NA
27.68
BURVT
5.40
NA
8.96
NA
NA
NA
4.45
6.29
12.67
NA
14.24
9.75
4.80
NA
NA
NA
NA
NA
6.73
4.10
NA
10.91
NA
NA
0.00
14.94
CHNJ
14.91
NA
4.78
NA
NA
NA
26.67
12.29
7.01
NA
NA
6.41
4.48
NA
NA
NA
NA
NA
NA
4.54
NA
3.43
NA
NA
NA
17.09
DEMI
4.24
NA
7.76
NA
NA
NA
5.66
5.62
6.96
5.05
NA
28.96
4.31
NA
NA
NA
NA
NA
0.00
3.83
NA
6.40
NA
NA
1.99
6.32
ELNJ
8.66
NA
7.74
NA
NA
NA
6.93
11.09
5.41
NA
NA
8.02
7.35
NA
NA
NA
NA
NA
7.41
6.44
NA
6.55
NA
NA
NA
19.58
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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.
30-5
-------
Table 30-2. VOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
1 ,2-Dichloropropane
cis- 1 ,3 -Dichloropropene
tr ans- 1 , 3 -D ichloropropene
Dichlorotetrafluoroethane
Ethyl Aery late
Ethyl fert-Butyl Ether
Ethylbenzene
Hexachloro- 1 ,3 -butadiene
Methyl Isobutyl Ketone
Methyl Methacrylate
Methyl ferMSutyl Ether
w-Octane
Propylene
Styrene
1 , 1 ,2,2-Tetrachloroethane
Tetrachloroethylene
Toluene
1 ,2,4-Trichlorobenzene
1,1,1 -Trichloroethane
1 , 1 ,2-Trichloroethane
Trichloroethylene
Trichlorofluoromethane
Trichlorotrifluoroethane
1 ,2,4-Trimethylbenzene
1 ,3 ,5-Trimethylbenzene
Vinyl chloride
7w,£>-Xylene
o-Xylene
Average by Site
ADOK
NA
NA
NA
8.03
NA
NA
14.60
NA
29.00
NA
NA
9.05
12.06
14.36
NA
NA
21.95
NA
NA
NA
NA
7.41
10.29
4.35
2.22
NA
15.14
15.93
11.69
BTUT
NA
NA
NA
10.27
NA
6.93
6.73
NA
14.18
NA
12.76
8.43
5.45
0.00
NA
17.80
5.18
NA
NA
NA
NA
3.94
3.12
5.43
6.21
NA
6.63
5.36
8.07
BURVT
NA
NA
NA
5.57
NA
NA
11.73
NA
25.03
NA
NA
12.42
12.15
14.18
NA
6.61
7.82
NA
NA
NA
NA
4.24
16.32
12.83
12.18
NA
12.00
12.76
9.97
CHNJ
NA
NA
NA
8.92
NA
21.99
13.42
NA
37.35
NA
4.48
11.22
21.94
7.39
NA
5.72
8.16
NA
NA
NA
NA
5.42
4.26
15.96
0.00
NA
23.57
22.69
11.64
DEMI
NA
NA
NA
2.99
NA
NA
10.35
NA
12.81
NA
NA
6.92
5.42
17.06
NA
4.70
15.27
NA
NA
NA
NA
3.48
4.47
8.74
8.60
NA
11.18
15.30
7.94
ELNJ
NA
NA
NA
9.85
NA
5.90
5.71
NA
24.44
7.55
10.04
10.90
11.74
8.51
NA
2.00
6.35
NA
NA
NA
0.00
7.87
5.77
3.68
4.01
NA
6.15
4.36
7.93
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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.
30-6
-------
Table 30-2. VOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
Acetylene
fer/-Amyl Methyl Ether
Benzene
Bromochloromethane
Bromodichloromethane
Bromoform
Bromomethane
1,3 -Butadiene
Carbon Tetrachloride
Chlorobenzene
Chloroethane
Chloroform
Chloromethane
Chloroprene
Dibromochloromethane
1 ,2-Dibromoethane
/w-Dichlorobenzene
o-Dichlorobenzene
£>-Dichlorobenzene
Dichlorodifluoromethane
1 , 1 -Dichloroethane
1 ,2-Dichloroethane
1 , 1 -Dichloroethene
cis- 1 ,2-Dichloroethylene
trans- 1 ,2-Dichloroethylene
Dichloromethane
1 ,2-Dichloropropane
cis- 1 ,3 -Dichloropropene
tr ans- 1 , 3 -D ichloropropene
Dichlorotetrafluoroethane
Ethyl Aery late
Ethyl fert-Butyl Ether
Ethylbenzene
Hexachloro- 1 ,3 -butadiene
Methyl Isobutyl Ketone
Methyl Methacrylate
Methyl ter/-Butyl Ether
w-Octane
Propylene
Styrene
GLKY
18.52
NA
10.55
NA
NA
NA
0.00
8.21
7.18
NA
NA
21.05
6.10
NA
NA
NA
NA
NA
NA
4.23
NA
7.34
NA
NA
NA
40.04
NA
NA
NA
11.76
NA
NA
4.81
NA
55.94
NA
NA
6.75
3.56
13.85
GPCO
2.50
NA
6.01
NA
NA
NA
3.52
4.22
15.48
NA
NA
7.44
1.97
NA
NA
NA
NA
NA
1.94
2.69
NA
7.56
NA
NA
NA
8.69
NA
NA
NA
6.04
NA
NA
3.70
NA
12.80
2.24
NA
4.24
2.68
3.24
NBIL
7.11
NA
10.94
NA
26.88
5.05
15.71
8.97
9.28
NA
5.81
20.77
5.87
NA
18.94
NA
NA
NA
6.12
6.20
NA
7.66
NA
NA
NA
14.18
NA
NA
NA
5.37
NA
4.65
5.57
NA
15.46
4.04
3.29
7.96
11.96
1.59
NBNJ
4.87
NA
9.03
NA
NA
NA
7.52
7.16
12.25
NA
4.08
3.87
7.06
NA
NA
NA
NA
NA
2.48
3.87
NA
4.52
NA
NA
NA
16.94
NA
NA
NA
3.68
NA
5.24
7.56
NA
17.40
NA
26.73
11.32
6.42
8.41
OCOK
9.56
NA
17.88
NA
NA
NA
NA
11.91
9.48
NA
NA
12.86
9.44
NA
NA
NA
NA
NA
NA
9.25
NA
11.48
NA
NA
NA
7.90
NA
NA
NA
11.44
NA
NA
6.24
NA
28.66
NA
NA
15.29
10.51
2.22
PROK
44.43
NA
27.25
NA
2.60
NA
NA
1.06
7.04
NA
NA
17.65
4.81
NA
NA
NA
NA
NA
NA
2.85
NA
11.13
NA
NA
NA
32.56
NA
NA
NA
3.34
NA
NA
14.29
NA
19.23
NA
NA
6.91
11.17
10.99
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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.
30-7
-------
Table 30-2. VOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
1 , 1 ,2,2-Tetrachloroethane
Tetrachloroethylene
Toluene
1 ,2,4-Trichlorobenzene
1,1,1 -Trichloroethane
1 , 1 ,2-Trichloroethane
Trichloroethylene
Trichlorofluoromethane
Trichlorotrifluoroethane
1 ,2,4-Trimethylbenzene
1 ,3 ,5-Trimethylbenzene
Vinyl chloride
7w,/>-Xylene
o-Xylene
Average by Site
GLKY
NA
NA
11.68
NA
NA
NA
NA
5.01
8.04
4.12
4.88
NA
22.10
4.70
12.19
GPCO
NA
5.90
5.01
NA
NA
NA
NA
3.02
5.50
2.49
4.33
NA
3.87
4.86
5.08
NBIL
NA
6.38
9.70
NA
NA
NA
10.48
5.37
5.64
5.35
4.02
NA
5.85
5.73
8.72
NBNJ
NA
7.86
7.28
NA
NA
NA
NA
4.28
3.51
15.51
9.56
NA
8.11
10.31
8.46
OCOK
NA
7.86
5.50
NA
NA
NA
7.97
8.22
9.16
7.08
5.63
NA
5.69
5.18
9.85
PROK
NA
NA
30.20
NA
NA
NA
NA
2.35
5.67
20.14
5.89
NA
33.15
16.55
14.40
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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.
30-8
-------
Table 30-2. VOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
Acetylene
fer/-Amyl Methyl Ether
Benzene
Bromochloromethane
Bromodichloromethane
Bromoform
Bromomethane
1,3 -Butadiene
Carbon Tetrachloride
Chlorobenzene
Chloroethane
Chloroform
Chloromethane
Chloroprene
Dibromochloromethane
1 ,2-Dibromoethane
/w-Dichlorobenzene
o-Dichlorobenzene
£>-Dichlorobenzene
Dichlorodifluoromethane
1 , 1 -Dichloroethane
1 ,2-Dichloroethane
1 , 1 -Dichloroethene
cis- 1 ,2-Dichloroethylene
/ra«s-l,2-Dichloroethylene
Dichloromethane
1 ,2-Dichloropropane
cis- 1 ,3 -Dichloropropene
tr cms- 1 , 3 -D ichloropropene
Dichlorotetrafluoroethane
Ethyl Aery late
Ethyl fert-Butyl Ether
Ethylbenzene
Hexachloro- 1 , 3 -butadiene
Methyl Isobutyl Ketone
Methyl Methacrylate
Methyl tert-Butyl Ether
w-Octane
Propylene
Styrene
PXSS
6.29
NA
2.24
NA
NA
NA
4.56
4.52
3.74
NA
8.21
2.86
4.58
NA
NA
NA
NA
NA
6.73
1.71
NA
3.37
NA
NA
NA
38.57
NA
NA
NA
3.63
NA
NA
5.95
NA
30.63
NA
NA
3.23
4.33
5.07
ROIL
9.13
NA
12.93
NA
NA
NA
3.70
7.45
15.73
NA
0.00
9.31
4.40
NA
NA
NA
NA
NA
NA
3.81
NA
2.83
NA
NA
NA
10.63
NA
NA
NA
4.01
NA
NA
13.28
NA
56.50
NA
NA
9.88
7.28
17.98
S4MO
5.24
NA
8.95
NA
NA
NA
9.68
7.81
17.93
NA
NA
4.85
2.66
NA
NA
NA
NA
NA
8.16
2.72
NA
8.81
NA
NA
5.73
11.23
NA
NA
NA
4.39
NA
NA
9.65
NA
15.50
NA
NA
18.08
6.39
16.99
SEWA
8.30
NA
6.32
NA
NA
NA
5.66
3.50
8.33
NA
NA
14.14
5.77
NA
NA
NA
NA
NA
NA
5.91
NA
0.00
NA
NA
NA
21.65
NA
NA
NA
19.40
NA
NA
9.37
NA
14.62
NA
NA
5.96
7.47
6.48
SPIL
18.82
NA
6.77
NA
NA
NA
NA
23.04
6.94
NA
NA
10.02
4.29
NA
NA
NA
NA
NA
NA
1.22
NA
4.23
NA
NA
NA
20.40
NA
NA
NA
0.00
NA
NA
4.59
NA
17.80
NA
NA
11.62
8.89
10.05
SSSD
11.75
NA
12.85
NA
NA
NA
0.00
5.35
12.21
NA
NA
2.70
7.85
NA
NA
NA
NA
NA
NA
6.80
NA
4.76
NA
NA
NA
16.46
NA
NA
NA
10.75
NA
NA
18.04
NA
21.64
NA
NA
8.77
9.15
15.94
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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.
30-9
-------
Table 30-2. VOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
1 , 1 ,2,2-Tetrachloroethane
Tetrachloroethylene
Toluene
1 ,2,4-Trichlorobenzene
1,1,1 -Trichloroethane
1 , 1 ,2-Trichloroethane
Trichloroethylene
Trichlorofluoromethane
Trichlorotrifluoroethane
1 ,2,4-Trimethylbenzene
1 , 3 ,5 -Trimethy Ibenzene
Vinyl chloride
7w,/>-Xylene
o-Xylene
Average by Site
PXSS
NA
4.37
4.93
NA
13.86
NA
NA
1.59
2.90
6.25
7.77
NA
5.88
5.76
7.17
ROIL
NA
NA
8.83
NA
NA
NA
NA
3.75
5.51
11.82
14.77
NA
10.94
14.11
10.77
S4MO
NA
10.65
8.09
NA
NA
NA
2.05
2.98
3.84
12.69
14.09
NA
10.42
11.66
8.93
SEWA
NA
2.32
7.49
NA
NA
NA
NA
5.49
5.31
9.63
1.66
NA
7.37
8.94
7.96
SPIL
NA
2.83
12.69
NA
NA
NA
12.50
1.41
3.45
5.57
8.08
NA
3.82
4.85
8.49
SSSD
NA
NA
15.72
NA
NA
NA
NA
6.77
8.19
11.78
9.32
NA
8.23
9.48
10.20
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designaled NATTS Site
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.
30-10
-------
Table 30-2. VOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
Acetylene
fer/-Amyl Methyl Ether
Benzene
Bromochloromethane
Bromodichloromethane
Bromoform
Bromomethane
1,3 -Butadiene
Carbon Tetrachloride
Chlorobenzene
Chloroethane
Chloroform
Chloromethane
Chloroprene
Dibromochloromethane
1 ,2-Dibromoethane
/w-Dichlorobenzene
o-Dichlorobenzene
£>-Dichlorobenzene
Dichlorodifluoromethane
1 , 1 -Dichloroethane
1 ,2-Dichloroethane
1 , 1 -Dichloroethene
cis- 1 ,2-Dichloroethylene
/ra«s-l,2-Dichloroethylene
Dichloromethane
1 ,2-Dichloropropane
cis- 1 ,3 -Dichloropropene
trans- 1 , 3 -Dichloropropene
Dichlorotetrafluoroethane
Ethyl Aery late
Ethyl fert-Butyl Ether
Ethylbenzene
Hexachloro- 1 ,3 -butadiene
Methyl Isobutyl Ketone
Methyl Methacrylate
Methyl tert-Butyl Ether
w-Octane
Propylene
Styrene
TMOK
3.16
NA
4.54
NA
NA
NA
27.90
4.30
4.01
NA
NA
6.29
5.28
NA
NA
NA
NA
NA
1.59
2.20
NA
5.46
NA
NA
NA
31.20
NA
NA
NA
5.53
NA
NA
6.19
NA
46.64
19.64
NA
7.55
12.01
9.85
TOOK
3.40
NA
5.70
NA
NA
NA
15.71
9.39
7.21
NA
NA
7.71
5.36
NA
NA
NA
NA
NA
13.69
3.70
NA
7.88
NA
NA
NA
3.60
NA
NA
NA
7.08
NA
NA
2.68
NA
35.11
NA
NA
5.81
5.35
6.54
TVKY
10.63
NA
12.79
NA
NA
NA
4.29
6.09
11.15
NA
11.59
13.80
6.90
NA
NA
NA
NA
NA
NA
6.09
5.26
34.06
8.76
14.14
1.99
35.31
NA
NA
NA
10.08
NA
NA
9.33
NA
45.92
NA
NA
13.65
37.54
16.16
#of
Pairs
149
0
148
0
7
1
28
135
148
1
11
93
149
0
3
0
0
0
20
149
2
104
1
1
4
149
0
0
0
149
0
19
140
0
129
6
16
130
149
93
Average by
Pollutant
9.96
NA
9.97
NA
14.74
5.05
8.68
7.85
9.51
5.05
7.32
11.00
5.59
NA
18.94
NA
NA
NA
5.48
4.45
5.26
7.76
8.76
14.14
2.43
19.66
NA
NA
NA
7.24
NA
8.94
8.75
NA
27.46
8.37
11.46
9.33
10.16
9.85
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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.
30-11
-------
Table 30-2. VOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
1 , 1 ,2,2-Tetrachloroethane
Tetrachloroethylene
Toluene
1 ,2,4-Trichlorobenzene
1,1,1 -Trichloroethane
1 , 1 ,2-Trichloroethane
Trichloroethylene
Trichlorofluoromethane
Trichlorotrifluoroethane
1,2,4-Trimethylbenzene
1 , 3 ,5 -Trimethy Ibenzene
Vinyl chloride
7w,/>-Xylene
o-Xylene
Average by Site
TMOK
NA
8.66
3.90
NA
NA
NA
NA
2.31
2.96
6.07
6.37
NA
4.14
5.24
9.35
TOOK
NA
5.62
4.79
NA
NA
NA
NA
4.15
4.74
5.69
4.93
NA
3.34
4.34
7.34
TVKY
NA
NA
19.90
NA
NA
6.82
NA
12.18
6.81
16.32
8.82
6.24
9.23
9.74
13.72
#of
Pairs
0
49
149
0
1
2
7
149
149
130
88
6
143
140
-
Average by
Pollutant
NA
6.62
10.50
NA
13.86
6.82
6.60
4.82
5.97
9.12
6.83
6.24
10.32
9.42
9.28
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designaled NATTS Site
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.
30.2.2 SNMOC Method Precision
The SNMOC method precision for duplicate and collocated samples is presented in
Table 30-3 as the CV per pollutant per site, the average CV per site, the average CV per
pollutant, and the overall average CV across all SNMOCs listed. The results from duplicate and
collocated samples exhibit low- to high-level variability among the pollutants and sites, ranging
from a CV of 0 percent (m,p-xy\ene for BRCO) to 71.98 percent (w-undecane for SSSD). The
pollutant-specific average CV ranged from 0.97 percent (c/s-2-butene) to 59.60 percent
(w-undecane). The site-specific average CV ranged from 4.31 percent (BRCO) to 11.02 percent
(SSSD), with an overall method average of 8.84 percent.
30-12
-------
Table 30-3. SNMOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant
Pollutant
Acetylene
Benzene
1,3 -Butadiene
w-Butane
c/s-2-Butene
/ra«s-2-Butene
Cyclohexane
Cyclopentane
Cyclopentene
w-Decane
1-Decene
/w-Diethylbenzene
/>-Diethylbenzene
2,2-Dimethylbutane
2,3 -Dimethylbutane
2,3 -Dimethylpentane
2,4-Dimethylpentane
w-Dodecane
1-Dodecene
Ethane
2-Ethyl-l-butene
Ethylbenzene
Ethylene
/w-Ethyltoluene
o-Ethyltoluene
£>-Ethyltoluene
w-Heptane
1-Heptene
w-Hexane
1-Hexene
c/s-2-Hexene
/ra«s-2-Hexene
Isobutane
Isobutene/ 1 -Butene
Isopentane
Isoprene
Isopropylbenzene
2-Methyl-l-butene
3 -Methyl- 1 -butene
2-Methyl- 1 -pentene
BMCO
9.43
6.47
NA
6.67
NA
NA
8.64
NA
NA
15.87
NA
NA
NA
NA
NA
NA
NA
NA
NA
5.81
NA
29.63
3.61
NA
NA
NA
11.19
NA
1.96
NA
NA
NA
8.76
NA
9.59
NA
NA
NA
NA
NA
BRCO
5.21
1.24
NA
5.89
NA
NA
5.06
3.43
NA
16.55
NA
NA
NA
7.82
NA
1.36
0.93
NA
NA
4.32
NA
NA
3.29
NA
NA
NA
1.27
6.33
5.36
NA
NA
NA
5.39
NA
5.22
NA
NA
NA
NA
NA
BTUT
6.56
3.76
4.98
5.85
0.97
4.98
2.85
3.69
NA
7.24
NA
8.43
NA
3.16
3.86
3.74
4.30
NA
NA
3.45
NA
9.79
3.80
6.70
13.31
5.03
8.07
NA
3.11
NA
NA
NA
13.19
NA
6.48
12.79
NA
5.63
NA
NA
NBIL
3.74
5.52
4.67
4.14
NA
NA
6.20
4.49
NA
20.15
4.24
47.02
NA
7.63
2.34
4.34
3.59
6.03
NA
10.40
NA
8.85
3.18
6.41
12.71
3.53
3.15
NA
5.01
NA
NA
NA
2.58
NA
8.87
3.10
NA
NA
NA
NA
SSSD
9.33
11.33
4.97
2.82
NA
8.91
6.66
30.90
NA
5.15
NA
3.00
39.27
10.38
2.56
9.00
3.41
NA
NA
3.35
NA
12.34
6.79
4.18
2.20
NA
5.08
NA
23.93
NA
NA
NA
12.12
NA
21.73
7.29
NA
5.01
NA
NA
#of
Pairs
22
21
8
22
2
9
18
17
0
16
1
5
1
14
13
16
15
1
0
22
0
20
22
14
5
4
20
1
22
0
0
0
21
0
14
9
0
7
0
0
Average by
Pollutant
6.86
5.66
4.88
5.08
0.97
6.95
5.88
10.63
NA
12.99
4.24
19.48
39.27
7.25
2.92
4.61
3.06
6.03
NA
5.46
NA
15.15
4.13
5.77
9.41
4.28
5.75
6.33
7.87
NA
NA
NA
8.41
NA
10.38
7.73
NA
5.32
NA
NA
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall averaj
from the pollutant-specific averages and is provided in the final column of the
;e CV for this method is calculated
table.
30-13
-------
Table 30-3. SNMOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
4-Methyl- 1 -pentene
2-Methyl-2-butene
Methylcyclohexane
Methylcyclopentane
2-Methylheptane
3-Methylheptane
2-Methylhexane
3-Methylhexane
2-Methylpentane
3-Methylpentane
w-Nonane
1-Nonene
w-Octane
1-Octene
w-Pentane
1 -Pentene
c/s-2-Pentene
trans-1 -Pentene
a-Pinene
6-Pinene
Propane
w-Propylbenzene
Propylene
Propyne
Styrene
Toluene
w-Tridecane
1-Tridecene
1 ,2,3 -Trimethylbenzene
1 ,2,4-Trimethylbenzene
1 ,3 ,5-Trimethylbenzene
2,2,3 -Trimethylpentane
2,2,4-Trimethylpentane
2,3 ,4-Trimethylpentane
w-Undecane
1-Undecene
/w-Xylene/p-Xylene
o-Xylene
SNMOC (Sum of Knowns)
Sum of Unknowns
Average by Site
BMCO
NA
NA
12.30
10.98
NA
NA
8.74
NA
13.72
11.72
10.41
NA
16.37
NA
1.59
NA
NA
NA
NA
NA
4.66
NA
0.57
NA
NA
6.05
NA
NA
NA
21.45
NA
NA
NA
NA
NA
NA
8.51
12.98
6.99
35.28
10.71
BRCO
NA
NA
4.20
3.55
0.76
1.39
4.15
4.36
4.68
4.46
0.26
NA
1.74
NA
5.29
NA
NA
NA
NA
10.34
4.67
NA
4.19
NA
NA
4.01
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
0.00
6.58
4.56
3.08
4.31
BTUT
NA
1.67
3.39
2.67
2.86
5.99
3.73
6.70
2.97
4.45
5.53
NA
3.84
NA
3.42
23.95
NA
5.08
6.63
13.28
2.10
3.15
3.95
NA
NA
4.78
NA
NA
9.98
2.66
3.55
NA
5.62
6.99
NA
NA
3.37
4.88
2.54
9.89
5.73
NBIL
NA
1.68
7.87
5.66
9.63
4.35
13.82
6.38
4.81
4.07
10.84
NA
32.66
14.67
23.74
8.00
NA
1.74
3.11
38.46
4.99
10.87
7.37
1.46
NA
10.75
NA
NA
2.97
10.27
16.50
NA
4.28
4.26
47.22
8.84
9.44
11.90
4.07
14.09
9.53
SSSD
NA
4.48
6.53
8.99
NA
0.47
9.40
14.60
3.00
6.82
3.13
11.89
10.59
NA
29.80
6.68
NA
6.01
NA
15.24
3.98
NA
6.51
NA
NA
11.89
NA
NA
3.84
4.23
NA
NA
8.70
11.12
71.98
NA
7.45
5.13
39.24
9.81
11.02
#of
Pairs
0
6
19
22
9
10
22
14
21
22
17
4
17
1
22
15
0
13
5
8
22
4
22
1
0
22
0
0
6
19
5
0
20
15
3
1
22
22
22
22
~
Average by
Pollutant
NA
2.61
6.86
6.37
4.42
3.05
7.97
8.01
5.84
6.31
6.03
11.89
13.04
14.67
12.77
12.88
NA
4.27
4.87
19.33
4.08
7.01
4.52
1.46
NA
7.50
NA
NA
5.60
9.65
10.03
NA
6.20
7.45
59.60
8.84
5.75
8.30
11.48
14.43
8.84
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall averaj
from the pollutant-specific averages and is provided in the final column of the
;e CV for this method is calculated
table.
30-14
-------
30.2.3 Carbonyl Compound Method Precision
Table 30-4 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 all carbonyl compounds listed. The duplicate and
collocated sample results exhibit low- to high-level variability, ranging from a CV of
0.94 percent (acetaldehyde for OCOK) to 120.31 percent (benzaldehyde for BMCO). The
pollutant-specific average CV ranged from 3.83 percent (acetaldehyde) to 18.84 percent
(tolualdehydes). The site-specific average CV ranged from 2.74 percent (OCOK) to
56.45 percent (BMCO). Only two sites have average CVs greater than 15 percent; while ROIL is
just outside the MQO (16.40 percent), BMCO is significantly higher (56.45 percent). The
precision for BMCO is based on a single pair of samples with poor precision. The overall
average method precision was 9.14 percent for carbonyl compounds.
Table 30-4. Carbonyl Compound Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant
Pollutant
Acetaldehyde
Acetone
Benzaldehyde
2-Butanone
Butyraldehyde
Crotonaldehyde
2,5-Dimethylbenzaldehyde
Formaldehyde
Hexaldehyde
Isovaleraldehyde
Propionaldehyde
Tolualdehydes
Valeraldehyde
Average by Site
ADOK
0.96
19.12
4.72
18.10
1.63
4.31
NA
4.98
13.99
NA
4.65
34.00
11.57
10.73
AZFL
4.18
9.15
5.70
9.78
10.35
5.21
NA
10.35
8.78
NA
4.43
34.61
9.02
10.14
BMCO
27.40
80.77
120.31
20.05
47.14
6.15
NA
13.17
106.79
NA
65.54
NA
77.14
56.45
BTUT
3.65
9.28
4.26
5.83
5.15
7.33
NA
3.88
9.97
NA
7.51
37.89
6.26
9.18
BURVT
1.99
2.26
5.67
3.25
4.69
2.40
NA
3.70
5.19
NA
3.22
4.94
4.04
3.76
CHNJ
2.92
7.80
5.32
5.79
6.67
7.69
NA
3.48
6.33
NA
7.04
7.59
7.18
6.16
DEMI
6.63
3.99
6.48
4.71
16.11
8.48
NA
11.19
7.82
NA
12.84
6.14
17.88
9.30
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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.
30-15
-------
Table 30-4. Carbonyl Compound Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
Acetaldehyde
Acetone
Benzaldehyde
2-Butanone
Butyraldehyde
Crotonaldehyde
2,5-Dimethylbenzaldehyde
Formaldehyde
Hexaldehyde
Isovaleraldehyde
Propionaldehyde
Tolualdehydes
Valeraldehyde
Average by Site
ELNJ
2.12
16.37
3.00
11.86
4.56
4.86
NA
2.51
7.06
NA
2.19
10.12
5.05
6.34
GLKY
1.58
3.97
5.78
3.49
3.28
7.08
NA
2.08
3.74
NA
2.38
10.82
10.65
4.98
GPCO
1.37
6.27
5.28
3.87
2.99
3.12
NA
0.97
4.72
NA
2.96
14.54
5.49
4.69
INDEM
8.04
6.71
9.06
7.05
6.19
6.86
NA
9.59
6.01
NA
8.35
6.03
7.53
7.40
NBIL
2.63
5.76
6.87
5.15
4.75
5.78
NA
3.86
5.02
NA
4.15
18.80
5.84
6.24
NBNJ
3.44
5.99
4.08
10.22
6.57
3.03
NA
10.25
7.22
NA
11.03
5.71
6.28
6.71
OCOK
0.94
2.90
3.62
3.63
1.48
2.69
NA
1.32
3.71
NA
1.85
4.77
3.17
2.74
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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.
Table 30-4. Carbonyl Compound Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
Acetaldehyde
Acetone
Benzaldehyde
2-Butanone
Butyraldehyde
Crotonaldehyde
2,5-Dimethylbenzaldehyde
Formaldehyde
Hexaldehyde
Isovaleraldehyde
Propionaldehyde
Tolualdehydes
Valeraldehyde
Average by Site
ORFL
1.32
3.90
4.85
6.30
2.48
5.87
NA
2.17
7.99
NA
2.85
35.79
7.74
7.39
PROK
1.82
1.63
5.91
2.96
6.15
8.46
NA
2.44
8.49
NA
4.72
6.70
5.75
5.00
PXSS
3.50
1.65
4.83
6.68
4.54
2.92
NA
11.00
7.09
NA
3.70
12.39
6.52
5.89
ROIL
1.57
11.51
32.86
17.81
23.25
7.19
NA
7.31
27.65
NA
6.84
21.77
22.65
16.40
S4MO
1.84
7.77
2.27
4.69
3.98
3.77
NA
2.50
1.99
NA
4.89
41.59
6.88
7.47
SEWA
1.73
1.54
9.39
2.63
3.92
4.85
NA
3.13
4.89
NA
3.70
20.28
4.73
5.53
SKFL
4.24
13.83
11.46
10.87
14.61
6.85
NA
5.33
12.15
NA
5.24
27.21
11.05
11.17
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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.
30-16
-------
Table 30-4. Carbonyl Compound Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
Acetaldehyde
Acetone
Benzaldehyde
2-Butanone
Butyraldehyde
Crotonaldehyde
2,5-Dimethylbenzaldehyde
Formaldehyde
Hexaldehyde
Isovaleraldehyde
Propionaldehyde
Tolualdehydes
Valeraldehyde
Average by Site
SPIL
4.59
8.46
8.18
5.36
7.37
9.78
NA
7.86
10.60
NA
2.00
8.89
6.94
7.28
SSSD
3.24
9.22
6.46
7.60
5.89
3.86
NA
3.38
6.02
NA
4.70
42.64
9.99
9.36
SYFL
4.34
16.57
7.70
11.24
8.16
6.22
NA
12.07
22.08
NA
6.71
33.07
18.71
13.35
TMOK
1.47
1.90
6.27
1.91
2.33
4.33
NA
1.71
3.12
NA
1.87
23.00
4.43
4.76
TOOK
1.64
4.27
5.94
4.14
3.05
2.97
NA
1.98
7.96
NA
4.21
7.81
5.50
4.50
WPIN
4.42
7.61
5.42
7.07
6.56
8.35
NA
4.35
10.50
NA
5.10
12.73
8.20
7.30
#of
Pairs
163
163
159
154
163
161
0
163
158
0
163
131
162
-
Average by
Pollutant
3.83
10.01
11.17
7.48
7.92
5.57
NA
5.43
12.11
NA
7.21
18.84
10.97
9.14
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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.
30.2.4 PAH Method Precision
The method precision results for collocated PAH samples are shown in Table 30-5 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 is based on data from five sites for 2012.
The results from collocated samples exhibit low- to high-level variability, ranging from a
CV of 2.97 percent (acenaphthene for DEMI) to 71.02 percent (anthracene for SDGA). The
pollutant-specific average CV ranged from 7.94 percent (phenanthrene) to 25.05 percent
(chrysene). The site-specific average CV ranged from 9.28 percent (SEWA) to 27.33 percent
(SDGA). SDGA was the only site with a site-specific average CV greater than 15 percent. The
overall average method precision was 14.72 percent.
30-17
-------
Table 30-5. PAH Method Precision: Coefficient of Variation
Based on Collocated Samples by Site and Pollutant
Pollutant
Acenaphthene
Acenaphthylene
Anthracene
Benzo(a)anthracene
Benzo(a)pyrene
Benzo(b)fluoranthene
Benzo(e)pyrene
Benzo(g,h,i)perylene
B enzo (k)fluoranthene
Chrysene
Coronene
Cyclopenta[cd]pyrene
Dibenz(a,h)anthracene
Fluoranthene
Fluorene
9-Fluorenone
Indeno( 1,2,3 -cd)pyrene
Naphthalene
Perylene
Phenanthrene
Pyrene
Retene
Average by Site
DEMI
2.97
18.88
8.46
15.20
14.89
10.18
10.43
11.73
14.34
11.71
18.10
NA
20.20
6.76
3.82
9.20
11.94
6.03
18.49
5.56
7.81
5.89
11.08
RUCA
7.93
32.88
18.79
9.89
NA
8.26
10.21
8.34
NA
15.36
NA
NA
NA
5.63
6.09
8.12
9.37
14.02
NA
5.08
6.68
52.67
13.71
SDGA
15.64
NA
71.02
NA
NA
26.95
13.08
13.50
NA
65.08
NA
NA
NA
34.77
18.59
16.28
18.91
14.72
NA
13.23
46.63
14.20
27.33
SEWA
5.66
19.66
10.44
10.09
12.62
5.26
8.99
7.13
18.84
7.12
13.21
NA
11.29
8.17
7.52
8.74
8.03
4.55
NA
4.90
8.47
4.87
9.28
SYFL
10.81
8.76
12.63
6.21
NA
8.66
17.98
18.23
NA
25.97
NA
NA
NA
13.28
11.02
15.76
27.33
10.73
NA
10.94
19.77
17.36
14.72
#of
Pairs
28
11
22
8
5
18
10
11
5
26
6
NA
2
29
29
29
9
29
1
29
29
15
~
Average by
Pollutant
8.60
20.05
24.27
10.35
13.75
11.86
12.14
11.79
16.59
25.05
15.66
NA
15.75
13.72
9.41
11.62
15.11
10.01
18.49
7.94
17.87
19.00
14.72
NA=No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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.
30.2.5 Metals Method Precision
The method precision for all collocated metals samples are presented in Table 30-6 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 high-level variability among sites, ranging
from a CV of 0 percent (antimony and cobalt for UNVT) to 33.91 percent (arsenic for UNVT).
The pollutant-specific average CV ranged from 4.38 percent (lead) to 16.87 percent (cadmium).
30-18
-------
The site-specific average CV ranged from 6.14 percent (NBIL) to 12.02 percent (S4MO). The
overall average method precision for metals was 9.12 percent.
Table 30-6. Metals Method Precision: Coefficient of Variation
Based on Collocated Samples by Site and Pollutant
Pollutant
Antimony
Arsenic
Beryllium
Cadmium
Chromium
Cobalt
Lead
Manganese
Mercury
Nickel
Selenium
Average by Site
ASKY-M
2.87
6.40
9.30
16.42
NA
5.27
4.15
3.99
9.14
7.00
5.44
7.00
BOMA
2.63
3.00
12.01
27.86
4.28
7.86
3.82
2.37
14.88
8.02
3.24
8.18
BTUT
3.78
11.79
14.29
10.59
NA
13.37
5.94
9.87
NA
11.20
NA
10.10
GLKY
7.12
16.87
NA
29.63
NA
17.71
4.07
3.44
NA
1.21
8.36
11.05
NBIL
5.90
3.61
9.43
4.95
NA
12.22
0.32
0.15
5.89
16.47
2.42
6.14
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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.
Table 30-6. Metals Method Precision: Coefficient of Variation
Based on Collocated Samples by Site and Pollutant (Continued)
Pollutant
Antimony
Arsenic
Beryllium
Cadmium
Chromium
Cobalt
Lead
Manganese
Mercury
Nickel
Selenium
Average by Site
S4MO
4.54
11.22
15.92
11.90
NA
15.15
4.25
5.23
27.67
17.00
7.32
12.02
TOOK
11.56
5.05
8.61
13.90
7.68
18.85
6.97
5.98
15.78
7.08
4.46
9.63
UNVT
0.00
33.91
NA
19.68
NA
0.00
5.51
6.39
NA
6.63
NA
10.30
#of
Pairs
191
191
94
197
61
163
199
199
102
165
172
~
Average by
Pollutant
4.80
11.48
11.59
16.87
5.98
11.30
4.38
4.68
14.67
9.33
5.21
9.12
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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.
30-19
-------
30.2.6 Hexavalent Chromium Method Precision
Table 30-7 presents the method precision results from collocated hexavalent chromium
samples as the CV per site and the overall average CV for the method. All samples evaluated in
this section are collocated samples. The site-specific CV ranged from 0.74 percent (SYFL) to
40.12 percent (WADC), with an overall average method precision of 13.77 percent.
Table 30-7. Hexavalent Chromium Method Precision: Coefficient of Variation
Based on Collocated Samples by Site
Site
BOMA
BTUT
BXNY
CAMS 35
CHSC
DEMI
GLKY
GPCO
HOW
MIWI
MONY
NBIL
PRRI
PXSS
RIVA
ROCH
S4MO
SDGA
SEWA
SKFL
SYFL
WADC
# of Pairs
Average by Site
Average CV
(%)
5.80
9.20
12.66
21.89
9.71
6.49
5.22
15.45
16.50
16.68
5.28
25.69
6.20
10.11
13.86
13.85
6.08
31.87
8.59
21.04
0.74
40.12
102
13.77
BOLD ITALICS = EPA-de signaled
NATTS Site
Orange shading indicates the average CV
for this method.
30-20
-------
30.3 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.
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 30.2.
Table 30-8 presents the 2012 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 six methods was less than 7 percent. This table also
includes the number of pairs that were included in the calculation of the analytical precision. The
number of pairs including those with concentrations less than the MDL is also included in
Table 30-8 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.
30-21
-------
Table 30-8. Analytical Precision by Analytical Method
Method/Pollutant
Group
voc
(TO-15)
SNMOC
Carbonyl Compounds
(TO-11A)
PAH
(TO-13)
Metals Analysis
(Method IO-3.5/FEM)
Hexavalent Chromium
(ASTMD7641)
MQO
Average
Coefficient of
Variation
(%)
6.88
5.29
2.91
3.93
4.94
6.65
Number of
Pairs Included
in the
Calculation
7,242
1,755
3,592
1,411
3,859
223
Total Number of
Pairs Without
the > MDL
exclusion
9,441
2,241
3,613
1,868
4,771
223
15. 00 percent CV
Tables 30-9 through 30-14 present analytical precision for VOCs, SNMOCs, carbonyl
compounds, PAHs, metals, 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 30-9 through 30-14, the number of pairs in
comparison to the respective tables listed for duplicate or collocated analyses in Tables 30-2
through 30-7, is higher, the reason for which is two-fold. One reason is because each 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 30-9
through 30-14 is higher than Tables 30-2 through 30-7. The replicate analyses of duplicate,
collocated, and individual samples indicate that the analytical precision level is within the
program MQOs.
30.3.1 VOC Analytical Precision
Table 30-9 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 was within the program MQO of 15 percent. The CV ranged from 0 percent
for several pollutants and several sites to 74.57 percent (methyl tert-butyl ether for SEW A). The
pollutant-specific average CV ranged from 0 percent (w-dichlorobenzene) to 20.02 percent
30-22
-------
(methyl tert-buty\ ether). The site-specific average CV ranged from 4.78 percent (ELNJ) to
9.25 percent (NBIL). The overall average analytical precision was 6.88 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.
Table 30-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant
Pollutant
Acetylene
fer/-Amyl Methyl Ether
Benzene
Bromochloromethane
Bromodichloromethane
Bromoform
Bromomethane
1,3 -Butadiene
Carbon Tetrachloride
Chlorobenzene
Chloroethane
Chloroform
Chloromethane
Chloroprene
Dibromochloromethane
1 ,2-Dibromoethane
/w-Dichlorobenzene
o-Dichlorobenzene
/>-Dichlorobenzene
Dichlorodifluoromethane
1 , 1 -Dichloroethane
1 ,2-Dichloroethane
1 , 1 -Dichloroethene
cis- 1 ,2-Dichloroethylene
trans- 1 ,2-Dichloroethylene
Dichloromethane
1 ,2-Dichloropropane
cis- 1 , 3 -D ichloropropene
trans- 1 ,3 -Dichloropropene
Dichlorotetrafluoroethane
Ethyl Aery late
Ethyl tert-Butyl Ether
Ethylbenzene
Hexachloro- 1 , 3 -butadiene
ADOK
7.82
NA
10.42
NA
NA
NA
8.91
6.36
6.96
NA
NA
7.85
6.99
NA
NA
NA
NA
NA
2.94
5.89
NA
5.82
NA
NA
NA
6.68
NA
NA
NA
7.60
NA
NA
8.46
NA
BTUT
5.21
NA
8.61
NA
NA
NA
13.69
7.99
9.36
NA
NA
9.33
5.73
NA
NA
NA
NA
NA
NA
5.78
NA
9.26
NA
NA
NA
5.58
NA
NA
NA
11.12
NA
9.16
5.70
NA
BURVT
4.31
NA
7.15
NA
NA
NA
7.68
4.70
6.00
NA
8.03
16.81
3.56
NA
NA
NA
NA
NA
6.90
3.34
NA
7.89
NA
NA
3.03
5.10
NA
NA
NA
7.54
NA
NA
8.32
NA
CHNJ
7.33
NA
5.63
NA
NA
NA
5.75
5.30
5.22
NA
NA
5.96
2.82
NA
NA
NA
NA
NA
NA
2.78
NA
6.21
NA
NA
NA
3.99
NA
NA
NA
5.59
NA
4.28
6.51
NA
DEMI
5.54
NA
4.20
NA
NA
NA
8.28
4.53
3.90
3.80
NA
3.52
3.63
NA
NA
NA
NA
NA
2.13
3.64
NA
6.44
NA
NA
0.53
2.99
NA
NA
NA
4.13
NA
NA
5.83
NA
ELNJ
3.66
NA
4.44
NA
NA
NA
5.00
3.58
5.49
NA
NA
5.15
3.27
NA
NA
NA
NA
NA
6.88
3.27
NA
7.34
NA
NA
NA
4.31
NA
NA
NA
5.22
NA
5.08
5.02
NA
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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
is calculated from the pollutant-specific averages and is provided in the final column of the
method
table.
30-23
-------
Table 30-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
Methyl Isobutyl Ketone
Methyl Methacrylate
Methyl fer/-Butyl Ether
w-Octane
Propylene
Styrene
1 , 1 ,2,2-Tetrachloroethane
Tetrachloroethylene
Toluene
1 ,2,4-Trichlorobenzene
1,1,1 -Trichloroethane
1 , 1 ,2-Trichloroethane
Trichloroethylene
Trichlorofluoromethane
Trichlorotrifluoroethane
1 ,2,4-Trimethylbenzene
1 ,3 ,5-Trimethylbenzene
Vinyl chloride
m,p-Xylene
o-Xylene
Average by Site
ADOK
11.80
NA
NA
5.34
8.08
5.07
NA
NA
11.61
NA
NA
NA
NA
5.64
6.46
5.21
4.88
NA
7.75
8.15
7.20
BTUT
10.38
NA
13.24
6.19
5.81
6.85
NA
13.25
6.02
NA
NA
NA
NA
6.16
5.85
4.51
4.64
NA
4.60
5.77
7.68
BURVT
7.71
NA
NA
9.74
4.27
9.65
NA
6.70
6.21
NA
NA
NA
NA
3.39
4.27
7.88
7.29
NA
6.15
7.10
6.69
CHNJ
8.76
NA
5.22
6.99
3.30
7.96
NA
4.03
4.75
NA
NA
NA
NA
2.82
2.96
6.70
9.23
NA
6.85
6.90
5.53
DEMI
3.53
4.56
NA
9.04
4.75
19.75
NA
3.02
4.01
NA
NA
NA
NA
2.72
5.56
4.46
6.68
NA
4.24
4.20
4.99
ELNJ
8.86
3.21
7.67
5.12
3.89
6.65
NA
3.31
3.59
NA
NA
NA
0.00
3.32
3.52
5.94
6.82
NA
3.98
5.06
4.78
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designaled NATTS Site
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
is calculated from the pollutant-specific averages and is provided in the final column of the
method
table.
30-24
-------
Table 30-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
Acetylene
tert-Amyl Methyl Ether
Benzene
Bromochloromethane
Bromodichloromethane
Bromoform
Bromomethane
1,3 -Butadiene
Carbon Tetrachloride
Chlorobenzene
Chloroethane
Chloroform
Chloromethane
Chloroprene
Dibromochloromethane
1 ,2-Dibromoethane
/w-Dichlorobenzene
o-Dichlorobenzene
£>-Dichlorobenzene
Dichlorodifluoromethane
1 , 1 -Dichloroethane
1 ,2-Dichloroethane
1 , 1 -Dichloroethene
cis- 1 ,2-Dichloroethylene
trans- 1 ,2-Dichloroethylene
Dichloromethane
1 ,2-Dichloropropane
cis- 1 , 3 -D ichloropropene
trans- 1 , 3 -Dichloropropene
Dichlorotetrafluoroethane
Ethyl Aery late
Ethyl fert-Butyl Ether
Ethylbenzene
Hexachloro- 1 , 3 -butadiene
Methyl Isobutyl Ketone
Methyl Methacrylate
Methyl tert-Butyl Ether
w-Octane
Propylene
Styrene
GLKY
7.91
NA
7.86
NA
NA
NA
5.08
5.51
7.82
NA
NA
13.22
4.15
NA
NA
NA
NA
NA
NA
3.95
NA
4.80
NA
NA
NA
6.16
NA
NA
NA
6.63
NA
NA
8.26
NA
11.90
NA
0.00
6.81
5.56
5.15
GPCO
3.99
NA
4.15
NA
NA
NA
9.37
5.10
3.71
NA
NA
7.03
4.01
NA
NA
NA
NA
NA
13.99
3.68
NA
8.83
NA
NA
NA
5.90
NA
NA
NA
7.99
NA
NA
2.51
NA
15.86
6.06
NA
3.43
4.26
3.06
NBIL
8.33
NA
10.44
NA
11.74
14.38
8.35
9.75
8.34
NA
14.85
5.52
5.89
NA
12.50
NA
NA
NA
8.67
5.27
NA
10.35
NA
NA
NA
8.10
NA
NA
NA
6.41
NA
5.94
8.00
NA
13.44
12.76
5.73
9.12
7.01
9.62
NBNJ
5.71
NA
5.50
NA
NA
NA
5.99
6.53
6.45
NA
5.55
6.42
4.47
NA
NA
NA
NA
NA
8.06
4.54
NA
8.50
NA
NA
NA
5.52
NA
NA
NA
6.69
NA
6.06
7.50
NA
8.20
NA
6.71
9.21
6.23
7.60
OCOK
6.76
NA
6.54
NA
NA
NA
4.00
8.84
6.79
NA
NA
9.60
5.68
NA
NA
NA
NA
NA
3.99
5.58
NA
9.28
NA
NA
NA
5.24
NA
NA
NA
8.03
NA
NA
8.45
NA
8.31
NA
NA
5.56
5.49
10.18
PROK
4.78
NA
6.86
NA
1.37
NA
NA
2.53
5.98
NA
NA
10.41
2.90
NA
NA
NA
NA
NA
8.50
2.78
NA
9.58
NA
NA
NA
5.09
NA
NA
NA
7.59
NA
NA
13.56
NA
8.47
NA
NA
13.01
4.37
17.55
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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
is calculated from the pollutant-specific averages and is provided in the final column of the
method
table.
30-25
-------
Table 30-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
1 , 1 ,2,2-Tetrachloroethane
Tetrachloroethylene
Toluene
1 ,2,4-Trichlorobenzene
1,1,1 -Trichloroethane
1 , 1 ,2-Trichloroethane
Trichloroethylene
Trichlorofluoromethane
Trichlorotrifluoroethane
1 ,2,4-Trimethylbenzene
1,3,5 -Trimethy Ibenzene
Vinyl chloride
7w,/>-Xylene
o-Xylene
Average by Site
GLKY
NA
NA
5.76
NA
NA
NA
NA
4.47
4.68
7.62
9.14
NA
14.47
6.50
6.81
GPCO
NA
7.41
2.39
NA
NA
NA
NA
3.84
4.46
2.87
4.24
NA
2.06
2.02
5.47
NBIL
NA
9.75
11.79
11.31
NA
NA
15.09
5.16
4.85
7.08
10.39
NA
9.40
9.23
9.25
NBNJ
NA
6.61
5.89
NA
3.45
NA
NA
4.48
5.26
10.43
8.00
NA
7.04
7.64
6.56
OCOK
NA
11.69
4.09
NA
NA
NA
5.67
5.29
6.44
7.66
8.38
NA
7.11
8.42
7.04
PROK
NA
0.53
5.61
NA
NA
NA
NA
2.84
8.40
20.04
13.14
NA
8.74
11.75
7.85
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designaled NATTS Site
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
is calculated from the pollutant-specific averages and is provided in the final column of the
method
table.
30-26
-------
Table 30-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
Acetylene
tert-Amyl Methyl Ether
Benzene
Bromochloromethane
Bromodichloromethane
Bromoform
Bromomethane
1,3 -Butadiene
Carbon Tetrachloride
Chlorobenzene
Chloroethane
Chloroform
Chloromethane
Chloroprene
Dibromochloromethane
1 ,2-Dibromoethane
/w-Dichlorobenzene
o-Dichlorobenzene
£>-Dichlorobenzene
Dichlorodifluoromethane
1 , 1 -Dichloroethane
1 ,2-Dichloroethane
1 , 1 -Dichloroethene
cis- 1 ,2-Dichloroethylene
trans- 1 ,2-Dichloroethylene
Dichloromethane
1 ,2-Dichloropropane
cis- 1 , 3 -D ichloropropene
trans- 1 ,3 -Dichloropropene
Dichlorotetrafluoroethane
Ethyl Aery late
Ethyl fert-Butyl Ether
Ethylbenzene
Hexachloro- 1 , 3 -butadiene
Methyl Isobutyl Ketone
Methyl Methacrylate
Methyl tert-Butyl Ether
w-Octane
Propylene
Styrene
PXSS
9.80
NA
4.01
NA
NA
NA
9.34
4.11
4.47
NA
3.59
4.24
3.53
NA
NA
NA
NA
NA
4.56
4.11
NA
7.10
NA
NA
NA
3.99
NA
3.82
NA
5.09
NA
58.23
3.38
NA
9.59
NA
45.92
3.74
3.56
5.05
ROIL
11.29
NA
4.77
NA
NA
NA
5.26
7.75
4.43
NA
7.47
5.48
5.30
NA
NA
NA
NA
NA
NA
5.34
NA
9.15
NA
NA
NA
5.70
NA
NA
NA
5.68
NA
NA
7.70
NA
8.65
NA
NA
7.63
4.75
10.69
RUVT
6.13
NA
4.31
NA
NA
NA
8.05
4.43
3.43
NA
NA
NA
2.60
NA
NA
NA
NA
NA
5.44
2.52
NA
6.53
NA
NA
5.66
3.19
NA
NA
NA
5.81
NA
NA
5.59
NA
5.91
NA
NA
6.31
3.53
11.30
S4MO
3.32
NA
4.65
NA
NA
NA
5.72
5.89
4.65
NA
NA
7.19
2.92
NA
NA
NA
NA
NA
4.26
2.87
NA
10.07
NA
NA
4.96
3.84
NA
NA
NA
8.06
NA
25.71
6.57
NA
5.52
2.67
21.13
7.79
3.46
10.68
SEWA
5.09
NA
6.99
NA
NA
NA
19.71
5.79
8.15
NA
NA
3.41
3.17
NA
NA
NA
NA
NA
NA
3.15
NA
5.01
NA
NA
NA
3.38
NA
NA
NA
4.50
NA
NA
5.78
NA
12.71
NA
74.57
7.95
3.00
3.34
SPAZ
6.94
NA
6.25
NA
NA
NA
4.56
3.66
5.15
NA
NA
3.60
2.85
NA
NA
NA
NA
NA
6.67
4.33
NA
7.20
NA
NA
NA
3.40
NA
NA
NA
4.05
NA
NA
5.21
NA
8.91
2.69
NA
7.17
6.19
7.75
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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
is calculated from the pollutant-specific averages and is provided in the final column of the
method
table.
30-27
-------
Table 30-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
1 , 1 ,2,2-Tetrachloroethane
Tetrachloroethylene
Toluene
1 ,2,4-Trichlorobenzene
1,1,1 -Trichloroethane
1 , 1 ,2-Trichloroethane
Trichloroethylene
Trichlorofluoromethane
Trichlorotrifluoroethane
1 ,2,4-Trimethylbenzene
1 ,3 ,5-Trimethylbenzene
Vinyl chloride
7w,/>-Xylene
o-Xylene
Average by Site
PXSS
NA
4.31
3.37
NA
4.10
NA
NA
4.03
5.81
3.61
4.77
NA
3.68
3.33
7.94
ROIL
NA
NA
5.98
NA
NA
NA
NA
4.87
6.14
8.31
11.52
NA
6.50
7.41
6.99
RUVT
NA
7.11
4.37
0.00
NA
NA
NA
2.27
4.63
6.69
11.17
NA
5.43
6.61
5.35
S4MO
NA
6.26
4.83
NA
NA
NA
7.61
3.11
4.21
6.59
10.38
NA
5.56
6.17
6.89
SEWA
NA
1.93
4.26
NA
NA
NA
NA
3.05
3.47
5.26
2.58
NA
4.21
5.92
8.26
SPAZ
NA
2.59
5.19
NA
NA
NA
NA
2.57
3.40
5.37
6.71
NA
4.90
5.71
5.12
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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
is calculated from the pollutant-specific averages and is provided in the final column of the
method
table.
30-28
-------
Table 30-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
Acetylene
fer/-Amyl Methyl Ether
Benzene
Bromochloromethane
Bromodichloromethane
Bromoform
Bromomethane
1,3 -Butadiene
Carbon Tetrachloride
Chlorobenzene
Chloroethane
Chloroform
Chloromethane
Chloroprene
Dibromochloromethane
1 ,2-Dibromoethane
/w-Dichlorobenzene
o-Dichlorobenzene
£>-Dichlorobenzene
Dichlorodifluoromethane
1 , 1 -Dichloroethane
1 ,2-Dichloroethane
1 , 1 -Dichloroethene
cis- 1 ,2-Dichloroethylene
trans- 1 ,2-Dichloroethylene
Dichloromethane
1 ,2-Dichloropropane
cis- 1 ,3 -Dichloropropene
trans- 1 , 3 -Dichloropropene
Dichlorotetrafluoroethane
Ethyl Aery late
Ethyl fert-Butyl Ether
Ethylbenzene
Hexachloro- 1 ,3 -butadiene
Methyl Isobutyl Ketone
Methyl Methacrylate
Methyl tert-Butyl Ether
w-Octane
Propylene
Styrene
SPIL
5.09
NA
5.31
NA
NA
NA
7.44
6.37
5.87
NA
NA
7.68
4.66
NA
NA
NA
NA
NA
NA
4.38
NA
7.84
NA
NA
NA
4.73
NA
NA
NA
3.37
NA
NA
6.96
NA
7.83
NA
NA
9.25
5.02
7.85
SSSD
6.68
NA
10.97
NA
NA
NA
7.80
8.70
9.84
NA
NA
7.59
3.94
NA
NA
NA
NA
NA
NA
3.87
NA
5.62
NA
NA
NA
6.69
NA
NA
NA
7.19
NA
NA
8.69
NA
12.78
NA
NA
12.58
3.94
10.04
TMOK
4.16
NA
4.27
NA
NA
NA
7.80
4.00
5.09
NA
5.82
7.03
3.54
NA
NA
NA
0.00
NA
13.34
3.49
NA
9.64
NA
NA
NA
4.57
NA
NA
NA
7.46
NA
NA
4.62
NA
6.71
25.33
NA
6.79
4.07
8.79
TOOK
5.00
NA
4.55
NA
NA
NA
14.08
6.14
6.21
NA
NA
10.01
5.43
NA
NA
NA
NA
NA
10.70
5.13
NA
6.66
NA
NA
NA
4.89
NA
NA
NA
7.91
NA
4.88
4.35
NA
7.23
NA
NA
5.08
4.84
10.37
TVKY
5.86
NA
6.69
NA
NA
NA
6.20
5.29
6.49
NA
6.39
5.45
4.49
NA
NA
NA
NA
NA
NA
4.04
4.39
4.01
5.17
16.01
3.84
5.40
NA
NA
NA
5.92
NA
NA
7.62
NA
11.82
NA
NA
8.72
4.19
7.75
UNVT
6.38
NA
7.48
NA
NA
NA
9.37
2.86
5.35
NA
NA
15.35
5.13
NA
NA
NA
NA
NA
NA
4.32
NA
5.77
NA
NA
NA
6.34
NA
NA
NA
8.37
NA
6.15
9.42
NA
7.89
NA
NA
7.21
6.25
11.76
#of
Pairs
338
0
334
0
14
2
104
295
337
2
29
215
338
0
7
0
1
0
68
338
4
239
2
2
11
338
0
1
0
338
0
43
318
0
304
17
38
300
338
219
Average by
Pollutant
6.13
NA
6.34
NA
6.56
14.38
8.15
5.65
6.05
3.80
7.39
7.73
4.19
NA
12.50
NA
0.00
NA
7.14
4.09
4.39
7.45
5.17
16.01
3.60
5.03
NA
3.82
NA
6.58
NA
13.94
6.83
NA
9.28
8.18
20.02
7.49
4.83
8.94
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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
is calculated from the pollutant-specific averages and is provided in the final column of the
method
table.
30-29
-------
Table 30-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
Propylene
Styrene
1 , 1 ,2,2-Tetrachloroethane
Tetrachloroethylene
Toluene
1 ,2,4-Trichlorobenzene
1,1,1 -Trichloroethane
1 , 1 ,2-Trichloroethane
Trichloroethylene
Trichlorofluoromethane
Trichlorotrifluoroethane
1 ,2,4-Trimethylbenzene
1,3,5 -Trimethy Ibenzene
Vinyl chloride
m,p-Xylene
o-Xylene
Average by Site
SPIL
5.02
7.85
NA
6.00
6.08
NA
NA
NA
5.56
4.23
4.09
6.70
6.62
NA
4.52
5.84
5.97
SSSD
3.94
10.04
NA
NA
8.35
NA
NA
NA
NA
4.23
5.09
11.82
15.15
NA
8.33
8.62
8.20
TMOK
4.07
8.79
NA
11.08
4.20
NA
NA
NA
NA
3.23
3.90
5.59
6.87
NA
3.90
4.65
6.43
TOOK
4.84
10.37
NA
11.02
3.67
NA
NA
NA
3.14
5.13
5.92
5.35
6.84
NA
3.87
3.95
6.38
TVKY
4.19
7.75
NA
4.71
6.50
NA
NA
3.27
NA
4.38
4.44
9.10
9.06
3.15
6.98
7.71
6.29
UNVT
6.25
11.76
NA
NA
6.46
NA
NA
NA
NA
3.74
5.54
9.33
3.42
NA
8.93
8.93
7.16
#of
Pairs
338
219
0
110
338
2
3
4
15
338
337
296
208
13
327
317
-
Average by
Pollutant
4.83
8.94
NA
6.39
5.62
5.66
3.78
3.27
6.18
3.96
4.97
7.25
7.83
3.15
6.22
6.57
6.88
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designaled NATTS Site
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
is calculated from the pollutant-specific averages and is provided in the final column of the
method
table.
30.3.2 SNMOC Analytical Precision
Table 30-10 presents analytical precision results from replicate analyses of duplicate,
collocated, and select individual samples as the CV per pollutant per site, the average CV per
site, the average CV per pollutant, and the overall average CV across the SNMOCs listed. The
CV ranged from 0 percent (methylcyclopentane and 3-methylhexane for BRCO) to 41.15 percent
(w-undecane for SSSD). The pollutant-specific average CV ranged from 0.94 percent (propyne)
to 27.12 percent (p-diethyIbenzene). The site-specific average CV ranged from 1.73 percent
(BRCO) to 8.61 percent (SSSD). The overall program average CV was 5.29 percent.
30-30
-------
Table 30-10. SNMOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant
Pollutant
Acetylene
Benzene
1,3 -Butadiene
w-Butane
c/s-2-Butene
/raws-2-Butene
Cyclohexane
Cyclopentane
Cyclopentene
w-Decane
1-Decene
/w-Diethylbenzene
/>-Diethylbenzene
2,2-Dimethylbutane
2,3-Dimethylbutane
2,3-Dimethylpentane
2,4-Dimethylpentane
w-Dodecane
1-Dodecene
Ethane
2-Ethyl-l-butene
Ethylbenzene
Ethylene
7w-Ethy toluene
o-Ethyltoluene
£>-Ethyltoluene
w-Heptane
1-Heptene
w-Hexane
1-Hexene
c/s-2-Hexene
/raws-2-Hexene
Isobutane
Isobutene/1 -Butene
Isopentane
Isoprene
Isopropylbenzene
2-Methyl-l-butene
3 -Methyl- 1 -butene
2-Methyl- 1 -pentene
BMCO
4.91
2.63
NA
0.13
NA
NA
1.43
2.08
NA
9.95
NA
NA
NA
NA
NA
2.43
NA
NA
NA
0.49
NA
5.66
2.68
NA
NA
NA
2.79
NA
1.10
NA
NA
NA
0.91
NA
1.66
NA
NA
NA
NA
NA
BRCO
0.77
1.66
NA
0.95
NA
NA
0.70
2.59
NA
3.06
NA
NA
NA
2.81
NA
4.14
2.51
NA
NA
0.26
NA
NA
0.59
NA
NA
NA
0.96
1.26
1.38
NA
NA
NA
0.67
NA
0.59
NA
NA
NA
NA
NA
BTUT
6.09
2.82
8.16
1.28
2.25
5.45
1.99
3.59
NA
5.86
NA
6.30
NA
5.85
2.24
2.87
4.58
NA
NA
2.16
NA
9.21
3.02
4.53
13.09
4.07
4.24
3.85
2.51
NA
NA
NA
0.88
NA
3.80
2.69
NA
5.34
NA
NA
NBIL
6.64
3.65
4.75
1.14
NA
1.56
3.59
5.53
NA
6.00
2.45
8.58
NA
6.56
1.90
3.10
3.51
3.00
13.47
1.80
NA
5.14
2.31
3.92
3.44
2.31
5.04
NA
2.65
NA
NA
NA
1.38
NA
1.25
3.55
NA
NA
NA
NA
SSSD
11.55
9.26
6.91
1.32
NA
7.16
8.48
7.63
NA
8.70
NA
9.53
27.12
14.29
10.12
6.12
9.10
NA
NA
3.91
NA
8.34
2.04
6.42
4.76
13.26
4.89
NA
4.85
NA
NA
NA
2.18
NA
6.23
9.87
NA
4.40
NA
NA
#of
Pairs
44
41
18
44
4
20
36
35
0
32
2
9
2
29
27
34
30
o
J
2
44
0
40
44
28
10
12
40
o
J
44
0
0
0
43
0
29
18
0
14
0
0
Average by
Pollutant
5.99
4.01
6.60
0.96
2.25
4.72
3.24
4.28
NA
6.71
2.45
8.14
27.12
7.38
4.75
3.73
4.92
3.00
13.47
1.72
NA
7.09
2.13
4.96
7.09
6.55
3.58
2.56
2.50
NA
NA
NA
1.20
NA
2.71
5.37
NA
4.87
NA
NA
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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
is calculated from the pollutant-specific averages and is provided in the final column of the
method
table.
30-31
-------
Table 30-10. SNMOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
4-Methyl-l-pentene
2-Methyl-2-butene
Methylcyclohexane
Methylcyclopentane
2-Methylheptane
3-Methylheptane
2-Methylhexane
3-Methylhexane
2-Methylpentane
3-Methylpentane
w-Nonane
1-Nonene
w-Octane
1-Octene
w-Pentane
1-Pentene
c/s-2-Pentene
/raws-2-Pentene
a-Pinene
6-Pinene
Propane
w-Propylbenzene
Propylene
Propyne
Styrene
Toluene
w-Tridecane
1-Tridecene
1 ,2,3 -Trimethylbenzene
1 ,2,4-Trimethylbenzene
1 ,3 ,5-Trimethylbenzene
2,2,3-Trimethylpentane
2,2,4-Trimethylpentane
2,3 ,4-Trimethylpentane
w-Undecane
1-Undecene
7w-Xylene//>-Xylene
o-Xylene
SNMOC (Sum of Knowns)
Sum of Unknowns
Average by Site
BMCO
NA
NA
3.12
0.93
9.80
2.77
2.07
3.00
1.10
1.69
4.33
NA
3.46
NA
0.59
NA
NA
NA
4.71
4.47
0.41
NA
3.53
NA
NA
2.13
NA
NA
NA
3.74
NA
NA
1.48
1.58
NA
NA
4.83
6.23
0.61
2.33
2.91
BRCO
NA
NA
1.50
0.00
2.42
4.22
2.24
0.00
1.00
0.82
2.76
NA
1.94
NA
0.74
NA
NA
NA
NA
1.65
0.38
NA
0.70
NA
NA
0.83
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
1.83
6.99
0.51
5.09
1.73
BTUT
NA
6.20
3.33
1.89
6.50
3.62
2.90
8.09
2.44
2.60
2.87
NA
2.98
NA
1.31
3.99
NA
3.00
4.66
2.71
1.09
10.16
2.21
NA
NA
3.10
NA
NA
6.33
3.53
5.62
NA
3.90
3.93
2.49
NA
2.99
3.21
1.36
18.43
4.32
NBIL
NA
1.08
8.35
5.51
6.59
4.23
3.86
7.63
2.86
4.34
5.56
NA
10.05
3.39
1.65
6.49
NA
5.42
4.21
2.09
0.98
4.31
3.44
NA
NA
5.28
NA
NA
3.18
3.92
11.12
NA
4.20
8.81
2.13
5.12
7.04
7.31
1.41
5.29
4.43
SSSD
NA
2.98
8.84
4.65
NA
13.79
7.80
4.39
2.75
6.23
8.13
12.79
8.82
NA
3.69
8.96
NA
8.64
26.57
3.96
2.09
NA
2.28
NA
NA
4.48
NA
NA
11.28
6.07
NA
NA
5.00
8.05
41.15
NA
5.67
6.91
24.57
6.19
8.61
#of
Pairs
0
14
38
44
20
21
44
29
43
44
35
8
37
2
44
32
0
27
13
18
44
6
44
2
0
44
0
0
13
39
10
0
41
32
8
2
44
44
44
44
-
Average by
Pollutant
NA
3.42
5.03
2.60
6.33
5.73
3.77
4.62
2.03
3.14
4.73
12.79
5.45
3.39
1.60
6.48
NA
5.69
10.04
2.98
0.99
7.23
2.43
0.94
NA
3.16
NA
NA
6.93
4.31
8.37
NA
3.64
5.59
15.26
5.12
4.47
6.13
5.69
7.46
5.29
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designaled NATTS Site
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
is calculated from the pollutant-specific averages and is provided in the final column of the
method
table.
30-32
-------
30.3.3 Carbonyl Compound Analytical Precision
Table 30-11 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 variability was 2.91 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 29.88 percent (crotonaldehyde for BRCO).
The pollutant-specific average CV ranged from 0.73 percent (acetone) to 5.69 percent
(tolualdehydes). The site-specific average CV ranged from 1.93 percent (BTUT) to 5.65 percent
(PACO).
Table 30-11. Carbonyl Compound Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant
Pollutant
Acetaldehyde
Acetone
Benzaldehyde
2-Butanone
Butyraldehyde
Crotonaldehyde
2,5-Dimethylbenzaldehyde
Formaldehyde
Hexaldehyde
Isovaleraldehyde
Propionaldehyde
Tolualdehydes
Valeraldehyde
Average by Site
ADOK
0.34
0.42
2.71
1.89
1.85
2.52
NA
0.84
3.10
NA
2.64
18.83
2.86
3.45
AZFL
1.47
1.85
3.52
2.31
3.66
1.29
NA
2.03
3.00
NA
3.07
7.57
4.15
3.08
BMCO
1.53
1.11
9.13
1.30
4.80
4.00
NA
1.50
0.00
NA
4.65
0.00
3.03
2.82
BRCO
0.60
0.61
0.00
0.83
2.48
29.88
NA
0.90
7.44
NA
1.89
0.00
0.00
4.06
BTUT
0.47
0.38
2.30
1.24
1.00
1.55
NA
0.32
4.60
NA
2.06
5.09
2.19
1.93
BURVT
1.10
0.95
2.51
3.21
2.43
2.56
NA
1.00
2.80
NA
2.73
3.57
3.22
2.37
CHNJ
0.63
0.39
5.54
2.28
2.96
4.55
NA
0.77
5.17
NA
3.46
5.09
4.21
3.19
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designaled NATTS Site
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
is calculated from the pollutant-specific averages and is provided in the final column of the
method
table.
30-33
-------
Table 30-11. Carbonyl Compound Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
Acetaldehyde
Acetone
Benzaldehyde
2-Butanone
Butyraldehyde
Crotonaldehyde
2,5-Dimethylbenzaldehyde
Formaldehyde
Hexaldehyde
Isovaleraldehyde
Propionaldehyde
Tolualdehydes
Valeraldehyde
Average by Site
DEMI
0.62
0.56
2.49
2.29
2.57
3.35
NA
1.27
3.52
NA
1.71
4.36
4.52
2.48
ELNJ
0.55
0.38
2.81
1.28
3.58
2.69
NA
0.79
4.55
NA
2.66
3.81
2.73
2.35
GLKY
0.70
0.78
3.17
3.53
3.42
1.03
NA
0.88
4.41
NA
1.96
6.41
2.06
2.58
GPCO
0.59
0.35
1.05
2.13
1.66
3.39
NA
0.52
4.41
NA
2.14
4.43
3.28
2.18
INDEM
1.48
0.49
3.87
1.56
3.99
3.77
NA
1.26
4.24
NA
2.65
4.17
3.90
2.85
NBIL
1.12
0.92
3.29
2.73
3.08
4.32
NA
1.87
4.16
NA
2.58
3.55
4.01
2.88
NBNJ
0.66
0.53
4.51
1.67
3.17
1.59
NA
1.20
3.95
NA
2.57
5.08
4.06
2.64
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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
is calculated from the pollutant-specific averages and is provided in the final column of the
method
table.
Table 30-11. Carbonyl Compound Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
Acetaldehyde
Acetone
Benzaldehyde
2-Butanone
Butyraldehyde
Crotonaldehyde
2,5-Dimethylbenzaldehyde
Formaldehyde
Hexaldehyde
Isovaleraldehyde
Propionaldehyde
Tolualdehydes
Valeraldehyde
Average by Site
OCOK
2.59
2.64
3.13
2.74
2.96
1.89
NA
2.63
4.16
NA
3.30
4.36
3.41
3.07
ORFL
0.45
1.43
2.64
2.95
3.81
3.11
NA
0.74
3.42
NA
2.68
4.78
3.80
2.71
PACO
0.00
0.00
6.73
4.26
12.86
NA
NA
0.22
8.32
NA
0.00
NA
18.45
5.65
PROK
0.68
0.43
2.32
1.90
3.53
2.10
NA
0.70
4.59
NA
3.75
3.81
4.29
2.56
PXSS
0.51
0.60
2.45
2.07
1.72
1.91
NA
0.90
3.26
NA
2.57
6.83
3.47
2.39
RICO
0.99
0.00
3.01
5.67
10.88
NA
NA
0.27
4.04
NA
5.98
18.45
0.00
4.93
ROIL
0.47
0.42
2.12
4.36
3.22
1.38
NA
1.19
5.22
NA
4.18
4.66
7.74
3.18
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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
is calculated from the pollutant-specific averages and is provided in the final column of the
method
table.
30-34
-------
Table 30-11. Carbonyl Compound Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
Acetaldehyde
Acetone
Benzaldehyde
2-Butanone
Butyraldehyde
Crotonaldehyde
2,5-Dimethylbenzaldehyde
Formaldehyde
Hexaldehyde
Isovaleraldehyde
Propionaldehyde
Tolualdehydes
Valeraldehyde
Average by Site
RRMI
0.47
0.45
2.37
0.52
2.82
3.70
NA
0.68
3.33
NA
2.17
2.80
3.23
2.05
S4MO
0.41
0.78
2.15
1.32
2.53
2.10
NA
0.86
3.72
NA
2.67
9.59
2.61
2.61
SEWA
1.57
0.60
3.22
1.06
3.25
3.84
NA
2.22
4.72
NA
3.06
4.68
3.72
2.90
SKFL
1.35
0.66
4.76
4.04
3.16
2.92
NA
1.76
3.85
NA
2.24
5.48
2.54
2.98
SPIL
1.05
0.65
3.44
1.66
2.15
3.26
NA
0.63
5.21
NA
3.53
4.83
5.26
2.88
SSSD
0.47
0.52
3.56
2.22
2.56
3.44
NA
1.20
4.24
NA
2.32
3.13
2.74
2.40
SWMI
0.00
0.00
1.89
3.63
4.60
0.00
NA
0.69
3.01
NA
4.68
4.56
3.45
2.41
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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
is calculated from the pollutant-specific averages and is provided in the final column of the
method
table.
Table 30-11. Carbonyl Compound Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
Acetaldehyde
Acetone
Benzaldehyde
2-Butanone
Butyraldehyde
Crotonaldehyde
2,5-Dimethylbenzaldehyde
Formaldehyde
Hexaldehyde
Isovaleraldehyde
Propionaldehyde
Tolualdehydes
Valeraldehyde
Average by Site
SYFL
0.52
0.99
4.05
3.90
3.12
2.00
NA
1.62
4.88
NA
1.90
11.31
3.48
3.43
TMOK
1.01
1.28
3.60
2.37
2.28
2.49
NA
0.91
3.97
NA
3.69
2.22
3.90
2.52
TOOK
0.49
0.40
2.93
1.82
2.65
2.84
NA
0.46
5.68
NA
3.42
5.07
3.42
2.65
WPIN
1.70
1.65
3.74
2.51
3.27
2.02
NA
1.44
4.24
NA
3.88
7.84
4.09
3.31
#of
Pairs
335
335
327
315
335
329
0
335
327
0
335
286
333
~
Average by
Pollutant
0.83
0.73
3.28
2.41
3.50
3.52
NA
1.07
4.23
NA
2.90
5.69
3.87
2.91
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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
is calculated from the pollutant-specific averages and is provided in the final column of the
method
table.
30-35
-------
30.3.4 PAH Analytical Precision
Table 30-12 presents analytical precision results from replicate analyses of collocated and
select individual PAH 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 analytical
precision results exhibit low- to mid-level variability, ranging from a CV of 0 percent (several
pollutants at different sites) to 36.70 percent (pyrene for SDGA, although fluoranthene had a
similar CV). The pollutant-specific average CV ranged from 1.56 percent (perylene) to
6.92 percent (acenaphthylene). The site-specific average CV ranged from 1.67 percent (UNVT)
to 10.63 percent (SDGA). The overall average CV was 3.93 percent.
Table 30-12. PAH Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant
Pollutant
Acenaphthene
Acenaphthylene
Anthracene
Benzo(a)anthracene
Benzo(a)pyrene
Benzo(b)fluoranthene
Benzo(e)pyrene
Benzo(g,h,i)perylene
Benzo(k)fluoranthene
Chrysene
Coronene
Cyclopenta[cd]pyrene
Dibenz(a,h)anthracene
Fluoranthene
Fluorene
9-Fluorenone
Indeno( 1,2,3 -cd)pyrene
Naphthalene
Perylene
Phenanthrene
Pyrene
Retene
Average by Site
BOMA
1.60
21.41
0.77
1.99
1.90
0.34
2.62
4.14
4.01
1.31
NA
NA
NA
0.17
0.41
0.99
4.60
3.54
NA
0.98
2.64
NA
3.14
BTUT
5.42
3.84
4.43
NA
NA
4.89
NA
NA
NA
4.29
NA
NA
NA
0.40
NA
0.60
NA
10.99
NA
1.18
0.74
0.81
3.42
BXNY
2.37
3.16
2.57
6.13
2.56
2.62
3.64
1.71
NA
0.75
1.42
NA
NA
1.41
1.27
0.68
1.72
6.90
NA
0.81
1.95
NA
2.45
CELA
1.14
0.37
1.90
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
0.62
2.57
1.96
NA
11.40
NA
1.99
0.73
4.88
2.75
CHSC
4.23
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
1.19
0.00
3.04
NA
2.53
NA
1.41
0.27
5.01
2.21
DEMI
2.81
3.61
3.28
2.73
4.08
2.42
4.06
2.08
8.81
1.38
3.43
NA
4.04
2.85
2.62
2.37
3.41
2.09
1.90
1.67
2.99
3.86
3.17
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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
is calculated from the pollutant-specific averages and is provided in the final column of the
method
table.
30-36
-------
Table 30-12. PAH Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
Acenaphthene
Acenaphthylene
Anthracene
Benzo(a)anthracene
Benzo(a)pyrene
Benzo(b)fluoranthene
Benzo(e)pyrene
Benzo(g,h,i)perylene
Benzo(k)fluoranthene
Chrysene
Coronene
Cyclopenta[cd]pyrene
Dibenz(a,h)anthracene
Fluoranthene
Fluorene
9-Fluorenone
Indeno( 1 ,2,3 -cd)pyrene
Naphthalene
Perylene
Phenanthrene
Pyrene
Retene
Average by Site
GLKY
4.46
1.01
4.66
1.23
1.99
3.67
2.85
3.42
3.03
5.18
NA
NA
NA
1.57
1.74
1.72
2.88
2.67
NA
1.04
2.68
3.87
2.76
LBHCA
3.82
NA
NA
NA
NA
NA
NA
NA
NA
4.71
NA
NA
NA
5.49
2.21
3.67
NA
5.33
NA
0.40
9.67
NA
4.41
MONY
4.60
8.87
4.31
NA
NA
NA
NA
NA
NA
3.50
NA
NA
NA
4.65
3.61
4.98
9.08
2.73
NA
3.35
5.60
NA
5.00
PRRI
1.91
11.21
1.50
NA
NA
NA
NA
NA
NA
1.39
NA
NA
NA
2.53
1.52
1.99
6.43
3.11
NA
0.67
2.62
7.24
3.66
PXSS
1.91
NA
5.44
NA
NA
NA
NA
NA
NA
5.46
NA
NA
NA
0.48
0.40
1.50
NA
1.05
NA
1.22
0.67
0.40
1.85
RIVA
1.21
5.00
6.23
1.78
10.10
0.94
1.66
2.15
8.71
5.24
4.63
NA
NA
2.91
1.92
2.57
0.33
0.47
NA
0.87
2.84
4.07
3.35
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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
is calculated from the pollutant-specific averages and is provided in the final column of the
method
table.
30-37
-------
Table 30-12. PAH Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
Acenaphthene
Acenaphthylene
Anthracene
Benzo(a)anthracene
Benzo(a)pyrene
Benzo(b)fluoranthene
Benzo(e)pyrene
Benzo(g,h,i)perylene
Benzo(k)fluoranthene
Chrysene
Coronene
Cyclopenta[cd]pyrene
Dibenz(a,h)anthracene
Fluoranthene
Fluorene
9-Fluorenone
Indeno( 1,2,3 -cd)pyrene
Naphthalene
Perylene
Phenanthrene
Pyrene
Retene
Average by Site
ROCH
2.08
NA
3.15
1.94
2.02
6.37
1.69
2.55
0.99
3.55
NA
NA
NA
3.87
2.11
2.85
8.57
1.93
NA
1.15
4.89
8.37
3.42
RUCA
5.25
12.79
10.55
6.54
4.92
7.65
4.48
2.30
1.96
5.41
0.46
NA
NA
2.85
3.30
3.87
3.40
12.77
NA
1.91
3.51
6.59
5.29
S4MO
2.83
2.96
4.33
1.48
7.42
3.92
1.41
4.80
2.50
2.31
2.68
6.31
4.14
1.51
3.22
1.45
5.38
4.75
1.10
1.04
1.33
1.45
3.11
SDGA
10.79
NA
12.09
3.84
3.11
3.92
2.24
3.32
3.70
5.06
4.86
NA
3.51
36.68
17.20
17.58
6.87
2.74
1.54
30.09
36.70
6.77
10.63
SEWA
3.30
10.69
11.83
1.47
3.06
4.44
4.40
5.24
7.76
3.62
5.35
NA
4.97
3.35
3.46
3.76
1.40
2.28
1.71
1.56
3.65
3.27
4.31
SJJCA
4.97
NA
8.71
NA
NA
NA
NA
NA
NA
9.56
NA
NA
NA
3.15
2.58
7.88
NA
1.18
NA
2.42
1.31
NA
4.64
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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
is calculated from the pollutant-specific averages and is provided in the final column of the
method
table.
30-38
-------
Table 30-12. PAH Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
Acenaphthene
Acenaphthylene
Anthracene
Benzo(a)anthracene
Benzo(a)pyrene
Benzo(b)fluoranthene
Benzo(e)pyrene
Benzo(g,h,i)perylene
Benzo(k)fluoranthene
Chrysene
Coronene
Cyclopenta[cd]pyrene
Dibenz(a,h)anthracene
Fluoranthene
Fluorene
9-Fluorenone
Indeno(l,2,3-cd)pyrene
Naphthalene
Perylene
Phenanthrene
Pyrene
Retene
Average by Site
SKFL
2.88
7.79
7.57
6.55
3.71
1.80
4.91
3.96
1.71
0.61
NA
NA
NA
1.75
0.79
2.41
3.66
0.88
NA
1.17
2.34
3.26
3.21
SYFL
2.01
4.22
7.31
NA
NA
3.12
1.77
3.16
2.10
5.27
NA
NA
NA
2.50
3.28
2.55
3.93
3.38
NA
1.23
3.29
3.18
3.41
UNVT
6.15
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
2.52
0.00
0.39
NA
0.68
NA
0.27
NA
NA
1.67
WADC
1.01
NA
11.21
NA
NA
0.00
NA
NA
NA
5.03
NA
NA
NA
0.88
4.56
1.23
NA
0.63
NA
0.50
0.00
NA
2.50
#of
Pairs
107
45
88
41
31
78
43
53
27
98
19
1
7
110
109
110
43
110
6
110
109
66
-
Average by
Pollutant
3.49
6.92
5.89
3.76
4.06
3.18
2.85
3.90
4.12
3.88
3.26
6.31
4.17
3.79
2.80
3.18
4.40
3.82
1.56
2.59
4.31
4.20
3.93
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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
is calculated from the pollutant-specific averages and is provided in the final column of the
method
table.
30.3.5 Metals Analytical Precision
Table 30-13 presents analytical precision results from replicate analyses of collocated and
select individual metals 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 results from
replicate analyses exhibit low- to mid-level variability among sites, ranging from a CV of
0 percent (beryllium for UNVT) to 28.61 percent (arsenic for UNVT). The pollutant-specific
average CV ranged from 1.59 percent (antimony) to 9.19 percent (beryllium). The site-specific
average CV ranged from 2.64 percent (SEWA) to 8.31 percent (UNVT). The overall average
analytical precision was 4.94 percent.
30-39
-------
Table 30-13. Metals Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant
Pollutant
Antimony
Arsenic
Beryllium
Cadmium
Chromium
Cobalt
Lead
Manganese
Mercury
Nickel
Selenium
Average by Site
ADOK
3.09
4.08
3.45
2.09
0.94
1.29
25.08
3.35
6.73
7.32
2.04
5.41
ASKY-M
0.95
9.57
8.70
3.37
NA
3.36
1.25
0.58
9.62
1.51
10.65
4.96
BAKY
0.61
4.53
28.28
3.13
NA
3.72
1.02
0.69
NA
1.53
10.52
6.00
BOMA
1.41
2.56
7.55
5.70
2.83
5.14
1.42
1.50
11.04
3.51
3.17
4.17
BTUT
0.85
13.63
11.15
3.86
NA
5.43
0.48
0.61
NA
1.14
NA
4.64
CCKY
1.19
11.93
NA
6.14
NA
8.31
0.58
1.75
NA
1.66
3.44
4.37
GLKY
1.70
13.68
NA
9.67
NA
20.15
0.75
1.28
NA
1.66
7.46
7.04
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designaled NATTS Site
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
is calculated from the pollutant-specific averages and is provided in the final column of the
method
table.
Table 30-13. Metals Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
Antimony
Arsenic
Beryllium
Cadmium
Chromium
Cobalt
Lead
Manganese
Mercury
Nickel
Selenium
Average by Site
LEKY
0.75
8.11
11.11
3.27
NA
1.60
3.09
0.37
NA
1.60
11.14
4.56
NBIL
1.31
1.50
10.63
1.63
1.98
4.98
2.38
1.29
4.20
3.02
1.43
3.12
OCOK
3.57
2.51
3.61
1.57
3.29
4.17
5.91
3.13
14.59
3.35
2.66
4.40
PAFL
2.97
3.15
7.69
4.18
NA
NA
1.62
2.28
5.26
2.53
4.14
3.76
PROK
1.82
3.88
4.34
9.83
0.57
5.04
2.18
3.01
2.21
1.82
3.40
3.46
PXSS
2.79
14.26
12.98
6.81
NA
4.61
3.09
2.37
NA
3.97
NA
6.36
S4MO
1.16
9.16
16.67
4.49
NA
18.92
0.97
0.92
7.14
10.42
6.26
7.61
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designated NATTS Site
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
is calculated from the pollutant-specific averages and is provided in the final column of the
method
table.
30-40
-------
Table 30-13. Metals Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
Antimony
Arsenic
Beryllium
Cadmium
Chromium
Cobalt
Lead
Manganese
Mercury
Nickel
Selenium
Average by Site
SEWA
0.45
1.16
NA
3.34
0.25
1.06
1.73
1.84
10.90
1.22
4.43
2.64
SJJCA
1.38
4.60
NA
7.66
NA
23.86
0.66
0.67
NA
1.37
18.74
7.37
TMOK
1.99
1.66
7.44
12.28
2.47
2.05
2.10
3.31
2.77
10.54
1.60
4.38
TOOK
1.46
2.48
4.22
3.37
1.64
3.49
1.94
1.66
6.44
2.65
2.63
2.91
UNVT
0.84
28.61
0.00
19.60
NA
12.65
2.61
1.01
NA
1.19
NA
8.31
#of
Pairs
425
420
224
436
140
363
441
441
222
373
375
-
Average by
Pollutant
1.59
7.42
9.19
5.89
1.75
7.21
3.10
1.66
7.36
3.26
5.86
4.94
NA = No pairs with concentrations greater than or equal to the MDL.
BOLD ITALICS = EPA-designaled NATTS Site
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
is calculated from the pollutant-specific averages and is provided in the final column of the
method
table.
30.3.6 Hexavalent Chromium Analytical Precision
Table 30-14 presents analytical precision results from replicate analyses of collocated and
select individual hexavalent chromium samples as the CV per site and the overall average CV for
hexavalent chromium. The range of variability for hexavalent chromium was 0.46 percent
(STMN) to 11.97 percent (GLKY), with an overall average analytical precision of 6.65 percent.
30-41
-------
Table 30-14. Hexavalent Chromium Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site
Site
BOMA
BTUT
BXNY
CAMS 35
CAMS 85
CHSC
DEMI
GLKY
GPCO
HOW
MIWI
MONY
NBIL
PRRI
PXSS
RIVA
ROCH
S4MO
SDGA
SEWA
SKFL
STMN
SYFL
UNVT
WADC
# of Pairs
Average by Site
Average CV
(%)
5.03
8.84
10.50
3.46
4.34
8.01
4.77
11.97
10.69
10.03
8.27
4.66
5.75
7.51
4.29
7.99
4.01
6.13
5.60
4.70
9.27
0.46
5.64
7.30
7.06
223
6.65
BOLD ITALICS = EPA-designated
NATTS Site
Orange shading indicates the average CV
for this method.
30-42
-------
30.4 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. However, due to a change in the external
source for the NATTS program in 2012, PT samples were not supplied for all methods within the
calendar year. For these methods, internal audit samples were prepared by task leaders and
analyzed by a separate analyst. Thus, Tables 30-15 through 30-20 present ERG's results for both
internal and external audit samples. Results for internal audit samples are provided for VOCs and
carbonyl compounds. Results for externally prepared NATTS PT audit samples are provided for
PAHs and metals. Results for both internal and external audit samples are provided for
hexavalent chromium. The program MQO for the percent difference from the true value is
± 25 percent, and the values exceeding this criterion are bolded in the tables. The percent
difference calculation is:
X —X
Percent Difference = — x 100
X2
Where:
Xi is the analytical result from the laboratory;
Xi is the true concentration of the audit sample
Note that for the PAH results in Table 30-17, the difference from the "true" value is based on the
mean value of all the participating laboratories' results rather than the "true" value from the
external source.
The results of the audit samples show that few of the pollutants for which audit samples
were analyzed exceed the MQO for accuracy. Of the 90 results provided in Tables 30-15 through
Table 30-20, only two exceed ± 25 percent. The two that exceeded ± 25 percent
30-43
-------
(1,2,4-trichlorobenzene and acetone) were internal audit samples. More than 60 percent of the
results were less than 10 percent different from the true value.
Table 30-15. VOC Internal PT Audit Samples-Percent Difference
from True Value
Pollutant
Acetonitrile
Acetylene
Acrolein
Acrylonitrile
fer/-Amyl Methyl Ether
Benzene
Bromochloromethane
Bromodichloromethane
Bromoform
Bromomethane
1,3 -Butadiene
Carbon Bisulfide
Carbon Tetrachloride
Chlorobenzene
Chloroethane
Chloroform
Chloromethane
Chloroprene
Dibromochloromethane
1 ,2-Dibromoethane
w-Dichlorobenzene
o-Dichlorobenzene
p-Dichlorobenzene
Dichlorodifluoromethane
c/5-l,2-Dichloroethylene
trans- 1 ,2-Dichloroethylene
1 , 1 -Dichloroethane
1 , 1 -Dichloroethene
1 ,2-Dichloroethane
Dichloromethane
1 ,2-Dichloropropane
cis- 1 ,3 -Dichloropropene
trans- 1 ,3 -Dichloropropene
Dichlorotetrafluoroethane
Ethyl Aery late
Ethyl fer/-Butyl Ether
Ethylbenzene
Hexachloro- 1 ,3 -butadiene
October 2012
-13.8
-23.6
-4.6
-13.2
-14.2
0.6
-1.1
5.5
-7.3
-2.2
-12.0
-3.5
1.5
-2.7
-3.0
-1.3
-3.4
-9.4
-3.7
-5.2
-8.5
-10.5
-10.2
-4.3
-1.3
-4.6
-2.9
-3.2
-3.2
-1.8
3.0
-8.0
-15.2
-0.8
-15.0
-20.3
-13.3
-16.2
30-44
-------
Table 30-15. VOC Internal PT Audit Samples-Percent Difference
from True Value (Continued)
Pollutant
Methyl Isobutyl Ketone
Methyl Methacrylate
Methyl fer/-Butyl Ether
rc-Octane
Propylene
Styrene
retrachloroethylene
Toluene
1 ,2,4-Trichlorobenzene
1,1,1 -Trichloroethane
1 , 1 ,2-Trichloroethane
Trichloroethylene
rrichlorofluoromethane
rrichlorotrifluoroethane
1 ,2,4-Trimethylbenzene
1,3,5 -Trimethy Ibenzene
1 , 1 ,2,2-Tetrachloroethane
Vinyl Chloride
w,/>-Xylene
o-Xylene
October 2012
-12.2
-13.4
-17.9
-12.8
-8.3
-17.0
0.3
-10.1
-28.2
-1.1
4.7
5.9
-2.2
-4.6
-22.3
-21.2
-5.4
-2.2
-15.5
-13.7
Table 30-16. Carbonyl Compound Internal PT Audit Sample-Percent Difference
from True Value
Pollutant
Acetaldehyde
Acetone
Benzaldehyde
2-Butanone
Butyraldehyde
Crotonaldehyde
2,5-Dimethylbenzaldehyde
Formaldehyde
Hexaldehyde
Isovaleraldehyde
Propionaldehyde
Tolualdehydes
Valeraldehyde
September
2012
-9.4
27.7
4.9
-5.6
-3.5
-9.5
-16.9
-6.9
-21.4
-20.8
-22.5
-9.7
-9.3
30-45
-------
Table 30-17. PAH NATTS PT Audit Samples-Percent Difference
from Mean
Pollutant
Acenaphthene
Anthracene
Benzo(a)pyrene
Fluoranthene
Fluorene
Naphthalene
Phenanthrene
Pyrene
February 2012
9.5
3.9
0.2
-1.3
5.1
15.5
-1.9
-1.1
Table 30-18. Metals NATTS PT Audit Samples-Percent Difference
from True Value
Pollutant
Arsenic
Beryllium
Cadmium
Cobalt
Lead
Manganese
Nickel
Selenium
February 2012
9.1
8.9
10.4
13.0
6.1
11.2
12.2
14.3
Table 30-19. Hexavalent Chromium NATTS PT Audit Samples-Percent Difference
from True Value
Pollutant
Hexavalent Chromium
February 2012
2.7
Table 30-20. Hexavalent Chromium Internal PT Audit Samples-Percent Difference
from True Value
Pollutant
Hexavalent Chromium
December 2012
Concentration #1
8.2
Concentration #2
15.2
In mid-2012, ERG's use of the ICP/MS to analyze speciated metals, in particular lead,
was approved as a Federal Equivalent Method (FEM) for the sampling and analysis of lead for
adherence to the National Ambient Air Quality Standards (NAAQS) (EPA 2012a). This approval
30-46
-------
requires additional quality assurance steps, including the analysis of quarterly audit strips. Table
30-21 provides the results of the quarterly NAAQS audit results for lead for ERG. All results are
within the ± 25 percent MQO.
Table 30-21. NAAQS Lead PT Audit Samples-Percent Difference
from True Value for Multiple Concentrations
Pollutant
Lead
Lead
Lead
June 2012
Concentration
#1
-14.9
-13.6
-12.4
Concentration
#2
-5.7
-6.4
-5.6
September 2012
Concentration
#1
-10.6
-4.8
-6.3
Concentration
#2
-4.8
-6.0
-5.2
December 2012
Concentration
#1
-12.5
-7.6
-10.9
Concentration
#2
-14.5
-8.5
-11.2
The accuracy of the 2012 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 in the 2012 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, all 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 2012 monitoring data accurately represent ambient air quality.
30-47
-------
31.0 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 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 and should ultimately
assist a wide range of audiences understand the complex nature of air pollution.
31.1 Summary of Results
Analyses of the 2012 monitoring data identified the following notable results,
observations, trends, and patterns in the program-level and state- and site-specific air monitoring
data.
31.1.1 National-level Results Summary
• Number of participating sites. Twenty-six of the 64 monitoring sites are EPA-
designated NATTS sites (BOMA, BTUT, BXNY, CAMS 35, CAMS 85, CELA,
CHSC, DEMI, GLKY, GPCO, HOWI, MONY, NBIL, PRRI, PXSS, RIVA, ROCH,
RUCA, S4MO, SDGA, SEW A, SJJCA, SKFL, SYFL, UNVT, and WADC). Thirty-
seven UATMP sites participated in 2012. Data from one CSATAM site (LBHCA) are
included in the 2012 NMP report.
• Total number of samples collected and analyzed. Over 9,600 samples were collected
yielding 233,600 valid measurements of air toxics.
• Detects. 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 53 percent of the reported measurements were
greater than the associated MDLs. Each of the 171 pollutants monitored were
detected at least once over the course of the 2012 monitoring effort. Quantification
below the MDL is possible; these results are considered valid measurements and are
therefore incorporated into the data analyses. These measurements account for
11 percent of concentrations. Non-detects account for the remaining 36 percent of
results.
• Program-level Pollutants of Interest. The pollutants of interest at the program-level
are based on the number of exceedances, or "failures," of the risk screening values.
Thirty-eight pollutants failed at least one risk screening value; of those pollutants, 15
were identified as program-level pollutants of interest.
31-1
-------
• Noncancer Risk-Based Screening using A TSDR MRLs. All of the preprocessed daily
measurements for which an MRL is available were less than the associated ATSDR
acute MRLs. Additionally, all of the quarterly or annual average concentrations of the
pollutants with available MRLs were less than the associated ATSDR intermediate or
chronic MRLs.
• Mobile Sources. Site-specific hydrocarbon concentrations had positive correlations
with county-level motor vehicle ownership data, traffic data, and VMT data. While
these correlations were not "strong", they do indicate that hydrocarbon concentrations
tend to increase with increasing motor vehicle activity data.
• Seasonal Trends. Formaldehyde concentrations tended to be highest during the third
quarter of 2012, or during the period from July to September. Acenaphthene,
acetaldehyde, and fluorene concentrations exhibit a similar pattern. Conversely,
benzene concentrations tended to be higher during the first or fourth quarters of the
year, or between January through March and October through December.
Concentrations of 1,3-butadiene tended to be higher during the fourth quarter of the
year, or from October to December. Arsenic concentrations for TSP metals tended to
be highest during the second quarter of 2012, during the period from April to June
(note however, that all of the sites monitoring TSP metals are located in Oklahoma).
31.1.2 State-level Results Summary
Arizona.
• The Arizona monitoring sites are located in Phoenix. PXSS is a NATTS site; SPAZ is
a UATMP site.
• PXSS sampled for VOCs, carbonyl compounds, PAHs, metals (PMio), and
hexavalent chromium. SPAZ sampled for VOCs only.
• Twenty-one pollutants failed screens for PXSS, 12 of which contributed to 95 percent
of failed screens. PXSS failed the second highest number of screens among all NMP
sites. Seven pollutants failed screens for SPAZ, six of which contributed to 95 percent
of failed screens.
• Of the pollutants of interest for PXSS, formaldehyde had 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 had the highest annual average concentration for SPAZ, and was the only
pollutant with an annual average concentration greater than 1 |ig/m3.
• PXSS had the highest annual average concentration of acetaldehyde and second
highest annual average concentrations of l,3-butadiene,/>-dichlorobenzene, and
manganese among NMP sites sampling these pollutants.
31-2
-------
• SPAZ had the highest annual average concentrations of 1,3-butadiene and
/>-dichlorobenzene and the second highest annual average concentrations of benzene
and ethylbenzene among NMP sites sampling these pollutants.
• 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. The most significant changes in recent years
are in the nickel concentrations measured at PXSS, which have been increasing over
the last few years. The detection rate of 1,2-dichloroethane at both sites has been
steadily increasing over the years, with a significant increase for 2012 at both sites.
• Formaldehyde and benzene had the highest cancer risk approximations for PXSS and
benzene had the highest cancer risk approximation for SPAZ. These are the only
pollutants of interest with cancer risk approximations greater than 10 in-a-million.
None of the pollutants of interest for either site had a noncancer hazard
approximation greater than 1.0.
• Benzene was the highest emitted pollutant with a cancer toxicity factor in Maricopa
County, while toluene was the highest emitted pollutant with a noncancer toxicity
factor. Formaldehyde had the highest cancer toxicity-weighted emissions, while
acrolein had the highest noncancer toxicity-weighted emissions for Maricopa County.
California.
• The four California monitoring sites are located in Los Angeles (CELA), Long Beach
(LBHCA), Rubidoux (RUCA), and San Jose (SJJCA). CELA, RUCA, and SJJCA are
NATTS sites; LBHCA is a CSATAM site.
• CELA, LBHCA, and RUCA sampled for PAHs only. SJJCA sampled for PAHs and
metals (PMio).
• Naphthalene failed the majority of screens at CELA, LBHCA, and RUCA.
Naphthalene and arsenic contributed almost equally to the total number of failed
screens at SJJCA, together accounting for nearly 70 percent of failed screens at the
site.
• Naphthalene had the highest annual average concentration for each site. The annual
average concentration of naphthalene for CELA was the second highest compared to
NMP sites sampling PAHs. Annual average concentrations could not be calculated
for LBHCA because sampling did not begin until July 2012.
• 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 and
fluorene increased at CELA for 2012; naphthalene concentrations have been
31-2
-------
increasing at RUCA as well. Concentrations of manganese and nickel measured at
SJJCA increased significantly between 2010 and 2011.
• Of the pollutants of interest for each site, naphthalene exhibited the highest cancer
risk approximation for all three California sites. The noncancer hazard approximation
for each pollutant of interest was considerably less than 1.0 for all three sites.
• Formaldehyde was the highest emitted pollutant with a cancer toxicity factor in Los
Angeles, Riverside, and Santa Clara Counties; formaldehyde also had the highest
cancer toxicity-weighted emissions for Los Angeles and Riverside Counties while
POM, Group la had the highest cancer toxicity-weighted emissions for Santa Clara
County.
• Toluene was the highest emitted pollutant with a noncancer toxicity factor in Los
Angeles, Riverside, and Santa Clara Counties, while acrolein had the highest
noncancer toxicity-weighted emissions for all three counties.
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).
• GPCO sampled for VOCs, carbonyl compounds, PAHs, and hexavalent chromium.
The Garfield County sites sampled for SNMOCs and carbonyl compounds.
• Nineteen pollutants failed at least one screen for GPCO, 13 of which contributed to
95 percent of failed screens. The number of pollutants that failed screens for the
Garfield County sites ranged from four (BRCO and RFCO) to five (BMCO, PACO,
and RICO).
• Of the pollutants of interest for GPCO, dichloromethane had the highest annual
average concentration, which was an order of magnitude higher than the next highest
annual average concentration.
• Formaldehyde had the highest annual average concentration for three of the Garfield
County sites (BRCO, PACO, and RICO). Although benzene had the highest annual
average for BMCO, it should be noted that annual averages could not be calculated
for the carbonyl compounds for this site. Annual average concentrations could not be
calculated for RFCO because sampling did not begin until June 2012.
• GPCO had the highest annual average concentrations of naphthalene, acenaphthene,
and fluorene among NMP sites sampling PAHs. GPCO also had the second highest
annual average concentrations of 1,2-dichloroethane and acetaldehyde among all
NMP sites sampling these pollutants.
31-4
-------
• 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. Benzene concentrations at GPCO
have an overall decreasing trend across the years of sampling, as do benzene
concentrations measured at BRCO and, in more recent years, RICO. Concentrations
ofp-dichlorobenzene have an increasing trend at GPCO. The range of concentrations
of naphthalene, fluorene, and acenaphthene measured at GPCO exhibit significant
increases for 2012. In addition, the detection rate of 1,2-dichloroethane at GPCO has
been increasing steadily over the last few years of sampling, particularly for 2012.
• Formaldehyde had the highest cancer risk approximation for each of the Colorado
sites, where an annual average could be calculated. All noncancer hazard
approximations were less than 1.0 for all five Colorado sites, where noncancer risk
approximations could be calculated.
• Benzene was the highest emitted pollutant with a cancer toxicity factor in Mesa
County, while formaldehyde was the highest emitted pollutant with a cancer toxicity
factor in Garfield County. Formaldehyde had the highest cancer toxicity-weighted
emissions for both counties.
• While toluene was the highest emitted pollutant with a noncancer toxicity factor for
both Mesa and Garfield Counties, acrolein had the highest noncancer toxicity -
emissions.
District of Columbia.
• The Washington, D.C. monitoring site (WADC) is a NATTS site.
• WADC sampled for hexavalent chromium and PAHs. Naphthalene accounted for
over 97 percent of failed screens for this site and was the only pollutant identified as a
pollutant of interest.
• The annual average concentration of naphthalene for WADC was fifth 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 not change significantly
since the onset of PAH sampling at WADC.
• The cancer risk approximation for naphthalene was 3.55 in-a-million. The noncancer
risk approximation for naphthalene was considerably less than 1.0.
31-5
-------
• Benzene was the highest emitted pollutant with a cancer toxicity factor in the District
of Columbia, while toluene was the highest emitted pollutant with a noncancer
toxicity factor. Formaldehyde had the highest cancer toxicity-weighted emissions,
while acrolein had the highest noncancer toxicity-weighted emissions in the District.
Florida.
• Three of the Florida monitoring sites are located in the Tampa-St. Petersburg-
Clearwater MSA (SYFL, AZFL, and SKFL) and two are located in the Orlando-
Kissimmee-Sanford MSA (ORFL and PAFL). SKFL and SYFL are NATTS sites
while the other three are UATMP sites.
• AZFL and ORFL sampled for carbonyl compounds only. SKFL and SYFL sampled
for hexavalent chromium and PAHs in addition to carbonyl compounds. PAFL
sampled for only metals (PMio).
• Acetaldehyde and formaldehyde were the only pollutants to fail screens for AZFL
and ORFL, where only carbonyl compounds were sampled. Eight pollutants failed
screens for SKFL and five pollutants failed screens for SYFL. Arsenic, manganese
and lead failed screens for PAFL, where only metals were sampled.
• Formaldehyde had the highest annual average concentrations for each of the Florida
sites where carbonyl compounds were sampled. The annual average concentration of
naphthalene for SKFL was more than twice the annual average concentration of
naphthalene for SYFL and ranked eighth highest among NMP sites sampling this
pollutant. The annual average concentration of arsenic for PAFL is the third highest
among NMP sites sampling PMio metals.
• Sampling for the site-specific pollutants of interest has occurred at the Florida sites
for at least 5 consecutive years; thus, a trends analysis was conducted for the site-
specific pollutants of interest. Concentrations of formaldehyde have an overall
decreasing trend at ORFL. A similar trend in formaldehyde concentrations is shown
at SKFL until recent years where an increasing trend is shown. Concentrations of
acetaldehyde decreased significantly between 2010 and 2012 at AZFL and SKFL
with a significant decrease also shown at ORFL from 2011 and 2012. Acetaldehyde
concentrations at SYFL increased significantly from 2011 to 2012. Concentrations of
naphthalene have not changed significantly at SKFL or SYFL. Both arsenic and
manganese exhibit increases at PAFL from 2011 to 2012.
• For the Florida sites sampling carbonyl compounds, formaldehyde had the highest
cancer risk approximations. Arsenic had the highest cancer risk approximation for
PAFL. All noncancer hazard approximations for the pollutants of interest for the
Florida sites were less than 1.0.
• Benzene was the highest emitted pollutant with a cancer toxicity factor in all three
Florida counties. Benzene also had the highest cancer toxicity-weighted emissions for
Pinellas County, while formaldehyde had the highest cancer toxicity-weighted
31-6
-------
emissions for Hillsborough County, and hexavalent chromium had the highest cancer
toxicity-weighted emissions for Orange County.
• Toluene was the highest emitted pollutant with a noncancer toxicity factor in all three
Florida counties. Acrolein had the highest noncancer toxicity-weighted emissions for
all three counties.
Georgia.
• The SDGA monitoring site located in Decatur, east of Atlanta, is a NATTS site.
• SDGA sampled for PAHs and hexavalent chromium, although sampling for PAHs
was discontinued in June. Naphthalene, hexavalent chromium, and benzo(a)pyrene
failed screens for SDGA, with naphthalene accounting for the majority of the total
failed screens.
• Hexavalent chromium was the only pollutant of interest for which an annual average
concentration could be calculated.
• Sampling for the site-specific pollutants of interest has occurred at SDGA for at least
5 consecutive years; thus, a trends analysis was conducted for the site-specific
pollutants of interest. This analysis shows that concentrations of hexavalent
chromium have not changed significantly at SDGA over the last few years of
sampling.
• The cancer risk approximation for hexavalent chromium for SDGA was considerably
less than 1 in-a-million. Hexavalent chromium's noncancer hazard approximation for
SDGA was considerably less than an HQ of 1.0.
• Tetrachloroethylene was the highest emitted pollutant with a cancer and noncancer
toxicity factor in DeKalb County. Formaldehyde had the highest cancer toxicity-
weighted emissions, while acrolein had the highest noncancer toxicity-weighted
emissions for DeKalb County.
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, near St. Louis.
• All three Illinois sites sampled for VOCs and carbonyl compounds. NBIL also
sampled for SNMOCs, PAHs, hexavalent chromium, and metals (PMi0), and is one of
two NMP sites sampling all six pollutant groups.
• Twenty-two pollutants failed screens for NBIL; 13 pollutants failed screens for SPIL;
and 11 pollutants failed screens for ROIL.
31-7
-------
• Of the pollutants of interest for NBIL and SPIL, formaldehyde and acetaldehyde had
the highest annual average concentrations. Annual averages could not be calculated
for ROIL due to the short sampling duration. The maximum acetaldehyde and
formaldehyde concentrations across the program were measured at SPIL. This was
also true of trichloroethylene.
• NBIL had the second highest annual average concentration of fluorene among NMP
sites sampling PAHs. The maximum fluorene and fluoranthene concentrations across
the program were measured at NBIL.
• 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. This analysis shows that concentrations of acetaldehyde
and manganese have an increasing trend at NBIL in recent years. 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, particularly for 2012.
• Formaldehyde had the highest cancer risk approximation for both NBIL and SPIL.
All noncancer hazard approximations for the pollutants of interest for the Chicago
sites were less than 1.0.
• Benzene was the highest emitted pollutant with a cancer toxicity factor in Cook
County, while coke oven emissions was the highest emitted "pollutant" with a cancer
toxicity factor in Madison County. Toluene was the highest emitted pollutant with a
noncancer toxicity factor for both counties, while acrolein had the highest noncancer
toxicity-weighted emissions for both counties.
Indiana.
• There are two Indiana monitoring sites, one located in Indianapolis (WPIN) and a
second located in Gary, near Chicago (INDEM). Both are UATMP sites.
• WPIN and INDEM sampled for carbonyl compounds only.
• Formaldehyde and acetaldehyde failed screens for both INDEM and WPIN;
propionaldehyde also failed a single screen for WPIN.
• Formaldehyde had the highest annual average concentration for both sites, although
concentrations were higher at WPIN than INDEM. WPIN's annual average
concentration of formaldehyde is the second highest annual average for this pollutant
among NMP sites sampling carbonyl compounds.
• 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. The most significant changes shown are for
INDEM. Both acetaldehyde and formaldehyde decreased dramatically at INDEM
31-8
-------
between 2008 and 2009, with little change since. These changes appear to be related
to a sampler contamination issue that was subsequently corrected.
• The cancer risk approximations for formaldehyde were an order of magnitude greater
than the cancer risk approximations for acetaldehyde for both sites. WPIN's cancer
risk approximation for formaldehyde is the second highest cancer risk approximation
calculated across the program.
• Benzene was the highest emitted pollutant with a cancer toxicity factor in both
Marion and Lake Counties. Coke oven emissions (PM) had the highest cancer
toxicity-weighted emissions for Lake County while formaldehyde had the highest
cancer toxicity-weighted emissions for Marion County.
• Toluene was the highest emitted pollutant with a noncancer toxicity factor in both
Lake and Marion Counties while acrolein had the highest noncancer toxicity-
weighted emissions for both counties.
Kentucky.
• Three Kentucky monitoring sites are located in northeast Kentucky, two in Ashland
(ASKY and ASKY-M) and one near Grayson Lake (GLKY). The Grayson Lake
monitoring site is aNATTS site. One monitoring site is located south of Evansville,
Indiana (BAKY). Five monitoring sites are located in or near the Calvert City area
(ATKY, CCKY, BLKY, LAKY, and TVKY). The final monitoring site is located in
Lexington, in north-central Kentucky (LEKY).
• Four monitoring sites (ASKY-M, BAKY, CCKY, and LEKY) began sampling
metals under the NMP in March 2012. Two monitoring sites (ASKY and LEKY)
began sampling carbonyl compounds in July. Seven monitoring sites also began
sampling VOCs in July. GLKY sampled VOCs, PAHs, carbonyl compounds,
metals and hexavalent chromium year-round.
• The number of pollutants failing screens for the Kentucky sites varies from three
(BAKY) to 15 (LEKY).
• Because the start dates for sampling were staggered, annual average concentrations
could only be calculated for GLKY and those Kentucky sites sampling PMio metals.
• The annual average concentrations for arsenic, manganese, lead, and nickel calculated
for ASKY-M were the highest annual averages among NMP sites sampling PMio
metals. The Kentucky sites have five of the 10 highest annual average concentrations
of arsenic among NMP sites; four of the highest annual average concentrations of
manganese; and two of the highest annual average concentrations of nickel.
• The Calvert city sites measured some of the highest concentrations of the VOCs,
particularly vinyl chloride, 1,2-dichloroethane, and carbon tetrachloride.
31-9
-------
• The highest cancer risk approximation (among the pollutants of interest for which
cancer risk approximations could be calculated) was calculated for formaldehyde for
GLKY. None of the pollutants of interest for which noncancer hazard approximations
could be calculated were greater than 1.0. The noncancer hazard approximation for
manganese for ASKY-M was the second highest noncancer hazard approximation
calculated across the program.
• Benzene and formaldehyde were the highest emitted pollutants with cancer toxicity
factors in all Kentucky counties with NMP sites, except Marshall County. Benzene
and ethylbenzene were the highest emitted pollutants with cancer toxicity factors in
Marshall County. Coke oven emissions and emissions of formaldehyde, benzene, and
POM Group la were among the pollutants with the highest cancer toxicity-weighted
emissions for the Kentucky counties with monitoring sites. Toluene was the highest
emitted pollutant with a noncancer toxicity factor in all Kentucky counties with NMP
sites, except Marshall County, where methanol emissions were higher than toluene
emissions. Acrolein had the highest noncancer toxicity-weighted emissions in four of
the Kentucky counties, but ranked second in Boyd County (behind manganese) and
Marshall County (behind chlorine).
Massachusetts.
• The Massachusetts monitoring site (BOMA) is a NATTS site located in Boston.
• BOMA sampled for metals (PMio), PAHs, and hexavalent chromium.
• Nine pollutants failed screens for BOMA. Arsenic and naphthalene each accounted
for roughly 40 percent of the site's failed screens.
• Of the pollutants of interest, naphthalene had the highest annual average
concentration.
• BOMA had the fifth 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. Concentrations of nickel have an overall decreasing trend at
BOMA over the years of sampling.
• The only pollutants of interest with cancer risk approximations greater than 1.0 in-a-
million were arsenic and naphthalene. None of the pollutants of interest for BOMA
had noncancer hazard approximations greater than 1.0.
• Formaldehyde was the highest emitted pollutant with a cancer toxicity factor in
Suffolk County and had the highest cancer toxicity-weighted emissions. Toluene was
31-10
-------
the highest emitted pollutant with a noncancer toxicity factor in Suffolk County,
while acrolein had the highest noncancer toxicity-weighted emissions.
Michigan.
• The three Michigan monitoring sites are located in the Detroit area. DEMI is a
NATTS site located in Dearborn. RRMI and SWMI are UATMP sites located in
River Rouge and Detroit, respectively.
• All three Michigan sites sampled carbonyl compounds; DEMI also sampled VOCs,
PAHs, and hexavalent chromium.
• Twenty-two pollutants failed screens for DEMI. Acetaldehyde and formaldehyde
both failed screens for RRMI and SWMI, contributing equally to the total number of
failed screens for each site.
• Formaldehyde had the highest annual average concentration for DEMI and SWMI.
Annual average concentrations could not be calculated for RRMI due to a collection
error that lead to the invalidation of 3 months of data.
• Compared to other NMP sites, DEMI had the highest annual average concentration of
carbon tetrachloride among sites sampling VOCs. DEMI also had the second highest
annual average concentration of acenaphthene and third highest annual averages of
fluorene and 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. The most notable trend is for benzene. Benzene concentrations
exhibit a steady decreasing trend although concentrations have leveled out in recent
years. In addition, the detection rate of 1,2-dichloroethane at DEMI has been
increasing steadily over the last few years of sampling, particularly for 2012.
• Formaldehyde had the highest cancer risk approximation for DEMI and SWMI. None
of the pollutants of interest for the Michigan sites had a noncancer hazard
approximation greater than 1.0.
• Benzene was the highest emitted pollutant with a cancer toxicity factor in Wayne
County, while coke oven emissions had the highest cancer toxicity-weighted
emissions. Toluene was the highest emitted pollutant with a noncancer toxicity factor
in Wayne County, while acrolein had the highest noncancer toxicity-weighted
emissions.
31-11
-------
Minnesota.
• The UATMP site in Minnesota (STMN) is located in St. Cloud.
• STMN sampled only hexavalent chromium.
• Hexavalent chromium failed six screens for STMN.
• Measured detections of hexavalent chromium span three orders of magnitude, ranging
from 0.0044 ng/m3 to 8.51 ng/m3, including 15 non-detects.
• The maximum hexavalent chromium concentration measured at STMN is the single
highest concentration measured under the NMP since this method was added to the
program in 2005.
• The cancer risk approximation for STMN for hexavalent chromium was the highest
cancer risk approximation calculated for this pollutant among NMP sites sampling
hexavalent chromium. The noncancer hazard approximation for hexavalent chromium
for STMN was still considerably less than 1.0.
• Bis(2-ethylhexyl)phthalate (DEHP), gas was the highest emitted pollutant with a
cancer toxicity factor in Stearns County, while formaldehyde had the highest cancer
toxicity-weighted emissions. Toluene was the highest emitted pollutant with a
noncancer toxicity factor in Stearns County, while acrolein had the highest noncancer
toxicity-weighted emissions.
Missouri.
• The NATTS site in Missouri (S4MO) is located in St. Louis.
• S4MO sampled for VOCs, carbonyl compounds, PAHs, metals (PMi0), and
hexavalent chromium.
• Twenty-four pollutants failed at least one screen for S4MO, 17 of which contributed
to 95 percent of failed screens. S4MO failed the greatest number of screens among
NMP sites.
• Of the pollutants of interest for S4MO, formaldehyde and acetaldehyde had the
highest annual average concentrations and were the only pollutants with annual
average concentrations greater than 1 |ig/m3.
• S4MO had the highest annual average concentration of 1,2-dichloroethane, the
second highest annual average concentrations of hexachloro-1,3-butadiene and
arsenic, and the third highest annual average concentrations of /?-dichlorobenzene and
manganese among NMP sites sampling these pollutants.
31-12
-------
• 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. The most significant changes in recent years are in the
acetaldehyde concentrations, which have decreased significantly since 2010. The
detection rate of 1,2-dichloroethane has been steadily increasing over the years, with
a significant increase for 2012.
• Formaldehyde had the highest cancer risk approximation for S4MO. None of the
pollutants of interest for S4MO had a noncancer hazard approximation greater
than 1.0.
• Formaldehyde was the highest emitted pollutant with a cancer toxicity factor in
St. Louis (city) and had the highest cancer toxicity-weighted emissions. Toluene was
the highest emitted pollutant with a noncancer toxicity factor, while acrolein had the
highest noncancer toxicity-weighted emissions in St. Louis (city).
New Jersey.
• The three UATMP sites in New Jersey are located in Chester (CLINT), Elizabeth
(ELNJ), and North Brunswick (NBNJ).
• CHNJ, ELNJ, and NBNJ sampled for VOCs and carbonyl compounds.
• Thirteen pollutants failed at least one screen for CHNJ; 14 pollutants failed at least
one screen for ELNJ; and 12 pollutants failed screens for NBNJ.
• Of the site-specific pollutants of interest, formaldehyde had the highest annual
average concentration for CHNJ, ELNJ, and NBNJ.
• NBNJ had the highest annual average concentration of hexachloro-1,3-butadiene
among NMP sites sampling VOCs. ELNJ had the fourth highest annual average
concentration of formaldehyde and fifth highest annual average concentration of
acetaldehyde among NMP sites sampling carbonyl compounds.
• Sampling for the site-specific pollutants of interest has occurred at each of the New
Jersey for at least 5 consecutive years; specifically, ELNJ is the longest running NMP
site still participating in the program. As such, a trends analysis was conducted for the
site-specific pollutants of interest. The most notable trend is for propionaldehyde at
ELNJ. Concentrations of propionaldehyde measured at ELNJ have a steady
increasing trend at this site. In addition, the detection rate of 1,2-dichloroethane at all
three New Jersey sites has been increasing steadily over the last few years of
sampling, particularly for 2012.
• Formaldehyde had the highest cancer risk approximation for CHNJ, ELNJ, and
NBNJ. None of the pollutants of interest for any of the New Jersey sites had
noncancer hazard approximations greater than 1.0.
31-13
-------
• Tetrachloroethylene was the highest emitted pollutant with a cancer URE in Union,
Middlesex, and Morris Counties. Formaldehyde also had the highest toxicity-
weighted emissions for Union and Middlesex Counties while benzene had the highest
toxicity-weighted emissions for Morris County.
• Toluene was the highest emitted pollutant with a noncancer toxicity factor in Morris
County, while tetrachloroethylene was the highest emitted pollutant with a noncancer
toxicity factor in Union and Middlesex Counties. Acrolein had the highest noncancer
toxicity-weighted emissions for each New Jersey county.
New York.
• The New York City monitoring sites are located in New York City (BXNY and
MONY) and Rochester (ROCH). All are NATTS monitoring sites.
• The instrumentation at BXNY were relocated to MONY due to roofing construction
in 2010. At the end of June 2012, the instrumentation was returned to the BXNY site
and sampling resumed at this location in July 2012. Thus, this report includes the
final 6 months of sampling at MONY and the initial 6 months of sampling after the
relocation back to BXNY.
• PAHs and hexavalent chromium were sampled at all three New York monitoring
sites.
• Five pollutants failed screens for BXNY, five pollutants failed screens for MONY,
and four pollutants failed screens for ROCH. Naphthalene failed the majority of
screens for all three sites.
• Naphthalene had the highest annual average concentration for ROCH. Sampling at
MONY ended in June, and sampling BXNY began in July. As a result, annual
average concentrations were not calculated for these two sites.
• ROCH has the third highest annual average concentration of acenaphthene and fourth
highest annual average of fluorene among NMP sites sampling PAHs.
• Cancer risk and noncancer hazard approximations could only be calculated for
ROCH. Naphthalene had the highest cancer risk approximation among the pollutants
of interest for ROCH. None of the pollutants of interest for ROCH had noncancer
hazard approximations greater than 1.0.
• Benzene was the highest emitted pollutant with a cancer toxicity factor for Bronx and
Monroe Counties while formaldehyde had the highest cancer toxicity-weighted
emissions for both counties.
31-14
-------
• Toluene was the highest emitted pollutant with a noncancer toxicity factor in Bronx
and Monroe Counties. Acrolein had the highest noncancer toxicity-weighted
emissions for both counties.
Oklahoma.
• There are five UATMP sites in Oklahoma: two located in Tulsa (TOOK and TMOK),
one in Pryor Creek (PROK), and two in Oklahoma City (ADOK and OCOK).
• Each of the Oklahoma sites sampled for VOCs, carbonyls compounds, and metals
(TSP). Sampling at PROK was discontinued at the end of October 2012.
• Twenty-one pollutants failed screens for TOOK; 19 failed screens for TMOK; 18
failed screens for PROK; 17 failed screens for ADOK; and 17 failed screens for
OCOK.
• Of the site-specific pollutants of interest, formaldehyde had the highest annual
average concentration for each Oklahoma site.
• TOOK had the highest annual average concentrations of acetaldehyde, benzene, and
ethylbenzene among NMP sites sampling these pollutants. TOOK has the highest
annual average concentration among NMP sites sampling manganese (PMio or TSP).
• Sampling for the site-specific pollutants of interest has occurred at TOOK for at least
5 consecutive years; thus, a trends analysis was conducted for the site-specific
pollutants of interest. Concentrations of ethylbenzene, manganese, and nickel exhibit
increasing trends at TOOK. In addition, the detection rate of 1,2-dichloroethane at
TOOK has been increasing steadily over the last few years of sampling, particularly
for 2012.
• Formaldehyde and benzene had the highest cancer risk approximations for all of the
Oklahoma monitoring sites. The benzene cancer risk approximation for TOOK is the
highest benzene cancer risk approximation program-wide. Arsenic had the highest
cancer risk approximations among the metals. None of the pollutants of interest for
the Oklahoma sites had a noncancer hazard approximation greater than 1.0.
• Benzene was the highest emitted pollutant with a cancer toxicity factor in Mayes,
Oklahoma, and Tulsa Counties. POM, Group la had the highest cancer toxicity-
weighted emissions for Oklahoma and Tulsa Counties while hexavalent chromium
had the highest cancer toxicity-weighted emissions for Mayes Counties.
• Toluene was the highest emitted pollutant with a noncancer toxicity factor in all three
counties, while acrolein had the highest noncancer toxicity-weighted emissions for all
three counties.
31-15
-------
Rhode Island.
• The Rhode Island monitoring site (PRRI) is located in Providence and is a NATTS
site.
• PRRI sampled for PAHs and hexavalent chromium.
• Three pollutants failed screens for PRRI, although 95 percent of failed screens were
attributable to naphthalene.
• Naphthalene had the highest annual average concentration among the other pollutants
of interest for PRRI.
• 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 exhibit little change over the
years. The range of hexavalent chromium concentrations has an increasing trend over
the last 4 years of sampling.
• The cancer risk approximation for naphthalene was considerably higher than the
cancer risk approximation for hexavalent chromium. All of the noncancer hazard
approximations for PRRI were less than 1.0.
• Benzene was the highest emitted pollutant with a cancer toxicity factor in Providence
County, while formaldehyde had the highest cancer toxicity-weighted emissions.
Toluene was the highest emitted pollutant with a noncancer toxicity factor, while
acrolein had the highest noncancer toxicity-weighted emissions for Providence
County.
South Carolina,
• The South Carolina monitoring site (CHSC) is located near Chesterfield and is a
NATTS site.
• CHSC sampled for hexavalent chromium and PAHs.
• Naphthalene was the only pollutant to fail screens for CHSC. Naphthalene failed
13 percent of screens for CHSC.
• Naphthalene concentrations measured at CHSC span an order of magnitude, ranging
from 5.61 ng/m3 to 58.3 ng/m3. Compared to other NMP sites sampling this pollutant,
CHSC had one of the lowest annual average concentrations of naphthalene.
• Sampling for the site-specific pollutants of interest has occurred at CHSC for at least
5 consecutive years; thus, a trends analysis was conducted for the site-specific
31-16
-------
pollutant of interest. Concentrations of naphthalene exhibit little change over the
years.
• The cancer risk approximation for naphthalene for CHSC was one of the lowest
among NMP sites sampling this pollutant; naphthalene's noncancer hazard
approximation for CHSC was considerably less than 1.0.
• Benzene was the highest emitted pollutant with a cancer toxicity factor in
Chesterfield County while formaldehyde had the highest cancer toxicity-weighted
emissions. Toluene was the highest emitted pollutant with a noncancer toxicity factor,
while acrolein had the highest noncancer toxicity-weighted emissions.
South Dakota.
• The UATMP site in South Dakota is located in Sioux Falls (SSSD).
• VOCs, SNMOCs, and carbonyl compounds were sampled for at SSSD.
• Eleven pollutants failed screens for SSSD, with six contributing to 95 percent of
failed screens for this site.
• Formaldehyde and acetaldehyde had the highest annual average concentrations for
SSSD and are the only two pollutants with annual averages greater than 1.0 |ig/m3.
• Sampling for the site-specific pollutants of interest has occurred at SSSD for at least
5 consecutive years; thus, a trends analysis was conducted for the site-specific
pollutants of interest. Concentrations of 1,3-butadiene increased significantly from
2011 to 2012 at SSSD. Conversely, formaldehyde concentrations measured at SSSD
exhibit a steady decreasing trend across the years, although the most significant
decreases were realized during the early years of sampling. In addition, the detection
rate of 1,2-dichloroethane at SSSD has been increasing steadily over the last few
years of sampling, particularly for 2012.
• Formaldehyde had the highest cancer risk approximation for SSSD. None of the
pollutants of interest for SSSD had a noncancer hazard approximation greater than
1.0.
• Formaldehyde was the highest emitted pollutant with a cancer toxicity factor in
Minnehaha County and had the highest toxicity-weighted emissions for this county.
• Toluene was the highest emitted pollutant with a noncancer toxicity factor in
Minnehaha County, while acrolein had the highest noncancer toxicity-weighted
emissions for this county.
31-17
-------
Texas.
• There are two NATTS sites in Texas: one in Deer Park (CAMS 35) and one in
Karnack (CAMS 85).
• The CAMS 35 site sampled for PAHs and hexavalent chromium, although sampling
for PAHs was discontinued in February 2012. CAMS 85 sampled for hexavalent
chromium only.
• Two pollutants failed screens for CAMS 35, naphthalene and hexavalent chromium.
Hexavalent chromium did not fail any screens for CAMS 85.
• Concentrations of hexavalent chromium measured at CAMS 35 ranged from
0.0044 ng/m3 to 0.195 ng/m3, including a single non-detect. The annual average
concentration for CAMS 35 is among the higher annual averages, ranking fifth
among the 22 NMP sites sampling hexavalent chromium. Because sampling for
PAHs was discontinued in February 2012, no quarterly or annual averages could be
calculated for naphthalene.
• The cancer risk approximation for hexavalent chromium for CAMS 35 is less than
1 in-a-million. The noncancer hazard approximation for hexavalent chromium for
CAMS 35 is considerably less than 1.0.
• Benzene was the highest emitted pollutant with a cancer toxicity factor in Harris
County, while 1,3-butadiene had the highest cancer toxicity-weighted emissions.
Formaldehyde was the highest emitted pollutant with a cancer toxicity factor in
Harrison County and had the highest cancer toxicity-weighted emissions.
• Toluene was the highest emitted pollutant with a noncancer toxicity factor in both
counties, while acrolein had the highest noncancer toxicity-weighted emissions.
Utah.
• The NATTS site in Utah (BTUT) is located in Bountiful, north of Salt Lake City.
• BTUT sampled for VOCs, carbonyl compounds, SNMOCs, PAHs, metals (PMio),
and hexavalent chromium and is one of two NMP sites sampling all six pollutant
groups.
• Twenty-one pollutants failed screens for BTUT, 13 of which contributed to
95 percent of this site's failed screens.
• Of the site-specific pollutants of interest, dichloromethane had the highest annual
average concentration for BTUT, similar to 2011. BTUT had the highest annual
average concentration of formaldehyde among NMP sites sampling carbonyl
compounds, for the second year in a row.
31-18
-------
• 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 an overall decreasing trend at BTUT. The 1-year average concentration
for 2012 is the lowest 1-year average concentration of benzene calculated since the
onset of sampling at BTUT. In addition, the detection rate of 1,2-dichloroethane at
BTUT has been increasing steadily over the last few years of sampling, particularly
for 2012.
• The pollutant with the highest cancer risk approximation for BTUT is formaldehyde;
this is the highest cancer risk approximation calculated across the program. None of
the pollutants of interest had noncancer hazard approximations greater than 1.0.
• Benzene was the highest emitted pollutant with a cancer toxicity factor in Davis
County while formaldehyde had the highest cancer toxicity-weighted emissions.
Toluene was the highest emitted pollutant with a noncancer toxicity factor, while
acrolein had the highest noncancer toxicity-weighted emissions for Davis County.
Vermont.
• Two Vermont monitoring sites are located in or near Burlington (BURVT and
UNVT); a third monitoring site is located in Rutland (RUVT). UNVT is a NATTS
monitoring site, while the remaining sites are UATMP sites.
• UNVT sampled year-round for VOCs, hexavalent chromium, PAHs, and metals
(PMio) while BURVT and RUVT sampled year-round for VOCs only. All three sites
began sampling carbonyl compounds under the NMP in July 2012.
• Eleven pollutants failed screens for BURVT; 12 pollutants failed screens for RUVT;
and 13 pollutants failed screens for UNVT.
• Benzene had the highest annual average concentrations for BURVT and RUVT,
while carbon tetrachloride had the highest annual average concentration for UNVT.
Annual averages for the carbonyl compounds could not be calculated due to the short
sampling duration.
• Annual average concentrations of the pollutants of interest for UNVT were among the
lowest compared to NMP sites sampling the same pollutants.
• Sampling for few of the site-specific pollutants of interest has occurred at UNVT for
at least 5 consecutive years; thus, a trends analysis was conducted for arsenic. Little
change is shown in the concentrations of arsenic since sampling under the NMP
began.
31-19
-------
• Benzene and carbon tetrachloride have the highest cancer risk approximations for the
Vermont monitoring sites (although not necessarily in that order). None of the
noncancer hazard approximations for these sites were greater than an HQ of 1.0.
• Benzene was the highest emitted pollutant with a cancer toxicity factor in Chittenden
and Rutland Counties. Benzene also had the highest cancer toxicity-weighted
emissions for Rutland County while formaldehyde had the highest cancer toxicity-
weighted emissions for Chittenden County.
• Toluene was the highest emitted pollutant with a noncancer toxicity factor in both
counties, while acrolein had the highest noncancer toxicity-weighted emissions.
Virginia.
• The NATTS site in Virginia is located near Richmond (RIVA).
• RIVA sampled for PAHs and hexavalent chromium.
• Four PAHs failed screens for RIVA, although naphthalene contributed to nearly
95 percent of the total failed screens. Hexavalent chromium did not fail any screens.
• Of the site-specific pollutants of interest, naphthalene had the highest annual average
concentration.
• Naphthalene had the highest cancer risk approximation for RIVA. None of the
pollutants of interest for RIVA had a noncancer hazard approximation greater than
1.0.
• Benzene was the highest emitted pollutant with a cancer toxicity factor in Henrico
County, while formaldehyde had the highest cancer toxicity-weighted emissions.
Toluene was the highest emitted pollutant with a noncancer toxicity factor in Henrico
County, while acrolein had the highest noncancer toxicity-weighted emissions.
Washington.
• The NATTS site in Washington is located in Seattle (SEWA).
• SEWA sampled for VOCs, carbonyl compounds, PAHs, metals (PMio), and
hexavalent chromium.
• Fifteen pollutants failed screens for SEWA, of which 10 were identified as pollutants
of interest for this site.
• Of the site-specific pollutants of interest for SEWA, acetaldehyde and carbon
tetrachloride had the highest annual average concentrations. The annual average
31-20
-------
concentration of formaldehyde for SEWA is the lowest among NMP sites sampling
this pollutant.
• SEWA had the second highest annual average concentration of nickel among NMP
sites sampling metals (PMio). This site had the highest annual average nickel
concentration for 2010 and 2011.
• 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 carbon tetrachloride exhibit a decreasing
trend over most of the sampling period, although this trend did not continue into
2012. In addition, the detection rate of 1,2-dichloroethane SEWA has been increasing
steadily over the last few years of sampling, particularly for 2012.
• Formaldehyde had 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 sites were
less than an HQ of 1.0.
• Benzene was the highest emitted pollutant with a cancer toxicity factor in King
County while formaldehyde had the highest cancer toxicity-weighted emissions.
Toluene was the highest emitted pollutant with a noncancer toxicity factor in King
County, while acrolein had the highest noncancer toxicity-weighted emissions.
Wisconsin.
• One Wisconsin monitoring site is located in Horicon (HOWI) and is a NATTS site.
The second site (MIWI) is located in Milwaukee and is a UATMP site.
• Both HOWI and MIWI sampled for hexavalent chromium only.
• Hexavalent chromium was detected in greater than half of samples collected at HOWI
but did not fail any screens. Hexavalent chromium was detected in nearly 80 percent
of samples collected at MIWI and failed nearly one-third of screens.
• Concentrations of hexavalent chromium measured at MIWI spanned three orders of
magnitude, ranging from 0.0045 ng/m3 to 2.30 ng/m3 (as well as 11 non-detects).
MIWI is one of only two NMP sites at which concentrations of hexavalent chromium
greater than 1 ng/m3 were measured.
• The cancer risk approximation for hexavalent chromium for MIWI is one of only two
cancer risk approximations for this pollutant greater than 1 in-a-million program-
wide. The noncancer hazard approximation for hexavalent chromium is considerably
less than an HQ of 1.0.
31-21
-------
• Benzene was the highest emitted pollutant with a cancer toxicity factor in Dodge
County and had the highest cancer toxicity-weighted emissions. Benzene was also the
highest emitted pollutant with a cancer toxicity factor in Milwaukee County, although
POM, Group la had the highest cancer toxicity-weighted emissions.
• Toluene was the highest emitted pollutant with a noncancer toxicity factor in Dodge
and Milwaukee Counties, while acrolein had the highest noncancer toxicity-weighted
emissions for each county.
31.1.3 Composite Site-level Results Summary
• Twenty-eight pollutants were identified as site-specific pollutants of interest, based
on the risk-based screening process. Acetaldehyde and formaldehyde were the two
most common pollutants of interest among the monitoring sites. Acetaldehyde and
formaldehyde were identified as pollutants of interest for all 37 sites that sampled
carbonyl compounds. Benzene, 1,3-butadiene, and carbon tetrachloride were the most
common VOC pollutants of interest. Benzene was identified as a pollutant of interest
for all 35 sites that sampled VOCs and/or SNMOCs. Twenty-three of the 25 sites that
sampled PAHs had naphthalene as a pollutant of interest (based on the risk-based
screening process). Arsenic was identified as a pollutant of interest for all 19 sites that
sampled metals. Hexavalent chromium was identified as a pollutant of interest for
seven of the 25 sites that sampled this pollutant.
• Concentrations from two sites, CAMS 85 and HOWI, did not fail any screens.
However, only hexavalent chromium was sampled at these two sites.
• 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 25 sites. Naphthalene had the next highest at nine.
• Formaldehyde tended to have the highest cancer risk approximations on a site-
specific basis. The cancer risk approximation calculated for BTUT (57.62 in-a-
million) from the annual average concentration of formaldehyde is the highest of all
annual average-based cancer risk approximations. Three other sites exhibited cancer
risk approximations greater than 50 in-a-million for formaldehyde (WPIN, PXSS, and
ELNJ). Benzene is the only other pollutant for which a cancer risk approximation
greater than 10 in-a-million was calculated (TOOK, SPAZ, and PXSS).
• 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 3.5 in-a-million and 4.5 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).
31-22
-------
• None of the noncancer hazard approximations were greater than 1.0. The noncancer
hazard approximation calculated for TOOK's annual average concentration of
manganese (an HQ of 0.77) was the highest of all annual average-based noncancer
hazard approximations. Formaldehyde and naphthalene along with manganese tended
to have the highest noncancer hazard approximations on a site-specific basis.
• Of those pollutants with cancer UREs, benzene, formaldehyde, ethylbenzene, and
acetaldehyde 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, and ethylene glycol 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 once among the highest
emitted pollutants for counties with NMP sites (Garfield County, Colorado).
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.
• For the 2012 NMP report, ethylene glycol emissions rank higher than they did for the
2011 NMP report. Emissions data provided in the 2012 NMP report are from
version 1 of the 2011 NEI while emissions data for the 2011 NMP report were from
the 2008 NEI. The movement in the ranking of ethylene glycol emissions may be
attributable to differences in the way these emissions were reported between the
different versions of the inventory.
• 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 county 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 measurements collected 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 10 percent detection rate across the program). The five
Calvert City, Kentucky sites account for nearly half of the measured detections of
vinyl chloride for 2012 yet sampling did not begin at these sites until July. The only
other monitoring site with a similar statistic is DEMI. The Calvert City sites also
account for the 50 highest concentrations of 1,2-dichloroethane measured across the
program. These ambient air measurements agree with corresponding emissions data
31-23
-------
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 one of the
highest emitted pollutants for any other county with an NMP site. From a quantitative
standpoint, the emissions of carbon tetrachloride and vinyl chloride in Marshall
County are higher than their emissions for any other county with an NMP site. The
emissions of 1,2-dichloroethane for Marshall County rank second highest (behind
only Harris County, Texas).
• For every NMP site for which 1,2-dichloroethane was a pollutant of interest (14
sites), and where a trends analysis could be conducted for this pollutant, a dramatic
increase in the number of measured detections is shown, particularly for 2012. 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, with a detection rate between 80 percent and 95 percent for most sites for 2012.
31.1.4 Data Quality Results Summary
Completeness, precision, and accuracy were assessed for the 2012 monitoring effort. The
quality assessments presented in this report show that the 2012 monitoring data are of a known
and high quality, based on the attainment of the established MQOs.
To the largest extent, ambient air concentration data sets met MQO for completeness.
Only five out of 144 site- and method-specific data sets failed to comply with the MQO of
85 percent completeness while 71 data sets achieved 100 percent completeness.
Method (sampling and analytical) precision and analytical precision were determined for
the 2012 NMP monitoring efforts using CV calculations based on duplicate, collocated, and
replicate samples. The precision for each analytical method utilized during the 2012 NMP was
within the MQO of 15 percent CV. The method precision presented in this report is based on
analytical results greater than or equal to the sample- and pollutant-specific MDL.
Analytical method accuracy is ensured by using proven methods, as demonstrated by
third-party analysis of proficiency test audit samples, and following strict quality control and
quality assurance guidelines. Most of the pollutants for which audit samples were analyzed met
the MQO for accuracy. Of the 90 pollutants analyzed for via audit samples, only two exceeded
the MQO of ± 25 percent.
31-24
-------
31.2 Conclusions
Conclusions resulting from the data analyses of the data generated from the 2012 NMP
monitoring efforts are presented below.
• There are a large number of concentrations that are greater than their respective risk
screening values, particularly for many of the NATTS MQO Core Analytes.
However, there were no instances where the preprocessed daily measurements or
time-period average concentrations were greater than the ATSDR MRL noncancer
health risk benchmarks.
• For those pollutants for which annual averages could be calculated and have available
cancer UREs, none of the cancer risk approximations were greater than 100 in-a-
million; 30 were greater than 10 in-a-million (27 for formaldehyde and three for
benzene); and roughly 80 percent were greater than 1.0 in-a-million.
• For those pollutants for which annual averages could be calculated and have available
noncancer RfCs, none of the noncancer hazard approximations were greater than 1.0.
• When comparing the highest emitted pollutants for a specific county to the pollutants
with the highest toxicity-weighted emissions, the listed pollutants were 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, but was not one of the pollutants with highest toxi city-
weighted emissions for any listed county. Conversely, while acrolein had the highest
noncancer toxicity-weighted emissions for most NMP counties, it was among the
highest emitted pollutants for only one county.
• The number of states and sites participating in the NMP changes from year to year.
The number of sites participating in the 2012 NMP increased rather significantly,
from 51 in 2011 to 64 for 2012. Yet, many of the data analyses utilized in this report
require data from year-round (or nearly year-round) sampling. Of the 64 sites whose
data are included in the 2012 report, 18 sites sampled for an abbreviated duration (due
to site initialization and/or site closure/relocation). Of the 144 site-method
combinations, 32 site-method combinations did not cover the entire year. As a result,
the number of time-period averages and subsequent risk-based analyses that could not
be calculated increased for 2012 compared to 2011. Fewer data gaps allow for more
complete results and inter-site comparisons. However, most of these abbreviated
durations are due to site initialization rather than site closure.
• Of the 64 monitoring sites participating in the 2012 NMP, only two sampled for all
six available pollutant groups under the national program (BTUT and NBIL). Another
five sites sampled all five pollutant groups required for NATTS sites (GLKY, PXSS,
S4MO, SEW A, and UNVT). The wide range of pollutant groups sampled among the
31-25
-------
sites, which is often a 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.
• This report strives to represent the best laboratory practices and utilize the best data
analysis techniques available. Examples for 2012 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. The major
difference between 2012 report and other reports in recent years is the determination
of the pollutants of interest. The NATTS MQO Core Analytes were not automatically
included as pollutants of interest for each site for 2012, allowing the data generated
for each site to be the primary driver of each site's pollutants of interest. These site-
specific pollutants of interest were then the same pollutants for which the trends
analysis and program vs. site-specific concentration comparisons were performed.
31.3 Recommendations
Based on the conclusions from the 2012 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 2012 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
31-26
-------
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. Two analytical methods were accepted by governing bodies as approved
method 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).
• Revise the pollutants targeted for 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/need to be reviewed and
revised periodically based experience with the collection and analysis methods and
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. 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 most recent version of the NATTS TAD, is integral to the
identification of trends and measuring the effectiveness of regulation.
• 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. Tetrachloride 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 identify state and local agencies pinpoint issues affecting
air quality in their area.
31-27
-------
• 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 several pollutants
sampled during the 2012 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.
• 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).
31-28
-------
32.0 References
ACE, 2014. U.S. Army Corps of Engineers. Huntington District, Grayson Lake website.
http://www.lrh.usace.army.mil/Missions/Recreation/Kentuckv/GravsonLake.aspx
Date Last Accessed: 3/4/2014.
ASTM, 2012. ASTM, International. ASTM D7614-12 Standard Test Method for Determination
of Total Suspended Particulate (TSP) Hexavalent Chromium in Ambient Air by Ion
Chromatography (1C) and Spectrophotometric Measurements.
http://www.astm.org/Standards/D7614.htm Date Last Accessed: 9/5/2013.
ASTM, 2013. ASTM, International. ASTM D6209 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: 9/3/2013.
ATSDR, 2013a. Agency for Toxics Substances and Disease Registry. July 2013. Minimal Risk
Levels (MRLs). http://www.atsdr.cdc.gov/mrls/pdfs/atsdr_mrlsjuly_2013.pdf
Date Last Accessed: 9/6/2013.
ATSDR, 2013b. Agency for Toxics Substances and Disease Registry. Toxic Substances Portal
website, Minimal Risk Levels (MRLs). http://www.atsdr.cdc.gov/mrls/index.asp
Date Last Accessed: 9/6/2013.
AZ DOT, 2010. Arizona Department of Transportation. Data Analysis website. 2010 AADTs.
http://mpd.azdot.gov/planning/DataandAnalysis/average-annual-daily-traffic
Date Last Accessed: 8/21/2013.
AZ DOT, 2011. Arizona Department of Transportation, Motor Vehicle Division. July 31, 2011.
Currently Registered Vehicles by County Listing (MV630419).
http://www.azdot.gov/mvd/statistics/registeredVehicles.asp Date Last Accessed:
7/31/2013.
AZ DOT, 2013. Arizona Department of Transportation, Motor Vehicle Division.
August 13, 2013. Daily Vehicle-Miles of Travel (DVMT) by County with Population
Estimates for the Year ended December 31, 2012.
http://mpd.azdot.gov/planning/DataandAnalvsis/highwav-performance-monitoring-
system Date Last Accessed: 8/21/2013.
Bair, 1992. Bair, Frank E. 1992. Weather of U.S. Cities, 4th ed. Gale Research Inc. Detroit, MI.
Boubel, et al., 1994. Boubel, R.W., D.L. Fox, D.B. Turner, and A.C. Stern. 1994. Fundamentals
of Air Pollution, 3rd ed. San Diego, CA: Academic Press, pp 263.
CA DMV, 2012. California Department of Motor Vehicles. Estimated Vehicles Registered by
County for the Period of January 1 through December 31, 2012.
http://www.dmv.ca.gov/about/profile/what_dmv_does.htm Date Last Accessed:
7/31/2013.
32-1
-------
CA DOT, 2012a. State of California Department of Transportation, Division of Traffic
Operations. 2012 Traffic Volumes on California State Highways. Sacramento, CA.
http://www.dot.ca.gov/hq/traffops/saferesr/trafdata/index.htm Date Last Accessed:
8/22/2013.
CA DOT, 2012b. State of California Department of Transportation. October 2012. 2011
California Public Road Data, http://www.dot.ca.gov/hq/tsip/hpms/index.php
Date Last Accessed: 11/16/2012.
Calvert City, 2014. Calvert City - Industry website, http://www.calvertcity.com/industry.htm
Date Last Accessed: 3/4/2014.
Carbondale, 2014. Carbondale, Colorado. The Town website, http://www.carbondale.com/town
Date Last Accessed: 2/4/2014.
CCRPC, 2013. Chittenden County Regional Planning Commission. Transportation Data-Annual
Average Daily Traffic website. 2007 and 2011 Data.
http://www.ccrpcvt.org/data/traffic.php?count=ATR Date Last Accessed: 8/27/2013.
Census Bureau, 2012. U.S. Census Bureau. Annual Estimates of the Population of the United
States, Region, States, and Puerto Rico: April 1, 2010 to July 1, 2012. Released
December 2012. http://www.census.gov/popest/data/state/totals/2012/index.html
Date Last Accessed: 8/1/2013.
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: 8/6/2013.
Census Bureau, 2013b. U.S. Census Bureau. Metropolitan and Micropolitan Statistical Areas
website, http://www.census.gov/population/metro/ Date Last Accessed: 8/28/2013.
Census Bureau, 2013c. U.S. Census Bureau. Annual Estimates of the Resident Population:
April 1, 2010 to July 1, 2012. Released March 2013.
http://www.census.gov/popest/data/counties/totals/2012/CO-EST2012-01.html
Date Last Accessed: 5/21/2013.
CO DOR, 2012. Colorado Department of Revenue. December 26, 2012. Colorado Department
of Revenue 2012 Annual Report July 1, 2011-June 30, 2012. Denver, CO.
http://www.colorado.gov/cs/Satellite?c=Page&cid=1213 86797503 5&pagename=Revenu
e-Main%2FXRMLavout Date Last Accessed: 7/31/2013.
CO DOT, 2011. Colorado Department of Transportation. Online Transportation Information
System (OTIS), Traffic Data Explorer. 2011 data.
http://dtdapps.coloradodot.info/Otis/TrafficData Date Last Accessed: 8/22/2013.
32-2
-------
CO DOT, 2012. Colorado Department of Transportation. Online Transportation Information
System (OTIS). 2012 State Highway Statistics - Daily Vehicle Miles of Travel (DVMT)
for All Vehicles by County.
http://dtdapps.coloradodot.info/staticdata/Statistics/dsp folder/Roadwav/2012/2012DVM
TbvCountv.pdf Date Last Accessed: 10/16/2013.
CoCoRaHS, 2011. Community Collaborative Rain, Hail, and Snow Network. CoCoRaHS "State
Climates" Series website. http://www.cocorahs.org/Content.aspx?page=50StatesClimates
Date Last Accessed: 5/7/2014.
Cornell, 2014. New York State Climate Office, Cornell University. The Climate of New York.
http://nysc.eas.cornell.edu/ Date Last Accessed: 4/24/2014.
DC DOT, 2012a. District Department of Transportation. March 2012. Traffic Volume Map
2010. Washington, D.C.
http ://ddot. dc. gov/DC/DDOT/About+DDOT/Maps/Traffic+Volume+Map+2010 Date
Last Accessed: 8/23/2013.
DC DOT, 2012b. Email from Nana Bailey-Thomas, District Department of Transportation,
FOIA Officer. September 11, 2012. 2011 Data, dated 7/9/12.
Draxler, R.R. and Rolph, G.D., 1997. Draxler, R.R. and Rolph, G.D. 1997. Description of the
HYSPLIT_4 modeling system. NOAA Air Resources Laboratory. NOAA Tech. Memo.
ERL ARL-224. Silver Spring, MD. 24 pp. http://www.arl.noaa.gov/HYSPLIT_pubs.php
Draxler, R.R. and Rolph, G.D., 1998. Draxler, R.R. and Rolph, G.D. 1998. An overview of the
HYSPLIT_4 modeling system of trajectories, dispersion, and deposition. Aust. Meteor.
Mag. 47, 295-308. http://www.arl.noaa.gov/HYSPLIT pubs.php
Draxler, R.R., 1999. Draxler, R.R. 1999. HYSPLIT4 user's guide. NOAA Air Resources
Laboratory. NOAA Tech. Memo. ERL ARL-230. Silver Spring, MD.
http://www.arl.noaa.gov/HYSPLIT_pubs.php
Draxler, R.R., et al, 2009. Draxler, R.R., et al. January 2009. HYSPLIT 4 User's Guide, Version
4.9. NOAA Air Resources Laboratory. Silver Spring, MD.
http://www.arl.noaa.gov/documents/reports/hysplit user guide.pdf
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/files/ambient/pams/newtad.pdfDate Last Accessed:
9/3/2013.
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: 9/3/2013.
32-3
-------
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: 9/3/2013.
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: 9/3/2013.
EPA, 1999d. U.S. EPA. June 1999. Compendium Method IO-3.5: Determination of Metals in
Ambient Particulate Matter Using Inductively Coupled Plasma/Mass Spectrometry
(ICP/MS). EPA/625/R-96/010a. Cincinnati, OH.
http://www.epa.gov/ttn/amtic/files/ambient/inorganic/mthd-3-5.pdf Date Last Accessed:
9/4/2013.
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.gOv/ttn/chief/conference/eil6/session6/a.pope.pdfDate Last Accessed:
12/10/2012.
EPA, 2009a. U.S. EPA. April 2009. UATMP, NATTS, CSATAM, PAMS, andNMOC Support,
Contract No. EP-D-09-48. Pages 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: 8/15/2013.
EPA, 2009c. U.S. EPA. September 10, 2009. Schools Air Toxics Monitoring Activity (2009)
Uses of Health Effects Information in Evaluating Sample Results.
http://www.epa.gov/schoolair/pdfs/UsesOfHealthEffectsInfoinEvalSampleResults.pdf
EPA, 2010. 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.
http://www.epa.gov/region4/air/airtoxic/Screening_l 11610_KMEL.pdf Date Last
Accessed: 9/5/2013.
EPA, 201 la. U.S. EPA. June 2011. Technical Report for School: Assessing Outdoor Air Near
Schools: Riggins School, North Birmingham Elemenary School, and Lewis Elementary
School (Birmingham, Alabama). http://www.epa.gov/schoolair/NorthBirmi.html
Date Last Accessed: 9/4/2013.
EPA, 201 Ib. U.S EPA. March 2011. The 2005 National-Scale Air Toxics Assessment (NATA)
glosssary website, http://www.epa.gov/ttn/atw/natamain/gloss 1 .html Date Last Accessed:
9/5/2013.
32-4
-------
EPA, 201 Ic. U.S. EPA. January 2011. An Overview of Methods for EPA's National-Scale Air
Toxics Assessment, Appendix I. http://www.epa.gov/ttn/atw/nata2005/methods.html
Date Last Accessed: 8/4/2014.
EPA, 2012a. U.S. EPA. June 2012. Ambient Air Monitoring Reference and Equivalent Methods:
Designation of Three New Equivalent Methods. June 1, 2012. Federal Register Volume
77, Number 106, page 32632. http://www.gpo.gov/fdsvs/pkg/FR-2012-06-01/html/2012-
13350.htm Date Last Accessed: 7/11/2014.
EPA, 2012b. U.S. EPA. September 2012. National Air Toxics Trends Stations (NATTS)
Network Assessment, Revised Draft. Prepared by Eastern Research Group, Inc.
EPA, 2013a. U.S. EPA. Air Toxics - National Air Toxics Trends Stations website.
http://www.epa.gov/ttn/amtic/natts.html Date Last Accessed: 8/15/2013.
EPA, 2013b. U.S. EPA. Technology Transfer Network, Air Quality System (AQS) website.
http://www.epa.gov/ttn/airs/airsaqs/ Date Last Accessed: 8/15/2013.
EPA, 2013c. U.S. EPA. September 30, 2013. 2011 National Emissions Inventory, Version 1.
http://www.epa.gov/ttnchiel/net/201 linventory.html Date Last Accessed: 10/14/2013.
EPA, 2013d. 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 O2.tpl Date Last Accessed:
8/29/2013.
EPA, 2013e. U.S. EPA. Technology Transfer Network, About Air Toxics website.
http://www.epa.gov/ttn/atw/allabout.html Date Last Accessed: 9/5/2013.
EPA, 2013f. Email from Ted Palma, U.S. EPA, OAQPS. September 9, 2013. Dose Response
Library spreadsheet.
EPA, 2013g. U.S. EPA. March 2013. Status of Acrolein Monitoring at Schools website.
http://www.epa.gov/schoolair/acrolein.html Date Last Accessed: 9/6/2013.
EPA, 2013h. U.S. EPA. Climate Change: Basic Information website.
http://epa.gov/climatechange/basics/ Date Last Accessed: 10/16/2013.
EPA, 2013L U.S. EPA. Overview of Greenhouse Gases website.
http://epa.gov/climatechange/ghgemissions/gases.html Date Last Accessed: 10/16/2013.
EPA, 2014. U.S. EPA. Envirofacts Multisystem Search website.
http://www.epa.gov/enviro/facts/multisystem.html Date Last Accessed: 7/16/2014.
ERG, 2012. Eastern Research Group, Inc. January 2012. Support for the EPA National
Monitoring Programs (UATMP, NATTS, CSATAM, PAMS, and NMOC Support),
Quality Assurance Project Plan, Category 1. Contract No. EP-D-09-048. Morrisville,
NC.
32-5
-------
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 28, 2007. Submitted to the U.S. EPA.
http://water.epa.gov/scitech/methods/cwa/det/ Date Last Accessed: 8/29/2013.
FCC, 2014. Florida Climate Center, Office of the State Climatologist. Florida State University.
Morton D. Winsberg. Climate of Florida, http://climatecenter.fsu.edu/topics/specials
Date Last Accessed: 1/15/2014.
FFIWA, 2013a. Federal Highway Administration. March 2013. Highway Statistics 2011. State
Motor Vehicle Registrations-2011 (Revised). Table MV-1.
http://www.fhwa.dot.gov/policyinformation/statistics.cfm Date Last Accessed:
7/31/2013.
FHWA, 2013b. Federal Highway Administration. March 2013. Highway Statistics 2011.
Glossary website.
http://www.fhwa.dot.gov/policyinformation/statistics/2011/16 glossary.cfm
Date Last Accessed: 8/28/2013.
FL DHSMV, 2012. Florida Department of Highway Safety and Motor Vehicles. Registered
Vehicles by County as of May 1, 2012.
http://flhsmv.gov/html/reports and statistics/reports and statistics.html
Date Last Accessed: 9/27/2013.
FL DOT, 2012a. Florida Department of Transportation. Florida Traffic Online Mapping
Application (2012). http://www.dot.state.fl.us/planning/statistics/trafficdata/
Date Last Accessed: 8/26/2013.
FL DOT, 2012b. Florida Department of Transportation, Transporation Statistics Office. Public
Road Mileage and Miles Traveled, 2012.
http://www.dot.state.fl.us/planning/statistics/mileage-rpts/public.shtm
Date Last Accessed: 8/9/2013.
GA DOR, 2011. Georgia Department of Revenue, Motor Vehicle Division. Statistics-Vehicle
Registration website. Registrations as of November 30, 2011.
http://motor.etax.dor.ga.gov/stats/renewalsstats.aspx Date Last Accessed: 12/5/2011.
GA DOT, 2012a. Georgia Department of Transportation. Georgia's State Traffic and Report
Statistics (STARS) mapping application. 2012 Published Data.
http://www.dot.ga.gov/informationcenter/statistics/stars/Pages/default.aspx
Date Last Accessed: 8/23/2013.
GA DOT, 2012b. Georgia Department of Transportation. Mileage by Route and Road System
Report 445 for 2012.
http://www.dot.ga.gov/informationcenter/statistics/RoadData/Pages/400Series.aspx
Date Last Accessed: 8/12/2013.
32-6
-------
GCPH, 2013. Garfleld County Public Heath Department. July 2013. Garfield County 2012 Air
Quality Monitoring Summary. Rifle, CO. Prepared by Air Resource Specialists, Inc.,
Ft. Collins, CO. http://www.garfield-county.com/air-quality/air-monitoring-2012.aspx
Date Last Accessed: 2/5/2014.
GCRBD, 2002. Garfield County Road and Bridge Department. 2002 Transportation Study,
County Road System, Average Daily Traffic, Garfield County, Colorado.
http://www.garfield-county.com/road-bridge/county-roads.aspx Date Last Accessed:
8/22/2013.
GSCO, 1998. Georgia State Climate Office, University of Georgia. 1998. Climatology of the
Georgia Piedmont. Athens, Georgia.
http://www.rivercenter.uga.edu/education/summit/general/climate/piedmont.pdf
Date Last Accessed: 4/16/2013.
HCPID, 2013. Harris County Public Infrastructure Department, Architecture and Engineering
Department, Traffic and Transportation Group. April 15, 2013. Traffic Volumes Report,
Update 2012. Data from December 2004. Houston, Texas.
http://www.eng.hctx.net/traffic/traffic.htm Date Last Accessed: 8/27/2013.
Henrico County, 2013. Email from Deborah Oxenreider, County of Henrico, Revenue Division.
August 1, 2013. 2012 Data.
IL DOT, 2011/2012. Illinois Department of Transportation. Getting Around Illinois website.
Average Annual Daily Traffic map viewer. 2011 and 2012 data.
http://www.gettingaroundillinois.com/gai.htm?mt=aadt Date Last Accessed: 8/26/2013.
IL DOT, 2012. Illinois Department of Transportation, Office of Planning and Programming.
Illinois Travel Statistics 2012. http://www.dot.il.gov/adttravelstats.html
Date Last Accessed: 8/12/2013.
IL SCO, 2014. State Climate Office for Illinois, Illinois State Water Survey, Prairie Research
Institute, University of Illinois Urbana-Champaign. Climate of Chicago - Description and
Normals website. http://www.isws.illinois.edu/atmos/statecli/General/chicago-climate-
narrative.htm Date Last Accessed: 2/24/2014.
IL SOS, 2012. Illinois Secretary of State, Vehicle Services Department. 2012 Registration
Counts by County.
http://www.cyberdriveillinois.com/departments/vehi cles/title_and_registration/countycou
nts.html Date Last Accessed: 8/1/2013.
IN BMV, 2012. Email from Julie Fletcher, Indiana Bureau of Motor Vehicles Communications
Department. October 19, 2012.
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: 8/26/2013.
32-7
-------
IN DOT, 2012. Indiana Department of Transportation. Mileage and DVMT by Year, County,
and System. Dated October 9, 2012. http://www.in.gov/indot/2469.htm
Date Last Accessed: 10/30/2012.
IPCC, 2012. Intergovernmental Panel on Climate Change. 2012. Errata for the Working Group I
Fourth Assesment Report. June 15, 2012.
http://www.ipcc.ch/publications and data/publications ipcc fourth assessment report
wgl_report_the_physical_science_basis.htm Date Last Accessed: 10/18/2013.
ISCO, 2002. Indiana State Climate Office, Purdue University. About Indiana Climate website.
December 2002. http://climate.agry.purdue.edu/climate/narrative.asp Date Last Accessed:
2/28/2014.
KY, 2014. Kentucky State Parks. Grayson Lake History website.
http://parks.kv.gov/parks/recreationparks/gravson-lake/history.aspx Date Last Accessed:
3/4/2014.
KYTC, 2013a. Email from Justin Carmack, Kentucky Transportation Cabinet, Division of Motor
Vehicle Licensing. August 5, 2013.
KYTC, 2013b. Kentucky Transportation Cabinet. Interactive Statewide Traffic Counts mapping
application. 2010-2012 Data. http://transportation.ky.gov/Planning/Pages/count-
maps.aspx Date Last Accessed: 8/26/2013.
KYTC, 2013c. Kentucky Transportation Cabinet. July 8, 2013. 2012 DVMT (in thousands) and
Mileage by County and Functional Class.
http://transportation.ky.gov/Planning/Pages/Roadway-Information-and-Data.aspx
Date Last Accessed: 8/12/2013.
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: 8/15/2013.
Lakes, 2011. Lakes Environmental. June 2011. WRPLOT View 7.0.0.
http://www.weblakes.com/products/wrplot/index.html
MA DOT, 2010. Massachusetts Department of Transportation, Highway Division.
Transportation Data Management System. 2010 Traffic Count.
http://www.mhd. state. ma.us/default.asp?pgid=content/traffic01&sid=about
Date Last Accessed: 8/26/2013.
MA DOT, 2013. Email from Bob Frey, Massachusetts Department of Transportation, Office of
Transportation Planning. August 20, 2013.
MA RMV, 2013. Email from Joan M. Valley, Massachusetts Department of Transportation,
Registry of Motor Vehicles. Data as of August 2012. October 7, 2013.
32-8
-------
MCC, 2014. Missouri Climate Center, University of Missouri. Wayne L. Decker. Climate of
Missouri website, http://climate.missouri.edu/climate.php Date Last Accessed:
3/24/2014.
MCWG, 2013. Minnesota Climatology Working Group, University of Minnesota. Normals,
Means, and Extremes St. Cloud, MN (STC).
http://climate.umn.edu/doc/online resources.htm Date Last Accessed: 8/30/2013.
MDS, 2013. Email from Arlene Rothwell, Michigan Department of State. September 17, 2013.
Non-Expired Total Registrations on File as of 11/10/2012.
MI DOT, 2012. Michigan Department of Transportation. Traffic Monitoring Information System
(TMIS) website. Annual Average Daily Traffic Report (2012).
http://mdotnetpublic.state.mi.us/tmispublic/Search.aspx Date Last Accessed: 8/26/2013.
MI DOT, 2013. Email from Don Howe, Michigan Department of Transportation, Bureau of
Transportation Planning. August 12, 2013. 2012 AVMT Data.
MN DOT, 2009. Minnesota Department of Transportation, Traffic Forecasting and Analysis.
Traffic Mapping Application. 2009 Data, http://www.dot.state.mn.us/traffic/data/tma-
cu5.html Date Last Accessed: 8/26/2013.
MN DOT, 2013. Minnesota Department of Transportation. October 14, 2013. Daily (Average)
and Annual (Total) Vehicle Miles - 2012. http://www.dot.state.mn.us/roadway/data/data-
products.html Date Last Accessed: 10/16/2013.
MN DPS, 2013. Minnesota Department of Public Safety, Driver and Vehicle Services.
February 12, 2013. 2012 Motor Vehicle County Summary Report (PSVH7301).
https://dps.mn.gov/divisions/dvs/forms-documents/Pages/motor-vehicle-forms.aspx
Date Last Accessed: 8/1/2013.
MO DOR, 2013. Missouri Department of Revenue. Total Motor Vehicles by Kind of Vehicle
within County (DM45080-01). Run Date 1/24/13.
http://dor.mo.gov/publicreports/index.php Date Last Accessed: 8/1/2013.
MO DOT, 2012. Missouri Department of Transportation, Transportation Planning. 2011
St. Louis District Traffic Volume and Commercial Vehicle Count Map.
http://www.modot.org/safety/trafficvolumemaps.htm Date Last Accessed: 10/19/2012.
MO DOT, 2013. Email from Tina Schmidtz, Missouri Department of Transportation.
May 17, 2013. Daily VMT by District, County and Functional Class for St. Louis City
and County, Traffic Year 2012.
MFC A, 2013. Minnesota Pollution Control Agency. Annual Air Monitoring Network Plan for
Minnesota 2013. St. Paul, MN. http://www.pca.state.mn.us/index.php/view-
document.html?gid=17855 Date Last Accessed: 3/24/2014.
32-9
-------
MSU, 2014. Michigan State Climatologist's Office, Michigan State University. Climate of
Dearborn Narrative, http://climate.geo.msu.edu/climate mi/stations/20157
Date Last Accessed: 3/18/2014.
NCDC, 2011. National Climatic Data Center. National Virtual Data System (NVDS). Quality
Controlled Local Climatological Data (QCLD/ULCD). 2011 Data.
https://nes.ncdc.noaa.gov/subscriptions.html Date Last Accessed: 8/23/2013.
NCDC, 2012. National Climatic Data Center. National Virtual Data System (NVDS). Quality
Controlled Local Climatological Data (QCLD/ULCD). 2012 Data.
https://nes.ncdc.noaa.gov/subscriptions.html Date Last Accessed: 8/23/2013.
NCDC, 2014. National Climatic Data Center. U.S. Climate Normals website, Climate of the
States, Climatography of the United States No. 60. http://hurricane.ncdc.noaa. gov/cgi-
bin/climatenormals/climatenormals.pl?directive=prod select2&prodtype=CLIM60&subr
num Date Last Accessed: 3/14/2014.
NJ DOT, 2011. New Jersey Department of Transportation. New Jersey's Roadway Mileage and
Daily VMT by Functional Classification Distributed by County. 2011 Data.
http://www.state.nj.us/transportation/refdata/roadwav/vmt.shtm Date Last Accessed:
8/12/2013.
NJ DOT, 2013. New Jersey Department of Transportation. Roadway Information and Traffic
Counts mapping application. 2009, 2011 Data.
http://www.state.nj.us/transportation/refdata/roadway/traffic counts/ Date Last Accessed:
8/26/2013.
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://dspace.nistatelib.org:8080/xmlui/bitstream/handle/10929/23355/f4912008d.pdf?seq
uence= 1 &isAllowed=y Date Last Accessed: 8/6/2014.
NOAA, 1999. NOAA, National Weather Service. April 1999. Miguel Miller. NOAA Technical
Memorandum, NWS-WR259, Climate of San Jose.
http://www.wrh.noaa.gov/mtr/sfd sic climate/sic/SJC CLIMATE3.php Date Last
Accessed: 8/4/2014.
NOAA, 2013. NOAA, National Weather Service. NWS glossary website.
http://weather.gov/glossary/ Date Last Accessed: 10/16/2013.
NOAA, 2014a. NOAA, Storm Prediction Center. Roger Edwards. The Online Tornado FAQ,
Frequently Asked Questions about Tornadoes website. Last modified 3/7/2014.
http://www.spc.noaa.gov/faq/tornado/ Date Last Accessed: 4/29/2014.
32-10
-------
NOAA, 2014b. NOAA, Climate Prediction Center. Climate Divisions with Counties website.
http://www.cpc.ncep.noaa.gov/products/analysis monitoring/regional monitoring/CLIM
_DIVS/states_counties_climate-divisions.shtml Date Last Accessed: 5/22/2014.
NYS DMV, 2012. New York State Department of Motor Vehicles. Vehicle Registrations in
Force (2012). http://www.dmv.ny.gov/stats.htmDate Last Accessed: 8/1/2013.
NYS DOT, 2011. New York State Department of Transportation. Traffic Data Viewer mapping
application. 2011 Data. https://www.dot.ny.gov/tdv Date Last Accessed: 8/27/2013.
NYS DOT, 2013. Email from Patrick Lentlie, New York State Department of Transportation,
Office of Environment. August 28, 2013. 2011 DVMT Data.
OK DEQ, 2013. Email from Randy Ward, Oklahoma Department of Environmental Quality.
August 29, 2013. ADOK site description additions.
OK DOT, 2011. Oklahoma Department of Transportation, Planning and Research Division. 2011
Annual Average Daily Traffic Oklahoma Highway System maps.
http://www.okladot.state.ok.us/maps/aadt/ Date Last Accessed: 8/27/2013.
OK DOT, 2013. Email from Ron Maxwell, Oklahoma Department of Transportation.
August 12, 2013. 2012 DVMT Data spreadsheet.
OKTC, 2012. Oklahoma Tax Commission, Motor Vehicle Division. Annual Vehicle
Registration Report, Fiscal Year July 1, 2011-June 30, 2012. Oklahoma City, Oklahoma.
http://www.tax.ok.gov/annrpts.html Date Last Accessed: 8/1/2013.
Pagano, P. and Gauvreau, K., 2000. Pagano, M. and Gauvreau, K. 2000. Principles of
Biostatistics, 2nd edition. Pacific Grove, CA: Duxbury.
Peltier, 2012. Peltier Techinical Services, Inc. Box and Whisker Plot Utility. Downloaded
April 5, 2012. http://peltiertech.com/Utilitv/
POLE, 2013. Port of Long Beach. About Us website, http://www.polb. com/about/default.asp
Date Last Accessed: 2/3/2014.
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/travelri/maps.asp
Date Last Accessed: 8/27/2013.
Rutgers, 2014. Office of the New Jersey State Climatologist, Rutgers University. New Jersey
Climate Overview website, http://climate.rutgers.edu/stateclimvl/njclimoverview.html
Date Last Accessed: 4/8/2014.
SC DMV, 2012. South Carolina Department of Motor Vehicles. Historical Information and
Statistics website. S.C. Registered Vehicles (12/2012).
http://www.scdmvonline.com/DMVNew/default.aspx?n=historical information
Date Last Accessed: 8/1/2013.
32-11
-------
SC DOT, 2012. South Carolina Department of Transportation. Interactive Traffic Count Map,
SCDOT Traffic Counts 2012. http://www.scdot.org/getting/annualtraffic.aspx
Date Last Accessed: 8/27/2013.
SC DOT, 2013. Email from Angela Hance, South Carolina Department of Transportation.
August 26, 2013. DVMT By County as of December 31, 2012.
SC SCO, 2014. South Carolina Department of Natural Resources, State Climate Office. South
Carolina Climate website.
http://www.dnr.sc.gov/climate/sco/ClimateData/cli sc climate.php Date Last Accessed:
5/9/2014.
SD DOR, 2012. South Dakota Department of Revenue, Motor Vehicle Division. Number of
Titles Awarded by Year by County website. 2012 Data.
http://www.state.sd.us/drr2/motorvehicle/title/history of titlesl2.htm Date Last
Accessed: 8/1/2013.
SD DOT, 2012. South Dakota Department of Transportation. Traffic website. Sioux Falls-
Brandon Map, 2012 Data.
http://www.sddot.com/transportation/highways/traffic/Default.aspx Date Last Accessed:
8/27/2013.
SD DOT, 2013. South Dakota Department of Transportation, Office of Transporation Inventory
Management. April 25, 2013. 2012 Vehicle Miles Travelled (VMT), All Vehicles by
County, http://sddot.com/transportation/highways/traffic/ Date Last Accessed: 8/12/2013.
TAMU, 2014. Office of the State Climatologist, Texas A&M University. John Nielsen-
Gammon. Climate of Texas, http://climatexas.tamu.edu/index.php/publications
Date Last Accessed: 5/13/2014.
TX DMV, 2013. Email from Diane Emrick-Dodson, Texas Department of Motor Vehicles,
Vehicle Titles and Registration Division. August 1, 2013. Vehicle Registration and
License Fees by County Fiscal Year September 1, 2011 to August 31, 2012 (spreadsheet).
TX DOT, 2011. Texas Department of Transportation. District Traffic Maps - 2011. Atlanta Base
Sheet, http://www.txdot.gov/inside-txdot/division/transportation-
planning/maps/traffic.html Date Last Accessed: 8/27/2013
TX DOT, 2012. Texas Department of Transportation. District and County Statistics (DISCOS)
website. Daily Vehicle Miles (9/1/11 through 8/31/12) for the Atlanta and Houston
Districts, http://www.txdot.gov/inside-txdot/division/fmance/discos.html Date Last
Accessed: 8/12/2013.
UT DOT, 2011. Utah Department of Transportation. AADT Traffic Map Updated Using 2011
Data (via Google Earth kmz file).
http://www.udot.utah.gov/main/f?p=100:pg:0::::V,T:,2256 Date Last Accessed:
8/27/2013.
32-12
-------
UT DOT, 2012. Utah Department of Transportation. September 2012. 2011 Vehicle Miles of
Travel (VMT) by Functional Class by County.
http://www.udot.utah.gov/main/f?p=100:pg:0::::V,T:,530 Date Last Accessed:
10/30/2012.
UT TC, 2012. Utah Tax Commission. Utah Current Registrations 2012, On-road Registrations
by County and Vehicle Type. Run date 1/31/12.
http://tax.utah.gov/econstats/mv/registrations-2010 Date Last Accessed: 8/1/2013.
VA DOT, 2012. Virginia Department of Transportation, Traffic Engineering Division. Average
Daily Traffic Volumes with Vehicle Classification Data on Interstate, Arterial, and
Primary Routes 2012. http://www.virginiadot.org/info/2012 traffic data.asp Date Last
Accessed: 8/27/2013.
VA DOT, 2013. Virginia Department of Transportation. August 2013. Daily Vehicle Miles
Traveled (DVMT) by Physical Jurisdiction, with Towns Combined into Counties
(VMT1206). 2012 data.
http://www.virginiadot.org/info/2012_traffic_data_daily_vehicle_miles_traveled.asp
Date Last Accessed: 9/17/2013.
VT DMV, 2012. Email from Judy Eastman, Vermont Department of Motor Vehicles.
October 19, 2012. 2012 Data.
Vtrans, 2013a. Vermont Agency of Transportation, Policy, Planning and Intermodal
Development Division, Traffic Research Unit. May 2013. 2012 (Route Log) AADTs
State Highways, http://vtransplanning.vermont.gov/research/traffic/publications
Date Last Accessed: 8/27/2013.
Vtrans, 2013b. Vermont Agency of Transportation, Highway Research. February 22, 2013. 2011
Miles and Annual Vehicle Miles Traveled (AVMT) by County and Functional Class.
http://www.aot.state.vt.us/planning/documents/highresearch/publications/pub.htm
Date Last Accessed: 8/12/2013.
WA DOL, 2012. Washington State Department of Licensing. Vehicle/Vessel Fee Collection and
Vehicle Counts Program website. (13) Motor Vehicle Registration by Class and County -
Calendar Year 2012. http://www.dol.wa.gov/about/vehvesselreports.html Date Last
Accessed: 8/1/2013.
WA DOT, 2012a. Washington State Department of Transportation. Statewide Travel and
Collision Data Office - Traffic Volume Map. 2012 Data.
http://www.wsdot.wa.gov/mapsdata/tools/traffictrends/ Date Last Accessed: 8/23/2013.
WA DOT, 2012b. Washington State Department of Transportation. 2012 State Highway
(Interstate, State Route and U.S. Route) Centerline Miles and Annual Vehicle Miles
Traveled (AVMT) in Thousands By Year, County, State Functional Class of Roadway
and Urban vs. Rural spreadsheet.
http://www.wsdot.wa.gov/mapsdata/travel/annualtrafficreport.htm Date Last Accessed:
8/12/2013.
32-13
-------
WI DOT, 2011. Wisconsin Department of Transportation. Roadrunner Interactive Mapping
Application. 2011 Data for HOWL
http://www.dot.wisconsin.gov/travel/counts/index.htm Date Last Accessed: 8/23/2013.
WI DOT, 2012a. Wisconsin Department of Transportation, Division of Motor Vehicles. 2012
Facts and Figures, A Reference Guide. Madison, WI.
http://www.dot.wisconsin.gov/drivers/facts.htm Date Last Accessed: 8/1/2013.
WI DOT, 2012b. Wisconsin Department of Transportation. September 2012. 2011 Vehicle Miles
of Travel (VMT) by County, http://www.dot.wisconsin.gov/travel/counts/vmt.htm
Date Last Accessed: 10/31/2012.
WI DOT, 2013. Wisconsin Department of Transportation. Roadrunner Interactive Mapping
Application. 2013 Preliminary Data for MIWI.
http://www.dot.wisconsin.gov/travel/counts/index.htm Date Last Accessed: 8/23/2013.
WI SCO, 2013a. Wisconsin State Climatology Office, University of Wisconsin-Madison.
Climate of Wisconsin webpage. http://www.aos.wisc.edu/~sco/clim-
history/state/stateclimate.html Date Last Accessed: 8/30/2013.
WI SCO, 2013b. Wisconsin State Climatology Office. Number of days of 1 inch or more of
snow cover per season: Milwaukee Airport (1948/49-2005/06 season) graphic.
http://www.aos.wisc.edu/~sco/clim-history/7cities/milwaukee.html Date Last Accessed:
8/30/2013.
Wood, 2004. Wood, Richard A. 1992. The Weather Almanac, 11th ed. The Gale Group, Inc.
Farmington Hills, MI.
WRCC, 2013. Western Region Climate Center. Climate Narrative of the States: Climate of
Arizona, California, Colorado, Utah, http://www.wrcc.dri.edu/CLEVIATEDATA.html
Date Last Accessed: 9/30/2013.
32-14
-------
United States Office of Air Quality Planning and Standards Publication No. EPA-454/R-14-006a
Environmental Protection Air Quality Assessment Division September 2014
Agency Research Triangle Park, NC
------- |