rn* UnitodSMHt
^*5#	Environmental Protection
^ I	Agsncy
2013 National Monitoring Programs Annual
Report (UATMP, NATTS, CSATAM)
October 2015
Final Report

-------
EPA-454/R-15-005a
October 2015
2013 National Monitoring Programs Annual Report (UATMP, NATTS, CSATAM)
By:
Eastern Research Group, Inc.
Morrisville, NC 27560
Prepared for:
Jeff Yane 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 46, 47, 48, 52, 55, 56, 60, 61, 62, & 63
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Air Quality Assessment Division
Research Triangle Park, NC 27711

-------
2013 National Monitoring Programs
Annual Report
(UATMP, NATTS, and CSATAM)
Final Report
EPA Contract No. EP-D-09-048
Delivery Orders 46, 47, 48, 52, 55, 56, 60, 61, 62, & 63
Prepared for:
Jeff Yane and David Shelow
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Prepared by:
Eastern Research Group, Inc.
601 Keystone Park Drive, Suite 700
Morrisville, NC 27560
October 2015

-------
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	xxxvii
List of Acronyms	xlvii
Abstract	xlix
1.0 Introduction	1-1
1.1	Background	1-1
1.2	The Report	1-2
2.0 The 2013 National Monitoring Programs Network	2-1
2.1	Monitoring Locations	2-1
2.2	Analytical Methods and Pollutants Targeted for Monitoring	2-14
2.2.1	VOC and SNMOC Concurrent Sampling and Analytical Methods ... 2-16
2.2.2	Carbonyl Compound Sampling and Analytical Method	2-20
2.2.3	PAH Sampling and Analytical Method	2-21
2.2.4	Metals Sampling and Analytical Method	2-23
2.2.5	Hexavalent Chromium Sampling and Analytical Method	2-24
2.3	Sample Collection Schedules	2-25
2.4	Completeness	2-32
3.0 Summary of the 2013 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	3-3
3.3	Additional Program-Level Analyses of the 2013 National Monitoring
Programs Dataset	3-7
3.3.1	The Contribution from Mobile Source Emissions on Spatial
Variations	3-7
3.3.2	Variability Analyses	3-9
3.3.3	Greenhouse Gas Assessment	3-9
in

-------
TABLE OF CONTENTS (Continued)
Page
3.4 Additional Site-Specific Analyses	3-10
3.4.1	Site Characterization	3-10
3.4.2	Meteorological Analysis	3-10
3.4.3	Preliminary Risk-Based Screening and Pollutants of Interest	3-12
3.4.3.1	Site-Specific Comparison to Program-level Average
Concentrations	3-12
3.4.3.2	Site Trends Analysis	3-13
3.4.3.3	Cancer Risk and Noncancer Hazard Approximations	3-14
3.4.3.4	Risk-Based Emissions Assessment	3-15
4.0 Summary of the 2013 National Monitoring Programs Data	4-1
4.1	Statistical Results	4-1
4.1.1	Target Pollutant Detection Rates	4-1
4.1.2	Concentration Range and Data Distribution	4-15
4.1.3	Central Tendency	4-15
4.2	Preliminary Risk-Based Screening and Pollutants of Interest	4-17
4.2.1 Concentrations of the Pollutants of Interest	4-23
4.3	The Contribution from Mobile Sources	4-30
4.3.1	Mobile Source Emissions	4-30
4.3.2	Hydrocarbon Concentrations	4-32
4.3.3	Motor Vehicle Ownership	4-34
4.3.4	Estimated Traffic Volume	4-35
4.3.5	Vehicle Miles Traveled	4-36
4.4	Variability Analysis	4-37
4.4.1	Inter-site Variability	4-37
4.4.2	Quarterly Variability Analysis	4-59
4.5	Greenhouse Gases from Method TO-15	4-82
5.0 Site in Alaska	5-1
5.1	Site Characterization	5-1
5.2	Meteorological Characterization	5-6
5.2.1	Climate Summary	5-6
5.2.2	Meteorological Summary	5-6
5.2.3	Wind Rose Comparison	5-8
iv

-------
TABLE OF CONTENTS (Continued)
Page
5.3	Pollutants of Interest	5-10
5.4	Concentrations	5-12
5.4.1	2013 Concentration Averages	5-12
5.4.2	Concentration Comparison	5-14
5.4.3	Concentration Trends	5-18
5.5	Additional Risk-Based Screening Evaluations	5-18
5.5.1	Cancer Risk and Noncancer Hazard Approximations	5-18
5.5.2	Risk-Based Emissions Assessment	5-20
5.6	Summary of the 2013 Monitoring Data for ANAK	5-24
6.0 Sites in Arizona	6-1
6.1	Site Characterization	6-1
6.2	Meteorological Characterization	6-7
6.2.1	Climate Summary	6-7
6.2.2	Meteorological Summary	6-7
6.2.3	Wind Rose Comparison	6-9
6.3	Pollutants of Interest	6-13
6.4	Concentrations	6-15
6.4.1	2013 Concentration Averages	6-15
6.4.2	Concentration Comparison	6-19
6.4.3	Concentration Trends	6-25
6.5	Additional Risk-Based Screening Evaluations	6-42
6.5.1	Cancer Risk and Noncancer Hazard Approximations	6-42
6.5.2	Risk-Based Emissions Assessment	6-44
6.6	Summary of the 2013 Monitoring Data for PXSS and SPAZ	6-48
7.0 Sites in California	7-1
7.1	Site Characterization	7-1
7.2	Meteorological Characterization	7-12
7.2.1	Climate Summary	7-13
7.2.2	Meteorological Summary	7-13
v

-------
TABLE OF CONTENTS (Continued)
Page
7.2.3 Wind Rose Comparison	7-15
7.3	Pollutants of Interest	7-22
7.4	Concentrations	7-24
7.4.1	2013 Concentration Averages	7-24
7.4.2	Concentration Comparison	7-28
7.4.3	Concentration Trends	7-31
7.5	Additional Risk-Based Screening Evaluations	7-40
7.5.1	Cancer Risk and Noncancer Hazard Approximations	7-40
7.5.2	Risk-Based Emissions Assessment	7-42
7.6	Summary of the 2013 Monitoring Data for the California Monitoring Sites .... 7-48
8.0 Sites in Colorado	8-1
8.1	Site Characterization	8-1
8.2	Meteorological Characterization	8-15
8.2.1	Climate Summary	8-15
8.2.2	Meteorological Summary	8-15
8.2.3	Wind Rose Comparison	8-19
8.3	Pollutants of Interest	8-27
8.4	Concentrations	8-30
8.4.1	2013 Concentration Averages	8-30
8.4.2	Concentration Comparison	8-35
8.4.3	Concentration Trends	8-46
8.5	Additional Risk-Based Screening Evaluations	8-71
8.5.1	Cancer Risk and Noncancer Hazard Approximations	8-71
8.5.2	Risk-Based Emissions Assessment	8-76
8.6	Summary of the 2013 Monitoring Data for the Colorado Monitoring Sites	8-84
9.0 Site in the District of Columbia	9-1
9.1	Site Characterization	9-1
9.2	Meteorological Characterization	9-6
9.2.1 Climate Summary	9-6
vi

-------
TABLE OF CONTENTS (Continued)
Page
9.2.2	Meteorological Summary	9-6
9.2.3	Wind Rose Comparison	9-8
9.3	Pollutants of Interest	9-10
9.4	Concentrations	9-11
9.4.1	2013 Concentration Averages	9-11
9.4.2	Concentration Comparison	9-12
9.4.3	Concentration Trends	9-13
9.5	Additional Risk-Based Screening Evaluations	9-14
9.5.1	Cancer Risk and Noncancer Hazard Approximations	9-14
9.5.2	Risk-Based Emissions Assessment	9-15
9.6	Summary of the 2013 Monitoring Data for WADC	9-19
10.0 Sites in Florida	10-1
10.1	Site Characterization	10-1
10.2	Meteorological Characterization	10-16
10.2.1	Climate Summary	10-16
10.2.2	Meteorological Summary	10-17
10.2.3	Wind Rose Comparison	10-20
10.3	Pollutants of Interest	10-29
10.4	Concentrations	10-31
10.4.1	2013 Concentration Averages	10-32
10.4.2	Concentration Comparison	10-37
10.4.3	Concentration Trends	10-42
10.5	Additional Risk-Based Screening Evaluations	10-55
10.5.1	Cancer Risk and Noncancer Hazard Approximations	10-55
10.5.2	Risk-Based Emissions Assessment	10-57
10.6	Summary of the 2013 Monitoring Data for the Florida Monitoring Sites	10-66
11.0 Site in Georgia	11-1
11.1 Site Characterization	11-1
vii

-------
TABLE OF CONTENTS (Continued)
Page
11.2	Meteorological Characterization	11-6
11.2.1	Climate Summary	11-6
11.2.2	Meteorological Summary	11-6
11.2.3	Wind Rose Comparison	11-8
11.3	Pollutants of Interest	11-10
11.4	Concentrations	11-11
11.4.1	2013 Concentration Averages	11-11
11.4.2	Concentration Trends	11-13
11.5	Additional Risk-Based Screening Evaluations	11-14
11.5.1 Risk-Based Emissions Assessment	11-14
11.6	Summary of the 2013 Monitoring Data for SDGA	11-18
12.0 Sites in Illinois	12-1
12.1	Site Characterization	12-1
12.2	Meteorological Characterization	12-10
12.2.1	Climate Summary	12-10
12.2.2	Meteorological Summary	12-11
12.2.3	Wind Rose Comparison	12-13
12.3	Pollutants of Interest	12-18
12.4	Concentrations	12-21
12.4.1	2013 Concentration Averages	12-21
12.4.2	Concentration Comparison	12-27
12.4.3	Concentration Trends	12-36
12.5	Additional Risk-Based Screening Evaluations	12-58
12.5.1	Cancer Risk and Noncancer Hazard Approximations	12-58
12.5.2	Risk-Based Emissions Assessment	12-61
12.6	Summary of the 2013 Monitoring Data for NBIL, SPIL, and ROIL	12-68
13.0 Sites in Indiana	13-1
13.1 Site Characterization	13-1
viii

-------
TABLE OF CONTENTS (Continued)
Page
13.2	Meteorological Characterization	13-8
13.2.1	Climate Summary	13-8
13.2.2	Meteorological Summary	13-9
13.2.3	Wind Rose Comparison	13-11
13.3	Pollutants of Interest	13-15
13.4	Concentrations	13-16
13.4.1	2013 Concentration Averages	13-16
13.4.2	Concentration Comparison	13-18
13.4.3	Concentration Trends	13-19
13.5	Additional Risk-Based Screening Evaluations	13-25
13.5.1	Cancer Risk and Noncancer Hazard Approximations	13-25
13.5.2	Risk-Based Emissions Assessment	13-26
13.6	Summary of the 2013 Monitoring Data for INDEM and WPIN	13-30
14.0 Sites in Kentucky	14-1
14.1	Site Characterization	14-1
14.2	Meteorological Characterization	14-22
14.2.1	Climate Summary	14-22
14.2.2	Meteorological Summary	14-22
14.2.3	Wind Rose Comparison	14-26
14.3	Pollutants of Interest	14-40
14.4	Concentrations	14-47
14.4.1	2013 Concentration Averages	14-48
14.4.2	Concentration Comparison	14-60
14.4.3	Concentration Trends	14-74
14.5	Additional Risk-Based Screening Evaluations	14-74
14.5.1	Cancer Risk and Noncancer Hazard Approximations	14-74
14.5.2	Risk-Based Emissions Assessment	14-82
14.6 Summary of the 2013 Monitoring Data for the Kentucky Monitoring Sites... 14-98
IX

-------
TABLE OF CONTENTS (Continued)
Page
15.0 Site in Massachusetts	15-1
15.1	Site Characterization	15-1
15.2	Meteorological Characterization	15-6
15.2.1	Climate Summary	15-6
15.2.2	Meteorological Summary	15-6
15.2.3	Wind Rose Comparison	15-8
15.3	Pollutants of Interest	15-10
15.4	Concentrations	15-11
15.4.1	2013 Concentration Averages	15-12
15.4.2	Concentration Comparison	15-13
15.4.3	Concentration Trends	15-15
15.5	Additional Risk-Based Screening Evaluations	15-19
15.5.1	Cancer Risk and Noncancer Hazard Approximations	15-19
15.5.2	Risk-Based Emissions Assessment	15-20
15.6	Summary of the 2013 Monitoring Data for BOMA	15-23
16.0 Site in Michigan	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	Wind Rose Comparison	16-8
16.3	Pollutants of Interest	16-10
16.4	Concentrations	16-12
16.4.1	2013 Concentration Averages	16-12
16.4.2	Concentration Comparison	16-14
16.4.3	Concentration Trends	16-20
16.5	Additional Risk-Based Screening Evaluations	16-33
16.5.1	Cancer Risk and Noncancer Hazard Approximations	16-33
16.5.2	Risk-Based Emissions Assessment	16-34
x

-------
TABLE OF CONTENTS (Continued)
Page
16.6 Summary of the 2013 Monitoring Data for DEMI	16-38
17.0 Site in Minnesota	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	Wind Rose Comparison	17-8
17.3	Pollutants of Interest	17-10
17.4	Concentrations	17-11
17.4.1 2013 Concentration Averages	17-11
17.5	Additional Risk-Based Screening Evaluations	17-12
17.5.1 Risk-Based Emissions Assessment	17-13
17.6	Summary of the 2013 Monitoring Data for STMN	17-17
18.0 Sites in Mississippi	18-1
18.1	Site Characterization	18-1
18.2	Meteorological Characterization	18-7
18.2.1	Climate Summary	18-7
18.2.2	Meteorological Summary	18-8
18.2.3	Wind Rose Comparison	18-8
18.3	Pollutants of Interest	18-13
18.4	Concentrations	18-15
18.4.1	2013 Concentration Averages	18-15
18.4.2	Concentration Comparison	18-18
18.4.3	Concentration Trends	18-24
18.5	Additional Risk-Based Screening Evaluations	18-24
18.5.1	Cancer Risk and Noncancer Hazard Approximations	18-24
18.5.2	Risk-Based Emissions Assessment	18-27
18.6	Summary of the 2013 Monitoring Data for KMMS and SSMS	18-31
XI

-------
TABLE OF CONTENTS (Continued)
Page
19.0 Site in Missouri	19-1
19.1	Site Characterization	19-1
19.2	Meteorological Characterization	19-6
19.2.1	Climate Summary	19-6
19.2.2	Meteorological Summary	19-6
19.2.3	Wind Rose Comparison	19-8
19.3	Pollutants of Interest	19-10
19.4	Concentrations	19-12
19.4.1	2013 Concentration Averages	19-12
19.4.2	Concentration Comparison	19-17
19.4.3	Concentration Trends	19-25
19.5	Additional Risk-Based Screening Evaluations	19-42
19.5.1	Cancer Risk and Noncancer Hazard Approximations	19-42
19.5.2	Risk-Based Emissions Assessment	19-44
19.6	Summary of the 2013 Monitoring Data for S4MO	19-48
20.0 Sites in New Jersey	20-1
20.1	Site Characterization	20-1
20.2	Meteorological Characterization	20-12
20.2.1	Climate Summary	20-13
20.2.2	Meteorological Summary	20-13
20.2.3	Wind Rose Comparison	20-15
20.3	Pollutants of Interest	20-21
20.4	Concentrations	20-24
20.4.1	2013 Concentration Averages	20-25
20.4.2	Concentration Comparison	20-31
20.4.3	Concentration Trends	20-41
20.5	Additional Risk-Based Screening Evaluations	20-69
20.5.1	Cancer Risk and Noncancer Hazard Approximations	20-69
20.5.2	Risk-Based Emissions Assessment	20-73
xii

-------
TABLE OF CONTENTS (Continued)
Page
20.6 Summary of the 2013 Monitoring Data for the New Jersey Monitoring
Sites	20-79
21.0 Sites in New York	21-1
21.1	Site Characterization	21-1
21.2	Meteorological Characterization	21-9
21.2.1	Climate Summary	21-9
21.2.2	Meteorol ogi cal Summary	21-10
21.2.3	Wind Rose Comparison	21-12
21.3	Pollutants of Interest	21-16
21.4	Concentrations	21-17
21.4.1	2013 Concentration Averages	21-17
21.4.2	Concentration Comparison	21-19
21.4.3	Concentration Trends	21-22
21.5	Additional Risk-Based Screening Evaluations	21-26
21.5.1	Cancer Risk and Noncancer Hazard Approximations	21 -27
21.5.2	Risk-Based Emissions Assessment	21-28
21.6	Summary of the 2013 Monitoring Data for BXNY and ROCH	21-32
22.0 Sites in Oklahoma	22-1
22.1	Site Characterization	22-1
22.2	Meteorological Characterization	22-14
22.2.1	Climate Summary	22-14
22.2.2	Meteorological Summary	22-14
22.2.3	Wind Rose Comparison	22-18
22.3	Pollutants of Interest	22-26
22.4	Concentrations	22-31
22.4.1	2013 Concentration Averages	22-31
22.4.2	Concentration Comparison	22-40
22.4.3	Concentration Trends	22-52
Xlll

-------
TABLE OF CONTENTS (Continued)
Page
22.5	Additional Risk-Based Screening Evaluations	22-84
22.5.1	Cancer Risk and Noncancer Hazard Approximations	22-84
22.5.2	Risk-Based Emissions Assessment	22-88
22.6	Summary of the 2013 Monitoring Data for the Oklahoma Monitoring Sites.. 22-97
23.0 Site in Rhode Island	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	Wind Rose Comparison	23-8
23.3	Pollutants of Interest	23-10
23.4	Concentrations	23-11
23.4.1	2013 Concentration Averages	23-12
23.4.2	Concentration Comparison	23-13
23.4.3	Concentration Trends	23-13
23.5	Additional Risk-Based Screening Evaluations	23-15
23.5.1	Cancer Risk and Noncancer Hazard Approximations	23-15
23.5.2	Risk-Based Emissions Assessment	23-16
23.6	Summary of the 2013 Monitoring Data for PRRI	23-19
24.0 Site in South Carolina	24-1
24.1	Site Characterization	24-1
24.2	Meteorological Characterization	24-6
24.2.1	Climate Summary	24-6
24.2.2	Meteorological Summary	24-6
24.2.3	Wind Rose Comparison	24-8
24.3	Pollutants of Interest	24-10
24.4	Concentrations	24-10
24.4.1	2013 Concentration Averages	24-11
24.4.2	Concentration Comparison	24-12
xiv

-------
TABLE OF CONTENTS (Continued)
Page
24.4.3 Concentration Trends	24-13
24.5	Additional Risk-Based Screening Evaluations	24-14
24.5.1	Cancer Risk and Noncancer Hazard Approximations	24-14
24.5.2	Risk-Based Emissions Assessment	24-15
24.6	Summary of the 2013 Monitoring Data for CHSC	24-19
25.0 Sites in Texas	25-1
25.1	Site Characterization	25-1
25.2	Meteorological Characterization	25-9
25.2.1	Climate Summary	25-9
25.2.2	Meteorological Summary	25-9
25.2.3	Wind Rose Comparison	25-11
25.3	Pollutants of Interest	25-15
25.4	Concentrations	25-16
25.4.1 2013 Concentration Averages	25-16
25.5	Additional Risk-Based Screening Evaluations	25-17
25.5.1 Risk-Based Emissions Assessment	25-18
25.6	Summary of the 2013 Monitoring Data for CAMS 35 and CAMS 85	25-21
26.0 Site in Utah	26-1
26.1	Site Characterization	26-1
26.2	Meteorological Characterization	26-6
26.2.1	Climate Summary	26-6
26.2.2	Meteorological Summary	26-6
26.2.3	Wind Rose Comparison	26-8
26.3	Pollutants of Interest	26-10
26.4	Concentrations	26-12
26.4.1	2013 Concentration Averages	26-12
26.4.2	Concentration Comparison	26-16
26.4.3	Concentration Trends	26-24
xv

-------
TABLE OF CONTENTS (Continued)
Page
26.5	Additional Risk-Based Screening Evaluations	26-39
26.5.1	Cancer Risk and Noncancer Hazard Approximations	26-39
26.5.2	Risk-Based Emissions Assessment	26-43
26.6	Summary of the 2013 Monitoring Data for BTUT	26-47
27.0 Sites in Vermont	27-1
27.1	Site Characterization	27-1
27.2	Meteorological Characterization	27-10
27.2.1	Climate Summary	27-10
27.2.2	Meteorological Summary	27-10
27.2.3	Wind Rose Comparison	27-13
27.3	Pollutants of Interest	27-18
27.4	Concentrations	27-20
27.4.1	2013 Concentration Averages	27-20
27.4.2	Concentration Comparison	27-25
27.4.3	Concentration Trends	27-31
27.5	Additional Risk-Based Screening Evaluations	27-48
27.5.1	Cancer Risk and Noncancer Hazard Approximations	27-48
27.5.2	Risk-Based Emissions Assessment	27-50
27.6	Summary of the 2013 Monitoring Data for the Vermont Monitoring Sites .... 27-56
28.0 Site in Virginia	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	Wind Rose Comparison	28-8
28.3	Pollutants of Interest	28-10
28.4	Concentrations	28-11
28.4.1	2013 Concentration Averages	28-11
28.4.2	Concentration Comparison	28-13
xvi

-------
TABLE OF CONTENTS (Continued)
Page
28.4.3 Concentration Trends	28-13
28.5	Additional Risk-Based Screening Evaluations	28-15
28.5.1	Cancer Risk and Noncancer Hazard Approximations	28-15
28.5.2	Risk-Based Emissions Assessment	28-16
28.6	Summary of the 2013 Monitoring Data for RIVA	28-19
29.0 Site in Washington	29-1
29.1	Site Characterization	29-1
29.2	Meteorological Characterization	29-6
29.2.1	Climate Summary	29-6
29.2.2	Meteorological Summary	29-6
29.2.3	Wind Rose Comparison	29-8
29.3	Pollutants of Interest	29-10
29.4	Concentrations	29-11
29.4.1	2013 Concentration Averages	29-12
29.4.2	Concentration Comparison	29-15
29.4.3	Concentration Trends	29-20
29.5	Additional Risk-Based Screening Evaluations	29-30
29.5.1	Cancer Risk and Noncancer Hazard Approximations	29-30
29.5.2	Risk-Based Emissions Assessment	29-31
29.6	Summary of the 2013 Monitoring Data for SEWA	29-35
30.0 Sites in Wisconsin	30-1
30.1	Site Characterization	30-1
30.2	Meteorological Characterization	30-9
30.2.1	Climate Summary	30-9
30.2.2	Meteorological Summary	30-10
30.2.3	Wind Rose Comparison	30-12
30.3	Pollutants of Interest	30-15
xvii

-------
TABLE OF CONTENTS (Continued)
Page
30.4	Concentrations	30-16
30.4.1 2013 Concentration Averages	30-17
30.5	Additional Risk-Based Screening Evaluations	30-18
30.5.1 Risk-Based Emissions Assessment	30-18
30.6	Summary of the 2013 Monitoring Data for HOWI and MIWI	30-22
31.0 Data Quality	31-1
31.1	Completeness	31-1
31.2	Method Preci si on	31-2
31.2.1	VOC Method Precision	31-5
31.2.2	SNMOC Method Precision	31-14
31.2.3	Carbonyl Compound Method Precision	31-17
31.2.4	PAH Method Precision	31-21
31.2.5	Metals Method Precision	31-23
31.2.6	Hexavalent Chromium Method Precision	31-25
31.3	Analytical Precision	31-26
31.3.1	VOC Analytical Precision	31-28
31.3.2	SNMOC Analytical Precision	31-39
31.3.3	Carbonyl Compound Analytical Precision	31-46
31.3.4	PAH Analytical Precision	31-52
31.3.5	Metal s Analy ti cal Preci si on	31-57
31.3.6	Hexavalent Chromium Analytical Precision	31-61
31.4	Accuracy	31-63
32.0 Results, Conclusions, and Recommendations	32-1
32.1 Summary of Results	32-1
32.1.1	National-level Results Summary	32-1
32.1.2	State-level Results Summary	32-2
xviii

-------
TABLE OF CONTENTS (Continued)
Page
32.1.3	Composite Site-level Results Summary	32-22
32.1.4	Data Quality Results Summary	32-28
32.2	Conclusions	32-28
32.3	Recommendations	32-31
33.0 References	33-1
xix

-------
List of Appendices
Appendix A	AQS Site Descriptions for the 2013 NMP Monitoring Sites
Appendix B	Program Method Detection Limits (MDLs)
Appendix C	2013 VOC Raw Data
Appendix D	2013 SNMOC Raw Data
Appendix E	2013 Carbonyl Compounds Raw Data
Appendix F	2013 PAH and PAH/Phenol Raw Data
Appendix G	2013 Metals Raw Data
Appendix H	2013 Hexavalent Chromium Raw Data
Appendix I	Summary of Invalidated 2013 Samples
Appendix J	2013 Summary Statistics for VOC Monitoring
Appendix K	2013 Summary Statistics for SNMOC Monitoring
Appendix L	2013 Summary Statistics for Carbonyl Compounds Monitoring
Appendix M	2013 Summary Statistics for PAH and PAH/Phenol Monitoring
Appendix N	2013 Summary Statistics for Metals Monitoring
Appendix O	2013 Summary Statistics for Hexavalent Chromium Monitoring
Appendix P	Glossary of Terms
Appendix Q	Risk Factors Used Throughout the 2013 NMP Report
xx

-------
LIST OF FIGURES
Page
2-1 Locations of the 2013 National Monitoring Programs Monitoring Sites	2-3
4-1 Inter-Site Variability for Acenaphthene	4-43
4-2 Inter-Site Variability for Acetaldehyde	4-44
4-3 Inter-Site Variability for Arsenic	4-45
4-4a Inter-Site Variability for Benzene - Method TO-15	4-46
4-4b Inter-Site Variability for Benzene - SNMOC	4-47
4-5a Inter-Site Variability for 1,3-Butadiene - Method TO-15	4-48
4-5b Inter-Site Variability for 1,3-Butadiene - SNMOC	4-49
4-6 Inter-Site Variability for Carbon Tetrachloride	4-50
4-7 Inter-Site Variability for/?-Dichlorobenzene	4-51
4-8 Inter-Site Variability for 1,2-Dichloroethane	4-52
4-9a Inter-Site Variability for Ethylbenzene - Method TO-15	4-53
4-9b Inter-Site Variability for Ethylbenzene - SNMOC	4-54
4-10 Inter-Site Variability for Formaldehyde	4-55
4-11 Inter-Site Variability for Hexachloro-1,3-butadiene	4-56
4-12 Inter-Site Variability for Naphthalene	4-57
4-13 Inter-Site Variability for Nickel	4-58
4-14 Comparison of Average Quarterly Acenaphthene Concentrations	4-64
4-15 Comparison of Average Quarterly Acetaldehyde Concentrations	4-65
4-16a Comparison of Average Quarterly Arsenic (PMio) Concentrations	4-66
4-16b Comparison of Average Quarterly Arsenic (TSP) Concentrations	4-67
4-17a Comparison of Average Quarterly Benzene (Method TO-15) Concentrations	4-68
4-17b Comparison of Average Quarterly Benzene (SNMOC) Concentrations	4-69
4-18a	Comparison of Average Quarterly 1,3-Butadiene (Method TO-15) Concentrations .... 4-70
4-18b Comparison of Average Quarterly 1,3-Butadiene (SNMOC) Concentrations	4-71
4-19 Comparison of Average Quarterly Carbon Tetrachloride Concentrations	4-72
4-20 Comparison of Average Quarterly p-Dichlorobenzene Concentrations	4-73
4-21 Comparison of Average Quarterly 1,2-Dichloroethane Concentrations	4-74
4-22a Comparison of Average Quarterly Ethylbenzene (Method TO-15) Concentrations	4-75
4-22b Comparison of Average Quarterly Ethylbenzene (SNMOC) Concentrations	4-76
4-23 Comparison of Average Quarterly Formaldehyde Concentrations	4-77
4-24 Comparison of Average Quarterly Hexachloro-1,3-Butadiene Concentrations	4-78
4-25 Comparison of Average Quarterly Naphthalene Concentrations	4-79
4-26a Comparison of Average Quarterly Nickel (PMio) Concentrations	4-80
4-26b	Comparison of Average Quarterly Nickel (TSP) Concentrations	4-81
5-1	Anchorage, Alaska (ANAK) Monitoring Site	5-2
5-2 NEI Point Sources Located Within 10 Miles of ANAK	5-3
5-3 Wind Roses for the Merrill Field Airport Weather Station near ANAK	5-9
5-4 Program vs. Site-Specific Average Benzene Concentration	5-14
5-5 Program vs. Site-Specific Average 1,3-Butadiene Concentration	5-14
5-6 Program vs. Site-Specific Average Carbon Tetrachloride Concentration	5-15
5-7 Program vs. Site-Specific Average/>-Dichlorobenzene Concentration	5-15
5-8 Program vs. Site-Specific Average 1,2-Dichloroethane Concentration	5-15
xxi

-------
LIST OF FIGURES (Continued)
Page
5-9 Program vs. Site-Specific Average Ethylbenzene Concentration	5-16
5-10 Program vs. Site-Specific Average Hexachloro-1,3-Butadiene Concentration	5-16
5-11	Program vs. Site-Specific Average Naphthalene Concentration	5-16
6-1	Phoenix, Arizona (PXSS) Monitoring Site	6-2
6-2	South Phoenix, Arizona (SPAZ) Monitoring Site	6-3
6-3	NEI Point Sources Located Within 10 Miles of PXSS and SPAZ	6-4
6-4	Wind Roses for the Phoenix Sky Harbor International Airport Weather Station
near PXSS	6-11
6-5 Wind Roses for the Phoenix Sky Harbor International Airport Weather Station
near SPAZ	6-12
6-6 Program vs. Site-Specific Average Acetaldehyde Concentration	6-19
6-7 Program vs. Site-Specific Average Arsenic (PMio) Concentration	6-19
6-8 Program vs. Site-Specific Average Benzene Concentrations	6-20
6-9 Program vs. Site-Specific Average 1,3-Butadiene Concentrations	6-20
6-10 Program vs. Site-Specific Average Carbon Tetrachloride Concentrations	6-21
6-11 Program vs. Site-Specific Average />-Dichlorobenzene Concentrations	6-21
6-12 Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations	6-22
6-13 Program vs. Site-Specific Average Ethylbenzene Concentrations	6-22
6-14 Program vs. Site-Specific Average Formaldehyde Concentration	6-22
6-15 Program vs. Site-Specific Average Naphthalene Concentration	6-23
6-16 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at PXSS	6-26
6-17 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at PXSS	6-27
6-18 Yearly Statistical Metrics for Benzene Concentrations Measured at PXSS	6-28
6-19 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at PXSS	6-29
6-20 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
PXSS	6-30
6-21 Yearly Statistical Metrics for /;-Dichlorobenzene Concentrations Measured at
PXSS	6-31
6-22 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
PXSS	6-32
6-23 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at PXSS	6-33
6-24 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at PXSS	6-34
6-25 Yearly Statistical Metrics for Naphthalene Concentrations Measured at PXSS	6-35
6-26 Yearly Statistical Metrics for Benzene Concentrations Measured at SPAZ	6-36
6-27 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at SPAZ	6-37
6-28 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
SPAZ	6-38
6-29 Yearly Statistical Metrics for /;-Dichlorobenzene Concentrations Measured at
SPAZ	6-39
6-30 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
SPAZ	6-40
6-31	Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at SPAZ	6-41
7-1	Los Angeles, California (CELA) Monitoring Site	7-2
7-2 Long Beach, California (LBHCA) Monitoring Site	7-3
xxii

-------
LIST OF FIGURES (Continued)
Page
7-3	NEI Point Sources Located Within 10 Miles of CELA and LBHCA	7-4
7-4	Rubidoux, California (RUCA) Monitoring Site	7-5
7-5	NEI Point Sources Located Within 10 Miles of RUCA	7-6
7-6	San Jose, California (SJJCA) Monitoring Site	7-7
7-7	NEI Point Sources Located Within 10 Miles of SJJCA	7-8
7-8	Wind Roses for the Downtown Los Angeles/USC Campus Weather Station
near CELA	7-17
7-9	Wind Roses for the Long Beach/Daugherty Field Airport Weather Station
near LBHCA	7-18
7-10	Wind Roses for the Riverside Municipal Airport Weather Station near RUCA	7-19
7-11	Wind Roses for the San Jose International Airport Weather Station near SJJCA	7-20
7-12	Program vs. Site-Specific Average Acenaphthene Concentration	7-28
7-13	Program vs. Site-Specific Average Arsenic (PMio) Concentration	7-28
7-14	Program vs. Site-Specific Average Benzo(a)pyrene Concentration	7-29
7-15	Program vs. Site-Specific Average Fluorene Concentration	7-29
7-16	Program vs. Site-Specific Average Naphthalene Concentrations	7-30
7-17	Program vs. Site-Specific Average Nickel (PMio) Concentration	7-30
7-18	Yearly Statistical Metrics for Acenaphthene Concentrations Measured at CELA	7-32
7-19	Yearly Statistical Metrics for Benzo(a)pyrene Concentrations Measured at CELA	7-33
7-20	Yearly Statistical Metrics for Fluorene Concentrations Measured at CELA	7-34
7-21	Yearly Statistical Metrics for Naphthalene Concentrations Measured at CELA	7-35
7-22	Yearly Statistical Metrics for Naphthalene Concentrations Measured at RUCA	7-36
7-23	Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at SJJCA	7-37
7-24	Yearly Statistical Metrics for Naphthalene Concentrations Measured at SJJCA	7-38
7-25	Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at SJJCA	7-39
8-1	Grand Junction, Colorado (GPCO) Monitoring Site	8-2
8-2	NEI Point Sources Located Within 10 Miles of GPCO	8-3
8-3	Battlement Mesa, Colorado (BMCO) Monitoring Site	8-4
8-4	Silt, Colorado (BRCO) Monitoring Site	8-5
8-5	Parachute, Colorado (PACO) Monitoring Site	8-6
8-6	Rifle, Colorado (RICO) Monitoring Site	8-7
8-7	NEI Point Sources Located Within 10 Miles of BMCO, BRCO, PACO, and RICO	8-8
8-8	Carbondale, Colorado (RFCO) Monitoring Site	8-9
8-9	NEI Point Sources Located Within 10 Miles of RFCO	8-10
8-10	Wind Roses for the Grand Junction Regional Airport Weather Station near GPCO .... 8-20
8-11	Wind Roses for the Garfield County Regional Airport Weather Station near BMCO.. 8-21
8-12	Wind Roses for the Garfield County Regional Airport Weather Station near BRCO... 8-22
8-13	Wind Roses for the Garfield County Regional Airport Weather Station near PACO... 8-23
8-14	Wind Roses for the Garfield County Regional Airport Weather Station near RICO.... 8-24
8-15	Wind Roses for the Aspen-Pitkin County Airport Weather Station near RFCO	8-25
8-16	Program vs. Site-Specific Average Acenaphthene Concentration	8-36
8-17	Program vs. Site-Specific Average Acetaldehyde Concentrations	8-36
8-18a	Program vs. Site-Specific Average Benzene (Method TO-15) Concentration	8-37
8-18b	Program vs. Site-Specific Average Benzene (SNMOC) Concentrations	8-37
8-19	Program vs. Site-Specific Average Benzo(a)pyrene Concentration	8-38
xxiii

-------
LIST OF FIGURES (Continued)
Page
8-20a	Program vs. Site-Specific Average 1,3-Butadiene (Method TO-15) Concentration	8-38
8-20b	Program vs. Site-Specific Average 1,3-Butadiene (SNMOC) Concentrations	8-39
8-21	Program vs. Site-Specific Average Carbon Tetrachloride Concentration	8-39
8-22	Program vs. Site-Specific Average 1,2-Dichloroethane Concentration	8-40
8-23a	Program vs. Site-Specific Average Ethylbenzene (Method TO-15) Concentration	8-40
8-23b	Program vs. Site-Specific Average Ethylbenzene (SNMOC) Concentrations	8-40
8-24	Program vs. Site-Specific Average Fluorene Concentration	8-41
8-25	Program vs. Site-Specific Average Formaldehyde Concentrations	8-41
8-26	Program vs. Site-Specific Average Hexachloro-l,3-Butadiene Concentration	8-42
8-27	Program vs. Site-Specific Average Naphthalene Concentration	8-42
8-28	Yearly Statistical Metrics for Acenaphthene Concentrations Measured at GPCO	8-47
8-29	Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at GPCO	8-48
8-30	Yearly Statistical Metrics for Benzene Concentrations Measured at GPCO	8-49
8-31	Yearly Statistical Metrics for Benzo(a)pyrene Concentrations Measured at GPCO	8-50
8-32	Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at GPCO	8-51
8-33	Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
GPCO	8-52
8-34	Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
GPCO	8-53
8-35	Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at GPCO	8-54
8-36	Yearly Statistical Metrics for Fluorene Concentrations Measured at GPCO	8-55
8-37	Yearly Statistical Metrics for Formaldehyde Concentrations Measured at GPCO	8-56
8-38	Yearly Statistical Metrics for Hexachloro-1,3-Butadiene Concentrations Measured
at GPCO	8-58
8-39	Yearly Statistical Metrics for Naphthalene Concentrations Measured at GPCO	8-59
8-40	Yearly Statistical Metrics for Benzene Concentrations Measured at BRCO	8-60
8-41	Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at PACO	8-61
8-42	Yearly Statistical Metrics for Benzene Concentrations Measured at PACO	8-62
8-43	Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at PACO	8-63
8-44	Yearly Statistical Metrics for Formaldehyde Concentrations Measured at PACO	8-65
8-45	Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at RICO	8-66
8-46	Yearly Statistical Metrics for Benzene Concentrations Measured at RICO	8-67
8-47	Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at RICO	8-68
8-48	Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at RICO	8-69
8-49	Yearly Statistical Metrics for Formaldehyde Concentrations Measured at RICO	8-70
8-50	Pollution Rose for Formaldehyde Concentrations Measured at GPCO	8-75
9-1	Washington, D.C. (WADC) Monitoring Site	9-2
9-2	NEI Point Sources Located Within 10 Miles of WADC	9-3
9-3	Wind Roses for the Ronald Reagan Washington National Airport Weather Station
near WADC	9-9
9-4	Program vs. Site-Specific Average Naphthalene Concentration	9-13
9-5	Yearly Statistical Metrics for Naphthalene Concentrations Measured at WADC	9-14
10-1	St. Petersburg, Florida (AZFL) Monitoring Site	10-2
10-2	Pinellas Park, Florida (SKFL) Monitoring Site	10-3
xxiv

-------
LIST OF FIGURES (Continued)
Page
10-3 NEI Point Sources Located Within 10 Miles of AZFL and SKFL	10-4
10-4 Valrico, Florida (SYFL) Monitoring Site	10-5
10-5 NEI Point Sources Located Within 10 Miles of SYFL	10-6
10-6 Winter Park, Florida (ORFL) Monitoring Site	10-7
10-7 Orlando, Florida (PAFL) Monitoring Site	10-8
10-8 NEI Point Sources Located Within 10 Miles of ORFL and PAFL	10-9
10-9 Belle Glade, Florida (WPFL) Monitoring Site	10-10
10-10 NEI Point Sources Located Within 10 Miles of WPFL	10-11
10-11 Wind Roses for the St. Petersburg/Whitted Airport Weather Station near AZFL	10-21
10-12 Wind Roses for the St. Petersburg/Clearwater International Airport Weather
Station near SKFL	10-22
10-13 Wind Roses for the Vandenberg Airport Weather Station near SYFL	10-23
10-14 Wind Roses for the Orlando Executive Airport Weather Station near ORFL	10-24
10-15 Wind Roses for the Orlando Executive Airport Weather Station near PAFL	10-25
10-16 Wind Roses for the Palm Beach International Airport Weather Station near WPFL.. 10-26
10-17 Program vs. Site-Specific Average Acenaphthene Concentration	10-37
10-18 Program vs. Site-Specific Average Acetaldehyde Concentrations	10-38
10-19 Program vs. Site-Specific Average Arsenic (PMio) Concentration	10-38
10-20 Program vs. Site-Specific Average Fluoranthene Concentration	10-39
10-21 Program vs. Site-Specific Average Fluorene Concentration	10-39
10-22 Program vs. Site-Specific Average Formaldehyde Concentrations	10-40
10-23 Program vs. Site-Specific Average Naphthalene Concentrations	10-40
10-24 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at AZFL	10-43
10-25 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at AZFL	10-44
10-26 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SKFL	10-45
10-27 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at SKFL	10-46
10-28 Yearly Statistical Metrics for Naphthalene Concentrations Measured at SKFL	10-47
10-29 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SYFL	10-48
10-30 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at SYFL	10-49
10-31 Yearly Statistical Metrics for Naphthalene Concentrations Measured at SYFL	10-50
10-32 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at ORFL	10-51
10-33 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at ORFL	10-52
10-34	Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at PAFL	10-54
11-1	Decatur, Georgia (SDGA) Monitoring Site	11-2
11-2 NEI Point Sources Located Within 10 Miles of SDGA	11-3
11-3 Wind Roses for the Hartsfield International Airport Weather Station near SDGA	11-9
11-4	Yearly Statistical Metrics for Hexavalent Chromium Concentrations Measured
at SDGA	11-13
12-1	Northbrook, Illinois (NBIL) Monitoring Site	12-2
12-2 Schiller Park, Illinois (SPIL) Monitoring Site	12-3
12-3 NEI Point Sources Located Within 10 Miles of NBIL and SPIL	12-4
12-4 Roxana, Illinois (ROIL) Monitoring Site	12-5
12-5 NEI Point Sources Located Within 10 Miles of ROIL	12-6
12-6 Wind Roses for the Palwaukee Municipal Airport Weather Station near NBIL	12-14
xxv

-------
LIST OF FIGURES (Continued)
Page
12-7 Wind Roses for the O'Hare International Airport Weather Station near SPIL	12-15
12-8 Wind Roses for the Lambert/St. Louis International Airport Weather Station near
ROIL	12-16
12-9 Program vs. Site-Specific Average Acenaphthene Concentration	12-28
12-10 Program vs. Site-Specific Average Acetaldehyde Concentrations	12-28
12-11 Program vs. Site-Specific Average Arsenic (PMio) Concentration	12-28
12-12 Program vs. Site-Specific Average Benzene Concentrations	12-29
12-13 Program vs. Site-Specific Average 1,3-Butadiene Concentrations	12-29
12-14 Program vs. Site-Specific Average Carbon Tetrachloride Concentrations	12-30
12-15 Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations	12-30
12-16 Program vs. Site-Specific Average Ethylbenzene Concentration	12-31
12-17 Program vs. Site-Specific Average Fluoranthene Concentration	12-31
12-18 Program vs. Site-Specific Average Fluorene Concentration	12-31
12-19 Program vs. Site-Specific Average Formaldehyde Concentrations	12-32
12-20 Program vs. Site-Specific Average Hexachloro-1,3-Butadiene Concentrations	12-32
12-21 Program vs. Site-Specific Average Naphthalene Concentration	12-33
12-22 Program vs. Site-Specific Average Trichloroethylene Concentration	12-33
12-23 Yearly Statistical Metrics for Acenaphthene Concentrations Measured at NBIL	12-37
12-24 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at NBIL	12-38
12-25 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at NBIL	12-39
12-26 Yearly Statistical Metrics for Benzene Concentrations Measured at NBIL	12-40
12-27 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at NBIL	12-41
12-28 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
NBIL	12-43
12-29 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
NBIL	12-44
12-30 Yearly Statistical Metrics for Fluoranthene Concentrations Measured at NBIL	12-45
12-31 Yearly Statistical Metrics for Fluorene Concentrations Measured at NBIL	12-46
12-32 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at NBIL	12-47
12-33 Yearly Statistical Metrics for Naphthalene Concentrations Measured at NBIL	12-48
12-34 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SPIL	12-49
12-35 Yearly Statistical Metrics for Benzene Concentrations Measured at SPIL	12-50
12-36 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at SPIL	12-51
12-37 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
SPIL	12-53
12-38 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
SPIL	12-54
12-39 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at SPIL	12-55
12-40 Yearly Statistical Metrics for Hexachloro-1,3-Butadiene Concentrations Measured
at SPIL	12-56
12-41	Yearly Statistical Metrics for Trichloroethylene Concentrations Measured at SPIL .. 12-57
13-1	Gary, Indiana (INDEM) Monitoring Site	13-2
13-2 NEI Point Sources Located Within 10 Miles of INDEM	13-3
13-3 Indianapolis, Indiana (WPIN) Monitoring Site	13-4
13-4 NEI Point Sources Located Within 10 Miles of WPIN	13-5
xxvi

-------
LIST OF FIGURES (Continued)
Page
13-5 Wind Roses for the Lansing Municipal Airport Weather Station near INDEM	13-12
13-6 Wind Roses for the Eagle Creek Airpark Weather Station near WPIN	13-13
13-7 Program vs. Site-Specific Average Acetaldehyde Concentrations	13-18
13-8 Program vs. Site-Specific Average Formaldehyde Concentrations	13-19
13-9	Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at INDEM.... 13-20
13-10	Yearly Statistical Metrics for Formaldehyde Concentrations Measured at INDEM... 13-21
13-11 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at WPIN	13-23
13-12	Yearly Statistical Metrics for Formaldehyde Concentrations Measured at WPIN	13-24
14-1	Ashland, Kentucky (ASKY) Monitoring Site	14-2
14-2	Ashland, Kentucky (ASKY-M) Monitoring Site	14-3
14-3	NEI Point Sources Located Within 10 Miles of ASKY and ASKY-M	14-4
14-4	Grayson, Kentucky (GLKY) Monitoring Site	14-5
14-5	NEI Point Sources Located Within 10 Miles of GLKY	14-6
14-6	Baskett, Kentucky (BAKY) Monitoring Site	14-7
14-7	NEI Point Sources Located Within 10 Miles of BAKY	14-8
14-8	Calvert City, Kentucky (ATKY) Monitoring Site	14-9
14-9	Smithland, Kentucky (BLKY) Monitoring Site	14-10
14-10	Calvert City, Kentucky (CCKY) Monitoring Site	14-11
14-11	Calvert City, Kentucky (LAKY) Monitoring Site	14-12
14-12	Calvert City, Kentucky (TVKY) Monitoring Site	14-13
14-13	NEI Point Sources Located Within 10 Miles of ATKY, BLKY, CCKY, LAKY,
and TVKY	14-14
14-14 Lexington, Kentucky (LEKY) Monitoring Site	14-15
14-15 NEI Point Sources Located Within 10 Miles of LEKY	14-16
14-16 Wind Roses for the Tri-State/M.J. Ferguson Field Airport Weather Station near
ASKY	14-28
14-17 Wind Roses for the Tri-State/M.J. Ferguson Field Airport Weather Station near
ASKY-M	14-29
14-18 Wind Roses for the Tri-State/M.J. Ferguson Field Airport Weather Station near
GLKY	14-30
14-19 Wind Roses for the Evansville Regional Airport Weather Station near BAKY	14-31
14-20 Wind Roses for the Barkley Regional Airport Weather Station near ATKY	14-33
14-21 Wind Roses for the Barkley Regional Airport Weather Station near BLKY	14-34
14-22 Wind Roses for the Barkley Regional Airport Weather Station near CCKY	14-35
14-23 Wind Roses for the Barkley Regional Airport Weather Station near LAKY	14-36
14-24 Wind Roses for the Barkley Regional Airport Weather Station near TVKY	14-37
14-25 Wind Roses for the Blue Grass Airport Weather Station near LEKY	14-39
14-26 Program vs. Site-Specific Average Acetaldehyde Concentrations	14-60
14-27 Program vs. Site-Specific Average Arsenic (PMio) Concentrations	14-61
14-28 Program vs. Site-Specific Average Benzene Concentrations	14-63
14-29 Program vs. Site-Specific Average 1,3-Butadiene Concentrations	14-64
14-30 Program vs. Site-Specific Average Cadmium Concentration	14-65
14-31 Program vs. Site-Specific Average Carbon Tetrachloride Concentrations	14-66
14-32 Program vs. Site-Specific Average/>-Dichlorobenzene Concentration	14-67
14-33 Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations	14-68
xxvii

-------
LIST OF FIGURES (Continued)
Page
14-34 Program vs. Site-Specific Average Formaldehyde Concentrations	14-69
14-35 Program vs. Site-Specific Average Hexachloro-1,3-Butadiene Concentrations	14-70
14-36 Program vs. Site-Specific Average Lead (PMio) Concentration	14-71
14-37 Program vs. Site-Specific Average Manganese (PMio) Concentration	14-71
14-38 Program vs. Site-Specific Average Nickel (PMio) Concentrations	14-72
14-39 Program vs. Site-Specific Average 1,1,2-Trichloroethane Concentrations	14-72
14-40 Program vs. Site-Specific Average Vinyl Chloride Concentrations	14-73
14-41	Pollution Rose for 1,2-Dichloroethane Concentrations Measured at TVKY	14-81
15-1	Boston, Massachusetts (BOMA) Monitoring Site	15-2
15-2 NEI Point Sources Located Within 10 Miles of BOMA	15-3
15-3 Wind Roses for the Logan International Airport Weather Station near BOMA	15-9
15-4 Program vs. Site-Specific Average Arsenic (PMio) Concentration	15-14
15-5 Program vs. Site-Specific Average Naphthalene Concentration	15-14
15-6 Program vs. Site-Specific Average Nickel (PMio) Concentration	15-14
15-7	Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at BOMA... 15-16
15-8 Yearly Statistical Metrics for Naphthalene Concentrations Measured at BOMA	15-17
15-9	Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at BOMA.... 15-18
16-1	Dearborn, Michigan (DEMI) Monitoring Site	16-2
16-2 NEI Point Sources Located Within 10 Miles of DEMI	16-3
16-3 Wind Roses for the Detroit City Airport Weather Station near DEMI	16-9
16-4 Program vs. Site-Specific Average Acenaphthene Concentration	16-15
16-5 Program vs. Site-Specific Average Acetaldehyde Concentration	16-15
16-6 Program vs. Site-Specific Average Benzene Concentration	16-15
16-7 Program vs. Site-Specific Average 1,3-Butadiene Concentration	16-16
16-8 Program vs. Site-Specific Average Carbon Tetrachloride Concentration	16-16
16-9 Program vs. Site-Specific Average 1,2-Dichloroethane Concentration	16-16
16-10 Program vs. Site-Specific Average Ethylbenzene Concentration	16-17
16-11 Program vs. Site-Specific Average Fluorene Concentration	16-17
16-12 Program vs. Site-Specific Average Formaldehyde Concentration	16-17
16-13 Program vs. Site-Specific Average Naphthalene Concentration	16-18
16-14 Yearly Statistical Metrics for Acenaphthene Concentrations Measured at DEMI	16-20
16-15 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at DEMI	16-22
16-16 Yearly Statistical Metrics for Benzene Concentrations Measured at DEMI	16-23
16-17 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at DEMI	16-24
16-18	Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
DEMI	16-25
16-19	Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
DEMI	16-26
16-20 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at DEMI	16-28
16-21 Yearly Statistical Metrics for Fluorene Concentrations Measured at DEMI	16-29
16-22 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at DEMI	16-30
16-23 Yearly Statistical Metrics for Naphthalene Concentrations Measured at DEMI	16-32
xxviii

-------
LIST OF FIGURES (Continued)
Page
17-1 St. Cloud, Minnesota (STMN) Monitoring Site	17-2
17-2 NEI Point Sources Located Within 10 Miles of STMN	17-3
17-3	Wind Roses for the St. Cloud Regional Airport Weather Station near STMN	17-9
18-1	Columbus, Mississippi (KMMS) Monitoring Site	18-2
18-2 Columbus, Mississippi (SSMS) Monitoring Site	18-3
18-3 NEI Point Sources Located Within 10 Miles of KMMS and SSMS	18-4
18-4	Wind Roses for the Columbus Air Force Base Airport Weather Station near
KMMS	18-11
18-5	Wind Roses for the Columbus Air Force Base Airport Weather Station near
SSMS	18-12
18-6 Program vs. Site-Specific Average Benzene Concentrations	18-19
18-7 Program vs. Site-Specific Average 1,3-Butadiene Concentrations	18-19
18-8 Program vs. Site-Specific Average Carbon Tetrachloride Concentrations	18-20
18-9 Program vs. Site-Specific Average/?-Dichlorobenzene Concentrations	18-20
18-10 Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations	18-21
18-11 Program vs. Site-Specific Average Ethylbenzene Concentrations	18-21
18-12 Program vs. Site-Specific Average Hexachloro-1,3-Butadiene Concentration	18-21
18-13	Program vs. Site-Specific Average Xylenes Concentration	18-22
19-1	St. Louis, Missouri (S4MO) Monitoring Site	19-2
19-2 NEI Point Sources Located Within 10 Miles of S4MO	19-3
19-3 Wind Roses for the St. Louis Downtown Airport Weather Station near S4MO	19-9
19-4 Program vs. Site-Specific Average Acenaphthene Concentration	19-17
19-5 Program vs. Site-Specific Average Acetaldehyde Concentration	19-17
19-6 Program vs. Site-Specific Average Arsenic (PMio) Concentration	19-18
19-7 Program vs. Site-Specific Average Benzene Concentration	19-18
19-8 Program vs. Site-Specific Average 1,3-Butadiene Concentration	19-18
19-9 Program vs. Site-Specific Average Cadmium (PMio) Concentration	19-19
19-10 Program vs. Site-Specific Average Carbon Tetrachloride Concentration	19-19
19-11 Program vs. Site-Specific Average/?-Dichlorobenzene Concentration	19-19
19-12 Program vs. Site-Specific Average 1,2-Dichloroethane Concentration	19-20
19-13 Program vs. Site-Specific Average Fluorene Concentration	19-20
19-14 Program vs. Site-Specific Average Formaldehyde Concentration	19-20
19-15 Program vs. Site-Specific Average Hexachloro-1,3-Butadiene Concentration	19-21
19-16 Program vs. Site-Specific Average Lead (PMio) Concentration	19-21
19-17 Program vs. Site-Specific Average Naphthalene Concentration	19-21
19-18 Program vs. Site-Specific Average Nickel (PMio) Concentration	19-22
19-19 Yearly Statistical Metrics for Acenaphthene Concentrations Measured at S4MO	19-26
19-20 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at S4MO	19-27
19-21	Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at S4MO.... 19-28
19-22 Yearly Statistical Metrics for Benzene Concentrations Measured at S4MO	19-29
19-23 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at S4MO	19-30
19-24	Yearly Statistical Metrics for Cadmium (PMio) Concentrations Measured at
S4MO	19-31
xxix

-------
LIST OF FIGURES (Continued)
Page
19-25 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
S4MO	19-32
19-26 Yearly Statistical Metrics for />Dichlorobenzene Concentrations Measured at
S4MO	19-33
19-27 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
S4MO	19-35
19-28 Yearly Statistical Metrics for Fluorene Concentrations Measured at S4MO	19-36
19-29 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at S4MO	19-37
19-30 Yearly Statistical Metrics for Hexachloro-1,3-Butadiene Concentrations Measured
at S4MO	19-38
19-31 Yearly Statistical Metrics for Lead (PMio) Concentrations Measured at S4MO	19-39
19-32 Yearly Statistical Metrics for Naphthalene Concentrations Measured at S4MO	19-40
19-33	Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at S4MO	19-41
20-1	Camden, New Jersey (CSNJ) Monitoring Site	20-2
20-2	NEI Point Sources Located Within 10 Miles of CSNJ	20-3
20-3	Chester, New Jersey (CHNJ) Monitoring Site	20-4
20-4	NEI Point Sources Located Within 10 Miles of CHNJ	20-5
20-5	Elizabeth, New Jersey (ELNJ) Monitoring Site	20-6
20-6	North Brunswick, New Jersey (NBNJ) Monitoring Site	20-7
20-7	NEI Point Sources Located Within 10 Miles of ELNJ and NBNJ	20-8
20-8	Wind Roses for the Philadelphia International Airport Weather Station near CSNJ.. 20-16
20-9	Wind Roses for the Summerville-Somerset Airport Weather Station near CHNJ	20-17
20-10	Wind Roses for the Newark International Airport Weather Station near ELNJ	20-18
20-11	Wind Roses for the Summerville-Somerset Airport Weather Station near NBNJ	20-19
20-12	Program vs. Site-Specific Average Acetaldehyde Concentrations	20-32
20-13	Program vs. Site-Specific Average Benzene Concentrations	20-33
20-14	Program vs. Site-Specific Average Bromomethane Concentration	20-33
20-15	Program vs. Site-Specific Average 1,3-Butadiene Concentrations	20-34
20-16	Program vs. Site-Specific Average Carbon Tetrachloride Concentrations	20-35
20-17	Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations	20-36
20-18	Program vs. Site-Specific Average Ethylbenzene Concentrations	20-36
20-19	Program vs. Site-Specific Average Formaldehyde Concentrations	20-37
20-20	Program vs. Site-Specific Average Hexachloro-1,3-Butadiene Concentrations	20-38
20-21	Program vs. Site-Specific Average Propionaldehyde Concentrations	20-38
20-22	Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at CHNJ	20-42
20-23	Yearly Statistical Metrics for Benzene Concentrations Measured at CHNJ	20-43
20-24	Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at CHNJ	20-44
20-25	Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
CHNJ	20-46
20-26 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
CHNJ	20-47
20-27 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at CHNJ	20-48
20-28 Yearly Statistical Metrics for Hexachloro-l,3-Butadiene Concentrations Measured
at CHNJ	20-50
20-29 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at ELNJ	20-51
XXX

-------
LIST OF FIGURES (Continued)
Page
20-30 Yearly Statistical Metrics for Benzene Concentrations Measured at ELNJ	20-52
20-31 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at ELNJ	20-53
20-32 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
ELNJ	20-55
20-33 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
ELNJ	20-56
20-34 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at ELNJ	20-57
20-35 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at ELNJ	20-58
20-36 Yearly Statistical Metrics for Hexachloro-1,3-Butadiene Concentrations Measured
at ELNJ	20-59
20-37 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at NBNJ	20-60
20-38 Yearly Statistical Metrics for Benzene Concentrations Measured at NBNJ	20-61
20-39 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at NBNJ	20-63
20-40 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
NBNJ	20-64
20-41 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
NBNJ	20-66
20-42 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at NBNJ	20-67
20-43	Yearly Statistical Metrics for Hexachloro-1,3-Butadiene Concentrations Measured
at NBNJ	20-68
21-1	New York City, New York (BXNY) Monitoring Site	21-2
21-2	NEI Point Sources Located Within 10 Miles of BXNY	21-3
21-3	Rochester, New York (ROCH) Monitoring Site	21-4
21-4	NEI Point Sources Located Within 10 Miles of ROCH	21-5
21-5	Wind Roses for the La Guardia Airport Weather Station near BXNY	21-13
21-6	Wind Roses for the Greater Rochester International Airport Weather Station near
ROCH	21-14
21-7 Program vs. Site-Specific Average Acenaphthene Concentrations	21-20
21-8 Program vs. Site-Specific Average Fluoranthene Concentration	21-20
21-9 Program vs. Site-Specific Average Fluorene Concentrations	21-21
21-10 Program vs. Site-Specific Average Naphthalene Concentrations	21-21
21-11 Yearly Statistical Metrics for Acenaphthene Concentrations Measured at ROCH	21-23
21-12 Yearly Statistical Metrics for Fluoranthene Concentrations Measured at ROCH	21-24
21-13 Yearly Statistical Metrics for Fluorene Concentrations Measured at ROCH	21-25
21-14	Yearly Statistical Metrics for Naphthalene Concentrations Measured at ROCH	21-26
22-1	Public Works, Tulsa, Oklahoma (TOOK) Monitoring Site	22-2
22-2	Fire Station, Tulsa, Oklahoma (TMOK) Monitoring Site	22-3
22-3	Riverside, Tulsa, Oklahoma (TROK) Monitoring Site	22-4
22-4	NEI Point Sources Located Within 10 Miles of TMOK, TOOK, and TROK	22-5
22-5	Air Depot, Oklahoma City, Oklahoma (ADOK) Monitoring Site	22-6
22-6	Oklahoma City, Oklahoma (OCOK) Monitoring Site	22-7
22-7	Yukon, Oklahoma (YUOK) Monitoring Site	22-8
22-8	NEI Point Sources Located Within 10 Miles of ADOK, OCOK, and YUOK	22-9
22-9	Wind Roses for the Richard Lloyd Jones Jr. Airport Weather Station near TOOK.... 22-19
xxxi

-------
LIST OF FIGURES (Continued)
Page
22-10 Wind Roses for the Tulsa International Airport Weather Station near TMOK	22-20
22-11 Wind Roses for the Tulsa International Airport Weather Station near TROK	22-21
22-12 Wind Roses for the Tinker Air Force Base Airport Weather Station near ADOK	22-22
22-13 Wind Roses for the Wiley Post Airport Weather Station near OCOK	22-23
22-14 Wind Roses for the Wiley Post Airport Weather Station near YUOK	22-24
22-15 Program vs. Site-Specific Average Acetaldehyde Concentrations	22-41
22-16 Program vs. Site-Specific Average Arsenic (TSP) Concentrations	22-42
22-17 Program vs. Site-Specific Average Benzene Concentrations	22-43
22-18 Program vs. Site-Specific Average 1,3-Butadiene Concentrations	22-44
22-19 Program vs. Site-Specific Average Carbon Tetrachloride Concentrations	22-45
22-20 Program vs. Site-Specific Average/?-Dichlorobenzene Concentrations	22-46
22-21 Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations	22-47
22-22 Program vs. Site-Specific Average Ethylbenzene Concentrations	22-48
22-23 Program vs. Site-Specific Average Formaldehyde Concentrations	22-49
22-24 Program vs. Site-Specific Average Hexachloro-1,3-Butadiene Concentrations	22-50
22-25 Program vs. Site-Specific Average Manganese (TSP) Concentration	22-51
22-26 Program vs. Site-Specific Average Nickel (TSP) Concentrations	22-51
22-27 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at TOOK	22-53
22-28 Yearly Statistical Metrics for Arsenic (TSP) Concentrations Measured at TOOK	22-54
22-29 Yearly Statistical Metrics for Benzene Concentrations Measured at TOOK	22-55
22-30 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at TOOK	22-56
22-31 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
TOOK	22-57
22-32 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
TOOK	22-58
22-33 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at TOOK	22-59
22-34 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at TOOK	22-60
22-35 Yearly Statistical Metrics for Hexachloro-1,3-Butadiene Concentrations Measured
at TOOK	22-61
22-36 Yearly Statistical Metrics for Manganese (TSP) Concentrations Measured at
TOOK	22-62
22-37 Yearly Statistical Metrics for Nickel (TSP) Concentrations Measured at TOOK	22-63
22-38 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at TMOK	22-64
22-39 Yearly Statistical Metrics for Arsenic (TSP) Concentrations Measured at TMOK .... 22-65
22-40 Yearly Statistical Metrics for Benzene Concentrations Measured at TMOK	22-66
22-41 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at TMOK	22-67
22-42 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
TMOK	22-68
22-43 Yearly Statistical Metrics for /;-Dichlorobenzene Concentrations Measured at
TMOK	22-69
22-44 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
TMOK	22-70
22-45 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at
TMOK	22-71
22-46 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at TMOK.... 22-72
XXXll

-------
LIST OF FIGURES (Continued)
Page
22-47 Yearly Statistical Metrics for Hexachloro-1,3-Butadiene Concentrations Measured
at TY10K	22-73
22-48 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at OCOK	22-74
22-49 Yearly Statistical Metrics for Arsenic (TSP) Concentrations Measured at OCOK	22-75
22-50 Yearly Statistical Metrics for Benzene Concentrations Measured at OCOK	22-76
22-51 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at OCOK	22-77
22-52 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured
at OCOK	22-78
22-53 Yearly Statistical Metrics for /;-Dichlorobenzene Concentrations Measured at
OCOK	22-79
22-54 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
OCOK	22-80
22-55 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at OCOK	22-81
22-56 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at OCOK	22-82
22-57	Yearly Statistical Metrics for Hexachloro-1,3-Butadiene Concentrations Measured
at OCOK	22-83
23-1	Providence, Rhode Island (PRRI) Monitoring Site	23-2
23-2	NEI Point Sources Located Within 10 Miles of PRRI	23-3
23-3	Wind Roses for the T.F. Green State Airport Weather Station near PRRI	23-9
23-4	Program vs. Site-Specific Average Naphthalene Concentration	23-13
23-5	Yearly Statistical Metrics for Naphthalene Concentrations Measured at PRRI	23-14
24-1	Chesterfield, South Carolina (CHSC) Monitoring Site	24-2
24-2	NEI Point Sources Located Within 10 Miles of CHSC	24-3
24-3	Wind Roses for the Richmond County Airport Weather Station near CHSC	24-9
24-4	Program vs. Site-Specific Average Naphthalene Concentration	24-12
24-5	Yearly Statistical Metrics for Naphthalene Concentrations Measured at CHSC	24-13
25-1	Deer Park, Texas (CAMS 35) Monitoring Site	25-2
25-2	NEI Point Sources Located Within 10 Miles of CAMS 35	25-3
25-3	Karnack, Texas (CAMS 85) Monitoring Site	25-4
25-4	NEI Point Sources Located Within 10 Miles of CAMS 85	25-5
25-5	Wind Roses for the William P. Hobby Airport Weather Station near CAMS 35 	25-12
25-6	Wind Roses for the Shreveport Regional Airport Weather Station near CAMS 85.... 25-13
26-1	Bountiful, Utah (BTUT) Monitoring Site	26-2
26-2 NEI Point Sources Located Within 10 Miles of BTUT	26-3
26-3 Wind Roses for the Salt Lake City International Airport Weather Station near
BTUT	26-9
26-4 Program vs. Site-Specific Average Acetaldehyde Concentration	26-17
26-5 Program vs. Site-Specific Average Arsenic (PMio) Concentration	26-17
26-6 Program vs. Site-Specific Average Benzene Concentration	26-17
26-7 Program vs. Site-Specific Average 1,3-Butadiene Concentration	26-18
26-8 Program vs. Site-Specific Average Carbon Tetrachloride Concentration	26-18
26-9 Program vs. Site-Specific Average/>-Dichlorobenzene Concentration	26-18
XXXlll

-------
LIST OF FIGURES (Continued)
Page
26-10 Program vs. Site-Specific Average 1,2-Dichloroethane Concentration	26-19
26-11 Program vs. Site-Specific Average Dichloromethane Concentration	26-19
26-12 Program vs. Site-Specific Average Ethylbenzene Concentration	26-19
26-13 Program vs. Site-Specific Average Formaldehyde Concentration	26-20
26-14 Program vs. Site-Specific Average Naphthalene Concentration	26-20
26-15 Program vs. Site-Specific Average Nickel (PMio) Concentration	26-20
26-16 Program vs. Site-Specific Average Propionaldehyde Concentration	26-21
26-17 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at BTUT	26-25
26-18	Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at BTUT.... 26-26
26-19 Yearly Statistical Metrics for Benzene Concentrations Measured at BTUT	26-27
26-20 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at BTUT	26-28
26-21	Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
BTUT	26-29
26-22	Yearly Statistical Metrics for /;-Dichlorobenzene Concentrations Measured at
BTUT	26-30
26-23	Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
BTUT	26-31
26-24	Yearly Statistical Metrics for Dichloromethane Concentrations Measured at
BTUT	26-32
26-25 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at BTUT	26-33
26-26 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at BTUT	26-35
26-27 Yearly Statistical Metrics for Naphthalene Concentrations Measured at BTUT	26-36
26-28 Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at BTUT	26-37
26-29	Yearly Statistical Metrics for Propionaldehyde Concentrations Measured at BTUT.. 26-38
26-30	Pollution Rose for Formaldehyde Concentrations Measured at BTUT	26-42
27-1	Burlington, Vermont (BURVT) Monitoring Site	27-2
27-2 Underhill, Vermont (UNVT) Monitoring Site	27-3
27-3 NEI Point Sources Located Within 10 Miles of BURVT and UNVT	27-4
27-4 Rutland, Vermont (RUVT) Monitoring Site	27-5
27-5 NEI Point Sources Located Within 10 Miles of RUVT	27-6
27-6	Wind Roses for the Burlington International Airport Weather Station near
BURVT	27-14
27-7 Wind Roses for the Rutland State Airport Weather Station near RUVT	27-15
27-8	Wind Roses for the Morrisville-Stowe State Airport Weather Station near UNVT.... 27-16
27-9 Program vs. Site-Specific Average Arsenic (PMio) Concentration	27-25
27-10 Program vs. Site-Specific Average Benzene Concentrations	27-26
27-11 Program vs. Site-Specific Average 1,3-Butadiene Concentrations	27-26
27-12 Program vs. Site-Specific Average Carbon Tetrachloride Concentrations	27-27
27-13 Program vs. Site-Specific Average/?-Dichlorobenzene Concentration	27-27
27-14 Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations	27-28
27-15 Program vs. Site-Specific Average Ethylbenzene Concentration	27-28
27-16 Program vs. Site-Specific Average Hexachloro-l,3-Butadiene Concentration	27-29
27-17 Yearly Statistical Metrics for Benzene Concentrations Measured at BURVT	27-32
27-18	Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at BURVT... 27-33
xxxiv

-------
LIST OF FIGURES (Continued)
Page
27-19 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured
at BURVT	27-34
27-20 Yearly Statistical Metrics for />Dichlorobenzene Concentrations Measured at
BURVT	27-35
27-21 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
BURVT	27-36
27-22 Yearly Statistical Metrics for Hexachloro-1,3-Butadiene Concentrations Measured
at BURVT	27-37
27-23 Yearly Statistical Metrics for Benzene Concentrations Measured at RUVT	27-38
27-24 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at RUVT	27-39
27-25 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
RUVT	27-40
27-26 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
RUVT	27-41
27-27 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at RUVT	27-42
27-28 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at UNVT ... 27-43
27-29 Yearly Statistical Metrics for Benzene Concentrations Measured at UNVT	27-44
27-30 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at UNVT	27-45
27-31 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
UNVT	27-46
27-32	Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
UNVT	27-47
28-1	East Highland Park, Virginia (RIVA) Monitoring Site	28-2
28-2	NEI Point Sources Located Within 10 Miles of RIVA	28-3
28-3	Wind Roses for the Richmond International Airport Weather Station near RIVA	28-9
28-4	Program vs. Site-Specific Average Naphthalene Concentration	28-13
28-5	Yearly Statistical Metrics for Naphthalene Concentrations Measured at RIVA	28-14
29-1	Seattle, Washington (SEWA) Monitoring Site	29-2
29-2 NEI Point Sources Located Within 10 Miles of SEWA	29-3
29-3 Wind Roses for the Boeing Field/King County International Airport Weather
Station near SEWA	29-9
29-4 Program vs. Site-Specific Average Acetaldehyde Concentration	29-16
29-5 Program vs. Site-Specific Average Arsenic (PMio) Concentration	29-16
29-6 Program vs. Site-Specific Average Benzene Concentration	29-16
29-7 Program vs. Site-Specific Average 1,3-Butadiene Concentration	29-17
29-8 Program vs. Site-Specific Average Carbon Tetrachloride Concentration	29-17
29-9 Program vs. Site-Specific Average 1,2-Dichloroethane Concentration	29-17
29-10 Program vs. Site-Specific Average Formaldehyde Concentration	29-18
29-11 Program vs. Site-Specific Average Naphthalene Concentration	29-18
29-12 Program vs. Site-Specific Average Nickel (PMio) Concentration	29-18
29-13 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SEWA	29-21
29-14	Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at
SEWA	29-22
29-15 Yearly Statistical Metrics for Benzene Concentrations Measured at SEWA	29-23
xxxv

-------
LIST OF FIGURES (Continued)
Page
29-16 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at SEWA	29-24
29-17 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
SEWA	29-25
29-18 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
SEWA	29-26
29-19 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at SEWA	29-27
29-20 Yearly Statistical Metrics for Naphthalene Concentrations Measured at SEWA	29-28
29-21	Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at SEWA	29-29
30-1	Horicon, Wisconsin (HOWI) Monitoring Site	30-2
30-2	NEI Point Sources Located Within 10 Miles of HOWI	30-3
30-3	Milwaukee, Wisconsin (MIWI) Monitoring Site	30-4
30-4	NEI Point Sources Located Within 10 Miles of MIWI	30-5
30-5	Wind Roses for the Dodge County Airport Weather Station near HOWI	30-13
30-6 Wind Roses for the Lawrence J. Timmerman Airport Weather Station near MIWI... 30-14
xxxvi

-------
LIST OF TABLES
Page
1-1	Organization of the 2013 National Monitoring Programs Report	1-4
2-1	2013 National Monitoring Programs Sites and Past Program Participation	2-4
2-2 Site Characterizing Information for the 2013 National Monitoring Programs Sites	2-8
2-3 2013 VOC Method Detection Limits	2-18
2-4 2013 SNMOC Method Detection Limits	2-19
2-5 2013 Carbonyl Compound Method Detection Limits	2-21
2-6 2013 PAH Method Detection Limits 	2-22
2-7 2013 PAH/Phenols Method Detection Limits	2-23
2-8 2013 Metals Method Detection Limits	2-24
2-9 2013 Hexavalent Chromium Method Detection Limit	2-25
2-10 2013 Sampling Schedules and Completeness Rates	2-26
2-11	Method Completeness Rates for 2013	2-33
3-1	Overview and Layout of Data Presented	3-1
3-2 NATTS MQO Core Analytes	3-7
3-3	POM Groups for PAHs and Phenols	3-17
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
4-4a Statistical Summaries of the PAH Concentrations	4-11
4-4b Statistical Summaries of the PAH/Phenols Concentrations	4-12
4-5 Statistical Summaries of the Metals Concentrations	4-13
4-6 Statistical Summary of the Hexavalent Chromium Concentrations	4-14
4-7 Results of the Program-Level Preliminary Risk-Based Screening Process	4-18
4-8 Site-Specific Risk-Based Screening Comparison	4-20
4-9	Annual Average Concentration Comparison of the VOC/SNMOC Pollutants of
Interest	4-24
4-10	Annual Average Concentration Comparison of the Carbonyl Compound Pollutants
of Interest	4-25
4-11 Annual Average Concentration Comparison of the PAH Pollutants of Interest	4-25
4-12 Annual Average Concentration Comparison of the Metals Pollutants of Interest	4-26
4-13 Summary of Mobile Source Information by Monitoring Site	4-31
4-14	Greenhouse Gases Measured by Method TO-15	4-83
5-1	Geographical Information for the Alaska Monitoring Site	5-4
5-2	Population, Motor Vehicle, and Traffic Information for the Alaska Monitoring
Site	5-5
5-3 Average Meteorological Conditions near the Alaska Monitoring Site	5-7
5-4 Risk-Based Screening Results for the Alaska Monitoring Site	5-11
5-5	Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Alaska Monitoring Site	5-13
5-6 Risk Approximations for the Alaska Monitoring Site	5-19
xxxvii

-------
LIST OF TABLES (Continued)
Page
5-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Alaska Monitoring Site	5-21
5-8	Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Alaska Monitoring
Site	5-22
6-1	Geographical Information for the Arizona Monitoring Sites	6-5
6-2 Population, Motor Vehicle, and Traffic Information for the Arizona Monitoring
Sites	6-6
6-3 Average Meteorological Conditions near the Arizona Monitoring Sites	6-8
6-4 Risk-Based Screening Results for the Arizona Monitoring Sites	6-13
6-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Arizona Monitoring Sites	6-16
6-6 Risk Approximations for the Arizona Monitoring Sites	6-43
6-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Arizona Monitoring Sites	6-45
6-8	Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Arizona Monitoring
Sites	6-46
7-1	Geographical Information for the California Monitoring Sites	7-9
7-2 Population, Motor Vehicle, and Traffic Information for the California Monitoring
Sites	7-12
7-3 Average Meteorological Conditions near the California Monitoring Sites	7-14
7-4 Risk-Based Screening Results for the California Monitoring Sites	7-23
7-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
California Monitoring Sites	7-25
7-6 Risk Approximations for the California Monitoring Sites	7-41
7-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the California Monitoring Sites	7-43
7-8	Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the California Monitoring
Sites	7-45
8-1	Geographical Information for the Colorado Monitoring Sites	8-11
8-2 Population, Motor Vehicle, and Traffic Information for the Colorado Monitoring
Sites	8-14
8-3 Average Meteorological Conditions near the Colorado Monitoring Sites	8-16
8-4 Risk-Based Screening Results for the Colorado Monitoring Sites	8-28
8-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Colorado Monitoring Sites	8-31
8-6 Risk Approximations for the Colorado Monitoring Sites	8-72
8-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Colorado Monitoring Sites	8-77
xxxviii

-------
LIST OF TABLES (Continued)
Page
8-8	Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Colorado Monitoring
Sites	8-80
9-1	Geographical Information for the Washington, D.C. Monitoring Site	9-4
9-2 Population, Motor Vehicle, and Traffic Information for the Washington, D.C.
Monitoring Site	9-5
9-3 Average Meteorological Conditions near the Washington, D.C. Monitoring Site	9-7
9-4 Risk-Based Screening Results for the Washington, D.C. Monitoring Site	9-10
9-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Washington, D.C. Monitoring Site	9-12
9-6 Risk Approximations for the Washington, D.C. Monitoring Site	9-15
9-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Washington, D.C. Monitoring Site	9-17
9-8	Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Washington, D.C.
Monitoring Site	9-18
10-1	Geographical Information for the Florida Monitoring Sites	10-12
10-2 Population, Motor Vehicle, and Traffic Information for the Florida Monitoring
Sites	10-15
10-3 Average Meteorological Conditions near the Florida Monitoring Sites	10-18
10-4 Risk-Based Screening Results for the Florida Monitoring Sites	10-30
10-5a Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Florida Monitoring Sites	10-33
10-5b Quarterly and Annual Average Concentrations of the Pollutants of Interest for
WPFL	10-34
10-6 Risk Approximations for the Florida Monitoring Sites	10-56
10-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Florida Monitoring Sites	10-58
10-8	Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Florida Monitoring
Sites	10-61
11-1	Geographical Information for the Georgia Monitoring Site	11-4
11-2 Population, Motor Vehicle, and Traffic Information for the Georgia Monitoring
Site	11-5
11-3 Average Meteorological Conditions near the Georgia Monitoring Site	11-7
11-4 Risk-Based Screening Results for the Georgia Monitoring Site	11-11
11-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Georgia Monitoring Site	11-12
11-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Georgia Monitoring Site	11-16
11-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Georgia Monitoring
Site	11-17
XXXIX

-------
LIST OF TABLES (Continued)
Page
12-1 Geographical Information for the Illinois Monitoring Sites	12-7
12-2 Population, Motor Vehicle, and Traffic Information for the Illinois Monitoring
Sites	12-9
12-3 Average Meteorological Conditions near the Illinois Monitoring Sites	12-12
12-4 Risk-Based Screening Results for the Illinois Monitoring Sites	12-19
12-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Illinois Monitoring Sites	12-22
12-6 Risk Approximations for the Illinois Monitoring Sites	12-59
12-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Illinois Monitoring Sites	12-62
12-8	Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Illinois Monitoring
Sites	12-64
13-1	Geographical Information for the Indiana Monitoring Sites	13-6
13-2 Population, Motor Vehicle, and Traffic Information for the Indiana Monitoring
Sites	13-8
13-3 Average Meteorological Conditions near the Indiana Monitoring Sites	13-10
13-4 Risk-Based Screening Results for the Indiana Monitoring Sites	13-15
13-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Indiana Monitoring Sites	13-17
13-6 Risk Approximations for the Indiana Monitoring Sites	13-25
13-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Indiana Monitoring Sites	13-28
13-8	Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Indiana Monitoring
Sites	13-29
14-1	Geographical Information for the Kentucky Monitoring Sites	14-17
14-2 Population, Motor Vehicle, and Traffic Information for the Kentucky Monitoring
Sites	14-21
14-3 Average Meteorological Conditions near the Kentucky Monitoring Sites	14-23
14-4 Overview of Sampling Performed at the Kentucky Monitoring Sites	14-41
14-5 Risk-Based Screening Results for the Kentucky Monitoring Sites	14-42
14-6 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Kentucky Monitoring Sites	14-48
14-7 Risk Approximations for the Kentucky Monitoring Sites	14-75
14-8 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Kentucky Monitoring Sites	14-83
14-9	Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Kentucky Monitoring
Sites	14-88
15-1	Geographical Information for the Massachusetts Monitoring Site	15-4
15-2 Population, Motor Vehicle, and Traffic Information for the Massachusetts
Monitoring Site	15-5
xl

-------
LIST OF TABLES (Continued)
Page
15-3 Average Meteorological Conditions near the Massachusetts Monitoring Site	15-7
15-4 Risk-Based Screening Results for the Massachusetts Monitoring Site	15-11
15-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Massachusetts Monitoring Site	15-12
15-6 Risk Approximations for the Massachusetts Monitoring Site	15-19
15-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Massachusetts Monitoring Site	15-21
15-8	Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Massachusetts
Monitoring Site	15-22
16-1	Geographical Information for the Michigan Monitoring Site	16-4
16-2 Population, Motor Vehicle, and Traffic Information for the Michigan Monitoring
Site	16-5
16-3 Average Meteorological Conditions near the Michigan Monitoring Site	16-7
16-4 Risk-Based Screening Results for the Michigan Monitoring Site	16-11
16-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Michigan Monitoring Site	16-13
16-6 Risk Approximations for the Michigan Monitoring Site	16-34
16-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Michigan Monitoring Site	16-35
16-8	Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Michigan Monitoring
Site	16-36
17-1	Geographical Information for the Minnesota Monitoring Site	17-4
17-2 Population, Motor Vehicle, and Traffic Information for the Minnesota Monitoring
Site	17-5
17-3 Average Meteorological Conditions near the Minnesota Monitoring Site	17-7
17-4 Risk-Based Screening Results for the Minnesota Monitoring Site	17-11
17-5 Quarterly and Annual Average Concentrations of Hexavalent Chromium for the
Minnesota Monitoring Site	17-12
17-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Minnesota Monitoring Site	17-14
17-7	Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Minnesota Monitoring
Site	17-15
18-1	Geographical Information for the Mississippi Monitoring Sites	18-5
18-2 Population, Motor Vehicle, and Traffic Information for the Mississippi Monitoring
Sites	18-7
18-3 Average Meteorological Conditions near the Mississippi Monitoring Sites	18-9
18-4 Risk-Based Screening Results for the Mississippi Monitoring Sites	18-14
18-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Mississippi Monitoring Sites	18-16
18-6 Risk Approximations for the Mississippi Monitoring Sites	18-26
xli

-------
LIST OF TABLES (Continued)
Page
18-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Mississippi Monitoring Sites	18-28
18-8	Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Mississippi
Monitoring Sites	18-29
19-1	Geographical Information for the Missouri Monitoring Site	19-4
19-2 Population, Motor Vehicle, and Traffic Information for the Missouri Monitoring
Site	19-5
19-3 Average Meteorological Conditions near the Missouri Monitoring Site	19-7
19-4 Risk-Based Screening Results for the Missouri Monitoring Site	19-11
19-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Missouri Monitoring Site	19-13
19-6 Risk Approximations for the Missouri Monitoring Site	19-43
19-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Missouri Monitoring Site	19-45
19-8	Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Missouri Monitoring
Site	19-46
20-1	Geographical Information for the New Jersey Monitoring Sites	20-9
20-2 Population, Motor Vehicle, and Traffic Information for the New Jersey Monitoring
Sites	20-12
20-3 Average Meteorological Conditions near the New Jersey Monitoring Sites	20-14
20-4 Risk-Based Screening Results for the New Jersey Monitoring Sites	20-22
20-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
New Jersey Monitoring Sites	20-26
20-6 Risk Approximations for the New Jersey Monitoring Sites	20-70
20-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the New Jersey Monitoring Sites	20-74
20-8	Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the New Jersey Monitoring
Sites	20-76
21-1	Geographical Information for the New York Monitoring Sites	21-6
21-2 Population, Motor Vehicle, and Traffic Information for the New York Monitoring
Sites	21-8
21-3 Average Meteorological Conditions near the New York Monitoring Sites	21-11
21-4 Risk-Based Screening Results for the New York Monitoring Sites	21-16
21-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
New York Monitoring Sites	21-18
21-6 Risk Approximations for the New York Monitoring Sites	21-27
21-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the New York Monitoring Sites	21-29
xlii

-------
LIST OF TABLES (Continued)
Page
21-8	Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the New York Monitoring
Sites	21-30
22-1	Geographical Information for the Oklahoma Monitoring Sites	22-10
22-2 Population, Motor Vehicle, and Traffic Information for the Oklahoma Monitoring
Sites	22-13
22-3 Average Meteorological Conditions near the Oklahoma Monitoring Sites	22-16
22-4 Risk-Based Screening Results for the Oklahoma Monitoring Sites	22-27
22-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Oklahoma Monitoring Sites	22-32
22-6 Risk Approximations for the Oklahoma Monitoring Sites	22-85
22-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Oklahoma Monitoring Sites	22-89
22-8	Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Oklahoma Monitoring
Sites	22-92
23-1	Geographical Information for the Rhode Island Monitoring Site	23-4
23-2 Population, Motor Vehicle, and Traffic Information for the Rhode Island Monitoring
Site	23-5
23-3 Average Meteorological Conditions near the Rhode Island Monitoring Site	23-7
23-4 Risk-Based Screening Results for the Rhode Island Monitoring Site	23-11
23-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Rhode Island Monitoring Site	23-12
23-6 Risk Approximations for the Rhode Island Monitoring Site	23-15
23-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Rhode Island Monitoring Site	23-17
23-8	Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Rhode Island
Monitoring Site	23-18
24-1	Geographical Information for the South Carolina Monitoring Site	24-4
24-2 Population, Motor Vehicle, and Traffic Information for the South Carolina
Monitoring Site	24-5
24-3 Average Meteorological Conditions near the South Carolina Monitoring Site	24-7
24-4 Risk-Based Screening Results for the South Carolina Monitoring Site	24-10
24-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
South Carolina Monitoring Site	24-11
24-6 Risk Approximations for the South Carolina Monitoring Site	24-15
24-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the South Carolina Monitoring Site	24-16
24-8 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the South Carolina
Monitoring Site	24-17
xliii

-------
LIST OF TABLES (Continued)
Page
25-1 Geographical Information for the Texas Monitoring Sites	25-6
25-2 Population, Motor Vehicle, and Traffic Information for the Texas Monitoring Sites... 25-8
25-3 Average Meteorological Conditions near the Texas Monitoring Sites	25-10
25-4 Risk-Based Screening Results for the Texas Monitoring Sites	25-15
25-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Texas Monitoring Sites	25-17
25-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Texas Monitoring Sites	25-19
25-7	Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Texas Monitoring
Sites	25-20
26-1	Geographical Information for the Utah Monitoring Site	26-4
26-2	Population, Motor Vehicle, and Traffic Information for the Utah Monitoring Site	26-5
26-3	Average Meteorological Conditions near the Utah Monitoring Site	26-7
26-4	Risk-Based Screening Results for the Utah Monitoring Site	26-11
26-5	Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Utah Monitoring Site	26-13
26-6 Risk Approximations for the Utah Monitoring Site	26-40
26-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Utah Monitoring Site	26-44
26-8	Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Utah Monitoring
Site	26-45
27-1	Geographical Information for the Vermont Monitoring Sites	27-7
27-2 Population, Motor Vehicle, and Traffic Information for the Vermont
Monitoring Sites	27-9
27-3 Average Meteorological Conditions near the Vermont Monitoring Sites	27-12
27-4 Risk-Based Screening Results for the Vermont Monitoring Sites	27-19
27-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Vermont Monitoring Sites	27-21
27-6 Risk Approximations for the Vermont Monitoring Sites	27-49
27-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Vermont Monitoring Sites	27-51
27-8	Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Vermont
Monitoring Sites	27-53
28-1	Geographical Information for the Virginia Monitoring Site	28-4
28-2 Population, Motor Vehicle, and Traffic Information for the Virginia Monitoring
Site	28-5
28-3 Average Meteorological Conditions near the Virginia Monitoring Site	28-7
28-4 Risk-Based Screening Results for the Virginia Monitoring Site	28-11
28-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Virginia Monitoring Site	28-12
xliv

-------
LIST OF TABLES (Continued)
Page
28-6 Risk Approximations for the Virginia Monitoring Site	28-15
28-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Virginia Monitoring Site	28-17
28-8	Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Virginia Monitoring
Site	28-18
29-1	Geographical Information for the Washington Monitoring Site	29-4
29-2 Population, Motor Vehicle, and Traffic Information for the Washington
Monitoring Site	29-5
29-3 Average Meteorological Conditions near the Washington Monitoring Site	29-7
29-4 Risk-Based Screening Results for the Washington Monitoring Site	29-11
29-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Washington Monitoring Site	29-13
29-6 Risk Approximations for the Washington Monitoring Site	29-31
29-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Washington Monitoring Site	29-32
29-8	Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Washington
Monitoring Site	29-33
30-1	Geographical Information for the Wisconsin Monitoring Sites	30-6
30-2 Population, Motor Vehicle, and Traffic Information for the Wisconsin
Monitoring Sites	30-8
30-3 Average Meteorological Conditions near the Wisconsin Monitoring Sites	30-11
30-4 Risk-Based Screening Results for the Wisconsin Monitoring Sites	30-16
30-5 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Wisconsin Monitoring Sites	30-17
30-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Wisconsin Monitoring Sites	30-19
30-7	Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Wisconsin
Monitoring Sites	30-20
31-1	Method Precision by Analytical Method	31-4
31-2 VOC Method Precision: Coefficient of Variation Based on Duplicate and
Collocated Samples by Site and Pollutant	31-6
31-3 SNMOC Method Precision: Coefficient of Variation Based on Duplicate Samples
by Site and Pollutant	31-14
31-4 Carbonyl Compound Method Precision: Coefficient of Variation Based on
Duplicate and Collocated Samples by Site and Pollutant	31-18
31-5 PAH Method Precision: Coefficient of Variation Based on Collocated Samples
by Site and Pollutant	31-22
31-6 Metals Method Precision: Coefficient of Variation Based on Collocated Samples
by Site and Pollutant	31-24
xlv

-------
LIST OF TABLES (Continued)
Page
31-7 Hexavalent Chromium Method Precision: Coefficient of Variation Based on
Collocated Samples by Site	31-26
31-8 Analytical Precision by Analytical Method	31 -27
31-9 VOC Analytical Precision: Coefficient of Variation Based on Replicate Analyses
by Site and Pollutant	31-29
31-10 SNMOC Analytical Precision: Coefficient of Variation Based on Replicate
Analyses by Site and Pollutant	31-40
31-11 Carbonyl Compound Analytical Precision: Coefficient of Variation Based on
Replicate Analyses by Site and Pollutant	31-47
31-12 PAH Analytical Precision: Coefficient of Variation Based on Replicate Analyses
by Site and Pollutant	31-53
31-13 PAH/Phenols Analytical Precision: Coefficient of Variation Based on Replicate
Analyses for KMMS	31-57
31-14 Metals Analytical Precision: Coefficient of Variation Based on Replicate Analyses
by Site and Pollutant	31-58
31-15 Hexavalent Chromium Analytical Precision: Coefficient of Variation Based on
Replicate Analyses by Site	31-62
31-16 TO-15 NATTS PT Audit Samples - Percent of True Value	31-64
31-17 TO-11A NATTS PT Audit Samples - Percent of True Value	31-64
31-18 TO-13A NATTS PT Audit Samples - Percent of True Value	31-64
31-19 Metals NATTS PT Audit Samples - Percent of True Value	31-65
31-20 Hexavalent Chromium NATTS PT Audit Samples - Percent of True Value	31-65
31-21	Metals NAAQS PT Audit Samples - Percent of True Value	31-65
32-1	Summary of Site-Specific Pollutants of Interest	32-24
xlvi

-------
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
CBS A
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 Ionization 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
IC
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
NAAQS
National Ambient Air Quality Standard
NATA
National-Scale Air Toxics Assessment
NATTS
National Air Toxics Trends Stations
NCDC
National Climatic Data Center
ND
Non-detect
xlvii

-------
LIST OF ACRONYMS (Continued)
NEI
National Emissions Inventory
ng/m3
Nanograms per cubic meter
NMOC
Non-Methane Organic Compound(s)
NMP
National Monitoring Programs
NO A A
National Oceanic and Atmospheric Administration
NOx
Oxides of Nitrogen
NWS
National Weather Service
PAH
Polycyclic Aromatic Hydrocarbon(s)
PAMS
Photochemical Assessment Monitoring Stations
PM
Particulate Matter
PMio
Particulate Matter less than 10 microns
POM
Polycyclic 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
UATMP
Urban Air Toxics Monitoring Program
Hg/m3
Micrograms per cubic meter
yiL
Microliter
URE
Unit Risk Estimate(s)
UTC
Universal Time Coordinated
UY
Ultraviolet
UY-VIS
Ultraviolet Visible
VMT
Vehicle Miles Traveled
VOC
Volatile Organic Compound(s)
WBAN
Weather Bureau/Army/Navy ID
xlviii

-------
Abstract
This report presents the results and conclusions from the ambient air monitoring conducted
as part of the 2013 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 2013 NMP includes data from
samples collected at 66 monitoring sites that collected 24-hour air samples, typically on a l-in-6
or l-in-12 day schedule. Thirty-four sites sampled for 59 volatile organic compounds (VOCs);
33 sites sampled for 15 carbonyl compounds; seven sites sampled for 80 speciated nonmethane
organic compounds (SNMOCs); 24 sites sampled for 22 polycyclic aromatic hydrocarbons
(PAHs) and one additional site sampled for a subset of PAHs and four phenols; 20 sites sampled
for 11 metals; and 24 sites sampled for hexavalent chromium. Nearly 263,000 ambient air
concentrations were measured during the 2013 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 from city-to-city and from season-to-season.
The ambient air monitoring data collected during the 2013 NMP serve a wide range of
purposes. Not only do these data allow for the characterization of the nature and extent of air
pollution close to the 66 individual monitoring sites participating in these programs, but they also
exhibit trends and patterns that may be common to urban and rural environments and across the
country. Therefore, this report presents results that are specific to particular monitoring locations
and presents other results that are common to all environments. The results presented provide
additional insight into the complex nature of air pollution. The raw data are included in the
appendices of this report.
xlix

-------
1.0	Introduction
Air pollution contains many components that originate from a wide range of stationary,
mobile, and natural emissions sources. Because some of these components include air toxics that
are known or suspected to have the potential for negative human health effects, the
U.S. Environmental Protection Agency (EPA) encourages state, local, and tribal agencies to
understand and appreciate the nature and extent of toxic air pollution in their respective
locations. To achieve this goal, EPA sponsors the National Monitoring Programs (NMP), which
include the Photochemical Assessment Monitoring Stations (PAMS) network, Urban Air Toxics
Monitoring Program (UATMP), National Air Toxics Trends Stations (NATTS) network,
Community-Scale Air Toxics Ambient Monitoring (CSATAM) Program, and monitoring for
other pollutants such as Non-Methane Organic Compounds (NMOCs). The UATMP, the
NATTS, and the CSATAM programs include longer-term monitoring efforts (durations of one
year or more) at specific locations. These programs have the following program-specific
objectives (EPA, 2009a):
•	The primary objective of the UATMP is to characterize the composition and
magnitude of air toxics pollution through ambient air monitoring.
http://www.epa.gov/ttnamtil/uatm.html
•	The primary objective of the NATTS network is to obtain a statistically significant
quantity of high-quality representative air toxics measurements such that long-term
trends can be identified, http://www.epa.gov/ttnamti 1 /natts.html
•	The primary objective of the CSATAM Program is to conduct local-scale
investigative ambient air toxics monitoring projects.
http://www.epa.gov/ttnamtil/local.html
1.1	Background
The UATMP was initiated by EPA to meet the increasing need for information on air
toxics. Over the years, the program has grown in both participation and targeted pollutants (EPA,
2009a). The program has allowed for the identification of compounds that are prevalent in
ambient air and for participating agencies to screen air samples for concentrations of air toxics
that could potentially result in adverse human health effects.
The NATTS network was created to generate long-term ambient air toxics concentration
data at specific fixed sites across the country. The 10-City Pilot Program (LADCO, 2003) was
developed and implemented during 2001 and 2002, leading to the development and initial
1-1

-------
implementation of the NATTS network during 2003 and 2004. The goal of the program is to
estimate the concentrations of air toxics on a national level from fixed sites that remain active
over an extended period of time such that concentration trends (i.e., any substantial increase or
decrease over a period of time) may be identified. The data generated are also used for validating
modeling results and emissions inventories, assessing current regulatory benchmarks, and
assessing the potential for developing cancerous and noncancerous health effects (EPA, 2014a).
The initial site locations were based on existing infrastructure of monitoring site locations
(e.g., PM2.5 network) and results from preliminary air toxics programs such as the 1996 National-
Scale Air Toxics Assessment (NATA), which used air toxics emissions data to model ambient
monitoring concentrations across the nation. Monitoring sites were placed in both urban and
rural locations. Urban areas were chosen to measure population exposure, while rural areas were
chosen to determine background levels of air pollution and to assess impacts to non-urban areas
(EPA, 2009b). Currently, 27 NATTS sites are strategically placed across the country (EPA,
2014a).
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 in 2013. 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
1-2

-------
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.
In past reports, measurements from UATMP, NATTS, and CSATAM monitoring sites
have been presented together and referred to as "UATMP sites." In more recent reports, a
distinction has been made among the three programs due to the increasing number of sites
covered under each program. Thus, it is appropriate to describe each program; to distinguish
among their purposes and scopes; and to integrate the data, which allows each program's
objectives and goals to complement one another.
Included in this report are data collected at 66 monitoring sites around the country. The
66 sites whose data are included in this report are located in or near 40 urban or rural locations in
25 states and the District of Columbia, including 38 metropolitan or micropolitan statistical areas
(collectively referred to as core-based statistical areas or CBSAs).
This report provides both a qualitative overview of air toxics pollution at participating
urban and rural locations and a quantitative data analysis of the factors that appear to most
significantly affect the behavior of air toxics in urban and rural areas. This report also focuses on
data characterizations for each of the 66 different air monitoring locations, a site-specific
approach that allows for a much more detailed evaluation of the factors (e.g., emissions sources,
natural sources, meteorological influences) that affect air quality differently from one location to
the next. Much of the data analysis and interpretation contained in this report focuses on
pollutant-specific risk potential.
This report offers participating agencies relevant information and insight into important
air quality issues. For example, participating agencies can use trends and patterns in the
monitoring data to determine whether levels of air pollution present public health concerns, to
identify which emissions sources contribute most to air pollution, or to forecast whether
proposed pollution control initiatives could significantly improve air quality. Monitoring data
may also be compared to modeling results, such as from EPA's NATA. Policy-relevant
questions that the monitoring data may help answer include the following:
• Which anthropogenic sources substantially affect air quality?
1-3

-------
•	Have pollutant concentrations decreased as a result of regulations (or increased
despite regulation)?
•	Which pollutants contribute the greatest health risk on a short-term, intermediate-
term, and long-term basis?
The data analyses contained in this report are applied to each participating UATMP,
NATTS, or C SAT AM monitoring site, depending upon pollutants sampled and duration of
sampling. Although many types of analyses are presented, state and local environmental agencies
are encouraged to perform additional evaluations of the monitoring data so that the many factors
that affect their specific ambient air quality can be understood fully.
To facilitate examination of the 2013 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, 2014b).
This report is organized into 33 sections and 18 appendices. While each state section is
designed to be a stand-alone section to allow those interested in a particular site or state to
understand the associated data analyses without having to read the entire report, it is
recommended that Sections 1 through 4 (Introduction, Monitoring Programs Network overview,
Data Treatments and Methods, and Summary of NMP Data) and Sections 31 and 32 (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 2013 National Monitoring Programs Report
Report
Section
Section Title
Overview of Contents
1
Introduction
This section serves as an introduction to the background
and scope of the NMP (specifically, the UATMP,
NATTS, and CSATAM Programs).
2
The 2013 National Monitoring
Programs Network
This section provides information on the 2013 NMP
monitoring effort, including:
•	Monitoring locations
•	Pollutants selected for monitoring
•	Sampling and analytical methods
•	Sampling schedules
•	Completeness of the air monitoring programs.
1-4

-------
Table 1-1. Organization of the 2013 National Monitoring Programs Report (Continued)
Report
Section
Section Title
Overview of Contents
3
Summary of the 2013 National
Monitoring Programs Data
Treatments and Methods
This section presents and discusses the data treatments
applied to the 2013 NMP data to determine significant
trends and relationships in the data, characterize data
based on how ambient air concentrations varied with
monitoring location and with time, interpret the
significance of the observed spatial and temporal
variations, and evaluate human health risk.
4
Summary of the 2013 National
Monitoring Programs Data
This section presents and discusses the results of the data
treatments from the 2013 NMP data.
5
Site in Alaska
Monitoring results for the site in the Anchorage, AK
CBS A (ANAK)
6
Sites in Arizona
Monitoring results for the sites in the Phoenix-Mesa-
Scottsdale, AZ CBSA (PXSS and SPAZ)
7
Sites in California
Monitoring results for the sites in the Los Angeles-Long
Beach-Anaheim, CA CBSA (CELA and LBHCA), the
Riverside-San Bernardino-Ontario, CA CBSA (RUCA),
and the San Jose-Sunnyvale-Santa Clara, CA CBSA
(SJJCA)
8
Sites in Colorado
Monitoring results for the sites in the Grand Junction, CO
CBSA (GPCO) and the Glenwood Springs, CO CBSA
(BMCO, BRCO, PACO, RFCO, and RICO)
9
Site in the District of Columbia
Monitoring results for the site in the Washington-
Arlington-Alexandria, DC-VA-MD-WV CBSA (WADC)
10
Sites in Florida
Monitoring results for the sites in the Miami-Fort
Lauderdale-West Palm Beach, FL CBSA (WPFL), the
Orlando-Kissimmee-Sanford, FL CBSA (ORFL and
PAFL), and the Tampa-St. Petersburg-Clearwater, FL
CBSA (AZFL, SKFL, and SYFL)
11
Site in Georgia
Monitoring results for the site in the Atlanta-Sandy
Springs-Roswell, GA CBSA (SDGA)
12
Sites in Illinois
Monitoring results for the sites in the Chicago-Naperville-
Elgin, IL-IN-WI CBSA (NBIL and SPIL) and the St.
Louis, MO-IL CBSA (ROIL)
13
Sites in Indiana
Monitoring results for the sites in the Chicago-
Naperville-Elgin, IL-IN-WI CBSA (INDEM) and the
Indianapolis-Carmel-Anderson, IN CBSA (WPIN)
14
Sites in Kentucky
Monitoring results for the sites in the Huntington-
Ashland, WV-KY-OH CBSA (ASKY and ASKY-M), the
Lexington-Fayette, KY CBSA (LEKY), the Evansville,
IN-KY CBSA (BAKY), the Paducah, KY-IL CBSA
(BLKY), and the sites in Marshall County (ATKY,
CCKY, LAKY, and TVKY) and Carter County (GLKY)
15
Site in Massachusetts
Monitoring results for the site in the Boston-Cambridge-
Newton, MA-NH CBSA (BOMA)
16
Site in Michigan
Monitoring results for the site in the Detroit-Warren-
Dearborn, MI CBSA (DEMI)
1-5

-------
Table 1-1. Organization of the 2013 National Monitoring Programs Report (Continued)
Report
Section
Section Title
Overview of Contents
17
Site in Minnesota
Monitoring results for the site in the St. Cloud, MN
CBSA (STMN)
18
Sites in Mississippi
Monitoring results for the sites in the Columbus, MS
CBSA (KMMS and SSMS)
19
Site in Missouri
Monitoring results for the site in the St. Louis, MO-IL
CBSA (S4MO)
20
Sites in New Jersey
Monitoring results for the sites in the New York-Newark-
Jersey City, NY-NJ-PA CBSA (CHNJ, ELNJ, and NBNJ)
and the Philadelphia-Camden-Wilmington, PA-NJ-DE-
MD CBSA (CSNJ)
21
Sites in New York
Monitoring results for the sites in the New York-Newark-
Jersey City, NY-NJ-PA CBSA (BXNY) and the
Rochester, NY CBSA (ROCH)
22
Sites in Oklahoma
Monitoring results for the sites in the Tulsa, OK CBSA
(TOOK, TMOK, and TROK), and the Oklahoma City,
OK CBSA (ADOK, OCOK, and YUOK)
23
Site in Rhode Island
Monitoring results for the site in the Providence-
Warwick, RI-MA CBSA (PRRI)
24
Site in South Carolina
Monitoring results for the site in Chesterfield County, SC
(CHSC)
25
Sites in Texas
Monitoring results for the sites in the Houston-The
Woodlands-Sugar Land, TX CBSA (CAMS 35) and the
Marshall, TX CBSA (CAMS 85)
26
Site in Utah
Monitoring results for the site in the Ogden-Clearfield,
UT CBSA (BTUT)
27
Sites in Vermont
Monitoring results for the sites in the Burlington-South
Burlington, VT CBSA (BURVT and UNVT) and the
Rutland, VT CBSA (RUVT)
28
Site in Virginia
Monitoring results for the site in the Richmond, VA
CBSA (RIVA)
29
Site in Washington
Monitoring results for the site in the Seattle-Tacoma-
Bellevue, WA CBSA (SEWA)
30
Sites in Wisconsin
Monitoring results for the sites in the Beaver Dam, WI
CBSA (HOWI) and the Milwaukee-Waukesha-West
Allis, WI CBSA (MIWI)
31
Data Quality
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 2013 NMP ambient air
monitoring data.
32
Results, Conclusions, and
Recommendations
This section summarizes the most significant findings of
the report and makes several recommendations for future
projects that involve ambient air monitoring.
33
References
This section lists the references cited throughout the
report.
1-6

-------
2.0	The 2013 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, North Carolina. Data from
66 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 for Speciated Nonmethane Organic
Compounds (SNMOCs) and/or Volatile Organic
Compounds (VOCs) using EPA Compendium Method
TO-15, carbonyl compounds from sorbent cartridge
samples using EPA Compendium Method TO-11 A,
polycyclic aromatic hydrocarbons (PAHs) from
polyurethane foam (PUF) and XAD-2® resin samples
and/or phenols from XAD-2® resin samples using EPA
Compendium Method TO-13 A, trace metals from
filters using EPA Compendium Method IO-3.5/Federal
Equivalency Methods (FEM) EQL-0512-201 and EQL-
0512-202, and hexavalent chromium from sodium bicarbonate-coated filters using ASTM
D7614. Section 2.2 provides additional information regarding each of the sampling
methodologies used to collect and analyze samples.
The following sections review the monitoring locations, pollutants selected for
monitoring, sampling and analytical methods, collection schedules, and completeness of the
2013 NMP dataset.
2.1	Monitoring Locations
For the NATTS network, monitor siting is based on the need to assess population
exposure and/or background-level concentrations. For the UATMP and 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, Illinois and Phoenix, Arizona), while others were placed in
moderately populated rural areas (e.g., Horicon, Wisconsin and Chesterfield, South Carolina).
2-1
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.

-------
Figure 2-1 shows the locations of the 66 monitoring sites participating in the 2013
programs, which encompass 40 different urban and rural areas. Outlined in Figure 2-1 are the
associated CBSAs, as designated by the U.S. Census Bureau, where each site is located (Census
Bureau, 2013a). A CBSA refers to either a metropolitan (an urban area with 50,000 or more
people) or micropolitan (an urban area with at least 10,000 people but less than 50,000 people)
statistical area (Census Bureau, 2013b).
Table 2-1 lists the respective monitoring program and the years of program participation
for the 66 monitoring sites. Sixty-one monitoring sites have been included in previous annual
reports, including two that are returning for the first time in five or more years; these two sites
are highlighted in purple in Table 2-1. Five monitoring sites are new to their respective programs
for 2013; these sites are highlighted in green in Table 2-1.
As Figure 2-1 and Table 2-1 show, the 2013 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. 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, five, or six-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, five, or six-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 2013 National Monitoring Programs Monitoring Sites
to
Seattle. WA
-^Underhill. VT
Burlington, VT
St. Cloud. MN
Rutland. VT
l_J/$Boston. MA
vj { }	Rochester, NY
Honcon, Wl |	) i
	A Si \ Dearborn', Ml	—S
-v	)	Milwaukee, Wl g J	f	Chester, NJf
NorthbrookT11 YTBi /	\
\ Schiller Park. IL'CfGary, IN	I Camder^NJ.
1	—E.a!
Lexington? KY .——* ^
¦Tulsa, OK (3) _ . ^	,
r V	Calvert City, KY (5)	1
Bountiful. UT
/	7"Rifle-CO_ I
Parachute. CO J m co
Battlement Mesa. CO~*&Catbonda\e.
I Grand Junction, CO
Los Angeles, CA
Rubidoux, CA
Yukon, OK
Phoenix. A1
Long Beach, CA
Chesterfield. SC
Columbus'MS (2)4
Decatur. GA
Karnack, TX £
Valrlco, FL
Pinellas Park. FL
St. Petersburg, FL
Belle Glade. FL
Anchorage. AK
Providence, Rl
New York, NY
Elizabeth. NJ
North Brunswick
Township, NJ
Washington. DC
Tr
Winter Park, FL
¦ Orlando. FL
Legend
Program
CSATAM
NATTS
J AT MP

CBSA


-------
Table 2-1. 2013 National Monitoring Programs Sites and Past Program Participation
Monitoring Location
and Site
Program
2003 and Earlier
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
Anchorage, AK (ANAK)
CSATAM





~




~
Ashland. KY (ASKY)
UATMP









~
~
Ashland, KY (ASKY-M)
UATMP









~
~
Basket! KY (BAKY)
UATMP









~
~
Battlement Mesa, CO (BMCO)
UATMP







~
~
~
~
Belle Glade, FL (WPFL)
UATMP
2002-2003









~
Boston, MA (BOMA)
NATTS
2003
~
~
~
~
~
~
~
~
~
~
Bountiful. UT (BTUT)
NATTS
2003
~
~
~
~
~
~
~
~
~
~
Burlington, VT (BURVT)
UATMP






~
~
~
~
~
Calvert City, KY (ATKY)
UATMP









~
~
Calvert City, KY (CCKY)
UATMP









~
~
Calvert City, KY (LAKY)
UATMP









~
~
Calvert City, KY (TVKY)
UATMP









~
~
Camden, NJ (CSNJ)
UATMP










~
Carbondale, CO (RFCO)
UATMP









~
~
Chester, NJ (CHNJ)
UATMP
2001-2003
~
~
~
~
~
~
~
~
~
~
Chesterfield, SC (CHSC)
NATTS


~
~
~
~
~
~
~
~
~
Green shading indicates new site participating in the NMP.
Purple shading indicates returning site with past NMP participation.
BOLD ITALICS = EPA-designated NATTS site

-------
Table 2-1. 2013 National Monitoring Programs Sites and Past Program Participation (Continued)
Monitoring Location
and Site
Program
2003 and Earlier
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
Columbus, MS (KMMS)
UATMP










~
Columbus, MS (SMMS)
UATMP










~
Dearborn MI (DEMI)
NATTS
2001-2003
~
~
~
~
~
~
~
~
~
~
Decatur, GA (SDGA)
NATTS


~
~
~
~
~
~
~
~
~
Deer Park, TX (CAMS 35)
NATTS




~
~
~
~
~
~
~
East Highland Park, VA (RIVA)
NATTS





~
~
~
~
~
~
Elizabeth NJ (ELNJ)
UATMP
1999-2003
~
~
~
~
~
~
~
~
~
~
Gary, IN (INDEM)
UATMP

~
~
~
~
~
~
~
~
~
~
Grand Junction, CO (GPCO)
NATTS

~
~
~
~
~
~
~
~
~
~
Grayson, KY (GLKY)
NATTS





~
~
~
~
~
~
Horicon, WI (HOWI)
NATTS







~
~
~
~
Indianapolis, IN (WPIN)
UATMP



~
~
~
~
~
~
~
~
Karnack, TX (CAMS 85)
NATTS




~


~
~
~
~
Lexington KY (LEKY)
UATMP









~
~
Long Beach, CA (LBHCA)
CSATAM









~
~
Los Angeles, CA (CELA)
NATTS




~
~
~
~
~
~
~
Milwaukee, WI (MIWI)
UATMP









~
~
Green shading indicates new site participating in the NMP.
Purple shading indicates returning site with past NMP participation.
BOLD ITALICS = EPA-designated NATTS site

-------
Table 2-1. 2013 National Monitoring Programs Sites and Past Program Participation (Continued)
Monitoring Location
and Site
Program
2003 and Earlier
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
New York, NY (BXNY)
NATTS



~
~
~
~
~

~
~
North Brunswick, NJ (NBNJ)
UATMP
2001-2003
~
~
~
~
~
~
~
~
~
~
Northbrook, IL (NBIL)
NATTS
2003
~
~
~
~
~
~
~
~
~
~
Oklahoma City, OK (ADOK)
UATMP









~
~
Oklahoma City, OK (OCOK)
UATMP






~
~
~
~
~
Orlando, FL (PAFL)
UATMP





~
~
~
~
~
~
Parachute, CO (PACO)
UATMP





~
~
~
~
~
~
Phoenix, AZ (PXSS)
NATTS
2001-2003
~

~
~
~
~
~
~
~
~
Phoenix, AZ (SPAZ)
UATMP
2001



~
~
~
~
~
~
~
Pinellas Park, FL (SKFL)
NATTS

~
~
~
~
~
~
~
~
~
~
Providence, RI (PRRI)
NATTS


~
~
~
~
~
~
~
~
~
Rifle, CO (RICO)
UATMP





~
~
~
~
~
~
Rochester, NY (ROCH)
NATTS



~
~
~
~
~
~
~
~
Roxana, IL (ROIL)
Special
Study









~
~
Rubidoux, CA (RUCA)
NATTS




~
~
~
~
~
~
~
Rutland, VT (RUVT)
UATMP
1995-1999, 2002





~
~
~
~
~
San Jose, CA (SJJCA)
NATTS





~
~
~
~
~
~
Green shading indicates new site participating in the NMP.
Purple shading indicates returning site with past NMP participation.
BOLD ITALICS = EPA-designated NATTS site

-------
Table 2-1. 2013 National Monitoring Programs Sites and Past Program Participation (Continued)
Monitoring Location
and Site
Program
2003 and Earlier
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
Schiller Park, IL (SPIL)
UATMP
2003
~
~
~
~
~
~
~
~
~
~
Seattle, WA (SEWA)
NATTS


~
~
~
~
~
~
~
~
~
Silt, CO (BRCO)
UATMP





~
~
~
~
~
~
Smithland, KY (BLKY)
UATMP









~
~
St. Cloud, MN (STMN)
UATMP









~
~
St. Louis, MO (S4MO)
NATTS
2002, 2003
~
~
~
~
~
~
~
~
~
~
St. Petersburg, FL (AZFL)
UATMP
1991-1992, 2001-
2003
~
~
~
~
~
~
~
~
~
~
Tulsa, OK (TMOK)
UATMP






~
~
~
~
~
Tulsa, OK (TOOK)
UATMP



~
~
~
~
~
~
~
~
Tulsa, OK (TROK)
UATMP










~
Underhill, VT (LINVI1)
NATTS
2002

~
~
~
~
~
~
~
~
~
Valrico, FL (SYFL)
NATTS

~
~
~
~
~
~
~
~
~
~
Washington, D.C. (WADC)
NATTS


~
~
~
~
~
~
~
~
~
Winter Park, FL (ORFL)
UATMP
1990-1991, 2003
~
~
~
~
~
~
~
~
~
~
Yukon, OK (YUOK)
UATMP










~
Green shading indicates new site participating in the NMP.
Purple shading indicates returning site with past NMP participation.
BOLD ITALICS = EPA-designated NATTS site

-------
Table 2-2. Site Characterizing Information for the 2013 National Monitoring Programs Sites
Site
Code
AQS
Code
Location
Land Use
Location
Setting
County-level
Population3
County-level
Vehicle
Registration,
# of Vehiclesb
(Year)
Estimated
Daily Traffic,
AADTb
(Year)
County-level
Stationary
Source HAP
Emissions0
(tpy)
County-level
Mobile Source
HAP
Emissions0
(tpy)
ADOK
40-109-0042
Oklahoma City, OK
Commercial
Urban/City
Center
755,245
835,642
(2013)
34,700
(2012)
2,156.08
3,425.17
ANAK
02-020-0018
Anchorage, AK
Residential
Suburban
300,950
358,999
(2013)
20,193
(2012)
684.58
2,749.31
ASKY
21-019-0017
Ashland, KY
Residential
Suburban
48,886
39,196
(2013)
7,230
(2011)
262.71
172.53
ASKY-M
21-019-0002
Ashland, KY
Industrial
Urban/City
Center
48,886
39,196
(2013)
12,842
(2012)
262.71
172.53
ATKY
21-157-0016
Calvert City, KY
Industrial
Suburban
31,107
30,254
(2013)
3,262
(2012)
1,119.74
476.37
AZFL
12-103-0018
St. Petersburg, FL
Residential
Suburban
929,048
879,683
(2013)
42,500
(2013)
2,132.17
3,217.48
BAKY
21-101-0014
Basket!, KY
Commercial
Rural
46,347
38,811
(2013)
922
(2012)
397.98
268.40
BLKY
21-139-0004
Smithland, KY
Agricultural
Rural
9,359
8,338
(2013)
2,510
(2013)
32.24
136.07
BMCO
08-045-0019
Battlement Mesa, CO
Commercial
Suburban
57,302
74,036
(2012)
1,880
(2014)
3,787.70
327.61
BOMA
25-025-0042
Boston, MA
Commercial
Urban/City
Center
755,503
410,436
(2014)
27,654
(2010)
851.81
1,015.72
BRCO
08-045-0009
Silt, CO
Agricultural
Rural
57,302
74,036
(2012)
1,182
(2014)
3,787.70
327.61
BTUT
49-011-0004
Bountiful, UT
Residential
Suburban
322,094
274,716
(2013)
130,950
(2012)
1,163.85
930.74
BURVT
50-007-0014
Burlington, VT
Commercial
Urban/City
Center
159,515
172,203
(2013)
14,200
(2009)
432.40
477.55
BOLD ITALICS = EPA-designated NATTS site
"¦Reference: Census Bureau, 2014
individual references provided in each state section.
°Reference: 2011 NEI version 2 (EPA, 2015a)
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.
f S4MO's county-level population and vehicle registration are the sum of the county- and city-level data.

-------
Table 2-2. Site Characterizing Information for the 2013 National Monitoring Programs Sites (Continued)
Site
Code
AQS
Code
Location
Land Use
Location
Setting
County-level
Population3
County-level
Vehicle
Registration,
# of Vehiclesb
(Year)
Estimated
Daily Traffic,
AADTb
(Year)
County-level
Stationary
Source HAP
Emissions0
(tpy)
County-level
Mobile Source
HAP
Emissions0
(tpy)
BXNY
36-005-0110
New York, NY
Residential
Urban/City
Center
1,418,733
254,752
(2013)
98,899
(2012)
3,796.74
840.39
CAMS 35
48-201-1039
Deer Park, TX
Residential
Urban/City
Center
4,336,853
3,401,957
(2013)
31,043
(2004)
13,524.71
8,643.58
CAMS 85
48-203-0002
Karnack, TX
Agricultural
Rural
66,886
72,689
(2013)
1,250
(2012)
879.41
346.43
CCKY
21-157-0018
Calvert City, KY
Residential
Suburban
31,107
30,254
(2013)
4,050
(2013)
1,119.74
476.37
CELA
06-037-1103
Los Angeles, CA
Residential
Urban/City
Center
10,017,068
7,609,517
(2013)
231,000
(2013)
21,804.55
14,773.30
CHNJ
34-027-3001
Chester, NJ
Agricultural
Rural
499,397
443,969
(Ratio)d
11,215
(2012)
680.93
1,278.46
CHSC
45-025-0001
Chesterfield, SC
Forest
Rural
46,197
41,728
(2013)
700
(2013)
166.35
206.82
CSNJ
34-007-0002
Camden, NJ
Industrial
Urban/City
Center
512,854
458,294
(Ratio)d
3,231
(2012)
577.27
953.66
DEMI
26-163-0033
Dearborn, MI
Industrial
Suburban
1,775,273
1,335,516
(2013)
94,600
(2013)
7,118.74
4,563.35
ELNJ
34-039-0004
Elizabeth, NJ
Industrial
Suburban
548,256
485,427
(Ratio)d
250,000
(2006)
814.19
1,017.46
GLKY
21-043-0500
Grayson, KY
Residential
Rural
27,202
25,487
(2013)
303
(2012)
75.96
145.24
GPCOe
08-077-0017
08-077-0018
Grand Junction CO
Commercial
Urban/City
Center
147,554
176,969
(2012)
11,000
(2013)
659.65
664.73
HO M I
55-027-0001
Horicon, WI
Agricultural
Rural
88,344
99,078
(2013)
5,100
(2011)
429.32
458.47
BOLD ITALICS = EPA-designated NATTS site
"¦Reference: Census Bureau, 2014
individual references provided in each state section.
°Reference: 2011 NEI version 2 (EPA, 2015a)
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.
f S4MO's county-level population and vehicle registration are the sum of the county- and city-level data.

-------
Table 2-2. Site Characterizing Information for the 2013 National Monitoring Programs Sites (Continued)
Site
Code
AQS
Code
Location
Land Use
Location
Setting
County-level
Population3
County-level
Vehicle
Registration,
# of Vehiclesb
(Year)
Estimated
Daily Traffic,
AADTb
(Year)
County-level
Stationary
Source HAP
Emissions0
(tpy)
County-level
Mobile Source
HAP
Emissions0
(tpy)
INDEM
18-089-0022
Gary, IN
Industrial
Urban/City
Center
491,456
425,854
(2013)
34,754
(2011)
1,603.10
1,607.33
KMMS
28-087-0002
Columbus, MS
Residential
Urban/City
Center
59,922
54,826
(2013)
9,900
(2013)
1,385.56
260.29
LAKY
21-157-0019
Calvert City, KY
Residential
Suburban
31,107
30,254
(2013)
1,189
(2012)
1,119.74
476.37
LBHCA
06-037-4002
Long Beach, CA
Residential
Suburban
10,017,068
7,609,517
(2013)
285,000
(2013)
21,804.55
14,773.30
LEKY
21-067-0012
Lexington, KY
Residential
Suburban
308,428
208,983
(2013)
10,083
(2012)
764.77
1,116.04
MIWI
55-079-0026
Milwaukee, WI
Commercial
Urban/City
Center
956,023
641,582
(2013)
12,400
(2013)
2,903.89
1,966.31
NBIL
17-031-4201
Northbrook, IL
Residential
Suburban
5,240,700
2,074,419
(2014)
115,700
(2013)
15,663.06
8,882.46
NBNJ
34-023-0006
North Brunswick, NJ
Agricultural
Rural
828,919
734,425
(Ratio)d
110,653
(2009)
1,038.26
1,577.17
OCOK
40-109-1037
Oklahoma City, OK
Residential
Suburban
755,245
835,642
(2013)
41,500
(2012)
2,156.08
3,425.17
ORFL
12-095-2002
Winter Park, FL
Commercial
Urban/City
Center
1,225,267
1,181,540
(2013)
29,500
(2013)
2,774.25
4,121.46
PACO
08-045-0005
Parachute, CO
Residential
Urban/City
Center
57,302
74,036
(2012)
15,000
(2013)
3,787.70
327.61
PAFL
12-095-1004
Orlando, FL
Commercial
Suburban
1,225,267
1,181,540
(2013)
49,000
(2013)
2,774.25
4,121.46
PRRI
44-007-0022
Providence, RI
Residential
Urban/City
Center
628,600
511,015
(Ratio)d
136,800
(2009)
1,362.28
1,350.29
BOLD ITALICS = EPA-designated NATTS site
"¦Reference: Census Bureau, 2014
individual references provided in each state section.
°Reference: 2011 NEI version 2 (EPA, 2015a)
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.
f S4MO's county-level population and vehicle registration are the sum of the county- and city-level data.

-------
Table 2-2. Site Characterizing Information for the 2013 National Monitoring Programs Sites (Continued)
Site
Code
AQS
Code
Location
Land Use
Location
Setting
County-level
Population3
County-level
Vehicle
Registration,
# of Vehiclesb
(Year)
Estimated
Daily Traffic,
AADTb
(Year)
County-level
Stationary
Source HAP
Emissions0
(tpy)
County-level
Mobile Source
HAP
Emissions0
(tpy)
PXSS
04-013-9997
Phoenix, AZ
Residential
Urban/City
Center
4,009,412
3,761,859
(2012)
29,515
(2010)
7,792.15
9,915.84
RFCO
08-045-0018
Carbondale, CO
Residential
Rural
57,302
74,036
(2012)
16,000
(2013)
3,787.70
327.61
RICO
08-045-0007
Rifle, CO
Commercial
Urban/City
Center
57,302
74,036
(2012)
15,000
(2013)
3,787.70
327.61
RIVA
51-087-0014
East Highland Park,
VA
Residential
Suburban
318,611
350,000
(2013)
72,000
(2012)
888.54
746.37
ROCH
36-055-1007
Rochester, NY
Residential
Urban/City
Center
749,606
558,063
(2013)
85,162
(2012)
2,959.44
1,742.27
ROIL
17-119-9010
Roxana, IL
Industrial
Suburban
267,225
267,302
(2014)
7,750
(2013)
1,359.86
815.08
RUCA
06-065-8001
Rubidoux, CA
Residential
Suburban
2,292,507
1,788,322
(2013)
150,000
(2013)
3,826.19
3,244.32
RUVT
50-021-0002
Rutland, VT
Commercial
Urban/City
Center
60,622
79,795
(2013)
10,400
(2013)
173.22
245.32
S4MO
29-510-0085
St. Louis, MO
Residential
Urban/City
Center
l,319,860f
1,117,375
(2013)f
100,179
(2013)
939.84
611.09
SDGA
13-089-0002
Decatur, GA
Residential
Suburban
713,340
479,533
(2013)
138,470
(2012)
1,358.69
1,814.77
SEWA
53-033-0080
Seattle, WA
Residential
Urban/City
Center
2,044,449
1,791,383
(2013)
176,000
(2013)
7,310.24
6,890.17
SJJCA
06-085-0005
San Jose, CA
Commercial
Urban/City
Center
1,862,041
1,575,973
(2013)
115,000
(2012)
4,177.14
3,634.86
SKFL
12-103-0026
Pinellas Park, FL
Residential
Suburban
929,048
879,683
(2013)
47,500
(2013)
2,132.17
3,217.48
BOLD ITALICS = EPA-designated NATTS site
"¦Reference: Census Bureau, 2014
individual references provided in each state section.
°Reference: 2011 NEI version 2 (EPA, 2015a)
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.
f S4MO's county-level population and vehicle registration are the sum of the county- and city-level data.

-------
Table 2-2. Site Characterizing Information for the 2013 National Monitoring Programs Sites (Continued)
Site
Code
AQS
Code
Location
Land Use
Location
Setting
County-level
Population3
County-level
Vehicle
Registration,
# of Vehiclesb
(Year)
Estimated
Daily Traffic,
AADTb
(Year)
County-level
Stationary
Source HAP
Emissions0
(tpy)
County-level
Mobile Source
HAP
Emissions0
(tpy)
SPAZ
04-013-4003
Phoenix, AZ
Residential
Urban/City
Center
4,009,412
3,761,859
(2012)
25,952
(2011)
7,792.15
9,915.84
SPIL
17-031-3103
Schiller Park, IL
Mobile
Suburban
5,240,700
2,074,419
(2014)
186,100
(2012)
15,663.06
8,882.46
SSMS
27-087-0003
Columbus, MS
Residential
Urban/City
Center
59,922
54,826
(2013)
19,000
(2013)
1,385.56
260.29
STMN
27-145-3053
St. Cloud, MN
Industrial
Suburban
152,092
221,636
(2013)
24,100
(2009)
1,217.04
1,275.96
SYFL
12-057-3002
Valrico, FL
Residential
Rural
1,291,578
1,157,057
(2013)
10,000
(2013)
3,155.70
4,260.15
TMOK
40-143-1127
Tulsa, OK
Residential
Urban/City
Center
622,409
614,543
(2013)
12,500
(2012)
1,902.81
4,149.89
TOOK
40-143-0235
Tulsa, OK
Industrial
Urban/City
Center
622,409
614,543
(2013)
64,424
(2012)
1,902.81
4,149.89
TROK
40-143-0179
Tulsa, OK
Industrial
Urban/City
Center
622,409
614,543
(2013)
56,200
(2012)
1,902.81
4,149.89
TVKY
21-157-0014
Calvert City, KY
Industrial
Suburban
31,107
30,254
(2013)
2,230
(2011)
1,119.74
476.37
UNVT
50-007-0007
Underhill, VT
Forest
Rural
159,515
172,203
(2013)
1,100
(2011)
432.40
477.55
WADC
11-001-0043
Washington, D.C.
Commercial
Urban/City
Center
646,449
322,350
(2012)
8,700
(2011)
933.45
829.76
WPFL
12-099-0008
Belle Glade, FL
Industrial
Rural
1,372,171
1,159,114
(2013)
6,600
(2013)
4,368.66
5,197.67
BOLD ITALICS = EPA-designated NATTS site
"¦Reference: Census Bureau, 2014
individual references provided in each state section.
°Reference: 2011 NEI version 2 (EPA, 2015a)
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.
f S4MO's county-level population and vehicle registration are the sum of the county- and city-level data.

-------
Table 2-2. Site Characterizing Information for the 2013 National Monitoring Programs Sites (Continued)
Site
Code
AQS
Code
Location
Land Use
Location
Setting
County-level
Population3
County-level
Vehicle
Registration,
# of Vehiclesb
(Year)
Estimated
Daily Traffic,
AADTb
(Year)
County-level
Stationary
Source HAP
Emissions0
(tpy)
County-level
Mobile Source
HAP
Emissions0
(tpy)
WPIN
18-097-0078
Indianapolis, IN
Residential
Suburban
928,281
830,851
(2013)
143,970
(2011)
2,627.90
4,042.65
YUOK
40-017-0101
Yukon, OK
Commercial
Suburban
126,123
106,000
(2013)
45,400
(2012)
680.10
447.57
BOLD ITALICS = EPA-designated NATTS site
"¦Reference: Census Bureau, 2014
individual references provided in each state section.
0 Reference: 2011 NEIversion2 (EPA, 2015a)
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.
f S4MO's county-level population and vehicle registration are the sum of the county- and city-level data.

-------
The proximity of the monitoring sites to different emissions sources, especially industrial
facilities and heavily traveled roadways, often explains the observed spatial variations in ambient
air quality. To provide a first approximation of the potential contributions of stationary and
mobile source emissions on ambient air quality at each site, Table 2-2 also lists the following:
•	The number of 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 version 2 of 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,
VOCs, metals, and particulate matter (PM). Because the sampling and analysis required to
monitor for every component of air pollution has been prohibitively expensive, the NMP focuses
on specific pollutants that are analyzed at the laboratory using methods based on modified
versions of EPA's Compendium methods, as listed below:
•	Compendium Method TO-15 was used to measure ambient air concentrations of
59 VOCs.
•	EPA-approved SNMOC Method was used to measure 80 ozone precursors. 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. For one site (KMMS), a subset (15) of these analytes was measured in
addition to four phenols.
•	A combination of Compendium Method IO-3.5 and EPA Federal Equivalency
Methods (FEM) EQL-0512-201 andEQL-0512-202 was used to measure ambient air
concentrations of 11 metals.
2-14

-------
• ASIMMethodD7614 was used to measure ambient air concentrations of hexavalent
chromium.
The target pollutants and methods utilized varied from monitoring site to monitoring site.
The sample collection equipment at each site was installed either as a stand-alone sampler or in a
temperature-controlled enclosure (usually a trailer or a shed) with the sampling probe inlet
exposed to the ambient air. With these common setups, most monitoring sites sampled ambient
air at heights approximately 5 feet to 20 feet above local ground level.
The detection limits of the analytical methods must be considered carefully when
interpreting the corresponding ambient air monitoring data. By definition, method detection
limits (MDLs) represent the lowest concentrations at which laboratory equipment have been
experimentally determined to reliably quantify concentrations of selected pollutants to a specific
confidence level. If a pollutant's concentration in ambient air is below the method sensitivity (as
gauged by the MDL), the analytical method might not differentiate the pollutant from other
pollutants in the sample or from the random "noise" inherent in the analyses. While
quantification below the MDL is possible, the measurement reliability is lower. Therefore, when
pollutants are present at concentrations below their respective detection limits, multiple analyses
of the same sample may lead to a wide range of measurement results, including highly variable
concentrations or "non-detect" observations (i.e., the pollutant was not detected by the
instrument). Data analysts should exercise caution when interpreting monitoring data with a high
percentage of reported concentrations at levels near or below the corresponding detection limits.
MDLs are determined annually at the ERG laboratory using 40 CFR, Part 136
Appendix B procedures (EPA, 2014c) in accordance with the specifications presented in the
NATTS Technical Assistance Document (TAD) (EPA, 2009b). This procedure involves
analyzing at least seven replicate standards spiked onto the appropriate sampling media and
extracted (per analytical method). Instrument-specific detection limits (replicate analysis of
standards in solution) are not determined because sampling media background and preparation
variability would not be considered.
2-15

-------
MDLs for metals samples were calculated using the procedure described by "Appendix
D: DQ FAC Single Laboratory Procedure v2.4" (FAC, 2007), with the exception of the arsenic
MDL for Teflon® filters. The Federal Advisory Committee (FAC) MDL procedure involves
using historical blank filter data to calculate MDLs for each pollutant. For arsenic, the procedure
described in 40 CFR was used to calculate the MDL rather than the FAC procedure because this
metal is not present at a high enough level in the background on the filters.
Tables 2-3 through 2-8 identify the specific target pollutants for each analytical method
and their corresponding MDLs, as determined for 2013. For the VOC and SNMOC analyses, the
experimentally-determined MDLs do not change within a given year unless the sample was
diluted. The 2013 VOC and SNMOC MDLs are presented in Tables 2-3 and 2-4, respectively.
For the rest of the analytical methods, the MDLs vary due to the actual volume pulled through
the sample or if the sample was diluted. For these analyses, the range and average MDL is
presented for each pollutant in Tables 2-5 through 2-8, based on valid samples. If the MDLs
presented in Tables 2-5 through 2-8 include an MDL for a diluted sample, the MDL may appear
elevated. Dilutions cause the MDL to increase by a factor of the dilution; MDLs affected by
dilution are denoted in the tables. ERG's published pollutant-specific MDLs are also presented in
Appendix B.
The following discussion presents an overview of the sampling and analytical methods.
For detailed descriptions of the methods, refer to EPA's original documentation of the
Compendium Methods (EPA, 1998; EPA, 1999a; EPA, 1999b; EPA, 1999c; EPA, 1999d; EPA
2012a; ASTM, 2012; ASTM, 2013).
2.2.1 VOC and SNMOC Concurrent Sampling and Analytical Methods
VOC and SNMOC sampling and analysis can be performed concurrently using a
combined methodology based on EPA Compendium Method TO-15 (EPA, 1999a) and the
procedure presented in EPA's "Technical Assistance Document for Sampling and Analysis of
Ozone Precursors" (EPA, 1998), respectively. When referring to SNMOC, this report may refer
to this method as the "concurrent SNMOC method" or "concurrent SNMOC analysis" because
both methods can be employed at the same time to analyze the same sample. Ambient air
samples for VOC and/or SNMOC analysis were collected in passivated stainless steel canisters.
The ERG laboratory distributed the prepared canisters (i.e., cleaned and evacuated) to the
2-16

-------
monitoring sites before each scheduled sample collection event, and site operators connected the
canisters to air sampling equipment prior to each sample day. Prior to field sampling, the
passivated canisters had internal pressures much lower than atmospheric pressure. Using this
pressure differential, ambient air flowed into the canisters automatically once an associated
system solenoid valve was opened. A mass flow controller on the sampling device inlet ensured
that ambient air entered the canister at an integrated constant rate across the collection period. At
the end of the 24-hour sampling period, the solenoid valve automatically closed and stopped
ambient air from flowing into the canister. Site operators recovered and returned the canisters,
along with the Chain of Custody (COC) forms and all associated documentation, to the ERG
laboratory for analysis.
By analyzing each sample with gas chromatography incorporating mass spectrometry
(operating in the Selected Ion Monitoring (SIM) mode) and flame ionization detection
(GC/MS-FID), laboratory staff determined ambient air concentrations of 59 VOCs and/or
80 SNMOCs, and calculated the total nonmethane organic compounds (TNMOC) concentration.
TNMOC is the sum of all hydrocarbon concentrations within the sample. Because w-xylene and
/^-xylene elute from the GC column at the same time, both the VOC and SNMOC analytical
methods report only the sum concentration for these two isomers, and not the separate
concentration for each isomer. Raw data for both methods are presented in Appendices C and D.
Table 2-3 presents the MDLs for the laboratory analysis of VOC samples with
Method TO-15 and Table 2-4 presents the MDLs for the analysis of SNMOC samples. The MDL
for every VOC is less than or equal to 0.047 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 or equal to 0.56 ppbC.
2-17

-------
Table 2-3. 2013 VOC Method Detection Limits
Pollutant
2013
MDL
(ppbv)
Pollutant
2013
MDL
(ppbv)
Acetonitrile
0.031
Dichloromethane
0.014
Acetylene
0.017
1,2-Dichloropropane
0.017
Acrolein
0.047
cis-1,3 -Dichloropropene
0.014
Acrylonitrile
0.025
trans-1,3 -Dichloropropene
0.016
fcrt-Amvl Methyl Ether
0.013
Dichlorotetrafluoroethane
0.011
Benzene
0.019
Ethyl Aery late
0.015
Bromochloromethane
0.016
Ethyl terf-Butyl Ether
0.014
Bromodichloromethane
0.019
Ethylbenzene
0.017
Bromofonn
0.021
Hexachloro-1,3 -Butadiene
0.028
Bromomethane
0.011
Methyl Isobutyl Ketone
0.018
1.3 -Butadiene
0.011
Methyl Methacrylate
0.013
Carbon Disulfide
0.011
Methyl tert-Butyl Ether
0.012
Carbon Tetrachloride
0.016
n-Octane
0.012
Chlorobenzene
0.018
Propylene
0.036
Chloroethane
0.011
Styrene
0.018
Clilorofonn
0.015
1.1.2,2-Tetrachloroethane
0.026
Cliloromethane
0.013
Tetracliloroethylene
0.014
Cliloroprene
0.012
Toluene
0.015
Dibromochloromethane
0.018
1,2,4-Trichlorobenzene
0.024
1,2-Dibromoethane
0.017
1.1.1 -Tricliloroethane
0.015
w-Dichlorobenzene
0.026
1,1,2-Trichloroethane
0.019
o-Diclilorobenzene
0.023
Tricliloroethylene
0.016
p-Dichlorobenzene
0.023
Triclilorofluoromethane
0.012
Diclilorodifluoromethane
0.011
Triclilorotrifluoroethane
0.013
1,1 -Dichloroethane
0.015
1,2,4-Trimethylbenzene
0.018
1,2-Dichloroethane
0.016
1,3,5-Trimethylbenzene
0.019
1,1 -Dichloroethene
0.011
Vinyl Cliloride
0.011
cis-1,2-Dichloroethylene
0.016
m,p-XXylene1
0.029
1.2-Dichloroetlnlene
0.012
o-Xylene
0.016
1 Because «/-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.
2-18

-------
Table 2-4. 2013 SNMOC Method Detection Limits

2013

2013

2013

MDL

MDL

MDL
Pollutant
(ppbC)1
Pollutant
(ppbC)1
Pollutant
(ppbC)1
Acetylene
0.22
1-Heptene
0.19
1-Pentene
0.18
Benzene
0.27
n-Hexane
0.30
67.Y-2-Pcntcnc
0.21
1.3 -Butadiene
0.19
1-Hexene
0.44
trans-2-Pentene
0.16
//-Butane
0.09
67.Y-2-Hc\cnc
0.19
fl-Pincnc
0.19
1-Butene
0.16
;ra/?.v-2-He\cne
0.19
/>-Pincnc
0.19
67.Y-2-Butcnc
0.10
Isobutane
0.09
Propane
0.11
;ra«.v-2-Butcne
0.10
Isobutylene
0.19
/7-Propylbenzene
0.14
Cyclohexane
0.20
Isopentane
0.09
Propylene
0.12
Cyclopentane
0.07
Isoprene
0.25
Propyne
0.19
Cyclopentene
0.19
Isopropylbenzene
0.17
Styrene
0.56
w-Decane
0.19
2-Methyl-1 -Butene
0.19
Toluene
0.20
1-Decene
0.19
3 -Methyl-1 -Butene
0.19
/7-Tridecane
0.19
///-Diethylbenzene
0.24
2-Methyl-1 -Pentene
0.19
1-Tridecene
0.19
p-Diethylbenzene
0.21
4-Methyl-1 -Pentene
0.19
1,2,3 -Trimethylbenzene
0.15
2,2-Dimethylbutane
0.10
2-Methyl-2-Butene
0.19
1,2,4-Trimethylbenzene
0.20
2,3 -Dimethylbutane
0.13
Methylcyclohexane
0.25
1,3,5 -T rime thy lbenzene
0.12
2,3 -Dimethylpentane
0.36
Methylcyclopentane
0.15
2,2,3 -T rime thy lpentane
0.19
2,4-Dimethylpentane
0.32
2-Methylheptane
0.22
2,2,4 -T rime thy lpentane
0.17
n-Dodecane
0.33
3-Methylheptane
0.17
2,3,4-Trimethy lpentane
0.14
1-Dodecene
0.19
2-Methylhexane
0.20
n-Undecane
0.22
Ethane
0.18
3-Methylhexane
0.45
1-Undecene
0.19
2-Ethyl-l-butene
0.19
2-Methylpentane
0.12
/w-Xylene/p-Xylene2
0.17
Ethylbenzene
0.14
3-Methylpentane
0.13
o-Xylene
0.10
Ethylene
0.09
n-Nonane
0.11
Sum of Knowns
NA
///-Ethyltoluene
0.11
1-Nonene
0.19
Sum of Unknowns
NA
o-Ethyltoluene
0.15
n-Octane
0.27
TNMOC
NA
p-Ethyltoluene
0.20
1-Octene
0.19

//-Heptane
0.17
n-Pentane
0.06
1	Concentration in ppbC = concentration in ppbv X number of carbon atoms in the compound.
2	Because ///-xylene and /^-xylene elute from the GC column at the same time, the SNMOC analytical method
reports the sum concentration for these two isomers and not concentrations of the individual isomers.
NA = Not applicable
2-19

-------
2.2.2 Carbonyl Compound Sampling and Analytical Method
Sampling and analysis for carbonyl compounds was performed using methodology based
on EPA Compendium Method TO-11A (EPA, 1999b). Ambient air samples for carbonyl
compound analysis were collected by passing ambient air through an ozone scrubber and then
through cartridges containing silica gel coated with 2,4-dinitrophenylhydrazine (DNPH), a
compound known to react selectively and reversibly with many aldehydes and ketones. Carbonyl
compounds in ambient air are retained in the sampling cartridge, while other compounds pass
through without reacting with the DNPH-coated matrix. The ERG laboratory distributed the
DNPH cartridges to the monitoring sites prior to each scheduled sample collection event and site
operators connected the cartridges to the air sampling equipment. After each 24-hour sampling
period, site operators recovered the cartridges and returned them, along with the COC forms and
all associated documentation, to the ERG laboratory for analysis.
To quantify concentrations of carbonyl compounds in the sampled ambient air, laboratory
analysts extracted the exposed DNPH cartridges with acetonitrile. High-performance liquid
chromatography (HPLC) analysis and ultraviolet (UV) detection of these solutions determined
the relative amounts of individual carbonyl compounds present in the original air sample.
Because the three tolualdehyde isomers elute from the HPLC column at the same time, the
carbonyl compound analytical method reports only the sum concentration for these isomers, and
not the separate concentrations for each isomer. Raw data for Method TO-11A are presented in
Appendix E.
Table 2-5 lists the MDLs reported by the ERG laboratory for measuring concentrations of
15 carbonyl compounds. Although the sensitivity varies from pollutant-to-pollutant and 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 or equal
to 0.016 ppbv.
2-20

-------
Table 2-5. 2013 Carbonyl Compound Method Detection Limits

Minimum
Maximum
Average

MDL
MDL
MDL
Pollutant
(ppbv)
(ppbv)
(ppbv)
Acetaldehyde
0.005
0.020
0.008
Acetone
0.010
0.0422
0.016
Benzaldehyde
0.002
0.008
0.003
2-Butanone
0.001
0.0092
0.002
Butyraldehyde
0.002
0.008
0.003
Crotonaldehyde
0.003
0.011
0.004
2,5-Dimethylbenzaldehyde
0.001
0.006
0.002
Formaldehyde
0.008
0.0362
0.013
Hexaldehyde
0.001
0.006
0.002
Isovaleraldehyde
0.001
0.006
0.002
Propionaldehyde
0.003
0.011
0.004
Tolualdehydes1
0.003
0.011
0.004
Valeraldehyde
0.002
0.008
0.003
1 The three tolualdehyde isomers elute from the HPLC column at the same time; thus,
the analytical method reports only the sum concentration for these three isomers and
not the individual concentrations.
indicates that sample dilution was required to perform analysis.
2.2.3 PAH Sampling and Analytical Method
PAH sampling and analysis was performed using methodology based on EPA
Compendium Method TO-13A (EPA, 1999c) and ASTM D6209 (ASTM, 2013). The ERG
laboratory prepared sampling media and supplied them to the sites before each scheduled sample
collection event. The clean sampling PUF/XAD-2® cartridge and glass fiber filter are installed in
a high volume sampler by the site operators and allowed to sample for 24 hours. Sample
collection modules and COC forms and all associated documentation were returned to the ERG
laboratory after sample collection. Within 14 days of sampling, the filter and cartridge are
extracted together using a toluene in hexane solution using the Dionex Accelerated Solvent
Extractor (ASE) 350 or ASE 300. The sample extract is concentrated to a final volume of
1.0 milliliter (mL). A volume of 1 microliter ([iL) 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
0.030 ng/m3 (chrysene) to 0.275 ng/m3 (naphthalene).
2-21

-------
Table 2-6. 2013 PAH Method Detection Limits

Minimum
Maximum
Average

MDL
MDL
MDL
Pollutant
(ng/m3)
(ng/m3)1
(ng/m3)
Acenaphthene
0.029
0.500
0.048
Acenaphthylene
0.029
0.513
0.049
Anthracene
0.021
0.363
0.035
Benzo(a)anthracene
0.058
1.01
0.097
Benzo(a)pyrene
0.039
0.675
0.065
Benzo(b)fluoranthene
0.030
0.531
0.051
Benzo(e)pyrene
0.039
0.679
0.065
Benzo(g,h,i)perylene
0.027
0.477
0.046
B enzo (k)fluoranthene
0.038
0.670
0.064
Chrysene
0.018
0.309
0.030
Coronene
0.037
0.654
0.063
Cvclopcnta|cd|pvrcnc
0.029
0.504
0.048
Dibenz(a,h)anthracene
0.028
0.485
0.046
Fluoranthene
0.034
0.589
0.056
Fluorene
0.039
0.692
0.066
9-Fluorenone
0.041
0.714
0.068
Indeno( 1,2,3 -cd)pyrene
0.025
0.438
0.042
Naphthalene
0.157
2.75
0.275
Perylene
0.028
0.483
0.046
Phenanthrene
0.030
0.531
0.051
Pyrene
0.036
0.623
0.060
Retene
0.072
1.26
0.121
indicates that sample dilution was required to perform analysis.
The PAH samples collected at the KMMS site were also performed using methodology
based on EPA Compendium Method TO-13A and ASTM D6209, although the media used and
extraction process were adjusted slightly in order to provide for the analysis of phenol and
cresols at the request of the monitoring agency. To achieve this, cartridges sent to the site
contained only XAD-2® and the filter and cartridge are extracted together using a
dichloromethane solution rather than a toluene in hexane solution. Raw data for KMMS are also
presented in Appendix F. Table 2-7 lists the MDLs for the 18 PAH target pollutants analyzed in
this manner. The average detection limit for valid samples reported by the ERG laboratory range
from 0.132 ng/m3 (phenanthrene) to 2.24 ng/m3 (phenol).
2-22

-------
Table 2-7. 2013 PAH/Phenols Method Detection Limits

Minimum
Maximum
Average

MDL
MDL
MDL
Pollutant
(ng/m3)
(ng/m3)
(ng/m3)
Acenaphthene
0.158
0.288
0.192
Anthracene
0.150
0.272
0.182
Benzo(a)anthracene
0.164
0.299
0.200
Benzo(a)pyrene
0.244
0.444
0.297
B enzo (b )fluoranthene
0.241
0.438
0.293
Benzo(g,h,i)perylene
0.200
0.365
0.244
B enzo (k)fluoranthene
0.142
0.259
0.173
Chrysene
0.208
0.378
0.253
«/,£>-Cresols23
1.30
2.38
1.59
o-Cresol2
0.678
1.24
0.825
Dibenz(a,h)anthracene
0.254
0.462
0.309
Fluoranthene
0.140
0.255
0.170
Fluorene
0.136
0.248
0.165
Indeno( 1,2,3 -cd)pyrene
0.226
0.412
0.275
Naphthalene
0.140
0.255
0.170
Phenanthrene
0.108
0.197
0.132
Phenol2
1.04
12.41
2.24
Pyrene
0.138
0.252
0.168
indicates that sample dilution was required to perform analysis,
identifies the pollutants not listed in Table 2-6.
3 Because w;-cresol and /?-crcsol elute from the GC column at the same time, the
analytical method reports the sum of w;-cresol and /?-crcsol concentrations and not
concentrations of the individual isomers.
2.2.4 Metals Sampling and Analytical Method
Ambient air samples for metals analysis were collected by passing ambient air through
either 47mm Teflon® filters or 8" x 10" quartz filters, depending on the separate and distinct
sampling apparatus used to collect the sample; the 47mm Teflon® filter is used for low-volume
samplers, whereas the 8" x 10" quartz filter is used for high-volume samplers. EPA provided the
filters to the monitoring sites. Sites sampled for either particulate matter less than 10 microns
(PMio) or total suspended particulate (TSP). Particulates in ambient air were collected on the
filters and, after a 24-hour sampling period, site operators recovered and returned the filters,
along with the COC forms and all associated documentation, to the ERG laboratory for analysis.
Extraction and analysis for the determination of speciated metals in or on particulate
matter was performed using a combination of EPA Compendium Method 10-3.5 and EPA FEM
Methods EQL-0512-201 and EQL-0512-202 (EPA, 1999d; EPA, 2012a). Upon receipt at the
laboratory, the whole filters (47mm Teflon®) or filter strips (8" x 10" quartz) were digested using
2-23

-------
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 2-8 lists the MDLs for the analysis of metals samples. Due to the difference in
sample volume/filter collection media, there are two sets of MDLs listed in Table 2-8, one for
each filter type. 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 0.003 ng/m3 (beryllium) to 2.49 ng/m3 (chromium) for the quartz filters and from
0.010 ng/m3 (cadmium) to 17.1 ng/m3 (chromium) for the Teflon® filters.
Table 2-8. 2013 Metals Method Detection Limits
Pollutant
Minimum
MDL
(ng/m3)
Maximum
MDL
(ng/m3)
Average
MDL
(ng/m3)
Pollutant
Minimum
MDL
(ng/m3)
Maximum
MDL
(ng/m3)
Average
MDL
(ng/m3)
8" X 10" Quartz Filters
47mm Teflon8 Filters
Antimony
0.011
0.019
0.015
Antimony
0.040
0.060
0.050
Arsenic
0.053
0.089
0.071
Arsenic
0.180
0.240
0.201
Beryllium
0.002
0.004
0.003
Beryllium
0.010
0.020
0.020
Cadmium
0.008
0.013
0.010
Cadmium
0.010
0.020
0.010
Chromium
1.87
3.12
2.49
Chromium
15.2
20.3
17.1
Cobalt
0.012
0.020
0.016
Cobalt
0.020
0.030
0.020
Lead
0.092
0.154
0.123
Lead
0.090
0.120
0.101
Manganese
0.099
0.165
0.132
Manganese
0.120
0.160
0.140
Mercury
0.003
0.005
0.004
Mercury
0.020
0.030
0.021
Nickel
0.945
1.58
1.26
Nickel
0.220
0.300
0.251
Selenium
0.021
0.034
0.027
Selenium
0.270
0.360
0.302
2.2.5 Hexavalent Chromium Sampling and Analytical Method
Hexavalent chromium was measured using the method described in ASTM D7614
(ASTM, 2012). Ambient air samples of hexavalent chromium on TSP were collected by passing
ambient air through sodium bicarbonate impregnated acid-washed cellulose filters. ERG
prepared and distributed the filters secured in Teflon® cartridges or in petri dishes, per the
specific sampler used at each site, to the monitoring sites prior to each scheduled sample
collection event. Site operators connected the cartridges (or installed the filters) to the air
sampling equipment. After a 24-hour sampling period, site operators recovered the cartridges (or
filters) and returned them, along with the COC forms and all associated documentation, to the
ERG laboratory for analysis. Upon receipt at the laboratory, the filters were extracted using a
2-24

-------
sodium bicarbonate solution. Ion chromatography (IC) analysis and ultraviolet-visible (UV-Vis)
detection of the extracts determined the amount of hexavalent chromium present in each sample.
Raw data for the hexavalent chromium method are presented in Appendix H.
Although the sensitivity varies from site-to-site due to the different volumes pulled
through the samples, Table 2-9 presents the range and average detection limit (0.0040 ng/m3) for
valid samples reported by the ERG laboratory across the program.
Table 2-9. 2013 Hexavalent Chromium Method Detection Limit
Pollutant
Minimum
MDL
(ng/m3)
Maximum
MDL
(ng/m3)
Average
MDL
(ng/m3)
Hexavalent Chromium
0.0032
0.0067
0.0040
2.3 Sample Collection Schedules
Table 2-10 presents the first and last date upon which sample collection occurred for each
monitoring site sampling under the NMP in 2013. 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 instrumentation at the Oklahoma City, Oklahoma site (ADOK) was relocated to a
new location in Yukon, Oklahoma (YUOK). Monitoring at ADOK was discontinued
at the end of June after which monitoring at YUOK began in July.
•	After June 30, 2013, sampling for hexavalent chromium under the NATTS program
was no longer required. As a result, several sites stopped sampling under the NMP at
this time as hexavalent chromium was the only pollutant sampled. These sites
include: the Decatur, Georgia site (SDGA), the Horicon, Wisconsin site (HOWI), the
Deer Park, Texas site (CAMS 35) and the Karnack, Texas site (CAMS 85).
•	The Milwaukee, Wisconsin (MIWI) monitoring site completed a 1-year hexavalent
chromium monitoring effort under the NMP in March. Similarly, the St. Cloud,
Minnesota (STMN) monitoring site completed a 1-year hexavalent chromium
monitoring effort under the NMP in May.
•	The Long Beach, California (LBHCA) monitoring site completed a 1-year monitoring
effort for PAHs under the NMP in July.
•	The Belle Glade, Florida monitoring site (WPFL) conducted a 1-year monitoring
effort for PAHs beginning in March 2013. To facilitate data analysis, the final
3 months of data from 2014 are included in Table 2-10 as well as the Florida state
section (Section 10).
2-25

-------
Table 2-10. 2013 Sampling Schedules and Completeness Rates
to
to
On
Site
Monitoring Period1
Carbonyl
Compounds
VOCs
Hexavalent
Chromium2
Metals
SNMOCs
PAHs
First
Sample
Last
Sample
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
ADOK
1/4/13
6/27/13
30
30
100
30
30
100
„
„
„
29
30
97
„
„
„
„
„
„
ANAK
1/4/13
12/30/13
„
„
„
61
61
100
„
„
„
„
„
„
„
„
„
62
61
>102
ASKY
1/4/13
12/30/13
61
61
100
61
61
100
„
„
„
„
„
„
„
„
„
„
„
„
ASKY-M
1/4/13
12/30/13
„
„
„
„
„
„
„
„
„
60
61
98
„
„
„
„
„
„
ATKY
1/4/13
12/30/13
„
„
„
61
61
100
„
„
„
„
„
„
„
„
„
„
„
„
AZFL
1/4/13
12/30/13
59
61
97
„
„
„
„
„
„
„
„
„
„
„
„
„
„
„
BAKY
1/4/13
12/30/13
„
„
„
„
„
„
„
„
„
60
61
98
„
„
„
„
„
„
BLKY
1/4/13
12/30/13
„
„
„
59
61
97
„
„
„
„
„
„
„
„
„
„
„
„
BMCO
1/4/13
12/24/13
28
30
933
„
„
„
„
„
„
„
„
„
55
61
90
„
„
„
BOMA
1/4/13
12/30/13
„
„
„
„
„
„
30
30
100
61
61
100
„
„
„
61
61
100
BRCO
1/4/13
12/30/13
26
30
873
„
„
„
„
„
„
„
„
„
57
61
93
„
„
„
BTUT
1/4/13
12/30/13
55
61
90
53
61
87
30
30
100
59
61
97
53
61
87
56
61
92
BURVT3
1/10/13
12/24/13
„
„
„
31
30
>100
„
„
„
„
„
„
„
„
„
„
„
„
BXNY
1/4/13
12/30/13
	
	
	
	
	
	
30
30
100
	
	
	
	
	
	
60
61
98
A = Number of valid samples collected.
B = Number of valid samples that should be collected in 2013 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	Hexavalent chromium sampling was discontinued as a required element under the NATTS program at the end of June 2013.
3	Sampling schedule was a l-in-12 day schedule rather than a l-in-6 schedule.
4	Sampling method was the adjusted TO-13 method for PAHs and phenols, and was performed for a 6-month period from March to October.
5	Sampling at WPFL was performed over a 1-year period from March 2013 to March 2014; thus, 3 months of data from 2014 are included in this table and selected parts of this report.
BOLD ITALICS = EPA-designated NATTS site.
Orange shading indicates that completeness is below the MQO of 85 percent.

-------
Table 2-10. 2013 Sampling Schedules and Completeness Rates (Continued)
to
to
-J
Site
Monitoring Period1
Carbonyl
Compounds
VOCs
Hexavalent
Chromium2
Metals
SNMOCs
PAHs
First
Sample
Last
Sample
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
CAMS 35
1/4/13
6/27/13
„
„
„
„
„
„
30
30
100
„
„
„
„
„
„
„
„
„
CAMS 85
1/4/13
6/27/13
„
„
„
„
„
„
30
30
100
„
„
„
„
„
„
„
„
„
CCKY
1/4/13
12/30/13
„
„
„
61
61
100
„
„
„
56
61
92
„
„
„
„
„
„
CELA
1/4/13
12/30/13
„
„
„
„
„
„
„
„
„
„
„
„
„
„
„
58
61
95
CHNJ
1/4/13
12/30/13
61
61
100
61
61
100
„
„
„
„
„
„
„
„
„
„
„
„
CHSC
1/4/13
12/30/13
„
„
„
„
„
„
27
30
90
„
„
„
„
„
„
58
61
95
CSNJ
1/4/13
12/30/13
59
61
97
57
61
93
„
„
„
„
„
„
„
„
„
„
„
„
DEMI
1/4/13
12/30/13
61
61
100
62
61
>100
30
30
100
„
„
„
„
„
„
60
61
98
ELNJ
1/4/13
12/30/13
61
61
100
61
61
100
„
„
„
„
„
„
„
„
„
„
„
„
GLKY
1/4/13
12/30/13
61
61
100
61
61
100
30
30
100
59
61
97
„
„
„
58
61
95
GPCO
1/4/13
12/30/13
58
61
95
61
61
100
28
30
93
„
„
„
„
„
„
56
61
92
HO M I
1/4/13
6/27/13
„
„
„
„
„
„
30
30
100
„
„
„
„
„
„
„
„
„
INDEM
1/4/13
12/30/13
61
61
100
„
„
„
„
„
„
„
„
„
„
„
„
„
„
„
KMMS
1/4/13
12/30/13
	
	
	
30
61
98
	
	
	
	
	
	
	
	
	
30
30
O
O
A = Number of valid samples collected.
B = Number of valid samples that should be collected in 2013 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	Hexavalent chromium sampling was discontinued as a required element under the NATTS program at the end of .Time 2013.
3	Sampling schedule was a l-in-12 day schedule rather than a l-in-6 schedule.
4	Sampling method was the adjusted TO-13 method for PAHs and phenols, and was performed for a 6-month period from March to October.
5	Sampling at WPFL was performed over a 1-year period from March 2013 to March 2014; thus, 3 months of data from 2014 are included in this table and selected parts of this report.
BOLD ITALICS = EPA-designated NATTS site.
Orange shading indicates that completeness is below the MQO of 85 percent.

-------
Table 2-10. 2013 Sampling Schedules and Completeness Rates (Continued)
to
to
00
Site
Monitoring Period1
Carbonyl
Compounds
VOCs
Hexavalent
Chromium2
Metals
SNMOCs
PAHs
First
Sample
Last
Sample
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
LAKY
1/4/13
12/30/13
„
„
„
60
61
98
„
„
„
„
„
„
„
„
„
„
„
„
LBHCA
1/4/13
7/27/13
„
„
„
„
„
„
„
„
„
„
„
„
„
„
„
29
35
83
LEKY
1/4/13
12/30/13
61
61
100
45
61
74
„
„
„
53
61
87
„
„
„
„
„
„
MIWI
1/4/13
3/11/13
„
„
„
„
„
„
11
12
92
„
„
„
„
„
„
„
„
„
NBIL
1/4/13
12/30/13
62
61
>100
61
61
100
30
30
100
59
61
97
61
61
100
58
61
95
NBNJ
1/4/13
12/30/13
62
61
>100
61
61
100
„
„
„
„
„
„
„
„
„
„
„
„
OCOK
1/4/13
12/30/13
61
61
100
61
61
100
„
„
„
61
61
100
„
„
„
„
„
„
ORFL
1/4/13
12/30/13
61
61
100
„
„
„
„
„
„
„
„
„
„
„
„
„
„
„
PACO
1/10/13
12/24/13
26
30
873
„
„
„
„
„
„
„
„
„
52
61
85
„
„
„
PAFL3
1/10/13
12/24/13
„
„
„
„
„
„
„
„
„
30
30
100
„
„
„
„
„
„
PRRI
1/4/13
12/30/13
„
„
„
„
„
„
30
30
100
„
„
„
„
„
„
59
61
97
PXSS
1/4/13
12/30/13
60
61
98
61
61
100
29
30
97
61
61
100
„
„
„
58
61
95
RFCO3
1/10/13
12/24/13
27
30
90
„
„
„
„
„
„
„
„
„
29
30
97
„
„
„
RICO
1/4/13
12/24/13
25
30
833
	
	
	
	
	
	
	
	
	
57
61
93
	
	
	
A = Number of valid samples collected.
B = Number of valid samples that should be collected in 2013 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	Hexavalent chromium sampling was discontinued as a required element under the NATTS program at the end of .Time 2013.
3	Sampling schedule was a l-in-12 day schedule rather than a l-in-6 schedule.
4	Sampling method was the adjusted TO-13 method for PAHs and phenols, and was performed for a 6-month period from March to October.
5	Sampling at WPFL was performed over a 1-year period from March 2013 to March 2014; thus, 3 months of data from 2014 are included in this table and selected parts of this report.
BOLD ITALICS = EPA-designated NATTS site.
Orange shading indicates that completeness is below the MQO of 85 percent.

-------
Table 2-10. 2013 Sampling Schedules and Completeness Rates (Continued)
to
to
VO
Site
Monitoring Period1
Carbonyl
Compounds
VOCs
Hexavalent
Chromium2
Metals
SNMOCs
PAHs
First
Sample
Last
Sample
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
RIVA
1/4/13
12/30/13
„
„
„
„
„
„
60
61
98
„
„
„
„
„
„
58
61
95
ROCH
1/4/13
12/30/13
„
„
„
„
„
„
30
31
97
„
„
„
„
„
„
56
61
92
ROIL
1/4/13
12/30/13
61
61
100
60
61
98
„
„
„
„
„
„
„
„
„
„
„
„
RUCA
1/4/13
12/30/13
„
„
„
„
„
„
„
„
„
„
„
„
„
„
„
58
61
95
RUVT3
1/10/13
12/24/13
„
„
„
31
30
>100
„
„
„
„
„
„
„
„
„
„
„
„
S4MO
1/4/13
12/30/13
61
61
100
61
61
100
59
61
97
61
61
100
„
„
„
60
61
98
SDGA
1/4/13
7/15/13
„
„
„
„
„
„
30
33
91
„
„
„
„
„
„
„
„
„
SEWA
1/4/13
12/30/13
57
61
93
57
61
93
29
30
97
60
61
98
„
„
„
57
61
93
SJJCA
1/4/13
12/30/13
„
„
„
„
„
„
„
„
„
60
61
98
„
„
„
59
61
97
SKFL
1/4/13
12/30/13
60
61
98
„
„
„
30
30
100
„
„
„
„
„
„
59
61
97
SPAZ3
1/4/13
12/30/13
„
„
„
31
31
100
„
„
„
„
„
„
„
„
„
„
„
„
SPIL
1/4/13
12/30/13
61
61
100
60
61
98
„
„
„
„
„
„
„
„
„
„
„
„
SSMS
1/4/13
12/30/13
„
„
„
61
61
100
„
„
„
„
„
„
„
„
„
„
„
„
STMN
1/4/13
5/28/13
	
	
	
	
	
	
24
25
96
	
	
	
	
	
	
	
	
	
A = Number of valid samples collected.
B = Number of valid samples that should be collected in 2013 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	Hexavalent chromium sampling was discontinued as a required element under the NATTS program at the end of .Time 2013.
3	Sampling schedule was a l-in-12 day schedule rather than a l-in-6 schedule.
4	Sampling method was the adjusted TO-13 method for PAHs and phenols, and was performed for a 6-month period from March to October.
5	Sampling at WPFL was performed over a 1-year period from March 2013 to March 2014; thus, 3 months of data from 2014 are included in this table and selected parts of this report.
BOLD ITALICS = EPA-designated NATTS site.
Orange shading indicates that completeness is below the MQO of 85 percent.

-------
Table 2-10. 2013 Sampling Schedules and Completeness Rates (Continued)
Site
Monitoring Period1
Carbonyl
Compounds
VOCs
Hexavalent
Chromium2
Metals
SNMOCs
PAHs
First
Sample
Last
Sample
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
SYFL
1/4/13
12/30/13
61
61
100
„
„
„
29
30
97
„
„
„
„
„
„
29
30
97
TOOK
1/4/13
12/30/13
61
61
100
61
61
100
„
„
„
58
61
95
„
„
„
„
„
„
TROK
1/4/13
12/30/13
61
61
100
61
61
100
„
„
„
56
61
92
„
„
„
„
„
„
TVKY
1/4/13
12/30/12
„
„
„
61
61
100
„
„
„
„
„
„
„
„
„
„
„
„
UNVT
1/4/13
12/30/13
„
„
„
60
61
98
28
30
93
60
61
98
„
„
„
59
61
97
WADC
1/4/13
12/30/13
„
„
„
„
„
„
30
30
100
„
„
„
„
„
„
60
61
98
WPFL35
3/11/13
3/30/14
„
„
„
„
„
„
„
„
„
„
„
„
„
„
„
30
33
91
WPIN
1/4/13
12/30/13
58
61
95
„
„
„
„
„
„
„
„
„
„
„
„
„
„
„
YUOK
7/3/13
12/30/13
30
31
97
30
31
97
—
—
—
31
31
100
—
—
—
—
—
—
A = Number of valid samples collected.
B = Number of valid samples that should be collected in 2013 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	Hexavalent chromium sampling was discontinued as a required element under the NATTS program at the end of .Time 2013.
3	Sampling schedule was a l-in-12 day schedule rather than a l-in-6 schedule.
4	Sampling method was the adjusted TO-13 method for PAHs and phenols, and was performed for a 6-month period from March to October.
5	Sampling at WPFL was performed over a 1-year period from March 2013 to March 2014; thus, 3 months of data from 2014 are included in this table and selected parts of this report.
BOLD ITALICS = EPA-designated NATTS site.
Orange shading indicates that completeness is below the MQO of 85 percent.

-------
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:
•	SNMOC samples were collected on a l-in-6 day schedule while carbonyl compounds
were collected on a l-in-12 day schedule at BMCO, BRCO, PACO, and RICO.
Sampling at RFCO was conducted on a l-in-12 day schedule for both methods.
•	The South Phoenix, Arizona site (SPAZ) collected VOC samples on a l-in-12 day
schedule.
•	The Orlando, Florida site (PAFL) collected metals samples on a l-in-12 day schedule.
•	The Belle Glade, Florida site (WPFL) collected PAH samples on a l-in-12 day
schedule.
•	The Burlington and Rutland, Vermont sites (BURVT and RUVT) collected VOC
samples on a l-in-12 day schedule.
Table 2-10 shows the following:
•	34 sites collected VOC samples.
•	33 sites collected carbonyl compound samples.
•	7 sites collected SNMOC samples.
•	24 sites collected PAH samples (with one additional site collecting PAHs/phenols).
•	20 sites collected metals samples.
•	24 sites collected hexavalent chromium samples.
As part of the sampling schedule, site operators were instructed to collect duplicate (or
collocated) samples on roughly 10 percent of the sample days for select methods when duplicate
(or collocated) samplers were available. Field blanks were collected once a month for carbonyl
compounds, hexavalent chromium, metals, and PAHs. Sampling calendars were distributed to
help site operators schedule the collection of samples, duplicates, and field blanks. In cases
where a valid sample was not collected on a given scheduled sample day, site operators were
instructed to reschedule or "make up" samples on other days. This practice explains why some
monitoring locations periodically strayed from the l-in-6 or l-in-12 day sampling schedule.
2-31

-------
The l-in-6 or l-in-12 day sampling schedule provides cost-effective approaches to data
collection for trends characterization of toxic pollutants in ambient air and ensures that sample
days are evenly distributed among the seven days of the week to allow weekday/weekend
comparison of air quality. Because the l-in-6 day schedule yields twice the number of
measurements than the l-in-12 day schedule, data characterization based on this schedule tends
to be more representative.
2.4 Completeness
Completeness refers to the number of valid samples collected and analyzed compared to
the number of total samples expected based on a l-in-6 or l-in-12 day sample schedule.
Monitoring programs that consistently generate valid samples have higher completeness than
programs that consistently have invalid samples. The completeness of an air monitoring
program, therefore, can be a qualitative measure of the reliability of air sampling and laboratory
analytical equipment as well as a measure of the efficiency with which the program is managed.
The completeness for each monitoring site and method sampled is presented in Table 2-10.
The measurement quality objective (MQO) for completeness based on the EPA-approved
Quality Assurance Project Plan (QAPP) specifies that at least 85 percent of samples from a given
monitoring site must be collected and analyzed successfully to be considered sufficient for data
trends analysis (ERG, 2013). The data in Table 2-10 show that three datasets from a total of 143
datasets from the 2013 NMP monitoring effort did not meet this MQO (orange shaded cells in
Table 2-10):
•	Sampler issues at RICO resulted in a carbonyl compound completeness less than
85 percent.
•	Sampler issues during the spring of 2013 combined with a shortened sampling
duration (monitoring was discontinued in July 2013) resulted in a PAH completeness
less than 85 percent for LBHCA.
•	A leak in the sample line was discovered at LEKY and resulted in the invalidation of
VOC samples collected between February 9, 2013 and May 4, 2013.
Appendix I identifies samples that were invalidated and lists the reason for invalidation,
based on the applied AQS null code.
2-32

-------
Table 2-11 presents method-specific completeness. Method-specific completeness was
greater than 90 percent for all methods performed under the 2013 NMP and ranged from
91.9 percent for SNMOCs to 100 percent for PAH/Phenols.
Table 2-11. Method Completeness Rates for 2013
Method
# of
Valid
Samples
# of
Samples
Scheduled
Method
Completeness
(%)
Minimum
Site-Specific
Completeness
(%)
Maximum
Site-Specific
Completeness
(%)
VOCs
1,883
1,921
98.0
74
(LEKY)
>100
(3 sites)
SNMOCs
364
396
91.9
85
(PACO)
100
(.NBIL)
Carbonyl Compounds
1,758
1,797
97.8
83
(RICO)
>100
(2 sites)
PAHs1
1,310
1,371
95.6
83
(LBHCA)
>100
(ANAK)
PAHs/Phenols
30
30
100
100
(KMMS)
Metals Analysis
1,090
1,128
96.6
87
(LEKY)
100
(6 sites)
Hexavalent Chromium
744
763
97.5
90
(¦CHSQ
100
(12 sites)
BOLD ITALICS = EPA-designated NATTS site.
'Excludes the eight samples collected at WPFL in 2014.
2-33

-------
3.0 Summary of the 2013 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 2013
NMP sampling year. These data were analyzed on
a program-wide basis as well as a site-specific
basis.
A total of 262,831 valid air toxics concentrations (including non-detects, duplicate
analyses, replicate analyses, and analyses for collocated samples) were produced from 9,418
valid samples collected at 66 monitoring sites during the 2013 reporting year. A tabular
presentation of the raw data and statistical summaries are found in Appendices C through O, as
presented in Table 3-1. Appendix P serves as the glossary for the NMP report and many of the
terms discussed and defined throughout the report are provided there.
Table 3-1. Overview and Layout of Data Presented
Pollutant Group
Number
of Sites
A
)|)cndix
Raw Data
Statistical Summary
VOCs
34
C
J
SNMOCs
7
D
K
Carbonyl Compounds
33
E
L
PAHs or PAHs/Phenols
24/1
F
M
Metals
20
G
N
Hexavalent Chromium
24
H
O
3.1 Approach to Data Treatment
This section examines the various statistical tools employed to characterize the data
collected during the 2013 sampling year. Certain data analyses were performed at the program-
level, other data analyses were performed at both the program-level and on a site-specific basis,
and still other approaches were reserved for site-specific data analyses only. Regardless of the
data analysis employed, it is important to understand how the concentration data were treated.
The following paragraphs describe techniques used to prepare this large quantity of
concentration data for data analysis.
Results from the program-wide data
analyses are presented in Section 4
while results from the site-specific
data analyses are presented in the
individual state sections, Sections 5
through 30.
	f?
3-1

-------
For each monitoring site, the primary, duplicate (or collocated), and replicate
measurements were averaged together for each pollutant in order to calculate a single
concentration per sample date and method. This is referred to as thepreprocesseddaily
measurement.
Concentrations of m,p-xylene and o-xylene were summed together and are referred to as
"total xylenes," or simply "xylenes" throughout the remainder of this report, with a few
exceptions. One exception is Section 4.1, which examines the results of basic statistical
calculations performed on the dataset. Table 4-1 and Table 4-2, which are the method-specific
statistics for VOCs and SNMOCs, respectively, present the xylenes results retained as
m,p-xylene and o-xylene species. Data for the isomers are also presented individually in the Data
Quality section (Section 31). Similarly, concentrations of /«,/;-cresol and o-cresol were also
summed together and are referred to as "cresols" throughout most of this report, with the same
exceptions as xylenes.
For the 2013 NMP, where statistical parameters are calculated based on the preprocessed
daily measurements, zeros have been substituted for non-detect results. This approach is
consistent with how data are loaded into AQS per the NATTS TAD (EPA, 2009b) as well as
other EPA air toxics monitoring programs, such as the School Air Toxics Monitoring Program
(SATMP) (EPA, 201 la), and other associated reports, 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 (|ig/m3).
However, whenever a particular sampling method is isolated from others, such as in Tables 4-1
through 4-6, the statistical parameters are presented in the unit of measure associated with the
particular sampling method. Thus, it is important to pay close attention to the unit of measure
associated with each data analysis discussed in this and subsequent sections of the report.
3-2

-------
In addition, this report presents various time-based averages to summarize the
measurements for a specific site; where applicable, quarterly and annual averages were
calculated for each site. The quarterly average of a particular pollutant is simply the average
concentration of the preprocessed daily measurements over a given calendar quarter. Quarterly
averages include the substitution of zeros for all non-detects. Quarterly averages for the first
quarter in the calendar year include measurements 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 (2013). Annual average concentrations were calculated for monitoring
sites where three quarterly averages could be calculated and where method completeness, as
presented in Section 2.4, is greater than or equal to 85 percent. Sites that do not meet these
requirements do not have an annual average concentration presented.
The concentration averages presented in this report are often provided with their
associated 95 percent confidence intervals. Confidence intervals represent the interval within
which the true average concentration falls 95 percent of the time. The confidence interval
includes an equal amount of quantities above and below the concentration average. 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).
3.2 Human Health Risk and the Pollutants of Interest
A practical approach to making an assessment on a large number of measurements is to
focus on a subset of pollutants based on the end-use of the dataset. Thus, a subset of pollutants is
selected for further data analyses for each annual NMP report. Health risk-based calculations
have been used to identify "pollutants of interest" in recent years. For the 2013 NMP report, the
3-3

-------
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 lb). Human
health risk can be further defined in terms of time. Chronic effects develop from repeated
exposure over long periods of time; acute effects develop from a single exposure or from
exposures over short periods of time (EPA, 2010a). Health risk is also route-specific; that is, risk
varies depending upon route of exposure (i.e., oral vs. inhalation). Because this report covers air
toxics in ambient air, only the inhalation route is considered. Hazardous air pollutants (HAPs)
are those pollutants "known or suspected to cause cancer or other serious health effects, such as
reproductive effects or birth defects, or adverse environmental effects" (EPA, 2014d).
Health risks are typically divided into cancer and noncancer effects when referring to
human health risk. Cancer risk is defined as the likelihood of developing cancer as a result of
exposure to a given concentration over a 70-year period, and is presented as the number of
people at risk for developing cancer per million people. Noncancer health effects include
conditions such as asthma; noncancer health risks are presented as a hazard quotient, the value
below which no adverse health effects are expected (EPA, 201 lb). Cancer risk is presented as a
probability while the hazard quotient is a ratio and thus, a unitless value.
In order to assess health risk, EPA and other agencies develop toxicity factors, such as
cancer unit risk estimates (UREs) and noncancer reference concentrations (RfCs), to estimate
cancer and noncancer risks and to identify (or screen) where air toxics concentrations may
present a human health risk. EPA has published a guidance document outlining a risk-based
screening approach for performing an initial screen of ambient air toxics monitoring datasets
(EPA, 2010a). The preliminary risk-based screening process provided in this report is an
adaption of that approach and is a risk-based methodology for analysts and interested parties to
identify which pollutants may pose a health risk in their area. Cancer UREs and noncancer RfCs
are converted into screening values. The cancer screening value is the cancer URE converted to
|ig/m3 and divided by one million. The noncancer screening value is one-tenth of the noncancer
RfC and converted from mg/m3 to |ig/m3. The final screening value used in this report is the
3-4

-------
lower of the two screening values. Not all pollutants analyzed under the NMP have screening
values; of the pollutants sampled under the NMP, 71 pollutants have screening values. The
screening values used in this analysis are presented in Appendix Q1.
The preprocessed daily measurements of the target pollutants were compared to these
chronic risk screening values in order to identify pollutants of interest across the program. The
following risk-based screening process was used to identify pollutants of interest:
1.	The TO-15 and SNMOC methods have 12 pollutants in common. If a pollutant was
measured by both the TO-15 and SNMOC methods at the same site, the TO-15
results were used. The purpose of this data treatment is to have one concentration per
pollutant for each sample day.
2.	Each preprocessed daily measurement was compared to its associated risk screening
value. Concentrations that are greater than the risk screening value are described as
"failing the screen."
3.	The number of failed screens was summed for each applicable pollutant.
4.	The percent contribution of the number of failed screens to the total number of failed
screens program-wide was calculated for each applicable pollutant.
5.	The pollutants contributing to the top 95 percent of the total failed screens were
identified as pollutants of interest.
In regards to Step 5 above, the actual cumulative contribution may exceed 95 percent in
order to include all pollutants contributing to the minimum 95 percent criteria (refer to nickel in
Table 4-7 for an example). In addition, if the 95 percent cumulative criterion is reached, but the
next pollutant contributed equally to the number of failed screens, that pollutant was also
designated as a pollutant of interest. Results of the program-wide risk-based screening process
are provided in Section 4.2.
Laboratory analysts have indicated that acetonitrile concentrations may be artificially
high (or non-existent) due to site conditions and potential cross-contamination with concurrent
sampling of carbonyl compounds using Method TO-11 A. The inclusion of acetonitrile in data
analyses must be determined on a site-specific basis by the agency responsible for the site. Thus,
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, 2014e).
3-5

-------
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 concentrations
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.
Acrolein was also excluded from the preliminary risk-based screening process due to
questions about the consistency and reliability of the measurements (EPA, 2010b). 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.
The NATTS TAD (EPA, 2009b) identifies 19 pollutants ("MQO Core Analytes") that
participating sites are required to sample and analyze for under the NATTS program. Table 3-2
presents these 19 NATTS MQO Core Analytes. Monitoring for these pollutants is required
because they are major health risk drivers according to EPA (EPA, 2009b). Many of the
pollutants listed in Table 3-2 are identified as pollutants of interest via the risk-based screening
process. Note that beginning in July 2013, hexavalent chromium was removed from the list of
required pollutants for which to sample under the NATTS program.
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 pollutants other than those targeted under the NMP. In these
cases, data are generated by sources other than ERG and are not included in the preliminary risk-
based screening process or any other data analysis contained in this report.
3-6

-------
Table 3-2. NATTS MQO Core Analytes
Pollutant
Class/Method
Acrolein
VOCs/TO-15
Benzene
1.3 -Butadiene
Carbon Tetrachloride
Chloroform
Tetrachloroethylene
Tricliloroethylene
Vinyl Chloride
Acetaldehyde
Carbonyl Compounds/
TO-11A
Formaldehyde
Naphthalene
PAHs or PAHs/Phenols/
TO-13A
Benzo(a)pyrene
Arsenic
Metals/
10-3.5 and EQL-0512-
201/202
Beryllium
Cadmium
Manganese
Lead
Nickel
Hexavalent chromium
Metals/ASTM D7614
3.3 Additional Program-Level Analyses of the 2013 National Monitoring Programs
Dataset
This section summarizes additional analyses performed on the 2013 NMP dataset at the
program level. Additional program-level analyses include an examination of the potential
contribution from 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 through 4.5.
3.3.1 The Contribution from 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 onroad (i.e., passenger vehicles) and nonroad 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
3-7

-------
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 latest version of 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 uncorrected 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.
3-8

-------
3.3.2	Variability Analyses
Variability refers to the degree of difference among values in a dataset. Two types of
variability are analyzed for this report and are discussed in Section 4.4. The first type of
variability assessed in this report is inter-site variability. For this analysis, the annual average
concentration for each site is plotted in the form of a bar graph for each program-wide pollutant
of interest. The criteria for calculating an annual average 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
compare to the program-level average for each pollutant. Note that the average concentrations
shown for VOCs, SNMOCs, and carbonyl compounds in Tables 4-1 through 4-3 are presented in
method-specific units, but have been converted to a common unit of measurement (|ig/m3) for
the purposes of this analysis.
Quarterly variability is the second type of variability assessed in this report. The
concentration data for each site were divided into the four quarters of the year, as described in
Section 3.1. The completeness criteria, also described in Section 3.1, are maintained here as well.
The site-specific quarterly averages are illustrated by bar graphs for each program-level pollutant
of interest. This analysis allows for a determination of a quarterly (or seasonal) correlation with
the magnitude of concentrations for a specific pollutant.
3.3.3	Greenhouse Gas Assessment
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, 2015b).
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
3-9

-------
phenomenon occur (NOAA, 2015a). 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 the predominant anthropogenic GHG in the atmosphere;
higher GWPs indicate a higher potential contribution to global warming (EPA, 2015c). 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.
3.4 Additional Site-Specific Analyses
In addition to the analyses described in the preceding sections, the state-specific sections
contain additional analyses that are applicable only at the local level. This section provides an
overview of these analyses but does not discuss their results. Results of these site-specific
analyses are presented in the individual state-specific sections (Sections 5 through 30).
3.4.1	Site Characterization
For each site participating in the 2013 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
registration, traffic data, and other characterizing information. For the 2013 NMP report, the
locations of point sources located near the monitoring sites were obtained from Version 2 of the
2011 NEI (EPA, 2015a). Sources for other site-characterizing data are provided in the individual
state sections.
3.4.2	Meteorological Analysis
Several site-specific meteorological analyses were performed in order to help readers
determine which meteorological factors may play a role in a given site's air quality. First, 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 2013 and one average for sample days only. These two averages provide an indication of how
3-10

-------
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, 2013 and 2014). 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 order to further characterize the meteorology at or near each monitoring site, wind
roses were constructed for each site. A wind rose shows the frequency at which a given wind
speed and direction are measured near the monitoring site, capturing day-to-day fluctuations at
the surface while allowing the predominant direction from which the wind blows to be identified.
Thus, a wind rose is often used in determining where to install an ambient monitoring site when
trying to capture emissions from an upwind source. A wind rose may also be useful in
determining whether high concentrations correlate with a specific wind direction. A wind rose
shows the frequency of wind directions as petals positioned around a 16-point compass, and uses
color or shading to represent wind speeds. Wind roses are constructed by uploading hourly 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, 2013 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. These wind roses are used
to determine if the meteorological conditions on days samples were collected were representative
3-11

-------
of conditions experienced throughout the sampling year and historically near each site. 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.
3.4.3 Preliminary Risk-Based Screening and Pollutants of Interest
The preliminary risk-based screening process described in Section 3.2 and applied at the
program-level was also completed for each individual monitoring site to determine site-specific
pollutants of interest. Once these were determined, the time-period averages (quarterly and
annual) described in Section 3.1 were calculated for each site and were used for various data
analyses at the site-specific level, as described below:
•	Comparison to the program-level concentrations
•	Trends analysis
•	The calculation of cancer risk and noncancer hazard approximations in relation to
cancer and noncancer health effects, including the emission tracer analysis
•	Risk-based emissions assessment.
3.4.3.1 Site-Specific Comparison to Program-level Average Concentrations
To better understand how an individual site's 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 each site-specific pollutant of interest
indentified via the risk-based screening process. This analysis is an extension of the analysis
discussed in Section 3.3.2 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-
level statistics, the first, second (median), third, and fourth (maximum) quartiles are shown as
colored segments on a "bar" where the color changes correspond to the exact numerical value of
3-12

-------
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 5-4. 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 30, and are grouped by pollutant within each state section. This
allows for both a "site vs. program" comparison as well as an inter-site comparison for sites
within a given state.
3.4.3.2 Site Trends Analysis
Table 2-1 presents current monitoring sites that have participated in the NMP in previous
years. A site-specific trends analysis was conducted for sites with at least 5 consecutive years of
method-specific data analyzed under the NMP. The trends analysis was conducted for each of
the site-specific pollutants of interest identified via the risk-based screening process. Forty-one
of the 66 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 30).
The trends figures and analyses are presented as 1-year statistical metrics. The following
criteria were used to calculate valid statistical metrics:
•	Analysis must have been performed under the NMP by the contract laboratory.
•	There must be a minimum of at least 5 years of consecutive data.
Five individual statistical metrics were calculated for this analysis and are presented as
box and whisker plots, an example of which can be seen in Figure 6-16. 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, 50th
percentile (or median), and 95th percentile (as shown by the y-values corresponding with the
bottom of the box, the blue line, and top of the box, respectively); and the average (or mean)
concentration (as denoted by the orange diamond). Each of the five metrics represents all
3-13

-------
measurements collected during that 1-year period. For each 1-year period, there must be a
minimum of 85 percent completeness, which corresponds to roughly 51 valid samples or
approximately 10 months of sampling (for a site sampling on a l-in-6 day sampling schedule) for
an average to be presented. For cases where sampling began mid-year, a minimum of 6 months
of sampling is required. In these cases, the 1-year average is not provided but the concentration
range and quartiles are still presented.
Data used in this analysis were downloaded from EPA's AQS database (EPA, 2014b),
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.
3.4.3.3 Cancer Risk and Noncancer Hazard Approximations
Risk was further examined by calculating cancer risk and noncancer hazard
approximations for each of the site-specific pollutants of interest. The cancer risk approximations
presented in this report estimate the cancer risk due to exposure to a given pollutant at the annual
average concentration over a 70-year period (not the risk resulting from exposure over the time
period covered in this report). A cancer risk approximation less than 1 in-a-million is considered
negligible; a cancer risk greater than 1 in-a-million but less than 100 in-a-million is generally
considered acceptable; and a cancer risk greater than 100 in-a-million is considered significant
(EPA, 2009c). The noncancer hazard approximation is presented as the Noncancer Hazard
Quotient (HQ), which is a unitless value. According to EPA, "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 lb).
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, 2014e). Cancer URE and noncancer RfC toxicity factors
can be applied to the annual averages to approximate risk based on ambient monitoring data.
3-14

-------
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 30).
To further this analysis, pollution roses were created for each of the site-specific
pollutants of interest that have cancer risk approximations greater than 75 in-a-million and/or a
noncancer hazard approximation greater than 1.0, where applicable. This analysis is performed
to help identify the geographical area where the emissions sources of these pollutants may have
originated. A pollution rose is a plot of the ambient concentration versus the wind speed and
direction; high concentrations may be shown in relation to the direction of potential emissions
sources.
There are, however, limitations to this analysis. NWS wind data are hourly observations
while concentrations from this report are 24-hour measurements. Thus, the wind data must be
averaged for comparison to the concentrations data. Wind speed and direction can fluctuate
throughout a given day or change dramatically if a frontal system moves through. Thus, the
average calculated wind data may not be completely representative of a given day. This can be
investigated more thoroughly if the need arises.
3.4.3.4 Risk-Based Emissions Assessment
A pollutant emitted in high quantities does not necessarily present a higher risk to human
health than a pollutant emitted in very low quantities. The more toxic the pollutant, the more risk
associated with its emissions in ambient air. The development of various health-based toxicity
factors, as discussed in previous sections, has allowed analysts to apply weight to the emissions
of pollutants based on toxicity rather than mass emissions. This approach considers both a
pollutant's toxicity potential and the quantity emitted.
This assessment compares county-level emissions to toxicity-weighted emissions based
on the EPA-approved approach described below (EPA, 2007). The 10 pollutants with the highest
total mass emissions and the 10 pollutants with the highest associated toxicity-weighted
3-15

-------
emissions for pollutants with cancer and noncancer toxicity factors are presented in each state
section. While the absolute magnitude of the pollutant-specific toxicity-weighted emissions is
not meaningful, the relative magnitude of toxicity-weighted emissions is useful in identifying the
order of potential priority for air quality managers. Higher values suggest greater priority;
however, even the highest values may not reflect potential cancer effects greater than the level of
concern (100 in-a-million) or potential noncancer effects above the level of concern
(e.g., HQ = 1.0). The pollutants exhibiting the 10 highest annual average-based risk
approximations for cancer and noncancer effects are also presented in each state section. The
results of this data analysis may help state, local, and tribal agencies better understand which
pollutants emitted, from a toxicity basis, are of the greatest concern and whether or not these
pollutants are already being monitoring or perhaps should be monitored in the future.
The toxicity-weighted emissions approach consists of the following steps:
1.	Obtain HAP emissions data for all anthropogenic sectors (nonpoint, point, onroad,
and nonroad) from the NEI. For point sources, sum the process-level emissions to the
county-level. Biogenic emissions are not included in this analysis.
2.	Apply the mass extraction speciation profiles to extract metal and cyanide mass.
3.	Apply weight to the emissions derived from the steps above based on their toxicity.
The results of the toxicity-weighting process are unitless.
a.	To apply weight based on cancer toxicity, multiply the emissions of each
pollutant by its cancer URE.
b.	To apply weight based on noncancer toxicity, divide the emissions of each
pollutant by its noncancer RfC.
The PAHs and/or phenols measured using Method TO-13A are a sub-group of 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 lc). Thus, emissions data and toxicity-weighted
emissions for many of the PAHs are presented by POM Groups for this analysis. Table 3-3
presents the 22 PAHs measured by Method TO-13A and their associated POM Groups, if
applicable. Table 3-3 also includes the additional phenols measured at KMMS (cresols and
phenol). The POM groups are sub-grouped in Table 3-3 because toxicity research has led to the
refining of UREs for certain PAHs (EPA, 2014e). Note that naphthalene, phenol, and cresols
emissions are reported to the NEI individually; therefore, they are not included in one of the
3-16

-------
POM Groups. Also note that four pollutants analyzed by Method TO-13A and listed in Table 3-3
do not have assigned POM Groups.
Table 3-3. POM Groups for PAHs and Phenols
Pollutant
POM
Group
POM
Subgroup
Acenaphthene
Group 2
Group 2b
Acenaphthylene
Group 2
Group 2b
Anthracene
Group 2
Group 2d
Benzo(a)anthracene
Group 6
Benzo(a)pyrene
Group 5
Group5a
Benzo(b)fluoranthene
Group 6
Benzo(e)pyrene
Group 2
Group 2b
Benzo(g,hi)perylene
Group 2
Group 2b
Benzo(k)fluoranthene
Group 6
Chrysene
Group 7
Coronene
NA
Cresols*
NA
Cyclopenta | cdlpy rene
NA
Dibenz(a,h)anthracene
Group 5
Group5b
Fluoranthene
Group 2
Group 2b
Fluorene
Group 2
Group 2b
9-Fluorenone
NA
Indeno( 1,2,3 -cd)pyrene
Group 6
Naphthalene*
NA
Perylene
Group 2
Group 2b
Phenantlirene
Group 2
Group 2d
Phenol*
NA
Pyrene
Group 2
Group 2d
Retene
NA
* Emissions for these pollutants are reported to the NEI individually;
therefore, they are not included in one of the POM Groups.
NA = no POM Group assigned.
3-17

-------
4.0	Summary of the 2013 National Monitoring Programs Data
This section summarizes the results of the data analyses performed on the NMP dataset,
as described in Section 3.
4.1	Statistical Results
This section examines the following statistical parameters for the target pollutants of each
analytical method: 1) detection rates, 2) concentration ranges and data distribution, and 3) central
tendency statistics. Tables 4-1 through 4-6 present statistical summaries for the target pollutants
and Sections 4.1.1 through 4.1.3 review the basic findings of these statistical calculations.
4.1.1 Target Pollutant Detection Rates
There is an experimentally determined MDL for every target pollutant, as described in
Section 2.2. Quantification below the MDL is possible, although the measurement's reliability is
lower. If a concentration does not exceed the MDL, it does not mean that the pollutant is not
present in the air. If the instrument does not generate a numerical concentration, the
measurement is marked as "ND," or "non-detect." As explained in Section 2.2, data analysts
should exercise caution when interpreting monitoring data with a high percentage of reported
concentrations at levels near or below the corresponding MDLs. A thorough review of the
number of measured detections, the number of non-detects, and the total number of samples is
beneficial to understanding the representativeness of the interpretations made.
Tables 4-1 through 4-6 summarize the number of times each target pollutant was detected
out of the number of valid samples collected and analyzed. Approximately 53 percent of the
reported measurements (based on the preprocessed daily measurements) were greater than the
MDLs across the program. The following list provides the percentage of measurements that were
above the MDLs for each analytical method:
•	42.2 percent for VO Cs
•	49.9 percent for SNMOCs
•	83.0 percent for carbonyl compounds
•	61.4 percent for PAHs and 54.4 percent for PAHs/Phenols
•	77.1 percent for metals
4-1

-------
• 38.7 percent for hexavalent chromium samples.
Some pollutants were detected in every sample collected while others were infrequently
detected or not detected at all. Among the carbonyl compounds, formaldehyde and acetone had
the greatest number of measured detections (1,758), based on the preprocessed daily
measurements. These pollutants were reported in every valid carbonyl compound sample
collected (1,758). Six VOCs, (benzene, toluene, chloromethane, dichlorodifluoromethane,
propylene, and trichlorotrifluoromethane) were detected in every valid VOC sample collected
(1,883). Thirteen pollutants, including acetylene, ethylene, ethane, and propylene, were detected
in every valid SNMOC sample collected (364). Naphthalene, phenanthrene, fluoranthene, and
pyrene were detected in every valid PAH or PAH/Phenol sample collected (1,340). Lead,
manganese, antimony, cadmium, and cobalt were detected in every valid metal sample collected
(1,090). Hexavalent chromium was detected in 290 samples (out of 744 valid samples).
Although NBIL and BTUT have the greatest number of measured detections (6,690 for
NBIL and 6,469 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 (61 percent and
67 percent, respectively) were not as high as other sites. Detection rates for sites that sampled
suites of pollutants that are frequently detected tended to be higher (refer to the list of method-
specific percentages of measurements above the MDL listed above). For example, metals were
rarely reported as non-detects. As a result, sites that sampled only metals (such as PAFL) would
be expected to have higher detection rates. PAFL's detection rate is 100 percent. Conversely,
VOCs had one of the lowest percentages of concentrations greater than the MDLs (42.2 percent).
A site measuring only VOCs would be expected to have lower detection rates, such as SPAZ
(48.9 percent).
4-2

-------
Table 4-1. Statistical Summaries of the VOC Concentrations



# of








#
# of
Measured


Arithmetic

First
Third
Standard

of Non-
Measured
Detections
Minimum2
Maximum
Mean
Median
Quartile
Quartile
Deviation
Pollutant
Detects1
Detections1

-------
Table 4-1. Statistical Summaries of the VOC Concentrations (Continued)
Pollutant
#
of Non-
Detects1
# of
Measured
Detections1
# of
Measured
Detections

-------
Table 4-1. Statistical Summaries of the VOC Concentrations (Continued)
Pollutant
#
of Non-
Detects1
# of
Measured
Detections1
# of
Measured
Detections

-------
Table 4-2. Statistical Summaries of the SNMOC Concentrations
Pollutant
#
of Non-
Detects1
# of
Measured
Detections1
# of
Measured
Detections

-------
Table 4-2. Statistical Summaries of the SNMOC Concentrations (Continued)
Pollutant
#
of Non-
Detects1
# of
Measured
Detections1
# of
Measured
Detections

-------
Table 4-2. Statistical Summaries of the SNMOC Concentrations (Continued)
Pollutant
#
of Non-
Detects1
# of
Measured
Detections1
# of
Measured
Detections

-------
Table 4-2. Statistical Summaries of the SNMOC Concentrations (Continued)



# of








#
# of
Measured


Arithmetic

First
Third
Standard

of Non-
Measured
Detections
Minimum2
Maximum
Mean
Median
Quartile
Quartile
Deviation
Pollutant
Detects1
Detections1

-------
Table 4-3. Statistical Summaries of the Carbonyl Compound Concentrations



# of








#
# of
Measured


Arithmetic

First
Third
Standard

of Non-
Measured
Detections
Minimum2
Maximum
Mean
Median
Quartile
Quartile
Deviation
Pollutant
Detects1
Detections1

-------
Table 4-4a. Statistical Summaries of the PAH Concentrations



# of








#
# of
Measured


Arithmetic

First
Third
Standard

of Non-
Measured
Detections
Minimum2
Maximum
Mean
Median
Quartile
Quartile
Deviation
Pollutant
Detects1
Detections1

-------
Table 4-4b. Statistical Summaries of the PAH/Phenols Concentrations
oo



# of








#
# of
Measured


Arithmetic

First
Third
Standard

of Non-
Measured
Detections
Minimum2
Maximum
Mean
Median
Quartile
Quartile
Deviation
Pollutant
Detects1
Detections1
-Cresols
0
30
0
4.13
25.8
9.88
8.22
5.70
11.5
5.90
o-Cresol
0
30
0
3.41
12.0
6.38
5.92
4.53
7.28
2.47
Dibenz(a,h)anthracene
30
0
0
Not Detected
Fluoranthene
0
30
0
0.558
12.1
5.46
4.85
3.45
7.42
3.20
Fluorene
0
30
0
1.20
23.6
10.6
9.46
6.28
13.4
6.13
Indeno(l,2,3-cd)pyrene
26
4
4
0.034
0.151
0.009
0
0
0
0.029
Naphthalene
0
30
0
22.1
281
116
107
59.8
148
69.1
Phenanthrene
0
30
0
2.71
52.8
24.5
21.9
15.4
31.3
13.6
Phenol
0
30
0
24.9
1245
250
174
92.2
249
264
Pyrene
0
30
0
0.285
4.64
1.98
1.70
1.30
2.65
1.11
1	Out of 30 valid samples.
2	Excludes zeros for non-detects.

-------
Table 4-5. Statistical Summaries of the Metals Concentrations



# of








#
# of
Measured


Arithmetic

First
Third
Standard

of Non-
Measured
Detections
Minimum2
Maximum
Mean
Median
Quartile
Quartile
Deviation
Pollutant
Detects1
Detections1

-------
Table 4-6. Statistical Summary of the Hexavalent Chromium Concentrations

#
# of
# of
Measured


Arithmetic

First
Third
Standard

of Non-
Measured
Detections
Minimum2
Maximum
Mean
Median
Quartile
Quartile
Deviation
Pollutant
Detects1
Detections1

-------
4.1.2	Concentration Range and Data Distribution
The concentrations measured during the 2013 NMP exhibit a wide range of variability.
The minimum and maximum concentration measured (excluding zeros substituted for non-
detects) for each target pollutant are presented in Tables 4-1 through 4-6 (in respective pollutant
group units). Some pollutants, such as dichloromethane, had a wide range of concentrations
measured, while other pollutants, such as dichlorotetrafluoroethane, did not, even though they
were both detected frequently. The pollutant for each method-specific pollutant group with the
largest range in measured concentrations is as follows:
•	For VOCs, dichloromethane (0.045 ppbv to 1,610 ppbv)
•	For SNMOCs, ethane (3.75 ppbC to 493 ppbC)
•	For carbonyl compounds, formaldehyde (0.010 ppbv to 17.8 ppbv)
•	For PAHs, naphthalene (1.95 ng/m3 to 748 ng/m3)
•	For PAHs/Phenols measured at KMMS, phenol (24.9 ng/m3 to 1,245 ng/m3)
•	For metals in PMio, cadmium (0.007 ng/m3 to 120 ng/m3)
•	For metals in TSP, manganese (0.835 ng/m3 to 75.58 ng/m3)
•	For hexavalent chromium, 0.0027 ng/m3 to 0.38 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, first and third quartiles, and standard deviation) for each of the
pollutants sampled during the 2013 NMP in respective pollutant group units. A multitude of
observations can be made from these tables. The pollutants with the three highest average
concentrations for each pollutant group are provided below, with respective confidence intervals
(although the 95 percent confidence intervals are not provided in the tables).
4-15

-------
The top three VOCs by average concentration, as presented in Table 4-1, are:
•	Acetonitrile (10.1 ± 2.29 ppbv)
•	Dichloromethane (2.35 ± 1.87 ppbv).
•	Acetylene (0.849 ± 0.064 ppbv)
The top three SNMOCs by average concentration, as presented in Table 4-2, are:
•	Ethane (56.2 ± 7.79 ppbC)
•	Propane (34.4 ± 4.55 ppbC)
•	w-Butane (13.7 ±1.71 ppbC).
The top three carbonyl compounds by average concentration, as presented in Table 4-3
are:
•	Formaldehyde (2.30 ± 0.089 ppbv)
•	Acetone (1.12 ± 0.041 ppbv).
•	Acetaldehyde (0.996 ± 0.033 ppbv)
The top three PAHs by average concentration, as presented in Tables 4-4a, are:
•	Naphthalene (75.3 ± 3.96 ng/m3)
•	Phenanthrene (9.86 ± 0.97 ng/m3)
•	Acenaphthene (4.49 ± 0.56 ng/m3).
The top three PAHs/Phenols by average concentration for KMMS, as presented in
Tables 4-4b, are:
•	Phenol (250 ± 98.6 ng/m3)
•	Naphthalene (116 ± 25.8 ng/m3)
•	Phenanthrene (24.5 ± 5.07 ng/m3).
4-16

-------
The top three metals by average concentration for both PMio and TSP fractions, as
presented in Table 4-5, are;
•	Manganese (PMio = 7.87 ± 0.65 ng/m3, TSP = 17.2 ± 1.56 ng/m3)
•	Lead (PMio = 3.57 ± 0.30 ng/m3, TSP = 3.48 ± 0.26 ng/m3)
•	Total chromium (PMio = 2.05 ± 0.16 ng/m3, TSP = 2.11 ± 0.11 ng/m3).
The average concentration of hexavalent chromium, as presented in Table 4-6, is
0.014 ± 0.002 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.
The results in Table 4-7 are listed in descending order by number of screens failed.
Table 4-7 shows that benzene failed the greatest number of screens (2,121), although carbon
tetrachloride, formaldehyde, acetaldehyde, 1,2-dichloroethane, and 1,3-butadiene each failed
greater than 1,500 screens. These pollutants were also among those with the greatest number of
measured detections among pollutants shown in Table 4-7. Conversely, four pollutants listed in
Table 4-7 failed only one screen each (bromoform, cz's-1,3-dichloropropene, styrene, and trans-
1,3-dichloropropene). The number of measured detections for these four pollutants varied
significantly. Styrene was detected in 1,555 samples (out of 2,247 samples) while bromoform
was detected less frequently (179 out of 1,883 valid samples) and cz's-l,3-dichloropropene and
trans-1,3-dic.hloropropene were rarely detected. Three pollutants exhibited a failure rate of
100 percent (1,2-dichloroethane, 1,2-dibromoethane, and chloroprene); however, chloroprene
and 1,2-dibromoethane were detected in less than 1 percent of samples collected. Thus, the
4-17

-------
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.
Table 4-7. Results of the Program-Level Preliminary Risk-Based Screening Process

Screening
# of
# of
%of
%of
Cumulative

Value
Failed
Measured
Failed
Total
%
Pollutant
(Ug/m3)
Screens
Detections
Screens
Failures
Contribution
Benzene
0.13
2,121
2,123
99.91
14.71
14.71
Carbon Tetrachloride
0.17
1,877
1,882
99.73
13.01
27.72
Formaldehyde
0.077
1,752
1,758
99.66
12.15
39.87
Acetaldehyde
0.45
1,678
1,756
95.56
11.63
51.50
1,2-Dichloroethane
0.038
1,605
1,605
100.00
11.13
62.63
1,3-Butadiene
0.03
1,574
1,651
95.34
10.91
73.54
Naphthalene
0.029
998
1,340
74.48
6.92
80.46
Arsenic
0.00023
925
1,077
85.89
6.41
86.88
Ethylbenzene
0.4
404
2,110
19.15
2.80
89.68
Hexachloro-1,3-butadiene
0.045
301
330
91.21
2.09
91.76
p-Dichlorobenzene
0.091
264
905
29.17
1.83
93.59
Acenaphthene
0.011
136
1,321
10.30
0.94
94.54
Nickel
0.0021
129
1,089
11.85
0.89
95.43
Fluorene
0.011
114
1,168
9.76
0.79
96.22
Propionaldehyde
0.8
91
1,744
5.22
0.63
96.85
Vinyl chloride
0.11
85
243
34.98
0.59
97.44
Manganese
0.03
61
1,090
5.60
0.42
97.86
Cadmium
0.00056
49
1,090
4.50
0.34
98.20
Fluoranthene
0.011
36
1,340
2.69
0.25
98.45
Trichloroethylene
0.2
34
308
11.04
0.24
98.69
Lead
0.015
27
1,090
2.48
0.19
98.88
Benzo(a)pyrene
0.00057
24
873
2.75
0.17
99.04
1,1,2-Trichloroethane
0.0625
22
29
75.86
0.15
99.20
Xylenes
10
22
2,132
1.03
0.15
99.35
Dichloromethane
60
17
1,791
0.95
0.12
99.47
Methyl terf-Butyl Ether
3.8
16
410
3.90
0.11
99.58
Hexavalent Chromium
0.000083
15
290
5.17
0.10
99.68
1,2-Dibromoethane
0.0017
14
14
100.00
0.10
99.78
Acenaphthylene
0.011
9
675
1.33
0.06
99.84
Bromomethane
0.5
8
1,404
0.57
0.06
99.90
Chloroform
9.8
4
1,478
0.27
0.03
99.92
Chloroprene
0.0021
3
3
100.00
0.02
99.94
1,1-Dichloroethane
0.625
2
30
6.67
0.01
99.96
T etrachloroethylene
3.8
2
1,453
0.14
0.01
99.97
Bromoform
0.91
1
179
0.56
0.01
99.98
cis-1,3-Dichloropropene
0.25
1
6
16.67
0.01
99.99
Styrene
100
1
1,555
0.06
0.01
99.99
trans-1, 3-Dichloropropene
0.25
1
3
33.33
0.01
100.00
Total
14,423
39,345
36.66

4-18

-------
The program-level pollutants of interest, as indicated by the shading in Table 4-7, are
identified as follows:
Acenaphthene
Acetaldehyde
Arsenic
Ethylbenzene
Formaldehyde
Hexachloro-1,3-butadiene
1,2-Dichloroethane
Benzene
1,3-Butadiene
Naphthalene
Carbon Tetrachloride
Nickel.
• /»-Dichlorobenzene
The pollutants of interest identified via the preliminary risk-based screening approach for
2013 are similar to the pollutants identified in previous years. Manganese and fluorene are the
only pollutants that were program-wide pollutants of interest for 2012 but are not on the list for
2013. The risk screening value for manganese was updated resulting in a significant decrease in
the number of failed screens for 2013. Fluorene is just outside the 95 percent criteria, as shown
in Table 4-7, and therefore is not a pollutant of interest for 2013.
Of the 71 pollutants sampled for under the NMP that have corresponding screening
values, concentrations of 38 pollutants failed at least one screen. Of these, a total of 14,423 out
of 39,345 concentrations (or nearly 37 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 (14,423 of 61,337, or nearly 24 percent). Note that these
percentages exclude acrolein, acetonitrile, acrylonitrile, and carbon disulfide measurements per
the explanations provided in Section 3.2; these pollutants are excluded from all risk-related
analyses contained in the report from this point forward.
Table 4-8 presents the total number of failed screens per site, in descending order, as a
means of comparing the results of the preliminary risk-based screening process across the sites.
In addition to the number of failed screens, Table 4-8 also provides the total number of screens
conducted (one screen per valid preprocessed daily measurement for each site for all pollutants
4-19

-------
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.
As shown, S4MO has the largest number of failed screens (574), followed by PXSS (554)
and NBIL (510); conversely, SDGA, CHSC, and CAMS 35 failed only one or two screens each.
Four additional sites did not fail any screens (STMN, HOWI, MIWI, and CAMS 85). These sites
sampled only hexavalent chromium and did not sample for the entire year. 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 due to the large number of pollutants
sampled. For sites sampling only one or two pollutant groups, it depends on the pollutant group
sampled as the number of compounds analyzed varies from one (hexavalent chromium) to 80
(SNMOCs). Sites sampling only hexavalent chromium, which was detected in less than
40 percent of the valid samples collected and has a failure rate of 5 percent across the program,
appear near the bottom of Table 4-8.
Table 4-8. Site-Specific Risk-Based Screening Comparison
Site
# of
Failed
Screens
Total # of
Measured
Detections1
%of
Failed
Screens
# of
Pollutant
Groups
Analyzed
S4MO
574
2,687
21.36
5
PXSS
554
2,411
22.98
5
NBIL
510
2,501
20.39
6
TOOK
494
1,644
30.05
3
DEMI
493
1,978
24.92
4
BTUT
487
2,188
22.26
6
TMOK
480
1,622
29.59
3
ASKY
477
1,049
45.47
2
GPCO
476
1,867
25.50
4
TROK
455
1,602
28.40
3
SEWA
446
2,170
20.55
5
OCOK
420
1,648
25.49
3
ELNJ
404
1,187
34.04
2
GLKY
403
2,128
18.94
5
CSNJ
402
1,163
34.57
2
!Total number of measured detections for all pollutants with
screening values, not just those failing screens. Also excludes
acrolein, acetonitrile, acrylonitrile, and carbon disulfide results.
BOLD ITALICS = EPA-designated NATTS Site
4-20

-------
Table 4-8. Site-Specific Risk-Based Screening Comparison (Continued)




# of

# of
Total # of
%of
Pollutant

Failed
Measured
Failed
Groups
Site
Screens
Detections1
Screens
Analyzed
SPIL
398
1,137
35.00
2
NBNJ
382
1,188
32.15
2
ROIL
372
1,090
34.13
2
CHNJ
363
1,107
32.79
2
LEKY
361
1,384
26.08
3
KMMS
334
1,175
28.43
2
ANAK
331
1,583
20.91
2
SSMS
287
959
29.93
1
TVKY
282
991
28.46
1
CCKY
281
1,464
19.19
2
ATKY
259
937
27.64
1
LAKY
257
911
28.21
1
BLKY
245
881
27.81
1
UNVT
226
1,733
13.04
4
ADOK
222
802
27.68
3
YUOK
203
810
25.06
3
SKFL
171
893
19.15
3
RICO
170
447
38.03
2
SPAZ
159
451
35.25
1
BURVT
134
483
27.74
1
SYFL
134
453
29.58
3
RUVT
126
441
28.57
1
ORFL
122
183
66.67
1
INDEM
121
183
66.12
1
AZFL
117
177
66.10
1
WPIN
116
174
66.67
1
BOMA
115
1,462
7.87
3
PACO
108
375
28.80
2
ASKY-M
104
595
17.48
1
SJJCA
103
1,198
8.60
2
BRCO
102
377
27.06
2
ROCH
102
691
14.76
2
BMCO
97
357
27.17
2
BXNY
87
861
10.10
2
RFCO
69
239
28.87
2
WADC
61
694
8.79
2
CELA
59
679
8.69
1
PRRI
56
839
6.67
2
RIVA
56
633
8.85
2
BAKY
54
592
9.12
1
RUCA
54
636
8.49
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
4-21

-------
Table 4-8. Site-Specific Risk-Based Screening Comparison (Continued)




# of

# of
Total # of
%of
Pollutant

Failed
Measured
Failed
Groups
Site
Screens
Detections1
Screens
Analyzed
PAFL
30
300
10.00
1
LBHCA
19
315
6.03
1
WPFL
11
238
4.62
1
CAMS 35
2
25
8.00
1
CHSC
2
395
0.51

SDGA
1
8
12.50
1
STMN
0
8
0
1
CAMS 85
0
7
0
1
HOWI
0
4
0
1
MIWI
0
8
0
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
Although 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 23 percent.
The following sections from this point forward focus primarily on those pollutants
designated as program-level pollutants of interest.
4-22

-------
4.2.1 Concentrations of the Pollutants of Interest
Concentrations of the program-level pollutants of interest vary significantly, among the
pollutants and among the sites. Tables 4-9 through 4-12 present the top 10 annual average
concentrations and 95 percent confidence intervals by site for each of the program-level
pollutants of interest (for VOC/SNMOCs, carbonyl compounds, PAHs, and metals,
respectively). As described in Section 3.1, an annual average is the average concentration of all
measured detections and zeros substituted for non-detects for a given year. Further, an annual
average is only calculated where at least three quarterly averages could be calculated and where
the site-specific method completeness is at least 85 percent. The annual average concentrations
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 VOC/SNMOCs and carbonyl
compounds, respectively, are reported in ^g/m3. Note that not all sites sampled each pollutant
group; thus, the list of possible sites presented in Tables 4-9 through 4-12 is limited to those sites
sampling each pollutant. For instance, only six sites sampled TSP metals; thus, these would be
the only sites to appear in Table 4-12 for each metal (TSP) pollutant of interest shown. However,
two of the sites only sampled for half of the year and as a result, only four sites are listed under
the TSP metals in Table 4-12.
4-23

-------
Table 4-9. Annual Average Concentration Comparison of the VOC/SNMOC Pollutants of Interest



Carbon
P-
1,2-

Hexachloro-1,3-

Benzene
1,3-Butadiene
Tetrachloride
Dichlorobenzene
Dichloro ethane
Ethylbenzene
Butadiene
Rank
fjiWm 3)
frig/m3)
Oig/m3)
frig/m3)
frig/m3)
frig/m3)
frig/m3)

PACO
TVKY
BLKY
SPAZ
TVKY
KMMS
SSMS
1
1.96 + 0.31
1.03 + 0.97
1.11 + 0.77
0.22 + 0.04
3.75 + 3.68
1.95+ 1.08
0.03 + 0.01

ANAK
LAKY
TVKY
PXSS
BLKY
ANAK
S4MO
2
1.56 + 0.34
0.66 + 0.80
0.80 + 0.08
0.20 + 0.03
1.28 + 0.75
0.89 + 0.22
0.02 + 0.01

RICO
BLKY
SEWA
TMOK
LAKY
SPAZ
NBNJ
3
1.52 + 0.26
0.63 + 0.49
0.69 + 0.03
0.10 + 0.02
0.70 + 0.45
0.68 + 0.15
0.02 + 0.01

ASKY
CCKY
LAKY
S4MO
ATKY
PXSS
OCOK
4
1.52+ 1.39
0.26 + 0.40
0.68 + 0.03
0.09 + 0.02
0.30 + 0.10
0.67 + 0.11
0.02 + 0.01

BMCO
SPAZ
CCKY
SSMS
CCKY
BTUT
CSNJ
5
1.26 + 0.19
0.22 + 0.07
0.67 + 0.02
0.08 + 0.02
0.24 + 0.08
0.49 + 0.24
0.02 + 0.01

TOOK
PXSS
DEMI
BURVT
BTUT
GPCO
CHNJ
6
1.21 + 0.17
0.21 + 0.05
0.67 + 0.02
0.06 + 0.01
0.11 + 0.03
0.49 + 0.07
0.02 + 0.01

BRCO
GPCO
ATKY
KMMS
CSNJ
TOOK
ROIL
7
1.14 + 0.20
0.15 + 0.03
0.67 + 0.02
0.05 + 0.01
0.09 + 0.01
0.45 + 0.06
0.02 + 0.01

SPAZ
ANAK
GLKY
ANAK
S4MO
TMOK
LAKY
8
1.07 + 0.21
0.15 + 0.04
0.67 + 0.03
0.05 + 0.01
0.09 + 0.01
0.43 + 0.07
0.02 + 0.01

TVKY
SPIL
KMMS
OCOK
NBNJ
ELNJ
BLKY
9
1.06 + 0.32
0.13 + 0.02
0.66 + 0.02
0.05 + 0.01
0.09 + 0.01
0.43 + 0.06
0.02 + 0.01

PXSS
ELNJ
SSMS
BTUT
BURVT
TROK
TMOK
10
1.06 + 0.18
0.11 + 0.01
0.66 + 0.02
0.05 + 0.03
0.08 + 0.01
0.39 + 0.06
0.02 + 0.01
BOLD ITALICS = EPA-designated NATTS Site

-------
Table 4-10. Annual Average Concentration Comparison of the
Carbonyl Compound Pollutants of Interest

Acetaldehyde
Formaldehyde
Rank
(ug/m3)
(ug/m3)

BTUT
BTUT
1
4.18 + 0.36
8.05 + 0.87

GPCO
GPCO
2
3.79 + 0.57
6.44+1.22

CSNJ
CSNJ
3
2.78 + 0.33
4.96 + 0.59

PXSS
ELNJ
4
2.78 + 0.29
4.90 + 0.67

ELNJ
PXSS
5
2.60 + 0.26
3.89 + 0.22

SPIL
WPIN
6
2.37 + 0.55
3.41 + 0.37

NBIL
SPIL
7
2.37 + 0.31
3.31 + 0.49

TOOK
S4MO
8
2.02 + 0.25
3.23 + 0.55

S4MO
TMOK
9
1.98 + 0.22
3.19 + 0.45

TMOK
ROIL
10
1.94 + 0.25
3.19 + 0.57
BOLD ITALICS = EPA-designated NATTS Site
Table 4-11. Annual Average Concentration Comparison of the PAH Pollutants of Interest

Acenaphthene
Naphthalene
Rank
(ng/m3)
(ng/m3)

NBIL
NBIL
1
25.12 + 8.19
155.94 + 44.27

ROCH
GPCO
2
19.37+5.35
136.93 + 23.05

DEMI
BXNY
3
9.62 + 2.72
126.77 + 12.63

GPCO
CELA
4
8.05+ 1.77
111.44+ 15.95

BXNY
DEMI
5
6.46+ 1.15
104.57 + 14.63

S4MO
SJJCA
6
5.02+ 1.16
93.97 + 22.27

WPFL
PXSS
7
4.62 + 2.72
93.36 + 18.63

CELA
RIVA
8
4.26 + 0.54
86.87+ 13.95

SEWA
WADC
9
3.68+ 1.10
83.14+ 13.50

RIVA
RUCA
10
3.45 + 0.66
81.40+ 16.05
BOLD ITALICS = EPA-designated NATTS Site
4-25

-------
Table 4-12. Annual Average Concentration Comparison of the Metals Pollutants of Interest

Arsenic
Arsenic
Nickel
Nickel

(PMio)
(TSP)
(PMio)
(TSP)
Rank
(ng/m3)
(ng/m3)
(ng/m3)
(ng/m3)

ASKY-M
TROK
ASKY-M
TOOK
1
1.24 + 0.29
0.80 + 0.11
2.40 + 0.89
2.09 + 0.42

BTUT
TOOK
SEWA
TROK
2
0.99 + 0.40
0.78 + 0.07
1.78 + 0.44
1.43 + 0.24

BAKY
TMOK
PXSS
TMOK
3
0.82 + 0.22
0.65 + 0.07
1.49 + 0.21
1.30 + 0.20

SEWA
OCOK
BTUT
OCOK
4
0.79 + 0.13
0.46 + 0.06
1.44 + 0.24
0.84 + 0.12

S4MO

BOMA

5
0.73 + 0.08

1.42 + 0.23


PAFL

SJJCA

6
0.72 + 0.22

1.40 + 0.21


LEKY

S4MO

7
0.68 + 0.12

1.06 + 0.26


NBIL

PAFL

8
0.62 + 0.11

0.76 + 0.10


CCKY

NBIL

9
0.61 + 0.15

0.75 + 0.07


SJJCA

BAKY

10
0.52 + 0.13

0.61 + 0.15

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 (8.05 ± 0.87 pg/m3). This was
also true for BTUT in 2012, although the concentration for 2013 is twice as high as it
was for 2012. Formaldehyde and acetaldehyde together account for 18 of the 19
annual average concentrations greater than 2.0 pg/m3 shown in Tables 4-9 through
4-12 (the one exception being for TVKY's annual average concentration of
1,2-dichloroethane).
•	All 10 annual average concentrations of benzene shown in Table 4-9 are greater than
1 pg/m3, the only pollutant for which this is true. PACO has the highest annual
average benzene concentration (1.96 ± 0.31 pg/m3) among sites sampling benzene,
with four of the five Garfield County, Colorado sites ranking among the 10 highest.
Only RFCO does not appear in Table 4-9, with this site's annual average benzene
concentration ranking among the lowest (0.57 ±0.12 pg/m3). Other sites ranking
among the highest benzene concentrations include ANAK, ASKY, TOOK, TVKY,
and the two Phoenix sites (SPAZ and PXSS). Note that the confidence intervals for
these sites span a relatively small range, with one exception. The annual average
concentration for ASKY is 1.52 ± 1.39 pg/m3. The large confidence interval for this
site indicates that this annual average is likely influenced by outlier (s) as opposed to
running on the higher side on a regular basis. The highest benzene concentration
measured across the program was measured at ASKY on November 6, 2013
4-26

-------
(43.5 pg/m3). The next highest benzene concentration measured across the program
was considerably less (9.38 pg/m3), which was measured at OCOK on the same date.
The four highest annual average concentrations of 1,3-butadiene were calculated for
Calvert City, Kentucky sites. The annual averages vary significantly by site, ranging
from 1.03 ± 0.97 pg/m3 for TVKY to 0.26 ± 0.40 pg/m3 for CCKY. The annual
1,3-butadiene average for the fifth Calvert City site (ATKY) is considerably less and
does not appear in Table 4-9. Note, however, the large confidence intervals associated
with each of the annual average concentrations for the Calvert City sites, indicating a
considerable amount of variability in the measurements. Concentrations of
1,3-butadiene measured at these four sites account for all 19 1,3-butadiene
concentrations greater than 1 pg/m3 measured across the program.
Calvert City sites also account for five of the 10 highest annual average
concentrations of carbon tetrachloride. Most of the annual average concentrations of
carbon tetrachloride do not vary significantly across the sites; less than 0.15 pg/m3
separates most of the annual average carbon tetrachloride concentrations. However,
this is not true for BLKY or TVKY. BLKY has the highest annual average
concentration of carbon tetrachloride across the sites by a large margin
(1.12 ± 0.77 pg/m3). The highest concentration of carbon tetrachloride across the
program was measured at BLKY (23.7 pg/m3) and is an order of magnitude higher
than the next highest measurement, which was measured at TVKY (2.33 pg/m3).
These two sites account for 11 of the 14 carbon tetrachloride concentrations greater
than 1 pg/m3 measured across the program (the other three were measured at LAKY,
SEWA, and SSMS).
The five Calvert City sites also account for the five highest annual average
concentrations of 1,2-dichloroethane, although the concentrations vary significantly.
Note the large confidence interval for the annual average for TVKY
(3.75 ± 3.68 pg/m3); the highest 1,2-dichloroethane concentration across the program
was measured at this site (111 pg/m3). Although the second highest
1,2-dichloroethane concentration was also measured at TVKY, it was significantly
less (19.3 pg/m3). These five sites account for all but one of the 77 measurements of
1,2-dichloroethane greater than 0.5 pg/m3 (the one exception was measured at BTUT,
which ranks sixth in Table 4-9 for 1,2-dichloroethane).
The highest annual average concentration of ethylbenzene (1.95 ± 1.08 pg/m3,
calculated for KMMS) is more than twice the next highest annual average
concentration of this pollutant (0.89 ± 0.22 pg/m3, calculated for ANAK). Nine of the
10 ethylbenzene concentrations greater than 5 pg/m3 across the program were
measured at KMMS (with the one additional concentration measured at BTUT).
Hexachloro-1,3-butadiene and/?-dichlorobenzene are the only two VOCs in Table 4-9
that do not have at least one annual average concentration greater than 1 pg/m3. The
annual average concentrations of /;-dic.hlorobenzene calculated for the two Phoenix
sites are significantly higher than the remaining annual averages shown for this
pollutant, although the range of annual average concentrations shown is less than
4-27

-------
0.18 ]ig/m3. The range of annual average concentrations for hexachloro-l,3-butadiene
is even less, varying by less than 0.01 pg/m3 across the sites shown.
ANAK, PXSS, SPAZ, LAKY, BLKY, and TVKY each appear in Table 4-9 a total of
four times.
The sites with the three highest annual average concentrations of acetaldehyde shown
in Table 4-10 are also the same three sites with the highest annual average
concentrations of formaldehyde (BTUT, GPCO, and CSNJ). Although their order
varies somewhat, most of the sites that appear for acetaldehyde also appear for
formaldehyde.
The maximum acetaldehyde concentration was measured at SPIL (14.2 pg/m3),
which ranks sixth for its annual average concentration. The next five highest
acetaldehyde concentrations (ranging from 7.53 pg/m3 to 10.7 pg/m3) were measured
at GPCO on consecutive sample days between June 9, 2013 and July 3, 2013.
Acetaldehyde concentrations measured at SPIL, GPCO, and BTUT account for all
11 acetaldehyde measurements greater than 7 pg/m3 measured across the program.
As shown in Table 4-10, four sites have annual average formaldehyde concentrations
greater than 4 pg/m3 (BTUT, GPCO, CSNJ, and ELNJ) and all 10 sites shown in
Table 4-10 have annual average concentrations of formaldehyde greater than 3 pg/m3.
Although BTUT has the highest annual average concentration of formaldehyde
(8.05 ± 0.87 pg/m3), the five highest concentrations measured across the program
were measured at GPCO. Of the eight formaldehyde concentrations greater than
15 pg/m3 measured across the program, all but one was measured at GPCO (with the
other one measured at ELNJ). The variability in GPCO's measurements of
formaldehyde is indicated by the confidence interval shown in Table 4-10.
Formaldehyde concentrations measured at GPCO range from 1.97 pg/m3 to
21.9 pg/m3, with a median concentration of 4.80 pg/m3.
Table 4-11 shows that NBIL has the highest annual average concentration for each of
the program-wide PAH pollutants of interest (acenaphthene and naphthalene). For
acenaphthene, the annual average concentrations for NBIL and ROCH are
considerably higher than the next highest annual averages and have relatively large
confidence intervals associated with them. Together, these two sites account for the
26 highest acenaphthene concentrations measured across the program. DEMI and
KMMS are the only other sites for which acenaphthene concentrations greater than
40 ng/m3were measured across the program. DEMI ranked third for its annual
average concentration of acenaphthene but KMMS did not sample long enough for
annual averages to be calculated. The confidence interval calculated for WPFL is
relatively large compared to its annual average concentration (4.62 ± 2.72 ng/m3);
concentrations measured at this site range from 0.665 ng/m3 to 39.9 ng/m3.
Naphthalene concentrations measured at NBIL account for three of the four
measurements greater than 500 ng/m3 measured across the program (with the other
measured at WPFL, whose annual average ranks 19th and therefore does not appear
4-28

-------
in Table 4-11). Another eight concentrations measured at NBIL were greater than
300 ng/m3. Other sites that measured naphthalene concentrations greater than
300 ng/m3 include GPCO (4), SJJCA (3) , SKFL (1), DEMI (1), CELA (1), RIVA (1),
RUCA (1). All of these sites, with the exception of SKFL, appear in Table 4-11.
ASKY-M has the highest annual average concentration for both of the program-wide
PMio metals pollutants of interest. Four of the five Kentucky sites sampling PMio
metals appear in Table 4-12 for arsenic while only two appear in Table 4-12 for
nickel. BTUT, SEWA, and S4MO round out the top five for arsenic. Annual averages
of arsenic for S4MO consistently rank among the highest in past annual reports.
Aside from ASKY-M, NATTS sites have the highest ranking annual averages for
nickel. For the last several years, the annual average nickel concentration for SEWA
has been at or near the top.
Although ASKY-M's annual arsenic concentration is the highest among NMP sites
sampling PMio metals (1.24 ± 0.29 ng/m3), the maximum arsenic concentration was
measured at BTUT (9.18 ng/m3). Arsenic concentrations greater than 3 ng/m3 were
measured at both BTUT and ASKY-M (four each). Compared to other sites,
ASKY-M has the greatest number of arsenic concentrations greater than 1 ng/m3 (26),
followed by BAKY (18), SEWA (14), and S4MO (13), with BTUT, NBIL, TROK,
and LEKY each measuring 10.
Among the Oklahoma sites sampling TSP metals, TROK has the highest annual
average concentration of arsenic (0.80 ±0.11 ng/m3), although the annual average
concentration for TOOK is similar (0.78 ± 0.07 ng/m3). The other Tulsa site, TMOK,
ranks third while the OCOK site has a significantly lower annual average
concentration of arsenic (0.46 ± 0.06 ng/m3). ADOK and YUOK are not shown in
Table 4-12 because these sites did not sample long enough for annual averages to be
calculated.
The two highest nickel concentrations program-wide were measured at ASKY-M
(21.2 ng/m3 and 17.1 ng/m3). The third highest concentration measured at this site is
considerably less (5.49 ng/m3) and the median nickel concentration for this site is
1.46 ng/m3, nearly 1 ng/m3 less than the annual average. This site has the largest
confidence interval associated with its annual average, although the confidence
interval for SEWA is also higher than the confidence intervals calculated for other
sites. Nickel concentrations measured at SEWA range from 0.17 ng/m3 to 9.75 ng/m3,
with a median concentration of 1.25 ng/m3.
Among the Oklahoma sites sampling TSP metals, the Tulsa sites ranked highest for
nickel while the Oklahoma City site has a significantly lower annual average
concentration of nickel. Nickel concentrations measured at TOOK range from
0.687 ng/m3 to 10.98 ng/m3, with a median concentration of 1.58 ng/m3. Nickel
concentrations measured at the other Oklahoma sites were less variable.
S4MO and PXSS appear on the top 10 list for eight of the 13 program-level pollutants
of interest shown in Tables 4-9 through 4-12; BTUT and TMOK appear in these
tables for seven of the 13 program-level pollutants of interest; and GPCO appears in
4-29

-------
the tables for six of the 13 program-level pollutants of interest. TOOK and NBIL each
appear in Tables 4-9 through 4-12 a total of five times.
4.3 The Contribution from Mobile Sources
Ambient air is significantly affected by mobile sources, as discussed in Section 3.3.1.
Table 4-13 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: onroad and nonroad. Onroad
emissions come from mobile sources such as automobiles, motorcycles, buses, and trucks that use
roadways; nonroad emissions come from the remaining mobile sources such as locomotives, lawn
mowers, airplanes, and boats (EPA, 2011b). Table 4-13 contains county-level onroad and nonroad
HAP emissions from the 2011 NEI, version 2. Total mobile source emissions for each county are
presented in Table 2-2.
Mobile source emissions tend to be highest in large urban areas and lowest in rural areas.
Estimated onroad county emissions were highest in Los Angeles County, California (where CELA
and LBHCA are located), followed by Harris County, Texas (where CAMS 35 is located),
Maricopa County, Arizona (where PXSS and SPAZ are located), and Cook County, Illinois (where
NBIL and SPIL are located). Estimated onroad emissions were lowest in five of the six counties in
Kentucky (the exception being Fayette County, where LEKY is located), Rutland County,
Vermont (RUVT), and Chesterfield County, South Carolina (CHSC). Estimated nonroad county
emissions were also highest in Los Angeles County, California; Cook County, Illinois; and
Maricopa County, Arizona. Estimated nonroad county emissions were lowest in Carter County,
Kentucky (GLKY); Boyd County, Kentucky (where ASKY and ASKY-M are located); Canadian
County, Oklahoma (where YUOK is located), and Chesterfield County, South Carolina.
4-30

-------
Table 4-13. Summary of Mobile Source Information by Monitoring Site
Site
County-level
Motor Vehicle
Registration1
(# of Vehicles)
Annual
Average Daily
Traffic1
(# of Vehicles)
County-level
Daily
VMT1
County-Level
Onroad HAP
Emissions2
(tpy)
County-Level
Nonroad HAP
Emissions2,3
(tpy)
Hydrocarbon
Average4
(ppbv)
ADOK
835,642
34,700
27,469,678
2,722.16
703.01
1.33
ANAK
358,999
20,193
5,301,813
895.95
1,853.36
5.92
ASKY
39,196
7,230
1,256,000
147.73
24.80
2.30
ASKY-M
39,196
12,842
1,256,000
147.73
24.80
NA
ATKY
30,254
3,262
1,241,000
124.82
351.54
3.98
AZFL
879,683
42,500
21,460,593
2,324.65
892.83
NA
BAKY
38,811
922
1,366,000
162.96
105.44
NA
BLKY
8,338
2,510
391,000
52.95
83.12
3.09
BMCO
74,036
1,880
2,171,019
249.78
77.83
NA
BOMA
393,252
27,654
10,963,634
594.94
420.78
NA
BRCO
74,036
1,182
2,171,019
249.78
77.83
NA
BTUT
274,716
130,950
6,950,795
669.34
261.40
4.86
BURVT
172,203
14,200
4,051,781
281.01
196.55
1.84
BXNY
254,752
98,899
8,170,256
596.68
243.71
NA
CAMS 35
3,401,957
31,043
56,245,209
6,834.27
1,809.32
NA
CAMS 85
72,689
1,250
2,511,619
245.30
101.13
NA
CCKY
30,254
4,050
1,241,000
124.82
351.54
2.55
CELA
7,609,517
231,000
214,482,440
10,307.83
4,465.47
NA
CHNJ
443,969
11,215
14,622,523
742.15
536.31
1.41
CHSC
41,728
700
1,265,439
149.54
57.28
NA
CSNJ
458,294
3,231
10,753,157
657.47
296.19
3.57
DEMI
1,335,516
94,600
41,554,962
3,483.29
1,080.06
3.24
ELNJ
485,427
250,000
12,081,401
678.06
339.40
3.91
GLKY
25,487
303
1,076,000
133.76
11.48
0.96
GPCO
176,969
11,000
3,355,813
511.27
153.45
3.99
HO M I
99,078
5,100
2,568,234
278.27
180.19
NA
INDEM
425,854
34,754
15,741,000
1,074.29
533.04
NA
KMMS
54,826
9,900
1,961,288
197.97
62.31
5.15
LAKY
30,254
1,189
1,241,000
124.82
351.54
3.69
LBHCA
7,609,517
285,000
214,482,440
10,307.83
4,465.47
NA
LEKY
208,983
10,083
7,490,000
773.37
342.67
1.84
MIWI
641,582
12,400
16,098,216
1,458.78
507.53
NA
NBIL
2,074,419
115,700
87,972,644
5,113.62
3,768.84
1.49
NBNJ
734,425
110,653
21,634,307
1,048.39
528.78
1.97
OCOK
835,642
41,500
27,469,678
2,722.16
703.01
1.73
ORFL
1,181,540
29,500
34,904,854
2,838.52
1,282.94
NA
PACO
74,036
15,000
2,171,019
249.78
77.83
NA
PAFL
1,181,540
49,000
34,904,854
2,838.52
1,282.94
NA
PRRI
511,015
136,800
11,670,714
1,043.74
306.55
NA
individual references provided in each state section.
Reference: 2011 NEI Version 2 (EPA, 2015a)
3Nonroad Emissions include Nonroad data as well as emissions from SCCs that were traditionally mobile categories,
such as aircraft, but have been included in Point or Nonpoint inventories in the 2011 NEI.
4This 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.
4-31

-------
Table 4-13. Summary of Mobile Source Information by Monitoring Site (Continued)
Site
County-level
Motor Vehicle
Registration1
(# of Vehicles)
Annual
Average Daily
Traffic1
(# of Vehicles)
County-level
Daily
VMT1
County-Level
Onroad HAP
Emissions2
(tpy)
County-Level
Nonroad HAP
Emissions2'3
(tpy)
Hydrocarbon
Average4
(ppbv)
PXSS
3,761,859
29,515
90,393,000
6,701.23
3,214.61
4.25
RFCO
74,036
16,000
2,171,019
249.78
77.83
NA
RICO
74,036
15,000
2,171,019
249.78
77.83
NA
RIVA
350,000
72,000
8,366,945
580.70
165.67
NA
ROCH
558,063
85,162
15,963,343
1,153.83
588.44
NA
ROIL
267,302
7,750
7,911,443
591.93
223.15
2.26
RUCA
1,788,322
150,000
55,336,730
2,271.03
973.29
NA
RUVT
79,795
10,400
1,736,164
116.35
128.97
2.40
S4MO
1,117,375
100,179
24,065,245
449.19
161.90
1.89
SDGA
479,533
138,470
20,900,748
1,539.38
275.40
NA
SEWA
1,791,383
176,000
23,266,320
4,541.86
2,348.31
1.87
SJJCA
1,575,973
115,000
41,478,310
2,984.29
650.57
NA
SKFL
879,683
47,500
21,460,593
2,324.65
892.83
NA
SPAZ
3,761,859
25,952
90,393,000
6,701.23
3,214.61
4.43
SPIL
2,074,419
186,100
87,972,644
5,113.62
3,768.84
2.42
SSMS
54,826
19,000
1,961,288
197.97
62.31
2.22
STMN
221,636
24,100
5,078,055
618.57
657.39
NA
SYFL
1,157,057
10,000
34,614,572
3,166.81
1,093.34
NA
TMOK
614,543
12,500
20,453,745
3,416.21
733.68
2.86
TOOK
614,543
64,424
20,453,745
3,416.21
733.68
3.29
TROK
614,543
56,200
20,453,745
3,416.21
733.68
3.13
TVKY
30,254
2,230
1,241,000
124.82
351.54
8.90
UNVT
172,203
1,100
4,051,781
281.01
196.55
0.70
WADC
322,350
8,700
9,786,301
580.26
249.51
NA
WPFL
1,159,114
6,600
33,617,131
3,051.13
2,146.53
NA
WPIN
830,851
143,970
31,727,000
3,351.30
691.35
NA
YUOK
106,000
45,400
4,457,374
395.05
52.52
1.24
individual references provided in each state section.
Reference: 2011 NEI Version 2 (EPA, 2015a)
3Nonroad Emissions include Nonroad data as well as emissions from SCCs that were traditionally mobile categories,
such as aircraft, but have been included in Point or Nonpoint inventories in the 2011 NEI.
4This 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.
4.3.2 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
4-32

-------
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
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 4-13 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 4-13 shows
that TVKY, ANAK, KMMS, BTUT, and SPAZ have the highest hydrocarbon averages among
the sites sampling VOCs. Interestingly, several of these sites are not the typical sites in past
reports. TVKY was among the higher sites in the 2012 report and is located in a highly
industrialized area in a moderately populated area. ANAK is a new site for 2013. ANAK is a site
with past participation in the NMP (2009) and had relatively high hydrocarbon concentrations
then, with the highest hydrocarbon average for that year. This site is located in Anchorage, the
most populous city in Alaska. KMMS, a new site in the NMP for 2013, is a source-oriented site
but is located in a moderately populated area. BTUT has higher concentrations compared to
previous years, and now ranks among the top five highest hydrocarbon averages. BTUT is
located in a suburb just north of the Salt Lake City area, less than one-half mile from 1-15. SPAZ
tends to have relatively high hydrocarbon concentrations, based on past reports, and is located in
a highly urbanized area (Phoenix), but not near a major roadway.
In past reports, TOOK and ELNJ have been among the sites with the highest hydrocarbon
averages. These sites are located in highly populated urban areas and in relatively close
proximity to heavily traveled roadways. TOOK is located near Exit 3A of 1-244 in Tulsa,
Oklahoma while ELNJ is location on Exit 13A of the New Jersey Turnpike. Both sites are also
located in close proximity to industry. Both of these sites exhibit a relatively substantial decrease
in their average hydrocarbon concentration from 2012 to 2013.
The sites with the lowest hydrocarbon averages are UNVT, GLKY, YUOK, and ADOK.
UNVT and GLKY are located in rural areas. Both YUOK and ADOK are located on the
periphery of a large urban area (Oklahoma City) and near major freeways.
4-33

-------
The average sum of hydrocarbon concentrations can be compared to other indicators of
mobile source activity to determine if correlations exist. Pearson correlation coefficients were
calculated between the average sum of hydrocarbon concentrations and the onroad (-0.01) and
nonroad (0.06) emissions. The Pearson correlation coefficients indicate virtually no correlation
between the emissions and the average hydrocarbon concentrations.
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 (FHWA, 2014). 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 4-13. As previously discussed, TVKY, ANAK, KMMS,
BTUT, and SPAZ have the highest average summed hydrocarbon concentrations, respectively,
while UNVT, GLKY, YUOK, and ADOK have the lowest. Table 4-13 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 0.03, a weak
correlation. CELA and LBHCA, which have the highest county-level vehicle ownership of all
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.
4-34

-------
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, 2013a). AADT counts obtained were based on
data from 2004 to 2013, primarily 2011 forward. The updated traffic counts are presented in
Table 4-13. The traffic data presented in Table 4-13 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 30).
For all monitoring sites (not just those sampling VOCs), the highest daily traffic volume
occurs near LBHCA, ELNJ, CELA, and SPIL. 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
SPIL is located near the Chicago-O'Hare International Airport, just west of 1-294. Of these, only
ELNJ and SPIL sampled VOCs. ELNJ has the second highest traffic volume but the ninth
4-35

-------
highest hydrocarbon average; SPIL has the fourth highest traffic volume but the 18th highest
hydrocarbon average.
Of the sites sampling VOCs, ELNJ, SPIL, SEWA, and BTUT have the highest daily
traffic volumes while GLKY, UNVT, and LAKY have the lowest, as shown in Table 4-13. A
Pearson correlation coefficient calculated between the average summed hydrocarbon calculations
and the traffic counts is -0.06, which is 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, 2013a). 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 provides this information for
state highways only. County-level VMT are presented in Table 4-13, where available. In the
absence of suitable VMT data, county-level VMT was obtained from the NEI (EPA, 2015a).
Of the sites sampling VOCs, county-level VMT was highest for PXSS and SPAZ, SPIL
and NBIL, and DEMI (Wayne County, Michigan). SPAZ and PXSS rank fifth and sixth,
respectively, for the average summed hydrocarbon concentration, SPIL and NBIL rank 18th and
29th, and DEMI ranks 13th, among the sites with the highest county-level VMT. The sites with
the lowest county-level VMT are BLKY, GLKY, and the sites in Marshall County, Kentucky.
TVKY, which is located in Marshall County, Kentucky ranks highest for its average summed
hydrocarbon concentration. A Pearson correlation coefficient calculated between the average
summed hydrocarbon concentrations and VMT is -0.04, 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).
4-36

-------
4.4 Variability Analysis
This section presents the results of the two variability analyses described in Section 3.3.2.
4.4.1 Inter-site Variability
Figures 4-1 through 4-13 are bar graphs depicting the site-specific annual averages (in
gray) overlain on the program-level averages (indicated by the solid shading), as presented in
Section 4.1. For each program-level pollutant of interest, the inter-site variability graphs allow
the reader to see how the individual site-specific annual averages feed into the program-level
averages (i.e., if a specific site(s) is driving the program average). In addition, the confidence
intervals provided on the inter-site variability graphs are an indication of the amount of
variability contained within the site-specific dataset and thus, annual averages. The published
MDL from the ERG laboratory is also plotted on the graph as an indication of the how the data
fall in relation to the MDL.
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. For the sites sampling metals, the program-level average for sites collecting PMio
samples is presented in green while the program-level average for sites collecting TSP samples is
presented in pink. For benzene, 1,3-butadiene, and ethylbenzene, the three pollutants sampled
and analyzed by two methods (VOC and SNMOC) and identified as program-level pollutants of
interest, two graphs are presented, one for each method. Note, however, that the Garfield County
sites have their canister samples analyzed using the SNMOC method only while BTUT and
NBIL have their canister samples analyzed using both methods, but only the VOCs results are
discussed throughout the remainder of this report, as described in Section 3.2.
Observations from Figures 4-1 through 4-13 include the following:
• The program-level average concentration of acenaphthene is 4.89 ng/m3, as shown in
orange in Figure 4-1. Site-specific annual average concentrations range from
0.30 ng/m3 (UNVT) to 25.12 ng/m3 (NBIL). The annual average concentrations for
ROCH and NBIL are three and four times greater than the program-level average for
acenaphthene, respectively, and have the most variability associated with the
measurements, as indicated by the large confidence intervals. Other sites with annual
average concentrations greater than the program-level average include DEMI, GPCO
and BXNY. Sites with relatively low annual average concentrations (less than
4-37

-------
1 ng/m3) other than UNVT include GLKY and CHSC. Annual averages could not be
calculated for LBHCA and SYFL.
•	The program-level average concentration of acetaldehyde is 1.80 pg/m3, as shown in
purple in Figure 4-2. Site-specific annual average concentrations range from
0.46 pg/m3 (BMCO) to 4.18 pg/m3 (BTUT). The annual average concentrations for
BTUT and GPCO are twice the program-level average for acetaldehyde. GPCO and
SPIL have the most variability associated with their measurements, as indicated by
the confidence intervals shown. Other sites with annual average concentrations
greater than the program-level average include CSNJ, ELNJ, NBIL, OCOK, PXSS,
ROIL, S4MO, TMOK, and TOOK. Sites with relatively low annual average
concentrations (less than 1 pg/m3) other than BMCO include BRCO, RFCO, GLKY,
SEWA, and PACO. Annual averages could not be calculated for ADOK, RICO, or
YUOK.
•	Figure 4-3 shows the inter-site variability graph for arsenic, which also includes a
comparison of PMio results and TSP results. Note that only sites from Oklahoma are
using TSP samplers. The program-level average concentration of arsenic (PMio) is
0.67 ng/m3, while the program-level average concentration of arsenic (TSP) is
0.61 ng/m3. There is more variability across the program associated with the
PMio measurements than the TSP measurements, as indicated by the range of annual
averages as well as confidence intervals shown. Site-specific annual average arsenic
concentrations range from 0.28 ng/m3 (UNVT) to 1.24 ng/m3 (ASKY-M) for PMio
and 0.46 ng/m3 (OCOK) to 0.80 ng/m3 (TOOK) for TSP. Annual averages could not
be calculated for ADOK and YUOK. BTUT has the most variability in the PMio
measurements, while TOOK has the most variability in the TSP measurements,
although the confidence intervals calculated for BTUT are nearly four times larger
than those for TOOK.
•	Figure 4-4a is the inter-site variability graph for benzene, as measured with
Method TO-15. The program-level average concentration of benzene is 0.78 pg/m3.
Site-specific annual average concentrations range from 0.37 pg/m3 (UNVT) to
1.56 pg/m3 (ANAK). Although the annual average concentrations for ANAK and
ASKY are similar, the variability associated with the measurements collected at
ASKY is considerably higher, as indicated by the confidence intervals shown in
Figure 4-4a. Other sites with annual average concentrations greater than 1 pg/m3
include TROK, PXSS, TVKY, SPAZ, and TOOK. Sites with relatively low annual
average concentrations (less than 0.5 pg/m3) other than UNVT include NBIL, GLKY,
CHNJ, and CCKY.
•	Figure 4-4b is the inter-site variability graph for benzene, as measured with the
concurrent SNMOC method. Canister samples collected at seven sites are analyzed
with this method. The program-level average concentration of benzene (SNMOC
only) is 1.20 pg/m3. Site-specific annual average concentrations range from
0.56 pg/m3 (NBIL) to 1.96 pg/m3 (PACO). The annual average concentrations for
PACO and RICO are greater than the program-level average; the annual average
concentrations for NBIL and RFCO are less than the program-level average; and the
annual average concentrations for BMCO, BRCO and BTUT are similar to the
4-38

-------
program-level average benzene concentration (SNMOC only). Note that canisters
from BTUT and NBIL are analyzed using both methods and their annual averages are
similar although slightly higher using the SNMOC method.
•	Figure 4-5a is the inter-site variability graph for 1,3-butadiene, as measured with
Method TO-15. The program-level average concentration of 1,3-butadiene is
0.15 pg/m3. Site-specific annual average concentrations range from 0.006 pg/m3
(UNVT) to 1.03 pg/m3 (TVKY). It is easy to see which sites' concentrations are
driving the program-level average concentration. While most sites' annual averages
are less than the program-level average, including some whose annual averages are
similar to or just greater than the MDL, the annual average concentrations for BLKY,
LAKY, and TVKY are four or more times greater than the program-level average
concentration of 1,3-butadiene. Another Calvert City, Kentucky site (CCKY) also has
an annual average concentration greater than the program-level average, but to a
lesser extent. Each of these sites has veiy large confidence intervals, indicating that
outliers are likely influencing these annual average concentrations. The fifth Calvert
City site, ATKY, has an annual average concentration that is significantly less than
the other sites in that area.
•	Figure 4-5b is the inter-site variability graph for 1,3-butadiene, as measured with the
concurrent SNMOC method. Canister samples collected at seven sites are analyzed
with this method. The program-level average concentration of 1,3-butadiene
(SNMOC only) is 0.037 pg/m3. Site-specific annual average concentrations range
from 0.004 pg/m3 (BMCO) to 0.11 pg/m3 (RICO). The annual average concentrations
for BTUT and RICO are greater than the program-level average, with the annual
average for RICO more than twice the program-level average, while the remaining
annual average concentrations are less than the program-level average. However, with
the exception of RICO, all of the annual average concentrations are less than the
MDL for 1,3-butadiene with the SNMOC method. This means that the annual average
concentrations shown incorporate data containing many zeroes substituted for non-
detects, many concentrations that are less than the MDL, or a combination of both.
The MDL for 1,3-butadiene is considerably higher for the SNMOC method
(0.104 pg/m3) than the TO-15 Method (0.024 pg/m3). Because so many of the results
are less than the MDL or non-detects, there is less certainty associated with the
SNMOC results for this pollutant.
•	The program-level average concentration of carbon tetrachloride is 0.66 pg/m3, as
shown in blue in Figure 4-6. For most sites, the annual average concentration is either
slightly less or slightly more than the program-level average concentration and the
associated confidence levels are relatively small. This indicates that there is little
variability in the carbon tetrachloride measurements across the program. This
uniformity is not unexpected. Carbon tetrachloride is a pollutant that was used
worldwide as a refrigerant. However, it was identified as an ozone-depleting
substance in the stratosphere and its use was banned by the Montreal Protocol (EPA,
2015d). This pollutant has a long lifetime in the atmosphere, but slowly degrades over
time. Today, its concentration in ambient air is fairly ubiquitous regardless of where it
is measured. The annual average carbon tetrachloride concentrations for BLKY and
TVKY are greater than annual averages for the remaining sites, particularly for
4-39

-------
BLKY. With the exceptions of these two sites, the annual average concentrations of
carbon tetrachloride range from 0.56 pg/m3 for BTUT to 0.69 pg/m3 for SEWA.
Further, the confidence intervals for these sites are less than ± 0.04 pg/m3. For
TVKY, the annual average concentration is 0.80 ± 0.08 pg/m3, which is somewhat
higher than the other NMP sites. For BLKY, the annual average concentration is
1.11 ± 0.77 pg/m3. The confidence interval for this site's average concentration
indicates that there is considerable variability in the carbon tetrachloride
concentrations measured at this site and will be discussed in more detail in the
Kentucky section (Section 14).
•	Figure 4-7 presents the program-level and annual average concentrations of
/7-dichlorobenzene. This figure shows that the program-level average concentration
(0.044 pg/m3) and most of the site-specific annual average concentrations are less
than the MDL for this pollutant (0.14 pg/m3), as indicated by the dashed blue line.
This indicates that many of the measurements are either non-detects or less than the
detection limit. Table 4-1 shows that roughly half of the 2013 measurements of
^-dichlorobenzene are non-detects and of the measured detections, 85 percent were
less than the MDL. Only two sites have annual average concentrations greater than
the MDL for this pollutant, PXSS and SPAZ. PXSS and SPAZ account for the
greatest number of /?-dichlorobenzene measurements greater than the MDL, 38 for
PXSS and 22 for SPAZ. These two sites also accounted for the two highest annual
average concentrations of this pollutant for the 2012 NMP report. Other sites with a
higher number of measurements greater than the MDL include S4MO (16), TMOK
(12), and ADOK (10). The maximum /?-dichlorobenzene concentration, though, was
measured at BTUT (0.681 pg/m3), which is more than twice the next highest
concentration and helps explain, at least partially, why the confidence interval is so
large for a site with an annual average concentration similar to the program average
concentration.
•	Figure 4-8 shows that the annual average concentrations of 1,2-dichloroethane
calculated for some of the Kentucky sites are significantly higher than the annual
averages for other NMP sites as well as the program-level average concentration.
Excluding the Calvert City sites, annual average concentrations of 1,2-dichloroethane
range from 0.06 pg/m3 (SPAZ) to 0.11 pg/m3 (BTUT), which are all similar to or just
greater than the MDL for this pollutant (0.063 pg/m3). The annual average
concentrations of 1,2-dichloroethane for the five Calvert City sites range from
0.24 pg/m3 (CCKY) to 3.75 pg/m3 (TVKY). The confidence intervals for these
annual average concentrations are relatively large, indicating there is considerable
variability in the measurements collected at these sites. These sites are driving the
program-level average concentration (0.26 pg/m3), which was a similar finding in the
2012 NMP report. Without the Calvert City sites, the program-level average
concentration would be 0.08 pg/m3.
•	Figure 4-9a is the inter-site variability graph for ethylbenzene, as measured with
Method TO-15. The program-level average concentration of ethylbenzene is
0.36 pg/m3. Site-specific annual average concentrations range from 0.07 pg/m3
(UNVT) to 1.95 pg/m3 (KMMS). The annual average concentration for KMMS is
considerably higher than the next highest annual average concentration (0.89 pg/m3
4-40

-------
for ANAK), and has a very large confidence interval associated with it. The only
other sites with annual average concentrations greater than 0.5 pg/m3 are PXSS and
SPAZ. Sites with relatively low annual average concentrations (less than 0.15 pg/m3)
other than UNVT include CCKY, ATKY, CHNJ, TVKY, GLKY, and BLKY.
•	Figure 4-9b is the inter-site variability graph for ethylbenzene, as measured with the
concurrent SNMOC method. Canister samples collected at seven sites are analyzed
with this method. The program-level average concentration of ethylbenzene (SNMOC
only) is 0.25 pg/m3. Site-specific annual average concentrations range from
0.11 pg/m3 (BRCO) to 0.56 pg/m3 (BTUT). The annual average concentrations for
BTUT and RICO are greater than the program-level average; the annual average
concentrations for the remaining sites are less than the program-level average
concentration. Note that canisters from BTUT and NBIL are analyzed using both
methods and their annual averages are similar although slightly higher using the
SNMOC method.
•	The program-level average concentration of formaldehyde is 2.83 pg/m3, as shown in
purple in Figure 4-10. Site-specific annual average concentrations range from
0.57 pg/m3 (SEWA) to 8.05 pg/m3 (BTUT). This is the third year in a row that BTUT
has had the highest annual average concentration of formaldehyde among NMP sites.
The annual average concentrations for BTUT and GPCO are more twice the program-
level average for formaldehyde, with all other NMP sites having annual average
concentrations less than 5 pg/m3. Sites with relatively low annual average
concentrations (less than 1 pg/m3) other than SEWA include BRCO, BMCO, and
RFCO. Annual averages could not be calculated for ADOK, RICO, or YUOK.
•	Figure 4-11 presents the program-level and site-specific annual average
concentrations of hexachloro-l,3-butadiene. This figure shows that the program-level
average concentration (0.014 pg/m3) and all of the site-specific annual average
concentrations are considerably less than the MDL for this pollutant (0.304 pg/m3), as
indicated by the dashed blue line. None of the hexachloro-l,3-butadiene
measurements collected in 2013 were greater than the detection limit, as indicated in
Table 4-1. Of the 1,883 valid VOC samples collected, only 330 (or 18 percent)
included measured detections of hexachloro-l,3-butadiene. This indicates that a large
number of substituted zeroes are included in the annual averages shown in
Figure 4-11, which generally pull the averages down.
•	Figure 4-12 presents the program-level and site-specific annual average
concentrations of naphthalene. The program-level average concentration
(75.26 ng/m3), as well as all of the annual average concentrations, where they could
be calculated, are considerably greater than the MDL for this pollutant. The site-
specific annual averages varied considerably, from 10.62 ng/m3 (UNVT) to
155.94 ng/m3 (NBIL). The sites with the highest variability in their measurements, as
indicated by the magnitude of their confidence intervals, are NBIL and WPFL.
Concentrations measured at WPFL range from 4.32 ng/m3 to 506 ng/m3;
concentrations measured at NBIL range from 2.87 ng/m3 to 748 ng/m3.
4-41

-------
• Figure 4-13 shows the inter-site variability graph for nickel, which also includes a
comparison of PMio results and TSP results. Note that only sites from Oklahoma are
using TSP samplers. The program-level average concentration of nickel (PMio) is
1.24 ng/m3, while the program-level average concentration of nickel (TSP) is
1.30 ng/m3. There is more variability across the program associated with the
PMio measurements than the TSP measurements, as indicated by the range of annual
averages as well as confidence intervals shown. Site-specific annual average nickel
concentrations range from 0.36 ng/m3 (GLKY) to 2.40 ng/m3 (ASKY-M) for PMio
and 0.84 ng/m3 (OCOK) to 2.09 ng/m3 (TOOK) for TSP. Annual averages could not
be calculated for ADOK and YUOK. ASKY-M has the most variability in the PMio
measurements, with nickel measurements spanning two orders of magnitude, ranging
from 0.20 ng/m3 to 21.2 ng/m3. TOOK has the most variability in the TSP nickel
measurements, ranging from 0.69 ng/m3 to 10.98 ng/m3.
4-42

-------
Figure 4-1. Inter-Site Variability for Acenaphthene
35.0
30.0
25.0
&o
c
c 20.0
Monitoring Site
] Program Average	i i Site-Specific Average	—— — • MDL (0.038 ng/m3)

-------
Figure 4-2. Inter-Site Variability for Acetaldehyde
5.0
4.5
4.0
3.5
I 3.0
c
o
c 2.5

<
1.5
1.0
0.5
ifl
0.0
Monitoring Site
Site-Specific Average	—— —-MDL (0.013 ng/m3)
~ Program Average

-------
Figure 4-3. Inter-Site Variability for Arsenic
1.8
1.6
1.4
sr 12
E
"5B
c
.1 1.0
+¦»
ro
c
CIJ
u
Monitoring Site
] Program PM10 Average 1^^] Program TSP Average	Site-Specific Average — — —-Teflon MDL 	Quartz MDL
(0.198 ng/m3)	(0.070 ng/m3)

-------
Figure 4-4a. Inter-Site Variability for Benzene - Method TO-15





¦¦





j


T

T
T


—





rh




-1-

-L







ft
¦¦

ft
ft
T

ft
I
¦¦
ft
¦¦

i—i i
ft

ft
-¦


-1-

T






























Monitoring Site
] Program Average	Site-Specific Average	———»MDL (0.061 ng/m3)

-------
Figure 4-4b. Inter-Site Variability for Benzene - SNMOC
3.0
2.5
2.0
E
"oB
c 1.5

<
1.0
0.5
0.0
BMCO
BRCO
~ Program Average
BTUT	NBIL	PACO
Monitoring Site
i i Site-Specific Average
RFCO	RICO
• MDL (0.144 ng/m3

-------
Figure 4-5a. Inter-Site Variability for 1,3-Butadiene - Method TO-15
2
1.75
1.5
2
c
CV
u
c
o
u
CIJ
w>
ro
i—

<
1.25
0.75
0.5
0.25
0
f
—M~H—¦M-I4,fjfci,-I—t4:—,r~n.4-—
Monitoring Site
i i Site-Specific Average
-T—li
I	ip	TJI T(1 |T	tp	 jJ—' i1 11'—V"	V—*i rr i m 11—
~ Program Average
»• MDL (0.024 ng/m3)

-------
Figure 4-5b. Inter-Site Variability for 1,3-Butadiene - SNMOC
hU
CO
0.14
0.12
BMCO
BRCO
BTUT
NBIL
Monitoring Site
PACO
RFCO
RICO
i i Program Average
] Site-Specific Average
• MDL (0.104 ng/m3)

-------
Figure 4-6. Inter-Site Variability for Carbon Tetrachloride
on
O
2.0
1.8
1.6
1.4
a 1.2
c
o
c 1.0

<
0.6
0.4
0.2
0.0
Monitoring Site
i i Site-Specific Average
J	LjJ	L,J	Ljl	lTl	Ljl	Lyi	Lpl	L,J	Ljl	lTJ	lTl	lTl		L,—
i i Program Average
MDL (0.101 ng/m3)

-------
Figure 4-7. Inter-Site Variability for /;-Dichlorobenzene
on
0.30
0.25
°-20
E
i
c
o
c

ro
i—

<
0.15
0.10
0.05
0.00
ntifil
ft
Monitoring Site
i i Site-Specific Average
r*~i
r*i
.J—v—v—M—M—M—M—M—M—V—M—M—v—v—h—
i i Program Average
MDL (0.140 ng/m3)

-------
Figure 4-8. Inter-Site Variability for 1,2-Dichloroethane
on
oo















L-t-t,-.x jl

""p — -1

Monitoring Site
] Program Average	Site-Specific Average	— — —-MDL (0.063 ng/m3)

-------
Figure 4-9a. Inter-Site Variability for Ethylbenzene - Method TO-15
on
CO
3.5
3.0


-Ji
a
Lffl
rr
Monitoring Site
i i Program Average	i i Site-Specific Average	—— — • MDL (0.075 ng/m3)

-------
Figure 4-9b. Inter-Site Variability for Ethylbenzene - SNMOC
0.9
0.8
0.7
0.6
I
o 0.5
3 0.4

-------
Figure 4-10. Inter-Site Variability for Formaldehyde
on
On
10.0
9.0
8.0
7.0
%
j£ 6.0
c
o
ro
E 5.0

-------
Figure 4-11. Inter-Site Variability for Hexachloro-l,3-butadiene
on
CT3
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00
Arh.rh,
m


ft
rh
El

Monitoring Site
i i Program Average	i i Site-Specific Average	—— — • MDL (0.304 ng/m3)

-------
Figure 4-12. Inter-Site Variability for Naphthalene
on
—a
250
200
1
1? 150
Monitoring Site
] Program Average	Site-Specific Average	———»MDL (0.207 ng/m3)

-------
Figure 4-13. Inter-Site Variability for Nickel
on
oo
3.5
3.0
2.5
w>
c
c 2.0
o
c
OJ
u
c
o
u 1.5

ro
i-
OJ
>
<
1.0 --
0.5 --
0.0
/ ^ *
^ / o
* 
-------
4.4.2 Quarterly Variability Analysis
Figures 4-14 through 4-26 provide a graphical display of the site-specific quarterly
average concentrations for each of the program-level pollutants of interest. Quarterly averages
are calculated based on the criteria specified in Section 3.1. The published MDL from the ERG
laboratory is also plotted on each graph, similar to the inter-site variability graphs. Note that the
scales on the PMio and TSP graphs are the same for a given speciated metal. The same is also
true for the air toxics measured by both Method TO-15 and the concurrent SNMOC method.
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. Another reason for missing quarterly averages in the figures
is due to the sampling duration of each site. Some sites started late or ended early in the year,
which may result in a lack of quarterly averages. 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). No quarterly average concentration is presented for sites that
did not meet this criterion.
Most of the program-level pollutants of interest were detected year-round. Few were
detected less frequently. For instance, hexachloro-l,3-butadiene was not detected at every site, as
shown in Figure 4-24. This pollutant was not detected at BTUT, SEWA, or SPAZ, and was
detected in two or fewer quarters at another three sites. However, comparing the quarterly
averages for sites with four valid quarterly averages in a year may reveal a temporal trend for
other pollutants. Examples of this include the following:
•	Quarterly averages of formaldehyde tend to be highest for the summer months,
based on previous reports. Figure 4-23 shows that 17 of the 33 sites sampling
formaldehyde exhibited the highest quarterly average for the third quarter (from
July through September), which is shown in green. In addition, another 13 sites
exhibited their highest quarterly formaldehyde average for the second quarter
(from April and June), which is shown in red. Thus, it appears that formaldehyde
concentrations tend to be highest during the warmer months of the year, although
there are exceptions.
•	Conversely, benzene averages tend to be higher for the winter months. As shown
in Figure 4- 17a, 21 sites have their highest quarterly benzene concentration for
the first quarter (shown in blue) and another nine exhibited their highest quarterly
4-59

-------
average for the fourth quarter (shown in purple). Similarly, two sites have their
highest quarterly benzene concentration for the first quarter and three sites
exhibited their highest quarterly average for the fourth quarter (shown in purple)
for those sampling benzene with the SNMOC method, as shown in Figure 4-17b.
Note, however, that for those sampling with the SNMOC method, few sites have a
quarterly average concentrations shown for all four quarters of the year.
•	Other notable trends include 1,3-butadiene with higher concentrations in the first
and fourth quarters, acenaphthene with higher concentrations in the second and
third quarters, and acetaldehyde with higher concentrations in the second and
third quarters.
•	Concentrations of some pollutants had a tendency to be higher in one quarter over
the others but the differences among the quarters were so small, it makes little
difference. For instance, 20 of the 34 sites sampling 1,2-dichlorothane have their
maximum quarterly average concentration for the second quarter of the year. But
a review of the quarterly average concentrations in Figure 4-21 shows that the
quarterly averages varied little for most of the sites. A similar observation can be
made for carbon tetrachloride in Figure 4-19. Twenty-seven of the 34 sites
sampling VOCs have their maximum quarterly average carbon tetrachloride
concentration for either the second or third quarter of 2013, but the quarterly
average concentrations for all but one monitoring site vaiy by less than
0.15 pg/m3.
The quarterly average concentration comparison also allows for the identification of sites
with unusually high concentrations of the pollutants of interest compared to other sites and when
those high concentrations were measured. The quarterly average graphs may also reveal if
concentrations measured at a particular site are significantly lower than other sites. These graphs
may also reveal when there is veiy little variability in the quarterly averages across other sites.
Inter-state trends may also be revealed. Examples include the following:
• Figure 4-14 for acenapthene shows that the second and third quarter average
concentrations for NBIL and ROCH were considerably higher than their other
quarterly averages, as well as those calculated for other sites.
•	Figure 4-15 is the quarterly average graph for acetaldehyde. This figures shows
that the quarterly averages are fairly variable. Of note, the quarterly averages for
the Garfield County, Colorado sites tended to be the lowest of NMP sites
sampling acetaldehyde. BMCO, BRCO, PACO, and RFCO, in addition to GLKY
and SEWA, are the only sites with all four quarterly average concentrations of
acetaldehyde less than 1 pg/m3.
•	ASKY-M is the only monitoring site for which all four quarterly average
concentrations of arsenic are greater than 1 ng/m3, as shown in Figures 4-16a and
4-60

-------
4-16b. BTUT's fourth quarter average concentration is the only quarterly average
among the sites sampling arsenic that is greater than 2 ng/m3.
ANAK's first and fourth quarter average benzene concentrations are both greater
than 2 pg/m3, as shown in Figures 4-17a and 4-17b, as are RICO's; PACO's
fourth quarter average concentration is greater than 2.50 pg/m3. ASKY's fourth
quarter average concentration is the only quarterly average among NMP sites
sampling benzene that is greater than 3 pg/m3.
Figures 4-18a and 4-18b are the quarterly average graphs for 1,3-butadiene.
Figure 4-18a shows that the second and fourth quarter average concentrations for
three of the Calvert City, Kentucky sites (BLKY, LAKY, and TVKY) are
considerably higher than the quarterly average concentrations calculated for other
NMP sites. This is also true for the fourth quarter average concentration for
CCKY. The ATKY monitoring site is also located in Calvert City, but does not
reflect this trend. The first and fourth quarter averages for PXSS and SPAZ are
also greater than most sites' quarterly average concentrations of 1,3-butadiene.
For sites sampling SNMOCs, only RICO has quarterly average concentrations
greater than the MDL for this pollutant.
Figure 4-19 is the quarterly average graph for carbon tetrachloride. Nearly all of
the quarterly average concentrations calculated for each site fall within a
relatively small range, generally between 0.50 pg/m3 and 0.75 pg/m3. However,
there are two exceptions to this. All four quarterly average concentrations for
TVKY are greater than 0.75 pg/m3, ranging from 0.76 pg/m3 for the fourth
quarter of 2013 to 0.85 pg/m3 for the first quarter of 2013. Three of the four
quarterly average concentration for BLKY are greater than 0.75 pg/m3, including
the first quarter average concentration, which is 2.21 pg/m3. The quarterly
average concentrations for the remaining Calvert City sites do not reflect this
trend.
Nearly all of the quarterly average concentrations of /;-dichlorobenzene are less
than the MDL for this pollutant, as shown in Figure 4-20. The MDL and detection
rate of this pollutant were discussed in the previous section. However, all four
quarterly average concentrations for SPAZ are greater than the MDL. In addition,
three of the four quarterly average concentrations for PXSS are also greater than
the MDL. The only other site for which a quarterly average concentration of
^-dichlorobenzene was greater than the MDL is for TMOK's third quarter
average.
As shown in Figure 4-21, most of the quarterly average concentrations for NMP
sites measuring 1,2-dichloroethane are similar to the MDL for this pollutant. The
exceptions to this are all for the Calvert City sites.
Most of the quarterly average concentrations of ethylbenzene are less than
0.75 pg/m3, as shown in Figures 4-22a and 4-22b. Exceptions to this include first
and fourth quarter average concentrations for ANAK, PXSS, and SPAZ. In
addition, all of the quarterly average concentrations of ethylbenzene for KMMS
4-61

-------
are greater than 0.75 pg/m3. Quarterly averages calculated for KMMS range from
0.85 pg/m3 for the fourth quarter of 2013 to 3.35 pg/m3 for the second quarter of
2013.
•	Figure 4-23 is the quarterly average concentration graph for formaldehyde. This
figure shows that most of the quarterly average concentrations are less than
6 pg/m3. With the exception of BTUT's first quarter average concentration, the
only quarterly averages greater than 6 pg/m3 are second and/or third quarter
averages. Similarly, all but four of the 23 quarterly averages greater than 4 pg/m3
were calculated for the second or third quarter. Only GPCO and BTUT have
quarterly average concentrations greater than 10 pg/m3, the first quarter for BTUT
and the second quarter for GPCO.
•	All of the quarterly average concentrations of hexachloro-1,3-butadiene are
roughly equal to or less than 0.05 pg/m3. However, the MDL for this pollutant is
0.304 pg/m3. As discussed previously, the detection rate for this pollutant is
relatively low. Of note, the fourth quarter averages, where they could be
calculated, were most often the highest quarterly average for sites where all four
are available. Of the 30 sites with four quarterly averages of hexachlor-1,3-
butadiene in Figure 4-24, the fourth quarter average concentration is the
maximum quarterly average for 26 of them.
•	Figure 4-25 for naphthalene shows that there is considerable variability in the
quarterly average concentrations calculated for NBIL. Quarterly average
concentrations for this site range from 33.16 ng/m3 for the fourth quarter of 2013
to 304.90 ng/m3 for the second quarter of 2013. This graph also shows that
naphthalene concentrations measured at CHSC, GLKY, and UNVT tended to be
lower than many of the other NMP sites.
•	Figures 4-26a and 4-26b show that concentrations of nickel tended to be highest
at ASKY-M, SEWA, and TOOK. These are the only NMP sites sampling nickel
with at least two quarterly average concentrations greater than 2 ng/m3. All four
quarterly averages of nickel are greater than 2 ng/m3 for ASKY-M, three are
greater than 2 ng/m3 for TOOK (all but the fourth quarter of 2014), and two are
greater than 2 ng/m3 for SEWA (second and third quarters only). The fourth
quarter average for BTUT is also greater than 2 ng/m3.
Graphing the data by method (TO-15 and SNMOC) or by particulate fraction (PMio and
TSP) may reveal other trends. Examples include the following:
•	Figures 4-18a and 4-18b show that there can be a difference in detection rates
between methods. 1,3-Butadiene was measured with both the TO-15 and SNMOC
methods. Figure 4-18a presents the quarterly averages for sites sampling VOCs
with Method TO-15 and Figure 4-18b presents the quarterly averages for sites
sampling using the SNMOC method. Most sites sampling 1,3-butadiene with
Method TO-15 detected this pollutant year-round. With those sampling with the
SNMOC method, this is harder to determine. Figure 4-18b has fewer bars on it. In
4-62

-------
some cases, such as the second and third quarter of 2013 for BMCO, this
compound was not detected; thus, the quarterly average for this site is zero. In
other cases, such as the first quarter for BMCO, the site does not meet the valid
sample criteria and thus, no quarterly average was calculated. The MDLs between
the two methods are quite different (0.024 pg/m3 for TO-15 and 0.104 pg/m3 for
SNMOC).
• Splitting the metals graphs based on particulate fraction isolates the Oklahoma
sites from sites in other states. For both arsenic and nickel, the Tulsa sites tended
to have higher concentrations of these pollutants than the Oklahoma City sites.
4-63

-------
Figure 4-14. Comparison of Average Quarterly Acenaphthene Concentrations
55.0
Monitoring Site
1st Quarter	2nd Quarter	3rd Quarter	4th Quarter	———•MDL
(0.038 ng/m3)

-------
Figure 4-15. Comparison of Average Quarterly Acetaldehyde Concentrations
6.00
5.00
E
"oB 4.00
o 3.00

<
>
1—

-------
Figure 4-16a. Comparison of Average Quarterly Arsenic (PMio) Concentrations
2.50
2.00
.1 1.50
nj
c

OJ
t
nj
3
a
0.50
¦
0.00
L


...
h
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
•Teflon MDL
(0.198 ng/m3)
• Quartz MDL
(0.070 ng/m3)

-------
Figure 4-16b. Comparison of Average Quarterly Arsenic (TSP) Concentrations
2.50
2.00
1.50
£
OJ
u
c
o
u
OJ
CxO
nj
i_
OJ
5
>
OJ
t
nj
3
a
1.00
0.50
0.00
ADOK
OCOK
TMOK	TOOK
Monitoring Site
TROK
YUOK
11st Quarter
I 2nd Quarter
3rd Quarter
14th Quarter
¦Ouartz MDL
(0.070 ng/m3)

-------
Figure 4-17a. Comparison of Average Quarterly Benzene (Method TO-15) Concentrations
3.50
3.00
ST 2-50
E
"oB
c
o
2	2.00
C

<
>
OJ
E
nj
3
® 1.00
0.50
0.00
,
II

11st Quarter
12nd Quarter
Monitoring Site
3rd Quarter
14th Quarter
---•MDL
(0.061 pg/m3)

-------
Figure 4-17b. Comparison of Average Quarterly Benzene (SNMOC) Concentrations
3.50
3.00
2.50
E
"oB
c
o
2 2.00
4-»
c

<
>•

-------
Figure 4-18a. Comparison of Average Quarterly 1,3-Butadiene (Method TO-15) Concentrations
2.25
2.00
1.75
E
"oB 1.50
? 1.25

<
>

-------
Figure 4-18b. Comparison of Average Quarterly 1,3-Butadiene (SNMOC) Concentrations
2.25
2.00
1.75
1.50
c
o
u

<
>

-------
Figure 4-19. Comparison of Average Quarterly Carbon Tetrachloride Concentrations
1.50
1st Quarter Average
Concentration for
BLKY is 2.21 |ig/m3
1.25
o 0.75
t 0.50
0.25
0.00
^ J? &
& &	.rf
Monitoring Site
1st Quarter	2nd Quarter	3rd Quarter	4th Quarter	———•MDL
(0.101 pg/m3)

-------
Figure 4-20. Comparison of Average Quarterly />-Dichlorobenzene Concentrations
0.40
0.35
0.30
E
"oB
c 0.25
o
c

<
> 0.15
i_
OJ
E
nj
3
a
o.io
0.05
0.00
J

Jt
I

11st Quarter
I 2nd Quarter
Monitoring Site
3rd Quarter
14th Quarter
---•MDL
(0.140 pg/m3)

-------
Figure 4-21. Comparison of Average Quarterly 1,2-Dichloroethane Concentrations
3.00
2.50
E
2.00
C

<
>
t 1.00
TO
3
a
0.50
4th Quarter Average
Concentration for
TVKY is 9.73 |jig/m3
¦———————————i
0.00
mJ
i
tirlUdllrlrliB
i i i i i i
J

i i i i i i	i i i i i i

11st Quarter
I 2nd Quarter
Monitoring Site
3rd Quarter
14th Quarter
---•MDL
(0.063 pg/m3)

-------
Figure 4-22a. Comparison of Average Quarterly Ethylbenzene (Method TO-15) Concentrations
2.50
2.00
2nd Quarter Average
Concentration for
KMMS is 3.35 |ig/m3
i	
E
"SB
c
¦2 1.50
C
OJ
u
£
O
u
OJ
U)
nj
< 1.00

-------
Figure 4-22b. Comparison of Average Quarterly Ethylbenzene (SNMOC) Concentrations
2.50
2.00
E
"op
c
o
•4= 1.50
£
OJ
u
c
o
u
OJ
W)
nj
OJ
5
>

-------
Figure 4-23. Comparison of Average Quarterly Formaldehyde Concentrations
12.0
Monitoring Site
1st Quarter	2nd Quarter	3rd Quarter	4th Quarter	———•MDL
(0.014 pg/m3)

-------
Figure 4-24. Comparison of Average Quarterly Hexachloro-l,3-Butadiene Concentrations
0.35
0.30
ST °-25
E
"SB
£ 0.20
c

-------
Figure 4-25. Comparison of Average Quarterly Naphthalene Concentrations
250
~ 150
o 125
< 100
2nd Quarter Average
Concentration for
NBIL is 304.90 ng/m3
in FT
[I n
I I.
mn iirrrin
/ / d? # / £	#	¦/ ^ f
Monitoring Site
11st Quarter	2nd Quarter	3rd Quarter
14th Quarter
---•MDL
(0.207 ng/m3)

-------
Figure 4-26a. Comparison of Average Quarterly Nickel (PMio) Concentrations
3.00
2.50
2.00
£

*_
OJ
£ 1.00
3
a
j
0.50
0.00
7

L IIJ j-L. i
ASKY-M BAKY BOMA BTUT CCKY GLKY LEKY NBIL
Monitoring Site
PAFL PXSS S4MO SEWA SJJCA UNVT
11st Quarter
12nd Quarter
I 3rd Quarter
14th Quarter
¦Teflon MDL
(0.250 ng/m3)
• Quartz MDL
(1.24 ng/m3)

-------
Figure 4-26b. Comparison of Average Quarterly Nickel (TSP) Concentrations
3.00
2.50
2.00
C
OJ
u
| 1-50
OJ
w>
ro
i-
OJ
>
k.
OJ
ro
3
a
1.00
0.50

0.00
ADOK
OCOK
TMOK
TOOK
TROK
YUOK
Monitoring Site
11st Quarter
12nd Quarter
3rd Quarter
14th Quarter
¦•Quartz MDL
(1.24 ng/m3)

-------
4.5 Greenhouse Gases from Method TO-15
Table 4-14 presents the program-level average concentrations for the 11 GHGs measured
using Method TO-15, in descending order by GWP. Also included in Table 4-14 is the alternate
name or acronym of each pollutant, where applicable. As shown, most of the GHGs were
detected frequently. The detection rate ranged from 73 percent to 100 percent, with only
chloroform, bromomethane, 1,2-dichloroethane, and 1,1,1-trichloroethane detected in less than
95 percent of VOC samples collected (out of a total 1,883 valid VOC samples).
Dichlorodifluoromethane (CFC-12) and dichlorotetrafluoroethane (CFC-114) have the highest
GWPs of the GHGs measured by Method TO-15 (10,200 and 8,590, respectively), while
bromomethane and 1,2-dichloroethane have the lowest GWPs (2 and <1, respectively). The
GWP for 1,2-dichloroethane is new for this report.
Dichloromethane has the highest program-level average concentration among the GHGs
measured (8.17 ± 6.51 pg/m3), although this average concentration is influenced by outliers, as
indicated by the confidence interval. A review of the data shows that three concentrations greater
than 1,000 pg/m3 were measured at BTUT; eight additional concentrations greater than
100 ]ig/m3 were also measured at this site. Dichloromethane concentrations greater than
50 ]ig/m3 were also measured at NBIL, DEMI, and GPCO. However, the median concentration
of this pollutant is less than 0.5 pg/m3, indicating that these high concentrations are the exception
rather than the rule.
Dichlorodifluoromethane, trichlorofluoromethane, and chloromethane are the only other
pollutants with program-level average concentrations greater than 1.0 pg/m3. With the exception
of chloroform and 1,2-dichloroethane, the confidence intervals for the remaining pollutants are
relatively small, indicating little variability in the measurements of these pollutants. The
variability in the chloroform concentration is primarily due to a few high concentrations
measured at NBIL (five concentrations ranging from 3.40 pg/m3 to 94.9 pg/m3) and BLKY (one
concentration of 29.0 pg/m3). All but 21 chloroform concentrations (out of 1,883) measured
across the program are less than 1 pg/m3. The variability in the 1,2-dichloroethane measurements
was discussed in Section 4.2 and is attributable to measurements collected at the Calvert City,
Kentucky sites.
4-82

-------
Table 4-14. Greenhouse Gases Measured by Method TO-15
Pollutant
Alternate Name
or Acronym
Global
Warming
Potential1
(100 yrs)
Total # of
Measured
Detections2
2013
Program
Average
frig/m3)
Dichlorodifluoromethane
CFC-12
10,200
1,883
2.52
+ 0.02
Dichlorotetrafluoroethane
CFC-114
8,590
1,882
0.12
+ <0.01
Trichlorotrifluoroethane
CFC-113
5,820
1,882
0.64
+ <0.01
T richlorofluoromethane
CFC-11
4,660
1,883
1.48
+ 0.02
Carbon Tetrachloride

1,730
1,882
0.66
+ 0.02
1,1,1 T richloroethane
Methyl chloroform
160
1,365
0.04
+ <0.01
Chloroform

16
1,478
0.24
± 0.11
Chloromethane
Methyl chloride
12
1,883
1.16
± 0.01
Dichloromethane3
Methylene chloride
9
1,791
8.17
± 6.51
Bromomethane
Methyl bromide
2
1,404
0.05
± 0.02
1,2-Dichloroethane

<1
1,605
0.26
± 0.12
'GWPs presented here are from the Intergovernmental Panel on Climate Change (IPCC) Fifth Assessment
Report (AR5) (IPCC, 2014).
2Out of 1,883 valid samples
3The total number of concentrations is not equal to 1,883 due to co-elution.
4-83

-------
5.0	Site in Alaska
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the CSATAM site in Alaska, 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 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 ANAK site is located in Anchorage, Alaska. Figure 5-1 is a composite satellite
image retrieved from ArcGIS Explorer showing the monitoring site and its immediate
surroundings. Figure 5-2 identifies nearby point source emissions locations by source category,
as reported in the 2011 NEI for point sources, version 2. Note that only sources within 10 miles
of the site are included in the facility counts provided in Figure 5-2. A 10-mile boundary was
chosen to give the reader an indication of which emissions sources and emissions source
categories could potentially have a direct effect on the air quality at the monitoring site. Further,
this boundary provides both the proximity of emissions sources to the monitoring site as well as
the quantity of such sources within a given distance of the site. Sources outside the 10-mile
boundary are still visible on the map for reference, but have been grayed out in order to
emphasize emissions sources within the boundary. Table 5-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates.
5-1

-------
Figure 5-1. Anchorage, Alaska (ANAK) Monitoring Site
to
w
V&l?
, rAv 	
dJLPL 4 X -rMk «L.^
*» Einrtlo A«1
w.
..MemK Fefcllk.
¦€*15t*"Aw--
> 10 li>	'I
	-K -A , -	^ jA	
. • ¦ -	«> Dibft^-Ra—
fjH
S 2.5-U3L 1«4
! 'A • I S n«C 1( -.7 * IIJ ¦ I" &
f *S ,	S
_ { iJiJ^Orf ,i, .
ts * i »'•  V t-™
• :* v *1 trj1
Jlu~& '
ut* C-f
iSr»wfP*fk
q
44

\3*
8'
O*
f ?4* Aw«
t*ng

In a
_ ¦ 4 I. .r.Mith *— *
r-T I
E 17lh Aw* «
I9lh Ay*
kM»*i^OlKAi«
«Ourc«: USGS
AntflOf Pj'k Source NASA NGA USGS
2008 Microsoft Corp,-
9*onimt|.
esri®

-------
Figure 5-2. NEI Point Sources Located Within 10 Miles of ANAK
Malanus*a-5usrt/ia
' County
Anchorage
County
•49 4'jrrw i«rsstrw WSBNI *48 25wv
Now Oue 1o fac«t»	and eolloeafton tha total toalrtw*
displayed n»y not rocrosont all f»ci«»s vnthai »ie a*oa of nterost.
Legend
ANAK CSATAM site	10 mile radius		 County boundary
Source Category Group (No. of Facilities)
T	A*porVAjrUne/Airport Support Operatrans (19)
B	Bulk Terminal&rfBulk Plants (2)
*	Etectnoty Generation via Combustion (3)
o	Institutional (school, hospital, prison etc ) (1)
It	Military Base'National Security Facility (2)
?	Miscellaneous Commercial/industrial Facility (1)
H	Plastic Resin, or Rubber Products Plant (1)
5-3

-------
Table 5-1. Geographical Information for the Alaska Monitoring Site
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Additional Ambient Monitoring Information1
ANAK
02-020-0018
Anchorage
Anchorage
Anchorage, AK
61.205861,
-149.824602
Residential
Suburban
CO, PMio, PM2 5
1 Data for additional pollutants are reported to AQS for ANAK (EPA, 2014b); however, these data are not generated by ERG and are therefore not included in this report.
-U

-------
Anchorage is located near the end of the Cook Inlet, on the landmass between the Knik
Arm and the Turnagain Arm. The city is surrounded primarily by mountains, including several
national parks. The monitoring site is located in the north-central portion of the city, on the roof
of Trinity Christian Reformed Church, off 16th Avenue. Figure 5-1 shows that residential
subdivisions surround the monitoring site. Merrill Field Airport and the Alaska Regional
Hospital are located just north of Debarr Road, both of which are shown in the top-left corner of
Figure 5-1.
Figure 5-2 shows that the monitoring site is located in close proximity to a number of
emissions sources. The source category with the greatest number of emissions sources near
ANAK is the airport/airline/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. Other nearby emissions sources include bulk terminals and bulk plants,
facilities generating electricity via combustion, institutions (which include schools, prisons,
and/or hospitals), military bases, and a plastic, resin, or rubber product plant. The closest sources
to ANAK are both in the "airport" category: the heliport at the Alaska Regional Hospital and
Merrill Field Airport.
Table 5-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Alaska monitoring site. Table 5-2 includes both county-level
population and vehicle registration information. Table 5-2 also contains traffic volume
information for ANAK as well as the location for which the traffic volume was obtained.
Additionally, Table 5-2 presents the county-level daily VMT for the Anchorage Municipality
from the 2011 NEI, version 2.
Table 5-2. Population, Motor Vehicle, and Traffic Information for the Alaska Monitoring
Site
Site
County
Estimated
County
Population1
County-level
Vehicle
Registration2
Annual
Average
Daily
Traffic3
Intersection
Used for
Traffic Data
County-
level Daily
VMT4 "
ANAK
Anchorage
300,950
358,999
20,193
Debarr Rd between Airport
Heights Dr and Bragaw St
5,301,813
bounty-level population estimate reflects 2013 data (Census Bureau, 2014)
2County-level vehicle registration reflects 2013 data (AK DMV, 2014)
3AADT reflects 2012 data (AK DOT, 2012)
4County-level VMT reflects 2011 data (EPA, 2015a)
5-5

-------
Observations from Table 5-2 include the following:
•	The population for the Anchorage Municipality is in the middle-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 ANAK is in the middle of the range compared to other NMP
sites. The traffic estimate provided is for Debarr Road between Airport Heights Drive
and Bragaw Street.
•	The daily VMT for the Anchorage Municipality is 5.3 million miles and ranks in the
bottom-third compared to other counties with NMP sites.
5.2 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
site in Alaska on sample days, as well as over the course of the year.
5.2.1	Climate Summary
The city of Anchorage is surrounded by the waters of the Cook Inlet to the north, west,
and south. The climate of Anchorage is considered a transition zone from maritime to continental
(WRCC, 2014). The Chugach Mountains to the south and east prevent warm, moist air from
moving northward from the Gulf of Alaska while the Alaska Range to the north and west acts as
a barrier to very cold air moving southward. Although there are four distinct seasons in
Anchorage, winters are long, extending from October through April, and snowfall is common.
Due to its high latitude, daylight lasts about 19 hours in June and only 6 hours in December.
Winds are generally light, although very strong winds off the surrounding mountains occur
occasionally during the winter. The prevailing wind direction in Anchorage is from the north
(Wood, 2004).
5.2.2	Meteorological Summary
Hourly meteorological data for 2013 were retrieved from NCDC for the weather station
closest to the Alaska monitoring site (NCDC, 2013), as described in Section 3.4.2. The closest
weather station is located at Merrill Field Airport (WBAN 26409). Additional information about
the Merrill Field Airport weather station, such as the distance between the site 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.
5-6

-------
Table 5-3. Average Meteorological Conditions near the Alaska Monitoring Site
Closest Weather
Station
Distance
and

Average
Maximum
Average
Average
Dew Point
Average
Wet Bulb
Average
Relative
Average
Sea Level
Average
Scalar Wind
(WBAN and
Coordinates)
Direction
from Site
Average
Type1
Temperature
(°F)
Temperature
(°F)
Temperature
(°F)
Temperature
(°F)
Humidity
(%)
Pressure
(mb)
Speed
(kt)
Anchorage, Alaska - ANAK
Merrill Field
1.3
miles
Sample
Days
44.4
38.4
28.6
34.5
70.4
1011.9
3.1
Airport
(64)
±4.7
±4.6
±4.4
±4.2
±3.4
±2.5
±0.4
26409
307°
(NW)








(61.22, -149.86)

44.9
38.8
28.8
34.9
70.1
1010.9
3.2

2013
± 1.9
± 1.9
± 1.8
± 1.7
± 1.5
± 1.1
±0.2
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
<1

-------
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 2013. 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 are very similar to conditions experienced throughout 2013.
The average maximum temperature and average daily temperature calculated for ANAK
for 2013 are the lowest among all NMP sites. This site also has the lowest average sea level
pressure for 2013 among NMP sites.
5.2.3 Wind Rose Comparison
Hourly surface wind data from the weather station at Merrill Field Airport near ANAK
were uploaded into a wind rose software program to produce customized wind roses, as
described in Section 3.4.2. A wind rose shows the frequency of wind directions using "petals"
positioned around a 16-point compass, and uses different colors to represent wind speeds.
Figure 5-3 presents a map showing the distance between the weather station and ANAK,
which may be useful for identifying topographical influences that can affect the meteorological
patterns experienced at this location. Figure 5-3 also presents three different wind roses for the
ANAK monitoring site. First, a historical wind rose representing 2003 to 2012 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 2013 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2013 is
presented. These can be used to identify the predominant wind speed and direction for 2013 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.
5-8

-------
Figure 5-3. Wind Roses for the Merrill Field Airport Weather Station near ANAK
Location of ANAK and Weather Station	2003-2012 Historical Wind Rose
WIND SPEED
(Kn ots)
~
~ 4-7
¦ 2-4
Calms: 37.61%
2013 Wind Rose
Sample Day Wind Rose
NORTH"--.,
WIND SPEED
(Knots)
~
H 17 -21
LI 11-17
r i 7-11
~~l 4-7
2-
Calms: 37.63%

WIND SPEED
(Kn ots)
~
¦I 11 -17
r i 7-11
~ 4-7
2- 4
Calms: 38.22%
| NORTH'
WEST
5-9

-------
Observations from Figure 5-3 for ANAK include the following:
•	The Merrill Field Airport weather station is located 1.3 miles northwest of ANAK.
Most of the airport property as well as a hospital lie between the weather station and
ANAK.
•	The historical wind rose shows that calm winds (those less than or equal to 2 knots)
were observed for nearly 40 percent of the hourly measurements over the last
10 years. For wind speeds greater than 2 knots, winds from the north were observed
most frequently (10 percent). Winds from the north-northeast and west to northwest
each account for another 5 percent to 6 percent of observations. With the exception of
southerly winds, winds from the southeast and southwest quadrants were observed
infrequently near ANAK. Wind speeds greater than 17 knots account for too few
observations near ANAK to be visible on the historical wind rose.
•	The wind patterns shown on the 2013 wind rose resemble the historical wind patterns,
although northerly winds account for a higher percentage of observations in 2013
(nearly 13 percent).
•	The sample day wind rose exhibits most of the same characteristics as the other wind
roses, with winds calm winds accounting for nearly 40 percent of observations and
north as the predominant wind direction. However, fewer north-northeasterly winds
were observed on sample days.
5.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for ANAK 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 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. VOCs and
PAHs were sampled for at ANAK.
5-10

-------
Table 5-4. Risk-Based Screening Results for the Alaska Monitoring Site
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Anchorage, Alaska - ANAK
Benzene
0.13
61
61
100.00
18.43
18.43
Carbon Tetrachloride
0.17
61
61
100.00
18.43
36.86
1.3 -Butadiene
0.03
51
54
94.44
15.41
52.27
1,2-Dichloroethane
0.038
49
49
100.00
14.80
67.07
Naphthalene
0.029
41
62
66.13
12.39
79.46
Ethylbenzene
0.4
38
61
62.30
11.48
90.94
p-Dichlorobenzene
0.091
11
38
28.95
3.32
94.26
Hexachloro -1,3 -butadiene
0.045
7
9
77.78
2.11
96.37
Xylenes
10
5
61
8.20
1.51
97.89
Benzo(a)pyrene
0.00057
4
40
10.00
1.21
99.09
Acenaphthylene
0.011
2
51
3.92
0.60
99.70
Acenaphthene
0.011
1
62
1.61
0.30
100.00
Total
331
609
54.35

Observations from Table 5-4 include the following:
•	Sixty-one valid VOC samples were collected at ANAK and concentrations of eight
VOCs failed screens. Sixty-two valid PAH samples were collected at ANAK and
concentrations of four PAHs failed. In total, 12 pollutants failed screens for ANAK.
More than half (54 percent) of all VOC and PAH concentrations measured at ANAK
were greater than their associated risk screening value, or failed screens (of those
pollutants for which a risk screening value is available).
•	Eight pollutants contributed to 95 percent of failed screens for ANAK and therefore
were identified as pollutants of interest for ANAK. These eight pollutants include
seven VOCs and one PAH (naphthalene).
•	Benzene and carbon tetrachloride were detected in every VOC sample collected at
ANAK and failed 100 percent of screens; 1,2-dichloroethane also failed 100 percent
of its screens but was detected less frequently. Naphthalene was detected in every
PAH sample collected at ANAK and failed 66 percent of screens, the highest of the
four PAHs that failed screens.
5-11

-------
5.4 Concentrations
This section presents various concentration averages used to characterize air toxics
pollution levels at the Alaska monitoring site. Where applicable, the following calculations and
data analyses were performed for each of the site-specific pollutants of interest:
•	Time period-based concentration averages (quarterly and annual) are provided for
each monitoring site.
•	Annual concentration averages are presented graphically for each site to illustrate
how the site's concentrations compare to the program-level averages, as presented in
Section 4.1.
•	Concentration averages and other statistical metrics are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at ANAK are provided in Appendices J and M.
5.4.1 2013 Concentration Averages
Quarterly and annual concentration averages were calculated for the pollutants of interest
for the Alaska 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 ANAK are presented
in Table 5-5, 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.
5-12

-------
Table 5-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest
for the Alaska Monitoring Site

# of






Measured
1st
2nd
3rd
4th


Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs. # of
Samples
Average
(Ug/m3)
Average
(Ug/m3)
Average
(Ug/m3)
Average
(Ug/m3)
Average
(Ug/m3)
Anchorage, Alaska - ANAK


2.17
1.01
0.72
2.36
1.56
Benzene
61/61
±0.75
±0.39
±0.26
±0.83
±0.34


0.19
0.08
0.06
0.26
0.15
1,3-Butadiene
54/61
±0.09
±0.04
±0.03
±0.10
±0.04


0.60
0.63
0.67
0.56
0.62
Carbon Tetrachloride
61/61
±0.04
±0.04
±0.04
±0.05
±0.02


0.06
0.05
0.02
0.07
0.05
p-Dichlorobenzene
38/61
±0.03
±0.03
±0.01
±0.02
±0.01


0.09
0.09
0.05
0.09
0.08
1,2-Dichloroethane
49/61
±0.02
±0.03
±0.02
±0.03
±0.01


1.21
0.53
0.40
1.42
0.89
Ethylbenzene
61/61
±0.46
±0.23
±0.15
±0.59
±0.22


0.01
<0.01
0.01
0.01
0.01
Hexachloro -1,3 -butadiene
9/61
±0.01
±0.01
±0.01
±0.01
±0.01


75.17
39.22
30.74
87.38
58.29
Naphthalene3
62/62
±32.95
± 13.53
±9.15
±33.79
± 13.23
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
Observations for ANAK from Table 5-5 include the following:
•	Benzene has the highest annual average concentration among the VOC pollutants of
interest (1.56 ± 0.34 |ig/m3) and is the only pollutant of interest with an annual
average greater than 1 |ig/m3. This annual average is the second highest benzene
concentration among NMP sites sampling this pollutant, as shown in Table 4-9.
•	The first and fourth quarter average concentrations of benzene are significantly
greater than the second and third quarter averages, indicating that benzene
concentrations tend to be higher during the colder months at ANAK. Concentrations
of benzene measured at ANAK range from 0.288 |ig/m3 to 6.18 |ig/m3; this maximum
benzene concentration is the fifth highest benzene concentration measured across the
program. Of the 29 concentrations less than 1 |ig/m3, 23 were measured during the
second or third quarter of 2013 and none were measured during February, November,
or December. Conversely, of the 10 concentrations greater than 3 |ig/m3 measured at
ANAK, all but one were measured during the first or fourth quarters of 2013.
•	Ethylbenzene, 1,3-butadiene, and p-dichlorobenzene also exhibit this seasonal
tendency but the differences among the quarterly averages are less significant.
•	Naphthalene is the only PAH identified as a pollutant of interest for ANAK.
Concentrations measured at ANAK were variable, ranging from 7.25 ng/m3 to
266 ng/m3. Naphthalene concentrations appear highest during the first and fourth
quarters, similar to several of the VOCs. All but one of the nine naphthalene
5-13

-------
measurements greater than 100 ng/m3 were measured during the first and fourth
quarters while all but one of the 12 measurements less than 20 ng/m3 were measured
during the second and third quarters.
5.4.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for the pollutants shaded in
gray in Table 5-4 for ANAK. Figures 5-4 through 5-11 overlay the site's minimum, annual
average, and maximum concentrations onto the program-level minimum, first quartile, median,
average, third quartile, and maximum concentrations, as described in Section 3.4.3.1.
Figure 5-4. Program vs. Site-Specific Average Benzene Concentration




-


Program Max Concentration = 43.5 ^ig/m3

u 1



0	2	4	6	8	10	12
Concentration {[ig/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range


Figure 5-5. Program vs. Site-Specific Average 1,3-Butadiene Concentration





Program Max Concentration = 21.5 jig/m3
,
C
c



H	1	1	r
0	0.3	0.6	0.9	1.2	1.5
Concentration {[ig/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range


5-14

-------
Figure 5-6. Program vs. Site-Specific Average Carbon Tetrachloride Concentration
Program Max Concentration = 23.7 ^ig/m3
0.25
0.5
0.75 1 1.25 1.5
Concentration {[ig/m3)
1.75
Program:
1st Quartile
¦
2nd Quartile 3rd Quartile 4th Quartile
~ ~ ~
Average
1
Site:
Site Average
o
Site Concentration Range

Figure 5-7. Program vs. Site-Specific Average />-Dichlorobenzene Concentration
0
0.1
0.2
0.3 0.4
Concentration {[jg/m3)
0.5
0.6

Program:
1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1

Site:
Site Average
o
Site Concentration Range


Figure 5-8. Program vs. Site-Specific Average 1,2-Dichloroethane Concentration


,
c
c

Program Max Concentration = 111 ^ig/m3





0
0.2
0.4 0.6
Concentration {[jg/m3)

0.8

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i

Site: Site Average
o
Site Concentration Range


5-15

-------
Figure 5-9. Program vs. Site-Specific Average Ethylbenzene Concentration
I
Program Max Concentration = 18.7 ^ig/m3
Concentration {[jg/m3;
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


Figure 5-10. Program vs. Site-Specific Average Hexachloro-l,3-Butadiene Concentration
0.15
Concentration {[jg/m3]
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


Figure 5-11. Program vs. Site-Specific Average Naphthalene Concentration
—
400	500
Concentration {ng/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


Observations from Figure 5-4 through 5-11 include the following:
• The program-level maximum benzene concentration (43.5 |ig/m3) is not shown
directly on the box plot in Figure 5-4 because the scale of the box plot would be
too large to readily observe data points at the lower end of the concentration
range. Thus, the scale of the box plot has been reduced to 12 |ig/m3. The
5-16

-------
maximum benzene concentration measured at ANAK is roughly one-seventh the
maximum concentration shown in Figure 5-4 and, as mentioned in the previous
section, is the fifth highest benzene concentration measured across the program.
The annual average concentration for ANAK is greater than the program-level
first, second, and third quartiles and is twice the program-level average
concentration. This site has the second highest annual average concentration of
benzene, second only to PACO.
Similar to benzene, the program-level maximum 1,3-butadiene concentration
(21.5 |ig/m3) is not shown directly on the box plot in Figure 5-5 because the scale
of the box plot would be too large to readily observe data points at the lower end
of the concentration range. Thus, the scale of the box plot has been reduced to
1.5 |ig/m3. Concentrations of 1,3-butadiene measured at ANAK range from zero
(non-detect) to 0.783 |ig/m3. The annual average concentration of 1,3-butadiene
for ANAK (0.15 ± 0.04 |ig/m3) is similar to the program-level average
concentration.
The scale of the box plot in Figure 5-6 has also been reduced to allow for the
observation of data points at the lower end of the concentration range, as the
program-level maximum carbon tetrachloride concentration (23.7 |ig/m3) is
considerably greater than the majority of measurements. Concentrations of carbon
tetrachloride measured at ANAK range from 0.359 |ig/m3 to 0.794 |ig/m3, with
the annual average concentration for ANAK just less than both the program-level
median and average concentrations.
Figure 5-7 is the box plot for /;-dichlorobenzene. Note that the program-level first
and second quartiles are both zero and therefore not visible on the box plot. All of
the />dichlorobenzene measurements collected at ANAK are less than
0.25 |ig/m3, including 23 non-detects. The annual average />dichlorobenzene
concentration for ANAK is just greater than the program-level average
concentration.
The scale of the box plot in Figure 5-8 has also been reduced to allow for the
observation of data points at the lower end of the concentration range, as the
program-level maximum 1,2-dichloroethane concentration (111 |ig/m3) is
considerably greater than the majority of measurements. Note that all of the
concentrations of 1,2-dichloroethane measured at ANAK are less than the
program-level average concentration of 0.26 |ig/m3. The annual average
concentration for ANAK is similar to the program-level median concentration of
0.085 |ig/m3.
The scale of the box plot in Figure 5-9 has also been reduced to allow for the
observation of data points at the lower end of the concentration range, as the
program-level maximum ethylbenzene concentration (18.7 |ig/m3) is considerably
greater than the majority of measurements. ANAK is one of only three NMP sites
to measure an ethylbenzene concentration greater than 3 |ig/m3 (KMMS and
BTUT are the others); in fact, two were measured at ANAK. The minimum
ethylbenzene concentration measured at ANAK is similar to the program-level
5-17

-------
first quartile. The annual average concentration for ANAK is more than twice the
program-level average concentration of 0.36 |ig/m3. As shown in Table 4-9,
ANAK has the second highest annual average concentration of ethylbenzene
among NMP sites sampling this pollutant, second only to KMMS.
•	Figure 5-10 is the box plot for hexachloro-l,3-butadiene. Note that the program-
level first, second, and third quartiles are zero and therefore not visible on the box
plot. Sixty-one valid VOC samples were collected at ANAK and of these,
hexachloro-1,3-butadiene was detected in only nine. Thus, many zeroes are
substituted into the annual average concentration of this pollutant. Yet, the annual
average for ANAK is less than the program-level average concentration of
hexachl oro-1,3 -butadi ene.
•	The maximum naphthalene concentration measured at ANAK is considerably less
than the maximum concentration measured across the program, as shown in
Figure 5-11. The annual average naphthalene concentration for ANAK is less than
the program-level average concentration but just greater than the program-level
median concentration.
5.4.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.3.2.
Although ANAK has sampled under the NMP previously, as shown in Table 2-1, sampling under
the NMP did not begin again until January 2013 at ANAK; thus, a trends analysis was not
conducted for this site.
5.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at the Alaska monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4 for
definitions and explanations regarding the various toxicity factors, time frames, and calculations
associated with these risk-based screenings.
5.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Alaska 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.
5-18

-------
Refer to Section 3.4.3.3 for an explanation of how cancer risk and noncancer hazard
approximations are calculated and what limitations are associated with them. Annual averages,
cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard approximations are
presented in Table 5-6, where applicable. Cancer risk approximations are presented as
probabilities while the noncancer hazard approximations are ratios and thus, unitless values.
Table 5-6. Risk Approximations for the Alaska Monitoring Site
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Anchorage, Alaska - ANAK
Benzene
0.0000078
0.03
61/61
1.56
±0.34
12.15
0.05
1,3-Butadiene
0.00003
0.002
54/61
0.15
±0.04
4.39
0.07
Carbon Tetrachloride
0.000006
0.1
61/61
0.62
±0.02
3.71
0.01
p-Dichlorobenzene
0.000011
0.8
38/61
0.05
±0.01
0.55
<0.01
1,2-Dichloroethane
0.000026
2.4
49/61
0.08
±0.01
2.09
<0.01
Ethylbenzene
0.0000025
1
61/61
0.89
±0.22
2.22
<0.01
Hexachloro-1,3 -butadiene
0.000022
0.09
9/61
0.01
±0.01
0.19
<0.01
Naphthalene3
0.000034
0.003
62/62
58.29
± 13.23
1.98
0.02
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
Observations for ANAK from Table 5-6 include the following:
•	The pollutants with the highest annual average concentrations for ANAK are
benzene, ethylbenzene, and carbon tetrachloride.
•	The pollutants with the highest cancer risk approximations for ANAK are benzene,
1,3-butadiene, and carbon tetrachloride. The cancer risk approximation for benzene is
12.15 in-a-million, the only cancer risk approximation greater than 10 in-a-million
calculated for ANAK. This is the second highest cancer risk approximation calculated
for benzene across the program.
•	The noncancer hazard approximations for ANAK's pollutants of interest are all
considerably less than 1.0. Noncancer hazard approximations less than 1.0 indicate
that no adverse noncancer health effects are expected from these individual
pollutants.
5-19

-------
5.5.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 5-7 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
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 3.4.3.4. Lastly,
Table 5-7 provides the pollutants with the highest cancer risk approximations (in-a-million) for
ANAK, 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 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 5.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
5-20

-------
Table 5-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Alaska Monitoring Site
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Anchorage, Alaska (Anchorage Municipality) - ANAK
Formaldehyde
253.72
Formaldehyde
3.30E-03
Benzene
12.15
Benzene
231.07
Benzene
1.80E-03
1,3-Butadiene
4.39
Acetaldehyde
111.90
1,3-Butadiene
1.33E-03
Carbon Tetrachloride
3.71
Ethylbenzene
106.03
POM, Group 2b
6.35E-04
Ethylbenzene
2.22
1.3 -Butadiene
44.36
Naphthalene
6.04E-04
1,2-Dichloroethane
2.09
Naphthalene
17.76
Ethylbenzene
2.65E-04
Naphthalene
1.98
POM, Group 2b
7.21
Arsenic, PM
2.47E-04
p-Dichlorobenzene
0.55
T etrachloroethy lene
2.36
Acetaldehyde
2.46E-04
Hexachloro-1,3 -butadiene
0.19
POM, Group 2d
2.05
POM, Group 2d
1.80E-04

T richloroethy lene
0.74
POM, Group 5a
1.42E-04

-------
Table 5-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Alaska Monitoring Site
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Anchorage, Alaska (Anchorage Municipality) - ANAK
Toluene
1,267.81
Acrolein
1,587,012.31
1,3-Butadiene
0.07
Xylenes
407.80
Formaldehyde
25,890.03
Benzene
0.05
Formaldehyde
253.72
1.3 -Butadiene
22,179.52
Naphthalene
0.02
Hexane
245.02
Acetaldehyde
12,433.63
Carbon Tetrachloride
0.01
Benzene
231.07
Benzene
7,702.29
Ethylbenzene
<0.01
Methanol
185.91
Lead, PM
6,641.46
Hexachloro-1,3 -butadiene
<0.01
Acetaldehyde
111.90
Naphthalene
5,919.53
p-Dichlorobenzene
<0.01
Ethylbenzene
106.03
Xylenes
4,078.00
1,2-Dichloroethane
<0.01
Ethylene glycol
61.50
Arsenic, PM
3,835.37

1.3 -Butadiene
44.36
Cadmium PM
3,056.13

-------
Observations from Table 5-7 include the following:
•	Formaldehyde, benzene, and acetaldehyde are the highest emitted pollutants with
cancer UREs within the Anchorage Municipality.
•	Formaldehyde and benzene are also the pollutants with the highest toxicity-weighted
emissions (of the pollutants with cancer UREs) for the Anchorage Municipality,
followed by 1,3-butadiene.
•	Eight of the highest emitted pollutants within the Anchorage Municipality also have
the highest toxicity-weighted emissions.
•	Benzene is the pollutant with the highest cancer risk approximation for ANAK and
ranks second for quantity emitted and its toxicity-weighted emissions. 1,3-Butadiene,
ethylbenzene, and naphthalene also appear on all three lists in Table 5-7. The
remaining four pollutants of interest for ANAK do not appear among the highest
emitted pollutants or those with the highest toxicity-weighted emissions.
•	Although formaldehyde tops both emissions-based lists in Table 5-7, carbonyl
compounds were not measured at ANAK under the NMP.
•	Several POM Groups rank among Anchorage's highest emitted pollutants and the
pollutants with the highest toxicity-weighted emissions. POM, Group 2b includes
acenapthene and acenaphthylene, both of which failed screens for ANAK but were
not identified as pollutants of interest. POM Group 5a, which ranks tenth for its
toxicity-weighted emissions, includes benzo(a)pyrene, which failed 10 percent of its
screens but was not identified as a pollutant of interest for ANAK. POM, Group 2d,
which appears on both emissions-based lists, includes anthracene, phenanthrene, and
pyrene, none of which failed screens for ANAK.
Observations from Table 5-8 include the following:
•	Toluene, xylenes, and formaldehyde are the highest emitted pollutants with noncancer
RfCs within the Anchorage Municipality. The emissions of toluene are considerably
greater than those of the other pollutants listed.
•	The pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) are acrolein, formaldehyde, and 1,3-butadiene. Although acrolein
was sampled for at ANAK, 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 of acrolein for Anchorage rank 11th.
•	Five of the highest emitted pollutants within the Anchorage Municipality also have
the highest toxicity-weighted emissions. Although toluene ranks highest for total
emissions, it ranks 16th for its toxicity-weighted emissions, which speaks to the
relative toxicity of toluene.
5-23

-------
• Benzene and 1,3-butadiene appear on all three lists in Table 5-8, although none of
ANAK's pollutant of interest have noncancer hazard approximations greater than 1.0.
5.6 Summary of the 2013 Monitoring Data for ANAK
Results from several of the data treatments described in this section include the
following:
~~~ VOCs and PAHs were sampledfor at ANAK throughout 2013.
~~~ Twelve pollutants failed screens for ANAK, eight VOCs andfour PAHs.
~~~ Of the site-specific pollutants of interest for ANAK, benzene had the highest annual
average concentration. ANAK has the second highest annual average concentrations
of benzene and ethylbenzene among NMP sites sampling these pollutants.
~~~ Concentrations of several VOCs, including benzene and ethylbenzene, tended to be
higher during the colder months of the year.
~~~ Benzene has the highest cancer risk approximation of the pollutants of interest for
ANAK. None of the pollutants of interest have noncancer hazard approximations
greater than an HQ of 1.0.
5-24

-------
6.0	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.
6.1	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 6-1 and 6-2 are
composite satellite images retrieved from ArcGIS Explorer showing the monitoring sites and
their immediate surroundings. Figure 6-3 identifies nearby point source emissions locations by
source category, as reported in the 2011 NEI for point sources, version 2. Note that only sources
within 10 miles of the sites are included in the facility counts provided in Figure 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 the
10-mile boundaries are still visible on the map for reference, but have been grayed out in order to
emphasize emissions sources within the boundary. Table 6-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates.
6-1

-------
Figure 6-1. Phoenix, Arizona (PXSS) Monitoring Site

-------
Figure 6-2. South Phoenix, Arizona (SPAZ) Monitoring Site




¦i
J
|u\4 • i
[•»¦ lAi wulh(j#tf Av# W -*i
\ i t

M *

| |.. Jp»4
•1
^ J'- ma*mT j
: Jt.irf* '« "«*
1
U*tr-
f

»
M|» |



¦ f *rwi*
^ i
f-fc >c^il
•


J
!¦ "¦ i »»**i^SSi3
»' Mwi

¦ «
i
.WJiivii
If'i V(flj
U' ',V*
'WU* —

i-.j.
ft i
gft '
i
C ty>. *

i_' * ^ ygt

•?
"?1
Lk

\mmi
u*
»f






1 K ' t
IBmc I
LA4


» '•%• •
• -

'


r k

if ^1


•9




I ml



- 1





f -f *4'i^*>5h ]
[ j|SM| 2'i>r * 1
1 » 3? *34


RR



U w«
w*
-•
j:'-
PL*
AC
r j
,T1j|
Eh !
5

¦¦

&»' p

PC i I
hjtyjfqBn
tA+r
i
<
t=jil


I

* .


uS
<
TO

s
IS"1
Mi r «kij
-


:-
k W?l

UL.V? '
' 4 "C1
* sf
* — —

k "•*¦^1
r/^l
I j ¦ i


» ^ it .
~
v-_ H

'" r j ,js
^«. •* 4
~ . f
ff-yBr .
SfftlJI
<•» * |J tJW»» 1
¦ < lT |
TV '
fit •
B


1 M wjit
1 |V«J
hr

pAw
r 1 ¦
u ] ,
t- * 5®

VfT
—
-
if-» 1
1 •
; f I
,
• a
¦ k
i »«

9
M 1> L.

I
f • f ^
9|
y w*- *
1 " ' *>*
~'

|r
Fi
di


J mh^*A


M *
A fe'
Alt'
2^*ik j
|1
65

b^t*1
¦ *
1 A^I
¦HJMB
. *

.j
o<> ftu
r • s
1% \

, i
. • • V" V

. -* •

^ ?.l


Jfe:
»>P M«WMf

¦ "if



> f ¦

"~ i

| ]


I JMii .
r.

L-tCh

Jr i
* ~'


i*" *¦¦

It fc A * * * i
K.. iM # «
W.mWu**"
\Vt
jip , §
¦v_^l 1

/L,
&
•" -rf" ¦', j
mo!

¦
¦
pi
~


lj— ,-g a
I 3 Ji
¦ t
rir:
- •
|»* W7 flftj
1 r'

1 -± Mi

^ I
• ^ 1
, • '


• ^ f r y
v i.i
F^i
• M t~.'

WM
I.M

' I
Lit-
.iiS.».


ij.



Ji

|
^Tj
IfPtlLi
|Y '"'1;



C
«.y
z.
1
v *4#, ^ i ¦


; |


«¦


•A* J
1 _ JL

i JL
a (
- - • -
^r.
1



1 *
¦ U"
^•1
'E
rk
Llii
mj * *


if Af
Lri«A^<
M
•¦¦¦¦



#h •
• I •
V# Atunt
Jft'l

tii n W *"g'
. to - 1 m
mTl
J
rT
. t
M»r*
iV Rom-.
1-
lUti^M




j * IM
i
"

PL£t
ffjj
- .

r - «>i

'
< 	5L.
t I
% • % 1
M IkNlri 1
;f * I'l
ar1
Vi • f'
M ¦

W rtcirv^t Rd11 J
|vk
i»!

3..' ¦
y- i,
¦n'
nw^M
"I
¦ ft
W .
Jw
K'i


p^i
j 4

•» I*pr-
'pJVI
f *r#>
J
P ¦Hfr x
te«J
1 #?'
U 1



jll TJ
1
1* 5
tag!

UA4
Krtf«#i .5 * <
J • «


-V/O'W
**• r

1^- • *
u
1
MU





\dr ». y




* | it •




fj
- *
rt^jj
k £§

Fs£|
*- It"' - *
«S!Ll3
iSwj
*.v
)"trlxf| »
» j{? .
(i» ' i i
t ' ' ^
MXtt!
v Kra
^2



vjhfl
' i
•n

Mi

fc

U
'S'.\
| p



-- ,v



ifr WfrTTL '
r'~
t j
1 Pi
j
Si
rf» • * A. -
I • rM
I v ) 1

•I • '
|U
¦ym»|
• V
•
[•
n;? 5?

-i
Wm£jr id|
i*» V?

9 .,'
r^j
i 1
A
^yi
fc-'-



;_ V |t
1 •fl
r—' ^WWTCTy i


w«

1 *" >#
I t L||
I • '5ou

NKi hr.&. iflr.c
K;
»-
t
#
'ft :
«¦£



¦B
¦ J]
• lfii

« 1
		



i .
™Qcn'
t*nc
J i- '-wLr,
¦ ¦ W
AndArfC


I • < jy

in


20 01
¦4*
J*M
Uf fliu<
1 »f> # » r{T*«
» COM

Rt

.:
uni4f
>rt-A-f. J1
*'1
'W I

[Ivjl

¦
¦s^'l



1
tTDI 1
ta

-------
Figure 6-3. NEI Point Sources Located Within 10 Miles of PXSS and SPAZ
i?-2nrw
Source Category Group (No. of Facilities)
«i<	Aeros«wc«.'Alrcfaft Manutacturmg FaoK'y (2|
T	Airport'Aaline'Alport Support Operations (36}
Aaphafl Prcducfconi'Ma? Mix Aaphai Plant (2)
X	Battery Manufacturing Facrity <1)
fi	Bu* Termlnali'Buft Plants <3)
C	Cnemlcal Manufacturing Facility (4>
6	Etectncal fcqufjment Manuteciumg Facility (Gi
I	Electnci»y Generation via Comouation <4)
E	Electroplating, P>atng, Poliahing Anorttrmg and Catering
1	Fount*** Iron and Steel (1)
&	Foundries. Non-terrous \ t)
¦	Mats* Can Bon and Otner Metal Container Manufacturing (1)
A	Matal Coating Engraving, and Allied Sarvioas to Manufacturers (1)
<•>	Metafe Proaeaamg/Faoncabon Faatty <4.1
*Ane.Qua>ryMkve density and coIIocjOop the total facilities
displayed rrajr not represent all f»cM«t within Pie a tea of oteieat
itri«rrw	*1mm	t^rtnrw	112WW
I Mancopa
I County
fi
f
- ^ _ J
^ - - - J
f	1
J	i
t	1
t	I
t	I
_ _ I	I
t ~	*
6-4

-------
Table 6-1. Geographical Information for the Arizona Monitoring Sites
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Additional Ambient Monitoring Information1
PXSS
04-013-9997
Phoenix
Maricopa
Phoenix-Mesa-
Scottsdale, AZ
33.503833,
-112.095767
Residential
Urban/City
Center
Haze, CO, S02, NO, N02, NOx, NOy, 03, S02,
Meteorological parameters, PMio, PM Coarse, PM2.5,
PM2.5 Speciation, IMPROVE Speciation.
SPAZ
04-013-4003
Phoenix
Maricopa
Phoenix-Mesa-
Scottsdale, AZ
33.40316,
-112.07533
Residential
Urban/City
Center
CO, O3, Meteorological parameters, PMn, PM Coarse,
PM2 5, PM2.5 Speciation.
1 Data for additional pollutants are reported to AQS for these sites (EPA, 2014b); however, these data are not generated by ERG and are therefore not included in this report.
BOLD ITALICS = EPA-designated NATTS Site
On

-------
PXSS is located in central Phoenix. Figure 6-1 shows that PXSS is located in a highly
residential area on North 17th Avenue. The Grand Canal is shown along the bottom of
Figure 6-1. The monitoring site is approximately three-quarters of a mile east of 1-17 and 2 miles
north of I-10. Figure 6-2 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 6-3.
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 6-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Arizona 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 Maricopa County.
Table 6-2. Population, Motor Vehicle, and Traffic Information for the Arizona
Monitoring Sites
Site
County
Estimated
County
Population1
County-level
Vehicle
Registration2
Annual
Average Daily
Traffic3
Intersection
Used for
Traffic Data
County-
level Daily
VMT4 "
PXSS
Maricopa
4,009,412
3,761,859
29,515
W Camelback Rd on either side
ofN 17th Ave
90,393,000
SPAZ
25,952
Central Ave, south of Tamarisk
bounty-level population estimate reflects 2013 data (Census Bureau, 2014)
2County-level vehicle registration reflects 2012 data (AZ DOT, 2011)
3AADT reflects 2010 data for PXSS and 2011 data for SPAZ (AZ DOT, 2014)
4County-level VMT reflects 2012 data (AZ DOT, 2013)
BOLD ITALICS = EPA-designated NATTS Site
6-6

-------
Observations from Table 6-2 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.
•	Although PXSS experiences a higher traffic volume compared to SPAZ, the traffic
volumes near these sites rank in the middle of the range compared to traffic volumes
near other NMP sites.
•	The daily VMT for Maricopa County is more than 90 million miles, which is the
second highest compared to other counties with NMP sites.
6.2 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.
6.2.1	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 1 inch. Winds are generally light and out of the east
for much of the year (Wood, 2004; WRCC, 2014).
6.2.2	Meteorological Summary
Hourly meteorological data for 2013 were retrieved from NCDC for the weather station
closest to the Arizona monitoring sites (NCDC, 2013), as described in Section 3.4.2. 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 6-3. These data were
used to determine how meteorological conditions on sample days vary from conditions
experienced throughout the year.
6-7

-------
Table 6-3. 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
(33.43, -112.00)
7.5
miles
135°
(SE)
Sample
Days
(66)
85.7
±4.0
75.2
±3.9
34.7
±3.6
54.8
±2.5
28.4
±3.7
1012.0
± 1.4
5.3
±0.5
2013
86.6
± 1.7
76.1
± 1.7
36.2
± 1.5
55.6
± 1.1
29.1
± 1.6
1011.5
±0.6
5.4
±0.2
South Phoenix, Arizona - SPAZ
Phoenix Sky
Harbor Intl.
Airport
23183
(33.43, -112.00)
4.5
miles
68°
(ENE)
Sample
Days
(35)
85.5
±6.1
75.2
±5.8
35.1
±4.5
54.6
±3.5
28.9
±5.2
1011.4
±2.1
5.4
±0.6
2013
86.6
± 1.7
76.1
± 1.7
36.2
± 1.5
55.6
± 1.1
29.1
± 1.6
1011.5
±0.6
5.4
±0.2
1 Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.

-------
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 2013. 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 were representative of average weather conditions experienced
throughout the year. The greatest difference between the sample day and full-year averages was
calculated for average dew point temperature for PXSS, although the difference is not
statistically significant.
The number of sample days for each site is provided in Table 6-3. Samples were
collected on a l-in-6 day schedule at PXSS while samples were collected on a l-in-12 day
schedule at SPAZ, yielding roughly half the number of collection events; thus, the number of
observations included in each sample day calculation for SPAZ is less. The difference in the
number of sample days is reflected in the larger confidence intervals for SPAZ (the fewer
observations, generally the larger the confidence intervals).
These sites experienced the warmest temperatures among NMP sites in 2013, based on
both average temperatures and average maximum temperatures for 2013. These sites also
experienced the lowest relative humidity levels among all NMP sites in 2013, based on average
relative humidity for 2013.
6.2.3 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 3.4.2. A wind rose shows the frequency of wind directions using "petals"
positioned around a 16-point compass, and uses different colors to represent wind speeds.
Figure 6-4 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 6-4 also presents three different wind roses for the
PXSS monitoring site. First, a historical wind rose representing 2003 to 2012 wind data is
presented, which shows the predominant surface wind speed and direction over an extended
6-9

-------
period of time. Second, a wind rose representing wind observations for all of 2013 is presented.
Next, a wind rose representing wind observations for days on which samples were collected in
2013 is presented. These can be used to identify the predominant wind speed and direction for
2013 and to determine if wind observations on sample days were representative of conditions
experienced over the entire year and historically. Figure 6-5 presents the distance map and three
wind roses for SPAZ.
Observations from Figures 6-4 and 6-5 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 7.5 miles southeast of PXSS and 4.5 miles east-northeast of SPAZ.
•	Because the Phoenix Sky Harbor weather station is the closest weather station to both
sites, the historical and 2013 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 19 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 (those less than
or equal to 2 knots) account for 16 percent of the hourly wind measurements from
2003 to 2012.
•	The 2013 wind patterns are similar to the historical wind patterns. Further, the sample
day wind patterns for each site resemble both the historical and 2013 wind patterns,
indicating that wind conditions on sample days were representative of those
experienced over the entire year and historically.
6-10

-------
Figure 6-4. Wind Roses for the Phoenix Sky Harbor International Airport Weather Station
near PXSS
Location of PXSS and Weather Station
2003-2012 Historical Wind Rose
nciM
; NORTH"'-.,
est;
WIND SPEED
(Knots)
~ =22
Esil 17-21
11 17
I I 7- 11
I 4-7
H 2-4
Calms: 16.34%
2013 Wind Rose
Sample Day Wind Rose
NORTH---.
WIN C SPEED
(Knots)
17 - 21
11 - 17
SOUTH
! NORTH

Calms: 16.12%
WIND SPEED
(Knots)
~ >=22
Ml 17-21
[H 11 - 17
I I 7- 11
~1 4-7
2-4
Calms: 17.49%
6-11

-------
Figure 6-5. Wind Roses for the Phoenix Sky Harbor International Airport Weather Station
near SPAZ
Location of SPAZ and Weather Station	2003-2012 Historical Wind Rose
2013 Wind Rose
VEST
WWD SPEED
[Kn ots >
SOUTH
¦ 2-4
Calms: 16.12%
Sample Day Wind Rose
WEST
WIN D S PE ED
(Kn ots)
SOUTH
Calms: 16.79%
6-12

-------
6.3 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 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. VOCs, carbonyl compounds, PAHs, metals (PMio), and hexavalent
chromium were sampled for at PXSS; VOCs were the only pollutants sampled for at SPAZ. Note
that hexavalent chromium sampling was discontinued at PXSS at the end of June 2013.
Table 6-4. Risk-Based Screening Results for the Arizona Monitoring Sites
Pollutant
Screening
Value
Oig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Phoenix, Arizona - PXSS
Benzene
0.13
61
61
100.00
11.01
11.01
Carbon Tetrachloride
0.17
61
61
100.00
11.01
22.02
Formaldehyde
0.077
60
60
100.00
10.83
32.85
Acetaldehyde
0.45
59
60
98.33
10.65
43.50
1.3 -Butadiene
0.03
58
60
96.67
10.47
53.97
Arsenic (PMio)
0.00023
52
61
85.25
9.39
63.36
/?-Dichlorobcnzcnc
0.091
51
58
87.93
9.21
72.56
Naphthalene
0.029
50
58
86.21
9.03
81.59
1,2-Dichloroethane
0.038
38
38
100.00
6.86
88.45
Ethylbenzene
0.4
37
61
60.66
6.68
95.13
Hexachloro-1,3 -butadiene
0.045
8
8
100.00
1.44
96.57
Nickel (PMio)
0.0021
8
61
13.11
1.44
98.01
Manganese
0.03
4
61
6.56
0.72
98.74
Benzo(a)pyrene
0.00057
3
36
8.33
0.54
99.28
Cadmium (PMio)
0.00056
1
61
1.64
0.18
99.46
cis-1,3 -Dichloropropene
0.25
1
4
25.00
0.18
99.64
trans-1,3 -Dichloropropene
0.25
1
2
50.00
0.18
99.82
Hexavalent Chromium
0.000083
1
27
3.70
0.18
100.00
Total
554
838
66.11

6-13

-------
Table 6-4. Risk-Based Screening Results for the Arizona Monitoring Sites (Continued)
Pollutant
Screening
Value
Oig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
South Phoenix, Arizona - SPAZ
Benzene
0.13
31
31
100.00
19.50
19.50
1.3 -Butadiene
0.03
31
31
100.00
19.50
38.99
Carbon Tetrachloride
0.17
31
31
100.00
19.50
58.49
p-Dichlorobenzene
0.091
27
30
90.00
16.98
75.47
Ethylbenzene
0.4
20
31
64.52
12.58
88.05
1,2-Dichloroethane
0.038
19
19
100.00
11.95
100.00
Total
159
173
91.91

Observations from Table 6-4 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.
•	Eighteen pollutants failed at least one screen for PXSS; 66 percent of concentrations
for these 18 pollutants were greater than their associated risk screening value (or
failed screens).
•	Ten pollutants contributed to 95 percent of failed screens for PXSS and therefore
were identified as pollutants of interest for PXSS. These 10 include two carbonyl
compounds, six VOCs, one PMio metal, and one PAH.
•	PXSS failed the second highest number of screens (554) among all NMP sites, behind
only S4MO with 574 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 23 percent. This is due primarily to the relatively high number of
pollutants sampled for at this site, as discussed in Section 4.2.
•	Six pollutants failed screens for SPAZ; approximately 92 percent of concentrations
for these six pollutants were greater than their associated risk screening value (or
failed screens). This percentage is greater than the percentage for PXSS. However,
nearly all of the measured detections for the pollutants listed for SPAZ failed screens;
for PXSS, the percentage of screens failed for each individual pollutant is more
varied.
•	All 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 and carbon tetrachloride were detected
in all valid samples and failed 100 percent of screens for each site. This was also true
for 1,3-butadiene for SPAZ. Other VOCs, such as 1,2-dibromoethane (for both sites)
6-14

-------
and hexachloro-1,3-butadiene (for PXSS), also failed 100 percent of screens, but were
detected less frequently.
•	Formaldehyde also failed 100 percent of screens for PXSS (and was detected in all of
the valid samples collected at this site).
6.4 Concentrations
This section presents various concentration averages used to characterize air toxics
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. Site-specific statistical summaries for all pollutants
sampled for at PXSS and SPAZ are provided in Appendices J, L, M, N, and O.
6.4.1 2013 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 6-5, where applicable. Note that concentrations
6-15

-------
of the PAHs and metals for PXSS are presented in ng/m3 for ease of viewing. Also note that if a
pollutant was not detected in a given calendar quarter, the quarterly average simply reflects "0"
because only zeros substituted for non-detects were factored into the quarterly average
concentration.
Table 6-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Arizona Monitoring Sites

# of






Measured
1st
2nd
3rd
4th


Detections
Quarter
Quarter
Quarter
Quarter
Annual

vs. # of
Average
Average
Average
Average
Average
Pollutant
Samples
(Hg/m3)
Oig/m3)
(Hg/m3)
(Hg/m3)
(Hg/m3)
Phoenix, Arizona - PXSS


2.71
2.51
2.24
3.57
2.78
Acetaldehyde
60/60
±0.74
±0.42
±0.48
±0.48
±0.29


1.42
0.56
0.55
1.66
1.06
Benzene
61/61
±0.38
±0.13
±0.14
±0.30
±0.18


0.28
0.08
0.07
0.40
0.21
1.3 -Butadiene
60/61
±0.10
±0.03
±0.02
±0.09
±0.05


0.65
0.61
0.63
0.58
0.62
Carbon Tetrachloride
61/61
±0.04
±0.04
±0.02
±0.02
±0.02


0.22
0.15
0.11
0.33
0.20
/?-Dichlorobcnzcnc
58/61
±0.07
±0.04
±0.04
±0.06
±0.03


0.09
0.09
0.03
0.05
0.07
1,2-Dichloroethane
38/61
±0.03
±0.02
±0.02
±0.03
±0.01


0.78
0.39
0.41
1.05
0.67
Ethylbenzene
61/61
±0.25
±0.09
±0.13
±0.19
±0.11


3.59
3.87
3.90
4.20
3.89
Formaldehyde
60/60
±0.61
±0.37
±0.28
±0.48
±0.22


0.57
0.31
0.51
0.54
0.49
Arsenic (PMi0)a
61/61
±0.18
±0.10
±0.18
±0.15
±0.08


115.56
51.89
41.75
157.35
93.36
Naphthalene3
58/58
±38.99
± 14.25
± 13.28
± 37.72
± 18.63
South Phoenix, Arizona - SPAZ


1.23
0.68
0.69
1.65
1.07
Benzene
31/31
±0.53
±0.12
±0.19
±0.33
±0.21


0.24
0.11
0.12
0.40
0.22
1.3 -Butadiene
31/31
±0.16
±0.02
±0.08
±0.13
±0.07


0.58
0.65
0.63
0.57
0.61
Carbon Tetrachloride
31/31
±0.06
±0.05
±0.04
±0.03
±0.02


0.17
0.16
0.18
0.35
0.22
/?-Dichlorobcnzcnc
30/31
±0.10
±0.03
±0.05
±0.09
±0.04


0.10
0.08
0.03
0.04
0.06
1,2-Dichloroethane
19/31
±0.02
±0.05
±0.03
±0.04
±0.02


0.70
0.50
0.45
1.05
0.68
Ethylbenzene
31/31
±0.37
±0.08
±0.17
±0.37
±0.15
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
6-16

-------
Observations for PXSS from Table 6-5 include the following:
•	The pollutants of interest with the highest annual average concentrations are
formaldehyde (3.89 ± 0.22 |ig/m3), acetaldehyde (2.78 ± 0.29 |ig/m3), and benzene
(1.06 ±0.18 |ig/m3). These are the only pollutants of interest with annual average
concentrations greater than 1 |ig/m3 for this site.
•	The first and fourth quarter average concentrations for benzene and 1,3-butadiene are
significantly greater than the second and third quarter average concentrations,
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 /;-dichlorobenzene and ethylbenzene exhibit a similar tendency but the
differences among the quarterly averages are not statistically significant.
•	The fourth quarter average concentrations of many of PXSS's pollutants of interest
(including acetaldehyde, formaldehyde, benzene, and naphthalene) are higher than the
other quarterly averages, yet the confidence intervals are highest for many of the first
quarter averages. A review of the data shows that the highest concentrations of
benzene, acetaldehyde, and formaldehyde were measured on January 22, 2013, while
many of the other higher concentrations were measured during the fourth quarter.
Several of PXSS's pollutant of interest were highest on December 18, 2013, including
1,3-butadiene, ethylbenzene, and naphthalene (and the second highest concentrations
of benzene and acetaldehyde were also measured on this date).
•	1,2-Dichloroethane concentrations appear highest during the first half of 2013. This
pollutant was detected in roughly 60 percent of the VOC samples collected. This
pollutant was detected in 25 of the 30 VOC samples collected from January through
June but only 13 of the 31 samples collected during the second half of the year.
•	Arsenic is the only metal pollutant of interest for PXSS. The second quarter average
concentration of arsenic (0.31 ± 0.10 ng/m3) is considerably less than the other
quarterly averages (all greater than 0.50 ng/m3). The second quarter is the only
quarter in which an arsenic concentration greater than 1 ng/m3 was not measured (the
maximum for the second quarter is 0.77 ng/m3).
•	Based on the quarterly averages shown in Table 6-5, measurements of naphthalene
measured at PXSS are highly variable. Concentrations span an order of magnitude,
ranging from 18.9 ng/m3 to 287 ng/m3. The first and fourth quarterly average
concentrations are significantly greater than the other quarterly average, similar to
several of the VOCs. All but two of the 22 concentrations of naphthalene greater than
100 ng/m3 were measured in the first and fourth quarters of 2013.
Observations for SPAZ from Table 6-5 include the following:
•	The pollutant of interest with the highest annual average concentration for SPAZ is
benzene (1.07 ± 0.21 |ig/m3), which is the only pollutant of interest with an annual
average concentration greater than 1 |ig/m3. The annual average concentration of
benzene for SPAZ is similar to the annual average benzene concentration of PXSS.
6-17

-------
•	Similar to PXSS, benzene and 1,3-butadiene concentrations were highest during the
first and fourth quarters of 2013 at SPAZ. This is also true for ethylbenzene.
However, the confidence intervals calculated for these averages indicate that the
concentrations included in the quarterly averages are variable, particularly for the first
quarter.
•	The fourth quarter average concentration of p-dichlorobenzene is roughly twice the
other quarterly averages for this pollutant. A review of the data shows that the only
two concentrations greater than 0.4 |ig/m3 were measured on October 19, 2013
(0.49 |ig/m3) and November 12, 2013 (0.47 |ig/m3). Further, seven of the nine
p-dichlorobenzene concentrations greater than 0.3 |ig/m3 were measured at SPAZ
during the fourth quarter.
•	Similar to PXSS, most of the measured detections of 1,2-dichloroethane were
measured at SPAZ during the first half of 2013. During the first half of the year,
1.2-dichloroethane	was detected in 13 of 15 valid samples; during the second half of
the year, this pollutant was detected in only six of the 16 valid samples collected.
Tables 4-9 through 4-12 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for PXSS and
SPAZ from those tables include the following:
•	PXSS and SPAZ appear in Tables 4-9 through 4-12 a total of 12 times.
•	SPAZ and PXSS have the highest annual average concentrations of
p-dichlorobenzene among all NMP sites sampling VOCs, similar to 2011 and 2012.
These annual average concentrations of />dichlorobenzene are roughly twice the next
highest concentration shown in Table 4-9. Of the 38 highest />dichlorobenzene
concentrations measured across the program (those greater than 0.25 |ig/m3), these
two sites account for 29 of them (23 for PXSS and 13 for SPAZ). By comparison, the
next highest site had three (S4MO).
•	SPAZ and PXSS also has the third and fourth highest annual average concentrations
of ethylbenzene, the fifth and sixth highest annual average concentrations of
1.3-butadiene;	and the eighth and tenth highest annual average concentrations of
benzene among NMP sites sampling these pollutants.
•	PXSS has the highest fourth annual average concentration of acetaldehyde and the
fifth highest annual average concentration of formaldehyde among NMP sites
sampling carbonyl compounds.
•	The annual average concentration of naphthalene for PXSS ranks seventh among
NMP sites sampling PAHs.
•	Even though arsenic is a pollutant of interest for PXSS, this site does not appear in
Table 4-12 for arsenic (it ranked 11th). However, PXSS ranks third highest for nickel
among NMP sites sampling PMio metals, similar to 2012.
6-18

-------
6.4.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for the pollutants shaded in
gray in Table 6-4 for PXSS and SPAZ. Figures 6-6 through 6-15 overlay the sites' minimum,
annual average, and maximum concentrations onto the program-level minimum, first quartile,
median, average, third quartile, and maximum concentrations, as described in Section 3.4.3.1.
Figure 6-6. Program vs. Site-Specific Average Acetaldehyde Concentration
0
3
6 9
Concentration {[jg/m3)

12
15

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i


Site: Site Average
o
Site Concentration Range



Figure 6-7. Program vs. Site-Specific Average Arsenic (PMio) Concentration
0
12 3
4 5 6
Concentration {ng/m3)
7
8

10

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i



Site: Site Average
o
Site Concentration Range




6-19

-------
Figure 6-8. Program vs. Site-Specific Average Benzene Concentrations
¦+>
Program Max Concentration = 43.5 ^ig/m3

•

-o	¦
Program Max Concentration = 43.5 jig/m3



0	2	4	6	8	10	12
Concentration {[ig/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


Figure 6-9. Program vs. Site-Specific Average 1,3-Butadiene Concentrations


¦
Ok ,
Program Max Concentration = 21.5 ^ig/m3

1
U 1




1
o ¦

Program Max Concentration = 21.5 ^ig/m3

1
U 1


0	0.3	0.6	0.9	1.2	1.5
Concentration {[ig/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


6-20

-------
Figure 6-10. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations
Program Max Concentration = 23.7 ^ig/m3
Program Max Concentration = 23.7 ^ig/m3
0.75	1
Concentration {[jg/m3]
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


Figure 6-11. Program vs. Site-Specific Average/>-Dichlorobenzene Concentrations
0.1
0.2
0.3 0.4
Concentration {[jg/m3)
0.5
0.6
Program:
IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i
Site:
Site Average
o
Site Concentration Range


6-21

-------
Figure 6-12. Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations
PXSS


¦
Program Max Concentration = 111 ^ig/m3





—
Program Max Concentration = 111 jig/m3


H	1	1	r
0	0.2	0.4	0.6	0.8	1
Concentration {[ig/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


Figure 6-13. Program vs. Site-Specific Average Ethylbenzene Concentrations



-
	0	¦
Program Max Concentration = 18.7 ^ig/m3
















Program Max Concentration = 18.7 ^ig/m3
,










H	1	1	1	r
0	1	2	3	4	5	6
Concentration {[ig/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


Figure 6-14. Program vs. Site-Specific Average Formaldehyde Concentration
0
3 6
9 12 15
Concentration (pg/m3)
18
21
24

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i


Site: Site Average
o
Site Concentration Range



6-22

-------
Figure 6-15. Program vs. Site-Specific Average Naphthalene Concentration
i
0
100
200
300 400 500
Concentration {ng/m3)
600
700
800

Program:
Site:
1st Quartile
¦
Site Average
o
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
Site Concentration Range
4th Quartile
~
Average
i

Observations from Figures 6-6 through 6-15 include the following:
•	Figure 6-6 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 fourth highest annual average concentration among NMP sites sampling this
pollutant, although the maximum acetaldehyde concentration measured at PXSS
is considerably less than the maximum concentration measured across the
program.
•	Figure 6-7 shows that the annual average arsenic (PMio) concentration for PXSS
is less than the program-level average concentration for arsenic (PMio) and is
similar to the program-level median concentration. Arsenic concentrations
measured at PXSS range from 0.03 ng/m3 to 1.66 ng/m3.
•	The program-level maximum benzene concentration (43.5 |ig/m3) is not shown
directly on the box plots in Figure 6-8 because the scale of the box plots would be
too large to readily observe data points at the lower end of the concentration
range. Thus, the scale of the box plots has been reduced to 12 |ig/m3. Figure 6-8
for benzene shows both Arizona sites, as both SPAZ and PXSS sampled VOCs.
Although the maximum benzene concentration measured at each Arizona site is
considerably less than the maximum benzene concentration measured across the
program, both sites' annual averages are greater than the program-level average
concentration. The annual average benzene concentration for these sites are very
similar to each other, although the range of measurements is greater for PXSS.
The minimum benzene concentration measured at SPAZ is just less than the
program-level first quartile.
•	Similar to benzene, the program-level maximum 1,3-butadiene concentration
(21.5 |ig/m3) is not shown directly on the box plots in Figure 6-9 because the
scale of the box plots would be too large to readily observe data points at the
lower end of the concentration range. Thus, the scale of the box plots has been
reduced to 1.5 |ig/m3. Figure 6-9 for 1,3-butadiene also shows both sites. The
range of 1,3-butadiene measurements is greater for PXSS than SPAZ, although
the annual average concentrations are similar to each other. The annual average
6-23

-------
concentrations for both sites are slightly greater than the program-level average
concentration. The minimum 1,3-butadiene concentration measured at SPAZ is
greater than the program-level first quartile. A single non-detect was measured at
PXSS.
Figure 6-10 presents the box plots for carbon tetrachloride for both sites. The
scale of the box plots in Figure 6-10 has also been reduced to allow for the
observation of data points at the lower end of the concentration range, as the
program-level maximum carbon tetrachloride concentration (23.75 |ig/m3) is
considerably greater than the majority of measurements. Figure 6-10 shows that
the annual average concentrations of carbon tetrachloride for both Arizona sites
are less than the program-level average and median concentrations. The range of
concentrations measured at SPAZ is less than the range for PXSS.
Figure 6-11 presents the box plots for /;-dichlorobenzene for both sites. Note that
the program-level first and second quartiles are both zero and therefore not visible
on the box plots. SPAZ and PXSS have the highest annual average concentrations
of />dichlorobenzene among NMP sites sampling VOCs. The annual averages for
SPAZ and PXSS are nearly five times the program-level average concentration.
Although the maximum concentrations measured at these sites are less than the
maximum concentration measured across the program, these two sites share the
second highest concentrations measured across the program (0.494 |ig/m3, both of
which were measured on the same day, October 19, 2013). A single non-detect of
/;-dichlorobenzene was measured at SPAZ while three non-detects were measured
at PXSS.
The scale of the box plots in Figure 6-12 has also been reduced to allow for the
observation of data points at the lower end of the concentration range, as the
program-level maximum 1,2-dichloroethane concentration (111 |ig/m3) is
considerably greater than the majority of measurements. Note that all of the
concentrations of 1,2-dichloroethane measured at PXSS and SPAZ are less than
the program-level average concentration of 0.26 |ig/m3, which is being driven by
the measurements at the upper end of the concentration range. The annual average
concentrations for PXSS and SPAZ are similar to the program-level first quartile
of 0.065 |ig/m3.
The scale of the box plots in Figure 6-13 has also been reduced to allow for the
observation of data points at the lower end of the concentration range, as the
program-level maximum ethylbenzene concentration (18.7 |ig/m3) is considerably
greater than the majority of measurements. Figure 6-13 show that the range and
annual average concentrations of ethylbenzene for SPAZ and PXSS are similar to
each other. The annual average concentrations of ethylbenzene for the two
Arizona sites are roughly twice the program-level averages; recall from the
previous section that these sites have the third and fourth highest annual average
concentrations of ethylbenzene among NMP sites sampling this pollutant. The
minimum ethylbenzene concentrations measured at PXSS and SPAZ are both
greater than the program-level first quartile.
6-24

-------
•	Figure 6-14 is the box plot for formaldehyde for PXSS. This figure shows that the
range of formaldehyde concentrations measured at PXSS falls within a relatively
small range (1.76 |ig/m3 to 5.41 |ig/m3) compared to the range of concentrations
measured across the program. However, 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 fifth highest annual
average concentration of formaldehyde among NMP sites sampling carbonyl
compounds.
•	Figure 6-15 is the box plot for naphthalene for PXSS. Figure 6-15 shows that the
annual average naphthalene concentration for PXSS is just less than 100 ng/m3,
which is greater than the program-level average concentration (75.26 ng/m3) and
similar to the program-level third quartile (94.65 ng/m3). However, the maximum
naphthalene concentration measured at PXSS (287 ng/m3) is considerably less
than the maximum concentration measured at the program-level (748 ng/m3).
6.4.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.3.2.
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 6-16 through 6-31 present the 1-year statistical metrics for each of the
pollutants of interest first for PXSS, then for SPAZ. The statistical metrics presented for
assessing trends include the substitution of zeros for non-detects. If sampling began mid-year, a
minimum of 6 months of sampling is required for inclusion in the trends analysis; in these cases,
a 1-year average concentration is not provided, although the range and percentiles are still
presented.
6-25

-------
Figure 6-16. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at PXSS
20071	2008	2009	20102	20112	2012	2013
0 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 July 2007.
2 Some statistical metrics are not presented because data from Feb 2010 to March 2011 was invalidated.
Observations from Figure 6-16 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 not changed much over the years. The median
concentrations have varied from 2.23 |ig/m3 (2011) to 3.24 |ig/m3 (2007).
•	The minimum concentration has decreased slightly every year, but considerably so
for 2013. This minimum concentration (0.20 |ig/m3) was measured on July 21, 2013
6-26

-------
and is oddly low for PXSS. This is not reflected in the formaldehyde concentrations
measured in this sample.
Figure 6-17. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at PXSS
2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	- Minimum	~ Median	- Maximum	O 95th Percentile	Average
Observations from Figure 6-17 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.
•	After several years of a slight decreasing trend, the 1-year average concentration
increased significantly from 2010 to 2011, after which a decreasing trend resumed.
The 1-year average concentration for 2013 (0.49 ng/m3) is the lowest one shown in
Figure 6-17.
•	The range of arsenic concentrations measured at PXSS is the smallest for 2013, with
less than 2 ng/m3 separating the minimum and maximum concentration measured.
Less than 1 ng/m3 separates the 5th and 95th percentiles for 2013. The difference
between the 1-year average and median concentrations for 2013 is less than
0.1 ng/m3. All of these metrics indicate a reduced level of variability in the arsenic
concentrations measured in 2013 compared to previous years of sampling.
6-27

-------
Figure 6-18. Yearly Statistical Metrics for Benzene Concentrations Measured at PXSS
20071	2008	2009	2010	2011	2012	2013
Year
0 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 July 2007.
Observations from Figure 6-18 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).
•	The 15 highest benzene concentrations (those greater than 3.5 |ig/m3) were all
measured in the first or fourth quarter of any given year. Further, of the 99 benzene
concentrations greater than 2 |ig/m3, all but 10 were measured during the first or
fourth quarters of a given year; those other 10 were all measured in either April or
September, or just outside the first or fourth quarters.
•	The median concentration increased significantly from 2008 to 2009 and is greater
than the 1-year average concentration for 2009. A review of the data shows that the
number of concentrations greater than 2 |ig/m3 increased from 15 to 24 from 2008 to
2009. For 2010, the number of benzene concentrations greater than 2 |ig/m3 decreased
to 12.
6-28

-------
• After the increase from 2008 to 2009, the 1-year average benzene concentration has a
decreasing trend, with the largest change from 2009 to 2010. The 1-year average
concentration for 2013 is the minimum average concentration shown in Figure 6-18.
This is also true for the median concentration; 2013 is the first year for which the
median concentration of benzene is less than 1 |ig/m3.
Figure 6-19. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at PXSS
0.75
2007 1	2008	2009	2010	2011	2012	2013
O 5th Percentile	- Minimum	~ Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 6-19 for 1,3-butadiene measurements collected at PXSS
include the following:
•	The maximum 1,3-butadiene concentration (1.09 |ig/m3) was measured on
December 11, 2011. The only other concentration greater than 1.0 |ig/m3 was
measured at PXSS on January 1, 2009. All but two of the 103 concentrations greater
than 0.30 |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 two not measured in the first or fourth quarters were
measured in September.
•	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, although the
median concentration for 2013 is the minimum shown (0.13 |ig/m3).
6-29

-------
• There have been nine non-detects of 1,3-butadiene measured at PXSS since the onset
of VOC sampling at PXSS under the NMP. Five of these were measured in 2011,
with one measured in 2007, two measured in 2010, and one measured in 2013. For
2011, the minimum and 5th percentile were both equal to zero. None of the non-
detects of 1,3-butadiene were measured during the first or fourth quarters of the year.
Figure 6-20. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at PXSS
£ 0.75
2007 1	2008	2009	2010	2011	2012	2013
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 6-20 for carbon tetrachloride measurements collected at PXSS
include the following:
•	Seven concentrations of carbon tetrachloride greater than 1.0 |ig/m3 have been
measured at PXSS since the onset of sampling in 2007, with five measured in 2008
and two measured in 2009.
•	For 2007, 2010, 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 concentration exhibits a decreasing trend through 2011. Although
the range of concentrations measured decreased for 2012, an increase is shown for the
1-year average and median concentrations for 2012. This is mostly a result of a
6-30

-------
change at the lower end of the concentration range. The number of concentrations less
than 0.6 |ig/m3 in 2011 was 24; the number of concentrations less than 0.6 |ig/m3 in
2012 was six. In addition, the maximum concentration measured is the same for both
years yet the 95th percentile exhibits a decrease from 2011 to 2012.
• All of the statistical parameters for carbon tetrachloride exhibit a decrease for 2013.
Figure 6-21. Yearly Statistical Metrics for/>-Dichlorobenzene Concentrations Measured at PXSS
20071	2008	2009	2010	2011	2012	2013
O 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 July 2007.
Observations from Figure 6-21 for p-dichlorobenzene measurements collected at PXSS
include the following:
•	The three highest concentrations of p-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. One non-detect was measured in
2012 and three were measured in 2013.
6-31

-------
• 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. This is also true for 2013.
Figure 6-22. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at PXSS
2007 1	2008	2009	2010	2011	2012	2013
O 5th Percentile	— Minimum	— Median	— Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 6-22 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, 47 in 2012, and 38 in
2013.
•	The median concentration is zero for all years except 2012 and 2013, indicating that
at least 50 percent of the measurements were non-detects for the first 5 years of
sampling.
•	As the number of measured detections increase, so do the corresponding statistical
metrics shown in Figure 6-22.
•	The number of measured detections increased dramatically for 2012, and the median
and 1-year average concentrations increased correspondingly. The median
concentration is greater than the 1-year average for both 2012 and 2013. This is
because there were still many non-detects (or zeros) factoring into the 1-year average
6-32

-------
concentration for 2012 (14) and 2013 (23), which drive the 1-year averages down in
the same manner that a maximum or outlier concentration can drive the average up.
Figure 6-23. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at PXSS




T



















—
I






i

nr

L-jJ




_2_

^r




20071	2008	2009	2010	2011	2012	2013
O 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 July 2007.
Observations from Figure 6-23 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 |ig/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 33 highest
concentrations (those greater than 1.40 |ig/m3) were measured between January and
March or October and December of any given year. The one exception was measured
in September.
•	The median ethylbenzene concentration has a decreasing trend through 2009, then
returns to 2008 levels for 2010. All of the statistical parameters shown increased from
2010 to 2011. Nearly twice the number of measurements greater than 1 |ig/m3 were
measured in 2011 (20) than the previous years (11 or less).
6-33

-------
• Although the range of measurements changed little, the 1-year average, median, and
95th percentile decreased from 2011 to 2012. Further decreases are shown for 2013,
except for the median, which increased slightly.
Figure 6-24. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at PXSS
20071	2008	2009	20102	20112	2012	2013
Year
O 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 July 2007.
2 Some statistical metrics are not presented because data from Feb 2010 to March 2011 was invalidated.
Observations from Figure 6-24 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.
6-34

-------
• Only one formaldehyde concentration less than 1 |ig/m3 has been measured at PXSS
(2012) and only 11 less than 2 |ig/m3 have been measured since 2007.
Figure 6-25. Yearly Statistical Metrics for Naphthalene Concentrations Measured at PXSS
v 200
2007 1	2008	2009	2010	2011	2012	2013
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 6-25 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 12 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 are highest for 2009. The median, or midpoint, for
2009	is 107 ng/m3. The median concentrations for the other years are less, ranging
from 62.15 ng/m3 (2013) to 84.1 ng/m3 (2010).
•	The difference between the 5th and 95th percentiles has an increasing trend between
2010	and 2012, indicating that the range of concentrations within which the majority
of concentrations lie increased. This range did not change between 2012 and 2013.
Conversely, the median concentration has a steady decreasing trend during this same
6-35

-------
period. This is mostly a result of an increase in the number of concentrations at both
the lower and higher magnitudes.
Figure 6-26. Yearly Statistical Metrics for Benzene Concentrations Measured at SPAZ
0 -I	1	1	1	1	1	1	
20071	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	- Minimum	~ Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 6-26 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 benzene 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 and 2013). Forty-eight of the
52 benzene concentrations greater than 2 |ig/m3 were measured during the first or
fourth quarters of any given year.
•	After several years of increasing, both the maximum and 95th percentile decreased
considerably for 2012 and again for 2013. The range of concentrations measured is at
a minimum for 2013, spanning less than 2 |ig/m3.
6-36

-------
• The 1-year average concentrations changed little from 2009 through 2011, then
decreased from 2011 to 2012 and again for 2013. The median concentration exhibits
more variability during this time frame.
Figure 6-27. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at SPAZ
20071	2008	2009	2010	2011	2012	2013
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 6-27 for 1,3-butadiene measurements collected at SPAZ
include the following:
•	The only 1,3-butadiene concentration greater than 1 |ig/m3 was measured on
January 27, 2011 (1.29 |ig/m3). Forty-five of the 47 concentrations greater than
0.35 |ig/m3 were measured at SPAZ during the first or fourth quarters of any given
year, similar to the trend seen in PXSS 1,3-butadiene measurements.
•	The maximum concentration and 95th percentile increased each year after 2008
through 2011, while the 5th percentile remained fairly static. This indicates that more
of the measurements collected were at the higher end of the concentration range for
each of these years. For 2012, the maximum concentration and 95th percentiles are
lower; with the maximum concentration for 2012 less than the 95th percentile for
2011. This is also true for 2013, where the maximum concentration is less than the
95th percentile for the preceding year. This is a pattern similar to that exhibited by
benzene in Figure 6-26.
•	The 1-year average concentrations exhibit a slight increasing trend between 2009 and
2011, followed by a return to 2010 levels for 2012 and 2008/2009 levels for 2013.
6-37

-------
However, the 1-year averages vary by less than 0.1 |ig/m3, ranging from 0.22 |ig/m3
(2008, 2009, and 2013) to 0.29 |ig/m3 (2011), and confidence intervals calculated
indicate these changes are not statistically significant.
Figure 6-28. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
SPAZ
20071	2008	2009	2010	2011	2012	2013
O 5th Percentile	- Minimum	~ Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 6-28 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 have
been measured at SPAZ, one in 2009 and one in 2011.
•	For the years 2009 through 2012, 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.
•	With the exception of 2012, the 1-year average exhibits a slight decreasing trend.
However, the differences represent an overall change of less than 0.11 |ig/m3 and,
based on the confidence intervals, are not statistically significant.
6-38

-------
• The range of concentrations measured is at a minimum for 2013, as is the difference
between the 1-year average and median for 2013 (less than 0.01 |ig/m3), indicating
the lowest level of variability in the measurements compared to other years. However,
the difference between the 1-year average and median concentrations is relatively low
for every year, with the difference for 2008 being the largest (0.04 |ig/m3).
Figure 6-29. Yearly Statistical Metrics for /7-Dichlorobenzene Concentrations Measured at SPAZ
.0
.8
.6
.4
.2
.0
2008
2009
2010
2011
2012
2013
Year
O 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 July 2007.
Observations from Figure 6-29 for p-dichlorobenzene measurements collected at SPAZ
include the following:
•	The widest range of p-dichlorobenzene concentrations measured is shown for 2008
(non-detect to 0.90 |ig/m3), while the smallest range is shown for the following year
(0.036 |ig/m3to 0.51 |ig/m3).
•	The 1-year average concentration decreased from 2008 to 2009, increased for 2010,
then decreased slightly each year between 2011 and 2013. However, confidence
intervals calculated for these averages indicate that the changes are not statistically
significant. The median concentrations exhibit larger fluctuations than the 1-year
average concentrations.
6-39

-------
Figure 6-30. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at SPAZ





I T
pL






















0.00 -






	




2007 1	2008	2009	2010	2011	2012	2013
Year
0 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 6-30 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, 26 in 2012, and 19 in
2013.
•	The median concentration is zero for all years until 2012, indicating that at least
50 percent of the measurements were non-detects.
•	As the number of measured detections increase, so do the corresponding central
tendency statistics shown in Figure 6-30.
•	The median concentration is greater than the 1-year average concentration for 2012.
This is because the four non-detects (or zeros) factored into the 1-year average
concentration are pulling the average down in the same manner that a maximum or
outlier concentration can pull the average up.
•	Even though the range of concentrations measured increased for 2013, both the
median concentration and the 1-year average concentration decreased. This is a result
of the increase of the number of non-detects for 2013 (12) compared to 2012.
6-40

-------
Figure 6-31. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at SPAZ
20071	2008	2009	2010	2011	2012	2013
Year
0 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 6-31 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 (four) or 2011 (four).
•	The median concentration is at a maximum for 2007, after which the median
decreases by half. Recall that 2007 includes only half a year's worth of samples. The
downward trend continues through 2009, followed by an increase that continues
through 2011. The median decreases somewhat for 2012 and again in 2013. The
1-year average concentration has a similar pattern, although no 1-year average
concentration is presented for 2007.
•	The only non-detects of ethylbenzene were measured during the first full-years of
sampling at SPAZ.
6-41

-------
6.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at each Arizona monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4
for definitions and explanations regarding the various toxicity factors, time frames, and
calculations associated with these risk-based screenings.
6.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Arizona monitoring sites and where annual average
concentrations could be calculated, risk was examined by calculating cancer risk and noncancer
hazard approximations. These approximations can be used as risk estimates for cancer and
noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers want to shift their air
monitoring priorities. Refer to Section 3.4.3.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 6-6, 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 6-6 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 (50.62 in-a-million), followed by benzene (8.23 in-a-million),
1,3-butadiene (6.34 in-a-million), and acetaldehyde (6.12 in-a-million).
•	Formaldehyde's cancer risk approximation for PXSS is the sixth 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 noncancer health effects are expected from
these individual pollutants. The pollutant with the highest noncancer hazard
approximation for PXSS is formaldehyde (0.40). This noncancer hazard
approximation is the eighth highest noncancer hazard approximation among all site-
specific pollutants of interest.
6-42

-------
Table 6-6. Risk Approximations for the Arizona Monitoring Sites
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Phoenix, Arizona - PXSS
Acetaldehyde
0.0000022
0.009
60/60
2.78
±0.29
6.12
0.31
Benzene
0.0000078
0.03
61/61
1.06
±0.18
8.23
0.04
1.3 -Butadiene
0.00003
0.002
60/61
0.21
±0.05
6.34
0.11
Carbon Tetrachloride
0.000006
0.1
61/61
0.62
±0.02
3.70
0.01
p-Dichlorobenzene
0.000011
0.8
58/61
0.20
±0.03
2.24
<0.01
1,2-Dichloroethane
0.000026
2.4
38/61
0.07
±0.01
1.73
<0.01
Ethylbenzene
0.0000025
1
61/61
0.67
±0.11
1.66
<0.01
Formaldehyde
0.000013
0.0098
60/60
3.89
±0.22
50.62
0.40
Arsenic (PMi0)a
0.0043
0.000015
61/61
0.49
±0.08
2.09
0.03
Naphthalene3
0.000034
0.003
58/58
93.36
± 18.63
3.17
0.03
South Phoenix, Arizona - SPAZ
Benzene
0.0000078
0.03
31/31
1.07
±0.21
8.36
0.04
1.3 -Butadiene
0.00003
0.002
31/31
0.22
±0.07
6.69
0.11
Carbon Tetrachloride
0.000006
0.1
31/31
0.61
±0.02
3.63
0.01
p-Dichlorobenzene
0.000011
0.8
30/31
0.22
±0.04
2.40
<0.01
1,2-Dichloroethane
0.000026
2.4
19/31
0.06
±0.02
1.58
<0.01
Ethylbenzene
0.0000025
1
31/31
0.68
±0.15
1.70
<0.01
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
Observations for SPAZ from Table 6-6 include the following:
•	The pollutants with the highest annual average concentrations for SPAZ are benzene,
ethylbenzene, and carbon tetrachloride. Only benzene has an annual average
concentration greater than 1 |ig/m3,
•	Based on the annual averages and cancer UREs, benzene has the highest cancer risk
approximation for SPAZ (8.36 in-a-million), followed by 1,3-butadiene
(6.69 in-a-million), and carbon tetrachloride (3.63 in-a-million). These cancer risk
6-43

-------
approximations are similar to the approximations calculated for these same pollutants
for PXSS.
• None of the pollutants of interest for SPAZ have noncancer hazard approximations
greater than 1.0, indicating no adverse noncancer health effects are expected from
these individual pollutants. The pollutant with the highest noncancer hazard
approximation for SPAZ is 1,3-butadiene (0.11).
6.5.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 6-7 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
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 3.4.3.4. Lastly,
Table 6-7 provides the 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 the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more
in-depth discussion of this analysis is provided in Section 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 6.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
6-44

-------
Table 6-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Arizona Monitoring Sites
Top 10 Total Emissions for Pollutants with
Cancer UREs
Top 10 Cancer Toxicity-Weighted
Emissions
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(County-Level)
(County-Level)

(Site-Specific)



Cancer

Cancer Risk
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(in-a-million)
Phoenix, Arizona (Maricopa County) - PXSS
Benzene
1,313.94
Formaldehyde
1.48E-02
Formaldehyde
50.62
Formaldehyde
1,141.02
Benzene
1.02E-02
Benzene
8.23
Ethylbenzene
862.37
1,3-Butadiene
5.42E-03
1,3-Butadiene
6.34
Acetaldehyde
576.27
Naphthalene
3.02E-03
Acetaldehyde
6.12
1.3 -Butadiene
180.82
Ethylbenzene
2.16E-03
Carbon Tetrachloride
3.70
Tetrachloroethylene
95.59
POM, Group 2b
1.48E-03
Naphthalene
3.17
Naphthalene
88.77
Acetaldehyde
1.27E-03
p-Dichlorobenzene
2.24
POM, Group 2b
16.83
POM, Group 2d
1.19E-03
Arsenic
2.09
POM, Group 2d
13.53
Arsenic, PM
1.03E-03
1,2-Dichloroethane
1.73
Dichloromethane
12.34
POM, Group 5a
7.15E-04
Ethylbenzene
1.66
South Phoenix, Arizona (Maricopa County) - SPAZ
Benzene
1,313.94
Formaldehyde
1.48E-02
Benzene
8.36
Formaldehyde
1,141.02
Benzene
1.02E-02
1,3-Butadiene
6.69
Ethylbenzene
862.37
1,3-Butadiene
5.42E-03
Carbon Tetrachloride
3.63
Acetaldehyde
576.27
Naphthalene
3.02E-03
p-Dichlorobenzene
2.40
1,3-Butadiene
180.82
Ethylbenzene
2.16E-03
Ethylbenzene
1.70
Tetrachloroethylene
95.59
POM, Group 2b
1.48E-03
1,2-Dichloroethane
1.58
Naphthalene
88.77
Acetaldehyde
1.27E-03


POM, Group 2b
16.83
POM, Group 2d
1.19E-03


POM, Group 2d
13.53
Arsenic, PM
1.03E-03


Dichloromethane
12.34
POM, Group 5a
7.15E-04



-------
Table 6-8. 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
Top 10 Noncancer Toxicity-Weighted
Emissions
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(County-Level)

(County-Level)
(Site-Specific)



Noncancer

Noncancer Hazard
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(HQ)
Phoenix, Arizona (Marico
)a County) - PXSS
Toluene
5,233.19
Acrolein
2,932,324.18
Formaldehyde
0.40
Xylenes
3,296.34
Formaldehyde
116,431.08
Acetaldehyde
0.31
Hexane
2,752.67
1,3-Butadiene
90,410.71
1,3-Butadiene
0.11
Methanol
2,399.14
Acetaldehyde
64,030.43
Benzene
0.04
Benzene
1,313.94
Benzene
43,798.12
Arsenic
0.03
Formaldehyde
1,141.02
Lead, PM
34,426.96
Naphthalene
0.03
Ethylene glycol
880.96
Xylenes
32,963.37
Carbon Tetrachloride
0.01
Ethylbenzene
862.37
Naphthalene
29,589.71
Ethylbenzene
<0.01
Acetaldehyde
576.27
Arsenic, PM
16,021.47
p-Dichlorobenzene
<0.01
Methyl isobutyl ketone
326.41
Propionaldehyde
10,771.78
1,2-Dichloroethane
<0.01
South Phoenix, Arizona (Maricopa County) - SPAZ
Toluene
5,233.19
Acrolein
2,932,324.18
1,3-Butadiene
0.11
Xylenes
3,296.34
Formaldehyde
116,431.08
Benzene
0.04
Hexane
2,752.67
1,3-Butadiene
90,410.71
Carbon Tetrachloride
0.01
Methanol
2,399.14
Acetaldehyde
64,030.43
Ethylbenzene
<0.01
Benzene
1,313.94
Benzene
43,798.12
p-Dichlorobenzene
<0.01
Formaldehyde
1,141.02
Lead, PM
34,426.96
1,2-Dichloroethane
<0.01
Ethylene glycol
880.96
Xylenes
32,963.37


Ethylbenzene
862.37
Naphthalene
29,589.71


Acetaldehyde
576.27
Arsenic, PM
16,021.47


Methyl isobutyl ketone
326.41
Propionaldehyde
10,771.78



-------
Observations from Table 6-7 include the following:
•	Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
cancer UREs in Maricopa County.
•	The pollutants with the highest toxicity-weighted emissions (of the pollutants with
cancer UREs) are formaldehyde, benzene, and 1,3-butadiene.
•	Eight of the highest emitted pollutants in Maricopa County also have the highest
toxicity-weighted emissions.
•	Formaldehyde has the highest cancer risk approximation for PXSS; carbonyl
compounds were not sampled for at SPAZ, thus, a cancer risk approximation is not
available for this pollutant for SPAZ. Formaldehyde has the second highest emissions
and the highest toxicity-weighted emissions for Maricopa County. Acetaldehyde,
which has the fourth highest cancer risk approximation for PXSS, also appears on
both emissions-based list for Maricopa County.
•	Among the VOCs, benzene, 1,3-butadiene, and carbon tetrachloride have highest
cancer risk approximations for PXSS and SPAZ. The cancer risk approximations for
these pollutants are similar between the two sites. While benzene and 1,3-butadiene
both appear among the pollutants with the highest emissions and highest toxicity-
weighted emissions for Maricopa County, carbon tetrachloride does not appear on
either list.
•	Naphthalene is among the highest emitted pollutants (seventh), has one of the highest
toxicity-weighted emissions (fourth), and has one of the highest cancer risk
approximations for PXSS (sixth). POM, Group 2b is the eighth highest emitted
"pollutant" in Maricopa County and ranks sixth for toxicity-weighted emissions.
POM, Group 2b includes several PAHs sampled for at PXSS including acenaphthene,
benzo(e)pyrene, fluoranthene, and perylene. Similarly, POM, Group 2d is the ninth
highest emitted "pollutant" and ranks eighth 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). POM,
Group 5a ranks tenth for toxicity-weighted emissions for Maricopa County. This
POM group includes benzo(a)pyrene, which failed three screens for PXSS but was
not identified as a pollutant of interest for this site.
•	Arsenic has the eighth highest cancer risk approximation among the pollutants of
interest for PXSS. This pollutant ranks ninth for its toxicity-weighted emissions but
does not appear among the highest emitted pollutants in Maricopa County (it ranks
20th).
Observations from Table 6-8 include the following:
•	Toluene, xylenes, and hexane are the highest emitted pollutants with noncancer RfCs
in Maricopa County.
6-47

-------
•	The pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) are acrolein, formaldehyde, and 1,3-butadiene.
•	Four of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Maricopa County.
•	Acrolein has the highest toxicity-weighted emissions for Maricopa County. Although
acrolein was sampled for at both sites, this pollutant was excluded from the pollutants
of interest designation, and thus subsequent risk-based screening evaluations, due to
questions about the consistency and reliability of the measurements, as discussed in
Section 3.2. The emissions for acrolein rank 16th.
•	Formaldehyde and acetaldehyde have the highest noncancer hazard approximations
for PXSS (although considerably less than an HQ of 1.0), both of which appear
among those with the highest emissions and toxicity-weighted emissions for
Maricopa County.
•	1,3-Butadiene and benzene have the highest noncancer hazard approximations among
the VOCs for both PXSS and SPAZ and are similar in magnitude between the two
sites. Benzene ranks fifth for both its emissions and its toxicity-weighted emissions.
1,3-Butadiene has the third highest toxicity-weighted emissions for Maricopa County
but is not one of the highest emitted pollutants in Maricopa County (with a noncancer
RfC), as it ranks 11th.
6.6 Summary of the 2013 Monitoring Data for PXSS and SPAZ
Results from several of the data treatments described in this section include the
following:
~~~ Eighteen pollutants failed screens for PXSS; six 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, particularly benzene and 1,3-butadiene, tended to
be higher during the colder months of the year. This was also reflected in the
concentration data from previous years of sampling.
~~~ SPAZ and PXSS have the highest and second highest annual average concentrations
of p-dichlorobenzene among NMP sites sampling VOCs. These sites also rank third
andfourth highest for ethylbenzene.
~~~ Concentrations of benzene appear to be decreasing at the Arizona sites. The detection
rate of 1,2-dichloroethane increased significantly during the later years of sampling.
Arsenic concentrations have decreased at PXSS the last few years.
6-48

-------
~~~ Formaldehyde has the highest cancer risk approximation of the pollutants of interest
for PXSS; benzene has the highest cancer risk approximation of the pollutants of
interest for SPAZ. None of the pollutants of interest for either site have noncancer
hazard approximations greater than an HQ of 1.0.
6-49

-------
7.0	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.
7.1	Site Characterization
This section characterizes the California monitoring sites by providing geographical and
physical information about the locations of the sites and the surrounding areas. This information
is provided to give the reader insight regarding factors that may influence the air quality near the
sites and assist in the interpretation of the ambient monitoring measurements.
Three monitoring sites are located in southern California cities: Los Angeles, Long
Beach, and Rubidoux. A fourth monitoring site is located in San Jose, which is in northern
California. Figures 7-1 and 7-2 are the composite satellite images retrieved from ArcGIS
Explorer showing the Los Angeles and Long Beach monitoring sites and their immediate
surroundings, respectively. Figure 7-3 identifies nearby point source emissions locations by
source category for each site, as reported in the 2011 NEI for point sources, version 2. Note that
only sources within 10 miles of the sites are included in the facility counts provided in
Figure 7-3. 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 boundary are still visible on the map for reference, but have been
grayed out in order to emphasize emissions sources within the boundary. Figures 7-4 through 7-7
are the composite satellite images and emissions maps for the Rubidoux and San Jose monitoring
sites. Table 7-1 provides supplemental geographical information such as land use, location
setting, and locational coordinates.
7-1

-------
Figure 7-1. Los Angeles, California (CELA) Monitoring Site
' N Ma
(HIP
-j
to

-------
Figure 7-2. Long Beach, California (LBHCA) Monitoring Site

-------
Figure 7-3. NEI Point Sources Located Within 10 Miles of CELA and LBHCA
Orange 1 -
Courvty f
1 t -
tiirnwi
iwwvr
CELA NATTS site	LBHCA CSATAM s»te	10 mile radius	County boundary
Source Category Group (No.
4* AamOUMMttaA W«n4actu*ra1 *»> | M)
T Mpant	r, i Openwcfw »M|
*	«t PTXWWvH* Mm Atprttft PW t!
& Arofeptir	tlr>«**v)T|
7*t *:WiUi«f "-jU WfK*Tr ''
& Auemcwesw nnwitir | Ji
X &**ivM4rut»rt>M>»f*alry ill
it Dt*Jn®C«r*mK*or |«)
0	Pi* ^HH»naa.0uk rAar*» c J11
C ClwmMi Manitfmimo * I6i
[ Cn»Wu.r Sllh>« l»|
- Cmmm*t Ainna*H*t*nill
1) D»v n»*r*fn rum i?i
G [WIMI f u*ir*fl	FHH. i
#	bIKl cn> ^^4U/n..»CuH£u*lk.f.iM,
£ fcte-AnpaArq ftatag Poatttag Ansbfi)
F *w: PftJcawr^'A^KiA^w r «- «ij 
of Facilities)
K OMefo«4iiM«i 8wvc» SUto* til
'MM Pwcifl
> new Mi»»viiy^i (3>
irwM*W«rf U« m«f at t w***1' • t
O imOUUr* kHoui iMib' ctftaan.ttfc i<41)
f2l

¦ WMCarditt iMOyiiwiiCnOAf n+*tacu*r31|
A MtIM CMln) i*D'avMg and Altai £*r»tc*s to W*iuWiUj«h 16)
® Met** »w*Ht"9»«erKHwn Faotfy i-IPi
A M»«ar> B»nUatn*« Sfr a*, rte+f,, >
v,	P>K«%aiq r sM| |4 I
? Mmatantout CenmerwilmuMttJ '*>
@ Uiirte** *f*»m CiMbtMu Mr
•	01 «mrc* G««	i4E'
Q "W MWtCmtre Mb	f cN| Hi
<	fUM ft |
#	KMA4iway r>ot fapf^sant alt f»cUt«s «^llh»n na a
-------
Figure 7-4. Rubidoux, California (RUCA) Monitoring Site
**Rutiidoux
jb * yi' ^ iii
L Sour<%flBi%CSf/i y t* >
I>tc NASA -NGA^VSGS
20 0 & M.^ro*oil^«rtp ^
<1

-------
Figure 7-5. NEI Point Sources Located Within 10 Miles of RUCA
Rlv*ra*Oft
County
Legend
RUCA NATTS site	10 mile radius	County boundary
Source Category Group (No. of Facilities)
*
m^uAkju ««t r*aitf H)
9
CJntMy Ohm* »¦ Cwftuwi i SQ
T
ArsortArfntnfcftiott SucpOft Cpcr*t on* I t?l
*
iNns) Bp^»vnH < »|
A
Antml or f«r*» STi
F
fonts Prcrmvo AQWtr* f k*> f Soma i"9i
¦
UK>'wUn«i W« 'itHon 11)
*
*,itHrnNk« '«*-> U»n, «=Kt.r--«j t i*J
*
frvMmM i v kwdV Mw*ll}
_
AuunvtMS'QV (3|
O
l«*OC< *K*K-*0 y**c* « ||M

&<» 3kuAMa> CWr ta Ctt, Ceanwc* Pum |t|
A

i
Bo o^QiWutiwi 11)
¦
UtolCm. Bo> o*aOV^MraiCon«M»iVUwUtocurn8l«|
B
EU* IbNMMMBall Plartt Hf
A
UMMI CcMlno arid A»«x 9«tM tu MviAklueit i' J
C
C>«(nul ManirtKtir »yj * witty jJJ
<•>
Umk HuOMfclOf MMKtfieil F*MCV 1 t?J
i
Can»«M«- Station 11)
A
WM«> B*s*N«ie*i# 5»cuiV ^«eMy i1>
e
L'«m» | u;cm(» KMn.twa. «m| 'K**V
H
WnaiQuWfMwM P«nj F«*t* |0|
»«•>
O*	(7:
P»rn »'C CiM*x	r •.»•»» (4)
Ftttik. Knr II *uob« U-*tuaCfe FS*« <*1
lS%wd CarMMt	f2i


E	Pwuli
X	*»« ra>rtP»» l«» OfwtwoA* jj|
V	IWIHl«
•	iVi-UPVMtrr Treat-wri f»o*v i J
f	*#*r TiMdmreFmtft  ••
7-6

-------
Figure 7-6. San Jose, California (SJJCA) Monitoring Site

-------
Figure 7-7. NEI Point Sources Located Within 10 Miles of SJJCA
i	i
I At>T»d» i
,	County f
31
Santa Oan
County
f
M CfUZ
County
i
Legend
~ SJJCA NATTS site
¦	b#0MI
Not# Du* to	«a«3 co«i&e«*©n tu total ficifrfi
Jisptayad n-«, nal represent all facitfcas *»th»n 4i» a»ra a< niciast
10 mile radius
County boundary
Source Category Group (No. of Facilities)
•Jl ¦» « ¦muni HnBicMT^trrw lit
^^.-oertAmi^Anvnnaimeri :<«ar«»o*nn»'
4	We	R*r# |3|
0 A<*> 9rair	S*n«e i?fi
*Ufc»r«Oriaf f>*«J	U («.*», | J|
«v~ttV ir C»i#T»ar!*s«
0	SkeamedrwUfCteff > -
B B«*	Pwna |T|
C	Nrt% Itf)
1	Uf^ww SUM* I?)
•Ovfioty) &%m»VMran •( i
C*»C»e«re»|F#r»YHlOt
0 #lermr»l f	jJTR i
f EMTOH CMMIfi M CMkMwMl
E DMUKfcWH »««»na Afinsrvtg an*C**M0H1|
^	NcrtttmMMh
¦	Qw|iere«eVn>«»niiMiiin
V	Qtotartirt M i
"• 'rwuuMel	6t	Pwit ,2J-j
O	'otpta» 0wuA ea • t S«i
A Lanffti
V	U**#' «»0 l«#««a Htiflurs f . •
¦	Mtnican P«» «^r.tt»»M»n»c=«»ttr»»wa*»»urMt^|i*i
Q •' 
^ y»J©*«im ftoatds «»fi
R K«r wH'XuerPTWusFiMiMdt
V °<*» toi W»T«» C»W>Wi I I
^	CaeanBtO^eT^tfralyvtrecilBf |4|
H t»rana>" pwr< :
X <*«'	tn» Diatalone |Ji
A	Cnprn*,, «UMiv.>«>WAM»^. n	f «.>, #\ I
• Ml
0 MnUb hiKuwVitiiUCfl r«_s», ,3«.
A Mmry	ml ten** PaMN|M t
A f^»*i»*<*iPw«»»ntricai( tin
0	,vm»Vwn»i f+mQme, »> tu-fl f»*y 11
? MiM«trwtMCximmn:M»lneMMiair vJt^!797
n(<)

^>•*19: aW1****** |2|
T «tUe •••
M ttoauo u«rbtoritrtrq .*>
«"¦ utxitu 
-------
Table 7-1. Geographical Information for the California Monitoring Sites
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Additional Ambient Monitoring Information1
CELA
06-037-1103
Los
Angeles
Los
Angeles
Los Angeles-Long
Beach-Anaheim, CA
34.06659,
-118.22688
Residential
Urban/City
Center
TSP, TSP Speciation, CO, S02, NO, N02, NOx, NOy,
PAMS, Carbonyl compounds, VOCs, SNMOC, O3,
Meteorological parameters, PM10, PM10 Speciation,
PM Coarse, PM2.5. PM2.5 Speciation, IMPROVE
Speciation, Methane.
LBHCA
06-037-4002
Long
Beach
Los
Angeles
Los Angeles-Long
Beach-Anaheim, CA
33.82376,
-118.18921
Residential
Suburban
TSP, TSP Speciation, CO, S02, NO, N02, NOx,
VOCs, Carbonyl compounds, O3, Meteorological
parameters, PM10, PM10 Speciation, PM2.5, PM2.5
Speciation
RUCA
06-065-8001
Rubidoux
Riverside
Riverside-San
B ernardino -Ontario,
CA
33.99958,
-117.41601
Residential
Suburban
Haze, TSP, TSP Speciation, CO, SO2, NO, NO2,
NOx, NOy, PAMS, VOCs, SNMOC, Carbonyl
compounds, O3, Meteorological parameters, PM10,
PM10 Speciation, PM Coarse, PM2 5,
PM2.5 Speciation, IMPROVE Speciation.
SJJCA
06-085-0005
San Jose
Santa
Clara
San Jose-Sunnyvale-
Santa Clara, CA
37.348497,
121.894898
Commercial
Urban/City
Center
TSP Speciation, CO, S02, NO, N02, NOx, NOy,
VOCs, Carbonyl compounds, O3, Meteorological
parameters, PM10, PM Coarse, PM2.5, PM2.5
Speciation, IMRPOVE Speciation.
1 Data for additional pollutants are reported to AQS for these sites (EPA, 2014b); however, these data are not generated by ERG and are therefore not included in this report.
BOLD ITALICS = EPA-designated NATTS Site

-------
CELA is located on the rooftop of a two-story building northeast of downtown Los
Angeles, just southeast of Dodgers' Stadium and Los Angeles State Historic Park, which are
prominent features in Figure 7-1. CELA is surrounded by major freeways, including 1-5 and
Route 110. Highway 101 is located farther south. Although the area is classified as residential, a
freight yard is located to the south of the site. The Los Angeles River runs north-south just east
of the site. This monitoring site was originally set up as an emergency response monitoring site.
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 Boulevard, as shown in Figure 7-2. Interstate-405 is
located approximately one-fifth of a mile from LBHCA and intersects with 1-710 just 1 mile
west of the site. This monitoring site is located approximately 4 miles north of the shores of
Long Beach and the Port of Long Beach, the second-busiest port in the U.S. (POLB, 2015).
LBHCA is nearly 17 miles south of CELA. Figure 7-3 shows that 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. A number of emissions sources are located immediately
around CELA, with a high density cluster of emissions sources located just to the west and
southwest 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 north of Riverside, in a residential area in the suburban town of
Rubidoux. Figure 7-4 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
7-10

-------
elementary school south of Riverview Drive, and a high school to the west of Pacific Avenue,
the football and baseball fields of which are prominent features in Figure 7-4. 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 44 miles west-southwest of CELA
and 45 miles northwest of LBHCA. Figure 7-5 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 in San Bernardino County. The point source located closest to RUCA is
Flabob Airport. Although the emissions source categories are varied, the emissions source
categories with the greatest number of sources within 10 miles of RUCA include airport
operations; metals processing and fabrication; auto 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 7-6 shows that SJJCA is located in a
commercial area surrounded by residential areas. A railroad is shown east of the monitoring site,
running north-south in Figure 7-6. 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 7-6. Figure 7-7
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
surrounding SJJCA 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 7-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the California 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 Los Angeles, Riverside, and
Santa Clara Counties.
7-11

-------
Table 7-2. Population, Motor Vehicle, and Traffic Information for the California
Monitoring Sites
Site
County
Estimated
County
Population1
County-level
Vehicle
Registration2
Annual
Average
Daily
Traffic3
Intersection
Used for
Traffic Data
County-
level Daily
VMT4 "
CELA
Los Angeles
10,017,068
7,609,517
231,000
1-5 between Main St. and
Broadway (exit 136 and 137)
214,482,440
LBHCA
285,000
1-405 between Wardlow Rd and
1-710 (exit 30 and 32)
RUCA
Riverside
2,292,507
1,788,322
150,000
60 (Mission Blvd) between
Rubidoux Blvd and Valley Way
55,336,730
SJJCA
Santa Clara
1,862,041
1,575,973
115,000
87 (Guadalupe Pkwy) between
Julian St and W Taylor St
41,478,310
bounty-level population estimates reflect 2013 data (Census Bureau, 2014)
2County-level vehicle registrations reflects 2013 data (CA DMV, 2013)
3AADT reflects 2013 data (CA DOT, 2013a)
4County-level VMT reflects 2012 data (CA DOT, 2013b)
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 7-2 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. The county-level
population for Los Angeles County is twice the population for the next highest county
(Cook County, IL) and the county-level vehicle registration for Los Angeles County
is twice the registration for the next highest county (Maricopa County, AZ).
•	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, in addition to ELNJ, are the only ones
with traffic volumes greater than 200,000. 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 12th compared to other
NMP sites.
•	Los Angeles County's daily VMT is the highest among all counties with NMP sites,
and is nearly double 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, ranking fifth and seventh, respectively.
7.2 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.
7-12

-------
7.2.1	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.
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,
2014).
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).
7.2.2	Meteorological Summary
Hourly meteorological data for 2013 were retrieved from NCDC for the weather stations
closest to the California monitoring sites (NCDC, 2013), as described in Section 3.4.2. 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 7-3. These data were used to determine
how meteorological conditions on sample days vary from conditions experienced throughout the
year.
7-13

-------
Table 7-3. 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
(34.02, -118.29)
4.7
miles
Sample
Days
(62)
74.9
±2.0
64.3
± 1.7
48.3
±3.0
56.0
± 1.9
60.6
±3.5
1015.2
± 1.1
1.2
±0.2
231°
(SW)
2013
74.3
±0.8
64.0
±0.7
48.0
± 1.2
55.8
±0.7
61.1
± 1.6
1014.9
±0.4
1.1
±0.1
Long Beach, California - LBHCA
Long
Beach/Daugherty
Field Airport
23129
(33.82, -118.15)
2.6
miles
Sample
Days
(35)
71.8
±2.4
62.7
±2.2
48.5
±3.5
55.3
±2.4
63.5
±4.2
1015.5
± 1.3
4.5
±0.6
109°
(ESE)
2013
73.9
±0.9
64.0
±0.7
48.7
± 1.1
56.0
±0.7
62.1
± 1.4
1015.1
±0.4
4.0
±0.2
Rubidoux, California - RUCA
Riverside Municipal
Airport
3.5
miles
Sample
Days
(61)
80.3
±2.9
66.0
±2.5
43.7
±3.5
54.6
±2.1
51.2
±4.1
1014.2
± 1.1
3.5
±0.3
03171
(33.95, -117.44)
202°
(SSW)
2013
79.7
± 1.3
66.0
± 1.0
43.2
± 1.4
54.5
±0.9
51.1
± 1.8
1013.9
±0.4
3.7
±0.1
San Jose, California - SJJCA
San Jose Intl.
Airport
1.8
miles
Sample
Days
(66)
71.2
±2.3
59.1
±2.0
45.1
±2.0
51.9
± 1.8
63.3
±2.2
1017.3
± 1.2
4.9
±0.6
23293
(37.36, -121.92)
295°
(WNW)
2013
70.3
± 1.0
58.8
±0.9
45.6
±0.9
52.0
±0.7
65.4
± 1.1
1017.0
±0.5
5.0
±0.2
1 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 2013. 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 CELA, RUCA, and SJJCA were representative of average
weather conditions experienced throughout the year. The largest difference in the table for these
sites is for average relative humidity for SJJCA, but is still less than 2 percent different.
The differences between the average meteorological conditions for 2013 and those
experienced on sample days near LBHCA are greater than the other sites, particularly for
temperature. A 1-year sampling effort at LBHCA was completed at the end of July 2013;
therefore, the sample day averages for this site include only data for the first half of 2013.
However, the differences between the full-year averages and the sample day averages are still
relatively small, with the largest difference calculated for average maximum temperature.
Table 7-3 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 in 2013 among all
NMP sites. As expected, conditions in 2013 were cooler near SJJCA than near the other sites.
For the southern California sites, average temperatures in 2013 tended to be slightly higher for
RUCA, which is farther inland than the other two sites.
7.2.3 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 3.4.2. A wind
rose shows the frequency of wind directions using "petals" positioned around a 16-point
compass, and uses different colors to represent wind speeds.
Figure 7-8 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 7-8 also presents three different wind roses for the
7-15

-------
CELA monitoring site. First, a historical wind rose representing 2003 to 2012 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 2013 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2013 is
presented. These can be used to identify the predominant wind speed and direction for 2013 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically. Figures 7-9 through 7-11 present the distance maps and
wind roses for LBHCA, RUCA, and SJJCA, respectively.
Observations from Figure 7-8 for CELA include the following:
•	The weather station at the Downtown Los Angeles/USC Campus is located 4.7 miles
southwest of CELA.
•	Historically, winds were generally light near this site, with calm winds (those less
than or equal to 2 knots) observed for more than 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 17 knots were
not measured at this weather station during this time frame.
•	The 2013 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 2013
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 2013 and on sample days were representative of
those experienced historically.
7-16

-------
Figure 7-8. Wind Roses for the Downtown Los Angeles/USC Campus Weather Station near
CELA
Location of CELA and Weather Station	2003-2012 Historical Wind Rose
2013 Wind Rose
6%s
4%.
2%	!
WIND SPEED
(Knots)
~
[IB 17-21
~ 4-7
H 2-4
Cslms: 70.79%
Sample Day Wind Rose
WIND SPEED
(Knots)
~ >=22
IWl 17-21
11 17
O 7- 11
4-7
2- 4
Cairns: 70.09%
7-17

-------
Figure 7-9. Wind Roses for the Long Beaeh/Daugherty Field Airport Weather Station near
LBHCA
Location of LBHCA and Weather Station
2003-2012 Historical Wind Rose

NORTH-^-,

20%

16%

12%

8%.

4% i :
•westF i \ M
/ •' ;EAS

SOUTH
WIND SPEED
(Knots)
I I »=22
~ 17-21
IH 11 1?
I -I 7- 11
~ 4-7
2- 4
Calms: 34.91%
2013 Wind Rose



westF


WIND SPEED
[Kn ots >
SOUTH
~ 4-7
H 2-4
Calms: 30.30%
Sample Day Wind Rose
WEST
WIND SPEED
(Kn ots)
SOUTH
Calms: 23.81%
7-18

-------
Figure 7-10. Wind Roses for the Riverside Municipal Airport Weather Station near RUCA
Locations of RUCA and Weather Station	2003-2012 Historical Wind Rose
2013 Wind Rose
Sample Day Wind Rose
NORTH"
WEST
v yPEtLJ
(Knots)
SOUTH
30%
24%
18%
12%
WIND SPEED
(Knots)
I I -22
I I 17-21
11
H 7-11
~l 4-7
2- 4
Calms: 36.63%
7-19

-------
Figure 7-11. Wind Roses for the San Jose International Airport Weather Station near
SJJCA
Location of SJJCA and Weather Station
2003-2012 Historical Wind Rose
EST
VWVlD SPEED
(Kn ots}
SOUTH
2013 Wind Rose
Sample Day Wind Rose
WEST
WWD SPEED
[Kn ots >
SOUTH
WEST
WIN D S PE ED
(Kn ots)
SOUTH
Calms: 31.63%
SJJCA
7-20

-------
Observations from Figure 7-9 for LBHCA include the following:
•	The weather station at the Long Beach/Daugherty Field Airport is located 2.6 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 2013 full-year wind rose are very similar to the historical
wind patterns, indicating that conditions in 2013 were representative of those
experienced historically.
•	The sample day wind rose has a lower percentage of calm winds than the historical
and full-year wind roses. The sample day wind rose also has fewer west-
northwesterly and northwesterly wind observations and a higher percentage of winds
from the south-southeast and south. Recall however, that sampling at LBHCA was
completed in July, and thus does not include wind observations from the second half
of 2013. The wind patterns on the sample day wind rose may be indicative of a
seasonal wind pattern.
Observations from Figure 7-10 for RUCA include the following:
•	The weather station at the Riverside Municipal Airport is located 3.5 miles south-
southwest of RUCA. The Santa Ana River and Wildlife Area lies between the airport
and the monitoring site.
•	Although calm winds were observed for approximately 32 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.
•	The full-year wind rose shows that west-northwesterly winds were observed less
frequently in 2013, as westerly winds account for more than 25 percent of
observations in 2013. As similar observation was noted in the 2011 and 2012 NMP
reports. Although still relatively low, the percentage of northeasterly winds shown on
the full-year wind rose is roughly double what is shown on the historical wind rose.
•	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 2013
were representative of those experienced over the entire year.
Observations from Figure 7-11 for SJJCA include the following:
•	The weather station at the San Jose International Airport is located 1.8 miles west-
northwest of SJJCA. Even though 1-880, the Guadalupe Parkway, and the Guadalupe
7-21

-------
River separate the airport and the monitoring site, this is one of the shortest distances
between a weather station and an NMP site.
•	Between 2003 and 2012, approximately 40 percent of winds near SJJCA were from
the west-northwest to north. Another 17 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 2013 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 2013 wind rose. This shift was also shown on the 2009 through
2012 wind roses in the 2008-2009, 2010, 2011, and 2012 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.
7.3 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 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. PAHs were sampled for at all four California sites; in addition, metals
(PMio) were also sampled for at SJJCA.
7-22

-------
Table 7-4. Risk-Based Screening Results for the California Monitoring Sites
Pollutant
Screening
Value
(Ug/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Los Angeles, California - CELA
Naphthalene
0.029
56
58
96.55
94.92
94.92
Acenaphthene
0.011
1
58
1.72
1.69
96.61
Benzo(a)pyrene
0.00057
1
32
3.13
1.69
98.31
Fluorene
0.011
1
57
1.75
1.69
100.00
Total
59
205
28.78

Long Beach, California - LBHCA
Naphthalene
0.029
19
29
65.52
100.00
100.00
Total
19
29
65.52

Rubidoux, California - RUCA
Naphthalene
0.029
53
58
91.38
98.15
98.15
Benzo(a)pyrene
0.00057
1
35
2.86
1.85
100.00
Total
54
93
58.06

San Jose, California - SJJCA
Naphthalene
0.029
48
59
81.36
46.60
46.60
Arsenic (PMi0)
0.00023
41
56
73.21
39.81
86.41
Nickel (PMio)
0.0021
11
60
18.33
10.68
97.09
Benzo(a)pyrene
0.00057
1
25
4.00
0.97
98.06
Cadmium (PMio)
0.00056
1
60
1.67
0.97
99.03
Lead (PMio)
0.015
1
60
1.67
0.97
100.00
Total
103
320
32.19

Observations from Table 7-4 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 95 percent (CELA) to 100 percent
(LBHCA).
•	Naphthalene was detected in all 58 valid PAH samples collected at CELA and failed
screens for 56 of these. Acenaphthene, benzo(a)pyrene, and fluorene also failed a
single screen each for CELA; because all three of these pollutants failed the same
number of screens, all three, in addition to naphthalene, were identified as pollutants
of interest for CELA.
•	Naphthalene was the only PAH to fail screens for LBHCA; thus naphthalene is the
only pollutant of interest for this site. Note that PAH sampling was discontinued in
July 2013.
7-23

-------
•	Naphthalene was detected in all 58 valid PAH samples collected at RUCA and failed
screens for 53 of these, accounting for 98 percent of this site's failed screens.
Benzo(a)pyrene also failed a single screen for RUCA, but was not identified as a
pollutant of interest.
•	SJJCA is the only site that sampled metals (PMio) in addition to PAHs. Although
naphthalene still accounts for the majority of failed screens for the site, arsenic also
contributed to a large number of the total failed screens. Together, these two
pollutants account for nearly 86 percent of SJJCA's total failed screens. Nickel
accounts for another 11 percent of the total failed screens for this site. Naphthalene,
arsenic, and nickel contributed to more than 95 percent of failed screens for SJJCA
and were therefore identified as pollutants of interest for this site. Benzo(a)pyrene,
cadmium, and lead also failed a single screen each for SJCCA but were not identified
as pollutants of interest.
7.4 Concentrations
This section presents various concentration averages used to characterize air toxics
pollution levels at the California monitoring sites. Where applicable, the following calculations
and data analyses were performed for each of the site-specific pollutants of interest:
•	Time period-based concentration averages (quarterly and annual) are provided for
each monitoring site.
•	Annual concentration averages are presented graphically for each site to illustrate
how the site's concentrations compare to the program-level averages, as presented in
Section 4.1.
•	Concentration averages and other statistical metrics are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at the California monitoring sites are provided in Appendices M and N.
7.4.1 2013 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
7-24

-------
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 7-5, 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 7-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest
for the California Monitoring Sites

# of






Measured
1st
2nd
3rd
4th


Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs. # of
Samples
Average
(ng/m3)
Average
(ng/m3)
Average
(ng/m3)
Average
(ng/m3)
Average
(ng/m3)
Los Angeles, California - CELA


3.95
3.56
4.72
4.71
4.26
Acenaphthene
58/58
± 1.35
±0.78
± 1.29
±0.86
±0.54


0.05
0.02
0.07
0.08
0.06
Benzo(a)pyrene
32/58
±0.04
±0.02
±0.10
±0.04
±0.03


5.04
5.46
6.48
5.68
5.67
Fluorene
57/58
± 1.02
± 1.45
± 1.41
±0.84
±0.57


127.85
84.26
88.87
141.17
111.44
Naphthalene
58/58
± 49.00
± 17.64
± 15.72
±27.21
± 15.95
Long Beach, California - LBHCA


120.00




Naphthalene
29/29
± 45.46
NA
NA
NA
NA
Rubidoux, California - RUCA


81.55
39.34
59.18
138.90
81.40
Naphthalene
58/58
± 18.46
± 10.71
± 11.11
±43.53
± 16.05
San Jose, California - SJJCA


0.49
0.20
0.28
1.06
0.52
Arsenic (PMio)
56/60
±0.19
±0.12
±0.09
±0.33
±0.13


148.08
41.70
32.48
149.43
93.97
Naphthalene
59/59
±57.34
± 11.59
±8.67
±41.24
± 22.27


1.52
1.07
1.24
1.77
1.40
Nickel (PMio)
60/60
±0.59
±0.30
±0.35
±0.41
±0.21
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
Observations for the California monitoring sites from Table 7-5 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 RUC A, based on the annual
averages. LBHCA does not have an annual average presented in Table 7-5 because
7-25

-------
sampling was discontinued at this site at the end of July 2013. However, summary
statistics for LBHCA covering the sampling period are provided in Appendix M.
For each site except LBHCA, naphthalene concentrations appear highest during the
first and fourth quarters of 2013, based on the quarterly averages. However, the
confidence intervals calculated for the quarterly averages indicate that there is
considerable variability in the measurements.
Naphthalene concentrations measured at CELA range from 23.20 ng/m3 to
332.5 ng/m3 with a median concentration of 92.2 ng/m3. All but one of CELA's 12
naphthalene concentrations greater than 150 ng/m3 was measured during either the
first or fourth quarter of 2013. However, both the maximum and minimum
naphthalene concentrations measured at CELA were measured in January. This helps
explain the large confidence interval shown for CELA's first quarter naphthalene
concentration.
The confidence interval for the third quarter average concentration of benzo(a)pyrene
for CELA is greater than the average itself, indicating a high level of variability
associated with the measurements. The two maximum concentrations measured at
CELA were measured during the third quarter (0.627 ng/m3 measured on July 27,
2013 and 0.371 ng/m3 measured on September 13, 2013). Only two additional
measured detections of benzo(a)pyrene were measured at CELA during the third
quarter; the other 11 were non-detects, which is the highest number of non-detects of
benzo(a)pyrene for a given quarter at CELA.
The maximum concentrations of acenaphthene and fluorene were measured at CELA
on the same date, September 7, 2013. Several of the higher measurements of these
two pollutants were measured on the same date. The magnitude of the concentrations
of these pollutants tend to track each other.
Concentrations of naphthalene measured at LBHCA range from 12.7 ng/m3 to
270 ng/m3, with a median concentration of 41.90 ng/m3. Because this site completed
sampling in July, an annual average could not be calculated. In addition, no second
quarter average is presented in Table 7-5 because there were a number of invalid
samples (five) and the quarterly criteria was not met. All seven naphthalene
concentrations greater than 100 ng/m3 were measured in January or February while
all 10 concentrations less than 30 ng/m3 were measured between late April and July.
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,
particularly the fourth quarter average concentration. All six naphthalene
concentrations greater than 150 ng/m3 measured at RUCA were measured during the
fourth quarter of 2013. Concentrations measured during the first and fourth quarters
range from 47.75 ng/m3 to 404 ng/m3 with a median concentration of 88.2 ng/m3.
Concentrations measured during the second and third quarters range from
20.85 ng/m3 to 100 ng/m3 with a median concentration of 43.2 ng/m3.
7-26

-------
•	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
variability associated with them. All 10 naphthalene concentrations greater than
150 ng/m3 measured at SJJCA were measured during the first or fourth quarters of
2013, with the three concentrations greater than 300 ng/m3 all measured in January.
Concentrations measured during the first and fourth quarters range from 41.0 ng/m3
to 377 ng/m3 with a median concentration of 121 ng/m3. Concentrations measured
during the second and third quarters range from 14.5 ng/m3 to 91.7 ng/m3 with a
median concentration of 34.5 ng/m3.
•	The fourth quarter average concentration of arsenic for SJJCA is significantly higher
than the other quarterly averages shown in Table 7-5. A review of the data shows that
all six arsenic concentrations greater than 1 ng/m3 were measured during the fourth
quarter, with the three highest concentrations all measured in late December. Further,
the 15 highest concentrations were all measured during the first or fourth quarters of
2013. Conversely, all eight concentrations less than or equal to 0.1 ng/m3 were
measured during the second or third quarters of 2013, as were three of the four non-
detects.
•	Concentrations of nickel measured at SJJCA range from 0.36 ng/m3 to 4.66 ng/m3,
with a median concentration of 1.14 ng/m3. Although the highest quarterly average
concentration is the fourth quarter average, the first quarter average has the highest
confidence interval associated with it, indicating considerable variability associated
with the measurements. The maximum nickel concentration measured at SJJCA
(4.66 ng/m3) was measured on February 27, 2013; the next highest concentration
measured that quarter was half as high (2.31 ng/m3 measured on January 4, 2013).
Three additional nickel concentrations greater than 1.50 ng/m3 were also measured
during the first quarter; yet, three of the 10 lowest concentrations were measured
during the first quarter.
Tables 4-9 through 4-12 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for the
California sites from those tables include the following:
•	CELA, RUCA, and SJJCA each appear in Table 4-11 for naphthalene, ranking fourth,
tenth, and sixth, respectively. CELA also ranks eighth for acenaphthene.
•	SJJCA appears twice in Table 4-12 for PMio metals. SJJCA has the sixth highest
annual average concentration of nickel and tenth highest annual average
concentration of arsenic among NMP sites sampling PMio metals.
7-27

-------
7.4.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for the pollutants shaded in
gray in Table 7-4 for CELA, RUCA, and SJJCA. Figures 7-12 through 7-17 overlay the sites'
minimum, annual average, and maximum concentrations onto the program-level minimum, first
quartile, median, average, third quartile, and maximum concentrations, as described in
Section 3.4.3.1. Because annual averages could not be calculated for LBHCA, box plots were not
created for this site.
Figure 7-12. Program vs. Site-Specific Average Acenaphthene Concentration


¦
—c

Program Max Concentration 123 ng/m3


T	1	1	1	1	1	T
0	10	20	30	40	50	60	70	80
Concentration (ng/m3)
Program: 1st Quartile
2nd Quartile 3rd Quartile
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



Figure 7-13. Program vs. Site-Specific Average Arsenic (PMio) Concentration
SJJCA
12 3
4 5 6
Concentration {ng/m3)
7
8
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site: Site Average
o
Site Concentration Range


7-28

-------
Figure 7-14. Program vs. Site-Specific Average Benzo(a)pyrene Concentration
—O
0.25
0.5
0.75 1 1.25
Concentration {ng/m3)
1.5 1.75
2
Program:
1st Quartile
¦
2nd Quartile 3rd Quartile
~ ~
4th Quartile
~
Average
i
Site:
Site Average
o
Site Concentration Range


Figure 7-15. Program vs. Site-Specific Average Fluorene Concentration
Ht I I I I I I I I
10
20 30
40 50 60
Concentration {ng/m3)
70 80 90
Program:
1st Quartile
2nd Quartile 3rd Quartile
4th Quartile Average

¦
~
~
~ 1
Site:
Site Average
Site Concentration Range


o


7-29

-------
Figure 7-16. Program vs. Site-Specific Average Naphthalene Concentrations
-
100
200
300 400 500
Concentration {ng/m3)
600
700
Program:
1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site:
Site Average
o
Site Concentration Range


Figure 7-17. Program vs. Site-Specific Average Nickel (PMio) Concentration
0
5
10 15
Concentration {ng/m3)

20
25

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i


Site: Site Average
o
Site Concentration Range



Observations from Figures 7-12 through 7-17 include the following:
• Figure 7-12 is the box plot for acenaphthene for CELA. Note that the program-
level maximum concentration (123 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
80 ng/m3. Figure 7-12 shows the maximum acenaphthene concentration measured
at CELA is an order of magnitude less than the program-level maximum
concentration. CELA's annual average concentration is just less than the
program-level average concentration and just greater than the program-level third
quartile.
7-30

-------
•	Figure 7-13 shows that the annual average arsenic (PMio) concentration for
SJJCA is less than the program-level average concentration and similar to the
median concentration of arsenic (PMio). The minimum concentration measured at
SJJCA is zero, indicating that at least one non-detect of arsenic was measured at
SJJCA; four non-detects of arsenic were measured at SJJCA.
•	Figure 7-14 is the box plot for benzo(a)pyrene for CELA. Note that the program-
level first quartile is zero and therefore not visible on the box plot. The annual
average benzo(a)pyrene for CELA falls right between the program-level median
and average concentrations of this pollutant. Note that CELA is one of only two
NMP sites for which benzo(a)pyrene is a pollutant of interest.
•	Figure 7-15 for fluorene shows that the range of fluorene concentrations measured
at CELA is relatively small compared to the range measured across the program,
yet CELA's annual average is greater than the program-level average and third
quartile. This is the result of non-detects. Of the 174 non-detects of fluorene
measured across the program, only one was measured at CELA. CELA ties with
three other sites for the second lowest number of fluorene non-detects (only
S4MO had none).
•	Figure 7-16 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 SJJCA and RUCA, although the
largest range of concentrations was measured at RUCA. All three annual average
naphthalene concentrations shown are greater than the program-level average
concentration. SJJCA's annual average is also similar to the program-level third
quartile and CELA's annual average concentration is greater than the program-
level third quartile. There were no non-detects of naphthalene measured at CELA,
RUCA, SJJCA, or across the program.
•	Figure 7-17 is the box plot for nickel for SJJCA. SJJCA's annual average nickel
concentration is greater than the program-level average concentration and just
greater than the program-level third quartile, although the maximum
concentration measured at SJJCA is considerably less than the maximum
concentration measured across the program. The minimum nickel concentration
measured at SJJCA is similar to the program-level first quartile.
7.4.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.3.2.
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 7-18 through 7-25 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
7-31

-------
of zeros for non-detects. If sampling began mid-year, a minimum of 6 months of sampling is
required for inclusion in the trends analysis; in these cases, a 1-year average concentration is not
provided, although the range and percentiles are still presented. A trends analysis was not
conducted for LBHCA because this site has not sampled under the NMP for at least
5 consecutive years.
Figure 7-18. Yearly Statistical Metrics for Acenaphthene Concentrations Measured at CELA
20071	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	- Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2007.
Observations from Figure 7-18 for acenaphthene measurements collected at CELA
include the following:
•	CELA began sampling PAHs under the NMP in April 2007. Because a full year's
worth of data is not available, a 1-year average concentration for 2007 is not
presented, although the range of measurements is provided.
•	The maximum acenaphthene concentration was measured at CELA on October 16,
2009 and is the only concentration greater than 25 ng/m3 measured at this site.
•	Acenaphthene concentrations measured at CELA increased significantly between
2007 and 2009, as indicated by nearly all of the statistical metrics shown. With the
exception of the minimum concentration, each of the statistical metrics exhibits a
decreasing trend between 2009 and 2011.
7-32

-------
• Excluding 2007, the maximum, 95th percentile, and 1-year average concentrations
are at a minimum for 2013. 2013 has the smallest difference between the 5th and 95th
percentiles since 2007, indicating that the majority of measurements are falling within
a smaller range of concentrations.
Figure 7-19. Yearly Statistical Metrics for Benzo(a)pyrene Concentrations Measured at CELA



1.0 -











I






—





T




o.o -






	













20071	2008	2009	2010	2011	2012	2013
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2007.
Observations from Figure 7-19 for benzo(a)pyrene measurements collected at CELA
include the following:
•	The maximum benzo(a)pyrene concentration was measured at CELA on January 1,
2009 and is the only concentration greater than 1 ng/m3 measured at this site,
although two benzo(a)pyrene concentrations close to 1 ng/m3 were measured at
CELA in 2010. The fourth highest benzo(a)pyrene concentration measured at CELA
is the maximum concentration for 2013 (0.627 ng/m3).
•	With the exception of the maximum concentration for 2009, the range of
concentrations measured at CELA in 2008 and 2009 is fairly similar. The increase in
the 1-year average concentration shown is mostly a result of the maximum
concentration. Excluding the maximum concentration from the 1-year average for
2009 results in an average concentration very similar to that of 2008.
•	The median concentration for 2010 is zero, indicating that at least half of the
measurements were non-detects. The number of non-detects increased each year from
7-33

-------
2007 through 2010, reaching a maximum of 41 non-detects in 2010. As a result, even
though the second and third highest concentrations of benzo(a)pyrene were measured
at CELA in 2010, each of the statistical parameters exhibits a decrease from the
previous year. Further decreases in the statistical metrics are shown for 2011 (even
though fewer non-detects (18) were measured) and 2012 (although the maximum
concentration is up).
• Several of the statistical metrics exhibit increases for 2013, although the median
concentration continues its decreasing trend. The decrease in the median
concentrations result from the increasing number of non-detects, which increased
from 18 in 2011 to 22 in 2012 and 26 in 2013 while the increase in the 1-year average
concentrations result from a higher number of concentrations at the upper end of the
range. For example, in 2012, only one concentration greater than 0.2 ng/m3 was
measured at CELA; in 2013, that number increased to five (the most since 2009).
Figure 7-20. Yearly Statistical Metrics for Fluorene Concentrations Measured at CELA
.2 20
2007 1	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2007.
Observations from Figure 7-20 for fluorene measurements collected at CELA include the
following:
• The smallest range of fluorene 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 until the a second non-detect was measured in 2013.
7-34

-------
•	The range of concentrations measured, and thus the statistical parameters shown,
increased through 2009, when the maximum fluorene concentration was measured
(on the same date that the maximum acenaphthene 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 have the smallest range of measurements besides 2007 (which was
a partial year).
•	All of the statistical parameters exhibit an increase from 2011 to 2012 then exhibit
decreases for 2013.
Figure 7-21. Yearly Statistical Metrics for Naphthalene Concentrations Measured at CELA
20071	2008	2009	2010	2011	2012	2013
Year
O 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 April 2007.
Observations from Figure 7-21 for naphthalene measurements collected at CELA include
the following:
•	The statistical parameters shown for naphthalene in Figure 7-21 exhibit a similar
pattern as the statistical parameters for fluorene shown in Figure 7-20 and, to a less
extent, acenaphthene in Figure 7-18.
•	The smallest range of concentrations was measured in 2007, although the statistical
metrics do not represent a full year of sampling. The minimum concentration
measured at CELA was measured in 2007 (1.30 ng/m3); in addition, 2007 is the only
7-35

-------
year in which a concentration less than 10 ng/m3 was measured (there were five in
total). The range of naphthalene measurements, and thus the statistical parameters
shown, increase through 2009, when the maximum concentration was measured (736
ng/m3 also on October 16, 2009). Concentrations greater than 500 ng/m3 were also
measured in 2008 and 2010. The maximum, 95th percentile, 1-year average, and
median concentrations decrease from 2009 to 2010 and again for 2011.
•	All of the statistical parameters shown in Figure 7-21 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.
•	With the exception of the minimum concentration, all of the statistical metric shown
in Figure 7-21 are at a minimum for 2013 since the first full year of sampling.
Figure 7-22. Yearly Statistical Metrics for Naphthalene Concentrations Measured at RUCA
2010
Year
O 5th Percentile	- Minimurr
O 95th Percentile
• Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2007.
Observations from Figure 7-22 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.
7-36

-------
•	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
(406 ng/m3), although another concentration greater than 400 ng/m3 was measured at
RUCA in 2013. Naphthalene concentrations greater than 300 ng/m3 have been
measured at least once every year since 2009.
•	The 1-year average concentration has an increasing trend over most of the years of
sampling through 2012, although 2010 was down slightly. The median concentration
has a similar pattern.
•	The range of concentrations measured at RUCA reflects a relatively high level of
variability in the measurements collected. For both 2009 and 2013, the maximum
concentration is twice the 95th percentile. Even though the majority of concentrations
measured in 2012 fall within a tighter range of measurements than preceding years,
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 shown in the
median concentration indicates that concentrations were higher overall for 2012.
•	Even with the second highest naphthalene concentration measured since the onset of
sampling at RUCA, most of the statistical parameters exhibit a decrease for 2013.
Figure 7-23. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at SJJCA
2010	2011
Year
O 5th Percentile	- Minimurr
O 95th Percentile
7-37

-------
Observations from Figure 7-23 for arsenic measurements collected at SJJCA include the
following:
•	The maximum concentration of arsenic was measured on the first day of sampling at
SJJCA (January 1, 2008). The second and third highest concentrations were measured
in 2013. All but one of the seven concentrations greater than 1.5 ng/m3 were
measured in 2008 (two) or 2013 (four).
•	The 1-year average arsenic concentration decreased from 2008 to 2009. Although this
is mostly due to the maximum concentration measured in 2008, all of the statistical
parameters exhibit a decrease from 2008 to 2009, indicating that the decrease is not
only due to the difference in the maximum concentrations. The number of
concentrations at the lower end of the concentration range increased for 2009. In
2009, two non-detects were measured at SJJCA, compared to none in 2008. In
addition, seven arsenic concentrations less than 0.1 ng/m3 were measured in 2009
compared to only two in 2008.
•	The 1-year average arsenic concentration changed little through 2012, ranging from
0.31 ng/m3 for 2009 to 0.39 ng/m3 for 2012. With the exception of the minimum and
5th percentile (which did not change), all of the statistical metrics exhibit an increase
for 2013. The 95th percentile for 2013 is greater than the maximum concentration
measured for all years except 2008.
Figure 7-24. Yearly Statistical Metrics for Naphthalene Concentrations Measured at SJJCA
o +-
20081
2009
2010
2011
Year
2012
2013

O 5th Percentile
— Minimum
— Median
- Maximum
O 95th Percentile

1A 1-year average is not presented because sampling under the NMP did not begin until May 2008.
7-38

-------
Observations from Figure 7-24 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). No additional concentrations greater than 400 ng/m3 have been
measured at SJJCA.
•	The median concentration has changed little over the years through 2012, ranging
from 43.00 ng/m3 (2010) to 49.90 ng/m3 (2011); 2013 is the first year with a median
concentration greater than 50 ng/m3 (57.70 ng/m3). The 1-year average concentration
exhibits more variability, ranging from 63.44 ng/m3 (2010) to 81.04 ng/m3 (2009)
through 2012, then increasing to 93.97 ng/m3 for 2013.
•	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 representing the upper end of the concentration range. For example, the
95th percentile increased by 70 percent from 2010 to 2011.
Figure 7-25. Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at SJJCA
2010	2011
Year
O 5th Percentile	— Minimurr
Maximum	O 95th Percentile
• Average
7-39

-------
Observations from Figure 7-25 for nickel measurements collected at SJJCA include the
following:
•	The maximum concentration of nickel was measured on February 27, 2013. All of the
14 measurements greater than or equal to 2.5 ng/m3 were measured after 2010,
specifically, five in 2011, three in 2012, and five in 2013.
•	After a significant decrease 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 maximum concentration increased from 2011 to 2012, most of the
statistical metrics exhibit decreases for 2012. Four of the five nickel concentrations
less than 0.30 ng/m3 were measured in 2012. The minimum concentration decreased
by half between 2009 and 2012.
•	Each of the statistical metrics shown in Figure 7-25 exhibits an increase for 2013.
Four of the six statistical parameters are at a maximum for 2013 (only the minimum
concentration and 5th percentile are not).
7.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at each California monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4
for definitions and explanations regarding the various toxicity factors, time frames, and
calculations associated with these risk-based screenings.
7.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the California monitoring sites and where annual
average concentrations could be calculated, risk was examined by calculating cancer risk and
noncancer hazard approximations. These approximations can be used as risk estimates for cancer
and noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers want to shift their air
monitoring priorities. Refer to Section 3.4.3.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 7-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
7-40

-------
Table 7-6. Risk Approximations for the California Monitoring Sites
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Los Angeles, California - CELA
Acenaphthene
0.000088

58/58
4.26
±0.54
0.37

Benzo(a)pyrene
0.00176

32/58
0.06
±0.03
0.10

Fluorene
0.000088

57/58
5.67
±0.57
0.50

Naphthalene
0.000034
0.003
58/58
111.44
± 15.95
3.79
0.04
Long Beach, California - LBHCA
Naphthalene
0.000034
0.003
29/29
NA
NA
NA
Rubidoux, California - RUCA
Naphthalene
0.000034
0.003
58/58
81.40
± 16.05
2.77
0.03
San Jose, California - SJJCA
Arsenic (PMi0)
0.0043
0.000015
56/60
0.52
±0.13
2.22
0.03
Naphthalene
0.000034
0.003
59/59
93.97
± 22.27
3.19
0.03
Nickel (PMio)
0.00048
0.00009
60/60
1.40
±0.21
0.67
0.02
— = 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 7-6 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 concentration CELA is the highest of the three
annual averages for naphthalene, followed by the annual average for SJJCA and
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 2.77 in-a-million for RUCA to 3.79 in-a-million for
CELA.
•	None of the other pollutants of interest for CELA have cancer risk approximations
greater than 1 in-a-million.
•	Even though the annual average concentration of nickel is nearly three times greater
than the annual average concentration of arsenic for SJJCA, arsenic has the higher
cancer risk approximation (2.22 in-a-million) compared to nickel (0.67 in-a-million).
7-41

-------
•	All of the noncancer hazard approximations for the pollutants of interest for the
California monitoring sites are less than 1.0, where noncancer RfCs are available,
indicating that no adverse noncancer health effects are expected from these individual
pollutants.
•	Cancer risk and noncancer hazard approximations could not calculated for LBHCA
due to the mid-year end date of sampling, as discussed in the previous sections.
7.5.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 7-7 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
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 3.4.3.4. Lastly,
Table 7-7 provides the 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.
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. Thus,
LBHCA does not have cancer risk and noncancer hazard approximations in Tables 7-7 and 7-8.
A more in-depth discussion of this analysis is provided in Section 3.4.3.4. Similar to the cancer
risk and noncancer hazard approximations provided in Section 7.5.1, this analysis may help
policy-makers prioritize their air monitoring activities.
7-42

-------
Table 7-7. 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)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Los Angeles, California (Los Angeles County) - CELA
Formaldehyde
2,221.45
Formaldehyde
2.89E-02
Naphthalene
3.79
Benzene
1,913.13
POM, Group la
1.49E-02
Fluorene
0.50
Dichloromethane
1,682.67
Benzene
1.49E-02
Acenaphthene
0.37
Ethylbenzene
1,101.33
1,3-Butadiene
9.87E-03
Benzo(a)pyrene
0.10
T etrachloroethy lene
1,076.88
POM, Group 2b
7.27E-03

Acetaldehyde
962.00
POM, Group 5a
6.02E-03
p-Dichlorobenzene
339.36
POM, Group 2d
5.84E-03
1.3 -Butadiene
328.83
Naphthalene
5.27E-03
POM, Group la
169.60
p-Dichlorobenzene
3.73E-03
Naphthalene
154.91
Hexavalent Chromium
3.03E-03
Long
Beach, California (Los Angeles County) - LBHCA
Formaldehyde
2,221.45
Formaldehyde
2.89E-02

Benzene
1,913.13
POM, Group la
1.49E-02
Dichloromethane
1,682.67
Benzene
1.49E-02
Ethylbenzene
1,101.33
1,3-Butadiene
9.87E-03
T etrachloroethy lene
1,076.88
POM, Group 2b
7.27E-03
Acetaldehyde
962.00
POM, Group 5a
6.02E-03
p-Dichlorobenzene
339.36
POM, Group 2d
5.84E-03
1.3 -Butadiene
328.83
Naphthalene
5.27E-03
POM, Group la
169.60
p-Dichlorobenzene
3.73E-03
Naphthalene
154.91
Hexavalent Chromium
3.03E-03

-------
Table 7-7. 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
Top 10 Cancer Toxicity-Weighted
Emissions
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(County-Level)

(County-Level)
(Site-Specific)



Cancer

Cancer Risk
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(in-a-million)
Rubidoux, California (Riverside County) - RUCA
Formaldehyde
418.81
Formaldehyde
5.44E-03
Naphthalene
2.77
Benzene
317.30
Benzene
2.47E-03


T etrachloroethy lene
214.39
Hexavalent Chromium
2.04E-03


Dichloromethane
200.68
POM, Group la
1.88E-03


Acetaldehyde
197.01
1.3 -Butadiene
1.47E-03


Ethylbenzene
191.03
POM, Group 2b
1.45E-03


p-Dichlorobenzene
70.48
POM, Group 5a
1.20E-03


1.3 -Butadiene
48.84
Naphthalene
1.19E-03


Naphthalene
34.99
POM, Group 2d
1.09E-03


1,3 -Dichloropropene
29.57
p-Dichlorobenzene
7.75E-04


San Jose, California (Santa Clara County) - SJJCA
Benzene
356.17
Formaldehyde
4.46E-03
Naphthalene
3.19
Formaldehyde
342.81
Benzene
2.78E-03
Arsenic
2.22
Ethylbenzene
232.74
POM, Group 2b
1.73E-03
Nickel
0.67
Dichloromethane
191.47
Hexavalent Chromium
1.67E-03


Acetaldehyde
171.62
POM, Group 5a
1.63E-03


T etrachloroethy lene
110.40
1,3-Butadiene
1.35E-03


p-Dichlorobenzene
60.37
POM, Group 2d
1.32E-03


1.3 -Butadiene
45.07
Naphthalene
1.26E-03


Naphthalene
37.18
POM, Group la
1.21E-03


T richloroethylene
29.51
/j-Dichlorobcnzcne
6.64E-04



-------
Table 7-8. 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)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Ha
Based on Annual Av<
(Site-S
izard Approximations
;rage Concentrations
>ecific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Los Angeles, California (Los Angeles County) - CELA
Toluene
8,265.39
Acrolein
6,797,409.70
Naphthalene
0.04
1,1,1 -T richloroethane
6,903.37
Chlorine
230,010.81

Xylenes
4,970.97
Formaldehyde
226,678.82
Hexane
4,520.90
1,3-Butadiene
164,416.69
Formaldehyde
2,221.45
Acetaldehyde
106,888.65
Benzene
1,913.13
Benzene
63,771.13
Dichloromethane
1,682.67
Cyanide Compounds, PM
63,440.92
Ethylene glycol
1,465.20
T richloroethy lene
56,352.54
Methanol
1,338.85
Naphthalene
51,636.02
Ethylbenzene
1,101.33
Xylenes
49,709.73
Long Beach, California (Los Angeles County) - LBHCA
Toluene
8,265.39
Acrolein
6,797,409.70

1,1,1 -T richloroethane
6,903.37
Chlorine
230,010.81
Xylenes
4,970.97
Formaldehyde
226,678.82
Hexane
4,520.90
1,3-Butadiene
164,416.69
Formaldehyde
2,221.45
Acetaldehyde
106,888.65
Benzene
1,913.13
Benzene
63,771.13
Dichloromethane
1,682.67
Cyanide Compounds, PM
63,440.92
Ethylene glycol
1,465.20
T richloroethy lene
56,352.54
Methanol
1,338.85
Naphthalene
51,636.02
Ethylbenzene
1,101.33
Xylenes
49,709.73

-------
Table 7-8. 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)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Ha
Based on Annual Av<
(Site-S
izard Approximations
;rage Concentrations
>ecific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Rubidoux, California (Riverside County) - RUCA
Toluene
1,541.54
Acrolein
1,151,923.43
Naphthalene
0.03
Xylenes
1,037.06
Chlorine
71,489.03

Hexane
1,034.89
Formaldehyde
42,736.21
1,1,1 -T richloroethane
617.84
1,3-Butadiene
24,417.60
Formaldehyde
418.81
Acetaldehyde
21,889.50
Benzene
317.30
Bromomethane
13,246.82
Ethylene glycol
241.17
Naphthalene
11,663.14
Methanol
218.85
Lead, PM
11,143.30
Tetrachloroethylene
214.39
Benzene
10,576.82
Dichloromethane
200.68
T richloroethy lene
10,486.48
San Jose, California (Santa Clara County) - SJJCA
Toluene
1,762.28
Acrolein
1,804,553.18
Arsenic
0.03
1,1,1 -T richloroethane
1,289.63
Chlorine
91,338.84
Naphthalene
0.03
Hexane
1,014.84
Formaldehyde
34,980.53
Nickel
0.02
Xylenes
987.31
1,3-Butadiene
22,537.16

Benzene
356.17
Acetaldehyde
19,068.78
Formaldehyde
342.81
T richloroethy lene
14,754.18
Ethylene glycol
280.57
Naphthalene
12,392.06
Ethylbenzene
232.74
Benzene
11,872.49
Methanol
216.21
Xylenes
9,873.13
Dichloromethane
191.47
Lead, PM
9,571.88

-------
Observations from Table 7-7 include the following:
•	Formaldehyde and benzene are the highest emitted pollutants with cancer UREs in
Los Angeles and Riverside Counties while benzene is emitted in slightly higher
quantities than formaldehyde in Santa Clara County. The quantity of emissions is
considerably greater for Los Angeles County than Riverside and Santa Clara
Counties.
•	Formaldehyde has the highest toxicity-weighted emissions (of the pollutants with
cancer UREs) for all three counties. POM, Group 1, benzene, and 1,3-butadiene rank
behind formaldehyde for Los Angeles County; benzene, hexavalent chromium, and
POM, Group 1 rank behind formaldehyde for Riverside County; and benzene, POM,
Group 2b, and hexavalent chromium rank behind formaldehyde for Santa Clara
County.
•	Six of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Los Angeles County, while there are five in common for Riverside and
Santa Clara Counties.
•	Naphthalene has the highest 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 and nickel, the other pollutants of interest for SJJCA, do not appear on either
emissions-based list (they rank lower than tenth). Hexavalent chromium is the only
metal shown for Santa Clara County, ranking fourth highest for its toxicity-weighted
emissions.
•	Several POM Groups appear among the pollutants with the highest toxicity-weighted
emissions for each county. POM, Group 2b includes acenaphthene and fluorene,
which were both identified as pollutants of interest for CELA. POM, Group 2d
includes several PAHs sampled for at the California sites, such as anthracene and
phenanthrene, although none of these failed screens. POM, Group 5a includes
benzo(a)pyrene, which failed screens for CELA and RUCA. POM, Group la, which
also appears among each county's toxicity-weighted emissions, includes unspeciated
compounds.
Observations from Table 7-8 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 fourth for
Riverside County. Xylenes are the second highest emitted pollutant in Riverside
County but ranks third and fourth for Los Angeles and Santa Clara Counties,
respectively. Hexane is also among the top four emitted pollutants in each of these
counties.
7-47

-------
•	Acrolein, chlorine, formaldehyde, 1,3-butadiene, and acetaldehyde are the five
pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) for all three counties. Although acrolein and chlorine rank highest
for toxicity-weighted emissions for each county, neither pollutant appears among the
highest emitted. This is also true for acetaldehyde, and 1,3-butadiene. Conversely,
formaldehyde has the fifth highest emissions for each county.
•	Three of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Los Angeles and Santa Clara Counties, while only two of the highest
emitted pollutants also have the highest toxicity-weighted emissions for Riverside
County.
•	Naphthalene is the only pollutant for which a noncancer hazard approximation could
be calculated for all three counties. Naphthalene does not appear among the highest
emitted pollutants (of those with a noncancer RfC) for any of the three counties.
Naphthalene ranks seventh for its toxicity-weighted emissions for Riverside and
Santa Clara Counties and ninth for Los Angeles County.
•	Arsenic and nickel are the only other pollutants of interest for SJJCA for which
noncancer hazard approximations could be calculated. Lead is the only metal that
appears on either emissions-based list for Santa Clara County in Table 7-8. This
pollutant failed a single screen for SJJCA but was not identified as a pollutant of
interest for this site.
7.6 Summary of the 2013 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. Three additional PAHs failed
screens for CELA and one additional PAH failed screens for RUCA. One additional
PAH andfour PMio 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 fourth
highest annual average concentration of naphthalene among NMP sites sampling
PAHs.
~~~ Concentrations of naphthalene were higher during the first andfourth quarters (or
the colder months) of 2013 for CELA, RUCA, and SJJCA.
~~~ Concentrations of each of the pollutants of interest for SJJCA increasedfrom 2012 to
2013, particularly for the metals.
~~~ Naphthalene has the highest cancer risk approximation of the pollutants of interest
for each site. None of the pollutants of interest for the California sites have
noncancer hazard approximations greater than an HQ of 1.0.
7-48

-------
8.0	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.
8.1	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 38 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 8-1 for GPCO is a composite satellite image retrieved from
ArcGIS Explorer showing the monitoring site and its immediate surroundings. Figure 8-2
identifies nearby point source emissions locations by source category, as reported in the 2011
NEI for point sources, version 2. Note that only sources within 10 miles of the site are included
in the facility counts provided in Figure 8-2. A 10-mile boundary was chosen to give the reader
an indication of which emissions sources and emissions source categories could potentially have
a direct effect on the air quality at the monitoring site. Further, this boundary provides both the
proximity of emissions sources to the monitoring site as well as the quantity of such sources
within a given distance of the site. Sources outside the 10-mile boundary are still visible on the
map for reference, but have been grayed out in order to emphasize emissions sources within the
boundary. Figures 8-3 through 8-9 are the composite satellite maps and emissions sources maps
for the Garfield County sites. Table 8-1 provides supplemental geographical information such as
land use, location setting, and locational coordinates.
8-1

-------
Figure 8-1. Grand Junction, Colorado (GPCO) Monitoring Site
00
to
fc
m
¦
*MlV t
P

-------
Figure 8-2. NEI Point Sources Located Within 10 Miles of GPCO
Masa County


Leaend	Note- out to taoMy dens*/ and collocation the total facilities
a	displayed nay not represent all tocOmes vwthm He a«ea of ntoreat
GPCO NATTS site	10 mile radius [	County boundary
Source Category Group (No. of Facilities)
T AlportAlftin*/Airport Support Opa rabona
jf Asphalt Production -Hot M«x Asphalt Plant <\2)
0	Auto Body S*>ociPaintan/Ai>iomotw« Stores <4I
Automo»v«'*lV Deaierswip 111
a nek Structural Clay or Clay Coramcs Plant \ 1)
B Bulk Terminals Bulk Plants i4>
C Chemical Manufacturing Pacifty 13)
1	Compressor station < 1 f
t< Crematory - AntmelHuran (4)
(J>	Dry Cleaning Facility (3|
6	Etece icet Eouiorrenl War ufacluung Facilfiv 111
E	EjecfroplMnu Pialrng Po<*hri|i Anod£>na anil Coloring <2 >
i	GasotaerDMrse • Service Stafeon |43)
•	ndustiiat Machinery or Equip»nert Plan! |3)
O	-nsnhjOonai iscnool hospital prison etc i (4)
•	LandWt 11|
A	Metal Coafirm fcng'avnj} and AH*a Serv cos to Mani/acsure's < 1
<•)	Metals Processn&'Fabncation facility (2)
Mwe-QuarryflMinefal ProcessnQ Facility <35)
?	Miscellaneous Commie'eiaA industrial Faulty t O
•	Ol and>c* Oas Production (2i
R	Risttc Rawp o« Rubber Products Plant 11)
X	R*» Ymtmu Line Ope rations (2}
T	Te * tile Yam. or Carpel Rant 111
•	Whatever uaetman* Faohty 11)
W	VAiodwoHL Fi#nrture. Milwo* & Vtood Preserving F aoniy. 1)
8-3

-------
Figure 8-3. Battlement Mesa, Colorado (BMCO) Monitoring Site
00
¦k

-------
Figure 8-4. Silt, Colorado (BRCO) Monitoring Site
00

-------
Figure 8-5. Parachute, Colorado (PACO) Monitoring Site

-------
Figure 8-6. Rifle, Colorado (RICO) Monitoring Site

-------
Figure 8-7. NEI Point Sources Located Within 10 Miles of BMCO, BRCO, PACO, and
RICO
lortsirW wwcnN wiwv. tortrrr* wrvmf) t#rsww vr*trr*i «rw* ottswv
io Stance
County I
Gart«la
County
TDKOTT-W	ujrxnrw w,'«irw tumnrw turwirw "jrjmrw !'!rv»-.-v*
NMe Duo to facility dansity and cctocatwn thototaf tastrtiM
Oeel*y*9 may no« f«9i*»*nt HI Utilities •HBIn tl'« «»• o> lnten>»t.
Leqend
BMCO UATMP site	BRCO UATMP site ^ PACO UATMP site ^ RICO UATMP site
10 mile radius | | County boundary
Source Category Group (No. of Facilities)
T
Airport'Airline/Alrport Support Operations (7)
¦
Gasoline/Diesel Service Station (19)
i
Asphalt Production/Hot Mix Asphalt Plant (1)
•
Landfill (1)
B
Bulk Terminals/Bulk Plants (1)
X
Mine/Quatry/Mineral Processing Facility (10)
I
Compressor Station (20)
f
Miscellaneous Commercial/Industrial Facility (1)

Dry Cleaning Facility (1)
•
Oil and/or Gas Production (1,057)
<
Electricity Generation via Combustion (1)
X
Rail Yard/Rail Una Operations (1)

Gas Plant (2)
TT
Telecommunications/Radio Facility (1)

-------
Figure 8-8. Carbondale, Colorado (RFCO) Monitoring Site

»1S» NG» use
M . Of t CO'p

-------
Figure 8-9. NEI Point Sources Located Within 10 Miles of RFC'O
GarSeio
County |
taa»
County
Pfiun
County
RFCO UATMP site	10 mile radius	County boundary
wrnnrw
Meu
County
^	,	j
«rwnni	'D7-2irtrw
Legend
Ht7t5irw	toruRrw	to; -st"*
Soto. Due 1o facut> d»ns«t> and collocation the total facilities
rx>f	al<	vMtMo en® of nl*rt«t
Source Category Group (No. of Facilities)
T AjrporVArlirve/Atrport Support Operations (6)
C BulldlngrCortttructton (1)
i Compressor Station (1)
! Crematory Animal/Human (1)
¦ Gasoline/Diesel Service Station (10)
O Institutional (school hospital pnson etc*(1>
Miri^OuairyMirieral Processing FaclMy (3)
8-10

-------
Table 8-1. Geographical Information for the Colorado Monitoring Sites
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Additional Ambient Monitoring Information1
GPCO
08-077-0017
08-077-0018
Grand
Junction
Mesa
Grand Junction,
CO
39.064289,
-108.56155
Commercial
Urban/City
Center
Meteorological parameters, CO, PMio, PMio
Speciation, PM Coarse, PM2.5, and PM2.5 Speciation,
IMPROVE Speciation.
BMCO
08-045-0019
Battlement
Mesa
Garfield
Glenwood Springs,
CO
39.438060,
-108.026110
Commercial
Suburban
Meteorological parameters, PMio, PM2.5, O3. NO,
NO2, NOx, TNMOC, and Total Hydrocarbons.
BRCO
08-045-0009
Silt
Garfield
Glenwood Springs,
CO
39.487755,
-107.659685
Agricultural
Rural
None.
PACO
08-045-0005
Parachute
Garfield
Glenwood Springs,
CO
39.453654,
-108.053259
Residential
Urban/City
Center
PMio.
RICO
08-045-0007
Rifle
Garfield
Glenwood Springs,
CO
39.531813,
-107.782298
Commercial
Urban/City
Center
PMio.
RFCO
08-045-0018
Carbondale
Garfield
Glenwood Springs,
CO
39.412278,
-107.230397
Residential
Rural
PMio.
1 Data for additional pollutants are reported to AQS for these sites (EPA, 2014b); however, these data are not generated by ERG and are therefore not included in this report.
BOLD ITALICS = EPA-designated NATTS Site

-------
The GPCO monitoring site is comprised of two locations. The first location is a small
1-story shelter that houses the VOC and carbonyl compound samplers, with the PAH sampler
located just outside the shelter. The second location, which is on the roof of an adjacent 2-story
building, is comprised of the hexavalent chromium samplers. As a result, two AQS codes are
provided in Table 8-1. Figure 8-1 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 a few
blocks to the south of the GPCO monitoring site, and merges with another railroad to the
southwest of the site. The Colorado River can be seen in the bottom left-hand corner of
Figure 8-1 near the junction with the Gunnison River. Grand Junction is located in the Grand
Valley, which lies north and northeast of the Colorado National Monument.
As Figure 8-2 shows, GPCO is located within 10 miles of numerous emissions sources.
Many of the sources are located along a diagonal line running roughly northwest to southeast
along Highways 6 and 50 and Business-70 and oriented along the mountain valley. Many of the
point sources near GPCO fall into the gasoline/diesel service station or the mine/quarry/mineral
processing source categories. The sources closest to GPCO are an industrial
machinery/equipment plant, a bulk terminal/bulk plant, a gasoline/diesel service station, and an
auto body shop.
Four of the five Garfield County monitoring sites are situated in towns located along a
river valley along the Colorado River and paralleling 1-70. The BMCO monitoring site is located
in Battlement Mesa, a rural community located to the southeast of Parachute. The monitoring site
is located on the roof of the Grand Valley Fire Protection District facility, near the intersection of
Stone Quarry Road and West Battlement Parkway, as shown in Figure 8-3. The site is
surrounded primarily by residential subdivisions. A gas station is location immediately to the
north of the site and a cemetery is located to the south.
The BRCO monitoring site is located on Bell/Melton Ranch, off Owens Drive,
approximately 4 miles south of the town of Silt. The site is both rural and agricultural in nature.
As shown in Figure 8-4, the closest major roadway is County Road 331, Dry Hollow Road.
8-12

-------
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.
The surrounding area is considered residential. Interstate-70 is less than a quarter of a mile from
the monitoring site, as shown in Figure 8-5. PACO is located 1.6 miles from BMCO, which are
the Garfield County sites that are the closest to each other.
RICO is located on the roof of the Henry Annex Building in downtown Rifle. This
location is near the crossroads of several major roadways through town, as shown in Figure 8-6.
Highway 13 and US-6/24 intersect just south of the site and 1-70 is just over a half-mile south of
the monitoring site, across the Colorado River. The surrounding area is 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 8-7. 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 air quality in the
surrounding areas (GCPH, 2014).
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, 2015). 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 8-8.
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 8-9. This figure
shows that the few emissions sources within 10 miles of RFCO are primarily gasoline and/or
diesel service stations. There is also a building/construction company, a compressor station, three
mine/quarry/mineral processing facilities, and an airport within a few miles of this site.
8-13

-------
Table 8-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Colorado monitoring sites. Table 8-2 includes both county-level
population and vehicle registration information. Table 8-2 also contains traffic volume
information for each site as well as the location for which the traffic volume was obtained.
Additionally, Table 8-2 presents the county-level daily VMT for Mesa and Garfield Counties.
Because VMT from the state of Colorado is available for state highways only, VMT presented in
this table is from the 2011 NEI, version 2.
Table 8-2. Population, Motor Vehicle, and Traffic Information for the Colorado
Monitoring Sites


Estimated
County-level
Annual
Intersection



County
Vehicle
Average Daily
Used for
County-level
Site
County
Population1
Registration2
Traffic3
Traffic Data
Daily VMT4
GPCO
Mesa
147,554
176,969
11,000
Bus-70 (Pitkin Ave) just E of 7th St
3,355,813
BMCO
Garfield
57,302
74,036
1,880
S Battlement Pkwy
2,171,019
BRCO
1,182
Dry Hollow Rd
PACO
15,000
1-70 near exit 75
RFCO
16,000
Rt 133 just south of Hwy 82
RICO
15,000
Rt 13 connecting US-6 and 1-70
bounty-level population estimates reflect 2013 data (Census Bureau, 2014)
2County-level vehicle registrations reflect 2012 data (CO DOR, 2013)
3AADT reflects 2013 data for GPCO, PACO, RFCO, and RICO (CO DOT, 2013a) and 2014 data for BMCO and BRCO
(GCRBD, 2014)
4County-level VMT reflects 2011 data (EPA, 2015a)
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 8-2 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 more than 1 million miles separate the Mesa County and Garfield County
VMTs, as obtained from the 2011 NEI, version 2, they are also both in the bottom-
third among VMTs for counties with NMP sites.
8-14

-------
8.2 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.
8.2.1	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 sunny and 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 canyons, the
Colorado River and its tributaries, and valley breezes (GCPH, 2014; WRCC, 2014).
8.2.2	Meteorological Summary
Hourly meteorological data for 2013 were retrieved from NCDC for the weather stations
closest to the Colorado monitoring sites (NCDC, 2013), as described in Section 3.4.2. 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 8-3. These data were
used to determine how meteorological conditions on sample days vary from conditions
experienced throughout the year.
8-15

-------
Table 8-3. 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/Grand
Junction Regional
Airport
23066
(39.13, -108.54)
5.0
miles
Sample
Day
(67)
61.3
±6.1
49.2
±5.7
27.1
±3.4
38.5
±4.0
52.7
±6.0
1016.5
±2.5
6.0
±0.7
13°
(NNE)
2013
61.7
±2.5
49.9
±2.3
28.7
± 1.5
39.6
± 1.7
54.0
±2.5
1016.3
± 1.0
5.9
±0.3
Battlement Mesa, Colorado - BMCO
Garfield County
Regional Airport
17.7
miles
Sample
Day
(58)
63.0
±6.1
48.4
±5.5
28.0
±3.5
38.4
±4.0
53.8
±4.8
1017.0
±2.4
4.9
±0.8
03016
(39.53, -107.72)
70°
(ENE)
2013
60.6
±2.4
46.8
±2.2
27.5
± 1.5
37.5
± 1.6
55.3
± 1.9
1017.8
±0.9
4.6
±0.3




Silt, Colorado
- BRCO




Garfield County
Regional Airport
4.2
miles
Sample
Day
(62)
62.2
±6.0
47.5
±5.4
27.3
±3.5
37.7
±4.0
53.8
±4.5
1017.6
±2.4
4.8
±0.7
03016
(39.53, -107.72)
311°
(NW)
2013
60.6
±2.4
46.8
±2.2
27.5
± 1.5
37.5
± 1.6
55.3
± 1.9
1017.8
±0.9
4.6
±0.3
Parachute, Colorado - PACO
Garfield County
Regional Airport
18.6
miles
Sample
Day
(56)
64.3
±6.1
49.0
±5.6
27.9
±3.6
38.7
±4.1
52.5
±4.7
1017.8
±2.4
4.7
±0.7
03016
(39.53, -107.72)
74°
(ENE)
2013
60.6
±2.4
46.8
±2.2
27.5
± 1.5
37.5
± 1.6
55.3
± 1.9
1017.8
±0.9
4.6
±0.3
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.

-------
Table 8-3. Average Meteorological Conditions near the Colorado Monitoring Sites (Continued)
Closest NWS
Distance
and

Average
Maximum
Average
Average
Dew Point
Average
Wet Bulb
Average
Relative
Average
Sea Level
Average
Scalar Wind
Station (WBAN
and Coordinates)
Direction
from Site
Average
Type1
Temperature
(°F)
Temperature
(°F)
Temperature
(°F)
Temperature
(°F)
Humidity
(%)
Pressure
(mb)
Speed
(kt)
Rifle, Colorado - RICO
Garfield County
3.4
miles
Sample
Day
61.7
46.9
26.8
37.2
53.8
1017.7
4.8
Regional Airport
(59)
±6.1
±5.5
±3.5
±4.0
±4.7
±2.4
±0.7
03016
95°
(E)








(39.53, -107.72)

60.6
46.8
27.5
37.5
55.3
1017.8
4.6

2013
±2.4
±2.2
± 1.5
± 1.6
± 1.9
±0.9
±0.3
Carbondale, Colorado - RFCO
Aspen-Pitkin
23.0
miles
Sample
Day
56.6
41.8
23.1
33.3
54.0
1017.8
5.1
County Airport
(31)
±7.4
±6.8
±5.0
±5.3
±5.3
±2.9
±0.4
93073
123°
(ESE)








(39.23, -106.87)

53.2
39.9
23.6
32.6
59.0
1016.1
4.9

2013
±2.1
± 1.9
± 1.5
± 1.5
± 1.8
±0.8
±0.2
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.

-------
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 2013. 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 near GPCO were representative of average weather conditions
experienced throughout the year. The parameter with the largest difference between the full-year
average and the sample day average for GPCO is dew point temperature.
Of the four Garfield County sites for which Garfield County Regional Airport is the
closest weather station, BMCO and PACO have the fewest sample days. Both of these sites
missed two sample days in January. BMCO also missed a sample day in March. PACO missed
two sample days in February and another in December. This may explain why temperatures on
sample days appear slightly higher at these sites compared to BRCO and RICO as all of the
missed sample dates are during colder months of the year.
RFCO 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 roughly half the number for the other Colorado sites. As a result, there is a higher level of
variability in the meteorological averages for this site, as indicated by the confidence intervals
shown. For RFCO, the temperature parameters on sample days appear higher than those shown
for the entire year. In addition, relative humidity levels were lower on sample days than over the
entire year, as shown in Table 8-3.
The average temperature shown for RFCO for 2013 is the second lowest average
temperature among NMP sites, behind only the Anchorage, Alaska site (ANAK). All six
Colorado sites account for the six lowest average dew point temperatures among NMP sites;
RFCO also has the lowest average wet bulb temperature. These sites also experienced some of
the lowest relative humidity levels among NMP sites.
8-18

-------
8.2.3 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 3.4.2. A wind rose shows the frequency of wind
directions using "petals" positioned around a 16-point compass, and uses different colors to
represent wind speeds.
Figure 8-10 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 8-10 also presents three different wind roses for the
GPCO monitoring site. First, a historical wind rose representing 2003 to 2012 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 2013 is presented.
Next, a wind rose representing wind observations for days on which samples were collected in
2013 is presented. These can be used to identify the predominant wind speed and direction for
2013 and to determine if wind observations on sample days were representative of conditions
experienced over the entire year and historically. Figures 8-11 through 8-15 present the distance
maps and wind roses for the five Garfield County sites.
Observations from Figure 8-10 for GPCO include the following:
•	The Walker Field Airport weather station is located 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 and east-southeasterly winds were
prevalent near GPCO over the last 10 years. Winds from the east-northeast to south-
southeast account for nearly half of the wind observations 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 (those less than or equal to 2 knots) were observed for approximately
16 percent of the hourly wind measurements.
•	The 2013 wind rose exhibits similar wind patterns as the historical wind rose. 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.
8-19

-------
Figure 8-10. Wind Roses for the Grand Junction Regional Airport Weather Station near
GPCO
Location of GPCO and Weather Station
2003-2012 Historical Wind Rose
8tMtoa

\\ I , :J
g 1
H


-xk
{GPCO
4-1
EST
VWVlD SPEED
(Kn ots}
SOUTH
2013 Wind Rose
VEST
WWD SPEED
[Kn ots >
SOUTH
~ 4-7
H 2-4
Calms: 19.40%
Sample Day Wind Rose
WEST
WIN D S PE ED
(Kn ots)
SOUTH
Calms: 18.10%
8-20

-------
EST!
Figure 8-11. Wind Roses for the Garfield County Regional Airport Weather Station near
BMCO
Location of BMCO and Weather Station	2003-2012 Historical Wind Rose
WIND SPEED
(Knots)
I I »22
[ B 17-21
IH 11 -1?
I I 7- 11
~ 4-7
2- 4
Calms: 35.02%
2013 Wind Rose
Sample Day Wind Rose
WESTl
west!
WIND SPEED
(Knots)
I I =-22
I 11 17-21
WIND SPEED
(Knots)
~ -22
I B 17-21
SOUTH
8-21

-------
EST!
Figure 8-12. Wind Roses for the Garfield County Regional Airport Weather Station near
BRCO
Location of BRCO and Weather Station	2003-2012 Historical Wind Rose
WIND SPEED
(Knots)
I I »22
[ B 17-21
IH 11 -1?
I I 7- 11
~ 4-7
2- 4
Calms: 35.02%
2013 Wind Rose
Sample Day Wind Rose
WESTl
west!
WIND SPEED
(Knots)
I I =-22
I 11 17-21
WIND SPEED
(Knots)
~ -22
I B 17-21
SOUTH
8-22

-------
Figure 8-13. Wind Roses for the Garfield County Regional Airport Weather Station near
PACO
Location of PACO and Weather Station
2003-2012 Historical Wind Rose
NORTH"''-.
I EA

WWC SPEED
(Knots)
17-21
11 - 17
SOUTH
Cams:
2013 Wind Rose
Sample Day Wind Rose
NORTH"*--,
15%
12%

WIND SPEED
(Knots)
~ >=22
HI 17-21
11-17
I 1 7- 11
I I 4-7
2- 4
Calms: 34.31%
NORTH"*--,
15%
12%
WIND SPEED
(Knots)
~	>=22
Ml	17-21
[H	11 - 17
I I	7- 11
~1	4-7
2- 4
Calms: 34.83%
8-23

-------
Figure 8-14. Wind Roses for the Garfield County Regional Airport Weather Station near
RICO
Location of RICO and Weather Station
2003-2012 Historical Wind Rose


NORTH-^-.


15%


12%


9%s


6%,
^3% : :
"KO 	 *4——


,1, mtttir-
" Stalk*




SOUTH
: 	

WIND SPEED
(Knots)
I I »=22
~ 17-21
IH 11 1?
I -I 7- 11
~ 4-7
2- 4
Calms: 35.02%
2013 Wind Rose
Sample Day Wind Rose
WEST!
WIND SPEED
(Knots)
~
I B 17-21
WIND SPEED
(Knots)
~ >=22
@| 17-21
|| 11-17
SOUTH
SOUTH
8-24

-------
WEST
Figure 8-15. Wind Roses for the Aspen-Pitkin County Airport Weather Station near RFCO
Location of RFCO and Weather Station	2003-2012 Historical Wind Rose
: EAS
WIND SPEED
(Kn ots)
~
~ 4-7
¦ 2-4
Calms: 18.85%
2013 Wind Rose
NORTH"
WEST
(Knots)
SOUTH
WIND SPEED
Sample Day Wind Rose
NORTH""-.,

WIND SPEED
(Kn ots)
~
I I 17-21
|H 11 -17
r i 7-11
TD 4-7
2- 4
Calms: 10.80%
8-25

-------
Observations from Figures 8-11 through 8-14 for BMCO, BRCO, PACO, and RICO,
respectively, 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 18 miles (PACO).
•	The historical and 2013 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 more than one-third of wind observations. For wind speeds greater
than 2 knots, westerly winds were prevalent, followed by southerly and south-
southwesterly winds. Winds from the northeast quadrant were rarely observed.
•	Calm winds were observed for 34 percent of the wind observations in 2013. Westerly
winds were again prevalent in 2013, although fewer southerly and south-
southwesterly winds were observed in 2013 near the Garfield County sites compared
to the historical wind rose. A similar observation was made in the 2011 and 2012
NMP reports.
•	The sample day wind patterns for each site resemble the full-year wind patterns,
indicating that conditions on sample days were representative of those experienced
over the entire year. With the exception of PACO, each Garfield County site's sample
day wind rose shows that a slightly higher number of west-northwesterly winds were
observed on sample days compared to all of 2013.
Observations from Figure 8-15 for RFCO include the following:
•	The Aspen-Pitkin County Airport weather station is located 23 miles east-southeast of
RFCO. The mountainous terrain surrounding the site and weather station is visible in
Figure 8-15.
•	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 roughly
20 percent of wind observations. Calm winds account for just less than one-fifth of
observations. Winds from the north-northeast to east-southeast and west-southwest to
northwest were rarely observed. The wind flow tends to follow the orientation of the
valley, similar to the wind observations near the other Garfield County sites.
•	The 2013 wind rose exhibits similar wind patterns as the historical wind rose,
indicating that conditions in 2013 were similar to conditions experienced over the last
10 years.
8-26

-------
• The sample day wind rose has similar wind patterns as the full-year and historical
wind roses, but has a higher percentage of south-southwesterly winds, accounting for
nearly 25 percent of observations on sample days.
8.3 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 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. VOCs, carbonyl compounds, PAHs, and hexavalent chromium were
sampled for at GPCO while SNMOCs and carbonyl compounds were sampled for at the Garfield
County sites.
8-27

-------
Table 8-4. Risk-Based Screening Results for the Colorado Monitoring Sites
Pollutant
Screening
Value
Oig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Grand Junction, Colorado - GPCO
Benzene
0.13
61
61
100.00
12.82
12.82
Carbon Tetrachloride
0.17
60
60
100.00
12.61
25.42
1.3 -Butadiene
0.03
59
61
96.72
12.39
37.82
Acetaldehyde
0.45
58
58
100.00
12.18
50.00
Formaldehyde
0.077
58
58
100.00
12.18
62.18
Naphthalene
0.029
54
56
96.43
11.34
73.53
1,2-Dichloroethane
0.038
45
45
100.00
9.45
82.98
Ethylbenzene
0.4
32
61
52.46
6.72
89.71
Acenaphthene
0.011
14
56
25.00
2.94
92.65
Hexachloro-1,3 -butadiene
0.045
11
13
84.62
2.31
94.96
Benzo(a)pyrene
0.00057
8
37
21.62
1.68
96.64
Fluorene
0.011
8
54
14.81
1.68
98.32
Acenaphthylene
0.011
3
36
8.33
0.63
98.95
Propionaldehyde
0.8
3
58
5.17
0.63
99.58
Dichloromethane
60
2
61
3.28
0.42
100.00
Total
476
775
61.42

Battlement Mesa, Colorado - BMCO
Benzene
0.13
54
54
100.00
55.67
55.67
Formaldehyde
0.077
26
28
92.86
26.80
82.47
Acetaldehyde
0.45
11
27
40.74
11.34
93.81
1.3 -Butadiene
0.03
3
3
100.00
3.09
96.91
Ethylbenzene
0.4
3
50
6.00
3.09
100.00
Total
97
162
59.88

Silt, Colorado - BRCO
Benzene
0.13
57
57
100.00
55.88
55.88
Formaldehyde
0.077
26
26
100.00
25.49
81.37
Acetaldehyde
0.45
16
26
61.54
15.69
97.06
1,3-Butadiene
0.03
2
2
100.00
1.96
99.02
Ethylbenzene
0.4
1
48
2.08
0.98
100.00
Total
102
159
64.15

Parachute, Colorado - PACO
Benzene
0.13
49
49
100.00
45.37
45.37
Formaldehyde
0.077
26
26
100.00
24.07
69.44
Acetaldehyde
0.45
19
25
76.00
17.59
87.04
1,3-Butadiene
0.03
9
9
100.00
8.33
95.37
Ethylbenzene
0.4
5
52
9.62
4.63
100.00
Total
108
161
67.08

8-28

-------
Table 8-4. Risk-Based Screening Results for the Colorado Monitoring Sites (Continued)
Pollutant
Screening
Value
Oig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Carbondale, Colorado - RFCO
Formaldehyde
0.077
25
27
92.59
36.23
36.23
Benzene
0.13
23
23
100.00
33.33
69.57
Acetaldehyde
0.45
15
27
55.56
21.74
91.30
1,3-Butadiene
0.03
6
6
100.00
8.70
100.00
Total
69
83
83.13

Rifle, Colorado - RICO
Benzene
0.13
57
57
100.00
33.53
33.53
1,3-Butadiene
0.03
52
52
100.00
30.59
64.12
Formaldehyde
0.077
24
25
96.00
14.12
78.24
Acetaldehyde
0.45
19
25
76.00
11.18
89.41
Ethylbenzene
0.4
18
57
31.58
10.59
100.00
Total
170
216
78.70

Observations from Table 8-4 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 the sites.
•	Fifteen pollutants failed at least one screen for GPCO; 61 percent of the
concentrations for these 15 pollutants were greater than their associated risk screening
value (or failed screens).
•	Twelve pollutants contributed to 95 percent of failed screens for GPCO and therefore
were identified as pollutants of interest for GPCO. These 12 include two carbonyl
compounds, six VOCs, and four PAHs. Although the 95 percent criteria is met with
benzo(a)pyrene, fluorene is also considered a pollutant of interest because it failed the
same number of screens as benzo(a)pyrene, per the steps described in Section 3.2.
•	The number of pollutants failing screens for the Garfield County sites range from four
(RFCO) to five (the remaining sites). The same four pollutants (benzene,
1,3-butadiene, formaldehyde, and acetaldehyde) failed screens for each Garfield
County site. Ethylbenzene also failed screens for all sites except RFCO.
•	Benzene, formaldehyde, and acetaldehyde were identified as pollutants of interest for
all five Garfield County sites. 1,3-Butadiene was also identified as a pollutant of
interest for all five sites except BRCO, and ethylbenzene was also identified as a
pollutant of interest for BMCO and RICO.
•	Benzene failed 100 percent of screens for all six Colorado sites.
8-29

-------
•	Note that carbonyl compound samples were collected on a l-in-12 day sampling
schedule at BMCO, BRCO, PACO, and RICO, while SNMOC samples were
collected on a 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 compound and SNMOC samples were collected on a l-in-12
day sampling schedule atRFCO.
8.4 Concentrations
This section presents various concentration averages used to characterize air toxics
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. Site-specific statistical summaries for all pollutants
sampled at the Colorado monitoring sites are provided in Appendices J through M and O.
8.4.1 2013 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 8-5, where applicable. Note that concentrations
8-30

-------
of the PAHs for GPCO are presented in ng/m3 for ease of viewing. Also note that if a pollutant
was not detected in a given calendar quarter, the quarterly average simply reflects "0" because
only zeros substituted for non-detects were factored into the quarterly average concentration.
Table 8-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest
for the Colorado Monitoring Sites

# of






Measured
1st
2nd
3rd
4th


Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs. # of
Samples
Average
Oig/m3)
Average
(Hg/m3)
Average
(Hg/m3)
Average
(Hg/m3)
Average
(Hg/m3)
Grand Junction, Colorado - GPCO


3.88
5.78
2.56
2.86
3.79
Acetaldehyde
58/58
±0.69
± 1.40
±0.99
±0.56
±0.57


1.35
0.59
0.59
1.42
0.99
Benzene
61/61
±0.38
±0.07
±0.10
±0.32
±0.16


0.18
0.07
0.08
0.26
0.15
1.3 -Butadiene
61/61
±0.06
±0.02
±0.02
±0.06
±0.03


0.52
0.61
0.65
0.56
0.59
Carbon Tetrachloride
60/61
±0.10
±0.04
±0.03
±0.03
±0.03


0.07
0.08
0.05
0.07
0.07
1,2-Dichloroethane
45/61
±0.02
±0.02
±0.02
±0.02
±0.01


0.52
0.33
0.41
0.67
0.49
Ethylbenzene
61/61
±0.16
±0.06
±0.13
±0.13
±0.07


5.30
11.77
4.70
3.72
6.44
Formaldehyde
58/58
±0.61
±2.85
±2.14
±0.66
± 1.22


0.03
0.01
0.01
0.02
0.02
Hexachloro-1,3 -butadiene
13/61
±0.03
±0.01
±0.01
±0.02
±0.01


4.07
9.99
13.85
4.30
8.05
Acenaphthene3
56/56
± 1.41
±3.98
±3.89
± 1.53
± 1.77


0.44
0.02
0.01
0.47
0.24
Benzo(a)pyrenea
37/56
±0.28
±0.01
±0.01
±0.20
±0.10


4.72
7.49
9.70
5.63
6.88
Fluorene3
54/56
± 1.42
±2.69
±2.52
± 1.22
± 1.09


162.62
100.86
108.54
175.72
136.93
Naphthalene3
56/56
±68.12
±30.39
± 26.77
± 45.42
±23.05
Battlement Mesa, Colorado - BMCO


0.41
0.50
0.54
0.37
0.46
Acetaldehyde
27/28
±0.23
±0.18
±0.36
±0.22
±0.11



1.00
0.98
1.74
1.26
Benzene
54/55
NA
±0.10
±0.13
±0.57
±0.19





0.01
<0.01
1.3 -Butadiene
3/55
NA
0
0
±0.02
±<0.01



0.10
0.07
0.21
0.14
Ethylbenzene
50/55
NA
±0.02
±0.02
±0.08
±0.03


0.64
0.83
1.20
0.56
0.82
Formaldehyde
28/28
±0.29
±0.09
±0.88
±0.27
±0.23
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.
8-31

-------
Table 8-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest
for the Colorado Monitoring Sites (Continued)

# of






Measured
1st
2nd
3rd
4th


Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs. # of
Samples
Average
(Ug/m3)
Average
(Ug/m3)
Average
(Ug/m3)
Average
(Ug/m3)
Average
(Ug/m3)

Silt, Colorado
- BRCO





0.52
0.62
0.70
0.39
0.56
Acetaldehyde
26/26
±0.23
±0.18
±0.45
±0.16
±0.12


1.50

0.82
1.23
1.14
Benzene
57/57
±0.51
NA
±0.12
±0.47
±0.20


0.75
0.87
1.28
0.62
0.87
Formaldehyde
26/26
±0.34
±0.18
±0.62
±0.27
±0.18
Parachute, Colorado - PACO


0.86
0.87

0.53
0.76
Acetaldehyde
25/26
±0.49
±0.24
NA
±0.35
±0.18



1.19
1.62
2.56
1.96
Benzene
49/52
NA
±0.28
±0.46
±0.60
±0.31



<0.01

0.03
0.01
1,3-Butadiene
9/52
NA
±0.01
0
±0.03
±0.01


1.35
1.38
1.28
1.08
1.28
Formaldehyde
26/26
±0.67
±0.13
±0.69
±0.69
±0.25
Carbondale, Colorado - RFCO


0.61
0.88
0.60
0.28
0.58
Acetaldehyde
27/27
±0.43
±0.42
±0.53
±0.16
±0.19


0.75
0.40

0.55
0.57
Benzene
23/29
±0.30
±0.10
NA
±0.24
±0.12


0.04
0.01
0.01
0.02
0.02
1,3-Butadiene
6/29
±0.04
±0.01
±0.03
±0.04
±0.01


0.79
0.98
0.98
0.35
0.75
Formaldehyde
27/27
±0.55
±0.36
±0.59
±0.20
±0.21
Rifle, Colorado - RICO


1.16
1.07
0.96
0.56

Acetaldehyde
25/25
±0.60
±0.16
±0.50
±0.45
NA


2.07
1.00
0.96
2.11
1.52
Benzene
57/57
±0.55
±0.12
±0.11
±0.70
±0.26


0.14
0.06
0.07
0.18
0.11
1,3-Butadiene
52/57
±0.04
±0.01
±0.03
±0.05
±0.02


0.38
0.29
0.29
0.45
0.35
Ethylbenzene
57/57
±0.08
±0.05
±0.04
±0.13
±0.04


1.56
1.19
2.00
0.87

Formaldehyde
25/25
±0.67
±0.16
± 1.12
±0.82
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.
Observations for GPCO from Table 8-5 include the following:
• The pollutants with the highest annual average concentrations for GPCO are
formaldehyde (6.44 ± 1.22 |ig/m3) and acetaldehyde (3.79 ± 0.57 |ig/m3). These are
also the only pollutants with annual average concentrations greater than 1 |ig/m3,
although benzene is very close (0.99 ± 0.16 |ig/m3). The annual average
8-32

-------
concentrations for these carbonyl compounds are considerably higher for 2013 than
they were for 2012.
The second quarter average concentrations for both acetaldehyde and formaldehyde
are considerably higher than the other quarterly averages and have relatively large
confidence intervals associated with them, particularly for formaldehyde. A review of
the data shows that all but one of the 13 formaldehyde concentrations greater than
7.50 |ig/m3 were measured at GPCO during the second quarter of 2013 (with the
exception being measured on the first sample day in the third quarter). These
measurements range from 7.74 |ig/m3 to 21.9 |ig/m3. The five highest concentrations
of formaldehyde measured across the program were all measured at GPCO (between
June 3, 2013 and July 3, 2013). Similar observations can be made for acetaldehyde
concentrations measured at GPCO. All but one of the six acetaldehyde concentrations
greater than 7.00 |ig/m3 were measured at GPCO during the second quarter of 2013
(with the exception being measured on the first sample day in the third quarter).
These measurements range from 7.00 |ig/m3 to 10.7 |ig/m3. Although the maximum
acetaldehyde concentration measured across the program was not measured at GPCO,
concentrations of acetaldehyde measured at GPCO account for the next five highest
acetaldehyde concentrations (and were measured in samples collected between
June 9, 2013 and July 3, 2013).
Concentrations of benzene and 1,3-butadiene appear highest during the colder months
of the year, based on the quarterly averages shown in Table 8-5. A review of the data
shows that all 18 of GPCO's benzene concentrations greater than 1 |ig/m3 were
measured during the first or fourth quarters of 2013. Conversely, all nine benzene
concentrations less than 0.5 |ig/m3 were measured during the second or third quarters
of 2013. Similarly, all 16 of GPCO's 1,3-butadiene concentrations greater than
0.2 |ig/m3 were measured during the first or fourth quarters of 2013. Other pollutants
of interest for GPCO exhibiting a similar tendency include ethylbenzene.
Of the PAH pollutants of interest, naphthalene has the highest annual average
concentration while benzo(a)pyrene has the lowest.
Concentrations of acenaphthene appear considerably higher during the warmer
months of the year, based on the quarterly averages shown in Table 8-5. A review of
the data shows that the nine highest concentrations of acenaphthene (those greater
than 15 ng/m3) were measured between June and August. A similar tendency is
shown for fluorene. The four highest concentrations of each of these pollutants were
measured on the same days in June and July.
Conversely, concentrations of benzo(a)pyrene appear higher during the colder months
of the year. A review of the data shows that all 15 concentrations of benzo(a)pyrene
greater than 0.3 ng/m3 were measured at GPCO during the first or fourth quarters of
the year, including the five measurements greater than 1 ng/m3. Conversely, all but
one of GPCO's 19 non-detects were measured during the second or third quarters of
2013.
8-33

-------
•	The maximum benzo(a)pyrene concentration measured at GPCO (1.49 ng/m3) is the
second highest benzo(a)pyrene concentration measured across the program. Further,
this site has the highest number of benzo(a)pyrene concentrations greater than
1 ng/m3 (five); no other NMP site sampling PAHs has more than one.
•	Although naphthalene concentrations also appear to be highest during the colder
months of the year at GPCO, the confidence intervals also indicate that there is
considerable variability in these measurements. A review of the naphthalene data
shows that concentrations measured at GPCO range from 27.2 ng/m3 to 368 ng/m3,
with a median concentration of 100 ng/m3. GPCO is one of only five NMP sites to
measure a naphthalene concentration greater than 350 ng/m3 and is one of only two
NMP sites to measure more than one of these higher concentrations (NBIL is the
other). All of GPCO's naphthalene concentrations greater than 250 ng/m3 were
measured during the first or fourth quarters of 2013. However, three of the four
lowest naphthalene concentrations measured at GPCO were also measured during the
first or fourth quarters of 2013.
Observations for the Garfield County sites from Table 8-5 include the following:
•	Acetaldehyde, benzene, and formaldehyde are pollutants of interest for each Garfield
County site. However, annual average concentrations of the carbonyl compounds
could not be calculated for RICO because the completeness for this method at this site
is less than 85 percent, as discussed in Section 2.4.
•	With the exception of RFCO, benzene has the highest annual average concentration
among the pollutants of interest for the Garfield County sites. Among the Garfield
County sites, annual average concentrations of benzene range from 0.57 ± 0.12 |ig/m3
(RFCO) to 1.93 ± 0.32 |ig/m3 (PACO).
•	RICO is the only Garfield County site for which four quarterly average
concentrations of benzene are available in Table 8-5. The quarterly average
concentrations for RICO show that benzene concentrations tended to be higher during
the colder months of the year, similar to the findings for GPCO.
•	Among the Garfield County sites, annual average concentrations of formaldehyde
range from 0.75 ± 0.21 |ig/m3 (RFCO) to 1.28 ± 0.25 |ig/m3 (PACO), where they
could be calculated. The Garfield County sites' annual averages of formaldehyde are
lower than the annual average for GPCO. Further, these sites' annual average
formaldehyde concentrations are among the lowest for NMP sites sampling carbonyl
compounds, as shown in Figure 4-10 in Section 4. BMCO, BRCO, and RFCO are the
only sites, in addition to SEW A, with annual average concentrations of formaldehyde
less than 1 |ig/m3. Similar observations can be made for acetaldehyde for the Garfield
County sites.
•	1,3-Butadiene was identified as a pollutant of interest for all of the Garfield County
sites except BMCO, although the detection rate of this pollutant varied significantly.
1,3-Butadiene was detected in 5 percent of samples collected at BMCO, 17 percent at
PACO, 21 percent at RFCO, and 91 percent at RICO, which is the closest to GPCO's
8-34

-------
100 percent detection rate. The annual average 1,3-butadiene concentrations for
GPCO and PACO are similar to each other. Quarterly average concentrations of
1,3-butadiene for RICO exhibit a similar seasonal tendency as GPCO's quarterly
averages of 1,3-butadiene. Note that GPCO samples were analyzed with Method TO-
15 while PACO's samples were analyzed with the SNMOC method.
Tables 4-9 through 4-12 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for the Colorado
sites from those tables include the following:
•	Annual average concentrations for the Colorado sites appear in Tables 4-9 through
4-12 a total of 10 times, with GPCO having the most (6).
•	PACO has the highest annual average concentration of benzene among all NMP sites
sampling this pollutant. All of the Garfield County sites rank the in the top 10 for
benzene in Table 4-9, with the exception of RFCO, which ranks 28th. None of the
Garfield County sites appear in Table 4-9 for any of the other pollutants listed.
•	GPCO's annual average concentrations of acetaldehyde and formaldehyde both rank
second highest among NMP sites sampling carbonyl compounds, as shown in
Table 4-10. Note that the confidence intervals calculated for the annual averages for
GPCO for both pollutants are the highest of those shown in Table 4-10.
•	GPCO has the second highest annual concentration of naphthalene and the fourth
highest annual concentration of acenaphthene among all NMP sites sampling PAHs,
as shown in Table 4-11. GPCO had the highest annual average concentration of
naphthalene in the 2010, 2011, and 2012 NMP reports.
8.4.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for 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 8-4 for each site. Note that the box plots for benzene, 1,3-butadiene, and
ethylbenzene 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 8-16 through 8-27 overlay the
sites' minimum, annual average, and maximum concentrations onto the program-level minimum,
first quartile, median, average, third quartile, and maximum concentrations, as described in
Section 3.4.3.1.
8-35

-------
Figure 8-16. Program vs. Site-Specific Average Acenaphthene Concentration




O i
Program Max Concentration = 123 ng/m3
,
u 1
1
1 1 1 1
1
0
10
20
30 40 50
Concentration {ng/m3)
60
70
80

Program:
Site:
1st Quartile
¦
Site Average
o
2nd Quartile 3rd Quartile
~ ~
Site Concentration Range
4th Quartile
~
Average
i

Figure 8-17. Program vs. Site-Specific Average Acetaldehyde Concentrations
id
0
3
6 9
Concentration {[jg/m3)

12

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i

Site: Site Average
o
Site Concentration Range


8-36

-------
Figure 8-18a. Program vs. Site-Specific Average Benzene (Method TO-15) Concentration


¦-
-o	¦
Program Max Concentration = 43.5 ^ig/m3

0	2	4	6	8	10	12
Concentration {[ig/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


Figure 8-18b. Program vs. Site-Specific Average Benzene (SNMOC) Concentrations
" ¦
D		
m ¦
	0		
K-f

h-«
	0		
0	1	2	3	4	5	6
Concentration {[jg/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


8-37

-------
Figure 8-19. Program vs. Site-Specific Average Benzo(a)pyrene Concentration
0.25
0.5
0.75 1 1.25
Concentration {ng/m3)
1.5 1.75
2
Program:
1st Quartile
¦
2nd Quartile 3rd Quartile
~ ~
4th Quartile
~
Average
i
Site:
Site Average
o
Site Concentration Range


Figure 8-20a. Program vs. Site-Specific Average 1,3-Butadiene (Method TO-15)
Concentration
E
Program Max Concentration = 21.5 ^ig/m3
0.6	0.9
Concentration {[jg/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



8-38

-------
Figure 8-20b. Program vs. Site-Specific Average 1,3-Butadiene (SNMOC) Concentrations
BMCO
o
PACO
-a
RFCO

0.3
Concentration {[jg/m3
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


Figure 8-21. Program vs. Site-Specific Average Carbon Tetrachloride Concentration
Program Max Concentration = 23.7 ^ig/m3
GPCO
0	0.25	0.5	0.75	1	1.25	1.5	1.75	2
Concentration {[ig/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


8-39

-------
Figure 8-22. Program vs. Site-Specific Average 1,2-Dichloroethane Concentration


¦
Program Max Concentration = 111 ^ig/m3


0	0.2	0.4	0.6	0.8	1
Concentration {[ig/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


Figure 8-23a. Program vs. Site-Specific Average Ethylbenzene (Method TO-15) Concentration














Program Max Concentration = 18.7 ^ig/m3











0	1	2	3	4	5	6
Concentration {[ig/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


Figure 8-23b. Program vs. Site-Specific Average Ethylbenzene (SNMOC) Concentrations
E
3	4
Concentration (|jg/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


8-40

-------
Figure 8-24. Program vs. Site-Specific Average Fluorene Concentration
40	50	60
Concentration {ng/m3)
Program: IstQuartile
2ndQuartile 3rdQuartile
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



Figure 8-25. Program vs. Site-Specific Average Formaldehyde Concentrations
w±
E
¦
E
D
0
3 6
9 12 15
Concentration {[jg/m3)
18
21

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i

Site: Site Average
o
Site Concentration Range


8-41

-------
Figure 8-26. Program vs. Site-Specific Average Hexachloro-l,3-Butadiene Concentration
0.05
0.1
0.15
Concentration {[jg/m3)
0.2
0.25
Program:
1st Qua rti le
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i
Site:
Site Average
o
Site Concentration Range


Figure 8-27. Program vs. Site-Specific Average Naphthalene Concentration
i
0
100
200
300 400 500
Concentration {ng/m3)
600
700

Program:
1st Qua rti le
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i

Site:
Site Average
o
Site Concentration Range


Observations from Figures 8-16 through 8-27 include the following:
•	Figure 8-16 is the box plot for acenaphthene for GPCO. The program-level
maximum concentration (123 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
80 ng/m3. The maximum concentration of acenaphthene measured at GPCO is
roughly one-fourth the magnitude of the program-level maximum concentration.
However, the annual average acenaphthene concentration for GPCO is greater
than the program-level average concentration and is the fourth highest annual
average concentration among NMP sites sampling this pollutant.
•	Figure 8-17 presents the acetaldehyde box plots for the five Colorado sites for
which annual averages could be calculated. The box plots show that the range of
acetaldehyde concentrations measured at GPCO is considerably larger than the
range of measurements collected at the Garfield County sites. Not surprisingly,
GPCO has the highest annual average acetaldehyde concentration among the
Colorado sites. The annual average for GPCO is more than five times greater than
the next highest annual average acetaldehyde concentration for a Garfield County
site (PACO), and is more than twice the program-level average concentration.
The minimum acetaldehyde concentration measured at GPCO is greater than the
8-42

-------
annual average concentrations for all of the Garfield County sites, an observation
that was also made in the 2012 NMP report. The maximum acetaldehyde
concentration for each Garfield County site is less than the program-level average,
with the exception of PACO.
•	Figures 8-18a and 8-18b present the box plots for benzene. Figure 8-18a
compares to the benzene concentrations measured at GPCO to those measured
across the program for NMP sites sampling VOCs with Method TO-15;
Figure 8-18b 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 3.4.3.1. Note that the scales are not the same in the figures.
•	The program-level maximum concentration (43.5 |ig/m3) is not shown directly on
the box plot in Figure 8-18a 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 12 |ig/m3. Figure 8-18a shows that the annual
average benzene concentration for GPCO is slightly higher than the program-level
average concentration as well as the third quartile for the program. The maximum
benzene concentration measured at GPCO is considerably less than the maximum
benzene concentration measured across the program.
•	Figure 8-18b includes a box plot for all five Garfield County sites. The maximum
benzene concentration measured at PACO is the maximum benzene concentration
measured among the seven sites sampling SNMOCs (5.45 |ig/m3). Of the Garfield
County sites, PACO has the highest annual average concentration of benzene,
followed by RICO then BMCO, BRCO, and RFCO. The range of benzene
concentrations measured at RFCO is considerably smaller than the ranges shown
for the other Garfield County sites. This sites annual average benzene
concentration is less than the program-level first quartile.
•	Figure 8-19 is the box plot for benzo(a)pyrene for GPCO. Note that the program-
level first quartile is zero and therefore not visible on the box plot. Although the
maximum benzo(a)pyrene concentration measured across the program was not
measured at GPCO, this site's maximum benzo(a)pyrene concentration is the
second highest concentration measured among NMP sites sampling PAHs. The
annual average benzo(a)pyrene concentration for GPCO is more than three times
the program-level average concentration. Note that GPCO is one of only two
NMP sites for which benzo(a)pyrene is a pollutant of interest.
•	Similar to the box plots for benzene, Figure 8-20a presents the minimum,
maximum, and annual average concentration of 1,3-butadiene for GPCO
compared to the 1,3-butadiene concentrations measured across the program for
NMP sites sampling VOCs with Method TO-15; Figure 8-20b presents the
minimum, maximum, and annual average 1,3-butadiene concentrations for the
Garfield County sites compared to the 1,3-butadiene concentrations measured
8-43

-------
across the program for NMP sites sampling SNMOCs. Note that the scales are not
the same in the figures.
The program-level maximum concentration (21.5 |ig/m3) is not shown directly on
the box plot in Figure 8-20a as the scale has been reduced to 1.5 |ig/m3 in order to
allow for the observation of data points at the lower end of the concentration
range. GPCO's annual average 1,3-butadiene concentration is similar to the
program-level average concentration. Even though the annual average
concentration of 1,3-butadiene for GPCO is among the higher annual averages for
this pollutant the maximum 1,3-butadiene concentration measured at GPCO
(0.426 |ig/m3) is considerably less than the maximum concentration measured
across the program.
The program-level first and second quartiles are both zero, and thus, not visible in
Figure 8-20b, indicating that at least half of the 1,3-butadiene concentrations
measured by sites sampling SNMOCs were non-detects. The box plots show that
non-detects were measured at each of the Garfield County sites. The maximum
1,3-butadiene concentration measured at RICO (0.344 |ig/m3) is twice the
maximum concentration measured among the remaining Garfield County sites. Of
the Garfield County sites shown, RICO has the highest annual average
concentration of 1,3-butadiene, followed by RFCO, PACO, and BMCO.
1,3-Butadiene is not a pollutant of interest for BRCO.
The scale of the box plot in Figure 8-21 has also been reduced to allow for the
observation of data points at the lower end of the concentration range, as the
program-level maximum carbon tetrachloride concentration (23.7 |ig/m3) is
considerably greater than the majority of measurements. Figure 8-21 shows that
maximum carbon tetrachloride concentration measured at GPCO is considerably
less than the program-level maximum concentration. The annual average carbon
tetrachloride concentration for GPCO is less than the program-level median and
average concentrations and similar to the program-level first quartile. A single
non-detect of carbon tetrachloride was measured at GPCO.
The scale of the box plot in Figure 8-22 has also been reduced to allow for the
observation of data points at the lower end of the concentration range, as the
program-level maximum 1,2-dichloroethane concentration (111 |ig/m3) is
considerably greater than the majority of measurements. Note that all of the
concentrations of 1,2-dichloroethane measured at GPCO are less than the
program-level average concentration of 0.26 |ig/m3. The annual average
concentration for GPCO is similar to the program-level first quartile.
Similar to the box plots for benzene and 1,3-butadiene, Figure 8-23a presents the
minimum, maximum, and annual average concentration of ethylbenzene for
GPCO compared to the ethylbenzene concentrations measured across the program
for NMP sites sampling VOCs with Method TO-15; Figure 8-23b presents the
minimum, maximum, and annual average ethylbenzene concentrations for the
Garfield County sites compared to the ethylbenzene concentrations measured
8-44

-------
across the program for NMP sites sampling SNMOCs. Note that the scales are not
the same in the figures.
The scale of the box plot in Figure 8-23a has also been reduced to allow for the
observation of data points at the lower end of the concentration range, as the
program-level maximum ethylbenzene concentration (18.7 |ig/m3) is considerably
greater than the majority of measurements. Figure 8-23a for ethylbenzene shows
that GPCO's range of ethylbenzene measurements spans approximately 1 |ig/m3.
GPCO's annual average concentration of ethylbenzene is greater than the
program-level average concentration. The minimum ethylbenzene concentration
measured at GPCO is similar the program-level first quartile, indicating that
roughly 25 percent of the ethylbenzene concentrations across the program are less
than GPCO's minimum concentration. Recall from the previous section that
GPCO has the sixth highest annual average ethylbenzene concentration among
NMP sites sampling ethylbenzene.
Figure 8-23b presents the box plots for only BMCO and RICO as these are the
only Garfield County sites for which ethylbenzene was identified as a pollutant of
interest. The range of ethylbenzene concentrations measured at RICO is larger
than the range of concentrations measured at BMCO, although both ranges are
relatively small compared to the range of concentrations measured by all seven
sites sampling SNMOCs. The annual average concentration for RICO is more
than twice the annual average concentration for BMCO and the program-level
average concentration lies between the two (0.25 |ig/m3).
Figure 8-24 is the box plot for fluorene for GPCO. The maximum fluorene
concentration across the program is considerably higher than the maximum
concentration measured at GPCO. Yet, GPCO's annual average concentration is
greater than the program-level average concentration and GPCO has the fifth
highest annual average concentration among NMP sites sampling PAHs. Two
non-detects of fluorene were measured at GPCO.
Figure 8-25 presents the box plots for formaldehyde for the five Colorado sites for
which annual averages could be calculated. These box plots share some of the
same characteristics as the box plots for acetaldehyde. The box plot for GPCO
shows that the maximum concentration of formaldehyde across the program was
measured at this site. 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 annual average for GPCO is more than twice the program-level average
concentration (2.83 |ig/m3). 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 measured at each Garfield County site is less than the
program-level third quartile and each annual average concentration is less than the
program-level first quartile. Similar observations were made in the 2012 NMP
report.
8-45

-------
•	Figure 8-26 is the box plot for hexachloro-l,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 (82 percent). Sixty-one valid VOC samples were
collected at GPCO and of these, hexachloro-1,3-butadiene was detected in only
13 of them. Thus, many zeroes are substituted into the annual average
concentration of this pollutant. The maximum hexachloro-1,3-butadiene
concentration measured at GPCO is among the higher hexachloro-l,3-butadiene
concentrations measured across the program. The annual average concentration
for GPCO is just greater than the program-level average concentration of
hexachl oro-1,3 -butadi ene.
•	Figure 8-27 is the box plot for naphthalene and shows that the maximum
concentration of naphthalene across the program is roughly twice the maximum
concentration measured at GPCO. The annual average naphthalene concentration
for GPCO is greater than the program-level average concentration and the
program-level third quartile. Recall from the previous section that GPCO has the
second highest annual average naphthalene concentration among NMP sites
sampling PAHs. The minimum concentration of naphthalene measured at GPCO
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 3.4.3.2.
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 8-28 through 8-49 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 concentration is not
provided, although the range and percentiles are still presented. BMCO began sampling
SNMOCs and carbonyl compounds under the NMP at the end of 2010 and RFCO began in 2012;
thus, the trends analysis was not conducted for these sites.
8-46

-------
Figure 8-28. Yearly Statistical Metrics for Acenaphthene Concentrations Measured at GPCO
120
20081	2009	2010	2011	2012	2013
0 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 April 2008.
Observations from Figure 8-28 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.
•	Five of the six highest concentrations of acenaphthene were measured at GPCO in the
spring of 2012 and ranged from 53.7 ng/m3 to 182 ng/m3. Concentrations measured in
2012 were higher overall compared to other years as nine of the 16 concentrations
greater than 30 ng/m3 were measured in 2012 while only one or two were measured
in each of the other years of sampling.
•	Concentrations of acenaphthene decreased significantly from 2009 to 2010, based on
the 1-year averages, after which a steady increasing trend is shown through 2012.
Even if the two highest concentrations measured in 2012 were removed from the
dataset, the 1-year average concentration for acenaphthene for 2012 would still
represent more than a 50 percent increase from 2011. Between 2010 and 2012, the
median concentration doubled.
•	All of the statistical metrics shown in Figure 8-28 exhibit a decrease for 2013. Both
the 1-year average and median concentrations decreased by more than half from 2012
to 2013.
8-47

-------
Figure 8-29. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at GPCO
Maximum
Concentration for
2004 is 93.0 M-g/m3
2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
Observations from Figure 8-29 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 years were significantly lower. The
two highest acetaldehyde concentrations (93.0 |ig/m3 and 54.9 |ig/m3) were both
measured in 2004 and the third highest acetaldehyde concentration (17.2 |ig/m3) was
measured in 2005. The remaining six measurements greater than 7 |ig/m3 were all
measured in 2013 and ranged from 7.00 to 10.7 |ig/m3.
•	Between 2005 and 2012, the 1-year average concentrations vary by less than 1 |ig/m3,
ranging from 2.00 |ig/m3 (2010) to 3.00 |ig/m3 (2005). The 1-year average and
median concentrations are both at a minimum for 2010, representing a statistically
significant decrease from 2009. The 1-year average concentration increases steadily
between 2010 and 2012. The median concentration exhibits a similar pattern.
•	An additional increase is also shown for 2013, where all of the statistical metrics
except the minimum concentration exhibit an increase. The 1-year average
concentration increases by nearly 1 |ig/m3 from 2012 to 2013.
8-48

-------
Figure 8-30. Yearly Statistical Metrics for Benzene Concentrations Measured at GPCO
2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
0 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
Observations from Figure 8-30 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 between 2004 and 2007, based on
the 1-year average and median concentrations. After a period of increasing for 2008
and 2009, a significant decrease is shown for 2010. This decreasing trend continues
through 2013, when most of the statistical metrics are at a minimum. 2013 is the first
year that the 1-year average benzene concentration is less than 1 |ig/m3. This is also
true for the median concentration.
•	Even though maximum concentration and 95th percentile increased slightly from
2012 to 2013, the decrease shown for the central tendency statistics is driven by the
higher number of concentrations at the lower end of the concentration range for 2013.
The number of benzene concentrations less than 1 |ig/m3 more than doubled from
2012 (19) to 2013 (43).
8-49

-------
Figure 8-31. Yearly Statistical Metrics for Benzo(a)pyrene Concentrations Measured at
GPCO
2.00
1.75 -





1.25 -
1.00 -
0.75 "
0.50 "
0.25 -












— 	







_J_










o.oo -









20081	2009	2010	2011	2012	2013
0 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 April 2008.
Observations from Figure 8-31 for benzo(a)pyrene measurements collected at GPCO
include the following:
•	The maximum benzo(a)pyrene concentration (1.72 ng/m3) was measured at GPCO on
January 13, 2009. Four of the five highest benzo(a)pyrene concentrations (those greater
than 1.50 ng/m3) were measured in 2009, with the fifth measured in 2011.
•	For each year where both could be calculated, the median concentration is considerably
less than the 1-year average concentration. This is a result of non-detects, for which
zeroes are substituted into the calculations. Figure 8-31 shows that the minimum and
5th percentile are zero for all years of sampling, indicating that at least 5 percent of the
measurements were non-detects. A review of the data shows that the percentage of non-
detects has ranged from 25 percent (2009) to 44 percent (2010). The percentage of non-
detects for 2013 is 34 percent.
•	The 1-year average concentration decreased by almost half from 2009 to 2010.
Between 2010 and 2013, the 1-year average concentration has varied by less than
0.1 ng/m3, ranging from 0.17 ng/m3 (2012) to 0.24 ng/m3 (2013).
8-50

-------
Figure 8-32. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at GPCO
£ 0.75
2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
0 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
Observations from Figure 8-32 for 1,3-butadiene measurements collected at GPCO
include the following:
•	The only 1,3-butadiene concentration greater than 1 |ig/m3 measured at GPCO was
measured on December 11, 2004. The second highest concentration was also
measured in 2004 (0.75 |ig/m3), although a similar concentration was measured in
2009 (0.71 |ig/m3).
•	The 1-year average concentrations have varied by less than 0.07 |ig/m3 over the years
of sampling, ranging from 0.132 |ig/m3 (2010) to 0.197 |ig/m3 (2006).
•	The increase in the 1-year average concentration from 2011 to 2012 represents the
largest year-to-year change (approximately 0.05 |ig/m3). The median also increased
by this much from 2011 to 2012. Not only are the measurements at the upper end of
the concentration range higher for 2012, as the number of 1,3-butadiene
concentrations greater than 0.35 |ig/m3 increased from one to six, there were also no
non-detects reported for 2012, while there were seven reported for 2011.
•	The largest year-to-year change in the median concentration is the decrease shown
from 2012 to 2013. Although non-detects were not measured in either year, the
number of measurements less than 0.1 |ig/m3 nearly doubled from 2012 (17) to 2013
(31), thus representing half of the measurements for 2013.
8-51

-------
Figure 8-33. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured
at GPCO
1.25 -







0.75 -
0.50 -
0.25 -









i

i

rh





—o-
li
i	
i..


r—o-
¦TP...
r
—





r~





T


1




..
¦¦

¦¦


2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile ~•"¦'/•••Average
Observations from Figure 8-33 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, 16 non-detects
have been measured (nine in 2004, five in 2005, and one each in 2006 and 2013).
•	The year with the least variability is 2012, with a difference of 0.38 |ig/m3 between
the minimum and maximum concentrations and a difference of 0.25 |ig/m3 between
the 5th and 95th percentiles. However, the year with the highest 1-year average and
median concentrations (0.67 |ig/m3 and 0.68 |ig/m3, respectively) is also 2012. Note
the difference between the minimum and 5th percentile for 2012 compared to other
years.
•	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 in the same manner
than an outlier can drive an average upward.
•	There is a significant increase in the 1-year average concentrations from 2007 to 2008
as the range of concentrations measured doubled from one year to the next. After
2008, a steady decreasing trend is shown through 2010, with little change in the
measurements from 2010 to 2011. These statistical parameters increased significantly
8-52

-------
from 2011 to 2012, and are at a maximum for the period of sampling. All of the
statistical metrics exhibit a decrease from 2012 to 2013, primarily as a result of the
higher number of concentrations at the lower end of the concentration range. The
number of carbon tetrachloride concentrations less than 0.5 |ig/m3 increased from one
in 2012 to 11 in 2013.
Figure 8-34. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations
Measured at GPCO
0.25

0.15 -





.0



o
—
0.05 -
0.00 -


r
o

-o-



		J

	¦¦¦' Or	




2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
Observations from Figure 8-34 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 and 2013, indicating that at least 50 percent of the measurements were
non-detects prior to 2012. The number of measured detections began to increase in
2009, from 12 percent for 2009 and 2010, to 27 percent in 2011, and 90 percent for
2012. The percentage of measured detections decreased slightly for 2013
(74 percent).
•	As the number of measured detections increases, so do each of the corresponding
statistical metrics shown in Figure 8-34. The number of measured detections
increased by 63 percent from 2011 to 2012, thus, the 1-year average and median
concentrations exhibit considerable increases.
8-53

-------
•	The median concentration is greater than the 1-year average concentration for 2012
and 2013. This is because there were still non-detects (or zeros) factoring into the
1-year average concentration for each year, which tend to pull the average down.
Excluding non-detects, the minimum concentration for 2012 would be 0.04 |ig/m3,
with a difference between the minimum and maximum concentrations measured for
2012 of less than 0.1 |ig/m3. This is also true for 2013.
•	Even though the maximum and 95th percentile increased from 2012 to 2013, the
1-year average and median concentrations decreased. This results from a greater
number of non-detects for 2013 (nearly three times as many).
Figure 8-35. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at GPCO
2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
Observations from Figure 8-35 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 also 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
8-54

-------
decrease in all of the statistical parameters is shown from 2005 to 2006, and the slight
decreasing trend 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 in the 1-year average concentration is shown
from 2010 through 2012. The median concentration exhibits a slight increasing trend
beginning with 2009 and continuing through 2012.
•	All of the statistical parameters exhibit a decrease from 2012 to 2013. The maximum
ethylbenzene concentration measured in 2013 is the lowest maximum concentration
for any given year of sampling shown in Figure 8-35.
Figure 8-36. Yearly Statistical Metrics for Fluorene Concentrations Measured at GPCO
20081	2009	2010	2011	2012	2013
Year
O 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 April 2008.
Observations from Figure 8-36 for fluorene measurements collected at GPCO include the
following:
•	The trends graph for fluorene resembles the trends graph for acenaphthene shown in
Figure 8-28.
•	The range of measurements collected at GPCO spans between 15 ng/m3 and 17 ng/m3
for each year of sampling until 2012. For 2012, the range of measurements is
8-55

-------
significantly higher, with a maximum concentration nearly four times higher than
those measured in previous years. Eight of the nine highest acenaphthene
concentrations (those greater than 20 ng/m3) were measured at GPCO in 2012, with
the one additional concentration measured in 2013.
• The 1-year average concentration decreased significantly from 2009 to 2010. The
slight increase from 2010 to 2011 is followed by a more significant increase for 2012.
The 1-year average concentration then decreased by half from 2012 to 2013. The
median concentration has a similar pattern. The number of concentrations at the upper
end of the concentration range decreased considerably for 2013; the number of
measurements greater than 10 ng/m3 decreased from 29 in 2012 to nine in 2013. In
addition, the only two non-detects of acenaphthene measured at GPCO over the
period of sampling were measured in 2013.
Figure 8-37. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at GPCO
Maximum
Concentration for
2004 is 40.5 [ig/m-
2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
Observations from Figure 8-37 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);
these two concentrations of formaldehyde were measured on the same days in 2004
8-56

-------
as the two highest acetaldehyde concentrations. The next eight highest formaldehyde
concentrations were measured in 2013 and ranged from 13.9 |ig/m3 to 21.9 |ig/m3,
•	Even with decreasing maximum concentrations, the 1-year average concentrations
have an increasing trend through 2006. The 1-year average concentration is
approximately 4 |ig/m3 for each year between 2006 and 2009. A significant decrease
in all of the statistical metrics is shown for 2010. Although an even smaller range of
concentrations was measured in 2011, there is little change in the 1-year average
concentration. With a few higher concentrations measured in 2012, the 1-year
average calculated for 2012 is slightly higher than the 1-year average concentrations
for the previous two years, although the increase is not statistically significant.
•	All of the statistical parameters exhibit increases for 2013, particularly those
representing concentrations at the upper end of the concentration range. The 1-year
average concentration for 2013 is greater than the maximum concentrations measured
in several of the previous years and is greater than the 95th percentile for each of the
previous years.
8-57

-------
Figure 8-38. Yearly Statistical Metrics for Hexachloro-l,3-Butadiene Concentrations
Measured at GPCO
2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
0 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile ~•"¦'/•••Average
Observations from Figure 8-38 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 11 (or 18 percent) for 2013. This explains why
the minimum, 5th percentile, and median concentrations (and in some cases, the
1-year averages) are all zero for each year of sampling. The detection rate has
increased slightly over the last few years. Additional years of sampling are needed to
determine if this trend continues.
•	The maximum hexachloro-l,3-butadiene concentration was measured during 2005
(0.26 |ig/m3). Although nine additional measurements greater than 0.20 |ig/m3 have
been measured at GPCO, all but one of these were measured between 2005 and 2007.
•	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.
8-58

-------
Figure 8-39. Yearly Statistical Metrics for Naphthalene Concentrations Measured at GPCO
900
800
700
600
500

-------
Figure 8-40. Yearly Statistical Metrics for Benzene Concentrations Measured at BRCO
0 H—
2008
2009
2010
Year
2011
2012
2013

O 5th Percentile
- Minimum
— Median ~
Maximum
O 95th Percentile

Observations from Figure 8-40 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 only other benzene concentration greater than
4 |ig/m3 was measured at BRCO in 2010.
•	The statistical parameters for benzene exhibit a steady decreasing trend over the years
of sampling at BRCO through 2012. Prior to 2013, the 1-year average concentration
decreased by roughly half, from a maximum of 1.39 |ig/m3 in 2009 to a minimum of
0.68 |ig/m3 in 2012. The median concentration 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 between the two of 0.43 |ig/m3 for
2009 to 0.03 |ig/m3 for 2012. This indicates a decreasing variability in the
measurements.
• All of the statistical metrics exhibit an increase from 2012 to 2013, returning to
concentration levels similar to 2010.
8-60

-------
Figure 8-41. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
PACO
.5
.0
.5
.0
.5
.0
2008
2009
2010
2012
2013
Year
0 5th Percentile	- Minimum	— Median	- Maximum	O 95th Percentile	Average
1A 1-year average is not presented due to low method completeness in 2011.
Observations from Figure 8-41 for acetaldehyde measurements collected at PACO
include the following:
•	PACO began sampling acetaldehyde under the NMP in February 2008. A 1-year
average concentration is not presented for 2011 due to low method completeness.
Note that carbonyl compounds are sampled on a l-in-12 sampling schedule at PACO.
•	The maximum acetaldehyde concentration (2.04 |ig/m3) was measured at PACO on
January 13, 2009 and is the only acetaldehyde concentration greater than 2 |ig/m3
measured at this site.
•	The 1-year average concentrations have a decreasing trend through 2012, 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 (albeit slight),
while the median and minimum concentrations increased. Even though the range of
measurements is at a minimum for 2011, the concentrations greater than 1 |ig/m3
represent a higher percentage of measurements for 2011 compared to the previous
year.
•	For 2013, both the 1-year average and median concentrations exhibit an increase. The
range within which the majority of the measurements fall, indicated by the 5th and
8-61

-------
95th percentiles, is at a maximum for 2013 over the years of sampling, indicating an
increase in the variability of the measurements.
Figure 8-42. Yearly Statistical Metrics for Benzene Concentrations Measured at PACO
u
2008
2009
2010
Year
2011
20121
2013

O 5th Percentile
— Minimum
— Median -
Maximum
O 95th Percentile

1A 1-year average is not presented due to low method completeness in 2012.
Observations from Figure 8-42 for benzene measurements collected at PACO include the
following:
•	PACO began sampling SNMOCs under the NMP in January 2008. A 1-year average
concentration is not presented for 2012 due to low method completeness resulting
sampler issues.
•	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 eight highest benzene
concentrations (those greater than 5.50 |ig/m3) were all measured in either 2008 or
2009.
•	Even though the maximum concentration decreased from 2008 to 2009, benzene
concentrations increased overall from 2008 to 2009, as indicated by the increases in
the 1-year average, median, and 95th percentile. Concentrations of benzene exhibit a
significant decreasing trend between 2009 and 2010, when the maximum and 95th
percentile decreased by nearly half. This decreasing trend continued into 2011 and
8-62

-------
2012. Although a 1-year average concentration could not be calculated for 2012, the
maximum, 95th percentile, and median concentrations are at a minimum for 2012. No
benzene concentrations greater than 3 |ig/m3 were measured in 2012.
• The range of benzene concentrations increased considerably from 2012 to 2013. The
range within which the majority of the measurements fall, indicated by the 5 th and
95th percentiles, is at its largest since 2009. Nine benzene concentrations greater than
the maximum concentration for 2012 (2.97 |ig/m3) were measured in 2013.
Figure 8-43. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at PACO
Maximum
Concentration for
2009 is 3.15 ng/m3
2010	2011
Year
O 5th Percentile
— Minimurr
O 95th Percentile
1A 1-year average is not presented due to low method completeness in 2012.
Observations from Figure 8-43 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 1,3-butadiene measurement greater than 1 |ig/m3
measured at this site.
The increase in the 1-year average concentration from 2008 to 2009 is a result of this
outlier concentration measured in 2009. The second highest concentration measured
in 2009 is substantially less (0.19 |ig/m3). Excluding the maximum concentration for
2009 would result is a 1-year average concentration of only 0.028 |ig/m3 (rather than
0.88 |ig/m3), and thus a decrease in the 1-year average concentration by almost half
8-63

-------
from 2008 to 2009. Note that the median 1,3-butadiene concentration for 2009 is
zero, indicating that at least half of the measurements for 2009 are non-detects.
•	The second, third, fourth, and fifth highest 1,3-butadiene concentrations measured at
PACO were all measured in December 2010 and ranged from 0.39 |ig/m3 to
0.66 |ig/m3. The next highest concentration for this year was also measured in
December but was considerably less (0.16 |ig/m3). The 95th percentile for 2010 is
greater than the maximum concentration measured for all other years except 2009 and
more than tripled from 2009 to 2010. Even though half of the measurements in 2010
were non-detects, the December measurements for 2010 are driving the top-end
statistical parameters upward.
•	Nearly all of the statistical parameters decreased from 2010 to 2011 except the
minimum and 5th percentile, which are zero for both of these years.
•	Prior to 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 these years. For 2012, the number of non-detects is
at a minimum (29 percent) and explains why the median increased considerably
although the range of measurements did not change much from 2011 and 2012.
•	For 2013, the median concentration returned to zero as the number of non-detects
increased from 29 percent in 2012 to 83 percent for 2013. The maximum and 95th
percentile decreased considerably for 2013 and are at a minimum for the period of
sampling, as is the 1-year average concentration.
8-64

-------
Figure 8-44. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at PACO
2008	2009	2010	20111	2012	2013
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile *•••¦¦>•*• Average
1A 1-year average is not presented due to low method completeness in 2011.
Observations from Figure 8-44 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 concentration range, as
indicated by the increases in the median and 95th percentile.
•	The data distribution statistics for 2010 resemble those for 2008, although the 1-year
average and median concentrations both exhibit decreases.
•	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 concentration. A
8-65

-------
similar observation can be made for 2013. 2012 and 2013 account for all nine of the
formaldehyde concentrations less than 0.5 |ig/m3 measured at PACO.
Figure 8-45. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at RICO
J.O H	
2008
2009
20101
20111
Year
2012
20131
O 5th Percentile
- Minimum
— Median
- Maximum
O 95th Percentile

1A 1-year average is not presented due to low method completeness in 2010, 2011, and 2013.
Observations from Figure 8-45 for acetaldehyde measurements collected at RICO include
the following:
•	RICO began sampling carbonyl compounds under the NMP in February 2008. A
1-year average concentration is not presented for 2010, 2011, or 2013 due to low
method completeness. However, the range of measurements is provided for each of
these years.
•	The maximum acetaldehyde concentration (2.91 |ig/m3) was measured at RICO in
July 2008, although a similar concentration was also measured two sample days prior.
•	Because few 1-year average concentrations are shown, a distinct trend is hard to
identify. However, the measurements appear to have an overall decreasing trend,
based on the decreases shown for nearly all of the other statistical parameters.
•	The minimum and 5th percentiles decreased considerably from 2011 to 2012 and into
2013. 2012 and 2013 account for the 10 lowest concentrations (those less than
0.45 |ig/m3) of acetaldehyde measured at RICO.
8-66

-------
Figure 8-46. Yearly Statistical Metrics for Benzene Concentrations Measured at RICO
5
£
1
3
2008
2009
2010
Year
2011
2012
2013
O 5th Percentile
- Minimum
— Median ~
Maximum
O 95th Percentile

Observations from Figure 8-46 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.
The six highest benzene concentrations measured at RICO were all measured in 2009.
•	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
shown from 2008 to 2009 as well as the subsequent decreases in the years that follow.
The median concentration is 0.96 |ig/m3 for 2012, indicating that nearly half of the
measurements are less than 1 |ig/m3. The 1-year average concentration is also less
than 1 |ig/m3 for 2012.
•	All of the statistical parameters exhibit increases for 2013. The number of
concentrations greater than 2 |ig/m3 increased six-fold from 2012 to 2013.
•	The statistical metrics shown for RICO's benzene concentrations resemble the ones
shown for benzene concentrations measured at PACO (and to a lesser extent BRCO),
as all three sites exhibit a decreasing trend through 2012 followed by a considerable
increase for 2013.
8-67

-------
Figure 8-47. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at RICO
.0
.8
.6
.4
.2
.0
2008
2009
2010
2011
2012
2013
Year
0 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
Observations from Figure 8-47 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/m3 to 0.98 |ig/m3. Higher 1,3-butadiene
concentrations were also measured at PACO during December 2010.
•	With the exception of the maximum concentration, the range of concentrations
measured in 2008 and 2009 were similar to each other, as indicated by most of the
statistical parameters shown. This was followed by an increase in the magnitude of
the 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 concentration
upward.
•	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 subtle increase.
•	Most of the statistical parameters exhibit decreases from 2012 to 2013 (the minimum
and 5th percentile both stay the same), with the median concentration decreasing by
8-68

-------
half. Overall, the concentrations measured were lower in 2013. The number of
concentrations greater than 0.25 |ig/m3 decreased from 17 in 2012 to five in 2013;
further, the number of concentrations less than 0.1 |ig/m3 increased from 15 in 2012
to 31 in 2013, accounting for more than half of the concentrations measured in 2013.
Figure 8-48. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at RICO
3.0 —|		
Maximum
Concentration for
2010 is 25.73 ng/m3
2.5
2.0
E
0.0 +-
2008
2009
2010
Year
2011
2012
2013

O 5th Percentile
- Minimum
— Median ~~
Maximum
O 95th Percentile

Observations from Figure 8-48 for ethylbenzene measurements collected at RICO include
the following:
•	The maximum ethylbenzene concentration measured at RICO was measured on
August 18, 2010 (25.7 |ig/m3). The next highest concentration was also measured in
2010 but was considerably less (6.72 |ig/m3). No other ethylbenzene concentrations
greater than 2 |ig/m3 have been measured at RICO and few greater than 1 |ig/m3 have
been measured at this site. This explains why the 1-year average concentration is
greater than the 95th percentile for 2010, it's skewed by the outlier. Excluding the
maximum concentration measured at RICO from the 1-year average calculation for
2010 would result in a 1-year average concentration similar to 2009.
•	Excluding the outlier, there is a decreasing trend in the statistical parameters
representing the upper end of the concentrations measured at RICO between 2009 and
2012. All of the statistical parameters are at a minimum for 2012.
•	Each of the statistical metrics shown in Figure 8-48 increased from 2012 to 2013,
returning to levels similar to 2011.
8-69

-------
• Aside from the outlier measured in 2010, the trends graph for ethylbenzene for RICO
resembles the trends graph for benzene shown in Figure 8-46.
Figure 8-49. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at RICO
2008	2009	20101	20111	2012	20131
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1A 1-year average is not presented due to low method completeness in 2010, 2011, and 2013.
Observations from Figure 8-49 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 only other formaldehyde concentration greater than 4 |ig/m3
was measured on August 26, 2013 (4.38 |ig/m3). Only three additional concentrations
measured at RICO are greater than 3 |ig/m3 (one each in 2008, 2010, and 2011).
•	Because few 1-year average concentrations are shown, a distinct trend is hard to
identify. However, the measurements appear to have an overall decreasing trend after
2010, based on the decreases shown for several of the other statistical parameters.
•	The minimum and 5th percentiles decreased considerably from 2011 to 2012 and into
2013, similar to acetaldehyde. 2012 and 2013 account for nine of the 10 lowest
concentrations (those less than 0.75 |ig/m3) of formaldehyde measured at RICO.
8-70

-------
8.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at each Colorado monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4
for definitions and explanations regarding the various toxicity factors, time frames, and
calculations associated with these risk-based screenings.
8.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Colorado monitoring sites and where annual average
concentrations could be calculated, risk was examined by calculating cancer risk and noncancer
hazard approximations. These approximations can be used as risk estimates for cancer and
noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers want to shift their air
monitoring priorities. Refer to Section 3.4.3.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 8-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
8-71

-------
Table 8-6. Risk Approximations for the Colorado Monitoring Sites
Pollutant
Cancer
URE
(jig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Grand Junction, Colorado - GPCO
Acetaldehyde
0.0000022
0.009
58/58
3.79
±0.57
8.33
0.42
Benzene
0.0000078
0.03
61/61
0.99
±0.16
7.73
0.03
1.3 -Butadiene
0.00003
0.002
61/61
0.15
±0.03
4.46
0.07
Carbon Tetrachloride
0.000006
0.1
60/61
0.59
±0.03
3.51
0.01
1,2-Dichloroethane
0.000026
2.4
45/61
0.07
±0.01
1.71
<0.01
Ethylbenzene
0.0000025
1
61/61
0.49
±0.07
1.21
<0.01
Formaldehyde
0.000013
0.0098
58/58
6.44
± 1.22
83.70
0.66
Hexachloro -1,3 -butadiene
0.000022
0.09
13/61
0.02
±0.01
0.38
<0.01
Acenaphthene3
0.000088

56/56
8.05
± 1.77
0.71

Benzo(a)pyrenea
0.00176

37/56
0.24
±0.10
0.42

F1 no re neH
0.000088

54/56
6.88
± 1.09
0.61

Naphthalene3
0.000034
0.003
56/56
136.93
±23.05
4.66
0.05
Battlement Mesa, Colorado - BMCO
Acetaldehyde
0.0000022
0.009
27/28
0.46
±0.11
1.00
0.05
Benzene
0.0000078
0.03
54/55
1.26
±0.19
9.86
0.04
1,3-Butadiene
0.00003
0.002
3/55
<0.01
±<0.01
0.11
<0.01
Ethylbenzene
0.0000025
1
50/55
0.14
±0.03
0.34
<0.01
Formaldehyde
0.000013
0.0098
28/28
0.82
±0.23
10.62
0.08
Silt, Colorado - BRCO
Acetaldehyde
0.0000022
0.009
26/26
0.56
±0.12
1.23
0.06
Benzene
0.0000078
0.03
57/57
1.14
±0.20
8.92
0.04
Formaldehyde
0.000013
0.0098
26/26
0.87
±0.18
11.36
0.09
— = A Cancer URE or Noncancer RfC is not available.
NA = Not available due to the criteria for calculating an annual average.
a Average concentrations provided for the pollutants below the blue line for GPCO are presented in ng/m3 for ease of
viewing.
8-72

-------
Table 8-6. Risk Approximations for the Colorado Monitoring Sites (Continued)
Pollutant
Cancer
URE
(jig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Parachute, Colorado - PACO
Acetaldehyde
0.0000022
0.009
25/26
0.76
±0.18
1.68
0.08
Benzene
0.0000078
0.03
49/52
1.96
±0.31
15.33
0.07
1.3 -Butadiene
0.00003
0.002
9/52
0.01
±0.01
0.43
0.01
Formaldehyde
0.000013
0.0098
26/26
1.28
±0.25
16.63
0.13
Carbondale, Colorado - RFCO
Acetaldehyde
0.0000022
0.009
27/27
0.58
±0.19
1.28
0.06
Benzene
0.0000078
0.03
23/29
0.57
±0.12
4.46
0.02
1,3-Butadiene
0.00003
0.002
6/29
0.02
±0.01
0.50
0.01
Formaldehyde
0.000013
0.0098
27/27
0.75
±0.21
9.76
0.08
Rifle, Colorado - RICO
Acetaldehyde
0.0000022
0.009
25/25
NA
NA
NA
Benzene
0.0000078
0.03
23/29
1.52
±0.26
11.88
0.05
1.3 -Butadiene
0.00003
0.002
52/57
0.11
±0.02
3.40
0.06
Ethylbenzene
0.0000025
1
57/57
0.35
±0.04
0.88
<0.01
Formaldehyde
0.000013
0.0098
25/25
NA
NA
NA
— = A Cancer URE or Noncancer RfC is not available.
NA = Not available due to the criteria for calculating an annual average.
a Average concentrations provided for the pollutants below the blue line for GPCO are presented in ng/m3 for ease of
viewing.
Observations for GPCO from Table 8-6 include the following:
•	Formaldehyde, acetaldehyde, and benzene have the highest annual average
concentrations among GPCO's pollutants of interest.
•	Formaldehyde has the highest cancer risk approximation for this site (83.70 in-a-
million), followed by acetaldehyde (8.33 in-a-million), benzene (7.73 in-a-million),
and naphthalene (4.66 in-a-million). GPCO's cancer risk approximation for
formaldehyde is an order of magnitude greater than the cancer risk approximation for
acetaldehyde and is the third highest cancer risk approximation calculated across the
program for 2013.
8-73

-------
•	None of the pollutants of interest for GPCO have noncancer hazard approximations
greater than 1.0, indicating that no adverse noncancer health effects are expected from
these individual pollutants. Acetaldehyde and formaldehyde have the highest
noncancer hazard approximations (0.66 and 0.42, respectively) among the pollutants
of interest for GPCO. The noncancer hazard approximation for formaldehyde for
GPCO is the second highest noncancer hazard approximation calculated across the
program for 2013.
Observations for the Garfield County sites from Table 8-6 include the following:
•	Benzene has the highest annual average concentration among the pollutants of
interest for each Garfield County site, with the exception of RFCO. For RFCO,
formaldehyde has the highest annual average concentration. Recall however, that
annual averages could not be calculated for the carbonyl compounds for RICO.
•	Formaldehyde and benzene have the highest cancer risk approximations for sites in
which annual averages could be calculated. Formaldehyde's cancer risk
approximations range from 9.76 in-a-million (RFCO) to 16.63 in-a-million (PACO).
All of these are considerably less than the cancer risk approximation for
formaldehyde for GPCO (83.70 in-a-million). Benzene's cancer risk approximations
range from 4.46 in-a-million (RFCO) to 15.02 in-a-million (PACO). All of these
benzene cancer risk approximations are slightly greater than the cancer risk
approximation for benzene for GPCO (7.73 in-a-million), with the exception of
RFCO.
•	None of the noncancer hazard approximations calculated for the Garfield County sites
are greater than 1.0, indicating that no adverse noncancer health effects are expected
from these individual pollutants. The highest noncancer hazard approximation was
calculated for formaldehyde for PACO (0.13).
•	Annual averages, and therefore cancer risk and noncancer hazard approximations,
could not be calculated for acetaldehyde and formaldehyde for RICO.
For each of the site-specific pollutants of interest that have cancer risk approximations
greater than 75 in-a-million and/or a noncancer hazard approximation greater than 1.0, a
pollution rose was created 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. Additional information about this analysis is presented
in Section 3.4.3.3. Figure 8-50 is GPCO's pollution rose for formaldehyde.
8-74

-------
Figure 8-50. Pollution Rose for Formaldehyde Concentrations Measured at GPCO
360/0
315
Hg/m3
270
225
135
180
®<5 |ig/m3 05-10 |ig/m3 O>10 |ig/m3
Observations from Figure 8-50 include the following:
•	The pollution rose shows that few formaldehyde concentrations are shown in relation
with samples days with winds with a westerly component. Most of the formaldehyde
concentrations are shown on the right-hand side of the wind rose, in relation to winds
with an easterly component. This matches the wind observations shown on the
sample day wind rose presented in Figure 8-10.
•	The facility map in Figure 8-2 shows that most of the point sources are located on the
right side of a diagonal line drawn northwest to southeast through the monitoring site
location.
•	The highest formaldehyde concentrations were all measured on sample days with
winds with a southerly component, and most often from the southeast quadrant.
•	If the formaldehyde concentrations are grouped by compass direction, the direction
with the most concentrations is southeast, followed by east and south. If the
formaldehyde concentrations are averaged by compass direction, the highest
concentrations would be calculated for the southwest and south directions. The
southwest direction only includes a single concentration; the average for the southerly
direction includes five concentrations.
8-75

-------
8.5.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 8-7 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
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 3.4.3.4. Lastly,
Table 8-7 provides the pollutants with the highest cancer risk approximations (in-a-million) for
each site, 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 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 8.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
8-76

-------
Table 8-7. 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
Top 10 Cancer Toxicity-Weighted
Emissions
Top 10 Cancer Risk Approximations
Based on Annual Average Concentrations
(County-Level)

(County-Level)
(Site-Specific)



Cancer

Cancer Risk
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(in-a-million)
Grand Junction, Colorado (Mesa County) - GPCO
Benzene
166.34
Formaldehyde
1.72E-03
Formaldehyde
83.70
Formaldehyde
131.94
Benzene
1.30E-03
Acetaldehyde
8.33
Ethylbenzene
55.92
1,3-Butadiene
4.48E-04
Benzene
7.73
Acetaldehyde
49.20
Naphthalene
2.34E-04
Naphthalene
4.66
1.3 -Butadiene
14.93
POM, Group 2b
1.55E-04
1,3-Butadiene
4.46
Naphthalene
6.89
Ethylbenzene
1.40E-04
Carbon Tetrachloride
3.51
Dichloromethane
5.44
Acetaldehyde
1.08E-04
1,2-Dichloroethane
1.71
T etrachloroethy lene
1.86
POM, Group 2d
1.00E-04
Ethylbenzene
1.21
POM, Group 2b
1.76
POM, Group 5a
6.90E-05
Acenaphthene
0.71
POM, Group 2d
1.14
Arsenic, PM
3.36E-05
Fluorene
0.61
Battlement Mesa, Colorado (Garfield County) - BMCO
Benzene
652.88
Formaldehyde
7.96E-03
Formaldehyde
10.62
Formaldehyde
612.56
Benzene
5.09E-03
Benzene
9.86
Acetaldehyde
112.59
1,3-Butadiene
3.78E-04
Acetaldehyde
1.00
Ethylbenzene
67.74
Acetaldehyde
2.48E-04
Ethylbenzene
0.34
1,3-Butadiene
12.62
Ethylbenzene
1.69E-04


Naphthalene
4.78
Naphthalene
1.62E-04


T etrachloroethy lene
1.01
POM, Group 2b
7.72E-05


POM, Group 2b
0.88
POM, Group 2d
5.42E-05


POM, Group 2d
0.62
POM, Group 5a
3.89E-05


Dichloromethane
0.25
Arsenic, PM
3.28E-05



-------
Table 8-7. 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
Top 10 Cancer Toxicity-Weighted
Emissions
Top 10 Cancer Risk Approximations
Based on Annual Average Concentrations
(County-Level)

(County-Level)
(Site-Specific)



Cancer

Cancer Risk
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(in-a-million)
Silt, Colorado (Garfield County) - BRCO
Benzene
652.88
Formaldehyde
7.96E-03
Formaldehyde
11.36
Formaldehyde
612.56
Benzene
5.09E-03
Benzene
8.92
Acetaldehyde
112.59
1,3-Butadiene
3.78E-04
Acetaldehyde
1.23
Ethylbenzene
67.74
Acetaldehyde
2.48E-04


1.3 -Butadiene
12.62
Ethylbenzene
1.69E-04


Naphthalene
4.78
Naphthalene
1.62E-04


T etrachloroethy lene
1.01
POM, Group 2b
7.72E-05


POM, Group 2b
0.88
POM, Group 2d
5.42E-05


POM, Group 2d
0.62
POM, Group 5a
3.89E-05


Dichloromethane
0.25
Arsenic, PM
3.28E-05


Parachute, Colorado (Garfield County) - PACO
Benzene
652.88
Formaldehyde
7.96E-03
Formaldehyde
16.63
Formaldehyde
612.56
Benzene
5.09E-03
Benzene
15.33
Acetaldehyde
112.59
1,3-Butadiene
3.78E-04
Acetaldehyde
1.68
Ethylbenzene
67.74
Acetaldehyde
2.48E-04
1,3-Butadiene
0.43
1,3-Butadiene
12.62
Ethylbenzene
1.69E-04


Naphthalene
4.78
Naphthalene
1.62E-04


T etrachloroethy lene
1.01
POM, Group 2b
7.72E-05


POM, Group 2b
0.88
POM, Group 2d
5.42E-05


POM, Group 2d
0.62
POM, Group 5a
3.89E-05


Dichloromethane
0.25
Arsenic, PM
3.28E-05



-------
Table 8-7. 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
Top 10 Cancer Toxicity-Weighted
Emissions
Top 10 Cancer Risk Approximations
Based on Annual Average Concentrations
(County-Level)

(County-Level)
(Site-Specific)



Cancer

Cancer Risk
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(in-a-million)

Carbondale, Colorado (Garfield
County) - RFCO

Benzene
652.88
Formaldehyde
7.96E-03
Formaldehyde
9.76
Formaldehyde
612.56
Benzene
5.09E-03
Benzene
4.46
Acetaldehyde
112.59
1,3-Butadiene
3.78E-04
Acetaldehyde
1.28
Ethylbenzene
67.74
Acetaldehyde
2.48E-04
1,3-Butadiene
0.50
1.3 -Butadiene
12.62
Ethylbenzene
1.69E-04


Naphthalene
4.78
Naphthalene
1.62E-04


T etrachloroethy lene
1.01
POM, Group 2b
7.72E-05


POM, Group 2b
0.88
POM, Group 2d
5.42E-05


POM, Group 2d
0.62
POM, Group 5a
3.89E-05


Dichloromethane
0.25
Arsenic, PM
3.28E-05


Rifle, Colorado (Garfield County) - RICO
Benzene
652.88
Formaldehyde
7.96E-03
Benzene
11.88
Formaldehyde
612.56
Benzene
5.09E-03
1,3-Butadiene
3.40
Acetaldehyde
112.59
1,3-Butadiene
3.78E-04
Ethylbenzene
0.88
Ethylbenzene
67.74
Acetaldehyde
2.48E-04


1,3-Butadiene
12.62
Ethylbenzene
1.69E-04


Naphthalene
4.78
Naphthalene
1.62E-04


T etrachloroethy lene
1.01
POM, Group 2b
7.72E-05


POM, Group 2b
0.88
POM, Group 2d
5.42E-05


POM, Group 2d
0.62
POM, Group 5a
3.89E-05


Dichloromethane
0.25
Arsenic, PM
3.28E-05



-------
Table 8-8. 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
Top 10 Noncancer Toxicity-Weighted
Emissions
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(County-Level)
(County-Level)
(Site-Specific)





Noncancer



Noncancer

Hazard
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(HQ)
Grand Junction, Colorado (Mesa County) - GPCO
Toluene
381.86
Acrolein
550,555.59
Formaldehyde
0.66
Xylenes
274.58
Formaldehyde
13,463.29
Acetaldehyde
0.42
Benzene
166.34
1,3-Butadiene
7,464.46
1,3-Butadiene
0.07
Formaldehyde
131.94
Benzene
5,544.61
Naphthalene
0.05
Hexane
120.83
Acetaldehyde
5,466.88
Benzene
0.03
Methanol
102.01
Xylenes
2,745.81
Carbon Tetrachloride
0.01
Ethylbenzene
55.92
Naphthalene
2,298.28
Ethylbenzene
<0.01
Acetaldehyde
49.20
Antimony, PM
1,050.63
Hexachloro-1,3 -butadiene
<0.01
Ethylene glycol
29.13
Lead, PM
767.25
1,2-Dichloroethane
<0.01
1.3 -Butadiene
14.93
Arsenic, PM
521.58

Battlement Mesa, Colorado (Garfield County) - BMCO
Toluene
1,190.11
Acrolein
3,464,518.24
Formaldehyde
0.08
Xylenes
730.99
Formaldehyde
62,505.94
Acetaldehyde
0.05
Benzene
652.88
Benzene
21,762.81
Benzene
0.04
Methanol
623.52
Acetaldehyde
12,509.99
1,3-Butadiene
<0.01
Formaldehyde
612.56
Xylenes
7,309.95
Ethylbenzene
<0.01
Hexane
169.35
1,3-Butadiene
6,308.09


Acetaldehyde
112.59
Naphthalene
1,592.72


Acrolein
69.29
Propionaldehyde
567.82


Ethylbenzene
67.74
Cadmium, PM
526.47


1,3-Butadiene
12.62
Arsenic, PM
508.98



-------
Table 8-8. 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
Top 10 Noncancer Toxicity-Weighted
Emissions
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(County-Level)
(County-Level)
(Site-Specific)





Noncancer



Noncancer

Hazard
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(HQ)
Silt, Colorado (Garfield County) - BRCO
Toluene
1,190.11
Acrolein
3,464,518.24
Formaldehyde
0.09
Xylenes
730.99
Formaldehyde
62,505.94
Acetaldehyde
0.06
Benzene
652.88
Benzene
21,762.81
Benzene
0.04
Methanol
623.52
Acetaldehyde
12,509.99


Formaldehyde
612.56
Xylenes
7,309.95


Hexane
169.35
1,3-Butadiene
6,308.09


Acetaldehyde
112.59
Naphthalene
1,592.72


Acrolein
69.29
Propionaldehyde
567.82


Ethylbenzene
67.74
Cadmium, PM
526.47


1.3 -Butadiene
12.62
Arsenic, PM
508.98


Parachute, Colorado (Garfield County) - PACO
Toluene
1,190.11
Acrolein
3,464,518.24
Formaldehyde
0.13
Xylenes
730.99
Formaldehyde
62,505.94
Acetaldehyde
0.08
Benzene
652.88
Benzene
21,762.81
Benzene
0.07
Methanol
623.52
Acetaldehyde
12,509.99
1,3-Butadiene
0.01
Formaldehyde
612.56
Xylenes
7,309.95


Hexane
169.35
1,3-Butadiene
6,308.09


Acetaldehyde
112.59
Naphthalene
1,592.72


Acrolein
69.29
Propionaldehyde
567.82


Ethylbenzene
67.74
Cadmium, PM
526.47


1,3-Butadiene
12.62
Arsenic, PM
508.98



-------
Table 8-8. 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)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Carbondale, Colorado (Garfield County) - RFCO
Toluene
1,190.11
Acrolein
3,464,518.24
Formaldehyde
0.08
Xylenes
730.99
Formaldehyde
62,505.94
Acetaldehyde
0.06
Benzene
652.88
Benzene
21,762.81
Benzene
0.02
Methanol
623.52
Acetaldehyde
12,509.99
1,3-Butadiene
0.01
Formaldehyde
612.56
Xylenes
7,309.95

Hexane
169.35
1,3-Butadiene
6,308.09
Acetaldehyde
112.59
Naphthalene
1,592.72
Acrolein
69.29
Propionaldehyde
567.82
Ethylbenzene
67.74
Cadmium, PM
526.47
1.3 -Butadiene
12.62
Arsenic, PM
508.98
Rifle, Colorado (Garfield County) - RICO
Toluene
1,190.11
Acrolein
3,464,518.24
1,3-Butadiene
0.06
Xylenes
730.99
Formaldehyde
62,505.94
Benzene
0.05
Benzene
652.88
Benzene
21,762.81
Ethylbenzene
<0.01
Methanol
623.52
Acetaldehyde
12,509.99

Formaldehyde
612.56
Xylenes
7,309.95
Hexane
169.35
1,3-Butadiene
6,308.09
Acetaldehyde
112.59
Naphthalene
1,592.72
Acrolein
69.29
Propionaldehyde
567.82
Ethylbenzene
67.74
Cadmium, PM
526.47
1,3-Butadiene
12.62
Arsenic, PM
508.98

-------
Observations from Table 8-7 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 the same pollutants listed for the pollutants with
the highest toxicity-weighted emissions.
•	Eight of the highest emitted pollutants also have the highest toxicity-weighted
emissions in Mesa County; the same eight pollutants have the highest emitted
pollutants and highest toxicity-weighted emissions for Garfield County.
•	For GPCO, eight of the 10 pollutants with the highest cancer risk approximations also
appear on both emissions-based lists for Mesa County. POM, Group 2b, which is the
ninth highest emitted "pollutant" in Mesa County and ranks fifth for toxicity-
weighted emissions, includes several PAHs sampled for at GPCO including
acenaphthene and fluorene, which have the ninth and tenth highest cancer risk
approximations, respectively, for GPCO. Carbon tetrachloride and 1,2-dichloroethane
do not appear on either emissions-based list for Mesa County, although they have the
sixth and seventh highest cancer risk approximations, respectively, for GPCO.
•	Each of the pollutants of interest identified for the Garfield County sites appear on
both emissions-based lists in Table 8-7.
Observations from Table 8-8 include the following:
•	Toluene is the highest emitted pollutant with a noncancer RfC in both Mesa and
Garfield Counties, although the emissions are considerably higher in Garfield County.
These two counties have an additional eight pollutants in common on their lists of
highest emitted pollutants with noncancer RfCs.
•	The pollutant with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) for both counties is acrolein. Although acrolein was sampled for at
GPCO, this pollutant was excluded from the pollutants of interest designation, and
thus subsequent risk-based screening evaluations, due to questions about the
consistency and reliability of the measurements, as discussed in Section 3.2. Acrolein
is not a target analyte for the SNMOC method. Although acrolein has the highest
toxicity-weighted emissions for all but one county with an NMP site, rarely does it
appear among the highest emitted pollutants. Garfield County is the only county with
an NMP site for which acrolein ranks among the highest emitted. A similar
observation was made in the 2011 and 2012 NMP reports.
•	Five of the highest emitted pollutants in Mesa County also have the highest toxicity-
weighted emissions. Six of the 10 highest emitted pollutants in Garfield County
8-83

-------
(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. Several metals appear near the bottom of each toxicity-
weighted emissions list.
•	Formaldehyde, acetaldehyde, benzene, and 1,3-butadiene appear on all three lists for
GPCO. Naphthalene appears among the pollutants with the highest noncancer hazard
approximations and highest toxicity-weighted emissions, but is not among the highest
emitted pollutants with a noncancer RfC in Mesa County. Ethylbenzene appears
among the pollutants with the highest noncancer hazard approximations for GPCO
and highest emissions in Mesa County, but is not among those with the highest
toxicity-weighted emissions.
•	Each of the pollutants of interest identified for the Garfield County sites appear on
both emissions-based lists in Table 8-8, with one exception. Ethylbenzene is a
pollutant of interest for RICO and BMCO. Ethylbenzene appears among the
pollutants with the highest emissions in Garfield County, but is not among those with
the highest toxicity-weighted emissions.
8.6 Summary of the 2013 Monitoring Data for the Colorado Monitoring Sites
Results from several of the data treatments described in this section include the
following:
~~~ Fifteen pollutants failed screens for GPCO. The number ofpollutants failing screens
for the Garfield County sites rangedfrom four to five.
~~~ Formaldehyde and acetaldehyde have highest annual average concentrations for
GPCO; these were the only pollutants with annual average concentrations greater
than 1 ng/m3. Benzene had the highest annual average concentration for each of the
Garfield County sites, except RFCO, where formaldehyde was highest.
~~~ GPCO has the second highest annual average concentrations of acetaldehyde,
formaldehyde, and naphthalene among all NMP sites sampling these pollutants. Each
of the Garfield County sites are among the sites with the highest annual average
concentrations of benzene except RFCO.
~~~ Benzene concentrations at GPCO have an overall decreasing trend across the years
of sampling, while acetaldehyde concentrations have been increasing in recent years
at this site. In addition, the detection rate of 1,2-dichloroethane at GPCO has been
increasing steadily over the last few years of sampling.
~~~ Formaldehyde has the highest cancer risk approximation of the pollutants of interest
for GPCO as well as four of the five Garfield County sites. None of the pollutants of
interest for the Colorado monitoring sites have noncancer hazard approximations
greater than an HQ of 1.0.
8-84

-------
9.0	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.
9.1	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 9-1 is a composite satellite image retrieved from ArcGIS Explorer showing the
monitoring site and its immediate surroundings. Figure 9-2 identifies nearby point source
emissions locations by source category, as reported in the 2011 NEI for point sources, version 2.
Note that only sources within 10 miles of the site are included in the facility counts provided in
Figure 9-2. A 10-mile boundary was chosen to give the reader an indication of which emissions
sources and emissions source categories could potentially have a direct effect on the air quality at
the monitoring site. Further, this boundary provides both the proximity of emissions sources to
the monitoring site as well as the quantity of such sources within a given distance of the site.
Sources outside the 10-mile boundary are still visible on the map for reference, but have been
grayed out in order to emphasize emissions sources within the boundary. Table 9-1 provides
supplemental geographical information such as land use, location setting, and locational
coordinates.
9-1

-------
Figure 9-1. Washington, D.C. (WADC) Monitoring Site

t ill J N

\Zf
L
IQoualM'^'Nl


iVif ft i
1


-------
Figure 9-2. NEI Point Sources Located Within 10 Miles of W A DC
DtSTOICT OF
COLUMBIA
VIRGINIA
10 mile radius	I	County boundary
Legend
~ WADC NATTS site
Que to facnt> dc-nsify and collocanon tKa total racings
diapteyed n«y not tepraMfrt all tocrtt*a wlhw tie a*?# qI ritewsi
Source Category Group (No. of Facilities)
T	Airport/Airline/Airport Support Operations (26)
tf	Asp^al' Production/Hot Mi* Asphalt Plant (5) ?
B	Bulk Terminals/Bulk Plants (1)	~
•	Eiectncity Generation via Combustion (3)	p
>	Hotete/Moiels.'Lodging (5)	x
o	Institutional (school, hospital, prison, etc.) (19) ~
A	Military Base/National Security Facility (11)
Mine/Quarry/Mineral Processing Facility (1)
Miscellaneous CommercaaVlndustnai Facility (6|
Paint and Coating Manufacturing Facility (1)
Pnnting'Pubiishirigi'Paper Product Manufacturing Facility (6)
Rai' Yard'Rall Line Operations (1)
Water Treatment Facility (2)
9-3

-------
Table 9-1. Geographical Information for the Washington, D.C. Monitoring Site
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Additional Ambient Monitoring Information1
WADC
11-001-0043
Washington
District
Of
Columbia
Washington-
Arlington-
Alexandria, DC-
VA-MD-WV
38.921847,
-77.013178
Commercial
Urban/City
Center
Arsenic, Lead, CO, VOCs, S02, NOy, NO, N02, NOx,
PAMS, Carbonyl compounds, O3, Meteorological
parameters, PM10, PM10 Speciation, Size fractionated
particulate. Black carbon PM Coarse, PM2 5, PM2 5
Speciation IMPROVE Speciation SNMOC
1 Data for additional pollutants are reported to AQS for WADC (EPA, 2014b); however, these data are not generated by ERG and are therefore not included in this report.
BOLD ITALICS = EPA-designated NATTS Site
vo

-------
Figure 9-1 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 9-2 shows, WADC is surrounded by a number
of sources, many of which are included in three sources categories: 1) the airport and airport
support operations source category, which includes airports and related operations as well as
small runways and heliports, such as those associated with hospitals or televisions stations; 2) the
institutions category, which includes hospital, schools, and prisons, etc.; and 3) the military bases
and national security facilities. The closest sources to WADC are a wastewater treatment facility,
hospitals, and heliports at hospitals.
Table 9-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Washington D.C. monitoring site. Table 9-2 includes both county-
level population and vehicle registration information. Table 9-2 also contains traffic volume
information for WADC, as well as the location for which the traffic volume was obtained.
Additionally, Table 9-2 presents the daily VMT for the District of Columbia.
Table 9-2. Population, Motor Vehicle, and Traffic Information for the Washington, D.C.
Monitoring Site
Site
County
Estimated
County
Population1
County-level
Vehicle
Registration2
Annual
Average Daily
Traffic3
Intersection
Used for
Traffic Data
County-
level Daily
VMT4 "
WADC
District of
Columbia
646,449
322,350
8,700
1st St between W St and
VSt
9,786,301
bounty-level population estimate reflects 2013 data (Census Bureau, 2014)
2County-level vehicle registration reflects 2012 data (FHWA, 2014)
3AADT reflects 2011 data (DC DOT, 2012)
4County-level VMT reflects 2012 data (FHWA, 2013b)
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 9-2 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 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.
9-5

-------
• The district-level VMT is nearly 10 million miles and is in the middle of the range
compared to VMTs for other counties with NMP sites.
9.2 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.
9.2.1	Climate Summary
Located on the Potomac River that divides Virginia and Maryland, Washington, D.C.
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 while southerly wind prevail
throughout the rest of the year. Precipitation is fairly evenly distributed across the seasons
(Wood, 2004).
9.2.2	Meteorological Summary
Hourly meteorological data for 2013 were retrieved from NCDC for the weather station
closest to the Washington, D.C. monitoring site (NCDC, 2013), as described in Section 3.4.2.
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 9-3. These data
were used to determine how meteorological conditions on sample days vary from conditions
experienced throughout the year.
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 2013. 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 were representative of average weather conditions experienced
throughout the year near WADC.
9-6

-------
Table 9-3. 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
(38.85, -77.03)
5.2
miles
193°
(SSW)
Sample
Days
(61)
66.7
±4.7
58.6
±4.5
44.0
±5.1
51.6
±4.2
61.0
±3.5
1019.2
± 1.7
7.5
±0.7
2013
66.2
± 1.8
58.5
± 1.7
44.2
± 1.9
51.6
± 1.6
61.8
± 1.4
1018.6
±0.7
7.2
±0.3
1 Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
vo

-------
9.2.3 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 3.4.2. A wind rose shows the frequency of wind directions using
"petals" positioned around a 16-point compass, and uses different colors to represent wind
speeds.
Figure 9-3 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 9-3 also presents three different wind roses for the
WADC monitoring site. First, a historical wind rose representing 2003 to 2012 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 2013 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2013 is
presented. These can be used to identify the predominant wind speed and direction for 2013 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.
Observations from Figure 9-3 for WADC include the following:
•	The weather station at Reagan National Airport is located 5.2 miles south-southwest
of WADC. Between WADC and Reagan National Airport is much of the city of
Washington and the Potomac River.
•	Historically, southerly to south-southwesterly 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 (those less than or equal
to 2 knots) were observed for just 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, with southerly and south-southwesterly winds accounting
for nearly 30 percent of the wind observations for 2013.
•	The sample day wind patterns resemble those on the full-year wind rose, although
there are some differences. Southerly winds account for an even higher percentage of
wind observations on sample days in 2013. Winds from the north-northwest also
account for a higher percentage of wind observations on sample days. Fewer calm
winds were observed on sample days in 2013, accounting for less than 8 percent of
observations. Overall, though, the similarities in the three wind roses indicate that
wind patterns in 2013 were similar to what is expected climatologically near this site.
9-8

-------
Figure 9-3. Wind Roses for the Ronald Reagan Washington National Airport Weather
Station near WADC
Location of WADC and Weather Station
2003-2012 Historical Wind Rose
NORTH----,
25%
"X 20%
15%
10%
WIND SPEED
(Knots)
~ -22
H 17-21
11 - 17
I I 7- 11
I 4-7
H 2-4
Calms: 9.68%
2013 Wind Rose
Sample Day Wind Rose
NORTH""--,
!NORTH~---„
25%
20%
15%
10%
WIND SPEED
(Knots)
~ >=22
F~1 17-21
¦ 11 -17
I I 7- 11
rzi 4-7
2- 4
Calms: 9.33%
25%
"X 20%
15%
10%
WIND SPEED
(Knots)
I I =22
Wl 17-21
¦ 11 - 17
I" 1 7- 11
4-7
2- 4
Calms: 7.79%
9-9

-------
9.3 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 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. Hexavalent chromium and PAHs were sampled for at WADC. Note that hexavalent
chromium sampling was discontinued at WADC at the end of June 2013.
Table 9-4. Risk-Based Screening Results for the Washington, D.C. Monitoring Site
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Washington, D.C. - WADC
Naphthalene
0.029
59
60
98.33
96.72
96.72
Acenaphthylene
0.011
1
25
4.00
1.64
98.36
Hexavalent Chromium
0.000083
1
8
12.50
1.64
100.00
Total
61
93
65.59

Observations from Table 9-4 include the following:
•	Three pollutants failed screens for WADC: naphthalene, acenaphthylene, and
hexavalent chromium.
•	While naphthalene failed 98 percent of its total screens, acenaphthylene and
hexavalent chromium failed a single screen each.
•	Naphthalene accounted for nearly 97 percent of the total failed screens for WADC;
thus, naphthalene is WADC's only pollutant of interest.
9-10

-------
9.4 Concentrations
This section presents various concentration averages used to characterize air toxics
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 for each site to illustrate
how the site's concentrations compare to the program-level averages, as presented in
Section 4.1.
•	Concentration averages and other statistical metrics are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at WADC are provided in Appendices M and O.
9.4.1 2013 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 9-5, 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.
9-11

-------
Table 9-5. 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
60/60
68.61
±20.37
100.85
±32.72
74.59
± 22.45
88.51
±33.64
83.14
± 13.50
Observations for WADC from Table 9-5 include the following:
•	Naphthalene was detected in every valid PAH sample collected at WADC.
•	Concentrations of naphthalene measured at WADC range from 18.9 ng/m3 to
280 ng/m3.
•	The second quarter average concentration of naphthalene is higher than the other
quarterly averages shown in Table 9-5, and the associated confidence interval
indicates that there is considerably variability in the measurements. The fourth quarter
average concentration has a similar confidence interval. The two highest
concentrations of naphthalene measured at WADC, 280 ng/m3 measured in
October and 274 ng/m3 measured in May, are both more than 100 ng/m3 higher than
the third highest naphthalene concentration measured at WADC (173 ng/m3 measured
in January). Of the 18 naphthalene concentrations greater than 100 ng/m3 measured at
WADC, the highest number of these were measured during the second quarter of
2013 (6), followed by the fourth quarter (5).
•	As shown in Table 4-11, WADC has the ninth highest annual average concentration
of naphthalene compared to other NMP sites sampling PAHs.
9.4.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for the site-specific pollutants of
interest, where applicable. Thus, a box plot was created for naphthalene for WADC. Figure 9-4
overlays the site's minimum, annual average, and maximum naphthalene concentrations onto the
program-level minimum, first quartile, median, average, third quartile, and maximum
concentrations, as described in Section 3.4.3.1.
9-12

-------
Figure 9-4. Program vs. Site-Specific Average Naphthalene Concentration

300	400
Concentration {ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



Observations from Figure 9-4 include the following:
• The maximum naphthalene concentration measured at WADC is considerably
less than the program-level maximum concentration. The annual average
concentration of naphthalene for WADC (83.14 ± 13.50 ng/m3) is greater than the
program-level average concentration (75.26 ng/m3) but less than the program-
level third quartile (94.65 ng/m3).
9.4.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.3.2.
WADC has sampled PAHs under the NMP since mid-2008. Thus, Figure 9-5 presents the 1-year
statistical metrics for naphthalene for WADC. The statistical metrics presented for assessing
trends include the substitution of zeros for non-detects. If sampling began mid-year, a minimum
of 6 months of sampling is required for inclusion in the trends analysis; in these cases, a 1-year
average concentration is not provided, although the range and percentiles are still presented.
Observations from Figure 9-5 for naphthalene measurements collected at WADC include
the following:
•	WADC began sampling PAHs under the NMP in late June 2008.
•	The maximum naphthalene concentration shown was measured in 2009 and is the
only concentration greater than 500 ng/m3 measured at this site (553 ng/m3).
Concentrations greater than 400 ng/m3 have been measured in each year of sampling
except 2008 (which included only half a year's worth of samples) and 2013.
•	The 1-year average concentrations exhibit an overall decreasing trend between 2009
and 2013. 2013 is the first year with a 1-year average concentration less than
9-13

-------
100 ng/m3. The median concentration also has an overall decreasing trend, although
the median increased from 2011 to 2012 before exhibiting further decreases from
2012 to 2013. The median concentration is less than 100 ng/m3 for each year shown
in Figure 9-5, and is at a minimum for 2013 (68.70 ng/m3).
• The difference between the 5th and 95th percentiles is at a minimum for 2013
(excluding 2008), indicating that the majority of concentrations measured fell within
a tighter range of measurements. A similar observation was made for 2012 in the
2012 NMP report.
Figure 9-5. Yearly Statistical Metrics for Naphthalene Concentrations Measured at WADC
E
"bJ
2010	2011
Year
0 5th Percentile
- Minimum
Median	~ Maximum	O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until late June 2008.
9.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at the WADC monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4 for
definitions and explanations regarding the various toxicity factors, time frames, and calculations
associated with these risk-based screenings.
9.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for WADC and where annual average concentrations could
be calculated, risk was examined by calculating cancer risk and noncancer hazard
9-14

-------
approximations. These approximations can be used as risk estimates for cancer and noncancer
effects attributable to the pollutants of interest. Although the use of these approximations is
limited, they may help identify where policy-makers want to shift their air monitoring priorities.
Refer to Section 3.4.3.3 for an explanation of how cancer risk and noncancer hazard
approximations are calculated and what limitations are associated with them. Annual averages,
cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard approximations are
presented in Table 9-6, where applicable. Cancer risk approximations are presented as
probabilities while the noncancer hazard approximations are ratios and thus, unitless values.
Table 9-6. Risk Approximations for the Washington, D.C. Monitoring Site
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections vs.
# of Samples
Annual
Average
(ng/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Washington, D.C. - WADC
Naphthalene
0.000034
0.003
60/60
83.14
± 13.50
2.83
0.03
Observations for WADC from Table 9-6 include the following:
•	As discussed in Section 9.4.1, the annual average concentration of naphthalene for
WADC is the ninth highest annual average concentration compared to other NMP
sites sampling this pollutant.
•	The cancer risk approximation for naphthalene is greater than 1.0 in-a-million
(2.83 in-a-million).
•	The noncancer hazard approximation for naphthalene is significantly less than
1.0, indicating that no adverse noncancer health effects are expected from this
individual pollutant.
9.5.2 Risk-Based Emissions Assessment
In addition to the risk-based screening 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 (version 2)
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 3.4.3.4. Lastly,
Table 9-7 provides the cancer risk approximation (in-a-million) for the pollutant of interest for
WADC, as presented in Table 9-6. The emissions, toxicity-weighted emissions, and cancer risk
9-15

-------
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 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 9.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
Observations from Table 9-7 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
toxicity-weighted emissions (of the pollutants with cancer UREs).
•	Eight of the highest emitted pollutants also have the highest toxicity-weighted
emissions.
•	Naphthalene is the only pollutant of interest for WADC. This pollutant appears on
both emissions-based lists. Naphthalene is the seventh highest emitted pollutant with
a cancer URE in the District of Columbia and has the fourth highest toxicity-weighted
emissions (of the pollutants with cancer UREs).
•	Several POM Groups are among the highest emitted "pollutants" in the District
and/or rank among the pollutants with the highest toxicity-weighted emissions. POM,
Group 2b includes several PAHs sampled for at WADC including acenaphthylene,
which failed a single screen for WADC. POM, Group 2d includes several PAHs
sampled for at WADC, such as anthracene and phenanthrene, but none of these failed
screens. POM, Group 5a includes benzo(a)pyrene, which did not fail screens for
WADC.
9-16

-------
Table 9-7. 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)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Washington, D.C. (District of Columbia) - WADC
Benzene
110.18
Formaldehyde
1.21E-03
Naphthalene
2.83
Formaldehyde
92.82
Benzene
8.59E-04

Acetaldehyde
52.06
1,3-Butadiene
5.06E-04
Ethylbenzene
51.75
Naphthalene
2.78E-04
Tetrachloroethylene
18.70
POM, Group 2b
2.21E-04
1.3 -Butadiene
16.86
Nickel, PM
1.51E-04
Naphthalene
8.18
POM, Group 2d
1.50E-04
POM, Group 2b
2.51
Ethylbenzene
1.29E-04
POM, Group 2d
1.71
Acetaldehyde
1.15E-04
Dichloro methane
0.82
POM, Group 5a
1.11E-04

-------
Table 9-8. 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)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Washington, D.C. (District of Columbia) - WADC
Toluene
363.94
Acrolein
229,665.41
Naphthalene
0.03
Methanol
352.82
Formaldehyde
9,471.05

Hexane
217.66
1.3 -Butadiene
8,432.47
Xylenes
213.36
Acetaldehyde
5,784.35
Ethylene glycol
123.11
Benzene
3,672.70
Benzene
110.18
Nickel, PM
3,505.21
Formaldehyde
92.82
Chlorine
3,176.67
Acetaldehyde
52.06
Naphthalene
2,725.10
Ethylbenzene
51.75
Xylenes
2,133.58
Methyl isobutyl ketone
26.88
Arsenic, PM
1,691.85

-------
Observations from Table 9-8 include the following:
•	Toluene, methanol, and hexane are the highest emitted pollutants with noncancer
RfCs in the District of Columbia.
•	The pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) are acrolein, formaldehyde, and 1,3-butadiene.
•	Four of the highest emitted pollutants in the District of Columbia also have the
highest toxicity-weighted emissions.
•	Naphthalene has the eighth highest toxicity-weighted emissions but is not one of the
10 highest emitted pollutants (of the pollutants with noncancer RfCs).
•	None of the other pollutants sampled for at WADC under the NMP appear in
Table 9-8.
9.6 Summary of the 2013 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 ninth among
NMP sites sampling this pollutant.
~~~ Concentrations of naphthalene have an overall decreasing trend at WADC.
9-19

-------
10.0	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.
10.1	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 six Florida sites are located in three different urban areas. Three sites (AZFL, SKFL,
and SYFL) are located in the Tampa-St. Petersburg-Clearwater, Florida CBSA. ORFL and PAFL
are located in the Orlando-Kissimmee-Sanford, Florida CBSA. Another site, WPFL, is located in
the Miami-Ft. Lauderdale-West Palm Beach, Florida CBSA. Figures 10-1 and 10-2 are
composite satellite images retrieved from ArcGIS Explorer showing the St. Petersburg area
monitoring sites and their immediate surroundings. Figure 10-3 identifies nearby point source
emissions locations that surround these two sites by source category, as reported in the 2011 NEI
for point sources, version 2. Note that only sources within 10 miles of the sites are included in
the facility counts provided in Figure 10-3. A 10-mile boundary was chosen to give the reader an
indication of which emissions sources and emissions source categories could potentially have a
direct effect on the air quality at the monitoring sites. Further, this boundary provides both the
proximity of emissions sources to the monitoring sites as well as the quantity of such sources
within a given distance of the sites. Sources outside the 10-mile boundaries are still visible on the
map for reference, but have been grayed out in order to emphasize emissions sources within the
boundary. Figures 10-4 through 10-10 are the composite satellite images and emissions sources
maps for the Tampa site, the two sites in the Orlando area, and the site in Belle Glade.
Table 10-1 provides supplemental geographical information such as land use, location setting,
and locational coordinates.
10-1

-------
Figure 10-1. St. Petersburg, Florida (AZFL) Monitoring Site

-------
Figure 10-2. Pinellas Park, Florida (SKFL) Monitoring Site

-------
Figure 10-3. NEI Point Sources Located Within 10 Miles of AZFL and SKFL
Pirmltas
County

I
Hillst»ro«jgh
County
Gulf of
M$vko
Legend
	
M|l»
m
B^secrw
Tampa flay
&•&irw	snstn*
Note Due to facility density and sotocatton. the total facilities
displayed may noi represent a> facilities the area of merest
site	10 mile radius 	 County boundary
AZFL UATMP site ^ SKFL NATTS
Source Category Group (No. of Facilities)
«i<	Aerospacft'Aircrafl Manufacturing Facility (1)	A
T	Airporl/Ainme.'Airport Support Operations (9)	®
it	Asphalt Production/Hot Mix Asphalt Plant (1)	?
B	Bulk Terminals/Bulk Plants (2)	t"!
C	Chemical Manufacturing Fatality (2)	Q
<9	Dry Cleaning Facility (1)	azi
6	Electrical Equipment Manufacturing Facility (4i r
f	Electricity Generation via Combustion (2)	p
F	Food Processing.'Agriculture Facility (21	A
*	Industrial Machinery or Equipment Plant (3)	•
o	Institutional (school, hospital prison, ele ) (1)	W
•	landfill (2)
Metal Coating, Engraving, and Allied Servces to Manufacturers (1)
Metals Piocessingi'Fabncation Facility (6)
Miscellaneous CommerclaMndustnai Facility (6)
Municipal Waste Combustor (1)
Paint and Coating Manufacturing Facility (2)
Pharmaceutical Manufactunng (1)
Plastic Resin, or Rubber Products Plant (4)
PnntingfPubiiBhing,'Paper Product Manufacturing Facility (9)
Ship/Boat Manufactunng or Repair Facisty (S)
Wastewater Treatment Facility (2)
Woodwork. Furniture, Millwork & Wood Preserving Facility (1)
10-4

-------
Figure 10-4. Valrico, Florida (SYFL) Monitoring Site
AftIney.Rd
O

-------
Figure 10-5. NEI Point Sources Located Within 10 Miles of SYFL
Hasoorough
County
Tampa
Bay
e-'svw
42-1CTTW
Legend
Note Due to tKllrty density and coiocafcon. tne total faoimee
displayed may not represent a* facilities wrthm the area of meekest
~ SYFL NATTS site
Source Category Group (No. of Facilities)
~I< Aerospace/Aircraft Manufactunng Faeihty {1)
"f AirportAirtine'Airporl Support Operations (7)
t Asphalt Production,'Hot Mi» Asphalt Plant (3)
0	Auto Body Shop,Painters/Automotive Stores (1)
M Automobile,'Truck Manufacturing Facility (1)
Srlc* Structural Clay or Clay Ceramics Plant 11)
1	Compressor Station (1)
e E lectncal Equipment Manufacturing Faculty (1)
Fertilizer Plant (1)
F Food Processn<}''Agncui!ure Facility (3)
¦	Landfill 1.1)
¦	Metal Can Box, and Other Metal Container Manufacturing (1)
Metal Coating, Engraving and Allied Services to Manufacturers (1)
Mela's Processing/Fabrication Facility (5)
Mine.'QuarryJMlneral Processing Facility (2)
M»scellaneous Go nvnerwal'Industrial Facility (3)
Municipal WSste Combustor I I)
Pesticide Manufacturing Plant (1)
Petroleum Refinery (1 >
Plasfcc. Resin, or Rubber Products Plant (2)
Printing Publishing'Paper Product Manufacturing Facility (2)
Ran YartVRail Line Operations (2)
Woodwork Furniture MillworV A Wood Preserving Facility (1)
10 mile radius	[ County boundary
A

-------
Figure 10-6. Winter Park, Florida (ORFL) Monitoring Site

-------
Figure 10-7. Orlando, Florida (PAFL) Monitoring Site
~ E Colonial Dr
-F— I A J • ¥
E Colonial C
O E Colonial Or
iivomMa.
CS3 E Robinson Si
6C1JII
Sour.tr US fti .J!
(Tof n.1, ojoViJcofjj
O
00

-------
Figure 10-8. NEI Point Sources Located Within 10 Miles of ORFL and PAFL
W3WW	*t WW	4t*90W
niJTYr	wtpttw	•» tsm	Ri-imrw
Note Du« to fK»ty (tonally *f*d colkicaBiyv tf»* total fooWfct.
aonia^d may not represent at fac Jitter. wtlh-n tnc area of interest
wi *WTW	(ii 301TW
Legend
ftTSVW
— "* \
Stmlnolt
County
Oninfli
County
ORFL UATMP site	PAFL UATMP site
Source Category Group (No. of Facilities)
T	AirporVAirtineMirport Support Operations (231	•
ill	Asphalt Production/Hoi Ml* Asphalt Plant (5)	¦
5	Auto Body Shop/'Paintere'Autornobve Stores (t)	A
A	Automobile/Truck Mamifacfunng Faculty (3)	®
B	Bulk Terrrvnals.'fiutti Plants (1)	¦
t	Compressor Station i t)	?
6	Electrical Equipment Manufacturing Facility (2)	Q
i	Electricity Generation via Combusuon (t)	R
f	Food Process
-------
Figure 10-9. Belle Glade, Florida (WPFL) Monitoring Site
p
o
Hooker Hwy

-------
Figure 10-10. NEI Point Sources Located Within 10 Miles of WPFL
i ft."* ¦ • A

Ptfen Bcacn
Cotint)
(c sSrw	mi avov#
Legend
~ WPFL UATMP site
ao*«ww	MM^nrw
Note Dua to tacitly danaity and collocation tha total r»c
displayed may not r*pr««nl all facJitias wilbn M>c area of tnta**sl
10 mite radius \ County boundary
Source Category Group (No. of Facilities)
T Airport/Anrline/Airport Support Operations (4)
6 Bulk Terminals/Bulk Plants (2)
F Food Prooessin^Agneultufe Facility (1)
10-11

-------
Table 10-1. Geographical Information for the Florida Monitoring Sites
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Additional Ambient Monitoring Information1
AZFL
12-103-0018
St.
Petersburg
Pinellas
Tampa-St.
Petersburg-
Clearwater, FL
27.785556,
-82.74
Residential
Suburban
NO, NO2, NOx, VOCs, O3, Meteorological
parameters, PM10, PM10 Speciation, PM2.5.
SKFL
12-103-0026
Pinellas
Park
Pinellas
Tampa-St.
Petersburg-
Clearwater, FL
27.850348,
-82.714465
Residential
Suburban
VOCs, Meteorological parameters, PM10 Speciation,
Black carbon PM2.5 Speciation, IMPROVE
Speciation.
SYFL
12-057-3002
Valrico
Hillsborough
Tampa-St.
Petersburg-
Clearwater, FL
27.96565,
-82.2304
Residential
Rural
CO, S02, NOy, NO, NOx, VOCs, 03,
Meteorological parameters, PM10, PM10 Speciation,
PM2.5. PM2.5 Speciation, PM Coarse, IMPROVE
Speciation.
ORFL
12-095-2002
Winter
Park
Orange
Orlando-
Kissimmee-
Sanford, FL
28.596389,
-81.3625
Commercial
Urban/City
Center
CO, SO2, NO, NO2, NOx, VOCs, 03,
Meteorological parameters, PM10, PIVLs PIVL s
Speciation.
PAFL
12-095-1004
Orlando
Orange
Orlando-
Kissimmee-
Sanford, FL
28.550833,
-81.345556
Commercial
Suburban
None.
WPFL
12-099-0008
Belle
Glade
Palm Beach
Miami-Ft.
Lauderdale-West
Palm Beach, FL
26.724444,
-80.666667
Industrial
Rural
PM2.5 Speciation.
1 Data for additional pollutants are reported to AQS for these sites (EPA, 2014b); 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 10-1 shows that the area
surrounding AZFL consists of mixed land use, including residential, commercial, and industrial
properties. The industrial property separated from Azalea Park by 72nd Street North is a former
electronics manufacturer and is a permanently closed facility (EPA, 2015e). Heavily traveled
roadways are located less than 1 mile from the monitoring site. AZFL is located less than 1 mile
east of Boca Ciega Bay, the edge of which can be seen in the bottom-left corner of Figure 10-1.
SKFL is located in Pinellas Park, north of St. Petersburg. This site is located on the
property of Skyview Elementary School at the corner of 86th Avenue North and 60th Street
North. Figure 10-2 shows that SKFL is located in a primarily residential area. However, a
railroad intersects the Pinellas Park Ditch near a construction company on the left-hand side of
Figure 10-2. Population exposure is the purpose behind monitoring at this location. This site is
the Pinellas County NATTS site.
Figure 10-3 shows the location of the St. Petersburg sites in relation to each other. AZFL
is located approximately 5 miles south-southwest of SKFL. Most of the emissions sources on the
Tampa Bay Peninsula are located north of SKFL. A small cluster of point sources is also located
southeast of SKFL. The airport source category, which includes airports and related operations
as well as small runways and heliports, such as those associated with hospitals or television
stations; printing, publishing, and paper product manufacturing; metals processing and
fabrication; and ship/boat manufacturing or repair are the source categories with the greatest
number of emissions sources in the St. Petersburg area (based on the areas covered by the
10-mile radii). The emissions source closest to AZFL is a plastic, resin, or rubber products plant.
While the emissions source closest to SKFL falls into the miscellaneous commercial/industrial
facility source category, a plastic, resin, or rubber products plant, a metals processing/fabrication
facility, and a ship/boat manufacturing or repair facility are also located within 2 miles of SKFL.
SYFL is located in Valrico, which is also part of the Tampa-St. Petersburg-Clearwater,
Florida CBSA, although it is on the eastern outskirts of the area. The SYFL monitoring site is
located in a rural area, although, as Figure 10-4 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 10-4) are tanks that are part of the local water treatment facility. This site
10-13

-------
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.
Figure 10-5 shows that most of the emissions sources surrounding SYFL are greater than
5 miles away from the site. The point sources shown cover a number of sources categories. The
airport source category and metals processing and fabrication are the source categories with the
greatest number of emissions sources near SYFL. The closest source to SYFL is the water
treatment facility pictured in Figure 10-4. However, this facility is not shown in Figure 10-5
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 10-6 shows that ORFL is
located near Lake Mendsen, just behind Community Playground. The site is east of Lake
Killarney and south of Winter Park Village. This site lies in a commercial area and serves as a
population exposure site.
PAFL is located in northeast Orlando, on the northwestern edge of the Orlando Executive
Airport property, as shown in Figure 10-7. 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 10-7). A large shopping
complex is located to the northeast of the site, just north of the airport, between Colonial Drive
and Maguire Boulevard. Interstate-4 runs north-south approximately 2 miles to the west of the
monitoring site.
Figure 10-8 shows that ORFL is located 3.3 miles north-northwest 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. The closest emissions source to PAFL is Orlando
Executive Airport, which is located under the star symbol for PAFL in Figure 10-8. The closest
emissions source to ORFL is a hospital, which falls into the institutions category, and the
heliport located at the hospital, which falls into the airport source category.
10-14

-------
The WPFL monitoring site is located north of Belle Glade, a rural town west of most of
the urbanized areas of Palm Beach County. The monitoring site is located on the property of the
Palm Beach County Health Center and is surrounded by medical and municipal services, as well
as the detention center for the sheriffs office. Lake Okeechobee is located about 4 miles
northwest of the site at its nearest point. The town is surrounded by various agricultural areas to
the east, south, and west, where sugar cane is the primary crop. A sugar mill is the closest major
source to WPFL. Although the Everglades are located roughly 65 miles to the south of the site,
various wildlife, conservation, and flood control areas are located between the site and the
national park.
Figure 10-10 shows that a total of seven point sources are located within 10 miles of
WPFL. More than half of the point sources are in the airport source category, including the
source located under the star symbol for WPFL in Figure 10-10, which is a hospital heliport. The
aforementioned sugar mill is the "F" symbol located to the southeast of WPFL in Figure 10-10.
Table 10-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Florida monitoring sites. Table 10-2 includes both county-level
population and vehicle registration information. Table 10-2 also contains traffic volume
information for each site as well as the location for which the traffic volume was obtained.
Additionally, Table 10-2 presents the county-level daily VMT for Pinellas, Hillsborough, Orange
and Palm Beach Counties.
Table 10-2. Population, Motor Vehicle, and Traffic Information for the Florida Monitoring Sites
Site
County
Estimated
County
Population1
County-level
Vehicle
Registration2
Annual
Average Daily
Traffic3
Intersection
Used for
Traffic Data
County-
level
Daily VMT4
AZFL
Pinellas
929,048
879,683
42,500
66th St N.N of 9th St
21,460,593
SKFL
47,500
Park Blvd. E of 66th St N
SYFL
Hillsborough
1,291,578
1,157,057
10,000
MLK, east of Mcintosh Rd
34,614,572
ORFL
Orange
1,225,267
1,181,540
29,500
Orlando Ave, N of Morse Blvd
34,904,854
PAFL
49,000
Colonial/MLK Blvd, b/w Primrose
Rd & Bumby Ave
WPFL
Palm Beach
1,372,171
1,159,114
6,600
Hwy 98 (Belle Glade Rd), north of
Hooker Hwy
33,617,131
bounty-level population estimates reflect 2013 data (Census Bureau, 2014)
2County-level vehicle registrations reflect 2013 data (FL DHSMV, 2013)
3AADT reflects 2013 data (FL DOT, 2013a)
4County-level VMT reflects 2013 data (FL DOT, 2013b)
BOLD ITALICS = EPA-designated NATTS Site
10-15

-------
Observations from Table 10-2 include the following:
•	Palm Beach County, where WPFL is located, is the most populous of the Florida
counties with NMP sites, although Orange County and Hillsborough County also
have more than 1 million people each and Pinellas County has just less 1 million
people. These counties rank close to each other compared to other counties with NMP
sites, ranking between 10th and 15th in population. Note that WPFL is located well to
the west of the center of population in Palm Beach County, which is oriented along
the eastern third of the county, near the coast.
•	The vehicle registration counts for three of the four Florida counties are greater than
1 million, with Orange 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 counties with NMP sites, ranking between
ninth and 14th.
•	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
PAFL. The traffic volume for PAFL ranks 21 st among other NMP sites, with the
traffic volumes for SKFL, AZFL, and ORFL in the middle of the range compared to
other NMP sites. The traffic near SYFL is in the bottom third compared to other NMP
sites.
•	The VMTs for Orange, Hillsborough, and Palm Beach Counties are fairly similar to
each other, around roughly 34 million miles and ranking eighth, ninth, and 10th
compared to the VMTs for other counties with NMP sites. The VMT for Pinellas
County is considerably less at roughly 21 million, but is still among the highest
VMTs, ranking 16th among counties with NMP sites.
10.2 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.
10.2.1 Climate Summary
The state of Florida is characterized by mild winters and warm, humid summers.
Temperatures below freezing are infrequent while temperatures reaching 90°F are common from
May to September. Florida receives more precipitation than any other state except Louisiana.
Precipitation tends to be concentrated during the summer months, as afternoon thunderstorms
occur almost daily. Semi-permanent high pressure over the Atlantic Ocean extends westward
towards Florida in the winter, resulting in reduced precipitation amounts and mainly sunny skies.
Land and sea breezes affect coastal locations and the proximity to the Atlantic Ocean or Gulf of
Mexico can have a marked effect on local meteorological conditions. Florida's orientation and
10-16

-------
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 coastal areas (Wood, 2004; FCC, 2015).
10.2.2 Meteorological Summary
Hourly meteorological data for 2013 were retrieved from NCDC for the weather stations
closest to the Florida monitoring sites (NCDC, 2013), as described in Section 3.4.2. The weather
station closest to the AZFL monitoring site is located at St. Petersburg/Whitted Airport (WBAN
92806); closest to SKFL is the St. Petersburg/Clearwater International Airport weather station
(WBAN 12873); closest to SYFL is the Vandenberg Airport weather station (WBAN 92816);
closest to both ORFL and PAFL is the Orlando Executive Airport weather station (WBAN
12841); and closest to WPFL is the Palm Beach International Airport weather station (WBAN
12844). Additional information about each of these weather stations, such as the distance
between the sites and the weather stations, 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.
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 2013. 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 in 2013 at the Florida monitoring sites were representative of average
weather conditions experienced throughout the entire year. The largest difference is shown for
WPFL for relative humidity. Note that sampling at WPFL took place on a l-in-12 day schedule
from March 2013 through March 2014, yielding roughly 13 months of sample days. This is the
only site where any 2014 data is incorporated into the 2013 NMP report.
10-17

-------
Table 10-3. 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/
Whitted Airport
7.1
miles
Sample
Days
(61)
81.6
±2.0
75.2
±2.0
66.5
±2.2
69.6
± 1.9
76.0
±2.1
1017.5
± 1.0
7.7
±0.6
92806
(27.76, -82.63)
103°
(ESE)
2013
81.5
±0.8
75.1
±0.8
65.9
± 1.0
69.3
±0.9
74.5
± 1.0
1017.3
±0.4
8.0
±0.3
Pinellas Park, Florida - SKFL
St Petersburg-
Clearwater Intl.
Airport
12873
(27.91, -82.69)
4.5
miles
Sample
Days
(63)
81.4
±2.2
73.2
±2.1
63.9
±2.5
67.4
±2.2
74.4
±2.1
1018.1
± 1.0
6.5
±0.5
22°
(NNE)
2013
81.7
±0.8
73.6
±0.9
64.1
± 1.1
67.8
±0.9
74.0
± 1.0
1017.8
±0.4
6.8
±0.3



Valrico, Florida
- SYFL




Vandenberg Airport
92816
(28.01, -82.35)
7.8
miles
Sample
Days
(61)
83.5
±2.0
72.4
±2.1
63.6
±2.5
67.0
±2.2
76.9
±2.1
NA
2.9
±0.4
295°
(WNW)
2013
83.6
±0.8
72.4
±0.9
63.3
± 1.1
66.8
±0.9
75.9
±0.8
NA
3.1
±0.2
Winter Park, Florida - ORFL
Orlando Executive
Airport
4.0
miles
Sample
Days
(61)
81.9
±2.0
72.8
± 1.9
62.9
±2.6
66.8
±2.1
73.6
±2.4
1018.6
± 1.0
6.0
±0.5
12841
(28.55, -81.33)
153°
(SSE)
2013
82.1
±0.8
72.8
±0.8
62.7
± 1.1
66.8
±0.9
73.4
± 1.1
1018.5
±0.4
6.0
±0.2
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 Vandenberg Airport.

-------
Table 10-3. Average Meteorological Conditions near the Florida Monitoring Sites (Continued)
Closest Weather
Station
Distance
and

Average
Maximum
Average
Average
Dew Point
Average
Wet Bulb
Average
Relative
Average
Sea Level
Average
Scalar
(WBAN and
Coordinates)
Direction
from Site
Average
Type1
Temperature
(°F)
Temperature
(°F)
Temperature
(°F)
Temperature
(°F)
Humidity
(%)
Pressure
(mb)
Wind Speed
(kt)



Orlando, Florida
- PAFL




Orlando Executive
0.9
miles
Sample
Days
82.1
73.2
63.3
67.1
73.4
1018.5
6.1
Airport
(30)
±2.9
±2.7
±3.7
±3.0
±3.3
± 1.3
±0.7
12841
117°
(ESE)








(28.55, -81.33)

82.1
72.8
62.7
66.8
73.4
1018.5
6.0

2013
±0.8
±0.8
± 1.1
±0.9
± 1.1
±0.4
±0.2
Belle Glade, Florida - WPFL
Palm Beach Intl.
35.2
miles
Sample
Days
82.8
76.3
65.8
69.6
74.5
1017.0
7.3
Airport
(34)
± 1.9
±2.2
±2.6
±2.2
±3.2
±0.9
±0.8
12844
94°
(E)








(26.68, -80.10)
Mar 2013 -
82.3
76.0
65.3
69.2
70.8
1016.9
8.0

Mar 2014
±0.6
±0.7
±0.8
±0.7
±0.9
±0.3
±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 Vandenberg Airport.

-------
The Florida sites have some of the highest daily average temperatures among the NMP
sites, behind only the Arizona sites. The highest average dew point and wet bulb temperatures
among NMP sites were calculated for the Florida monitoring sites. The Tampa/St. Petersburg
sites and the Orlando sites also experienced some of the highest relative humidity levels among
NMP sites, behind only the Mississippi sites. While AZFL ranks among the windier locations,
with an average wind speed around 8 knots, SYFL ranks among the least windy locations, with
an average wind speed around 3 knots.
10.2.3 Wind Rose Comparison
Hourly surface wind data from the weather stations nearest the Florida sites, as presented
in Section 10.2.2, were uploaded into a wind rose software program to produce customized wind
roses, as described in Section 3.4.2. A wind rose shows the frequency of wind directions using
"petals" positioned around a 16-point compass, and uses different colors to represent wind
speeds.
Figure 10-11 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 10-11 also presents three different
wind roses for the AZFL monitoring site. First, a historical wind rose representing 2003 to 2012
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 2013 is
presented. Next, a wind rose representing wind data for days on which samples were collected in
2013 is presented. These can be used to identify the predominant wind speed and direction for
2013 and to determine if wind observations on sample days were representative of conditions
experienced over the entire year and historically. Figures 10-12 through 10-16 present the three
wind roses and distance maps for SKFL, SYFL, ORFL, PAFL, and WPFL, respectively. Note
that the full-year wind rose for WPFL includes data from March 2013 through March 2014.
10-20

-------
Figure 10-11. Wind Roses for the St. Petersburg/Whitted Airport Weather Station near
AZFL
Location of AZFL and Weather Station
2003-2012 Historical Wind Rose
WIND SPEED
(Knots)
I I »22
[ B 17-21
IH 11 17
I I 7- 11
~ 4-7
2- 4
Calms: 823%
2013 Wind Rose
e ~:
WWD SPEED
[Kn ots >
~ 4-7
H 2-4
Calms: 5.23%
Sample Day Wind Rose
WEST
(Kn ots}
SOUTH
WIND SPEED
10-21

-------
Figure 10-12. Wind Roses for the St. Petersburg/Clearwater International Airport Weather
Station near SKFL
Location of SKFL and Weather Station
2003-2012 Historical Wind Rose
WIND SPEED
(Knots)
I I »22
[ B 17-21
IH 11 -1?
I I 7- 11
~ 4-7
2- 4
Calms: 10.22%
2013 Wind Rose
WEST
WWD SPEED
[Kn ots >
~ 4-7
H 2-4
Calms: 9.25%
Sample Day Wind Rose
WEST
(Kn ots}
SOUTH
WIND SPEED
10-22

-------
Figure 10-13. Wind Roses for the Vandenberg Airport Weather Station near SYFL
Location of SYFL and Weather Station	2005-2012 Historical Wind Rose
;NORTH"--.,
;WEST
WIND SPEED
(Kn ots)
~ >=22
[ id 17-21
2013 Wind Rose
Sample Day Wind Rose
WIND SPEED
(Knots)
~
H 17-21
LI 11-17
r i 7-11
~~l 4-7
2-
Calms: 44.20%
WEST
(Kn ots}
SOUTH
WIND SPEED
SYFL
NORTH'
WEST
10-23

-------
Figure 10-14. Wind Roses for the Orlando Executive Airport Weather Station near ORFL
Location of ORFL and Weather Station	2003-2012 FHstorical Wind Rose
west:
: EAS
WIND SPEED
(Kn ots)
~
~ 4-7
¦ 2-4
Calms: 14.57%
2013 Wind Rose
Sample Day Wind Rose
; NORTH""-,
WIND SPEED
(Knots)
~
H 17 -21
LI 11-17
r i 7-11
~~l 4-7
2-
Calms: 1474%

WIND SPEED
(Kn ots)
~
¦I 11 -17
r i 7-11
~ 4-7
2- 4
Calms: 15.03%
NORTH'
EST
10-24

-------
Figure 10-15. Wind Roses for the Orlando Executive Airport Weather Station near PAFL
Location of PAFL and Weather Station
2002-2012 Historical Wind Rose
WEST
WIND SPEED
(Kn ots)
SOUTH
2013 Wind Rose
Sample Day Wind Rose
NORTHS-
WIND SPEED
(Knots)
I I >=22
I 1 17-21
L| 11-17
r i 7-11
~~l 4-7
2-
Calms: 14.74%

WWC SPEED
(Kn ots.)
17-21
11 - 17
SOUTH
Calms: 13.75%
NORTH'
WEST
10-25

-------
Figure 10-16. Wind Roses for the Palm Beach International Airport Weather Station near
WPFL
Location of WPFL and Weather Station
2003-2012 Historical Wind Rose
\
I * &
38 2 milM
SMHOfY
EST
WIN C SPEED
(Kn ots}
SOUTH

March 2013-March 2014 Wind Rose
WEST
WIND SPEED
(Knots)
~
[IB 17-21
~ 4-7
H 2-4
Calms: 8.99%
Sample Day Wind Rose
WEST
(Kn ots}
SOUTH
WIND SPEED
10-26

-------
Observations from Figure 10-11 for AZFL include the following:
•	The weather station at St. Petersburg/Whitted Airport is located 7.1 miles east-
southeast of AZFL. Between them is most of the city of St. Petersburg. Note that the
Whitted Airport is located on the Tampa Bay coast while AZFL is on the west side of
the peninsula near the Boca Ciega Bay.
•	The historical wind rose shows that winds from the north, northeast quadrant, and
east were the most commonly observed wind directions near AZFL, accounting for
approximately 40 percent of observations. Winds from the western quadrants were
observed less frequently than winds from the eastern quadrants. Calm winds (those
less than or equal to 2 knots) account for roughly 8 percent of the hourly wind
measurements over the last 10 years.
•	The full-year wind rose shows that fewer calm winds were observed in 2013. In
addition, a higher percentage of winds from the east-northeast, east, east-southeast,
and southeast were observed in 2013.
•	The sample day wind patterns resemble the full-year wind patterns, with east-
northeasterly and easterly winds observed most often. There were even fewer calm
winds (less than 5 percent) observed on sample days.
Observations from Figure 10-12 for SKFL include the following:
•	The weather station at St. Petersburg/Clearwater International Airport is located
4.5 miles north-northeast of SKFL. The St. Petersburg/Clearwater Airport is located
on Old Tampa Bay while SKFL is located farther inland.
•	The historical wind rose shows that winds from a variety of directions were observed
near SKFL, although winds from the north and northeast to east-southeast were the
most commonly observed wind directions. Winds from the southwest quadrant were
observed infrequently. Calm winds account for approximately 10 percent of the
hourly wind measurements.
•	The 2013 wind rose resembles the historical wind rose in that winds from the
northeast to east-southeast account for a majority of the wind observations. Winds
from the southeast account for slightly more wind observations while winds from the
north account for slightly less. The percentage of calm winds for 2013 is less than
10 percent.
•	Northeast to east-southeast winds account for a majority of the wind observations on
sample days near SKFL, but northerly and north-northeasterly winds were observed
nearly as often. Together, winds from the north to east-southeast account for nearly
50 percent of the wind observations on sample days. None of the other directions
account for more than 6 percent of wind observations. The percentage of calm winds
on sample days is similar to the percentage of calm winds over the course of 2013.
10-27

-------
Observations from Figure 10-13 for SYFL include the following:
•	The weather station at Vandenberg Airport is located 7.8 miles west-northwest of
SYFL. Note that the closest weather station to SYFL is actually located at Plant City
Municipal Airport; however, data from the second closest weather station is used for
this report because the meteorological sensors at Plant City Municipal Airport were
down for part of 2013.
•	The historical wind rose shows that calm winds account for approximately 43 percent
of the hourly wind measurements between 2005 and 2012. Winds from the east were
prevalent during this time period, although winds from the north, northeast quadrant,
and east were observed often, together accounting for about one-quarter of the
observations. Winds from the southeast quadrant were observed the least. Note that
among the sites in the Tampa-St. Petersburg area, winds were lightest near SYFL, as
few wind observations greater than 11 knots were measured.
•	Calm winds account for a slightly higher percentage of observations on the full-year
wind rose (44 percent). East was the prevailing wind direction again for 2013,
accounting for a slightly higher percentage of winds (8 percent). Northerly winds
were observed less frequently in 2013 while winds from the north-northeast to east-
northeast each account for roughly 5 percent of wind observations.
•	While easterly winds prevailed, winds from the north-northeast and northeast account
for a higher percentage of wind observations on sample days near SYFL. Conversely,
winds from the east-northeast account for fewer observations on sample days. Calm
winds account for 45 percent of observations on sample days.
Observations from Figures 10-14 and 10-15 for ORFL and PAFL include the following:
•	The closest weather station to both ORFL and PAFL is located at the Orlando
Executive Airport. The weather station is located 4 miles south-southeast of ORFL
and less than 1 mile east-southeast of PAFL, as PAFL is located on the edge of the
Orlando Executive Airport property. Thus, the historical and full-year wind roses
presented for these sites are identical. Note that the distance between PAFL and the
weather station at Orlando Executive Airport is the shortest distance from a weather
station among NMP sites.
•	The historical wind roses show that winds from all directions were observed near
these sites, with easterly winds 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 for 2013 resemble the wind patterns on the historical wind rose,
although winds from the north, with an easterly component, or south were observed
even more frequently in 2013.
10-28

-------
•	The sample day wind rose for ORFL shows that winds from the north prevailed,
accounting for approximately 12 percent of wind observations, followed by easterly
winds, which account for roughly 11 percent of observations. Winds from the
northwest quadrant were observed even less frequently on sample days.
•	The sample day wind rose for PAFL shares the northerly and easterly prominence of
ORFL's sample day wind rose but with a higher percentage wind observations from
the northeast quadrant and slightly fewer calm winds. Note that although the sample
days are fairly standardized, samples are collected at PAFL on a l-in-12 day sampling
schedule, leading to roughly half the sample days included in the sample day wind
rose as ORFL.
Observations from Figure 10-16 for WPFL include the following:
•	The weather station at Palm Beach International Airport is located more than 35 miles
east of WPFL. The weather station is located near the Atlantic coast while WPFL is
located well inland, just southeast of Lake Okeechobee. Note that the distance between
WPFL and the weather station at Palm Beach International Airport is the longest
distance among NMP sites.
•	The historical wind rose shows that east winds are also prevalent near WPFL,
accounting for 16 percent of observations. Winds from the southeast quadrant were
also observed frequently nearly WPFL. Winds from the southwest quadrant were
observed the least between 2003 and 2012. Calm winds account for nearly 11 percent
of the hourly wind measurements.
•	The wind patterns on the wind rose representing observations between March 2013
and March 2014 resemble those on the historical wind rose.
•	Although winds from the east and southeast account for the greatest number of wind
observations on sample days, the percentage is reduced compared to the full-year
wind rose. Winds from the northeast, south, northwest, and north were observed more
frequently on sample days near WPFL compared to the full-year wind rose, as were
calm winds.
10.3 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 10-4.
Pollutants of interest are those for which the individual pollutant's total failed screens contribute
10-29

-------
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. Carbonyl compounds were sampled for at AZFL and ORFL. Hexavalent chromium
and PAHs were sampled for in addition to carbonyl compounds at SKFL and SYFL. Only PMio
metals were sampled for at PAFL and only PAHs were sampled for at WPFL. Note that
hexavalent chromium sampling was discontinued at SKFL and SYFL at the end of June 2013. In
addition, PAH sampling was also discontinued at SYFL at the end of June 2013.
Table 10-4. 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
0.077
59
59
100.00
50.43
50.43
Acetaldehyde
0.45
58
59
98.31
49.57
100.00
Total
117
118
99.15

Pinellas Park, Florida - SKFL
Acetaldehyde
0.45
60
60
100.00
35.09
35.09
Formaldehyde
0.077
60
60
100.00
35.09
70.18
Naphthalene
0.029
50
59
84.75
29.24
99.42
Hexavalent Chromium
0.000083
1
8
12.50
0.58
100.00
Total
171
187
91.44

Valrico, Florida - SYFL
Acetaldehyde
0.45
61
61
100.00
45.52
45.52
Formaldehyde
0.077
61
61
100.00
45.52
91.04
Naphthalene
0.029
12
29
41.38
8.96
100.00
Total
134
151
88.74

Winter Park, Florida - ORFL
Acetaldehyde
0.45
61
61
100.00
50.00
50.00
Formaldehyde
0.077
61
61
100.00
50.00
100.00
Total
122
122
100.00

Orlando, Florida - PAFL
Arsenic (PMio)
0.00023
30
30
100.00
100.00
100.00
Total
30
30
100.00

Belle Glade, Florida - WPFL
Naphthalene
0.029
5
30
16.67
45.45
45.45
Acenaphthene
0.011
2
30
6.67
18.18
63.64
Fluoranthene
0.011
2
30
6.67
18.18
81.82
Fluorene
0.011
2
27
7.41
18.18
100.00
Total
11
117
9.40

10-30

-------
Observations from Table 10-4 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 100 percent of screens. For ORFL, acetaldehyde also failed
100 percent of screens. For AZFL, acetaldehyde failed one less screen than
formaldehyde. Among the carbonyl compounds, only acetaldehyde, formaldehyde,
and propionaldehyde have risk screening values. Propionaldehyde did not fail any
screens for these two sites.
•	Four pollutants failed at least one screen for SKFL; 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. Hexavalent
chromium was the only other pollutant to fail screens for SKFL; this pollutant was
detected in eight samples collected at SFKL and failed only one screen. Acetaldehyde
and formaldehyde both failed 100 percent of screens for SKFL, contributing equally
to the total number of failed screens for SKFL.
•	Three pollutants failed at least one screen for SYFL (acetaldehyde, formaldehyde,
and naphthalene) and each of these was identified as a pollutant of interest for this
site. Similar to SKFL, acetaldehyde and formaldehyde both failed 100 percent of
screens for SYFL, contributing equally to the total number of failed screens.
•	Arsenic is the only PMio metal to fail screens for PAFL. This pollutant was detected
in every metals sample collected at PAFL and failed 100 percent of screens.
•	Four PAHs failed screens for WPFL. Less than 10 percent of concentrations of these
four pollutants failed screens. Naphthalene failed the highest number of screens (five)
compared to the other three pollutants, all of which failed only two screens (which
were all for the same two sample days). Compared to the other Florida monitoring
sites sampling PAHs, this site failed relatively few screens (11 total), although the
number of PAHs failing screens was greater for WPFL (four) than the other Florida
sites.
10.4 Concentrations
This section presents various concentration averages used to characterize air toxics
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.
10-31

-------
• Concentration averages and other statistical metrics are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at the Florida monitoring sites are provided in Appendices L, M, N, and O.
10.4.1 2013 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 AZFL, SKFL, SYFL, ORFL, and PAFL
are presented in Table 10-5a, 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. Quarterly and annual
average concentrations for WPFL are presented in Table 10-5b due to the differences in the
sampling period for this site. Quarterly average concentrations are presented beginning with the
second quarter of 2013 and continuing through the first quarter of 2014 to match the period of
sampling. Note that the results presented in Table 10-5b are in ng/m3.
10-32

-------
Table 10-5a. Quarterly and Annual Average Concentrations of the Pollutants of Interest
for the Florida Monitoring Sites

# of






Measured
1st
2nd
3rd
4th


Detections
Quarter
Quarter
Quarter
Quarter
Annual

vs. # of
Average
Average
Average
Average
Average
Pollutant
Samples
(Ug/m3)
(Ug/m3)
(Ug/m3)
(Ug/m3)
(Ug/m3)


St. Petersburg, Florida -
AZFL




2.27
1.51
1.00
1.64
1.63
Acetaldehyde
59/59
±0.47
±0.29
±0.14
±0.32
±0.19


1.31
1.30
2.31
2.19
1.77
Formaldehyde
59/59
±0.17
±0.14
±0.27
±0.31
±0.16


Pinellas Park, Florida -
SKFL




1.30
0.96
1.78
1.50
1.39
Acetaldehyde
60/60
±0.23
±0.21
±0.25
±0.30
±0.14


2.21
2.27
0.49
1.19
1.54
Formaldehyde
60/60
±0.36
±0.41
±0.12
±0.17
±0.24


95.15
52.76
66.59
61.45
69.26
Naphthalene1
59/59
±43.83
± 14.67
± 20.70
± 22.49
± 13.80
Valrico, Florida - SYFL


1.75
1.32
1.08
1.05
1.30
Acetaldehyde
61/61
±0.52
±0.26
±0.17
±0.18
±0.16


1.54
2.08
2.01
1.67
1.82
Formaldehyde
61/61
±0.19
±0.32
±0.31
±0.27
±0.14


31.84
24.76



Naphthalene1
29/29
±8.16
±5.55
NA
NA
NA
Winter Park, Florida - ORFL


2.66
1.31
1.13
1.12
1.55
Acetaldehyde
61/61
±0.86
±0.36
±0.20
±0.16
±0.28


0.83
2.35
2.29
1.87
1.84
Formaldehyde
61/61
±0.33
± 1.04
±0.34
±0.28
±0.31
Orlando, Florida - PAFL


0.98
0.58
0.78
0.58
0.72
Arsenic (PMi0)a
30/30
±0.84
±0.20
±0.52
±0.22
±0.22
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for
ease of viewing.
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
10-33

-------
Table 10-5b. Quarterly and Annual Average Concentrations of the Pollutants of Interest
for WPFL

# of






Measured
2nd Quarter
3rd Quarter
4th Quarter
1st Quarter


Detections
2013
2013
2013
2014
Period

vs. # of
Average
Average
Average
Average
Average
Pollutant
Samples
(ng/m3)
(ng/m3)
(ng/m3)
(ng/m3)
(ng/m3)
Belle Glade, Florida - WPFL


7.39
4.82
2.29
3.71
4.62
Acenaphthene
30/30
± 10.07
±4.90
± 1.51
± 1.96
±2.72


10.44
6.34
2.26
2.82
5.62
Fluoranthene
30/30
± 13.15
±6.80
±2.28
± 1.09
±3.64


6.01
3.64
2.20
2.34
3.66
Fluorene
27/30
±7.70
±3.91
± 1.18
± 1.43
±2.10


13.65
10.44
103.00
19.86
35.74
Naphthalene
30/30
±6.56
±5.28
± 180.15
±6.17
±33.34
Observations from Table 10-5a include the following:
•	The annual average concentration of formaldehyde is similar to the annual average
concentration of acetaldehyde for AZFL.
•	For acetaldehyde, the three acetaldehyde concentrations greater than 3 |ig/m3 were
measured during the first quarter and all but one of the 14 acetaldehyde
concentrations greater than 2 |ig/m3 were measured during the first and fourth
quarters of 2013. This is reflected in the quarterly average concentrations for AZFL
shown in Table 10-5a. Concentrations of formaldehyde do not follow this trend. The
third and fourth quarter average concentrations of formaldehyde are significantly
higher than the first and second quarter averages. The 18 highest concentrations of
formaldehyde measured at AZFL were measured between July and December, with
the two highest measurements collected in October (those greater than 3.0 |ig/m3).
•	The annual average concentration of formaldehyde is similar to the annual average
concentration of acetaldehyde for SKFL.
•	The quarterly average concentrations of formaldehyde for SKFL show that the
concentrations measured during the first half of the year were considerably higher
than those measured during the second half of the year. Concentrations measured
between January and June range from 1.19 |ig/m3 to 4.06 |ig/m3, with a median
concentration of 2.14 |ig/m3; concentrations measured between July and December
range from 0.297 |ig/m3 to 1.74 |ig/m3, with a median concentration of 0.84 |ig/m3.
The instrumentation at SKFL was changed at the beginning of July 2013, which may
have something to do with the change in measurements. For the first half of the year,
only two acetaldehyde concentrations greater than 2 |ig/m3 were measured; nine were
measured during the second half of the year.
•	Concentrations of naphthalene measured at SKFL range from 17.1 ng/m3 to
357 ng/m3, with a median concentration of 50.6 ng/m3. The maximum concentration,
measured on January 16, 2013, was nearly 100 ng/m3 greater than the next highest
10-34

-------
concentration, which was also measured during the first quarter of 2013. Half of the
12 naphthalene concentrations greater than 100 ng/m3 were measured during the first
quarter of 2013 (with no more than three measured in the other quarters of the year).
This explains both the higher first quarter average as well as the variability associated
with that average concentration of naphthalene for SKFL.
Concentrations of formaldehyde were higher than concentrations of acetaldehyde
measured at SYFL, based on the annual average concentrations.
The maximum acetaldehyde concentration measured at SYFL (4.87 |ig/m3) was
measured on March 29, 2013 and is the only measurement greater than 3 |ig/m3. Nine
of the 10 highest acetaldehyde concentrations were measured at SYFL between
January and May, explaining the differences shown in the quarterly averages.
The nine highest formaldehyde concentrations were measured between April and
October while all but one of the 20 lowest formaldehyde concentrations were
measured between January and April or September and December. Although
formaldehyde concentrations appear higher during the warmer months of the year,
confidence intervals indicate that the differences are not statistically significant.
Concentrations of formaldehyde were just slightly higher than concentrations of
acetaldehyde measured at ORFL, based on the annual average concentrations.
Concentrations of acetaldehyde at ORFL were highest during the first quarter of 2013
and appear to decrease throughout the year, based on the quarterly average
concentrations. A review of the data shows that the three highest concentrations were
measured in March and that nine of the 10 concentrations were measured between
February and April. Conversely, most of the concentrations less than 1 |ig/m3 were
measured after the first quarter (two during the first quarter, six during the second
quarter, and seven each during the third and fourth quarters).
Thirteen concentrations of formaldehyde less than 1 |ig/m3 were measured at ORFL,
all of which were measured between January and April. This explains why the first
quarter average concentration of formaldehyde is significantly less than the other
quarterly averages. The maximum concentration of formaldehyde was measured at
ORFL on June 9, 2013 (8.92 |ig/m3) and is more twice the next highest concentration
(3.66 |ig/m3). With some of the lowest and highest concentrations measured during
the second quarter of 2013, it makes sense that the confidence interval for the second
quarter is nearly three times greater than the other confidences intervals, reflecting the
variability in the measurements.
Concentrations of naphthalene measured at SYFL range from 9.52 ng/m3 to
57.1 ng/m3, with a median concentration of 23.3 ng/m3. PAH sampling was
discontinued at SYFL at the end of June 2013; thus, only first and second quarter
average concentrations could be calculated.
PAFL is the only Florida monitoring site that did not sample carbonyl compounds or
PAHs. PMio metals were sampled for at PAFL and arsenic is the only pollutant
10-35

-------
identified as a pollutant of interest for this site. Concentrations of arsenic measured at
PAFL range from 0.264 ng/m3 to 3.20 ng/m3, with a median concentration of
0.53 ng/m3.
•	The confidence interval for the first quarter average concentration of arsenic is just
less than the average itself, indicating the potential for outliers. The maximum arsenic
concentration was measured at PAFL on February 3, 2013 (3.20 ng/m3). The next
highest measurement collected during this quarter was significantly less (1.03 ng/m3)
and all other concentrations measured at this site are less than 2 ng/m3.
Observations from Table 10-5b for WPFL include the following:
•	The period average concentrations for the pollutants of interest for WPFL show that
the period average for naphthalene is an order of magnitude greater than the period
averages of the other PAHs. The period averages also indicate that there is
considerable variability in the measurements of these pollutants.
•	Concentrations of naphthalene measured at WPFL range from 4.32 to 506 ng/m3,
spanning two orders of magnitude. However, the median concentration is only
16.75 ng/m3, less than half the period average concentration. A review of the data
shows that the two highest naphthalene concentrations, 506 ng/m3 measured on
October 25, 2013 and 113 ng/m3 measured on March 11, 2013, are considerably
higher than most of the measurements, the remainder of which are less than 33 ng/m3.
WPFL's maximum naphthalene concentration is the fourth highest naphthalene
measurement across the program; WPFL is one of only three NMP sites with a
naphthalene measurement greater than 400 ng/m3. This explains the significant
increase in the quarterly average shown for the fourth quarter of 2013 as well as the
variability associated with it.
•	For each of the three remaining pollutants of interest, the second quarter 2013 average
concentration is considerably higher than the other quarterly average concentrations;
further, the confidence intervals shown are all greater than each individual average,
indicating possible outliers. The same is true for the third quarter average
concentrations, but to a lesser extent. A review of the data shows that the maximum
concentration of each of these pollutants was measured on June 15, 2013. In each
case, the maximum concentration is more than twice the next highest concentration,
which were also measured on the same day, August 2, 2013. The concentrations
measured on August 2nd were also more than twice the third highest concentration.
For example, the maximum concentration of fluorene measured at WPFL is
30.6 ng/m3 (June 15, 2013), followed by 13.7 ng/m3 (August 2, 2013), and 5.55 ng/m3
(May 10, 2013). All other fluorene concentrations range from 0.507 ng/m3 to
5.39 ng/m3.
10-36

-------
Tables 4-9 through 4-12 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for the Florida
sites from those tables include the following:
•	None of the Florida monitoring sites appear in Table 4-10 for carbonyl compounds.
•	None of the Florida monitoring sites appear in Table 4-11 for naphthalene. WPFL has
the seventh highest annual average concentration of acenaphthene among NMP sites
sampling this pollutant. Note that WPFL is the only non-NATTS site that appears in
Table 4-11.
•	The annual average concentration of arsenic for PAFL ranks sixth highest among
NMP sites sampling PMio metals.
10.4.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for the pollutants shaded in
gray in Table 10-4 for each of the Florida monitoring sites. Figures 10-17 through 10-23 overlay
the sites' minimum, annual average, and maximum concentrations onto the program-level
minimum, first quartile, median, average, third quartile, and maximum concentrations, as
described in Section 3.4.3.1.
Figure 10-17. Program vs. Site-Specific Average Acenaphthene Concentration



H

Program Max Concentration = 123 ng/m3

i
i i i i
1
0	10	20	30	40	50	60	70	80
Concentration {ng/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range


10-37

-------
Figure 10-18. Program vs. Site-Specific Average Acetaldehyde Concentrations
¦
)		
¦
>	¦
K

K

0
3
6 9
Concentration {[jg/m3)

12

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i

Site: Site Average
o
Site Concentration Range


Figure 10-19. Program vs. Site-Specific Average Arsenic (PMio) Concentration
0
12 3
4 5 6
Concentration {ng/m3)
7
8
9
10

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i



Site: Site Average
o
Site Concentration Range




10-38

-------
Figure 10-20. Program vs. Site-Specific Average Fluoranthene Concentration
-
0
5 10 15
20 25 30
Concentration {ng/m3)
35 40
45

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i

Site: Site Average
o
Site Concentration Range


Figure 10-21. Program vs. Site-Specific Average Fluorene Concentration
C
40	50	60
Concentration {ng/m3)
Program: 1st Qua rti le
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


10-39

-------
Figure 10-22. Program vs. Site-Specific Average Formaldehyde Concentrations








0
3 6
9 12 15
Concentration {[jg/m3)
18
21

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i

Site: Site Average
o
Site Concentration Range


Figure 10-23. Program vs. Site-Specific Average Naphthalene Concentrations
0
100
200
300 400 500
Concentration {ng/m3)
600
700

Program:
1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i

Site:
Site Average
o
Site Concentration Range


10-40

-------
Observations from Figures 10-17 through 10-23 include the following:
•	Figure 10-17 presents the box plot for acenaphthene for WPFL, the only Florida site for
which acenaphthene is a pollutant of interest. Note that the program-level maximum
concentration (123 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 80 ng/m3. Figure 10-17 shows
that concentrations of acenaphthene measured at WPFL range from 0.665 ng/m3 to
39.9 ng/m3. The maximum concentration measured at PAFL is considerably less than the
maximum concentration measured among NMP sites sampling PAHs. The annual
average concentration for WPFL is just less than the program-level average and more
than twice the program-level median concentration. Compared to other NMP sites
sampling PAHs, this site's annual average concentration ranks seventh (out of 22), as
discussed in the previous section.
•	Figure 10-18 presents the box plots for acetaldehyde for AZFL, ORFL, SKFL, and
SYFL. The box plots show that the range of acetaldehyde concentrations measured is
smallest for SKFL and largest for ORFL. All of the annual average concentrations
calculated for the Florida sites for which acetaldehyde is a pollutant of interest are less
than the program-level average concentration. The annual averages for SKFL and SYFL
are also less than the program-level median concentration.
•	Figure 10-19 presents the box plot for arsenic for PAFL. The maximum arsenic
concentration measured at PAFL is roughly one-third the maximum arsenic (PMio)
concentration measured across the program. The minimum concentration of arsenic
measured at PAFL is similar to the program-level first quartile and is the highest
minimum arsenic concentration among NMP sites sampling PMio metals. The annual
average concentration of arsenic for PAFL is just greater than the program-level average
concentration (0.67 ng/m3).
•	Figure 10-20 presents the box plot for fluoranthene for WPFL. This box plot shows that
the maximum fluoranthene concentration measured across the program was measured at
WPFL. The minimum fluoranthene concentration measured at WPFL (0.77 ng/m3) is
greater than the program-level first quartile. The annual average concentration for WPFL
is twice the program-level average concentration. WPFL is one of only three NMP sites
for which fluoranthene is a pollutant of interest.
•	Figure 10-21 presents the box plot for fluorene for WPFL. This box plot shows that the
maximum fluorene concentration measured at WPFL is considerably less than the
maximum concentration measured across the program. Three non-detects of fluorene
were measured at WPFL. The annual average concentration for WPFL is 1 ng/m3 less
than program-level average concentration.
•	Figure 10-22 presents the box plots for formaldehyde for AZFL, ORFL, SKFL, and
SYFL. The box plots show that the range of formaldehyde concentrations measured is
smallest for SYFL and largest for ORFL. The maximum concentration measured at
ORFL is more than twice the next highest formaldehyde concentration measured at a
Florida site. All of the annual average concentrations calculated for the Florida sites for
10-41

-------
which formaldehyde is a pollutant of interest are less than the program-level average
concentration as well as the program-level median concentration.
• Figure 10-23 presents the box plots for naphthalene for SKFL and WPFL. Although
naphthalene was also sampled for at SYFL, sampling was discontinued at the end of
June 2013 and thus, no annual average concentration could be calculated. This figure
shows that although the range of concentrations measured was greater for WPFL,
SKFL's annual average concentration (69.26 ± 13.80 ng/m3) is nearly twice WPFL's
(35.74 ± 33.34 ng/m3). However, WPFL's period average is influenced by an outlier, as
discussed in the previous section. Both averages, though, are less than the program-level
average 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 3.4.3.2.
AZFL, ORFL, SKFL, and SYFL have sampled carbonyl compounds under the NMP for at least
5 consecutive years; in addition, sampling for PAHs at SKFL and SYFL and PMio metals at
PAFL began in 2008. Thus, Figures 10-24 through 10-34 present the 1-year statistical metrics for
each of the pollutants of interest for each of these Florida monitoring sites. The statistical metrics
presented for assessing trends include the substitution of zeros for non-detects. If sampling began
mid-year, a minimum of 6 months of sampling is required for inclusion in the trends analysis; in
these cases, a 1-year average concentration is not provided, although the range and percentiles
are still presented. A trends analysis was not conducted for WPFL because this sampling at this
site began in March 2013 and ended in March 2014.
10-42

-------
Figure 10-24. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at AZFL
1

I
1
A
T
t
¦-2-i
I

T

T
2001 2002 2003
2004 2005
2006
2007
Year
2008
2009 2010 2011
2012 2013
O 5th Percentile
- Minimum
~ Median
"
Maximum
O 95th Percentile

Observations from Figure 10-24 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 similar concentrations were also measured in 2003 (8.00 |ig/m3) and 2009
(7.74 |ig/m3).
•	The 1-year average and median concentrations did not change significantly during the
first 2 years of sampling, although the range of measurements is twice as large for
2001 compared to 2002. The 1-year average and median concentrations increased
significantly from 2002 to 2003, stayed elevated through 2004, then began to
decrease significantly, a trend that continued through 2008.
•	The 1-year average and median concentrations began to increase again in 2009. This
increase cannot be attributed to an outlier here or there because nearly all of the
statistical metrics exhibit this increase and the trend continued into 2010. The 95th
percentile more than doubled from 2008 to 2009, and the 1-year average and median
concentrations exhibit a similar increase. A significant decrease is shown for 2011
and continues into 2012, despite the increase in the maximum concentration measured
in 2012. Slight increases in the central tendency statistics are shown for 2013, even
though the range of measurements decreases.
10-43

-------
Figure 10-25. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at AZFL










—






J-
L2-

I rh



\h
L-o-l



^2"







[*i rh rh
		
	



'	E,
J 1-2
V T
		"" t ™
|^J L^J L^J L^J
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
Observations from Figure 10-25 for formaldehyde measurements collected at AZFL
include the following:
•	The maximum formaldehyde concentration was measured in 2001, after which the
highest concentration measured in any given year 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 (although the median concentration
increased for 2010). Little change is shown for the last 3 years of sampling.
•	The trends shown for formaldehyde in Figure 10-25 are almost the opposite of the
trends shown for acetaldehyde in Figure 10-24, 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 less than 2 |ig/m3 between 2011 and 2013,
indicating decreased variability in the measurements collected at AZFL compared to
other years.
10-44

-------
Figure 10-26. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SKFL
2009
Year
0 5th Percentile
- Minimum
~ Maximurr
O 95th Percentile
Observations from Figure 10-26 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 10-26 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. Of the 18 acetaldehyde concentrations greater than 5 |ig/m3,
11 were measured in 2010.
•	Even though the range of concentrations measured decreased by half from 2005 to
2006, the change in the 1-year average concentration is not statistically significant.
After 2006, the 1-year average acetaldehyde concentration increased steadily,
reaching a maximum in 2010. A significant decrease is shown for 2011 and continues
into 2012.
• Although the range of concentrations measured decreased by half for 2013, the 1-year
average concentration changed little.
10-45

-------
Figure 10-27. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at SKFL

Maximum
Concentration for
2005 is 91.7 ng/m3
2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
Observations from Figure 10-27 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). No other concentrations greater than
6 |ig/m3 have been measured at SKFL.
•	For 2005, the 1-year average concentration is greater than the 95th percentile,
reflecting the effect that an outlier can have on statistical measurements. The second
highest concentration measured in 2005 was 4.07 |ig/m3.
•	The 1-year average and median concentrations exhibit 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 measurements.
•	The range of concentrations measured increased significantly from 2010 to 2011 and
the range within which the majority of the concentrations fall, as indicated by the
difference between the 5th and 95th percentiles, more than doubled.
•	All of the statistical parameters increased from 2011 to 2012, indicating that
concentrations of formaldehyde were higher overall at SKFL for 2012. Conversely,
10-46

-------
all of the statistical parameters exhibit a decrease for 2013. Both the minimum and
5th percentile are at a minimum for 2013.
Figure 10-28. Yearly Statistical Metrics for Naphthalene Concentrations Measured at SKFL
20081	2009	2010	2011	2012	2013
Year
O 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 March 2008.
Observations from Figure 10-28 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
concentration 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). Three additional measurements greater than 300 ng/m3 have been
measured at SKFL (one each in 2008, 2010, and 2013).
•	The range within which the majority of naphthalene concentrations fall changed little
through 2011. There is an increase shown for 2012 as this year has the greatest
number of measurements greater than 200 ng/m3 (seven). This increase is followed by
a considerable decrease for 2013, which has the fewest measurements greater than
200 ng/m3 (one).
•	Prior to 2013, the 1-year average concentrations ranged from 82.22 ng/m3 (2011) to
96.91 ng/m3 (2012). For 2013, the 1-year average concentration of naphthalene is at a
10-47

-------
minimum (69.26 ng/m3). Confidence intervals calculated for these averages indicate
that the year-to-year changes shown are not statistically significant.
Figure 10-29. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SYFL
2004	2005	2006	2007
2008	2009
Year
2010	2011	2012	2013
O 5th Percentile	- Minimurr
- Maximurr
O 95th Percentile
Observations from Figure 10-29 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 at SYFL 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
higher 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 two highest
measurements driving this 1-year average concentration.
10-48

-------
• 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 between 2009 and 2012 indicate that the year-to-year changes
are statistically significant, although the undulating pattern indicates no specific trend.
Though most of the statistics exhibit a slight decrease for 2013, the change is not
significant.
Figure 10-30. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at SYFL
2004	2005	2006	2007
2008	2009
Year
2010	2011	2012	2013
O 5th Percentile
O 95th Percentile
~ Average
Observations from Figure 10-30 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 second highest
concentration measured that year is considerably less (4.17 |ig/m3). The 1-year
average concentration exhibits a slight increase from 2004 to 2005 while the median
concentration decreased slightly. The outlier measured in 2005 is mostly reflected in
the confidence intervals calculated for this 1-year average concentration.
10-49

-------
•	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 increased four-fold 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.
•	The 1-year average formaldehyde concentration has fluctuated over the years, ranging
from 1.58 |ig/m3 (2006) to 3.19 |ig/m3 (2007). The 1-year average concentration for
2013 is the lowest since 2006 and has the smallest range of measurements of any year
shown.
Figure 10-31. Yearly Statistical Metrics for Naphthalene Concentrations Measured at SYFL
80
60
40
20
0
2008-
2009
2010
2011
2012
2013
Year
O 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 April 2008.
2 PAH sampling was discontinued at SYFL in June 2013.
Observations from Figure 10-31 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
concentration is not presented for 2008 because a full year's worth of data is not
available, although the range of measurements is provided. In addition, PAH
sampling was discontinued at SYFL in June 2013; thus, a 1-year average
concentration is not presented.
10-50

-------
•	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, changed little
between 2009 and 2012. 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.
•	The 1-year average concentrations have varied from 36.75 ng/m3 (2012) to
43.38 ng/m3 (2010) and confidence intervals calculated for these averages indicate
that the changes over the years are not statistically significant.
•	The range of concentrations measured decreased considerably for 2013, although
only 6 months of measurements are included in the statistical metrics shown.
Figure 10-32. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at ORFL
20031 2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 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 April 2003.
Observations from Figure 10-32 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 concentration is not presented for 2003 because a full year's worth of
data is not available, although the range of measurements is provided.
10-51

-------
•	The maximum acetaldehyde concentration was measured in 2006 (9.55 |ig/m3). The
next three highest concentrations, each between 6 |ig/m3 and 7 |ig/m3, were measured
in 2007, 2008, and 2013.
•	Between 2004 and 2011, the 1-year average concentrations have varied by just less
than 1 |ig/m3, ranging from 1.45 |ig/m3 (2010) to 2.41 |ig/m3 (2006).
•	The 1-year average concentration is at a minimum for 2012 (1.08 |ig/m3) and the
median concentration decreased by almost half from 2011 to 2012. Only one
concentration less than 1 |ig/m3 was measured in 2011 compared to 38 for 2012 (and
no other year has more than 30).
•	All of the statistical metrics exhibit increases for 2013.
Figure 10-33. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at ORFL


































*>
LgJ












TT









¦TW
-©-

r-5-|
L-q-I

r^-
ULm
L2"

—<5—








I


LgJ






20031	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 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 April 2003.
Observations from Figure 10-33 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.
10-52

-------
•	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, starting at 3.27 |ig/m3 for 2004 and reaching a minimum of 1.89 |ig/m3 for
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
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 (9) than 2011 (5).
•	Even though the maximum concentration more than doubled from 2012 to 2013 and
the 95th percentile did not change, the remaining statistical parameters exhibit
decreases for 2013. This is mostly due to a higher number of measurements at the
lower end of the concentration range. The number of concentrations less than 1 |ig/m3
increased from none in 2012 to 13 in 2013.
Figure 10-34. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at PAFL
0.0 +-
2008
2009
2010
Year
2011
2012
2013

O 5th Percentile
- Minimum
— Median ~~
Maximum
O 95th Percentile

10-53

-------
Observations from Figure 10-34 for arsenic measurements collected at PAFL include the
following:
•	Four of the five arsenic concentrations greater than 2 ng/m3 were measured at PAFL
in 2012, and ranged from 2.08 ng/m3 to 3.86 ng/m3. The fifth was measured on
February 3, 2013 (3.20 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 for 2012. The
median concentration exhibits a decreasing trend through 2011, even though the
range of measurements increases for 2011, then increases for 2012.
•	The number of measurements at the upper end of the concentration range has been
increasing at PAFL since 2010, as the number of measurements greater than 1 ng/m3
increased from two in 2010 to five in 2011 to nine in 2012.
•	With the exception of the minimum and 5th percentile, most of the statistical
parameters exhibit a decrease from 2012 to 2013, with the 95th percentile decreasing
by almost half from 2012 to 2013. The number of arsenic measurements greater than
1 ng/m3 returned to five in 2013.
10.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at each Florida monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4 for
definitions and explanations regarding the various toxicity factors, time frames, and calculations
associated with these risk-based screenings.
10.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Florida sites and where annual average
concentrations could be calculated, risk was examined by calculating cancer risk and noncancer
hazard approximations. These approximations can be used as risk estimates for cancer and
noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers want to shift their air
monitoring priorities. Refer to Section 3.4.3.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
10-54

-------
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 Florida Monitoring Sites
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections vs.
# of Samples
Annual
Average
frig/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
St. Petersburg, Florida - AZFL
Acetaldehyde
0.0000022
0.009
59/59
1.63
±0.19
3.58
0.18
Formaldehyde
0.000013
0.0098
59/59
1.77
±0.16
22.95
0.18
Pinellas Park, Florida - SKFL
Acetaldehyde
0.0000022
0.009
60/60
1.39
±0.14
3.05
0.15
Formaldehyde
0.000013
0.0098
60/60
1.54
±0.24
20.03
0.16
Naphthalene3
0.000034
0.003
59/59
69.26
± 13.80
2.35
0.02
Valrico, Florida - SYFL
Acetaldehyde
0.0000022
0.009
61/61
1.30
±0.16
2.85
0.14
Formaldehyde
0.000013
0.0098
61/61
1.82
±0.14
23.69
0.19
Naphthalene3
0.000034
0.003
29/29
NA
NA
NA
Winter Park, Florida - ORFL
Acetaldehyde
0.0000022
0.009
61/61
1.55
±0.28
3.41
0.17
Formaldehyde
0.000013
0.0098
61/61
1.84
±0.31
23.86
0.19
Orlando, Florida - PAFL
Arsenic (PMio)3
0.0043
0.000015
30/30
0.72
±0.22
3.10
0.05
Belle Glade, Florida - WPFL1
Acenaphthene3
0.000088

30/30
4.62
±2.72
0.41

Fluoranthene3
0.000088

30/30
5.62
±3.64
0.49

Fluorene3
0.000088

27/30
3.66
±2.10
0.32

Naphthalene3
0.000034
0.003
30/30
35.74
±33.34
1.22
0.01
— = A Cancer URE or Noncancer RfC is not available.
NA = Not available due to the criteria for calculating an annual average.
3 Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of viewing.
Period averages are provided for WPFL.
10-55

-------
Observations for the Florida sites from Table 10-6 include the following:
•	Formaldehyde has the highest cancer risk approximations among the various
pollutants of interest for the Florida sites. These cancer risk approximations span a
relatively small range (20.03 in-a-million for SKFL to 23.86 in-a-million for ORFL).
•	The cancer risk approximations for acetaldehyde are an order of magnitude less than
the cancer risk approximations for formaldehyde, ranging from 2.85 in-a-million for
SYFL to 3.58 in-a-million for AZFL.
•	For PAFL, arsenic has a cancer risk approximation of 3.10 in-a-million.
•	For the sites sampling naphthalene, the cancer risk approximations range from
1.22 in-a-million (WPFL) to 2.35 in-a-million (SKFL). As previously discussed, an
annual average concentration, and therefore cancer risk and noncancer hazard
approximations, could not be calculated for SYFL for naphthalene.
•	The cancer risk approximations for WPFL's remaining PAH pollutants of interest are
all less than 1 in-a-million.
•	All of the noncancer hazard approximations for the site-specific pollutants of interest
are less than 1.0, indicating that no adverse noncancer health effects are expected
from these individual pollutants. The highest noncancer hazard approximation was
calculated for formaldehyde (0.19), based on the annual average concentration for
SYFL and ORFL.
10.5.2 Risk-Based Emissions Assessment
In addition to the risk-based screening 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 (version 2)
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 3.4.3.4. Lastly,
Table 10-7 provides the pollutants with the highest cancer risk approximations (in-a-million) for
each 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.
10-56

-------
Table 10-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Florida Monitoring Sites
Top 10 Total Emissions for Pollutants
with Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
St. Petersburg, Florida (Pinellas County) - AZFL
Benzene
423.95
Benzene
3.31E-03
Formaldehyde
22.95
Ethylbenzene
262.81
Formaldehyde
2.68E-03
Acetaldehyde
3.58
Formaldehyde
206.42
1,3-Butadiene
1.77E-03

Acetaldehyde
133.62
Naphthalene
6.91E-04
1.3 -Butadiene
58.94
Ethylbenzene
6.57E-04
Naphthalene
20.33
POM, Group 2b
3.19E-04
Dichloromethane
3.85
Acetaldehyde
2.94E-04
POM, Group 2b
3.63
POM, Group 2d
2.68E-04
POM, Group 2d
3.04
Arsenic, PM
2.28E-04
Tetrachloroethylene
1.67
Nickel, PM
1.48E-04
Pinellas Park, Florida (Pinellas County) - SKFL
Benzene
423.95
Benzene
3.31E-03
Formaldehyde
20.03
Ethylbenzene
262.81
Formaldehyde
2.68E-03
Acetaldehyde
3.05
Formaldehyde
206.42
1,3-Butadiene
1.77E-03
Naphthalene
2.35
Acetaldehyde
133.62
Naphthalene
6.91E-04

1,3-Butadiene
58.94
Ethylbenzene
6.57E-04
Naphthalene
20.33
POM, Group 2b
3.19E-04
Dichloromethane
3.85
Acetaldehyde
2.94E-04
POM, Group 2b
3.63
POM, Group 2d
2.68E-04
POM, Group 2d
3.04
Arsenic, PM
2.28E-04
Tetrachloroethylene
1.67
Nickel, PM
1.48E-04

-------
Table 10-7. 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)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Valrico, Florida (Hillsborough County) - SYFL
Benzene
439.99
Formaldehyde
3.47E-03
Formaldehyde
23.69
Ethylbenzene
294.34
Benzene
3.43E-03
Acetaldehyde
2.85
Formaldehyde
266.66
1,3-Butadiene
1.89E-03

Acetaldehyde
166.39
Nickel, PM
1.47E-03
1.3 -Butadiene
63.16
Cadmium PM
1.37E-03
Naphthalene
27.75
Arsenic, PM
1.23E-03
Methyl tert-butyl ether
7.67
Naphthalene
9.43E-04
POM, Group 2b
5.34
Ethylbenzene
7.36E-04
POM, Group 2d
4.24
Hexavalent Chromium
6.15E-04
Nickel, PM
3.07
POM, Group 2b
4.70E-04
Winter Park, Florida (Orange County) - ORFL
Benzene
557.93
Hexavalent Chromium
5.22E-03
Formaldehyde
23.86
Formaldehyde
373.01
Formaldehyde
4.85E-03
Acetaldehyde
3.41
Ethylbenzene
343.02
Benzene
4.35E-03

Acetaldehyde
198.71
1,3-Butadiene
2.41E-03
1,3-Butadiene
80.46
Naphthalene
1.03E-03
Naphthalene
30.26
Ethylbenzene
8.58E-04
POM, Group 2b
6.43
POM, Group 2b
5.65E-04
POM, Group 2d
4.63
Acetaldehyde
4.37E-04
Tetrachloroethylene
2.91
POM, Group 2d
4.08E-04
Dichloromethane
1.09
Arsenic, PM
3.86E-04

-------
Table 10-7. 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)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Orlando, Florida (Orange County) - PAFL
Benzene
557.93
Hexavalent Chromium
5.22E-03
Arsenic
3.10
Formaldehyde
373.01
Formaldehyde
4.85E-03

Ethylbenzene
343.02
Benzene
4.35E-03
Acetaldehyde
198.71
1,3-Butadiene
2.41E-03
1.3 -Butadiene
80.46
Naphthalene
1.03E-03
Naphthalene
30.26
Ethylbenzene
8.58E-04
POM, Group 2b
6.43
POM, Group 2b
5.65E-04
POM, Group 2d
4.63
Acetaldehyde
4.37E-04
Tetrachloroethylene
2.91
POM, Group 2d
4.08E-04
Dichloromethane
1.09
Arsenic, PM
3.86E-04
Belle Glade, Florida (Palm Beach County) - WPFL
Formaldehyde
955.60
Formaldehyde
1.24E-02
Naphthalene
1.22
Benzene
643.30
Naphthalene
1.10E-02
Fluoranthene
0.49
Acetaldehyde
441.40
Benzene
5.02E-03
Acenaphthene
0.41
Ethylbenzene
326.91
1,3-Butadiene
4.13E-03
Fluorene
0.32
Naphthalene
322.43
Acetaldehyde
9.71E-04

1,3-Butadiene
137.62
Ethylbenzene
8.17E-04
POM, Group 2d
8.01
POM, Group 2d
7.05E-04
POM, Group 2b
7.49
POM, Group 2b
6.60E-04
Tetrachloroethylene
4.71
Arsenic, PM
5.84E-04
Dichloromethane
3.63
Nickel, PM
5.20E-04

-------
Table 10-8. 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)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard
Approximations Based on Annual
Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
St. Petersburg, Florida (Pinellas County) - AZFL
Toluene
1,691.40
Acrolein
529,170.22
Formaldehyde
0.18
Xylenes
1,112.81
1,3-Butadiene
29,468.37
Acetaldehyde
0.18
Hexane
837.02
Formaldehyde
21,063.17

Methanol
533.81
Acetaldehyde
14,846.98
Benzene
423.95
Benzene
14,131.55
Ethylbenzene
262.81
Xylenes
11,128.12
Formaldehyde
206.42
Naphthalene
6,776.34
Ethylene glycol
183.89
Lead, PM
4,834.15
Acetaldehyde
133.62
Arsenic, PM
3,541.06
Methyl isobutyl ketone
85.23
Nickel, PM
3,431.37
Pinellas Park, Florida (Pinellas County) - SKFL
Toluene
1,691.40
Acrolein
529,170.22
Formaldehyde
0.16
Xylenes
1,112.81
1,3-Butadiene
29,468.37
Acetaldehyde
0.15
Hexane
837.02
Formaldehyde
21,063.17
Naphthalene
0.02
Methanol
533.81
Acetaldehyde
14,846.98

Benzene
423.95
Benzene
14,131.55
Ethylbenzene
262.81
Xylenes
11,128.12
Formaldehyde
206.42
Naphthalene
6,776.34
Ethylene glycol
183.89
Lead, PM
4,834.15
Acetaldehyde
133.62
Arsenic, PM
3,541.06
Methyl isobutyl ketone
85.23
Nickel, PM
3,431.37

-------
Table 10-8. 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)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard
Approximations Based on Annual
Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Valrico, Florida (Hillsborough County) - SYFL
Toluene
1,908.71
Acrolein
839,881.94
Formaldehyde
0.19
Xylenes
1,141.28
Cadmium, PM
76,216.14
Acetaldehyde
0.14
Hexane
974.23
Nickel, PM
34,087.06

Methanol
723.09
1,3-Butadiene
31,578.65
Benzene
439.99
Formaldehyde
27,210.45
Hydrochloric acid
356.26
Arsenic, PM
19,144.38
Ethylbenzene
294.34
Acetaldehyde
18,488.27
Ethylene glycol
287.12
Hydrochloric acid
17,813.03
Formaldehyde
266.66
Benzene
14,666.49
Acetaldehyde
166.39
Xylenes
11,412.81
Winter Park, Florida (Orange County) - ORFL
Toluene
2,144.16
Acrolein
1,048,114.49
Formaldehyde
0.19
Xylenes
1,437.17
1,3-Butadiene
40,232.28
Acetaldehyde
0.17
Hexane
985.39
Formaldehyde
38,061.79

Methanol
678.41
Hexamethylene-l,6-diisocyanate, gas
30,043.31
Benzene
557.93
Acetaldehyde
22,079.12
Formaldehyde
373.01
Benzene
18,597.79
Ethylbenzene
343.02
Xylenes
14,371.73
Ethylene glycol
268.80
Naphthalene
10,085.90
Acetaldehyde
198.71
Arsenic, PM
5,985.06
Styrene
101.41
Hydrochloric acid
4,682.79

-------
Table 10-8. 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)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard
Approximations Based on Annual
Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Orlando, Florida (Orange County) - PAFL
Toluene
2,144.16
Acrolein
1,048,114.49
Arsenic
0.05
Xylenes
1,437.17
1,3-Butadiene
40,232.28

Hexane
985.39
Formaldehyde
38,061.79
Methanol
678.41
Hexamethylene-l,6-diisocyanate, gas
30,043.31
Benzene
557.93
Acetaldehyde
22,079.12
Formaldehyde
373.01
Benzene
18,597.79
Ethylbenzene
343.02
Xylenes
14,371.73
Ethylene glycol
268.80
Naphthalene
10,085.90
Acetaldehyde
198.71
Arsenic, PM
5,985.06
Styrene
101.41
Hydrochloric acid
4,682.79
Belle Glade, Florida (Palm Beach County) - WPFL
Toluene
2,191.25
Acrolein
1,137,704.46
Naphthalene
0.01
Xylenes
1,392.77
Naphthalene
107,477.97

Hexane
977.01
Formaldehyde
97,510.06
Formaldehyde
955.60
1,3-Butadiene
68,809.22
Methanol
902.57
Chlorine
60,272.30
Benzene
643.30
Acetaldehyde
49,044.32
Acetaldehyde
441.40
Benzene
21,443.18
Ethylbenzene
326.91
Xylenes
13,927.73
Naphthalene
322.43
Nickel, PM
12,047.05
Ethylene glycol
262.17
Manganese, PM
10,798.23

-------
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 10.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
Observations from Table 10-7 include the following:
•	Benzene, ethylbenzene, formaldehyde, and acetaldehyde are the highest emitted
pollutants with cancer UREs in Pinellas, Hillsborough, Orange, and Palm Beach
Counties, although not necessarily in that order.
•	Benzene, formaldehyde, and 1,3-butadiene have the highest toxicity-weighted
emissions for Pinellas County. The same three pollutants have the highest toxicity-
weighted emissions for Hillsborough County but the order is different. Hexavalent
chromium has the highest toxicity-weighted emissions for Orange County, followed
by the other three pollutants. Formaldehyde and naphthalene have the highest
toxicity-weighted emissions for Palm Beach County, followed by benzene and
1,3-butadiene.
•	Eight of the highest emitted pollutants in Pinellas, Orange, and Palm Beach Counties
also have the highest toxicity-weighted emissions while seven of the highest emitted
pollutants in Hillsborough County also have the highest toxicity-weighted emissions.
•	Formaldehyde, which has the highest cancer risk approximations for 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, Orange, and Palm Beach Counties, but
acetaldehyde does not appear among those pollutants with the highest toxicity-
weighted emissions for Hillsborough County (it ranks 12th).
•	Naphthalene, which is a pollutant of interest for SFKL, SYFL, and WPFL, is one of
the highest emitted pollutants in all three counties and has one of the highest toxicity-
weighted emissions for each county. Naphthalene ranks second highest for toxicity-
weighted emissions for Palm Beach County, one of only two counties with NMP sites
where naphthalene ranks this high.
•	Arsenic is the only pollutant of interest for PAFL. Arsenic ranks 10th for toxicity-
weighted emissions for Orange County, but is not among the highest emitted
pollutants, ranking 23rd for quantity emitted. This is an indication of the relative
10-63

-------
toxicity of even a low quantity of emissions. Arsenic appears among those with the
highest toxicity-weighted emissions for all four Florida counties with NMP sites.
•	Fluoranthene, acenaphthene, and fluorene are pollutants of interest for WPFL. These
pollutants are part of POM, Group 2b. POM Group 2b appears on both emissions-
based lists for Palm Beach County, ranking eighth for total emissions and seventh for
toxicity-weighted emissions. POM, Group 2b appears on both emissions-based lists
for all four Florida counties with NMP sites.
•	POM, Group 2d is also among the highest emitted "pollutants" in all four counties
and appears among the pollutants with the highest toxicity-weighted emissions for
three of the four counties. POM, Group 2d includes several PAHs sampled for at
SKFL, SYFL, and WPFL including anthracene, phenanthrene, and pyrene, none of
which failed screens for these sites.
Observations from Table 10-8 include the following:
•	Toluene, xylenes, and hexane are the highest emitted pollutants with noncancer RfCs
in all four Florida counties.
•	Acrolein has the highest toxicity-weighted emissions of the pollutants with noncancer
RfCs for each county, but is not among the highest emitted pollutants in the four
Florida counties. None of the Florida sites sampled VOCs under the NMP.
•	Four of the highest emitted pollutants in Pinellas and Orange Counties also have the
highest toxicity-weighted emissions. Five of the highest emitted pollutants in
Hillsborough and Palm Beach Counties also have the highest toxicity-weighted
emissions. Four of these pollutants are in common amongst the counties:
formaldehyde, acetaldehyde, benzene, and xylenes.
•	Formaldehyde and acetaldehyde appear on both emissions-based lists for each
county. Naphthalene is among the pollutants with the highest toxicity-weighted
emissions for each county (except Hillsborough County) but is not among the highest
emitted (with a noncancer RfC) in any of the counties except Palm Beach County.
For Palm Beach County, naphthalene ranks second highest behind acrolein for
toxicity-weighted emissions, its highest ranking among counties with NMP sites, and
ranks ninth highest for its total emissions. Compared to other counties with NMP
sites, Palm Beach County has the highest naphthalene emissions, which are more than
twice the next highest emissions (Los Angeles County).
•	Several metals appear among those pollutants with the highest toxicity-weighted
emissions for each Florida county, ranking highest for Hillsborough County, but these
metals are not among the highest emitted. Metals were sampled for only at PAFL
under the NMP. Arsenic is the only metal that appears among the pollutants with the
highest toxicity-weighted emissions for Orange County (ranking ninth).
10-64

-------
10.6 Summary of the 2013 Monitoring Data for the Florida Monitoring Sites
Results from several of the data treatments described in this section include the
following:
~~~ Acetaldehyde andformaldehyde failed screens for AZFL and ORFL, where only
carbonyl compounds were sampled. Four pollutants (formaldehyde, acetaldehyde,
naphthalene, and hexavalent chromium) failed screens for SKFL. Three pollutants
(formaldehyde, acetaldehyde, naphthalene) failed screens for SYFL. Arsenic failed
screens for PAFL. Four PAHs failed screens for WPFL.
~~~ Concentrations of acetaldehyde andformaldehyde did not vary significantly among
the Florida sites where carbonyl compounds were sampled. The annual average
concentration of naphthalene for SKFL was nearly twice the annual average
concentration for WPFL, the two sites where annual average concentrations of
naphthalene could be calculated. Arsenic was the only metals identified as a pollutant
of interest for PAFL; its annual average ranked sixth highest among NMP sites
sampling PMw metals.
~~~ After several years of decreasing, concentrations of acetaldehyde appear to be
leveling off at SKFL while concentrations offormaldehyde appear to be decreasing.
Formaldehyde concentrations measured in 2013 at SYFL exhibit the least amount of
variability over the 10 years of sampling.
~~~ Formaldehyde has the highest cancer risk approximation among the pollutants of
interest for each Florida site, where carbonyl compounds were sampled. None of the
pollutants of interest have noncancer hazard approximations greater than an HQ of
1.0.
10-65

-------
11.0	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.
11.1	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 11-1 is a composite satellite image retrieved from ArcGIS Explorer showing the
monitoring site and its immediate surroundings. Figure 11-2 identifies nearby point source
emissions locations by source category, as reported in the 2011 NEI for point sources, version 2.
Note that only sources within 10 miles of the site are included in the facility counts provided in
Figure 11-2. A 10-mile boundary was chosen to give the reader an indication of which emissions
sources and emissions source categories could potentially have a direct effect on the air quality at
the monitoring site. Further, this boundary provides both the proximity of emissions sources to
the monitoring site as well as the quantity of such sources within a given distance of the site.
Sources outside the 10-mile boundary are still visible on the map for reference, but have been
grayed out in order to emphasize emissions sources within the boundary. Table 11-1 provides
supplemental geographical information such as land use, location setting, and locational
coordinates.
11-1

-------
Figure 11-1. Decatur, Georgia (SDGA) Monitoring Site
G\iKon'


i Source: USGS
our.tc NASA NGA USGS
_	Microsoft Corp.

-------
Figure 11-2. NEI Point Sources Located Within 10 Miles of SDGA
DeKalt
County
Rockdale
I Cbyton
County
arioirvY
Not® Ou« 10	a no ;.ollo«i4&on IN» total toCJlffits
dnp«yad -ray not rr»p*rsani all focitteft wtthm ma a {1)
A
Animal Feediof or Farm (1)
a
Landfill (3)
X
Battery Manufacturing Facility (1)
0
Paint and Coating Manufactvnng Facility (2)
c
Chemical Manufacturing Facility (2)
R
Plastic. Res-n, or Rubber Products Plant (3)
1
Compressor Station (1)
F
Prmtlng/Puoltshlng/Paper Product Manufacturing Facility (1)
F
Food Pfocessing/Agrleullurc Facility (3)
X
Rail Yard/Rail Llrve Operations (3)
V
Glass Ptant (1)


11-3

-------
Table 11-1. Geographical Information for the Georgia Monitoring Site
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Additional Ambient Monitoring Information1
SDGA
13-089-0002
Decatur
DeKalb
Atlanta-Sandy
Springs-Roswell,
GA
33.68797,
-84.29048
Residential
Suburban
CO, S02, NOy, NO, N02, NOx, PAMS, Carbonyl
compounds, VOCs, O3, Meteorological parameters, PM10,
PM Coarse, PM10 Speciation, Black carbon PM2.5, and
PM2.5 Speciation, Haze, IMPROVE Speciation, SNMOC
1 Data for additional pollutants are reported to AQS for this site (EPA, 2014b); however, these data are not generated by ERG and are therefore not included in this report.
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, athletic fields, and a middle school surround the monitoring site. A golf course backs
up against the school property on the south and east sides. Interstate-285 is located about one-
half mile north of the site, as shown in Figure 11-1. As Figure 11-2 shows, only one point source
(a food processing facility) is located in close proximity to SDGA. Additional point 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 11-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Georgia monitoring site. Table 11-2 includes both county-level
population and vehicle registration information. Table 11-2 also contains traffic volume
information for SDGA as well as the location for which the traffic volume was obtained.
Additionally, Table 11-2 presents the county-level daily VMT for DeKalb County.
Table 11-2. Population, Motor Vehicle, and Traffic Information for the Georgia Monitoring Site




Annual




Estimated
County-level
Average
Intersection
County-


County
Vehicle
Daily
Used for
level Daily
Site
County
Population1
Registration2
Traffic3
Traffic Data
VMT4
SDGA
DeKalb
713,340
479,533
138,470
1-285, north of Clifton Springs Rd
20,900,748
bounty-level population estimate reflects 2013 data (Census Bureau, 2014)
2County-level vehicle registration reflects 2013 data (GA DOR, 2013)
3AADT reflects 2012 data (GA DOT, 2012)
4County-level VMT reflects 2013 data (GA DOT, 2013)
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 11-2 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 eighth highest compared to other
NMP sites. The traffic estimate provided is for 1-285, north of Clifton Springs Road.
•	The daily VMT for DeKalb County is in the top third compared to other counties with
NMP sites.
11-5

-------
11.2 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.
11.2.1	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. These 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 over the
Atlantic Ocean is a dominant weather feature affecting the Atlanta area, which pulls warm, moist
air into the region. Precipitation is plentiful. Although autumn is the driest season, monthly
rainfall generally ranges between 3 inches and 5 inches. Westerly and northwesterly winds
prevail throughout much of the year, although east winds tend to be more common in the late
summer and fall (Wood, 2004; GSCO, 1998; NCDC, 2015).
11.2.2	Meteorological Summary
Hourly meteorological data for 2013 were retrieved from NCDC for the weather station
closest to the Georgia monitoring site (NCDC, 2013), as described in Section 3.4.2. The closest
weather station to SDGA is located at W. B. Hartsfield/Atlanta International Airport (WBAN
13874). Additional information about the Hartsfield Airport weather station, such as the distance
between the site and the weather station, 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.
11-6

-------
Table 11-3. 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
(33.63, -84.44)
9.6
miles
245°
(WSW)
Sample
Days
(33)
70.0
±4.3
61.5
±4.6
49.4
±5.5
55.2
±4.5
67.5
±5.2
1018.5
± 1.7
7.4
±0.9
2013
70.3
± 1.4
61.7
± 1.4
50.1
± 1.7
55.7
± 1.4
68.5
± 1.5
1018.8
±0.5
6.8
±0.3
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.

-------
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 2013. 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 SDGA were representative of average weather conditions
experienced throughout the year. This is true even though sampling was discontinued at SDGA
in mid-July 2013 and the sample day averages shown in Table 11-3 include only 7 months of
sample days.
11.2.3 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 3.4.2. A wind rose shows the frequency of wind directions using "petals"
positioned around a 16-point compass, and uses different colors to represent wind speeds.
Figure 11-5 presents 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 11-5 also presents three different wind roses for the
SDGA monitoring site. First, a historical wind rose representing 2003 to 2012 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 2013 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2013 is
presented. These can be used to identify the predominant wind speed and direction for 2013 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.
11-8

-------
Figure 11-3. Wind Roses for the Hartsfield International Airport Weather Station near
six; A
Location of SDGA and Weather Station
2003-2012 Historical Wind Rose
A" n,»..

NORTH"

20%

Vsv 16%

12%

8%.

:
'.west:
; ¦ : : EAST



SOUTH
WIND SPEED
(Knots)
I I =22
W1 17-21
B 11 -17
I: "I 7-11
\^3 4-7
¦ 2-4
Calms: 9.11%
2013 Wind Rose
Sample Day Wind Rose
e-:- ,
WIND SPEED
i, Knots)
SOUTH
west:
WIND SPEED
(Knots)
I I >-22
Ml 17-21
LI 11 - 17
O 7- 11
~ 4-7
2- 4
Calms: 5.56%
11-9

-------
Observations from Figure 11-5 for SDGA include the following:
•	The weather station at Hartsfield International Airport is the closest weather station to
SDGA and is located 9.6 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, with northwesterly winds observed the
most. Easterly winds were also common and account for the second highest
percentage of observations. Winds from the north to northeast were rarely observed.
Calm winds (less than or equal to 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, although winds from the east and northwest account for a higher percentage of
wind observations, particularly east winds, which account for nearly 16 percent of
observations in 2013.
•	Easterly winds were prevalent on sample days, as shown on the sample day wind
rose, accounting for more than 16 percent of wind observations. Although west-
northwesterly to north-northwesterly winds were observed, the percentage is
considerably reduced, while westerly winds account for a greater percentage of
observations on the sample day wind rose compared to the full-year wind rose. Winds
from the southeast and southwest quadrants also account for a greater percentage of
wind observations near SDGA. Finally, the percentage of calm winds on sample days
is reduced by nearly half. Recall, though, that sampling was discontinued in July at
SDGA; thus, a wind rose for a full year of sample days may look different.
11.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for SDGA 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 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. Hexavalent
chromium was sampled for at SDGA, although sampling was discontinued at SDGA in mid-July.
11-10

-------
Table 11-4. Risk-Based Screening Results for the Georgia Monitoring Site
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Decatur, Georgia - SDGA
Hexavalent Chromium
0.000083
1
8
12.50
100.00
100.00
Total
1
8
12.50

Observations from Table 11-4 for SDGA include the following:
•	Thirty valid hexavalent chromium samples were collected at SDGA between
January 4, 2013 and July 15, 2013, in which hexavalent chromium was detected in
eight samples.
•	A single measurement failed a screen for SDGA, which represents a 12.50 percent
failure rate.
11.4 Concentrations
This section typically presents various concentration averages used to characterize
pollution levels at the monitoring site for each of the site-specific pollutants of interest. However,
the short sampling duration at SDGA prevents an annual average concentration for hexavalent
chromium to be calculated. In order to facilitate a review of the data collected at SDGA in 2013,
a few statistical calculations are provided in the sections that follow. Site-specific statistical
summaries for SDGA are also provided in Appendix O. Concentration averages and statistical
metrics are also presented from previous years of sampling in order to characterize concentration
trends at the site. The concentration comparison analysis was not performed.
11.4.1 2013 Concentration Averages
Quarterly concentration averages were calculated for hexavalent chromium for SDGA
site, as described above. 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, which includes all
measured detections and substituted zeros for non-detects for the entire year of sampling, could
not be calculated as sampling at SDGA was discontinued at the end of July 2013. Quarterly
average concentrations for SDGA are presented in Table 11-5, where applicable. Note that if a
11-11

-------
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 11-5. 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
Hexavalent Chromium
8/30
0.002
± 0.003
0.015
±0.016
NA
NA
NA
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
Observations for SDGA from Table 11-5 include the following:
•	Hexavalent chromium was detected in roughly 27 percent of the samples collected at
SDGA.
•	Measured detections of hexavalent chromium range from 0.0068 ng/m3 to
0.103 ng/m3.
•	There were only two measured detections during the first quarter of 2013 and both
were measured in samples collected in January; thus, only non-detects were measured
in February and March. For the second quarter, there were no measured detections in
April, two measured detections were measured in May samples, including the
maximum concentration measured on May 22, 2013, and three were collected in June
samples. The final measured detection was measured in a sample collected on the
final sample day, July 15, 2015.
•	The relatively large number of non-detects included in the available quarterly average
concentration calculations explains why the confidence interval is greater than the
average itself for each quarterly average.
•	Third and fourth quarter averages, as well as an annual average, could not be
calculated because hexavalent chromium sampling was discontinued in July.
However, statistical summaries for the entire period of sampling at SDGA are
provided in Appendix O.
11-12

-------
11.4.2 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.3.2.
SDGA has sampled hexavalent chromium under the NMP since 2005. Thus, Figure 11-4
presents the 1-year statistical metrics for this pollutant 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 concentration is not provided, although the range and
percentiles are still presented.
Figure 11-4. Yearly Statistical Metrics for Hexavalent Chromium Concentrations
Measured at SDGA
0.35 -
0.30 -













0.05 -
0.00 -



T




¦ Hi Hi r
pL —i—
—
9—
2005 1	2006	20072	20082	2009	2010	2011	2012	20133
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until late February 2005.
21-Year averages are not presented because there was a break in sampling between Sept 2007 and May 2008.
3 A 1-year average is not presented because sampling under the NMP was discontinued in July 2013.
11-13

-------
Observations from Figure 11-4 for hexavalent chromium measurements collected at
SDGA include the following:
•	Because hexavalent chromium sampling under the NMP began at SDGA in late
February 2005, a 1-year average is not presented for 2005. 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. A 1-year
average is also not presented for 2013 as sampling at SDGA was discontinued in
July 2013.
•	The maximum concentration was measured in 2006 (0.300 ng/m3). Five of the six
concentrations greater than 0.1 ng/m3 measured at SDGA were measured in either
2005 or 2006. The sixth was measured on May 22, 2013.
•	The difference between the 5th and 95th percentiles exhibits little change over the
years of sampling after 2006, indicating that a majority of the measurements fall
within a fairly similar range.
•	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 the onset of sampling in 2005, the number
of non-detects has varied from 5 percent (2007) to 73 percent (2013). Note,
however, that both 2007 and 2013 were partial sampling years. The median
concentration increased considerably from 2009 to 2010, then changed little
through 2012. This is also true for the 1-year average concentrations for 2010
through 2012. The median concentration returned to zero for 2013 even though
the range of measurements is at its largest since 2006.
11.5 Additional Risk-Based Screening Evaluations
In order to characterize risk at participating monitoring sites, additional risk-based
screening evaluations were conducted. Because an annual average could not be calculated for the
pollutant sampled at SDGA, cancer risk and noncancer hazard approximations, as described in
Section 3.4.3.3, were not calculated. The risk-based emissions assessment described in
Section 3.4.3.4 was still conducted, at least in part, as the emissions can be reviewed independent
of concentrations measured.
11.5.1 Risk-Based Emissions Assessment
This section presents an evaluation of county-level emissions based on cancer and
noncancer toxicity, respectively, and is intended to help policy-makers prioritize their air
monitoring activities. Table 11-6 presents the 10 pollutants with the highest emissions from the
2011 NEI (version 2) that have cancer toxicity factors. Table 11-6 also presents the 10 pollutants
11-14

-------
with the highest toxicity-weighted emissions, based on the weighting schema described in
Section 3.4.3.4. The emissions and toxicity-weighted emissions are shown in descending order in
Table 11-6. Table 11-7 presents similar information, but is limited to those pollutants with
noncancer toxicity factors. Because not all pollutants have both cancer and noncancer toxicity
factors, the highest emitted pollutants in the cancer table may be different from the noncancer
table, although the actual quantity of emissions is the same. A more in-depth discussion of this
analysis is provided in Section 3.4.3.4.
Observations from Table 11-6 include the following:
•	Benzene, ethylbenzene, and formaldehyde are the highest emitted pollutants with
cancer UREs in DeKalb 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 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. Hexavalent
chromium ranks 28th for total emissions and 12th for its toxicity-weighted emissions.
Observations from Table 11-7 include the following:
•	Toluene, hexane, and xylenes are the highest emitted pollutants with noncancer RfCs
in DeKalb County.
•	The pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) are acrolein, 1,3-butadiene, and formaldehyde.
•	Four of the highest emitted pollutants in DeKalb County also have the highest
toxicity-weighted emissions.
•	Hexavalent chromium is not among the highest emitted pollutants in DeKalb County
nor is it among those with the highest toxicity-weighted emissions (of the pollutants
with noncancer RfCs). Hexavalent chromium ranks 55th for total emissions and 29th
for its toxicity-weighted emissions.
11-15

-------
Table 11-6. 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)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Decatur, Georgia (DeKalb County) - SDGA
Benzene
188.80
Formaldehyde
1.54E-03

Ethylbenzene
128.35
Benzene
1.47E-03
Formaldehyde
118.30
1,3-Butadiene
9.04E-04
Acetaldehyde
79.46
Naphthalene
4.67E-04
1.3 -Butadiene
30.14
Ethylbenzene
3.21E-04
T etrachloroethylene
19.09
POM, Group 2b
2.45E-04
Naphthalene
13.72
POM, Group 2d
2.07E-04
POM, Group 2b
2.78
Acetaldehyde
1.75E-04
POM, Group 2d
2.35
POM, Group 5a
1.56E-04
T richloroethylene
2.32
Arsenic, PM
1.54E-04

-------
Table 11-7. 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)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Decatur, Georgia (DeKalb County) - SDGA
Toluene
834.60
Acrolein
336,255.68

Hexane
484.49
1.3 -Butadiene
15,072.00
Xylenes
474.91
Formaldehyde
12,071.56
Methanol
395.10
Acetaldehyde
8,828.36
Benzene
188.80
Benzene
6,293.24
Ethylene glycol
137.48
Xylenes
4,749.05
Ethylbenzene
128.35
Naphthalene
4,574.95
Formaldehyde
118.30
Lead, PM
3,306.94
Acetaldehyde
79.46
Arsenic, PM
2,388.76
Methyl isobutyl ketone
55.33
T richloroethy lene
1,159.47

-------
11.6 Summary of the 2013 Monitoring Data for SDGA
Results from several of the data treatments described in this section include the
following:
~~~ Hexavalent chromium was the only pollutant sampled for at SDGA in 2013. Sampling
was discontinued at this location in mid-July.
~~~ Hexavalent chromium was detected in about one-quarter of samples collected.
Concentrations of this pollutant failed only one screen for SDGA.
11-18

-------
12.0	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.
12.1	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 CBSA. Figures 12-1 and 12-2 are composite satellite images
retrieved from ArcGIS Explorer showing the Chicago monitoring sites and their immediate
surroundings. Figure 12-3 identifies the nearby point source emissions locations by source
category, as reported in the 2011 NEI for point sources, version 2, for NBIL and SPIL. Note that
only sources within 10 miles of the sites are included in the facility counts provided in
Figure 12-3. A 10-mile boundary was chosen to give the reader an indication of which emissions
sources and emissions source categories could potentially have a direct effect on the air quality at
the monitoring sites. Further, this boundary provides both the proximity of emissions sources to
the monitoring sites as well as the quantity of such sources within a given distance of the sites.
Sources outside the 10-mile boundaries are still visible on the map for reference, but have been
grayed out in order to emphasize emissions sources within the boundary. Figures 12-4 and 12-5
present the composite satellite image and facility map for ROIL, respectively. Table 12-1
provides supplemental geographical information such as land use, location setting, and locational
coordinates for each site.
12-1

-------
Figure 12-1. Northbrook, Illinois (Mill,) Monitoring Site
to
to

fcdensExpy Spur Edfins ^ $ £
or t
w \
\\
\\\


•««.' «f»V
,N=\ Edens £Xp
}ys,
*V
Jf L V j«. »
\ in
\l
£ $
\
\ ~ \ l \	X	i
\\ w\	3
n \r ^ifT W 
- 1 ?Y A ¦ \

\
•r i\
t si
t .*
p
'iJ?

£,5Jp,
S
9
idee Rd 4a
Dundee Rd
*Jr I	F| \
¦I'n v	p ¦ \
1.	i
T I	, ,
\ \ %
'*1
k	«¦*
\ £
\ I.
\
<1
p-
Ji
*
j-
1/
%

SI-
I JR.i »,JUtl
a Dundee Rd	Dundee Rd
/ i J
I ¦ Wffl|glf|> l	J0\
/ /I'
/JB/ r
/	V
/	*1
$
\\ »1
W
-------
Figure 12-2. Schiller Park, Illinois (SPIL) Monitoring Site

-------
Figure 12-3. NEI Point Sources Located Within 10 Miles of NBIL and SPIL
Lake
County
La ht
M>crvy»r?
Cook
County
U7 4U &-W 87 4Str\V
19 to faeftty	and <
4atrw
"T	%mmo* C«inln>t • V,
^	flDdj£»Oi.rt* MO AuMI FWI .!>
0	tnc* *»*V vn 11)	V
«"l*we*1fu;e f k«v 1*1	^
^ 111	*
—u iwmwon >o»>Ciiiwiiwwti)	~
f oiMm »i)	O
B	A
C C*«rt»r» (»l	¦
1	A
@
^ D7 Omw] ***n fV3|	¦
6 Bttt»o(&|»
# gnWH^WIIWM •^OffWOXtW ,7|	v
E ?pcrxw«l) ^ni'J -njsnr^ inyjz'v ana Cow»fJ :#T|
F Ego* Piecu^ Hniigifcff F«t»^ i«S>	U
Rwwm uehmi&mfl'	Lrf
Hwfwm >—1 >"«* t Ki	R
UNMtM S«Aw» fUAlo Of	t
P
l k9M*>UoMai I T ,7)	ffi
w^itwwi Mechwv IT t »#rwrt PWt i)B
UOui*f*i hmpM (MCr M: l ffft	~
MtMlM	IT
MM Sut «il MM CtMft	Oi ^
trn* fl3*t»*9 ft.lt-.03 M Vwe £«mca* Ic U«xUiuwi »2tl J
tMttt Htom, 1 m	^
^11'UiWfVtM t'l-r—m * acMr i»l
•Mc0«a*cat» ;«rN» n»n»«nwi ai	4
(Hart irK- r^Mt«Q W»r»«t»r».i»Yj » 11 i	W
* W«cw»ic* M«n»»»**krtH T>
^««U A«MI * XiDtm hMKU fw ,14 .
Gnttr« A OiMf u* N»«t» f*** Q|
*<•«««]PtttfaU MMttUmi Fu» 1M1
"I
•Ul iMHNUaCWMMtl
9MMA
%al^Ubu*k»«kli|
W«m ¦•m aC4-^nwi'i
f*.m Uait/Kl4in} ('}
thWii>rn#w Caitndwi r?1
MMMM' Tf^wrc F«r** ||H
*4)
ftuoffKM MAit MIko* « mom Mc*»m«dt?»i
NBIL NATTS site ^ SPIL UATMP site	10 mile radius 	 County boundary
Source Category Group (No. of Facilities)
•frMOTW fH 4UV#
Legend
BT'49VW
	T	
DuPage
Coun^
¦ -

Soft Que to faeftty dansity a*hJ coilncaBon the total facilitie?
displayed rr®v not 'epreseit all focittes v%tth»n tfie afB of nteresl
srww
inrmv
12-4

-------
Figure 12-4. Roxana, Illinois (ROIL) Monitoring Site

-------
Figure 12-5. NEI Point Sources Located Within 10 Miles of ROIL
ILLINOIS
St Cham*
County
Missouri
SL Clan
County
Koto. Due to focftty density and coUacabon the total facilities
Legend	may not rapfwit all wtMn en« a<»« of nmms t
ROIL UATMP site jjf S4MO NATTS site	10 mile radius 	County boundary
Source Category Group (No. of Facilities)
•J" Aerospace.'Airoaft Manufacturing Facility (1}
TAuport/AlrftneJAirport Support Operations (7]
» Asphalt Production/Hot Mi* Asphalt Plant (4)
Bnck. Structural Clay or Qay Ceramics Plant (t)
B Bulk Terminals/Bulk Plants <11J
C Chemical Manufacturing f acitily (3)
i Compressor Stabon (7)
Crematory • Animal'Hurnan (3)
'I'Dry Cleaning Facility (1)
*	Electricity Generation via Combusbon (3)
F Food Processing/AgncuItLn® FacSty (3)
£. Foundries Monferroua (1)
¦ Gaeoime'Dies^ Service Staiieo (1>
•	Industrial Machinery or Equipment Plant (1)
O Institutional (schoa. hospital, pnson. etc.) (5)
•	Landfill (2)
A Metal Coating Engraving and Allied Services to Manufacturers (2)
<•> Metals Processlng'Fabrication Faculty (3)
- Mine/Quarry'Mineral Processing Facility (12)
? Miscellaneous Commercial/Industrial Facility (12)
< Pesticide Manufacturing Plant (1)
•	Petroleum Products Manufacturing (1)
4 Petroleum Refinery (1)
•	Port and Harbor Operations (2)
P Pnnbng>'Publishingi'Paper Product Manufacturing Faculty (1)
IS Putp and Paper Plant (1)
X Rail Yard/Rail Line Operations (1)
V steel MID {2)
•	Wbstevrater Treatment Facility fit
12-6

-------
Table 12-1. Geographical Information for the Illinois Monitoring Sites
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Additional Ambient Monitoring Information1
NBIL
17-031-4201
Northbrook
Cook
Chicago-
Naperville-Elgin
IL-IN-WI
42.139996,
-87.799227
Residential
Suburban
TSP, TSP Metals, CO, S02, NO, N02, NOx, NOy, 03,
Meteorological parameters, PMio, PM2 5, PM2 5
Speciation IMPROVE Speciation.
SPIL
17-031-3103
Schiller
Park
Cook
Chicago-
Naperville-Elgin
IL-IN-WI
41.965193,
-87.876265
Mobile
Suburban
TSP, TSP Metals, NO, N02, NOx, O3 Meteorological
parameters, PM2 5, PM2 5 Speciation.
ROIL
17-119-9010
Roxana
Madison
St. Louis, MO-IL
38.848382,
-90.076413
Industrial
Suburban
IMPROVE Speciation Meteorological parameters,
PM2 5 Speciation.
1 Data for additional pollutants are reported to AQS for these sites (EPA, 2014b); however, these data are not generated by ERG and are therefore not included in this report.
BOLD ITALICS = EPA-designated NATTS Site

-------
NBIL is located on the property of the Northbrook Water Filtration Station. Figure 12-1
shows that NBIL is located off State Highway 68 (Dundee Road), near Exit 30 on 1-94. A
railway runs north-south in front of the water filtration station, separating the municipal
buildings and nearby residential subdivision from a business complex to the east, and intersects
Dundee Road just south of the monitoring site. The surrounding area is classified as suburban
and residential. Commercial, residential, and forested areas surround the site, 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 one-
half mile from the site. The surrounding area is classified as suburban and mobile. Commercial
and residential areas are located to the east of the airport and 1-294. The rail yard located to the
east of 1-294 is an intermodal terminal/facility that has been closed (Podmolik, 2015).
NBIL and SPIL are located within 13 miles of each other. Each site is located within
10 miles of numerous point sources, although the quantity of emissions sources is higher near
SPIL than NBIL, as shown in Figure 12-3. 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 12-3; 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 12-4 shows. Just north of the monitoring site are a junior high school and a
high school, whose track and tennis courts are shown across the street from the monitoring site.
North of the schools is a community park. Ambient monitoring data from this location will be
used to assess near-field concentrations in the neighboring community, with emphasis on
comparing and contrasting these data to the St. Louis NATTS site (S4MO), which is also
12-8

-------
pictured in Figure 12-5. The Mississippi River, which is the border between Missouri and
Illinois, is just over a mile and a half west of the ROIL monitoring site.
In addition to showing the ROIL monitoring site's location relative to the S4MO
monitoring site, Figure 12-5 also shows the point sources within 10 miles of each site (although
only the facilities within 10 miles of ROIL are included in the facility counts below the map).
There is a large cluster of emissions sources surrounding and mostly to the south and northwest
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 12-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Illinois 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 Cook County and Madison
County.
Table 12-2. Population, Motor Vehicle, and Traffic Information for the Illinois Monitoring Sites
Site
County
Estimated
County
Population1
County-level
Vehicle
Registration2
Annual
Average
Daily
Traffic3
Intersection
Used for
Traffic Data
County-
level Daily
VMT4 "
NBIL
Cook
5,240,700
2,074,419
115,700
1-94 north of Dundee Rd
87,972,644
SPIL
186,100
1-294, just south of Lawrence Ave
ROIL
Madison
267,225
267,302
7,750
S Central Ave at Hawthorne Ave
7,911,443
bounty-level population estimates reflect 2013 data (Census Bureau, 2014)
2County-level vehicle registrations reflect 2014 data (IL SOS, 2014)
3AADT reflects 2012 data for SPIL and 2013 data for NBIL and ROIL (IL DOT, 2012/2013)
4County-level VMT reflects 2013 data (IL DOT, 2013)
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 12-2 include the following:
• Cook County has the second highest county-level population (behind only Los
Angeles County, California) and fourth highest county-level vehicle registration
compared to other counties with NMP sites.
12-9

-------
•	Both the county-level population and vehicle registration for Madison County are an
order of magnitude less than those for Cook County and rank 33rd for population and
31st for vehicle registration compared to other counties with NMP sites. Note the
difference between the population and vehicle registration for these two counties.
There is a nearly one-to-one ratio of vehicles to people in Madison County while the
population of Cook County is more than double the number of registered vehicles.
•	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 fourth highest among all NMP sites. 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 interstate highways while the traffic volume for ROIL is not.
•	The Cook County daily VMT is nearly 88 million miles and ranks third highest
among counties with NMP sites, behind only Los Angeles County, California and
Maricopa County, Arizona. 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.
12.2 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.
12.2.1 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 (IL SCO, 2015; Wood,
2004).
12-10

-------
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
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 occurred within the region. Southerly winds prevail in the summer and fall while
northwesterly winds are prevalent during the colder months of the year. (Wood, 2004; MCC,
2015).
12.2.2 Meteorological Summary
Hourly meteorological data for 2013 were retrieved from NCDC for the weather stations
closest to the Illinois monitoring sites (NCDC, 2013), as described in Section 3.4.2. 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 12-3. These data were
used to determine how meteorological conditions on sample days vary from conditions
experienced throughout the year.
12-11

-------
Table 12-3. Average Meteorological Conditions near the Illinois Monitoring Sites
IO
to
Closest Weather
Station
Distance
and

Average
Maximum
Average
Average
Dew Point
Average
Wet Bulb
Average
Relative
Average
Sea Level
Average
Scalar Wind
(WBAN and
Coordinates)
Direction
from Site
Average
Type1
Temperature
(°F)
Temperature
(°F)
Temperature
(°F)
Temperature
(°F)
Humidity
(%)
Pressure
(mb)
Speed
(kt)
Northbrook, Illinois - NBIL
Palwaukee Municipal
5.6
miles
Sample
Days
56.5
48.5
38.4
43.8
70.2
1017.9
6.2
Airport
(62)
±5.7
±5.4
±5.4
±5.0
±2.8
±2.0
±0.5
04838
(42.12, -87.90)
256°
(WSW)

57.2
49.3
38.7
44.3
69.1
1017.5
6.5

2013
±2.2
±2.1
±2.0
± 1.9
± 1.1
±0.7
±0.3
Schiller Park, Illinois - SPIL
O'Hare International
3.6
miles
Sample
Days
57.1
49.3
39.4
44.7
70.8
1017.3
8.0
Airport
(62)
±5.9
±5.5
±5.5
±5.1
±3.0
±2.0
±0.6
94846
305°
(NW)








(42.00, -87.93)

57.3
49.5
39.2
44.7
69.8
1016.8
8.4

2013
±2.2
±2.1
±2.0
± 1.9
± 1.2
±0.7
±0.3
Roxana, Illinois - ROIL
Lambert/
St. Louis International
Airport
13994
(38.75, -90.37)
17.4
miles
Sample
Days
(61)
63.4
±5.5
55.4
±5.3
42.6
±5.4
49.1
±4.8
64.7
±3.0
1018.5
±2.0
7.0
±0.6
248°
(WSW)
2013
65.2
±2.1
56.5
±2.0
43.1
±2.0
49.8
± 1.8
63.3
± 1.1
1017.9
±0.7
7.3
±0.3
1 Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.

-------
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 2013. 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 near NBIL, SPIL, and ROIL were representative of average weather
conditions experienced throughout the year near these sites. The largest difference shown in
Table 12-3 is for ROIL and the temperature parameters, although the difference is not
significant. 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.
12.2.3 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 3.4.2. A wind rose shows the frequency of wind directions using "petals"
positioned around a 16-point compass, and uses different colors to represent wind speeds.
Figure 12-6 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 12-6 also presents three different wind roses for the
NBIL monitoring site. First, a historical wind rose representing 2003 to 2012 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 2013 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2013 is
presented. These can be used to identify the predominant wind speed and direction for 2013 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically. Figures 12-7 and 12-8 present the distance map and three
wind roses for SPIL and ROIL, respectively.
12-13

-------
westI
Figure 12-6. Wind Roses for the Palwaukee Municipal Airport Weather Station near NBIL
Location of NBIL and Weather Station	2003-2012 Historical Wind Rose
WIND SPEED
(Knots)
~
~1 4-7
2- 4
Calms: 16.14%
2013 Wind Rose
VEST
(Knots)
11 -17
WIND SPEED
Calms: 16.88%
Sample Day Wind Rose
NORTH""-.
WEST
'A'INC S HE EC
(Kn ots i
11 -17
SOUTH
Calms: 16.88%
12-14

-------
Figure 12-7. Wind Roses for the O'Hare International Airport Weather Station near SPIL
Location of SPIL and Weather Station	2003-2012 Historical Wind Rose
¦NORTH"--.,
west:
WIND SPEED
(Knots)
I I	>-22
EZI	17-21
|	11 - 17
[ |	7- 11
~1	4-7
IH 2- 4
2013 Wind Rose
WEST
(Knots)
11 -17
SOUTH
WIND SPEED
Calms: 6.33%
Sample Day Wind Rose
N O RT H""
WEST
(Knots)
SOUTH
WIND SPEED
WMthar
5'sUort
12-15

-------
Figure 12-8. Wind Roses for the Lambert/St. Louis International Airport Weather Station
near ROIL
Location of ROIL and Weather Station	2003-2012 Historical Wind Rose
WIND SPEED
(Knots)
I I =22
n 17-21
H 11 - 17
I ll 7- 11
\^3 4-7
H 2-4
Calms: 11.51%
west:
! EAS
2013 Wind Rose
Sample Day Wind Rose
15%
12%
west:
west:
WIND SPEED
(Knots)
~ «22
F~B 17-21
II 11-17
WIND SPEED
(Knots)
~ >=22
EH 17-21
| 11-17
SOUTH
12-16

-------
Observations from Figure 12-6 for NBIL include the following:
•	The Palwaukee Municipal Airport weather station is located 5.6 miles west-southwest
of NBIL, and about four times as far from Lake Michigan as NBIL.
•	The historical wind rose shows that winds from a variety of directions were observed
near NBIL. Winds from the south, south-southwest, and west together account for
one-quarter of wind observations while winds with a northerly component each
account for 5 percent to 6 percent of observations. Winds from the east-southeast to
south-southeast were observed the least often. Calm winds (those less than or equal to
2 knots) were observed for approximately 16 percent of the hourly measurements.
•	The 2013 wind rose exhibits similar patterns in wind speed and direction as the
historical wind rose.
•	The sample day wind patterns resemble the full-year wind patterns in that the
majority of wind observations are associated with a direction on the left-hand side of
the wind rose. However, the percentages are more variable. For instance, fewer
southerly winds were observed on sample days while a greater percentage of south-
southwesterly to west-southwesterly winds were observed. Also, winds appear lighter
on sample days; winds speeds greater than 11 knots account for fewer observations
on sample days than throughout the year. However, the calm rate is the same for both
wind roses.
Observations from Figure 12-7 for SPIL include the following:
•	The O'Hare International Airport weather station is located 3.6 miles 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 near this site, 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 were observed for less than
8 percent of the hourly measurements.
•	The 2013 wind rose exhibits similar patterns in wind speed and direction as the
historical wind rose, although winds from the west accounted for a higher percentage
of the wind observations. The strongest winds were from the southwest quadrant and
west.
•	The sample day wind patterns resemble those of the full-year wind rose, with the
winds from the south to southwest to west accounting for nearly half of the wind
observations. Wind speeds appear lower on sample days. A review of the wind data
shows that only one of the windiest days in 2013, based on average scalar wind
speed, was a sample day at SPIL.
12-17

-------
Observations from Figure 12-8 for ROIL include the following:
•	The Lambert/St. Louis International Airport weather station is located 17.4 miles
west-southwest of ROIL. The airport lies on the northwest side 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 were observed for less than 12 percent of the hourly
measurements.
•	The 2013 wind rose exhibits similar patterns in wind speed and direction as the
historical wind rose, although the calm rate is slightly less.
•	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-northeast and east
account for a greater percentage of winds on sample days while winds from the east-
southeast and southeast account for fewer wind observations. There were also fewer
northwesterly wind observations on sample days. The percentage of calm winds is
also less than 10 percent.
12.3 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 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. VOCs, carbonyl compounds, SNMOCs, metals (PMio), PAHs, and hexavalent
chromium were sampled for atNBIL, while only VOCs and carbonyl compounds were sampled
for at SPIL and ROIL.
12-18

-------
Table 12-4. Risk-Based Screening Results for the Illinois Monitoring Sites
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Northbrook, Illinois - NBIL
Acetaldehyde
0.45
62
62
100.00
12.16
12.16
Formaldehyde
0.077
62
62
100.00
12.16
24.31
Benzene
0.13
61
61
100.00
11.96
36.27
Carbon Tetrachloride
0.17
61
61
100.00
11.96
48.24
1,2-Dichloroethane
0.038
53
53
100.00
10.39
58.63
Arsenic (PMio)
0.00023
47
59
79.66
9.22
67.84
Naphthalene
0.029
45
58
77.59
8.82
76.67
1.3 -Butadiene
0.03
33
38
86.84
6.47
83.14
Acenaphthene
0.011
29
58
50.00
5.69
88.82
Fluorene
0.011
27
57
47.37
5.29
94.12
Fluoranthene
0.011
13
58
22.41
2.55
96.67
Ethylbenzene
0.4
4
61
6.56
0.78
97.45
Clilorofonn
9.8
3
61
4.92
0.59
98.04
/?-Dichlorobcnzcnc
0.091
2
8
25.00
0.39
98.43
Dicliloromethane
60
2
61
3.28
0.39
98.82
T richloroethylene
0.2
2
14
14.29
0.39
99.22
Bromofonn
0.91
1
6
16.67
0.20
99.41
Hexacliloro -1,3 -butadiene
0.045
1
1
100.00
0.20
99.61
Hexavalent Chromium
0.000083
1
13
7.69
0.20
99.80
T etrachloroethylene
3.8
1
48
2.08
0.20
100.00
Total
510
900
56.67

Schiller Park, Illinois - SPIL
Acetaldehyde
0.45
60
61
98.36
15.08
15.08
Benzene
0.13
60
60
100.00
15.08
30.15
Carbon Tetrachloride
0.17
60
60
100.00
15.08
45.23
Formaldehyde
0.077
60
61
98.36
15.08
60.30
1.3 -Butadiene
0.03
59
59
100.00
14.82
75.13
1,2-Dichloroethane
0.038
57
57
100.00
14.32
89.45
T richloroethylene
0.2
18
44
40.91
4.52
93.97
Hexacliloro -1,3 -butadiene
0.045
9
10
90.00
2.26
96.23
/?-Dichlorobcnzcnc
0.091
5
28
17.86
1.26
97.49
Propionaldehyde
0.8
5
60
8.33
1.26
98.74
Ethylbenzene
0.4
4
60
6.67
1.01
99.75
1,2-Dibromoethane
0.0017
1
1
100.00
0.25
100.00
Total
398
561
70.94

12-19

-------
Table 12-4. Risk-Based Screening Results for the Illinois Monitoring Sites (Continued)
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Roxana, Illinois - ROIL
Acetaldehyde
0.45
61
61
100.00
16.40
16.40
Formaldehyde
0.077
61
61
100.00
16.40
32.80
Benzene
0.13
60
60
100.00
16.13
48.92
Carbon Tetrachloride
0.17
60
60
100.00
16.13
65.05
1.3 -Butadiene
0.03
51
54
94.44
13.71
78.76
1,2-Dichloroethane
0.038
45
45
100.00
12.10
90.86
Hexachloro -1,3 -butadiene
0.045
15
15
100.00
4.03
94.89
Ethylbenzene
0.4
13
60
21.67
3.49
98.39
Propionaldehyde
0.8
3
61
4.92
0.81
99.19
/?-Dichlorobcnzcnc
0.091
2
28
7.14
0.54
99.73
1,2-Dibromoethane
0.0017
1
1
100.00
0.27
100.00
Total
372
506
73.52

Observations from Table 12-4 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 pollutants measured at each
site.
•	Twenty pollutants failed at least one screen for NBIL; 57 percent of concentrations
for these 20 pollutants were greater than their associated risk screening value (or
failed screens).
•	Eleven pollutants contributed to 95 percent of failed screens for NBIL and therefore
were identified as pollutants of interest for this site. These 11 include two carbonyl
compounds, four VOCs, one PMio metal, and four PAHs.
•	NBIL failed the third highest number of screens (510) 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 20 percent. This is due
primarily to the relatively high number of pollutants sampled for at this site. 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.
•	Twelve pollutants failed screens for SPIL; approximately 71 percent of
concentrations for these 12 pollutants were greater than their associated risk screening
value (or failed screens).
12-20

-------
Eight pollutants contributed to 95 percent of failed screens for SPIL and therefore
were identified as pollutants of interest for this site. These eight include two carbonyl
compounds and six VOCs.
•	Eleven pollutants failed screens for ROIL; approximately 74 percent of
concentrations for these 11 pollutants were greater than their associated risk screening
value (or failed screens).
•	Eight pollutants contributed to 95 percent of failed screens for ROIL and therefore
were identified as pollutants of interest for this site. These eight include two carbonyl
compounds and six VOCs.
•	The Illinois monitoring sites have six pollutants of interest in common: two carbonyl
compounds (acetaldehyde and formaldehyde) and four VOCs (benzene,
1,3 -butadiene, carbon tetrachloride, and 1,2-dichloroethane). Of these, benzene,
carbon tetrachloride, and 1,2-dichloroethane failed 100 percent of screens for each
site.
12.4 Concentrations
This section presents various concentration averages used to characterize air toxics
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. Site-specific statistical summaries for all pollutants
sampled for at NBIL, SPIL, and ROIL are provided in Appendices J through O.
12.4.1 2013 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
12-21

-------
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 12-5, 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.
Table 12-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest
for the Illinois Monitoring Sites

# of






Measured
1st
2nd
3rd
4th


Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs. # of
Samples
Average
(Hg/m3)
Average
(Hg/m3)
Average
(Hg/m3)
Average
(Hg/m3)
Average
(ng/m3)
Northbrook, Illinois - NBIL


1.75
2.10
2.61
2.93
2.37
Acetaldehyde
62/62
±0.35
±0.49
±0.71
±0.76
±0.31


0.60
0.44
0.42
0.43
0.47
Benzene
61/61
±0.07
±0.08
±0.09
±0.10
±0.05


0.03
0.03
0.02
0.04
0.03
1.3 -Butadiene
38/61
±0.02
±0.01
±0.01
±0.02
±0.01


0.58
0.62
0.63
0.57
0.60
Carbon Tetrachloride
61/61
±0.04
±0.02
±0.02
±0.05
±0.02


0.07
0.08
0.04
0.06
0.06
1,2-Dichloroethane
53/61
±0.01
±0.01
±0.02
±0.01
±0.01


1.85
2.79
1.95
1.39
1.98
Formaldehyde
62/62
±0.25
±0.58
±0.62
±0.34
±0.26


9.41
49.98
39.81
1.91
25.12
Acenaphthene3
58/58
±7.75
± 18.65
± 17.66
±0.84
±8.19


0.34
0.77
0.81
0.57
0.62
Arsenic (PMi0)a
59/59
±0.09
±0.25
±0.27
±0.23
±0.11


1.59
12.12
14.67
1.44
7.47
Fluoranthene3
58/58
±0.68
±4.85
±6.62
±0.53
±2.52


6.08
38.83
30.39
2.10
19.24
Fluorcne"
57/58
±4.36
± 15.04
± 12.39
±0.87
±6.19


97.57
304.90
194.17
33.16
155.94
Naphthalene3
58/58
±68.12
± 119.66
±68.05
±6.41
± 44.27
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
12-22

-------
Table 12-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest
for the Illinois Monitoring Sites (Continued)

# of






Measured
1st
2nd
3rd
4th


Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs. # of
Samples
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Schiller Park, Illinois - SPIL


4.59
1.35
1.35
2.20
2.37
Acetaldehyde
61/61
± 1.69
±0.33
±0.22
±0.74
±0.55


0.95
0.74
0.67
0.61
0.74
Benzene
60/60
±0.26
±0.14
±0.09
±0.07
±0.08


0.16
0.11
0.11
0.13
0.13
1,3-Butadiene
59/60
±0.07
±0.03
±0.02
±0.02
±0.02


0.61
0.70
0.67
0.59
0.64
Carbon Tetrachloride
60/60
±0.03
±0.04
±0.03
±0.05
±0.02


0.08
0.10
0.07
0.07
0.08
1,2-Dichloroethane
57/60
±0.02
±0.01
±0.01
±0.02
±0.01


5.18
2.31
3.28
2.51
3.31
Formaldehyde
61/61
± 1.39
±0.55
±0.64
±0.52
±0.49


<0.01
0.01
0.01
0.03
0.01
Hexachloro-1,3 -butadiene
10/60
±0.01
±0.01
±0.01
±0.02
±0.01


0.32
0.30
0.28
0.14
0.26
T richloroethylene
44/60
±0.44
±0.20
±0.25
±0.12
±0.13
Roxana, Illinois - ROIL


1.14
2.02
2.60
1.61
1.84
Acetaldehyde
61/61
±0.17
±0.48
±0.51
±0.21
±0.22


1.05
0.86
1.05
0.90
0.97
Benzene
60/60
±0.17
±0.17
±0.27
±0.27
±0.11


0.05
0.05
0.06
0.07
0.06
1,3-Butadiene
54/60
±0.02
±0.02
±0.01
±0.02
±0.01


0.62
0.70
0.67
0.64
0.66
Carbon Tetrachloride
60/60
±0.05
±0.06
±0.03
±0.02
±0.02


0.07
0.10
0.04
0.08
0.07
1,2-Dichloroethane
45/60
±0.03
±0.03
±0.02
±0.02
±0.01


0.27
0.32
0.32
0.31
0.31
Ethylbenzene
60/60
±0.10
±0.09
±0.07
±0.11
±0.04


1.74
3.74
5.40
1.95
3.19
Formaldehyde
61/61
±0.24
± 1.30
± 1.27
±0.27
±0.57


<0.01
0.03
0.01
0.04
0.02
Hexachloro-1,3 -butadiene
15/60
±0.01
±0.02
±0.01
±0.03
±0.01
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
Observations for NBIL from Table 12-5 include the following:
•	The pollutants with the highest annual average concentrations are acetaldehyde
(2.37 ± 0.31 |ig/m3) and formaldehyde (1.98 ± 0.26 |ig/m3). The annual average
concentrations for the remaining pollutants of interest are less than 1 |ig/m3.
•	The third and fourth quarter average acetaldehyde concentrations are higher than the
first and second quarter averages and have relatively large confidence intervals
associated with them. A review of the data shows that acetaldehyde concentrations
measured at NBIL range from 0.86 |ig/m3 to 6.10 |ig/m3. The three highest
12-23

-------
acetaldehyde concentrations measured at NBIL were measured during the fourth
quarter of 2013 and that all but one of the nine acetaldehyde concentrations greater
than 4 |ig/m3 were measured between August and December.
While acetaldehyde concentrations were highest at NBIL during the fourth quarter,
formaldehyde concentrations were at their lowest. A review of the data shows that
formaldehyde concentrations measured at NBIL range from 0.44 |ig/m3 to
5.09 |ig/m3. The 10 concentrations greater than 3 |ig/m3 were all measured between
April and September, with seven of them measured between the end of May and mid-
July. Conversely, the 10 concentrations less than 1 |ig/m3 were all measured between
August and November.
Of the VOCs, carbon tetrachloride and benzene have the highest annual average
concentrations for NBIL. Quarterly average concentrations of carbon tetrachloride are
fairly consistent. This is also true for benzene with the exception of the first quarter as
the first quarter average is slightly higher than the other quarterly averages. Of the 23
benzene concentrations greater than 0.5 |ig/m3, 11 were measured during the first
quarter, with four each measured during the remaining calendar quarters.
Arsenic concentrations measured at NBIL range from 0.057 ng/m3 to 2.07 ng/m3.
Concentrations greater than 1 ng/m3 were not measured during the first quarter of
2013 while at least two were measured during each of the remaining calendar
quarters.
Of the PAHs, naphthalene has the highest annual average concentration.
Concentrations of each of the PAH pollutants of interest were significantly higher
during the warmer months of the year and exhibit a relatively large amount of
variability, based on the confidence intervals. Concentrations of naphthalene
measured at NBIL range from 2.87 ng/m3 to 748 ng/m3. The maximum concentration
measured at NBIL is the highest naphthalene concentration measured across the
program. This explains the large confidence interval associated with NBIL's second
quarter naphthalene concentration, when concentrations span an order of magnitude.
Five of the seven highest naphthalene concentrations measured across the program
(those greater than 400 ng/m3) were measured at NBIL and these were measured
between April and July.
Some of the highest concentrations of acenaphthene, fluorene, and fluoranthene
measured across the program were also measured at NBIL. Concentrations of
acenaphthene measured at NBIL range from 0.145 ng/m3 to 123 ng/m3, accounting
for eight of the nine highest acenaphthene measurements across the program and the
only two greater than 100 ng/m3. Concentrations of fluorene measured at NBIL range
from 0.357 ng/m3 to 99.1 ng/m3, accounting for 10 of the 12 highest fluorene
measurements across the program. Concentrations of fluoranthene range from
0.163 ng/m3 to 43.7 ng/m3, with the second, third, and fourth highest fluoranthene
concentrations across the program measured at NBIL.
Many of the higher PAH concentrations were measured on the same days. The
highest naphthalene and fluorene concentrations were measured at NBIL on May 16,
12-24

-------
2013, along with the second highest acenaphthene and third highest fluoranthene
concentrations. The highest acenaphthene concentration was measured atNBIL on
July 15, 2013, along with the second highest fluorene and fluoranthene concentrations
and the fourth highest naphthalene concentration.
Observations for SPIL from Table 12-5 include the following:
•	The pollutants with the highest annual average concentrations are formaldehyde
(3.31 ± 0.49 |ig/m3) and acetaldehyde (2.37 ± 0.55 |ig/m3). These are the only
pollutants with annual average concentrations greater than 1 |ig/m3. Of the VOCs,
benzene (0.74 ± 0.08 |ig/m3) and carbon tetrachloride (0.64 ± 0.02 |ig/m3) have the
highest annual average concentrations for SPIL.
•	Several of the pollutants of interest for SPIL were highest during the first quarter of
2013, in particular acetaldehyde and formaldehyde. Concentrations of acetaldehyde
measured at SPIL span three orders of magnitude, ranging from 0.0108 |ig/m3 to
14.2 |ig/m3. The maximum acetaldehyde concentration measured at SPIL is the
highest acetaldehyde concentration measured across the program. The second highest
acetaldehyde concentration measured at SPIL (7.17 |ig/m3) is roughly half as high but
is still one of the 10 highest acetaldehyde concentrations measured across the
program. Both of these measurements were from samples collected in February 2013.
Of the 16 acetaldehyde concentrations greater than 3 |ig/m3 measured at SPIL, 12
were measured in January, February, and March, with the other four in November and
December. Concentrations of formaldehyde measured at SPIL also span three orders
of magnitude, ranging from 0.0148 |ig/m3 to 10.5 |ig/m3. Five of the seven highest
formaldehyde concentrations measured at SPIL were measured in January 2013, with
one each in February and March.
•	Benzene and 1,3-butadiene concentrations also appear higher during the first quarter
of 2013, although the difference among the quarterly averages for these pollutants is
smaller. For benzene, four of the five benzene concentrations greater than 1 |ig/m3
were measured during the first quarter (with the fifth measured on the first sample
day of the second quarter). For 1,3-butadiene, the three highest concentrations were
measured in January and March.
•	The first quarter average concentration of trichloroethylene has a confidence interval
larger than the average itself. In addition, all of the quarterly average concentrations
shown in Table 12-5 have relatively large confidence intervals associated with them.
A review of the data shows that trichloroethylene was detected in roughly 73 percent
of the samples collected, with measured detections ranging from 0.0404 |ig/m3 to
3.21 |ig/m3. The five highest trichloroethylene concentrations measured across the
program were all measured at SPIL; further, 15 of the 18 highest trichloroethylene
concentrations (those greater than 0.30 |ig/m3) were measured at SPIL. SPIL is the
only NMP site for which trichloroethylene is a pollutant of interest. Similar
observations were also made in the 2011 and 2012 NMP reports.
12-25

-------
•	Hexachloro-1,3-butadiene was detected in less than 20 percent of the samples
collected at SPIL. Six of the 10 measured detections were measured during the fourth
quarter, with one each during the first and second quarters, and two in the third
quarter.
Observations for ROIL from Table 12-5 include the following:
•	The pollutants with the highest annual average concentrations are formaldehyde
(3.19 ± 0.57 |ig/m3) and acetaldehyde (1.84 ± 0.22 |ig/m3). These are the only
pollutants with annual average concentrations greater than 1 |ig/m3. Of the VOCs,
benzene (0.97 ± 0.11 |ig/m3) and carbon tetrachloride (0.66 ± 0.02 |ig/m3) have the
highest annual average concentrations for ROIL. ROIL's annual average
concentration of benzene is higher than the annual average concentrations for the
Chicago sites.
•	The second and third quarter average concentrations for formaldehyde are
significantly higher than the first and fourth quarter averages. A review of the data
shows that formaldehyde concentrations measured at ROIL range from 0.874 |ig/m3
to 10.7 |ig/m3, ROIL is one of only six NMP sites at which formaldehyde
concentrations greater than 10 |ig/m3 were measured; two were measured at ROIL
(and a third was measured at SPIL). The 20 highest formaldehyde measurements
collected at ROIL (those greater than 3 |ig/m3) were all measured between April and
September. Although the minimum formaldehyde concentration was measured in
May, most of the lower formaldehyde concentrations were measured during the first
and fourth quarters of 2013. Of the 25 formaldehyde measurements less than 2 |ig/m3,
12 were measured during the first quarter and 10 were measured during the fourth
quarter of 2013.
•	A similar observation can be made for acetaldehyde, in that the second and third
quarter average concentrations are higher than the remaining quarterly averages,
although the difference is considerably less. Of the 19 highest acetaldehyde
concentrations measured at ROIL (those greater than 2 |ig/m3), 17 were measured
between April and September. Conversely, of the eight concentrations of
acetaldehyde less than 1 |ig/m3, six were measured between January and March.
•	Ethylbenzene is the only pollutant of interest for ROIL that is not a pollutant of
interest for at least one of the Chicago sites. Ethylbenzene concentrations measured at
ROIL range from 0.091 |ig/m3 to 0.857 |ig/m3, with a median concentration of
0.27 |ig/m3; the quarterly average concentrations of ethylbenzene shown in
Table 12-5 are consistent from quarter to quarter. This is true for most of the VOCs
listed.
12-26

-------
Tables 4-9 through 4-12 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for NBIL, SPIL,
and ROIL from those tables include the following:
•	The Illinois monitoring sites appear in Tables 4-9 through 4-12 a total of 10 times,
with NBIL appearing five times, SPIL appearing three times, and ROIL appearing
twice.
•	Table 4-9 for VOCs shows that SPIL ranks ninth for its annual average concentration
of 1,3-butadiene while ROIL ranks seventh for its annual average concentration of
hexachloro-1,3-butadiene. NBIL does not appear in Table 4-9.
•	SPIL and NBIL both appear in Table 4-10, ranking sixth and seventh, respectively,
for their annual average concentrations of acetaldehyde, which were similar in
magnitude. SPIL and ROIL both appear in Table 4-10, ranking seventh and tenth,
respectively, for their annual average concentrations of formaldehyde.
•	NBIL ranks first for its annual average concentrations of acenaphthene and
naphthalene among NMP sites sampling PAHs, as shown in Table 4-11. In addition
to having the highest annual averages, NBIL's confidence intervals are also the
largest shown, a reflection of the variability within the measurements.
•	As shown in Table 4-12, NBIL's annual average concentration of arsenic ranks eighth
among NMP sites sampling PMio metals. Although not a pollutant of interest for
NBIL, this site's annual average concentration of nickel ranks ninth highest among
NMP sites sampling PMio metals.
12.4.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for the pollutants shaded in
gray in Table 12-4 for NBIL, SPIL, and ROIL. Figures 12-9 through 12-22 overlay the sites'
minimum, annual average, and maximum concentrations onto the program-level minimum, first
quartile, median, average, third quartile, and maximum concentrations, as described in
Section 3.4.3.1.
12-27

-------
Figure 12-9. Program vs. Site-Specific Average Acenaphthene Concentration



Program Max Concentration = 123 ng/m3



,





0	10	20	30	40	50	60	70	80
Concentration {ng/m3)
Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



Figure 12-10. Program vs. Site-Specific Average Acetaldehyde Concentrations
m±
¦+
0
3
6 9
Concentration {[jg/m3)

12

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i

Site: Site Average
o
Site Concentration Range


Figure 12-11. Program vs. Site-Specific Average Arsenic (PMio) Concentration

0
12 3
4 5 6
Concentration {ng/m3)
7
8

10

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i



Site: Site Average
o
Site Concentration Range




12-28

-------
Figure 12-12. Program vs. Site-Specific Average Benzene Concentrations
NBIL

1+
Program Max Concentration = 43.5 ^ig/m3


SPIL

U	


Program Max Concentration = 43.5 ^ig/m3
1 1 1



¦-
-o		
Program Max Concentration = 43.5 ^ig/m3



0	2	4	6	8	10	12
Concentration {[ig/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


Figure 12-13. Program vs. Site-Specific Average 1,3-Butadiene Concentrations


t
—
—¦ Program Max Concentration = 21.5 ^ig/m3




I 0

Program Max Concentration = 21.5 ^ig/m3
¦ 0





-
—
Program Max Concentration = 21.5 ^ig/m3


1	1	1	T
0	0.3	0.6	0.9	1.2	1.5
Concentration {[ig/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


12-29

-------
Figure 12-14. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations
Program Max Concentration = 23.7 ^ig/m3
Program Max Concentration = 23.7 ^ig/m3
Program Max Concentration = 23.7 ^ig/m3
Concentration {[jg/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


Figure 12-15. Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations


¦

Program Max Concentration = 111 ^ig/m3




—
—¦

Program Max Concentration = 111 jig/m3
Program Max Concentration = 111 ^ig/m3
0.2	0.4	0.6	0.8	1
Concentration (jig/m3)
I
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


12-30

-------
Figure 12-16. Program vs. Site-Specific Average Ethylbenzene Concentration
P—
Program Max Concentration = 18.7 ^ig/m3
0
1 2
3
Concentration {[jg/m3)
4
5

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i

Site: Site Average
o
Site Concentration Range


Figure 12-17. Program vs. Site-Specific Average Fluoranthene Concentration
25	30
Concentration {ng/m3]
Program: 1st Qua rti le
2nd Qua rti le 3rd Qua rti le
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



Figure 12-18. Program vs. Site-Specific Average Fluorene Concentration
40	50	60
Concentration {ng/m3)
Program: IstQuartile
2ndQuartile 3rdQuartile
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



12-31

-------
Figure 12-19. Program vs. Site-Specific Average Formaldehyde Concentrations
¦-+
9	12	15
Concentration (jig/m3)
Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



Figure 12-20. Program vs. Site-Specific Average Hexachloro-l,3-Butadiene Concentrations
0
0.05
0.1
0.15
Concentration {[ig/m3)
0.2
0.25
0.3

Program:
Site:
1st Qua rti le
¦
Site Average
o
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
Site Concentration Range
4th Qua rti le
~
Average
i

12-32

-------
Figure 12-21. Program vs. Site-Specific Average Naphthalene Concentration
-
0
100
200
300 400 500
Concentration {ng/m3)
600
700
800

Program:
Site:
1st Quartile
¦
Site Average
o
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
Site Concentration Range
4th Quartile
~
Average
i

Figure 12-22. Program vs. Site-Specific Average Trichloroethylene Concentration
0.5
l
1.5 2
Concentration {[jg/m3)
2.5
3
Program:
1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site:
Site Average
o
Site Concentration Range


Observations from Figures 12-9 through 12-22 include the following:
•	Figure 12-9 is the box plot for acenaphthene for NBIL. Note that the program-
level maximum concentration (123 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
80 ng/m3. The maximum acenaphthene concentration measured at NBIL is the
maximum concentration measured across the program. NBIL's annual average
acenaphthene concentration is more than five times the program-level average
concentration. More than half of NBIL's acenaphthene measurements are greater
than the program-level average concentration. Note that the program-level
average is greater than the program-level third quartile, an indication that the
measurements at the upper end of the concentration range are driving the
program-level average. Although non-detects were measured across the program,
none were measured at NBIL.
•	Figure 12-10 presents the acetaldehyde box plots for all three Illinois sites. The
box plots show that the maximum acetaldehyde concentration across the program
was measured at SPIL; a similar observation was made in the 2012 NMP report.
However, the minimum acetaldehyde concentration measured across the program
was also measured at SPIL. The annual average acetaldehyde concentration for
12-33

-------
SPIL is the same as the annual average acetaldehyde concentration for NBIL.
These annual averages are greater than the program-level average concentration
and just greater than the program-level third quartile. The range of acetaldehyde
concentrations measured at ROIL is smaller than the range of measurements from
the Chicago sites and ROIL's annual average is similar to the program-lev el
average concentration.
Figure 12-11 is the box plot for arsenic, which was measured at NBIL only. The
box plot shows the maximum concentration measured at NBIL is considerably
less than the maximum concentration measured across the program. The annual
average concentration for NBIL is just less than the program-level average
concentration. While a few non-detects of arsenic were measured among sites
sampling PMio metals, none were measured at NBIL.
Figure 12-12 presents the box plots for benzene for all three sites. Similar to the
box plot for acenaphthene, the program-level maximum benzene concentration
(43.5 |ig/m3) is not shown directly on the box plots as the scale has been reduced
to 12 |ig/m3 in Figure 12-12 to allow for the observation of data points at the
lower end of the concentration range. The range of concentrations measured at
SPIL and ROIL are similar to each other and twice the range of concentrations
measured at NBIL. NBIL's annual average benzene concentration is less than the
program-level median concentration and is the third lowest among NMP sites
sampling this pollutant. SPIL's annual average benzene concentration is similar to
the program-level average concentration while ROIL's annual average is greater
than the program-level average concentration and third quartile. Among the NMP
sites sampling benzene, ROIL's annual average ranks 13th.
Figure 12-13 presents the box plots for 1,3-butadiene for all three sites. Again, the
program-level maximum 1,3-butadiene concentration (21.5 |ig/m3) is not shown
directly on the box plots as the scale has been reduced to 1.5 |ig/m3 in
Figure 12-13 to allow for the observation of data points at the lower end of the
concentration range. The range of 1,3-butadiene concentrations is largest for SPIL
and smallest for ROIL. In fact, all of the 1,3-butadiene concentrations measured at
ROIL are less than the program-level average concentration. However, the
program-level average concentration 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. The annual average concentration of
1,3-butadiene is highest for SPIL and lowest for NBIL among the Illinois sites.
ROIL's annual average concentration is similar to the program-lev el median
concentration while NBIL's annual average is similar to the program-level first
quartile.
Figure 12-14 presents the box plots for carbon tetrachloride. The scale of these
box plots have also been reduced to allow for the observation of data points at the
lower end of the concentration range, as the program-level maximum carbon
tetrachloride concentration (23.7 |ig/m3) is considerably greater than the majority
of measurements. Figure 12-14 shows that maximum carbon tetrachloride
concentrations measured at the Illinois sites are considerably less than the
12-34

-------
program-level maximum concentration. The minimum concentrations measured at
NBIL and SPIL are roughly half the minimum concentration measured at ROIL.
The annual average carbon tetrachloride concentration for ROIL is similar to the
program-level average concentration. While the annual averages for NBIL and
SPIL are both less than the program-level average concentration, NBIL's annual
average concentration is similar to the program-level first quartile.
The scale of the box plot in Figure 12-15 has also been reduced to allow for the
observation of data points at the lower end of the concentration range, as the
program-level maximum 1,2-dichloroethane concentration (111 |ig/m3) is
considerably greater than the majority of measurements. Note that all of the
concentrations of 1,2-dichloroethane measured at the Illinois sites are less than the
program-level average concentration of 0.26 |ig/m3. The annual average
concentrations for the three Illinois sites are less than the program-level median
concentration, with NBIL's annual average concentration just less than the
program-level first quartile. This is another example of measurements at the upper
end of the concentration range driving the program-level average concentration,
as the program-level average is more than twice the program-level third quartile.
Figure 12-16 is the box plot for ethylbenzene for ROIL, the only Illinois site for
which this is a pollutant of interest. The scale of the box plot in Figure 12-16 has
also been reduced to allow for the observation of data points at the lower end of
the concentration range, as the program-level maximum ethylbenzene
concentration (18.7 |ig/m3) is considerably greater than the majority of
measurements. Figure 12-16 shows that all of the ethylbenzene concentrations
measured at ROIL are less than 1 |ig/m3. ROIL's annual average concentration of
ethylbenzene is just less than the program-level average concentration.
Figure 12-17 presents the box plot for fluoranthene for NBIL. This box plot
shows that the maximum concentration of fluoranthene across the program was
not measured at NBIL, although several of the highest fluoranthene
concentrations across the program were measured at NBIL, as discussed in the
previous section. The annual average concentration of fluoranthene for NBIL is
more than three times the program-level average concentration.
Figure 12-18 presents the box plot for fluorene for NBIL. This box plot shows
that the maximum concentration of fluorene across the program was measured at
NBIL, as discussed in the previous section. The annual average concentration of
fluorene for NBIL is more than four times the program-level average
concentration. NBIL is one of only two NMP sites with fluorene concentrations
greater than 35 ng/m3; of the 20 fluorene concentrations greater than 35 ng/m3,
concentrations measured at NBIL account for 14 of them.
Figure 12-19 presents the box plots for formaldehyde for all three sites. The
maximum formaldehyde concentration measured at ROIL is similar to the
maximum formaldehyde concentration measured at SPIL. The annual average
formaldehyde concentrations for these two sites are similar to each other and both
are greater than the program-level average concentration but less than the
12-35

-------
program-level third quartile. The range of formaldehyde concentrations measured
at NBIL is considerably smaller. NBIL's annual average formaldehyde
concentration is less than the program-level average concentration as well as the
program-level median.
•	Figure 12-20 presents the box plots for hexachloro-1,3-butadiene for SPIL and
ROIL. 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 (82 percent). Sixty valid VOC
samples were collected at SPIL and ROIL; 10 measured detections were
measured at SPIL and 15 at ROIL. Thus, many zeroes are substituted into the
annual average concentrations for this pollutant. The annual average for SPIL is
slightly less than the program-level average while the annual average
concentration for ROIL is slightly greater than the program-level average
concentrati on of hexachl oro-1,3 -butadi ene.
•	Figure 12-21 is the box plot for naphthalene for NBIL. The maximum
naphthalene concentration measured at NBIL (748 ng/m3) is the maximum
concentration measured across the program. The fourth lowest naphthalene
concentration across the program was also measured at NBIL. Thus, this site has
the largest range of naphthalene measurements across the program. The annual
average concentration for NBIL is more than twice the program-level average
concentration of naphthalene.
•	The first, second, and third quartiles for trichloroethylene are all zero in the box
plot for SPIL presented in Figure 12-22 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 (0.26 |ig/m3) is more than five times greater than the next
highest annual average concentration for this pollutant (calculated for S4MO,
0.05 |ig/m3) and an order of magnitude higher than the program-level average
concentration (0.02 |ig/m3).
12.4.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.3.2.
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 12-23 through 12-52 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 concentration is not provided, although the range
12-36

-------
and percentiles are still presented. Because sampling at ROIL began in 2012, a trends analysis
was not performed.
Figure 12-23. Yearly Statistical Metrics for Acenaphthene Concentrations Measured at NBIL
2010	2011
Year
0 5th Percentile	— Minimum
O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until June 2008.
Observations from Figure 12-23 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.
•	The three highest acenaphthene concentrations measured at NBIL were all measured
in 2013, and all but one of the six acenaphthene concentrations greater than 75 ng/m3
were measured in 2013, with the other measured in 2008 (93.5 ng/m3).
•	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 through 2012, after
which the median doubles for 2013.
•	The 1-year average concentration increases between 2009 and 2011, nearly doubling
over this time frame. However, confidence intervals calculated for these averages
12-37

-------
indicate that the increase is not statistically significant due to the relatively large
amount of variability in the measurements. The 1-year average decreased slightly for
2012, although the median continued to increase. For 2013, the 1-year average
concentration more than doubled, with similar increases for the median, 95th
percentile, and maximum concentration. Even if the two highest measurements from
2013 were excluded from the calculation, the increase in the 1-year average
concentrations from 2012 to 2013 would still represent a nearly 90 percent increase;
thus, concentrations were higher overall for 2013.
Figure 12-24. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at NBIL
20051	2006	2007	2008	2009	2010	2011	2012	2013
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until March 2005.
Observations from Figure 12-24 for acetaldehyde measurements collected at NBIL
include the following:
•	Carbonyl compound sampling at NBIL under the NMP began in March 2005;
because a full year's worth of data is not available for 2005, a 1-year average
concentration is not presented, although the range of measurements is provided.
•	The maximum acetaldehyde concentration measured at NBIL since the onset of
sampling (6.10 |ig/m3) was measured in 2013; the seven highest concentrations were
all measured at NBIL in 2013. The highest acetaldehyde concentrations were
measured in the most recent years; of the 25 acetaldehyde concentrations greater than
3 |ig/m3 measured at NBIL, one was measured in 2010, three in 2011, six in 2012,
and 15 were measured in 2013.
12-38

-------
•	Prior to 2010, the 1-year average concentrations were all less than 1 |ig/m3,
fluctuating between 0.69 |ig/m3 (2009) and 0.98 |ig/m3 (2006). After 2009,
acetaldehyde concentrations measured at NBIL increase significantly as all of the
statistical metrics exhibit an increase from 2009 to 2010 and again for 2011, 2012,
and 2013 (although the minimum concentration decreased for 2012). The 95th
percentile for 2013 is greater than the maximum concentrations measured for all
previous years of sampling. The 5th percentile for 2013 is greater than the 1-year
average concentrations for each of the earlier years of sampling.
•	The increase in the 1-year average concentration of acetaldehyde between 2009 and
2013 represents a 244 percent increase.
Figure 12-25. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at NBIL
2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
Observations from Figure 12-25 for arsenic (PMio) measurements collected at NBIL
include the following:
•	Metals sampling at NBIL began in January 2005.
•	The maximum arsenic concentration was measured at NBIL on July 12, 2009,
although a similar concentration was also measured in 2010. Only four concentrations
equal to or greater than 3 ng/m3 have been measured at NBIL (one in 2006, one in
2009, and two in 2010).
12-39

-------
•	Although the statistical parameters representing the upper end of the concentration
range have fluctuated somewhat each year, the 1-year average concentrations exhibit
relatively little significant change over the course of sampling. The 1-year average
concentration increased from 2005 to 2006, reached a maximum for 2007
(0.86 ng/m3), decreased slightly for 2008, after which the 1-year average
concentration remained steady through 2012. Between 2008 and 2012, the 1-year
average concentrations ranged from 0.73 ng/m3 (2012) to 0.75 ng/m3 (2010). Most of
the statistical parameters are at a minimum for 2013, with the 1-year average
concentration (0.62 ng/m3) at its lowest since the first year of sampling.
•	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 at NBIL.
Figure 12-26. Yearly Statistical Metrics for Benzene Concentrations Measured at NBIL
20031 20042 2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1	A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
2	A 1-year average is not presented because there was a gap in sampling from late October 2004 until late
December 2004.
Observations from Figure 12-26 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 2003, a
1-year average is not presented, although the range of measurements is provided. In
addition, sampling for VOCs was discontinued in October 2004 through the end of
the year. Thus, a 1-year average is not presented for 2004 either.
12-40

-------
•	The maximum benzene concentration (4.51 |ig/m3) was measured on January 9, 2011
and is the only benzene measurement greater than 4 |ig/m3 measured at NBIL. Three
additional benzene concentrations greater than 3 |ig/m3 were measured in 2004 and
2005 and most of the measurements greater than 2 |ig/m3 were measured in 2004.
•	The 1-year average concentration decreased significantly from 2005 to 2006, and
decreased slightly for 2007, then remained steady through 2009. All of the statistical
parameters exhibit increases from 2009 to 2010. Although the maximum
concentration nearly doubled from 2010 to 2011, the rest of the statistical parameters
decreased for 2011. This decreasing continued into 2012 (although the median
concentration actually increased slightly) and 2013.
•	With the exceptions of the minimum and 5th percentile, the statistical parameters are
each at a minimum for 2013; 2013 is the first year the 1-year average concentration is
less than 0.5 |ig/m3.
Figure 12-27. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at NBIL
u 0.50
Maximum
Concentration for
2011 is 2.68 n.g/m3
2003 1 20042 2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1	A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
2	A 1-year average is not presented because there was a gap in sampling from late October 2004 until late
December 2004.
12-41

-------
Observations from Figure 12-27 for 1,3-butadiene measurements collected atNBIL
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 concentrations of 1,3-butadiene measured at NBIL are less
than 0.35 |ig/m3.
•	For each year shown, the minimum and 5th percentile are zero, indicating the
presence of non-detects (at least 5 percent of the measurements). For the first 2 years
of sampling, the median concentration is also zero, indicating that at least half of the
measurements were non-detects. The 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), although the percentage of non-detects has been increasing slightly each year
since 2007, with 38 percent of measurements as non-detects for 2013.
•	The 1-year average concentration decreased slightly between 2005 and 2009,
although the changes are not significant. From 2009 to 2010, the 1-year average
doubled, and then nearly doubled again for 2011. However, there is a significant
amount of variability associated with these measurements, based on the confidence
intervals. Even with the relatively high concentrations measured in 2010 and 2011,
the 95th percentile changed only slightly, indicating that the majority of the
measurements were within the same range. If the three outlier concentrations
measured in 2010 and 2011 were excluded from the calculations, the 1-year average
concentrations would still exhibit increasing trend between 2009 and 2012, but they
would be less dramatic.
•	The range within which the majority of concentrations fall, as determined by the 5th
and 95th percentiles, is at a minimum for 2013, as is the 1-year average
concentration.
12-42

-------
Figure 12-28. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at NBIL
2.5
Maximum
Concentration for
2004 is 4.81 ng/m3
2003 x	2004 " 2005	2006	2007	2008	2009	2010	2011	2012	2013
0 5th Percentile	— Minimum
- Maximum	O 95th Percentile	Average
1	A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
2	A 1-year average is not presented because there was a gap in sampling from late October 2004 until late
December 2004.
Observations from Figure 12-28 for carbon tetrachloride measurements collected at NBIL
include the following:
•	The maximum concentration of carbon tetrachloride was measured in 2004
(4.81 |ig/m3). Only one additional measurement greater than 1.5 |ig/m3 has been
measured (1.88 |ig/m3 in 2012).
•	Five non-detects of carbon tetrachloride have been measured at NBIL. All of these
were measured during the first 2 years of sampling (two in 2003 and three in 2004).
•	After a slight decreasing trend between 2005 and 2007, the 1-year average
concentration increased significantly for 2008. The 1-year average concentration
exhibits a decreasing trend after 2008 that continued through 2011. After exhibiting
an increase for 2012, the 1-year average concentration is at a minimum for 2013
(0.60 |ig/m3). The 1-year average concentrations presented range from 0.60 |ig/m3
(2013) to 0.83 |ig/m3 (2008), with most of the 1-year averages falling between
0.65 |ig/m3 and 0.75 |ig/m3. The median concentration exhibits a similar pattern.
•	The difference between the minimum and maximum concentrations is at a minimum
for 2013 as is the difference between the 5th and 95th percentiles. The differences in
these parameters has generally been decreasing over the last few years of sampling,
12-43

-------
indicating that the level of variability within the carbon tetrachloride measurements is
decreasing.
Figure 12-29. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations
Measured at NBIL
0.20










pL

o




1
o
—
0.04 -
o.oo -






o


7?

— , — , 2
~"	"\
L-i	J
			
o




20031 2004 2 2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	~••• Average
1	A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
2	A 1-year average is not presented because there was a gap in sampling from late October 2004 until late
December 2004.
Observations from Figure 12-29 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 between 2003 and 2008. The median concentration is zero
through 2011, indicating that at least half of the measurements are non-detects.
•	The number of non-detects began to decrease starting with 2009 and continued
through 2012. The 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.
•	The median concentration is greater than zero for the first time for 2012 and is also
greater than the 1-year average concentration. This is because the eight non-detects
(or zeros) factored into the 1-year average concentration are pulling the average down
12-44

-------
(in the same manner that a maximum or outlier concentration can drive the average
up) and are not contributing to the majority of measurements. This is also true for
2013.
• Each of the statistical parameters except the minimum and 5th percentile exhibit
slight decreases for 2013.
Figure 12-30. Yearly Statistical Metrics for Fluoranthene Concentrations Measured at NBIL
20081	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until June 2008.
Observations from Figure 12-30 for fluoranthene measurements collected at NBIL
include the following:
• The maximum fluoranthene concentration was measured at NBIL on July 21, 2013
(43.7 ng/m3), although two similar concentrations were also measured in 2012 and
2013. All but one of the seven fluoranthene concentrations greater than 30 ng/m3 have
been measured since 2011.
• 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, despite the higher maximum concentration measured in
2009. The number of measurements less than 2 ng/m3 tripled from 2008 to 2009,
accounting for 25 percent of measurements in 2008 compared to 45 percent for 2009.
Recall, however, that 2008 does not include a full year's worth of sampling. The
median fluoranthene concentrations shown after 2009 vary little.
12-45

-------
• Like acenaphthene, the 1-year average concentration of fluoranthene increases
between 2009 and 2011, decreases slightly for 2012, then increases slightly for 2013.
However, confidence intervals calculated for these averages indicate that the changes
are not statistically significant due to the relatively large amount of variability in the
measurements.
Figure 12-31. Yearly Statistical Metrics for Fluorene Concentrations Measured at NBIL
120 -|	
100
80
60
40
20
0
20081	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until June 2008.
Observations from Figure 12-31 for fluorene measurements collected at NBIL include the
following:
•	The statistical patterns for fluorene resemble the statistical patterns shown on the
trends graph for 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.
•	Like acenaphthene and fluoranthene, the 1-year average concentration of fluorene
increases between 2009 and 2011, decreases slightly for 2012, then increases for
2013. Confidence intervals calculated for these averages indicate that the changes are
not statistically significant due to the relatively large amount of variability in the
measurements. The range of fluorene measurements spans two orders of magnitude
12-46

-------
for each year. For example, the minimum and maximum concentrations for 2012 are
0.93 ng/m3 and 93.4 ng/m3, respectively.
• Since 2009, the maximum concentration of fluorene measured at NBIL has nearly
doubled. The 95th percentile also has an increasing trend between 2009 and 2013,
although a decrease is shown for 2012.
Figure 12-32. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at NBIL

Maximum
Concentration for
2006 is 91.7 ng/m3















—6—i
— g
3 :			;	li

i—o ; 
-------
fold increase in the 95th percentile. Although difficult to discern in Figure 12-32, 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
range increased. The number of measurements between 2 |ig/m3 and 4 |ig/m3 nearly
doubled from 2011 to 2012.
•	The range of formaldehyde concentrations measured at NBIL in 2013 is at its
smallest in four years. The difference between the 1-year average and median
concentrations is at a minimum for 2013, indicating less variability in the
measurements than the preceding years.
Figure 12-33. Yearly Statistical Metrics for Naphthalene Concentrations Measured at NBIL
20081	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until June 2008.
Observations from Figure 12-33 for naphthalene measurements collected at NBIL include
the following:
• The maximum naphthalene concentration was measured on September 23, 2010
(869 ng/m3). The second highest concentration measured was measured on October 6,
2011 (779 ng/m3). The next six highest concentrations of naphthalene were all
12-48

-------
measured at NBIL in 2013, making 2013 the only year in which multiple naphthalene
measurements greater than 400 ng/m3 were measured.
•	The central tendency parameters for naphthalene exhibit a similar pattern of changes
as those shown on the trends graphs for the other PAH pollutants of interest for
NBIL.
•	With the exception of the minimum concentration, the statistical parameters exhibit
increases for 2013. The 1-year average and maximum concentrations doubled from
2012 to 2013 while the 95th percentile increased by more than 200 ng/m3. These
changes from 2012 to 2013 are almost the opposite of those shown from 2011 to
2012, where many of the parameters exhibited substantial decreases.
Figure 12-34. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SPIL
20051	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because consistent sampling did not begin until March 2005.
Observations from Figure 12-34 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.
12-49

-------
•	The maximum acetaldehyde concentration was measured at SPIL on
November 17, 2012 (20.4 |ig/m3). Twenty-three of the 25 concentrations of
acetaldehyde greater than 5 |ig/m3 were measured in 2011 (eight), 2012 (eight), or
2013 (seven), with the other two measured in 2006.
•	The 1-year average concentration decreased significantly from 2006 to 2007, as did
most of the other statistical parameters. Between 2007 and 2009, the 1-year average
concentration changed little, hovering between 1.25 |ig/m3 and 1.50 |ig/m3. The
1-year average concentration increased in 2010 then increased significantly in 2011.
All of the statistical metrics increased for 2011, particularly the maximum and 95th
percentile, indicating that the increases shown are not attributable to a few of outliers.
As an illustration, the number of measurements greater than 2 |ig/m3 increased from
three in 2009 to 15 for 2010 to 40 in 2011.
•	The profile of acetaldehyde concentrations measured at SPIL in 2012 and 2013 is
more similar to 2011 than other years of sampling, although the measurements exhibit
a slight decrease in the magnitude of most of the measurements, as indicated by the
slight decrease in the 1-year average concentrations and difference in the 5th and 95th
percentiles. Yet these measurements still reflect considerable variability, based on the
range of concentrations measured.
Figure 12-35. Yearly Statistical Metrics for Benzene Concentrations Measured at SPIL
20031	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
12-50

-------
Observations from Figure 12-35 for benzene measurements collected at SPIL include the
following:
•	Sampling for VOCs at SPIL under the NMP began in April 2003; because a full
year's worth of data is not available for 2003, a 1-year average is not presented,
although the range of measurements is provided.
•	The only two concentrations of benzene greater than 5 |ig/m3 were both measured in
2005.
•	The 1-year average benzene concentration has a significant decreasing trend over the
years between 2004 and 2009. During the last 5 years of sampling, the 1-year average
benzene concentration has an undulating pattern, fluctuating between 0.68 |ig/m3
(2009) and 0.95 |ig/m3 (2012). The median concentration has a similar pattern.
•	The majority of benzene concentrations measured at SPIL, as indicated by the 5th and
95th percentiles, fell within roughly the same range between 2010 and 2012. The
range of benzene concentrations measured in 2013 is slightly smaller than other
recent years.
Figure 12-36. Yearly Statistical Metrics for 1,3-Butadiene Concentrations
Measured at SPIL











-I




T I



¦¦



*	


n h I










UJ
Qm
l—o—1 Q—'
T

h5

J U

2003 1 2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile •*•*¦¦/••• Average
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
12-51

-------
Observations from Figure 12-36 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 seven concentrations greater than 0.5 |ig/m3 have been measured at SPIL,
one in 2004, two in 2005, two in 2011, and one each in 2012 and 2013.
•	The detection rate for 1,3-butadiene has increased over time, ranging from
approximately 45 percent non-detects in 2003 and 2004 to zero in 2008 and 2009,
with one non-detect each measured in each of the following years.
•	The 1-year average concentrations of 1,3-butadiene changed little between 2004 and
2006, then decreased between 2006 and 2009. The increase in the 1-year average
concentration from 2009 to 2010 represents a 67 percent increase and a return to 2006
levels. Although a slight decreasing trend is shown after 2011, there is more
variability in the 1,3-butadiene concentrations measured during the last few years of
sampling. Despite these changes, most of the 1-year average concentrations shown
fall between 0.10 |ig/m3 and 0.15 |ig/m3, with only the minimum (0.08 |ig/m3 for
2009) and maximum (0.16 |ig/m3 for 2011) falling outside this range.
•	The 5th and 95th percentiles indicate the range within which the majority of
concentrations fall. This range decreased considerably between 2004 and 2009,
increased for 2010 and 2011, then began to decrease again. The difference between
these two parameters is at a minimum for 2013.
12-52

-------
Figure 12-37. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at SPIL
1.2 -


I


T

4-
¦p.''.-"
T


0.8 -

—o-






rl


r





hr






¦


T.l?




r
•o

0.4 -
0.2 -









—i
p—

r



-







1

20031 2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile •*•*¦¦/••• Average
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
Observations from Figure 12-37 for carbon tetrachloride measurements collected at SPIL
include the following:
•	The maximum concentration of carbon tetrachloride was measured three times, once
in 2005 and twice in 2008 (1.20 |ig/m3).
•	Six non-detects of carbon tetrachloride have been measured at SPIL. All of these
were measured during the first 2 years of sampling (four in 2003 and two in 2004).
•	The 1-year average concentration changed very little between 2004 and 2007, varying
between 0.65 |ig/m3 and 0.70 |ig/m3. The 1-year average then increased significantly
for 2008 (0.84 |ig/m3). The 1-year average concentration exhibits a decreasing trend
after 2008 that continued through 2011, when the 1-year average is at a minimum
(0.58 |ig/m3). The increase shown for 2012 brings the 1-year average carbon
tetrachloride concentration near 2010 levels. A similar change was exhibited by the
carbon tetrachloride concentrations measured at NBIL for 2012.
•	With the exception of the 5th percentile, all the of the statistical parameters exhibit a
decrease for 2013. Although the 5th percentile increased considerably in 2013, this is
an indication that the variability in the measurements is decreasing as the difference
between the 5th and 95th percentiles is at a minimum for 2013.
12-53

-------
Figure 12-38. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations
Measured at SPIL

Maximum
Concentration for
2003 is 0.75 ^g/m3













-c
y-

-c
}—
—









yr

7F




<0







i——	'¦
n""'

.0



2003	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	— Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
Observations from Figure 12-38 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 is also zero for
2010 and 2011, indicating that at least half the measurements are non-detects. The
percentage of non-detects decreased to 80 percent for 2010 and 73 percent for 2011.
For 2012, the percentage of non-detects decreased to 8 percent of samples collected
and was at a minimum of 5 percent for 2013, which is the first year that the 5th
percentile is greater than zero.
•	The 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. No other 1,2-dichloroethane concentrations greater than 0.15 |ig/m3
have been measured at SPIL.
•	As the number of non-detects decreases and the number of measured detections
increases, 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 a
12-54

-------
sharp increase in the 1-year average concentration shown for 2012. A similar range of
1,2-dichloroethane measurements was collected at SPIL in 2013.
Figure 12-39. Yearly Statistical Metrics for Formaldehyde Concentrations
Measured at SPIL


Maximum
Concentration for
2006 is 162 n.g/m3











—o-









""""'¦'irs	i		sL	
20051	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because consistent sampling did not begin until March 2005.
Observations from Figure 12-39 for formaldehyde measurements collected at SPIL
include the following:
•	The maximum formaldehyde concentration (162 |ig/m3) was measured at SPIL on
May 29, 2006 and is more than 10 times the maximum concentration for any of the
other years shown in Figure 12-39 other than 2005. Of the 29 formaldehyde
concentrations greater than 15 |ig/m3, 12 were measured at SPIL in 2005, 17 were
measured in 2006, and none were measured in the years that followed.
•	The 1-year average concentration for 2006 is 13.76 |ig/m3. After 2006, the 1-year
average concentration decreased each year, reaching a minimum of 1.85 |ig/m3 for
2009. Although difficult to discern in Figure 12-39, there is an increasing trend in the
1-year average concentration between 2009 and 2011, after which little change is
shown.
•	The majority of formaldehyde concentrations measured at SPIL fell within roughly
the same range between 2011 and 2013, as indicated by the 5th and 95th percentiles.
12-55

-------
Figure 12-40. Yearly Statistical Metrics for Hexachloro-l,3-Butadiene Concentrations
Measured at SPIL
.40
Maximum
Concentration for
2011 is 0.68 ng/m-
.35
.30
.25
.20
.15
.10
.05
"V"'
2009
2008
2004
.00
2003
2005
2006
2007
2010
2011
2012
2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	0 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
Observations from Figure 12-40 for hexachloro-l,3-butadiene measurements collected at
SPIL include the following:
•	The trends graph for hexachloro-1,3-butadiene measurements resembles the trends
graph for 1,2-dichloroethane in that the statistical parameters reflect that non-detects
make up the majority of measurements of this pollutant.
•	There were no measured detections of hexachloro-1,3-butadiene measured at SPIL
during the first 2 years of sampling. Non-detects made up 83 percent of
measurements in 2005 and 93 percent in 2006. Between 2007 and 2010, the
percentage of non-detects was constant at 98 percent. After 2010, the percentage of
non-detects began to fall slightly each year, returning to 83 percent by 2013.
•	The maximum hexachloro-l,3-butadiene concentration measured at SPIL was
measured on December 11, 2011 (0.68 |ig/m3) and is the only measurement of this
pollutant greater than 0.25 |ig/m3. Only 39 total measured detections have been
measured at SPIL since the onset of sampling. The effect of the non-detects (zeros)
factored into the statistical calculations can be seen in the scale of the trends graph
and by noting that none of the 1-year average concentrations shown are greater than
0.025 |ig/m3.
12-56

-------
Figure 12-41. Yearly Statistical Metrics for Trichloroethylene Concentrations
Measured at SPIL
Maximum
Concentration for
2003 is 110 n.g/m3
20031	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	0 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
Observations from Figure 12-41 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 concentration
(17.5 |ig/m3), which was measured in 2012. No other trichloroethylene concentrations
greater than 10 |ig/m3 have been measured at SPIL.
•	The concentrations of trichloroethylene exhibit considerable variability, as indicated
by confidence intervals calculated for the 1-year average concentrations, particularly
for 2012, when the maximum concentration was nearly four times the next highest
concentration measured that year and non-detects made up more than a quarter of the
measurements.
•	The 1-year average concentrations have fluctuated between 0.26 |ig/m3 (2013) to
0.79 |ig/m3 (2010), with no distinct trend in the concentrations. It should be noted
however, that the difference between the 5th and 95th percentiles is at a minimum for
12-57

-------
2013, indicating that most of the trichloroethylene measurements collected in 2013 at
SPIL fell within a tighter range.
12.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at each Illinois monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4 for
definitions and explanations regarding the various toxicity factors, time frames, and calculations
associated with these risk-based screenings.
12.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Illinois sites and where annual average
concentrations could be calculated, risk was examined by calculating cancer risk and noncancer
hazard approximations. These approximations can be used as risk estimates for cancer and
noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers want to shift their air
monitoring priorities. Refer to Section 3.4.3.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 12-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
12-58

-------
Table 12-6. Risk Approximations for the Illinois Monitoring Sites



# of






Measured


Noncancer

Cancer
Noncancer
Detections
Annual
Cancer Risk
Hazard
Pollutant
URE
frig/m3)1
RfC
(mg/m3)
vs. # of
Samples
Average
frig/m3)
Approximation
(in-a-million)
Approximation
(HQ)
Northbrook, Illinois - NBIL




2.37


Acetaldehyde
0.0000022
0.009
62/62
±0.31
5.20
0.26




0.47


Benzene
0.0000078
0.03
61/61
±0.05
3.68
0.02




0.03


1,3-Butadiene
0.00003
0.002
38/61
±0.01
0.88
0.01




0.60


Carbon Tetrachloride
0.000006
0.1
61/61
±0.02
3.59
0.01




0.06


1,2-Dichloroethane
0.000026
2.4
53/61
±0.01
1.62
<0.01




1.98


Formaldehyde
0.000013
0.0098
62/62
±0.26
25.68
0.20




25.12


Acenaphthene3
0.000088
--
58/58
±8.19
2.21
--




0.62


Arsenic (PMio)3
0.0043
0.000015
59/59
±0.11
2.66
0.04




7.47


Fluoranthene3
0.000088
--
58/58
±2.52
0.66
--




19.24


Fluorene3
0.000088
--
57/58
±6.19
1.69
--




155.94


Naphthalene3
0.000034
0.003
58/58
± 44.27
5.30
0.05
Schiller Park, Illinois - SPIL




2.37


Acetaldehyde
0.0000022
0.009
61/61
±0.55
5.21
0.26




0.74


Benzene
0.0000078
0.03
60/60
±0.08
5.78
0.02




0.13


1,3-Butadiene
0.00003
0.002
59/60
±0.02
3.80
0.06




0.64


Carbon Tetrachloride
0.000006
0.1
60/60
±0.02
3.85
0.01




0.08


1,2-Dichloroethane
0.000026
2.4
57/60
±0.01
2.13
<0.01




3.31


Formaldehyde
0.000013
0.0098
61/61
±0.49
43.00
0.34




0.01


Hexachloro-1,3 -butadiene
0.000022
0.09
10/60
±0.01
0.24
<0.01




0.26


T richloroethylene
0.0000048
0.002
44/60
±0.13
1.26
0.13
- = a Cancer URE or Noncancer RfC is not available.
3 Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of
viewing.
12-59

-------
Table 12-6. Risk Approximations for the Illinois Monitoring Sites (Continued)
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Roxana, Illinois - ROIL
Acetaldehyde
0.0000022
0.009
61/61
1.84
±0.22
4.05
0.20
Benzene
0.0000078
0.03
60/60
0.97
±0.11
7.54
0.03
1,3-Butadiene
0.00003
0.002
54/60
0.06
±0.01
1.71
0.03
Carbon Tetrachloride
0.000006
0.1
60/60
0.66
±0.02
3.94
0.01
1,2-Dichloroethane
0.000026
2.4
45/60
0.07
±0.01
1.93
<0.01
Ethylbenzene
0.0000025
1
60/60
0.31
±0.04
0.76
<0.01
Formaldehyde
0.000013
0.0098
61/61
3.19
±0.57
41.43
0.33
Hexachloro-1,3 -butadiene
0.000022
0.09
15/60
0.02
±0.01
0.45
<0.01
- = a Cancer URE or Noncancer RfC is not available.
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of
viewing.
Observations for the Illinois sites from Table 12-6 include the following:
•	Formaldehyde and acetaldehyde are the pollutants with the highest annual average
concentrations for all three sites.
•	Formaldehyde has the highest cancer risk approximation for all three sites, ranging
from 25.68 in-a-million for NBIL to 43.00 in-a-million for SPIL. 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, SPIL, or ROIL have noncancer hazard
approximations greater than 1.0, indicating that no adverse noncancer health effects
are expected from these individual pollutants. The pollutant with the highest
noncancer hazard approximation among the pollutants of interest for the Illinois sites
is formaldehyde (0.34 for SPIL).
12-60

-------
12.5.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 12-7 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
that have cancer toxicity factors. Table 12-7 also presents the 10 pollutants with the highest
toxicity-weighted emissions, based on the weighting schema described in Section 3.4.3.4. Lastly,
Table 12-7 provides the pollutants with the highest cancer risk approximations (in-a-million) for
each Illinois site, as presented in Table 12-6. The emissions, toxicity-weighted emissions, and
cancer risk approximations are shown in descending order in Table 12-7. Table 12-8 presents
similar information, but is limited to those pollutants with noncancer toxicity factors.
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 12.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
12-61

-------
Table 12-7. 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)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Northbrook, Illinois (Cook County) - NBIL
Benzene
1,391.32
Formaldehyde
1.48E-02
Formaldehyde
25.68
Formaldehyde
1,135.39
Benzene
1.09E-02
Naphthalene
5.30
Ethylbenzene
756.81
1,3-Butadiene
6.47E-03
Acetaldehyde
5.20
Acetaldehyde
623.34
Hexavalent Chromium
4.02E-03
Benzene
3.68
1.3 -Butadiene
215.66
Naphthalene
3.60E-03
Carbon Tetrachloride
3.59
T etrachloroethylene
187.87
Arsenic, PM
2.64E-03
Arsenic
2.66
Naphthalene
105.84
Ethylbenzene
1.89E-03
Acenaphthene
2.21
T richloroethylene
99.56
POM, Group 2b
1.81E-03
Fluorene
1.69
Dichloromethane
35.41
Acetaldehyde
1.37E-03
1,2-Dichloroethane
1.62
POM, Group 2b
20.53
POM, Group 2d
1.19E-03
1,3-Butadiene
0.88
Schiller Park, Illinois (Cook County) - SPIL
Benzene
1,391.32
Formaldehyde
1.48E-02
Formaldehyde
43.00
Formaldehyde
1,135.39
Benzene
1.09E-02
Benzene
5.78
Ethylbenzene
756.81
1,3-Butadiene
6.47E-03
Acetaldehyde
5.21
Acetaldehyde
623.34
Hexavalent Chromium
4.02E-03
Carbon Tetrachloride
3.85
1.3 -Butadiene
215.66
Naphthalene
3.60E-03
1,3-Butadiene
3.80
T etrachloroethylene
187.87
Arsenic, PM
2.64E-03
1,2-Dichloroethane
2.13
Naphthalene
105.84
Ethylbenzene
1.89E-03
T richloroethylene
1.26
T richloroethylene
99.56
POM, Group 2b
1.81E-03
Hexachloro-1,3 -butadiene
0.24
Dichloromethane
35.41
Acetaldehyde
1.37E-03

POM, Group 2b
20.53
POM, Group 2d
1.19E-03

-------
Table 12-7. 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)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Roxana, Illinois (Madison County) - ROIL
Formaldehyde
117.39
Coke Oven Emissions, PM
1.58E-02
Formaldehyde
41.43
Benzene
116.81
Formaldehyde
1.53E-03
Benzene
7.54
Ethylbenzene
56.77
Hexavalent Chromium
1.29E-03
Acetaldehyde
4.05
Acetaldehyde
50.30
Arsenic, PM
1.03E-03
Carbon Tetrachloride
3.94
Coke Oven Emissions, PM
15.95
Benzene
9.11E-04
1,2-Dichloroethane
1.93
Naphthalene
14.00
Naphthalene
4.76E-04
1,3-Butadiene
1.71
1.3 -Butadiene
12.69
1,3-Butadiene
3.81E-04
Ethylbenzene
0.76
Dichloromethane
12.11
Nickel, PM
3.20E-04
Hexachloro-1,3 -butadiene
0.45
T etrachloroethylene
3.60
POM, Group 5a
2.42E-04

POM, Group 2b
1.85
POM, Group 2b
1.63E-04

-------
Table 12-8. 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


Top 10 Noncancer Hazard Approximations
with Noncancer RfCs
Top 10 Noncancer Toxicity-Weighted Emissions
Based on Annual Average Concentrations
(County-Level)
(County-Level)

(Site-Specific)





Noncancer



Noncancer

Hazard
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(HQ)
Northbrook, Illinois (Cook County) - NBIL
Toluene
10,461.60
Acrolein
4,043,182.36
Acetaldehyde
0.26
Xylenes
3,369.01
Formaldehyde
115,856.06
Formaldehyde
0.20
Methanol
3,041.83
1,3-Butadiene
107,829.46
Naphthalene
0.05
Hexane
2,784.74
Cyanide Compounds, gas
86,974.16
Arsenic
0.04
Benzene
1,391.32
Acetaldehyde
69,259.50
Benzene
0.02
Formaldehyde
1,135.39
T richloroethy lene
49,780.32
1,3-Butadiene
0.01
Ethylene glycol
1,052.17
Benzene
46,377.32
Carbon Tetrachloride
0.01
Ethylbenzene
756.81
Arsenic, PM
40,902.71
1,2-Dichloroethane
<0.01
Acetaldehyde
623.34
Naphthalene
35,279.80


Methyl isobutyl ketone
342.65
Xylenes
33,690.12


Schiller Park, Illinois (Cook County) - SPIL
Toluene
10,461.60
Acrolein
4,043,182.36
Formaldehyde
0.34
Xylenes
3,369.01
Formaldehyde
115,856.06
Acetaldehyde
0.26
Methanol
3,041.83
1,3-Butadiene
107,829.46
T richloroethy lene
0.13
Hexane
2,784.74
Cyanide Compounds, gas
86,974.16
1,3-Butadiene
0.06
Benzene
1,391.32
Acetaldehyde
69,259.50
Benzene
0.02
Formaldehyde
1,135.39
T richloroethy lene
49,780.32
Carbon Tetrachloride
0.01
Ethylene glycol
1,052.17
Benzene
46,377.32
Hexachloro-1,3 -butadiene
<0.01
Ethylbenzene
756.81
Arsenic, PM
40,902.71
1,2-Dichloroethane
<0.01
Acetaldehyde
623.34
Naphthalene
35,279.80


Methyl isobutyl ketone
342.65
Xylenes
33,690.12



-------
Table 12-8. 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


Top 10 Noncancer Hazard Approximations
with Noncancer RfCs
Top 10 Noncancer Toxicity-Weighted Emissions
Based on Annual Average Concentrations
(County-Level)
(County-Level)

(Site-Specific)





Noncancer



Noncancer

Hazard
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(HQ)
Roxana, Illinois (Madison County) - ROIL
Toluene
635.77
Acrolein
274,415.42
Formaldehyde
0.33
Xylenes
208.66
Chlorine
95,420.68
Acetaldehyde
0.20
Hexane
195.81
Hexamethylene-l,6-diisocyanate, gas
25,000.00
Benzene
0.03
Methanol
178.11
Manganese, PM
16,632.19
1,3-Butadiene
0.03
Hydrochloric acid
128.20
Arsenic, PM
16,022.05
Carbon Tetrachloride
0.01
Formaldehyde
117.39
Lead, PM
14,477.27
Ethylbenzene
<0.01
Benzene
116.81
Formaldehyde
11,978.64
Hexachloro-1,3 -butadiene
<0.01
Ethylbenzene
56.77
Cyanide Compounds, gas
7,490.07
1,2-Dichloroethane
<0.01
Ethylene glycol
53.93
Nickel, PM
7,414.81


Acetaldehyde
50.30
Hydrochloric acid
6,410.24



-------
Observations from Table 12-7 include the following:
•	Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
cancer UREs in Cook County. These same pollutants are the highest emitted
pollutants with cancer UREs in Madison County, although the order differs. The
quantity of emissions is considerably different between the two counties, with the
emissions for Cook County an order of magnitude greater than Madison County.
•	The pollutants with the highest toxicity-weighted emissions (of the pollutants with
cancer UREs) for Cook County are formaldehyde, benzene, and 1,3-butadiene. Coke
oven emissions top Madison County's toxicity-weighted emissions, followed by
formaldehyde and hexavalent chromium.
•	Seven of the highest emitted pollutants in Cook County also have the highest toxicity-
weighted emissions while six of the highest emitted pollutants in Madison County
also have the highest toxicity-weighted emissions.
•	For NBIL and SPIL, formaldehyde is the pollutant with the highest cancer risk
approximation. This pollutant also has the highest toxicity-weighted emissions and
ranks second for quantity emitted. Benzene, acetaldehyde, and 1,3-butadiene also
appear on all three lists for both sites. For ROIL, formaldehyde is also the pollutant
with the highest cancer risk approximation. This pollutant also has the highest
emissions in Madison County and the second highest toxicity-weighted emissions.
Benzene and 1,3-butadiene also appear on all three lists for ROIL.
•	Carbon tetrachloride, which has the fifth highest cancer risk approximation for NBIL
and fourth highest cancer risk approximation for SPIL and ROIL, does not appear on
either county's emissions-based list. Similarly, 1,2-dichloroethane appears on neither
emissions-based list though it ranks among the pollutants with the highest cancer risk
approximations for all three sites.
•	Naphthalene has the second highest cancer risk approximation for NBIL. This
pollutant also has the fifth highest toxicity-weighted emissions for Cook County and
ranks seventh for quantity emitted. POM, Group 2b ranks 10th for quantity emitted
and eighth for toxicity-weighted emissions in Cook County. POM, Group 2b includes
acenaphthene, fluorene, and fluoranthene, all three of which are pollutants of interest
for NBIL.
•	Trichloroethylene has the seventh highest cancer risk approximation for SPIL and is
the eighth highest emitted pollutant in Cook County, but does not appear among the
pollutants with the highest toxicity-weighted emissions (this pollutant ranks 13th).
•	Arsenic has the sixth highest cancer risk approximation for NBIL (SPIL did not
sample metals). Arsenic ranks sixth for Cook County for its toxicity-weighted
emissions, but does not appear among the highest emitted pollutants (this pollutant
ranks 17th).
12-66

-------
•	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.
Observations from Table 12-8 include the following:
•	Toluene and xylenes are the highest emitted pollutants with noncancer RfCs in both
Cook and Madison Counties, although the quantity emitted is significantly higher in
Cook County.
•	The pollutant with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) for both counties is acrolein. Although acrolein was sampled for at
all three sites, this pollutant was excluded from the pollutants of interest designation,
and thus subsequent risk-based screening evaluations, due to questions about the
consistency and reliability of the measurements, as discussed in Section 3.2.
•	Only four of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Cook County (formaldehyde, benzene, xylenes, and acetaldehyde). The
highest emitted pollutants and the pollutants with the highest toxicity-weighted
emissions for Madison County have only two pollutants in common (formaldehyde
and hydrochloric acid). This speaks to the relative toxicity of a pollutant; a pollutant
does not have to be emitted in high quantities to be hazardous to human health.
•	Formaldehyde and acetaldehyde have the highest noncancer hazard approximations
for all three Illinois monitoring sites (albeit less than an HQ of 1.0). These two
pollutants appear on both emissions-based lists for Cook County but only
formaldehyde appears on both lists for Madison County (acetaldehyde ranks 12th for
its toxicity-weighted emissions).
•	Naphthalene, arsenic, benzene, and 1,3-butadiene are pollutants of interest for NBIL
and are among those with the highest toxicity-weighted emissions in Cook County
but are not among the highest emitted. Trichloroethylene, benzene, and 1,3-butadiene
are pollutants of interest for SPIL and are among those with the highest toxicity-
weighted emissions but are not among the highest emitted in Cook County.
•	While six 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. Several metals appear among the
pollutants with the highest toxicity-weighted emissions in Madison County, although
none are among the highest emitted. Metals were not sampled for at ROIL under the
NMP.
12-67

-------
12.6 Summary of the 2013 Monitoring Data for NBIL, SPIL, and ROIL
Results from several of the data treatments described in this section include the
following:
~~~ Twenty pollutants (two carbonyl compounds, 12 VOCs, four PAHs, one speciated
metal, and hexavalent chromium) failed screens for NBIL; 12 pollutants (three
carbonyl compounds and nine VOCs) failed screens for SPIL; and 11 pollutants
(three carbonyl compounds and eight VOCs) failed screens for ROIL.
~~~ Formaldehyde had the highest annual average concentration among the pollutants of
interest for SPIL and ROIL, while acetaldehyde had the highest annual average
concentration among the pollutants of interest for NBIL. None of the other site-
specific pollutants of interest had annual average concentrations greater than
1 ng/m3.
~~~ The maximum concentrations of several pollutants across the program were
measured at the Chicago sites. The maximum concentrations of acetaldehyde and
trichloroethylene program-wide were measured at SPIL. The maximum
concentrations of acenaphthene, fluorene, and naphthalene program-wide were
measured at NBIL.
~~~ Concentrations of acetaldehyde have been increasing significantly in recent years at
NBIL. Several of NBIL's pollutants of interest, including benzene, 1,3-butadiene,
arsenic, andformaldehyde, exhibited less variability in 2013, as measurements fell
within their smallest range. Like many NMP sites, a significant decrease in the
number of non-detects reportedfor 1,2-dichloroethane has occurred at both Chicago
sites. While no longer increasing, some of the highest acetaldehyde measurements
have been measured at SPIL in recent years.
~~~ Formaldehyde has the highest cancer risk approximation among the pollutants of
interest for all three sites. None of the pollutants of interest have noncancer hazard
approximations greater than an HQ of 1.0.
12-68

-------
13.0	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.
13.1	Site Characterization
This section characterizes the Indiana monitoring sites by providing geographical and
physical information about the location of the sites and the surrounding areas. This information
is provided to give the reader insight regarding factors that may influence the air quality near the
sites and assist in the interpretation of the ambient monitoring measurements.
One Indiana monitoring site (INDEM) is located in the Chicago-Naperville-Elgin, IL-IN-
WI CBSA, and another site (WPIN) is located in the Indianapolis-Carmel-Anderson, IN CBSA.
Figures 13-1 and 13-3 are composite satellite images retrieved from ArcGIS Explorer showing
the monitoring sites and their immediate surroundings. Figures 13-2 and 13-4 identify nearby
point source emissions locations by source category near INDEM and WPIN, respectively, as
reported in the 2011 NEI for point sources, version 2. Note that only sources within 10 miles of
the sites are included in the facility counts provided in Figures 13-2 and 13-4. A 10-mile
boundary was chosen to give the reader an indication of which emissions sources and emissions
source categories could potentially have a direct effect on the air quality at the monitoring sites.
Further, this boundary provides both the proximity of emissions sources to the monitoring sites
as well as the quantity of such sources within a given distance of the sites. Sources outside the
10-mile boundary are still visible on each map for reference, but have been grayed out in order to
emphasize emissions sources within the boundary. Table 13-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates.
13-1

-------
Figure 13-1. Gary, Indiana (INDEM) Monitoring Site

-------
Figure 13-2. NEI Point Sources Located Within 10 Miles of INDEM
grwonw
Uke
kicrvpan
Portei \
County |
Lake
County
ILLINOIS
Hal* Ou« lo	dan»»ty a ntf cotloeaBon the total facilitM*
Legend	dtsptayed T«y not represent all fucrtbeB Mithtn the area of wt»rest
~ INDEM UATMP Sfte	10
Legend
T Airport/Airkne'Airport Support Operators (14\
j) Asphalt Produdlon'Hot Mm Asphalt Ptarvl (1)
Bnc*. Structural Clay, of Clay Ceramics Plant (1)
9 Bulk Term-na-a'Bulk Plants <5)
C Comical Manufactunng Faciity (11)
L Coke Battery (2)
i Ccrrpr»t*o* Station (7)
* Electricity Generation Via Combustion (6)
E Electroplating Plating Poliiftng Arxxtemg ana Cotonng (2)
A LanflPH (2)
¦ Metal Can Do* and OtKe» Metal Conitalnar Manufacturing <1>
iile radius	County boundary
A Metal Coating Engrav»ng and Allied Services to Manufacture's <1}
0 Metals Processing/Fabrication Facility '71
Mine-'QuanyMineral P'ocessjng Facility <16)
? Wtscelaneous Commercial'! Must^t Facfxty (20)
~ ^int and Coating Manufacturing Faciity t\)
Petroleum Products Manufacturing (2)
d Petroleum Refinery (2 j
R PtaHic. Resin or Rubber Products Plant (1)
P Piint^ng^Publmhing'Pape" Product Manufacturing Facility (1)
X Rail YanS/Ratl Une Operations 16)
V Steei Mi <10)
13-3

-------
Figure 13-3. Indianapolis, Indiana (WPIN) Monitoring Site

ifcdiMfcSti

-------
Figure 13-4. NEI Point Sources Located Within 10 Miles of WPIN
sv&mM
1 Ha«oock
, Comity
Manon '
Cflurty |
I Kwdrtcua
County
arwrw	«*im»	«r WW	WWW	www	vr-v-rrv*
NoW Dj« to toeitry  total factlilM*
Legend	d»sptaye
-------
Table 13-1. Geographical Information for the Indiana Monitoring Sites
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Additional Ambient Monitoring Information1
INDEM
18-089-0022
Gary
Lake
Chicago-
Naperville-Elgin
IL-IN-WI
41.606680,
-87.304729
Industrial
Urban/City
Center
Black carbon VOCs, S02, NO, N02, NOx, PAMS,
O3, Meteorological parameters, PM10, PM2.5, PM2.5
Speciation, SNMOC, IMPROVE Speciation.
WPIN
18-097-0078
Indianapolis
Marion
Indianapolis-
Cannel-Anderson,
IN
39.811097,
-86.114469
Residential
Suburban
Black carbon TSP Metals, CO, VOCs, SNMOCs,
SO2, NOy, NOx, NO2, NO, O3, Meteorological
parameters, PM10, PM2.5. PM2.5 Speciation PM
Coarse, IMPROVE Speciation.
1 Data for additional pollutants are reported to AQS for these sites (EPA, 2014b); however, these data are not generated by ERG and are therefore not included in this report.
u>
6\

-------
INDEM is located in Gary, Indiana, roughly 11 miles east of the Indiana-Illinois border
and 25 miles southeast of Chicago. Gary is located on the southernmost bank of Lake Michigan.
The site is located just north of 1-90, the edge of which can be seen in the bottom left portion of
Figure 13-1, and 1-65. Although INDEM resides on the Indiana Dunes National Lakeshore, about
1 mile south of the Lake Michigan shoreline, the surrounding area is highly industrialized, as
shown in Figure 13-1, and several railroads transverse the area. Figure 13-2 shows that the
majority of point sources within 10 miles of INDEM are located to the west of the site. There is
also a second cluster of facilities located to the east of INDEM in Porter County. The emissions
source categories with the highest number of sources within 10 miles of INDEM include steel
mills; aircraft operations, which includes airports and related operations as well as small runways
and heliports, such as those associated with hospitals or TV stations; chemical manufacturing;
and mine/quarry/mineral processing. The sources closest to INDEM include a steel mill; an
industrial complex that includes several facilities that fall into the miscellaneous
commercial/industrial category as well as two mines/quarries and another steel mill; and a
heliport at a police station and a hospital.
WPIN is located in the parking lot of a police station across from George Washington
Park, near East 30th Street in northeast Indianapolis. Figure 13-3 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 13-4 shows that the majority of point sources are located to the south
and southwest of WPIN, towards the center of Marion County. The source category with the
highest number of sources near WPIN is the airport operations source category. The sources
closest to WPIN are a painting and coating manufacturer, a metals processing/fabrication facility,
and a heliport. Each of these facilities is located within 2 miles of WPIN.
Table 13-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Indiana 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 Marion and Lake Counties.
13-7

-------
Table 13-2. Population, Motor Vehicle, and Traffic Information for the Indiana Monitoring
Sites




Annual




Estimated
County-level
Average
Intersection
County-


County
Vehicle
Daily
Used for
level Daily
Site
County
Population1
Registration2
Traffic3
Traffic Data
VMT4
INDEM
Lake
491,456
425,854
34,754
1-90 N of 1-65 Interchange
15,741,000
WPIN
Marion
928,281
830,851
143,970
1-70 b/w Exit 85 & 87
31,727,000
bounty-level population estimates reflect 2013 data (Census Bureau, 2014)
2County-level vehicle registrations reflect 2013 data (IN BMV, 2014)
3AADT reflects 2011 data (IN DOT, 2011)
4County-level VMT reflects 2013 data (IN DOT, 2013)
Observations from Table 13-2 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 less than 1 mile south of WPIN. Traffic data were not available for a location
closer to WPIN. The traffic volume near WPIN is the seventh highest among NMP
sites.
•	The traffic volume for INDEM is based on data from the 1-90 toll road north of the
1-65 interchange. Traffic near INDEM is in the middle of the range among traffic
volumes for NMP sites.
•	The VMT for Marion County is roughly twice the VMT for Lake County. The
Marion County VMT ranks 11th among counties with NMP sites, while the VMT for
Lake County is in the middle of the range, ranking 21st.
13.2 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.
13.2.1 Climate Summary
The city of Gary is located to the southeast of Chicago, at the southern-most tip of Lake
Michigan. Climate of the region is characterized by warm, humid summers, cloudy and cold
13-8

-------
winters, and frequently changing weather as storm systems regularly track across the region.
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 late fall and winter can
provide cloudiness and 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 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).
13.2.2 Meteorological Summary
Hourly meteorological data for 2013 were retrieved from NCDC for the weather stations
closest to the Indiana monitoring sites (NCDC, 2013), as described in Section 3.4.2. 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 13-3. These data were used to determine how meteorological conditions on
sample days vary from conditions experienced throughout the year.
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 2013. Also included in Table 13-3 is the 95 percent
confidence interval for each parameter. As shown in Table 13-3, average meteorological
conditions on sample days at WPIN and INDEM were representative of average weather
conditions experienced throughout the year near these locations.
13-9

-------
Table 13-3. Average Meteorological Conditions near the Indiana Monitoring Sites
Closest Weather
Station
Distance
and

Average
Maximum
Average
Average
Dew Point
Average
Wet Bulb
Average
Relative
Average
Sea Level
Average
Scalar Wind
(WBAN and
Coordinates)
Direction
from Site
Average
Type1
Temperature
(°F)
Temperature
(°F)
Temperature
(°F)
Temperature
(°F)
Humidity
(%)
Pressure
(mb)
Speed
(kt)
Gary, Indiana - INDEM
Lansing Municipal
Airport
12.7
miles
Sample
Days
(61)
56.8
±5.6
48.6
±5.3
38.8
±5.3
44.1
±4.9
71.6
±3.0
NA
6.7
±0.7
04879
(41.54, -87.53)
249°
(WSW)

57.6
49.2
39.2
44.5
71.2

7.0

2013
±2.2
±2.0
± 1.9
± 1.8
± 1.1
NA
±0.3
Indianapolis, Indiana - WPIN
Eagle Creek
Airpark
9.7
miles
Sample
Days
(65)
59.4
±5.4
51.6
±5.1
42.3
±5.2
47.1
±4.8
73.0
±2.9
1018.2
± 1.8
5.5
±0.5
53842
(39.83, -86.30)
276°
(W)

59.8
51.7
42.3
47.2
72.4
1017.7
5.7

2013
±2.1
±2.0
±2.0
± 1.8
± 1.1
±0.7
±0.3
1 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.

-------
13.2.3 Wind Rose Comparison
Hourly surface wind data from the weather stations nearest the Indiana 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 3.4.2. A wind rose shows the frequency of wind directions using
"petals" positioned around a 16-point compass, and uses different colors to represent wind
speeds.
Figure 13-5 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 13-5 also presents three different
wind roses for the INDEM monitoring site. First, a historical wind rose representing 2003 to
2012	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
2013	is presented. Next, a wind rose representing wind data for days on which samples were
collected in 2013 is presented. These can be used to identify the predominant wind speed and
direction for 2013 and to determine if wind observations on sample days were representative of
conditions experienced over the entire year and historically. Figure 13-6 presents the distance
map and three wind roses for WPIN.
13-11

-------
west:
Figure 13-5. Wind Roses for the Lansing Municipal Airport Weather Station near INDEM
Location of INDEM and Weather Station	2003-2012 Historical Wind Rose
WIND SPEED
(Knots)
~
~1 4-7
2- 4
Calms: 19.23%
2013 Wind Rose
WEST
(Knots)
11 - 17
SOUTH
WIND SPEED
~ 4-7
Calms: 1459%
Sample Day Wind Rose
NORTH"--.,
WEST
WIND SPEED
(Knots)
~ >=22
~
7- 11
4- 7
Calms: 15.11%
13-12

-------
west:
Figure 13-6. Wind Roses for the Eagle Creek Airpark Weather Station near WPIN
Location of WPIN and Weather Station	2003-2012 Historical Wind Rose
WIND SPEED
(Knots)
~
~1 4-7
2- 4
Calms: 18.78%
2013 Wind Rose
WEST
(Knots)
11 - 17
SOUTH
WIND SPEED
~ 4-7
Calms: 17.20%
Sample Day Wind Rose
NORTH"--.,
ES ,
WIND SPEED
(Knots)
~ >=32
GO 17-21
IH 11 -17
f I 7- 11
~ 4-7
2- 4
Calms: 16.86%
13-13

-------
Observations from Figure 13-5 for INDEM include the following:
•	The weather station at Lancing Municipal Airport is the closest weather station to
INDEM, although it is located nearly 13 miles west-southwest of INDEM. The
location of the weather station is just over 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 2003-2012 time
frame. Northerly to northeasterly winds off Lake Michigan account for less than
20 percent of the wind measurements, as do calm winds (those less than or equal
to 2 knots). Winds from the southeast and northwest quadrants were less frequently
observed. The strongest winds were those from the south to southwest to west.
•	The wind patterns shown on the 2013 wind rose generally resemble the wind patterns
shown on the historical wind rose. There were, however, fewer calm winds and a
higher percentage of winds from the west in 2013.
•	The sample day wind patterns resemble the full-year wind patterns, although there
were fewer observations of northerly and north-northeasterly winds.
Observations from Figure 13-6 for WPIN include the following:
•	The weather station at Eagle Creek Airpark is the closest weather station to WPIN
and is located approximately 10 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 55 percent) of wind observations from 2003 to 2012, while winds
from the eastern quadrants were observed for approximately one-quarter of the
observations. Calm winds were observed for roughly 19 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 winds from the western quadrants account for an even higher percentage of
wind observations. The calm rate was slightly lower in 2013 (accounting for roughly
17 percent of observations).
•	The sample day wind patterns resemble the full-year wind patterns but with a higher
percentage of winds from the south to southwest and fewer calm wind observations.
13-14

-------
13.3 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 13-4. 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 13-4. It is important to note which pollutants were sampled for at each site when reviewing
the results of this analysis. Carbonyl compounds were sampled for at both INDEM and WPIN.
Table 13-4. Risk-Based Screening Results for the Indiana Monitoring Sites
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Gary, Indiana - INDEM
Formaldehyde
0.077
61
61
100.00
50.41
50.41
Acetaldehyde
0.45
60
61
98.36
49.59
100.00
Total
121
122
99.18

Indianapolis, Indiana - WPIN
Acetaldehyde
0.45
58
58
100.00
50.00
50.00
Formaldehyde
0.077
58
58
100.00
50.00
100.00
Total
116
116
100.00

Observations from Table 13-4 include the following:
•	Formaldehyde, acetaldehyde, and propionaldehyde are the only carbonyl compounds
with risk screening values.
•	Acetaldehyde and formaldehyde are the only pollutants to fail screens for INDEM
and WPIN.
•	Formaldehyde failed 100 percent of screens for both sites. Acetaldehyde failed
100 percent of screens for WPIN and 98 percent of screens for INDEM. Both
pollutants were identified as pollutants of interest for each site.
13-15

-------
13.4 Concentrations
This section presents various concentration averages used to characterize air toxics
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. Site-specific statistical summaries for all pollutants
sampled for at the Indiana sites are provided in Appendix L.
13.4.1 2013 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 13-5, 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.
13-16

-------
Table 13-5. 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
frig/m3)
2nd
Quarter
Average
frig/m3)
3rd
Quarter
Average
frig/m3)
4th
Quarter
Average
frig/m3)
Annual
Average
frig/m3)
Gary, Indiana - INDEM
Acetaldehyde
61/61
0.88
±0.11
1.16
±0.18
1.11
±0.26
1.02
±0.11
1.04
±0.09
Formaldehyde
61/61
1.47
±0.13
2.87
±0.98
2.77
±0.57
1.49
±0.19
2.14
±0.32
Indianapolis, Indiana - WPIN
Acetaldehyde
58/58
1.53
±0.26
2.33
±0.31
1.83
±0.22
1.32
±0.19
1.78
±0.16
Formaldehyde
58/58
2.48
±0.33
4.64
±0.57
4.15
±0.63
2.05
±0.42
3.41
±0.37
Observations for the Indiana sites from Table 13-5 include the following:
•	For both sites, acetaldehyde and formaldehyde were detected in all of the valid
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 roughly 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. A review
of the data shows that the 14 highest formaldehyde concentrations (those greater than
2.50 |ig/m3) were measured between April and September and ranged from
2.69 |ig/m3 to 8.90 |ig/m3; conversely, all but two of the 22 lowest concentrations
(those less than 1.50 |ig/m3) were measured between January and March or October
and December. This supports the trend identified in Section 4.4.2 indicating that
formaldehyde concentrations tended to be higher during the warmer months of the
year.
•	This trend in seasonality continues at WPIN. All but one of the 20 formaldehyde
concentrations greater than 4.0 |ig/m3 were measured between April and September
while all 11 measurements less than 2.00 |ig/m3 were measured in February or
between October and December.
13-17

-------
Tables 4-9 through 4-12 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for the Indiana
sites from those tables include the following:
•	WPIN's annual average concentration of formaldehyde is the sixth highest annual
average of this pollutant among NMP sites sampling carbonyl compounds. WPIN
does not appear in Table 4-10 for acetaldehyde (it ranks 13th).
•	INDEM does not appear in Table 4-10. Its annual average concentration of
formaldehyde ranks 19th and its annual average concentration of acetaldehyde ranks
24th among NMP sites sampling carbonyl compounds.
13.4.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for the pollutants shaded in
gray in Table 13-4 for INDEM and WPIN. Figures 13-7 and 13-8 overlay the sites' minimum,
annual average, and maximum concentrations onto the program-level minimum, first quartile,
median, average, third quartile, and maximum concentrations, as described in Section 3.4.3.1.
Figure 13-7. Program vs. Site-Specific Average Acetaldehyde Concentrations
INDEM
WPIN
0
6
9
12
15
Concentration {[jg/m3)
Prog ra m: 1st Qu a r ti I e
2nd Quartile 3rd Quartile 4th Quartile Average
~ ~ ~
Site:
Site Average Site Concentration Range
o
13-18

-------
Figure 13-8. Program vs. Site-Specific Average Formaldehyde Concentrations
E
¦
0	3	6	9
12	15
Concentration (pg/m3)
18	21	24
Progra m: 1st Qua rti 1 e
2nd Quartile 3rd Quartile
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



Observations from Figures 13-7 and 13-8 include the following:
• Figure 13-7 presents the box plots for both sites for acetaldehyde. 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 just greater than the program-level average concentration as well as the
annual average concentration for WPIN. WPIN's annual average concentration is
similar to the program-level average concentration. The minimum concentration
measured at WPIN is just less than the program-level first quartile.
• Figure 13-8 presents the box plots for formaldehyde for both sites. Although the
range of formaldehyde concentrations measured at INDEM is larger than the
range measured at WPIN, INDEM's annual average concentration is less than
WPIN's. The annual average concentration for INDEM is less than the program-
level average concentration and similar to the program-level median
concentration while the annual average for WPIN is greater than the program-
level average but less than the third quartile. The minimum formaldehyde
concentration measured at WPIN is greater than the program-level first quartile.
13.4.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.3.2.
INDEM and WPIN have sampled carbonyl compounds under the NMP since 2004 and 2007,
respectively. Thus, Figures 13-9 through 13-12 present the 1-year statistical metrics for each of
the pollutants of interest first for INDEM, then for WPIN. The statistical metrics presented for
assessing trends include the substitution of zeros for non-detects. If sampling began mid-year, a
13-19

-------
minimum of 6 months of sampling is required for inclusion in the trends analysis; in these cases,
a 1-year average concentration is not provided, although the range and percentiles are still
presented.
Figure 13-9. Yearly Statistical Metrics for Acetaldehyde Concentrations
Measured at INDEM
>•
2004	20051	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 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 13-9 for acetaldehyde measurements collected at INDEM
include the following:
•	Although carbonyl compound sampling under the NMP began in 2003, samples were
only collected for 3 months. Carbonyl compound sampling began in earnest at
INDEM at the beginning of 2004; thus, Figure 13-9 begins with 2004. However, a
1-year average concentration 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).
13-20

-------
•	Although the maximum and 95th percentile increased from 2007 to 2008, the 1-year
average, median, 5th percentile, and minimum concentrations of acetaldehyde all
exhibit decreases from 2007 to 2008. Although three concentrations greater than
10 |ig/m3 were measured in 2008 (compared to zero in 2007), the number of
measurements at the lower end of the concentration range increased significantly. The
number of acetaldehyde concentrations less than 2 |ig/m3 increased seven-fold (from
three in 2007 to 21 for 2008).
•	With the exception of the minimum and 5th percentile, the statistical parameters
decreased significantly from 2008 to 2009. The 1-year average and median
concentrations decreased by more than half and the 95th percentile decreased by more
than 80 percent during this time. The carbonyl compound samplers were switched out
in 2009, which seems to have had a significant effect on the concentrations measured,
particularly with respect to formaldehyde, which is discussed in more detail below.
•	Most of the statistical parameters exhibit a slight decreasing trend between 2010 and
2013, with many of them at a minimum for 2013. The median concentration for 2013
is less than 1.00 |ig/m3 for 2013 and the 1-year average concentration is close behind.
Figure 13-10. Yearly Statistical Metrics for Formaldehyde Concentrations
Measured at INDEM
3 300
o..
2004	20051	2006	2007	2008 J 2009	2010	2011	2012	2013
O 5th Percentile	— Minimum	— Median	— Maximum	O 95th Percentile	Average
1 A 1-year average is not presented due to a break in sampling between September 2005 and November
2005."
13-21

-------
Observations from Figure 13-10 for formaldehyde measurements collected at INDEM
include the following:
•	Five formaldehyde concentrations greater than 400 |ig/m3 were measured in the
summer of 2008 (ranging from 414 |ig/m3 to 500 |ig/m3). While these are extremely
high values of formaldehyde, concentrations of formaldehyde have been historically
high at this site, as shown by the statistics in Figure 13-10. There have been 38
concentrations of formaldehyde greater than 100 |ig/m3 measured at INDEM.
•	Prior to 2009, the maximum concentration for each year is greater than 150 |ig/m3. In
addition, the median concentrations for 2004, 2006, and 2007 are greater than
30 |ig/m3, indicating that at least half of the concentrations were greater than
30 |ig/m3 for these years.
•	Although the 1-year average concentration doubled from 2007 to 2008, the median
concentration decreased by more than half. This means that although the magnitude
of those higher measurements is driving the 1-year average concentration upward,
there were a larger number of concentrations at the lower end of the concentration
range as well. There were 24 formaldehyde concentrations measured in 2008 that
were less than the minimum concentration measured in 2007; those 24 measurements
represent 40 percent of the concentrations measured in 2008. The last "high"
concentration was measured on August 10, 2008, after which no formaldehyde
concentrations greater than 4 |ig/m3 were measured in 2008.
•	All the statistical metrics decreased significantly for 2009 and the years that follow,
with the 1-year average concentrations ranging from 2.14 |ig/m3 (2013) to 2.58 |ig/m3
(2009). In contrast to the previous bullet, the number of measurements greater than
4 |ig/m3 ranged from two to seven for each year between 2009 and 2013 (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 median concentrations shown before and
after 2009, as discussed in the previous bullets.
13-22

-------
Figure 13-11. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at WPIN
2010
Year
0 5th Percentile	— Minimurr
0 95th Percentile
Observations from Figure 13-11 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 13-11 begins with 2007.
•	The three highest acetaldehyde concentrations were measured at WPIN in 2010 and
ranged from 5.96 |ig/m3 to 6.72 |ig/m3. Three additional concentrations greater than
5 |ig/m3 have been measured at WPIN (two in 2007 and one in 2012).
•	The 1-year average concentration has a decreasing trend through 2009, after which a
significant increase is shown. For 2010, all of the statistical parameters increased,
particularly the maximum (which doubled) and the 95th percentile (which increased
by nearly 60 percent). The 1-year average concentration has a slight decreasing trend
again after 2010, with the 1-year average concentration at a minimum for 2013 over
the years of sampling.
13-23

-------
Figure 13-12. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at WPIN
2007	2008	2009	2010	2011	2012	2013
Year
0 5th Percentile	— Minimum	— Median	- Maximum	0 95th Percentile	Average
Observations from Figure 13-12 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/m3 to 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 2012. Although the
1-year average concentration did not change significantly, the median concentration
for 2012 decreased considerably. 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 in 2011 to 11 in 2012). These
changes explain the change in the median concentration while a few additional
measurements in the upper end of the concentration range explain the increase in the
95th percentile.
•	Nearly all of the statistical parameters exhibit decreases for 2013, particularly the
95th percentile.
13-24

-------
13.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at each Indiana monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.3 for
definitions and explanations regarding the various toxicity factors, time frames, and calculations
associated with these risk-based screenings.
13.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Indiana sites and where annual average
concentrations could be calculated, risk was examined by calculating cancer risk and noncancer
hazard approximations. These approximations can be used as risk estimates for cancer and
noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers want to shift their air
monitoring priorities. Refer to Section 3.4.3.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 13-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Table 13-6. Risk Approximations for the Indiana Monitoring Sites
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Gary, Indiana - INDEM
Acetaldehyde
0.0000022
0.009
61/61
1.04
±0.09
2.29
0.12
Formaldehyde
0.000013
0.0098
61/61
2.14
±0.32
27.77
0.22
Indianapolis, Indiana - WPIN
Acetaldehyde
0.0000022
0.009
58/58
1.78
±0.16
3.92
0.20
Formaldehyde
0.000013
0.0098
58/58
3.41
±0.37
44.34
0.35
13-25

-------
Observations for the Indiana sites from Table 13-6 include the following:
•	For both sites, the annual average concentration of formaldehyde is greater than the
annual average concentration of acetaldehyde. The annual average concentrations 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 (44.34 in-a-million) is the
sixth highest cancer risk approximation for formaldehyde program-wide and the
seventh highest among all site-specific pollutants of interest across the program.
•	Neither pollutant of interest for INDEM or WPIN have noncancer hazard
approximations greater than 1.0, indicating that no adverse noncancer health effects
are expected from these individual pollutants. The noncancer hazard approximation
for WPIN ranks 10th highest among all pollutants of interest program-wide.
13.5.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 13-7 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
that have cancer toxicity factors. Table 13-7 also presents the 10 pollutants with the highest
toxicity-weighted emissions, based on the weighting schema described in Section 3.4.3.4. Lastly,
Table 13-7 provides the pollutants with the highest cancer risk approximations (in-a-million) for
each site, as presented in Table 13-6. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 13-7. Table 13-8 presents similar
information, but is limited to those pollutants with noncancer toxicity factors.
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.3.4. Similar to the cancer risk and
13-26

-------
noncancer hazard approximations provided in Section 13.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
Observations from Table 13-7 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, POM, Group lb, and formaldehyde, are the pollutants with the
highest toxicity-weighted emissions (of the pollutants with cancer UREs) for Lake
County. Formaldehyde, benzene, and 1,3-butadiene are the pollutants with the highest
toxicity-weighted emissions for Marion County.
•	Six of the highest emitted pollutants in Lake County also have the highest toxicity-
weighted emissions; seven of the highest emitted pollutants in Marion County also
have the highest toxicity-weighted emissions.
•	Acetaldehyde and formaldehyde are the only pollutants of interest for INDEM and
WPIN. Acetaldehyde and formaldehyde appear among the highest emitted pollutants
for both counties, with only formaldehyde appearing among the pollutants with the
highest toxicity-weighted emissions.
•	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 under the NMP.
13-27

-------
Table 13-7. 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)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Gary, Indiana (Lake County) - INDEM
Benzene
177.07
Coke Oven Emissions, PM
2.38E-03
Formaldehyde
27.77
Formaldehyde
145.37
POM, Group lb
1.92E-03
Acetaldehyde
2.29
Ethylbenzene
94.06
Formaldehyde
1.89E-03

Acetaldehyde
84.11
Benzene
1.38E-03
1.3 -Butadiene
27.28
Hexavalent Chromium
9.67E-04
POM, Group lb
21.84
1,3-Butadiene
8.18E-04
Naphthalene
13.35
Arsenic, PM
6.53E-04
Tetrachloroethylene
9.35
Naphthalene
4.54E-04
POM, Group 2b
2.78
POM, Group 2b
2.45E-04
POM, Group 2d
2.68
Nickel, PM
2.38E-04
Indianapolis, Indiana (Marion County) - WPIN
Benzene
421.74
Formaldehyde
4.14E-03
Formaldehyde
44.34
Formaldehyde
318.24
Benzene
3.29E-03
Acetaldehyde
3.92
Ethylbenzene
268.73
1,3-Butadiene
1.87E-03

Acetaldehyde
189.64
Naphthalene
1.11E-03
1,3-Butadiene
62.21
Arsenic, PM
1.05E-03
Tetrachloroethylene
33.59
Ethylbenzene
6.72E-04
Naphthalene
32.73
POM, Group 2b
6.51E-04
POM, Group 2b
7.40
Nickel, PM
5.08E-04
POM, Group 2d
5.22
POM, Group 2d
4.60E-04
Propylene oxide
4.72
Hexavalent Chromium
4.20E-04

-------
Table 13-8. 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)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Gary, Indiana (Lake County) - INDEM
Toluene
669.66
Acrolein
476,219.32
Formaldehyde
0.22
Xylenes
436.89
Lead, PM
52,690.40
Acetaldehyde
0.12
Hexane
427.58
Manganese, PM
22,492.64

Methanol
328.00
Hydrochloric acid
16,187.23
Hydrochloric acid
323.74
Formaldehyde
14,834.03
Benzene
177.07
1,3-Butadiene
13,641.59
Formaldehyde
145.37
Chlorine
12,016.67
Ethylene glycol
98.80
Arsenic, PM
10,126.34
Ethylbenzene
94.06
Acetaldehyde
9,345.56
Acetaldehyde
84.11
Benzene
5,902.45
Indianapolis, Indiana (Marion County) - WPIN
Toluene
1,660.99
Acrolein
1,224,556.10
Formaldehyde
0.35
Xylenes
1,008.89
Formaldehyde
32,473.78
Acetaldehyde
0.20
Hexane
773.82
1,3-Butadiene
31,104.81

Methanol
532.81
Hydrochloric acid
23,337.36
Hydrochloric acid
466.75
Acetaldehyde
21,070.71
Benzene
421.74
Arsenic, PM
16,282.89
Formaldehyde
318.24
Benzene
14,057.94
Ethylbenzene
268.73
Lead, PM
13,691.58
Ethylene glycol
203.01
Nickel, PM
11,766.86
Acetaldehyde
189.64
Naphthalene
10,909.09

-------
Observations from Table 13-8 include the following:
•	Toluene, xylenes, and hexane 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.
•	Acrolein is the pollutant with the highest toxicity-weighted emissions (of the
pollutants with noncancer RfCs) for both counties. Lead and manganese rank second
and third for Lake County, while formaldehyde and 1,3-butadiene rank second and
third for Marion County.
•	Four of the highest emitted pollutants in Lake County also have the highest toxicity-
weighted emissions (formaldehyde, acetaldehyde, benzene, and hydrochloric acid).
The same four pollutants appear on both emissions-based lists for Marion County.
•	Several metals are among the pollutants with the highest toxicity-weighted emissions
for Lake and Marion Counties, although none of these appear among the highest
emitted pollutants.
•	Formaldehyde and acetaldehyde appear in all three columns in Table 13-8 for both
sites.
13.6 Summary of the 2013 Monitoring Data for INDEM and WPIN
Results from several of the data treatments described in this section include the
following:
~~~ Carbonyl compounds were sampled for at INDEM and WPIN in 2013. Acetaldehyde
andformaldehyde failed screens for each site and were identified as pollutants of
interest for each site.
~~~ The annual average concentration of formaldehyde is greater than the annual
average concentration of acetaldehyde for both sites, with the annual averages for
WPIN greater than the annual averages for INDEM.
~~~ Concentrations of formaldehyde and acetaldehyde exhibited a significant decreasing
trend at INDEMfrom 2008 to 2009; these changes may be at least partially explained
by a sampler replacement.
13-30

-------
14.0	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.
14.1	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, east of Paducah, Kentucky. The final monitoring site is located in
Lexington, in north-central Kentucky. A composite satellite image and facility map is provided
for each site in Figures 14-1 through 14-15. The composite satellite images were retrieved from
ArcGIS Explorer and show each monitoring site in its respective location. The facility maps
identify nearby point source emissions locations by source category, as reported in the 2011 NEI
for point sources, version 2. Note that only sources within 10 miles of each site are included in
the facility counts provided. A 10-mile boundary was chosen to give the reader an indication of
which emissions sources and emissions source categories could potentially have a direct effect
on the air quality at each monitoring site. Further, this boundary provides both the proximity of
emissions sources to each monitoring site as well as the quantity of such sources within a given
distance of the sites. Sources outside the 10-mile boundaries are still visible on the maps for
reference, but have been grayed out in order to emphasize emissions sources within the
boundary. Table 14-1 provides supplemental geographical information such as land use, location
setting, and locational coordinates for each site.
14-1

-------
Figure 14-1. Ashland, Kentucky (ASKY) Monitoring Site

-------
Figure 14-2. Ashland, Kentucky (ASKY-M) Monitoring Site
-p*
>o o r ic e N A
TL ^j'oOa M

-------
Figure 14-3. NEI Point Sources Located Within 10 Miles of ASKY and ASKY-M
82"MOV#	B/-43JTW	tOMSWv	tt ttYVk	XTWH	eCZIfW

OHIO
County
Boyd
County
Carter
County
V^ST
VIRGINIA '
Note Due to facility density end cclk>c*tie« the total bc*te«
Legend	displayed may not represent afe facrttttes ttMs ansa of interest
^ ASKY UATMP site	ASKY-M UATMP srte	10 mile radius I	County boundary
Source Category Group (No. of Facilities)
T
Airport/Airline/Airport Support Operations (6)
*
Mine/Quarry/Mineral Processing Facility (8)
*
Asphalt Production/Hot Ml* Asptialt Plant (1)
o
Miscellaneous CoirimerciaWlndustrial Facility (3)
D
Bulk Terrmnals/Bu'k Plants (1)
D
Paint and Coating Manufacturing Facility (1)
C
Chemical Manufacturing Facility (5)
<
Pesticide Manufacturing Plant < 1)
E.
Coke Battery (1)
4
Petroleum Refinery (1)
1
Compressor Station (3)
R
Plastic, Rusin, or Rubber Products Plant (2)
«
Electricity Generation via Combustion (4)
V
Port and Harbor Operations (2)
F
Food Processing/Agriculture Facility (2)
X
Rail Yard/Rail Line Operations (2)
¦
Gasoline/Diesel Service Station (1)
A
Ship/Boat Manufactunng or Repair Facility (1)
¦
Landfill (2)
V
Steel Milt (2)

Metals Processing/Fabrication Facility (2)
*
Testing Laboratories (1)
14-4

-------
Figure 14-4. Grayson, Kentucky (GLKY) Monitoring Site

-------
Figure 14-5. NEI Point Sources Located Within 10 Miles of GLKY
ttl'OV*/	SfTTM	BTWVK	B"550-W	(ETBtTW
OrMrhip
count) I
Boyd
County

Carter
Gotintv
Rowan
Cour anc sollucaboit IM total facJMWi
disptoy^d aiay r>o( rapie»Bnl all r»ciib»a Mdthin a>»a or 4if»r*9(
Legend
~ GLKY NATTS site	10 mi(e radius	| County boundary
Source Category Group (No. of Facilities)
* Asphalt Production/Hot Mix Asphalt Plant (1)
= Brick. Structural Clay, or Clay Ceramics Plant (2)
B Bulk Terminals/Bulk Plants (1)
F Food Processing/Agnculture Facility (1)
o Institutional (school, hospital, prison, etc.) (1)
Mlne/Quany/Mineral Processing Facility (2)
14-6

-------
Figure 14-6, Baskett, Kentucky (BAKY) Monitoring Site
*091*
esri
SourcV U5G5
ircc NASA NGA OSGS
;OOK M,cr a\a(t Co'p


-------
Figure 14-7. NEI Point Sources Located Within 10 Miles of BAKY
ar*nrw	tf-ww	nr-wcn*	irsmyi	wjw
i>mmm
I He*i*d miy not r»cr«**r*l *» <" BAKY UATMP site	10 mile radius	[ County boundary
Source Category Group (No. of Facilities)
T
AirporVAirtinei'Airport Support Operations (6)
M
Mlne/Quarry'Minerai Processing Facility (3)
it
Asphalt Production'Hot Mm Asphalt Plant (1)
?
Miscellaneous CommerciaWndustiiai Faculty (3)
0
BuiK Terminals/Bulk Plants (1)
c
Paint and Coating Manufacturing Facility (1)
C
Chemical Manufacturing Facility (2)
R
Plastic Resin or Rubber Products Plant (1)
«i
Dry Cleaning Facility (2)
IB
Pulp and Paper Plant (2)
1
Electricity Generation via Combustion (4)
K
Rail Yard/Rail Line Operations 11)
E
Electroplating Plating, Polishing. Anodizing and Coloring (1)
A
Ship/Boat Manufacturing or Repair Facility (1)
F
Food Processing;Agriculture Facility (3)
«
Testing Laboratories 111
®
Metals Processing,'Fabrication Facility (6)


14-8

-------
Figure 14-8. Calvert City, Kentucky (ATKY) Monitoring Site
•f*
'O
r
\vjhIs
\fjml
f/ v




\ *
iU\
* A
G3 Q
' $E
« *
SW./5
%3Sji
n5 x r*t
fiHbk
w kV V
' *

Source U5GS «l*
Sdurce NASA NGAJu?
2008 MiAoiofl Ifnrc

-------
Figure 14-9. Smithland, Kentucky (BLKY) Monitoring Site

-------
Figure 14-10. Calvert City, Kentucky (CCKY) Monitoring Site

-------
Figure 14-11. Calvert City, Kentucky (LAKY) Monitoring Site

-------
Figure 14-12. Calvert City, Kentucky (TVKY) Monitoring Site
'Sflftrtg: USG5
iourtf NASA NGA, USGS
JJ'OO ».Mi c r o » o r t C Of p

-------
Figure 14-13. NEI Point Sources Located Within 10 Miles of ATKY, BLKY, CCKY,
LAKY, and TVKY
Lwngttpn
County
OHIO
Lyon
County
Maiahai
County
Kentucky
Latm t
04YM
County
& ATKY UATMP site yf CCKY UATMP site
^ BLKY UATMP site ^ LAKY UATM P site
Source Category Group (No. of Facilities)
^ TVKY UATMP site
10 mile radius ~ County boundary
T
Airport'Airline/Airport Support Operations (1)

Metals Processmg/Fabncation Facility (3)
»
Asphalt Produc1lon/Hot Ml* Asphalt Plant (1)
*
Mine/Quarry/Mineral Processing Facility (7)
B
Bulk Temmats.'Bulk Plants 11)
¦9
Miscellaneous Commercsali'lndustnal Facility (S)
C
Chemical Manufacturing Facility (10)
R
Plastic Resin or Rubber Products Plant (4)
I
Compressor Station (1)
4
Sh>P''Boat Manufacturing or Repair Facility (4)
t
Electricity Generation via Combustion (1)
V
Steel M3I (1)
~
Industrial Machinery or Equipment Plant (1)


s-xnrw	«*»irw	sa ^rrvi	e8 2»v>n	m-upirw	»icu-w
Note Oufl to	and collocjoorv th« ioui facilmes
Legend	displayed may mil represent all (»(;«**# within the 3'ofl of filaiertt
MeCracken i
County .
Crwenoefi
County
14-14

-------
Figure 14-14. Lexington, Kentucky (LEKY) Monitoring Site
had Dr

JnghLS^
ourcc U5GS	.
I; NASA NGA. USG
jVw itrotdM Corp
4^

-------
Figure 14-15. NET Point Sources Located Within 10 Miles of LEKY
Bourbon
County
» Soafl
I County
Fayette
County
Legend
t*
^ *A«odtorct
County
[
i
V- » "
I
1
I
4
I
Net# Ou« to •¦cMy d»n*ily «rvd colWctOon, ih* total
3itniarocJ may not reprsMnl al tariiUes tha ana rl •nto'ast
~ LEKY UATMP site _ 10 mil
Source Category Group (No. of Facilities)
T	AirpoiVAirtme/Airpgtf Support Op(Maliens (B)
*	Animal »edlM or Farm (1>
4	Aspftalt Production-Hot MM A&cNW Plan! (4)
0	AuVa Body SHopiPamlersJAiiomotlye Stoiea (2)
mi	AutomoBlle'Truc* Manufacturing Faculty (4|
B	Bdlk '•rrr'nals'B*ill< Plants (3)
I-'-	Crematory - AnimalfHurnarH 11
'I'	Dry Cleaning Facility |l)
e	Flecmoa* Equipment Manufacturing Faculty (3)
1	EJsanclty Generation vie Combustion i1)
fc	BecJioplaLng Plating, Polishing Arioit(2ing and Coloring (1)
F	Fo«l Processing,'Apncultuie Facility (2)
%	(Sass Plant (1)
radius 	 County boundary
•	Industrial Machinery or Equipment Plant (B)
O	Irwuituttonal ;#cftool, hoapitai pnact etc i (10)
•	UnsiM (2)
©	tMtftfs Ptooeasirrg.'Faorealion Facility (2i
A	«A«tar, Base National Security ^acntti (1)
~AneiQuarryMltwral Processng *acMy (13i
?	M«a»iian»ou» Comma rcmlinduvnal Faoiltv 1*1
Q	Paint and Coaling Manufacturing Facility fl)
R	Plaattc. He ten. ot Rubber Products Plant |3(
P	Prlntlnj'PiiDllsWng.'Paper Product Manufacturing Facility ti)
n	l^iecomrpunlcaoonsi'aadio Fae«y j 11
M	To bacon Manutaetunng 0}
W	WbodWDtV Furniture, MiSwnrk & Mtocd Preserving Faculty (4)
14-16

-------
Table 14-1. Geographical Information for the Kentucky Monitoring Sites
Site Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude and
Longitude
Land Use
Location
Setting
Additional Ambient Monitoring Information1
ASKY
21-019-0017
Ashland
Boyd
Huntington-
Ashland, WV-KY-
OH
38.45934,
-82.64041
Residential
Suburban
SO2, NO, NO2, O3, Meteorological parameters,
PM2.5. PM2.5 Speciation IMPROVE Speciation.
ASKY-M
21-019-0002
Ashland
Boyd
Huntington-
Ashland, WV-KY-
OH
38.476,
-82.63137
Industrial
Urban/City
Center
PM10.
GLKY
21-043-0500
Grayson
Carter
None
38.23887,
-82.9881
Residential
Rural
O3, Meteorological parameters, PM10, PM2.5,
PM2.5 Speciation, IMPROVE Speciation.
BAKY
21-101-0014
Baskett
Henderson
Evansville, IN-KY
37.8712,
-87.46375
Commercial
Rural
SO2, O3, Meteorological parameters, PM10, PM2.5,
PM2.5 Speciation.
ATKY
21-157-0016
Calvert
City
Marshall
None
37.04176,
-88.35407
Industrial
Suburban
None.
BLKY
21-139-0004
Smithland
Livingston
Paducah KY-IL
37.07151,
-88.33389
Agricultural
Rural
Meteorological parameters.
CCKY
21-157-0018
Calvert
City
Marshall
None
37.02702,
-88.34387
Residential
Suburban
Meteorological parameters, PM10.
LAKY
21-157-0019
Calvert
City
Marshall
None
37.03718,
-88.33411
Residential
Suburban
None.
TVKY
21-157-0014
Calvert
City
Marshall
None
37.0452,
-88.33087
Industrial
Suburban
None.
LEKY
21-067-0012
Lexington
Fayette
Lexington-Fayette,
KY
38.06503,
-84.49761
Residential
Suburban
SO2, NO, NO2, O3, Meteorological parameters,
PM10, PM25, PM2.5 Speciation, IMPROVE
Speciation.
1 Data for additional pollutants are reported to AQS for these sites (EPA, 2014b); however, these data are not generated by ERG and are therefore not included in this report.
BOLD ITALICS = EPA-designated NATTS Site

-------
There are two Kentucky monitoring sites in the town of Ashland. Ashland is located on
the Ohio River, just north of where the borders of Kentucky, West Virginia, and Ohio meet, and
is part of the Huntington-Ashland, WV-KY-OH CBSA. The ASKY site is located behind the
county health department, which is nestled in a residential area in the center of town, as shown in
Figure 14-1. The ASKY-M site is located on the roof of an oil company complex in the north-
central part of Ashland, which is more industrial. The monitoring site is located less than one-
quarter mile from the Ohio River, and a rail yard, a scrap yard, and other industries are located
between ASKY-M and the river, as shown in Figure 14-2. The ASKY-M monitoring site is
located on Greenup Road (Route 60/23), a major thoroughfare through downtown Ashland.
ASKY and ASKY-M are approximately 1.25 miles apart, as shown in Figure 14-3. Most
of the emissions sources near these sites are located along the Ohio River and its tributary to the
south, the Big Sandy River. These emissions sources 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, to name a few. A cluster of emissions sources is located very close to ASKY-M,
within a half-mile, such that the symbol for the site hides the symbols for the facilities. This
cluster includes a testing laboratory, a miscellaneous commercial/industrial facility, a
mine/quarry, and a heliport at a hospital. There are no emissions sources within a half-mile of
ASKY. The closest sources to ASKY are the same ones under the symbol for ASKY-M,
although a metals processing/fabrication facility and coke battery are located a little farther to the
east of ASKY.
Grayson Lake is located in northeast Kentucky, south of the town of Grayson, and
southwest of the Huntington-Ashland, WV-KY-OH CBSA. The Little Sandy River feeds into
Grayson Lake, which is a U.S. Army Corps of Engineers-managed project, and part of the
Kentucky State Parks system. The lake is narrow and winding, with sandstone cliffs rising to up
to 200 feet above the lake surface (KY, 2015; ACE, 2015). The closest road to the monitoring
site is a service road feeding into Camp Grayson, as shown in Figure 14-4. This site serves as the
Grayson Lake NATTS site. Figure 14-5 shows that few point sources surround GLKY and that
most of them are on the outer periphery of the 10-mile boundary around GLKY. This is not
surprising given the rural nature of the area and that Grayson Lake is located roughly in the
center of the 10-mile radius in Figure 14-5. Sources within 10 miles of GLKY are involved in
14-18

-------
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 14-6, and there are no emissions sources within a
few miles of BAKY, as shown in Figure 14-7. 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 pseudo-dividing line. The city is home to some 17 industrial plants, including
metal, steel, and chemical plants (Calvert City, 2015).
The ATKY monitoring site is located off Main Street (State Road 95), just south of the
entrance to a chemical manufacturing plant. The majority of the city's industry lies north and
east of ATKY. Approximately 1 mile east-southeast down Gilbertsville Highway is the LAKY
monitoring site. LAKY is located behind a mobile home park. Although located in a residential
area, industrial areas are located to the west, northwest, and north. Just over one-half mile north
of LAKY is the TVKY monitoring site. This monitoring site is located at a power substation just
south of another chemical manufacturing plant. The fourth monitoring site in Calvert City is
located at Calvert City Elementary School. The CCKY site is located behind the school, which
backs up to a forested area just south of the aforementioned railroad and to the south of most of
the industry. The BLKY site is located across the Tennessee River, north of Calvert City, in
Smithland. The site is located on a residential property in an agricultural area. This site is
potentially downwind of the Calvert City industrial area. The composite satellite images for these
sites are provided in alphabetical order by site in Figures 14-8 through 14-12.
14-19

-------
Figure 14-13 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, LAKY, and the Tennessee River. Many of the emissions sources closest
to the Calvert City sites are in the chemical manufacturing source category. There are also
several plastic, resin, or rubber product plants located between these sites. Industries located
farther away from the sites but within 10 miles include ship/boat manufacturing or repair; mine,
quarry, or mineral processing; a steel mill; metals processing/fabrication, and an asphalt
production/hot mix asphalt plant.
The LEKY monitoring site is located in the city of Lexington in north-central Kentucky.
The site is located on the property of the county health department in a primarily residential area
of northern Lexington. A YMCA is located adjacent to the health department along W. Loudon
Avenue and a community college is located immediately to the south. The mental health facility
formerly located on the property has been demolished after relocating. Although the area is
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 14-14. LEKY is located just over a half-mile south of New Circle Road
(4/421), a loop encircling the city of Lexington.
Figure 14-15 shows that most of the emissions sources within 10 miles of LEKY are
within a few miles of the site. Emissions sources within 1 mile of LEKY include a food
processing plant, the aforementioned electrical equipment manufacturing plant, a crematory, a
metals processing and fabrication facility, and an automobile/truck manufacturing facility.
Table 14-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Kentucky monitoring sites. Table 14-2 includes both county-level
population and vehicle registration information. Table 14-2 also contains traffic volume
information for each site as well as the location for which the traffic volume was obtained.
Additionally, Table 14-2 presents the county-level daily VMT for Boyd, Carter, Henderson,
Marshall, Livingston, and Fayette Counties.
14-20

-------
Table 14-2. Population, Motor Vehicle, and Traffic Information for the Kentucky
Monitoring Sites
Site
County
Estimated
County
Population1
County-level
Vehicle
Registration2
Annual
Average
Daily Traffic3
Intersection Used for
Traffic Data
County-level
Daily VMT4
ASKY
Boyd
48,886
39,196
7,230
29th St between Newman St
and Lynwood Ave
1,256,000
ASKY-M
12,842
Greenup (23rd) between
16th St and 17th St
GLKY
Carter
27,202
25,487
303
Rd 1496, S of Camp
Webb Rd
1,076,000
BAKY
Henderson
46,347
38,811
922
Rte 1078 N of Hwy 60
1,366,000
ATKY
Marshall
31,107
30,254
3,262
Main St (Rte 95), S of
Johnson Riley Rd
1,241,000
CCKY
4,050
Industrial Pkwy, S of E 5th
Ave
LAKY
1,189
Rte 282 (Gilbertsville Hwy),
E of Industrial Ln
TVKY
2,230
Industrial Pkwy (Rte 1523),
E of Plant Cut-off Rd
BLKY
Livingston
9,359
8,338
2,510
Rte 93/453
391,000
LEKY
Fayette
308,428
208,983
10,083
W Loudon Ave, E of Newton
Pike
7,490,000
bounty-level population estimates reflect 2013 data (Census Bureau, 2014)
2County-level vehicle registrations reflect 2013 data (KYTC, 2014a)
3AADT reflects 2011 data for ASKY & TVKY; 2012 data for ASKY-M, GLKY, LEKY, BAKY, ATKY, and LAKY; and
2013 data for BLKY and CCKY (KYTC, 2014b)
4County-level VMT reflects 2013 data (KYTC, 2014c)
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 14-2 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 31st 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) fifth, and Boyd County
(ASKY/ASKY-M) sixth.
•	All of the Kentucky counties with NMP sites rank among the bottom third for county-
level vehicle ownership, with Fayette County ranking 34th and the remaining five
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.
14-21

-------
• 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, while the other five Kentucky counties
account for five of the six lowest county-level VMTs. Livingston County (BLKY)
has the lowest county-level VMT among NMP sites.
14.2 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.
14.2.1	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, 2015).
14.2.2	Meteorological Summary
Hourly meteorological data for 2013 were retrieved from NCDC for the weather stations
closest to the Kentucky monitoring sites (NCDC, 2013), as described in Section 3.4.2. The
closest weather station to each site is as follows: For ASKY, ASKY-M, and GLKY, Tri-
State/M.J. 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 14-3. These data were used to determine how meteorological conditions on
sample days vary from conditions experienced throughout the year.
14-22

-------
Table 14-3. 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/M.J.
Ferguson Field
Airport
03860
(38.37, -82.56)
8.0
miles
145°
(SE)
Sample
Davs
(66)
63.3
±4.8
54.3
±4.4
43.4
±4.8
49.0
±4.2
69.7
±3.6
1019.1
± 1.6
4.3
±0.5
2013
64.7
± 1.9
55.4
± 1.7
44.7
± 1.9
50.2
± 1.7
70.5
± 1.4
1018.3
±0.6
4.1
±0.2
21st and Greenup, Ashland, Kentucky - ASKY-M
Tri-St/M.J.
Ferguson Field
Airport
03860
(38.37, -82.56)
8.7
miles
152°
(SE)
Sample
Davs
(62)
64.2
±5.0
55.1
±4.6
44.8
±4.9
50.0
±4.3
71.1
±3.4
1018.9
± 1.7
4.3
±0.5
2013
64.7
± 1.9
55.4
± 1.7
44.7
± 1.9
50.2
± 1.7
70.5
± 1.4
1018.3
±0.6
4.1
±0.2
Grayson, Kentucky - GLKY
Tri-St/M.J.
Ferguson Field
Airport
03860
(38.37, -82.56)
25.1
miles
70°
(ENE)
Sample
Days
(66)
63.6
±4.8
54.6
±4.5
44.1
±4.8
49.5
±4.2
70.7
±3.3
1018.9
± 1.6
4.4
±0.5
2013
64.7
± 1.9
55.4
± 1.7
44.7
± 1.9
50.2
± 1.7
70.5
± 1.4
1018.3
±0.6
4.1
±0.2
Baskett, Kentucky - BAKY
Evansville
Regional
Airport
93817
(38.04, -87.52)
12.3
miles
345°
(NNW)
Sample
Days
(62)
63.0
±5.4
55.0
±5.0
44.5
±5.1
49.8
±4.6
69.9
±2.7
1019.2
± 1.9
5.6
±0.6
2013
64.9
±2.0
55.9
± 1.9
45.3
± 1.9
50.6
± 1.8
69.6
± 1.0
1018.5
±0.7
5.3
±0.3
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.

-------
Table 14-3. 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)
23.3
miles
273°
(W)
Sample
Days
(61)
64.7
±5.1
56.5
±4.8
47.0
±5.0
51.7
±4.5
72.6
±2.9
1018.4
± 1.8
5.7
±0.7
2013
66.3
± 1.9
57.1
± 1.8
47.2
± 1.9
52.1
± 1.7
71.8
± 1.1
1018.0
±0.7
5.3
±0.3
Smithland, Kentucky - BLKY
Barkley
Regional
Airport
03816
(37.06, -88.77)
24.4
miles
268°
(W)
Sample
Days
(61)
64.3
±5.0
56.0
±4.8
46.4
±5.1
51.2
±4.5
72.3
±2.8
1018.4
± 1.8
5.6
±0.7
2013
66.3
± 1.9
57.1
± 1.8
47.2
± 1.9
52.1
± 1.7
71.8
± 1.1
1018.0
±0.7
5.3
±0.3
Calvert City Elementary, Calvert City, Kentucky - CCKY
Barkley
Regional
Airport
03816
(37.06, -88.77)
23.9
miles
275°
(W)
Sample
Days
(63)
65.3
±5.0
56.9
±4.7
47.3
±5.0
52.0
±4.5
72.2
±2.7
1018.4
± 1.8
5.6
±0.7
2013
66.3
± 1.9
57.1
± 1.8
47.2
± 1.9
52.1
± 1.7
71.8
± 1.1
1018.0
±0.7
5.3
±0.3
Lazy Daze, Calvert City, Kentucky - LAKY
Barkley
Regional
Airport
03816
(37.06, -88.77)
24.4
miles
273°
(W)
Sample
Days
(63)
64.1
±4.9
55.7
±4.7
46.0
±4.9
50.8
±4.4
72.0
±2.8
1018.5
± 1.8
5.6
±0.7
2013
66.3
± 1.9
57.1
± 1.8
47.2
± 1.9
52.1
± 1.7
71.8
± 1.1
1018.0
±0.7
5.3
±0.3
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.

-------
Table 14-3. 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)
24.5
miles
Sample
Days
(61)
64.6
±5.1
56.3
±4.8
46.7
±5.1
51.5
±4.6
72.3
±2.8
1018.4
± 1.8
5.6
±0.7
272°
(W)
2013
66.3
± 1.9
57.1
± 1.8
47.2
± 1.9
52.1
± 1.7
71.8
± 1.1
1018.0
±0.7
5.3
±0.3
Lexington, Kentucky - LEKY
Blue Grass
Airport
6.1
miles
Sample
Days
(62)
62.9
±5.1
54.7
±4.8
45.2
±4.9
49.9
±4.5
72.5
±3.0
1019.1
± 1.7
7.2
±0.7
93820
(38.04, -84.61)
254°
(WSW)
2013
64.2
± 1.9
55.4
± 1.8
45.6
± 1.9
50.5
± 1.7
71.9
± 1.2
1018.5
±0.6
6.7
±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 2013. Also included in Table 14-3 is the 95 percent
confidence interval for each parameter. Table 14-3 shows that average meteorological conditions
on sample days near the Kentucky monitoring sites were generally representative of average
weather conditions experienced throughout the year. The largest difference between the sample
day average and the average for 2013 was calculated for the average maximum temperature for
each site, the difference for which is largest for LAKY.
It should be noted that even though sample days are generally standardized, the need for
making up invalid samples leads to additional sample days. This is why although the data are
from the same weather station, the sample day averages are often different from each other, such
as the case with ASKY, ASKY-M, and GLKY, for which the closest weather station is Tri-
State/M. J. Ferguson Field Airport.
14.2.3 Wind Rose Comparison
Hourly surface wind data from the weather stations nearest the Kentucky sites, as
presented in Section 14.2.2, were uploaded into a wind rose software program to produce
customized wind roses, as described in Section 3.4.2. A wind rose shows the frequency of wind
directions using "petals" positioned around a 16-point compass, and uses different colors to
represent wind speeds.
Figure 14-16 presents a map showing the distance between the Tri-State/M.J. 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 14-16
also presents three different wind roses for the ASKY monitoring site. First, a historical wind
rose representing 2003 to 2012 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 2013 is presented. Next, a wind rose representing wind data for days
on which samples were collected in 2013 is presented. These can be used to identify the
predominant wind speed and direction for 2013 and to determine if wind observations on sample
days were representative of conditions experienced over the entire year and historically.
14-26

-------
Figures 14-17 through 14-25 present the distance maps and wind roses for the remaining
Kentucky monitoring sites.
Observations from Figures 14-16 through 14-18 for ASKY, ASKY-M, and GLKY,
respectively, include the following:
•	The Tri-State/M.J. Ferguson Field weather station is the closest weather station to
both Ashland sites and GLKY. The weather station is located between 8 miles and
9 miles southeast of the Ashland sites and 25 miles to the east-northeast of GLKY.
This weather station is located 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 more than 40 percent of the wind observations near these sites,
particularly those from south-southwest. Calm winds (those less than or equal to
2 knots) account for roughly one-quarter of the hourly measurements. Winds from the
southeast quadrant were observed the least.
•	The wind patterns on the full-year wind roses are similar to those on the historical
wind roses.
•	The sample day wind rose for ASKY resembles both the historical and full-year wind
roses, although there is a slightly higher percentage of north-northwesterly winds as
well as southwesterly and west-southwesterly winds. Conversely, the calm rate is
slightly less.
•	The sample day wind rose for ASKY-M has similar wind patterns as the sample day
wind rose for ASKY.
•	The sample day wind rose for GLKY has similar wind patterns as the sample day
wind roses for ASKY and ASKY-M, although winds from the southwest to west
account for an even higher percentage of winds on sample days while the percentage
of winds from the south to south-southwest is slightly less.
14-27

-------
Figure 14-16. Wind Roses for the Tri-State/iYl.J. Ferguson Field Airport Weather Station
near ASKY
Location of ASKY and Weather Station	2003-2012 Historical Wind Rose
WIND SPEED
(Knots)
I I =22
n 17-21
H 11 - 17
I ll 7- 11
\^3 4-7
H 2-4
Calms: 25.54%
west:
2013 Wind Rose
VEST
WIND SPEED
i, Knots)
SOUTH
Sample Day Wind Rose
E = ~
WW C SPEEC
(Knots)
SOUTH
14-28

-------
Figure 14-17. Wind Roses for the Tri-State/iYl.J. Ferguson Field Airport Weather Station
near ASKY-M
Location of ASKY-M and Weather Station
2003-2012 Historical Wind Rose
ES.
WW D SPEED
(Kn ots}
SOUTH
2013 Wind Rose
Sample Day Wind Rose
VEST
WIND SPEED
(Knots)
~
I? 1	17-21
HI	11 -17
I 1	7-11
I'D	4-7
H	2-4
Calms: 27.38%
ES~
WWC SPEED
(Knots)
SOUTH
14-29

-------
Figure 14-18. Wind Roses for the Tri-State/iYl.J. Ferguson Field Airport Weather Station
near GLKY
Location of GLKY and Weather Station
2003-2012 Historical Wind Rose
ES.
WW D SPEED
(Kn ots}
SOUTH
2013 Wind Rose
Sample Day Wind Rose
VEST
WIND SPEED
i, Knots)
SOUTH
ES~
WWC SPEED
(Knots)
SOUTH
14-30

-------
west:
Figure 14-19. Wind Roses for the Evansville Regional Airport Weather Station near BAKY
Location of BAKY and Weather Station	2003-2012 Historical Wind Rose
WIND SPEED
(Knots)
~
~1 4-7
2- 4
Calms: 2409%
2013 Wind Rose
WEST
(Knots)
11 - 17
SOUTH
WIND SPEED
~ 4-7
Calms: 2230%
Sample Day Wind Rose
NORTH
west:
(Knots)
SOUTH
WIND SPEED
14-31

-------
Observations from Figure 14-19 for BAKY include the following:
•	The Evansville Regional Airport weather station is located approximately 12 miles
north-northwest of BAKY. This weather station is in Ohio, with most of the city of
Evansville between the site and the station.
•	The historical wind rose shows that winds from a variety of directions were observed
near BAKY, although winds from the south and southwest quadrant were observed
the most and winds from the southeast quadrant were 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-southwest accounting for the highest percentage of
winds greater than 2 knots and calm winds accounting for approximately 22 percent
of the observations. Winds from the south, south-southwest, and northwest were
observed slightly more often in 2013 compared to the historical wind rose.
•	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 and
south-southwesterly winds were still prevalent, the percentage of these winds was
higher on sample days, as were most of the observations from the southwest quadrant.
Northerly winds were also observed more often on sample days. Conversely, wind
observations from the northwest and southeast quadrants were observed less
frequently and calm winds accounted for one-fifth of the observations.
14-32

-------
Figure 14-20. Wind Roses for the Barkley Regional Airport Weather Station near ATKY
Location of ATKY and Weather Station
2003-2012 Historical Wind Rose
Sutton
rrtitet
-w
+
NORTH
WEST
WWC SPEED
i Knots)
SOUTH
Sample Day Wind Rose
NORTH
west:
WEST
WIND SPEED
(Knots)
I I >=22
n 17-21
WIND SPEED
(Knots)
I I >-22
[ I 17-21
14-33

-------
Figure 14-21. Wind Roses for the Barkley Regional Airport Weather Station near BLKY
Location of BLKY and Weather Station
2003-2012 Historical Wind Rose
24.4 m|M«
NORTH
WEST
WWD SPEED
i Knots)
SOUTH
2013 Wind Rose
Sample Day Wind Rose
VEST
'A'INC SPEED
(Knots)
11 -17
SOUTH
Calms: 21.16%
N O RT H ~"
WEST
WIND SPEED
(Knots)
SOUTH
14-34

-------
Figure 14-22. Wind Roses for the Barkley Regional Airport Weather Station near CCKY
Location of CCKY and Weather Station
2003-2012 Historical Wind Rose
23 9 mi|,.8
CCKY
NORTH
WEST
WWC SPEED
i Knots)
SOUTH
Sample Day Wind Rose
NORTH
west:
WEST
WIND SPEED
(Knots)
I I >=22
n 17-21
WIND SPEED
(Knots)
I I >-22
[ I 17-21
14-35

-------
Figure 14-23. Wind Roses for the Barkley Regional Airport Weather Station near LAKY
Location of LAKY and Weather Station
2003-2012 Historical Wind Rose
i.e. •
NORTH
WEST
WWC SPEED
i Knots)
SOUTH
Sample Day Wind Rose
NORTH
west:
WEST
WIND SPEED
(Knots)
I I >=22
n 17-21
WIND SPEED
(Knots)
I I >-22
[ I 17-21
Calms: 19.44%
14-36

-------
Figure 14-24. Wind Roses for the Barkley Regional Airport Weather Station near TVKY
Location of TVKY and Weather Station
2003-2012 Historical Wind Rose
'V
j['i ~
_ K o—	
24.5 miles TVKV
StMion
J V . -.fU.

\ |

S \?

+
NORTH
WEST
WWC SPEED
i Knots)
SOUTH
Sample Day Wind Rose
NORTH
WEST
WEST
WIND SPEED
(Knots)
I I >=22
n 17-21
WIND SPEED
(Knots)
I I >-22
[ I 17-21
14-37

-------
Observations from Figures 14-20 through 14-24 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 23 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 Calvert City 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 roses resemble the historical wind roses, but with a higher
percentage of southerly winds and fewer calm winds (21 percent).
•	The sample day wind roses for the Calvert City sites resemble each other as well as
the full-year and historical wind roses. The sample day wind roses show that
southerly winds were prevalent on sample days near the Calvert City sites, with winds
from the south, southwest quadrant and north accounting for the highest percentage of
wind observations. Calm winds account for 19 percent to 20 percent of the wind
observations on sample days.
14-38

-------
Figure 14-25. Wind Roses for the Blue Grass Airport Weather Station near LEKY
Location of LEKY and Weather Station
2003-2012 Historical Wind Rose
20%
"16%
12%
WIND SPEED
(Knots)
I I >-22
luJ]	17-21
| 11 - 17
[~JH	7- 11
~1 4-7
¦ 2- 4
Calms: 1272%
2013 Wind Rose
Sample Day Wind Rose

NORTH"*--^

X 20%

16%

12%

8%,

;WESTf
IIE7 I EAST |
WIND SPEED
(Knots)
I I >=22
F 1 17-21
11
r i 7-11
l ~~l 4-7
2-4
Calms: 9.87%
NORTH
WEST
WIND SPEED
(Knots)
SOUTH
14-39

-------
Observations from Figure 14-25 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 roughly 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 a higher
percentage of southerly winds were observed while the percentage of calm wind
observations was less than 10 percent.
•	The wind patterns on the sample day wind rose for LEKY resemble the wind patterns
on full-year wind rose, with an even higher percentage of winds from the south to
southwest and an even lower percentage of calm winds (6 percent).
14.3 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 14-4 provides an overview of which analyses were performed at
each site. The site-specific results of the risk-based screening process are presented in
Table 14-5, with the pollutants of interest for each site shaded in gray.
14-40

-------
Table 14-4. Overview of Sampling Performed at the Kentucky Monitoring Sites
Site
VOCs
Carbonyl
Compounds
PAHs
PMio
Metals
Hexavalent
Chromium
ASKY
V
V
—
—
—
ASKY-M
—
—
—
V
—
GLKY



V

BAKY
—
—
—
V
—
ATKY
V
—
—
—
—
BLKY
V
—
—
—
—
CCKY
V
—
—

—
LAKY
V
—
—
—
—
TVKY
V
—
—
—
—
LEKY
V

--

--
-- = This pollutant group was not sampled for at this site.
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 14-4 include the following:
•	Carbonyl compounds, VOCs, PAHs, PMio metals, and hexavalent chromium
were sampled for at GLKY throughout 2013.
•	Additional sites sampling PMio metals include ASKY-M, BAKY, CCKY, and
LEKY.
•	Additional sites sampling VOCs include ASKY, ATKY, BLKY, CCKY, LAKY,
TVKY, and LEKY.
•	Additional sites sampling carbonyl compounds include ASKY and LEKY.
•	No additional sites sampled PAHs or hexavalent chromium.
14-41

-------
Table 14-5. Risk-Based Screening Results for the Kentucky Monitoring Sites
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Health Department, Ashland, Kentucky - ASKY
Acetaldehyde
0.45
61
61
100.00
16.90
16.90
Benzene
0.13
61
61
100.00
16.90
33.80
Carbon Tetrachloride
0.17
61
61
100.00
16.90
50.69
Formaldehyde
0.077
61
61
100.00
16.90
67.59
1,2-Dichloroethane
0.038
53
53
100.00
14.68
82.27
1.3 -Butadiene
0.03
43
47
91.49
11.91
94.18
Hexacliloro -1,3 -butadiene
0.045
8
8
100.00
2.22
96.40
Ethylbenzene
0.4
7
61
11.48
1.94
98.34
/?-Dichlorobcnzcnc
0.091
5
27
18.52
1.39
99.72
Propionaldehyde
0.8
1
60
1.67
0.28
100.00
Total
361
500
72.20

21st and Greenup, Ashland, Kentucky - ASKY-M
Arsenic (PMio)
0.00023
53
59
89.83
50.96
50.96
Nickel (PMio)
0.0021
19
60
31.67
18.27
69.23
Manganese (PMio)
0.03
13
60
21.67
12.50
81.73
Lead (PMio)
0.015
10
60
16.67
9.62
91.35
Cadmium (PMio)
0.00056
9
60
15.00
8.65
100.00
Total
104
299
34.78

Grayson, Kentucky - GLKY
Benzene
0.13
61
61
100.00
15.14
15.14
Carbon Tetrachloride
0.17
61
61
100.00
15.14
30.27
Formaldehyde
0.077
61
61
100.00
15.14
45.41
1,2-Dichloroethane
0.038
56
56
100.00
13.90
59.31
Acetaldehyde
0.45
55
61
90.16
13.65
72.95
1.3 -Butadiene
0.03
45
48
93.75
11.17
84.12
Arsenic
0.00023
44
58
75.86
10.92
95.04
Hexacliloro -1,3 -butadiene
0.045
12
13
92.31
2.98
98.01
Naphthalene
0.029
5
58
8.62
1.24
99.26
Cadmium
0.00056
1
59
1.69
0.25
99.50
Chloroprene
0.0021
1
1
100.00
0.25
99.75
1,2-Dibromoethane
0.0017
1
1
100.00
0.25
100.00
Total
403
538
74.91

Baskett, Kentucky - BAKY
Arsenic (PMio)
0.00023
53
59
89.83
98.15
98.15
Nickel (PMio)
0.0021
1
60
1.67
1.85
100.00
Total
54
119
45.38

14-42

-------
Table 14-5. Risk-Based Screening Results for the Kentucky Monitoring Sites (Continued)
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Atmos Energy, Calvert City, Kentucky - ATKY
Benzene
0.13
61
61
100.00
23.55
23.55
Carbon Tetrachloride
0.17
61
61
100.00
23.55
47.10
1,2-Dichloroethane
0.038
60
60
100.00
23.17
70.27
1.3 -Butadiene
0.03
42
44
95.45
16.22
86.49
Vinyl chloride
0.11
24
38
63.16
9.27
95.75
Hexacliloro -1,3 -butadiene
0.045
8
9
88.89
3.09
98.84
1.1,2-Trichloroethane
0.0625
3
3
100.00
1.16
100.00
Total
259
276
93.84

Smithland, Kentucky - BLKY
Benzene
0.13
59
59
100.00
24.08
24.08
Carbon Tetrachloride
0.17
59
59
100.00
24.08
48.16
1,2-Dichloroethane
0.038
57
57
100.00
23.27
71.43
1.3 -Butadiene
0.03
36
39
92.31
14.69
86.12
Vinyl chloride
0.11
16
35
45.71
6.53
92.65
Hexacliloro -1,3 -butadiene
0.045
7
8
87.50
2.86
95.51
1.1,2-Trichloroethane
0.0625
7
7
100.00
2.86
98.37
Chloroform
9.8
1
55
1.82
0.41
98.78
1,2-Dibromoethane
0.0017
1
1
100.00
0.41
99.18
/j-Dichlorobcnzcne
0.091
1
10
10.00
0.41
99.59
1,1 -Dichloroethane
0.625
1
7
14.29
0.41
100.00
Total
245
337
72.70

Calvert City Elementary School, Calvert City, Kentucky - CCKY
Benzene
0.13
61
61
100.00
21.71
21.71
Carbon Tetrachloride
0.17
61
61
100.00
21.71
43.42
1,2-Dichloroethane
0.038
59
59
100.00
21.00
64.41
Arsenic (PMio)
0.00023
42
55
76.36
14.95
79.36
1,3-Butadiene
0.03
40
45
88.89
14.23
93.59
Hexacliloro -1,3 -butadiene
0.045
10
11
90.91
3.56
97.15
Vinyl chloride
0.11
6
26
23.08
2.14
99.29
Nickel (PMio)
0.0021
1
56
1.79
0.36
99.64
1.1,2-Trichloroethane
0.0625
1
1
100.00
0.36
100.00
Total
281
375
74.93

14-43

-------
Table 14-5. Risk-Based Screening Results for the Kentucky Monitoring Sites (Continued)
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Lazy Daze, Calvert City, Kentucky - LAKY
Benzene
0.13
60
60
100.00
23.35
23.35
Carbon Tetrachloride
0.17
60
60
100.00
23.35
46.69
1,2-Dichloroethane
0.038
60
60
100.00
23.35
70.04
1.3 -Butadiene
0.03
47
48
97.92
18.29
88.33
Vinyl chloride
0.11
14
29
48.28
5.45
93.77
Hexacliloro -1,3 -butadiene
0.045
11
12
91.67
4.28
98.05
1.1,2-Trichloroethane
0.0625
3
5
60.00
1.17
99.22
1,2-Dibromoethane
0.0017
1
1
100.00
0.39
99.61
Ethylbenzene
0.4
1
60
1.67
0.39
100.00
Total
257
335
76.72

TVA Substation, Calvert City, Kentucky - TVKY
Benzene
0.13
61
61
100.00
21.63
21.63
Carbon Tetrachloride
0.17
61
61
100.00
21.63
43.26
1,2-Dichloroethane
0.038
61
61
100.00
21.63
64.89
1.3 -Butadiene
0.03
45
46
97.83
15.96
80.85
Vinyl chloride
0.11
22
36
61.11
7.80
88.65
Hexacliloro -1,3 -butadiene
0.045
11
11
100.00
3.90
92.55
/?-Dichlorobcnzcnc
0.091
7
24
29.17
2.48
95.04
1.1,2-Trichloroethane
0.0625
7
9
77.78
2.48
97.52
1,2-Dibromoethane
0.0017
3
3
100.00
1.06
98.58
T richloroethylene
0.2
2
12
16.67
0.71
99.29
1,1 -Dichloroethane
0.625
1
11
9.09
0.35
99.65
Ethylbenzene
0.4
1
61
1.64
0.35
100.00
Total
282
396
71.21

14-44

-------
Table 14-5. Risk-Based Screening Results for the Kentucky Monitoring Sites (Continued)
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Lexin
gton, Kentucky - LEKY
Acetaldehyde
0.45
61
61
100.00
16.90
16.90
Formaldehyde
0.077
61
61
100.00
16.90
33.80
Arsenic (PMio)
0.00023
46
52
88.46
12.74
46.54
Benzene
0.13
45
45
100.00
12.47
59.00
Carbon Tetrachloride
0.17
45
45
100.00
12.47
71.47
1,2-Dichloroethane
0.038
43
43
100.00
11.91
83.38
1.3 -Butadiene
0.03
40
41
97.56
11.08
94.46
Hexacliloro -1,3 -butadiene
0.045
9
9
100.00
2.49
96.95
Ethylbenzene
0.4
4
45
8.89
1.11
98.06
/?-Dichlorobcnzcnc
0.091
3
28
10.71
0.83
98.89
Nickel (PMio)
0.0021
3
52
5.77
0.83
99.72
Manganese (PMio)
0.03
1
53
1.89
0.28
100.00
Total
361
535
67.48

Observations for the Ashland sites from Table 14-5 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 14-4. VOCs and carbonyl compounds were sampled for at ASKY while only
PMio metals were sampled for at ASKY-M.
•	Ten pollutants failed at least one screen for ASKY, with 72 percent of concentrations
for these 10 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.
•	Five metals failed at least one screen for ASKY-M, with 35 percent of concentrations
for these five pollutants greater than their associated risk screening value (or failed
screens).
•	All five metals contributed to 95 percent of failed screens for ASKY-M and therefore
were identified as pollutants of interest. ASKY-M is one of only two NMP sites with
manganese as a pollutant of interest (TOOK is the other). This is also true for lead
and cadmium (S4MO is the other).
14-45

-------
Observations for GLKY from Table 14-5 include the following:
•	GLKY sampled for all five pollutant groups shown in Table 14-4.
•	Twelve pollutants failed at least one screen for GLKY, with nearly 75 percent of
concentrations for these 12 pollutants greater than their associated risk screening
value (or failed screens).
•	Seven pollutants contributed to 95 percent of failed screens for GLKY and therefore
were identified as pollutants of interest. These include two carbonyl compounds, four
VOCs, and one metal.
Observations for BAKY from Table 14-5 include the following:
•	Like ASKY-M, BAKY sampled for PMio metals only.
•	Arsenic and nickel failed at least one screen for BAKY, with 45 percent of
concentrations for these two pollutants greater than their associated risk screening
value (or failed screens).
•	Arsenic contributed to 98 percent of the failed screens for BAKY and therefore was
identified as BAKY's sole pollutant of interest.
Observations for the Calvert City sites from Table 14-5 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) to 12 (TVKY).
•	Five pollutants contributed to 95 percent of failed screens for ATKY and therefore
were identified as pollutants of interest for this site.
•	Seven pollutants contributed to 95 percent of failed screens for BLKY 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 BLKY, 1,1,2-trichloroethane failed the same number of screens as
hexachloro-1,3-butadiene; thus, 1,1,2-trichloroethane was added as a pollutant of
interest for BLKY, per the procedure described in Section 3.2.
•	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
CCKY include one speciated metal and five VOCs.
•	Six pollutants contributed to 95 percent of failed screens for LAKY and therefore
were identified as pollutants of interest for this site.
14-46

-------
•	Eight pollutants contributed to 95 percent of failed screens for TVKY and therefore
were identified as pollutants of interest for this site. Similar to BLKY,
1,1,2-trichloroethane was added as pollutants of interest for TVKY as this pollutant
failed the same number of screens as />dichlorobenzene, which is the first pollutant to
reach the 95 percent contribution level.
•	Benzene, carbon tetrachloride, 1,2-dichloroethane, and 1,3-butadiene were identified
as pollutants of interest for all five Calvert City sites. Vinyl chloride was identified as
a pollutant of interest for four of the five sites (CCKY was the exception), as was
hexachloro-1,3-butadiene (ATKY was the exception). ATKY, BLKY, LAKY, and
TVKY are the only NMP sites with vinyl chloride as a pollutant of interest.
Observations for LEKY from Table 14-5 include the following:
•	Aside from GLKY, LEKY sampled for the most pollutant groups. Carbonyl
compounds, VOCs, and PMio metals were sampled for at LEKY.
•	Twelve pollutants failed at least one screen for LEKY, with 67 percent of
concentrations for these 12 pollutants greater than their associated risk screening
value (or failed screens).
•	Eight pollutants contributed to 95 percent of failed screens for LEKY and therefore
were identified as pollutants of interest. These include two carbonyl compounds, five
VOCs, and one metal.
14.4 Concentrations
This section presents various concentration averages used to characterize air toxics
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.
•	Concentration averages and other statistical metrics are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at the Kentucky monitoring sites are provided in Appendices J, L, M, N, and O.
14-47

-------
14.4.1 2013 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 14-6, where applicable. Note that
concentrations of the PAHs and metals are presented in ng/m3 for ease of viewing. Also note that
if a pollutant was not detected in a given calendar quarter, the quarterly average simply reflects
"0" because only zeros substituted for non-detects were factored into the quarterly average
concentration.
Table 14-6. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
	the Kentucky Monitoring Sites	

# of






Measured
1st
2nd
3rd
4th


Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs. # of
Samples
Average
(Hg/m3)
Average
(Hg/m3)
Average
(Hg/m3)
Average
(Hg/m3)
Average
(Hg/m3)
Health Department, Ashland, Kentucky - ASKY


1.04
1.69
1.04
1.00
1.19
Acetaldehyde
61/61
±0.19
±0.43
±0.14
±0.17
±0.14


0.86
0.71
1.03
3.36
1.52
Benzene
61/61
±0.12
±0.32
±0.42
±5.50
± 1.39


0.06
0.04
0.05
0.08
0.06
1.3 -Butadiene
47/61
±0.03
±0.02
±0.02
±0.03
±0.01


0.65
0.69
0.66
0.62
0.65
Carbon Tetrachloride
61/61
±0.04
±0.06
±0.04
±0.03
±0.02


0.09
0.09
0.04
0.08
0.08
1,2-Dichloroethane
53/61
±0.01
±0.02
±0.02
±0.01
±0.01


1.42
3.57
2.98
1.29
2.30
Formaldehyde
61/61
±0.30
± 1.04
±0.49
±0.22
±0.38
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.
14-48

-------
Table 14-6. 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
frig/m3)
2nd
Quarter
Average
frig/m3)
3rd
Quarter
Average
frig/m3)
4th
Quarter
Average
frig/m3)
Annual
Average
frig/m3)
21st and Greenup, Ashland, Kentucky - ASKY-M
Arsenic (PMi0)a
59/60
1.08
±0.65
1.46
±0.46
1.39
±0.48
1.04
±0.77
1.24
±0.29
Cadmium (PMi0)a
60/60
0.60
±0.72
0.28
±0.10
0.37
±0.23
0.36
±0.30
0.40
±0.19
Lead (PMi0)a
60/60
8.21
±5.82
9.37
±3.31
9.24
±3.18
6.32
±3.38
8.25
± 1.89
Manganese (PMi0)a
60/60
18.53
±9.94
26.42
±9.60
21.51
±6.17
13.32
±6.96
19.86
±4.06
Nickel (PMi,:,)a
60/60
2.61
±3.02
2.29
±0.76
2.45
±0.85
2.27
±2.15
2.40
±0.89
Grayson, Kentucky - GLKY
Acetaldehyde
61/61
0.62
±0.13
0.94
±0.15
0.56
±0.06
0.62
±0.09
0.68
±0.06
Benzene
61/61
0.70
±0.16
0.33
±0.05
0.55
±0.35
0.38
±0.06
0.49
±0.10
1.3 -Butadiene
48/61
0.06
±0.01
0.03
±0.02
0.03
±0.01
0.04
±0.01
0.04
±0.01
Carbon Tetrachloride
61/61
0.65
±0.05
0.72
±0.07
0.70
±0.04
0.59
±0.05
0.67
±0.03
1,2-Dichloroethane
56/61
0.09
±0.01
0.10
±0.01
0.05
±0.02
0.06
±0.01
0.08
±0.01
Formaldehyde
61/61
0.80
±0.18
2.08
±0.44
1.67
±0.26
0.73
±0.15
1.31
±0.20
Arsenic (PMi0)a
58/59
0.50
±0.30
0.59
±0.16
0.57
±0.17
0.30
±0.13
0.48
±0.10
Baskett, Kentucky - BAKY
Arsenic (PMi0)a
59/60
0.68
±0.22
0.89
±0.30
1.24
±0.80
0.48
±0.18
0.82
±0.22
Nickel (PMi,:,)a
60/60
0.47
±0.19
0.74
±0.29
0.74
±0.51
0.48
±0.18
0.61
±0.15
Atmos Energy, Calvert City, Kcntuck;
- ATKY
Benzene
61/61
0.75
±0.10
0.42
±0.09
0.48
±0.09
0.49
±0.08
0.54
±0.05
1.3 -Butadiene
44/61
0.06
±0.04
0.08
±0.09
0.05
±0.02
0.07
±0.03
0.07
±0.02
Carbon Tetrachloride
61/61
0.65
±0.02
0.72
±0.06
0.69
±0.03
0.63
±0.04
0.67
±0.02
1,2-Dichloroethane
60/61
0.33
±0.25
0.28
±0.20
0.36
±0.22
0.22
±0.13
0.30
±0.10
Vinyl chloride
38/61
0.99
±0.79
0.48
±0.54
0.50
±0.58
0.42
±0.62
0.60
±0.31
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.
14-49

-------
Table 14-6. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Kentucky Monitoring Sites (Continued)

# of






Measured
1st
2nd
3rd
4th


Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs. # of
Samples
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Smithland, Kentucky - BLKY


0.63
0.78
0.66
0.63
0.67
Benzene
59/59
±0.13
±0.31
±0.35
±0.39
±0.15


0.03
2.05
0.13
0.29
0.63
1.3 -Butadiene
39/59
±0.02
± 1.82
±0.16
±0.32
±0.49


2.21
0.76
0.82
0.65
1.11
Carbon Tetrachloride
59/59
±3.16
±0.07
±0.13
±0.03
±0.77


1.16
2.03
0.95
0.94
1.27
1,2-Dichloroethane
57/59
±2.11
± 1.89
±0.74
± 1.06
±0.75


0.02
0.01
<0.01
0.04
0.02
Hexachloro-1,3 -butadiene
8/59
±0.03
±0.02
±0.01
±0.05
±0.01


0.01
0.01
0.05

0.02
1,1,2 -T richloroethane
7/59
±0.01
±0.02
±0.06
0
±0.02


0.10
0.10
0.10
0.43
0.18
Vinyl chloride
35/59
±0.17
±0.06
±0.05
±0.67
±0.16
Calvert City Elementary School, Calvert City, Kentucky - CCKY


0.67
0.37
0.46
0.48
0.50
Benzene
61/61
±0.09
±0.05
±0.09
±0.11
±0.05


0.04
0.06
0.06
0.85
0.26
1.3 -Butadiene
45/61
±0.02
±0.04
±0.03
± 1.58
±0.40


0.64
0.72
0.69
0.64
0.67
Carbon Tetrachloride
61/61
±0.03
±0.05
±0.06
±0.04
±0.02


0.17
0.24
0.30
0.25
0.24
1,2-Dichloroethane
59/61
±0.06
±0.23
±0.22
±0.11
±0.08


0.01

0.02
0.03
0.02
Hexachloro-1,3 -butadiene
11/61
±0.01
0
±0.02
±0.03
±0.01


0.52
0.64
0.76
0.54
0.61
Arsenic (PMi0)a
55/56
±0.18
±0.21
±0.32
±0.41
±0.15

Lazy Daze, Calvert City, Kentucky
LAKY




0.73
0.52
0.53
0.73
0.62
Benzene
60/60
±0.18
±0.16
±0.14
±0.37
±0.11


0.07
0.88
0.05
1.57
0.66
1.3 -Butadiene
48/60
±0.05
± 1.57
±0.02
±2.86
±0.80


0.65
0.70
0.69
0.67
0.68
Carbon Tetrachloride
60/60
±0.04
±0.10
±0.05
±0.04
±0.03


0.23
0.39
0.55
1.62
0.70
1,2-Dichloroethane
60/60
±0.14
±0.31
±0.50
± 1.73
±0.45


0.03
<0.01
0.02
0.03
0.02
Hexachloro-1,3 -butadiene
12/60
±0.03
±0.01
±0.02
±0.03
±0.01


0.09
0.04
0.05
0.09
0.07
Vinyl chloride
29/60
±0.13
±0.04
±0.05
±0.06
±0.04
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.
14-50

-------
Table 14-6. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Kentucky Monitoring Sites (Continued)

# of






Measured
1st
2nd
3rd
4th


Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs. # of
Samples
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)

TVA Substation, Calvert
City, Kentucky - TVKY




0.93
0.65
0.88
1.74
1.06
Benzene
61/61
±0.39
±0.59
±0.56
±0.88
±0.32


0.09
1.72
0.17
2.08
1.03
1.3 -Butadiene
46/61
±0.08
±2.84
±0.13
±2.72
±0.97


0.85
0.77
0.83
0.76
0.80
Carbon Tetrachloride
61/61
±0.23
±0.09
±0.20
±0.11
±0.08


<0.01
0.01
0.08
0.03
0.03
/?-Dichlorobcnzcnc
24/61
±0.01
±0.01
±0.06
±0.02
±0.02


2.69
0.88
1.29
9.73
3.75
1,2-Dichloroethane
61/61
±3.05
±0.88
± 1.86
± 13.97
±3.68


<0.01
0.01
0.02
0.02
0.01
Hexachloro-1,3 -butadiene
11/61
±0.01
±0.02
±0.02
±0.01
±0.01


0.01

0.03
0.16
0.05
1,1,2 -T richloroethane
9/61
±0.02
0
±0.04
±0.27
±0.07


0.54
0.10
0.10
0.41
0.29
Vinyl chloride
36/61
±0.82
±0.11
±0.13
±0.40
±0.22
Lexington, Kentucky - LEKY


0.87
1.51
1.68
1.87
1.49
Acetaldehyde
61/61
±0.12
±0.30
±0.19
±0.30
±0.15




0.50
0.81

Benzene
45/45
NA
NA
±0.07
±0.16
NA




0.06
0.08

1.3 -Butadiene
41/45
NA
NA
±0.02
±0.02
NA




0.66
0.63

Carbon Tetrachloride
45/45
NA
NA
±0.04
±0.06
NA




0.05
0.09

1,2-Dichloroethane
43/45
NA
NA
±0.01
±0.01
NA


1.59
3.94
4.65
1.56
2.91
Formaldehyde
61/61
±0.30
± 1.05
±0.77
±0.52
±0.49




<0.01
0.04

Hexachloro-1,3 -butadiene
9/45
NA
NA
±0.01
±0.03
NA


0.52
0.75
0.77
0.67
0.68
Arsenic (PMi0)a
52/53
±0.19
±0.15
±0.25
±0.33
±0.12
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
Observations for the Ashland sites from Table 14-6 include the following:
•	VOCs and carbonyl compounds were sampled for at ASKY and PMio metals were
sampled for at ASKY-M. Thus, these sites have no pollutants of interest in common.
•	With the exception of 1,3-butadiene and 1,2-dichloroethane, each of the pollutants of
interest for ASKY was detected in all the valid VOC samples collected.
14-51

-------
The pollutants of interest with the highest annual average concentrations for ASKY
are formaldehyde (2.30 ± 0.38 |ig/m3), benzene (1.52 ± 1.39 |ig/m3), and
acetaldehyde (1.19 ± 0.14 |ig/m3). Note the high confidence interval for the annual
average concentration of benzene.
The second 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 the maximum acetaldehyde concentration was
measured at ASKY on June 21, 2013 (4.08 |ig/m3) and is the only acetaldehyde
concentration greater than 2.5 |ig/m3 measured at this site. The six highest
acetaldehyde concentrations measured at ASKY were all measured between April and
June, with the seventh and eighth highest measured at the end of March.
The second quarter average concentration of formaldehyde is also the highest of the
quarterly averages for ASKY, although the third quarter average is also relatively
high. The maximum formaldehyde concentration was also measured on June 21, 2013
at ASKY (9.05 |ig/m3). All but three of the 29 formaldehyde concentrations greater
than 2 |ig/m3 were measured at ASKY during the second and third quarters of 2013
(with two of the three measured at the end of March, and the third in October).
Conversely, the 14 lowest formaldehyde concentrations were all measured during the
first or fourth quarters of 2013.
The fourth quarter average benzene concentration is more than three times greater
than the other quarterly averages and has a confidence interval greater than the
average itself. This indicates that outliers may be affecting this quarterly average. A
review of the data shows that the maximum benzene concentration was measured at
ASKY on November 6, 2013 (43.5 |ig/m3). This measurement is nearly 14 times
greater than the next highest benzene concentrations measured at this site
(3.19 |ig/m3) and nearly five times greater than the next highest benzene
concentration measured across the program (9.38 |ig/m3 measured at OCOK). All
other benzene concentrations measured at ASKY during the fourth quarter are less
than 1.5 |ig/m3, explaining the large confidence interval associated with this quarterly
average.
Table 4-9 presents the NMP sites with the 10 highest annual average concentrations
for each of the program-level VOC pollutants of interest. This table shows that ASKY
has the fourth highest annual average concentration of benzene calculated across the
program. However, this site has the largest confidence interval among the sites shown
(at least four or more times greater), indicating that this annual average is influenced
by outliers while the other annual averages likely run higher on a more consistent
basis. Excluding the maximum concentration from the calculation would result in an
annual average concentration for ASKY nearly half as high and in the middle of the
site-specific annual average concentrations of benzene.
With the exception of arsenic, the metal pollutants of interest were detected in all of
the valid samples collected at ASKY-M. Arsenic was detected in all but one of
samples collected.
14-52

-------
The pollutant of interest with the highest annual average concentration for ASKY-M
is manganese (19.86 ± 4.06 ng/m3), followed by lead (8.25 ± 1.89 ng/m3), and nickel
(2.40 ± 0.89 ng/m3).
Many of the quarterly average concentrations for the pollutants of interest for
ASKY-M have relatively large confidence intervals, particularly for the first quarter
of 2013, indicating the measurements collected at ASKY-M are highly variable. For
instance, concentrations of nickel span two orders of magnitude, ranging from
0.200 ng/m3 to 21.2 ng/m3, with a median concentration of 1.46 ng/m3, nearly
1 ng/m3 less than the annual average concentration. Both the maximum and minimum
nickel concentrations measured at ASKY-M were measured during the first quarter of
2013. The maximum concentration of nickel measured at ASKY-M is the maximum
nickel concentration measured across the program. This explains why the confidence
interval for the first quarter is greater than the average itself. The second highest
nickel concentration across the program was also measured at ASKY-M (17.1 ng/m3)
and was measured on November 30, 2013. The second highest and second lowest
nickel concentrations were measured at ASKY-M during the fourth quarter of 2013,
which explains the relatively large confidence interval shown in Table 14-6 for this
pollutant.
Concentrations of manganese measured at ASKY-M range from 1.46 ng/m3 to
56.0 ng/m3, with a median concentration of 15.59 ng/m3. Five measurements greater
than 50 ng/m3 were measured at ASKY-M in 2013, the most of any other NMP site
sampling PMio metals. Three of these five were measured on back-to-back sample
days in April 2013. At least four manganese measurements greater than 25 ng/m3
were measured at ASKY-M in each quarter of 2013 while at least one manganese
concentration less than 5 ng/m3 was also measured in each quarter. This explains the
relatively large confidence intervals shown for each quarterly average of manganese
in Table 14-6.
Concentrations of lead measured at ASKY-M range from 0.94 ng/m3 to 40.5 ng/m3,
with a median concentration of 5.53 ng/m3. The first quarter average concentration,
while not the highest quarterly average, has the largest confidence interval associated
with it. The maximum concentration of lead was measured at ASKY-M on
March 23, 2013 and is nearly three times greater than the next highest lead
concentration measured during the first quarter of 2013 at this site. The maximum
concentration of lead measured at ASKY-M is the second highest lead measurement
across the program. ASKY-M has the second highest number of lead measurements
greater than 15 ng/m3 (10), second only to S4MO.
The first quarter average concentration of cadmium is greater than the other quarterly
averages and the associated confidence interval is greater than the average itself.
Although this indicates the likely presence of outliers factoring into the average
concentration, the confidence intervals for each of the quarterly averages are
relatively large. Concentrations of cadmium measured at ASKY-M range from
0.040 ng/m3 to 5.05 ng/m3, with a median concentration of 0.19 ng/m3. The maximum
cadmium concentration was measured at ASKY-M on the same day the maximum
lead concentration was measured and is the second highest cadmium concentration
14-53

-------
measured across the program. Both the maximum and minimum cadmium
concentrations were measured at ASKY-M during the first quarter of 2013. Three
cadmium concentrations greater than 1 ng/m3 were measured at ASKY-M and are
spread across the calendar quarters, with the exception of the second quarter.
•	Concentrations of arsenic measured at ASKY-M span three orders of magnitude,
ranging from 0.003 ng/m3 to 5.97 ng/m3 plus one non-detect, with a median
concentration of 0.94 ng/m3. The maximum arsenic concentration was measured at
ASKY-M on the same day as the second highest manganese concentration was
measured at this site. Although the first and fourth quarter average concentrations are
less than the other two quarterly averages, they have larger confidence intervals,
indicating more variability in their measurements. While the two highest arsenic
concentrations were measured during the fourth (5.97 ng/m3 on November 30, 2013)
and first (4.40 ng/m3 on March 29, 013) quarters of 2013 at ASKY-M, all eight
arsenic concentrations less than 0.30 ng/m3 were also measured during the first or
fourth quarters of 2013.
•	Table 4-12 presents the NMP sites with the 10 highest annual average concentrations
for each of the program-level metal pollutants of interest. This table shows that the
highest annual averages for arsenic and nickel calculated across the program were
calculated for ASKY-M.
Observations for GLKY from Table 14-6 include the following:
•	GLKY sampled VOCs, carbonyl compounds, metals, PAHs, and hexavalent
chromium. However, most of the pollutants of interest identified for GLKY are
VOCs.
•	The only pollutant of interest with an annual average concentration greater than
1 |ig/m3 is formaldehyde (1.31 ± 0.20 |ig/m3). However, this is one of the lowest
annual averages of formaldehyde calculated among NMP sites sampling carbonyl
compounds.
•	Concentrations of formaldehyde were considerably higher during the warmer months
of the year, based on the quarterly averages. The 18 highest measurements (those
greater than 1.50 |ig/m3) were measured during the second or third quarters of 2013.
Conversely, all 27 concentrations less than 1.0 |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. The second
quarter is the quarter during which the greatest number of concentrations greater than
1 |ig/m3 were measured (six), with one each measured during the first and fourth
quarters. A review of the median concentration for each quarter shows that the
median concentrations follow as similar pattern as the quarterly averages, ranging
from 0.55 |ig/m3 to 0.61 |ig/m3 for the first, third, and fourth quarters and 0.79 |ig/m3
for the second quarter.
14-54

-------
•	Although the first quarter average concentration of benzene is the highest of the
quarterly averages, the third quarter average has the highest associated confidence
interval. A review of the data shows that the maximum benzene concentration was
measured on August 8, 2013 (2.75 |ig/m3). A second benzene concentration of
1.29 |ig/m3 was measured in September. All other benzene concentrations measured
during the third quarter of 2013 are less than 0.5 |ig/m3. Of the 16 benzene
measurements greater than 0.5 |ig/m3 measured at GLKY, 11 were measured during
the first quarter, one was measured during the second quarter, and two each were
measured in the third and fourth quarters.
•	Arsenic is the only other pollutant of interest for GLKY that is not a VOC or carbonyl
compound. Concentrations of arsenic measured at GLKY range from 0.005 ng/m3 to
2.07 ng/m3, plus a single non-detect. Both the minimum and maximum concentrations
of arsenic were measured at GLKY during the first quarter of 2013, explaining the
relatively large confidence interval associated with this quarterly average. Fifteen of
the 16 lowest concentrations measured at GLKY, including the non-detect, were
measured in during the first or fourth quarters of 2013.
Observations for BAKY from Table 14-6 include the following:
•	Only speciated metals were sampled for at BAKY; only two pollutants of interest
were identified for BAKY: arsenic and nickel.
•	Nickel was measured in all 60 valid metals samples collected at BAKY, while a
single non-detect of arsenic was measured.
•	Concentrations of arsenic tended to be higher than the concentrations of nickel
measured at this site, as evident from the annual averages and quarterly averages
shown in Table 14-6 (although the fourth quarter averages of these metals are equal to
each other). For both metals, concentrations appear higher during the warmer months
of the year, although the differences are not statistically significant.
•	Arsenic concentrations measured at BAKY range from 0.01 ng/m3 to 6.37 ng/m3, plus
the one non-detect, with a median concentration of 0.66 ng/m3. The maximum arsenic
concentration was measured on September 7, 2013 and is more than three times
greater than the next highest arsenic measurement collected at BAKY. The maximum
and non-detect concentrations were both measured at BAKY during the third quarter
of 2013. This explains the relatively high third quarter average concentration and
associated confidence interval.
•	Nickel concentrations measured at BAKY range from 0.11 ng/m3 to 4.14 ng/m3, with
a median concentration of 0.46 ng/m3. The maximum nickel concentration was
measured on July 21, 2013 and is more than twice the next highest nickel
measurement collected at BAKY (1.96 ng/m3 measured on May 28, 2013). At least
one nickel concentration greater than 1 ng/m3 was measured during each calendar
quarter.
14-55

-------
•	As shown in Table 14-6, the first and fourth quarter nickel averages are similar to
each other and have similar confidence intervals. The second and third quarter
averages are similar to each other but their confidence intervals are considerably
different. If median concentrations are calculated for each quarter's data, the median
concentrations for the first and fourth quarters are both just less than 0.40 ng/m3, but
the median for the second quarter (0.61 ng/m3) is considerably higher than the median
for the third quarter (0.49 ng/m3). This indicates the nickel concentrations ran higher
overall during the second quarter while the maximum concentration measured during
the third quarter is influencing the quarterly average nickel concentration for the third
quarter.
•	Among NMP sites sampling PMio metals, BAKY has the third highest annual
average concentration of arsenic and the 10th highest annual average concentration of
nickel, as shown in Table 4-12.
Observations for the Calvert City monitoring sites from Table 14-6 include the following:
•	With the exception of CCKY, only VOC samples were collected at the Calvert City
sites; PMio metals were sampled for at CCKY in addition to VOCs.
•	Some of the highest concentrations of VOCs were measured at the Calvert City sites
and these data are reviewed in the bullets that follow.
•	Vinyl chloride is an infrequently detected pollutant under the NMP in typical urban
atmospheres. Across the program, this pollutant was detected in less than 15 percent
of the total samples collected. Together, the five Calvert City sites account for more
than 67 percent of the 243 measured detections of this pollutant. The Calvert City
sites account for all 41 concentrations of vinyl chloride greater than 0.30 |ig/m3
measured across the program, including the 15 measurements greater than 1 |ig/m3.
The maximum concentration of vinyl chloride across the program was measured at
TVKY (6.07 |ig/m3), although additional measurements greater than 4 |ig/m3 were
measured at ATKY and BLKY.
•	Vinyl chloride is a pollutant of interest for four of the five Calvert City sites (with
CCKY as the exception). As shown in Table 14-6, annual average concentrations for
these sites range from 0.07 ± 0.04 |ig/m3 for LAKY to 0.60 ± 0.31 |ig/m3 for TVKY.
All of the annual average and quarterly average concentrations of vinyl chloride for
these sites have relatively large confidence intervals, indicating the relatively large
amount of variability associated with these measurements, including substitutions for
non-detects.
•	Another pollutant for which the highest concentrations program-wide were measured
at the Calvert City sites is 1,2-dichloroethane. The 77 highest concentrations of
1,2-dichloroethane across the program were all measured at the Calvert City sites.
This includes all 49 measurements greater than 1 |ig/m3, the eight greater than
10 |ig/m3 and the one greater than 100 |ig/m3 (111 |ig/m3 measured at TVKY on
November 18, 2013).
14-56

-------
1,2-Dichloroethane is a pollutant of interest for all five Calvert City sites. Annual
average concentrations for these sites range from 0.24 ± 0.08 |ig/m3 for CCKY to
3.75 ± 3.68 |ig/m3 for TVKY. With the exception of CCKY and perhaps ATKY, the
annual average and quarterly average concentrations of 1,2-dichloroethane for these
sites have relatively large confidence intervals, indicating the relatively large amount
of variability associated with these measurements.
Some of the highest measurements of carbon tetrachloride were also measured at the
Calvert City sites, particularly TVKY. Of the 14 carbon tetrachloride concentrations
greater than or equal to 1 |ig/m3 measured across the program, 12 were measured at
the Calvert City sites (eight at TVKY, one at LAKY, and three at BLKY). The
maximum carbon tetrachloride concentration measured at BLKY on January 28, 2013
(23.7 |ig/m3) is an order of magnitude greater than the next highest carbon
tetrachloride measurement across the program (2.33 |ig/m3 measured at TVKY on
January 16, 2013) and the maximum carbon tetrachloride concentration measured at
an NMP site since 2001.
Carbon tetrachloride is a pollutant of interest for all five Calvert City sites. Annual
average concentrations for ATKY, CCKY, and LAKY were similar to each other,
ranging from 0.67 ± 0.02 |ig/m3 for ATKY and CCKY to 0.68 ± 0.03 |ig/m3 for
LAKY while the annual averages were greater for TVKY (0.80 ± 0.08 |ig/m3) and
BLKY (1.11 ± 0.77 |ig/m3). Quarterly averages for ATKY, CCKY, and LAKY span
less than 0.1 |ig/m3, ranging from 0.63 |ig/m3 to 0.72 |ig/m3, which are fairly typical
for this pollutant among NMP sites. A review of the quarterly average concentrations
for BLKY provides insight into when the higher concentrations of carbon
tetrachloride were measured. The first quarter average concentration for BLKY
(2.21 ±3.16 |ig/m3) is more than twice the other quarterly averages and the
confidence interval is considerably greater than the average itself, indicating during
which quarter the maximum concentration was measured. The third quarter average
concentration for BLKY is also higher than the other quarterly averages and has a
relatively large confidence interval (0.82 ±0.13 |ig/m3) for this pollutant. Although
this average is nothing like the first quarter average, it is greater than quarterly
averages typically calculated for this pollutant, which generally range from
0.55 |ig/m3 to 0.75 |ig/m3, with a central tendency around 0.65 |ig/m3. Even the
second quarter average concentration for BLKY is above this range. For TVKY, all
of the quarterly average concentrations fall outside (and greater than) this range.
All 19 1,3-butadiene concentrations greater than 1 |ig/m3 measured across the
program were measured at the Calvert City sites, including three greater than
20 |ig/m3. Three of the four highest 1,3-butadiene concentrations across the program
were measured on the same date at TVKY, LAKY, and CCKY (October 13, 2013)
and ranged from 12.4 |ig/m3 (CCKY) to 21.5 |ig/m3 (TVKY) while the measurement
at ATKY was considerably less (0.05 |ig/m3). No sample was collected at BLKY on
this day. The COCs for this particular day indicate that on-going construction
activities were occurring near TVKY, although this is noted on all samples collected
between August and December.
14-57

-------
1,3-Butadiene is a pollutant of interest for all five Calvert City sites. The annual
average concentration of 1,3-butadiene for ATKY exhibits the least variability
(0.07 ± 0.02 |ig/m3) compared to the other Calvert City sites, which range from
0.26 ± 0.40 |ig/m3 for CCKY to 1.03 ± 0.97 |ig/m3 for TVKY. However, each of
these five sites has at least one quarterly average concentration of 1,3-butadiene
where the confidence interval is larger than the average itself. For ATKY, it's the
second quarter average (0.08 ± 0.09 |ig/m3), where 1,3-butadiene measurements
range from 0.022 |ig/m3 to 0.652 |ig/m3 plus six non-detects. For other sites, the
difference among the quarterly averages is larger and the calendar quarter during
which an outlier was measured easily identified. For example, the quarterly average
concentrations of 1,3-butadiene for CCKY range from 0.04 ± 0.02 |ig/m3 (first
quarter) to 0.85 ± 1.58 |ig/m3 (fourth quarter). Each of the Calvert City sites has at
least one quarterly average concentration like CCKY's fourth quarter average
concentration, except ATKY.
Hexachloro-1,3-butadiene is another infrequently detected pollutant that is a pollutant
of interest for four of the five Calvert City sites (ATKY is the exception). The
maximum hexachloro-l,3-butadiene concentration measured across the program was
measured at BLKY (0.29 |ig/m3), along with the fourth and eighth highest
concentrations of this pollutant. However, this pollutant was detected in no more than
20 percent of valid VOC sampled collected at each site; thus, zeros substituted for
non-detects make up the majority of the measurements incorporated into the quarterly
and annual averages shown in Table 14-6. As a result, the annual averages are not
significantly different across the sites, ranging from 0.011 ± 0.007 |ig/m3 for ATKY
to 0.019 ± 0.010 |ig/m3 for LAKY.
1,1,2-Trichloroethane is a pollutant of interest for two of the Calvert City sites,
BLKY and TVKY. Together, these sites account for more than half (16) of the 29
measured detections of this pollutant across the program. Along with measurements
from ATKY, CCKY, and LAKY, the Calvert City sites account for all but four of the
29 measured detections of 1,1,2-trichloroethane measured across the program. The
program-level maximum concentration of this pollutant (2.15 |ig/m3) was measured at
TVKY on November 18, 2013. No other concentration of 1,1,2-trichloroethane
greater than 0.4 |ig/m3 was measured at an NMP site in 2013.
Benzene is the only other VOC that is a pollutant of interest across the Calvert City
sties. Annual average concentrations of benzene range from 0.50 ± 0.05 |ig/m3 for
CCKY to 1.06 ± 0.32 |ig/m3 for TVKY. Benzene concentrations measured at TVKY
exhibit the most variability, ranging from 0.20 |ig/m3 to 6.40 |ig/m3. The maximum
benzene concentration was measured on October 13, 2013, the same day that the
highest 1,3-butadiene concentration was measured at this site. This is also the fourth
highest benzene concentration measured across the program.
Table 4-9 presents the NMP sites with the 10 highest annual average concentrations
for each of the program-level VOC pollutants of interest. This table shows that
TVKY has the ninth highest annual average benzene concentration among sites
sampling this pollutant. Calvert City sites account for four of the five highest annual
average concentrations of 1,3-butadiene across the program (with ATKY as the
14-58

-------
exception). All five Calvert City sites rank among the sites with the highest annual
average concentrations of carbon tetrachloride across the program, ranging from
highest (BLKY) to seventh highest (ATKY). Calvert City sites account for the five
highest annual average concentrations of 1,2-dichloroethane across the program.
LAKY and BLKY rank eighth and ninth highest, respectively, among NMP sites for
their annual average concentrations of hexachloro-1,3-butadiene.
•	Arsenic is the only non-VOC pollutant of interest for CCKY. Concentrations of
arsenic measured at CCKY range from 0.08 ng/m3 to 3.38 ng/m3, plus a single non-
detect, with a median concentration of 0.49 ng/m3. The maximum concentration
measured at CCKY is the ninth highest arsenic (PMio) concentration measured across
the program. This site has the ninth highest annual average concentration of arsenic
among NMP sites sampling PMio metals, as shown in Table 4-12.
Observations for LEKY from Table 14-6 include the following:
•	Although VOCs were sampled for at LEKY year-round, a leak in the sample line was
discovered at the site and resulted in the invalidation of VOC samples between
February 9, 2013 and May 4, 2013; thus, first and second quarterly averages and
annual averages for the VOC pollutants of interest could not be calculated. However,
Appendix J provides the pollutant-specific average concentrations for all valid VOC
samples collected at LEKY for the entire year.
•	In most cases, the fourth quarter average concentrations were greater than the third
quarter averages of the VOC pollutants of interest.
•	The annual average concentration for formaldehyde is roughly twice the annual
average concentration of acetaldehyde, the two carbonyl compound pollutants of
interest for LEKY.
•	The second and third quarter average concentrations of formaldehyde are
significantly higher than the first and fourth quarter averages, indicating that
formaldehyde concentrations tended to be higher during the warmer months of the
year at this site. Concentrations of formaldehyde measured at LEKY range from
0.609 |ig/m3 to 8.33 |ig/m3, with the 17 highest measurements (those greater than
4 |ig/m3) measured between May and September and the 19 lowest measurements
(those less than 1.50 |ig/m3) measured between January and March or October and
December.
•	Acetaldehyde concentrations appear lowest during the first quarter of 2013 and
highest during the fourth quarter of 2013, although the differences between the
second, third, and fourth quarterly averages are not statistically significant. Ten of the
12 lowest acetaldehyde concentrations (those less than 1 |ig/m3) were measured
between January and March. Three of the four highest acetaldehyde concentrations
measured at LEKY were measured in October and November with nearly half (5) of
the 11 acetaldehyde concentrations greater than 2 |ig/m3 measured during the fourth
quarter (and none measured during the first quarter and each three in the second and
third quarters).
14-59

-------
•	Concentrations of arsenic measured at LEKY range from 0.08 ng/m3 to 2.53 ng/m3,
including a single non-detect, with a median concentration of 0.67 ng/m3. Among
NMP sites sampling PMio metals, LEKY has the seventh highest annual average
concentration of arsenic, as shown in Table 4-12.
•	It should be noted that during the second half of 2013, demolition of a nearby mental
health hospital was completed.
14.4.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where annual averages are available. Thus, box plots were created for the
pollutants of interest for each of the Kentucky monitoring sites. Figures 14-26 through 14-40
overlay the sites' minimum, annual average, and maximum concentrations onto the program-
level minimum, first quartile, median, average, third quartile, and maximum concentrations, as
described in Section 3.4.3.1. Figures 14-26 through 14-40 and their associated observations are
as follows:
Figure 14-26. Program vs. Site-Specific Average Acetaldehyde Concentrations
¦

k

0	3	6	9	12	15
Concentration {[ig/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range


14-60

-------
• Figure 14-26 is the box plot for acetaldehyde for ASKY, GLKY, and LEKY, the
only Kentucky sites at which carbonyl compounds were sampled. The range of
acetaldehyde concentrations measured was largest for ASKY and smallest for
GLKY, with all of the acetaldehyde concentrations measured at GLKY less than
the program-level median concentration. Among these three sites, GLKY has the
lowest annual average concentration while LEKY has the highest, although each
of the annual average concentrations is less than the program-level average
concentration. LEKY's annual average concentration is similar to the program-
level median concentration; ASKY's annual average is less than the program-
level median but greater than the first quartile; and GLKY's annual average is less
than the program-level first quartile.
Figure 14-27. Program vs. Site-Specific Average Arsenic (PMio) Concentrations
E
it
i±
i±
4	5	6
Concentration {ng/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range


14-61

-------
• Figure 14-27 presents the box plots for the five Kentucky sites for which arsenic
is a pollutant of interest. The box plots show that the range of arsenic
concentrations measured is smallest for GLKY and largest for BAKY (although a
similar range was measured at ASKY-M). The annual average concentrations of
arsenic for ASKY-M and BAKY are greater than the program-level average
concentration; the annual average concentration for LEKY is similar to the
program-level average concentration; the annual average concentration for CCKY
is just less than the program-level average concentration; and the annual average
concentration for GLKY is less than the program-level average concentration but
similar to the program-level median. The maximum arsenic concentration across
the program was not measured at any of the Kentucky sites, even though these
sites account for four of the highest annual average concentrations among NMP
sites sampling PMio metals.
14-62

-------
Figure 14-28. Program vs. Site-Specific Average Benzene Concentrations


¦
	^	


Program Max Concentration = 43.5 ^ig/m3 |


I
Program Max Concentration = 43.5 ^ig/m3

Program Max Concentration = 43.5 ^ig/m3
H
Program Max Concentration = 43.5 ^ig/m3
Wh
Program Max Concentration = 43.5 ^ig/m3
-
Program Max Concentration = 43.5 ^ig/m3
¦+
Program Max Concentration = 43.5 ^ig/m3
4	6
Concentration {[jg/m3]
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


• Figure 14-28 presents the box plots for the seven Kentucky sites for which
benzene is a pollutant of interest. Note that the program4evel maximum
concentration (43.5 |ig/m3) is not shown directly on the box plot because the scale
of the box plot would be too large to readily observe data points at the lower end
of the concentration range. Thus, the scale of the box plots has been reduced to
12 |ig/m3. The box plots show that the maximum benzene concentration measured
across the program was measured at ASKY. All other benzene concentrations
14-63

-------
measured at a Kentucky site fall well within the range of benzene concentrations
shown on the box plots. After ASKY, the range of benzene concentrations is
largest at TVKY and smallest at ATKY (although a similar range was measured
at CCKY). The annual average concentrations of benzene across all the Kentucky
sites range from 0.49 ± 0.10 |ig/m3 (GLKY) to 1.52 ± 1.39 |ig/m3 (ASKY). The
annual average benzene concentrations for ASKY and TVKY are the only ones
greater than the program-level average concentration.
Figure 14-29. Program vs. Site-Specific Average 1,3-Butadiene Concentrations
H+
t
*

¦

I
Program Max Concentration = 21.5 ^ig/m3
Program Max Concentration = 21.5 ^ig/m3
Program Max Concentration = 21.5 ^ig/m3
BLKY Max Concentration = 10.1 ^ig/m3
Program Max Concentration = 21.5 ^ig/m3
CCKY Max Concentration = 12.4 ^ig/m3
Program Max Concentration = 21.5 ^ig/m3
Program Max Concentration = 21.5 ^ig/m3
LAKY Max Concentration = 21.0 ^ig/m3
Program Max Concentration = 21.5 ^ig/m3
TVKY Max Concentration = 21.5 ^ig/m3
0.6	0.9
Concentration {[jg/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


14-64

-------
• Figure 14-29 presents the box plots for the seven Kentucky sites for which
1,3-butadiene is a pollutant of interest. Note that the program-level maximum
concentration (21.5 |ig/m3) is not shown directly on the box plots because the
scale of the box plots would be too large to readily observe data points at the
lower end of the concentration range. Thus, the scale of the box plot has been
reduced to 1.5 |ig/m3. Also, since the maximum 1,3-butadiene concentration for
several sites is greater than the scale of the box plots, the site-specific maximum
concentrations are labeled for these sites. The maximum 1,3-butadiene
concentration measured across the program was measured at TVKY, although
maximum concentrations greater than the scale of the box plots were also
measured at BLKY, CCKY, and LAKY. In each of these cases, the maximum
concentrations were a full order of magnitude greater than the scale on the box
plots. The annual average concentrations of 1,3-butadiene for TVKY, LAKY, and
BLKY are at least twice the next highest annual average concentration (which
was calculated for CCKY). With the exception of ATKY, each Calvert City site's
annual average concentration is greater than the program-level concentration.
Note that the program-level average concentration is considerably higher than the
third quartile and more than twice the program-level median concentration,
indicating that the 1,3-butadiene concentrations on the upper end of the
concentration range are driving the program-level average upward. The annual
average 1,3-butadiene concentrations for the two sites not located in Calvert City
are less than or similar to the program-level median concentration.
Figure 14-30. Program vs. Site-Specific Average Cadmium Concentration

-
^ 1
Program Max Concentration = 120 ng/m3
LJ 1

0	2	4	6	8	10
Concentration {ng/m3)
Program: 1st Quartile
2nd Quartile 3rd Quartile
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



• Figure 14-30 is the box plot for cadmium for ASKY-M, the only Kentucky site
for which this is a pollutant of interest. Similar to other pollutants, the program-
level maximum concentration (120 ng/m3) is not shown directly on the box plot
for cadmium as the scale of the box plot has been reduced to 10 ng/m3 in order to
allow for the observation of data points at the lower end of the concentration
range. Although the maximum concentration across the program was not
measured at ASKY-M, the second highest cadmium concentration was measured
at this site (although considerably less). The annual average concentration of
cadmium for ASKY-M is just greater than the program-level average
concentration. This site has the third highest annual average concentration of
cadmium, behind only SEWA and S4MO. Note that the program-level average
14-65

-------
cadmium concentration is more than twice the third quartile and nearly four times
the program-level median concentration, indicating that the cadmium
concentrations on the upper end of the concentration range, particularly the
maximum concentration, are driving the program-level average concentration
upward.
Figure 14-31. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations





Program Max Concentration = 23.7 ^ig/m3





D		
Program Max Concentration = 23.7 ^ig/m3

BLKY







Q




BLKY Max Concentration = 23.7 ^ig/m3





CCKY
Program Max Concentration = 23.7 ^ig/m3
GLKY
Program Max Concentration = 23.7 ^ig/m3
LAKY
Program Max Concentration = 23.7 ^ig/m3

Program Max Concentration = 23.7 ^ig/m3








TVKY Max Concentration = 2.33 ^ig/m3



0	0.25	0.5	0.75	1	1.25	1.5	1.75	2
Concentration {[ig/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



14-66

-------
• Figure 14-31 presents the box plots for the seven Kentucky sites for which carbon
tetrachloride is a pollutant of interest. Similar to other pollutants, the program-
level maximum concentration (23.7 |ig/m3) is not shown directly on the box plots
for carbon tetrachloride as the scale of the box plots has been reduced to 2 |ig/m3
in order to allow for the observations data points at the lower end of the
concentration range. Also, since the maximum carbon tetrachloride concentration
for several sites is greater than the scale of the box plots, the site-specific
maximum concentrations are labeled for these sites. The maximum carbon
tetrachloride concentration measured across the program was measured at BLKY.
Although the maximum carbon tetrachloride concentration measured at TVKY is
also greater than the scale of the box plot, the measurement was an order of
magnitude less (2.33 |ig/m3). The annual average concentrations for TVKY and
BLKY are higher than the annual averages for the other sites as well as the
program-level average, which all fall between the program-level median and third
quartile.
Figure 14-32. Program vs. Site-Specific Average /7-Dichlorobenzene Concentration
0.3	0.4
Concentration {[jg/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range


• Figure 14-32 is the box plot for /;-dichlorobenzene for TVKY, the only Kentucky
site for which this is a pollutant of interest. Note that the first and second quartiles
are both zero for this pollutant, indicating that at least half of the measurements
are non-detects and thus, are not visible on the box plot. The maximum
concentration measured at TVKY is less than the maximum concentration
measured across the program but still among the higher measurements. The
annual average p-dichlorobenzene concentration for TVKY is less than the
program-level average concentration and is the lowest annual average
concentration among NMP sites for which this is a pollutant of interest.
14-67

-------
Figure 14-33. Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations
¦
¦
¦
Program Max Concentration = 111 ^ig/m3
Program Max Concentration = 111 ^ig/m3
ATKY Max Concentration = 1.86 ^ig/m3
Program Max Concentration = 111 ^ig/m3
¦
in
¦
BLKY Max Concentration = 15.5 (ig/m3; Average Concentration = 1.27 [ig/m3
Program Max Concentration = 111 ^ig/m3
CCKY Max Concentration = 1.78 fig/m3
Program Max Concentration = 111 ^ig/m3
Program Max Concentration = 111 ^ig/m3
LAKY Max Concentration = 11.2 ^ig/m3
¦
Program Max Concentration = 111 ^ig/m3
TVKY Max Concentration = 111 |ig/m3; Average Concentration = 3.75 ^ig/m3
0.4	0.6
Concentration {[jg/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


• Figure 14-33 presents the box plots for the seven Kentucky sites for which
1,2-dichloroethane is a pollutant of interest. Similar to other pollutants, the
program-level maximum concentration (111 |ig/m3) is not shown directly on the
box plots for 1,2-dichloroethane as the scale of the box plots has been reduced to
1 |ig/m3 in order to allow for the observations data points at the lower end of the
concentration range. Also, since the maximum 1,2-dichloroethane concentration
for several sites is greater than the scale of the box plots, the site-specific
14-68

-------
maximum concentrations are labeled for these sites. The range of
1,2-dichloroethane concentrations measured at each Calvert City site exceeds the
scale of the box plots; by comparison, the entire range of 1,2-dichloroethane
concentrations measured at ASKY and GLKY is less than the first gridline on the
box plots (0.2 |ig/m3). In addition, the annual average 1,2-dichloroethane
concentrations for BLKY and TVKY exceed the scale of the box plots and are
also labeled on the box plots for these sites. Note that the program-level average
1,2-dichloroethane concentration is considerably higher than the program-level
median and third quartile, indicating that the concentrations on the upper end of
the concentration range are driving the program-level average upward. Recall
from the previous section that the annual average concentrations for the Calvert
City sites account for the five highest 1,2-dichloroethane concentrations among
NMP sites sampling VOCs. The only Calvert City site with an annual average
concentration less than the program-level average is CCKY, although it is still
more than twice the annual average for the NMP monitoring site with the next
highest annual average (BTUT), as shown in Table 4-9.
Figure 14-34. Program vs. Site-Specific Average Formaldehyde Concentrations




¦-<
^	
0	3	6	9	12	15	18	21	24
Concentration {[jg/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range


• Figure 14-34 is the box plot for formaldehyde for ASKY, GLKY, and LEKY, the
only Kentucky sites at which carbonyl compounds were sampled. The range of
formaldehyde concentrations measured was largest for ASKY and smallest for
GLKY. Among these three sites, GLKY has the lowest annual average
concentration while LEKY has the highest. LEKY's annual average is similar to
the program-level average concentration; ASKY's annual average falls between
the program-level median and average concentrations; and GLKY's is similar to
the program-level first quartile.
14-69

-------
Figure 14-35. Program vs. Site-Specific Average Hexachloro-l,3-Butadiene Concentrations



Kj 1





?



O |

U 1



	¦
0.05
0.1
0.15
Concentration {[jg/m3)
0.2
0.25
Program:
IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i
Site:
Site Average
o
Site Concentration Range


• Figure 14-35 is the box plot for hexachloro-1,3-butadiene for BLKY, CCKY,
LAKY, and TVKY, the Kentucky sites for which this is a pollutant of interest.
Note that the first, second, and third quartiles are all zero for this pollutant,
indicating that at least 75 percent of the measurements are non-detects and thus,
are not visible on the box plots. The maximum hexachloro-l,3-butadiene
concentration measured at BLKY is the maximum hexachloro-l,3-butadiene
concentration measured across the program. Even though the annual average
hexachloro-1,3-butadiene concentrations for all of the sites shown are similar to
the program-level average concentration, the annual averages of this pollutant for
all NMP sites sampling VOCs fall within a relatively tight range across the
program (spanning less than 0.03 |ig/m3 for all NMP sites sampling VOCs).
14-70

-------
Figure 14-36. Program vs. Site-Specific Average Lead (PMio) Concentration

20	30
Concentration {ng/m3)
Progra m: 1st Qua rti 1 e
2ndQuartile 3rdQuartile
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



• Figure 14-36 is the box plot for lead for ASKY-M, the only Kentucky site for
which lead was identified as a pollutant of interest. Although the maximum lead
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 two times greater than the program-level average
concentration and is the second highest annual average concentration of lead
calculated among NMP sites sampling PMio metals (behind S4MO). Note that
ASKY-M is one of only two NMP sites sampling metals for which lead is a
pollutant of interest (S4MO is the other).
Figure 14-37. Program vs. Site-Specific Average Manganese (PMio) Concentration

60
Concentration {ng/m3
Program: 1st Qua rti le
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


• Figure 14-37 is the box plot for manganese (PMio) for ASKY-M, the only
Kentucky site for which manganese was identified as a pollutant of interest.
Although the maximum manganese concentration across the program was not
measured at ASKY-M, this site does have one of the higher measurements,
including the fourth through 13th highest concentrations. The annual average
concentration of manganese for ASKY-M is nearly five times greater than the
program-level average concentration and is the highest annual average
concentration of manganese calculated among NMP sites sampling PMio metals.
Note that ASKY-M is the only NMP site sampling PMio metals for which
manganese is a pollutant of interest.
14-71

-------
Figure 14-38. Program vs. Site-Specific Average Nickel (PMio) Concentrations
I
0
5
10 15
Concentration (ng/m3)

20

Program: IstQuartile
¦
2nd Quartile 3rdQuartile
~ ~
4th Quartile
~
Average
i

Site: Site Average
o
Site Concentration Range


• Figure 14-38 presents the box plots for nickel for ASKY-M and BAKY. The box
plots show that the maximum nickel concentration measured across the program
was measured at ASKY-M. The maximum nickel concentration measured at
BAKY is considerably less. The annual average concentration of nickel for
ASKY-M is more than two times greater than the program-lev el average
concentration and is the highest annual average concentration of nickel calculated
among NMP sites sampling PMio metals. By comparison, the annual average
concentration of nickel for BAKY is roughly one-fourth as high and ranks 10th
across the program.
Figure 14-39. Program vs. Site-Specific Average 1,1,2-Trichloroethane Concentrations



i O i
Program Max Concentration = 2.15 ^ig/m3
i u •




Program Max Concentration = 2.15 ^ig/m3
0.25
0.5
0.75

Concentration {[jg/m3)

Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile Average
¦
~
~
~ 1
Site: Site Average
Site Concentration Range

o


14-72

-------
• Figure 14-39 presents the box plots for 1,1,2-trichloroethane for BLKY and
TVKY, the only Kentucky sites for which this is a pollutant of interest. The
program-level maximum concentration (2.15 |ig/m3) is not shown directly on the
box plots for 1,1,2-trichloroethane as the scale of the box plots has been reduced
to 1 |ig/m3 in order to allow for the observations data points at the lower end of
the concentration range. Also, the first, second, and third quartiles are all zero for
this pollutant, indicating that at least 75 percent of the measurements are non-
detects and thus, are not visible on the box plots. The maximum
1,1,2-trichloroethane concentration measured at TVKY is the maximum
1,1,2-trichloroethane concentration measured across the program. The annual
average 1,1,2-trichloroethane concentrations for these two sites are an order of
magnitude greater than the program-level average concentration. As discussed
previously most of the measured detections of this pollutant were measured at the
Calvert City sites, TVKY and BLKY in particular, and these are the only two
NMP sites across the program with 1,1,2-trichloroethane as a pollutant of interest.
ATKY
BLKY
LAKY
TVKY
1
2
3 4
Concentration {[jg/m3)
5
Program:
IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4thQuartile Average

¦
~
~
~ 1
Site:
Site Average
Site Concentration Range


o


• Figure 14-40 presents the box plots for vinyl chloride for ATKY, BLKY, LAKY,
and TVKY, the only Kentucky sites for which this is a pollutant of interest. The
first, second, and third quartiles are all zero for this pollutant, indicating that at
least 75 percent of the measurements are non-detects and thus, are not visible on
14-73
Figure 14-40. Program vs. Site-Specific Average Vinyl Chloride Concentrations
O
>o
y

-------
the box plots. The maximum vinyl chloride concentration measured at TVKY is
the maximum concentration measured across the program, although several
concentrations greater than 4 |ig/m3 were also measured at ATKY and BLKY.
The annual average vinyl chloride concentrations for these sites range from
0.07 ± 0.04 |ig/m3 for LAKY to 0.60 ± 0.31 |ig/m3 for ATKY, all of which are
greater than the program-level average concentration of 0.04 |ig/m3. The number
of measured detections ranges from 29 for LAKY to 38 for ATKY for the sites
shown, with 26 measured at the fifth Calvert City site (CCKY). The other NMP
sites combined measured no more than 14 measured detections of this pollutant,
with most measuring three or less.
14.4.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.3.2. The
only pollutant group for which GLKY has sampled under the NMP since at least 2009 is
hexavalent chromium and PAHs (sampling of VOCs at GLKY began in 2010, and carbonyl
compounds and PMio metals in 2011); 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. The remaining Kentucky sites did not
begin sampling under the NMP until 2012.
14.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at the Kentucky monitoring sites. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4
for definitions and explanations regarding the various toxicity factors, time frames, and
calculations associated with these risk-based screenings.
14.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Kentucky monitoring sites and where annual average
concentrations could be calculated, risk was examined by calculating cancer risk and noncancer
hazard approximations. These approximations can be used as risk estimates for cancer and
noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers want to shift their air
monitoring priorities. Refer to Section 3.4.3.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
14-74

-------
approximations are presented in Table 14-7, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Table 14-7. Risk Approximations for the Kentucky Monitoring Sites
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Health Department, Ashland, Kentucky - ASKY
Acetaldehyde
0.0000022
0.009
61/61
1.19
±0.14
2.62
0.13
Benzene
0.0000078
0.03
61/61
1.52
± 1.39
11.85
0.05
1.3 -Butadiene
0.00003
0.002
47/61
0.06
±0.01
1.71
0.03
Carbon Tetrachloride
0.000006
0.1
61/61
0.65
±0.02
3.92
0.01
1,2 -Dichloroethane
0.000026
2.4
53/61
0.08
±0.01
1.95
<0.01
Formaldehyde
0.000013
0.0098
61/61
2.30
±0.38
29.87
0.23
21st and Greenup, Ashland, Kentucky - ASKY-M
Arsenic (PMi0)a
0.0043
0.000015
59/60
1.24
±0.29
5.34
0.08
Cadmium (PMi0)a
0.0018
0.00001
60/60
0.40
±0.19
0.72
0.04
Lead (PMi0)a

0.00015
60/60
8.25
± 1.89

0.06
Manganese (PMi0)a

0.0003
60/60
19.86
±4.06

0.07
Nickel (PMi,:,)a
0.00048
0.00009
60/60
2.40
±0.89
1.15
0.03
Grayson, Kentucky - GLKY
Acetaldehyde
0.0000022
0.009
61/61
0.68
±0.06
1.50
0.08
Benzene
0.0000078
0.03
61/61
0.49
±0.10
3.79
0.02
1.3 -Butadiene
0.00003
0.002
48/61
0.04
±0.01
1.24
0.02
Carbon Tetrachloride
0.000006
0.1
61/61
0.67
±0.03
3.99
0.01
1,2 -Dichloroethane
0.000026
2.4
56/61
0.08
±0.01
1.96
<0.01
Formaldehyde
0.000013
0.0098
61/61
1.31
±0.20
17.05
0.13
Arsenic (PMi0)a
0.0043
0.000015
58/59
0.48
±0.10
2.08
0.03
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of viewing.
NA = Not available due to the criteria for calculating an annual average.
— = A Cancer URE or Noncancer RfC is not available.
14-75

-------
Table 14-7. Risk Approximations for the Kentucky Monitoring Sites (Continued)
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Baskett, Kentucky - BAKY
Arsenic (PMi0)a
0.0043
0.000015
59/60
0.82
±0.22
3.53
0.05
Nickel (PMi,;,)a
0.00048
0.00009
60/60
0.61
±0.15
0.29
0.01
Atmos Energy, Calvert City, Kentucky - ATKY
Benzene
0.0000078
0.03
61/61
0.54
±0.05
4.18
0.02
1.3 -Butadiene
0.00003
0.002
44/61
0.07
±0.02
1.98
0.03
Carbon Tetrachloride
0.000006
0.1
61/61
0.67
±0.02
4.02
0.01
1,2 -Dichloroethane
0.000026
2.4
60/61
0.30
±0.10
7.71
<0.01
Vinyl chloride
0.0000088
0.1
38/61
0.60
±0.31
5.24
0.01
Smithland, Kentucky - BLKY
Benzene
0.0000078
0.03
59/59
0.67
±0.15
5.26
0.02
1.3 -Butadiene
0.00003
0.002
39/59
0.63
±0.49
18.98
0.32
Carbon Tetrachloride
0.000006
0.1
59/59
1.11
±0.77
6.69
0.01
1,2 -Dichloroethane
0.000026
2.4
57/59
1.27
±0.75
33.15
<0.01
Hexachloro -1,3 -butadiene
0.000022
0.09
8/59
0.02
±0.01
0.40
<0.01
1,1,2-Trichloroethane
0.000016
0.4
7/59
0.02
±0.02
0.29
<0.01
Vinyl chloride
0.0000088
0.1
35/59
0.18
±0.16
1.56
<0.01
Calvert City Elementary School, Calvert City, Kentucky - CCKY
Benzene
0.0000078
0.03
61/61
0.50
±0.05
3.87
0.02
1.3 -Butadiene
0.00003
0.002
45/61
0.26
±0.40
7.80
0.13
Carbon Tetrachloride
0.000006
0.1
61/61
0.67
±0.02
4.03
0.01
1,2 -Dichloroethane
0.000026
2.4
59/61
0.24
±0.08
6.20
<0.01
Hexachloro -1,3 -butadiene
0.000022
0.09
11/61
0.02
±0.01
0.34
<0.01
Arsenic (PMi0)a
0.0043
0.000015
55/56
0.61
±0.15
2.64
0.04
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of viewing.
NA = Not available due to the criteria for calculating an annual average.
— = A Cancer URE or Noncancer RfC is not available.
14-76

-------
Table 14-7. Risk Approximations for the Kentucky Monitoring Sites (Continued)
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Lazy Daze, Calvert City, Kentucky - LAKY
Benzene
0.0000078
0.03
60/60
0.62
±0.11
4.86
0.02
1.3 -Butadiene
0.00003
0.002
48/60
0.66
±0.80
19.67
0.33
Carbon Tetrachloride
0.000006
0.1
60/60
0.68
±0.03
4.07
0.01
1,2 -Dichloroethane
0.000026
2.4
60/60
0.70
±0.45
18.21
<0.01
Hexachloro -1,3 -butadiene
0.000022
0.09
12/60
0.02
±0.01
0.41
<0.01
Vinyl chloride
0.0000088
0.1
29/60
0.07
±0.04
0.59
<0.01
TVA Substation, Calvert City, Kentucky - TVKY
Benzene
0.0000078
0.03
61/61
1.06
±0.32
8.29
0.04
1,3-Butadiene
0.00003
0.002
46/61
1.03
±0.97
30.97
0.52
Carbon Tetrachloride
0.000006
0.1
61/61
0.80
±0.08
4.80
0.01
/?-Dichlorobcnzcnc
0.000011
0.8
24/61
0.03
±0.02
0.35
<0.01
1,2 -Dichloroethane
0.000026
2.4
61/61
3.75
±3.68
97.42
<0.01
Hexachloro -1,3 -butadiene
0.000022
0.09
11/61
0.01
±0.01
0.29
<0.01
1,1,2-Trichloroethane
0.000016
0.4
9/61
0.05
±0.07
0.82
<0.01
Vinyl chloride
0.0000088
0.1
36/61
0.29
±0.22
2.57
<0.01
Lexington, Kentucky - LEKY
Acetaldehyde
0.0000022
0.009
61/61
1.49
±0.15
3.27
0.17
Benzene
0.0000078
0.03
45/45
NA
NA
NA
1,3-Butadiene
0.00003
0.002
41/45
NA
NA
NA
Carbon Tetrachloride
0.000006
0.1
45/45
NA
NA
NA
1,2 -Dichloroethane
0.000026
2.4
43/45
NA
NA
NA
Formaldehyde
0.000013
0.0098
61/61
2.91
±0.49
37.85
0.30
Hexachloro -1,3 -butadiene
0.000022
0.09
9/45
NA
NA
NA
Arsenic (PMi0)a
0.0043
0.000015
52/53
0.68
±0.12
2.93
0.05
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of viewing.
NA = Not available due to the criteria for calculating an annual average.
— = A Cancer URE or Noncancer RfC is not available.
14-77

-------
Observations for the Kentucky monitoring sites from Table 14-7 include the following:
•	The pollutants with the highest annual average concentrations for ASKY are
formaldehyde, benzene, and acetaldehyde. Formaldehyde and benzene are the only
two pollutants with cancer risk approximations greater than 10 in-a-million for ASKY
(29.87 in-a-million and 11.85 in-a-million, respectively). All of the noncancer hazard
approximations for the pollutants of interest for ASKY are considerably less than an
HQ of 1.0 (0.23 or less), indicating that no adverse noncancer health effects are
expected from these individual pollutants.
•	The pollutants with the highest annual average concentrations for ASKY-M are
manganese, lead, and nickel. Arsenic has the highest cancer risk approximation
among ASKY-M's pollutants of interest (5.34 in-a-million). All of the noncancer
hazard approximations for the pollutants of interest for ASKY-M are considerably
less than an HQ of 1.0 (0.08 or less), indicating that no adverse noncancer health
effects are expected from these individual pollutants.
•	Formaldehyde is the only pollutant of interest for GLKY with an annual average
concentration greater than 1 |ig/m3. This pollutant also has the only cancer risk
approximation greater than 10 in-a-million for GLKY (17.05 in-a-million). All of the
noncancer hazard approximations for the pollutants of interest for GLKY are
considerably less than an HQ of 1.0 (0.13 or less), indicating that no adverse
noncancer health effects are expected from these individual pollutants.
•	Arsenic and nickel are the only pollutants of interest for BAKY. Arsenic has a cancer
risk approximation greater than 1 in-a-million for BAKY (3.53 in-a-million) while
nickel does not. The noncancer hazard approximations for the two pollutants of
interest for BAKY are considerably less than an HQ of 1.0 (0.05 or less), indicating
that no adverse noncancer health effects are expected from these individual
pollutants.
•	With the exception of ATKY, 1,3-butadiene and 1,2-dichloroethane have the highest
cancer risk approximations among the pollutants of interest for the Calvert City sites.
Cancer risk approximations for 1,3-butadiene range from 1.98 in-a-million for ATKY
to 30.97 in-a-million for TVKY, with the cancer risk approximations for TVKY,
LAKY, BLKY, and CCKY ranking highest among all sites sampling this pollutant.
Cancer risk approximations for 1,2-dichloroethane range from 6.20 in-a-million for
CCKY to 97.42 in-a-million for TVKY, with the cancer risk approximations for
TVKY, BLKY, LAKY, ATKY, and CCKY ranking highest among all sites sampling
this pollutant. Further, the cancer risk approximation for TVKY is the second highest
among all cancer risk approximations calculated for the site-specific pollutants of
interest (behind only BTUT's formaldehyde cancer risk approximation).
•	Benzene and carbon tetrachloride have the next highest cancer risk approximations
among the pollutants of interest for the Calvert City sites. Cancer risk approximations
for benzene range from 3.87 in-a-million for CCKY to 8.29 in-a-million for TVKY.
Cancer risk approximations for carbon tetrachloride range from 4.02 in-a-million for
ATKY to 6.69 in-a-million for BLKY. The cancer risk approximations for BLKY and
TVKY rank highest among all sites sampling this carbon tetrachloride.
14-78

-------
•	Vinyl chloride is another pollutant of interest that most of the Calvert City sites have
in common (CCKY is the exception). Cancer risk approximations for vinyl chloride
range from 0.59 in-a-million for LAKY to 5.24 in-a-million for ATKY.
•	With the exception of arsenic, which is a pollutant of interest for CCKY, the
remaining pollutants of interest for the Calvert City sites have cancer risk
approximations less than 1.0 in-a-million. The cancer risk approximation for arsenic
for CCKY is 2.64 in-a-million.
•	All of the noncancer hazard approximations for the pollutants of interest for the
Calvert City sites are less than an HQ of 1.0, indicating that no adverse noncancer
health effects are expected from these individual pollutants. For each of these sites,
the pollutant with the highest noncancer hazard approximation is 1,3-butadiene,
which ranged from 0.03 for ATKY to 0.52 for TVKY, which is the third highest
noncancer hazard approximation among all noncancer hazard approximations
calculated for the site-specific pollutants of interest (behind only BTUT and GPCO's
formaldehyde noncancer hazard approximations).
•	Formaldehyde, acetaldehyde, and arsenic are the only pollutants of interest for LEKY
for which annual average concentrations could be calculated. The cancer risk
approximation for formaldehyde (37.85 in-a-million) is an order of magnitude greater
than the cancer risk approximations for the other two pollutants of interest. All of the
noncancer hazard approximations for the pollutants of interest for LEKY are
considerably less than an HQ of 1.0 (0.30 or less), indicating that no adverse
noncancer health effects are expected from these individual pollutants, where they
could be calculated.
As an extension of this analysis, pollution roses were created for each of the site-specific
pollutants of interest that have a cancer risk approximation greater than 75 in-a-million and/or a
noncancer hzard approximation greater than 1.0, where applicable. Thus, a pollution rose was
created for TVKY's 1,2-dichloroethane measurements. A pollution rose is a plot of the ambient
concentration versus the wind speed and direction; the magnitude of the concentration is
indicated using different colored dots and are shown in relation to the average wind direction
oriented about a 16-point compass, similar to the wind roses presented in Section 14.2.3. Thus,
high concentrations may be shown in relation to the direction of potential emissions sources.
Hourly NWS wind observations used in this analysis were averaged (using vector averaging
techniques) to compute daily wind direction averages for comparison to the 24-hour
concentration data. This analysis is intended to help identify the geographical area where the
emissions sources of these pollutants may have originated. Additional information regarding this
analysis is also presented in Section 3.4.3.3.
14-79

-------
Figure 14-41 presents the pollution rose for all 61 1,2-dichloroethane concentrations
measured at TVKY. However, the magnitude of the maximum concentration (111 |ig/m3) is such
that all of the lower concentrations are plotted nearly on top of each other. As a result, three
pollution roses were created for TVKY. One that shows all measurements, one that shows all
1,2-dichloroethane measurements that fall between 1 and 20 |ig/m3, and one that shows all
1,2-dichloroethane measurements less than 1 |ig/m3.
Observations for Figure 14-41 include the following:
•	The pollution rose with all 1,2-dichloroethane measurements plotted on it shows
that the maximum 1,2-dichloroethane (shown in green) measured at TVKY is
greater than 100 |ig/m3 but all other measurements are less than 20 |ig/m3. This
maximum concentration was measured on November 18, 2013, a day with an
average wind direction of roughly 315° or northwest. Individual hourly wind
directions observed that day ranged from 260° (west) to 350° (north-northwest)
with wind speeds ranging from 3 knots to 14 knots; in addition, calm winds were
measured for six of the hourly observations.
•	The pollution rose with 1,2-dichloroethane measurements ranging from 1 |ig/m3
to 20 |ig/m3 plotted on it shows that most of these concentrations were measured
on days with average wind directions between 225° (southwest) and 45°
(northeast) and are shown above the diagonal northeast-southwest line. In
addition, only four of these measurements are shown in association with wind
directions with a southerly component (or below the horizontal east-west line).
•	The pollution rose with 1,2-dichloroethane measurements less 1 |ig/m3 plotted on
it shows that these concentrations were measured on days with average wind
directions between 45° (northeast) and 225° (southwest) and are shown below the
diagonal northeast-southwest line. In addition, only four of these measurements
are shown in association with wind directions between 270° (west) and 45°
(northeast), which is nearly the opposite of the previous pollution rose.
14-80

-------
Figure 14-41. Pollution Rose for 1,2-Dichloroethane Concentrations Measured at TVKY
All Measurements	Measurements 1-20 |ig/m3
180
• <0.1 jig/m3 O 0.1 -1.0 jig/m3 01-20 jig/m3 O >100 jig/m3
180
01-20 jig/m3
Measurements less than 10 |ig/m
360/0
180
• <0.1 ng/m3 O 0.1 -1.0 fig/m3
14-81

-------
14.5.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, Tables 14-8 and 14-9 present an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 14-8 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
that have cancer toxicity factors. Table 14-8 also presents the 10 pollutants with the highest
toxicity-weighted emissions, based on the weighting schema described in Section 3.4.3.4. Lastly,
Table 14-8 provides the pollutants with the highest cancer risk approximations (in-a-million) for
each site, as presented in Table 14-7. The emissions, toxicity-weighted emissions, and cancer
risk approximations are shown in descending order in Table 14-8. Table 14-9 presents similar
information, but is limited to those pollutants with noncancer toxicity factors.
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 14.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
14-82

-------
Table 14-8. 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)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Health Department, Ashland, Kentucky (Boyd County) - ASKY
Benzene
61.79
Coke Oven Emissions, PM
7.25E-03
Formaldehyde
29.87
Formaldehyde
20.35
Hexavalent Chromium
9.84E-04
Benzene
11.85
Ethylbenzene
13.36
Nickel, PM
6.71E-04
Carbon Tetrachloride
3.92
Acetaldehyde
11.59
Benzene
4.82E-04
Acetaldehyde
2.62
Coke Oven Emissions, PM
7.32
Formaldehyde
2.64E-04
1,2-Dichloroethane
1.95
1.3 -Butadiene
3.65
2,4-Dinitrotoluene
1.96E-04
1,3-Butadiene
1.71
2,4-Dinitrotoluene
2.20
1,3-Butadiene
1.10E-04

Tetrachloroethylene
2.00
Naphthalene
6.77E-05
Naphthalene
1.99
Cadmium, PM
5.96E-05
Nickel, PM
1.40
POM, Group 2b
4.51E-05
21st and Greenup, Ashland, Kentucky (Boyd County) - ASKY-M
Benzene
61.79
Coke Oven Emissions, PM
7.25E-03
Arsenic
5.34
Formaldehyde
20.35
Hexavalent Chromium
9.84E-04
Nickel
1.15
Ethylbenzene
13.36
Nickel, PM
6.71E-04
Cadmium
0.72
Acetaldehyde
11.59
Benzene
4.82E-04

Coke Oven Emissions, PM
7.32
Formaldehyde
2.64E-04
1,3-Butadiene
3.65
2,4-Dinitrotoluene
1.96E-04
2,4-Dinitrotoluene
2.20
1,3-Butadiene
1.10E-04
Tetrachloroethylene
2.00
Naphthalene
6.77E-05
Naphthalene
1.99
Cadmium, PM
5.96E-05
Nickel, PM
1.40
POM, Group 2b
4.51E-05

-------
Table 14-8. 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)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Grayson, Kentucky (Carter County) - GLKY
Benzene
20.12
Formaldehyde
1.78E-04
Formaldehyde
17.05
Formaldehyde
13.70
Benzene
1.57E-04
Carbon Tetrachloride
3.99
Acetaldehyde
9.15
1,3-Butadiene
6.86E-05
Benzene
3.79
Ethylbenzene
9.14
Naphthalene
5.57E-05
Arsenic
2.08
1.3 -Butadiene
2.29
POM, Group 2b
3.60E-05
1,2-Dichloroethane
1.96
Naphthalene
1.64
POM, Group 2d
2.66E-05
Acetaldehyde
1.50
POM, Group 2b
0.41
Ethylbenzene
2.29E-05
1,3-Butadiene
1.24
POM, Group 2d
0.30
POM, Group 5a
2.23E-05

POM, Group 6
0.04
Acetaldehyde
2.01E-05
Trichloroethylene
0.03
POM, Group 6
7.36E-06
Baskett, Kentucky (Henderson County) - BAKY
Formaldehyde
52.75
Formaldehyde
6.86E-04
Arsenic
3.53
Benzene
42.14
Naphthalene
5.72E-04
Nickel
0.29
Acetaldehyde
27.10
POM, Group 2d
3.74E-04

Naphthalene
16.81
Benzene
3.29E-04
Ethylbenzene
16.17
Hexavalent Chromium
2.83E-04
Tetrachloroethylene
6.71
Nickel, PM
2.73E-04
1,3-Butadiene
6.59
POM, Group 2b
2.52E-04
POM, Group 2d
4.25
1,3-Butadiene
1.98E-04
POM, Group 2b
2.87
Acetaldehyde
5.96E-05
Dichloro methane
0.83
Cadmium, PM
5.03E-05

-------
Table 14-8. 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)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Atmos Energy, Calvert City, Kentucky (Marshall County) - ATKY
Benzene
139.16
Benzene
1.09E-03
1,2-Dichloroethane
7.71
Ethylbenzene
100.03
Hexavalent Chromium
6.36E-04
Vinyl chloride
5.24
Formaldehyde
36.34
Formaldehyde
4.72E-04
Benzene
4.18
Acetaldehyde
33.61
1,3-Butadiene
4.26E-04
Carbon Tetrachloride
4.02
Vinyl chloride
30.93
Vinyl chloride
2.72E-04
1,3-Butadiene
1.98
1.3 -Butadiene
14.20
Ethylbenzene
2.50E-04

1,2-Dichloroethane
9.25
1,2-Dichloroethane
2.41E-04
Naphthalene
3.45
POM, Group la
2.34E-04
POM, Group la
2.66
Naphthalene
1.17E-04
Carbon tetrachloride
2.32
Nickel, PM
7.66E-05
Calvert City Elementary, Calvert City, Kentucky (Marshall County) - CCKY
Benzene
139.16
Benzene
1.09E-03
1,3-Butadiene
7.80
Ethylbenzene
100.03
Hexavalent Chromium
6.36E-04
1,2-Dichloroethane
6.20
Formaldehyde
36.34
Formaldehyde
4.72E-04
Carbon Tetrachloride
4.03
Acetaldehyde
33.61
1,3-Butadiene
4.26E-04
Benzene
3.87
Vinyl chloride
30.93
Vinyl chloride
2.72E-04
Arsenic
2.64
1.3 -Butadiene
14.20
Ethylbenzene
2.50E-04
Hexachloro-1,3 -butadiene
0.34
1,2-Dichloroethane
9.25
1,2-Dichloroethane
2.41E-04

Naphthalene
3.45
POM, Group la
2.34E-04
POM, Group la
2.66
Naphthalene
1.17E-04
Carbon tetrachloride
2.32
Nickel, PM
7.66E-05

-------
Table 14-8. 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)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Lazy Daze, Calvert City, Kentucky (Marshall County) - LAKY
Benzene
139.16
Benzene
1.09E-03
1,3-Butadiene
19.67
Ethylbenzene
100.03
Hexavalent Chromium
6.36E-04
1,2-Dichloroethane
18.21
Formaldehyde
36.34
Formaldehyde
4.72E-04
Benzene
4.86
Acetaldehyde
33.61
1,3-Butadiene
4.26E-04
Carbon Tetrachloride
4.07
Vinyl chloride
30.93
Vinyl chloride
2.72E-04
Vinyl chloride
0.59
1.3 -Butadiene
14.20
Ethylbenzene
2.50E-04
Hexachloro-1,3 -butadiene
0.41
1,2-Dichloroethane
9.25
1,2-Dichloroethane
2.41E-04

Naphthalene
3.45
POM, Group la
2.34E-04
POM, Group la
2.66
Naphthalene
1.17E-04
Carbon tetrachloride
2.32
Nickel, PM
7.66E-05
TVA Substation, Calvert City, Kentucky (Marshall County) - TVKY
Benzene
139.16
Benzene
1.09E-03
1,2-Dichloroethane
97.42
Ethylbenzene
100.03
Hexavalent Chromium
6.36E-04
1,3-Butadiene
30.97
Formaldehyde
36.34
Formaldehyde
4.72E-04
Benzene
8.29
Acetaldehyde
33.61
1,3-Butadiene
4.26E-04
Carbon Tetrachloride
4.80
Vinyl chloride
30.93
Vinyl chloride
2.72E-04
Vinyl chloride
2.57
1.3 -Butadiene
14.20
Ethylbenzene
2.50E-04
1,1,2-Trichloroethane
0.82
1,2-Dichloroethane
9.25
1,2-Dichloroethane
2.41E-04
p-Dichlorobenzene
0.35
Naphthalene
3.45
POM, Group la
2.34E-04
Hexachloro-1,3 -butadiene
0.29
POM, Group la
2.66
Naphthalene
1.17E-04

Carbon tetrachloride
2.32
Nickel, PM
7.66E-05

-------
Table 14-8. 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)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Smithland, Kentucky (Livingston County) - BLKY
Benzene
14.04
Formaldehyde
1.50E-04
1,2-Dichloroethane
33.15
Formaldehyde
11.52
Benzene
1.09E-04
1,3-Butadiene
18.98
Acetaldehyde
6.66
1,3-Butadiene
5.45E-05
Carbon Tetrachloride
6.69
Ethylbenzene
5.39
Naphthalene
2.27E-05
Benzene
5.26
1.3 -Butadiene
1.82
POM, Group 2b
1.63E-05
Vinyl chloride
1.56
Naphthalene
0.67
Acetaldehyde
1.47E-05
Hexachloro-1,3 -butadiene
0.40
POM, Group 2b
0.18
Ethylbenzene
1.35E-05
1,1,2-Trichloroethane
0.29
POM, Group 2d
0.15
POM, Group 2d
1.31E-05

POM, Group 6
0.03
Nickel, PM
1.16E-05
Trichloroethylene
0.03
POM, Group 5a
1.14E-05
Lexington, Kentucky (Fayette County) - LEKY
Benzene
135.46
Formaldehyde
1.20E-03
Formaldehyde
37.85
Formaldehyde
92.28
Benzene
1.06E-03
Acetaldehyde
3.27
Ethylbenzene
82.26
1,3-Butadiene
5.57E-04
Arsenic
2.93
Acetaldehyde
54.60
Naphthalene
3.50E-04

1,3-Butadiene
18.57
POM, Group 2b
2.32E-04
Tetrachloroethylene
13.04
Ethylbenzene
2.06E-04
Naphthalene
10.31
POM, Group 2d
1.52E-04
POM, Group 2b
2.63
Hexavalent Chromium
1.34E-04
Trichloroethylene
1.94
Arsenic, PM
1.28E-04
POM, Group 2d
1.73
Acetaldehyde
1.20E-04

-------
Table 14-9. 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)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Health Department, Ashland, Kentucky (Boyd County) - ASKY
Toluene
89.31
Acrolein
63,687.70
Formaldehyde
0.23
Benzene
61.79
Chlorine
45,169.74
Acetaldehyde
0.13
Xylenes
52.93
Manganese, PM
33,849.17
Benzene
0.05
Hexane
49.05
Nickel, PM
15,539.17
1,3-Butadiene
0.03
Methanol
39.10
Lead, PM
11,227.89
Carbon Tetrachloride
0.01
Hydrochloric acid
27.65
Cadmium, PM
3,311.70
1,2-Dichloroethane
<0.01
Formaldehyde
20.35
Formaldehyde
2,076.07

Ethylbenzene
13.36
Benzene
2,059.62
Acetaldehyde
11.59
1,3-Butadiene
1,826.21
Manganese, PM
10.15
Hydrochloric acid
1,382.51
21st and Greenup, Ashland, Kentucky (Boyd County) - ASKY-M
Toluene
89.31
Acrolein
63,687.70
Arsenic
0.08
Benzene
61.79
Chlorine
45,169.74
Manganese
0.07
Xylenes
52.93
Manganese, PM
33,849.17
Lead
0.06
Hexane
49.05
Nickel, PM
15,539.17
Cadmium
0.04
Methanol
39.10
Lead, PM
11,227.89
Nickel
0.03
Hydrochloric acid
27.65
Cadmium, PM
3,311.70

Formaldehyde
20.35
Formaldehyde
2,076.07
Ethylbenzene
13.36
Benzene
2,059.62
Acetaldehyde
11.59
1,3-Butadiene
1,826.21
Manganese, PM
10.15
Hydrochloric acid
1,382.51

-------
Table 14-9. 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


Top 10 Noncancer Hazard Approximations
Noncancer RfCs

Top 10 Noncancer Toxicity-Weighted Emissions
Based on Annual Average Concentrations
(County-Level)

(County-Level)

(Site-Specific)





Noncancer



Noncancer

Hazard
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(HQ)


Grayson, Kentucky (Carter County
- GLKY


Toluene
58.15
Acrolein
45,189.54
Formaldehyde
0.13
Xylenes
35.59
Formaldehyde
1,397.58
Acetaldehyde
0.08
Hexane
25.95
Cyanide Compounds, gas
1,278.36
Arsenic
0.03
Benzene
20.12
1.3 -Butadiene
1,144.01
1,3-Butadiene
0.02
Methanol
15.68
Acetaldehyde
1,016.31
Benzene
0.02
Formaldehyde
13.70
Benzene
670.52
Carbon Tetrachloride
0.01
Acetaldehyde
9.15
Naphthalene
545.72
1,2-Dichloroethane
<0.01
Ethylbenzene
9.14
Xylenes
355.95


Ethylene glycol
5.50
Arsenic, PM
91.58


1.3 -Butadiene
2.29
Propionaldehyde
84.60


Baskett, Kentucky (Henderson County) - BAKY
Carbonyl sulfide
128.78
Acrolein
76,864.06
Arsenic
0.05
Toluene
112.00
Manganese, PM
7,205.03
Nickel
0.01
Xylenes
78.62
Nickel, PM
6,326.84


Hexane
54.97
Naphthalene
5,604.26


Formaldehyde
52.75
Formaldehyde
5,383.11


Benzene
42.14
1,3-Butadiene
3,295.39


Methanol
28.37
Chlorine
3,245.91


Acetaldehyde
27.10
Acetaldehyde
3,010.82


Naphthalene
16.81
Cadmium, PM
2,795.50


Ethylbenzene
16.17
4,4'-Methylenediphenyl diisocyanate, gas
2,483.57



-------
Table 14-9. 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)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Atmos Energy, Calvert City, Kentucky (Marshall County) - ATKY
Methanol
677.58
Chlorine
210,803.93
1,3-Butadiene
0.03
Xylenes
522.49
Acrolein
125,961.44
Benzene
0.02
Toluene
480.91
1.3 -Butadiene
7,098.77
Carbon Tetrachloride
0.01
Benzene
139.16
Xylenes
5,224.89
Vinyl chloride
0.01
Hexane
100.70
Benzene
4,638.60
1,2-Dichloroethane
<0.01
Ethylbenzene
100.03
Hydrochloric acid
4,173.99

Hydrochloric acid
83.48
Acetaldehyde
3,734.64
Vinyl acetate
73.28
Formaldehyde
3,708.16
Formaldehyde
36.34
Acrylic acid
2,916.21
Acetaldehyde
33.61
Nickel, PM
1,773.75
Calvert City Elementary, Calvert City, Kentucky (Marshall County) - CCKY
Methanol
677.58
Chlorine
210,803.93
1,3-Butadiene
0.13
Xylenes
522.49
Acrolein
125,961.44
Arsenic
0.04
Toluene
480.91
1.3 -Butadiene
7,098.77
Benzene
0.02
Benzene
139.16
Xylenes
5,224.89
Carbon Tetrachloride
0.01
Hexane
100.70
Benzene
4,638.60
Hexachloro-1,3 -butadiene
0.00
Ethylbenzene
100.03
Hydrochloric acid
4,173.99
1,2-Dichloroethane
<0.01
Hydrochloric acid
83.48
Acetaldehyde
3,734.64

Vinyl acetate
73.28
Formaldehyde
3,708.16
Formaldehyde
36.34
Acrylic acid
2,916.21
Acetaldehyde
33.61
Nickel, PM
1,773.75

-------
Table 14-9. 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


Top 10 Noncancer Hazard Approximations
Noncancer RfCs

Top 10 Noncancer Toxicity-Weighted Emissions
Based on Annual Average Concentrations
(County-Level)

(County-Level)

(Site-Specific)





Noncancer



Noncancer

Hazard
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(HQ)
Lazy Daze, Calvert City, Kentucky (Marshall County) - LAKY
Methanol
677.58
Chlorine
210,803.93
1,3-Butadiene
0.33
Xylenes
522.49
Acrolein
125,961.44
Benzene
0.02
Toluene
480.91
1.3 -Butadiene
7,098.77
Carbon Tetrachloride
0.01
Benzene
139.16
Xylenes
5,224.89
Vinyl chloride
<0.01
Hexane
100.70
Benzene
4,638.60
1,2-Dichloroethane
<0.01
Ethylbenzene
100.03
Hydrochloric acid
4,173.99
Hexachloro-1,3 -butadiene
<0.01
Hydrochloric acid
83.48
Acetaldehyde
3,734.64


Vinyl acetate
73.28
Formaldehyde
3,708.16


Formaldehyde
36.34
Acrylic acid
2,916.21


Acetaldehyde
33.61
Nickel, PM
1,773.75


TVA Substation, Calvert City, Kentucky (Marshall County) - TVKY
Methanol
677.58
Chlorine
210,803.93
1,3-Butadiene
0.52
Xylenes
522.49
Acrolein
125,961.44
Benzene
0.04
Toluene
480.91
1.3 -Butadiene
7,098.77
Carbon Tetrachloride
0.01
Benzene
139.16
Xylenes
5,224.89
Vinyl chloride
<0.01
Hexane
100.70
Benzene
4,638.60
1,2-Dichloroethane
<0.01
Ethylbenzene
100.03
Hydrochloric acid
4,173.99
Hexachloro-1,3 -butadiene
<0.01
Hydrochloric acid
83.48
Acetaldehyde
3,734.64
1,1,2-Trichloroethane
<0.01
Vinyl acetate
73.28
Formaldehyde
3,708.16
/?-Dichlorobcnzcnc
<0.01
Formaldehyde
36.34
Acrylic acid
2,916.21


Acetaldehyde
33.61
Nickel, PM
1,773.75



-------
Table 14-9. 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


Top 10 Noncancer Hazard Approximations
Noncancer RfCs

Top 10 Noncancer Toxicity-Weighted Emissions
Based on Annual Average Concentrations
(County-Level)

(County-Level)

(Site-Specific)





Noncancer



Noncancer

Hazard
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(HQ)
Smithland, Kentucky (Livingston County) - BLKY
Toluene
43.04
Acrolein
20,492.52
1,3-Butadiene
0.32
Xylenes
38.50
Formaldehyde
1,175.61
Benzene
0.02
Benzene
14.04
1,3-Butadiene
909.04
Carbon Tetrachloride
0.01
Hexane
12.08
Acetaldehyde
740.44
Vinyl chloride
<0.01
Formaldehyde
11.52
Cyanide Compounds, gas
527.46
1,2-Dichloroethane
<0.01
Acetaldehyde
6.66
Benzene
467.89
Hexachloro-1,3 -butadiene
<0.01
Ethylbenzene
5.39
Xylenes
384.98
1,1,2-Trichloroethane
<0.01
Methanol
5.38
Nickel, PM
268.68


Ethylene glycol
1.89
Naphthalene
222.51


1.3 -Butadiene
1.82
Manganese, PM
201.76


Lexington, Kentucky (Fayette County) - LEKY
Toluene
487.75
Acrolein
277,725.18
Formaldehyde
0.30
Xylenes
315.94
Formaldehyde
9,416.72
Acetaldehyde
0.17
Hexane
246.40
1,3-Butadiene
9,286.39
Arsenic
0.05
Methanol
176.71
Acetaldehyde
6,066.40


Benzene
135.46
Benzene
4,515.24


Formaldehyde
92.28
Naphthalene
3,436.18


Ethylbenzene
82.26
Xylenes
3,159.41


Ethylene glycol
59.08
Hexamethylene-1,6-diisocyanate, gas
2,051.30


Acetaldehyde
54.60
Arsenic, PM
1,982.63


Methyl isobutyl ketone
29.90
4,4'-Methylenediphenyl diisocyanate, gas
1,757.48



-------
Observations from Table 14-8 include the following:
•	Among the Kentucky counties with monitoring sites, emissions (for pollutants with
cancer UREs) are highest in Fayette County (LEKY) and Marshall County (Calvert
City) and lowest in Livingston County (BLKY) and Carter County (GLKY).
•	Benzene, formaldehyde, ethylbenzene are the highest emitted pollutants with cancer
UREs in Boyd County, where the Ashland sites are located. Coke oven emissions,
hexavalent chromium, and nickel are the pollutants with the highest toxicity-weighted
emissions (of the pollutants with cancer UREs) for Boyd County. Seven of the
highest emitted pollutants also have the highest toxicity-weighted emissions for Boyd
County.
•	For ASKY, formaldehyde, benzene, and 1,3-butadiene are among the pollutants with
the highest cancer risk approximations and appear on both emissions-based lists.
Acetaldehyde, which has the fourth highest cancer risk approximation for ASKY, has
the fourth highest emissions for Boyd County but is not among the pollutants with the
highest toxicity-weighted emissions (acetaldehyde ranks 13th for toxicity-weighted
emissions). Carbon tetrachloride and 1,2-dichloroethane, the other pollutants of
interest for ASKY, appear on neither emissions-based list.
•	Nickel is the only pollutant of interest for ASKY-M to appear on both emissions-
based lists for Boyd County. While cadmium ranks ninth in Boyd County for its
toxicity-weighted emissions, it is not among the highest emitted (ranking 18th).
Arsenic, which has the highest cancer risk approximation for ASKY-M, appears on
neither emissions-based list (ranking 24th for total emissions and 15th for toxicity-
weighted emissions).
•	Benzene, formaldehyde, acetaldehyde, and ethylbenzene are the highest emitted
pollutants with cancer UREs in Carter County, where GLKY is located.
Formaldehyde, benzene, 1,3-butadiene, and naphthalene are the pollutants with the
highest toxicity-weighted emissions (of the pollutants with cancer UREs) for this
county. Nine of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Carter County (all of which are sampled for at GLKY).
•	Formaldehyde has the highest cancer risk approximation for GLKY, and ranks first
for its toxicity-weighted emissions and second for it total emissions in Table 14-8.
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, 5a, and 6). Many
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. Formaldehyde,
naphthalene, and POM Group 2d are the pollutants with the highest toxicity-weighted
14-93

-------
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 and nickel are the only pollutants of interest for BAKY. Arsenic appears on
neither emissions-based list for Henderson County (arsenic ranks 22nd for total
emissions and 13th for toxicity-weighted emissions). Nickel ranks sixth for its
toxicity-weighted emissions but is not among the highest emitted (ranking 11th for
total emissions). Cadmium is another speciated metal that appears among those with
the highest toxicity-weighted emissions for Henderson County (ranking 10th), but it
is not among the highest emitted (ranking 15th). Cadmium did not fail any screens for
BAKY.
Benzene, ethylbenzene, and formaldehyde are the highest emitted pollutants with
cancer UREs in Marshall County, where four of the five Calvert City sites are
located. Benzene, hexavalent chromium, and formaldehyde are the pollutants with the
highest toxicity-weighted emissions (of the pollutants with cancer UREs) for this
county. Eight of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Marshall County.
Marshall County is the only county with 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 (1.26 tpy), and Harrison
County, Texas (1.08 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).
Marshall County is the only county for which vinyl chloride and 1,2-dichloroethane
appear among the pollutants with the highest toxicity-weighted emissions.
Most of the VOC pollutants of interest for the Calvert City sites for which cancer risk
approximations could be calculated appear on both emissions-based lists for Marshall
County. Carbon tetrachloride is an exception, appearing among the highest emitted
but not those with the highest toxicity-weighted emissions. Hexachloro-1,3-
butadiene, 1,1,2-trichloroethane, and />dichlorobenzene, which are pollutants of
interest for at least one of the four Marshall County sites, do not appear on either
emissions-based list.
Arsenic is the only pollutant of interest among the speciated metals sampled for at
CCKY. Arsenic appears on neither emissions-based list for Marshall County (arsenic
ranks 25th for total emissions and 13th for toxicity-weighted emissions).
14-94

-------
•	Benzene, formaldehyde, and acetaldehyde are the highest emitted pollutants with
cancer UREs in Livingston County, where BLKY is located. Formaldehyde, benzene,
and 1,3-butadiene are the pollutants with the highest toxicity-weighted emissions (of
the pollutants with cancer UREs) for this county. Eight of the highest emitted
pollutants also have the highest toxicity-weighted emissions for Livingston County.
•	Few of BLKY's pollutants of interest appear among the pollutants on the emissions-
based lists for Livingston County (only 1,3-butadiene and benzene).
•	Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
cancer UREs in Fayette County, where LEKY is located. Formaldehyde, benzene,
and 1,3-butadiene are the pollutants with the highest toxicity-weighted emissions (of
the pollutants with cancer UREs) for this county. Eight of the highest emitted
pollutants also have the highest toxicity-weighted emissions for Fayette County.
•	Cancer risk approximations could only be calculated for formaldehyde, acetaldehyde,
and arsenic. All three of these pollutant appear among those with the highest toxicity-
weighted emissions, and both carbonyl compounds appear among the highest emitted
(with arsenic ranking 22nd for total emissions).
Observations from Table 14-9 include the following:
•	Among the Kentucky counties with monitoring sites, emissions (for pollutants with
noncancer RfCs) are highest in Marshall County (Calvert City) and Fayette County
(LEKY) and lowest in Livingston County (BLKY).
•	Toluene, benzene, and xylenes are the highest emitted pollutants with noncancer
RfCs in Boyd County. Acrolein, chlorine, and manganese are the pollutants with the
highest toxicity-weighted emissions (of the pollutants with noncancer RfCs) for Boyd
County. Four of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Boyd County.
•	Although acrolein was sampled for at ASKY, this pollutant was excluded from the
pollutants of interest designation, and thus subsequent risk-based screening
evaluations, due to questions about the consistency and reliability of the
measurements, as discussed in Section 3.2. Acrolein does not appear among Boyd
County's highest emitted pollutants.
•	Of the pollutants of interest for which noncancer hazard approximations could be
calculate for ASKY, two (formaldehyde and benzene) also appear on both emissions-
based lists. Acetaldehyde, the pollutant with the second-highest noncancer hazard
approximation for ASKY, is among the highest emitted but not among those with the
highest toxicity-weighted emissions. 1,3-Butadiene, the pollutant with the fourth-
highest noncancer hazard approximation for ASKY, is among those with the highest
toxicity-weighted emissions but is not among the highest emitted.
•	Nonancer hazard approximations could be calculated for all five metal pollutants of
interest for ASKY-M. Manganese, which has the second highest nonancer hazard
14-95

-------
approximation, also has the third highest toxicity-weighted emissions and tenth
highest total emissions for Boyd County. Nickel, lead, and cadmium are also among
the pollutants with the highest toxicity-weighted emissions, although none of these
are among the highest emitted in Boyd County.
Toluene, xylenes, and hexane are the highest emitted pollutants with noncancer RfCs
in Carter County. Acrolein, formaldehyde, and cyanide compounds (gaseous) are the
pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) for Carter County. Five of the highest emitted pollutants also have
the highest toxicity-weighted emissions for Carter County.
Although acrolein was sampled for at GLKY, this pollutant was excluded from the
pollutants of interest designation, and thus subsequent risk-based screening
evaluations, due to questions about the consistency and reliability of the
measurements, as discussed in Section 3.2. Acrolein does not appear among Carter
County's highest emitted pollutants.
Formaldehyde and acetaldehyde have the highest nonancer hazard approximations for
GLKY and appear on both emissions-based lists. Benzene and 1,3-butadiene also
appears on all three lists. Arsenic has the third highest nonancer hazard
approximation for GLKY and is among the pollutants with the highest toxicity-
weighted emissions but is not among the highest emitted in Carter County (its
emissions rank 32nd). Carbon tetrachloride and 1,2-dichloroethane, the remaining
two pollutants of interest for GLKY, appear on neither emissions-based list for Carter
County.
Carbonyl sulfide, toluene, and xylenes are the highest emitted pollutants with
noncancer RfCs in Henderson County. Henderson County is the only county with an
NMP site for which carbonyl sulfide appears among the 10 highest emitted pollutants.
Acrolein, manganese, and nickel are the pollutants with the highest toxicity-weighted
emissions (of the pollutants with noncancer RfCs) for this county. Three of the
highest emitted pollutants also have the highest toxicity-weighted emissions for
Henderson County.
Arsenic and nickel are the pollutants of interest for BAKY. Nickel has the third
highest toxicity-weighted emissions for Henderson County but is not among the
highest emitted (ranking 27th among pollutants with noncancer RfCs). Arsenic
appears on neither emissions-based list (ranking 44th for total emissions and 18th for
toxicity-weighted emissions). Manganese and cadmium, which were sampled for at
BAKY but did not fail any screens, rank second and ninth, respectively for their
toxicity-weighted emissions for Henderson County.
Methanol, xylenes, toluene are the highest emitted pollutants with noncancer RfCs in
Marshall County. Chlorine, acrolein, and 1,3-butadiene are the pollutants with the
highest toxicity-weighted emissions (of the pollutants with noncancer RfCs) for this
county. This is the only county with an NMP site for which acrolein was not the
pollutant with the highest toxicity-weighted emissions. Five of the highest emitted
pollutants also have the highest toxicity-weighted emissions for Marshall County.
14-96

-------
Benzene is the only pollutant of interest for the Calvert City sites to appear on all
three lists. 1,3-Butadiene has the highest nonancer hazard approximation for all four
Calvert City sites located in Marshall County (as well as the one located in Livingston
County). This pollutant has the third highest toxicity-weighted emissions but is not
among the highest emitted (ranking 14th). None of the other VOC pollutants of
interest for the Calvert City sites appear on either emissions-based list for Marshall
County. This is also true for arsenic, the only other pollutant of interest for CCKY.
Toluene, xylenes, and benzene are the highest emitted pollutants with noncancer
RfCs in Livingston County. Acrolein, formaldehyde, and 1,3-butadiene are the
pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) for this county. Five of the highest emitted pollutants also have the
highest toxicity-weighted emissions for Livingston County.
Although acrolein was sampled for at BLKY, this pollutant was excluded from the
pollutants of interest designation, and thus subsequent risk-based screening
evaluations, due to questions about the consistency and reliability of the
measurements, as discussed in Section 3.2. Acrolein does not appear among
Livingston County's highest emitted pollutants.
1,3-Butadiene and benzene have the highest noncancer hazard approximations for
BLKY. These pollutants appear on both emissions-based lists for Livingston County
but are the only pollutants of interest for BLKY to do so.
Toluene, xylenes, and hexane are the highest emitted pollutants with noncancer RfCs
in Fayette County. Acrolein, formaldehyde, and 1,3-butadiene are the pollutants with
the highest toxicity-weighted emissions (of the pollutants with noncancer RfCs) for
this county. Four of the highest emitted pollutants also have the highest toxicity-
weighted emissions for Fayette County.
Although acrolein was sampled for at LEKY, this pollutant was excluded from the
pollutants of interest designation, and thus subsequent risk-based screening
evaluations, due to questions about the consistency and reliability of the
measurements, as discussed in Section 3.2. Acrolein does not appear among Fayette
County's highest emitted pollutants.
Noncancer hazard approximations could only be calculated for formaldehyde,
acetaldehyde, and arsenic. All three of these pollutant appear among those with the
highest toxicity-weighted emissions, and both carbonyl compounds also appear
among the highest emitted (with arsenic ranking 42nd for total emissions).
14-97

-------
14.6 Summary of the 2013 Monitoring Data for the Kentucky Monitoring Sites
Results from several of the data treatments described in this section include the
following:
~~~ Eight monitoring sites sampledfor VOCs; five monitoring sites sampledfor PMio
metals; three monitoring sites sampledfor carbonyl compounds; PAHs and
hexavalent chromium were also sampledfor at GLKY (although hexavalent
chromium sampling was discontinued in June 2013).
~~~ The number ofpollutants failing screens for the Kentucky sites varies from two
(BAKY) to 12 (GLKY, TVKY, and LEKY).
~~~ ASKY-M had the highest annual average concentrations of arsenic and nickel among
NMP sites sampling PMio metals. Three additional Kentucky sites (BAKY, LEKY, and
CCKY) were among the sites with the highest annual average concentrations of
arsenic and BAKY was also among the sites with the highest annual average
concentrations of nickel.
~~~ The maximum benzene concentration measured across the program was measured at
ASKY, which had the fourth highest annual average concentration of benzene among
NMP sites sampling this pollutant.
~~~ Some of the highest concentrations of VOCs were measured at the Calvert City sites,
particularly vinyl chloride, carbon tetrachloride, 1,3-butadiene, and
1,2-dichloroethane.
~~~ The cancer risk approximation 1,2-dichloroethane for TVKY is the second highest
among cancer risk approximations calculated for all site-specific pollutants of
interest.
14-98

-------
15.0 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.
15.1 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 15-1 is a composite satellite
image retrieved from ArcGIS Explorer showing the monitoring site and its immediate
surroundings. Figure 15-2 identifies nearby point source emissions locations by source category,
as reported in the 2011 NEI for point sources, version 2. Note that only sources within 10 miles
of the site are included in the facility counts provided in Figure 15-2. A 10-mile boundary was
chosen to give the reader an indication of which emissions sources and emissions source
categories could potentially have a direct effect on the air quality at the monitoring site. Further,
this boundary provides both the proximity of emissions sources to the monitoring site as well as
the quantity of such sources within a given distance of the site. Sources outside the 10-mile
boundary are still visible on the map for reference, but have been grayed out in order to
emphasize emissions sources within the boundary. Table 15-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates.
15-1

-------
Figure 15-1. Boston, Massachusetts (BOMA) Monitoring Site
alcolm-Xsg^
»Vhjcmi p

\SaurPr U5GS . jt-
SdMfgWi^, hqa.Jjsc
-c ifios MiOMfrCutp
to

-------
Figure 15-2. NEI Point Sources Located Within 10 Miles of BO MA
EiSei
County
Mddtem \
t County
Suite* ,
County jJ \
BOMA NATTS site	10 mile radius	J County boundary
Source Category Group (No. of Facilities)
*
t
*
B
C
Awca*NK**Al««n	ftng Pat**, <11
4 mar I'Ar lArpwt Sufipc'tOpe-flbefW I •*)
A»|*4» PrariuctaitWoi Mm Asphalt |«}
AnAomoWrT'iHk MaruAaouirg Fa*tv i !
Bom Tr.rr»rMU>&uft> PlanK 0i
Cwnui W(Mit(KH«inj' *tm< <||
t>^ C>*nerr> An*nal*4t«min Jll
® Oty ClMnnf Pat wr» <2l
6 E «Klrc« E<)t»p*at'
E f «Sln*il»»nt F«H»b*ig AnoWlJ Mtf COkVMQ
F ci>f*J F*0C*Mr»o'Ag'iatft«'« F«««y C)i
I FotMUM llMl AflUdlMill*>
>	I KMiarq (i I
0 inr>i»»r * Mjriinifj ni bQw LiMlr-g fcj«^v«>y tAdAI«tS«fp>jfrnr»«i»f»»(«' PiaOo^ M*^rwwma f«*i#v «i I
R*i VmRiI law Otariilmu I f»
MaiultUuiiiq 
-------
Table 15-1. Geographical Information for the Massachusetts Monitoring Site
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Additional Ambient Monitoring Information1
BOMA
25-025-0042
Boston
Suffolk
Boston-
Cambridge-
Newton MA-NH
42.3295,
-71.0826
Commercial
Urban/City
Center
CO, VOCs, S02, NO, N02, NOx, NOy, 03, PM10,
PAMS/NMOCs, Carbonyl compounds.
Meteorological parameters. Black carbon PM coarse,
PM2 5, PM2 5 Speciation, IMPROVE Speciation.
1 Data for additional pollutants are reported to AQS for this site (EPA, 2014b); however, these data are not generated by ERG and are therefore not included in this report.
BOLD ITALICS = EPA-designated NATTS Site
-U

-------
The BOMA monitoring site is located at Dudley Square in Roxbury, a southwest
neighborhood of Boston and is the Roxbury NATTS site. The surrounding area is commercial as
well as residential, as shown in Figure 15-1. Immediately to the east of the monitoring site are
town homes, to the north is a parking lot and to the west are commercial properties. The original
purpose for the location of this site was to measure population exposure to a city bus terminal
located another block west of the monitoring site. In recent years, the buses servicing the area
were converted to compressed natural gas (CNG). The monitoring site is 1.3 miles south of 1-90
and 1 mile west of 1-93. As Figure 15-2 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; bulk terminals and bulk plants; and electricity generating units (via combustion).
Sources located within 1 mile of BOMA include several hospitals, a heliport at one of the
hospitals, a university, and a dry cleaning facility. Figure 15-2 shows that BOMA is located less
than 2 miles from the shoreline (Dorchester Bay).
Table 15-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Massachusetts monitoring site. Table 15-2 includes both county-
level population and vehicle registration information. Table 15-2 also contains traffic volume
information for BOMA as well as the location for which the traffic volume was obtained.
Additionally, Table 15-2 presents the county-level daily VMT for Suffolk County.
Table 15-2. Population, Motor Vehicle, and Traffic Information for the Massachusetts
Monitoring Site
Site
County
Estimated
County
Population1
County-level
Vehicle
Registration2
Annual
Average
Daily
Traffic3
Intersection
Used for
Traffic Data
County-
level Daily
VMT4 "
BOMA
Suffolk
755,503
393,252
27,654
Melnea Cass Blvd near
Shawmut Ave
10,963,634
bounty-level population estimate reflects 2013 data (Census Bureau, 2014)
2County-level vehicle registration reflects 2013 data (MA RMV, 2014)
3AADT reflects 2010 data (MA DOT, 2010)
4County-level VMT reflects 2013 data (MA DOT, 2014)
BOLD ITALICS = EPA-designated NATTS Site
15-5

-------
Observations from Table 15-2 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
26th among other counties with NMP sites.
•	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. The VMT for Suffolk County ranks 25th.
15.2 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.
15.2.1	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, 2015).
15.2.2	Meteorological Summary
Hourly meteorological data for 2013 were retrieved from NCDC for the weather station
closest to the Massachusetts monitoring site (NCDC, 2013), as described in Section 3.4.2. 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 15-3. These data were used to determine how
meteorological conditions on sample days vary from conditions experienced throughout the year.
15-6

-------
Table 15-3. Average Meteorological Conditions near the Massachusetts Monitoring Site
Closest Weather
Station
Distance
and

Average
Maximum
Average
Average
Dew Point
Average
Wet Bulb
Average
Relative
Average
Sea Level
Average
Scalar Wind
(WBAN and
Coordinates)
Direction
from Site
Average
Type1
Temperature
(°F)
Temperature
(°F)
Temperature
(°F)
Temperature
(°F)
Humidity
(%)
Pressure
(mb)
Speed
(kt)
Boston, Massachusetts - BOMA
Logan
International
Airport
14739
(42.36, -71.01)
4.3
miles
Sample
Dav
58.5
51.8
39.5
46.3
65.7
1016.7
9.3
(63)
±4.7
±4.4
±4.7
±4.1
±3.8
± 1.8
±0.8
60°
(ENE)
2013
58.9
± 1.9
51.8
± 1.8
39.2
±2.0
46.3
± 1.7
65.1
± 1.6
1016.6
±0.8
9.1
±0.3
1 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 2013. Also included in Table 15-3 is the 95 percent
confidence interval for each parameter. As shown in Table 15-3, average meteorological
conditions on sample days are very similar to conditions experienced throughout 2013. BOMA is
among the windier locations with a NMP site, with an average scalar wind speed around 9 knots.
15.2.3 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 3.4.2. A wind rose shows the frequency of wind directions using "petals"
positioned around a 16-point compass, and uses different colors to represent wind speeds.
Figure 15-3 presents 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 15-3 also presents three different
wind roses for the BOMA monitoring site. First, a historical wind rose representing 2003 to 2012
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 2013 is
presented. Next, a wind rose representing wind data for days on which samples were collected in
2013 is presented. These can be used to identify the predominant wind speed and direction for
2013 and to determine if wind observations on sample days were representative of conditions
experienced over the entire year and historically.
15-8

-------
est:
Figure 15-3. Wind Roses for the Logan International Airport Weather Station near BOM A
Location of BOMA and Weather Station	2003-2012 Historical Wind Rose
WIND SPEED
(Knots)
~
~1 4-7
2- 4
Calms: 2.52%
2013 Wind Rose
WEST
(Knots)
11 -17
SOUTH
WIND SPEED
~ 4-7
Calms: 438%
Sample Day Wind Rose
NORTH"--.,
ES ,
WIND SPEED
(Knots)
~ >=22
~
7- 11
4- 7
Calms: 4.04%
15-9

-------
Observations from Figure 15-3 for BOMA include the following:
•	The Logan International Airport weather station is located 4.3 miles east-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 (those less than or equal to 2 knots)
account for less than 4 percent of wind observations. Winds with a westerly
component (south-southwest to north-northwest) make up the majority (more than
60 percent) of winds greater than 2 knots, with westerly and west-northwesterly
winds observed the most.
•	The wind patterns shown on the 2013 wind rose resemble the historical wind patterns,
indicating that wind conditions during 2013 were typical of conditions experienced
historically near BOMA. Westerly and west-northwesterly winds account for an even
higher percentages of wind observations in 2013.
•	The sample day wind patterns generally resemble the full-year and historical wind
patterns, although the percentage of westerly, west-northwesterly, and northwesterly
winds observed was more similar, with each direction accounting for between
9 percent and 11 percent of observations. The percentage of easterly winds was also
higher on sample days while the number of northerly wind observations was lower.
15.3 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 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. PMio metals, PAHs, and hexavalent chromium were sampled for at BOMA.
15-10

-------
Table 15-4. Risk-Based Screening Results for the Massachusetts Monitoring Site
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
Boston, Massachusetts - BOMA
Naphthalene
0.029
54
61
88.52
46.96
46.96
Arsenic (PMio)
0.00023
47
61
77.05
40.87
87.83
Nickel (PMio)
0.0021
9
61
14.75
7.83
95.65
Hexavalent Chromium
0.000083
2
14
14.29
1.74
97.39
Acenaphthene
0.011
1
61
1.64
0.87
98.26
Benzo(a)pyrene
0.00057
1
61
1.64
0.87
99.13
Fluorene
0.011
1
57
1.75
0.87
100.00
Total
115
376
30.59

Observations from Table 15-4 include the following:
•	Seven pollutants failed at least one screen for BOMA; approximately 31 percent of
concentrations for these seven pollutants were greater than their associated risk
screening value (or failed screens).
•	Most of the pollutants that failed screens were detected in all or most of the valid
samples collected at BOMA, hexavalent chromium being the exception. However,
hexavalent chromium sampling was discontinued at BOMA at the end of June 2013.
•	Three pollutants contributed to 95 percent of failed screens for BOMA and therefore
were identified as pollutants of interest for this site. These include two PMio metals
(arsenic and nickel) and one PAH (naphthalene).
•	Naphthalene and arsenic each account for more than 40 percent of the total failed
screens for BOMA while nickel accounts for just less than 8 percent of failed screens.
15.4 Concentrations
This section presents various concentration averages used to characterize air toxics
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
each monitoring site.
•	Annual concentration averages are presented graphically to illustrate how each site's
concentrations compare to the program-level averages, as presented in Section 4.1.
•	Concentration averages and other statistical metrics are presented from previous years
of sampling in order to characterize concentration trends at the site.
15-11

-------
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at BOMA are provided in Appendices M through O.
15.4.1 2013 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 15-5,
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 15-5. 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 (PMio)
61/61
0.39
±0.13
0.50
±0.20
0.58
±0.16
0.46
±0.18
0.48
±0.08
Naphthalene
61/61
44.21
± 13.33
54.69
± 16.97
57.50
±9.87
60.48
± 17.09
54.32
±7.09
Nickel (PMio)
61/61
1.23
±0.21
1.25
±0.31
1.80
±0.65
1.40
±0.57
1.42
±0.23
Observations for BOMA from Table 15-5 include the following:
• Naphthalene is the pollutant with the highest annual average concentration
(54.32 ± 7.09 ng/m3). The annual average concentrations for the remaining pollutants
of interest are at least an order of magnitude lower.
15-12

-------
•	Concentrations of naphthalene measured at BOMA range from 19 ng/m3 to
164 ng/m3. Concentrations tended to be lowest in the first quarter and highest in the
fourth quarter, based on the quarterly averages, although the differences are not
statistically significant. Four naphthalene concentrations greater than 100 ng/m3 were
measured at BOMA: one in January (101 ng/m3), one in May (157 ng/m3), and two in
October (103 ng/m3 and 164 ng/m3).
•	Concentrations of arsenic measured at BOMA range from 0.06 ng/m3 to 1.41 ng/m3.
The maximum concentration of arsenic was measured on the same day as the
maximum naphthalene concentration, October 31, 2013. The quarterly average
concentrations do not vary significantly across the calendar quarters. Four arsenic
concentrations greater than 1 ng/m3 were measured at BOMA: one in April
(1.38 ng/m3), one in May (1.32 ng/m3), one in September (1.16 ng/m3), and one in
October (1.41 ng/m3).
•	Concentrations of nickel measured at BOMA range from 0.43 ng/m3 to 5.26 ng/m3.
The third and fourth quarterly averages have more variability associated with their
individual measurements, as their confidence intervals are roughly twice the
confidence intervals for the first and second quarterly averages. The five highest
concentrations (those greater than 2.5 ng/m3) were measured at BOMA between July
and December and of the 10 measurements greater than 2 ng/m3, seven were
measured during the second half of the year.
•	Table 4-12 present the NMP sites with the 10 highest annual average concentrations
for each of the program4evel speciated metals pollutants of interest. This table shows
that BOMA has the fifth highest annual average concentrations of nickel among NMP
sites sampling PMio metals.
15.4.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for the pollutants shaded in
gray in Table 15-4 for BOMA. Figures 15-4 through 15-6 overlay the site's minimum, annual
average, and maximum concentrations onto the program-level minimum, first quartile, median,
average, third quartile, and maximum concentrations, as described in Section 3.4.3.1.
15-13

-------
Figure 15-4. Program vs. Site-Specific Average Arsenic (PMio) Concentration
W
0
12 3
4 5 6
Concentration {ng/m3)
7
8

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i

Site: Site Average
o
Site Concentration Range


Figure 15-5. Program vs. Site-Specific Average Naphthalene Concentration
400	500
Concentration {ng/m3)
Program: 1st Qua rti le
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



Figure 15-6. Program vs. Site-Specific Average Nickel (PMio) Concentration
E
10	15
Concentration {ng/m3)
Program: IstQuartile
2ndQuartile 3rdQuartile
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



Observations from Figures 15-4 through 15-6 include the following:
• Figure 15-4 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.
15-14

-------
•	Figure 15-5 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. There were no non-detects of naphthalene
measured at BOMA or across the program.
•	Figure 15-6 is the box plot for nickel (PMio). This box plot shows that BOMA's
annual average concentration of nickel is greater than the program-level average
concentration as well as the program-level third quartile. The minimum nickel
concentration measured at BOMA is greater than the program-level first quartile.
Although the maximum nickel concentration measured at BOMA is about one-
fourth the magnitude of the maximum nickel concentration measured across the
program, it is among the higher measurements (ranking 10th highest).
15.4.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.3.2.
BOMA has sampled PMio metals under the NMP since 2003 and PAHs since 2008. Thus,
Figures 15-7 through 15-9 present the 1-year statistical metrics for each of the pollutants of
interest for BOMA. The statistical metrics presented for assessing trends include the substitution
of zeros for non-detects. If sampling began mid-year, a minimum of 6 months of sampling is
required for inclusion in the trends analysis; in these cases, a 1-year average concentration is not
provided, although the range and percentiles are still presented.
15-15

-------
Figure 15-7. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at BOMA
20041	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
0 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because there were breaks in sampling during portions of 2004.
Observations from Figure 15-7 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 concentration is not presented, although the range of measurements is
provided.
•	The maximum arsenic concentration shown was measured on July 5, 2008
(5.45 ng/m3). The next highest concentration measured is approximately half as high
(2.52 ng/m3) and was measured on July 4, 2006.
•	The 1-year average concentrations of arsenic have fluctuated over the years, ranging
from 0.36 ng/m3 (2010) to 0.61 ng/m3 (2008). For 2008, the maximum concentration
is driving the 1-year average upward, which is evident from the median
concentration, which hardly changed between 2007 and 2008, even though the
smallest range of measurements was collected in 2007. If the maximum concentration
for 2008 was removed from the dataset, the 1-year average concentration for 2008
would fall from 0.61 ng/m3 to 0.53 ng/m3, making the changes in the 1-year averages
between 2007 and 2009 more subtle.
15-16

-------
•	All of the statistical metrics exhibit a decrease from 2008 to 2009 and again for 2010.
Conversely, all of the statistical metrics exhibit an increase from 2010 to 2011 and
again for 2012.
•	For 2013, a higher number of concentrations at the lower end of the concentration
range were measured while concentrations at the top of the range changed little. The
number of arsenic concentrations less than 0.2 ng/m3 increased from one in 2012 to
13 for 2013. This is explains the considerable decrease in the minimum, 5th
percentile, and median concentration shown for 2013, as well as the slight decrease in
the 1-year average concentration, although the change is not statistically significant.
Figure 15-8. Yearly Statistical Metrics for Naphthalene Concentrations Measured at BOMA
0 +-
2008 1
2009
2010
2011
Year
2012
2013

O 5th Percentile
- Minimum
~ Median
~ Maximum
O 95th Percentile
~~•«~••• Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2008.
Observations from Figure 15-8 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 concentration 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 also collected in 2012. Only two
additional concentrations greater than 200 ng/m3 have been measured at BOMA (one
each in 2008 and 2009).
15-17

-------
•	The difference between the 5th and 95th percentiles (the range of concentrations
within which 90 percent of the measurements lie) decreased each year through 2011.
The range increased somewhat for 2012, and is more similar to the range shown for
2010, before decreasing further for 2013.
•	The median concentration decreased significantly from 2008 to 2009, from
84.00 ng/m3 to 56.30 ng/m3. Little change is shown after 2008, with the median
varying by only 11 ng/m3 between 2009 and 2013. Similarly, the 1-year average
concentration varies by only 16 ng/m3 for the years shown, ranging from 54.32 ng/m3
for 2013 to 70.33 ng/m3 for 2009. Both statistical parameters are at a minimum for
2013.
Figure 15-9. Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at BOMA
2004 1	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because there were breaks in sampling during portions of 2004.
Observations from Figure 15-9 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 12 highest nickel concentrations (those greater than 7.50 ng/m3) were
measured in 2004 or 2005 (with the other measured in 2012).
•	A steady decreasing trend in the nickel measurements collected at BOMA is shown
through 2010. Concentrations for 2011 increased just slightly, returning to 2009
levels. Even with the higher concentrations measured in 2012 and 2013, the 1-year
15-18

-------
average concentration did not change significantly from 2011 (ranging from
1.38 ng/m3 for 2011 to 1.42 ng/m3 for 2013).
15.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at the BOMA monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4 for
definitions and explanations regarding the various toxicity factors, time frames, and calculations
associated with these risk-based screenings.
15.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for BOMA and where annual average concentrations could
be calculated, risk was examined by calculating cancer risk and noncancer hazard
approximations. These approximations can be used as risk estimates for cancer and noncancer
effects attributable to the pollutants of interest. Although the use of these approximations is
limited, they may help identify where policy-makers want to shift their air-monitoring priorities.
Refer to Section 3.4.3.3 for an explanation of how cancer risk and noncancer hazard
approximations are calculated and what limitations are associated with them. Annual averages,
cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard approximations are
presented in Table 15-6, where applicable. Cancer risk approximations are presented as
probabilities while the noncancer hazard approximations are ratios and thus, unitless values.
Table 15-6. Risk Approximations for the Massachusetts Monitoring Site
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Boston, Massachusetts - BOMA
Arsenic (PMio)
0.0043
0.000015
61/61
0.48
±0.08
2.06
0.03
Naphthalene
0.000034
0.003
61/61
54.32
±7.09
1.85
0.02
Nickel (PMio)
0.00048
0.00009
61/61
1.42
±0.23
0.68
0.02
Observations for BOMA from Table 15-6 include the following:
• Among the pollutants of interest for BOMA, naphthalene has the highest annual
average concentration while arsenic has the lowest annual average concentration.
15-19

-------
•	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 (2.06 in-a-million for
arsenic and 1.85 in-a-million for naphthalene). This speaks to the relative toxicity of
one pollutant compared to the other.
•	None of the pollutants of interest for BOMA have noncancer hazard approximations
greater than 1.0; in fact, none of the pollutants of interest have noncancer hazard
approximations greater than 0.05. This indicates that no adverse noncancer health
effects are expected due to these individual pollutants.
15.5.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 15-7 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
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 3.4.3.4. Lastly,
Table 15-7 provides the pollutants with the highest cancer risk approximations (in-a-million) for
BOMA, 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 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 15.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
15-20

-------
Table 15-7. 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)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Boston, Massachusetts (Suffolk County) - BOMA
Formaldehyde
143.05
Formaldehyde
1.86E-03
Arsenic
2.06
Benzene
137.55
Nickel, PM
1.22E-03
Naphthalene
1.85
Acetaldehyde
66.22
Benzene
1.07E-03
Nickel
0.68
Ethylbenzene
64.30
1,3-Butadiene
7.34E-04

1.3 -Butadiene
24.47
Arsenic, PM
4.62E-04
T etrachloroethylene
19.26
Hexavalent Chromium
4.41E-04
Naphthalene
10.82
Naphthalene
3.68E-04
POM, Group 2b
3.41
POM, Group 2b
3.00E-04
Nickel, PM
2.53
Ethylbenzene
1.61E-04
POM, Group 2d
1.66
POM, Group 2d
1.46E-04

-------
Table 15-8. 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)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Boston, Massachusetts (Suffolk County) - BOMA
Toluene
511.53
Acrolein
501,247.48
Arsenic
0.03
Hexane
399.62
Nickel, PM
28,149.17
Naphthalene
0.02
Xylenes
271.73
Formaldehyde
14,596.91
Nickel
0.02
Formaldehyde
143.05
1.3 -Butadiene
12,234.00

Benzene
137.55
Acetaldehyde
7,357.30
Acetaldehyde
66.22
Arsenic, PM
7,162.73
Ethylbenzene
64.30
Benzene
4,584.90
Methyl isobutyl ketone
55.81
Naphthalene
3,605.47
1.3 -Butadiene
24.47
Cadmium, PM
3,035.10
T etrachloroethylene
19.26
Xylenes
2,717.32

-------
Observations from Table 15-7 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, nickel, and benzene.
•	Eight of the highest emitted pollutants also have the highest toxicity-weighted
emissions.
•	All three of BOMA's pollutants of interest appear among the pollutants with the
highest toxicity-weighted emissions for Suffolk County. Nickel and naphthalene are
also among those with the highest total emissions in Suffolk County while arsenic is
not among the highest emitted (it ranks 16th).
•	POM, Group 2b ranks eighth for both quantity emitted and its toxicity-weighted
emissions. POM, Group 2b includes several PAHs sampled for at BOMA including
acenaphthene and fluorene, both of which failed a single screen but were not
identified as pollutants of interest. POM, Group 2d ranks tenth for both quantity
emitted and its toxicity-weighted emissions. POM, Group 2d includes several PAHs
sampled for at BOMA, including anthracene and phenanthrene, although none of
these failed screens for BOMA.
Observations from Table 15-8 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, nickel, and formaldehyde.
•	Five of the highest emitted pollutants also have the highest toxicity-weighted
emissions.
•	All three of BOMA's pollutants of interest appear among the pollutants with the
highest toxicity-weighted emissions for Suffolk County, although none of these
appear among the highest emitted pollutants. Cadmium, which was also sampled for
at BOMA but did not fail any screens, also appears among the pollutants with the
highest toxicity-weighted emissions for Suffolk County
15.6 Summary of the 2013 Monitoring Data for BOMA
Results from several of the data treatments described in this section include the
following:
~~~ Seven pollutants failed screens for BOMA, with naphthalene and arsenic accounting
for a majority of the failed screens.
15-23

-------
~~~ Naphthalene had the highest annual average concentration among the pollutants of
interest for BOMA.
~~~ Even though concentrations of nickel have a decreasing trend over most of the years
of sampling, BOMA has the fifth highest annual average concentration of nickel for
2013 among NMP sites sampling PMio metals.
15-24

-------
16.0	Site in Michigan
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS site in Michigan, and integrates these concentrations
with emissions, meteorological, and risk information. Data generated by sources other than ERG
are not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) for detailed discussions and definitions
regarding the various data analyses presented below.
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 DEMI monitoring site is located in the Detroit-Warren-Dearborn, Michigan CBSA.
Figure 16-1 is the 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, version 2.
Note that only sources within 10 miles of the site are included in the facility counts provided in
Figure 16-2. A 10-mile boundary was chosen to give the reader an indication of which emissions
sources and emissions source categories could potentially have a direct effect on the air quality at
the monitoring site. Further, this boundary provides both the proximity of emissions sources to
the monitoring site as well as the quantity of such sources within a given distance of the site.
Sources outside the 10-mile boundary are still visible on the map for reference, but have been
grayed out in order to emphasize the emissions sources within the boundary. Table 16-1 provides
supplemental geographical information such as land use, location setting, and locational
coordinates.
16-1

-------
Figure 16-1. Dearborn, Michigan (DEMI) Monitoring Site

-------
Figure 16-2. NEI Point Sources Located Within 10 Miles of DEMI
Macomb 1
County 1
County
CANADA
DEMI NATTS site	10 mite radius	County boundary
C-50tf-W
	A	
Source Category Group (No. of Facilities)
f	Airport/Airline.'Airport Support Operators (10)
it	Asphalt P'oducllwvHot Mm Asphalt Plant (4)
M Automobila/Truck Manufacturing Facility (5)
B	But* TermmalsiRudi Plants (101
C	Chemical Manufacturing Facility (5)
i	Compressor Station (2)
6	Electrwal Equipment Manufacturing Facility (1|
*	Electricity Generation via Combustion (8)
E	Electroplating. Plating Poffihmg, Anodizing, and Coloring |4I
F	Food Processing/Agriculture Facility (2)
I	Foundnes. Iron and Steer (1)
*	industrial Machinery of Equipment Plant (2)
o	institutional (school hospital, pnson elc.l (9)
*	Metal Can Qoj and CMher Metal Contwiet Manufacturing (1)
A	Metal Coabng, Engraving, and Allied Services to Manufacturers
®	Melsis Processing/Fabrication Facility (4)
».	MineiOuarryMmerai Processing FaciMy (101
?	Miscellaneous Commercial.'lndustrial Facility (18|
0	Munlopal Waste Comixistor (1)
Q	Paint and Coaling Manutactunng Facility (2)
i*	Petroleum Products Manufacturing (1)
a	Petroleum Refinery (1)
-	Pharmaceutical Manufaclunng (1)
R	Plasbc Resin or Rubber Products Plant (3|
P	PnnlingfPublishlng'Paper Product Manufaclunng Fac»'vr	wts'av*
Legend
ss^^rw	aj'oqv.	a'&irv*
N Jto 3-jfi to tactffy dc-nsit, and ccllocafeon tha total raciimas
ditpas -ray not reppBMrri All facias	the area of ntenp*»t
Lake
St CHiir
16-3

-------
Table 16-1. Geographical Information for the Michigan Monitoring Site
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Additional Ambient Monitoring Information1
DEMI
26-163-0033
Dearborn
Wayne
Detroit-Warren-
Dearborn, MI
42.306666,
-83.148889
Industrial
Suburban
TSP Metals, Meteorological parameters, PMio, PMio
Speciation, PM2 5, PM2.5 Speciation, IMPROVE
Speciation, Black carbon.
1 Data for additional pollutants are reported to AQS for this site (EPA, 2014b); however, these data are not generated by ERG and are therefore not included in this report.
BOLD ITALICS = EPA-designated NATTS Site
On

-------
DEMI is located in the parking lot of Salina Elementary School in Dearborn, just
southwest of Detroit, and is the Detroit NATTS site. The surrounding area is both suburban and
industrial in nature. Figure 16-1 shows that a freight yard is located just west of the site and a
residential neighborhood is located to the east. Industrial sources such as automobile and steel
manufacturing facilities are also located in the vicinity. The monitoring site lies between two
heavily traveled roadways, 1-75 (1.4 miles to the east) and 1-94 (1.2 miles to the west).
Figure 16-2 shows that DEMI is surrounded by numerous point sources. A cluster of
sources is located just west of DEMI. Another cluster of sources is located farther south. The
source categories with the most point sources within 10 miles of the site include the airport
source category, which includes airports and related operations as well as small runways and
heliports, such as those associated with hospitals or television stations; bulk terminals and bulk
plants; mines, quarries, and mineral processing facilities; and institutional facilities (schools,
prisons, and/or hospitals). Although difficult to discern in Figure 16-2, the closest sources to
DEMI are just west of the site and include a steel mill, an automobile/truck manufacturing
facility, a facility generating electricity via combustion, a metal coatings facility, and a rail yard.
Note that DEMI is located approximately 3 miles from the Canadian border, and that no
emission sources information is provided for Canada.
Table 16-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Michigan monitoring site. Table 16-2 includes both county-level
population and vehicle registration information. Table 16-2 also contains traffic volume
information for DEMI as well as the location for which the traffic volume was obtained.
Additionally, Table 16-2 presents the county-level daily VMT for Wayne County.
Table 16-2. Population, Motor Vehicle, and Traffic Information for the Michigan
Monitoring Site




Annual




Estimated
County-level
Average
Intersection
County-


County
Vehicle
Daily
Used for
level Daily
Site
County
Population1
Registration2
Traffic3
Traffic Data
VMT4
DEMI
Wayne
1,775,273
1,335,516
94,600
1-94 from Ford Plant to Rotunda Dr
41,554,962
bounty-level population estimate reflects 2013 data (Census Bureau, 2014)
2County-level vehicle registration reflects 2013 data (MDS, 2014)
3AADT reflects 2013 data (MI DOT, 2013)
4County-level VMT reflects 2013 data (MI DOT, 2014)
BOLD ITALICS = EPA-designated NATTS Site
16-5

-------
Observations from Table 16-2 include the following:
•	Wayne County's population and vehicle registration both rank eighth highest among
counties with NMP sites.
•	The traffic volume near DEMI ranks 16th among NMP sites. Traffic for DEMI is
provided for 1-94, between the Ford Plant and Rotunda Drive.
•	The Wayne County daily VMT is the sixth highest VMT compared to other counties
with NMP sites.
16.2 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
site in Michigan on sample days, as well as over the course of the year.
16.2.1	Climate Summary
Detroit is located in southeast Michigan, where the Detroit River serves as the
U.S./Canadian border, and is situated directly across from Windsor, Canada's southernmost city.
The river separates the two cities and is a channel between Lake St. Clair to the east and Lake
Erie to the south. Detroit is located in a region of active weather. Winters tend to be cold and
wet, with snowfall amounts around 35 inches per year. Summers are generally mild, although
temperatures exceeding 90°F are common. 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 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, 2015a and 2015b).
16.2.2	Meteorological Summary
Hourly meteorological data for 2013 were retrieved from NCDC for the weather station
closest to the Michigan monitoring site (NCDC, 2013), as described in Section 3.4.2. The closest
weather station to DEMI is located at Detroit City Airport (WBAN 14822). Additional
information about this 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 Michigan Monitoring Site
Closest Weather
Station
Distance
and

Average
Maximum
Average
Average
Dew Point
Average
Wet Bulb
Average
Relative
Average
Sea Level
Average
Scalar
Wind
(WBAN and
Coordinates)
Direction
from Site
Average
Type1
Temperature
(°F)
Temperature
(°F)
Temperature
(°F)
Temperature
(°F)
Humidity
(%)
Pressure
(mb)
Speed
(kt)
Dearborn, Michigan - DEMI
Detroit City Airport
14822
(42.41, -83.01)
10.0
miles
Sample
Day
(68)
55.9
±5.3
48.9
±4.9
38.3
±4.9
44.0
±4.5
69.0
±3.1
1017.5
± 1.7
7.2
±0.7
45°
(NE)

57.4
50.0
39.3
45.0
68.8
1017.3
6.8

2013
±2.1
±2.0
± 1.9
± 1.8
± 1.1
±0.7
±0.3
1 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 2013. Also included in Table 16-3 is the 95 percent
confidence interval for each parameter. Average meteorological conditions on sample days near
DEMI appear slightly cooler than conditions experienced throughout the year, although the
difference is not statistically significant. A number of make-up samples were collected at DEMI
throughout the year, most of which were collected during cooler parts of the year (one each in
March, April, September, October, and November, and two in December).
16.2.3 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 3.4.2. A wind rose shows the frequency of wind directions using "petals" positioned
around a 16-point compass, and uses different colors to represent wind speeds.
Figure 16-3 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 16-3 also presents three different wind roses for the
DEMI monitoring site. First, a historical wind rose representing 2003 to 2012 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 2013 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2013 is
presented. These can be used to identify the predominant wind speed and direction for 2013 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.
16-8

-------
Figure 16-3. Wind Roses for the Detroit City Airport Weather Station near DEMI
Location of DEMI and Weather Station	2003-2012 Historical Wind Rose
WIND SPEED
(Knots)
~ >=22
~
7- 11
4- 7
2013 Wind Rose
NORTH"''-
' ES ,
WIND SPEED
(Knots)
17-21
SOUTH
Calms: 11.58%
Sample Day Wind Rose
EST
WW C SPEED
i, Knots)
SOUTH
16-9

-------
Observations from Figure 16-3 include the following:
•	The weather station at Detroit City Airport is located 10 miles to the northeast of
DEMI. Most of the city of Detroit lies between the weather station and the monitoring
site.
•	The historical wind roses show that winds from a variety of directions were observed
near DEMI, 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 (those less than or equal to 2 knots) were observed for
approximately 11 percent of the hourly measurements.
•	The wind patterns on the 2013 wind rose resemble the historical wind patterns,
although there was a higher percentage of wind observations from the south-
southwest and west.
•	The sample day wind rose for DEMI bears some resemblance to the full-year wind
rose, although there are also differences. Winds from the west-southwest account for
an even greater number of observations on sample days while the number of northerly
wind observations was down considerably. The percentage of calm winds is lower on
the sample day wind rose, accounting for less than 9 percent of the hourly
measurements.
16.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for DEMI in
order to identify site-specific "pollutants of interest," which allows analysts and readers to focus
on a subset of pollutants through the context of risk. Each pollutant's preprocessed daily
measurement was compared to its associated risk screening value. If the concentration was
greater than the risk screening value, then the concentration "failed the screen." The site-specific
results of this risk-based screening process are presented in Table 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. VOCs,
carbonyl compounds, PAHs, and hexavalent chromium were sampled for at DEMI. Note that
hexavalent chromium sampling was discontinued at DEMI at the end of June 2013.
16-10

-------
Table 16-4. Risk-Based Screening Results for the Michigan Monitoring Site
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Dearborn, Michigan - DEMI
Benzene
0.13
62
62
100.00
12.58
12.58
Carbon Tetrachloride
0.17
62
62
100.00
12.58
25.15
Acetaldehyde
0.45
61
61
100.00
12.37
37.53
Formaldehyde
0.077
61
61
100.00
12.37
49.90
Naphthalene
0.029
60
60
100.00
12.17
62.07
1.3 -Butadiene
0.03
58
58
100.00
11.76
73.83
1,2-Dichloroethane
0.038
57
57
100.00
11.56
85.40
Ethylbenzene
0.4
19
62
30.65
3.85
89.25
Acenaphthene
0.011
18
60
30.00
3.65
92.90
Fluorene
0.011
17
59
28.81
3.45
96.35
Fluoranthene
0.011
6
60
10.00
1.22
97.57
Hexavalent Chromium
0.000083
4
22
18.18
0.81
98.38
Hexachloro-1,3 -butadiene
0.045
3
3
100.00
0.61
98.99
Dichloromethane
60
1
62
1.61
0.20
99.19
Propionaldehyde
0.8
1
61
1.64
0.20
99.39
Styrene
100
1
62
1.61
0.20
99.59
T etrachloroethylene
3.8
1
62
1.61
0.20
99.80
T richloroethylene
0.2
1
8
12.50
0.20
100.00
Total
493
942
52.34

Observations from Table 16-4 for DEMI include the following:
•	Eighteen pollutants failed at least one screen for DEMI; greater than 50 percent of
concentrations for these 18 pollutants were greater than their associated risk screening
value (or failed screens).
•	Ten pollutants contributed to 95 percent of failed screens for DEMI and therefore
were identified as pollutants of interest for DEMI. These 10 include two carbonyl
compounds, five VOCs, and three PAHs.
•	The first seven pollutants listed in Table 16-4 each failed 100 percent of screens, with
each contributing to roughly 12 percent to the total number of failed screens; together
these seven pollutants account for more than 85 percent of the total failed screens.
•	Four VOCs listed in Table 16-4 failed a single screen. The concentrations of each of
these pollutants that failed the screen were measured on the same day, October 15,
2013. The highest concentrations of 1,3-butadiene and 1,2-dichloroethane were also
measured on this day at DEMI.
16-11

-------
16.4 Concentrations
This section presents various concentration averages used to characterize air toxics
pollution levels at the Michigan monitoring site. Where applicable, the following calculations
and data analyses were performed for each of the site-specific pollutants of interest:
•	Time period-based concentration averages (quarterly and annual) are provided for
each monitoring site.
•	Annual concentration averages are presented graphically for each site to illustrate
how the site's concentrations compare to the program-level averages, as presented in
Section 4.1.
•	Concentration averages and other statistical metrics are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each data analysis is performed where the data meet the applicable criteria specified in
the appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at DEMI are provided in Appendices J, L, M, and O.
16.4.1 2013 Concentration Averages
Quarterly and annual concentration averages were calculated for the pollutants of interest
for the Michigan site, as described in Section 3.1. The quarterly average 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 site are presented in Table 16-5, where applicable. Note that concentrations of the
PAHs are presented in ng/m3 for ease of viewing. Also note that if a pollutant was not detected in
a given calendar quarter, the quarterly average simply reflects "0" because only zeros substituted
for non-detects were factored into the quarterly average concentration.
16-12

-------
Table 16-5. Quarterly and Annual Average Concentrations of the Pollutants of
Interest for the Michigan Monitoring Site

# of






Measured
1st
2nd
3rd
4th


Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs. # of
Samples
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Dearborn, Michigan - DEMI


1.77
2.01
1.85
1.43
1.76
Acetaldehyde
61/61
±0.26
±0.22
±0.25
±0.21
±0.12


0.83
0.58
0.65
0.55
0.65
Benzene
62/62
±0.11
±0.09
±0.14
±0.11
±0.06


0.09
0.06
0.08
0.08
0.08
1.3 -Butadiene
58/62
±0.03
±0.02
±0.02
±0.03
±0.01


0.65
0.72
0.68
0.64
0.67
Carbon Tetrachloride
62/62
±0.05
±0.06
±0.04
±0.02
±0.02


0.08
0.09
0.06
0.07
0.08
1,2-Dichloroethane
57/62
±0.01
±0.01
±0.02
±0.02
±0.01


0.31
0.26
0.68
0.31
0.39
Ethylbenzene
62/62
±0.11
±0.06
±0.36
±0.13
±0.10


2.38
3.44
4.22
2.20
3.05
Formaldehyde
61/61
±0.32
±0.57
±0.75
±0.22
±0.31


2.83
14.38
17.47
3.80
9.62
Acenaphthene3
60/60
±0.97
±6.76
±5.90
± 1.67
±2.72


2.95
12.36
14.49
3.58
8.35
Fluorene3
59/60
±0.91
±5.05
±4.75
± 1.17
±2.13


99.17
114.16
137.51
67.45
104.57
Naphthalene3
60/60
±39.51
± 24.02
± 27.64
± 14.28
± 14.63
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
Observations for DEMI from Table 16-5 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 greater than
the other quarterly averages, supporting the seasonal trend identified in Section 4.4.2.
A review of the data shows that concentrations of formaldehyde measured at DEMI
range from 1.42 |ig/m3 to 7.22 |ig/m3, with the three highest concentrations of
formaldehyde measured in August 2013. The 15 highest concentrations measured at
DEMI (those greater than 3.5 |ig/m3) were measured between May and September.
Conversely, all but two of the 13 formaldehyde concentrations less than 2 |ig/m3 were
measured during the first or fourth quarters of 2013. Concentrations of acetaldehyde
measured at DEMI do not exhibit the same seasonal tendency as formaldehyde.
•	The third quarter average concentration of ethylbenzene is roughly twice the other
quarterly averages and 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 August 2, 2013 (2.32 |ig/m3), with the second highest concentration
measured on the previous sample day (1.88 |ig/m3, measured on July 27, 2013). All
other concentrations measured were less than 1 |ig/m3. The maximum ethylbenzene
16-13

-------
concentration measured at DEMI is also the highest ethylbenzene concentration
measured among NMP sites sampling this pollutant. Only five NMP sites measured
concentrations of ethylbenzene greater than 2 |ig/m3.
•	Of the PAHs, naphthalene has the highest annual average concentration for DEMI.
Naphthalene concentrations appear to be highest during the warmer months, based on
the quarterly average concentrations of naphthalene, although all four have relatively
large confidence intervals associated with them, indicating that the measurements are
highly variable. Note, however, the confidence interval is highest for the first quarter
average. The maximum concentration of naphthalene (314 ng/m3) was measured on
January 10, 2013, with the other three concentrations greater than 200 ng/m3
measured during the second and third quarters of 2013. At least one concentration
greater than 100 ng/m3 was measured during each quarter of 2013: four during the
first quarter, nine during the second, 11 during the third, and two during the fourth.
•	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 two
pollutants were measured on the same day, June 9, 2013. 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 and 2012
NMP reports.
Tables 4-9 through 4-12 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for DEMI from
those tables include the following:
•	DEMI appears in Table 4-9 for VOCs only once, having the sixth highest annual
average concentration of carbon tetrachloride. However, with the exception of two
sites (BLKY and TVKY), and the difference among the annual average
concentrations of this pollutant varies little.
•	DEMI does not appear in Table 4-10 among the NMP sites with the highest annual
average concentrations of acetaldehyde and formaldehyde.
•	The annual average concentration of acenaphthene for DEMI is the third highest
among NMP sites sampling PAHs, as shown in Table 4-11. DEMI's annual average
concentration of naphthalene ranks fifth among NMP sites sampling PAHs.
16.4.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for the pollutants shaded in
16-14

-------
gray in Table 16-4. Figures 16-4 through 16-13 overlay the Michigan site's minimum, annual
average, and maximum concentrations onto the program-level minimum, first quartile, median,
average, third quartile, and maximum concentrations, as described in Section 3.4.3.1.
Figure 16-4. Program vs. Site-Specific Average Acenaphthene Concentration



Pi 1
Program Max Concentration = 123 ng/m3
,
U 1


0	10	20	30	40	50	60	70	80
Concentration {ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



Figure 16-5. Program vs. Site-Specific Average Acetaldehyde Concentration
0
3
6 9
Concentration {[jg/m3)

12

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i

Site: Site Average
o
Site Concentration Range


Figure 16-6. Program vs. Site-Specific Average Benzene Concentration
¦+-
Program Max Concentration = 43.5 ^ig/m3
6
Concentration {[jg/m3]
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range


16-15

-------
Figure 16-7. Program vs. Site-Specific Average 1,3-Butadiene Concentration

,
O
0

Program Max Concentration = 21.5 ^ig/m3




0	0.3	0.6	0.9	1.2	1.5
Concentration {[ig/m3)
Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



Figure 16-8. Program vs. Site-Specific Average Carbon Tetrachloride Concentration







1
Program Max Concentration = 23.7 ^ig/m3


1




0	0.25	0.5	0.75	1	1.25	1.5	1.75	2
Concentration {[ig/m3)
Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



Figure 16-9. Program vs. Site-Specific Average 1,2-Dichloroethane Concentration



¦
<-
Program Max Concentration = 111 ^ig/m3



0	0.2	0.4	0.6	0.8	1
Concentration {[ig/m3)
Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



16-16

-------
Figure 16-10. Program vs. Site-Specific Average Ethylbenzene Concentration














Program Max Concentration = 18.7 ^ig/m3
1

J










0	1	2	3	4	5	6
Concentration {[ig/m3)
Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



Figure 16-11. Program vs. Site-Specific Average Fluorene Concentration
0	10	20	30	40	50	60	70	80	90	100
Concentration {ng/m3)
Program: IstQuartile
2ndQuartile 3rdQuartile
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



Figure 16-12. Program vs. Site-Specific Average Formaldehyde Concentration

0
3 6
9 12 15
Concentration {[jg/m3)
18
21
24

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i


Site: Site Average
o
Site Concentration Range



16-17

-------
Figure 16-13. Program vs. Site-Specific Average Naphthalene Concentration
I
0
100
200
300 400 500
Concentration (ng/m3)
600
700

Program:
1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i

Site:
Site Average
o
Site Concentration Range


Observations from Figures 16-4 through 16-13 include the following:
•	Figure 16-4 is the box plot for acenaphthene for DEMI. Note that the program-
level maximum concentration (123 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
80 ng/m3. This box plot shows that the maximum acenaphthene concentration
measured at DEMI is considerably less than the maximum concentration across
the program. The annual average acenaphthene concentration for DEMI
(9.62 ± 2.72 ng/m3) is nearly than twice the program-level average concentration
(4.88 ng/m3). There were no non-detects of acenaphthene measured at DEMI.
•	Figure 16-5 is the box plot for acetaldehyde. The box plot shows that the
maximum acetaldehyde concentration measured at DEMI is significantly less than
the program-level maximum concentration while the minimum concentration
measured at DEMI is just less than the first quartile for the program. The annual
average concentration of acetaldehyde for DEMI is similar to the program-level
average concentration.
•	Figure 16-6 is the box plot for benzene. Similar to acenaphthene, the program-
level maximum benzene concentration (43.5 |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 12 |ig/m3. This box plot shows that the range of benzene
concentrations measured at DEMI spans just over 1 |ig/m3. DEMI's annual
average benzene concentration (0.65 ± 0.06 |ig/m3) is less than the program-level
average concentration (0.78 |ig/m3) but greater than the program-level median
concentration (0.60 |ig/m3).
•	Figure 16-7 is the box plot for 1,3-butadiene. The program-level maximum
concentration (21.5 |ig/m3) is not shown directly on the box plot as the scale has
been reduced to 1.5 |ig/m3 to allow for the observation of data points at the lower
end of the concentration range. Figure 16-7 shows that the range of 1,3-butadiene
concentrations measured at DEMI is relatively small compared to the range
measured across the program. However, the concentrations at the upper end of the
16-18

-------
concentration range are driving the program-level average, as more than 75
percent of the 1,3-butadiene measurements are less than 0.1 |ig/m3. The annual
average concentration for DEMI (0.08 ± 0.01 |ig/m3) is roughly half the program-
level average concentration of this pollutant (0.15 |ig/m3).
Figure 16-8 presents the box plot for carbon tetrachloride for DEMI. The scale of
the box plot has also been reduced to allow for the observation of data points at
the lower end of the concentration range, as the program-level maximum carbon
tetrachloride concentration (23.7 |ig/m3) is considerably greater than the majority
of measurements. Figure 16-8 shows that the range of carbon tetrachloride
concentrations measured at DEMI spans approximately 0.5 |ig/m3, with a
maximum concentration that is considerably less than the program-level
maximum concentration. DEMI's annual average concentration of carbon
tetrachloride (0.67 ± 0.02 |ig/m3) falls between the program-level average
concentration and third quartile.
The scale of the box plot in Figure 16-9 has also been reduced to allow for the
observation of data points at the lower end of the concentration range, as the
program-level maximum 1,2-dichloroethane concentration (111 |ig/m3) is
considerably greater than the majority of measurements. All of the concentrations
of 1,2-dichloroethane measured at DEMI are less than the program-level average
concentration of 0.26 |ig/m3. The annual average concentration for DEMI is just
less than the program-level median concentration. This is another example of
measurements at the upper end of the concentration range driving the program-
level average concentration, as the program-level average is more than twice the
program-level third quartile.
Figure 16-10 is the box plot for ethylbenzene for DEMI. The scale of this box plot
has also been reduced to allow for the observation of data points at the lower end
of the concentration range, as the program-level maximum ethylbenzene
concentration (18.7 |ig/m3) is considerably greater than the majority of
measurements. Figure 16-10 shows that all of the ethylbenzene concentrations
measured at DEMI are less than 3 |ig/m3. DEMI's annual average concentration
of ethylbenzene is just greater than the program-level average concentration.
The box plot for fluorene presented in Figure 16-11 shows that the maximum
fluorene concentration measured at DEMI is about one-third the maximum
concentration of fluorene measured across the program. Yet, the annual average
concentration for DEMI is just less than twice the program-level average
concentration of this pollutant.
Figure 16-12 presents the box plot for formaldehyde for DEMI. The maximum
formaldehyde concentration measured at DEMI is about one-third the maximum
concentration measured across the program while the minimum concentration
measured at DEMI is greater than the program-level first quartile. The annual
average concentration for DEMI is just greater than the program-level average
concentration.
16-19

-------
• Figure 16-13 is the box plot for naphthalene. The maximum naphthalene
concentration measured at DEMI is considerably less than the maximum
concentration measured across the program while the minimum concentration
measured at DEMI is greater than the program-level first quartile. The annual
average concentration of naphthalene for DEMI is greater than the program-level
average concentration and program-level third quartile.
16.4.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.3.2.
DEMI has sampled VOCs and carbonyl compounds under the NMP since 2003, and PAHs since
2008. Thus, Figures 16-14 through 16-23 present the 1-year statistical metrics for each of the
pollutants of interest for DEMI. The statistical metrics presented for assessing trends include the
substitution of zeros for non-detects. If sampling began mid-year, a minimum of 6 months of
sampling is required for inclusion in the trends analysis; in these cases, a 1-year average
concentration is not provided, although the range and percentiles are still presented.
Figure 16-14. Yearly Statistical Metrics for Acenaphthene Concentrations Measured at DEMI
0
20081 2009
2010
2011
Year
2012
2013

O 5th Percentile — Minimum
~ Median
~ Maximum
O 95th Percentile
~~•«~••• Average
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2008.
16-20

-------
Observations from Figure 16-14 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 concentration 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 concentrations greater than 100 ng/m3 measured at
DEMI were measured in either July or August; further, all 46 measurements greater
than 20 ng/m3 were measured during the second or third quarters of a given year
(during 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
concentration. The median is influenced less by a few concentrations at the upper end
of the concentration range than the 1-year average concentration, such as the two
concentrations greater than 100 ng/m3 that were measured in 2010. The third highest
concentration measured in 2010 was considerably less (55.1 ng/m3) but still among
the higher measurements collected at this site.
•	Although the 95th percentile increased considerably from 2010 to 2011, several of the
other statistical metrics exhibit decreases (however slight). The number of
concentrations greater than 20 ng/m3 increased from five to 12 from 2010 to 2011,
accounting for one-fifth of the measurements collected in 2011.
•	The range of concentrations measured has a decreasing trend between 2010 and 2013,
with the lowest maximum concentration (since 2009) measured in 2013. Confidence
intervals calculated for these 1-year average concentrations indicate that the
measurements collected are highly variable, particularly between 2010 and 2012.
16-21

-------
Figure 16-15. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at DEMI
2003	2004	2005	2006	20071 20081 2009	2010	2011	2012	2013
Year
0 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because data from March 2007 to March 2008 was invalidated.
Observations from Figure 16-15 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 between
March 13, 2007 and March 25, 2008 by the state of Michigan due to a leak in the
sample line. With only 12 valid samples in 2007, no statistical metrics are provided.
Because less than 75 percent of the samples were valid in 2008, a 1-year average is
not presented for 2008, although the range of measurements is provided.
•	The maximum acetaldehyde concentration was measured 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 (and none in the
years that follow).
•	The 1-year average concentration exhibits a decreasing trend after 2004 that
continues through 2006. The median concentration, which is available for 2008,
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.
16-22

-------
• The smallest range of acetaldehyde concentrations was measured at DEMI in 2013,
yet the median exhibits a considerable increase and is greater than the 1-year average
concentration, which is at its highest since 2005. The number of measurements in the
1.75 |ig/m3 to 2.75 |ig/m3 range increased from 21 to 32 from 2012 to 2013
(accounting for more than half of the measurements in 2013) while the number
greater than 2.75 |ig/m3 decreased from five to one.
Figure 16-16. Yearly Statistical Metrics for Benzene Concentrations Measured at DEMI
20031	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented due to low completeness for 2003.
Observations from Figure 16-16 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 invalid
samples resulted in low completeness; as a result, a 1-year average concentration 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
between 2004 and 2009. Between 2009 and 2012, the 1-year average concentration
fluctuated between 0.81 |ig/m3 (2009) and 0.94 |ig/m3 (2010).
16-23

-------
• The smallest range of benzene concentrations was measured at DEMI in 2013, for
which all of the statistical metrics decreased except the minimum concentration. Both
the 1-year average and median concentrations are at a minimum for 2013,
representing a significant decrease from previous years.
Figure 16-17. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at DEMI
20031 2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
0 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented due to low completeness for 2003.
Observations from Figure 16-17 for 1,3-butadiene measurements collected at DEMI
include the following:
•	The maximum 1,3-butadiene concentration (1.04 |ig/m3) was measured on
October 18, 2004 and is the only 1,3-butadiene concentration greater than 1 |ig/m3
measured at DEMI, although concentrations greater than 0.90 |ig/m3 were measured
in 2004 and 2006.
•	For 2004, the minimum, 5th percentile, and median concentrations are all zero,
indicating that at least half of the measurements were non-detects. Yet, two of the
three highest concentrations were also measured at DEMI in 2004; in addition, the
maximum 95th percentile was calculated for 2004. This indicates there is a high level
of variability within 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
16-24

-------
60 percent in 2004 to 2 percent in 2008, then fluctuated between 2 percent and
8 percent for the years that follow. The number of non-detects measured in 2013
(five) is the highest number of non-detects since 2006.
•	Even as the number of non-detects decreased (and thus, the number of zeros factored
into the calculated decreased), the 1-year average concentration decreased by almost
half between 2006 and 2009. This was followed by an increasing trend between 2009
and 2012.
•	The 1-year average concentration decreased significantly from 2012 to 2013, as did
the median, both of which are at their lowest since 2010.
Figure 16-18. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured
at DEMI



r-O-

T







T




1
ml.Jm
T




	

•<:

i

i-i-

—
r^"



¦ 		

i


¦ „ti
		 j
I
			
* ik.ii



1


7

£


-


2003 1 2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile
— Minimum
— Median
— Maximum
O 95th Percentile
Average

1 A 1-year average is not presented due to low completeness for 2003.
Observations from Figure 16-18 for carbon tetrachloride measurements collected at
DEMI include the following:
•	In 2003, the measured detections ranged from 0.32 |ig/m3 to 0.76 |ig/m3, plus two
non-detects. This is the only year of sampling for which nearly half the measurements
were less than 0.5 |ig/m3.
•	The range of concentrations measured in 2004 doubled from 2003 levels. The number
of measurements greater than 1 |ig/m3 increased from none in 2003 to 12 for 2004.
16-25

-------
•	The 1-year average concentration decreased by more than 0.1 |ig/m3 from 2004 to
2005, as the range of concentrations measured decreased substantially. Little change
in the 1-year average concentration is shown from 2005 to 2007, despite the
differences in the ranges of concentrations measured.
•	With the exception of the 5th percentile, all of the statistical metrics increased
significantly for 2008, with the 1-year average and median concentrations for 2008
similar to the 95th percentile for 2007.
•	A steady decreasing trend in the 1-year average concentration is shown between 2008
and 2011. Between these years, the majority of concentrations fell within a tighter
concentration range, as indicated by the difference between the 5th and 95th
percentiles. For 2012, the difference between the 5th and 95th percentiles is less than
0.25 |ig/m3, even though an increase in the 1-year average and median concentrations
is shown.
•	Most of the statistical parameters exhibit a slight decrease from 2012 to 2013.
Figure 16-19. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured
at DEMI

Maximum
Concentration for
2006 is 3.45 ^g/m3








1
yr


—


>,
rL
&

o








o® —



>


£ m
o


20031 2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	™ Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented due to low completeness for 2003.
16-26

-------
Observations from Figure 16-19 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. Through 2011, the median concentration is zero for all years, indicating that at
least half of the measurements are non-detects: there was only one measured
detection in 2005, three in 2006, four in 2009, 12 in 2010, and 11 in 2011. The
number of measured detections increased by a factor of five for 2012, with a similar
percentage in 2013.
•	As the number of measured detections increase, so do each of the corresponding
statistical metrics shown in Figure 16-19.
•	As the number of measured detections increased dramatically for 2012, the 1-year
average and median concentrations increased correspondingly. The median
concentration is greater than the 1-year average concentration for 2012. This is
because there were still 10 non-detects (or zeros) factoring into the 1-year average
concentration for the year, which can pull down an average in the same manner an
outlier can drive an average upward.
•	The statistical metrics for 2013 resemble those calculated for 2012. The maximum
concentration measured in 2013 is very similar to the 95th percentile, such that it
appears there is no maximum concentration indicator for 2013.
•	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 |ig/m3). A similar concentration was also
measured in 2005. All of the 10 remaining concentrations greater than 0.1 |ig/m3 were
measured between 2011 and 2013.
16-27

-------
Figure 16-20. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at DEMI
20031 2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
0 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented due to low completeness for 2003.
Observations from Figure 16-20 for ethylbenzene measurements collected at DEMI
include the following:
•	The maximum ethylbenzene concentration was measured at DEMI in September
2004 (4.35 |ig/m3). Only two other ethylbenzene concentrations greater than 3 |ig/m3
have been measured at DEMI (one each in 2011 and 2012). Only 11 concentrations
greater than 2 |ig/m3 have been measured at DEMI.
•	A steady decreasing trend in the 1-year average concentration is shown after 2004,
although the rate of decrease levels out after 2006, with the 1-year average reaching a
minimum for 2008 (0.30 |ig/m3). Little change is shown for 2009.
•	The maximum concentration measured exhibits a steady increasing trend between
2008 and 2012, with all of the statistical parameters exhibiting increases for 2010,
with most continuing this increase for 2011.
•	For 2012, the minimum concentration decreased (as one non-detect was measured)
while the maximum concentration increased. The number of concentrations at the
lower end of the concentration range (those less than 0.25 |ig/m3) nearly doubled
from 2011 to 2012 (up from 10 to 19), resulting in the slight decreases shown in the
central tendency statistics for 2012.
16-28

-------
For 2013, all of the statistical metrics exhibit decreases, with the exception of the
minimum concentration, as there were no non-detects measured in 2013. The
concentrations less than 0.25 |ig/m3 account for an even greater percentage of the
measurements, accounting for 26 of the measurements (or more than 40 percent) for
2013.
Figure 16-21. Yearly Statistical Metrics for Fluorene Concentrations Measured at DEMI
20081	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under did not begin until April 2008.
Observations from Figure 16-21 for fluorene measurements collected at DEMI include
the following:
•	The maximum fluorene concentration (152 ng/m3) was measured at DEMI on
August 18, 2010 (on the same day as the maximum acenaphthene concentration was
measured). Only two other measurements greater than 100 ng/m3 have been measured
at DEMI (one in August 2008 and another in August 2010). All eight concentrations
greater than 50 ng/m3 have been measured in June, July, or August and all 38
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.
•	The trends graph for fluorene resembles the trends graph for acenaphthene in
Figure 16-14.
16-29

-------
•	The median concentrations have varied less than 2 ng/m3 over the years, ranging from
4.91 ng/m3 (2013) to 6.82 ng/m3 (2010). The 1-year average concentrations exhibit
more variability, ranging from 7.68 ng/m3 (2009) to 12.62 ng/m3 (2010).
•	All of the statistical metrics increased (at least slightly) from 2009 to 2010. The
1-year average concentration is being driven by the two highest concentrations
measured in 2010 (both greater than 100 ng/m3). The next highest concentration
measured in 2010 is considerably less (44.8 ng/m3). If the two highest concentrations
were excluded from the calculation, the 1-year average concentration would fall 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.
•	All of the statistical parameters exhibit decreases from 2012 to 2013. The maximum
concentration measured in 2013 is less than the 95th percentile for 2012 (similar to
acenaphthene).
Figure 16-22. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at DEMI
2003	2004	2005	2006	2007
2008
Year
2009	2010	2011	2012	2013
O 5th Percentile	— Minimum
~ Maximurr
O 95th Percentile
1 A 1-year average is not presented because data from March 2007 to March 2008 was invalidated.
16-30

-------
Observations from Figure 16-22 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 concentration is not presented for
2008, although the range of measurements is provided.
•	The five highest concentrations measured at DEMI were measured in 2005 and
ranged from 13.3 |ig/m3 to 33.1 |ig/m3. The nine highest formaldehyde concentrations
(those greater than 9 |ig/m3) were measured during the first 3 years of sampling.
•	The decrease in the 1-year average concentration shown between 2005 and 2006 is
significant (from 5.35 |ig/m3 to 2.92 |ig/m3). The 1-year average concentrations for
the years following 2006 (where they could be calculated) did not vary significantly
through 2011.
•	All of the statistical parameters exhibit increases for 2012. A review of the data
shows that the 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.
•	While most of the statistical parameters exhibit decreases for 2013, the minimum
concentration measured in 2013 is at its highest since the onset of sampling.
16-31

-------
Figure 16-23. Yearly Statistical Metrics for Naphthalene Concentrations Measured at DEMI
300
20081	2009	2010	2011	2012	2013
0 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2008.
Observations from Figure 16-23 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 except 2013).
•	With the exception of the maximum concentration, all of the statistical parameters
exhibit increases from 2009 to 2010. Little change is shown in the naphthalene
concentrations measured at DEMI between 2010 and 2012.
•	The smallest range of naphthalene concentrations was measured in 2013, with all of
the statistical parameters exhibiting decreases except the minimum concentration.
Both the 1-year average and median concentrations are at a minimum for 2013, with
the median concentration falling below 100 ng/m3 for the first time.
16-32

-------
16.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at the Michigan monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4
for definitions and explanations regarding the various toxicity factors, time frames, and
calculations associated with these risk-based screenings.
16.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Michigan site and where annual average
concentrations could be calculated, risk was examined by calculating cancer risk and noncancer
hazard approximations. These approximations can be used as risk estimates for cancer and
noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers want to shift their air
monitoring priorities. Refer to Section 3.4.3.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 16-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Observations from Table 16-6 include the following:
•	Formaldehyde has the highest annual average concentration for DEMI, followed by
acetaldehyde, carbon tetrachloride, and benzene.
•	These four pollutants also have the highest cancer risk approximations for this site,
although the order varies. Formaldehyde's cancer risk approximation is the highest
(39.59 in-a-million), with all other cancer risk approximations an order of magnitude
lower.
•	None of the pollutants of interest for DEMI have noncancer hazard approximations
greater than 1.0, indicating that no adverse noncancer health effects are expected from
these individual pollutants. The pollutant with the highest noncancer hazard
approximation for DEMI is formaldehyde (0.31).
16-33

-------
Table 16-6. Risk Approximations for the Michigan Monitoring Site
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Dearborn, Michigan - DEMI
Acetaldehyde
0.0000022
0.009
61/61
1.76
±0.12
3.87
0.20
Benzene
0.0000078
0.03
62/62
0.65
±0.06
5.04
0.02
1,3-Butadiene
0.00003
0.002
58/62
0.08
±0.01
2.29
0.04
Carbon Tetrachloride
0.000006
0.1
62/62
0.67
±0.02
4.02
0.01
1,2 -Dichloroethane
0.000026
2.4
57/62
0.08
±0.01
2.00
<0.01
Ethylbenzene
0.0000025
1
62/62
0.39
±0.10
0.96
<0.01
Formaldehyde
0.000013
0.0098
61/61
3.05
±0.31
39.59
0.31
Acenaphthene3
0.000088

60/60
9.62
±2.72
0.85

Fluorene3
0.000088

59/60
8.35
±2.13
0.73

Naphthalene1
0.000034
0.003
60/60
104.57
± 14.63
3.56
0.03
- = A Cancer URE or Noncancer RfC is not available.
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of
viewing.
16.5.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 16-7 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
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 3.4.3.4. Lastly,
Table 16-7 provides the 10 pollutants with the highest cancer risk approximations (in-a-million)
for DEMI, 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.
16-34

-------
Table 16-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Michigan Monitoring Site
Top 10 Total Emissions for Pollutants
with Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Dearborn, Michigan (Wayne County) - DEMI
Benzene
524.56
Coke Oven Emissions, PM
8.62E-03
Formaldehyde
39.59
Formaldehyde
438.33
Formaldehyde
5.70E-03
Benzene
5.04
Ethylbenzene
338.52
Benzene
4.09E-03
Carbon Tetrachloride
4.02
Acetaldehyde
254.42
POM, Group 5a
3.22E-03
Acetaldehyde
3.87
1.3 -Butadiene
79.05
Hexavalent Chromium
2.53E-03
Naphthalene
3.56
Naphthalene
45.78
1,3-Butadiene
2.37E-03
1,3-Butadiene
2.29
T etrachloroethylene
30.63
Arsenic, PM
2.06E-03
1,2-Dichloroethane
2.00
T richloroethylene
17.05
Naphthalene
1.56E-03
Ethylbenzene
0.96
Dichloromethane
10.97
Nickel, PM
9.22E-04
Acenaphthene
0.85
POM, Group 2b
9.34
Ethylbenzene
8.46E-04
Fluorene
0.73

-------
Table 16-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Michigan Monitoring Site
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Dearborn, Michigan (Wayne County) - DEMI
Hydrochloric acid
3,022.43
Acrolein
1,456,276.15
Formaldehyde
0.31
Toluene
2,046.58
Hydrochloric acid
151,121.26
Acetaldehyde
0.20
Hexane
1,276.18
Formaldehyde
44,727.33
1,3-Butadiene
0.04
Xylenes
1,255.32
1,3-Butadiene
39,523.56
Naphthalene
0.03
Methanol
1,113.64
Arsenic, PM
31,862.61
Benzene
0.02
Benzene
524.56
Acetaldehyde
28,268.80
Carbon Tetrachloride
0.01
Formaldehyde
438.33
Nickel, PM
21,350.40
Ethylbenzene
<0.01
Ethylene glycol
384.08
Manganese, PM
21,158.92
1,2-Dichloroethane
<0.01
Ethylbenzene
338.52
Benzene
17,485.46

Acetaldehyde
254.42
Naphthalene
15,259.58

-------
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 16.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
Observations from Table 16-7 include the following:
•	Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
cancer UREs in Wayne County.
•	The pollutants with the highest toxicity-weighted emissions (of the pollutants with
cancer UREs) for Wayne County are coke oven emissions, formaldehyde, and
benzene.
•	Five of the highest emitted pollutants in Wayne County also have the highest toxicity-
weighted emissions.
•	Formaldehyde has the highest cancer risk approximation for DEMI. This pollutant
also appears on both emissions-based lists, ranking second for both its quantity
emitted and its toxicity-weighted emissions. Benzene, naphthalene, 1,3-butadiene,
and ethylbenzene also appear on both emissions-based lists.
•	Acetaldehyde has the fourth highest cancer risk approximation for DEMI and is one
of the highest emitted pollutants in Wayne County but does not appear among those
with the highest toxicity-weighted emissions. This is also true for acenaphthene and
fluorene, which are included as part of POM, Group 2b in the NEI.
•	Carbon tetrachloride and 1,2-dichloroethane, the two remaining pollutants of interest
shown in Table 16-7, do not appear on either emissions-based list.
•	Hexavalent chromium has the fifth highest toxicity-weighted emissions for Wayne
County. Although this pollutant was sampled for at DEMI (through June 2013), it
was not identified as a pollutant of interest for this site.
16-37

-------
Observations from Table 16-8 include the following:
•	Hydrochloric acid, toluene, and hexane are the highest emitted pollutants with
noncancer RfCs in Wayne County. Wayne County is one of the few counties with an
NMP site where toluene is the not the highest emitted pollutant in the noncancer
table. The quantity of emissions for the highest ranking pollutants in Table 16-8 is an
order of magnitude higher than the quantity of emissions for the highest ranking
pollutants in Table 16-7.
•	The pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) for Wayne County are acrolein, hydrochloric acid, and
formaldehyde. Although acrolein was sampled for at DEMI, this pollutant was
excluded from the pollutants of interest designation and thus subsequent risk-based
screening evaluations due to questions about the consistency and reliability of the
measurements, as discussed in Section 3.2.
•	Four of the highest emitted pollutants in Wayne County also have the highest
toxicity-weighted emissions.
•	Formaldehyde has the highest noncancer hazard approximation for DEMI (although
none of the pollutants of interest have associated noncancer hazard approximations
greater than 1.0). Formaldehyde emissions rank seventh highest for Wayne County
while the toxicity-weighted emissions rank third (of the pollutants with noncancer
RfCs). Acetaldehyde and benzene also appear on all three lists for DEMI.
•	Several metals appear among the pollutants with the highest toxicity-weighted
emissions for Wayne County. (This was also true for the pollutants with cancer UREs
in Table 16-7.) Speciated metals were not sampled for under the NMP through the
contract laboratory.
16.6 Summary of the 2013 Monitoring Data for DEMI
Results from several of the data treatments described in this section include the
following:
~~~ Eighteen pollutants failed screens for DEMI, including three carbonyl compounds,
10 VOCs, andfour PAHs, and hexavalent chromium.
~~~ Of the site-specific pollutants of interest, formaldehyde and acetaldehyde had the
highest annual average concentrations for DEMI. None of the other site-specific
pollutants of interest had annual average concentrations greater than 1 ng/m3.
~~~ DEMI has the third highest annual average concentration of acenaphthene and the
fifth highest annual average concentration naphthalene among NMP sites sampling
PAHs.
~~~ Concentrations of formaldehyde, acenaphthene, andfluorene measured at DEMI
were highest during the warmer months of the year.
16-38

-------
~~~ 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.
16-39

-------
17.0	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.
17.1	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 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, version 2. 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
boundary are still visible on the map for reference, but have been grayed out in order to
emphasize emissions sources within the boundary. Table 17-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates.
17-1

-------
Figure 17-1. St. Cloud, Minnesota (ST.VIN) Monitoring Site
to

-------
Figure 17-2. NEI Point Sources Located Within 10 Miles of STMN
Benton
Count |
Sterns
County
•m ttfCY#	•rtnrw
I t—¦ v ¦—:—,—.—
- I		 I
9A-2WyYf	M77T"W
Legend
~ STMN UATMP site
Source Category Group (No. of Facilities)
T AirporVAirtine/Airport Support Operation (3)
it Asphalt Production/Ho! Mi* Asphalt Plant (1)
K Automotnte'Tnick Manufacturing Faculty (2)
c Chemical Manufacturing Facility (1)
Crematory • AnimalfHuman (3)
H) Dry Cleaning Facility (3)
* Electricity Generation via Combustion (1)
F Food Processing/Agriculture Facility (1)
I Foundries. Iron and Steel (1)
» industrial Machinery or Equipment Plant (3)
o institutional (school, hospita I. prison, etc I (9)
«« wffyi	OMttrcnH	ursirw
Note Due to density and e«lloea*or the total facKlwa
dispayad i\a» not represent all taciMack within tno ansa at oterest
10 mile radius	|	 County boundary
a Landfill 11)
<•> Metals Processing/Fabrication Facility (3)
A Military Base/National Security Facility (1)
Mine'Quarry/M me rat Processing Facility (5)
? Miscellaneous Goimvertiairtndustnal Facikfy (2)
P Piinting/PublisJiing'Papef Product Manufacturing Facillty |3>
IB Pulp and Paper Plant (1)
n Telecommunications/Radio Facility (1 >
T Textile. Yam. or Carpet Plant (1)
• Wastewater Treatment Facility (S)
W Woodwork Furniture. MithvorK 8 Wood Preserving Facility (12)
17-3

-------
Table 17-1. Geographical Information for the Minnesota Monitoring Site
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Additional Ambient Monitoring Information1
STMN
27-145-3053
St. Cloud
Stearns
St. Cloud, MN
45.564637,
-94.226345
Industrial
Suburban
TSP, TSP Metals.
1 Data for additional pollutants are reported to AQS for STMN (EPA, 2014b); 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 17-2 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, a foundry, iron, and steel facility, and an industrial
machinery/equipment plant. Additional facilities are located to the southwest and south.
Table 17-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Minnesota monitoring site. Table 17-2 includes both county-level
population and vehicle registration information. Table 17-2 also contains traffic volume
information for STMN as well as the location for which the traffic volume was obtained.
Additionally, Table 17-2 presents the county-level daily VMT for Stearns County.
Table 17-2. Population, Motor Vehicle, and Traffic Information for the Minnesota
Monitoring Site
Site
County
Estimated
County
Population1
County-level
Vehicle
Registration2
Annual
Average
Daily
Traffic3
Intersection
Used for
Traffic Data
County-
level Daily
VMT4 "
STMN
Stearns
152,092
221,636
24,100
8th St N (Route 4/Vctcrans Dr), at
Anderson Ave
5,078,055
bounty-level population estimate reflects 2013 data (Census Bureau, 2014)
2County-level vehicle registration reflects 2013 data (MN DPS, 2014)
3AADT reflects 2009 data (MN DOT, 2012)
4County-level VMT reflects 2013 data (MN DOT, 2014)
17-5

-------
Observations from Table 17-2 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 greater than 5 million miles and ranks 34th
compared to other counties with NMP sites.
17.2 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.
17.2.1	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, with blizzard conditions
developing twice per winter (on average). A northwest wind is predominant in St. Cloud most of
the year, although a southerly wind occurs during the summer months (NCDC, 2015; MCWG,
2015).
17.2.2	Meteorological Summary
Hourly meteorological data for 2013 were retrieved from NCDC for the weather station
closest to the Minnesota monitoring site (NCDC, 2013), as described in Section 3.4.2. The
closest weather station to STMN 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 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 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)
8.6
miles
Sample
Davs
(25)
38.2
±7.4
30.6
±7.0
20.8
±6.3
27.0
±6.2
70.4
±5.7
1018.9
±2.5
7.0
±0.9
100°
(E)
2013
50.4
+ 2.6
41.3
±2.5
32.0
±2.3
37.3
±2.3
72.3
± 1.1
1017.4
±0.8
6.9
±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 2013. Also included in Table 17-3 is the 95 percent
confidence interval for each parameter. As shown in Table 17-3, average meteorological
conditions on sample days appear cooler and drier than conditions experienced throughout 2013.
Sampling at STMN under the NMP was discontinued at the end of May 2013, thereby missing
the warmest and wettest months of the year. Based on the full-year averages, STMN is located in
one of the coldest locations, with the second lowest average maximum temperature and third
lowest average temperature. This location is also among the driest locations, based on the
average dew point and wet bulb temperatures, although the average relative humidity ranks
among the higher averages.
17.2.3 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 3.4.2. A wind rose shows the frequency of wind directions using "petals"
positioned around a 16-point compass, and uses different colors to represent wind speeds.
Figure 17-3 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 17-3 also presents three different wind roses for the
STMN monitoring site. First, a historical wind rose representing 2003 to 2012 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 2013 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2013 is
presented. These can be used to identify the predominant wind speed and direction for 2013 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.
17-8

-------
Figure 17-3. Wind Roses for the St. Cloud Regional Airport Weather Station near STMN
Location of STMN and Weather Station
2003-2012 Historical Wind Rose
NORTH
west:
WWC SPEED
i Knots)
SOUTH
2013 Wind Rose
VEST
(Knots)
11 -17
SOUTH
WIND SPEED
~ 4-7
Calms: 14.62%
Sample Day Wind Rose
! NORTH""-,
ES ,
WIND SPEED
(Knots)
~ >=22
~
7- 11
4- 7
Calms: S.68%
17-9

-------
Observations from Figure 17-3 for STMN include the following:
•	The St. Cloud Regional Airport weather station is located 8.6 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 (although this is more evident in a satellite-type
map).
•	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 quadrants. Winds from
northwest and south each account for just less than 10 percent of observations. The
strongest wind speeds were most often associated with westerly to northwesterly
winds. Calm winds (those less than or equal to 2 knots) were observed for more than
15 percent of the hourly measurements.
•	The wind patterns shown on the 2013 wind rose resemble the historical wind patterns,
indicating that wind observations in 2013 were similar to those observed historically.
•	The sample day wind rose exhibits few of the characteristics of the other wind roses,
with winds from the north accounting for the greatest percentage of observations.
Winds from the northeast quadrant were observed nearly as often as winds from the
southeast quadrant and the percentage of calm winds was reduced by nearly half.
However, the sample day wind rose includes observations from January through May
only and a wind rose with a full year's worth of sample day observations may look
different.
17.3 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 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. Only
hexavalent chromium was sampled for at STMN.
17-10

-------
Table 17-4. Risk-Based Screening Results for the Minnesota Monitoring Site
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
St. Cloud, Minnesota - STMN
Hexavalent Chromium
0.000083
0
8
0.00
0.00
0.00
Total
0
8
0.00

Observations from Table 17-4 include the following:
•	Hexavalent chromium was detected in eight of the 24 valid samples collected at
STMN, representing a 33 percent detection rate.
•	The eight measured detections of hexavalent chromium measured in 2013 at STMN
did not fail any screens.
•	By comparison, hexavalent chromium was detected in 39 of the 54 valid samples
collected at STMN during the 2012 portion of the monitoring effort (from February to
December 2012) and failed a total of six screens.
17.4 Concentrations
This section typically presents various concentration averages used to characterize
pollution levels at the monitoring site for each of the site-specific pollutants of interest. However,
because there were no failed screens for STMN, this site has no pollutants of interest based on
the risk screening process. The short sampling duration also prevents an annual average
concentration for hexavalent chromium to be calculated. In order to facilitate a review of the data
collected at STMN in 2013, a few statistical calculations are provided in the section that follows.
A statistical summary for all of the hexavalent chromium measurements collected at STMN is
provided in Appendix O. The concentration comparison and trend analysis were not performed.
17.4.1 2013 Concentration Averages
Quarterly concentration averages were calculated for hexavalent chromium for the
Minnesota site, as described above. 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, which
includes all measured detections and substituted zeros for non-detects for the entire year of
17-11

-------
sampling, could not be calculated as sampling at STMN was discontinued at the end of May
2013. Quarterly average concentrations for STMN are presented in Table 17-5, 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 17-5. Quarterly Average Concentrations of Hexavalent Chromium 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 - STMT*

Hexavalent Chromium
8/24
0.007
± 0.006
0.003
± 0.004
NA
NA
NA
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
Observations for STMN from Table 17-5 include the following:
•	The eight measured detections of hexavalent chromium range from 0.008 ng/m3 to
0.039 ng/m3. This dataset also includes 16 non-detects, which account for two-thirds
of the valid samples collected in 2013.
•	Six of the eight measured detections were measured in samples collected in January
and February (with the other two measured in May). The measurements from all five
samples collected in January resulted in measured detections.
•	The hexavalent chromium concentrations measured in 2013 are considerably lower
than those measured during the 2012 portion of the monitoring effort, when several of
the highest measurements of hexavalent chromium were collected. In total, five
hexavalent chromium measurements greater than 1 ng/m3 have been collected under
the NMP between 2005 and 2012, with three collected at STMN in 2012.
•	In June 2012, a nearby facility manufacturing stainless steel storage and processing
tanks installed new air pollution control equipment to reduce its emissions, which
resulted in a 97 percent decrease in its hexavalent chromium emissions (MPCA,
2015).
17.5 Additional Risk-Based Screening Evaluations
In order to characterize risk at participating monitoring sites, additional risk-based
screening evaluations were conducted. Because there were no pollutants of interest identified for
STMN and because annual averages could not be calculated for the pollutant sampled for at
17-12

-------
STMN, cancer risk and noncancer hazard approximations, as described in Section 3.4.3.3, were
not calculated. The risk-based emissions assessment described in Section 3.4.3.4 was still
conducted, at least in part, as the emissions can be reviewed independent of concentrations
measured.
17.5.1 Risk-Based Emissions Assessment
This section presents an evaluation of county-level emissions based on cancer and
noncancer toxicity, respectively, and is intended to help policy-makers prioritize their air
monitoring activities. Table 17-6 presents the 10 pollutants with the highest emissions from the
2011 NEI (version 2) that have cancer toxicity factors. Table 17-6 also presents the 10 pollutants
with the highest toxicity-weighted emissions, based on the weighting schema described in
Section 3.4.3.4. The emissions and toxicity-weighted emissions are shown in descending order in
Table 17-6. Table 17-7 presents similar information, but is limited to those pollutants with
noncancer toxicity factors. Because not all pollutants have both cancer and noncancer toxicity
factors, the highest emitted pollutants in the cancer table may be different from the noncancer
table, although the actual quantity of emissions is the same. A more in-depth discussion of this
analysis is provided in Section 3.4.3.4.
17-13

-------
Table 17-6. 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)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
St. Cloud, Minnesota (Stearns County) - STMN
Formaldehyde
200.69
Formaldehyde
2.61E-03

Benzene
187.61
Benzene
1.46E-03
Acetaldehyde
110.84
1,3-Butadiene
8.61E-04
Ethylbenzene
63.20
Naphthalene
5.41E-04
1.3 -Butadiene
28.71
POM, Group 5a
2.62E-04
Naphthalene
15.90
POM, Group 2b
2.62E-04
1,3 -Dichloropropene
12.08
Acetaldehyde
2.44E-04
Dichloro methane
3.07
POM, Group 2d
1.92E-04
POM, Group 2b
2.98
POM, Group 3
1.67E-04
Tetrachloroethylene
2.69
Ethylbenzene
1.58E-04

-------
Table 17-7. 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)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard
Approximations Based on Annual Average
Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer
Hazard
Approximation
(HQ)
St. Cloud, Minnesota (Stearns County) - STMN
Toluene
813.91
Acrolein
434,785.66

Xylenes
315.44
Formaldehyde
20,478.35
Hexane
206.04
1.3 -Butadiene
14,356.53
Formaldehyde
200.69
Acetaldehyde
12,315.70
Benzene
187.61
Cyanide Compounds, PM
6,419.33
Acetaldehyde
110.84
Benzene
6,253.79
Methanol
95.35
Lead, PM
6,189.57
Ethylbenzene
63.20
Naphthalene
5,300.31
1.1.1 -T richloroethane
36.41
Chlorine
5,240.71
Ethylene glycol
32.26
Manganese, PM
3,727.64

-------
Observations from Table 17-6 include the following:
•	Formaldehyde, benzene, and acetaldehyde are the highest emitted pollutants with
cancer UREs in Stearns County.
•	Formaldehyde, benzene, and 1,3-butadiene are the pollutants with the highest
toxicity-weighted emissions (of the pollutants with cancer UREs) for Stearns County.
•	Seven of the highest emitted pollutants in Stearns County also have the highest
toxicity-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 shown in Table 17-6. Hexavalent chromium ranks 29th for total emissions
and 12th for toxicity-weighted emissions.
•	Naphthalene and several POM Groups rank among Stearns County's highest toxicity-
weighted emissions. PAHs were not sampled for at STMN.
•	In the 2012 NMP report, emissions of bis(2-ethylhexyl)phthalate (DEHP) gas ranked
highest in Stearns County. The quantity of emissions of this pollutant decreased
substantially with the corrections from version 1 to version 2 of the 2011 NEI,
ranking 20th for quantity emitted in version 2.
Observations from Table 17-7 include the following:
•	Toluene, xylenes, and hexane are the highest emitted pollutants with noncancer RfCs
in Stearns County.
•	The pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) are acrolein, formaldehyde, and 1,3-butadiene.
•	Three 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 60th and its
toxicity-weighted emissions rank 29th (among the pollutants with noncancer RfCs).
•	Similar to Table 17-6, the rankings of bis(2-ethylhexyl)phthalate (DEHP) gas
decreased substantially with the corrections from version 1 to version 2 of the 2011
NEI.
17-16

-------
17.6 Summary of the 2013 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. Sampling was
discontinued at this site in May 2013.
~~~ Non-detects account for two-thirds of the hexavalent chromium measurements
collected at STMN in 2013, with the measured detections ranging from 0.008 ng/m3
to 0.039 ng/m3.
17-17

-------
18.0	Sites in Mississippi
This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the UATMP sites in Mississippi, 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 Mississippi 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 Mississippi monitoring sites are located in Columbus, Mississippi. Figures 18-1 and
18-2 are composite satellite images retrieved from ArcGIS Explorer showing the monitoring
sites and their immediate surroundings. Figure 18-3 identifies nearby point source emissions
locations by source category, as reported in the 2011 NEI for point sources, version 2. Note that
only sources within 10 miles of the sites are included in the facility counts provided in
Figure 18-3. 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 boundary are still visible on the map for reference, but have been
grayed out in order to emphasize emissions sources within the boundary. Table 18-1 provides
supplemental geographical information such as land use, location setting, and locational
coordinates.
18-1

-------
Figure 18-1. Columbus, Mississippi (KMMS) Monitoring Site

-------
Figure 18-2. Columbus, Mississippi (SSMS) Monitoring Site
00
tK Ave N
Main St

-------
Figure 18-3. NEI Point Sources Located Within 10 Miles of KM MS and SSMS
Manrce
County
ALABAMA
Lovvnoes
County


Note Que to 'acnty density and coltacatiofi the total fac
c Chemical Manufacturing Facility (1)
I Compressor Station (1)
R Plastic, Resl n, or Rubber Products Plant (t)
P Prmting.'Publrshlng/Paper Product Manufacturing Facility {1 >
W Woodwork, Furniture. Millwork & Wood Preserving Facility (2)
18-4

-------
Table 18-1. Geographical Information for the Mississippi Monitoring Sites
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Additional Ambient Monitoring Information
KMMS
28-087-0002
Columbus
Lowndes
Columbus, MS
33.50944,
-88.40889
Residential
Urban/City
Center
None.
SSMS
28-087-0003
Columbus
Lowndes
Columbus, MS
33.499588,
-88.403648
Residential
Urban/City
Center
None.

-------
The KMMS and SSMS monitoring sites are located in Columbus, Mississippi. Columbus
is located about 7 miles from the Mississippi/Alabama border. One of the sites (KMMS) is
located on the property of Kerr-McGee, a former chemical manufacturing facility that pressure-
treated wood products such as railroad ties in the northern part of town. The Kerr-McGee
property consists of 90 acres. Operations ceased in 2003. The property is a Superfund site and
remediation is on-going. Monitoring at these sites is part of an investigation into risk potential
associated with soil contamination. More information can be found at EPA's Region 4 Superfund
website (EPA, 2015f).
The KMMS site is located on the east side of the Kerr-McGee property. VOC samples
were collected at this site throughout the year. PAH sampling occurred at two locations over
6 months: at the primary location for 3 months and at the nearby pine yard for 3 months, a
location where treated wood awaited shipment, on the northeast side of the property.
Immediately to the south of the property lies a cemetery. Additional industrial facilities are
located to the south, northwest, and northeast, with residential areas surrounding the
aforementioned areas, as shown in Figure 18-1.
The SSMS monitoring site is located to south of the KMMS site at Stokes-Beard
Elementary School, the ball fields of which are prominent features in Figure 18-2. The school is
located on Main Street (Route 69/182), a primary thoroughfare through the center of Columbus.
Magby Creek runs roughly north-south to the east of both sites. A rail line runs roughly north-
south through Columbus, lies just to the east Kerr-McGee property, and about three blocks west
of SSMS.
Table 18-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Mississippi monitoring sites. Table 18-2 includes both county-
level population and vehicle registration information. 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 Lowndes County.
18-6

-------
Table 18-2. Population, Motor Vehicle, and Traffic Information for the Mississippi
Monitoring Sites
Site
County
Estimated
County
Population1
County-level
Vehicle
Registration2
Annual
Average
Daily Traffic3
Intersection
Used for Traffic Data
County-
level Daily
VMT4 "
KMMS
Lowndes
59,922
54,826
9,900
N 14th Ave east of N 21st St
1,961,288
SSMS
19,000
Main St east of N 23rd St
bounty-level population estimate reflects 2013 data (Census Bureau, 2014)
2County-level vehicle registration reflects 2013 data (MS DOR, 2014)
3AADT reflects 2013 data (MS DOT, 2013)
4County-level VMT reflects 2012 data (MS DOT, 2014)
Observations from Table 18-2 include the following:
•	Lowndes County has one of the lower county-level populations compared to other
counties with NMP sites. The same is true for the county-level vehicle registration for
Lowndes County.
•	SSMS experiences a higher traffic volume compared to KMMS. The traffic volume
near SSMS is in the middle of the range compared to traffic volumes near other NMP
sites, with the traffic volume near KMMS ranking in the bottom third.
•	The daily VMT for Lowndes County is nearly 2 million miles, ranking among the
lower VMT compared to other counties with NMP sites.
18.2 Meteorological Characterization
The following sections characterize the meteorological conditions near the monitoring
sites in Mississippi on sample days, as well as over the course of the year.
18.2.1 Climate Summary
The state of Mississippi has a humid subtropical climate, with mild winters and long hot
summers. Southerly winds prevail much of the year, bringing warm moist air out of the Gulf of
Mexico, contributing to high humidity levels. During the winter, warm moist air out of the Gulf
of Mexico alternates with cooler, drier air from the north, although cold spells tend to be short-
lived. Mississippi is one of the wettest states in the country, with state-wide annual precipitation
levels greater than 55 inches. Afternoon thunderstorms occur with regularity and tornadoes are
not uncommon (NCDC, 2015).
18-7

-------
18.2.2	Meteorological Summary
Hourly meteorological data for 2013 were retrieved from NCDC for the weather station
closest to the Mississippi monitoring sites (NCDC, 2013), as described in Section 3.4.2. The
closest weather station to both KMMS and SSMS is located at the Columbus Air Force Base
Airport (WBAN 13825). Additional information about the Columbus Air Force Base weather
station, such as the distance between the sites and the weather station, 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.
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 2013. 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 representative of average weather conditions experienced
throughout the year. These sites experienced the highest relative humidity levels among all NMP
sites, one of only two locations with an average relative humidity greater than 75 percent.
18.2.3	Wind Rose Comparison
Hourly surface wind data from the weather station at the Columbus Air Force Base
Airport were uploaded into a wind rose software program to produce customized wind roses, as
described in Section 3.4.2. A wind rose shows the frequency of wind directions using "petals"
positioned around a 16-point compass, and uses different colors to represent wind speeds.
18-8

-------
Table 18-3. Average Meteorological Conditions near the Mississippi 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)
Columbus, Mississippi - KMMS
Columbus Air
Force Base
Airport
13825
(33.65, -88.45)
10.0
miles
346°
(NNW)
Sample
Days
(63)
72.3
±4.0
62.2
±4.0
53.4
±4.2
57.3
±3.8
76.0
±2.5
1018.9
± 1.6
4.9
±0.6
2013
72.1
± 1.6
61.5
± 1.6
52.7
± 1.7
56.8
± 1.5
76.1
± 1.1
1018.8
±0.6
4.6
±0.3
Columbus, Mississippi - SSMS
Columbus Air
Force Base
Airport
13825
(33.65, -88.45)
10.7
miles
346°
(NNW)
Sample
Days
(60)
72.0
±4.2
61.7
±4.1
52.8
±4.4
56.9
±3.9
75.7
±2.6
1019.0
± 1.6
5.0
±0.7
2013
72.1
± 1.6
61.5
± 1.6
52.7
± 1.7
56.8
± 1.5
76.1
± 1.1
1018.8
±0.6
4.6
±0.3
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.

-------
Figure 18-4 presents a map showing the distance between the weather station and
KMMS, which may be useful for identifying topographical influences that may affect the
meteorological patterns experienced at this location. Figure 18-4 also presents three different
wind roses for the KMMS monitoring site. First, a historical wind rose representing 2003 to 2012
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 2013 is
presented. Next, a wind rose representing wind observations for days on which samples were
collected in 2013 is presented. These can be used to identify the predominant wind speed and
direction for 2013 and to determine if wind observations on sample days were representative of
conditions experienced over the entire year and historically. Figure 18-5 presents the distance
map and three wind roses for SSMS.
Observations from Figures 18-4 and 18-5 for the Mississippi monitoring sites include the
following:
•	The weather station at Columbus Air Force Base Airport is the closest weather station
to both KMMS and SSMS. The Columbus Air Force Base is located well north of
Columbus, 10 miles north-northwest of KMMS and nearly 11 miles north-northwest
of SSMS.
•	Because the Columbus Air Force Base weather station is the closest weather station
to both sites, the historical and 2013 wind roses for KMMS are the same as those for
SSMS.
•	The historical wind rose shows that winds from the south prevail near KMMS and
SSMS. Winds from the southeast to south and northwest to north are the most
commonly observed winds observed near the Mississippi sites. Winds from the
northeast and southwest quadrants were infrequently observed. Calm winds (those
less than or equal to 2 knots) account for 27 percent of the hourly wind measurements
from 2003 to 2012.
•	The 2013 wind patterns are similar to the historical wind patterns.
•	The sample day wind patterns for each site resemble both the historical and 2013
wind patterns, indicating that wind conditions on sample days were representative of
those experienced over the entire year and historically. However, northerly winds
were observed more frequently on sample days, while calm winds were observed less
frequently.
18-10

-------
Figure 18-4. Wind Roses for the Columbus Air Force Base Airport Weather Station near
KMMS
Location of KMMS and Weather Station
2003-2012 Historical Wind Rose
•NORTH""-'.
west:
E- S~
WW C SPEED
(Knots)
17-21
11 - 17
SO UTH ' - '
Calms: 27.17%
2013 Wind Rose
Sample Day Wind Rose
NORTH"---
WEST
WIN C S PE EC
(Kn ots)
17 - 21
11 - 17
SOUTH
7- 11
Calms: 2727%
NORTH
El--;
WIND SPEED
(Knots)
I I >-22
liBl	17-21
¦	11 17
[ ill	7- 11
I ~1 4-7
2-4
Calms: 23.38%
18-11

-------
Figure 18-5. Wind Roses for the Columbus Air Force Base Airport Weather Station near
SSMS
Location of SSMS and Weather Station
2003-2012 Historical Wind Rose
If

\\
\


-W- , N


LO-
WEST
WW D SPEED
(Kn ots}
SOUTH
2013 Wind Rose
Sample Day Wind Rose
WEST
WIND SPEED
i, Knots)
SOUTH

WWC SPEED
(Knots)
SOUTH
18-12

-------
18.3 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each
Mississippi 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. VOCs were sampled for at both Mississippi sites year-round; PAHs
were also sampled for at KMMS for a 6-month period from May 10, 2013 to October 31, 2013.
However, the PAH instrumentation was moved from near the VOC sampler to the pine yard after
August 8, 2013, as mentioned in Section 18.1, which is located north of N 14th Avenue, just
west of the rail road (EPA, 2014f). The PAH samples collected at the KMMS site were analyzed
using an adjusted methodology in order to provide for the analysis of phenol and cresols at the
request of the monitoring agency, as discussed in Section 2.2.3. As a result, the PAH analyte list
is slightly different and includes results for phenols and cresols.
Observations from Table 18-4 include the following:
•	The number of pollutants failing screens is higher for KMMS than SSMS; this is
expected given that PAHs were not sampled for at SSMS.
•	Fourteen pollutants failed at least one screen for KMMS; 61 percent of concentrations
for these 14 pollutants were greater than their associated risk screening value (or
failed screens).
•	Ten pollutants contributed to 95 percent of failed screens for KMMS and therefore
were identified as pollutants of interest for KMMS. These 10 pollutants include seven
VOCs and three PAHs. KMMS is the only NMP site for which xylenes are a pollutant
of interest. Xylenes concentrations measured at KMMS account for half of the failed
screens of this pollutant across the program.
•	Ten pollutants failed screens for SSMS; approximately 64 percent of concentrations
for these 10 pollutants were greater than their associated risk screening value (or
failed screens).
18-13

-------
•	Seven pollutants contributed to 95 percent of failed screens for SSMS and therefore
were identified as pollutants of interest for this site.
•	These sites have six pollutants of interest in common: benzene, 1,3-butadiene, carbon
tetrachloride, /?-dichlorobenzene, ethylbenzene, and 1,2-dichloroethane.
•	Benzene, carbon tetrachloride, and 1,2-dichloroethane each failed 100 percent of
screens for each site.
•	Although both phenol and cresols have screening values, none of these failed screens
for KMMS.
Table 18-4. Risk-Based Screening Results for the Mississippi Monitoring Sites
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Kerr-McGee, Columbus, Mississippi - KMMS
Benzene
0.13
60
60
100.00
17.96
17.96
Carbon Tetrachloride
0.17
60
60
100.00
17.96
35.93
1,2-Dichloroethane
0.038
57
57
100.00
17.07
52.99
1.3 -Butadiene
0.03
43
51
84.31
12.87
65.87
Naphthalene
0.029
29
30
96.67
8.68
74.55
Acenaphthene
0.011
20
30
66.67
5.99
80.54
Ethylbenzene
0.4
17
60
28.33
5.09
85.63
/?-Dichlorobcnzcnc
0.091
11
39
28.21
3.29
88.92
Fluorene
0.011
11
30
36.67
3.29
92.22
Xylenes
10
11
60
18.33
3.29
95.51
Hexacliloro -1,3 -butadiene
0.045
10
12
83.33
2.99
98.50
Fluoranthene
0.011
3
30
10.00
0.90
99.40
T richloroethylene
0.2
1
16
6.25
0.30
99.70
Vinyl chloride
0.11
1
10
10.00
0.30
100.00
Total
334
545
61.28

Stokes-Beard Elementary School, Columbus, Mississippi - SSMS
Benzene
0.13
61
61
100.00
21.25
21.25
Carbon Tetrachloride
0.17
61
61
100.00
21.25
42.51
1,2-Dichloroethane
0.038
58
58
100.00
20.21
62.72
1.3 -Butadiene
0.03
54
58
93.10
18.82
81.53
/?-Dichlorobcnzcnc
0.091
18
48
37.50
6.27
87.80
Hexacliloro -1,3 -butadiene
0.045
16
19
84.21
5.57
93.38
Ethylbenzene
0.4
14
61
22.95
4.88
98.26
Xylenes
10
3
61
4.92
1.05
99.30
Methyl tort-Butyl Ether
3.8
1
7
14.29
0.35
99.65
Vinyl chloride
0.11
1
11
9.09
0.35
100.00
Total
287
445
64.49

18-14

-------
18.4 Concentrations
This section presents various concentration averages used to characterize air toxics
pollution levels at the Mississippi monitoring sites. Where applicable, the following calculations
and data analyses were performed for each of the site-specific pollutants of interest:
•	Time period-based concentration averages (quarterly and annual) are provided for
each monitoring site.
•	Annual average concentrations are presented graphically for each site to illustrate
how the site's concentrations compare to the program-level averages, as presented in
Section 4.1.
•	Concentration averages and other statistical metrics are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at KMMS and SSMS are provided in Appendices J and M.
18.4.1 2013 Concentration Averages
Quarterly and annual concentration averages were calculated for the pollutants of interest
for each Mississippi 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
Mississippi monitoring sites are presented in Table 18-5, where applicable. Note that
concentrations of the PAHs for KMMS are presented in ng/m3 for ease of viewing. Also note
that if a pollutant was not detected in a given calendar quarter, the quarterly average simply
reflects "0" because only zeros substituted for non-detects were factored into the quarterly
average concentration.
18-15

-------
Table 18-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Mississippi Monitoring Sites
Pollutant
# of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
frig/m3)
2nd
Quarter
Average
frig/m3)
3rd
Quarter
Average
frig/m3)
4th
Quarter
Average
frig/m3)
Annual
Average
frig/m3)
Kerr-McGee, Columbus, Mississippi - KMMS
Benzene
60/60
0.74
±0.12
0.39
±0.05
0.45
±0.12
0.55
±0.11
0.53
±0.06
1,3-Butadiene
51/60
0.07
±0.03
0.03
±0.02
0.04
±0.02
0.09
±0.03
0.06
±0.01
Carbon Tetrachloride
60/60
0.65
±0.03
0.68
±0.05
0.66
±0.04
0.65
±0.02
0.66
±0.02
/?-Dichlorobcnzcnc
39/60
0.06
±0.04
0.04
±0.02
0.05
±0.02
0.06
±0.03
0.05
±0.01
1,2-Dichloroethane
57/60
0.10
±0.01
0.08
±0.01
0.05
±0.02
0.08
±0.01
0.08
±0.01
Ethylbenzene
60/60
1.78
±2.07
3.35
±3.01
1.82
±2.43
0.85
±0.99
1.95
± 1.08
Xylenes
60/60
11.22
± 14.04
21.66
± 19.79
11.57
± 16.35
4.88
±6.43
12.33
±7.20
Acenaphthene3
30/30
NA
NA
19.34
±5.21
NA
NA
Fluorene3
30/30
NA
NA
11.27
±2.77
NA
NA
Naphthalene3
30/30
NA
NA
116.30
±33.22
NA
NA
Stokes-Beard Elementary School, Columbus, Mississippi - SSMS
Benzene
61/61
0.80
±0.16
0.47
±0.08
0.52
±0.11
0.63
±0.18
0.60
±0.07
1,3-Butadiene
58/61
0.10
±0.04
0.05
±0.02
0.05
±0.02
0.11
±0.04
0.08
±0.02
Carbon Tetrachloride
61/61
0.62
±0.04
0.69
±0.06
0.69
±0.03
0.65
±0.04
0.66
±0.02
p-Dichlorobenzene
48/61
0.08
±0.05
0.06
±0.02
0.06
±0.02
0.10
±0.03
0.08
±0.02
1,2-Dichloroethane
58/61
0.10
±0.01
0.09
±0.01
0.07
±0.01
0.07
±0.02
0.08
±0.01
Ethylbenzene
61/61
0.52
±0.35
0.35
±0.25
0.24
±0.07
0.46
±0.24
0.39
±0.12
Hexachloro-1,3 -butadiene
19/61
0.04
±0.03
0.02
±0.02
0
0.05
±0.02
0.03
±0.01
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 KMMS from Table 18-5 include the following:
• The pollutants of interest with the highest annual average concentrations are xylenes
(13.22 ± 7.20 |ig/m3) and ethylbenzene (1.95 ± 1.08 |ig/m3). These are the only
pollutants of interest with an annual average concentration greater than 1 |ig/m3 for
this site.
18-16

-------
•	Each of the quarterly average concentrations for xylenes have relatively large
confidence intervals associated with them, some of which are greater than the
averages themselves. A review of the xylenes data shows that concentrations span
three orders of magnitude and range from 0.19 |ig/m3 to 117 |ig/m3. The 10 highest
xylenes concentrations measured across the NMP (those greater than 25 |ig/m3) were
all measured at KMMS and were measured at various time throughout the year. Half
of the concentrations greater than 25 |ig/m3 were measured during the second quarter
of 2013, in April and May, with two measured during the first quarter, two during the
third quarter, and one during the fourth quarter. The maximum xylenes concentration
was measured on April 10, 2013, with a second xylenes concentrations greater than
100 |ig/m3 measured on August 8, 2013. A third concentration approaching
100 |ig/m3 was also measured on April 16, 2013.
•	The annual and quarterly average concentrations of ethylbenzene also have large
confidence intervals. Concentrations of ethylbenzene measured at KMMS range from
0.0696 |ig/m3 to 18.7 |ig/m3, with nine of the 10 ethylbenzene concentrations greater
than 5 |ig/m3 across the program measured at KMMS. The 10 highest ethylbenzene
concentrations were measured at KMMS on the same days as the 10 highest xylenes
concentrations.
•	Concentrations of benzene and 1,3-butadiene measured at KMMS appear higher
during the colder months of the year and lower during the warmer months of the year,
although the differences among the quarterly average concentrations are not
statistically significant.
•	Due to the duration of the PAH sampling and the criteria for quarterly averages, only
third quarter average concentrations could be calculated, as shown in Table 18-5.
•	Of the PAH pollutants of interest, naphthalene had the highest concentrations,
spanning an order of magnitude and ranging from 22.1 ng/m3 to 281 ng/m3.
Naphthalene concentrations greater than 200 ng/m3 were measured five times in
samples collected between May and August. Concentrations of fluorene ranged from
1.20 ng/m3 to 23.6 ng/m3, with fluorene concentrations greater than 20 ng/m3
measured four times between May and August. Concentrations of acenaphthene
ranged from 1.56 ng/m3 to 40.3 ng/m3, with acenaphthene concentrations greater than
30 ng/m3 measured five times between May and August. The maximum concentration
of all three PAHs was measured on June 21, 2013. The top five concentrations of
each were measured on the same 5 days, although the order varied, and were all
measured prior to the instrument relocation to the pine yard.
Observations for SSMS from Table 18-5 include the following:
•	None of the pollutants of interest for SSMS have an annual average concentration
greater than 1 |ig/m3. The pollutant of interest with the highest annual average
concentration for SSMS is carbon tetrachloride (0.66 ± 0.02 |ig/m3), which is
equivalent to the annual average concentration of this pollutant for KMMS.
18-17

-------
•	The annual average concentration of ethylbenzene calculated for SSMS is
considerably less than the annual average concentration of ethylbenzene calculated
for KMMS. Although several of the quarterly average concentrations have relatively
large confidence intervals, they are all considerably less than the quarterly averages
calculated for KMMS. Concentrations of ethylbenzene measured at SSMS range from
0.052 |ig/m3 to 2.72 |ig/m3, with concentrations greater than 1 |ig/m3 measured on
two different days in January, one in June, and another in November. None of these
dates correspond with "high" days for KMMS.
•	Similar to KMMS, concentrations of benzene and 1,3-butadiene measured at SSMS
appear higher during the colder months of the year, although the differences among
the quarterly average concentrations are not statistically significant. A review of the
data shows that the five benzene concentrations greater than 1 |ig/m3 were measured
at SSMS between January and February or November and December. For
1,3-butadiene, the six highest concentrations were measured in January, February,
November or December. Note that the four highest concentrations of benzene and
1,3-butadiene were measured on the same days, although the order varied, with the
two highest concentrations of both pollutants measured on December 18, 2013 and
January 4, 2013.
Tables 4-9 through 4-12 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for KMMS and
SSMS from those tables include the following:
•	KMMS and SSMS appear in Table 4-9 a total of six times.
•	KMMS has the highest annual average ethylbenzene concentration among NMP sites
sampling this pollutant. This site's annual average concentration is more than twice
the next highest concentration of ethylbenzene shown and has the largest confidence
interval (by a considerable margin).
•	SSMS has the highest annual average concentration of hexachloro-1,3-butadiene
among NMP sites sampling this pollutant. This site has the most measured detections
of this pollutant (19); in addition, the second and third highest concentrations of this
pollutant across the program were measured at SSMS.
•	SSMS and KMMS rank fifth and seventh, respectively, for their annual average
concentrations of />dichlorobenzene.
18.4.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for the pollutants shaded in
gray in Table 18-4 for KMMS and SSMS. Figures 18-6 through 18-13 overlay the sites'
18-18

-------
minimum, annual average, and maximum concentrations onto the program-level minimum, first
quartile, median, average, third quartile, and maximum concentrations, as described in
Section 3.4.3.1.
Figure 18-6. Program vs. Site-Specific Average Benzene Concentrations

Wr
Program Max Concentration = 43.5 ^ig/m3
1 1 1



-o

Program Max Concentration = 43.5 jig/m3



0	2	4	6	8	10	12
Concentration {[ig/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range


Figure 18-7. Program vs. Site-Specific Average 1,3-Butadiene Concentrations

Program Max Concentration = 21.5 ^ig/m3





Program Max Concentration = 21.5 ^ig/m3
,
O
0



0	0.3	0.6	0.9	1.2	1.5
Concentration {[ig/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range


18-19

-------
Figure 18-8. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations
Program Max Concentration = 23.7 ^ig/m3
Program Max Concentration = 23.7 ^ig/m3
0.75	1
Concentration {[jg/m3]
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


Figure 18-9. Program vs. Site-Specific Average />-Dichlorobenzene Concentrations
0
0.1
0.2
0.3 0.4
Concentration {[jg/m3)
0.5
0.6

Program:
1st Qua rti le
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i

Site:
Site Average
o
Site Concentration Range


18-20

-------
Figure 18-10. Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations
¦
Program Max Concentration = 111 ^ig/m3
¦
Program Max Concentration = 111 ^ig/m3
0
0.2
0.4 0.6
Concentration {[jg/m3)

0.8

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1

Site: Site Average
o
Site Concentration Range


Figure 18-11. Program vs. Site-Specific Average Ethylbenzene Concentrations





Program Max Concentration = 18.7 ^ig/m3
1
u	



Program Max Concentration = 18.7 ^ig/m3
0
1 2
3
Concentration {[jg/m3)
4
5

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i

Site: Site Average
o
Site Concentration Range


Figure 18-12. Program vs. Site-Specific Average Hexachloro-l,3-Butadiene Concentration
0.05
0.1
0.15
Concentration {[jg/m3)
0.2
0.25
Program:
1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site:
Site Average
o
Site Concentration Range


18-21

-------
Figure 18-13. Program vs. Site-Specific Average Xylenes Concentration
Program Max Concentration = 117 ^ig/m3
KMMS
o
o
5
10
15
Concentration {[jg/m3)
20
25
30
Program: IstQuartile
2ndQuartile 3rdQuartile 4thQuartile Average
~ ~ ~
Site:
Site Average Site Concentration Range
o
Observations from Figures 18-6 through 18-13 include the following:
•	Figure 18-6 presents the box plots for benzene for both sites. The program-level
maximum benzene concentration (43.5 |ig/m3) is not shown directly on the box
plots as the scale has been reduced to 12 |ig/m3 in Figure 18-6 to allow for the
observation of data points at the lower end of the concentration range. The range
of benzene concentrations measured at SSMS is greater than those measured at
KMMS, although the entire range of benzene concentrations measured at both
sites is less than 2 |ig/m3. The annual average concentration for SSMS is slightly
higher than the annual average concentration for KMMS, although both are less
than the program-level average concentration.
•	Figure 18-7 presents the box plots for 1,3-butadiene for both sites. Similar to
benzene, the program-level maximum 1,3-butadiene concentration (21.5 |ig/m3) is
not shown directly on the box plots as the scale has been reduced to 1.5 |ig/m3 in
Figure 18-7 to allow for the observation of data points at the lower end of the
concentration range. The range of 1,3-butadiene concentrations measured is larger
for SSMS than KMMS, although all 1,3-butadiene concentrations measured at
these sites are less than 0.4 |ig/m3. The annual average concentration of
1,3-butadiene is slightly higher for SSMS than KMMS, although both are less
than the program-level average concentration. However, the program-level
average concentration 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.
•	Figure 18-8 presents the box plots for carbon tetrachloride. The scale of these box
plots has also been reduced to allow for the observation of data points at the lower
end of the concentration range, as the program-level maximum carbon
tetrachloride concentration (23.7 |ig/m3) is considerably greater than the majority
of measurements. Figure 18-8 shows that the range of carbon tetrachloride
concentrations measured at the Mississippi sites are considerably less than the
range of concentrations measured across the program. Even though the range of
concentrations measured at SSMS is larger than the range of concentrations
measured at KMMS, the annual average carbon tetrachloride concentrations are
18-22

-------
the same for both sites, and both are negligibly greater than the program-level
average concentration.
Figure 18-9 presents the box plots for/>-dichlorobenzene for KMMS and SSMS.
The program-level first and second (median) quartiles are both zero and therefore
not visible on the box plots. This is due to the large number of non-detects of this
pollutant across the program. Similar to several other VOCs, the range of
concentrations measured was larger for SSMS than KMMS. The annual average
concentration was slightly higher for SSMS than for KMMS, although both are
greater than the program-level average concentration.
The scale of the box plots in Figure 18-10 have also been reduced to allow for the
observation of data points at the lower end of the concentration range, as the
program-level maximum 1,2-dichloroethane concentration (111 |ig/m3) is
considerably greater than the majority of measurements. All of the concentrations
of 1,2-dichloroethane measured at the Mississippi sites are less than the program-
level average concentration of 0.26 |ig/m3. The annual average concentrations for
both Mississippi sites are just less than the program-level median concentration.
This is another example of measurements at the upper end of the concentration
range driving the program-level average concentration, as the program-level
average is more than twice the program-level third quartile.
Figure 18-11 presents the box plots for ethylbenzene for KMMS and SSMS. The
scale of the box plots in Figure 18-11 have also been reduced to allow for the
observation of data points at the lower end of the concentration range, as the
program-level maximum ethylbenzene concentration (18.7 |ig/m3) is considerably
greater than the majority of measurements. The maximum concentration of
ethylbenzene across the program was measured at KMMS and is more than six
times greater than the maximum ethylbenzene concentration measured at SSMS.
The annual average concentration for SSMS is just greater than the program-level
average concentration while the annual average for KMMS is more than five
times greater than the program-level average concentration.
Figure 18-12 is the box plot for hexachloro-1,3-butadiene for SSMS (this
pollutant is not a pollutant of interest for KMMS). 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
(82 percent). Although the maximum hexachloro-1,3-butadiene concentration was
not measured at SSMS, the second and third highest concentrations were. Recall
from the previous section that this site has the highest annual average hexachloro-
1,3-butadiene concentration among NMP sites sampling VOCs. The annual
average concentration for SSMS is nearly twice the program-level average
concentration. It should be noted, however, that none of the measured detections
of hexachloro-1,3-butadiene across the program were greater than the MDL for
this pollutant.
18-23

-------
• Figure 18-13 is the box plot for xylenes for KMMS, the only NMP site for which
this is a pollutant of interest. The scale of the box plot in Figure 18-13 has also
been reduced to allow for the observation of data points at the lower end of the
concentration range, as the program-level maximum xylenes concentration
(117 |ig/m3) is considerably greater than the majority of measurements. The
maximum xylenes concentration measured across the program was measured at
KMMS, and a second similar concentration was also measured at this site. The
annual average xylenes concentration for KMMS is more than eight times greater
than the program-level average concentration. This is another example of
measurements at the upper end of the concentration range driving the program-
level average concentration, as the program-level average concentration is greater
than the program-level third quartile and more than twice the program-level
median concentration.
18.4.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.3.2.
Because the two Columbus, Mississippi sites are part of a 1-year monitoring effort completed at
the end of the 2013, a trends analysis could not be conducted for these sites.
18.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at each Mississippi monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4
for definitions and explanations regarding the various toxicity factors, time frames, and
calculations associated with these risk-based screenings.
18.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Mississippi monitoring sites and where annual
average concentrations could be calculated, risk was examined by calculating cancer risk and
noncancer hazard approximations. These approximations can be used as risk estimates for cancer
and noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers want to shift their air
monitoring priorities. Refer to Section 3.4.3.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 18-6, where applicable. Cancer risk approximations are
18-24

-------
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Observations for KMMS from Table 18-6 include the following:
•	The pollutants of interest with the highest annual average concentrations for KMMS
are xylenes, ethylbenzene, carbon tetrachloride, and benzene.
•	Based on the annual averages and cancer UREs, ethylbenzene has the highest cancer
risk approximation (4.87 in-a-million), followed by benzene (4.15 in-a-million), and
carbon tetrachloride (3.98 in-a-million). Note that xylenes do not have a cancer URE.
•	None of the pollutants of interest for KMMS have noncancer hazard approximations
greater than 1.0, indicating that no adverse noncancer health effects are expected from
these individual pollutants. The pollutant with the highest noncancer hazard
approximation for KMMS is xylenes (0.12).
Observations for SSMS from Table 18-6 include the following:
•	The pollutants with the highest annual average concentrations are carbon
tetrachloride, benzene, and ethylbenzene, although none of these pollutants have an
annual average concentration greater than 1 |ig/m3.
•	Based on the annual averages and cancer UREs, benzene has the highest cancer risk
approximation for SSMS (4.71 in-a-million), followed by carbon tetrachloride
(3.97 in-a-million), and 1,3-butadiene (2.37 in-a-million).
•	None of the pollutants of interest for SSMS have noncancer hazard approximations
greater than 1.0, indicating that no adverse noncancer health effects are expected from
these individual pollutants. The pollutant with the highest noncancer hazard
approximation for SSMS is 1,3-butadiene (0.04).
18-25

-------
Table 18-6. Risk Approximations for the Mississippi Monitoring Sites
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections vs.
# of Samples
Annual
Average
frig/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Kerr-McGee, Columbus, Mississippi - KMMS
Benzene
0.0000078
0.03
60/60
0.53
±0.06
4.15
0.02
1.3 -Butadiene
0.00003
0.002
51/60
0.06
±0.01
1.74
0.03
Carbon Tetrachloride
0.000006
0.1
60/60
0.66
±0.02
3.98
0.01
/?-Dichlorobcnzcnc
0.000011
0.8
39/60
0.05
±0.01
0.58
<0.01
1,2 -Dichloroethane
0.000026
2.4
57/60
0.08
±0.01
2.02
<0.01
Ethylbenzene
0.0000025
1
60/60
1.95
± 1.08
4.87
<0.01
Xylenes
_
0.1
60/60
12.33
±7.20
_
0.12
Acenaphthene3
0.000088

30/30
NA
NA
NA
Fluorene3
0.000088

30/30
NA
NA
NA
Naphthalene1
0.000034
0.003
30/30
NA
NA
NA
Stokes-Beard Elementary School, Columbus, Mississippi - SSMS
Benzene
0.0000078
0.03
61/61
0.60
±0.07
4.71
0.02
1.3 -Butadiene
0.00003
0.002
58/61
0.08
±0.02
2.37
0.04
Carbon Tetrachloride
0.000006
0.1
61/61
0.66
±0.02
3.97
0.01
/?-Dichlorobcnzcnc
0.000011
0.8
48/61
0.08
±0.02
0.83
<0.01
1,2 -Dichloroethane
0.000026
2.4
58/61
0.08
±0.01
2.15
<0.01
Ethylbenzene
0.0000025
1
61/61
0.39
±0.12
0.98
<0.01
Hexachloro -1,3 -butadiene
0.000022
0.09
19/61
0.03
±0.01
0.60
<0.01
— = A Cancer URE or Noncancer RfC is not available.
NA = Not available due to the criteria for calculating an annual average.
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of
viewing.
18-26

-------
18.5.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 18-7 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
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 3.4.3.4. Lastly,
Table 18-7 provides the 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 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 18.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
18-27

-------
Table 18-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Mississippi Monitoring Sites
Top 10 Total Emissions for Pollutants with
Cancer UREs
Top 10 Cancer Toxicity-Weighted
Emissions
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(County-Level)
(County-Level)

(Site-Specific)



Cancer

Cancer Risk
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(in-a-million)
Kerr-McGee, Columbus, Mississippi (Lowndes County) - KMMS
Ethylbenzene
50.88
Formaldehyde
4.89E-04
Ethylbenzene
4.87
Acetaldehyde
47.06
Nickel, PM
3.75E-04
Benzene
4.15
Benzene
38.47
Benzene
3.00E-04
Carbon Tetrachloride
3.98
Formaldehyde
37.64
Naphthalene
2.68E-04
1,2-Dichloroethane
2.02
Naphthalene
7.88
Hexavalent Chromium
2.19E-04
1,3-Butadiene
1.74
1.3 -Butadiene
5.07
Arsenic, PM
1.60E-04
/?-Dichlorobcnzcnc
0.58
T etrachloroethylene
3.81
1,3-Butadiene
1.52E-04


Dichloro methane
1.23
Ethylbenzene
1.27E-04


Nickel, PM
0.78
Acetaldehyde
1.04E-04


POM, Group 2b
0.46
POM, Group 2b
4.02E-05


Stokes-Beard Elementary School, Columbus, Mississippi (Lowndes County) - SSMS
Ethylbenzene
50.88
Formaldehyde
4.89E-04
Benzene
4.71
Acetaldehyde
47.06
Nickel, PM
3.75E-04
Carbon Tetrachloride
3.97
Benzene
38.47
Benzene
3.00E-04
1,3-Butadiene
2.37
Formaldehyde
37.64
Naphthalene
2.68E-04
1,2-Dichloroethane
2.15
Naphthalene
7.88
Hexavalent Chromium
2.19E-04
Ethylbenzene
0.98
1,3-Butadiene
5.07
Arsenic, PM
1.60E-04
/?-Dichlorobcnzcnc
0.83
T etrachloroethylene
3.81
1,3-Butadiene
1.52E-04
Hexachloro-1,3 -butadiene
0.60
Dichloro methane
1.23
Ethylbenzene
1.27E-04


Nickel, PM
0.78
Acetaldehyde
1.04E-04


POM, Group 2b
0.46
POM, Group 2b
4.02E-05



-------
Table 18-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Mississippi Monitoring Sites
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
Top 10 Noncancer Toxicity-Weighted
Emissions
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(County-Level)

(County-Level)
(Site-Specific)



Noncancer

Noncancer Hazard
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(HQ)
Kerr-McGee, Columbus, Mississippi (Lowndes County) - KMMS
Methanol
881.01
Acrolein
113,667.50
Xylenes
0.12
Xylenes
231.91
Chlorine
66,619.28
1,3-Butadiene
0.03
Toluene
150.47
Manganese, PM
19,091.49
Benzene
0.02
Hexane
83.91
Nickel, PM
8,690.62
Carbon Tetrachloride
0.01
Ethylbenzene
50.88
Acetaldehyde
5,229.22
Ethylbenzene
<0.01
Acetaldehyde
47.06
Formaldehyde
3,840.61
/?-Dichlorobcnzcnc
<0.01
Hydrochloric acid
38.52
Naphthalene
2,627.71
1,2-Dichloroethane
<0.01
Benzene
38.47
1,3-Butadiene
2,533.29


Formaldehyde
37.64
Arsenic, PM
2,474.52


Methyl isobutyl ketone
16.64
Xylenes
2,319.10


Stokes-Beard Elementary School, Columbus, Mississippi (Lowndes County) - SSMS
Methanol
881.01
Acrolein
113,667.50
1,3-Butadiene
0.04
Xylenes
231.91
Chlorine
66,619.28
Benzene
0.02
Toluene
150.47
Manganese, PM
19,091.49
Carbon Tetrachloride
0.01
Hexane
83.91
Nickel, PM
8,690.62
Ethylbenzene
<0.01
Ethylbenzene
50.88
Acetaldehyde
5,229.22
Hexachloro-1,3 -butadiene
<0.01
Acetaldehyde
47.06
Formaldehyde
3,840.61
/?-Dichlorobcnzcnc
<0.01
Hydrochloric acid
38.52
Naphthalene
2,627.71
1,2-Dichloroethane
<0.01
Benzene
38.47
1,3-Butadiene
2,533.29


Formaldehyde
37.64
Arsenic, PM
2,474.52


Methyl isobutyl ketone
16.64
Xylenes
2,319.10



-------
Observations from Table 18-7 include the following:
•	Ethylbenzene is the highest emitted pollutants with a cancer URE in Lowndes
County, followed by acetaldehyde and benzene. Although the quantity of emissions is
not very high, this is the only county with an NMP site for which ethylbenzene ranks
highest for quantity emitted.
•	The pollutants with the highest toxicity-weighted emissions (of the pollutants with
cancer UREs) are formaldehyde, nickel, and benzene.
•	Eight of the highest emitted pollutants in Lowndes County also have the highest
toxicity-weighted emissions.
•	Ethylbenzene has the highest cancer risk approximation for KMMS, which ranks
highest for total emissions and eighth highest for toxicity-weighted emissions.
Benzene and 1,3-butadiene also appear among the pollutants with the highest cancer
risk approximations and on both emissions-based lists. The remaining pollutants of
interest for KMMS appear on neither emissions-based list. Similar observations can
be made for SSMS.
•	Naphthalene is the fifth highest emitted pollutant in Lowndes County and ranks
fourth for its toxicity-weighted emissions. Naphthalene was sampled for at KMMS
and was identified as a pollutant of interest, but PAHs were not long enough for an
annual average concentration, and thus, risk approximations, to be calculated. POM,
Group 2b ranks 10th highest for both its total emissions and its toxicity-weighted
emissions. POM, Group 2b includes several PAHs sampled for at KMMS, including
acenaphthene and fluorene, both of which were also identified as pollutants of interest
for KMMS.
•	Several metals and carbonyl compounds appear among the highest emitted pollutants
in Lowndes County and have some of the highest toxicity-weighted emissions.
Speciated metals and carbonyl compounds were not sampled for at KMMS or SSMS
as part of this monitoring effort.
Observations from Table 18-8 include the following:
•	Methanol, xylenes, and toluene are the highest emitted pollutants with noncancer
RfCs in Lowndes County.
•	The pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) are acrolein, chlorine, and manganese.
•	Three of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Lowndes County.
•	Acrolein has the highest toxicity-weighted emissions for Lowndes 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
18-30

-------
questions about the consistency and reliability of the measurements, as discussed in
Section 3.2.
•	Xylenes have the highest noncancer hazard approximation for KMMS, followed by
1,3-butadiene and benzene. 1,3-Butadiene and benzene have the highest noncancer
hazard approximations for SSMS (xylenes were not identified as a pollutant of
interest for SSMS). However, all noncancer hazard approximations calculated for the
Mississippi sites are considerably less than an HQ of 1.0 (all are less than 0.15).
•	Of the pollutants of interest for KMMS and SSMS, xylenes, benzene, and
ethylbenzene appear among the highest emitted pollutants in Lowndes County while
only xylenes and 1,3-butadiene appear among those with the highest toxicity-
weighted emissions (of the pollutants with noncancer RfCs).
•	Several metals and carbonyl compounds appear among the highest emitted pollutants
in Lowndes County and have some of the highest toxicity-weighted emissions.
Speciated metals and carbonyl compounds were not sampled at KMMS or SSMS as
part of this monitoring effort.
18.6 Summary of the 2013 Monitoring Data for KMMS and SSMS
Results from several of the data treatments described in this section include the
following:
~~~ Fourteen pollutants failed screens for KMMS; 10 pollutants failed screens for SSMS.
~~~ Of the site-specific pollutants of interest for KMMS, xylenes had the highest annual
average concentration. KMMS is the only pollutant for which xylenes was identified
as a pollutant of interest, with several of the highest concentrations of this pollutant
across the program measured at this site. For SSMS, carbon tetrachloride had the
highest annual average concentration among this site's pollutants of interest.
~~~ Compared to other sites sampling VOCs, KMMS has the highest annual average
concentration of ethylbenzene, with several of the highest concentrations of this
pollutant across the program measured at this site. SSMS has the highest annual
average concentration of hexachloro-1,3-butadiene among NMP sites sampling
VOCs.
~~~ Ethylbenzene has the highest cancer risk approximation of the pollutants of interest
for KMMS while benzene has the highest cancer risk approximation of the pollutants
of interest for SSMS. None of the pollutants of interest for the Mississippi sites have
noncancer hazard approximations greater than an HQ of 1.0.
18-31

-------
19.0	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.
19.1	Site Characterization
This section characterizes the S4MO monitoring site by providing geographical and
physical information about the location of the site and the surrounding area. This information is
provided to give the reader insight regarding factors that may influence the air quality near the
site and assist in the interpretation of the ambient monitoring measurements.
The S4MO monitoring site is located in the St. Louis, MO-IL CBSA. Figure 19-1 is a
composite satellite image retrieved from ArcGIS Explorer showing the monitoring site and its
immediate surroundings. Figure 19-2 identifies nearby point source emissions locations by
source category, as reported in the 2011 NEI for point sources, version 2. Note that only sources
within 10 miles of the site are included in the facility counts provided in Figure 19-2. A 10-mile
boundary was chosen to give the reader an indication of which emissions sources and emissions
source categories could potentially have a direct effect on the air quality at the monitoring site.
Further, this boundary provides both the proximity of emissions sources to the monitoring site as
well as the quantity of such sources within a given distance of the site. Sources outside the
10-mile boundary are still visible on the map for reference, but have been grayed out in order to
emphasize emissions sources within the boundary. Table 19-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates.
19-1

-------
Figure 19-1. St. Louis, Missouri (S4MO) Monitoring Site
to

-------
Figure 19-2. NEI Point Sources Located Within 10 Miles of S4MO
S? loiw
County
Y	WSippor OtmtmtQU
j)	P«wy£*-XVHM V* AM# P*" •«
~	111
0	Bu IdrrtiLcra Pacini 4<
B	M f^n—aPVMli 4
C uttuiKWD) r«»* CZJ1
<«M8tft«ryi2v
, - | :.m>emr>	i|1>
 G«*crMoFac«<\
0	D«rw»fft»
f	Baano* CtwitMHo* in GoabuMc* 1|
E »»liw»| »w«(g «««
f	ftUMS	• I
F	coou Prco»«nj'*3TOj«v» Fkx» ir>
i	ctmiw> «o* ma 9w« lil
/iK	hmtai *tonlairai» f»|
%	CmammOmm S4ivo» win 471
•	WWW WwM» ) (V (quern*! n»« 0f
O	mmum** itcAoei acuin *tun Mt S
«	:«*«£}
qp	J«f»c«rd lmVm« PiMjcfeF«?*t Iti
¦	Mtto Cm Eo« «id Orm HUM Ctn—i M*-ukwui«<,j. 1 •¦
A	to+m) fc«4M tIM	». M4»uWUi«-» M C«p*I fW III
A	t^rl'fKiWT>i>»no1g>iwi
4	«» TIB«»nff 1 Ffltfl?
Source Category Group (No. of Facilities)
S4MO NATTS site
10 mile radius
County boundary
mm	wismv	»tmrw	arwvi	k-wa
MM
^~~	T m
80"25WV	«yw*rw
Legend
acicnrw	arstrw
Hots. Due to faculty density anc oolloc-abon the total faolwea
4i«p*4y»
-------
Table 19-1. Geographical Information for the Missouri Monitoring Site
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Additional Ambient Monitoring Information1
S4MO
29-510-0085
St. Louis
St. Louis
City
St. Louis, MO-IL
38.656449,
-90.198548
Residential
Urban/City
Center
TSP Lead, CO, S02, N02, NOx, NOy, NO, 03,
Meteorological parameters, PMio, PM Coarse, Black
carbon PM2.5, PM2.5 Speciation, SO2, IMPROVE
Speciation.
1 Data for additional pollutants are reported to AQS for this site (EPA, 2014b); however, these data are not generated by ERG and are therefore not included in this report.
BOLD ITALICS = EPA-designated NATTS Site
vo

-------
S4M0 is located in central St. Louis. Figure 19-1 shows that the S4MO monitoring site is
located less than one-quarter mile west of 1-70. The Mississippi River, which separates Missouri
and Illinois, is less than 1 mile east of the site. Although the area directly around the monitoring
site is primarily residential, industrial facilities are located nearby, primarily just on the other
side of 1-70. Figure 19-2 shows that a large number of point sources are located within 10 miles
of S4MO, particularly on the east side of the Missouri/Illinois border. The source categories with
the greatest number of point sources surrounding S4MO include chemical manufacturing
facilities; airport and airport support operations, which include airports and related operations as
well as small runways and heliports, such as those associated with hospitals or television
stations; mines, quarries, and mineral processing facilities; and rail yard/rail line operations.
Within 1 mile of S4MO are a pharmaceutical manufacturing facility, a printing and publishing
facility, a leather products facility, a metals processing/fabrication facility, and a chemical
manufacturing facility.
Table 19-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Missouri monitoring site. Table 19-2 includes both county-level
population and vehicle registration information. Table 19-2 also contains traffic volume
information for S4MO as well as the location for which the traffic volume was obtained.
Additionally, Table 19-2 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 19-2 includes the combination of the city and county data for county-level statistics in
order to compare these statistics with other NMP sites' county-level data.
Table 19-2. Population, Motor Vehicle, and Traffic Information for the Missouri
Monitoring Site
Site
County
Estimated
County
Population1
County-level
Vehicle
Registration2
Annual
Average Daily
Traffic3
Intersection
Used for Traffic Data
County-
level Daily
VMT4 "
S4MO
St. Louis City
+ County
1,319,860
1,117,375
100,179
1-70 at 1-44 split (at bridge)
24,065,245
bounty-level population estimate reflects county and city data for 2013 (Census Bureau, 2014)
2County-level vehicle registration reflects county and city data for 2013 (MO DOR, 2014)
3 AADT reflects 2013 data (MO DOT, 2013)
4County-level VMT reflects county and city data for 2013 (MO DOT, 2014)
BOLD ITALICS = EPA-designated NATTS Site
19-5

-------
Observations from Table 19-2 include the following:
•	S4MO's county-level population and vehicle registration rank 11th and 12th highest,
respectively, compared to other counties with NMP sites.
•	The traffic volume experienced near S4MO ranks 14th, which falls in the upper third
of the range compared to other NMP sites. The traffic estimate provided is for 1-70
near the split with 1-44 (at the bridge).
•	The VMT for S4MO is roughly 24 million miles, ranking 13th among counties with
NMP sites.
19.2 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.
19.2.1	Climate Summary
The city of St. Louis is located along the Mississippi River, which is 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 moving southward from Canada and the northern U.S.,
resulting in weather patterns that are relatively short in duration. 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, 2015).
19.2.2	Meteorological Summary
Hourly meteorological data for 2013 were retrieved from NCDC for the weather station
closest to the Missouri monitoring site (NCDC, 2013), as described in Section 3.4.2. 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 19-3. These data were used to determine how meteorological
conditions on sample days vary from conditions experienced throughout the year.
19-6

-------
Table 19-3. 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)
6.3
miles
Sample
Davs
(62)
62.5
±5.5
53.4
±5.2
43.1
±5.2
48.3
±4.8
70.5
±2.6
1018.5
± 1.9
5.4
±0.6
159°
(SSE)
2013
64.8
±2.1
54.8
± 1.9
43.8
± 1.9
49.3
± 1.8
69.2
± 1.1
1017.9
±0.7
5.6
±0.3
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.

-------
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 all of 2013. Also included in Table 19-3 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 2013, the temperatures appear slightly cooler on sample days, as shown in
Table 19-3. Few of the hottest days in 2013 were sample days at S4MO.
19.2.3 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 3.4.2. A wind rose shows the frequency of wind directions using "petals"
positioned around a 16-point compass, and uses different colors to represent wind speeds.
Figure 19-3 presents 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 19-3 also presents three different wind roses for the
S4MO monitoring site. First, a historical wind rose representing 2003 to 2012 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 2013 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2013 is
presented. These can be used to identify the predominant wind speed and direction for 2013 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.
19-8

-------
Figure 19-3. Wind Roses for the St. Louis Downtown Airport Weather Station near S4MO
Location of S4MO and Weather Station
2003-2012 Historical Wind Rose


\w

N* 	/AJ ]
¦P
1*^^. '*' * \ im**'
*2» St Loul* \ £

" / w

• / / V

/ \f "



ff / / \ /

f \
J iktfrt

WOJBWH
Station
L
Ti""
NORTH""-.
west:
'A'INC SPEEC
(Kn ots i
11 -17
SOUTH

3 aim; 19.28%
2013 Wind Rose
Sample Day Wind Rose
NORTH"''-
ES"
WIN C S PE EC
(Knots)
17-21
11 - 17
SOUTH
Calms: 21.82%
;
WW C SPEED
i, Knots)
SOUTH
19-9

-------
Observations from Figure 19-3 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 Cahokia, Illinois.
•	The historical wind rose shows that winds from the southeast, south-southeast, and
south were frequently observed near S4MO, with prevailing winds from the south-
southeast. Winds from these directions account for approximately 28 percent of
observations. Calm winds (those less than or equal to 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 2013 wind rose generally resemble those shown on
the historical wind rose, although the percentage of calm winds was slightly higher in
2013 (22 percent).
•	The wind patterns on the sample day wind rose mostly resemble the historical and
full-year wind roses, although there are a few differences. Fewer winds from the
southeast, northwest, and north-northwest were observed while winds from the south
were observed more frequently.
19.3 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 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. VOCs, PAHs, carbonyl compounds, metals (PMio), and hexavalent chromium were
sampled for at S4MO. S4MO is one of two NATTS sites that sampled hexavalent chromium
year-round.
19-10

-------
Table 19-4. Risk-Based Screening Results for the Missouri Monitoring Site
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
St. Louis, Missouri - S4MO
Acetaldehyde
0.45
61
61
100.00
10.63
10.63
Benzene
0.13
61
61
100.00
10.63
21.25
Carbon Tetrachloride
0.17
61
61
100.00
10.63
31.88
Formaldehyde
0.077
61
61
100.00
10.63
42.51
Arsenic (PMio)
0.00023
60
61
98.36
10.45
52.96
1,2-Dichloroethane
0.038
58
58
100.00
10.10
63.07
1.3 -Butadiene
0.03
56
57
98.25
9.76
72.82
Naphthalene
0.029
54
60
90.00
9.41
82.23
p-Dichlorobenzene
0.091
21
47
44.68
3.66
85.89
Cadmium (PMio)
0.00056
20
61
32.79
3.48
89.37
Hexacliloro -1,3 -butadiene
0.045
16
17
94.12
2.79
92.16
Lead (PMio)
0.015
12
61
19.67
2.09
94.25
Acenaphthene
0.011
7
60
11.67
1.22
95.47
Fluorene
0.011
7
60
11.67
1.22
96.69
Nickel (PMio)
0.0021
7
61
11.48
1.22
97.91
Ethylbenzene
0.4
3
61
4.92
0.52
98.43
Manganese (PMio)
0.03
3
61
4.92
0.52
98.95
Propionaldehyde
0.8
3
61
4.92
0.52
99.48
Acenaphthylene
0.011
1
36
2.78
0.17
99.65
Benzo(a)pyrene
0.00057
1
57
1.75
0.17
99.83
1,2-Dibromoethane
0.0017
1
1
100.00
0.17
100.00
Total
574
1124
51.07


Observations from Table 19-4 include the following:
•	Twenty-one pollutants failed at least one screen for S4MO; 51 percent of
concentrations for these 21 pollutants were greater than their associated risk screening
value (or failed screens). S4MO tied with BTUT for the highest number of individual
pollutants failing screens.
•	Fifteen pollutants contributed to 95 percent of failed screens for S4MO and therefore
were identified as pollutants of interest for this site. These 15 include two carbonyl
compounds, six VOCs, four PMio metals, and three PAHs. Although the 95 percent
criteria is met with acenaphthene, fluorene and nickel are also considered pollutants
of interest because they failed the same number of screens as acenaphthene, per the
steps described in Section 3.2.
•	S4MO has the greatest number of pollutants of interest among NMP sites. Similar to
previous years, S4MO failed the highest number of screens (574) among all NMP
sites (refer to Table 4-8 of Section 4.2). However, the failure rate for S4MO, when
19-11

-------
incorporating all pollutants with screening values, is approximately 21 percent. This
is due primarily to the relatively large 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 also failed 100 percent of screens but was
detected only once.
•	Lead and cadmium were identified as pollutants of interest for only two NMP sites
sampling metals: S4MO and ASKY-M.
19.4 Concentrations
This section presents various concentration averages used to characterize air toxics
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
each monitoring site.
•	Annual average concentrations are presented graphically for each site to illustrate
how the site's concentrations compare to the program-level averages, as presented in
Section 4.1.
•	Concentration averages and other statistical metrics are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at S4MO are provided in Appendices J, L, M, N, and O.
19.4.1 2013 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
19-12

-------
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 19-5, 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.
Table 19-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Missouri Monitoring Site

# of






Measured
1st
2nd
3rd
4th


Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs. # of
Samples
Average
(Ug/m3)
Average
(Ug/m3)
Average
(Ug/m3)
Average
(Ug/m3)
Average
(Ug/m3)
St. Louis, Missouri - S4MO


1.21
2.01
2.73
1.96
1.98
Acetaldehyde
61/61
±0.16
±0.49
±0.37
±0.32
±0.22


0.78
0.50
0.56
0.59
0.61
Benzene
61/61
±0.08
±0.09
±0.08
±0.11
±0.05


0.06
0.05
0.06
0.08
0.07
1.3 -Butadiene
57/61
±0.02
±0.02
±0.01
±0.02
±0.01


0.64
0.69
0.67
0.56
0.64
Carbon Tetrachloride
61/61
±0.05
±0.04
±0.04
±0.06
±0.03


0.06
0.09
0.12
0.10
0.09
p-Dichlorobenzene
47/61
±0.06
±0.04
±0.05
±0.04
±0.02


0.10
0.11
0.06
0.09
0.09
1,2-Dichloroethane
58/61
±0.01
±0.02
±0.02
±0.01
±0.01


1.65
3.40
5.73
2.22
3.23
Formaldehyde
61/61
±0.29
± 1.02
± 1.24
±0.31
±0.55


0.01
0.02
0.01
0.05
0.02
Hexachloro-1,3 -butadiene
17/61
±0.02
±0.02
±0.02
±0.03
±0.01


2.64
5.45
9.68
2.32
5.02
Acenaphthene3
60/60
± 1.25
± 1.83
±2.68
± 1.31
± 1.16


0.65
0.75
1.01
0.53
0.73
Arsenic (PMi0)a
61/61
±0.13
±0.13
±0.16
±0.14
±0.08


0.33
0.56
0.75
0.60
0.56
Cadmium (PMi0)a
61/61
±0.15
±0.36
±0.52
±0.33
±0.18


3.50
6.34
10.50
2.80
5.79
Fluorene3
60/60
± 1.00
± 1.69
±2.27
±0.80
± 1.07


7.12
8.96
14.13
7.53
9.40
Lead (PMi0)a
61/61
±2.18
±3.93
±5.99
±3.00
±2.01


73.68
61.16
90.49
62.34
71.92
Naphthalene3
60/60
±24.15
± 16.74
±23.65
± 15.98
± 10.08


1.07
1.24
1.36
0.61
1.06
Nickel (PMi,;,)a
61/61
±0.45
±0.58
±0.78
±0.15
±0.26
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
19-13

-------
Observations for S4MO from Table 19-5 include the following:
•	The pollutants with the highest annual average concentrations are formaldehyde
(3.23 ± 0.55 |ig/m3) and acetaldehyde (1.98 ± 0.22 |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 range from
0.90 |ig/m3 to 9.86 |ig/m3. The 19 highest concentrations were measured between
May and September, or the warmest months of the year, with six measured during the
second quarter and 13 measured during the third quarter. Conversely, 11 of the 13
formaldehyde concentrations less than 1.50 |ig/m3 were measured during the first or
fourth quarters and four of the five lowest concentrations were measured during the
months of February and March.
•	Concentrations of acetaldehyde were markedly less during the first quarter of 2013.
Concentrations of acetaldehyde range from 0.711 |ig/m3 to 3.88 |ig/m3, with a median
concentration 1.75 |ig/m3. None of the concentrations measured during the first
quarter were greater than the median concentration.
•	Concentrations of benzene appear highest during the first quarter while
concentrations measured during the fourth quarter have a higher level of variability.
A review of the data shows that the maximum benzene concentration (1.01 |ig/m3)
was measured at S4MO in November, while the next five highest concentrations were
all measured between January and March. Two of the three lowest benzene
concentrations were also measured during the fourth quarter of 2013.
•	The third quarter average concentration of />dichlorobenzene is twice the magnitude
of the first quarter average, although the confidence interval is largest for the first
quarter average. A review of the data shows that concentrations of />dichlorobenzene
measured at S4MO range from 0.0301 |ig/m3 to 0.458 |ig/m3 and include 14 non-
detects. The maximum p-dichlorobenzene concentration was measured at S4MO in
January and is the seventh highest/>dichlorobenzene concentration measured across
the program. Only one of the 19 /;-dichlorobenzene concentrations greater than
0.1 |ig/m3 was measured during the first quarter (compared to five or more for each of
the remaining calendar quarters). In addition, half of the 14 non-detects were
measured during the first quarter, while no more than three were measured during the
remaining calendar quarters.
•	Concentrations of 1,2-dichloroethane appear higher during the first half of the year.
Of the 21 concentrations of 1,2-dichloroethane greater than 0.1 |ig/m3 measured at
S4MO, 16 were measured during the first and second quarters of 2013, including
every sample day in April. By comparison, only one concentration greater than
0.1 |ig/m3 was measured during the third quarter and four were measured during the
fourth quarter.
19-14

-------
The fourth quarter average concentration of hexachloro-1,3-butadiene is greater than
the other quarterly averages. This pollutant was detected in roughly one-quarter of the
samples collected, with nearly half of the measured detections measured during the
fourth quarter. In addition, four of the five hexachloro-1,3-butadiene concentrations
greater than 0.1 |ig/m3 were measured in November and December, including the
maximum concentration (0.171 |ig/m3), which is the fourth highest hexachloro-1,3-
butadiene concentration measured across the program.
Lead has the highest annual average concentration (9.40 ± 2.01 ng/m3) among the
PMio metals identified as pollutants of interest for S4MO, followed by nickel
(1.06 ± 0.26 ng/m3).
The confidence intervals associated with the quarterly average concentrations of lead
are relatively large, indicating that there is a high level of variability in the
measurements. Concentrations of lead measured at S4MO range from 1.31 ng/m3 to
49.9 ng/m3. The maximum concentration measured at S4MO is the highest
concentration of lead measured among all NMP sites sampling metals and four of the
seven concentrations greater than 20 ng/m3 across the program were measured at
S4MO (of which three of the four were measured in July and August).
The quarterly averages of cadmium and nickel also reflect a high level of variability,
particularly the third quarter, based on the associated confidence intervals.
Concentrations of cadmium measured at S4MO span two orders of magnitude,
ranging from 0.055 ng/m3 to 4.08 ng/m3. The maximum concentration of cadmium
measured at S4MO is the third highest cadmium concentration measured across the
program and more than one-third of the 21 cadmium concentrations greater than
1 ng/m3 measured across the program were measured at S4MO (eight), the most of
any NMP site.
Concentrations of nickel measured at S4MO range from 0.188 ng/m3 to 6.37 ng/m3.
The maximum concentration of nickel measured at S4MO is among the highest nickel
concentrations measured across the program. The fourth quarter average
concentration of nickel is considerably less than the other quarterly averages and has
a considerably smaller confidence interval. The fourth quarter includes the fewest
number of nickel concentrations greater than 1 ng/m3 (one), compared to the other
quarterly averages, which include between five and six each. Conversely, the fourth
quarter includes the largest number of nickel concentrations less than 1 ng/m3 (15),
compared to the other quarterly averages, which include between nine and 10 each.
Concentrations of arsenic appear highest during the third quarter of 2013 at S4MO. A
review of the data shows that arsenic concentrations measured at S4MO range from
0.055 ng/m3 to 1.47 ng/m3. The three highest concentrations were all measured during
the third quarter and of the 13 arsenic concentrations greater than 1 ng/m3 measured
at S4MO, eight were measured between July and September.
Naphthalene has the highest annual average concentration among the PAHs identified
as pollutants of interest for S4MO. The highest concentrations were measured during
the third quarter at S4MO, although the confidence interval calculated for the first
19-15

-------
quarter average is larger than the confidence interval for the third quarter.
Concentrations of naphthalene measured at S4MO range from 10.1 ng/m3 to
182 ng/m3. Although six of the 11 naphthalene concentrations greater than 100 ng/m3
measured at S4MO were measured during the third quarter, three others were
measured during the first three sample days of 2013 (January 4th, 10th, and 16th),
including the two highest concentrations. Conversely, two of the four lowest
concentrations were also measured during the first quarter of 2013.
•	Concentrations of acenaphthene and fluorene appear to be highest during the warmer
months of the year, particularly the third quarter, although each of the quarterly
averages exhibit a considerable level of variability associated with them. A review of
the data shows that the maximum concentration of each pollutant was measured on
July 15, 2012 (20.2 ng/m3 and 18.2 ng/m3, respectively). Of the eight concentrations
of acenaphthene greater than 10 ng/m3, all but two were measured between July and
September (with the other two in April and May). Conversely, all but two of the 20
acenaphthene concentrations less than 2 ng/m3 were measured during the first or
fourth quarters of 2013 (eight during the first quarter and 10 during the fourth). For
fluorene, the highest 16 concentrations were measured at S4MO during the second or
third quarters of 2013, with 11 measured between July and September. Conversely,
all nine of the fluorene measurements less than 2.0 ng/m3 were measured in either the
first or fourth quarter of the year.
Tables 4-9 through 4-12 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for S4MO from
those tables include the following:
•	S4MO appears in Tables 4-9 through 4-12 a total of eight times, which ties with
PXSS for the most of any NMP site.
•	S4MO has the second highest annual average concentration of hexachloro-1,3-
butadiene, the fourth highest annual average concentration of />dichlorobenzene, and
the eighth highest annual average concentration of 1,2-dichloroethane, as shown in
Table 4-9.
•	S4MO appears in Table 4-10 for both formaldehyde and acetaldehyde, ranking eighth
and ninth, respectively, among NMP sites sampling carbonyl compounds.
•	S4MO's annual average concentration of acenaphthene ranks sixth highest among
NMP sites sampling PAHs, while this site's annual average concentration of
naphthalene does not rank among the top 10.
•	S4MO has the fifth highest annual average concentration of arsenic and the seventh
highest annual average concentration of nickel among NMP sites sampling PMio
metals.
19-16

-------
19.4.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for the pollutants shaded in
gray in Table 19-4 for S4MO. Figures 19-4 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, as described in Section 3.4.3.1.
Figure 19-4. Program vs. Site-Specific Average Acenaphthene Concentration
E
Program Max Concentration = 123 ng/m3
40	50
Concentration {ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



Figure 19-5. Program vs. Site-Specific Average Acetaldehyde Concentration
s±=
6
Concentration {[jg/m3]
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



19-17

-------
Figure 19-6. Program vs. Site-Specific Average Arsenic (PMio) Concentration
F
0
12 3
4 5 6
Concentration {ng/m3)
7
8

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i

Site: Site Average
o
Site Concentration Range


Figure 19-7. Program vs. Site-Specific Average Benzene Concentration


-
Program Max Concentration = 43.5 ^ig/m3
1 1 1
1 1
0
2 4
6
Concentration {[jg/m3)
8
10

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1

Site: Site Average
o
Site Concentration Range


Figure 19-8. Program vs. Site-Specific Average 1,3-Butadiene Concentration





->1
Program Max Concentration = 21.5 ^ig/m3
,

J



0
0.3
0.6 0.9
Concentration {[jg/m3)

1.2

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i

Site: Site Average
o
Site Concentration Range


19-18

-------
Figure 19-9. Program vs. Site-Specific Average Cadmium (PMio) Concentration
Program Max Concentration = 120 ng/m3
2
4 6
Concentration {ng/m3)


Program: 1st Quartile
2nd Quartile 3rd Quartile
4th Quartile
Average
¦
~
~
~
i
Site: Site Average
Site Concentration Range


o



Figure 19-10. Program vs. Site-Specific Average Carbon Tetrachloride Concentration
Program Max Concentration = 23.7 ^ig/m3
0
0.25
0.5
0.75 1 1.25 1.5
Concentration {[ig/m3)
1.75

Program:
1st Quartile
¦
2nd Quartile 3rd Quartile 4th Quartile
~ ~ ~
Average
i

Site:
Site Average
o
Site Concentration Range

Figure 19-11. Program vs. Site-Specific Average />-Dichlorobenzene Concentration
0
0.1
0.2
0.3 0.4
Concentration {[jg/m3)
0.5
0.6

Program:
1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i

Site:
Site Average
o
Site Concentration Range


19-19

-------
Figure 19-12. Program vs. Site-Specific Average 1,2-Dichloroethane Concentration
¦
Program Max Concentration = 111 ^ig/m3
0.4	0.6
Concentration {[jg/m3)
Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



Figure 19-13. Program vs. Site-Specific Average Fluorene Concentration

40	50	60
Concentration {ng/m3)
Program: IstQuartile
2ndQuartile 3rdQuartile
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



Figure 19-14. Program vs. Site-Specific Average Formaldehyde Concentration
3
6
9 12 15
Concentration {[jg/m3)
18
21
Program:
IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i
Site:
Site Average
o
Site Concentration Range


19-20

-------
Figure 19-15. Program vs. Site-Specific Average Hexachloro-l,3-Butadiene Concentration
0
0.05
0.1
0.15
Concentration {[jg/m3)
0.2
0.25
0.3

Program:
Site:
1st Qua rti le
¦
Site Average
o
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
Site Concentration Range
4th Qua rti le
~
Average
i

Figure 19-16. Program vs. Site-Specific Average Lead (PMio) Concentration
-
10
20 30
Concentration {ng/m3)
40
Program: IstQuartile
2ndQuartile 3rdQuartile
4thQuartile Average
¦
~
~
~ 1
Site: Site Average
Site Concentration Range

o


Figure 19-17. Program vs. Site-Specific Average Naphthalene Concentration
0
100
200
300 400 500
Concentration {ng/m3)
600
700

Program:
1st Qua rti le
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i

Site:
Site Average
o
Site Concentration Range


19-21

-------
Figure 19-18. Program vs. Site-Specific Average Nickel (PMio) Concentration
S4MO
0
5
10
15
20
25
Concentration {ng/m3)
Program: IstQuartile
2ndQuartile 3rdQuartile 4thQuartile Average
~ ~ ~
Site:
Site Average Site Concentration Range
o
Observations from Figures 19-4 through 19-18 include the following:
•	Figure 19-4 is the box plot for acenaphthene. Note that the program-level
maximum concentration (123 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
80 ng/m3. Figure 19-4 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 similar to the program-level average concentration.
Although a few non-detects of acenaphthene were measured across the program,
none were measured at S4MO.
•	Figure 19-5 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 19-6 shows that the maximum arsenic (PMio) concentration measured at
S4MO is about one-sixth the maximum concentration measured across the
program. S4MO's annual average arsenic (PMio) concentration falls between the
program-level average concentration and third quartile. Recall from the previous
section that this site has the fifth highest annual average arsenic concentration
among NMP sites sampling PMio metals.
•	Figure 19-7 is the box plot for benzene. Note that the program-level maximum
concentration (43.5 |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 12 |ig/m3. This
box plot shows that the range of benzene concentrations measured at S4MO is
relatively small compared to the range measured at the program-level. In fact, the
range of benzene concentrations for S4MO is the third smallest among NMP sites
sampling benzene with Method TO-15. The annual average benzene
19-22

-------
concentration for S4MO is less than the program-level average concentration and
similar to the program-level median concentration.
Similar to acenaphthene and benzene, the program-level maximum 1,3-butadiene
concentration (21.5 |ig/m3) is not shown directly on the box plot as the scale has
been reduced to 1.5 |ig/m3 in Figure 19-8 to allow for the observation of data
points at the lower end of the concentration range. Figure 19-8 for 1,3-butadiene
shows that the program-level average concentration is being driven by the highest
concentrations measured at a few sites, as the program average is greater than the
third quartile. The maximum concentration measured at S4MO is similar to the
program-level average concentration. The annual average 1,3-butadiene
concentration for S4MO is just greater than the program-level median
concentration.
The program-level maximum cadmium concentration (120 ng/m3) is not shown
directly on the box plot in Figure 19-9 as the scale has been reduced to 10 ng/m3
to allow for the observation of data points at the lower end of the concentration
range. Even though Figure 19-9 shows that the maximum cadmium (PMio)
concentration measured at S4MO is considerably less than the maximum
concentration measured across the program (4.08 ng/m3), it is the third highest
cadmium concentration among NMP sites sampling PMio metals. (Note that there
is a considerable difference between the highest (120 ng/m3) and second highest
(5.05 ng/m3) cadmium concentrations measured across the program). Of the 21
concentrations greater than 1 ng/m3 measured across the program, eight were
measured at S4MO. S4MO's annual average cadmium concentration is greater
than the program-level average concentration and is the second highest annual
average among NMP sites sampling PMio metals. The minimum concentration
measured at S4MO is greater than the program-level first quartile.
The program-level maximum carbon tetrachloride concentration (23.7 |ig/m3) is
not shown directly on the box plot in Figure 19-10 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 19-10 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 falls between
the program-level median and average concentrations, although only 0.02 |ig/m3
separates these parameters.
Figure 19-11 is the box plot for /;-dichlorobenzene. Note that the first and second
quartiles are zero and therefore not visible on the graph. Although the maximum
/>dichlorom ethane concentration across the program was not measured at S4MO,
the maximum concentration measured at S4MO is among the higher
measurements. The annual average concentration of this pollutant for S4MO is
more than twice the program-level average concentration. S4MO has the fourth
highest annual average concentration of />dichlorobenzene among NMP sites
sampling this pollutant.
19-23

-------
Figure 19-12 is the box plot for 1,2-dichloroethane. The program-level maximum
concentration (111 |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 19-12 shows that the entire range of
1,2-dichloromethane measurements collected at S4MO is less than the program-
level average concentration. The program-level average concentration is being
driven by higher measurements collected at a handful of monitoring sites. The
annual average concentration for S4MO is similar to the median concentration at
the program level. Recall from the previous section that S4MO has the eighth
highest annual average concentration among NMP sites sampling this pollutant.
Figure 19-13 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 5.03 ng/m3. The maximum
concentration measured across the program is significantly higher (99.1 ng/m3).
The annual average concentration of fluorene for S4MO is greater than the
program-level average and third quartile, although the maximum fluorene
concentration measured at S4MO (18.2 ng/m3) is considerably less than the
program-level maximum concentration.
Figure 19-14 for formaldehyde shows that the maximum formaldehyde
concentration measured at S4MO is roughly half the maximum concentration
measured across the program. The annual average concentration for S4MO is
greater than the program-level average concentration but less than the program-
level third quartile and ranks eighth highest among NMP sites sampling
formaldehyde.
Figure 19-15 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, due to the large number of non-detects. The box plot shows that the
maximum concentration measured at S4MO is less than the program-level
maximum concentration, although it is one of the higher measurements across the
program. S4MO ranks second for the most measured detections of this pollutant
(17). The annual average concentration of hexachloro-l,3-butadiene for S4MO is
slightly greater than the program-level average concentration. Recall from the
previous section that S4MO has the second highest annual average concentration
among NMP sites sampling this pollutant, similar to 2012.
Figure 19-16 shows that the majority of the lead measurements program-wide fall
within a relatively small concentration range as indicated by the first, second
(median), and third quartiles, which are relatively close together. The maximum
lead concentration measured at S4MO is the maximum concentration measured
across the program. The minimum lead concentration measured at S4MO
(1.31 ng/m3) is just less than the program-level first quartile and is the only site-
specific minimum lead concentration greater than 1 ng/m3. The annual average
lead (PMio) concentration for S4MO is more than two and half times greater than
19-24

-------
the program-level average concentration. This site has the highest annual average
lead concentration among NMP sites sampling metals.
•	Figure 19-17 is the box plot for naphthalene and shows that the maximum
naphthalene concentration measured at S4MO is considerably less than the
program-level maximum concentration. The annual average naphthalene
concentration for S4MO is just less than the program-level average concentration.
•	Figure 19-18 is the box plot for nickel. The maximum nickel concentration
measured at S4MO is roughly one-third the program-level maximum
concentration, although it is among the higher nickel concentrations measured
across the program. S4MO's annual average nickel concentration is similar to the
program-level average 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 3.4.3.2.
S4MO has sampled VOCs and carbonyl compounds under the NMP since 2002, PMio metals
since 2003, and PAHs since 2008. Thus, Figures 19-19 through 19-33 present the 1-year
statistical metrics for each of the pollutants of interest for S4MO. The statistical metrics
presented for assessing trends include the substitution of zeros for non-detects. If sampling began
mid-year, a minimum of 6 months of sampling is required for inclusion in the trends analysis; in
these cases, a 1-year average concentration is not provided, although the range and percentiles
are still presented.
19-25

-------
Figure 19-19. Yearly Statistical Metrics for Acenaphthene Concentrations Measured at S4MO
20081	2009	2010	2011	2012	2013
Year
0 5th Percentile	— Minimum	— Median	- Maximum	0 95th Percentile ••"¦"/"•Average
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2008.
Observations from Figure 19-19 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 concentration 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 and again for
2011, the median concentration decreased slightly each year.
•	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 5 ng/m3 increased from 20 to 32 from 2011 to 2012 while the number of
concentrations less than 1 ng/m3 decreased from 10 to three from 2011 to 2012.
•	All of the statistical parameters exhibit decreases for 2013.
19-26

-------
Figure 19-20. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at S4MO
Maximum
Concentration for
2004 is 32.5 ^g/m3
2008
Year
O 5th Percentile
- Minimurr
0 95th Percentile
Observations from Figure 19-20 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, 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, with little change for 2013. The range of concentrations
measured is at a minimum for 2013 while the difference between the 5th and 95th
percentiles, or the range within which 90 percent of the measurements fall, is at a
minimum for 2012. The 2-year period from 2012 to 2013 exhibits the least year-to-
year variability in concentrations measured since the onset of sampling.
19-27

-------
Figure 19-21. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at S4MO
Maximum
Concentration for
2007 is 44.1 ng/m-
20031	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2003.
Observations from Figure 19-21 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 concentration is not presented,
although the range of measurements is provided.
•	The maximum arsenic concentration was measured at S4MO on December 26, 2007
(44.1 ng/m3). Five additional arsenic concentrations greater than 10 ng/m3 have been
measured at S4MO (three in 2005 and one each in 2003, 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.
Less variability in the measurements is shown in the last few years of sampling.
•	Most of the statistical parameters are at a minimum for 2013. The range of
measurements, the difference between the 5th and 95th percentiles, and the difference
between the median and 1-year average concentrations are all at a minimum for 2013
(less than 0.05 ng/m3 separates these two parameters, indicating a decrease in
variability associated with the arsenic concentrations measured at S4MO in 2013.
19-28

-------
Figure 19-22. Yearly Statistical Metrics for Benzene Concentrations Measured at S4MO



































¦iiiiii

—o—
¦C'jm

-I
3-









pL
-c
!>—




L-qJ

T

L-qJ





L2"

L2"
LjJ
—Q-
*1£>
2003	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
0 5th Percentile	— Minimum	— Median	- Maximum	0 95th Percentile	Average
Observations from Figure 19-22 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 between 2007 and 2011, when
the 1-year average concentration has a slight undulating pattern, the 1-year average
varied between 0.80 |ig/m3 (2011) and 1.03 |ig/m3 (2010).
•	From 2011 to 2012, the statistical parameters representing the upper end of the
concentration range (the maximum and 95th percentile) increased while the statistical
parameters representing the lower end of the concentration range (the minimum and
5th percentile) decreased, indicating a widening of concentrations measured. Yet, the
1-year average concentration did not change and the median decreased. The number
of concentrations greater than 1 |ig/m3 doubled (from six in 2011 to 12 in 2012) while
the number of concentrations less than 0.5 |ig/m3 increased from two to 11 from 2011
to 2012.
19-29

-------
• With the exception of the minimum concentration, all of the statistical parameters are
at a minimum for 2013. The change in the 1-year average concentrations between
2003 and 2013 represents a 66 percent decrease in the 1-year average concentrations.
Figure 19-23. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at S4MO
£ 0.75
2008
Year
O 5th Percentile
O 95th Percentile
• Average
Observations from Figure 19-23 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 [j,g/m3 that have been measured at
S4MO.
The minimum, 5th percentile, and median concentrations are all zero for 2003 and
2004, indicating that at least 50 percent of the measurements were non-detects. The
number of non-detects decreased after 2004, from a maximum of 43 non-detects in
2004 to a minimum of zero in 2010 and 2012. After 2006, no more than five non-
detects of 1,3-butadiene have been measured at S4MO for any given year.
Between 2004 and 2008, the 1-year average concentration changed very little,
ranging from 0.078 [j,g/m3 (2005) to 0.095 [j,g/m3 (2006). Greater fluctuations are
shown in the years that follow. Years with a higher number of non-detects, as
indicated by a minimum and 5th percentile of zero, such as 2009 and 2011 and 2013,
alternate with years without any non-detects (2010 and 2012) and concentrations that
19-30

-------
are higher in magnitude, as indicated by the 95th percentile and maximum
concentration.
Figure 19-24. Yearly Statistical Metrics for Cadmium (PMio) Concentrations Measured at
S4MO
20031	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2003.
Observations from Figure 19-24 for cadmium measurements collected at S4MO include
the following:
•	The maximum cadmium concentration was measured in 2009 (9.71 ng/m3). The five
cadmium concentrations greater than 5 ng/m3 were measured at S4MO in 2004 (one),
2008 (two), and 2009 (two).
•	A steady decreasing trend is shown in the 1-year average and median concentrations
through 2006. Even though the 1-year average concentration exhibits an increasing
trend between 2006 and 2009, the median concentration does not, and actually
continued decreasing during most of this time. This indicates that concentrations at
the upper end of the concentration range are driving the 1-year averages, particularly
for 2008 and 2009, while the concentrations at the lower end of the concentration
range are accounting for a higher percentage of measurements.
•	The range of concentrations measured decreased significantly from 2009 to 2010.
19-31

-------
• Even though the range of concentrations increased every year between 2010 and
2013, the 1-year average concentration changed little while the median exhibits a
slight decreasing trend, reaching a minimum for 2013.
Figure 19-25. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
S4MO
2003	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimurr
Maximum	O 95th Percentile
• Average
Observations from Figure 19-25 for carbon tetrachloride measurements collected at
S4MO include the following:
•	Twenty of the 21 non-detects of carbon tetrachloride were measured at S4MO in
2003, 2004, or 2005, with a single non-detect measured in 2007.
•	A steady increasing trend in the 1-year average concentration is shown through 2006.
Although the maximum concentration decreased substantially from 2006 to 2007 (and
a non-detect was measured), the change in the 1-year average concentration is slight
and the median concentration did not change at all. In fact, the median concentration
is steady between 2005 and 2007.
•	All of the statistical parameters exhibit increases from 2007 to 2008.
• Both the median and 1-year average concentrations have a decreasing trend between
2008 and 2010, with the largest change shown for 2010.
19-32

-------
•	Between 2010 and 2012, the 1-year average concentrations have a significant
increasing trend even as the majority of concentrations measured are falling into a
tighter range, as indicated by the decreasing difference between the 5th and 95th
percentiles for these years.
•	Nearly all of the statistical parameters exhibit decreases for 2013, which is mostly a
result of a larger number of concentrations at the lower end of the concentration
range. Three concentrations measured in 2013 are less than the minimum
concentration measured in 2012. Further, the number of concentrations less than
0.6 [j,g/m3 increased from 2012 to 2013 (from nine in 2012 to 20 in 2013).
Figure 19-26. Yearly Statistical Metrics for />-Dichlorobenzene Concentrations Measured at
S4MO
Maximum
Concentration for
2008 is 6.18 ng/m-
2003	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimurr
Maximum	O 95th Percentile
• Average
Observations from Figure 19-26 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 percentage of non-detects was at a maximum in 2003 (90 percent), after which
the percentage decreased, reaching a minimum of 5 percent for 2009. The percentage
of non-detects varied between 10 percent and 20 percent each year following 2009.
•	After little change in the early years, the 1-year average and median concentrations
exhibit a steady increasing trend between 2005 and 2008. However, the relatively
19-33

-------
large number of non-detects (zeros) combined with the range of measured detections
result in a relatively high level of variability, based on the confidence intervals
calculated for the 1-year averages. This is particularly true for 2008, when the
maximum /;-dichlorobenzene concentration was measured (6.18 (J,g/m3). The
difference between the median and 1-year average concentration is also an indicator
of this variability. During this period, the 1-year average was at least three times
greater than the median.
•	The concentrations measured decreased considerably from 2008 to 2009 then
increased again in 2010. Concentrations measured in 2010 were higher and more
variable than those measured in 2009. Five concentrations measured in 2010 were
greater than the maximum concentration measured in 2009 and the number of
concentrations greater than 0.5 [j,g/m3 more than doubled, from four in 2009 to 10 for
2010. At the same time, the number of non-detects increased from three in 2009 to
eight in 2010.
•	Although the range of concentrations measured in 2011 is similar to the range of
concentrations measured in 2010, the 95th percentile and 1-year average
concentration decreased. Further decreases are shown for these parameters for 2012.
Yet, the median concentration increased slightly for 2011 and then did not change for
2012.
•	Several of the statistical parameters are at a minimum for 2013, including the 1-year
average concentration, which is less than 0.1 [j.g/m3 for the first time. This year has
the smallest range of concentrations measured by a considerable margin.
19-34

-------
Figure 19-27. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
S4MO
o
Maximum
Concentration for
2008 is 0.41 ng/m3
—O*-	i		;		o
2003	2004	2005	2006	2007	2008	2009	2010
Year
2011	2012
0 5th Percentile
— Minimurr
0 95th Percentile
Observations from Figure 19-27 for 1,2-dichloroethane measurements collected at S4MO
include the following:
• With the exception of 2012 and 2013, the median concentration is zero for all years,
indicating that at least 50 percent of the measurements were non-detects. There were
no measured detections of 1,2-dichloroethane in 2003, 2004, or 2007, one measured
detection in 2005, and two each in 2006 and 2008. Beginning in 2009, the number of
measured detections increased steadily, from five in 2009, to 10 in 2010, 18 in 2011,
56 in 2012, and 58 in 2013.
As the number of measured detections increased in the later years of sampling, each
of the corresponding statistical metrics shown in Figure 19-27 also increased. The 5th
percentile and median concentrations are greater than zero for the first time in 2012,
when measured detections accounted for a majority of the measurements for the first
time.
19-35

-------
Figure 19-28. Yearly Statistical Metrics for Fluorene Concentrations Measured at S4MO
20081	2009	2010	2011	2012	2013
0 5th Percentile	— Minimum	— Median	- Maximum	0 95th Percentile ••"¦"/"•Average
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2008.
Observations from Figure 19-28 for fluorene measurements collected at S4MO include
the following:
•	The box and whisker plots for fluorene measurements resemble the plots for
acenaphthene presented in Figure 19-23.
•	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).
•	Several of the statistical parameters shown exhibit decreases from 2008 to 2009.
From 2009 to 2010, the range of concentrations measured increased but the median
concentration decreased, a trend that continued into 2011. A similar observation was
made for acenaphthene.
•	With the exception of the maximum concentration, the statistical parameters exhibit
increases from 2011 to 2012. This is because the number of measurements at the
upper end of the range increased while the number of measurements at the lower end
of the concentration range decreased. The number of concentrations greater than
10 ng/m3 increased from 13 to 22 from 2011 to 2012; conversely, the number of
concentrations less than 2 ng/m3 decreased from 11 to three from 2011 to 2012.
•	All of the statistical parameters exhibit decreases for 2013.
19-36

-------
Figure 19-29. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at S4MO
Maximum
Concentration for
2004 is 43.8 p.g/m3
2003	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	0 95th Percentile	Average
Observations from Figure 19-29 for formaldehyde measurements collected at S4MO
include the following:
•	The maximum formaldehyde concentration (43.8 (J,g/m3) was measured in 2004 on
the same day that the maximum acetaldehyde concentration was measured
(August 31, 2004). This concentration is more than twice the next highest
concentration (17.8 (j,g/m3), which was measured in 2011. The six highest
concentrations of formaldehyde were all measured in 2004 (2) or 2011 (4).
•	The 1-year average concentration has a decreasing trend between 2004 and 2006.
After the increase shown for 2007, the decreasing trend resumed through 2009, when
the 1-year average was at a minimum (2.46 |ig/m3). The 1-year average concentration
did not change significantly between 2009 and 2010, even though the smallest range
of concentrations was measured in 2010.
•	Most of the statistical parameters exhibit considerable 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.
•	The central tendency statistics exhibit little change from 2012 to 2013.
19-37

-------
Figure 19-30. Yearly Statistical Metrics for Hexachloro-l,3-Butadiene Concentrations
Measured at S4MO




0.3 -




o



0.1 -
00 -



o

pL
o




	<
>	


2003	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
0 5th Percentile	— Minimum	— Median	- Maximum	0 95th Percentile	Average
Observations from Figure 19-30 for hexachloro-l,3-butadiene measurements collected at
S4MO include the following:
•	The median concentration of hexachloro-1,3-butadiene for each year of sampling is
zero, indicating that at least 50 percent of the measurements were non-detects. For
2003, 2004, and 2007 through 2010, 100 percent of the measurements were non-
detects.
•	For 2005 and 2006, the percentage of measured detections was less than 15 percent.
For 2011, measured detections accounted for 16 percent of the measurements. For
2012, that number increased to 22 percent and then up to 26 percent for 2013.
Additional years of sampling are needed to determine if the number of measured
detections will continue to increase.
•	The 1-year average concentration has varied little over the last 3 years of sampling,
from 0.018 |ig/m3 (2012) to 0.023 |ig/m3 (2013).
19-38

-------
Figure 19-31. Yearly Statistical Metrics for Lead (PMio) Concentrations Measured at S4MO
2003 1 2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
0 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2003.
Observations from Figure 19-31 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, as there
were five measurements in 2012 greater than the maximum concentration measured
in 2011, the median concentration decreased. This is due to an increase in the number
of measurements at the lower end of the concentration range. For example,
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 average concentrations have ranged from
9.40 ng/m3 (2013) 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. The 1-year average concentration is at a minimum concentration
for 2013.
19-39

-------
Figure 19-32. Yearly Statistical Metrics for Naphthalene Concentrations Measured at S4MO
*
0 +-
2008 1
2009
2010
2011
Year
2012
2013

O 5th Percentile
- Minimum
— Median
- Maximum
O 95th Percentile

1 A 1-year average is not presented because sampling under the NMP did not begin until April 2008.
Observations from Figure 19-32 for naphthalene measurements collected at S4MO
include the following:
•	Naphthalene concentrations measured at S4MO exhibit considerable variability,
ranging from 10.1 ng/m3 (2013) to 784 ng/m3 (2010).
•	The 1-year average concentration has ranged from 71.92 ng/m3 (2013) to 135 ng/m3
(2010). The median varies less, ranging from 66.30 ng/m3 (2013) to 89.85 ng/m3
(2010). All of the statistical parameters are at a minimum for 2013.
•	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.
19-40

-------
Figure 19-33. Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at S4MO
20031	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
0 5th Percentile	— Minimum	— Median	- Maximum	0 95th Percentile ••"¦"/"•Average
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2003.
Observations from Figure 19-33 for nickel measurements collected at S4MO include the
following:
•	The two highest nickel concentrations 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, indicating that the change is
not statistically significant.
•	Several of the lowest nickel concentrations were measured at S4MO in 2013,
including the lowest minimum concentration. This year has the greatest number of
nickel concentrations less than 0.5 ng/m3 (nine).
19-41

-------
19.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at the S4MO monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4 for
definitions and explanations regarding the various toxicity factors, time frames, and calculations
associated with these risk-based screenings.
19.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for S4MO and where annual average concentrations could
be calculated, risk was examined by calculating cancer risk and noncancer hazard
approximations. These approximations can be used as risk estimates for cancer and noncancer
effects attributable to the pollutants of interest. Although the use of these approximations is
limited, they may help identify where policy-makers want to shift their air monitoring priorities.
Refer to Section 3.4.3.3 for an explanation of how cancer risk and noncancer hazard
approximations are calculated and what limitations are associated with them. Annual averages,
cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard approximations are
presented in Table 19-6, where applicable. Cancer risk approximations are presented as
probabilities while the noncancer hazard approximations are ratios and thus, unitless values.
Observations for S4MO from Table 19-6 include the following:
•	The pollutants with the highest annual average concentrations for S4MO are
formaldehyde, acetaldehyde, carbon tetrachloride, and benzene.
•	The same four pollutants have the highest cancer risk approximations for S4MO,
although the order is different. However, formaldehyde's cancer risk approximation
for S4MO (42.05 in-a-million) is an order of magnitude higher than the cancer risk
approximations for these other pollutants.
•	Benzene has the highest cancer risk approximation for S4MO among the VOCs
(4.74 in-a-million); arsenic has the highest cancer risk approximation for S4MO
among the metals (3.14 in-a-million); and naphthalene has the highest cancer risk
approximation for S4MO among the PAHs (2.45 in-a-million).
•	None of the pollutants of interest for S4MO have noncancer hazard approximations
greater than 1.0, indicating that no adverse noncancer health effects are expected from
these individual pollutants. The pollutant with the highest noncancer hazard
approximation is formaldehyde (0.33).
19-42

-------
Table 19-6. Risk Approximations for the Missouri Monitoring Site



# of






Measured

Cancer
Noncancer

Cancer
Noncancer
Detections
Annual
Risk
Hazard
Pollutant
URE
frig/m3)1
RfC
(mg/m3)
vs. # of
Samples
Average
frig/m3)
Approximation
(in-a-million)
Approximation
(HQ)
St. Louis, Missouri - S4MO




1.98


Acetaldehyde
0.0000022
0.009
61/61
±0.22
4.35
0.22




0.61


Benzene
0.0000078
0.03
61/61
±0.05
4.74
0.02




0.07


1,3-Butadiene
0.00003
0.002
57/61
±0.01
1.97
0.03




0.64


Carbon Tetrachloride
0.000006
0.1
61/61
±0.03
3.84
0.01




0.09


/?-Dichlorobcnzcnc
0.000011
0.8
47/61
±0.02
1.03
<0.01




0.09


1,2-Dichloroethane
0.000026
2.4
58/61
±0.01
2.27
<0.01




3.23


Formaldehyde
0.000013
0.0098
61/61
±0.55
42.05
0.33




0.02


Hexachloro-1,3 -butadiene
0.000022
0.09
17/61
±0.01
0.53
<0.01




5.02


Acenaphthene3
0.000088
--
60/60
± 1.16
0.44
--




0.73


Arsenic (PMi0)a
0.0043
0.000015
61/61
±0.08
3.14
0.05




0.56


Cadmium (PMi0)a
0.0018
0.00001
61/61
±0.18
1.01
0.06




5.79


Fluorene3
0.000088
--
60/60
± 1.07
0.51
--




9.40


Lead (PMi0)a
--
0.00015
61/61
±2.01
--
0.06




71.92


Naphthalene1
0.000034
0.003
60/60
± 10.08
2.45
0.02




1.06


Nickel (PMi,;,)a
0.00048
0.00009
61/61
±0.26
0.51
0.01
- = A Cancer URE or Noncancer RfC is not available.
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of
viewing.
19-43

-------
19.5.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 19-7 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
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 3.4.3.4. Lastly,
Table 19-7 provides the pollutants with the highest cancer risk approximations (in-a-million) for
S4MO, 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.
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 19.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
19-44

-------
Table 19-7. 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)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
St. Louis, Missouri (St. Louis City) - S4MO
Formaldehyde
86.19
Formaldehyde
1.12E-03
Formaldehyde
42.05
Benzene
85.02
Hexavalent Chromium
7.89E-04
Benzene
4.74
Ethylbenzene
48.46
Benzene
6.63E-04
Acetaldehyde
4.35
Acetaldehyde
46.53
1,3-Butadiene
3.78E-04
Carbon Tetrachloride
3.84
T richloroethylene
15.45
Naphthalene
3.26E-04
Arsenic
3.14
1.3 -Butadiene
12.60
Arsenic, PM
2.49E-04
Naphthalene
2.45
Naphthalene
9.59
POM, Group 2b
1.80E-04
1,2-Dichloroethane
2.27
T etrachloroethylene
5.83
POM, Group 2d
1.44E-04
1,3-Butadiene
1.97
Dichloromethane
3.65
Ethylbenzene
1.21E-04
/?-Dichlorobcnzcnc
1.03
POM, Group 2b
2.05
Acetaldehyde
1.02E-04
Cadmium
1.01

-------
Table 19-8. 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)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
St. Louis, Missouri (St. Louis City) - S4MO
Toluene
313.06
Acrolein
268,853.09
Formaldehyde
0.33
Hexane
226.91
Formaldehyde
8,795.13
Acetaldehyde
0.22
Methanol
208.08
T richloroethylene
7,726.86
Lead
0.06
Xylenes
196.63
1.3 -Butadiene
6,302.24
Cadmium
0.06
Formaldehyde
86.19
Acetaldehyde
5,170.07
Arsenic
0.05
Benzene
85.02
Arsenic, PM
3,864.62
1,3-Butadiene
0.03
Hydrochloric acid
70.78
Hydrochloric acid
3,539.11
Naphthalene
0.02
Ethylene glycol
64.32
Cadmium, PM
3,474.08
Benzene
0.02
Ethylbenzene
48.46
Lead, PM
3,349.14
Nickel
0.01
Acetaldehyde
46.53
Naphthalene
3,195.33
Carbon Tetrachloride
0.01

-------
Observations from Table 19-7 include the following:
•	Formaldehyde, benzene, and ethylbenzene are the highest emitted pollutants with
cancer UREs in the city of St. Louis.
•	The pollutants with the highest toxicity-weighted emissions (of the pollutants with
cancer UREs) are formaldehyde, hexavalent chromium, and benzene.
•	Seven of the highest emitted pollutants also have the highest toxicity-weighted
emissions.
•	Formaldehyde tops all three lists, with the highest quantity emitted, the highest
toxicity-weighted emissions, and the highest cancer risk approximation. Benzene,
acetaldehyde, naphthalene, and 1,3-butadiene also appear on all three lists.
•	Arsenic has the fifth highest cancer risk approximation for S4MO. While arsenic is
not one of the highest emitted pollutants in the city of St. Louis, it ranks sixth for its
toxicity-weighted emissions. Carbon tetrachloride, 1,2-dichloroethane,
/;-dichlorobenzene, and cadmium also appear among the pollutants of interest with
the highest cancer risk approximations for S4MO but none of these appear on either
emissions-based list.
•	POM, Group 2b is the 10th highest emitted "pollutant" in St. Louis and ranks seventh
for toxicity-weighted emissions. POM, Group 2b includes several PAHs sampled for
at S4MO including acenaphthene and fluorene, which are pollutants of interest for
S4MO. These pollutants are not among those with the highest cancer risk
approximations for S4MO.
Observations from Table 19-8 include the following:
•	Toluene, hexane, and methanol are the highest emitted pollutants with noncancer
RfCs in the city of St. Louis.
•	The pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) are acrolein, formaldehyde, and trichloroethylene. Although
acrolein was sampled for at S4MO, this pollutant was excluded from the pollutants of
interest designation, and thus subsequent risk-based screening evaluations, due to
questions about the consistency and reliability of the measurements, as discussed in
Section 3.2.
•	Three of the highest emitted pollutants in the city of St. Louis also have the highest
toxicity-weighted emissions.
•	Formaldehyde, the pollutant with highest noncancer hazard approximation, has the
second highest toxicity-weighted emissions and the fifth highest total emissions.
Acetaldehyde also appears on all three lists.
19-47

-------
• Three of S4MO's metal pollutants of interest (lead, arsenic, and cadmium) appear
among the pollutants with the highest toxicity-weighted emissions, but none of these
are among the highest emitted.
19.6 Summary of the 2013 Monitoring Data for S4MO
Results from several of the data treatments described in this section include the
following:
~~~ Twenty-one pollutants failed screens for S4MO. S4MO failed the highest number of
screens among all NMP sites, similar to previous years.
~~~ Formaldehyde and acetaldehyde have the highest annual average concentrations for
S4MO. These are the only pollutants of interest with annual averages greater than
1 ng/m3.
~~~ S4MO has the second highest annual average concentration of hexachloro-1,3-
butadiene and the fourth highest annual average concentration of p-dichlorobenzene
among NMP sites sampling VOCs. S4MO also has the fifth highest annual average
concentration of arsenic among sites sampling PMio metals.
~~~ Concentrations of acetaldehyde measured at S4MO have decreased significantly
since 2010. Concentrations of benzene have an overall decreasing trend as well, with
some of the lowest concentrations measured in 2013. Some of the lowest
concentrations of pollutants such as arsenic, p-dichlorobenzene, and naphthalene
were measured in 2013. In addition, the detection rate of 1,2-dichloroethane has been
increasing steadily at S4MO over the last few years of sampling.
~~~ Formaldehyde has the highest cancer risk approximation of the pollutants of interest
for S4MO. None of the pollutants of interest have noncancer hazard approximations
greater than an HQ of 1.0.
19-48

-------
20.0	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.
20.1	Site Characterization
This section characterizes the New Jersey monitoring sites by providing geographical and
physical information about the location of the sites and the surrounding areas. This information
is provided to give the reader insight regarding factors that may influence the air quality near the
sites and assist in the interpretation of the ambient monitoring data.
One New Jersey monitoring site (CSNJ) is located in the Philadelphia-Camden-
Wilmington, PA-NJ-DE-MD CBSA while the other three New Jersey sites are located within the
New York-Newark-Jersey City, NY-NJ-PA CBSA. Figure 20-1 is a composite satellite image
retrieved from ArcGIS Explorer showing the CSNJ monitoring site and its immediate
surroundings. Figure 20-2 identifies nearby point source emissions locations by source category,
as reported in the 2011 NEI for point sources, version 2. Note that only sources within 10 miles
of the site are included in the facility counts provided in Figure 20-2. A 10-mile boundary was
chosen to give the reader an indication of which emissions sources and emissions source
categories could potentially have a direct effect on the air quality at the monitoring site. Further,
this boundary provides both the proximity of emissions sources to the monitoring site as well as
the quantity of such sources within a given distance of the site. Sources outside the 10-mile
boundary are still visible on the map for reference, but have been grayed out in order to
emphasize emissions sources within the boundary. Figures 20-3 through 20-7 are the composite
satellite maps and emissions source maps for CHNJ, ELNJ, and NBNJ. Table 20-1 provides
supplemental geographical information such as land use, location setting, and locational
coordinates.
20-1

-------
Figure 20-1. Camden, New Jersey (CSNJ) Monitoring Site
-IJSourc*' U5G5
Sour c«	NC.A USGS
£ 206% MurmcM Cftrp.

-------
Figure 20-2. NEI Point Sources Located Within 10 Miles of CSNJ
n nrarw
n	_ B s ^ J * »
* u f * i • %
^ / p \ N Burlington
0	l O	X	\ l County
PENNSYLVANIA
CJrndftn
County 1
Oouwfi«ef
Lwawanp_	^
County
s *nrw
Legend
wrismw	ft iottw	nmrfi
ttof 0u« to fac«ty	and ooHocatfon ft* total fidWta
chap-ayed may not represent all tocMfces *tm«n arte area of ntoreat
CSNJ UATMP site	10 mile radius
Source Category Group (No. of Facilities)
County boundary
IT	A* (MbAvrvAjipol Supptrt Oo*f 4(ksu i>r •
£ PTOdUdlKAHOt Ml.	(1"
TC	Au»fnoD4»True* Wanutacwrig r*ity < 11
ff	bu*Of*»Comiiw<»o«DH!ri12j
i Gamia«*tur Saabc*» (5 »
tl?ca*«| G«fiefa»on >ci ComcuMKm <91
Food Proo»**iv»yc«lfejr» Facliy 
M«t* £a» Fto« wdClNrMtiiil u»«»ar«ii Mtrw>lncUir«^p|1)
Men* OaalMB E*a»avma 4f>a *Ji»d Sarvcat id tomubeiuMet i2 t
(7* Mhm MtK(»f.!4itcnraiwn f «a*t| rtJi
A	SrfM-NtrioruJ SafeiM* F^uty Pi
•, Mm»'Quin) lMmilt>n»mt*QrKaif (2^
I
F
~
o
A
7	GWMWfllfllMMNMViMlt B1
(m]	lAtntcra* W«Ma CAfrctjMr r? i
•	0*	Cat. Piad^turi S>
^	Rift ana CootnaUantiKacainng <*.•*> 11
ft	Patio Mum Product* Mnavfocfiimg i21
.4	Pairateufn RaftMnv
RinpnKt.Mi I Marx/aeturmg i21
R	-* Frudustt PttMl |41
H	pou ar.auartjor opr*8l«n» |2j
P	^ n1fnj»P>i»»»ilir'yt-»p«« > r-«nni ^mutechJfrq huttr il4>
S	Pulp and Faotf PUnt »3|
ii	Sftofeoat u*iut»cu>«*>a a Recur *»di> < t:
V	SUWWfil)
T1	T0tooemm«m?-abnns*i3«3o F«cHy if)
•rfV	in* k	Uncial- •• r»|
>	MMl*
-------
Figure 20-3. Chester, New Jersey (CHNJ) Monitoring Site

~
XM
to
o
~
ci
US'-
V
m
v
\
slt>
5oufC«rU5G5
Source NASA NGA. U5GS
£ iOUS'HicrotoM Cttrp.

-------
Figure 20-4. NEI Point Sources Located Within 10 Miles of CHNJ
5us&e>
County
Wmv* r<
Gour*y
Moms
County
Sorvwaet
Count)
Hunterdon
County

WW***
Note- Duo to tacuty density ana oollocabon tt\o total facilities
Legend	:l»&p«aya4 may not i«p»a*tnt all facte** wtthin tn« o<»a ot ntaiait
CHNJ UATMP site	10 mile radius	County boundary
Source Category Group (No. of Facilities)
4"	Aerospace/Aircraft Manufacturing Facility (1)
T	Airport/Airlina/Airport Support Operations (12)
4	Aspnalt Production/Hot Mix Asphalt Plant (1)
C	Chemical Manufacturing FacHrty (3)
e	Electrical Equipment Manufacturing Facility (1)
F	Food Processing/Agriculture Facility (1)
?	Miscellaneous Commerciai'lndustriat Facility (2)
W	vVocxtvvor*. Furniture Mil(*orV & 'Afood Preserving FaciUty (1)
20-5

-------
Figure 20-5. Elizabeth, New Jersey (ELNJ) Monitoring Site

-------
Figure 20-6. North Brunswick, New Jersey (NBNJ) Monitoring Site

-------
Figure 20-7. NEI Point Sources Located Within 10 Miles of ELM J and NBNJ
Hussor
Courtr
Essex
County
Morns
County
\ Richmond
" y Counly
Urnon
County
*Addm%e«
County
ELNJ UATMP site
NBNJ UATMP s>te
10 mile radius
County boundary
Source Category Group (No. of Facilities)
^ MpMiitivtocot 1uwiiapnt9n 4	0 v%
«
3	«k*	ClB.tt.^1
I	Owwum t3i
Q	Bktm £«(nwi VMMkhim ^ih (tr n '
f	Slaancrt	. w C'a«i»n («tm -I i
A
•	««<»¦« WHMO> H E«(l««ll (S
O	««(MKnH iMtacf Km Hit ?»cr «
A
¦	**U' L »t Mai m*t J*«> Ucu CfMmr	1 -1
n-aktr Fit*) '
f««* iii
(flf <*»***• MID COMMA) tit
J	15
R	Rnk Dm 3Hin*«w '¦uwwmMrin
*	mm TwMnwi
7**3TU"v* t4 wctv nrsnrw wnwwi WMinm fncnnAi rj«rtrw ?•* yc
I' NEW i
il YORK A
Swncwl
*tj :.3'

Morrooutn
1 County
7r4Sirv» 74-40 trw 74-jsirvv Tirs** uiswi ?»w
-------
Table 20-1. Geographical Information for the New Jersey Monitoring Sites
Site
Code
AQS Code
Location
County
Micro- or Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Additional Ambient Monitoring
Information1
CSNJ
34-007-0002
Camden
Camden
Philadelphia-Camden-
Wilmington, PA-NJ-DE-
MD
39.934446,
-75.125291
Industrial
Urban/City
Center
CO, IMPROVE Speciation Meteorological
parameters, NO, NO2, NOx, O3, PM2.5, SO2,
PM2.5 Speciation.
CHNJ
34-027-3001
Chester
Morris
New York-Newark-
Jersey City, NY-NJ-PA
40.787628,
-74.676301
Agricultural
Rural
SO2, NO, NO2, O3, Meteorological
parameters, PM2.5, PM2.5 Speciation
IMPROVE Speciation.
ELNJ
34-039-0004
Elizabeth
Union
New York-Newark-
Jersey City, NY-NJ-PA
40.64144,
-74.208365
Industrial
Suburban
CO, SO2, NO2, Meteorological parameters,
PM2.5. PM2.5 Speciation IMPROVE
Speciation.
NBNJ
34-023-0006
North
Brunswick
Middlesex
New York-Newark-
Jersey City, NY-NJ-PA
40.472825,
-74.422403
Agricultural
Rural
Meteorological parameters, PM2.5, PM2.5
Speciation IMPROVE Speciation.
1 Data for additional pollutants are reported to AQS for these sites (EPA, 2014b); however, these data are not generated by ERG and are therefore not included in this
report.

-------
The CSNJ monitoring site is located just outside Philadelphia, across the state line, in
the city of Camden in southwest New Jersey. The monitoring site is in an industrial area a few
blocks east of the Delaware River, as shown in Figure 20-1. Residential areas are located to the
east between the site and 1-676. Figure 20-2 shows that the large number of point sources located
within 10 miles of CSNJ are involved in a variety of industries. The source categories with the
largest number of facilities include institutions (such as schools, hospitals, and prisons); airports
and airport support operations, which include airports and related operations as well as small
runways and heliports, such as those associated with hospitals or television stations; printing,
publishing, and paper product manufacturing; chemical manufacturing; and bulk terminals and
bulk plants. The sources closest to CSNJ include a metals processing and fabrication facility; a
mine/quarry/minerals processing facility; an airport/airport operations facility; and a metal can,
box, and other container manufacturing facility.
CHNJ is located in northern New Jersey, in the town of Chester, west of the New York
City metropolitan area. Figure 20-3 shows that CHNJ is located in an open area near Building 1
of the Department of Public Works off Routes 513 and 510. The surrounding area is rural and
agricultural with a rolling topography, but surrounded by small neighborhoods. Two schools are
located on the other site of Route 510 to the south-southwest of CHNJ. Although the location is
considered part of the New York City metro area, the site's location is outside most of the
urbanized areas. Figure 20-4 shows that few sources are located within a few miles of CHNJ.
The source category with the greatest number of emissions sources within 10 miles of CHNJ is
the airport source category. The sources closest to CHNJ include a privately owned heliport to
the south and a wood work, furniture, millwork, and wood preserving facility to the west.
ELNJ is located in the city of Elizabeth, which lies just south of Newark and west of
Newark Bay and Staten Island, New York. As Figure 20-5 shows, the monitoring site is located
near the toll plaza just off Exit 13 of the New Jersey Turnpike (1-95). Interstate-278 intersects the
Turnpike here as well. The surrounding area is highly industrialized, with an oil refinery located
just southwest of the site. Additional industry is located to the southwest and west, as well as on
the east side of the Turnpike, while residential neighborhoods are located to the north and
northwest of ELNJ.
20-10

-------
NBNJ is located in North Brunswick, approximately 16 miles southwest of Elizabeth.
The monitoring site is located on the property of Rutgers University's Cook-Douglass campus,
on a horticultural farm. The surrounding area is agricultural and rural, although residential
neighborhoods are located to the east, across a branch of the Raritan River, as shown in
Figure 20-6. County Road 617 (Ryders Lane) and US-1 intersect just west of the site and 1-95
runs northeast-southwest less than 1 mile east of the site, part of which can be seen on the right
hand side of Figure 20-6.
Figure 20-7 shows that the outer portions of the 10-mile boundaries for ELNJ and NBNJ
intersect; these sites are located approximately 17 miles apart. Many emissions sources surround
these two sites. The majority of the emissions sources are located in northern Middlesex County
and northeastward toward New York City and northern New Jersey. The source categories with
the greatest number of emissions sources in the vicinity of these sites include airport operations,
chemical manufacturing, bulk terminals and bulk plants, and electricity generation via
combustion. The emissions sources in closest proximity to the ELNJ monitoring site are in the
wastewater treatment, chemical manufacturing, bulk terminals/bulk plant, petroleum refining,
and electricity generation via combustion source categories. The emissions sources in closest
proximity to the NBNJ monitoring site are involved in plastic, resin, or rubber products
manufacturing, airport and airport support operations, and pharmaceutical manufacturing.
Table 20-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the New Jersey monitoring sites. Table 20-2 includes a county-level
population for each site. County-level vehicle registration data for Camden, 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 (FHWA), was
allocated to the county level using the county-level proportion of the state population from the
U.S. Census Bureau. Table 20-2 also contains traffic volume information for each site as well as
the location for which the traffic volume was obtained. Additionally, Table 20-2 presents the
county-level daily VMT for the four New Jersey counties.
20-11

-------
Table 20-2. Population, Motor Vehicle, and Traffic Information for the New Jersey
Monitoring Sites
Site
County
Estimated
County
Population1
County-level
Vehicle
Registration2
Annual
Average
Daily Traffic3
Intersection
Used for Traffic Data
County-level
Daily VMT4
CSNJ
Camden
512,854
458,294
3,231
S 2nd St. south of Walnut St.
10,753,157
CHNJ
Morris
499,397
443,969
11,215
Mendham Rd (510/24) east of Fox
Chase Rd
14,622,523
ELNJ
Union
548,256
485,427
250,000
Between Exits 13 & 13AonI-95
12,081,401
NBNJ
Middlesex
828,919
734,425
110,653
US-1 west of 617
21,634,307
bounty-level population estimates reflect 2013 data (Census Bureau, 2014)
2County-level vehicle registrations reflect ratios based on 2012 state-level vehicle registration data from the FHWA and the
2012 county-level proportion of the state population data (FHWA, 2014 and Census Bureau, 2013c)
3AADT for ELNJ reflects 2006 data from NJ Department of Treasury; AADT reflect 2009 data for NBNJ and 2012 data for
CSNJ and CHNJ from the NJ DOT (NJ DOTr, 2008 and NJ DOT, 2014)
4County-level VMT reflects 2012 data (NJ DOT, 2012)
Observations from Table 20-2 include the following:
•	Middlesex County, where NBNJ is located, has the highest county-level population
for the New Jersey sites while Morris County, where CHNJ is located, has the least
(although the populations for the counties other than Middlesex are not that different).
Compared to NMP monitoring sites in other locations, the county-level populations
are in the middle of the range, ranking 17th, then 25th through 27th.
•	The estimated county-level vehicle registration is also highest for NBNJ and least for
CHNJ (although the vehicle registrations for the counties other than Middlesex are
not that different). The county-level registration estimates for these sites have similar
rankings as the county-level populations among NMP sites.
•	ELNJ and NBNJ experience a significantly higher traffic volume than CHNJ and
CSNJ. Traffic data for ELNJ are provided for 1-95, between Exit 13 and 13A; this is
the second highest traffic volume among all NMP sites (behind only LBHC A).
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); and
traffic data for CSNJ are provided for South 2nd Street, south of Walnut Street.
•	The daily VMT is highest for Middlesex County (NBNJ) and lowest for Camden
County (CSNJ). The VMT for Middlesex County ranks 15th compared to other
counties with NMP sites while the other New Jersey counties are in the middle of the
range.
20.2 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.
20-12

-------
20.2.1	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, generally between 3 inches and 4 inches per month, and
relatively high humidity. Temperatures tend to be higher and precipitation lower in the southwest
part of the state than the northern and coastal portions 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, 2015).
20.2.2	Meteorological Summary
Hourly meteorological data for 2013 were retrieved from NCDC for the weather stations
closest to the New Jersey monitoring sites (NCDC, 2013), as described in Section 3.4.2. The
closest weather stations are located at Philadelphia International Airport (near CSNJ),
Somerville-Somerset Airport (near CHNJ and NBNJ), and Newark International Airport (near
ELNJ), WBANs 13738, 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 20-3. These data were used to determine how meteorological conditions on sample days
vary from conditions experienced throughout the year.
20-13

-------
Table 20-3. Average Meteorological Conditions near the New Jersey Monitoring Sites
to
p
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)
Camden, New Jersey - CSNJ
Philadelphia Intl
Airport
13738
(39.87, -75.23)
7.2
miles
231°
(SW)
Sample
Days
(66)
64.5
±4.7
56.4
±4.4
42.1
±4.8
49.6
±4.1
61.5
±3.1
1018.6
± 1.7
7.5
±0.7
2013
64.0
± 1.9
56.3
± 1.8
42.3
±2.0
49.7
± 1.7
62.2
± 1.4
1018.2
±0.7
7.5
±0.3
Chester, New Jersey - CHNJ
Somerville, New
Jersey/Somerset
Airport
54785
(40.62, -74.67)
11.3
miles
178°
(S)
Sample
Days
(61)
63.1
±5.0
52.6
±4.5
41.9
±5.1
47.7
±4.4
70.3
±3.4
1017.7
± 1.7
3.4
±0.7
2013
62.6
± 1.9
52.4
± 1.8
41.9
±2.0
47.6
± 1.7
70.8
± 1.4
1017.2
±0.7
3.2
±0.2
Elizabeth, New Jersey - ELNJ
Newark International
Airport
14734
(40.68, -74.17)
3.5
miles
36°
(NE)
Sample
Days
(61)
63.5
±5.0
55.5
±4.7
41.8
±5.0
49.0
±4.3
62.9
±3.6
1017.6
± 1.6
7.8
±0.8
2013
62.9
± 1.9
55.4
± 1.8
41.7
±2.0
49.0
± 1.7
63.1
± 1.5
1017.5
±0.7
8.0
±0.3
North Brunswick, New Jersey - NBNJ
Somerville, New
Jersey/Somerset
Airport
54785
(40.62, -74.67)
16.7
miles
309°
(NW)
Sample
Days
(66)
63.0
±4.6
52.3
±4.2
42.0
±4.6
47.5
±4.0
71.1
±3.0
1017.3
± 1.8
3.3
±0.6
2013
62.6
± 1.9
52.4
± 1.8
41.9
±2.0
47.6
± 1.7
70.8
± 1.4
1017.2
±0.7
3.2
±0.2
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.

-------
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 2013. 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 near CSNJ, CHNJ, ELNJ, and NBNJ. The largest difference between a
sample day and a full-year average is for relative humidity at CSNJ, although the difference is
less than 1 percent and not statistically significant.
20.2.3 Wind Rose Comparison
Hourly surface wind data from the weather stations nearest the New Jersey sites, as
presented in Section 20.2.2, were uploaded into a wind rose software program to produce
customized wind roses, as described in Section 3.4.2. A wind rose shows the frequency of wind
directions using "petals" positioned around a 16-point compass, and uses different colors to
represent wind speeds.
Figure 20-8 presents a map showing the distance between the weather station and CSNJ,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 20-8 also presents three different wind roses for the
CSNJ monitoring site. First, a historical wind rose representing 2003 to 2012 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 2013 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2013 is
presented. These can be used to identify the predominant wind speed and direction for 2013 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically. Figures 20-9 through 20-11 present the distance maps and
wind roses for CHNJ, ELNJ and NBNJ, respectively.
20-15

-------
Figure 20-8. Wind Roses for the Philadelphia International Airport Weather Station near
CSNJ
Location of CSNJ and Weather Station	2003-2012 Historical Wind Rose
WIND SPEED
(Knots)
I I =22
n 17-21
H 11 - 17
I I 7- 11
\^3 4-7
H 2-4
Calms: 7.72%
west:

2013 Wind Rose
•VEST
WIND SPEED
i, Knots)
SOUTH
Sample Day Wind Rose
WEST
WW C SPEEC
(Knots)
SOUTH
20-16

-------
Figure 20-9. Wind Roses for the Somerville-Somerset Airport Weather Station near CHNJ
Location of CHNJ and Weather Station
\

\
W
y
k
6
n»Hhtr

4-
2003-2012 Historical Wind Rose
15%
12%
9%,
WIND SPEED
(Knots)
I I >=22
n 17-21
| 11 -17
[ n 7-11
~l 4-7
¦ 2- 4
Calms: 44.32%
2013 Wind Rose
Sample Day Wind Rose
west:
'A'INC SPEED
(Knots)
11 -17
SOUTH
Calms: 45.03%
NORTH
west:
(Knots)
SOUTH
WIND SPEED
20-17

-------
NORTH
est:
Figure 20-10. Wind Roses for the Newark International Airport Weather Station near
ELNJ
Location of ELNJ and Weather Station	2003-2012 Historical Wind Rose
WIND SPEED
(Knots)
I I «22
[~B 17-21
IH 11 17
m 7-11
l~D 4-7
¦ 2- 4
Calms: 6.38%
2013 Wind Rose
[NORTH"'-.
WEST
WViu SPEED
(Knots)
17-21
11 - 17
SOUTH
Calms. 7.14%
Sample Day Wind Rose
NORTHS-
WIND SPEED
(Knots)
17-21
SOUTH
Calms: 7 ei%
20-18

-------
west:
Sample Day Wind Rose
WIND SPEED
(Knots)
~
H 17-21
LI 11 - 17
[3 7- 11
I ~1 4-7
2- 4
Calms: 45.39%
Figure 20-11. Wind Roses for the Somerville-Sonierset Airport Weather Station near
NBNJ
Location of NBNJ and Weather Station	2003-2012 Historical Wind Rose
2013 Wind Rose
WIND SPEED
(Knots)
I I	=22
O	17-21
H	11 -17
I: .1	7-11
\^3	4-7
H	2-4
Calms: 44.32%
WIND SPEED
(Knots)
~ «22
F~B 17-21
11 -17
I I 7- 11
\~3 4-7
H 2-4
Calms: 45.03%
west:
NORTH
west:
20-19

-------
Observations from Figure 20-8 for CSNJ include the following:
•	The Philadelphia International Airport weather station is located 7.2 miles southwest
of CSNJ. Both the site and the weather station are located near the Delaware River,
which separates Pennsylvania from New Jersey, and runs along the east and south
sides of Philadelphia.
•	The historical wind rose shows that winds from a variety of directions were observed
near CSNJ, with westerly and southwesterly winds observed the most and north-
northeasterly winds and winds from the southeast quadrant observed the least. Calm
winds (those less than or equal to 2 knots) account for nearly 8 percent of
observations. The strongest winds were most often observed with westerly to
northwesterly winds.
•	The wind patterns shown on the 2013 wind rose are similar to the historical wind
patterns, indicating that wind conditions observed throughout 2013 were similar to
those observed historically, although a slightly higher percentage of calm winds were
observed.
•	The sample day wind rose shows that fewer winds with an easterly component were
observed on sample days; thus, an even higher percentage of winds with a westerly
component were observed. Observations of winds from the northwest and south
increased the most, from about 7.5 percent to 10 percent on sample days. The
percentage of calm winds is also slightly higher on the sample day wind rose.
Observations from Figures 20-9 and 20-11 for CHNJ and NBNJ include the following:
•	The weather station at Somerville/Somerset Airport is the closest weather station to
both CHNJ and NBNJ. The Somerville/Somerset Airport weather station is located
11.3 miles south of CHNJ and 16.7 miles northwest of NBNJ.
•	The historical and full-year wind roses for CHNJ 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 account for nearly
45 percent of observations. For wind speeds greater than 2 knots, northerly winds
were observed most frequently, accounting for nearly 9 percent of the observations,
while winds from the southwest quadrant were rarely observed.
•	Calm winds account for 45 percent of the wind observations in 2013. Winds from the
west-northwest to north account for another one-quarter of wind observations. With
the exception of southerly winds, which account for roughly 5 percent of the
observations, winds from the other directions were observed infrequently.
•	The sample day wind roses for CHNJ and NBNJ are similar to the full-year wind
roses. However, the number of northerly wind observations was less on sample days,
while southerly and northwesterly winds were observed more often.
20-20

-------
•	While the 2013 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 is a higher percentage of wind observations
from the northwest quadrant. Similar observations have been made in NMP reports
going back to 2009.
Observations from Figure 20-10 for ELNJ include the following:
•	The Newark International Airport weather station is located 3.5 miles northeast of
ELNJ. Both the site and the weather station are located in close proximity to the New
Jersey Turnpike.
•	The 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 2013 wind rose are similar to the historical wind
patterns, indicating that wind conditions observed throughout 2013 were similar to
those observed historically. However, a higher percentage of southwesterly winds
was observed in 2013 (about 10 percent compared to less than 7 percent historically),
while a lower percentage of northeasterly winds was observed (less than 3 percent in
2013 compared to nearly 7 percent historically).
•	The sample day wind rose shows that winds from the southeast to south as well as
those from the western quadrants accounted for a higher percentage of wind
observations on sample days while fewer wind observations from the north to east
were observed.
20.3 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 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. VOCs and carbonyl compounds were sampled for at all four New
Jersey sites.
20-21

-------
Table 20-4. Risk-Based Screening Results for the New Jersey Monitoring Sites
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Camden, New Jersey - CSNJ
Acetaldehyde
0.45
59
59
100.00
14.68
14.68
Formaldehyde
0.077
59
59
100.00
14.68
29.35
Benzene
0.13
57
57
100.00
14.18
43.53
Carbon Tetrachloride
0.17
57
57
100.00
14.18
57.71
1.3 -Butadiene
0.03
56
56
100.00
13.93
71.64
1,2-Dichloroethane
0.038
53
53
100.00
13.18
84.83
Hexachloro-1,3 -butadiene
0.045
15
15
100.00
3.73
88.56
Ethylbenzene
0.4
14
57
24.56
3.48
92.04
Propionaldehyde
0.8
11
59
18.64
2.74
94.78
Bromomethane
0.5
7
53
13.21
1.74
96.52
p-Dichlorobenzene
0.091
4
31
12.90
1.00
97.51
T richloroethylene
0.2
4
23
17.39
1.00
98.51
Methyl tert-Butyl Ether
3.8
3
55
5.45
0.75
99.25
1,2-Dibromoethane
0.0017
1
1
100.00
0.25
99.50
1,1,2 -T richloroethane
0.0625
1
1
100.00
0.25
99.75
Vinyl cliloride
0.11
1
12
8.33
0.25
100.00
Total
402
648
62.04

Chester, New Jersey - CHNJ
Acetaldehyde
0.45
61
61
100.00
16.80
16.80
Formaldehyde
0.077
61
61
100.00
16.80
33.61
Benzene
0.13
60
61
98.36
16.53
50.14
Carbon Tetrachloride
0.17
60
61
98.36
16.53
66.67
1,2-Dichloroethane
0.038
54
54
100.00
14.88
81.54
1.3 -Butadiene
0.03
40
46
86.96
11.02
92.56
Hexachloro-1,3 -butadiene
0.045
14
14
100.00
3.86
96.42
Methyl tert-Butyl Ether
3.8
12
59
20.34
3.31
99.72
Ethylbenzene
0.4
1
61
1.64
0.28
100.00
Total
363
478
75.94

20-22

-------
Table 20-4. Risk-Based Screening Results for the New Jersey Monitoring Sites (Continued)
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Elizabeth, New Jersey - ELNJ
Acetaldehyde
0.45
61
61
100.00
15.10
15.10
Benzene
0.13
61
61
100.00
15.10
30.20
1.3 -Butadiene
0.03
61
61
100.00
15.10
45.30
Carbon Tetrachloride
0.17
61
61
100.00
15.10
60.40
Formaldehyde
0.077
61
61
100.00
15.10
75.50
1,2-Dichloroethane
0.038
47
47
100.00
11.63
87.13
Ethylbenzene
0.4
31
61
50.82
7.67
94.80
Hexachloro-1,3 -butadiene
0.045
11
13
84.62
2.72
97.52
Propionaldehyde
0.8
7
61
11.48
1.73
99.26
/?-Dichlorobcnzcnc
0.091
2
29
6.90
0.50
99.75
T richloroethylene
0.2
1
18
5.56
0.25
100.00
Total
404
534
75.66

North Brunswick, New Jersey - NBNJ
Acetaldehyde
0.45
62
62
100.00
16.23
16.23
Formaldehyde
0.077
62
62
100.00
16.23
32.46
Benzene
0.13
61
61
100.00
15.97
48.43
Carbon Tetrachloride
0.17
61
61
100.00
15.97
64.40
1,2-Dichloroethane
0.038
59
59
100.00
15.45
79.84
1.3 -Butadiene
0.03
57
58
98.28
14.92
94.76
Hexachloro-1,3 -butadiene
0.045
14
16
87.50
3.66
98.43
Ethylbenzene
0.4
3
61
4.92
0.79
99.21
1,2-Dibromoethane
0.0017
1
1
100.00
0.26
99.48
/?-Dichlorobcnzcnc
0.091
1
28
3.57
0.26
99.74
T richloroethylene
0.2
1
12
8.33
0.26
100.00
Total
382
481
79.42

Observations from Table 20-4 include the following:
•	Sixteen pollutants failed at least one screen for CSNJ; 62 percent of concentrations
for these 16 pollutants were greater than their associated risk screening value (or
failed screens).
•	Ten pollutants contributed to 95 percent of failed screens for CSNJ and therefore
were identified as pollutants of interest for this site. These 10 include three carbonyl
compounds and seven VOCs.
•	Nine pollutants failed at least one screen for CHNJ; 76 percent of concentrations for
these nine pollutants were greater than their associated risk screening value (or failed
screens).
20-23

-------
•	Seven pollutants contributed to 95 percent of failed screens for CHNJ and therefore
were identified as pollutants of interest for this site. These seven include two carbonyl
compounds and five VOCs.
•	Eleven pollutants failed at least one screen for ELNJ, with nearly 76 percent of
concentrations for these 11 pollutants greater than their associated risk screening
value.
•	Eight pollutants contributed to 95 percent of failed screens for ELNJ and therefore
were identified as pollutants of interest for this site. These eight include two carbonyl
compounds and six VOCs.
•	Eleven pollutants failed at least one screen for NBNJ, with 79 percent of
concentrations for these 11 pollutants greater than their associated risk screening
value.
•	Seven pollutants contributed to 95 percent of failed screens for NBNJ and therefore
were identified as pollutants of interest for this site. These seven include two carbonyl
compounds and five VOCs.
•	The New Jersey sites have seven pollutants of interest in common: acetaldehyde,
formaldehyde, benzene, carbon tetrachloride, 1,3-butadiene, 1,2-dichloroethane, and
hexachloro-1,3 -butadiene.
•	CSNJ is the only NMP site with bromomethane as a site-specific pollutant of interest;
CSNJ is one of only two NMP sites with propionaldehyde as a site-specific pollutant
of interest.
20.4 Concentrations
This section presents various concentration averages used to characterize air toxics
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.
20-24

-------
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at the New Jersey monitoring sites are provided in Appendices J and L.
20.4.1 2013 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 20-5, 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.
20-25

-------
Table 20-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the New Jersey Monitoring Sites

# of






Measured
1st
2nd
3rd
4th


Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs. # of
Samples
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Camden, New Jersey - CSNJ


2.15
3.96
3.06
1.86
2.78
Acetaldehyde
59/59
±0.46
±0.76
±0.48
±0.23
±0.33


0.99
0.88
0.75
0.78
0.85
Benzene
57/57
±0.23
±0.24
±0.17
±0.18
±0.10


0.45
0.52
0.11
0.91
0.52
Bromo methane
53/57
±0.45
±0.72
±0.08
± 1.79
±0.50


0.10
0.09
0.09
0.11
0.10
1.3 -Butadiene
56/57
±0.02
±0.02
±0.03
±0.03
±0.01


0.55
0.62
0.67
0.60
0.61
Carbon Tetrachloride
57/57
±0.05
±0.06
±0.03
±0.04
±0.03


0.10
0.11
0.06
0.08
0.09
1,2-Dichloroethane
53/57
±0.02
±0.01
±0.02
±0.02
±0.01


0.28
0.29
0.36
0.30
0.31
Ethylbenzene
57/57
±0.08
±0.07
±0.06
±0.11
±0.04


3.34
6.69
6.29
3.44
4.96
Formaldehyde
59/59
±0.80
± 1.07
± 1.06
±0.36
±0.59


0.01
0.02
0.01
0.04
0.02
Hexacliloro -1,3 -butadiene
15/57
±0.01
±0.02
±0.02
±0.03
±0.01


0.46
0.88
0.67
0.36
0.60
Propionaldehyde
59/59
±0.09
±0.16
±0.11
±0.04
±0.07
Chester, New Jersey - CHNJ


1.29
1.58
1.18
1.18
1.31
Acetaldehyde
61/61
±0.39
±0.39
±0.21
±0.23
±0.15


0.60
0.39
0.57
0.42
0.49
Benzene
61/61
±0.09
±0.06
±0.36
±0.07
±0.09


0.03
0.04
0.03
0.05
0.04
1.3 -Butadiene
46/61
±0.01
±0.02
±0.02
±0.02
±0.01


0.60
0.59
0.64
0.62
0.61
Carbon Tetrachloride
61/61
±0.04
±0.08
±0.04
±0.03
±0.02


0.09
0.09
0.06
0.06
0.08
1,2-Dichloroethane
54/61
±0.01
±0.02
±0.02
±0.02
±0.01


1.09
2.88
3.22
1.41
2.14
Formaldehyde
61/61
±0.32
±0.56
±0.66
±0.46
±0.34


0.01
0.02
0.01
0.04
0.02
Hexacliloro -1,3 -butadiene
14/61
±0.02
±0.02
±0.01
±0.03
±0.01
20-26

-------
Table 20-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the New Jersey Monitoring Sites (Continued)

# of






Measured
1st
2nd
3rd
4th


Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs. # of
Samples
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)

Elizabeth, New
Jersey - ELNJ




1.89
3.26
3.02
2.26
2.60
Acetaldehyde
61/61
±0.32
±0.65
±0.40
±0.41
±0.26


0.84
0.80
0.76
0.82
0.81
Benzene
61/61
±0.17
±0.09
±0.11
±0.17
±0.07


0.11
0.11
0.10
0.13
0.11
1,3-Butadiene
61/61
±0.03
±0.02
±0.02
±0.03
±0.01


0.60
0.64
0.66
0.61
0.63
Carbon Tetrachloride
61/61
±0.04
±0.05
±0.03
±0.03
±0.02


0.09
0.10
0.05
0.06
0.07
1,2-Dichloroethane
47/61
±0.01
±0.03
±0.02
±0.03
±0.01


0.25
0.54
0.51
0.43
0.43
Ethylbenzene
61/61
±0.07
±0.10
±0.10
±0.13
±0.06


2.89
6.87
6.35
3.56
4.90
Formaldehyde
61/61
±0.35
± 1.83
±0.93
±0.69
±0.67


<0.01
0.01
0.01
0.04
0.02
Hexacliloro -1,3 -butadiene
13/61
±0.01
±0.01
±0.01
±0.02
±0.01
North Brunswick, New Jersey - NBNJ


1.08
1.77
1.98
1.73
1.66
Acetaldehyde
62/62
±0.16
±0.31
±0.28
±0.26
±0.15


0.82
0.63
0.52
0.64
0.65
Benzene
61/61
±0.11
±0.06
±0.06
±0.09
±0.05


0.08
0.06
0.06
0.08
0.07
1,3-Butadiene
58/61
±0.02
±0.01
±0.01
±0.01
±0.01


0.57
0.64
0.65
0.61
0.62
Carbon Tetrachloride
61/61
±0.04
±0.03
±0.03
±0.02
±0.02


0.09
0.10
0.06
0.09
0.09
1,2-Dichloroethane
59/61
±0.01
±0.01
±0.01
±0.01
±0.01


1.17
2.38
3.47
1.91
2.24
Formaldehyde
62/62
±0.25
±0.51
±0.70
±0.29
±0.30


0.03
0.01
<0.01
0.05
0.02
Hexacliloro -1,3 -butadiene
16/61
±0.03
±0.02
±0.01
±0.03
±0.01
Observations for CSNJ from Table 20-5 include the following:
•	The pollutants of interest with the highest annual average concentrations are
formaldehyde (4.96 ± 0.59 |ig/m3) and acetaldehyde (2.78 ± 0.33 |ig/m3). These are
the only two pollutants with annual average concentrations greater than 1 |ig/m3. Of
the VOCs, benzene has the highest annual average concentration (0.85 ± 0.10 |ig/m3).
•	Concentrations of formaldehyde were highest during the second and third quarters of
2013, based on the quarterly averages shown, and nearly double the magnitude of the
first and fourth quarter averages. A review of the data shows that formaldehyde
concentrations measured at CSNJ range from 1.44 |ig/m3 to 11.6 |ig/m3. All but two
of the 25 highest formaldehyde concentrations (those greater than 5 |ig/m3) were
20-27

-------
measured between April and September (with the other two measured in late March).
Further, the 11 lowest concentrations (those less than 3 |ig/m3) were measured
between January and March or November and December. This supports the seasonal
tendency of formaldehyde discussed in Section 4.4.2. Quarterly average
concentrations of acetaldehyde and propionaldehyde exhibit a similar tendency.
•	The quarterly average concentrations of bromomethane each have relatively large
confidence intervals compared to the averages themselves, indicating that
concentrations of bromomethane measured at CSNJ are highly variable. A review of
the data shows that concentrations of bromomethane measured at CSNJ range from
0.04 |ig/m3 to 13.1 |ig/m3, including four non-detects, with a median concentration of
0.07 |ig/m3. This site has the six highest bromomethane concentrations measured
across the program, including all five measurements were greater than 1 |ig/m3.
Observations for CHNJ from Table 20-5 include the following:
•	The pollutants of interest with the highest annual average concentrations are
formaldehyde (2.14 ± 0.34 |ig/m3) and acetaldehyde (1.31 ± 0.15 |ig/m3). These are
the only two pollutants with annual average concentrations greater than 1 |ig/m3. Of
the VOCs, carbon tetrachloride has the highest annual average concentration
(0.61 ± 0.02 |ig/m3).
•	Similar to CSNJ, concentrations of formaldehyde were highest during the second and
third quarters of 2013 at CHNJ, based on the quarterly averages shown, although
considerably less than those measured at CSNJ. A review of the data shows that
formaldehyde concentrations measured at CHNJ range from 0.48 |ig/m3 to
5.34 |ig/m3, with the five highest concentrations all measured in June and July and the
10 highest concentrations all measured between April and September. Conversely, the
lowest 17 formaldehyde concentrations (those less than 1 |ig/m3) were measured
between January and March or October and December. Quarterly average
concentrations of acetaldehyde do not exhibit a similar tendency as the highest
concentrations were measured between January and April and at least one of the 10
highest concentrations was measured in every calendar quarter.
•	The quarterly average concentrations of benzene are fairly similar to each other, but
the confidence interval for the third quarter average is four to six times higher than
the other confidence intervals. A review of the data shows that the two highest
benzene concentrations were both measured during the third quarter and are the only
two benzene concentrations greater than 1 |ig/m3 measured at CHNJ (2.88 |ig/m3
measured on September 13, 2013 and 1.09 |ig/m3 measured on July 3, 2013). The
next highest 13 concentrations were all measured during the first or fourth quarters of
the year. Benzene concentrations measured during the third quarter span an order of
magnitude, ranging from 0.23 |ig/m3 to 2.88 |ig/m3, with a median concentration of
0.35 |ig/m3. This range is greater than the range of concentrations measured in each
of the other quarters yet this quarter has the lowest median benzene concentration.
A similar observation regarding third quarter benzene concentrations was made in the
2012 NMP report.
20-28

-------
Observations for ELNJ from Table 20-5 include the following:
•	The pollutants of interest with the highest annual average concentrations are
formaldehyde (4.90 ± 0.67 |ig/m3) and acetaldehyde (2.60 ± 0.26 |ig/m3). These are
the only two pollutants with annual average concentrations greater than 1 |ig/m3. Of
the VOCs, benzene has the highest annual average concentration (0.81 ± 0.07 |ig/m3).
•	Similar to CSNJ and CHNJ, concentrations of formaldehyde measured at ELNJ were
higher during the warmer months of the year, as indicated by the second and third
quarter average concentrations. A review of the data shows that formaldehyde
concentrations measured at ELNJ range from 1.33 |ig/m3 to 15.9 |ig/m3. The
maximum concentration was measured on May 10, 2013 and is tied for the fifth
highest formaldehyde concentration measured across the program. All 11 highest
formaldehyde concentrations (those greater than 7 |ig/m3) were measured at ELNJ
between April and August while all but one of the 18 formaldehyde concentrations
less than 3 |ig/m3 were measured between January and March or November and
December. Quarterly average concentrations of acetaldehyde exhibit a similar
tendency but the differences are not statistically significant.
•	Concentrations of ethylbenzene were lowest during the first quarter of 2013, as
indicated by the quarterly averages of this pollutant shown in Table 20-5. Eight of the
10 ethylbenzene concentrations less than or equal to 2 |ig/m3 were measured during
the first quarter of 2013. By contrast, only one concentration greater than 5 |ig/m3 was
measured at ELNJ during the first quarter compared to between five and eight during
the other calendar quarters.
Observations for NBNJ from Table 20-5 include the following:
•	The pollutants of interest with the highest annual average concentrations are
formaldehyde (2.24 ± 0.30 |ig/m3) and acetaldehyde (1.66 ± 0.15 |ig/m3). These are
the only two pollutants with annual average concentrations greater than 1 |ig/m3. Of
the VOCs, benzene has the highest annual average concentration (0.65 ± 0.05 |ig/m3),
although the annual average for carbon tetrachloride is similar (0.62 ± 0.02 |ig/m3).
•	Similar to the other New Jersey sites, concentrations of formaldehyde appear higher
during the warmer months of the year, although the differences among the quarterly
averages are not statistically significant for NBNJ. A review of the data shows that
formaldehyde concentrations ranged from 0.613 |ig/m3 to 5.82 |ig/m3, with the three
highest concentrations of formaldehyde (those greater than 5 |ig/m3) measured in
July, and all 14 formaldehyde concentrations greater than 3 |ig/m3 measured during
the second and third quarters of the year.
•	Concentrations of acetaldehyde do not follow the same tendency as formaldehyde at
NBNJ. Concentrations measured during the first quarter of the year were the lowest,
based on the quarterly averages, while concentrations measured during the fourth
quarter were similar to those measured during the second and third quarters. The
maximum concentration of acetaldehyde was measured on November 21, 2013
(2.96 |ig/m3). Concentrations greater than 2 |ig/m3 were measured in all calendar
20-29

-------
quarters except the first quarter (with six measured during the second quarter, nine in
the third, and five in the fourth).
•	Concentrations of benzene appear highest during the first quarter of 2013 at NBNJ
(although the differences among the quarterly averages are not statistically
significant). Five of the six highest concentrations of benzene were measured at
NBNJ in January and February.
Additional observations for the New Jersey sites from Table 20-5 include:
•	Formaldehyde and acetaldehyde were the pollutants of interest with the highest
annual average concentrations for each New Jersey site. Concentrations of these
pollutants were higher at CSNJ and ELNJ than CHNJ and NBNJ. Concentrations of
formaldehyde were higher during the warmer months of the year at each site, as
indicated by the quarterly averages.
•	Of the VOC pollutants of interest, benzene has the highest annual average
concentration for three of the four sites, ranking second to carbon tetrachloride for
CHNJ. Concentrations of benzene were also highest at CSNJ and ELNJ compared to
CHNJ and NBNJ.
•	Concentrations of hexchloro-1,3-butadiene were highest during the fourth quarter of
2013 at each New Jersey site, based on the quarterly averages shown in Table 20-5.
This pollutant was detected in only 13 samples collected at ELNJ, with eight of them
measured during the fourth quarter (and only one measured during the first quarter
and two each in the second and third quarters of 2013). This is true across all of the
New Jersey sites, with the number of measured detections of hexchloro-l,3-butadiene
for the fourth quarter similar to or greater than the number of measured detections for
the other three quarters combined. However, all measured detections of this pollutant
are less than the MDL.
Tables 4-9 through 4-12 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for the New
Jersey sites from those tables include the following:
•	The New Jersey sites appear in Table 4-9 for VOCs a total of seven times (CSNJ,
twice; CHNJ, once; ELNJ, twice; and NBNJ, twice).
•	Three New Jersey sites appear in Table 4-9 for hexachloro-1,3-butadiene, with NBNJ,
CSNJ, and CHNJ ranking third, fifth, and sixth, respectively, for this pollutant.
ELNJ's annual average concentration is similar to the other sites, although it ranks
13th among NMP sites sampling VOCs. Most of the annual average concentrations of
this pollutant across NMP sites are within 0.015 |ig/m3 of each other.
20-30

-------
•	ELNJ has the ninth highest annual average concentration of ethylbenzene and the
10th highest annual average concentration of 1,3-butadiene among NMP sites
sampling VOCs.
•	CSNJ and NBNJ rank seventh and ninth, respectively, for their annual average
concentrations of 1,2-dichloroethane.
•	CSNJ and ELNJ both appear in Table 4-10 for both carbonyl compounds. CSNJ has
the third highest annual average concentrations of both acetaldehyde and
formaldehyde among NMP sites sampling carbonyl compounds. ELNJ has the fourth
highest annual average concentration of formaldehyde and the fifth highest annual
average concentration of acetaldehyde, the same rankings this site had for these
pollutants in the 2012 NMP report.
20.4.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for the pollutants shaded in
gray in Table 20-4 for each of the New Jersey sites. Figures 20-12 through 20-21 overlay the
sites' minimum, annual average, and maximum concentrations onto the program-level minimum,
first quartile, median, average, third quartile, and maximum concentrations, as described in
Section 3.4.3.1.
20-31

-------
Figure 20-12. Program vs. Site-Specific Average Acetaldehyde Concentrations
K

-
	0	¦
h
	0		
¦
-o	¦
0	3	6	9	12	15
Concentration {[jg/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


20-32

-------
Figure 20-13. Program vs. Site-Specific Average Benzene Concentrations
I
Program Max Concentration = 43.5 ^ig/m3
H
Program Max Concentration = 43.5 ^ig/m3
¦4
Program Max Concentration = 43.5 ^ig/m3


Program Max Concentration = 43.5 ^ig/m3



0
2 4
6
Concentration {[jg/m3)
8
10

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i

Site: Site Average
o
Site Concentration Range


Figure 20-14. Program vs. Site-Specific Average Bromomethane Concentration
0
2
4
6 8
Concentration {[jg/m3)
10
12
14

Program:
Site:
1st Qua rti le
¦
Site Average
o
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
Site Concentration Range
4th Qua rti le
~
Average
i

20-33

-------
Figure 20-15. Program vs. Site-Specific Average 1,3-Butadiene Concentrations
«
Program Max Concentration = 21.5 ^ig/m3

Program Max Concentration = 21.5 ^ig/m3
>-
Program Max Concentration = 21.5 ^ig/m3

Program Max Concentration = 21.5 ^ig/m3
0.6	0.9
Concentration {[jg/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


20-34

-------
Figure 20-16. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations
CHNJ
Program Max Concentration = 23.7 ^ig/m3
CSNJ








ELNJ
Program Max Concentration = 23.7 ^ig/m3
Program Max Concentration = 23.7 ^ig/m3
0	0.25	0.5	0.75	1	1.25	1.5	1.75	2
Concentration {[ig/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


20-35

-------
Figure 20-17. Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations
¦
¦
¦
¦
Program Max Concentration = 111 ^ig/m3
Program Max Concentration = 111 ^ig/m3
Program Max Concentration = 111 ^ig/m3
Program Max Concentration = 111 ^ig/m3
0.4	0.6
Concentration {[jg/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


Figure 20-18. Program vs. Site-Specific Average Ethylbenzene Concentrations


h

Program Max Concentration = 18.7 ^ig/m3
















Program Max Concentration = 18.7 ^ig/m3










H	1	r
0	1	2	3	4	5	6
Concentration {[ig/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


20-36

-------
Figure 20-19. Program vs. Site-Specific Average Formaldehyde Concentrations
N i -
N i ¦
I
0	3	6	9	12	15	18	21	24
Concentration {[ig/m3)
Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



20-37

-------
Figure 20-20. Program vs. Site-Specific Average Hexachloro-l,3-Butadiene Concentrations
CSNJ
O
ELNJ
>
0
0.05
0.1
0.15
Concentration {[ig/m3)
0.2
0.25
0.3

Program:
Site:
IstQuartile
¦
Site Average
o
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
Site Concentration Range
4th Qua rti le
~
Average
i

Figure 20-21. Program vs. Site-Specific Average Propionaldehyde Concentration
0
0.25
0.5
0.75 1 1.25
Concentration {[jg/m3)
1.5
1.75
2

Program:
Site:
1st Qua rti le
¦
Site Average
o
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
Site Concentration Range
4th Qua rti le
~
Average
i

Observations from Figures 20-12 through 20-21 include the following:
• Figure 20-12 presents the box plots for acetaldehyde for all four New Jersey sites.
The range of acetaldehyde concentrations measured is largest for CSNJ and
smallest for NBNJ. The annual average concentration of acetaldehyde is highest
for CSNJ and lowest for CHNJ. The annual average concentrations for CSNJ and
ELNJ are greater than the program-level average concentration as well as the
20-38

-------
program-level third quartile. The annual average concentration for NBNJ is less
than the program-level average concentration but greater than the program-level
median concentration while the annual average for CHNJ is less than both the
program-level average and median concentrations. The minimum concentrations
measured at CSNJ and ELNJ are greater than the program-lev el first quartile.
•	Figure 20-13 presents the box plots for benzene. Note that the program-level
maximum benzene concentration (43.5 |ig/m3) is not shown directly on the box
plots as the scale has been reduced to 12 |ig/m3 to allow for the observation of
data points at the lower end of the concentration range. The range of benzene
concentrations measured at each New Jersey site is largest for CHNJ and smallest
for NBNJ. The annual average benzene concentration for CSNJ is greater than the
program-level average concentration; the annual average concentration for ELNJ
is similar to the program-level average concentration; the annual average
concentration for NBNJ is less than the program-level average concentration but
just greater than the program-level median concentration; and CHNJ's annual
average benzene concentration is less than both the program-level average and
median concentrations. Even though the maximum benzene concentration among
the New Jersey sites was measured at CHNJ, the minimum benzene concentration
measured at an NMP site sampling benzene with Method TO-15 was also
measured at CHNJ.
•	Figure 20-14 presents the box plot for bromomethane for CSNJ, the only NMP
site for which bromomethane is a pollutant of interest. Note that the first, second,
and third quartiles for bromomethane are zero at the program-level and therefore
not visible on the box plot due to the large number of non-detects. This box plot
shows that the maximum concentration of bromomethane across the program was
measured at CSNJ. All concentrations of bromomethane greater than 1 |ig/m3
across the program were measured at CSNJ. CSNJ's annual average concentration
of bromomethane (0.52 ± 0.50 |ig/m3) is an order of magnitude greater than the
program-level average concentration (0.054 |ig/m3).
•	Figure 20-15 presents the box plots for 1,3-butadiene. Similar to benzene, the
program-level maximum concentration (21.5 |ig/m3) is not shown directly on the
box plots as the scale has been reduced to 1.5 |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 CHNJ
while the largest range was measured at ELNJ, although a similar range was
measured at CSNJ. ELNJ is the only site that did not measure any non-detects of
this pollutant; in fact, the minimum concentration measured at ELNJ is greater
than the program-level first quartile. The annual average 1,3-butadiene
concentrations for all four sites are less than the program-level average
concentration. However, the program-level average concentration is being driven
by the higher measurements collected at a few sites. Among the New Jersey sites,
the annual average concentration is lowest for CHNJ and highest for ELNJ,
although only 0.07 |ig/m3 separates them.
20-39

-------
Figure 20-16 presents the box plots for carbon tetrachloride. The scale of the box
plots in Figure 20-16 has also been reduced to allow for the observation of data
points at the lower end of the concentration range, as the program-level maximum
carbon tetrachloride concentration (23.7 |ig/m3) is considerably greater than the
majority of measurements. The range of carbon tetrachloride measurements is
largest for CHNJ and lowest for ELNJ, although this is being driven by the
minimum concentration measured at CHNJ. Excluding this data point would
result in CHNJ's range of measurement resembling CSNJ's range of
measurements. The annual average concentrations vary little among the New
Jersey sites, although they are all less than the program-level average and median
concentrations.
Figure 20-17 presents the box plots for 1,2-dichloroethane. Similar to other
VOCs, the program-level maximum concentration (111 |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 being driven by the higher measurements
collected at a few monitoring sites. Figure 20-17 shows that the maximum
1,2-dichloroethane concentrations measured at the New Jersey sites are all less
than 0.2 |ig/m3, and thus, the entire range of measurements collected at each site
is less than the average concentration across the program. The annual average
concentrations for these sites are shown on either side of the program-level
median concentration.
Figure 20-18 presents the box plots for ethylbenzene for CSNJ and ELNJ, the
only sites for which this is a pollutant of interest. The scale of the box plots in
Figure 20-18 have also been reduced to allow for the observation of data points at
the lower end of the concentration range. All of the ethylbenzene concentrations
measured at these two sites are less than or equal to 1.0 |ig/m3. The annual
average concentration for ELNJ is just greater than the program-level average
concentration and program-level third quartile. The range of measurements
collected at CSNJ is smaller and its annual average concentration is less than the
program-level average concentration.
Figure 20-19 presents the box plots for formaldehyde for all four sites. The range
of formaldehyde concentrations is smallest for CHNJ and largest for ELNJ. The
annual average concentration of formaldehyde for ELNJ is similar to the annual
average for CSNJ, both of which are greater than the program-level average
concentration and third quartile. The annual average concentrations for CHNJ and
NBNJ are similar to each other and the program-level median concentration. The
minimum concentration measured at CSNJ is greater than the program-level first
quartile.
Figure 20-20 presents the box plots for hexchloro-1,3-butadiene for all four sites.
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-l,3-butadiene
concentrations for all four sites are greater than the program-level average
20-40

-------
concentration, though by only a small margin. Roughly one-quarter of the
measurements of these pollutants were measured detections, although none were
greater than the MDL.
• Figure 20-21 presents the box plot for propionaldehyde for CSNJ, the only New
Jersey site for which this is a pollutant of interest. The minimum concentration
measured at CSNJ is greater than the program-level first quartile and just less than
the program-level median concentration. The annual average concentration for
CSNJ is nearly twice the program-level average concentration. CSNJ is one of
only two NMP sites sampling carbonyl compounds with propionaldehyde as a
pollutant of interest (BTUT is the other).
20.4.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.3.2.
CHNJ, ELNJ, and NBNJ have sampled VOCs and carbonyl compounds under the NMP for
many years. ELNJ has sampled under the NMP since 2000 and CHNJ and NBNJ since 2001.
Thus, Figures 20-22 through 20-43 present the 1-year statistical metrics for each of the pollutants
of interest first for CHNJ, then for ELNJ and NBNJ. The statistical metrics presented for
assessing trends include the substitution of zeros for non-detects. If sampling began mid-year, a
minimum of 6 months of sampling is required for inclusion in the trends analysis; in these cases,
a 1-year average concentration is not provided, although the range and percentiles are still
presented. CSNJ began sampling under the NMP is 2013; thus, a trends analysis was not
performed for this site.
20-41

-------
Figure 20-22. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at CHNJ
Maximum
Concentration for
2004 is 29.1 n.g/m3
£
I
T
t
Tr

I
L-2"1

(J ¦+
20011 2002 2003
2004
2005 2006
2007
Year
2008
2009 2010 2011
2012 2013

O 5th Percentile
- Minimum
— Median
-
Maximum
O 95th Percentile

1 A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
Observations from Figure 20-22 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
concentration is not presented, although the range of measurements is provided.
•	The two highest acetaldehyde concentrations were measured at CHNJ in 2004
(29.1 |ig/m3 and 11.5 |ig/m3). All other concentrations measured in 2004 were less
than 3 |ig/m3. Only two additional acetaldehyde concentrations greater than 5 |ig/m3
have been measured at CHNJ, one in 2005 (8.38 |ig/m3) and one in 2012
(5.38 |ig/m3).
•	An overall decreasing trend in the 1-year average and median concentrations is shown
though 2006, with the exception of 2004, when the maximum concentrations were
measured. Between 2006 and 2010, the 1-year average and median concentrations
changed little, with the 1-year average concentrations varying by less than 0.25 |ig/m3
over these years.
•	All of the statistical metrics exhibit an increase from 2010 to 2011. Although the
maximum concentration increased again for 2012, the 95th percentile decreased
nearly 1 |ig/m3, indicating that fewer concentrations at the upper end of the range
were measured in 2012. The second highest concentration measured in 2012 is half
20-42

-------
the magnitude of the maximum concentration for 2012. Additional decreases for all
of the statistical parameters are shown for 2013.
Figure 20-23. Yearly Statistical Metrics for Benzene Concentrations Measured at CHNJ





T












r-»
Lt

o-
r



H
!>-

-t

T T

r-O-











LqJ

r

-o


I

|—o

-a

-c
1
>-
>-
T

—



-<
p-





-o-








20011 2002 2003 2004 20052 2006 2007 2008 2009 2010 2011 2012 2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1	A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
2	A 1-year average is not presented due to low completeness in 2005.
Observations from Figure 20-23 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
concentration is not presented, although the range of measurements is provided. In
addition, a 1-year average concentration for 2005 is not provided due to low
completeness.
•	The maximum benzene concentration measured at CHNJ was measured on
September 13, 2013 (2.88 |ig/m3). Only nine 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, 2012, and 2013).
•	The 1-year average and median concentrations exhibit a decreasing trend through
2007, although a 1-year average concentration is not provided for 2001 or 2005.
20-43

-------
•	Even though an increase in the 1-year average concentration is shown from 2007 to
2008, this increase is being driven less by the two measurements greater than 2 |ig/m3
and more by the measurements in the mid- to upper-end of the concentration range.
This evident from the increase shown in the median concentration. The number of
concentrations between 0.5 |ig/m3 and 1 |ig/m3 nearly doubled from 2007 to 2008
(from 15 to 28).
•	The difference between the 5th and 95th percentiles, or the range within which the
majority of concentrations fall, is at a maximum for 2009, indicating an increase in
variability of the concentrations measured. Conversely, the difference between the 5th
and 95th percentiles is at a minimum for the following year.
•	An increase in the 1-year average, median, 95th percentile, and maximum
concentrations is shown from 2010 to 2011 and again for 2012. Although the range of
concentrations measured is at a maximum for 2013, all of the statistical metrics
exhibit decreases for 2013.
•	Although an undulating pattern is shown in the 1-year average concentrations of
benzene between 2006 and 2013, the averages have varied by less than 0.2 |ig/m3,
ranging from 0.47 |ig/m3 (2007) to 0.64 |ig/m3 (2012).
Figure 20-24. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at CHNJ
0.5 -




0.3 -









0.1 -










I
T

r
r-d

¦-i-.
b	^
	 ¦
™ ail
- ^ r
	 ..v		
m
rl
		 r"".i
20011 2002 2003 2004 20052 2006 2007 2008 2009 2010 2011 2012 2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile •*•*¦¦/••• Average
1	A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
2	A 1-year average is not presented due to low completeness in 2005.
20-44

-------
Observations from Figure 20-24 for 1,3-butadiene measurements collected at CHNJ
include the following:
•	The maximum 1,3-butadiene concentration was measured in 2003 (0.58 |ig/m3) and is
the only concentration greater than 0.5 |ig/m3 measured at CHNJ. Only five
1,3-butadiene concentrations measured at CHNJ are greater than 0.2 |ig/m3.
•	For 2001 and 2004, the minimum, 5th percentile, median, and 95th percentile are all
zero. This is because the percentage of non-detects was greater than 95 percent for
these years. More than 50 percent of the measurements were non-detects between
2001 and 2005 (as well as 2010), as indicated by the median concentration. The
percentage of non-detects decreased steadily between 2004 (96 percent) and 2008
(17 percent), when the percentage of non-detects reached a minimum. After 2008, the
percentage of non-detects reported varied considerably, from as low as 18 percent
(2012) and as high as 70 percent (2010).
•	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.
•	Despite the increase in the number of non-detects, and thus zero substitutions, from
2012 to 2013 (from 11 to 17) and the smaller range of concentrations measured, the
changes in the 1-year average and median concentrations are minimal.
20-45

-------
Figure 20-25. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
CHNJ
1.2 -






T



T


"c-
T
—o-
W*





T

rl
C""


0.6 -

r^~
T

c-




—


r



T


..






























¦¦




20011 2002 2003 2004 20052 2006 2007 2008 2009 2010 2011 2012 2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile •*•*¦¦/••• Average
1	A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
2	A 1-year average is not presented due to low completeness in 2005.
Observations from Figure 20-25 for carbon tetrachloride measurements collected at
CHNJ include the following:
•	The range of carbon tetrachloride concentrations measured appear to increase
significantly from 2001 to 2002, with fairly similar ranges measured between 2003
and 2005. This apparent increase is predominantly due to a few non-detects that were
measured between 2002 and 2005. After 2005, only one non-detect was reported
(2007).
•	All of the statistical parameters exhibit an increase from 2007 to 2008. The 95th
percentile for 2007 is just greater than the 1-year average and median concentrations
calculated for 2008. 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 more than doubled from 2007 to 2008.
•	The minimum concentration measured in 2009 increased by an order of magnitude
from 2008 and the maximum concentration increased as well. Yet the 1-year average
increased only slightly from 2008 to 2009 and the median concentration decreased
slightly.
20-46

-------
•	The minimum carbon tetrachloride concentration decreased every year after 2009, as
did the maximum concentration, with the exception of 2011. Between 2009 and 2013,
the range within which most of the concentrations fell, as indicated by the difference
between the 5th and 95th percentiles, decreased each year (except for 2011), and is at
a minimum for 2013 for all the years of sampling.
•	Between 2008 and 2013, the 1-year average concentrations varied by less than
0.11 |ig/m3.
Figure 20-26. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
CHNJ
.7
Maximum
Concentration for
2008 is 1.27 n.g/m3
.6
.5
.4
.3
.2
.1
.0
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1	A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
2	A 1-year average is not presented due to low completeness in 2005.
Observations from Figure 20-26 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 concentrations shown for the
later years of sampling. The number of measured detections decreased slightly for
2013 but still account for the majority of measurements.
20-47

-------
•	2012 is the first year that the median concentration and 5th percentile are greater than
zero. Aside from the three non-detects, the range of measurements collected in 2012
is relatively small, ranging from 0.0527 |ig/m3 to 0.121 |ig/m3. The 1-year average
and median concentrations calculated for 2012 are less than 0.001 |ig/m3 apart,
indicating little variability associated with the measurements collected in 2012.
•	The 5th percentile returned to zero for 2013, as six additional non-detects were
measured in 2013. However, the 1-year average and median concentrations did not
change. The effects of the additional non-detects are balanced by the additional
concentrations measured at the upper end of the concentration range. The number of
1,2-dichloroethane concentrations greater than 0.1 |ig/m3 doubled from 2012 to 2013,
increasing from four to nine.
Figure 20-27. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at CHNJ
Maximum
Concentration for
2004 is 57.2 p.g/m-
200r 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Year
O 5th Percentile	— Minimurr
Maximum	O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
Observations from Figure 20-27 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.
20-48

-------
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
changed little through 2009. Less than 0.5 |ig/m3 separates the 1-year average
concentrations calculated for the period between 2006 and 2009.
The 1-year and median concentrations decreased significantly for 2010, when both
statistical parameters are at a minimum. This is due primarily to the measurements at
the lower end of the concentration range. The number of concentrations less than
1 |ig/m3 increased from two in 2009 to 21 in 2010.
Similar to acetaldehyde, all of the statistical metrics calculated for formaldehyde
exhibit an increase from 2010 to 2011, including the 95th percentile, which is greater
than the maximum concentration measured in 2010. Four formaldehyde
concentrations measured in 2011 are greater than the maximum concentration
measured in 2010. Although the range of measurements decreased for 2012, little
change is shown in the 1-year average concentration and the median continued to
increase. This is due to a lower number of concentrations at the lower end of the
concentration range. The number of measurements less than 1 |ig/m3 fell from 19 in
2011 to four in 2012.
With the exception of the minimum concentration, all of the statistical parameters
exhibit decreases for 2013. The maximum formaldehyde concentration measured at
CHNJ in 2013 is the lowest maximum for any given year.
20-49

-------
Figure 20-28. Yearly Statistical Metrics for Hexachloro-l,3-Butadiene Concentrations
Measured at CHNJ

2001 2002 2003 2004 20052 2006 2007 2008 2009 2010 2011 2012 2013
Year
0 5th Percentile	— Minimurr
0 95th Percentile
1	A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
2	A 1-year average is not presented due to low completeness in 2005.
Observations from Figure 20-28 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 began
increasing again after 2009, with one measured in 2010, four in 2011, 12 in 2012, and
14 in 2013, which is the maximum number of measured detections since sampling
began.
20-50

-------
Figure 20-29. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at ELNJ









-o-
T



£

o-


T

















-
1 i








T





-a-

i i
-H3 J











1




r
	

| i T





20001 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Year
0 5th Percentile	- Minimum	— Median	- Maximum	O 95th Percentile •~•^'~~•Average
1 A 1-year average is not presented due to low completeness in 2000.
Observations from Figure 20-29 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 concentration is not presented for 2000, although
the range of measurements is provided.
•	The maximum acetaldehyde concentration was measured at ELNJ 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 concentrations measured between 2003 and 2007 is considerably higher
than those collected during the first 3 years of sampling. The 1-year average
concentration increased significantly from 2002 to 2003. This increasing trend
continued through 2007, although the rate of change slowed over the years. A
significant decrease in the measurements is shown from 2007 to 2008, where the
1-year average decreased by more than half. The range of measurements collected in
2008 is more similar to the range shown before 2003.
20-51

-------
•	Although an increasing trend is also shown between 2008 and 2011, the 1-year
average concentrations are roughly half the magnitude of those shown before 2008.
•	All of the statistical parameters exhibit decreases from 2011 to 2012 with additional
decreases shown for some of the parameters for 2013. The range of measurements
collected in 2013 is the smallest since the onset of sampling at ELNJ. The maximum
concentration measured in 2013 is at its lowest since 2002.
Figure 20-30. Yearly Statistical Metrics for Benzene Concentrations Measured at ELNJ

Maximum
Concentration for
2008 is 34.3 ng/m3















|H
5-

-c
>1

—<
>-

rO-
L2-

--|

-o-




L-«-



¦Ub
T


T




LgJ

-o-

(—
>-|



-c
5-
r^i
—o—

-a-1
LgJ


LgJ
lip
20001 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Year
O 5th Percentile	- Minimum	— Median	- Maximum	O 95th Percentile •~•^'~~•Average
1 A 1-year average is not presented due to low completeness in 2000.
Observations from Figure 20-30 for benzene measurements collected at ELNJ include the
following:
•	VOC sampling under the NMP also began at ELNJ in January 2000. However, a
1-year average concentration is not presented for 2000 due to low completeness,
although the range of measurements is provided.
•	The maximum benzene concentration (34.3 |ig/m3) was measured in 2008 and is
more than four times higher than the next highest concentration (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 fairly steady decreasing trend in the 1-year average and median concentrations is
shown through 2007.
20-52

-------
•	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. The median increased by less than 0.02 |ig/m3
between 2007 and 2008 further indicating that this outlier is the primary driver
pulling the 1-year average concentration upward. However, the minimum
concentration doubled from 2007 to 2008 and the 5th percentile increased as well,
indicating that the outlier may not be the only factor.
•	Even though two of the three highest concentrations were measured at ELNJ in 2009,
the 1-year average concentration decreased from 2008 to 2009, likely a result of the
magnitude of the outlier affecting the 2008 calculations.
•	Figure 20-30 shows that benzene concentrations measured in 2010, 2011, and 2012
were fairly consistent. The difference in the 1-year average concentrations for these
years is less than 0.02 |ig/m3.
•	Additional decreases are shown for 2013, as no benzene concentrations greater than
2 |ig/m3 were measured in 2013, the only year for which this is true. The 1-year
average benzene concentration is at a minimum for 2013, and is the only 1-year
average concentration less than 1 |ig/m3.
Figure 20-31. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at ELNJ









—



1 _ T _ T









ri
				

lift
	-d&i —•~••• i&	r
L^II U
3 (5 (^3 e5
20001 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Year
O 5th Percentile	- Minimum	— Median	- Maximum	O 95th Percentile •~•^'~~•Average
1 A 1-year average is not presented due to low completeness in 2000.
20-53

-------
Observations from Figure 20-31 for 1,3-butadiene measurements collected atELNJ
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
and only 15 concentrations measured at ELNJ are greater than 0.5 |ig/m3.
•	The minimum and 5th percentile are zero for the first 6 years of sampling, indicating
that at least 5 percent of the measurements were non-detects. For 2004, the median
concentration is also zero, indicating that at least half of the measurements were non-
detects. Between 2000 and 2006, the percentage of non-detects ranged from 5 percent
(2006) to 57 percent (2004). After 2006, only two non-detects have been measured
(both in 2011).
•	There is a decreasing trend in the 1-year average concentration through 2004, after
which the 1-year average concentration remains fairly static. Even with the higher
concentration measured in 2009, the 1-year average concentration for 2009 is similar
to the 1-year average concentration for 2008. Between 2005 and 2012, the 1-year
average concentration has ranged from 0.12 |ig/m3 (2010) to 0.16 |ig/m3 (2006 and
2009).
•	Concentrations of 1,3-butadiene measured at ELNJ have become less variable in
recent years, with concentrations measured in 2010 and 2013 exhibiting the least
variability. These two years have the smallest range of concentrations measured and
the smallest differences between the 5th and 95th percentiles, the range within which
the majority of concentrations fall.
20-54

-------
Figure 20-32. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
ELNJ
1.50
1.25 -
T





T T

0.75 -
0.50 -
0.25 -
o.oo -

r-O

1

-o-
T7"

-o
-o
I
O-i
-0
¦h
-o

|-J
-o-

i
<¦
-0

r1
Y
I


r
rf"
-O


-6-



..p
T

y	
i
		
















r






1




-






20001 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented due to low completeness in 2000.
Observations from Figure 20-32 for carbon tetrachloride measurements collected at ELNJ
include the following:
•	The trends graph for carbon tetrachloride concentrations measured at ELNJ resembles
the trends graph for CHNJ.
•	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 roughly 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 2013 are greater than 0.6 |ig/m3.
•	The difference between the 5th percentile and 95th percentile, or the range within
which the majority of measurements fall, has been decreasing each year since 2005
and is at a minimum for 2013. Less than 0.25 |ig/m3 separates these parameters for
2013.
20-55

-------
Figure 20-33. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
ELNJ















<5

O

1
—



-X-
£

-L
o


	
¦ o ¦ -O-	O	-O-	¦O"" ¦c
L	*
	o	


>'

2000 1 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Year
0 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile •~~^"~~•Average
1 A 1-year average is not presented due to low completeness in 2000.
Observations from Figure 20-33 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/m3 to 0.144 |ig/m3. The 1-year average and median
concentrations calculated for 2012 are approximately 0.0015 |ig/m3 apart, indicating
relatively little variability associated with the measurements collected in 2012.
•	For 2013, the number of non-detects more than doubled (from six in 2012 to 14 in
2013), accounting for nearly one-quarter of the measurements collected. Yet, the
1-year average concentration changed little and the median concentration increased.
Although the maximum concentration increased only slightly from 2012 to 2013, the
number of 1,2-dichloroethane concentrations greater than 0.1 |ig/m3 measured at
ELNJ increased from seven in 2012 to 20 in 2013.
20-56

-------
Figure 20-34. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at
ELNJ
1
LgJ L_£J
T 1
T
r^i
LgJ

&
	
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Year
0 5th Percentile	— Minimurr
O 95th Percentile
• Average
1 A 1-year average is not presented due to low completeness in 2000.
Observations from Figure 20-34 for ethylbenzene measurements collected at ELNJ
include the following:
•	The trends graph for ELNJ's ethylbenzene concentrations resembles the trends graph
for ELNJ's benzene concentrations.
•	There is an overall decreasing trend in the 1-year average and median concentrations
between 2001 and 2007.
•	A significant increase in the statistical parameters is shown for 2008. The median
concentration for 2008 is greater than the 95th percentile for 2007. The number of
ethylbenzene measurements greater than 1 |ig/m3 increased from one in 2007 to 16 in
2008.
•	The measurements collected in 2009 more closely 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 collected spanning less than 0.75 |ig/m3.
20-57

-------
• Between 2009 and 2013, the majority of concentrations fell within a fairly similar
range and the 1-year average concentrations did not change significantly, ranging
from 0.41 |ig/m3 (2012) to 0.51 |ig/m3 (2011).
Figure 20-35. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
ELNJ














I 1



—r®"
i
r-O

-o-
T







T


—











T


Lr










	

-r








T
















*

T

L
2"1


T 1
20001 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile •~~^"~~•Average
1 A 1-year average is not presented due to low completeness in 2000.
Observations from Figure 20-35 for formaldehyde measurements collected at ELNJ
include the following:
•	The maximum formaldehyde concentration was measured at ELNJ in 2013
(15.88 |ig/m3), A total of 14 concentrations greater than 10 |ig/m3 have been
measured at ELNJ, with the most measured in 2007 (three).
•	After a decreasing trend through 2002, there was a significant increase in
formaldehyde concentrations from 2002 to 2003, as shown by the median
concentration, 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
concentrations of formaldehyde, which ranged from 4.52 |ig/m3 (2006) to 4.70 |ig/m3
(2005) during this time period.
20-58

-------
• Similar to acetaldehyde, the 1-year average concentration of formaldehyde decreased
significantly between 2007 and 2008, as the magnitude of concentrations measured
decreased considerably. Afterward, an increasing trend is shown through 2010. While
Figure 20-29 for acetaldehyde shows a continued increase for 2011 followed by a
decrease for 2012, formaldehyde concentrations exhibit a decrease for 2011 followed
by increases for 2012 and 2013. The 1-year average concentration of formaldehyde
for ELNJ for 2013 is the highest 1-year average calculated since the onset of
sampling.
Figure 20-36. Yearly Statistical Metrics for Hexachloro-l,3-Butadiene Concentrations
Measured at ELNJ
0.5 -




0.3 -
0.2 -




r
H
|-<
1 1 I
0.0 -
...«
>	<

20001 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented due to low completeness in 2000.
Observations from Figure 20-36 for hexachloro-l,3-butadiene measurements collected at
ELNJ include the following:
•	There were no measured detections of hexachloro-1,3-butadiene measured at ELNJ
during the first 5 years of sampling.
•	The number of measured detections increased to 13 for 2005, representing 22 percent
of measurements, then decreased to five for 2006. Between 2007 and 2010, a single
measured detection was measured (2008). Beginning in 2010, the number of
measured detections began increasing again, from five for 2011 to seven for 2012,
and 13 in 2013.
20-59

-------
Figure 20-37. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at NBNJ

Maximum
Concentration for
2004 is 111 |ig/m3





-o-
i

-O-
T"

1
-Q-

i
-2-



X
L2"

|H
>-

pL
LgJ
T
fin
tijk"
L^J



¦Or"
-
I
LgJ
	£>.. |-5-| 1-5-1
		
^ L£J La-1
20011 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Year
0 5th Percentile	— Minimum	- Median	- Maximum	0 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
Observations from Figure 20-37 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
concentration is not presented, although the range of measurements is provided.
•	The maximum acetaldehyde concentration was measured in 2004 (111 |ig/m3). This
concentration is nearly seven times higher, and an order of magnitude higher, than the
next highest concentration (16.2 |ig/m3, measured in 2005).
•	Of the 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 concentration,
however, does not. If the outlier was removed from the data set for 2004, the 1-year
average concentration for 2004 would be less than the 1-year average concentration
for 2005.
•	The 1-year average concentration decreases significantly between 2005 and 2007, as
do all of the other statistical parameters. This is followed by a significant increase in
20-60

-------
the concentrations measured for 2008 as the range of concentrations measured
doubled.
•	Between 2008 and 2011, the 1-year average concentrations have an undulating
pattern, fluctuating between 2 |ig/m3 and 3 |ig/m3.
•	The concentrations decreased significantly for 2012, when the 1-year average
concentration is at a minimum (1.41 |ig/m3).
•	The smallest range of acetaldehyde concentrations was measured at NBNJ in 2013,
although slight increases are shown for the 1-year average and median concentrations.
Concentrations were higher overall in 2013 compared to 2012, although this is
obscured somewhat by the compact range of concentrations measured. The minimum
acetaldehyde concentration measured in 2012 is an order of magnitude less than the
minimum concentration measured in 2013. Further, the number of measurements less
than 1 |ig/m3 decreased from 17 in 2012 to eight in 2013. Differences are also evident
at the upper end of the concentration range. Although the maximum concentration
measured in 2013 is less than the maximum concentration measured in 2012, the
number of concentrations greater than 2 |ig/m3 is higher for 2013: eight acetaldehyde
concentrations greater than 2 |ig/m3 were measured in 2012 compared to 21 for 2013.
Figure 20-38. Yearly Statistical Metrics for Benzene Concentrations Measured at NBNJ
















T






—
L9-








T



T







1-a-



£

l l


1 rh










i-4-
	
	










20011 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Year
O 5th Percentile	— Minimum	- Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
20-61

-------
Observations from Figure 20-38 for benzene measurements collected atNBNJ include
the following:
•	Sampling for VOCs under the NMP also began at NBNJ in May 2001. Because a full
year's worth of data is not available for 2001, a 1-year average concentration is not
presented, although the range of measurements is provided.
•	The maximum benzene concentration was measured in 2012 (4.00 |ig/m3); aside from
this measurement, only three additional concentrations of benzene greater than
3 |ig/m3 have been measured at NBNJ.
•	Although a slight decreasing trend in the 1-year average concentration is shown
between 2002 and 2004, a significant decrease is shown between 2005 and 2007,
when both the median and 1-year average concentrations are 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 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
measurements at the upper-end of the concentration increased substantially for 2012.
In addition to a higher maximum concentration, the number of benzene measurements
greater than 0.75 |ig/m3 increased from 11 in 2011 to 31 in 2012, accounting for more
than half of the measurements.
•	The range of benzene concentrations measured at NBNJ in 2013 spans less than
1 |ig/m3 and is very similar to the levels of benzene measured in 2010. The 1-year
average concentration decreased significantly from 2012 to 2013 and is similar to the
1-year averages calculated for the period between 2008 and 2011.
20-62

-------
Figure 20-39. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at NBNJ


¦¦





r-O-













X


i-o-









1

7


















T

r
1
H

20011 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Year
0 5th Percentile	— Minimum	- Median	- Maximum	0 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
Observations from Figure 20-39 for 1,3-butadiene measurements collected at NBNJ
include the following:
•	The maximum 1,3-butadiene concentration was measured at NBNJ in 2005
(0.47 |ig/m3) and is the only measurement greater than 0.40 |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 only two in 2010 to 17 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. Three non-detects were measured in
2013.
•	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 concentrations measured. The number of non-detects
increased from 35 in 2003 to 56 in 2004 (accounting for more than 93 percent of the
20-63

-------
samples collected in 2004). Thus, many zeros were substituted into this average. The
increase in the 1-year average concentration shown from 2004 to 2005 results from a
combination of fewer non-detects and a larger range of measurements. The number of
non-detects decreased to 27 for 2005, accounting for fewer than half of the
measurements for the first time.
•	The 1-year average concentration exhibits little change between 2005 and 2011,
ranging from 0.046 |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 largely
due to the decrease in non-detects (and thus, zeroes substituted for non-detects in the
calculations) from 17 non-detects in 2011 to zero for 2012. The number of
concentrations at the upper end of the concentration range increased as well; the
number of measurements greater than 0.1 |ig/m3 doubled, increasing from eight in
2011 to 18 in 2012.
•	The 1,3-butadiene concentrations measured in 2013 decreased from 2012 levels but
were still higher than those measured in the previous years. These two years have the
only 1-year average and median concentrations greater than 0.06 |ig/m3.
Figure 20-40. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
NBNJ
1.50
1.25
o.oo -I	1—L-e-J—i—L-e-J—i—L-e-J—¦—L-e-J—¦	¦	¦	¦	¦	J-	¦—
20011 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
20-64

-------
Observations from Figure 20-40 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 for each year
during this time frame.
•	The 1-year average concentration changed little between 2002 and 2005, ranging
from 0.49 |ig/m3 to 0.53 |ig/m3. An increase in the 1-year average concentration is
shown from 2005 to 2006, although the change is not statistically significant. This is a
result of higher concentrations at both the lower and upper end of the concentration
range. Between 2004 and 2007, the median concentration varied by only
0.003 |ig/m3.
•	All of the statistical parameters exhibit increases for 2008. The minimum
concentration increased six-fold from 2007 to 2008. In addition, there were 20
measurements collected in 2008 that were greater than the maximum concentration
measured in 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 differences between the 5th percentile and 95th percentile decrease
each year, the 1-year average and median concentrations exhibit an increasing trend
through 2012.
•	Carbon tetrachloride concentrations measured in 2013 exhibit the least amount of
variability. The smallest range of carbon tetrachloride concentrations was measured
in 2013, the difference between the 5th and 95th percentiles is at a minimum, and the
difference between the 1-year average and median concentrations is less than
0.001 |ig/m3.
20-65

-------
Figure 20-41. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
NBNJ




0.15 -





,0

-c
¦c
-c
(-
r-O

0.05 -
o.oo -





i
o


p
I—'
-. ^	<—
L	«
r-4
—-G
a	¦




20011 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Year
0 5th Percentile	— Minimum	— Median	- Maximum	0 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
Observations from Figure 20-41 for 1,2-dichloroethane measurements collected at NBNJ
include the following:
•	There were no measured detections of 1,2-dichloroethane between 2001 and 2004.
Between one and four measured detections were measured between 2005 and 2007,
after which 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, 58 for 2012, and 59 in 2013. This increase in the number of measured
detections is very similar to what was exhibited by the measurements collected at
CHNJ and ELNJ. This also 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 two non-detects, the range of measurements collected in 2012 is relatively small,
ranging from 0.053 |ig/m3 to 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. A
similar observation can be made for 2013, although slight increases are shown for the
1-year average and median concentrations.
20-66

-------
Figure 20-42. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
NBNJ

Maximum
Concentration for
2004 is 96.1 ng/m3
20011 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
Observations from Figure 20-42 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 13 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 1-year average concentration for 2004 would fall between those of 2003
and 2005, exhibiting lesser increases. However, concentrations were higher overall in
2004 compared to 2003 as the number of concentrations greater than 3 |ig/m3 doubled
from 2003 to 2004, from 17 to 34. At the lower end of the concentration range, five
concentrations measured in 2003 are less than the minimum concentration measured
in 2004.
•	After 2005, concentrations of formaldehyde measured at NBNJ decreased
significantly, with the 1-year average and median concentrations decreasing each year
and reaching a minimum for 2008. This year also has the smallest range of
20-67

-------
formaldehyde concentrations measured, although a similar range was also measured
in 2010.
•	Between 2008 and 2012, a year with more variability in the measurements alternates
with a year with less variability. The measurements for 2011 exhibit a considerable
amount of variability compared to the rest of the years within this period. The 95th
percentile for 2011 is more than double the 95th percentile for the other years within
this period. Yet, the median concentrations are nearly the same for 2011 and 2012.
•	Most of the statistical parameters exhibit at least a slight increase for 2013. The
minimum concentration measured in 2013 is an order of magnitude greater than
minimum concentration measured in 2012. In addition, the number of formaldehyde
concentrations greater than 2 |ig/m3 measured at NBNJ in 2013 increased
considerably, from 18 in 2012 to 33 in 2013, accounting for more than half of the
measurements.
Figure 20-43. Yearly Statistical Metrics for Hexachloro-l,3-Butadiene Concentrations
Measured at NBNJ
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
O 5th Percentile	— Minimurr
O 95th Percentile
* Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
20-68

-------
Observations from Figure 20-43 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 between 2007 and 2009. A single measured
detection was reported for 2010. The number of measured detections increased to
eight for 2011, 11 for 2012, and 16 for 2013, the most since the onset of VOC
sampling at NBNJ, accounting for roughly one-fourth of the measurements.
20.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at each New Jersey monitoring site. Refer to Sections 3.2, 3.4.3.3, and
3.4.3.4 for definitions and explanations regarding the various toxicity factors, time frames, and
calculations associated with these risk-based screenings.
20.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the New Jersey sites and where annual average
concentrations could be calculated, risk was examined by calculating cancer risk and noncancer
hazard approximations. These approximations can be used as risk estimates for cancer and
noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers want to shift their air
monitoring priorities. Refer to Section 3.4.3.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 20-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
20-69

-------
Table 20-6. Risk Approximations for the New Jersey Monitoring Sites
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Camden, New Jersey - CSNJ
Acetaldehyde
0.0000022
0.009
59/59
2.78
±0.33
6.12
0.31
Benzene
0.0000078
0.03
57/57
0.85
±0.10
6.65
0.03
Bromomethane

0.005
53/57
0.52
±0.50

0.10
1.3 -Butadiene
0.00003
0.002
56/57
0.10
±0.01
2.92
0.05
Carbon Tetrachloride
0.000006
0.1
57/57
0.61
±0.03
3.65
0.01
1,2-Dichloroethane
0.000026
2.4
53/57
0.09
±0.01
2.33
<0.01
Ethylbenzene
0.0000025
1
57/57
0.31
±0.04
0.76
<0.01
Formaldehyde
0.000013
0.0098
59/59
4.96
±0.59
64.54
0.51
Hexachloro-1,3 -butadiene
0.000022
0.09
15/57
0.02
±0.01
0.48
<0.01
Propionaldehyde

0.008
59/59
0.60
±0.07

0.07
Chester, New Jersey - CHNJ
Acetaldehyde
0.0000022
0.009
61/61
1.31
±0.15
2.88
0.15
Benzene
0.0000078
0.03
61/61
0.49
±0.09
3.84
0.02
1.3 -Butadiene
0.00003
0.002
46/61
0.04
±0.01
1.17
0.02
Carbon Tetrachloride
0.000006
0.1
61/61
0.61
±0.02
3.67
0.01
1,2-Dichloroethane
0.000026
2.4
54/61
0.08
±0.01
1.97
<0.01
Formaldehyde
0.000013
0.0098
61/61
2.14
±0.34
27.82
0.22
Hexachloro-1,3 -butadiene
0.000022
0.09
14/61
0.02
±0.01
0.45
<0.01
— = A Cancer URE or Noncancer RfC is not available.
20-70

-------
Table 20-6. Risk Approximations for the New Jersey Monitoring Sites (Continued)
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Elizabeth, New Jersey - ELNJ
Acetaldehyde
0.0000022
0.009
61/61
2.60
±0.26
5.72
0.29
Benzene
0.0000078
0.03
61/61
0.81
±0.07
6.28
0.03
1.3 -Butadiene
0.00003
0.002
61/61
0.11
±0.01
3.42
0.06
Carbon Tetrachloride
0.000006
0.1
61/61
0.63
±0.02
3.76
0.01
1,2-Dichloroethane
0.000026
2.4
47/61
0.07
±0.01
1.93
<0.01
Ethylbenzene
0.0000025
1
61/61
0.43
±0.06
1.08
<0.01
Formaldehyde
0.000013
0.0098
61/61
4.90
±0.67
63.67
0.50
Hexachloro-1,3 -butadiene
0.000022
0.09
13/61
0.02
±0.01
0.36
<0.01
North Brunswick, New Jersey - NBNJ
Acetaldehyde
0.0000022
0.009
62/62
1.66
±0.15
3.64
0.18
Benzene
0.0000078
0.03
61/61
0.65
±0.05
5.08
0.02
1.3 -Butadiene
0.00003
0.002
58/61
0.07
±0.01
2.10
0.03
Carbon Tetrachloride
0.000006
0.1
61/61
0.62
±0.02
3.71
0.01
1,2-Dichloroethane
0.000026
2.4
59/61
0.09
±0.01
2.24
<0.01
Formaldehyde
0.000013
0.0098
62/62
2.24
±0.30
29.15
0.23
Hexachloro-1,3 -butadiene
0.000022
0.09
16/61
0.02
±0.01
0.53
<0.01
— = A Cancer URE or Noncancer RfC is not available.
Observations from Table 20-6 include the following:
• For CSNJ, the pollutants with the highest annual average concentrations are
formaldehyde, acetaldehyde, and benzene. Formaldehyde has the highest cancer risk
approximation for this site (64.54 in-a-million), followed by benzene and
acetaldehyde. The cancer risk approximation for formaldehyde is at least an order of
magnitude higher than the cancer risk approximations for the other pollutants of
interest for CSNJ. CSNJ's cancer risk approximation for formaldehyde is the highest
cancer risk approximation among the pollutants of interest for the New Jersey sites
and the fourth highest among all NMP sites. None of the pollutants of interest for
CSNJ have noncancer hazard approximations greater than 1.0, indicating that no
adverse noncancer health effects are expected from these individual pollutants.
20-71

-------
Formaldehyde is the pollutant with the highest noncancer hazard approximation for
CSNJ (0.51).
•	For CHNJ, the pollutants with the highest annual average concentrations are
formaldehyde, acetaldehyde, and carbon tetrachloride. Formaldehyde has the highest
cancer risk approximation for this site (27.82 in-a-million), followed by benzene and
carbon tetrachloride. The cancer risk approximation for formaldehyde is at least an
order of magnitude higher than the approximations for the other pollutants of interest
for CHNJ. None of the pollutants of interest for CHNJ have noncancer hazard
approximations greater than 1.0, indicating that no adverse noncancer health effects
are expected from these individual pollutants. Formaldehyde is the pollutant with the
highest noncancer hazard approximation for CHNJ (0.22).
•	For ELNJ, the pollutants with the highest annual average concentrations are
formaldehyde, acetaldehyde, and benzene. These three pollutants also have the
highest cancer risk approximations for this site, although the cancer risk
approximation for benzene is greater than the cancer risk approximation for
acetaldehyde. ELNJ's cancer risk approximation for formaldehyde (63.67 in-a-
million) is similar to the cancer risk approximation calculated for CSNJ and is the
fifth highest cancer risk approximation among all NMP sites. None of the pollutants
of interest for ELNJ have noncancer hazard approximations greater than 1.0,
indicating that no adverse noncancer health effects are expected from these individual
pollutants. Formaldehyde is the pollutant with the highest noncancer hazard
approximation for ELNJ (0.50).
•	For NBNJ, the pollutants with the highest annual average concentrations are
formaldehyde, acetaldehyde, and benzene. Formaldehyde has the highest cancer risk
approximation for NBNJ (29.15 in-a-million), 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 noncancer health effects
are expected from these individual pollutants. Formaldehyde is the pollutant with the
highest noncancer hazard approximation for NBNJ (0.23).
20-72

-------
20.5.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 20-7 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
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 3.4.3.4. Lastly,
Table 20-7 provides the pollutants with the highest cancer risk approximations (in-a-million) for
each New Jersey 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.
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 20.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
20-73

-------
Table 20-7. 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)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Camden, New Jersey (Camden County) - CSNJ
Benzene
130.73
Formaldehyde
1.24E-03
Formaldehyde
64.54
Formaldehyde
95.20
Benzene
1.02E-03
Benzene
6.65
Ethylbenzene
64.14
1,3-Butadiene
6.27E-04
Acetaldehyde
6.12
Acetaldehyde
55.11
Naphthalene
3.58E-04
Carbon Tetrachloride
3.65
1.3 -Butadiene
20.89
POM, Group 2b
2.36E-04
1,3-Butadiene
2.92
Tetrachloroethylene
11.72
Nickel, PM
2.02E-04
1,2-Dichloroethane
2.33
Naphthalene
10.52
POM, Group 2d
1.65E-04
Ethylbenzene
0.76
POM, Group 2b
2.68
Ethylbenzene
1.60E-04
Hexachloro-1,3 -butadiene
0.48
POM, Group 2d
1.87
Arsenic, PM
1.38E-04

Trichloroethylene
1.20
POM, Group 5a
1.23E-04
Chester, New Jersey (Morris County) - CHNJ
Benzene
161.55
Benzene
1.26E-03
Formaldehyde
27.82
Formaldehyde
95.57
Formaldehyde
1.24E-03
Benzene
3.84
Ethylbenzene
86.05
1,3-Butadiene
7.62E-04
Carbon Tetrachloride
3.67
Acetaldehyde
58.64
Naphthalene
3.43E-04
Acetaldehyde
2.88
1,3-Butadiene
25.41
Ethylbenzene
2.15E-04
1,2-Dichloroethane
1.97
Tetrachloroethylene
11.82
POM, Group 2b
2.05E-04
1,3-Butadiene
1.17
Naphthalene
10.09
Nickel, PM
1.97E-04
Hexachloro-1,3 -butadiene
0.45
Dichloro methane
5.27
POM, Group 2d
1.45E-04

POM, Group 2b
2.33
POM, Group 5a
1.31E-04
POM, Group 2d
1.64
Arsenic, PM
1.30E-04

-------
Table 20-7. 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)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Elizabeth, New Jersey (Union County) - ELNJ
Benzene
138.53
Formaldehyde
1.30E-03
Formaldehyde
63.67
Formaldehyde
99.64
Benzene
1.08E-03
Benzene
6.28
Ethylbenzene
74.03
1,3-Butadiene
6.29E-04
Acetaldehyde
5.72
Acetaldehyde
59.30
Nickel, PM
4.27E-04
Carbon Tetrachloride
3.76
1.3 -Butadiene
20.96
Naphthalene
3.75E-04
1,3-Butadiene
3.42
Tetrachloroethylene
14.36
Arsenic, PM
2.03E-04
1,2-Dichloroethane
1.93
Naphthalene
11.04
Ethylbenzene
1.85E-04
Ethylbenzene
1.08
Dichloro methane
2.96
POM, Group 2b
1.84E-04
Hexachloro-1,3 -butadiene
0.36
POM, Group 2b
2.09
Hexavalent Chromium
1.55E-04

Trichloroethylene
1.77
POM, Group 2d
1.33E-04
North Brunswick, New Jersey (Middlesex County) - NBNJ
Benzene
213.63
Formaldehyde
1.81E-03
Formaldehyde
29.15
Formaldehyde
139.48
Benzene
1.67E-03
Benzene
5.08
Ethylbenzene
110.60
1,3-Butadiene
9.59E-04
Carbon Tetrachloride
3.71
Acetaldehyde
83.83
Naphthalene
5.42E-04
Acetaldehyde
3.64
1.3 -Butadiene
31.96
Hydrazine
4.38E-04
1,2-Dichloroethane
2.24
Tetrachloroethylene
24.38
POM, Group 2b
2.82E-04
1,3-Butadiene
2.10
Naphthalene
15.95
Ethylbenzene
2.77E-04
Hexachloro-1,3 -butadiene
0.53
POM, Group 2b
3.20
POM, Group 2d
2.03E-04

Trichloroethylene
3.19
Arsenic, PM
1.86E-04
Dichloro methane
3.03
POM, Group 5a
1.85E-04

-------
Table 20-8. 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)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Camden, New Jersey (Camden County) - CSNJ
Toluene
422.02
Acrolein
598,846.68
Formaldehyde
0.51
Hexane
272.19
1,3-Butadiene
10,445.69
Acetaldehyde
0.31
Xylenes
249.52
Formaldehyde
9,713.82
Bromo methane
0.10
Benzene
130.73
Acetaldehyde
6,122.91
Propionaldehyde
0.07
Formaldehyde
95.20
Nickel, PM
4,680.35
1,3-Butadiene
0.05
Ethylbenzene
64.14
Benzene
4,357.53
Benzene
0.03
Acetaldehyde
55.11
Naphthalene
3,506.47
Carbon Tetrachloride
0.01
Methyl isobutyl ketone
32.86
Xylenes
2,495.17
Ethylbenzene
<0.01
Hydrochloric acid
29.17
Arsenic, PM
2,139.62
Hexachloro-1,3 -butadiene
<0.01
1.3 -Butadiene
20.89
Cadmium, PM
1,996.78
1,2-Dichloroethane
<0.01
Chester, New Jersey (Morris County) - CHNJ
Toluene
528.02
Acrolein
251,595.35
Formaldehyde
0.22
Xylenes
342.26
1,3-Butadiene
12,707.28
Acetaldehyde
0.15
Hexane
314.43
Formaldehyde
9,751.59
1,3-Butadiene
0.02
Benzene
161.55
Acetaldehyde
6,515.84
Benzene
0.02
Formaldehyde
95.57
Benzene
5,385.08
Carbon Tetrachloride
0.01
Ethylbenzene
86.05
Nickel, PM
4,561.47
Hexachloro-1,3 -butadiene
<0.01
Ethylene glycol
81.36
Xylenes
3,422.56
1,2-Dichloroethane
<0.01
Acetaldehyde
58.64
Naphthalene
3,363.00

Methyl isobutyl ketone
44.35
Lead, PM
2,402.90
Methanol
38.69
Arsenic, PM
2,017.68

-------
Table 20-8. 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)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Elizabeth, New Jersey (Union County) - ELNJ
Toluene
482.50
Acrolein
306,476.99
Formaldehyde
0.50
Hexane
351.57
Cyanide Compounds, PM
37,500.01
Acetaldehyde
0.29
Xylenes
279.98
1,3-Butadiene
10,478.70
1,3-Butadiene
0.06
Benzene
138.53
Formaldehyde
10,167.13
Benzene
0.03
Formaldehyde
99.64
Nickel, PM
9,894.58
Carbon Tetrachloride
0.01
Ethylbenzene
74.03
Acetaldehyde
6,588.93
Ethylbenzene
<0.01
Acetaldehyde
59.30
Benzene
4,617.66
Hexachloro-1,3 -butadiene
<0.01
Ethylene glycol
45.18
Chlorine
4,370.00
1,2-Dichloroethane
<0.01
Methyl isobutyl ketone
44.98
Naphthalene
3,678.63

Cyanide Compounds, PM
30.00
Lead, PM
3,167.18
North Brunswick, New Jersey (Middlesex County) - NBNJ
Toluene
721.66
Acrolein
424,778.44
Formaldehyde
0.23
Hexane
499.90
1,3-Butadiene
15,980.24
Acetaldehyde
0.18
Xylenes
432.41
Formaldehyde
14,232.86
1,3-Butadiene
0.03
Benzene
213.63
Acetaldehyde
9,314.11
Benzene
0.02
Formaldehyde
139.48
Benzene
7,120.91
Carbon Tetrachloride
0.01
Ethylbenzene
110.60
Naphthalene
5,317.45
Hexachloro-1,3 -butadiene
<0.01
Acetaldehyde
83.83
Lead, PM
5,099.62
1,2-Dichloroethane
<0.01
Methyl isobutyl ketone
58.79
Titanium tetrachloride
4,535.00

Ethylene glycol
35.26
Xylenes
4,324.07
1.3 -Butadiene
31.96
Arsenic, PM
2,886.75

-------
Observations from Table 20-7 include the following:
•	Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
cancer UREs in all four New Jersey counties.
•	Formaldehyde, benzene, and 1,3-butadiene are the pollutants with the highest
toxicity-weighted emissions (of the pollutants with cancer UREs) for all four New
Jersey counties, although the order is different for Morris County (CHNJ).
•	Six of the 10 highest emitted pollutants in Union and Middlesex Counties also have
the highest toxicity-weighted emissions. Seven of the highest emitted pollutants in
Camden and Morris Counties also have the highest toxicity-weighted emissions.
•	Formaldehyde, benzene, ethylbenzene, and 1,3-butadiene are among the pollutants
with the highest cancer risk approximations for CSNJ and also appear on both
emissions-based lists. Acetaldehyde is also among the pollutants with the highest
cancer risk approximations for CSNJ; this pollutant also appears among the highest
emitted pollutants in Camden County but does not appear among those with the
highest toxicity-weighted emissions. These observations are also true for ELNJ.
•	Formaldehyde, benzene, and 1,3-butadiene are among the pollutants with the highest
cancer risk approximations for CHNJ and also appear on both emissions-based lists.
Acetaldehyde is also among the pollutants with the highest cancer risk
approximations for CHNJ; this pollutant also appears among the highest emitted
pollutants in Morris County but does not appear among those with the highest
toxicity-weighted emissions. These observations are also true for NBNJ.
•	Carbon tetrachloride, 1,2-dichloroethane, and hexachloro-1,3-butadiene are among
the pollutants with the highest cancer risk approximations for each New Jersey site.
Yet these pollutants do not appear on either emissions-based list for any of the four
counties.
•	Arsenic and several POM Groups appear among the pollutants with the highest
toxicity-weighted emissions for each New Jersey county with an NMP site. Neither
speciated metals nor PAHs were sampled for under the NMP.
Observations from Table 20-8 include the following:
•	Toluene, hexane, and xylenes are the highest emitted pollutants with noncancer RfCs
in Camden, Union, and Middlesex Counties. In Morris County (CHNJ), toluene is
also the highest emitted pollutant, but the xylenes emissions are greater than the
hexane emissions.
•	Acrolein is the pollutant with the highest toxicity-weighted emissions (of the
pollutants with noncancer RfCs) for all four New Jersey counties but is not among the
highest emitted pollutants for any of the New Jersey counties (acrolein ranks between
11th and 17th for these counties). Although acrolein was sampled for at all four sites,
this pollutant was excluded from the pollutant of interest designation, and thus
20-78

-------
subsequent risk-based screening evaluations, due to questions about the consistency
and reliability of the measurements, as discussed in Section 3.2. 1,3-Butadiene and
formaldehyde are the pollutants with the second and third highest toxicity-weighted
emissions in three of the four counties. For Union County (ELNJ), cyanide
compounds rank higher than 1,3-butadiene and formaldehyde for this county's
toxicity-weighted emissions.
•	Between four and five of the 10 highest emitted pollutants also have the highest
toxicity-weighted emissions for each of the New Jersey counties.
•	Formaldehyde, acetaldehyde, and benzene are pollutants of interest for all four New
Jersey sites and appear on both emissions-based lists for their respective counties.
1,3-Butadiene, another pollutant of interest for all four sites, appears among those
with the highest toxicity-weighted emissions for all four counties, but does not rank
among the highest emitted in Morris County (CHNJ) or Union County (ELNJ).
Ethylbenzene, a pollutant of interest for CSNJ and ELNJ, appears among the highest
emitted pollutants (with a noncancer RfC) but not among those with the highest
toxicity-weighted emissions.
•	Carbon tetrachloride, 1,2-dichloroethane, and hexachloro-1,3-butadiene are among
the pollutants with the highest noncancer hazard approximations for each site. Yet
these pollutants do not appear on either emissions-based list for any of the New
Jersey counties. Bromomethane and propionaldehyde, pollutants of interest for CSNJ,
appear on neither emissions-based list for Camden County.
•	Several speciated metals and naphthalene appear among the pollutants with the
highest toxicity-weighted emissions for each New Jersey county with an NMP site.
Neither speciated metals nor PAHs were sampled for under the NMP.
20.6 Summary of the 2013 Monitoring Data for the New Jersey Monitoring Sites
Results from several of the data treatments described in this section include the
following:
~~~ Sixteen pollutants failed at least one screen for CSNJ; nine failed screens for CHNJ;
11 failed screens for ELNJ; and 11 failed screens for NBNJ.
~~~ Formaldehyde and acetaldehyde had the highest annual average concentrations for
each of the New Jersey sites.
~~~ NBNJ has the third highest annual average concentration of
hexachloro-1,3-butadiene among NMP sites sampling VOCs, with the annual
averages for CSNJ and CHNJ ranking fifth and sixth, respectively. CSNJ has the
third highest annual average concentrations of both acetaldehyde andformaldehyde
among NMP sites sampling carbonyl compounds.
20-79

-------
ELNJ is the longest running NMP site still participating in the program.
Concentrations of benzene have decreased significantly at this site since the onset of
sampling. This is also true of ethylbenzene, although concentrations have leveled out
in the last few years. The detection rates of 1,2-dichloroethane and hexachloro-1,3-
butadiene at CHNJ, ELNJ, and NBNJ have been increasing steadily over the last few
years of sampling.
Formaldehyde has the highest cancer risk approximations of the pollutants of interest
for each of the New Jersey sites. None of the site-specific pollutants of interest have
noncancer hazard approximations greater than an HQ of 1.0.
20-80

-------
21.0	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.
21.1	Site Characterization
This section characterizes the New York monitoring sites by providing geographical and
physical information about the locations of the sites and the surrounding areas. This information
is provided to give the reader insight regarding factors that may influence the air quality near the
sites and assist in the interpretation of the ambient monitoring measurements.
One New York monitoring site is located in New York City (BXNY) and one is located
in Rochester (ROCH). Figure 21-1 is a composite satellite image retrieved from ArcGIS
Explorer showing the New York City monitoring site and its immediate surroundings.
Figure 21-2 identifies nearby point source emissions locations by source category, as reported in
the 2011 NEI for point sources, version 2. Note that only sources within 10 miles of BXNY are
included in the facility counts provided in Figure 21-2. A 10-mile boundary was chosen to give
the reader an indication of which emissions sources and emissions source categories could
potentially have a direct effect on the air quality at the monitoring site. Further, this boundary
provides both the proximity of emissions sources to the monitoring site as well as the quantity of
such sources within a given distance of the site. Sources outside the 10-mile boundary are still
visible on the map for reference, but have been grayed out in order to emphasize emissions
sources just within the boundary. Figures 21-3 and 21-4 are the composite satellite image and
emissions sources map for ROCH. Table 21-1 provides supplemental geographical information
such as land use, location setting, and locational coordinates.
21-1

-------
Figure 21-1. New York City, New York (BXNY) Monitoring Site

-------
Figure 21-2. NEI Point Sources Located Within 10 Miles of BXNY

n*45Tmv
Westchester
\ County
iong
Wwd
SOU no1
O BfOfl*
County
Nassau
County
£§stR»f
Queens
County
New Yqi^ Co^nt ,
Kngs
County
n utttw
Legend
wwv»	trifrw	nww	ry^nv
Note Dub h> tocJHy density and venocefcon ihe total fadtotea
displayed may not represent all facias writMn tne area of ntercit
~>V BXNY NATTS site	10 mile radius [ ] County boundary
Source Category Group (No. of Facilities)
T	AirportAirline/Airport Support Operations (26)
B	Bulk TerminalaiBulk Plants (?)
C	Chemical Manufacturing Facility <4|
J	Compressor Station \2)
*	Electricity Generation vta Combustion (IS)
E	Electroplating Plating Po«ish	Hotete/Motels'Lodgmg (1)
O	institutional (school, hospital prison, etc ) (27\
A	Metal Coaling Engraving and Allied Services to Manulaclviets (2)

-------
Figure 21-3. Rochester, Mew York (ROCH) Monitoring Site
-Blossomed!'
Blossom Rd
I	056s /
Ci K»5A NGA.'uSCS
to
¦k

-------
Figure 21-4. NET Point Sources Located Within 10 Miles of ROCH
¦CTTVf
VWyrc
County
Monro*
Coonty
Ontario
County
Note: Quo 10 tacMty don sir, and collocation tho local fiadHHta
t	.	dupayixJ -raj1 rot feprtkftrri all facAUJh WTthm (ha a-ea of rrtara«l
Legend
ROCH NATTS site	10 mile radius [ County boundary
Source Category Group {No. of Facilities)
+
Airport/Airline/Airport Support Operations (6)
a
Landfll (1)
B
Bulk Terminals/Bulk Plants (4)
®
Metals Processing/Fabrication Facility (2)
C
Chemical Manufacturing Facility (4)
¦»
Miscellaneous Commeraatflndustnal Facility (3)
e
Electrical Equipment Manufacturing Facility (2)
Cr
Pharmaceutical Manufacturing (1)
Si
Glass Plant (1)
R
Plastic, Resin, or Rubber Products Plant (1)
•
Industrial Machinery or Equipment Plan! (1)
P
Printmg.'Publish'ng/Paper Product Manufacturing Facility (3)
21-5

-------
Table 21-1. Geographical Information for the New York Monitoring Sites
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Additional Ambient Monitoring Information1
BXNY
36-005-0110
New York
Bronx
New York-Newark-
Jersey City,
NY-NJ-PA
40.81618,
-73.902
Residential
Urban/City
Center
Haze, S02, NO, N02, NOx, VOCs, Carbonyl
compounds, O3, Meteorological Parameters, PM
coarse. Black Carbon PM10, PM10 Speciation, PM2.5,
PM2.5 Speciation, IMPROVE Speciation.
ROCH
36-055-1007
Rochester
Monroe
Rochester, NY
43.14618,
-77.54817
Residential
Urban/City
Center
CO, SO2, NO, NO2, NOy, VOCs, Carbonyl
compounds, O3, Meteorological parameters. Black
Carbon, PM10. PM10 Speciation PM2.5, PM2.5
Speciation, IMPROVE Speciation.
1 Data for additional pollutants are reported to AQS for these sites (EPA, 2014b); however, these data are not generated by ERG and are therefore not included in this report.
BOLD ITALICS = EPA-designated NATTS Site

-------
BXNY is located on the property of Public School 52 (PS 52) in the Bronx Borough of
New York City, northeast of Manhattan. The site was established in 1999 and is considered one
of the premier particulate sampling sites in New York City and is the Bronx (#1) NATTS site.
The surrounding area is urban and residential, as shown in Figure 21-1. 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 21-1. BXNY is less than one-half mile from the East River at its closest point.
Figure 21-2 shows the numerous point sources that are located within 10 miles of BXNY,
with a majority of the emissions sources located to the south and west of the site. The source
categories with the greatest number of emissions sources surrounding the site include institutions
such as hospitals, schools, and prisons; airport and airport support operations, which include
airports and related operations as well as small runways and heliports, such as those associated
with hospitals or television stations; electricity generation via combustion; and printing,
publishing, and paper product manufacturing. The point source closest to BXNY is a compressor
station.
ROCH is located at a power substation on the east side of Rochester, in western New
York. Rochester is approximately halfway between Syracuse and Buffalo, with Lake
Ontario situated to the north. Although the area north and west of the site is primarily residential,
as shown in Figure 21-3, 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 21-4 shows, the relatively few point sources within 10 miles of ROCH
are located primarily on the west side of the 10-mile boundary. The airport and airport support
operations source category is the source category with the greatest number of emissions sources
surrounding ROCH, although there are also bulk plants/bulk terminals, chemical manufacturers,
metals processors/fabricators, and printing, publishing, and paper product manufacturers nearby,
to name a few. The closest source to ROCH is an electrical equipment manufacturer.
21-7

-------
Table 21-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the New York monitoring sites. Table 21-2 includes both county-level
population and vehicle registration information. Table 21-2 also contains traffic volume
information for each site as well as the location for which the traffic volume was obtained.
Additionally, Table 21-2 presents the county-level daily VMT for Bronx and Monroe Counties
from the 2011 NEI.
Table 21-2. Population, Motor Vehicle, and Traffic Information for the New York
Monitoring Sites
Site
County
Estimated
County
Population1
County-level
Vehicle
Registration2
Annual
Average
Daily Traffic3
Intersection
Used for Traffic Data
County-level
Daily VMT4
BXNY
Bronx
1,418,733
254,752
98,899
1-278 between 1-87 & 1-895
8,170,256
ROCH
Monroe
749,606
558,063
85,162
1-490 at 1-590
15,963,343
bounty-level population estimates reflect 2013 data (Census Bureau, 2014)
2County-level vehicle registrations reflect 2013 data (NYS DMV, 2013)
3AADT reflects 2012 data (NYS DOT, 2012)
4County-level VMT reflects 2011 data (EPA, 2015a)
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 21-2 include the following:
•	Bronx County has the ninth highest county-level population among counties with
NMP sites. The population of Rochester County is roughly half the Bronx County
population and ranks 20th among NMP sites.
•	County-level vehicle ownership for Bronx County ranks 32nd 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 19th 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.
•	Traffic volume is higher near BXNY, which ranks 15th among NMP sites, than
ROCH, which ranks 17th, although their rankings are similar. The traffic data for
BXNY is for 1-278 between 1-87 and 1-895; the traffic data for ROCH are provided
for 1-490 at 1-590.
21-8

-------
• 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.
21.2 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.
21.2.1 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 monthly snow accumulations generally range from 3 inches to 10 inches. The
proximity to the Atlantic Ocean offers a moderating influence from cold air 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 of the year (Wood, 2004; NCDC, 2015).
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. Lake Ontario acts as a moderating influence on the city's temperatures, both in the
summer and the winter, as the lake 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.
Ninety inches of snow can fall in the city during the average winter. 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).
21-9

-------
21.2.2 Meteorological Summary
Hourly meteorological data for 2013 were retrieved from NCDC for the weather stations
closest to the New York monitoring sites (NCDC, 2013), as described in Section 3.4.2. The
closest weather station to 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 21-3. These data were used to determine how
meteorological conditions on sample days vary from conditions experienced throughout the year.
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 2013. Also included in Table 21-3 is the 95 percent
confidence interval for each parameter. Average meteorological conditions on sample days near
BXNY and ROCH were representative of average weather conditions experienced throughout
the year at each location. As expected, Table 21-3 shows that temperatures were cooler in
western New York than in New York City. BXNY is among the windier locations with an NMP
site, based on the 2013 average scalar wind speed.
21-10

-------
Table 21-3. 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)
2.80
miles
156°
(SSE)
Sample
Davs
(61)
62.0
±4.8
55.5
±4.6
40.8
±4.8
48.6
±4.2
59.9
±3.3
1017.7
± 1.7
9.3
±0.9
2013
61.8
+ 1.9
55.6
+ 1.8
40.5
+ 1.9
48.5
+ 1.6
59.4
+ 1.4
1017.3
±0.7
9.1
±0.3
Rochester, New York - ROCH
Greater
Rochester Intl.
Airport
14768
(43.12, -77.68)
6.84
miles
253°
(WSW)
Sample
Davs
(61)
57.0
±5.5
49.2
±4.9
39.2
±4.9
44.6
±4.6
70.8
±3.1
1017.5
± 1.7
7.5
±0.9
2013
57.2
±2.1
49.2
+ 1.9
39.1
+ 1.9
44.5
+ 1.8
70.6
+ 1.1
1017.2
±0.7
7.4
±0.3
1 Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.

-------
21.2.3 Wind Rose Comparison
Hourly surface wind data from the weather stations at La Guardia Airport (for BXNY)
and Greater Rochester International Airport (for ROCH) were uploaded into a wind rose
software program to produce customized wind roses, as described in Section 3.4.2. A wind rose
shows the frequency of wind directions using "petals" positioned around a 16-point compass,
and uses different colors to represent wind speeds.
Figure 21-5 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 21-5 also presents three different wind roses for the
BXNY monitoring site. First, a historical wind rose representing 2003 to 2012 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 2013 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2013 is
presented. These can be used to identify the predominant wind speed and direction for 2013 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically. Figure 21-6 presents the distance map and wind roses for
ROCH.
21-12

-------
Figure 21-5. Wind Roses for the La Guardia
Location of BXNY and Weather Station
Airport Weather Station near BXNY
2003-2012 Historical Wind Rose
WIND SPEED
(Knots)
I I =-22
[I~| 17-21
11 - 17
CT 7- 11
H 4-7
2- 4
Calms: 4.35%
Sample Day Wind Rose
WIND SPEED
(Knots)
~
7- 11
4- 7
2- 4
Calms: 4.64%
~
2013 Wind Rose
WIND SPEED
(Knots)
I I >=22
F~1 17-21
| 11 -17
~
4- 7
2- 4
Calms: 4.26%
21-13

-------
Figure 21-6. Wind Roses for the Greater Rochester International Airport Weather Station
near ROCH
Location of ROCH and Weather Station
2003-2012 Historical Wind Rose


ROCH
/



SMfOO




,4
INORTH"^-
i EAS
ES :
WHO SPEED
(Knots)
17-21
11 - 17
SOUTH
Calms asm
2013 Wind Rose
NORTH"---
W NC S PE EC
(Kn ots)
17 - 21
11 - 17
SOUTH
7- 1'
Calms: 13.85%
Sample Day Wind Rose
NORTH
WIND SPEED
(Knots)
17-21
SOUTH
Calms: 11.21%
21-14

-------
Observations from Figure 21-5 for BXNY include the following:
•	The weather station at La Guardia Airport is located 2.8 miles southeast of BXNY.
The East River and Rikers Island separate the site and the weather station.
•	The historical wind rose shows that winds from a variety of directions are observed
near BXNY, although winds from the southeast quadrant were rarely observed.
Winds from the west to northwest to north account for 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 (those less than or equal to 2 knots) were observed for less than 5 percent
of the hourly measurements near BXNY.
•	The full-year wind rose for 2013 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.
•	The sample day wind patterns resemble the wind patterns on the other wind roses in
that northwesterly and southerly winds prevail, although these directions account for
a higher percentage of winds on samples days (17 percent and 15 percent,
respectively) compared to the historical and full-year wind roses. Fewer northerly,
northeasterly, and east-northeasterly winds were observed on sample days.
Observations from Figure 21-6 for ROCH include the following:
•	The Greater Rochester International Airport weather station is located 6.8 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.
Winds from most other directions individually count for less than 4 percent of
observations each. Calm winds were observed for less than 10 percent of the hourly
measurements near ROCH, while the strongest winds were most frequently observed
with west-southwesterly and westerly winds.
•	The wind patterns shown on the 2013 wind rose are similar to the historical wind
patterns for ROCH, although westerly winds account for an even higher percentage of
wind observations in 2013 (nearly 17 percent).
•	The sample day wind patterns are similar to those shown on the full-year and
historical wind roses, although the percentage of calm winds is slightly higher.
21-15

-------
21.3 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 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. Hexavalent chromium and PAHs were sampled for at both New York
sites although hexavalent chromium sampling was discontinued in June 2013 at BXNY and July
2013 atROCH.
Table 21-4. Risk-Based Screening Results for the New York Monitoring Sites
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
New York City, New York - BXNY
Naphthalene
0.029
60
60
100.00
68.97
68.97
Fluorene
0.011
12
57
21.05
13.79
82.76
Acenaphthene
0.011
11
60
18.33
12.64
95.40
Fluoranthene
0.011
2
60
3.33
2.30
97.70
Benzo(a)pyrene
0.00057
1
58
1.72
1.15
98.85
Hexavalent Chromium
0.000083
1
19
5.26
1.15
100.00
Total
87
314
27.71

Rochester, New York - ROCH
Naphthalene
0.029
39
56
69.64
38.24
38.24
Acenaphthene
0.011
28
56
50.00
27.45
65.69
Fluorene
0.011
25
54
46.30
24.51
90.20
Fluoranthene
0.011
10
56
17.86
9.80
100.00
Total
102
222
45.95

Observations from Table 21-4 include the following:
• Six pollutants failed screens for BXNY; 28 percent of concentrations for these six
pollutants were greater than their associated risk screening value (or failed screens).
21-16

-------
•	Three pollutants, naphthalene, fluorene, and acenaphthene, were identified as
pollutants of interest for BXNY.
•	Four pollutants failed screens for ROCH; 46 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 for ROCH;
therefore, all four were identified as pollutants of interest for this site.
•	For both sites, naphthalene, acenaphthene, and fluorene were identified as pollutants
of interest. Naphthalene failed the majority of screens for each site, accounting for
69 percent of failed screens for BXNY and 38 percent of failed screens for ROCH.
21.4 Concentrations
This section presents various concentration averages used to characterize air toxics
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. Site-specific statistical summaries for all pollutants
sampled for at BXNY and ROCH are provided in Appendices M and O.
21.4.1 2013 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
21-17

-------
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 York monitoring sites 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 New York Monitoring Sites

# of






Measured
1st
2nd
3rd
4th


Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs. # of
Samples
Average
(ng/m3)
Average
(ng/m3)
Average
(ng/m3)
Average
(ng/m3)
Average
(ng/m3)

New York City,
New York - BXNY




2.45
8.05
11.04
4.19
6.46
Acenaphthene
60/60
±0.55
± 1.93
± 1.85
± 1.65
± 1.15


3.34
8.95
11.02
4.63
7.01
Fluorene
57/60
±0.92
±2.22
± 1.89
± 1.51
± 1.14


108.15
136.59
137.49
123.81
126.77
Naphthalene
60/60
± 27.24
±28.35
±23.66
±25.23
± 12.63
Rochester, New York - ROCH


4.07
40.71
32.43
6.56
19.37
Acenaphthene
56/56
±3.43
±9.51
± 10.54
±4.27
±5.35


1.37
8.42
10.30
2.27
5.18
Fluoranthene
56/56
±0.36
±2.92
±2.72
±0.82
± 1.33


2.73
27.22
23.21
4.82
13.40
Fluorene
54/56
± 1.48
±6.82
±7.36
±2.80
±3.65


36.71
101.06
85.51
37.03
62.20
Naphthalene
56/56
±8.41
±28.92
± 18.63
± 10.89
± 11.15
Observations for BXNY from Table 21-5 include the following:
•	Acenaphthene and naphthalene were detected in all of the valid PAH samples
collected at BXNY, while three non-detects of fluorene were measured.
•	Of the pollutants of interest for BXNY, naphthalene has the highest annual average
concentration, while the annual averages for acenaphthene and fluorene are similar to
each other. Concentrations of naphthalene measured at BXNY range from 39.7 ng/m3
to 231 ng/m3. Concentrations of acenaphthene range from 1.11 ng/m3 to 19.1 ng/m3
while concentrations of fluorene range from 1.94 ng/m3 to 18.7 ng/m3 plus three non-
detects.
•	Concentrations of acenaphthene and fluorene are significantly higher during the
warmer months than the cooler months, based on the quarterly average
concentrations. Concentrations measured during the second and third quarters of 2013
also exhibit more variability, based on the confidence intervals shown. A similar
21-18

-------
observation can be made for naphthalene although the confidence intervals indicate
the differences are not statistically significant.
Observations for ROCH from Table 21-5 include the following:
•	Acenaphthene, fluoranthene, and naphthalene were detected in all of the valid PAH
samples collected at ROCH, while two non-detects of fluorene were measured.
•	Of the pollutants of interest for ROCH, naphthalene has the highest annual average
concentration, followed by acenaphthene, fluorene, and then fluoranthene.
•	Concentrations of naphthalene measured at ROCH range from 13.4 ng/m3 to
216 ng/m3. Concentrations of acenaphthene range from 0.383 ng/m3 to 69.7 ng/m3;
concentrations of fluoranthene range from 0.483 ng/m3 to 19.1 ng/m3; and
concentrations of fluorene range from 0.631 ng/m3 to 53.4 ng/m3 plus two non-
detects.
•	Quarterly average concentrations of each of the pollutants of interest for ROCH were
considerably higher during the second and third quarters of the year. For example, all
nine acenaphthene concentrations less than 1 ng/m3 were measured during the first or
fourth quarters of 2013 while all but two of the 20 concentrations greater than
25 ng/m3 were measured during the second and third quarters of 2013.
Tables 4-9 through 4-12 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for BXNY and
ROCH from those tables include the following:
•	ROCH and BXNY have the second and fifth highest annual average concentrations of
acenaphthene among NMP sites sampling PAHs, as shown in Table 4-11.
•	The annual average concentration of naphthalene for BXNY ranks third compared to
other NMP sites sampling PAHs while the annual average concentration for ROCH
does not appear in Table 4-11 for naphthalene (it ranks 14th).
21.4.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created each of the pollutants of
interest for BXNY and ROCH. Figures 21-7 through 21-10 overlay the site's minimum, annual
average, and maximum concentrations onto the program-level minimum, first quartile, median,
average, third quartile, and maximum concentrations for each pollutant, as described in
Section 3.4.3.1.
21-19

-------
Figure 21-7. Program vs. Site-Specific Average Acenaphthene Concentrations
Program Max Concentration = 123 ng/m3
Program Max Concentration = 123 ng/m3
0
10
20
30 40
Concentration {ng/m3)
50 60
70

Program:
1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i

Site:
Site Average
o
Site Concentration Range


Figure 21-8. Program vs. Site-Specific Average Fluoranthene Concentration
20	25	30
Concentration {ng/m3)
Program: 1st Quartile
2nd Quartile 3rd Quartile
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



21-20

-------
Figure 21-9. Program vs. Site-Specific Average Fluorene Concentrations
BXNY
ROCH
0	10	20	30	40	50	60	70	80	90	100
Concentration (ng/m3)
Program: IstQuartile
¦
2ndQuartile 3rdQuartile
~ ~
4thQuartile
~
Average
1
Site: Site Average
o
Site Concentration Range


Figure 21-10. Program vs. Site-Specific Average Naphthalene Concentrations
H
400	500
Concentration (ng/m3)
Program: IstQuartile
¦
2ndQuartile 3rdQuartile
~ ~
4thQuartile
~
Average
1
Site: Site Average
o
Site Concentration Range


Observations from Figures 21-7 through 21-10 include the following:
• Figure 21-7 presents the box plots for acenaphthene for both sites. Note that the
program-level maximum concentration (123 ng/m3) is not shown directly on the
box plots because the scale of the box plots would be too large to readily observe
data points at the lower end of the concentration range. Thus, the scale has been
reduced to 80 ng/m3. The box plots show that although the maximum
acenaphthene concentration measured across the program was not measured at
either New York site, the maximum concentration measured at ROCH is among
the higher concentrations. The entire range of acenaphthene concentrations
measured at BXNY is less than the annual average concentration for ROCH. The
annual average concentrations for both sites are greater than the program-level
21-21

-------
average concentration, although the annual average for ROCH is three times
greater the annual average concentration for BXNY. Recall that ROCH has the
second highest annual average concentration of acenaphthene among NMP sites
sampling PAHs (behind NBIL).
•	Figure 21-8 presents the box plot for fluoranthene for ROCH, the only New York
site for which this is a pollutant of interest. Although the maximum concentration
of fluoranthene measured at ROCH is less than the maximum concentration
measured across the program, it is the seventh highest concentration measured
among NMP sites sampling PAHs. The annual average concentration for ROCH
is more than three times greater the program-level average concentration. This site
is one of only three NMP sites sampling PAHs with fluoranthene as a pollutant of
interest.
•	Figure 21-9 presents the box plots for fluorene for both New York sites. The
maximum concentration of fluorene measured at ROCH is considerably greater
than the maximum concentration measured at BXNY but is roughly half the
maximum concentration measured across the program (although the maximum
concentration measured at ROCH is among the higher measurements). The annual
average concentrations for both sites are greater than the program-level average,
although the annual average for BXNY is roughly half the annual average
concentration for ROCH.
•	Figure 21-10 presents the box plots for naphthalene for both sites. In contrast to
the box plots for the other pollutants of interest in common for the New York
sites, Figure 21-10 shows that the naphthalene concentrations measured at ROCH
are less than the ones measured at BXNY. The annual average naphthalene
concentration for ROCH is half the annual average for BXNY and is less than the
program-level average concentration. The annual average concentration for
BXNY is greater than the program-level average and third quartile. Recall that
BXNY has the third highest annual average concentration of naphthalene among
NMP sites sampling PAHs, even though the range of measurements is not that
large. The minimum naphthalene concentration measured at BXNY is greater
than the program-level first quartile. The minimum naphthalene concentration
measured at BXNY is the highest minimum concentration of this pollutant
measured at an NMP site.
21.4.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.3.2. In
June 2010, the monitoring instruments at BXNY were relocated to a new, temporary location due
to roofing construction near the BXNY site. Two years later, the instrumentation was returned to
the BXNY site and sampling resumed at this location in July 2012. A trends analysis was not
performed for BXNY because sampling did not occur consecutively at the same location.
21-22

-------
Sampling for PAHs at ROCH began in July 2008, so a trends analysis was performed for
ROCH. However, due to the mid-year start, a 1-year average concentration for 2008 is not
presented, although the range of measurements is provided. In addition, a collection error was
discovered at the site, resulting in the invalidation of nearly one and one-half years' worth of
samples between July 2009 and December 2010. Thus, the range of measurements is provided
for 2009, although a 1-year average concentration is not provided and no statistical metrics are
provided for 2010. This, combined with the mid-year start in 2008, results in the calculation of
few 1-year average concentrations for the ROCH monitoring site.
Figure 21-11. Yearly Statistical Metrics for Acenaphthene Concentrations Measured at ROCH
2010 z	2011
Year
O 5th Percentile
- Minimum
- Maximum
O 95th Percentile
1	A 1-year average is not presented because sampling under the NMP did not begin until July 2008.
2	Some statistical metrics are not presented because data from July 2009 to Dec 2010 was invalidated.
Observations from Figure 21-11 for acenaphthene measurements collected at ROCH
include the following:
•	The range of acenaphthene concentrations appears to have decreased by half from 2008
to 2009, although 2008 includes data from July through December while 2009 includes
data from January through June.
•	The measurements collected in 2011 are similar to the measurements collected in 2012.
21-23

-------
• The range of concentrations increased considerably from 2012 to 2013. The median
concentration nearly doubled from 2012 to 2013 while the 1-year average concentration
increased by 58 percent.
Figure 21-12. Yearly Statistical Metrics for Fluoranthene Concentrations Measured at ROCH
0 -I—
2008 1
2009 2
2010 2
Year
2011
2012
2013

O 5th Percentile
- Minimum
— Median -
Maximum
O 95th Percentile

1	A 1-year average is not presented because sampling under the NMP did not begin until July 2008.
2	Some statistical metrics are not presented because data from July 2009 to Dec 2010 was invalidated.
Observations from Figure 21-12 for fluoranthene measurements collected at ROCH
include the following:
• With the exception of the 95th percentile, each of the statistical parameters exhibits a
decrease from 2008 to 2009, although 2008 includes data from July through December
while 2009 includes data from January through June.
The median concentration decreased considerably from 2008 to 2009, after which little
change is shown. Between 2009 and 2013, the median concentration varied by less than
0.35 ng/m3, ranging from 2.66 ng/m3 (2009) to 2.99 ng/m3 (2012). Similarly, the 1-year
average concentrations have changed little, with 0.50 ng/m3 separating them, even though
the range of concentrations measured at ROCH has decreased each year since 2011.
21-24

-------
Figure 21-13. Yearly Statistical Metrics for Fluorene Concentrations Measured at ROCH
50
40
£
20081	20092	20102	2011	2012	2013
Year
O 5th Percentile	- Minimum	— Median	- Maximum	O 95th Percentile •~•^'~~•Average
1	A 1-year average is not presented because sampling under the NMP did not begin until July 2008.
2	Some statistical metrics are not presented because data from July 2009 to Dec 2010 was invalidated.
Observations from Figure 21-13 for fluorene measurements collected at ROCH include
the following:
•	The trends graph for fluorene resembles the trends graph for acenaphthene.
•	The range of fluorene concentrations appears to have decreased from 2008 to 2009 and
the median concentration decreased by more than half during this time frame, although
2008 includes data from July through December while 2009 includes data from January
through June.
•	The measurements collected in 2011 are similar to the measurements collected in 2012.
•	The range of concentrations increased from 2012 to 2013, when the maximum fluorene
concentration (53.4 ng/m3) since the onset of sampling at ROCH was measured. The
median increased by 67 percent from 2012 to 2013 while the 1-year average
concentration increased by about half that percentage.
21-25

-------
Figure 20-14. Yearly Statistical Metrics for Naphthalene Concentrations Measured at ROCH
20102	2011
Year
0 5th Percentile	- Minimurr
~ Maximurr
O 95th Percentile
1	A 1-year average is not presented because sampling under the NMP did not begin until July 2008.
2	Some statistical metrics are not presented because data from July 2009 to Dec 2010 was invalidated.
Observations from Figure 20-14 for naphthalene measurements collected at ROCH
include the following:
•	Similar to the other pollutants of interest, the range of naphthalene concentrations appears
to have decreased from 2008 to 2009.
•	Even though the maximum concentration has increased each year since 2011, the 1-year
average naphthalene concentrations calculated for 2011, 2012, and 2013 exhibit little
change, varying by less than 1 ng/m3 across the time period.
21.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at the New York monitoring sites. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4
for definitions and explanations regarding the various toxicity factors, time frames, and
calculations associated with these risk-based screenings.
21-26

-------
21.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the New York sites and where annual average
concentrations could be calculated, risk was examined by calculating cancer risk and noncancer
hazard approximations. These approximations can be used as risk estimates for cancer and
noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers want to shift their air
monitoring priorities. Refer to Section 3.4.3.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 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 New York Monitoring Sites
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
New York City, New York - BXNY
Acenaphthene
0.000088

60/60
6.46
± 1.15
0.57

Fluorene
0.000088

57/60
7.01
± 1.14
0.62

Naphthalene
0.000034
0.003
60/60
126.77
± 12.63
4.31
0.04
Rochester, New York - ROCH
Acenaphthene
0.000088

56/56
19.37
±5.35
1.70

Fluoranthene
0.000088

56/56
5.18
± 1.33
0.46

Fluorene
0.000088

54/56
13.40
±3.65
1.18

Naphthalene
0.000034
0.003
56/56
62.20
± 11.15
2.11
0.02
— = A Cancer URE or Noncancer RfC is not available.
Observations for the New York sites from Table 21-6 include the following:
• Naphthalene has the highest annual average concentration among the pollutants of
interest for each site, although the annual average concentration for BXNY is
significantly higher than the annual average for ROCH.
21-27

-------
•	Naphthalene also has the highest cancer risk approximation for each site
(4.31 in-a-million for BXNY and 2.11 in-a-million for ROCH). The cancer risk
approximations for the other pollutants of interest for each site are all less than
2 in-a-million.
•	Only naphthalene has a noncancer RfC. The noncancer hazard approximations for
naphthalene for each site are both less than 0.05, considerably less than 1.0,
indicating that no adverse noncancer health effects are expected from this individual
pollutant.
21.5.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 21-7 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
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 3.4.3.4. Lastly,
Table 21-7 provides the pollutants with the highest cancer risk approximations (in-a-million) for
each site, 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 each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 21.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
21-28

-------
Table 21-7. 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)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
New York City, New York (Bronx County) - BXNY
Benzene
127.66
Formaldehyde
1.04E-03
Naphthalene
4.31
Ethylbenzene
92.28
Benzene
9.96E-04
Fluorene
0.62
Tetrachloroethylene
81.66
1,3-Butadiene
4.42E-04
Acenaphthene
0.57
Formaldehyde
80.26
Naphthalene
2.76E-04

Acetaldehyde
47.43
Ethylbenzene
2.31E-04
1.3 -Butadiene
14.74
Arsenic, PM
2.19E-04
Naphthalene
8.11
POM, Group 2b
1.53E-04
POM, Group 2b
1.74
Nickel, PM
1.40E-04
POM, Group 2d
1.53
POM, Group 2d
1.35E-04
Trichloroethylene
1.05
POM, Group 5a
1.11E-04
Rochester, New York (Monroe County) - ROCH
Benzene
257.25
Formaldehyde
2.24E-03
Naphthalene
2.11
Formaldehyde
172.39
Benzene
2.01E-03
Acenaphthene
1.70
Ethylbenzene
140.93
1,3-Butadiene
1.24E-03
Fluorene
1.18
Acetaldehyde
98.59
Naphthalene
6.88E-04
Fluoranthene
0.46
Dichloro methane
46.10
POM, Group 2b
5.08E-04

1,3-Butadiene
41.31
Arsenic, PM
3.81E-04
Tetrachloroethylene
24.16
Ethylbenzene
3.52E-04
Naphthalene
20.23
POM, Group 2d
3.16E-04
Trichloroethylene
6.40
Hexavalent Chromium
2.69E-04
POM, Group 2b
5.77
POM, Group 5a
2.69E-04

-------
Table 21-8. 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)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer
Hazard
Approximation
(HQ)
New York City, New York (Bronx County) - BXNY
Toluene
2,161.13
Acrolein
203,787.77
Naphthalene
0.04
Methanol
793.11
Formaldehyde
8,190.17

Hexane
479.04
1,3-Butadiene
7,368.79
Xylenes
293.26
Acetaldehyde
5,270.29
Ethylene glycol
275.15
Benzene
4,255.29
Benzene
127.66
Cadmium, PM
3,946.26
Ethylbenzene
92.28
Arsenic, PM
3,399.31
Tetrachloroethylene
81.66
Nickel, PM
3,238.72
Formaldehyde
80.26
Xylenes
2,932.64
Methyl isobutyl ketone
63.81
Naphthalene
2,702.28
Rochester, New York (Monroe County) - ROCH
Toluene
1,679.94
Acrolein
492,322.38
Naphthalene
0.02
Methanol
510.18
1,3-Butadiene
20,653.74

Xylenes
507.26
Formaldehyde
17,591.01
Hexane
498.21
Acetaldehyde
10,954.81
Benzene
257.25
Hydrochloric acid
10,479.37
Hydrochloric acid
209.59
Cadmium, PM
9,067.59
Formaldehyde
172.39
Benzene
8,575.15
Ethylene glycol
149.53
Naphthalene
6,742.12
Ethylbenzene
140.93
Arsenic, PM
5,913.63
Acetaldehyde
98.59
Nickel, PM
5,849.56

-------
Observations from Table 21-7 include the following:
•	Benzene, ethylbenzene, and tetrachloroethylene are the highest emitted pollutants
with cancer UREs in Bronx County while benzene, formaldehyde, and ethylbenzene
are the highest emitted pollutants in Monroe County.
•	Formaldehyde, benzene, and 1,3-butadiene are the pollutants with the highest
toxicity-weighted emissions (of the pollutants with cancer UREs) for both New York
counties.
•	Seven of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Bronx County; six of the highest emitted pollutants also have the
highest toxicity-weighted emissions for Monroe County.
•	Naphthalene, which is a pollutant of interest for both sites and has the highest cancer
risk approximation for each site, appears on both emissions-based lists for Bronx and
Monroe Counties.
•	Emissions of several POM Groups rank among the highest emitted pollutants as well
as the pollutants with the highest toxicity-weighted emissions for Bronx County.
POM, Group 2b appears on both emissions-based lists for Bronx County and includes
several PAHs sampled for at BXNY, including acenaphthene, fluoranthene, and
fluorene. POM, Group 2d also appears on both emissions-based lists for Bronx
County and includes anthracene, phenanthrene, and pyrene. None of these pollutants
failed screens for BXNY. POM, Group 5a also appears among those with the highest
toxicity-weighted emissions for Bronx County and includes benzo(a)pyrene, which
failed a single screen for BXNY.
•	POM, Groups 2b, 2d, and 5a also appear among the pollutants with the highest
toxicity-weighted emissions for Monroe County while only POM, Group 2b appears
among the highest emitted pollutants for Monroe County.
Observations from Table 21-8 include the following:
•	Toluene and methanol are the highest emitted pollutants with noncancer RfCs in both
Bronx and Monroe Counties. The emissions of toluene are considerably higher than
the other pollutants listed for both Bronx and Monroe Counties.
•	The pollutant with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) is acrolein for both counties. Formaldehyde and 1,3-butadiene round
out the top three for both counties, although the order varies.
•	Three of the highest emitted pollutants in Bronx County are also among the pollutants
with the highest toxicity-weighted emissions; four of the highest emitted pollutants in
Monroe County are also among the pollutants with the highest toxicity-weighted
emissions.
21-31

-------
• Naphthalene is the only pollutant of interest for each site for which a noncancer
hazard approximation could be calculated. Naphthalene is among the pollutants with
the highest toxicity-weighted emissions for each county, but is not among the highest
emitted pollutants with a noncancer toxicity factor for either county.
21.6 Summary of the 2013 Monitoring Data for BXNY and ROCH
Results from several of the data treatments described in this section include the
following:
~~~ Six pollutants failed screens for BXNY, of which three were identified as pollutants of
interest. Four pollutants failed screens for ROCH, all of which were identified as
pollutants of interest. Naphthalene, acenaphthene, andfluorene were identified as
pollutants of interest for both New York monitoring sites.
~~~ Naphthalene had the highest annual average concentration for both sites, although
the annual average for BXNY is twice the annual average for ROCH.
~~~ Concentrations of acenaphthene andfluorene for both sites andfluoranthene for
ROCH were highest during the warmer months of the year.
~~~ ROCH and BXNY have the second and fifth highest annual average concentrations of
acenaphthene (respectively) among NMP sites sampling PAHs. BXNY has the third
highest annual average concentration of naphthalene among NMP sites sampling
PAHs.
~~~ Naphthalene has the highest cancer risk approximation among the pollutants of
interest for both BXNY and ROCH. None of the pollutants of interest have noncancer
hazard approximations greater than an HQ of 1.0.
21-32

-------
22.0	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.
22.1	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.
Three Oklahoma sites (TOOK, TMOK, and TROK) are located in Tulsa, Oklahoma. Two
sites are located in Oklahoma City, Oklahoma (ADOK and OCOK), although the
instrumentation at ADOK was moved mid-year to a new location in Yukon, Oklahoma, just west
of Oklahoma City (YUOK).
Figures 22-1 through 22-3 are composite satellite images retrieved from ArcGIS Explorer
showing the Tulsa monitoring sites and their immediate surroundings. Figure 22-4 identifies
nearby point source emissions locations by source category, as reported in the 2011 NEI for
point sources, version 2. Note that only sources within 10 miles of the sites are included in the
facility counts provided in Figure 22-4. A 10-mile boundary was chosen to give the reader an
indication of which emissions sources and emissions source categories could potentially have a
direct effect on the air quality at the monitoring sites. Further, this boundary provides both the
proximity of emissions sources to the monitoring sites as well as the quantity of such sources
within a given distance of the sites. Sources outside the 10-mile boundary are still visible on the
map for reference, but have been grayed out in order to emphasize emissions sources within the
boundary. Figures 22-5 through 22-8 are the composite satellite maps and emissions source map
for the Oklahoma City sites. Table 22-1 provides supplemental geographical information such as
land use, location setting, and locational coordinates for each site.
22-1

-------
Figure 22-1. Public Works, Tulsa, Oklahoma (TOOK) Monitoring Site
W-23rd St S


-------
Figure 22-2. Fire Station, Tulsa, Oklahoma (TIY1QK) Monitoring Site
K>
to
< f'- -if-T r* m
< ¦ • f . -F7
P ¦ i' 'r VKiT J I
r	I 'lH. *1 PI I in 1 IV I ¦ 1
/	-is 'jrTF» *.# / i!
jf a .«. I * w ¦ k ? ¦ »K • I
'4?	.7,' /f ••¦r, >	J. (
Iff a	Sf*
•,	- sj	.il; •- ¦ - - - a, I		
*- .	^	- ' t-l	^ I w r>" J
• »T. t	'[• .
A
J • •
fj
-E36th'StN-
HI4.H T* 1
BS g
Jliw
»¦ «!r'
/* g:	M I-F h i tig* A ;: I
jfafsJi r i .f j 9Kf% i muI, | /- f V
1 'litiQ" rjp i 1 ,4* 1Rj r
JSP SS# • I a | if vPJU ///
^ ! ®£ 1-1 $ -i '¦
G/a. J0 ;f? j / Vf; *f I »M*»*.r*«skSHi. Ijf
'Hm*.	/C^»ac '	' HMV.-N-1J '-'fa f. ;~	ytf
£»og |	*	--rfL	
K* Roc * Gwfc M
E-36th St N O
^ r
* Jrl f *
>0 *•	«=| a] V t->¦
i j ,'¦ L.7h. '
¦Srru	v.,
jj- * z J fc _ .. V - .r '¦ ij*
¦ »i -I? )H1 PI N'nl	C—TililPIIM		¦, ™
r . r •	•••!
6\\cTe§ *

-------
Figure 22-3. Riverside, Tulsa, Oklahoma (TROK) Monitoring Site
f
M ii ¦ ¦ W Edison-St-—	w fcdison'St-
I	I " • I t . m
,<	rtL.ii.v.ci— -«	i	i
ffl *-A» J	.	.1"
i' k(>|. 4 n • «#./». Jf ¦. 1 j t
p f'Tt*"" * ¦] r I -¦ - *¦ j	Xi	"ft "—<
>— im, - jr K
W-Edison-St
¦/	> y
"F L	-
#	«v
/<	J *
s	11	_¦ " e	* | n jyn * 'iiL ' ^"U
. C	^MUt
¦aXWiUH*tl(HVj>ti«aadn
© . V, wHrtdyAi-
*" O
Or
J

Q ffl mmm	— ..Mh8n
to
llUb&IJjHJ
¦ •' |!
t 4 vi
i. -1 4
II
ISC
5 J
• -f, >«•
...
1 •'
t Q md¦'¦
- V« A^trril dI ? M I —
I , ¦ „
t IM
	ATii* 1A.
u\v^vd' —s "TJFH
»s paQe ® (n^ «"?
Wf a
"""'ftw/es.p
.- * »V Ad«r.. rfki fllvil -
o !
I i £
I * 'li
p#-' ¦ 3; J"	g	-I	I	*•'1/
-IW.lnSt!.	» jg	.3	§	5	j^Jr '
~l *• I	*	a	•?	•} -if
L J	*	V I
Lir JN / l_ Iff
»1 aUv ° r

]	WJnd V -
•». ..
r.nlH
sriv

X -T K
wNST OL
O/ka»
t-%
, ® #
^Charks P^ye^Blvd W.3rCi St S
* v
t>*ng
Source US G S
Source NASA, NGA. USGS
_ 2008 Microsoft C«»p.
\ r
t ! Am
H L ri
\ f-Li
^"sri
l^r HiT L n

-------
Figure 22-4. NEI Point Sources Located Within 10 Miles of TMOK, TOOK, and TROK

Oiage
County
Roue's
County
Tuna
County
Kaystan*
Creek
County
irKrcrw
wrocniv
Nolo. Duo to faculty density and collDcaUon. II: e total faciHies
Legend	dttpl#^»d may pot raprirant all facltm withn rh« S'O- ot mtanst
TMOK UATMP site ~ TOOK UATMP site ^ TROK UATMP site
10 mile radius j	| County boundary
Source Category Group (No. of Facilities)
^ **tcuftitg rji
GMftng engorvn} iiujiIv Sfwtw to Uanirt>rlu*n •?!
WaUHi P-soiiMnij'.-ilftjitrriKm (?,
F**e«»»0 f-ltMy HI
wOl
•	Ok Mid IV 
o	-iiiru.'j	UifiA^ciuum f«jr» i • t
22-5

-------
Figure 22-5. Air Depot, Oklahoma City, Oklahoma (ADOK) Monitoring Site

ft«3 r»
r%5ource USGS f
[r fit NASA NGA. USGS
0 0 H Mu ro»oM loip
to
K>
ON

-------
Figure 22-6. Oklahoma City, Oklahoma (OCOK) Monitoring Site
Jg Memorial Rd
E.Memorialed-
¦t Memorial 
-------
Figure 22-7. Yukon, Oklahoma (YUOK) Monitoring Site
to
K>
00
VX
Mihl
V-' »
§
V
/
7'
^•i|

,I|lj
. VI
9
J

cL-4
V f
m
IS* 1


•I kf

V
1.
V I
v
%
i
\
N.v K \r
NW 10th St
a .M
NW 10th S

Ui ' T'-' ' "
, *i WUr^i V< 0'

iu ,laif**,.I F	c'l
~	^ ' 1*
F Hi
lisaJ|lif'PtiRlHMk4MP
^ , .. "» «?
^ -«•
* l a imftihmil
, v4£i^ W^W' ki Mi(i Di
i j*!M,1 »•**<><'
U—CP^kljfxi t>t /i y


~U'O r~~i,'
V>l Si^'I.OrX
lliH
:r
*
ft' *
' Ml
* & x Jr:
at:
t /¦ //.ra
©

cii
-v J. .
NW 10th S;
WVV^ifgPH £,
• ii If



/1
.5
T
J
III
4-'
T5
OC
f
ft' £
tang
4>
I *
<4L
4#L. W Mh SI.
-NWSta-Sfe- <5
J
iDur c
200R Microioft Corp.
'—	'Skfur c« U5GS
Source ' N*ASA NGA. USGS
«
cr- n4

9
a>
T3
'»> .
esrb

-------
Figure 22-8. NET Point Sources Located Within 10 Miles of ADOK, OCOK, and YUOK
North Canatfan
Oklahoma
County
f a*» sr*nn»y
i Cleveland
. County
¦ wvw


¦r v j w
v HI
Logati I
County
Kingfisher
1 Cowtty
Canadian
I County
Grady ,
County
- -#
__ _C*7*cfl»r I m —
Rjvw t
trsFC~*» cat»©n tha total tBciUfn
Legend	di&pta>-ed w>a> not repfesarrt alt 
O	Railroad Engmes'Parts Manufactunng FaciMy (1)
22-9

-------
Table 22-1. Geographical Information for the Oklahoma Monitoring Sites
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude and
Longitude
Land Use
Location
Setting
Additional Ambient Monitoring Information1
TOOK
40-143-0235
Tulsa
Tulsa
Tulsa, OK
36.126945,
-95.998941
Industrial
Urban/City
Center
SO2, H2S, Meteorological parameters.
TMOK
40-143-1127
Tulsa
Tulsa
Tulsa, OK
36.204902,
-95.976537
Residential
Urban/City
Center
CO, SO2, NOy, NO, NO2, NOx, O3, Meteorological
parameters, PM0, PM Coarse, PM2.5, and PM2 5
Speciation, IMPROVE Speciation.
TROK
40-143-0179
Tulsa
Tulsa
Tulsa, OK
36.154830,
-96.015845
Industrial
Urban/City
Center
SO2, H2S, Meteorological parameters.
ADOK
40-109-0042
Oklahoma
City
Oklahoma
Oklahoma City,
OK
35.380316,
-97.405720
Commercial
Urban/City
Center
None.
OCOK
40-109-1037
Oklahoma
City
Oklahoma
Oklahoma City,
OK
35.614131,
-97.475083
Residential
Suburban
CO, SO2, NO, NO2, NOx, O3, Meteorological
parameters, PM coarse, PM10, PM2 5, PM2 5
Speciation, IMPROVE Speciation.
YUOK
40-017-0101
Yukon
Canadian
Oklahoma City,
OK
35.479215, -
97.751503
Commercial
Suburban
Meteorological parameters, NO, NO2, NOx, O3.
1 Data for additional pollutants are reported to AQS for these sites (EPA, 2014b); however, these data are not generated by ERG and are therefore not included in this report.

-------
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. The site is located near the City of Tulsa West Maintenance Yard, which includes a
public access CNG station. As shown in Figure 22-1, an oil refinery is located just south of West
25th Street South. Another refinery is located to the northwest of the site, on the other side of
1-244. A rail yard is also located on the west side of 1-244, which can be seen on left-hand side of
Figure 22-1.
TMOK is located in north Tulsa on the property of Fire Station Number 24. As shown in
Figure 22-2, the intersection of North Peoria Avenue (Highway 11) and East 36th Street North
lies just to the northeast of the site. The surrounding area is primarily residential, with wooded
areas just to the east, an early childhood education facility and an elementary school to the south,
and a park to the west.
The TROK monitoring site is located west of downtown Tulsa, less than one-half mile
north of the Arkansas River and north-northwest of the TOOK site. Although the area
surrounding the TROK monitoring site is classified as "industrial", the site is immediately
adjacent to a residential dwelling, less than one-quarter mile south of Highway 412/51 (Sand
Springs Expressway). The site is elevated above the river, and a wooded area separates the
residential area from the industrial areas west of Newblock Park, as shown in Figure 22-3.
Figure 22-4 shows that the Tulsa sites are located roughly 5 miles apart, with TMOK
farthest north and TOOK farthest south. Many of the emissions sources are clustered around
TOOK, while there are no point sources within 2 miles of TMOK. There are a variety of
industries in the area although the source category with the greatest number of sources
surrounding the Tulsa sites is the airport source category, which includes airports and related
operations as well as small runways and heliports, such as those associated with hospitals or
television stations. Point sources closest to TOOK include two petroleum refineries (including
one directly under the star symbol for TOOK); a rail yard; a municipal waste combustor; a
compressor station; a metal coating, engraving, and allied services to manufacturers facility; an
airport/airport support operation; and a facility generating electricity via combustion. The closest
22-11

-------
point source to TROK is a refinery located on the other side of the Arkansas River, according to
Figure 22-4. However, several industrial facilities are located between the site and river but are
not included in the NEI for point sources.
The ADOK monitoring site 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 22-5. This site lies northwest of Stanley
Draper Lake and is surrounded by grasslands, with little activity or traffic in the immediate
vicinity. The monitoring site was established at this location to capture any influence from
Tinker Air Force Base and to collect background data (OK DEQ, 2013), although sampling was
discontinued in June 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 22-6.
The instrumentation at ADOK was relocated to the YUOK location in July 2013. This is
the location of an existing monitoring site for Oklahoma Department of Environmental Quality
(ODEQ) in Yukon, a town to the west of Oklahoma City and in neighboring Canadian County.
The monitoring site is located at the Integris water tower, just south of 1-40. The site is located in
a primarily commercial area, although the area north of 1-40 is highly residential while the area
to the south is of mixed usage. An oil well pump jack is located to the southwest of YUOK,
which is shown in the middle of the green field to the southwest of YUOK in Figure 22-7.
Yukon is a rapidly growing area, with both commercial and residential development.
Figure 22-8 shows that the locations of the ADOK, OCOK, and YUOK monitoring sites
form a triangle around Oklahoma City. The outer boundary of each site's 10-mile radius
intersects the other two sites. Most of the point sources located within 10 miles of the three sites
are located between the sites in the center of Oklahoma City (northwest of ADOK, south of
OCOK, and west of YUOK). The source category with the greatest number of sources
22-12

-------
surrounding these sites is the airport source category. The point source closest to ADOK is
Tinker Air Force Base, which lies just on the other side of 1-240. The source closest to OCOK is
involved in brick, structural clay, or clay ceramics. The source closest to YUOK is an oil and gas
production facility, although a chemical manufacturing facility is located roughly the same
distance away.
Table 22-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Oklahoma monitoring sites. Table 22-2 includes both county-level
population and vehicle registration information. Table 22-2 also contains traffic volume
information for each site as well as the location for which the traffic volume was obtained.
Additionally, Table 22-2 presents the county-level daily VMT for Tulsa, Oklahoma, and
Canadian Counties.
Table 22-2. Population, Motor Vehicle, and Traffic Information for the Oklahoma
Monitoring Sites
Site
County
Estimated
County
Population1
County-level
Vehicle
Registration2
Annual
Average
Daily
Traffic3
Intersection
Used for
Traffic Data
County-
level Daily
VMT4 "
TOOK
Tulsa
622,409
614,543
64,424
1-244 at Southwest Blvd
20,453,745
TMOK
12,500
Near E 36th St N and N Peoria
Ave intersection
TROK
56,200
64/51/412 just west of 1-244
ADOK
Oklahoma
755,245
835,642
34,700
1-240 between 1-35 and 1-40
27,469,678
OCOK
41,500
US-77 north of Turnpike
YUOK
Canadian
126,123
106,000
45,400
1-40 west of Hwy 4
4,457,374
bounty-level population estimates reflect 2013 data (Census Bureau, 2014)
2County-level vehicle registrations reflect 2013 data (OKTC, 2013)
3AADT reflects 2012 data (OK DOT, 2012)
4County-level VMT reflects 2013 data (OK DOT, 2014)
Observations from Table 22-2 include the following:
• The Canadian 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 Canadian
County's population is on the lower end.
• The Canadian County vehicle registration is also significantly less than vehicle
registrations for Tulsa and Oklahoma Counties. Compared to other NMP sites, the
Oklahoma County vehicle ownership is in the top third while the vehicle ownership
22-13

-------
for Tulsa County is in the middle third and Canadian County's vehicle ownership is
in the bottom third.
•	The traffic volume passing the TMOK site is the lowest among the Oklahoma
monitoring sites while the traffic passing by TOOK is the highest of the six sites. The
traffic data for TOOK is in the top third while the traffic volumes for the remaining
Oklahoma sites are in the middle third compared to other NMP sites.
•	County-level VMT is greatest for Oklahoma County and ranks 12th compared to
other NMP sites. VMT is the least for Canadian County and ranks 35th compared to
other NMP sites. The VMT for Tulsa County ranks 18th.
22.2 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.
22.2.1	Climate Summary
Tulsa is located in northeast Oklahoma, just southeast of the Osage Indian Reservation,
and along the Arkansas River. 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. Drought conditions are not
uncommon. 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, 2015; NOAA, 2015b).
22.2.2	Meteorological Summary
Hourly meteorological data for 2013 were retrieved from NCDC for the weather stations
closest to the Oklahoma monitoring sites (NCDC, 2013), as described in Section 3.4.2. The
closest weather stations to the Tulsa sites are located at Richard Lloyd Jones Jr. Airport (near
TOOK) and Tulsa International Airport (near TMOK and TROK), WBANs 53908 and 13968,
respectively. The two closest weather stations to the Oklahoma City sites are located at Tinker
22-14

-------
Air Force Base Airport (near ADOK) and Wiley Post Airport (near OCOK and YUOK),
WBANs 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 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 2013. Also included in Table 22-3 is the 95 percent
confidence interval for each parameter. As shown in Table 22-3, average meteorological
conditions on sample days appear cooler and slightly more humid at the Tulsa monitoring sites,
although the differences are not statistically significant. Among the Oklahoma City sites, the
differences are greatest for ADOK, where sample days appear cooler than conditions
experienced throughout the year, but the difference is not statistically significant. Sampling was
discontinued at ADOK at the end of June 2013, thereby missing the warmest months of the year.
The opposite is true for YUOK, where sampling did not begin until July 2013, thereby missing
the coldest months of the year. The average wind speed on sample days near YUOK is lower
than the full-year average wind speed. Near OCOK, where sampling occurred year-round,
sample days appear slightly cooler and more humid, similar to the Tulsa sites. The Oklahoma
City area is the windiest location among NMP sites in 2013.
22-15

-------
Table 22-3. 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)
Public Works, Tulsa, Oklahoma - TOOK
Richard Lloyd
Jones Jr.
Airport
53908
(36.04, -95.98)
6.1
miles
Sample
Days
(64)
68.0
±4.9
57.7
±4.7
46.0
±4.8
51.7
±4.3
68.0
±2.9
1018.5
± 1.8
5.1
±0.6
172°
(S)
2013
70.5
+ 1.9
59.5
+ 1.9
46.9
+ 1.9
52.9
+ 1.7
66.6
+ 1.2
1017.9
±0.7
5.2
±0.3



Fire Station, Tulsa,
Oklahoma - TMOK



Tulsa
International
Airport
13968
(36.20, -95.89)
5.0
miles
Sample
Days
(63)
67.3
±5.0
57.9
±4.8
46.1
±4.8
51.7
±4.4
68.0
±3.3
1017.6
± 1.9
7.9
±0.7
94°
(E)
2013
69.6
+ 1.9
59.6
+ 1.9
46.8
+ 1.9
52.9
+ 1.7
65.9
+ 1.3
1016.9
±0.7
8.1
±0.3
Riverside, Tulsa, Oklahoma - TROK
Tulsa
International
Airport
13968
(36.20, -95.89)
7.8
miles
Sample
Days
(64)
68.2
±5.0
58.7
±4.9
46.7
±4.8
52.4
±4.4
67.6
±3.2
1017.6
± 1.9
7.7
±0.7
67°
(ENE)
2013
69.6
+ 1.9
59.6
+ 1.9
46.8
+ 1.9
52.9
+ 1.7
65.9
+ 1.3
1016.9
±0.7
8.1
±0.3
Air De
)ot, Oklahoma City, Oklahoma
- ADOK



Tinker
AFB/Airport
2.8
miles
Sample
Days
(31)
63.9
±6.2
54.9
±6.1
45.1
±6.3
49.9
±5.6
72.9
±5.5
1015.9
±2.1
9.8
± 1.1
13919
(35.42, -97.38)
27°
(NNE)
2013
69.9
+ 1.9
59.5
+ 1.8
47.7
+ 1.9
53.2
+ 1.7
68.5
+ 1.5
1017.2
±0.7
9.4
±0.4
1 Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.

-------
Table 22-3. Average Meteorological Conditions near the Oklahoma Monitoring Sites (Continued)
Closest









Weather
Station
Distance
and

Average
Maximum
Average
Average
Dew Point
Average
Wet Bulb
Average
Relative
Average
Sea Level
Average
Scalar Wind
(WBAN and
Coordinates)
Direction
from Site
Average
Type1
Temperature
(°F)
Temperature
(°F)
Temperature
(°F)
Temperature
(°F)
Humidity
(%)
Pressure
(mb)
Speed
(kt)
Oklahoma City, Oklahoma - OCOK
Wiley Post
Airport
11.1
miles
Sample
Davs
(63)
66.5
±5.0
57.1
±4.9
42.9
±4.6
49.7
±4.3
62.1
±3.1
1017.4
± 1.9
9.4
±0.8
03954
240°
(WSW)








(35.53, -97.65)
2013
69.7
+ 1.9
59.5
+ 1.9
44.4
+ 1.8
51.6
+ 1.7
60.6
+ 1.3
1016.5
±0.7
10.1
±0.4
Yukon, Oklahoma - YUOK
Wiley Post
7.0
miles
Sample
Davs
72.2
62.5
46.7
53.8
59.2
1019.0
8.4
Airport
(33)
±7.5
±7.4
±6.9
±6.3
±3.9
±2.8
± 1.0
03954
57°
(ENE)








(35.53, -97.65)

69.7
59.5
44.4
51.6
60.6
1016.5
10.1

2013
+ 1.9
+ 1.9
+ 1.8
+ 1.7
+ 1.3
±0.7
±0.4
1 Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.

-------
22.2.3 Wind Rose Comparison
Hourly surface wind data from the weather stations at Richard Lloyd Jones Jr. Airport
(for TOOK), Tulsa International Airport (for TMOK and TROK), Tinker Air Force Base (for
ADOK), and Wiley Post Airport (for OCOK and TUOK) were uploaded into a wind rose
software program to produce customized wind roses, as described in Section 3.4.2. A wind rose
shows the frequency of wind directions using "petals" positioned around a 16-point compass,
and uses different colors to represent wind speeds.
Figure 22-9 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 22-9 also presents three different wind roses for the
TOOK monitoring site. First, a historical wind rose representing 2003 to 2012 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 2013 is presented.
Next, a wind rose representing wind observations for days on which samples were collected in
2013 is presented. These can be used to identify the predominant wind speed and direction for
2013 and to determine if wind observations on sample days were representative of conditions
experienced over the entire year and historically. Figures 22-10 through 22-14 present the
distance maps and wind roses for the remaining Oklahoma sites.
22-18

-------
Figure 22-9. Wind Roses for the Richard Lloyd Jones Jr. Airport Weather Station near
TOOK
Location of TOOK and Weather Station
2003-2012 Historical Wind Rose
; NORTH""'-,
25%
20%
•15%
10%
WIND SPEED
(Knots)
~ -22
H 17-21
11 - 17
I I 7- 11
I 4-7
H 2-4
Calms: 24.58%
2013 Wind Rose
Sample Day Wind Rose
INORTH"---,
NORTH"-'-,
25%
20%
15%
10%
WIND SPEED
(Knots)
~ >=22
F~1 17-21
¦ 11 -17
I I 7- 11
rzi 4-7
2- 4
Calms: 23.17%
25%
"X 20%
15%
10%
WIND SPEED
(Knots)
I I =22
B3 17-21
¦ 11 - 17
I I 7- 11
~ 4-7
2- 4
Calms: 25.41%
22-19

-------
Figure 22-10. Wind Roses for the Tulsa International Airport Weather Station near
TMOK
Location of TMOK and Weather Station
2003-2012 Historical Wind Rose
5 0 mi (m
; NORTH""'-,
25%
"X 20%
15%
10%
WIND SPEED
(Knots)
~ -22
H 17-21
11 - 17
I I 7- 11
I 4-7
H 2-4
Calms: 8 71%
2013 Wind Rose
Sample Day Wind Rose
INORTH"---,
! NORTH"-'-,
25%
20%
15%
10%
WIND SPEED
(Knots)
~ >=22
F~1 17-21
¦ 11 -17
I I 7- 11
rzi 4-7
2- 4
Calms: 7.17%
25%
"X 20%
15%
10%
WIND SPEED
(Knots)
I I =22
Wl 17-21
¦ 11 - 17
I" 1 7- 11
4-7
2- 4
Calms: 6.9*%
22-20

-------
Figure 22-11. Wind Roses for the Tulsa International Airport Weather Station near TROK
Location of TROK and Weather Station
2003-2012 Historical Wind Rose
! NORTH""'-,
25%
20*
15*
10%
WIND SPEED
(Knots)
I I >-22
~ 17-21
11 - 17
E3 7-11
I ~l 4-7
IH 2-4
Calms: 8.71%
2013 Wind Rose
Sample Day Wind Rose
NORTH
El- ,
WIND SPEED
(Kn OtS)
17-21
SOUTH
Calms: 717%
25%
20%t
15%
10%
WIND SPEED
(Knots)
I I >-22
liBI 17-21
¦ 11 17
0	7- 11
1	~1 4-7
2-4
Calms: 716%
22-21

-------
Figure 22-12. Wind Roses for the Tinker Air Force Base Airport Weather Station near
ADOK
Location of ADOK and Weather Station
2006-2012 Historical Wind Rose
; NORTH""'-,
30%
24%
WIND SPEED
(Knots)
~ -22
H 17-21
11 - 17
I I 7- 11
I 4-7
H 2-4
Calms: 3.41%
2013 Wind Rose
Sample Day Wind Rose
INORTH"---,
NORTH"-'-,
30%
24%
18%
12%
WIND SPEED
(Knots)
~ >=22
F~1 17-21
¦ 11 -17
I I 7- 11
rzi 4-7
2- 4
Calms: 1.75%
30%
"X 24%
18%
12%
WIND SPEED
(Knots)
I I =22
Wl 17-21
¦ 11 - 17
I" 1 7- 11
4-7
2- 4
Calms: 1.22%
22-22

-------
Figure 22-13. Wind Roses for the Wiley Post Airport Weather Station near OCOK
Location of OCOK and Weather Station
2003-2012 Historical Wind Rose
+
Oklahoma
City

NORTH"-'-,^

-,._v X 30%

24%

1 18%

|-N 12%

fc. 6%
•west:" i ^ A
K} :

SOUTH
WIND SPEED
(Knots)
~ >=22
mm 17-21
IH 11 -17
f I 7- 11
\~~3 4-7
2- 4
Calms: 5.15%
2013 Wind Rose
Sample Day Wind Rose
NORTH"''-
(Knots)
17-21
11 - 17
SOUTH
C=!n-=
NORTH"*--,
WIND SPEED
30%
24%
18%
12%
WIND SPEED
(Knots)
I I >-22
S3 17-21
¦ 11 -17
HI 7- 11
i ~l 4-7
2-4
Calms: 5.85%
22-23

-------
Figure 22-14. Wind Roses for the Wiley Post Airport Weather Station near YUOK
Location of YUOK and Weather Station	2003-2012 Historical Wind Rose
west:
WIND SPEED
(Knots)
~ >=22
~
7- 11
4- 7
2013 Wind Rose
Sample Day Wind Rose
NORTH"''-
(Knots)
17-21
11 - 17
SOUTH
C=!n-=
NORTH"
WIND SPEED
30%
24%
18%
12%
WIND SPEED
(Knots)
I I >-22
S3 17-21
¦ 11 -17
HI 7- 11
i ~l 4-7
2-4
Calms: 7.39%
22-24

-------
Observations from Figures 22-9 through 22-14 include the following:
•	The 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 11.1 miles between
OCOK and the Wiley Post Airport.
•	Even though the historical data are from four 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 (those less than or equal to 2 knots) observed, ranging
from as little as 3 percent at the Tinker Air Force Base (ADOK) to as high as
25 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 2013 wind patterns are similar to the historical wind patterns, as are
the sample day wind patterns, although there are slightly fewer southerly winds and
winds from the north to northeast were observed more evenly 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 2013 wind patterns are similar to the historical wind patterns,
although a higher percentage of south-southeasterly winds was observed in 2013. The
sample day wind rose also resembles the historical and full-year wind roses, although
a slightly higher percentage of northwesterly winds and lower percentage of north-
northwesterly winds were observed on sample days. The percentage of calm winds
shown for the full-year and sample day wind roses is slightly less than the percentage
shown on the historical wind rose.
•	The weather station closest to TROK is also located at the Tulsa International
Airport; as such, the historical and full-year wind roses for TROK are identical to
those shown for TMOK. The sample day wind patterns near TROK are also similar to
those shown for TMOK.
•	For ADOK, the historical wind rose includes 7 years of data, starting with 2006. The
2013 wind patterns resemble the historical wind patterns, although there were slightly
more southeasterly to south-southeasterly wind observations in 2013 and fewer south-
southwesterly winds. The calm wind rate also decreased by almost half. The sample
day wind patterns exhibit additional differences, although southerly winds still
prevailed. The percentage of northwesterly to north-northwesterly winds on sample
days is more than twice the percentage shown on the full-year and historical wind
roses. Winds from the south were observed less frequently on sample days and the
calm rate is just greater than 1 percent. Recall, however, that sampling was
discontinued at this site at the end of June; thus, the sample day wind rose includes
observations from the first half of the year only.
22-25

-------
•	For OCOK, the wind patterns shown on the 2013 wind rose resemble the historical
wind patterns. The sample day wind rose for OCOK is similar to both the historical
and full-year wind roses, although winds greater than 22 knots account for a lower
percentage of wind observations on sample days.
•	The weather station closest to YUOK is also located at the Wiley Post Airport; as
such, the historical and full-year wind roses for YUOK are identical to those shown
for OCOK. The sample day wind rose for YUOK is similar to the sample day wind
rose for OCOK, with winds greater than 22 knots also accounting for a lower
percentage of wind observations on sample days. In addition, winds from the
southeast and south-southeast were observed nearly equally and the calm rate is
higher. Recall, however, that sampling at YUOK did not begin until July; thus, the
sample day wind rose includes observations from the second half of the year only.
22.3 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 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. VOCs, carbonyl compounds, and metals (TSP) were sampled for at
each Oklahoma monitoring site.
22-26

-------
Table 22-4. Risk-Based Screening Results for the Oklahoma Monitoring Sites
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Public Works, Tulsa, Oklahoma - TOOK
Acetaldehyde
0.45
61
61
100.00
12.35
12.35
Benzene
0.13
61
61
100.00
12.35
24.70
Carbon Tetrachloride
0.17
61
61
100.00
12.35
37.04
Formaldehyde
0.077
61
61
100.00
12.35
49.39
Arsenic (TSP)
0.00023
57
58
98.28
11.54
60.93
1.3 -Butadiene
0.03
57
57
100.00
11.54
72.47
1,2-Dichloroethane
0.038
34
34
100.00
6.88
79.35
Ethylbenzene
0.4
29
61
47.54
5.87
85.22
Manganese (TSP)
0.03
22
58
37.93
4.45
89.68
Nickel (TSP)
0.0021
22
58
37.93
4.45
94.13
Hexacliloro -1,3 -butadiene
0.045
10
13
76.92
2.02
96.15
Propionaldehyde
0.8
8
61
13.11
1.62
97.77
/?-Dichlorobcnzcnc
0.091
5
40
12.50
1.01
98.79
Cadmium (TSP)
0.00056
3
58
5.17
0.61
99.39
Chloroprene
0.0021
1
1
100.00
0.20
99.60
1,2-Dibromoethane
0.0017
1
1
100.00
0.20
99.80
Lead (TSP)
0.015
1
58
1.72
0.20
100.00
Total
494
802
61.60

Fire Station, Tulsa, Oklahoma - TMOK
Acetaldehyde
0.45
61
61
100.00
12.71
12.71
Formaldehyde
0.077
61
61
100.00
12.71
25.42
Benzene
0.13
60
60
100.00
12.50
37.92
Carbon Tetrachloride
0.17
60
60
100.00
12.50
50.42
1.3 -Butadiene
0.03
57
59
96.61
11.88
62.29
Arsenic (TSP)
0.00023
54
56
96.43
11.25
73.54
1,2-Dichloroethane
0.038
35
35
100.00
7.29
80.83
/?-Dichlorobcnzcnc
0.091
34
51
66.67
7.08
87.92
Ethylbenzene
0.4
25
60
41.67
5.21
93.13
Hexacliloro -1,3 -butadiene
0.045
13
14
92.86
2.71
95.83
Nickel (TSP)
0.0021
7
56
12.50
1.46
97.29
Propionaldehyde
0.8
7
61
11.48
1.46
98.75
Manganese (TSP)
0.03
3
56
5.36
0.63
99.38
Cadmium (TSP)
0.00056
2
56
3.57
0.42
99.79
T richloroethylene
0.2
1
8
12.50
0.21
100.00
Total
480
754
63.66

22-27

-------
Table 22-4. Risk-Based Screening Results for the Oklahoma Monitoring Sites (Continued)
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Riverside, Tulsa, Oklahoma - TROK
Acetaldehyde
0.45
61
61
100.00
13.41
13.41
Benzene
0.13
61
61
100.00
13.41
26.81
Formaldehyde
0.077
61
61
100.00
13.41
40.22
Carbon Tetrachloride
0.17
60
61
98.36
13.19
53.41
1.3 -Butadiene
0.03
54
56
96.43
11.87
65.27
Arsenic (TSP)
0.00023
53
56
94.64
11.65
76.92
1,2-Dichloroethane
0.038
40
40
100.00
8.79
85.71
Ethylbenzene
0.4
28
61
45.90
6.15
91.87
Nickel (TSP)
0.0021
11
56
19.64
2.42
94.29
Hexacliloro -1,3 -butadiene
0.045
10
11
90.91
2.20
96.48
Manganese (TSP)
0.03
6
56
10.71
1.32
97.80
Propionaldehyde
0.8
4
61
6.56
0.88
98.68
Cadmium (TSP)
0.00056
3
56
5.36
0.66
99.34
/?-Dichlorobcnzcnc
0.091
3
28
10.71
0.66
100.00
Total
455
725
62.76

Air Depot, Oklahoma City, Oklahoma - ADOK
Acetaldehyde
0.45
30
30
100.00
13.51
13.51
Benzene
0.13
30
30
100.00
13.51
27.03
Carbon Tetrachloride
0.17
30
30
100.00
13.51
40.54
Formaldehyde
0.077
30
30
100.00
13.51
54.05
1,2-Dichloroethane
0.038
28
28
100.00
12.61
66.67
/?-Dichlorobcnzcnc
0.091
27
29
93.10
12.16
78.83
Arsenic (TSP)
0.00023
25
29
86.21
11.26
90.09
1.3 -Butadiene
0.03
11
16
68.75
4.95
95.05
Hexacliloro -1,3 -butadiene
0.045
5
5
100.00
2.25
97.30
Nickel (TSP)
0.0021
3
29
10.34
1.35
98.65
Manganese (TSP)
0.03
2
29
6.90
0.90
99.55
Propionaldehyde
0.8
1
30
3.33
0.45
100.00
Total
222
315
70.48

22-28

-------
Table 22-4. Risk-Based Screening Results for the Oklahoma Monitoring Sites (Continued)
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Oklahoma City, Oklahoma - OCOK
Acetaldehyde
0.45
61
61
100.00
14.52
14.52
Benzene
0.13
61
61
100.00
14.52
29.05
Carbon Tetrachloride
0.17
61
61
100.00
14.52
43.57
Formaldehyde
0.077
61
61
100.00
14.52
58.10
Arsenic (TSP)
0.00023
51
61
83.61
12.14
70.24
1,2-Dichloroethane
0.038
51
51
100.00
12.14
82.38
1.3 -Butadiene
0.03
37
40
92.50
8.81
91.19
Hexacliloro -1,3 -butadiene
0.045
15
16
93.75
3.57
94.76
p-Dichlorobenzene
0.091
6
39
15.38
1.43
96.19
Ethylbenzene
0.4
6
61
9.84
1.43
97.62
Manganese (TSP)
0.03
3
61
4.92
0.71
98.33
Nickel (TSP)
0.0021
2
61
3.28
0.48
98.81
Propionaldehyde
0.8
2
61
3.28
0.48
99.29
Bromo methane
0.5
1
47
2.13
0.24
99.52
Cadmium (TSP)
0.00056
1
61
1.64
0.24
99.76
T richloroethylene
0.2
1
7
14.29
0.24
100.00
Total
420
810
51.85

Yukon, Oklahoma - YUOK
Acetaldehyde
0.45
30
30
100.00
14.78
14.78
Benzene
0.13
30
30
100.00
14.78
29.56
Carbon Tetrachloride
0.17
30
30
100.00
14.78
44.33
Formaldehyde
0.077
30
30
100.00
14.78
59.11
Arsenic (TSP)
0.00023
26
31
83.87
12.81
71.92
1.3 -Butadiene
0.03
21
22
95.45
10.34
82.27
1,2-Dichloroethane
0.038
20
20
100.00
9.85
92.12
Hexacliloro -1,3 -butadiene
0.045
7
9
77.78
3.45
95.57
/?-Dichlorobcnzcnc
0.091
5
15
33.33
2.46
98.03
Manganese (TSP)
0.03
3
31
9.68
1.48
99.51
1,2-Dibromoethane
0.0017
1
1
100.00
0.49
100.00
Total
203
249
81.53

Observations from Table 22-4 include the following:
• Seventeen pollutants failed at least one screen for TOOK; nearly 62 percent of
concentrations for these 17 pollutants were greater than their associated risk screening
value (or failed screens).
22-29

-------
Eleven pollutants contributed to 95 percent of failed screens for TOOK and therefore
were identified as pollutants of interest for this site. These 11 include two carbonyl
compounds, six VOCs, and three TSP metals. TOOK is one of only two NMP sites
for which manganese was identified as a pollutant of interest.
Fifteen pollutants failed at least one screen for TMOK; nearly 64 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 TMOK and therefore
were identified as pollutants of interest for this site. These 10 include two carbonyl
compounds, seven VOCs, and one TSP metal.
Fourteen pollutants failed at least one screen for TROK; 63 percent of concentrations
for these 14 pollutants were greater than their associated risk screening value (or
failed screens).
Ten pollutants contributed to 95 percent of failed screens for TROK and therefore
were identified as pollutants of interest for this site. These 10 include two carbonyl
compounds, six VOCs, and two TSP metals.
Twelve pollutants failed at least one screen for ADOK; 70 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 ADOK and therefore
were identified as pollutants of interest for this site. These eight include two carbonyl
compounds, five VOCs, and one TSP metal.
Sixteen pollutants failed at least one screen for OCOK; 52 percent of concentrations
for these 16 pollutants were greater than their associated risk screening value (or
failed screens).
Ten pollutants contributed to 95 percent of failed screens for OCOK and therefore
were identified as pollutants of interest for this site. These 10 include two carbonyl
compounds, seven VOCs, and one TSP metal. Note that because />dichlorobenzene
and ethylbenzene failed the same number of screens, both pollutants were identified
as pollutants of interest for OCOK.
Eleven pollutants failed at least one screen for YUOK; nearly 82 percent of
concentrations for these 11 pollutants were greater than their associated risk screening
value (or failed screens).
Eight pollutants contributed to 95 percent of failed screens for YUOK and therefore
were identified as pollutants of interest for this site. These eight include two carbonyl
compounds, five VOCs, and one TSP metal.
22-30

-------
•	The number of pollutants identified as pollutants of interest range from eight to 11
among the Oklahoma sites. These sites have seven pollutants of interest in common:
acetaldehyde, arsenic, benzene, 1,3-butadiene, carbon tetrachloride,
1,2-dichloroethane, and formaldehyde.
•	TOOK failed the fourth highest number of screens among NMP sites, with other
Oklahoma sites ranking seventh (TMOK), 10th (TROK), and 12th (OCOK), as shown
in Table 4-8. The ADOK and YUOK sites failed fewer screens, ranking much lower.
However, sampling at these sites includes only half of a year's worth of samples.
22.4 Concentrations
This section presents various concentration averages used to characterize air toxics
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. Site-specific statistical summaries for all pollutants
sampled for at the Oklahoma sites are provided in Appendices J, L, and N.
22.4.1 2013 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
22-31

-------
to 85 percent, as presented in Section 2.4. Quarterly and annual average concentrations for the
Oklahoma monitoring sites are presented in Table 22-5, 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"
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 Oklahoma Monitoring Sites

# of






Measured
1st
2nd
3rd
4th


Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs. # of
Samples
Average
(Hg/m3)
Average
(Hg/m3)
Average
(Hg/m3)
Average
(Hg/m3)
Average
(Hg/m3)

Public Works, Tulsa,
Oklahoma - TOOK




1.45
2.19
2.93
1.54
2.02
Acetaldehyde
61/61
±0.37
±0.56
±0.44
±0.30
±0.25


1.22
1.07
1.64
0.93
1.21
Benzene
61/61
±0.40
±0.26
±0.40
±0.24
±0.17


0.09
0.05
0.06
0.10
0.07
1.3 -Butadiene
57/61
±0.04
±0.01
±0.02
±0.03
±0.01


0.61
0.69
0.64
0.55
0.63
Carbon Tetrachloride
61/61
±0.03
±0.05
±0.04
±0.06
±0.03


0.09
0.05
0.05
0.07
0.06
1,2-Dichloroethane
34/61
±0.04
±0.03
±0.03
±0.03
±0.02


0.47
0.42
0.47
0.42
0.45
Ethylbenzene
61/61
±0.16
±0.10
±0.11
±0.13
±0.06


2.05
3.29
4.58
1.63
2.87
Formaldehyde
61/61
±0.36
± 1.12
±0.72
±0.32
±0.44


0.02
0.01
0.01
0.03
0.02
Hexachloro-1,3 -butadiene
13/61
±0.02
±0.02
±0.01
±0.02
±0.01


0.74
0.76
0.88
0.71
0.78
Arsenic (TSP)
58/58
±0.13
±0.14
±0.15
±0.17
±0.07


23.25
28.89
31.88
25.94
27.59
Manganese (TSP)
58/58
± 10.12
±9.66
±9.45
± 10.78
±4.75


2.05
2.05
2.50
1.70
2.09
Nickel (TSP)
58/58
±0.74
±0.69
± 1.29
±0.55
±0.42
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.
22-32

-------
Table 22-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Oklahoma Monitoring Sites (Continued)

# of






Measured
1st
2nd
3rd
4th


Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs. # of
Samples
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Average
frig/m3)
Fire Station, Tulsa, Oklahoma - TMOK


1.25
2.14
2.84
1.55
1.94
Acetaldehyde
61/61
±0.30
±0.62
±0.44
±0.28
±0.25


1.08
0.69
1.12
1.00
0.96
Benzene
60/60
±0.27
±0.09
±0.23
±0.29
±0.12


0.11
0.06
0.10
0.15
0.11
1.3 -Butadiene
59/60
±0.05
±0.01
±0.03
±0.05
±0.02


0.62
0.66
0.63
0.56
0.62
Carbon Tetrachloride
60/60
±0.05
±0.06
±0.03
±0.04
±0.02


0.04
0.10
0.15
0.11
0.10
/?-Dichlorobcnzcnc
51/60
±0.02
±0.05
±0.04
±0.02
±0.02


0.09
0.08
0.03
0.04
0.06
1,2-Dichloroethane
35/60
±0.03
±0.03
±0.02
±0.03
±0.01


0.48
0.28
0.52
0.47
0.43
Ethylbenzene
60/60
±0.23
±0.05
±0.13
±0.15
±0.07


2.07
3.54
4.59
2.59
3.19
Formaldehyde
61/61
±0.38
± 1.30
±0.82
±0.36
±0.45


0.01
0.01
0.02
0.03
0.02
Hexachloro-1,3 -butadiene
14/60
±0.02
±0.01
±0.02
±0.02
±0.01


0.56
0.65
0.75
0.61
0.65
Arsenic (TSP)
56/56
±0.13
±0.15
±0.14
±0.16
±0.07
Riverside, Tulsa, Oklahoma - TROK


1.14
1.55
2.38
1.48
1.63
Acetaldehyde
61/61
±0.25
±0.39
±0.34
±0.34
±0.20


1.32
0.80
0.99
0.92
1.00
Benzene
61/61
±0.91
±0.16
±0.20
±0.28
±0.23


0.09
0.05
0.06
0.10
0.07
1.3 -Butadiene
56/61
±0.03
±0.01
±0.01
±0.03
±0.01


0.59
0.68
0.63
0.55
0.61
Carbon Tetrachloride
61/61
±0.04
±0.05
±0.02
±0.07
±0.03


0.08
0.07
0.04
0.08
0.07
1,2-Dichloroethane
40/61
±0.03
±0.03
±0.02
±0.03
±0.01


0.32
0.35
0.46
0.45
0.39
Ethylbenzene
61/61
±0.11
±0.09
±0.10
±0.19
±0.06


1.71
2.98
4.51
1.94
2.77
Formaldehyde
61/61
±0.36
± 1.00
±0.58
±0.33
±0.41


0.01
0.01
0.01
0.03
0.01
Hexachloro-1,3 -butadiene
11/61
±0.01
±0.01
±0.01
±0.02
±0.01


0.77
0.72
0.83
0.88
0.80
Arsenic (TSP)
56/56
±0.24
±0.26
±0.21
±0.27
±0.11


1.40
1.14
1.73
1.45
1.43
Nickel (TSP)
56/56
±0.55
±0.39
±0.51
±0.50
±0.24
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.
22-33

-------
Table 22-5. 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
frig/m3)
2nd
Quarter
Average
frig/m3)
3rd
Quarter
Average
frig/m3)
4th
Quarter
Average
frig/m3)
Annual
Average
frig/m3)
Oklahoma City, Oklahoma - OCOK
Acetaldehyde
61/61
1.15
±0.23
1.93
±0.49
2.72
±0.40
1.60
±0.33
1.85
±0.23
Benzene
61/61
0.83
±0.18
0.50
±0.06
0.61
±0.13
1.14
± 1.14
0.78
±0.29
1.3 -Butadiene
40/61
0.05
±0.03
0.02
±0.01
0.04
±0.02
0.06
±0.04
0.04
±0.01
Carbon Tetrachloride
61/61
0.63
±0.06
0.67
±0.05
0.63
±0.05
0.62
±0.02
0.64
±0.02
/?-Dichlorobcnzcnc
39/61
0.06
±0.02
0.04
±0.02
0.05
±0.02
0.03
±0.02
0.05
±0.01
1,2-Dichloroethane
51/61
0.09
±0.02
0.10
±0.03
0.05
±0.02
0.06
±0.02
0.08
±0.01
Ethylbenzene
61/61
0.24
±0.09
0.20
±0.04
0.28
±0.07
0.22
±0.07
0.23
±0.03
Formaldehyde
61/61
1.53
±0.30
2.80
±0.84
4.47
±0.63
1.78
±0.50
2.63
±0.41
Hexachloro-1,3 -butadiene
16/61
0.02
±0.02
0.02
±0.02
0.01
±0.01
0.03
±0.02
0.02
±0.01
Arsenic (TSP)
61/61
0.32
±0.08
0.53
±0.13
0.60
±0.13
0.39
±0.10
0.46
±0.06
Air Depot, Oklahoma City, Oklahoma - ADOK
Acetaldehyde
30/30
1.21
±0.19
1.97
±0.50
NA
NA
NA
Benzene
30/30
0.62
±0.09
0.43
±0.06
NA
NA
NA
1.3 -Butadiene
16/30
0.03
±0.02
0.02
±0.02
NA
NA
NA
Carbon Tetrachloride
30/30
0.62
±0.05
0.65
±0.06
NA
NA
NA
p-Dichlorobenzene
29/30
0.12
±0.02
0.14
±0.01
NA
NA
NA
1,2-Dichloroethane
28/30
0.09
±0.02
0.10
±0.03
NA
NA
NA
Formaldehyde
30/30
1.67
±0.30
2.82
±0.78
NA
NA
NA
Arsenic (TSP)
29/29
0.32
±0.06
0.44
±0.12
NA
NA
NA
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
22-34

-------
Table 22-5. 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
frig/m3)
2nd
Quarter
Average
frig/m3)
3rd
Quarter
Average
frig/m3)
4th
Quarter
Average
frig/m3)
Annual
Average
frig/m3)
Yukon, Oklahoma - YUOK
Acetaldehyde
30/30
NA
NA
2.30
±0.27
1.30
±0.29
NA
Benzene
30/30
NA
NA
0.48
±0.05
0.52
±0.07
NA
1,3-Butadiene
22/30
NA
NA
0.04
±0.01
0.04
±0.02
NA
Carbon Tetrachloride
30/30
NA
NA
0.65
±0.04
0.59
±0.02
NA
1,2-Dichloroethane
20/30
NA
NA
0.06
±0.02
0.05
±0.02
NA
Formaldehyde
30/30
NA
NA
4.18
±0.52
1.78
±0.40
NA
Hexachloro-1,3 -butadiene
9/30
NA
NA
0.02
±0.02
0.02
±0.02
NA
Arsenic (TSP)
31/31
NA
NA
0.51
±0.07
0.36
±0.11
NA
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 the Oklahoma sites from Table 22-5 include the following:
•	Formaldehyde has the highest annual average concentration for each site, where
annual averages could be calculated, followed by acetaldehyde and benzene. With the
exception of TOOK and TROK, these were the only two pollutants of interest with
annual average concentrations greater than or equal to 1 |ig/m3 for each site. For
TOOK and TROK, benzene also has an annual average concentration greater than or
equal to 1 |ig/m3.
•	Annual average concentrations of formaldehyde range from 2.63 ± 0.41 |ig/m3 for
OCOK to 3.19 ± 0.45 |ig/m3 for TMOK. The annual average concentrations of
acetaldehyde range from 1.63 ± 0.20 |ig/m3 for TROK to 2.02 ± 0.25 |ig/m3 for
TOOK. The annual average concentrations of benzene range from 0.78 ± 0.29 |ig/m3
for OCOK to 1.21 ±0.17 |ig/m3 for TOOK. TOOK has had the highest annual
average benzene concentration among the Oklahoma sites for several years (and is
usually one of the highest across the program), including 2013, although the
difference is becoming less significant.
•	Concentrations of the carbonyl compounds, formaldehyde in particular, tended to be
highest in the warmer months and lowest in the cooler months. While concentrations
were highest during the third quarter, the confidence intervals associated with the
second quarter averages tended to be larger.
22-35

-------
•	The risk screening value for manganese increased from 0.005 |ig/m3 to 0.03 |ig/m3
for the 2013 report. As a result, this pollutant has significantly fewer failures in the
2013 report than in previous reports. For example, manganese failed only 61 screens
for 2013 while failing 706 in the 2012 NMP report. Of the 61 failures for 2013, more
than half were for concentrations measured at the Oklahoma sites (39). Of these 39,
22 were measured at TOOK, the only Oklahoma site (and one of only two NMP sites)
for which manganese is a pollutant of interest. The other 17 break out as follows: two
for ADOK, three for OCOK, three for TMOK, six for TROK, and three for YUOK.
Observations for TOOK from Table 22-5 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. The two highest formaldehyde concentrations were measured at
TOOK on June 21, 2013 (7.91 |ig/m3) and June 27, 2013 (7.77 |ig/m3). Of the 13
formaldehyde concentrations greater than 4 |ig/m3 measured at TOOK, three were
measured during the second quarter and 10 were measured during the third quarter.
•	A similar observation can be made for the quarterly average concentrations of
acetaldehyde. The maximum acetaldehyde concentration was measured at TOOK on
June 27, 2013 (4.42 |ig/m3), while the next four highest concentrations were
measured during the third quarter.
•	The third quarter average concentration of nickel is the highest among TOOK's
quarterly average concentrations of nickel and has a relatively large confidence
interval associated with it. A review of the data shows that the maximum
concentration of nickel was measured at TOOK on July 3, 2013 (11.0 ng/m3). This
measurement is nearly twice the next highest concentration (5.74 ng/m3) measured at
TOOK and more than three times greater than the next highest concentration
measured during the third quarter. If the maximum concentration measured at TOOK
was removed from the dataset, the third quarter average would be less than 2 ng/m3.
This measurement is the third highest nickel concentration measured across the
program.
•	The third quarter average concentrations of the other two speciated metals are greater
than each of their other quarterly averages, although not significantly so. A review of
the data shows that the maximum concentrations of manganese (75.6 ng/m3) and
arsenic (1.51 ng/m3) were measured at TOOK on the same day, September 25, 2013.
Observations for TMOK from Table 22-5 include the following:
•	The second and third quarter average concentrations of formaldehyde are higher than
the other quarterly averages for TMOK. Although the third quarter average
concentration is the highest, the confidence interval is highest for the second quarter
average concentration. A similar observation was made for concentrations of this
pollutant measured at TOOK. A review of the data shows that the two highest
formaldehyde concentrations were also measured at TMOK on June 21, 2013
22-36

-------
(9.43 |ig/m3) and June 27, 2013 (7.78 |ig/m3). Of the 17 formaldehyde concentrations
greater than 4 |ig/m3 measured at TMOK, five were measured during the second
quarter and 10 were measured during the third quarter (with one each in the other
calendar quarters). The minimum concentration of formaldehyde, the only one less
than 1 |ig/m3, was also measured during the second quarter of 2013.
•	A similar observation can be made for the quarterly average concentrations of
acetaldehyde. The maximum acetaldehyde concentration was measured at TMOK on
June 27, 2013 (4.75 |ig/m3), while the next five highest concentrations were measured
during the third quarter. The minimum concentration of acetaldehyde was also
measured during the second quarter of 2013.
•	Several of the lowest quarterly average concentrations of the VOC pollutants of
interest for TMOK (benzene, 1,3-butadiene, and ethylbenzene in particular) were
calculated for the second quarter of 2013. In addition, these averages exhibit the least
variability. A review of the benzene data shows that the range of measurements was
smallest for the second quarter, that the median concentration was the lowest for the
second quarter, and that the second quarter of 2013 had the fewest benzene
concentrations greater than 1 |ig/m3 (only one, compared to five for the first quarter,
eight for the third, and six for the fourth). For ethylbenzene, no concentrations greater
than 0.5 |ig/m3 were measured during the second quarter while between four and six
were measured in the other quarters. Similar observations can be made for
1,3-butadiene, where no concentrations greater than 0.1 |ig/m3 were measured at
TMOK during the second quarter and between five and nine were measured during
the other calendar quarters.
•	Concentrations of />dichlorobenzene were lowest during the first quarter of 2013 for
TMOK. Six of the nine non-detects were measured during the first quarter of 2013
and none were measured at TMOK after April. In addition, five of the eight lowest
measured detections (those less than 0.05 |ig/m3) were measured during the first
quarter (and the other three were measured during the second quarter).
•	Arsenic concentrations appear highest during the third quarter, although the quarterly
averages shown in Table 22-5 are not significantly different. The two highest arsenic
concentrations were measured at TMOK in July and August. Conversely, of the 18
arsenic concentrations less than 0.5 ng/m3 measured at TMOK, only two were
measured during the third quarter, compared to seven for the first quarter, five for the
second, and four for the fourth.
Observations for TROK from Table 22-5 include the following:
•	The second and third quarter average concentrations of formaldehyde are higher than
the other quarterly averages for TROK. Although the third quarter average
concentration is the highest, the confidence interval is highest for the second quarter.
A similar observation was made for concentrations of this pollutant measured at
TOOK and TMOK. A review of the data shows that the highest formaldehyde
concentration was measured at TROK on June 27, 2013 (7.78 |ig/m3), the same date
that the second highest concentrations of formaldehyde were measured at TOOK and
22-37

-------
TMOK. Of the 15 formaldehyde concentrations greater than 4 |ig/m3 measured at
TROK, five were measured in June and 10 were measured during the third quarter.
The minimum concentration of formaldehyde (0.631 |ig/m3) was also measured
during the second quarter of 2013.
•	Concentrations of acetaldehyde were highest during the third quarter of 2013 at
TROK. The maximum acetaldehyde concentration was measured at TROK on
September 1, 2013 (3.90 |ig/m3), which is also a day higher concentrations were
measured at TOOK and TMOK. Of the 18 acetaldehyde concentrations greater than
2 |ig/m3 measured at TROK, 11 were measured during the third quarter while three or
less were measured during the other calendar quarters.
•	The first quarter benzene concentration for TROK is higher than the other quarterly
averages and has a relatively large confidence interval associated with it. The benzene
concentration measured at TROK on January 16, 2013 (7.43 |ig/m3) is more than
three times greater than the next highest benzene concentration measured at TROK
(2.31 |ig/m3, measured on December 18, 2013). The maximum concentration
measured at TROK is the third highest benzene concentration measured across the
program.
•	Several of the lowest quarterly average concentrations of the VOC pollutants of
interest for TROK (benzene and 1,3-butadiene in particular) were calculated for the
second quarter of 2013, similar to TMOK. In addition, these averages generally
exhibit the least variability. A review of the benzene data shows that the range of
measurements was smallest for the second quarter, that the median concentration was
the lowest for the second quarter, and that the second quarter of 2013 had the fewest
benzene concentrations greater than 1 |ig/m3 (only two, compared to five for the first
quarter, eight for the third, and four for the fourth). Similar observations can be made
for 1,3-butadiene, where no concentrations greater than 0.1 |ig/m3 were measured at
TROK during the second quarter (or third quarter) while four were measured during
the first quarter and seven were measured during the fourth quarter. There were three
concentrations measured during the third quarter that are greater than the maximum
concentration measured during the second quarter of 2013.
Observations for OCOK from Table 22-5 include the following:
• Similar to the Tulsa sites, the second and third quarter average concentrations of
formaldehyde are higher than the other quarterly averages for OCOK, and although
the third quarter average concentration is the highest, the confidence interval is
highest for the second quarter. A review of the data shows that the two formaldehyde
concentrations greater than 6 |ig/m3 were measured at OCOK during the third quarter,
while the next three highest were all measured in June (second quarter). Of the 13
formaldehyde concentrations greater than 4 |ig/m3 measured at OCOK, three were
measured in June and nine were measured during the third quarter (with the
additional one in the fourth quarter). No formaldehyde concentrations greater than
4 |ig/m3 were measured before June.
22-38

-------
•	The maximum acetaldehyde concentration was measured at OCOK on the same day
as the maximum formaldehyde concentration (September 7, 2013). Acetaldehyde
concentrations greater than 3 |ig/m3 were all measured between June and October.
•	The confidence interval associated with fourth quarter average benzene concentration
for OCOK is similar in magnitude to the average itself, indicating that outliers may be
affecting this average. A review of the data shows that the benzene concentration
measured at OCOK on November 6, 2013 (9.38 |ig/m3) is nearly six times greater
than the next highest benzene concentration measured at OCOK (1.81 |ig/m3,
measured on January 4, 2013). This is the second highest benzene concentration
measured across the program.
•	Concentrations of 1,3-butadiene appear lowest during the second quarter of 2013,
similar to the Tulsa sites. No concentrations greater than 0.1 |ig/m3 were measured
during the second quarter, while between one and three were measured in the other
quarters. In addition, the second quarter has the highest number of non-detects at
seven, although this is the same number measured during the fourth quarter.
•	The quarterly average concentrations of arsenic are highest during the warmer months
and have a higher level of variability associated with them. The only arsenic
concentration greater than 1 ng/m3 was measured at OCOK on July 3, 2013
(1.03 ng/m3) and seven of the eight arsenic concentrations greater than 0.75 ng/m3
were measured between April and September.
Observations for ADOK and YUOK from Table 22-5 include the following:
•	The instrumentation at ADOK was moved to YUOK mid-year; as a result, ADOK has
only first and second quarter averages, YUOK has only third and fourth quarter
averages, and neither site has annual averages in Table 22-5. However, site-specific
statistical summaries for all pollutants sampled for at these sites for the time frame of
sampling are available in Appendices J, L, and N.
•	Formaldehyde and acetaldehyde have the highest quarterly averages among the
pollutants of interest for these two sites, with the highest of the two available
quarterly averages calculated for the quarter including samples collected during the
warmer months of the year. Of the VOCs, carbon tetrachloride and benzene have the
highest quarterly averages for these two sites.
Tables 4-9 through 4-12 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for the
Oklahoma sites include the following:
•	The Oklahoma sites appear in Tables 4-9 through 4-12 a total of 19 times. However,
because they are the only sites sampling TSP metals, four of the sites appear for each
metal, accounting for eight of the appearances.
22-39

-------
All three Tulsa sites appear in Table 4-9 among the sites with the highest annual
average concentrations of ethylbenzene, with TOOK ranking seventh, TMOK ranking
eighth, and TROK ranking 10th.
•	TOOK has the sixth highest annual average of concentration of benzene. OCOK
appears in Table 4-9 for hexachloro-1,3-butadiene, ranking fourth highest. TMOK
appears for /;-dichlorobenzene, ranking third highest.
•	TOOK and TMOK rank eighth and 10th, respectively for their annual average
concentrations of acetaldehyde. Only TOOK appears in Table 4-10 for formaldehyde
(ranking ninth).
•	The Tulsa sites rank higher than OCOK for the two TSP metals shown in Table 4-12.
TROK has the highest annual average arsenic concentration among the Oklahoma
sites while TOOK highest annual average nickel concentration among the Oklahoma
sites.
22.4.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for the pollutants shaded in
gray in Table 22-4 for the three Tulsa sites and OCOK. Figures 22-15 through 22-26 overlay
these sites' minimum, annual average, and maximum concentrations onto the program-level
minimum, first quartile, median, average, third quartile, and maximum concentrations, as
described in Section 3.4.3.1. Figures 22-15 through 22-26 and their associated observations are
as follows:
22-40

-------
Figure 22-15. Program vs. Site-Specific Average Acetaldehyde Concentrations
¦
-o	¦
¦
-0	¦
¦
)	'
¦
+—'
0	3	6	9	12	15
Concentration {[ig/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


• Figure 22-15 presents the box plots for acetaldehyde for the Tulsa sites and
OCOK. The range of acetaldehyde concentrations measured at TOOK was similar
to the range measured at TMOK, while the range of concentrations measured at
TROK and OCOK were slightly smaller. The annual average concentrations for
these four sites fall between the program-level second and third quartiles, with the
annual averages for TOOK, TMOK, and OCOK slightly greater than the
program-level average concentration and the annual average for TROK slightly
less than the program-level average.
22-41

-------
Figure 22-16. Program vs. Site-Specific Average Arsenic (TSP) Concentrations
TOOK
TMOK
TROK
OCOK
Concentration {ng/m3)
Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



• Because the Oklahoma sites are the only sites sampling TSP metals, Figure 22-16
compares the individual Oklahoma site arsenic data against the combined
Oklahoma data. Figure 22-16 shows that the maximum arsenic concentration
among the Oklahoma sites was measured at TROK. The annual average arsenic
(TSP) concentration is greatest for TROK (although the annual average for
TOOK is similar) and lowest for OCOK. This figure also shows that arsenic
concentrations were higher at the Tulsa sites, based on the range of measurements
as well as the annual average concentrations.
22-42

-------
Figure 22-17. Program vs. Site-Specific Average Benzene Concentrations
I
Program Max Concentration = 43.5 ^ig/m3
Ho
Program Max Concentration = 43.5 ^ig/m3
Ho
Program Max Concentration = 43.5 ^ig/m3
H
Program Max Concentration = 43.5 ^ig/m3
4	6
Concentration {[jg/m3
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range


• Figure 22-17 presents the box plots for benzene. Note that the program-level
maximum concentration (43.5 |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 12 |ig/m3. The box plots show that the smallest range of
benzene concentrations was measured at TMOK while the largest range was
measured at OCOK. The maximum benzene concentration measured at OCOK is
the second highest benzene concentration measured across the program, yet this
site has the lowest annual average concentration of this pollutant compared to the
other Oklahoma sites. The annual average concentrations of benzene for the Tulsa
sites are greater than the program-level average concentration while the annual
average for OCOK is similar to it. The minimum concentration measured at
TROK is similar to the program-level first quartile concentration, while the
minimum benzene concentrations measured at TMOK and TROK are just less
than program-level first quartile.
22-43

-------
Figure 22-18. Program vs. Site-Specific Average 1,3-Butadiene Concentrations

*
Program Max Concentration = 21.5 ^ig/m3
1
Program Max Concentration = 21.5 ^ig/m3
h+:
Program Max Concentration = 21.5 ^ig/m3
Program Max Concentration = 21.5 ^ig/m3
0.6	0.9
Concentration {[jg/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range


• Figure 22-18 presents the box plots for 1,3-butadiene. Note that the program-level
maximum concentration (21.5 |ig/m3) is not shown directly on the box plots
because the scale of the box plots would be too large to readily observe data
points at the lower end of the concentration range. Thus, the scale of the box plots
has been reduced to 1.5 |ig/m3. Note that the program-level average concentration
is greater than the program-level third quartile, indicating that the concentrations
on the upper end of the range of measurements are driving the program-level
average. All of the annual average concentrations of 1,3-butadiene for the
Oklahoma sites are less than the program-level average concentration. The annual
average concentration of 1,3-butadiene is highest for TMOK and lowest for
OCOK, although the range of measurements is smallest for TROK.
22-44

-------
Figure 22-19. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations
TOOK
Program Max Concentration = 23.7 ^ig/m3
TMOK
Program Max Concentration = 23.7 ^ig/m3
TROK
Program Max Concentration = 23.7 ^ig/m3
Program Max Concentration = 23.7 ^ig/m3
OCOK
0	0.25	0.5	0.75	1	1.25	1.5	1.75	2
Concentration {[ig/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


• Figure 22-19 presents the box plots for carbon tetrachloride. Similar to other
VOCs, the program-level maximum concentration (23.7 |ig/m3) is not shown
directly on the box plots for carbon tetrachloride as the scale of the box plots 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. The range of carbon tetrachloride
concentrations measured at each site was largest for TROK and smallest for
OCOK. The annual average concentrations of carbon tetrachloride did not vary
much across the Oklahoma sites, ranging between 0.60 |ig/m3 and 0.65 |ig/m3 for
each site, all of which are just less than the program level average concentration
of 0.66 |ig/m3.
22-45

-------
Figure 22-20. Program vs. Site-Specific Average /7-Dichlorobenzene Concentrations

0
0.1
0.2
0.3 0.4
Concentration {[jg/m3)
0.5
0.6

Program:
1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i

Site:
Site Average
o
Site Concentration Range


• Figure 22-20 presents the box plots for p-dichlorobenzene for TMOK and OCOK,
the only Oklahoma sites for which this is a pollutant of interest. Note that the first
and second quartiles are both zero for this pollutant, indicating that at least half of
the measurements are non-detects and thus, are not visible on the box plots. The
maximum concentration measured at TMOK is twice the maximum concentration
measured at OCOK, although both are less than the maximum concentration
measured across the program. The annual average p-dichlorobenzene
concentration for TMOK is more than twice the program-level average
concentration while the annual average for OCOK is similar to it. The number of
non-detects measured at OCOK (22) is two and half times greater than the
number measured at TMOK (9).
22-46

-------
Figure 22-21. Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations
¦
Program Max Concentration = 111 ^ig/m3
¦
Program Max Concentration = 111 ^ig/m3
¦
Program Max Concentration = 111 ^ig/m3




Program Max Concentration = 111 ^ig/m3
,
0
0
1
	1	1	1	
0.4	0.6
Concentration {[jg/m3)
Program: 1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range


• Figure 22-21 presents the box plots for 1,2-dichloroethane. Similar to other
VOCs, the program-level maximum concentration (111 |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 being driven by the higher measurements
collected at a few monitoring sites. Figure 22-21 shows that the entire range of
1,2-dichloroethane concentrations measured at the Oklahoma sites are less than
the average concentration across the program. The annual average concentrations
of 1,2-dichloroethane for these sites are all less than the program-level median
concentration, with the annual averages for the Tulsa sites similar to the program-
level first quartile.
22-47

-------
Figure 22-22. Program vs. Site-Specific Average Ethylbenzene Concentrations
F=
Program Max Concentration = 18.7 ^ig/m3

Program Max Concentration = 18.7 ^ig/m3
E
Program Max Concentration = 18.7 ^ig/m3
Fh
Program Max Concentration = 18.7 ^ig/m3
0
1
2 3
Concentration {[jg/m3)
4
5

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i

Site: Site Average
o
Site Concentration Range


• Figure 22-22 presents the box plots for ethylbenzene. The scale of these box plots
has also been reduced to allow for the observation of data points at the lower end
of the concentration range. The range of ethylbenzene concentrations measured is
largest for TMOK and smallest for OCOK. The annual average concentrations for
the Tulsa sites are greater than the program-level average concentration while the
annual average concentration for OCOK is closer to the program-level median
concentration.
22-48

-------
Figure 22-23. Program vs. Site-Specific Average Formaldehyde Concentrations
	
¦
¦o	¦
¦ >
	1


3
6
9 12 15
Concentration {[ig/m3)
18
21
Program:
IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i
Site:
Site Average
o
Site Concentration Range


• Figure 22-23 presents the box plots for formaldehyde for the Tulsa sites and
OCOK. The range of formaldehyde concentrations measured at these sites is in
descending order from top to bottom in Figure 22-23, as are the annual average
concentrations, which vary little from the program-level average concentrations.
22-49

-------
Figure 22-24. Program vs. Site-Specific Average Hexachloro-l,3-Butadiene Concentrations
TOOK
o
0
0.05
0.1
0.15
Concentration {[jg/m3)
0.2
0.25
0.3

Program:
Site:
1st Quartile
¦
Site Average
o
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
Site Concentration Range
4th Quartile
~
Average
i

• Figure 22-24 presents the box plots for hexchloro-1,3-butadiene for the Oklahoma
sites. Note that the first, second, and third quartiles for hexchloro-l,3-butadiene
are zero at the program-level and therefore not visible on the box plots due to the
large number of non-detects. For these sites, roughly one-quarter or less of the
measurements of this pollutant were measured detections, although none were
greater than the MDL. OCOK has the highest annual average hexchloro-1,3-
butadiene concentration while TROK has the lowest, though less than 0.01 |ig/m3
separates them.
22-50

-------
Figure 22-25. Program vs. Site-Specific Average Manganese (TSP) Concentration
10
20
30 40 50
Concentration {ng/m3)
60
70
Program:
1st Quartile
¦
2nd Quartile
~
3rd Quartile
~
4th Quartile
~
Average
i
Site:
Site Average
o
Site Concentration Range


• Figure 22-25 presents the manganese data for TOOK, the only Oklahoma site for
which manganese is a pollutant of interest. Because the Oklahoma sites are the
only sites sampling TSP metals, Figure 22-25 compares the arsenic measurements
collected at TOOK against the combined Oklahoma data. Figure 22-25 shows that
the maximum manganese concentration among the Oklahoma sites was measured
at TOOK. The annual average manganese concentration for TOOK is greater than
the program-level manganese concentration and third quartile (TSP only).
Figure 22-26. Program vs. Site-Specific Average Nickel (TSP) Concentrations
n—

0	2	4	6	8	10	12
Concentration {ng/m3)
Program: 1st Quartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Quartile
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



• Figure 22-26 presents the nickel concentration data for the two Oklahoma sites for
which nickel was identified as a pollutant of interest. The maximum concentration
of nickel among the Oklahoma sites was measured at TOOK while the minimum
nickel concentration measured at TOOK is greater than the program-level first
quartile (TSP only). The range of nickel measurements is greater for TOOK than
TROK. The annual average nickel concentration for TOOK is nearly 50 percent
higher than the annual average concentration for TROK.
22-51

-------
22.4.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.3.2.
TOOK has sampled TSP metals, carbonyl compounds, and VOCs under the NMP since 2006 and
TMOK and OCOK have sampled these pollutants since 2009. Thus, Figures 22-27 through
22-57 present the 1-year statistical metrics for each of the pollutants of interest first for TOOK,
followed by TMOK and OCOK. The statistical metrics presented for assessing trends include the
substitution of zeros for non-detects. If sampling began mid-year, a minimum of 6 months of
sampling is required for inclusion in the trends analysis; in these cases, a 1-year average
concentration is not provided, although the range and percentiles are still presented.
22-52

-------
Figure 22-27. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at TOOK
20061	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	- Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because issues at the onset of sampling resulted in low completeness.
Observations from Figure 22-27 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 concentration is not presented for 2006,
although the range of measurements is provided.
•	The maximum concentration of acetaldehyde was measured in 2011 (8.95 |ig/m3),
although a similar concentration was also measured in 2012 (8.59 |ig/m3). The 12
highest concentrations were all measured in 2011 or 2012. Of the 33 acetaldehyde
concentrations greater than 4 |ig/m3 measured at TOOK, 12 were measured in 2012,
eight were measured in 2011, five were measured in 2010, and three or fewer were
measured in all other years.
•	The statistical metrics exhibit an increasing trend between 2009 and 2011. The 95th
percentiles 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 while a
significant decrease in acetaldehyde concentrations is shown for 2013.
22-53

-------
Figure 22-28. Yearly Statistical Metrics for Arsenic (TSP) Concentrations Measured at TOOK
I
-2-1 LjJ	2	2
2007
2008
2009 2010 2011
Year
2012
2013
O 5th Percentile
- Minimum
— Median - Maximum
O 95th Percentile

Observations from Figure 22-28 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 22-28 excludes data from 2006 per the criteria specified in
Section 3.4.3.2.
•	The two highest concentrations of arsenic were measured at TOOK in September
2007 and are the only two concentrations greater than 4 ng/m3 measured at TOOK.
All eight concentrations of arsenic greater than 2 ng/m3 were measured in either 2007
or 2008.
•	The 1-year average and median concentrations exhibit a decreasing trend between
2007 and 2010, although the difference is relatively small between 2009 and 2010.
The 1-year average and median concentrations increased for 2011, an increase that
continued into 2012.
•	The smallest range of arsenic concentrations was measured at TOOK in 2013. The
difference between the 1-year average and median concentrations is at a minimum for
2013. Both of these indicate a decreasing level of variability in the arsenic
measurements collected in 2013. All of the statistical parameters exhibit decreases
from 2012 to 2013.
22-54

-------
Figure 22-29. Yearly Statistical Metrics for Benzene Concentrations Measured at TOOK
20061	2007	2008	2009	2010	2011	2012	2013
O 5th Percentile	- Minimum	— Median	- Maximum	O 95th Percentile •*~*$~~~ Average
1 A 1-year average is not presented because issues at the onset of sampling resulted in low completeness.
Observations from Figure 22-29 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 concentration 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). All four
of the benzene concentrations greater than 10 |ig/m3 measured at TOOK were
measured in 2011. The 95th percentile for 2011 is greater than the maximum
concentration for each of the other years shown.
•	The 1-year average benzene concentration has fluctuated over the years. After a
significant decrease from 2008 to 2009, an increasing trend through 2011 occurred.
After 2011, a significant decrease in benzene concentrations is exhibited, particularly
for 2013. All of the statistical parameters are at a minimum for 2013. The maximum
concentration measured in 2013 is less than the 95th percentiles for all previous years
and is less than the 1-year average concentration for 2011.
22-55

-------
Figure 22-30. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at TOOK
0.35
0.30
0.25
£
1
•| 0.20
c

-------
Figure 22-31. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
TOOK
2006 1	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because issues at the onset of sampling resulted in low completeness.
Observations from Figure 22-31 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). Four additional concentrations greater than 1 |ig/m3
have been measured at TOOK.
•	With the exception of 2011, the range of carbon tetrachloride measurements spans
roughly 1 |ig/m3 or less. The range of measurements is at a minimum for 2012, when
the difference between the minimum and maximum concentrations is less than
0.45 |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 average
concentrations range from 0.61 |ig/m3 to 0.63 |ig/m3. A slight increase is shown for
2012 (0.66 |ig/m3), even though the measurements for this year exhibit the least
variability. For 2013, the majority of concentrations fall into a similar range as 2012,
although the range of concentrations measured widened. Across the years of
sampling, the 1-year average concentration of carbon tetrachloride has varied by only
0.10 |ig/m3.
22-57

-------
• For each year shown, the 1-year average concentration is very similar to the median
concentration. The difference between these two parameters is greatest for 2009, yet
only 0.02 |ig/m3 separates them. This indicates that there is relatively little variability
in the central tendency of these measurements.
Figure 22-32. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
TOOK
0.40
0.35
0.30
0.25
£
1
I °-20
2006
2007
2008
2009
2010
2011
2012
2013
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because issues at the onset of sampling resulted in low completeness.
Observations from Figure 22-32 for 1,2-dichloroethane measurements collected at TOOK
include the following:
•	The median concentration for all years through 2011 is zero, indicating that at least
half of the measurements were non-detects. In 2006, there was one measured
detection of 1,2-dichloroethane. In 2007 and 2008 there were none. Between 2009
and 2011, the number of measured detections varied from five to six. The number of
measured detections increased significantly for 2012, up from six in 2011 to 38 in
2012. Greater than 30 measured detections were measured in 2013 as well.
•	The 1-year average concentration for 2012 is less than the median concentration,
which is a little unusual. The 1-year average concentration is more susceptible to
outliers (on either end of the concentration range) than the median concentration,
which represents the midpoint of a group of measurements. The 1-year average
concentration for 2012 is less than the median, indicating that concentrations on the
lower end of the concentration range (the many zeroes representing non-detects) are
22-58

-------
pulling the average down (just like a maximum or outlier concentration can drive the
average upward). This is also true for 2013, although the difference between the two
statistical parameters is less.
Figure 22-33. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at TOOK
2006 1	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because issues at the onset of sampling resulted in low completeness.
Observations from Figure 22-33 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 concentration would still exhibit a slight increase. The number of
ethylbenzene concentrations greater than 1 |ig/m3 nearly doubled from 2007 (seven)
to 2008 (13).
22-59

-------
•	Most of the statistical parameters are at a minimum for 2009. The 1-year average and
median concentrations decreased by more than half from 2008 to 2009. There were
no ethylbenzene concentrations greater than 1 |ig/m3 measured at TOOK in 2009.
•	After 2009, concentrations of ethylbenzene measured at TOOK exhibit a significant
increasing trend through 2012. The 95th percentile, 1-year average concentration, and
the median concentration are all at a maximum for 2012. The 95th percentile for 2012
is greater than the maximum concentration for all other years except 2008. The 1-year
average concentration for 2012 is approaching 1 |ig/m3.
•	Ethylbenzene concentrations measured in 2013 decreased significantly from 2012, as
the 1-year average concentration decreased by more than half.
Figure 22-34. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at TOOK
20061	2007
2009	2010
Year
O 5th Percentile
Maximum	O 95th Percentile
1 A 1-year average is not presented because issues at the onset of sampling resulted in low completeness.
Observations from Figure 22-34 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 (10.2 |ig/m3 measured in 2011).
•	All but two of the 79 formaldehyde concentrations greater than 5 |ig/m3 were
measured during the period between May and September, regardless of year.
22-60

-------
•	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. Further decreases are shown
for all of the statistical parameters for 2013.
Figure 22-35. Yearly Statistical Metrics for Hexachloro-l,3-Butadiene Concentrations
Measured at TOOK
ST 0.15
o
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because issues at the onset of sampling resulted in low completeness.
Observations from Figure 22-35 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 several
years of sampling at TOOK.
•	The median concentration is zero for all years of sampling, indicating that at least half
of the measurements were non-detects for each year. Between 2006 and 2010, there
were a total of four measured detections. In 2011, five measured detections were
reported. This number doubled for 2012 and is at a maximum for 2013 (13).
22-61

-------
Figure 22-36. Yearly Statistical Metrics for Manganese (TSP) Concentrations Measured at
TOOK
300
250
200
E
"m
c
0 -I	1	1	1	1	1	1	
2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
Observations from Figure 22-36 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 a dust storm (October 18, 2012). Measurements greater than 100 ng/m3
were also measured in 2007 (131 ng/m3) and 2011 (104 ng/m3).
•	A decreasing trend in the concentrations is shown through 2009, which was followed
by an increasing trend through 2012. Even if the maximum concentration measured in
2012 was excluded from the calculations, the 1-year average and median
concentrations would still exhibit an increasing trend for 2012. This is because there
were more concentrations at the upper end of the concentration range for 2012 (the
number of manganese measurements greater than 50 ng/m3 increased from 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 20 ng/m3
decreased from 17 in 2011 to 11 in 2012).
22-62

-------
Figure 22-37. Yearly Statistical Metrics for Nickel (TSP) Concentrations Measured at TOOK
2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	- Minimum	— Median	- Maximum	O 95th Percentile	Average
Observations from Figure 22-37 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 a dust storm).
Only two nickel concentrations greater than 10 ng/m3 have been measured at TOOK,
with the other on July 3, 2013 (11.0 ng/m3). Six of the eight nickel concentrations
greater than 5 ng/m3 were measured at TOOK in either 2012 or 2013 (with the other
two in 2007).
•	A significant decreasing trend in the nickel concentrations measured at TOOK is
shown through 2009. A slight increase is shown for 2010, which was followed by
significant increases for 2011 and 2012. The minimum concentration shown for 2012
is greater than the 5th percentile for the four previous years.
•	All of the statistical parameters exhibit slight decreases for 2013.
22-63

-------
Figure 22-38. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at TMOK
20091
2010
2011
2012
2013


Year


O 5th Percentile
- Minimum
— Median - Maximum
O 95th Percentile

1 A 1-year average is not presented because sampling under the NMP did not begin until April 2009.
Observations from Figure 22-38 for acetaldehyde measurements collected at TMOK
include the following:
•	Sampling for carbonyl compounds began at TMOK under the NMP in April 2009. A
1-year average concentration is not presented for 2009 because a full year's worth of
data is not available, although the range of measurements is provided.
•	The maximum acetaldehyde concentration (7.00 |ig/m3) was measured at TMOK on
August 19, 2011. All seven acetaldehyde concentrations greater than 5 |ig/m3 were
measured in either 2011 or 2012.
The range of acetaldehyde concentrations measured increased considerably from
2010 to 2011, after which the range of measurements has decreased each year.
Although a decreasing trend is shown in the 1-year average and median
concentrations between 2011 and 2013, the difference is not statistically significant.
22-64

-------
Figure 22-39. Yearly Statistical Metrics for Arsenic (TSP) Concentrations Measured at TMOK
3.0	
2009 1
2010
2011
2012 2013


Year

O 5th Percentile
- Minimum
— Median - Maximum
O 95th Percentile •*~*#~~~ Average
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2009.
Observations from Figure 22-39 for arsenic (TSP) measurements collected at TMOK
include the following:
•	Sampling for TSP metals began at TMOK under the NMP in April 2009. A 1-year
average concentration is not presented for 2009 because a full year's worth of data is
not available, although the range of measurements is provided.
•	The three highest arsenic concentrations measured at TMOK were all measured in
2009 and all but one of the six arsenic concentrations greater than 2 ng/m3 were
measured in 2009. The entire range of concentrations measured in other years is less
than the 95th percentile for 2009 and the median concentration is at a maximum for
2009.
•	Although a slight increasing trend is shown between 2010 and 2012, the changes in
the 1-year average concentrations are not statistically significant.
•	The smallest range of arsenic concentrations was measured in 2013. The difference
between the 1-year average and median concentrations is at a minimum for 2013, as
less than 0.01 ng/m3 separates them. Both of these indicate a decreasing level of
variability in the arsenic measurements for 2013. A similar observation was made for
TOOK.
22-65

-------
Figure 22-40. Yearly Statistical Metrics for Benzene Concentrations Measured at TMOK
~ 2.0
3.0	
2009 1
2010
2011
2012
2013


Year


O 5th Percentile
- Minimum
— Median - Maximum
O 95th Percentile

1 A 1-year average is not presented because sampling under the NMP did not begin until April 2009.
Observations from Figure 22-40 for benzene measurements collected at TMOK include
the following:
•	Sampling for VOCs began at TMOK under the NMP in April 2009. A 1-year average
concentration is not presented for 2009 because a full year's worth of data is not
available, although the range of measurements is provided.
•	The maximum benzene concentration (3.91 |ig/m3) was measured at TMOK on
May 7, 2009, although benzene concentrations greater than 3 |ig/m3 have been
measured in all years of sampling except 2013.
•	The 1-year average and median benzene concentrations have a significant decreasing
trend between 2010 and 2013, with the largest decrease shown for 2013. The 1-year
average and median concentrations have both decreased by more than 0.5 |ig/m3 since
the onset of sampling. The maximum concentration measured in 2013 is less than the
95th percentile for all the previous years of sampling.
22-66

-------
Figure 22-41. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at TMOK
2011
Year
O 5th Percentile
— Minimum
— Median
— Maximum
O 95th Percentile


1 A 1-year average is not presented because sampling under the NMP did not begin until April 2009.
Observations from Figure 22-41 for 1,3-butadiene measurements collected at TMOK
include the following:
•	The range of 1,3-butadiene concentrations measured at TMOK is at a minimum for
2009, with all concentrations measured spanning less than 0.2 |ig/m3, with the range
of concentrations measured increasing each year through 2012.
•	Even with the increasing range of measurements, the 1-year average concentration of
1,3 butadiene has varied by less than 0.025 |ig/m3 across the years of sampling.
•	The number of non-detects has varied across the years of sampling, from a few as
none (2009) to as many as nine (2011).
22-67

-------
Figure 22-42. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
TMOK
.25
.00
.75
.50
.25
.00
1
2009
2010
2011
2012
2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2009.
Observations from Figure 22-42 for carbon tetrachloride measurements collected at TMOK
include the following:
•	The maximum carbon tetrachloride concentration was measured on August 17, 2009
and is the only concentration greater than 1 |ig/m3 measured at TMOK.
•	All of the statistical parameters exhibit decreases from 2009 to 2010, with little
change in the carbon tetrachloride measurements at TMOK shown from 2010 to
2011.
•	Even though the range of concentrations measured decreased considerably for 2012,
all of the statistical parameters exhibit increases. The 1-year average and median
concentrations are equivalent for 2012 (0.68 |ig/m3), which represents a statistically
significant increase from 2011.
•	All of the statistical parameters exhibit decreases again from 2012 to 2013.
•	All of the 1-year average carbon tetrachloride concentrations shown fall between
0.60 |ig/m3 and 0.70 |ig/m3.
22-68

-------
Figure 22-43. Yearly Statistical Metrics for o-Dichlorobenzene Concentrations Measured at
TMOK
0.8
0.7
0.6
3.0	
20091
2010
2011
Year
2012
2013
O 5th Percentile
— Minimum
— Median - Maximum
O 95th Percentile
.

1 A 1-year average is not presented because sampling under the NMP did not begin until April 2009.
Observations from Figure 22-43 for p-dichlorobenzene measurements collected at
TMOK include the following:
•	The maximum /;-dichlorobenzene concentration was measured on June 30, 2009
(0.75 |ig/m3) and is the only measurement greater than 0.5 |ig/m3 measured at
TMOK.
•	The range of concentrations measured at TMOK has decreased every year through
2012. Most of the statistical parameters exhibit a decrease from 2009 to 2010, with
many exhibiting further decreases for 2011.
•	The 1-year average concentration decreased significantly from 2010 to 2011 with
little change shown from 2011 to 2012. Although an increase in the 1-year average
concentration is shown for 2013, the difference is not statistically significant.
•	The minimum and 5th percentile are both zero for all years of sampling except 2009,
indicating the presence of non-detects. Aside from 2009, the number of non-detects
has ranged from as few as four in 2010 to as many as 13 in 2011.
22-69

-------
Figure 22-44. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
TMOK
0.15
2009 1	2010	2011	2012	2013
0 5th Percentile	— Minimum	— Median	- Maximum	0 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2009.
Observations from Figure 22-44 for 1,2-dichloroethane measurements collected at
TMOK include the following:
•	The median concentration for 2009, 2010, and 2011 is zero, indicating that at least
half of the measurements were non-detects. In 2009, there were three measured
detections of 1,2-dichloroethane. In 2010 and 2011, there were 10. For 2012, the
number of measured detections increased by a factor of four. Measured detections
also accounted for more than half of measurements in 2013.
•	The 1-year average concentration for 2012 is less than the median concentration,
which is similar to what was shown for TOOK in 2012. The 1-year average
concentration is more susceptible to outliers (on either end of the concentration range)
than the median concentration. The 1-year average concentration for 2012 is less than
the median, indicating that concentrations on the lower end of the concentration range
are pulling the 1-year average down (just like a maximum or outlier concentration can
drive the average upward). This is also true for 2013, although the difference between
the two statistical parameters is less.
22-70

-------
Figure 22-45. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at TMOK
4.0
3.5
3.0
2.5
£
3.0	
20091
2010
2011
2012
2013


Year


O 5th Percentile
- Minimum
— Median - Maximum
O 95th Percentile

1 A 1-year average is not presented because sampling under the NMP did not begin until April 2009.
Observations from Figure 22-45 for ethylbenzene measurements collected at TMOK
include the following:
•	The maximum ethylbenzene concentration was measured in 2010 (3.63 |ig/m3) and is
the only measurement greater than 2 |ig/m3 measured at TMOK.
•	Despite the fluctuation in the maximum concentrations shown between 2010 and
2012,	little change is shown for most of the statistical parameters. Less than
0.05 |ig/m3 separates the median concentrations for these years and roughly
0.01 |ig/m3 separates the 1-year average concentrations during this period.
•	With the exception of the maximum concentration, all of the statistical parameters
exhibit decreases for 2013. The 1-year average concentration is at a minimum for
2013,	although confidence intervals calculated for these averages indicate that the
decrease from 2012 to 2013 is not statistically significant.
22-71

-------
Figure 22-46. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
TMOK
12
0 H—
2009 1
2010
2011
Year
2012
2013

O 5th Percentile
— Minimum
— Median - Maximum
O 95th Percentile

1 A 1-year average is not presented because sampling under the NMP did not begin until April 2009.
Observations from Figure 22-46 for formaldehyde measurements collected at TMOK
include the following:
•	The maximum formaldehyde concentration was measured on August 19, 2011
(10.8 |ig/m3), the same date that the maximum acetaldehyde concentration was
measured at TMOK. Two additional formaldehyde concentrations greater than
10 |ig/m3 were measured at TMOK in 2012.
•	The 1-year average concentration increased from 2010 to 2011, then has a decreasing
trend through 2013, when the 1-year average is at a minimum. However, these
changes are not statistically significant. The 1-year average concentrations have
ranged from 3.19 |ig/m3 (2013) to 3.94 |ig/m3 (2011). The median concentration is
also at a minimum for 2013, ranging from 2.63 |ig/m3 (2013) to 3.09 |ig/m3 (2012)
across the years of sampling.
•	The 1-year average concentrations for formaldehyde exhibit a similar pattern as the
1-year average concentrations for acetaldehyde for TMOK.
22-72

-------
Figure 22-47. Yearly Statistical Metrics for Hexachloro-l,3-Butadiene Concentrations
Measured at TMOK
ST 0.12
2009 1	2010	2011	2012	2013
Year
0 5th Percentile	— Minimum	— Median	- Maximum	0 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2009.
Observations from Figure 22-47 for hexachloro-l,3-butadiene measurements collected at
TMOK include the following:
•	There were few measured detections in the first few years of sampling at TMOK. The
median concentration is zero for all years of sampling, indicating that at least half of
the measurements were non-detects for each year. There were no measured detections
in 2009, two in 2010, three in2011,ninein2012, and 14 in 2013.
•	For 2013, more than three-quarters of the measurements are still non-detects and all
of the measured detections are less than the MDL.
22-73

-------
Figure 22-48. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at OCOK
20091	2010	2011	2012	2013
Year
O 5th Percentile	- Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2009.
Observations from Figure 22-48 for acetaldehyde measurements collected at OCOK
include the following:
•	Sampling for carbonyl compounds began at OCOK under the NMP in May 2009. A
1-year average concentration is not presented for 2009 because a full year's worth of
data is not available, although the range of measurements is provided.
•	The maximum acetaldehyde concentration was measured on May 9, 2011
(6.68 |ig/m3). Only one additional acetaldehyde concentration greater than 6 |ig/m3
has been measured at OCOK (6.16 |ig/m3 in 2012).
•	The smallest range of acetaldehyde concentrations was measured in 2009, after which
the range of measurements increased considerably. The 1-year average concentration
increased significantly from 2010 to 2011, with the median concentration exhibiting a
similar increase. Little change is shown from 2011 to 2012.
•	All of the statistical parameters exhibit decreases from 2012 to 2013.
22-74

-------
Figure 22-49. Yearly Statistical Metrics for Arsenic (TSP) Concentrations Measured at OCOK
1-
2011
Year
0 5th Percentile
0 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2009.
Observations from Figure 22-49 for arsenic (TSP) measurements collected at OCOK
include the following:
•	Sampling for TSP metals began at OCOK under the NMP in May 2009. A 1-year
average concentration is not presented for 2009 because a full year's worth of data is
not available, although the range of measurements is provided.
•	The maximum concentration of arsenic was measured at OCOK in 2009 (3.11 ng/m3).
The maximum concentration measured after 2009 has been steadily decreasing,
reaching a minimum for 2013 (1.03 ng/m3). At the same time, the minimum
concentration measured each year has been steadily increasing, reaching a maximum
in 2012 (0.21 ng/m3).
•	Although a slight increasing trend is shown in the 1-year average concentrations
between 2010 and 2012 and is followed by a slight decrease for 2013, the 1-year
average concentrations have varied by less than 0.15 ng/m3, ranging from 0.44 ng/m3
(2010) to 0.57 ng/m3 (2012).
22-75

-------
Figure 22-50. Yearly Statistical Metrics for Benzene Concentrations Measured at OCOK
2009
2010
2011
2012
2013
Maximum
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2009.
Observations from Figure 22-50 for benzene measurements collected at OCOK include
the following:
•	Sampling for VOCs began at OCOK under the NMP in May 2009. A 1-year average
concentration is not presented for 2009 because a full year's worth of data is not
available, although the range of measurements is provided.
•	The maximum benzene concentration was measured at OCOK on November 6, 2013
(9.38 |ig/m3). The next highest concentration was measured in 2011 (6.80 |ig/m3). No
other benzene concentrations greater than 3 |ig/m3 have been measured at OCOK.
•	Although the range of benzene concentrations measured at OCOK is highly variable,
the majority of concentrations (as indicated by the 5th and 95th percentiles) are
falling into a tighter range each year (excluding 2009).
•	Even though the maximum benzene concentration was measured at OCOK in 2013,
both the 1-year average and median concentrations are at a minimum for 2013.
•	A decreasing trend in the 1-year average concentrations is shown between 2010 and
2013, although the variability in the measurements results in relatively large
confidence intervals, reducing the statistical significance of the changes.
22-76

-------
Figure 22-51. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at OCOK
.25
Maximum
Concentration for
2011 is 9.52 M-g/m-
.00
.75
.50
.25
.00
l
2009
2010
2011
2012
2013
Year
O 5th Percentile	— Minimum	— Median	— Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2009.
Observations from Figure 22-51 for 1,3-butadiene measurements collected at OCOK
include the following:
•	The maximum 1,3-butadiene concentration was measured at OCOK on
September 18, 2011 (9.52 |ig/m3). The next highest concentration was measured in
2012 (1.09 |ig/m3). No other 1,3-butadiene concentrations greater than 0.5 |ig/m3
have been measured at OCOK.
•	The range of concentrations measured at OCOK increased exponentially between
2009 and 2011. Although the range within which the majority of concentrations fall
increased as well, as indicated by the difference between the 5th and 95th percentiles,
the change is less dramatic.
•	The 1-year average concentration for 2011 is being driven by the outlier; if this
measurement was excluded from the calculation, the 1-year average concentration
would decrease from 0.20 |ig/m3 to 0.05 |ig/m3, resulting in only a slight decrease
from 2010 levels.
•	The median concentrations shown between 2010 and 2013 have varied by less than
0.01 |ig/m3 over the period, ranging from 0.035 |ig/m3 (2013) to 0.044 |ig/m3 (2012).
22-77

-------
Figure 22-52. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
OCOK
1.25 -i	
1.00
E '
1
c
0
1
c
OJ
5 0.50
0.25
0.00
2009 1	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile '•**~*** Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2009.
Observations from Figure 22-52 for carbon tetrachloride measurements collected at OCOK
include the following:
•	The two highest concentrations of carbon tetrachloride were measured at OCOK in
2009, including one greater than 1 |ig/m3 (1.10 |ig/m3). The maximum concentrations
measured in other years fall between 0.80 |ig/m3 and 0.90 |ig/m3.
•	The range of carbon tetrachloride concentrations measured at OCOK has decreased
each year since the onset of sampling, reaching a minimum in 2013, when all carbon
tetrachloride concentrations measured spanned less than 0.50 |ig/m3.
•	The 1-year average concentrations of carbon tetrachloride have varied by less than
0.1 |ig/m3, ranging from 0.58 |ig/m3 (2011) to 0.66 |ig/m3 (2012). The median
concentrations have a similar pattern, ranging from 0.59 |ig/m3 (2011) to 0.67 |ig/m3
(2012).
•	With the exception of 2009, the maximum and 95th percentiles have changed little
over the years of sampling.
•	With the exception of 2013, the median concentration is greater than the 1-year
average concentration, which can be attributed to the few concentrations on the lower


pU-

—5—









1

	•	





I
22-78

-------
end of the range, particularly for 2010 and 2011, which can pull an average down
similar to an outlying high concentration.
Figure 22-53. Yearly Statistical Metrics for />-Dichlorobenzene Concentrations Measured at
OCOK
.5
.0
.5
.0
.5
.0
.5
.0
1
2009
2010
2011
2012
2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2009.
Observations from Figure 22-53 for p-dichlorobenzene measurements collected at OCOK
include the following:
•	The maximum p-dichlorobenzene concentration was measured at OCOK on
September 22, 2009 (3.18 |ig/m3). Three concentrations greater than 1 |ig/m3 were
measured in 2010. No other/?-dichlorobenzene concentrations greater than 1.0 |ig/m3
were measured at OCOK in the years that follow.
•	/;-Dichlorobenzene concentrations measured at OCOK in 2010 were higher than any
other years of sampling. The 1-year average concentration calculated for 2010 is
greater than the maximum concentration measured for any of the following years. In
addition, the median concentration for 2010 is greater than the 95th percentile for any
of the following years.
•	The range of concentrations measured, the range within which the majority of
concentrations fall, and the 1-year average concentration are all at a minimum for
2013.
22-79

-------
Figure 22-54. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
OCOK
0.45
2009 1	2010	2011	2012	2013
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2009.
Observations from Figure 22-54 for 1,2-dichloroethane measurements collected at OCOK
include the following:
•	The median concentration for 2009, 2010, and 2011 is zero, indicating that at least
half of the measurements were non-detects. In 2009, there were four measured
detections of 1,2-dichloroethane, which increased to 11 for 2010 and 13 for 2011. For
2012, the number of measured detections increased by a factor of four (up to 52) and
there was a similar number of measured detections for 2013.
•	The increase in the measured detections results in an increase in the 1-year average
concentrations shown for each year (and for the median concentration for the later
years).
•	The 95th percentiles changed little between 2010 and 2013, even as a greater number
of measured detections were measured.
22-80

-------
Figure 22-55. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at OCOK
3.0
2.5
2.0
3.0	
2009 1
2010
2011
2012
2013


Year


O 5th Percentile
- Minimum
— Median - Maximum
O 95th Percentile

1 A 1-year average is not presented because sampling under the NMP did not begin until May 2009.
Observations from Figure 22-55 for ethylbenzene measurements collected at OCOK
include the following:
•	The maximum ethylbenzene concentration was measured at OCOK in 2012
(2.94 |ig/m3); only one additional concentration greater than 1 |ig/m3 has been
measured at OCOK (1.71 |ig/m3 measured in 2009).
•	The range of concentrations within which the majority of concentrations fall, as
indicated by the difference between the 5th and 95th percentiles, increased between
2009 and 2011 and then decreased through 2013.
•	The 1-year average ethylbenzene concentration is at a minimum for 2013 (although it
has not changed significantly over the years shown).
22-81

-------
Figure 22-56. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
OCOK
20091	2010	2011	2012	2013
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2009.
Observations from Figure 22-56 for formaldehyde measurements collected at OCOK
include the following:
•	The maximum formaldehyde concentration was measured at OCOK in 2011
(19.57 |ig/m3); the only other concentration greater than 10 |ig/m3 was also measured
at OCOK in 2011 (10.60 |ig/m3). All 17 formaldehyde concentrations greater than
7 |ig/m3 were measured in either 2011 or 2012.
•	With the exception of the 5th percentile, all of the statistical parameters exhibit an
increase from 2010 to 2011. This is not just a result of the highest concentrations
measured in 2011, as concentrations were higher overall. There were 12
measurements collected in 2011 that were greater than the maximum concentration
measured in 2010. The median concentration increased by more than 1 |ig/m3 and the
1-year average concentration increased by more than 60 percent for 2011.
•	Concentrations measured in the years following 2011 are lower, as all of the
statistical parameters exhibit decreases, particularly at the upper end of the
concentration range.
22-82

-------
Figure 22-57. Yearly Statistical Metrics for Hexachloro-l,3-Butadiene Concentrations
Measured at OCOK
0.20
0.16
























0.00
¦ o ¦	T	T "fc
¦	t	r
B	•	


20091	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2009.
Observations from Figure 22-57 for hexachloro-l,3-butadiene measurements collected at
OCOK include the following:
•	Few measured detections were measured during the first few years of sampling at
OCOK. The median concentration is zero for all years of sampling, indicating that at
least half of the measurements were non-detects for each year. There were no
measured detections in 2009, two in 2010, three in 2011, seven in 2012, and 14 in
2013.
•	For 2013, more than three-quarters of the measurements are non-detects and all of the
measured detections are less than the MDL.
22-83

-------
22.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at each Oklahoma monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4
for definitions and explanations regarding the various toxicity factors, time frames, and
calculations associated with these risk-based screenings.
22.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Oklahoma monitoring sites and where annual
average concentrations could be calculated, risk was examined by calculating cancer risk and
noncancer hazard approximations. These approximations can be used as risk estimates for cancer
and noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers want to shift their air
monitoring priorities. Refer to Section 3.4.3.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 22-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Observations from Table 22-6 include the following:
•	Formaldehyde and acetaldehyde have the highest annual average concentrations for
each site (where annual average concentrations could be calculated). Arsenic was the
only TSP metal that was identified as a pollutant of interest for all of the Oklahoma
sites. Annual average arsenic concentrations were all less than 1 ng/m3.
•	Formaldehyde and benzene have the highest cancer risk approximations among the
pollutants of interest for the Oklahoma monitoring sites (where annual average
concentrations could be calculated). Cancer risk approximations for formaldehyde
range from 34.19 in-a-million for OCOK to 41.44 in-a-million for TMOK. TMOK's
cancer risk approximation for formaldehyde ranks 10th highest among all cancer risk
approximations program-wide. Benzene cancer risk approximations for the Oklahoma
monitoring sites range from 6.05 in-a-million for OCOK to 9.44 in-a-million for
TOOK.
•	For arsenic, the cancer risk approximations range from 0.69 in-a-million for TROK to
3.33 in-a-million for TOOK.
22-84

-------
•	None of the pollutants of interest have noncancer hazard approximations greater than
1.0, indicating that no adverse noncancer health effects are expected from these
individual pollutants. The highest noncancer hazard approximation was calculated for
formaldehyde for TMOK (0.33).
•	Cancer risk and noncancer hazard approximations could not be calculated for ADOK
and YUOK.
Table 22-6. Risk Approximations for the Oklahoma Monitoring Sites
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Public Works, Tulsa, Oklahoma - TOOK
Acetaldehyde
0.0000022
0.009
61/61
2.02
±0.25
4.44
0.22
Benzene
0.0000078
0.03
61/61
1.21
±0.17
9.44
0.04
1.3 -Butadiene
0.00003
0.002
57/61
0.07
±0.01
2.23
0.04
Carbon Tetrachloride
0.000006
0.1
61/61
0.63
±0.03
3.75
0.01
1,2 -Dichloroethane
0.000026
2.4
34/61
0.06
±0.02
1.67
<0.01
Ethylbenzene
0.0000025
1
61/61
0.45
±0.06
1.11
<0.01
Formaldehyde
0.000013
0.0098
61/61
2.87
±0.44
37.26
0.29
Hexachloro -1,3 -butadiene
0.000022
0.09
13/61
0.02
±0.01
0.38
<0.01
Arsenic (TSP)a
0.0043
0.000015
58/58
0.78
±0.07
3.33
0.05
Manganese (TSP)a

0.0003
58/58
27.59
±4.75

0.09
Nickel (TSP)a
0.00048
0.00009
58/58
2.09
±0.42
1.00
0.02
— = A Cancer URE or Noncancer RfC is not available.
NA = Not available due to the criteria for calculating an annual average.
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of
viewing.
22-85

-------
Table 22-6. Risk Approximations for the Oklahoma Monitoring Sites (Continued)
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Fire Station, Tulsa, Oklahoma - TMOK
Acetaldehyde
0.0000022
0.009
61/61
1.94
±0.25
4.26
0.22
Benzene
0.0000078
0.03
60/60
0.96
±0.12
7.53
0.03
1.3 -Butadiene
0.00003
0.002
59/60
0.11
±0.02
3.17
0.05
Carbon Tetrachloride
0.000006
0.1
60/60
0.62
±0.02
3.69
0.01
/?-Dichlorobcnzcnc
0.000011
0.8
51/60
0.10
±0.02
1.12
<0.01
1,2 -Dichloroethane
0.000026
2.4
35/60
0.06
±0.01
1.58
<0.01
Ethylbenzene
0.0000025
1
60/60
0.43
±0.07
1.08
<0.01
Formaldehyde
0.000013
0.0098
61/61
3.19
±0.45
41.44
0.33
Hexachloro -1,3 -butadiene
0.000022
0.09
14/60
0.02
±0.01
0.40
<0.01
Arsenic (TSP)a
0.0043
0.000015
56/56
0.65
±0.07
2.78
0.04
Riverside, Tulsa, Oklahoma - TROK
Acetaldehyde
0.0000022
0.009
61/61
1.63
±0.20
3.60
0.18
Benzene
0.0000078
0.03
61/61
1.00
±0.23
7.84
0.03
1,3-Butadiene
0.00003
0.002
56/61
0.07
±0.01
2.17
0.04
Carbon Tetrachloride
0.000006
0.1
61/61
0.61
±0.03
3.66
0.01
1,2 -Dichloroethane
0.000026
2.4
40/61
0.07
±0.01
1.76
<0.01
Ethylbenzene
0.0000025
1
61/61
0.39
±0.06
0.99
<0.01
Formaldehyde
0.000013
0.0098
61/61
2.77
±0.41
36.03
0.28
Hexachloro -1,3 -butadiene
0.000022
0.09
11/61
0.01
±0.01
0.29
<0.01
Arsenic (TSP)a
0.0043
0.000015
56/56
0.80
±0.11
3.42
0.05
Nickel (TSP)a
0.00048
0.00009
56/56
1.43
±0.24
0.69
0.02
— = A Cancer URE or Noncancer RfC is not available.
NA = Not available due to the criteria for calculating an annual average.
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of
viewing.
22-86

-------
Table 22-6. Risk Approximations for the Oklahoma Monitoring Sites (Continued)
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Oklahoma City, Oklahoma - OCOK
Acetaldehyde
0.0000022
0.009
61/61
1.85
±0.23
4.06
0.21
Benzene
0.0000078
0.03
61/61
0.78
±0.29
6.05
0.03
1.3 -Butadiene
0.00003
0.002
40/61
0.04
±0.01
1.32
0.02
Carbon Tetrachloride
0.000006
0.1
61/61
0.64
±0.02
3.83
0.01
/?-Dichlorobcnzcnc
0.000011
0.8
39/61
0.05
±0.01
0.51
<0.01
1,2 -Dichloroethane
0.000026
2.4
51/61
0.08
±0.01
2.01
<0.01
Ethylbenzene
0.0000025
1
61/61
0.23
±0.03
0.59
<0.01
Formaldehyde
0.000013
0.0098
61/61
2.63
±0.41
34.19
0.27
Hexachloro -1,3 -butadiene
0.000022
0.09
16/61
0.02
±0.01
0.48
<0.01
Arsenic (TSP)a
0.0043
0.000015
61/61
0.46
±0.06
1.97
0.03
Air Depot, Oklahoma City, Oklahoma - ADOK
Acetaldehyde
0.0000022
0.009
30/30
NA
NA
NA
Benzene
0.0000078
0.03
30/30
NA
NA
NA
1,3-Butadiene
0.00003
0.002
16/30
NA
NA
NA
Carbon Tetrachloride
0.000006
0.1
30/30
NA
NA
NA
/?-Dichlorobcnzcnc
0.000011
0.8
29/30
NA
NA
NA
1,2 -Dichloroethane
0.000026
2.4
28/30
NA
NA
NA
Formaldehyde
0.000013
0.0098
30/30
NA
NA
NA
Arsenic (TSP)a
0.0043
0.000015
29/29
NA
NA
NA
— = A Cancer URE or Noncancer RfC is not available.
NA = Not available due to the criteria for calculating an annual average.
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of
viewing.
22-87

-------
Table 22-6. Risk Approximations for the Oklahoma Monitoring Sites (Continued)
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Yukon, Oklahoma - YUOK
Acetaldehyde
0.0000022
0.009
30/30
NA
NA
NA
Benzene
0.0000078
0.03
30/30
NA
NA
NA
1.3 -Butadiene
0.00003
0.002
22/30
NA
NA
NA
Carbon Tetrachloride
0.000006
0.1
30/30
NA
NA
NA
1,2 -Dichloroethane
0.000026
2.4
20/30
NA
NA
NA
Formaldehyde
0.000013
0.0098
30/30
NA
NA
NA
Hexachloro -1,3 -butadiene
0.000022
0.09
9/30
NA
NA
NA
Arsenic (TSP)a
0.0043
0.000015
31/31
NA
NA
NA
— = A Cancer URE or Noncancer RfC is not available.
NA = Not available due to the criteria for calculating an annual average.
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of
viewing.
22.5.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 22-7 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
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 3.4.3.4. Lastly,
Table 22-7 provides the pollutants with the highest cancer risk approximations (in-a-million) for
each 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.
22-88

-------
Table 22-7. 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)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Public Works, Tulsa, Oklahoma (Tulsa County) - TOOK
Benzene
642.74
Benzene
5.01E-03
Formaldehyde
37.26
Ethylbenzene
397.71
Hexavalent Chromium
4.29E-03
Benzene
9.44
Formaldehyde
314.78
Formaldehyde
4.09E-03
Acetaldehyde
4.44
Acetaldehyde
183.16
1,3-Butadiene
2.69E-03
Carbon Tetrachloride
3.75
1.3 -Butadiene
89.52
Naphthalene
1.08E-03
Arsenic
3.33
Tetrachloroethylene
54.93
Ethylbenzene
9.94E-04
1,3-Butadiene
2.23
Naphthalene
31.71
POM, Group 2b
5.18E-04
1,2-Dichloroethane
1.67
Trichloroethylene
16.89
POM, Group 2d
4.29E-04
Ethylbenzene
1.11
Dichloro methane
8.60
Acetaldehyde
4.03E-04
Nickel
1.00
POM, Group 2b
5.89
Nickel, PM
3.15E-04
Hexachloro-1,3 -butadiene
0.38
Fire Station, Tulsa, Oklahoma (Tulsa County) - TMOK
Benzene
642.74
Benzene
5.01E-03
Formaldehyde
41.44
Ethylbenzene
397.71
Hexavalent Chromium
4.29E-03
Benzene
7.53
Formaldehyde
314.78
Formaldehyde
4.09E-03
Acetaldehyde
4.26
Acetaldehyde
183.16
1,3-Butadiene
2.69E-03
Carbon Tetrachloride
3.69
1.3 -Butadiene
89.52
Naphthalene
1.08E-03
1,3-Butadiene
3.17
Tetrachloroethylene
54.93
Ethylbenzene
9.94E-04
Arsenic
2.78
Naphthalene
31.71
POM, Group 2b
5.18E-04
1,2-Dichloroethane
1.58
Trichloroethylene
16.89
POM, Group 2d
4.29E-04
/j-Dichlorobcnzcne
1.12
Dichloro methane
8.60
Acetaldehyde
4.03E-04
Ethylbenzene
1.08
POM, Group 2b
5.89
Nickel, PM
3.15E-04
Hexachloro-1,3 -butadiene
0.40

-------
Table 22-7. 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)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Riverside, Tulsa, Oklahoma (Tulsa County) - TROK
Benzene
642.74
Benzene
5.01E-03
Formaldehyde
36.03
Ethylbenzene
397.71
Hexavalent Chromium
4.29E-03
Benzene
7.84
Formaldehyde
314.78
Formaldehyde
4.09E-03
Carbon Tetrachloride
3.66
Acetaldehyde
183.16
1,3-Butadiene
2.69E-03
Acetaldehyde
3.60
1.3 -Butadiene
89.52
Naphthalene
1.08E-03
Arsenic
3.42
Tetrachloroethylene
54.93
Ethylbenzene
9.94E-04
1,3-Butadiene
2.17
Naphthalene
31.71
POM, Group 2b
5.18E-04
1,2-Dichloroethane
1.76
Trichloroethylene
16.89
POM, Group 2d
4.29E-04
Ethylbenzene
0.99
Dichloro methane
8.60
Acetaldehyde
4.03E-04
Nickel
0.69
POM, Group 2b
5.89
Nickel, PM
3.15E-04
Hexachloro-1,3 -butadiene
0.29
Air Depot, Oklahoma City, Oklahoma (Oklahoma County) - ADOK
Benzene
469.97
Benzene
3.67E-03

Ethylbenzene
297.38
Formaldehyde
3.63E-03
Formaldehyde
279.17
1,3-Butadiene
1.78E-03
Acetaldehyde
149.46
Hexavalent Chromium
9.52E-04
1.3 -Butadiene
59.23
Naphthalene
8.47E-04
Tetrachloroethylene
48.47
Ethylbenzene
7.43E-04
Naphthalene
24.91
POM, Group 2b
4.40E-04
Dichloro methane
14.77
POM, Group 2d
3.52E-04
POM, Group 2b
5.01
Acetaldehyde
3.29E-04
POM, Group 2d
3.99
Arsenic, PM
2.40E-04

-------
Table 22-7. 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)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Oklahoma City, Oklahoma (Oklahoma County) - OCOK
Benzene
469.97
Benzene
3.67E-03
Formaldehyde
34.19
Ethylbenzene
297.38
Formaldehyde
3.63E-03
Benzene
6.05
Formaldehyde
279.17
1,3-Butadiene
1.78E-03
Acetaldehyde
4.06
Acetaldehyde
149.46
Hexavalent Chromium
9.52E-04
Carbon Tetrachloride
3.83
1.3 -Butadiene
59.23
Naphthalene
8.47E-04
1,2-Dichloroethane
2.01
Tetrachloroethylene
48.47
Ethylbenzene
7.43E-04
Arsenic
1.97
Naphthalene
24.91
POM, Group 2b
4.40E-04
1,3-Butadiene
1.32
Dichloro methane
14.77
POM, Group 2d
3.52E-04
Ethylbenzene
0.59
POM, Group 2b
5.01
Acetaldehyde
3.29E-04
/?-Dichlorobcnzcnc
0.51
POM, Group 2d
3.99
Arsenic, PM
2.40E-04
Hexachloro-1,3 -butadiene
0.48
Yukon, Oklahoma (Canadian County) - YUOK
Formaldehyde
153.30
Formaldehyde
1.99E-03

Benzene
69.83
Benzene
5.45E-04
Acetaldehyde
50.89
1,3-Butadiene
3.53E-04
Ethylbenzene
34.53
Naphthalene
1.61E-04
1,3-Butadiene
11.75
Acetaldehyde
1.12E-04
Naphthalene
4.72
Ethylbenzene
8.63E-05
Tetrachloroethylene
2.35
POM, Group 2b
8.61E-05
POM, Group 2b
0.98
POM, Group 2d
7.69E-05
Dichloro methane
0.88
POM, Group 5a
5.54E-05
POM, Group 2d
0.87
Arsenic, PM
2.78E-05

-------
Table 22-8. 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)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Public Works, Tulsa, Oklahoma (Tulsa County) - TOOK
Toluene
2,096.03
Acrolein
869,130.16
Formaldehyde
0.29
Xylenes
1,502.33
1,3-Butadiene
44,761.98
Acetaldehyde
0.22
Hexane
862.22
Formaldehyde
32,120.04
Manganese
0.09
Benzene
642.74
Benzene
21,424.75
Arsenic
0.05
Ethylbenzene
397.71
Acetaldehyde
20,350.75
Benzene
0.04
Methanol
360.45
Xylenes
15,023.31
1,3-Butadiene
0.04
Formaldehyde
314.78
Naphthalene
10,570.68
Nickel
0.02
Acetaldehyde
183.16
T richloroethy lene
8,445.87
Carbon Tetrachloride
0.01
Ethylene glycol
120.05
Nickel, PM
7,292.24
Ethylbenzene
<0.01
1.3 -Butadiene
89.52
Lead, PM
5,904.21
Hexachloro-1,3 -butadiene
<0.01
Fire Station, Tulsa, Oklahoma (Tulsa County) - TMOK
Toluene
2,096.03
Acrolein
869,130.16
Formaldehyde
0.33
Xylenes
1,502.33
1,3-Butadiene
44,761.98
Acetaldehyde
0.22
Hexane
862.22
Formaldehyde
32,120.04
1,3-Butadiene
0.05
Benzene
642.74
Benzene
21,424.75
Arsenic
0.04
Ethylbenzene
397.71
Acetaldehyde
20,350.75
Benzene
0.03
Methanol
360.45
Xylenes
15,023.31
Carbon Tetrachloride
0.01
Formaldehyde
314.78
Naphthalene
10,570.68
Ethylbenzene
<0.01
Acetaldehyde
183.16
T richloroethy lene
8,445.87
Hexachloro-1,3 -butadiene
<0.01
Ethylene glycol
120.05
Nickel, PM
7,292.24
/?-Dichlorobcnzcnc
<0.01
1,3-Butadiene
89.52
Lead, PM
5,904.21
1,2-Dichloroethane
<0.01

-------
Table 22-8. 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)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Riverside, Tulsa, Oklahoma (Tulsa County) - TROK
Toluene
2,096.03
Acrolein
869,130.16
Formaldehyde
0.28
Xylenes
1,502.33
1,3-Butadiene
44,761.98
Acetaldehyde
0.18
Hexane
862.22
Formaldehyde
32,120.04
Arsenic
0.05
Benzene
642.74
Benzene
21,424.75
1,3-Butadiene
0.04
Ethylbenzene
397.71
Acetaldehyde
20,350.75
Benzene
0.03
Methanol
360.45
Xylenes
15,023.31
Nickel
0.02
Formaldehyde
314.78
Naphthalene
10,570.68
Carbon Tetrachloride
0.01
Acetaldehyde
183.16
T richloroethylene
8,445.87
Ethylbenzene
<0.01
Ethylene glycol
120.05
Nickel, PM
7,292.24
Hexachloro-1,3 -butadiene
<0.01
1.3 -Butadiene
89.52
Lead, PM
5,904.21
1,2-Dichloroethane
<0.01
Air De
)ot, Oklahoma City, Oklahoma (Oklahoma County) - ADOK
Toluene
1,716.89
Acrolein
825,550.98

Xylenes
1,179.06
1,3-Butadiene
29,617.41
Hexane
800.82
Formaldehyde
28,486.41
Benzene
469.97
Acetaldehyde
16,606.84
Methanol
444.71
Benzene
15,665.51
Ethylbenzene
297.38
Xylenes
11,790.55
Formaldehyde
279.17
Naphthalene
8,303.68
Ethylene glycol
202.30
Arsenic, PM
3,725.81
Acetaldehyde
149.46
Nickel, PM
3,115.38
Methyl isobutyl ketone
71.17
Propionaldehyde
2,411.31

-------
Table 22-8. 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
Top 10 Noncancer Toxicity-Weighted
Emissions
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(County-Level)
(County-Level)
(Site-Specific)





Noncancer



Noncancer

Hazard
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(HQ)
Oklahoma City, Oklahoma (Oklahoma County) - OCOK
Toluene
1,716.89
Acrolein
825,550.98
Formaldehyde
0.27
Xylenes
1,179.06
1,3-Butadiene
29,617.41
Acetaldehyde
0.21
Hexane
800.82
Formaldehyde
28,486.41
Arsenic
0.03
Benzene
469.97
Acetaldehyde
16,606.84
Benzene
0.03
Methanol
444.71
Benzene
15,665.51
1,3-Butadiene
0.02
Ethylbenzene
297.38
Xylenes
11,790.55
Carbon Tetrachloride
0.01
Formaldehyde
279.17
Naphthalene
8,303.68
Hexachloro-1,3 -butadiene
<0.01
Ethylene glycol
202.30
Arsenic, PM
3,725.81
Ethylbenzene
<0.01
Acetaldehyde
149.46
Nickel, PM
3,115.38
/?-Dichlorobcnzcnc
<0.01
Methyl isobutyl ketone
71.17
Propionaldehyde
2,411.31
1,2-Dichloroethane
<0.01
Yukon, Oklahoma (Canadian County) - YUOK
Xylenes
270.61
Acrolein
960,699.47


Toluene
218.04
Formaldehyde
15,642.59


Formaldehyde
153.30
1,3-Butadiene
5,876.61


Hexane
119.27
Acetaldehyde
5,654.44


Methanol
80.71
Xylenes
2,706.07


Benzene
69.83
Benzene
2,327.52


Acetaldehyde
50.89
Naphthalene
1,574.09


Ethylbenzene
34.53
Cyanide Compounds, gas
1,510.86


Ethylene glycol
22.94
Lead, PM
1,020.34


Acrolein
19.21
Arsenic, PM
430.77



-------
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 22.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
Observations from Table 22-7 include the following:
•	Benzene is the highest emitted pollutant with a cancer URE in Tulsa and Oklahoma
Counties, followed by ethylbenzene and formaldehyde. The highest emitted
pollutants in Canadian County are formaldehyde, benzene, and acetaldehyde. The
quantity of emissions is highest in Tulsa County and lowest in Canadian County.
•	The pollutant with the highest toxicity-weighted emissions (of the pollutants with
cancer UREs) for Tulsa County is benzene, followed by hexavalent chromium and
formaldehyde. The pollutant with the highest toxicity-weighted emissions (of the
pollutants with cancer UREs) for Oklahoma County is also benzene, followed by
formaldehyde and 1,3-butadiene. The pollutant with the highest toxicity-weighted
emissions (of the pollutants with cancer UREs) for Canadian County is formaldehyde,
followed by benzene and 1,3-butadiene.
•	Seven of the highest emitted pollutants in Tulsa County also have the highest
toxicity-weighted emissions. Eight of the highest emitted pollutants in Oklahoma
County also have the highest toxicity-weighted emissions. Eight of the highest
emitted pollutants in Canadian County also have the highest toxicity-weighted
emissions.
•	Formaldehyde and benzene have the highest cancer risk approximations among the
Oklahoma sites' pollutants of interest (where risk approximations could be
calculated). Both of these pollutants appear at or near the top of both emissions-based
lists for each county. Acetaldehyde, 1,3-butadiene, and ethylbenzene also appear on
all three lists for TOOK, TMOK, TROK, and OCOK.
•	Nickel is a pollutant of interest for TOOK and TROK and has one of the higher
cancer risk approximations for each site. Nickel has the 10th highest toxicity-
weighted emissions for Tulsa County but is not among the highest emitted (its
emissions rank 12th).
•	Carbon tetrachloride, arsenic, 1,2-dichloroethane, and hexachloro-1,3-butadiene are
pollutants of interest for each site and have one of the higher cancer risk
22-95

-------
approximations for each site but do not appear on either emissions-based site.
/;-Dichlorobenzene is a pollutant of interest for TMOK and OCOK that appears on
neither emissions-based list.
• Naphthalene and several POM Groups appear in Table 22-7 for quantity emitted and
toxicity-weighted emissions. PAHs were not sampled for under the NMP at the
Oklahoma sites.
Observations from Table 22-8 include the following:
•	Toluene and xylenes are the highest emitted pollutants with noncancer RfCs in Tulsa
and Oklahoma Counties, while the order was reversed for Canadian County.
Emissions were generally highest in Tulsa County and lowest in Canadian County.
•	Acrolein is the pollutant with the highest toxicity-weighted emissions (of the
pollutants with noncancer RfCs) for all three counties. Acrolein has the highest
toxicity-weighted emissions for almost all counties with NMP sites but appears
among the highest emitted for only two. Canadian County is one of those counties,
with acrolein ranking 10th among those with the highest emissions. Compared to
other counties with NMP sites, Canadian County's acrolein emissions are not
exceedingly high (19.21 tpy), but are the 15th highest for counties for NMP sites and
are slightly higher than the emissions for Tulsa County (17.38 tpy) and Oklahoma
County (16.51 tpy). Acrolein was sampled for at all of the Oklahoma sites, but this
pollutant was excluded from the pollutants of interest designation, and thus
subsequent risk-based screening evaluations, due to questions about the consistency
and reliability of the measurements, as discussed in Section 3.2.
•	Four of the highest emitted pollutants in Oklahoma County also have the highest
toxicity-weighted emissions; five of the highest emitted pollutants in Tulsa County
and Canadian County also have the highest toxicity-weighted emissions. Note that
although toluene is one of, if not the highest emitted pollutant in all three counties,
this pollutant does not appear among those with the highest toxicity-weighted
emissions.
•	Formaldehyde and acetaldehyde have the highest noncancer hazard approximations
among the Oklahoma sites (where they could be calculated). These pollutants appear
on both emissions-based lists for each county. Benzene and 1,3-butadiene also appear
on all three lists for TOOK, TMOK, and TROK. 1,3-Butadiene has one of the highest
noncancer hazard approximations for OCOK, and has some of the highest toxicity-
weighted emissions for Oklahoma County but is not one of the 10 highest emitted in
that county (but is just outside the list at 11th highest).
•	Several metals appear among the pollutants with the highest toxicity-weighted
emissions for each county but no metals are listed among the highest emitted
pollutants for any of the three counties. This speaks to the relative toxicity of the
speciated metals. Note that for the metals, the emissions-based lists are PMio while
the Oklahoma sites sampled TSP metals.
22-96

-------
22.6 Summary of the 2013 Monitoring Data for the Oklahoma Monitoring Sites
Results from several of the data treatments described in this section include the
following:
~~~ Seventeen pollutants failed at least one screen for TOOK; 15 pollutants failed screens
for TMOK; 14 pollutants failed screens for TROK; 12 pollutants failed screens for
ADOK; 16 pollutants failed screens for OCOK; and 11 pollutants failed screens for
YUOK. Sampling at ADOK and YUOK includes only half of a year's worth of
sampling due to the mid-year relocation of the sampling instrumentation from ADOK
to YUOK
~~~ Formaldehyde and acetaldehyde had the highest annual average concentrations for
each site. Concentrations of these carbonyl compounds tended to be higher during
the warmer months of the year.
~~~ After several years of increasing, concentrations of acetaldehyde, ethylbenzene, and
manganese decreased at TOOK for 2013. Other pollutants exhibit this trend as well
but the difference is less significant. Benzene concentrations measured at TOOK have
been decreasing over the last few years. Benzene concentrations have also been
decreasing at TMOK. In addition, the detection rates of 1,2-dichloroethane and
hexachloro-1,3-butadiene have been increasing at TOOK, TMOK, and OCOK over
the last few years of sampling, particularly for 1,2-dichloroethane. Concentrations of
the acetaldehyde andformaldehyde have also been decreasing at OCOK.
~~~ Formaldehyde has the highest cancer risk approximation among the site-specific
pollutants of interest for each site. None of the pollutants of interest have noncancer
hazard approximations greater than an HQ of 1.0.
22-97

-------
23.0	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.
23.1	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 23-1 is a composite
satellite image retrieved from ArcGIS Explorer showing the monitoring site and its immediate
surroundings. Figure 23-2 identifies nearby point source emissions locations by source category,
as reported in the 2011 NEI for point sources, version 2. Note that only sources within 10 miles
of the site are included in the facility counts provided in Figure 23-2. A 10-mile boundary was
chosen to give the reader an indication of which emissions sources and emissions source
categories could potentially have a direct effect on the air quality at the monitoring site. Further,
this boundary provides both the proximity of emissions sources to the monitoring site as well as
the quantity of such sources within a given distance of the site. Sources outside the 10-mile
boundary are still visible on the map for reference, but have been grayed out in order to
emphasize emissions sources within the boundary. Table 23-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates.
23-1

-------
Figure 23-1. Providence, Rhode Island (PRR1) Monitoring Site
rBfklgham>5t
Dod\ey
*¦
.'•VJutma'ih St
*>uin
lire#: U4GS
h * iV ,rtf onuses
MiiflSiofl Cu'liJ.
b&
OJ
to

-------
Figure 23-2. NEI Point Sources Located Within 10 Miles of PRRI

MASSACHUSETTS
* Prmtwnce
\ County
an mo i
County
P/avtdtncv
naiear
Kent
County
PRRI NATTS site
10 mite radius [	| County boundary
Source Category Group (No. of Facilities)
T
*
0
B
C
AwrAttnaMU'HHi bupport O^mtuon <11)
Prooidnn^w M« *acr«ft P»«nl •*
hJnHrtr S»w^ Plwitef**
3J* T«WiNiBuh »W#» IS)
CltttTM*	tu • '*} Fwlt) 110)
I Cawp»tsvtr SUUK) 17}
# Oiy ClCMIQ z*tWi l4t I
0	fc*KK«jj	Mvwac».
HmewyoieH<.oflgf»e 1

*
»
**UM
nut Mae***** e< Pt»i (111
P
o
MMU
NUN«I IkMM h'jipa^t y».
MM ¦*&( *25>
4
•
Ltndf
win

n

UMM
I P-nr—ii|)yM«aBer f

*
*
W*ia<
iu.r»> kAMO* |,'UuaU<|j

T

Um>
IttlMMiA CflWIItMUlt*.
iititl Fatiav (75t
A
•
dMI
¦rv 0*h f^otiuejan > • |

«
~

nj W»-ahrJii»«ri
J »•*»«,, (J)
w
MMVttttrflt* »/)
Rj*U£ Retflt c« ^oc«t PrwJut*»R»»«v1«/
P»m«ig CMing A Qyvng trt "-at*** Fac^ty l2>
Rtrhn9*M>tenn^Papet Prrjdvo MourtMjrog f <«0
Vipftaoi Mvuftitiimg rjt	» «Mly «4I
Teeoo*nnmif»c*imwR*t« '«u«ry .4|
Immg labmiKsn* n i
Ttntfc o* OpK Plmi t?t
Trtcfc^uaTrarooraocn Cswwfli 111
Wlwewetf Ti»
-------
Table 23-1. Geographical Information for the Rhode Island Monitoring Site
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Additional Ambient Monitoring Information1
PRRI
44-007-0022
Providence
Providence
Providence-
Warwick, RI-MA
41.807776,
-71.415105
Residential
Urban/City
Center
PAMS, VOCs, Carbonyl Compounds, Meteorological
parameters, PMio, PMio Speciation, Black Carbon
PM2 5, PM2 5 Speciation, SNMOC.
1 Data for additional pollutants are reported to AQS for PRRI (EPA, 2014b); however, these data are not generated by ERG and are therefore not included in this report.
BOLD ITALICS = EPA-designated NATTS Site
to
-L

-------
Figure 23-1 shows that the areas to the west and south of PRRI are primarily residential,
but areas to the north and east are commercial. A hospital lies to the northeast of the site, just
north of Dudley Street. Interstate-95 runs north-south about one-half mile to the east of the site,
then turns northwestward, entering downtown Providence. The Providence Harbor is just on the
other side of 1-95 and can be seen on the right-hand side of Figure 23-1. Figure 23-2 shows that a
large number of point sources are located within 10 miles of PRRI, most of which are within
about 5 miles of the site. The source categories with the greatest number of point sources within
10 miles of PRRI include dry cleaners; institutions (such as schools, prisons, and hospitals);
metals processing and fabrication facilities; electroplating, plating, polishing, anodizing, and
coloring facilities; and auto body shops, painters, and automotive stores. Sources within one-half
mile of PRRI include several hospitals, a heliport at a hospital, a bulk terminal/bulk plant, an
electroplating, plating, polishing, anodizing, and coloring facility, and a facility that falls into the
miscellaneous commercial and industrial source category.
Table 23-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Rhode Island monitoring site. Table 23-2 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 23-2 also
contains traffic volume information for PRRI as well as the location for which the traffic volume
was obtained. Additionally, Table 23-2 presents county-level daily VMT for Providence County
from the 2011 NEI.
Table 23-2. Population, Motor Vehicle, and Traffic Information for the Rhode Island
Monitoring Site




Annual




Estimated
County-level
Average
Intersection
County-


County
Vehicle
Daily
Used for
level Daily
Site
County
Population1
Registration2
Traffic3
Traffic Data
VMT4
PRRI
Providence
628,600
511,015
136,800
1-95 near 1-195
11,670,714
bounty-level population estimate reflects 2013 data (Census Bureau, 2014)
2County-level vehicle registration is a ratio based on 2012 state-level vehicle registration data from the FHWA and
the 2012 county-level proportion of the state population data (FHWA, 2014 and Census Bureau, 2013c)
3AADT reflects 2009 data (RI DOT, 2009)
4County-level VMT reflects 2011 data (EPA, 2015a)
BOLD ITALICS = EPA-designated NATTS Site
23-5

-------
Observations from Table 23-2 include the following:
•	Providence County's population is in the middle of the range compared to other
counties with NMP sites.
•	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 ninth highest compared to traffic
volumes near other NMP monitoring sites. The traffic estimate provided is for 1-95
near the 1-195 interchange.
•	The daily VMT for Providence County is 11.7 million miles and ranks in the middle
of the range compared to other counties with NMP sites.
23.2 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.
23.2.1	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 these bodies of water temper cold air
outbreaks, and breezes off the ocean moderate summertime heat. On average, southerly and
southwesterly winds in the summer become west-northwesterly in the winter. Storm systems
frequently affect the New England region, producing variable weather. Precipitation occurs in
Providence about once every 3 days and is distributed rather evenly throughout the year.
Thunderstorms are common between May and August, while coastal storms during the cooler
months tend to produce the greatest amounts of rain and snow. Thirty inches of snow is typical in
winter (Wood, 2004; CoCoRaHS, 2011).
23.2.2	Meteorological Summary
Hourly meteorological data for 2013 were retrieved from NCDC for the weather station
closest to the Rhode Island monitoring site (NCDC, 2013), as described in Section 3.4.2. 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 23-3. These data were used to determine how
meteorological conditions on sample days vary from conditions experienced throughout the year.
23-6

-------
Table 23-3. 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
Sample
Days
(64)
59.6
±4.7
51.4
±4.4
39.8
±4.8
46.2
±4.2
67.5
±3.4
1016.8
± 1.8
7.3
±0.7
189°
(S)
2013
60.2
+ 1.9
51.9
+ 1.8
40.4
±2.0
46.8
+ 1.7
67.9
+ 1.5
1017.0
±0.8
7.2
±0.3
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
to
<1

-------
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 all of 2013. 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 are representative of average weather conditions experienced
throughout the year near PRRI.
23.2.3 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 3.4.2. A wind rose shows the frequency of wind directions using "petals" positioned
around a 16-point compass, and uses different colors to represent wind speeds.
Figure 23-3 presents 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 23-3 also presents three different wind roses for the
PRRI monitoring site. First, a historical wind rose representing 2003 to 2012 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 2013 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2013 is
presented. These can be used to identify the predominant wind speed and direction for 2013 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.
23-8

-------
Figure 23-3. Wind Roses for the T.F. Green State Airport Weather Station near PRRI
Location of PRRI and Weather Station
2003-2012 Historical Wind Rose
NORTH
ES .
'A'INC SPEEC
(Kn ots i
11 -17
SOUTH

Calms: 9.19%
2013 Wind Rose
Sample Day Wind Rose
NORTH ---
ES"
WIN 0 S PE EC
(Knots)
17-21
11 - 17
SOUTH
Calms: 10.99%
E = ~
WIND SPEEC
(Knots)
SOUTH
23-9

-------
Observations from Figure 23-3 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 westerly winds were observed most, accounting
for approximately 11 percent of observations. Winds from the western quadrants, due
north, and due south were often observed near PRRI while wind from the east-
northeast to southeast were infrequently observed. Calm winds (those less than or
equal to 2 knots) account for less than 10 percent of the hourly measurements.
•	The wind patterns shown on the 2013 wind rose are similar to the historical wind
patterns, as winds from the western quadrants, due north, and due south were
observed most often. Westerly winds prevailed near PRRI in 2013, accounting for
13 percent of observations. The calm rate for 2013 is 11 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 wind
observations from the north is reduced. There are also fewer observations from the
north-northeast and northeast. There is also a higher number of winds observations
from the northwest, such that winds from the west to northwest account for nearly
one-third of the wind observations on sample days near PRRI.
23.3 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 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. PAHs and
hexavalent chromium were sampled for at PRRI, although hexavalent chromium sampling was
discontinued at the end of June.
23-10

-------
Table 23-4. Risk-Based Screening Results for the Rhode Island Monitoring Site
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Providence, Rhode Island - PRRI
Naphthalene
0.029
54
59
91.53
96.43
96.43
Benzo(a)pyrene
0.00057
1
58
1.72
1.79
98.21
Hexavalent Chromium
0.000083
1
12
8.33
1.79
100.00
Total
56
129
43.41

Observations from Table 23-4 include the following:
•	Three pollutants failed at least one screen for PRRI; 43 percent of concentrations for
these three pollutants were greater than their associated risk screening value (or failed
screens).
•	Concentrations of naphthalene failed 54 of the 59 screens, with benzo(a)pyrene and
hexavalent chromium failing one screen each.
•	Naphthalene accounted for 96 percent of the total failed screens for PRRI. Thus,
naphthalene is the only pollutant of interest for PRRI.
23.4 Concentrations
This section presents various concentration averages used to characterize air toxics
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:
•	Time period-based concentration averages (quarterly and annual) are provided for
each monitoring site.
•	Annual concentration averages are presented graphically to illustrate how each site's
concentrations compare to the program-level averages, as presented in Section 4.1.
•	Concentration averages and other statistical metrics are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at PRRI are provided in Appendices M and O.
23-11

-------
23.4.1 2013 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 23-5, 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 23-5. 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
Naphthalene
59/59
67.85
± 26.22
53.83
± 12.14
53.11
± 11.81
70.74
±23.93
61.57
±9.53
Observations for PRRI from Table 23-5 include the following:
•	Naphthalene was detected in all of the valid PAH samples collected at PRRI.
•	Concentrations of naphthalene measured at PRRI span an order of magnitude, ranging
from 17.5 ng/m3 to 187 ng/m3.
•	The second and third quarter average concentrations of naphthalene are very similar
to each other.
•	The first and fourth quarter average concentrations are also similar to each other and
are slightly higher than the quarterly averages for the warmer months (although not
significantly so), and exhibit more variability. Both the minimum and maximum
naphthalene concentrations were measured at PRRI in November. Of the eight
naphthalene concentrations greater than 100 ng/m3, four were measured during the
23-12

-------
first quarter and two were measured during the fourth quarter (with one each
measured during the other two calendar quarters).
23.4.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, a box plot was created for the pollutant shaded in
gray in Table 23-4 for PRRI. Figure 23-4 overlays PRRI's minimum, annual average, and
maximum naphthalene concentrations onto the program-level minimum, first quartile, median,
average, third quartile, and maximum concentrations, as described in Section 3.4.3.1.
Figure 23-4. Program vs. Site-Specific Average Naphthalene Concentration
—
300	400	500
Concentration {ng/m3)
Program: 1st Quartile
¦
2nd Quartile 3rd Quartile
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range


Observations from Figure 23-4 include the following:
• The maximum naphthalene concentration measured at PRRI is one-fourth the
maximum concentration measured at the program-level. There were no non-
detects of naphthalene measured at PRRI (or across the program). The annual
average naphthalene concentration for PRRI falls between the program-level
median and average concentrations. PRRI's annual average concentration of
naphthalene is in the middle of the range compared to other NMP sites sampling
PAHs.
23.4.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.3.2.
PRRI has sampled PAHs under the NMP since 2008. Thus, Figure 23-5 presents the 1-year
statistical metrics for the pollutant of interest for PRRI. The statistical metrics presented for
assessing trends include the substitution of zeros for non-detects. If sampling began mid-year, a
23-13

-------
minimum of 6 months of sampling is required for inclusion in the trends analysis; in these cases,
a 1-year average concentration is not provided, although the range and percentiles are still
presented.
Figure 23-5. Yearly Statistical Metrics for Naphthalene Concentrations Measured at PRRI
1 200
0 +-
20081
2009
2010
2011
Year
2012
2013

O 5th Percentile
- Minimum
— Median
- Maximum
O 95th Percentile

1 A 1-year average is not presented because sampling under the NMP did not begin until July 2008.
Observations from Figure 23-5 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 concentration is not presented for 2008,
although the range of measurements is provided.
•	The maximum naphthalene concentration was measured at PRRI in 2011
(301 ng/m3). In total, 10 naphthalene concentrations greater than 200 ng/m3 have been
measured at PRRI, of which seven were measured in November of any given year. In
fact, the maximum concentration for all years except 2008 was measured in
November. Of the 61 naphthalene concentrations greater than 100 ng/m3 measured at
PRRI, more than half were measured during the fourth quarter of any given year.
23-14

-------
•	Although the range of concentrations measured has varied between 2009 and 2012,
the 1-year average concentrations of naphthalene exhibit little variability, ranging
from 71.39 ng/m3 (2010) to 77.73 ng/m3 (2009). This is also true for the median
concentration, which ranges from 59.35 ng/m3 (2012) to 64.80 ng/m3 (2009).
•	Several of the statistical parameters, including the 1-year average and median
concentrations (61.57 ng/m3 and 52.30 ng/m3, respectively), are at a minimum for
2013.
23.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at the PRRI monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4 for
definitions and explanations regarding the various toxicity factors, time frames, and calculations
associated with these risk-based screenings.
23.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Rhode Island monitoring site and where annual
average concentrations could be calculated, risk was examined by calculating cancer risk and
noncancer hazard approximations. These approximations can be used as risk estimates for cancer
and noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers want to shift their air
monitoring priorities. Refer to Section 3.4.3.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 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 Rhode Island Monitoring Site
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections vs.
# of Samples
Annual
Average
(ng/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Providence, Rhode Island - PRRI
Naphthalene
0.000034
0.003
59/59
61.57
±9.53
2.09
0.02
23-15

-------
Observations for PRRI from Table 23-6 include the following:
•	Naphthalene has both a cancer URE and a noncancer RfC.
•	The cancer risk approximation for naphthalene is 2.09 in-a-million.
•	The noncancer hazard approximation for naphthalene is negligible (0.02 in-a-
million), indicating that no adverse noncancer health effects are expected from this
individual pollutant.
23.5.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 23-7 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
that have cancer toxicity factors. Table 23-7 also presents the 10 pollutants with the highest
toxicity-weighted emissions, based on the weighting schema described in Section 3.4.3.4. Lastly,
Table 23-7 provides the pollutants with the highest cancer risk approximations (in-a-million) for
PRRI, 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 the site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 23.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
23-16

-------
Table 23-7. 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)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Providence, Rhode Island (Providence County) - PRRI
Benzene
196.93
Formaldehyde
1.89E-03
Naphthalene
2.09
Formaldehyde
145.48
Benzene
1.54E-03

Ethylbenzene
94.74
1.3 -Butadiene
9.40E-04
Acetaldehyde
76.01
Naphthalene
5.27E-04
1.3 -Butadiene
31.32
POM, Group 2b
4.00E-04
Tetrachloroethylene
17.48
POM, Group 2d
2.39E-04
Naphthalene
15.50
Ethylbenzene
2.37E-04
Trichloroethylene
6.49
POM, Group 5a
2.30E-04
POM, Group 2b
4.54
Arsenic, PM
1.83E-04
Dichloro methane
4.12
Acetaldehyde
1.67E-04

-------
Table 23-8. 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)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Providence, Rhode Island (Providence County) - PRRI
Toluene
636.76
Acrolein
336,121.99
Naphthalene
0.02
Xylenes
390.67
1.3 -Butadiene
15,660.35

Methanol
386.43
Formaldehyde
14,844.73
Hexane
324.64
Acetaldehyde
8,445.40
Benzene
196.93
Benzene
6,564.42
Formaldehyde
145.48
Naphthalene
5,167.59
Ethylene glycol
130.02
Xylenes
3,906.66
Ethylbenzene
94.74
Nickel, PM
3,332.59
Acetaldehyde
76.01
T richloroethy lene
3,243.04
Methyl isobutyl ketone
41.59
Arsenic, PM
2,840.49

-------
Observations from Table 23-7 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 toxicity-weighted emissions (of the
pollutants with cancer UREs), followed by benzene and 1,3-butadiene.
•	Seven of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Providence County.
•	Naphthalene, which is the only pollutant of interest for PRRI, has the seventh highest
emissions and the fourth highest toxicity-weighted emissions for Providence County.
•	Several POM Groups appear among the pollutants with the highest toxicity-weighted
emissions for Providence County. POM, Groups 2b and 2d rank fifth and sixth for
their toxicity-weighted emissions, respectively and POM, Group 2b also ranks ninth
for its quantity emitted. POM, Groups 2b and 2d include several PAHs sampled for at
PRRI, although none of these pollutants failed screens.
•	POM, Group 5a ranks eighth for toxicity-weighted emissions. POM, Group 5a
includes benzo(a)pyrene, which failed a single screen for PRRI. POM, Group 5a is
not among the highest emitted "pollutants" in Providence County.
Observations from Table 23-8 include the following:
•	Toluene, xylenes, and methanol are the highest emitted pollutants with noncancer
RfCs in Providence County.
•	The pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) are acrolein, 1,3-butadiene, and formaldehyde.
•	Four of the highest emitted pollutants in Providence County also have the highest
toxicity-weighted emissions.
•	Although naphthalene ranks sixth among the pollutants with the highest toxicity-
weighted emissions, it is not one of the highest emitted pollutants (with a noncancer
RfC) in Providence County (it ranks 15th).
23.6 Summary of the 2013 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 concentrations of
naphthalene accounting for more than 95 percent of the failed screens. As such,
naphthalene is PRRI's only pollutant of interest.
23-19

-------
•	Concentrations of naphthalene measured at PRRI span an order of magnitude,
ranging from 17.5 ng/m3 to 187 ng/m3.
•	The highest concentrations of naphthalene measured at PRRI tended to be measured
during the fourth quarter of the year, based on data collected since sampling
commenced in 2008.
23-20

-------
24.0	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.
24.1	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 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, as reported in the 2011 NEI for point sources, version 2. Note that only sources
within 10 miles of the site are included in the facility counts provided in Figure 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 boundary are still visible on the map for reference, but have been grayed out in order to
emphasize emissions sources within the boundary. Table 24-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates.
24-1

-------
Figure 24-1. Chesterfield, South Carolina (CHSC) Monitoring Site
Source: US6S
irCc NASA, NCA. USGS
JOOM Microvolt Jj a t p

-------
Figure 24-2. NEI Point Sources Located Within 10 Miles of CHSC
com
County
Dyiingto
County
wafowi
Legend
CHSC NATTS site	10 mile radius
irffvvv
Note Due to faciKj den»«y and cototaton the fatal facilities
dteplay»d rnay not represent all fadlate* wiffttn the area of interest
County boundary
Source Category Group (No. of Facilities)
T Airport/Airtine/Airport Support Operations (1)
~ Industrial Machinery or Equipment Plant (1)
24-3

-------
Table 24-1. Geographical Information for the South Carolina Monitoring Site
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land
Use
Location
Setting
Additional Ambient Monitoring Information1
CHSC
45-025-0001
Not in a
city
Chesterfield
None
34.615367,
-80.198787
Forest
Rural
VOCs, Carbonyl Compounds, O3, Meteorological
parameters, PM10, PM10 Speciation, PM2.5,
PM2.5 Speciation Black Carbon, IMPROVE Speciation.
1 Data for additional pollutants are reported to AQS for CHSC (EPA, 2014b); however, these data are not generated by ERG and are therefore not included in this report.
BOLD ITALICS = EPA-designated NATTS Site
to

-------
CHSC is located about 14 miles south of the North Carolina/South Carolina border, about
halfway between the towns of McBee and Chesterfield. The monitoring site is located near the
Ruby fire tower and, as Figure 24-1 shows, is located just off State Highway 145. The
surrounding area is rural in nature and is part of the Carolina Sandhills National Wildlife Refuge.
Figure 24-2 shows that few point sources are located within 10 miles of CHSC, the closest of
which is the Wild Irish Rose Airport.
Table 24-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the South Carolina monitoring site. Table 24-2 includes both county-
level population and vehicle registration information. Table 24-2 also contains traffic volume
information for CHSC as well as the location for which the traffic volume was obtained.
Additionally, Table 24-2 presents the daily VMT for Chesterfield County.
Table 24-2. Population, Motor Vehicle, and Traffic Information for the South Carolina
Monitoring Site
Site
County
Estimated
County
Population1
County-level
Vehicle
Registration2
Annual
Average Daily
Traffic3
Intersection
Used for Traffic Data
County-level
Daily VMT4
CHSC
Chesterfield
46,197
41,728
700
Hwy 145 between US-1
and Hwy 109
1,265,439
bounty-level population estimate reflects 2013 data (Census Bureau, 2014)
2County-level vehicle registration reflects 2013 data (SC DMV, 2013)
3AADT reflects 2013 data (SC DOT, 2013)
4County-level VMT reflects 2013 data (SC DOT, 2014)
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 24-2 include the following:
•	Chesterfield County's population is among the lowest compared to other counties
with NMP sites, ranking fourth lowest. This is also true for the county-level vehicle
ownership for Chesterfield County, which is the sixth lowest among NMP sites.
•	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 among the lowest VMT compared to other
counties with NMP sites, ranking fifth lowest.
24-5

-------
24.2 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.
24.2.1	Climate Summary
Chesterfield County is located along 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. More than 2 inches
of precipitation can be expected any given month, with the maximum typically occurring in July
and the minimum occurring during the fall months. Chesterfield County leads the state in the
average number of sleet and freezing rain events per year (Bair, 1992; SC SCO, 2015).
24.2.2	Meteorological Summary
Hourly meteorological data for 2013 were retrieved from NCDC for the weather station
closest to the South Carolina monitoring site (NCDC, 2013), as described in Section 3.4.2. The
closest weather station with adequate data is located at the Richmond County Airport in
Rockingham, North Carolina (WBAN 03738). Additional information about the Richmond
County Airport weather station, such as the distance between the site and the weather station, 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), and wind (average scalar wind speed) information for days samples were collected
and for all of 2013 (sea level pressure was not recorded at the Richmond County Airport). Also
included in Table 24-3 is the 95 percent confidence interval for each parameter. As shown in
Table 24-3, average meteorological conditions experienced on sample days were representative
of average weather conditions experienced throughout the year near CHSC. The largest
difference is shown for relative humidity, although the difference is not statistically significant.
24-6

-------
Table 24-3. Average Meteorological Conditions near the South Carolina Monitoring Site
Closest









Weather
Station
Distance
and

Average
Maximum
Average
Average
Dew Point
Average
Wet Bulb
Average
Relative
Average
Sea Level
Average
Scalar Wind
(WBAN and
Coordinates)
Direction
from Site
Average
Type1
Temperature
(°F)
Temperature
(°F)
Temperature
(°F)
Temperature
(°F)
Humidity
(%)
Pressure
(mb)
Speed
(kt)
Chesterfield, South Carolina - CHSC
Riclunond
34.9
miles
Sample
Days
71.6
61.8
51.3
56.3
72.1

4.1
County Airport
(61)
±3.7
±3.6
±4.4
±3.7
±3.6
NA
±0.5
03738
53°
(NE)








(34.89, -79.76)
2013
71.2
+ 1.5
61.5
+ 1.5
51.5
+ 1.7
56.2
+ 1.5
73.3
+ 1.4
NA
4.2
±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 Richmond County Airport.
to

-------
24.2.3 Wind Rose Comparison
Hourly surface wind data from the weather station at Richmond County Airport near
CHSC were uploaded into a wind rose software program to produce customized wind roses, as
described in Section 3.4.2. A wind rose shows the frequency of wind directions using "petals"
positioned around a 16-point compass, and uses different colors to represent wind speeds.
Figure 24-3 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 24-3 also presents three different wind roses for the
CHSC monitoring site. First, a historical wind rose representing 2005 to 2012 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 2013 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2013 is
presented. These can be used to identify the predominant wind speed and direction for 2013 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.
Observations from Figure 24-3 for CHSC include the following:
•	The Richmond County Airport weather station is located across the North
Carolina/South Carolina border, approximately 31 miles northeast of CHSC.
•	The historical wind rose for CHSC shows that calm winds (those less than or equal to
2 knots) account for nearly 30 percent of the hourly measurements. For wind speeds
greater than 2 knots, winds from the south are most common, accounting for
10 percent of observations. Winds from the southwest and northeast quadrants
(including north) are also observed frequently, while winds from the northwest and
southeast quadrants are infrequently observed.
•	The wind patterns shown on the 2013 wind rose for CHSC are similar to the historical
wind patterns, indicating that wind conditions in 2013 were similar to what is
expected climatologically near this site, although there was a slightly higher
percentage of winds observations from the north to northeast.
•	The sample day wind rose shows that southerly winds were still prevalent on sample
days. Northerly winds were also observed frequently, while slightly fewer north-
northeasterly were observed.
24-8

-------
Figure 24-3. Wind Roses for the Richmond County Airport Weather Station near CHSC
Location of CHSC and Weather Station
2005-2012 Historical Wind Rose
. - -
E- :i
EST
Win C S RE EC
(Kn ots >
17-21
11 - 17
SOUTH
3 2J:::
2013 Wind Rose
NORTH ---
; EAST
ri--
WVfD SPEEC
(Knots)
17-21
SOUTH
-sir-;.: ^ gJ-S
Sample Day Wind Rose
EST
WIND SPEED
(Kn ots)
17-21
SOUTH
'I sir- =
CHSC
24-9

-------
24.3 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 24-4. 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 24-4. It is important to note which
pollutants were sampled for at each site when reviewing the results of this analysis. Hexavalent
chromium and PAHs were sampled for at CHSC, although hexavalent chromium sampling was
discontinued in June.
Table 24-4. Risk-Based Screening Results for the South Carolina Monitoring Site
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Chesterfield, South Carolina - CHSC
Naphthalene
0.029
2
58
3.45
100.00
100.00
Total
2
58
3.45

Observations from Table 24-4 include the following:
•	Naphthalene was the only pollutant to fail screens for CHSC.
•	This pollutant was detected in all 58 valid samples collected at CHSC and failed two
screens, or approximately 3 percent of screens.
•	This site has the second lowest number of failed screens (2) among NMP sites
(excluding the four NMP sites with no pollutants failing screens).
24.4 Concentrations
This section presents various concentration averages used to characterize air toxics
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
each monitoring site.
24-10

-------
•	Annual concentration averages are presented graphically for each site to illustrate
how the site's concentrations compare to the program-level averages, as presented in
Section 4.1.
•	Concentration averages and other statistical metrics are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at CHSC are provided in Appendices M and O.
24.4.1 2013 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 the
pollutants of interest for CHSC 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.
Table 24-5. 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
58/58
19.69
±8.57
10.81
±3.31
10.38
±2.78
10.80
± 1.97
12.69
±2.33
24-11

-------
Observations for CHSC from Table 24-5 include the following:
•	Naphthalene concentrations measured at CHSC span an order of magnitude, ranging
from 4.46 ng/m3 to 51.8 ng/m3, with a median concentration of 10.25 ng/m3.
•	The annual average concentration of naphthalene is 12.69 ± 2.33 ng/m3. This is the
second lowest annual average concentration of naphthalene among NMP sites
sampling PAHs.
•	The first quarter average concentration of naphthalene (19.69 ± 8.57 ng/m3) is higher
than the other quarterly averages and has a relatively large confidence interval
associated with it, while the other quarterly averages fall between 10 ng/m3 and
11 ng/m3. The two highest naphthalene concentrations are both around 50 ng/m3 and
were measured at CHSC in January. All other naphthalene concentrations measured
at CHSC were less than 30 ng/m3. Four of the six naphthalene concentrations greater
than 20 ng/m3 were measured during the first quarter of 2013.
24.4.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, a box plot was created for the pollutant shaded in
gray in Table 24-4 for CHSC. Figure 24-4 overlays the site's minimum, annual average, and
maximum naphthalene concentrations onto the program-level minimum, first quartile, median,
average, third quartile, and maximum concentrations of naphthalene, as described in
Section 3.4.3.1.
Figure 24-4. Program vs. Site-Specific Average Naphthalene Concentration
CHSC
0
100
200
300
400
Concentration {ng/m3)
500
600
700
800
Program: 1st Quartile
2nd Quartile 3rd Quartile 4th Quartile Average
~ ~ ~
Site:
SiteAverage Site Concentration Range
o
24-12

-------
Observations from Figure 24-4 include the following:
• All of the naphthalene measurements collected at CHSC are less than the median
naphthalene concentration at the program-level. The annual average concentration
of naphthalene for CHSC is less than the program-level first quartile and roughly
one-sixth the program-level average concentration. There were no non-detects of
naphthalene measured at CHSC or across the program.
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 3.4.3.2.
CHSC has sampled PAHs under the NMP since 2008. Thus, Figure 24-5 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 concentration is not provided, although the range and percentiles are still
presented.
Figure 24-5. Yearly Statistical Metrics for Naphthalene Concentrations Measured at CHSC
Maximum
Concentration for
2009 is 323 ng/m3
2010	2011
Year
O 5th Percentile
O 95th Percentile
• Average
1A 1-year average is not presented because sampling under the NMP did not begin until March 2008.
24-13

-------
Observations from Figure 24-5 for naphthalene measurements collected at CHSC include
the following:
•	CHSC began sampling PAHs under the NMP in late March 2008. Because a full
year's worth of data is not available, a 1-year average concentration is not presented
for 2008, although the range of measurements is provided.
•	The maximum 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
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 1-year average concentration of naphthalene has a slight overall decreasing trend
over the period of sampling, although the confidence intervals calculated for CHSC's
1-year averages are relatively large, particularly for 2009, when the outlier was
measured.
•	With the exception of 2012, the 1-year average concentration has decreased slightly
each year, from 21.71 ng/m3 for 2009 to 12.69 ng/m3 for 2013. The slight bump in the
1-year average concentration for 2012 results from fewer concentrations at the lower
end of the concentration range (the number of naphthalene concentrations less than
10 ng/m3 decreased from 21 to 16 from 2011 to 2012) and more concentrations at the
upper end of the concentration range (the number of concentrations greater than
30 ng/m3 increased from three to seven).
•	All of the statistical parameters shown in Figure 24-5 are at a minimum for 2013.
24.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at the CHSC monitoring site. Refer to Sections 3.2, 3.4.3.3 and 3.4.3.4 for
definitions and explanations regarding the various toxicity factors, time frames, and calculations
associated with these risk-based screenings.
24.5.1 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 3.4.3.3 for an explanation of how cancer risk and
24-14

-------
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 South Carolina Monitoring Site
Pollutant
Cancer
URE
(Ug/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections vs.
# of Samples
Annual
Average
(ng/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Chesterfield, South Carolina - CHSC
Naphthalene
0.000034
0.003
58/58
12.69
±2.33
0.43
<0.01
Observations for CHSC from Table 24-6 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
(0.43 in-a-million).
•	The noncancer hazard approximation for naphthalene is low (less than 0.01),
indicating that no adverse noncancer health effects are expected from this individual
pollutant.
24.5.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 24-7 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
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 3.4.3.4. Lastly,
Table 24-7 provides the pollutants with the highest cancer risk approximations (in-a-million) for
CHSC, 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.
24-15

-------
Table 24-7. 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)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Chesterfield, South Carolina (Chesterfield County) - CHSC
Benzene
26.67
Formaldehyde
2.95E-04
Naphthalene
0.43
Formaldehyde
22.71
Benzene
2.08E-04

Ethylbenzene
13.66
1,3-Butadiene
1.19E-04
Acetaldehyde
12.77
Naphthalene
4.64E-05
1.3 -Butadiene
3.96
POM, Group 2b
3.48E-05
Naphthalene
1.36
Ethylbenzene
3.42E-05
POM, Group 2b
0.40
Arsenic, PM
3.20E-05
POM, Group 2d
0.34
POM, Group 5a
3.01E-05
T etrachloroethy lene
0.31
POM, Group 2d
2.95E-05
T richloroethy lene
0.30
Acetaldehyde
2.81E-05

-------
Table 24-8. 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)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Chesterfield, South Carolina (Chesterfield County) - CHSC
Toluene
98.52
Acrolein
37,454.84
Naphthalene
<0.01
Xylenes
57.96
Formaldehyde
2,316.92

Hexane
45.68
Cyanide Compounds, gas
2,002.83
Methanol
30.61
1.3 -Butadiene
1,978.47
Benzene
26.67
Acetaldehyde
1,419.39
Formaldehyde
22.71
Benzene
888.84
Ethylene glycol
16.72
Xylenes
579.57
Ethylbenzene
13.66
Lead, PM
568.45
Acetaldehyde
12.77
Arsenic, PM
495.85
Methyl isobutyl ketone
4.44
Naphthalene
454.80

-------
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 24.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
Observations from Table 24-7 include the following:
•	Benzene, formaldehyde, and ethylbenzene 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, also appears on both emissions-based lists. POM, Group 5a, which includes
benzo(a)pyrene, ranks eighth for its 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.
Observations from Table 24-8 include the following:
•	Toluene, xylenes, and hexane 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).
24-18

-------
•	Four of the highest emitted pollutants in Chesterfield County also have the highest
toxicity-weighted emissions.
•	Naphthalene ranks 10th for its toxicity-weighted emissions but does not appear
among the highest emitted pollutants in Chesterfield County (of the pollutants with
noncancer RfCs).
24.6 Summary of the 2013 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 second
lowest number of failed screens among NMP sites.
~~~ CHSC has the second lowest annual average concentration of naphthalene among
NMP sites sampling PAHs.
~~~ Concentrations of naphthalene measured at CHSC in 2013 were the lowest since the
onset of sampling in 2008.
24-19

-------
25.0	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.
25.1	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 CBSA and CAMS 85 is part of the Marshall, Texas CBSA. 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
for the site, as reported in the 2011 NEI for point sources, version 2. Note that only sources
within 10 miles of the site are included in the facility counts provided in Figure 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 boundary are still visible on the map for reference, but have been grayed out in order to
emphasize emissions sources within the boundary. Figures 25-3 and 25-4 are the composite
satellite image and emissions sources map for CAMS 85. Table 25-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates.
25-1

-------
Figure 25-1. Deer Park, Texas (CAMS 35) Monitoring Site

-------
Figure 25-2. NEI Point Sources Located Within 10 Miles of CAMS 35
Itarrta
County
1 Houston
SNp CtHHwef
Braio'ia
County
Galveston
County
•5*2l7trw
WW
Legend
Mote: Oue to tac.it/ density and ooHocabor the total facilities
Ji*p*y«d fay r>ot repratanl «li fircittjet *ithm tna e«ea of
10 mile radius 	 County boundaries
~ CAMS 35 NATTS site
Source Category Group (No. of Facilities)
f	AiiBOftt Aalne.Arpoit Seaport Operaitarcs J24)
£	Aspnstt PiodudonfHo* Ml* Asphalt Plant (1 }
B	Buk Tcmtnat*i Suit Planfts 11B |
C	Cnemlcal Manulaceur^g Facility (851
•	Compressor Sutton (9)
6	Elactrical Equfpnent Manutactjrog Faalily «f I
t	ElactfGeneration s>ia Combustion |7>
E	Electrcolating PlaSiog Pclistiing Aroaumg and Coloong 11»
rert«p» Plant 'it
F	Food ProeftSfcng Agneuture Facility 111
f	Ga^P'arUlU
•	ataCTP»irt
•3fr	**trnieum Proouds Manufactunrg 11)
J	Petroleuir R» finery t'5)
R	Ptastie Resin or Rubber Proouds Pant <111
V	»oft and HartoorOperational)
IB	Pulo and Paper Plant 1t1
X I v«rdiRaH L^na Operation Hi
4i	S*i'p3oa! Manutaotunrg or ftapatt Pac«»> i4>
•	Wastewater Treatment Facrtty <21
25-3

-------
Figure 25-3. Karnack, Texas (CAMS 85) Monitoring Site
/

I

w x % \
\A \
S&-
x,, Ik.
m.» * ^ v
O	o


fc>ing
a

IS|\
A'
i, \

i ' |vF\
'f	o, *1	s.
r i-
V •
:0 .
v ^:

,I«v9"
-r*
v ¦
*
>*
a
/
/
Soum'jisf *6» OSfc1
zpo^MleroiofUcoi |>.
\
A
^ »25>t
~\	'
esrr
ej

-------
Figure 25-4. NEI Point Sources Located Within 10 Miles of CAMS 85
LOUISIANA
Hanison
County
Kote. Dub to fac*It> densit) and collocation. the total 'acillwa
dtapayad may not raprasent all r»cHc«a *ithw tha a»aa of ntarasr
Legend
CAMS 85 NATTS site	10 mile radius		| County boundaries
Source Category Group (No. of Facilities)
f AirporVAirline/Alrporl Support Operations (3)
? Miscellaneous Commercial/industrial Facility (1)
25-5

-------
Table 25-1. Geographical Information for the Texas Monitoring Sites
Site Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Additional Ambient Monitoring Information1
CAMS 35
48-201-1039
Deer Park
Harris
Houston-The
Woodlands-Sugar
Land, TX
29.670025,
-95.128508
Residential
Urban/City
Center
Haze, TSP Lead, CO, S02, NOy, NO, N02, NOx,
PAMS/SNMOCs, VOCs, Carbonyl compounds, O3,
Meteorological parameters, PM10, PM Coarse, PM10
Speciation, PM2 5, PM2 5 Speciation, Black Carbon,
IMPROVE Speciation, SVOCs.
CAMS 85
48-203-0002
Karnack
Harrison
Marshall. TX
32.668987,
-94.167457
Agricultural
Rural
SVOCs, N02, NO, NOx, PAMS/SNMOCs, Caibonyl
Compounds, VOCs, 03, Meteorological parameters,
PM10, PM10 Speciation, PM2 5, PM2 5 Speciation,
IMPROVE Speciation.
1 Data for additional pollutants are reported to AQS for these sites (EPA, 2014b); however, these data are not generated by ERG and are therefore not included in this report.
BOLD ITALICS = EPA-designated NATTS Site
to
6\

-------
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 the Spencerview
Athletic Complex (formerly Brown Memorial Field), a 10-acre park with several baseball fields
(Deer Park, 2015). The surrounding area is primarily residential area, as shown in Figure 25-1.
Beltway 8, a major thoroughfare around Houston, is located 1.6 miles to the west of CAMS 35.
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 roughly 4 miles 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 25-2 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 (85) 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; and plastic, resin, or rubber products plants. The point
source located closest to the CAMS 35 monitoring site is a heliport at San Jacinto College's
Central Campus in Pasadena, just over 1 mile southeast of the site. There are no other point
sources within 2 miles of CAMS 35.
The CAMS 85 NATTS site is located in Karnack, in northeast Texas. The monitoring site
is about 12 miles northeast of Marshall, Texas and about 7 miles west of the Texas-Louisiana
border. This site is located on the property of the former Longhorn Army Ammunition Plant near
the intersection of FM Road 134 and Spur Road 449 (Taylor Avenue), as shown in Figure 25-3.
The plant ceased manufacturing munitions in the late 1990s. The property was identified by EPA
as a Superfund site in 1990. Ownership of the property was later transferred from the Army to
the U.S. Fish and Wildlife Service, where the Caddo Lake National Wildlife Refuge was
established (EPA, 2015g). The surrounding area is rural and agricultural. As Figure 25-4 shows,
there are few point sources within 10 miles of CAMS 85 and most these sources all fall into the
airport source category. The closest source to CAMS 85 is the Fly-N-Fish Lodge Airport near
Caddo Lake.
25-7

-------
Table 25-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Texas monitoring sites. Table 25-2 includes both county-level
population and vehicle registration information. Table 25-2 also contains traffic volume
information for each site as well as the location for which the traffic volume was obtained.
Additionally, Table 25-2 presents the county-level daily VMT for Harris and Harrison Counties.
Table 25-2. Population, Motor Vehicle, and Traffic Information for the Texas
Monitoring Sites
Site
County
Estimated
County
Population1
County-level
Vehicle
Registration2
Annual
Average Daily
Traffic3
Intersection
Used for
Traffic Data
County-
level Daily
VMT4 "
CAMS 35
Harris
4,336,853
3,401,957
31,043
Spencer Hwy, from Red Bluff
Rd to Underwood Rd
56,245,209
CAMS 85
Harrison
66,886
72,689
1,250
FM 134 at Spur Road 449
2,511,619
bounty-level population estimates reflect 2013 data (Census Bureau, 2014)
2County-level vehicle registrations reflect 2013 data (TX DMV, 2014)
3AADT reflects 2004 data for CAMS 35 and 2012 data for CAMS 85 (HCPID, 2014 and TX DOT, 2013a)
4County-level VMT reflects 2013 data (TX DOT, 2013b)
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 25-2 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 ranks on the low end 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.
25-8

-------
25.2 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.
25.2.1	Climate Summary
The eastern third of Texas, where the CAMS 35 and CAMS 85 sites are located, 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. 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 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, 2015; TWDR, 1983).
25.2.2	Meteorological Summary
Hourly meteorological data for 2013 were retrieved from NCDC for the weather stations
closest to the Texas monitoring sites (NCDC, 2013), as described in Section 3.4.2. 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 25-3. These data were used to determine how meteorological conditions on sample days
vary from conditions experienced throughout the year.
25-9

-------
Table 25-3. 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
(29.64, -95.28)
9.5
miles
257°
(WSW)
Sample
Days
(30)
76.2
±4.2
67.7
±4.1
56.0
±5.4
61.3
±4.2
69.5
±5.1
1016.5
±2.0
7.4
± 1.0
2013
78.7
+ 1.4
69.8
+ 1.3
58.7
+ 1.5
63.4
+ 1.3
70.7
+ 1.2
1017.4
±0.6
6.6
±0.3
Karnack, Texas - CAMS 85
Shreveport
Regional
Airport
13957
(32.45, -93.82)
25.2
miles
127°
(SE)
Sample
Days
(30)
72.5
±5.1
62.4
±5.1
51.6
±5.5
56.5
±4.8
70.5
±4.1
1016.0
±2.0
6.4
± 1.0
2013
76.1
+ 1.7
65.6
+ 1.6
53.7
+ 1.6
58.8
+ 1.4
69.0
+ 1.3
1017.2
±0.6
6.0
±0.3
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.

-------
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 2013. 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 appear cooler than average weather conditions experienced
throughout the year near both sites. Sampling under the NMP was discontinued at both Texas
monitoring sites at the end of June 2013. Thus, the sample day averages include only samples
days from the first half of the year, thereby missing some of the warmest months of the year.
25.2.3 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 3.4.2. A wind rose shows the frequency
of wind directions using "petals" positioned around a 16-point compass, and uses different colors
to represent wind speeds.
Figure 25-5 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 25-5 also presents three different
wind roses for the CAMS 35 monitoring site. First, a historical wind rose representing 2003 to
2012	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
2013	is presented. Next, a wind rose representing wind data for days on which samples were
collected in 2013 is presented. These can be used to identify the predominant wind speed and
direction for 2013 and to determine if wind observations on sample days were representative of
conditions experienced over the entire year and historically. Figure 25-6 presents the distance
map and three wind roses for CAMS 85.
25-11

-------
Figure 25-5. Wind Roses for the William P. Hobby Airport Weather Station near CAMS 35
Location of CAMS 35 and Weather Station
2003-2012 Historical Wind Rose
	1	:	 	\	





„	" CAMS »

			

ttmathw —


Station





NORTHW-
EST
WIND SPEED
(Knots)
17 - 21
SOUTH
Calms: 14.79%
2013 Wind Rose
Sample Day Wind Rose
NORTH'--.
WIND SPEED
(Kn ots )
17-21
SOUTH
Calms: 15.38%
EST
WIND SPEED
(Knots)
SOUTH
25-12

-------
NORTH'
;SO UTH
Figure 25-6. Wind Roses for the Shreveport Regional Airport Weather Station near
CAMS 85
Location of CAMS 85 and Weather Station	2003-2012 Historical Wind Rose
WIND SPEED
(Knots)
~ »22
P3 17-21
|| 11-17
~
7- 11
4-7
H 2-4
Calms: 15.15%
2013 Wind Rose
NORTH"'-.
WIN C SPEED
(Knots)
17 - 21
11 - 17
SOUTH
Calms: 16.95%
Sample Day Wind Rose
INORTH"^.
est;
WND SPEED
f Kn ots)
17-21
11 - 17
SOUTH *'
Calms: 14.58%
25-13

-------
Observations from Figure 25-5 for CAMS 35 include the following:
•	The Hobby Airport weather station is located 9.5 miles west-southwest of CAMS 35.
•	The historical wind rose shows that southerly and south-southeasterly winds prevail
near CAMS 35, although winds from the southeast quadrant, including easterly
winds, are commonly observed. Northerly winds were also observed often. Calm
winds (those less than or equal to 2 knots) were observed for approximately
15 percent of the wind measurements. Winds from the western quadrants were
infrequently observed.
•	The wind patterns on the wind rose for 2013 resemble the historical wind patterns;
however, the percentage of northerly winds is slightly higher for 2013.
•	The wind patterns shown on the sample day wind rose exhibit some differences from
the wind patterns shown on the full-year and historical wind roses, with fewer
southerly wind observations and a higher percentage of southeasterly and south-
southeasterly wind observations. In addition, winds from the north-northwest were
observed as often as winds from the north. Calm winds were also observed less
frequently. Due to the shortened sampling duration, the sample day wind rose
includes sample day wind data through the first half of 2013 only; a wind rose with a
full year's worth of sample days may look different.
Observations from Figure 25-6 for CAMS 85 include the following:
•	The Shreveport Regional Airport weather station is located across the Texas-
Louisiana border, approximately 25 miles southeast of CAMS 85.
•	The wind patterns on the historical wind rose for CAMS 85 resemble 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 often. Calm winds were
observed for approximately 15 percent of the wind measurements.
•	The wind patterns shown on the 2013 wind rose are similar to the historical wind
patterns, indicating that wind conditions observed in 2013 are similar to those
observed historically.
•	Although southerly winds still prevailed, the wind patterns shown on the sample day
wind rose exhibit some differences from the wind patterns shown on the full-year and
historical wind roses. The primary difference is that winds from the northwest
quadrant were observed more frequently, with winds from the west-northwest to
north together accounting for one-quarter of the observations. Similar to CAMS 35,
the sample day wind rose for CAMS 85 includes wind data through the first half of
2013 only; a wind rose with a full year's worth of sample days may look different.
25-14

-------
25.3 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 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 were sampled for at each site when reviewing the results of
this analysis. Hexavalent chromium was the only pollutant sampled for at both CAMS 35 and
CAMS 85, although sampling was discontinued at the end of June 2013.
Table 25-4. Risk-Based Screening Results for the Texas Monitoring Sites
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Deer Park, Texas - CAMS 35
Hexavalent Chromium
0.000083
2
25
8.00
100.00
100.00
Total
2
25
8.00

Karnack, Texas - CAMS 85
Hexavalent Chromium
0.000083
0
7
0.00
0.00
0.00
Total
0
7
0.00

Observations from Table 25-4 include the following:
•	Hexavalent chromium was detected in 25 of the 30 valid samples collected at
CAMS 35. This pollutant failed two screens, representing an 8 percent failure rate.
•	Hexavalent chromium was detected in seven of the 30 valid samples collected at
CAMS 85. This pollutant did not fail any screens for CAMS 85.
•	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 25.5.1).
25-15

-------
25.4 Concentrations
This section typically presents various concentration averages used to characterize
pollution levels at the monitoring site for each of the site-specific pollutants of interest. However,
the short sampling duration at CAMS 35 prevents an annual average concentration for
hexavalent chromium from being calculated. In order to facilitate a review of the data collected
at CAMS 35 in 2013, a few statistical calculations are provided in the sections that follow. Site-
specific statistical summaries for CAMS 35 (and CAMS 85) are also provided in Appendix O.
The concentration comparison analysis was not performed due to the lack of an annual average
concentration for CAMS 35. The trends analysis was not conducted for this site because
hexavalent chromium sampling under the NMP did not begin at CAMS 35 until 2010 and was
discontinued in June 2013 and therefore does not meet the criteria specified for this data analysis.
25.4.1 2013 Concentration Averages
Quarterly concentration averages were calculated for hexavalent chromium for
CAMS 35, as described above. 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, which
includes all measured detections and substituted zeros for non-detects for the entire year of
sampling, could not be calculated as sampling at CAMS 35 was discontinued at the end of June
2013. Quarterly average concentrations for CAMS 35 are presented in Table 25-5, 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.
25-16

-------
Table 25-5. 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
25/30
0.06
±0.05
0.04
±0.01
NA
NA
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 25-5 include the following:
•	Concentrations of hexavalent chromium measured at CAMS 35 range from
0.0167 ng/m3 to 0.38 ng/m3, including five non-detects.
•	The maximum hexavalent chromium concentration measured at CAMS 35 is the
maximum concentration of this pollutant measured across the program, and is more
than twice the next highest hexavalent chromium concentration.
•	The first quarter average concentration has a relatively large confidence interval
associated with it. A review of the data shows that the range of concentrations
measured is wider for the first quarter. Excluding non-detects, the concentrations
measured during the first quarter range from 0.0167 ng/m3 to 0.38 ng/m3 (which are
the minimum and maximum measured detections), while the concentrations measured
during the second quarter range from 0.0334 ng/m3 to 0.0869 ng/m3. The first quarter
includes only one non-detect while four were measured during the second quarter.
•	Because sampling for hexavalent chromium was discontinued in June 2013, an annual
average concentration could not be calculated.
25.5 Additional Risk-Based Screening Evaluations
In order to characterize risk at participating monitoring sites, additional risk-based
screening evaluations were conducted. Because an annual average concentration could not be
calculated for the pollutant sampled for at CAMS 35, cancer risk and noncancer hazard
approximations, as described in Section 3.4.3.3, were not calculated. The risk-based emissions
assessment described in Section 3.4.3.4 was still conducted, at least in part, as the emissions can
be reviewed independent of concentrations measured.
25-17

-------
25.5.1 Risk-Based Emissions Assessment
This section presents an evaluation of county-level emissions based on cancer and
noncancer toxicity, respectively, and is intended to help policy-makers prioritize their air
monitoring activities. Table 25-6 presents the 10 pollutants with the highest emissions from the
2011 NEI (version 2) that have cancer toxicity factors. Table 25-6 also presents the 10 pollutants
with the highest toxicity-weighted emissions, based on the weighting schema described in
Section 3.4.3.4. The emissions and toxicity-weighted emissions are shown in descending order in
Table 25-6. This information is provided for both counties in which each of the Texas
monitoring sites are located. Table 25-7 presents similar information, but is limited to those
pollutants with noncancer toxicity factors. Because not all pollutants have both cancer and
noncancer toxicity factors, the highest emitted pollutants in the cancer table may be different
from the noncancer table, although the actual quantity of emissions is the same. A more in-depth
discussion of this analysis is provided in Section 3.4.3.4.
Observations from Table 25-6 include the following:
•	Benzene, ethylbenzene, and formaldehyde 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 substantially higher in Harris County than Harrison County.
•	1,3-Butadiene is the pollutant with the highest toxicity-weighted emissions (of the
pollutants with cancer UREs) for Harris County, followed by benzene and benzidine
(gas). Harris County is the only county with an NMP site for which 1,3-butadiene
ranks this high. The pollutants with the highest toxicity-weighted emissions for
Harrison County are formaldehyde, benzene, and ethylene oxide.
•	Five of the highest emitted pollutants in Harris County also have the highest toxicity-
weighted emissions (1,3-butadiene, formaldehyde, benzene, ethylbenzene, and
naphthalene).
•	Formaldehyde and benzene top both emissions-based lists for Harrison County. Four
additional pollutants appear among the highest emitted pollutants in Harrison County
and also are among those with the highest toxicity-weighted emissions (naphthalene,
1,3-butadiene, ethylene oxide, and acetaldehyde).
•	Hexavalent chromium, the only pollutant sampled for at the Texas monitoring sites,
ranks fifth for its toxicity-weighted emissions for Harris County (CAMS 35) and
seventh highest for Harrison County (CAMS 85). This pollutant is not one of the
highest emitted in either county (its emissions rank 29th and 30th, for each respective
county).
25-18

-------
Table 25-6. 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)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Deer Park, Texas (Harris County) - CAMS 35
Benzene
1,159.59
1,3-Butadiene
1.02E-02

Ethylbenzene
736.97
Benzene
9.04E-03
Formaldehyde
665.99
Benzidine, gas
8.83E-03
Acetaldehyde
415.50
Formaldehyde
8.66E-03
1.3 -Butadiene
341.18
Hexavalent Chromium
7.25E-03
Methyl tert butyl ether
109.34
Naphthalene
3.51E-03
Naphthalene
103.35
Nickel, PM
2.28E-03
Propylene oxide
59.07
Ethylene oxide
2.23E-03
Dichloromethane
49.16
Ethylbenzene
1.84E-03
Trichloroethylene
27.72
Acrylonitrile
1.79E-03
Karnack, Texas (Harrison County) - CAMS 85
Formaldehyde
127.17
Formaldehyde
1.65E-03

Benzene
82.70
Benzene
6.45E-04
Acetaldehyde
55.09
Ethylene oxide
5.91E-04
Ethylbenzene
35.08
Naphthalene
5.14E-04
Naphthalene
15.11
Nickel, PM
3.50E-04
1,3-Butadiene
11.35
1,3-Butadiene
3.40E-04
Ethylene oxide
6.72
Hexavalent Chromium
1.88E-04
Dichloromethane
2.61
Arsenic, PM
1.39E-04
Tetrachloroethylene
1.71
Acetaldehyde
1.21E-04
Benzyl chloride
1.37
POM, Group 2b
9.98E-05

-------
Table 25-7. 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)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Deer Park, Texas (Harris County) - CAMS 35
Toluene
4,895.12
Acrolein
2,446,074.13

Hexane
4,300.06
1,3-Butadiene
170,588.08
Methanol
2,807.01
Chlorine
120,849.93
Xylenes
2,750.29
Titanium tetrachloride
77,090.00
Benzene
1,159.59
Formaldehyde
67,958.53
Ethylene glycol
824.11
Nickel, PM
52,732.60
Ethylbenzene
736.97
Acetaldehyde
46,166.51
Formaldehyde
665.99
Benzene
38,653.10
Methyl isobutyl ketone
609.10
Cadmium PM
37,230.00
Acetaldehyde
415.50
Naphthalene
34,450.67
Karnack, Texas (Harrison County) - CAMS 85
Toluene
189.46
Acrolein
647,248.74

Xylenes
172.55
Hexamethylene-1,6-diisocyanate, gas
48,091.54
Formaldehyde
127.17
Chlorine
22,538.67
Hexane
111.73
Formaldehyde
12,977.02
Ethylene glycol
91.50
Cyanide Compounds, PM
9,151.68
Benzene
82.70
Nickel, PM
8,094.58
Acetaldehyde
55.09
Maleic anhydride
7,969.71
Methanol
50.13
Acetaldehyde
6,120.78
Chloromethane
48.40
1,3-Butadiene
5,674.42
Ethylbenzene
35.08
Naphthalene
5,037.99

-------
Observations from Table 25-7 include the following:
•	Toluene is the highest emitted pollutant with a noncancer RfC in both Harris and
Harrison Counties. The magnitude of the emissions is substantially 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.
•	Three of the highest emitted pollutants also have the highest toxicity-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.
•	Hexavalent chromium appears on neither emissions-based list for Harris County,
ranking 62nd for its total emissions and 23rd for its toxicity-weighted emissions (of
the pollutants with noncancer RfCs).
•	Hexavalent chromium appears on neither emissions-based list for Harrison County,
ranking 60th for its total emissions and 30th for its toxicity-weighted emissions (of
the pollutants with noncancer RfCs).
25.6 Summary of the 2013 Monitoring Data for CAMS 35 and CAMS 85
Results from several of the data treatments described in this section include the
following:
~~~ Hexavalent chromium was the only pollutant sampled for at CAMS 35 and CAMS 85
in 2013. Sampling was discontinued at these locations at the end of June.
~~~ Hexavalent chromium was detected in more than 80 percent of samples collected at
CAMS 35 andfailed two screens.
~~~ Hexavalent chromium was detected in fewer than 25 percent of samples collected at
CAMS 85. Concentrations of hexavalent chromium did not fail any screens for
CAMS 85.
~~~ The highest concentration of hexavalent chromium across the program was measured
at CAMS 35.
25-21

-------
26.0	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.
26.1	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 26-1 is a
composite satellite image retrieved from ArcGIS Explorer showing the monitoring site and its
immediate surroundings. Figure 26-2 identifies nearby point source emissions locations by
source category, as reported in the 2011 NEI for point sources, version 2. Note that only sources
within 10 miles of the site are included in the facility counts provided in Figure 26-2. A 10-mile
boundary was chosen to give the reader an indication of which emissions sources and emissions
source categories could potentially have a direct effect on the air quality at the monitoring site.
Further, this boundary provides both the proximity of emissions sources to the monitoring site as
well as the quantity of such sources within a given distance of the site. Sources outside the
10-mile boundary are still visible on the map for reference, but have been grayed out in order to
emphasize emissions sources within the boundary. Table 26-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates.
26-1

-------
Figure 26-1. Bountiful, Utah (BTUT) Monitoring Site
ittuaaoN'

{ II »-'1 A .-o
10 Common*
£ Pages Ln
L&OQHu

-------
Figure 26-2. NEI Point Sources Located Within 10 Miles of BTUT
1M-S6V*
Morfljn
County
' Davn
i County
Salt Lake
County
Wot# Due to tacilty 
-------
Table 26-1. Geographical Information for the Utah Monitoring Site
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Additional Ambient Monitoring Information1
BTUT
49-011-0004
Bountiful
Davis
Ogden-Clearfield,
UT
40.902967,
-111.884467
Residential
Suburban
SO2, NO, NO2, NOx, O3, Meteorological parameters,
PM10, PM2 5, PM2 5 Speciation, Black Carbon,
IMPROVE Speciation.
1 Data for additional pollutants are reported to AQS for BTUT (EPA, 2014b); however, these data are not generated by ERG and are therefore not included in this report.
BOLD ITALICS = EPA-designated NATTS Site
to
On

-------
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 26-1 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 26-2 shows that most of the
point sources near BTUT are located to the south of the site and run parallel to 1-15. The
facilities surrounding BTUT are involved in a variety of industries, although the source
categories with the greatest number of point sources surrounding BTUT are the airport and
airport support operations category, 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, a painting and coatings manufacturer, and a landfill.
Table 26-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Utah monitoring site. Table 26-2 includes both county-level
population and vehicle registration information. Table 26-2 also contains traffic volume
information for BTUT as well as the location for which the traffic volume was obtained.
Additionally, Table 26-2 presents the county-level daily VMT for Davis County.
Table 26-2. Population, Motor Vehicle, and Traffic Information for the Utah Monitoring
Site




Annual

County-


Estimated
County-level
Average
Intersection
level


County
Vehicle
Daily
Used for
Daily
Site
County
Population1
Registration2
Traffic3
Traffic Data
VMT4
BTUT
Davis
322,094
274,716
130,950
1-15, N of Hwy 89 junction
6,950,795
bounty-level population estimate reflects 2013 data (Census Bureau, 2014)
2County-level vehicle registration reflects 2013 data (UT TC. 2013)
3AADT reflects 2012 data (UT DOT, 2012)
4County-level VMT reflects 2013 data (UT DOT, 2014)
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 26-2 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.
26-5

-------
•	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.
26.2 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.
26.2.1	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, 2014).
26.2.2	Meteorological Summary
Hourly meteorological data for 2013 were retrieved from NCDC for the weather station
closest to the Utah monitoring site (NCDC, 2013), as described in Section 3.4.2. 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 26-3. These data were
used to determine how meteorological conditions on sample days vary from conditions
experienced throughout the year.
26-6

-------
Table 26-3. Average Meteorological Conditions near the Utah Monitoring Site
Closest Weather
Station
Distance
and

Average
Maximum
Average
Average
Dew Point
Average
Wet Bulb
Average
Relative
Average
Sea Level
Average
Scalar Wind
(WBAN and
Coordinates)
Direction
from Site
Average
Type1
Temperature
(°F)
Temperature
(°F)
Temperature
(°F)
Temperature
(°F)
Humidity
(%)
Pressure
(mb)
Speed
(kt)
Bountiful, Utah - BTUT
Salt Lake City
9.7
miles
Sample
Days
64.1
54.3
31.6
43.0
48.8
1016.5
6.0
International
(78)
±5.6
±5.1
±3.0
±3.5
±4.3
±2.1
±0.6
24127
207°
(SSW)








(40.78, -111.97)
2013
63.0
+ 2.5
53.3
±2.3
31.3
+ 1.3
42.5
+ 1.6
50.2
±2.1
1016.4
±0.9
6.1
±0.3
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
to
On

-------
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 2013. Also included in Table 26-3 is the 95 percent
confidence interval for each parameter. As shown in Table 26-3, average meteorological
conditions on sample days near BTUT were representative of average weather conditions
experienced throughout the year. Note that the number of sample days (78) is larger than a
l-in-6 day sampling schedule would typically present; a number of make-up samples were
collected at BTUT, primarily between April and September as well as December.
As indicated in the previous section, BTUT is located in a relatively dry climate. The
average relative humidity shown in Table 26-3 is the second lowest, second only to the relative
humidity calculated for the Phoenix, Arizona sites. The average dew point temperature
calculated for BTUT is also among the lowest compared to other NMP sites.
26.2.3 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 3.4.2. A wind rose shows the frequency of wind directions using "petals"
positioned around a 16-point compass, and uses different colors to represent wind speeds.
Figure 26-3 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 26-3 also presents three different wind roses for the
BTUT monitoring site. First, a historical wind rose representing 2003 to 2012 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 2013 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2013 is
presented. These can be used to identify the predominant wind speed and direction for 2013 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.
26-8

-------
Figure 26-3. Wind Roses for the Salt Lake City International Airport Weather Station near
BTUT
Location of BTUT and Weather Station	2003-2012 Historical Wind Rose
WIND SPEED
(Knots)
I I »22
[ B 17-21
IH 11 -1?
I. I 7- 11
~ 4-7
2- 4
Calms: 11.82%
est:
2013 Wind Rose
Sample Day Wind Rose

NORTH""--,^

v\ 20%

1 16%

[ 12%

8%.

>¦ 4% : :
WESTf » \ #l
; ; east
WIND SPEED
(Knots)
~	=22
n 17-21
11 -17
I 1 7- 11
~	4-7
H 2-4
Calms: 14.84%
WEST
WIND SPEED
(Kn ots)
SOUTH
Calms: 15.60%
26-9

-------
Observations from Figure 26-3 for BTUT include the following:
•	The Salt Lake City International Airport weather station is located 9.7 miles south-
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 commonly
observed. Winds from the northeast and southwest quadrants were rarely observed.
Calm winds (those less than or equal to 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 2013 wind rose are similar to the historical wind
patterns, indicating that wind conditions in 2013 were similar to wind conditions
experienced historically near BTUT. There are, however, a few differences between
the historical and the 2013 wind roses. The 2013 wind rose has a higher percentage of
calm winds than the historical wind rose, with nearly 15 percent of the observations
identified as calm. Also, the number of south-southeasterly and southerly winds
observed is less than on the historical wind rose.
•	The wind patterns shown on the sample day wind rose are very similar to the full-year
wind patterns.
26.3 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 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 each site sampled for when reviewing the results of this
analysis. VOCs, carbonyl compounds, SNMOCs, PAHs, metals (PMio), and hexavalent
chromium were sampled for at BTUT, although sampling for hexavalent chromium was
discontinued in June 2013. BTUT is one of only two NMP sites sampling the entire suite of
pollutants under the NMP (NBIL is the other).
26-10

-------
Table 26-4. Risk-Based Screening Results for the Utah Monitoring Site
Pollutant
Screening
Value
Oig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Bountiful, Utah - BTUT
Acetaldehyde
0.45
55
55
100.00
11.29
11.29
Formaldehyde
0.077
55
55
100.00
11.29
22.59
Benzene
0.13
53
53
100.00
10.88
33.47
Carbon Tetrachloride
0.17
52
53
98.11
10.68
44.15
Arsenic (PMio)
0.00023
49
59
83.05
10.06
54.21
1.3 -Butadiene
0.03
47
50
94.00
9.65
63.86
Naphthalene
0.029
42
56
75.00
8.62
72.48
1,2-Dichloroethane
0.038
38
38
100.00
7.80
80.29
Propionaldehyde
0.8
35
55
63.64
7.19
87.47
Ethylbenzene
0.4
13
53
24.53
2.67
90.14
Dichloromethane
60
12
53
22.64
2.46
92.61
Nickel (PMio)
0.0021
10
59
16.95
2.05
94.66
/?-Dichlorobcnzcnc
0.091
8
19
42.11
1.64
96.30
Cadmium (PMio)
0.00056
3
59
5.08
0.62
96.92
Hexachloro -1,3 -butadiene
0.045
3
3
100.00
0.62
97.54
Lead (PMK,)
0.015
3
59
5.08
0.62
98.15
Xylenes
10
3
53
5.66
0.62
98.77
Acenaphthylene
0.011
2
34
5.88
0.41
99.18
Benzo(a)pyrene
0.00057
2
22
9.09
0.41
99.59
Chloroprene
0.0021
1
1
100.00
0.21
99.79
1,2-Dibromoethane
0.0017
1
1
100.00
0.21
100.00
Total
487
890
54.72

Observations from Table 26-4 include the following:
•	Twenty-one pollutants failed at least one screen for BTUT; nearly 55 percent of
concentrations for these 21 pollutants were greater than their associated risk screening
value (or failed screens). BTUT tied with S4MO for the highest number of individual
pollutants failing 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, seven VOCs, two PMio metals, and one PAH.
•	Acetaldehyde, formaldehyde, and benzene were detected in every valid carbonyl
compound and VOC sample collected at BTUT and failed 100 percent of screens.
Other pollutants also failed 100 percent of screens but were detected less frequently.
26-11

-------
•	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.
26.4 Concentrations
This section presents various concentration averages used to characterize air toxics
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
each monitoring site.
•	Annual concentration averages are presented graphically for each site to illustrate
how the site's concentrations compare to the program-level averages, as presented in
Section 4.1.
•	Concentration averages and other statistical metrics are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at BTUT are provided in Appendix J through Appendix O.
26.4.1 2013 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
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 26-5, where applicable. Fourth quarter average concentrations could not be
calculated for the VOCs because fewer than 75 percent of the samples collected were valid. First
26-12

-------
quarter average concentrations could not be calculated for the PAHs because the sampler was not
operating properly and was not repaired in time for make-up samples to be collected. Note that
concentrations of the PAHs and PMio metals are presented in ng/m3 in Table 26-5 for ease of
viewing. Also note that if a pollutant was not detected in a given calendar quarter, the quarterly
average concentration simply reflects "0" because only zeros substituted for non-detects were
factored into the quarterly average concentration.
Table 26-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Utah Monitoring Site

# of






Measured
1st
2nd
3rd
4th


Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs. # of
Samples
Average
(Hg/m3)
Average
(Hg/m3)
Average
(Hg/m3)
Average
(Hg/m3)
Average
(Hg/m3)
Bountiful, Utah - BTUT


5.44
3.99
4.26
3.11
4.18
Acetaldehyde
55/55
±0.85
±0.48
±0.56
±0.58
±0.36


1.19
0.58
0.68

0.94
Benzene
53/53
±0.32
±0.13
±0.17
NA
±0.16


0.11
0.04
0.04

0.09
1.3 -Butadiene
50/53
±0.04
±0.01
±0.02
NA
±0.02


0.57
0.58
0.58

0.56
Carbon Tetrachloride
53/53
±0.06
±0.04
±0.02
NA
±0.03


0.02
0.06
0.04

0.05
p-Dichlorobenzene
19/53
±0.02
±0.08
±0.04
NA
±0.03


0.06
0.10
0.02

0.11
1,2-Dichloroethane
38/53
±0.03
±0.03
±0.02
NA
±0.03


17.77
332.69
451.06

225.03
Dichloromethane
53/53
±26.41
± 674.34
± 274.86
NA
±219.72


0.32
0.24
0.28

0.49
Ethylbenzene
53/53
±0.10
±0.08
±0.06
NA
±0.24


10.27
9.15
8.71
3.78
8.05
Formaldehyde
55/55
± 1.72
± 1.03
± 1.15
±0.86
±0.87


1.09
0.93
1.00
0.53
0.89
Propionaldehyde
55/55
±0.13
±0.10
±0.11
±0.11
±0.08


0.96
0.41
0.45
2.23
0.99
Arsenic (PMi0)a
59/59
±0.58
±0.11
±0.13
± 1.46
±0.40



28.11
40.70
79.93
55.48
Naphthalene3
56/56
NA
±7.18
±5.85
±20.41
± 11.39


1.40
0.89
1.37
2.17
1.44
Nickel (PMi,;,)a
59/59
±0.49
±0.21
±0.26
±0.69
±0.24
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.
26-13

-------
Observations for BTUT from Table 26-5 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 for
2013, but is considerably higher than the annual averages calculated for 2012 and
2011. The annual average concentration for 2013 has a very large confidence interval
associated it, indicating the likely presence of outliers, as do the quarterly average
concentrations. A review of the data shows that concentrations of dichloromethane
measured at BTUT in 2013 range from 0.585 |ig/m3 to 5,604 |ig/m3. The maximum
concentration of this pollutant was measured on June 3, 2013 and is one of three
dichloromethane concentrations greater than 1,000 |ig/m3 measured at this site.
Eleven of the 12 dichloromethane concentrations greater than 100 |ig/m3 measured
across the program were measured at BTUT. The median concentration of
dichloromethane for BTUT is 5.85 |ig/m3, which is greater than all but one of the
other NMP sites annual average dichloromethane concentrations, indicating that the
statistics for this site are not being thrown off just by one or two outliers. Four of the
five highest dichloromethane concentrations measured at BTUT were measured
between August 1, 2013 and September 1, 2013 and 10 of the 12 concentrations
greater than 50 |ig/m3 were measured between June and September (with the other
two in January and December). All of the concentrations less than 2 |ig/m3 were
measured during the first half of 2013 and predominantly during the second quarter of
the year (with two measured during the first quarter, nine measured during the
second, and one measured during the third).
•	Based on the quarterly average concentrations of formaldehyde, concentrations
measured during the fourth quarter of the year are significantly lower than those
measured during the rest of the year. A review of the data shows that formaldehyde
concentrations measured at BTUT range from 2.13 |ig/m3 to 14.9 |ig/m3. All but one
of the 13 concentrations less than 5 |ig/m3 were measured between October and
December, with the maximum concentration measured during this period less than the
median formaldehyde concentration calculated for the year. Conversely, the five
highest concentrations were measured at BTUT between the end of January and the
end of February and were greater than or equal to 12 |ig/m3. Similar observations can
be made for propionaldehyde and, to a lesser extent, acetaldehyde.
•	Based on the three quarterly average concentrations of benzene available for BTUT,
concentrations are significantly higher during the first quarter. A review of the data
shows that the maximum concentration of benzene (3.65 |ig/m3) was actually
measured in December. Concentrations measured in December account for five of the
nine highest benzene concentrations measured at this site, with the other four
measured in either January or February. Of the 18 benzene concentrations greater
than 1 |ig/m3 measured at BTUT, most were measured during the colder months of
the year (six during the first quarter, one during the second quarter, two during the
third quarter, and nine during the fourth quarter). Similar observations can be made
for 1,3-butadiene and, to a lesser extent, ethylbenzene. Note that for ethylbenzene, the
annual average concentration shown in Table 26-5 is greater than the three available
26-14

-------
quarterly average concentrations. The three ethylbenzene concentrations measured
between December 13, 2013 and December 18, 2013 are the only ethylbenzene
concentrations greater than 1 |ig/m3 measured at BTUT and range from 2.85 |ig/m3 to
5.53 |ig/m3, all three of which are among the higher ethylbenzene concentrations
measured across the program.
Similar to ethylbenzene, the annual average concentration of 1,2-dichloroethane
shown in Table 26-5 is greater than the three available quarterly average
concentrations. Five of the six highest 1,2-dichloroethane concentrations measured at
BTUT were measured in December, with concentrations measured during the fourth
quarter comprising 11 of the 13 highest 1,2-dichloroethane concentrations measured
at this site.
The available quarterly average concentrations of />dichlorobenzene each have
relatively large confidence intervals, particularly the second quarter, for which the
confidence interval is larger than the average itself. A review of the data shows that
this pollutant was detected in 36 percent of the samples collected, such that many
zeroes substituted for non-detects are included in each quarterly average (the number
of measured detections ranges from three to four for the first three calendar quarters
and eight for fourth quarter). The two highest concentrations of />dichlorobenzene
were measured at BTUT during the second quarter, on May 16, 2013 (0.681 |ig/m3)
and April 28, 2013 (0.205 |ig/m3), the first of which is the maximum
/;-dichlorobenzene concentration measured across the program.
The fourth quarter average arsenic concentration is considerably higher than the other
quarterly averages and has a relatively large confidence interval associated with it
(although this is also true for the first quarter average). A review of the data shows
that the three highest arsenic concentrations measured at BTUT were measured in
November and December and range from 5.22 ng/m3 to 9.18 ng/m3, accounting for
three of the five highest arsenic concentrations measured across the program,
including the maximum. All 11 arsenic concentrations greater than or equal to
1 ng/m3 were measured at BTUT during the first (five) or fourth (six) quarters of the
year.
Concentrations of nickel measured at BTUT range from 0.31 ng/m3 to 5.08 ng/m3,
with seven of the nine highest concentrations of nickel measured at BTUT between
October and December (with the other two measured in January and February). Of
the 20 highest nickel concentrations measured at BTUT, six were measured during
the first quarter, none were measured during the second quarter, four were measured
during the third quarter, and 10 were measured during the fourth quarter.
Concentrations of naphthalene appear highest during the fourth quarter of the year,
based on the available quarterly average concentrations. Concentrations of
naphthalene measured at BTUT range from 8.29 ng/m3 to 242 ng/m3, with the
maximum concentration of naphthalene measured on January 4, 2013. The seven
concentrations of naphthalene greater than 100 ng/m3 were all measured in January,
November, or December. Of the 15 concentrations greater than 65 ng/m3 measured at
26-15

-------
BTUT, five were measured during the first quarter and 10 were measured during the
fourth quarter.
Tables 4-9 through 4-12 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for BTUT from
those tables include the following:
•	BTUT appears in Table 4-9 through 4-12 a total of seven times for the program-level
pollutants of interest.
•	BTUT is listed for three of the program-level VOC pollutants of interest shown in
Table 4-9. BTUT ranks highest for ethylbenzene, ranking fifth among other NMP
sites sampling this pollutant. BTUT also ranks sixth for 1,2-dichloroethane and 10th
for />dichlorobenzene.
•	For the third 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 ranks highest for its 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 second highest for its annual average concentration of arsenic, as shown
in Table 4-12. The annual average arsenic concentration calculated for BTUT has the
highest confidence interval shown in Table 4-12 for arsenic. BTUT's annual average
concentration ranks fourth highest for nickel.
26.4.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for the pollutants shaded in
gray in Table 26-4 for BTUT. Figures 26-4 through 26-16 overlay the site's minimum, annual
average, and maximum concentrations onto the program-level minimum, first quartile, median,
average, third quartile, and maximum concentrations, as described in Section 3.4.3.1.
26-16

-------
Figure 26-4. Program vs. Site-Specific Average Acetaldehyde Concentration
0
3
6 9
Concentration {[jg/m3)

12
15

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i


Site: Site Average
o
Site Concentration Range



Figure 26-5. Program vs. Site-Specific Average Arsenic (PMio) Concentration
E
0
12 3
4 5 6
Concentration {ng/m3)
7
8
9
10

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i



Site: Site Average
o
Site Concentration Range




Figure 26-6. Program vs. Site-Specific Average Benzene Concentration
¦+
Program Max Concentration = 43.5 ^ig/m3
0
2 4
6
Concentration {[jg/m3)
8
10

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i

Site: Site Average
o
Site Concentration Range


26-17

-------
Figure 26-7. Program vs. Site-Specific Average 1,3-Butadiene Concentration

,


Program Max Concentration = 21.5 ^ig/m3





0	0.3	0.6	0.9	1.2	1.5
Concentration {[ig/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


Figure 26-8. Program vs. Site-Specific Average Carbon Tetrachloride Concentration
Program Max Concentration = 23.7 ^ig/m3
0	0.25	0.5	0.75	1	1.25	1.5	1.75	2
Concentration {[ig/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


Figure 26-9. Program vs. Site-Specific Average />-Dichlorobenzene Concentration
0.3	0.4
Concentration {[jg/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


26-18

-------
Figure 26-10. Program vs. Site-Specific Average 1,2-Dichloroethane Concentration


¦


Program Max Concentration = 111 ^ig/m3
0
0



0	0.2	0.4	0.6	0.8	1
Concentration {[ig/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


Figure 26-11. Program vs. Site-Specific Average Dichloromethane Concentration

Program Max Concentration = 5,604 jig/m3







0	50	100	150	200	250	300	350
Concentration {[ig/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


Figure 26-12. Program vs. Site-Specific Average Ethylbenzene Concentration




Program Max Concentration = 18.7 ^ig/m3





		





H	1	1	1	r
0	1	2	3	4	5	6
Concentration {[ig/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


26-19

-------
Figure 26-13. Program vs. Site-Specific Average Formaldehyde Concentration
0
3 6
9 12 15
Concentration {[jg/m3)
18
21

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i

Site: Site Average
o
Site Concentration Range


Figure 26-14. Program vs. Site-Specific Average Naphthalene Concentration
—
0
100
200
300 400 500
Concentration {ng/m3)
600
700

Program:
1st Qua rti le
¦
2nd Qua rti le 3rd Qua rti le
~ ~
4th Qua rti le
~
Average
i

Site:
Site Average
o
Site Concentration Range


Figure 26-15. Program vs. Site-Specific Average Nickel (PMio) Concentration
0
5
10 15
Concentration {ng/m3)

20

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i

Site: Site Average
o
Site Concentration Range


26-20

-------
Figure 26-16. Program vs. Site-Specific Average Propionaldehyde Concentration
0.25
0.5
0.75 1 1.25
Concentration {[jg/m3)
1.5
1.75
Program:
1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site:
Site Average
o
Site Concentration Range


Observations from Figures 26-4 through 26-16 include the following:
•	Figure 26-4 shows that the maximum acetaldehyde concentration measured at
BTUT is roughly half the maximum acetaldehyde concentration measured at the
program-level. The box plot also shows that the minimum acetaldehyde
concentration measured at BTUT (1.49 |ig/m3) is equal to the program-level
median concentration. BTUT is one of only two NMP sites' whose minimum
concentration is greater than 1 |ig/m3. The annual average acetaldehyde
concentration for BTUT is more than twice the program-level average
concentration, is greater than the program-level third quartile, and is the highest
annual average concentration among NMP sites sampling carbonyl compounds.
•	Figure 26-5 for arsenic shows that BTUT's maximum arsenic concentration is the
maximum arsenic concentration measured at the program-level. The annual
average concentration calculated for BTUT is greater than the program-level
average concentration and third quartile and is the second highest among NMP
sites sampling PMio metals.
•	Figure 26-6 presents the box plot for benzene. Note that the program-level
maximum benzene concentration (43.5 |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 12 |ig/m3. Figure 26-6 shows that the maximum concentration of
benzene measured at BTUT is an order of magnitude less than the maximum
concentration measured across the program. The annual average concentration for
BTUT is just greater than the program-level average concentration and third
quartile.
•	Figure 26-7 presents the box plot for 1,3-butadiene. Note that the program-level
maximum 1,3-butadiene concentration (21.5 |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 1.5 |ig/m3. Figure 26-7 shows that the program-level average is greater
than the program-level third quartile, indicating that concentrations at the upper
end of the concentration range are driving the program average upward. The
26-21

-------
annual average concentration for BTUT is similar to the program-level third
quartile. There were three non-detects of 1,3-butadiene measured at BTUT in
2013.
Figure 26-8 presents the box plot for carbon tetrachloride. Similar to benzene and
1,3-butadiene, the program-level maximum concentration (23.7 |ig/m3) is not
shown directly on the box plot 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. The minimum carbon tetrachloride concentration measured at BTUT is the
second lowest concentration of this pollutant measured across the program. The
annual average concentration of carbon tetrachloride for BTUT is just less than
the program-level first quartile and is the lowest annual average concentration
among NMP sites sampling this pollutant, although the range of averages is
relatively small for most of the sites.
Figure 26-9 presents the box plot for /;-dichlorobenzene. Note that the program-
level first and second quartiles are both zero and therefore not visible on the box
plot. The maximum />dichlorobenzene concentration measured at BTUT is the
maximum concentration measured across the program, although more than half of
the measurements collected at BTUT were non-detects. The annual average
concentration for BTUT is just greater than the program-level average
concentration and ranks 10th highest among NMP sampling this pollutant.
Figure 26-10 is the box plot for 1,2-dichloroethane. Similar to other pollutants,
the program-level maximum concentration (111 |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 more than twice the program third quartile
for this pollutant and is greater than 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 26-10 shows that the maximum 1,2-dichloroethane concentration
measured at BTUT is greater than the program-level average concentration
(BTUT is the only non-Calvert City, Kentucky site for which this is true) but is
still half the scale of the box plot. The annual average concentration for BTUT is
similar to the program-level third quartile and ranks sixth highest among NMP
sites sampling this pollutant. Fifteen non-detects of 1,2-dichloroethane were
measured at BTUT.
Although the maximum dichloromethane concentration across the program was
measured at BTUT (5,604 |ig/m3), the scale of the box plot in Figure 26-11 was
reduced, although the first, second, and third quartiles are still relatively unclear.
What is clear, though is that a high percentage of the dichloromethane
concentrations measured across the program fall below the minimum
concentration measured at BTUT. As discussed in the previous section,
dichloromethane concentrations measured at BTUT account for all but one of the
12 measurements greater than 100 |ig/m3 across the program. The maximum
dichloromethane concentration measured at BTUT is more than 20 times the next
26-22

-------
highest concentration measured at another NMP site. Concentrations of
dichloromethane measured at BTUT typically run high compared to other NMP
sites, but concentrations measured in 2013 are particularly high compared to past
years. The program-level average concentration (8.17 |ig/m3) is an order of
magnitude greater than third quartile (0.90 |ig/m3), indicating that while most of
the dichloromethane concentrations measured across the program are less than
1 |ig/m3, the concentrations at the upper end of the range are driving that
program-level average. BTUT is the only site for which dichloromethane is a
pollutant of interest and has the highest annual average concentration of
dichloromethane among sites sampling this pollutant (the annual average
concentration for BTUT is 15 times greater than the next highest annual average
for an NMP sites sampling VOCs).
Similar to many of the other VOCs, the program-level maximum ethylbenzene
concentration (18.7 |ig/m3) is not shown directly on the box plot in Figure 26-12
as the scale has been reduced to 6 |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
program-level average concentration is similar to the program-level third quartile,
both of which are less than the annual average ethylbenzene concentration for
BTUT. The annual average ethylbenzene concentration for BTUT is the fifth
highest annual average concentration among NMP sites sampling this pollutant.
Figure 26-13 shows that the annual average formaldehyde concentration for
BTUT is nearly three times greater than the program-level average and more than
twice the program-level third quartile. As discussed in the previous section,
BTUT has the highest annual average formaldehyde concentration among NMP
sites sampling carbonyl compounds. The minimum concentration measured at
BTUT is just less than the program-level median concentration, meaning that
nearly half of the formaldehyde concentrations measured across the program are
less than BTUT's minimum formaldehyde concentration. Even though the
maximum formaldehyde concentration was not measured at BTUT, this site has
the greatest number of formaldehyde concentrations greater than 10 |ig/m3 among
NMP sites sampling carbonyl compounds (15, compared eight for GPCO and two
or less for four additional sites).
Figure 26-14 is the box plot for naphthalene, which shows that the annual average
naphthalene concentration for BTUT is less than the program-level average
concentration and similar to the program-level median concentration. The annual
average concentration of naphthalene for BTUT ranks 17th among the 22 sites for
which annual averages could be calculated.
Figure 26-15 is the box plot for nickel (PMio). The maximum concentration of
nickel measured at BTUT is about one-quarter of the program-level maximum
concentration. The annual average concentration of nickel for BTUT is greater
than the program-level average concentration and just greater than the program-
level third quartile. The minimum concentration of nickel measured at BTUT is
just less than the program-level first quartile.
26-23

-------
• Figure 26-16 shows that concentrations of propionaldehyde measured at BTUT
are on the higher end of the concentration range, as the entire range of
concentrations measured at BTUT is greater than the program-level median
concentration. BTUT is one of only two NMP sites for which propionaldehyde is
a pollutant of interest and has the highest annual average concentration of this
pollutant among NMP sites sampling carbonyl compounds. The annual average
propionaldehyde concentration for BTUT is twice the program-level average
concentration, although the maximum propionaldehyde concentration was not
measured at BTUT.
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 3.4.3.2.
BTUT has sampled carbonyl compounds, VOCs, metals, and SNMOCs under the NMP since
2003 and PAHs since 2008. Thus, Figures 26-17 through 26-29 present the 1-year statistical
metrics for each of the pollutants of interest for BTUT. The statistical metrics presented for
assessing trends include the substitution of zeros for non-detects. If sampling began mid-year, a
minimum of 6 months of sampling is required for inclusion in the trends analysis; in these cases,
a 1-year average concentration is not provided, although the range and percentiles are still
presented.
26-24

-------
Figure 26-17. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at BTUT
—J—i




rh

rh








—®-
	0	
¦0"

L-2-J

—Q—

tium
I
2004	2005	2006	2007
2008	2009
Year
2011	2012
0 5th Percentile
0 95th Percentile
Observations from Figure 26-17 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 26-17 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 2005.
•	After 2005, the 1-year average concentration exhibits a steady decreasing trend
through 2009, when the 1-year average concentration reaches a minimum
(1.97 |ig/m3), although the most significant changes occurred between 2005 and
2007. Between 2007 and 2011, the 1-year average concentration varied by less than
0.30 |ig/m3, ranging from 1.97 |ig/m3 (2009) to 2.25 |ig/m3 (2010).
•	Although the range of concentrations measured in 2012 is smaller than the range
measured in 2011, a slight increase is shown in both the 1-year average and median
concentrations for 2012. This slight increase is followed by a significant increase for
2013, when both the 1-year average and median concentrations are at a maximum for
the period of sampling.
26-25

-------
Figure 26-18. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at BTUT
Maximum
Concentration for
2004 is 33.0 ng/nv
2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	— Maximum	O 95th Percentile	Average
Observations from Figure 26-18 for arsenic measurements collected at BTUT include the
following:
•	Sampling for PMio metals under the NMP began at BTUT in late July 2003. Because
this represents less than half of the sampling year, Figure 26-18 excludes data from
2003.
•	The maximum arsenic concentration was 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 14 highest concentrations of arsenic (those greater than 5 ng/m3)
were measured in 2004.
•	Of the 24 highest arsenic concentrations measured at BTUT, 13 were measured
during the first quarter of the calendar year and 11 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) and 1.13 ng/m3 (2009). However, the statistical parameters for 2007 and 2009
are being driven primarily by a single "high" measurement. If the maximum
26-26

-------
concentrations measured in 2007 and 2009 were removed from the data sets, the
1-year average concentrations for this period would all be less than 1 ng/m3.
•	The smallest range of arsenic concentrations was measured at BTUT in 2012, when
the 1-year average concentration is at a minimum. The maximum arsenic
concentration measured in 2012 is less than 2 ng/m3, the only year for which this is
true, and is less than the 95th percentile for several other years of sampling.
•	Concentrations of arsenic measured at BTUT increased significantly for 2013, as
indicated by the increase shown in all of the statistical parameters. Although the
1-year average concentration doubled from 2012 to 2013, the increase in the median
concentration is less dramatic.
Figure 26-19. Yearly Statistical Metrics for Benzene Concentrations Measured at BTUT
2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	— Maximum	O 95th Percentile ••••v—* Average
Observations from Figure 26-19 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 26-19 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.
26-27

-------
•	Concentrations of benzene appear to be higher during the colder months of the year,
as 50 of the 54 highest concentrations (those greater than 2.50 |ig/m3) were measured
during the first (28) or fourth (22) quarters of the calendar year.
•	The 1-year average and median benzene concentrations have a decreasing trend
through 2007. An increasing trend in the 1-year average concentration is then shown
through 2009, after which another decreasing trend follows. The 1-year average
benzene concentration is at a minimum for 2013 (0.95 |ig/m3), the first time since the
onset of sampling that the 1-year average concentration is less than 1 |ig/m3. The
median concentration exhibits a similar trend, except it did not exhibit the same
increase for 2009 as the 1-year average concentration.
Figure 26-20. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at BTUT
2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
Observations from Figure 26-20 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 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
26-28

-------
concentrations for 2005 and 2006 and then the 5th percentile for 2007. The
percentage of non-detects decreased from 75 percent for 2004 to 0 percent for 2008
and 2009. The percentage of non-detects increased to 7 percent for 2010 and
18 percent for 2011, explaining why the 5th percentile returned to zero. There was a
single non-detect of this pollutant in 2012 and three in 2013.
• 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 2013, the 1-year average concentration
has changed little, ranging from 0.093 |ig/m3 (2013) to 0.118 |ig/m3 (2012). The
median concentration varies a little more, although both the 1-year average and
median concentrations are at a minimum for 2013.
Figure 26-21. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
BTUT

I T









ft-

r-i-
*7"

Ml.'M
T

J
Mb
—
L









pU
T






J
L






















2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
Observations from Figure 26-21 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 |ig/m3 were measured in
2006 (two), 2008 (three), and 2011 (one).
26-29

-------
•	A significant increasing trend is shown in the 1-year average concentrations between
2004 and 2008, with the exception of 2007. The range and magnitude of
concentrations measured decreased substantially for 2007, which is reflected in the
dip in the 1-year average concentration. A slight decreasing trend in the carbon
tetrachloride measurements is shown between 2008 and 2010, after which an
increasing trend is shown through 2012.
•	A significant decrease in the 1-year average concentration, and the other statistical
parameters, is shown for 2013. This year has the lowest maximum concentration
since 2007 and the lowest minimum concentration since 2006. In addition, the
difference between the 5th and 95th percentiles, or the range within which a majority
of concentrations fall, is also at a minimum for 2013.
Figure 26-22. Yearly Statistical Metrics for />-Dichlorobenzene Concentrations Measured at
BTUT
Maximum Concentration
for 2010 is 21.2 ng/m3
Maximum Concentration
for 2008 is 7.59 ng/m3
o
2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile
— Minimum
— Median
— Maximum
O 95th Percentile


Observations from Figure 26-22 for p-dichlorobenzene measurements collected at BTUT
include the following:
• The minimum, 5th percentile, and median concentrations are all zero for 2004 and
2005, indicating that at least half of the measurements were non-detects. In 2004, all
but one measurement was a non-detect. The detection rate of p-dichlorobenzene then
increased each year through 2008 when the fewest non-detects were measured (nine).
The percentage of non-detects has then increased each year since, reaching a secondary
maximum for 2013 (63 percent of measurements were non-detects in 2013, the most
26-30

-------
since 2005), explaining why the median concentration returned to zero for 2011, 2012,
and 2013.
•	The maximum p-dichlorobenzene concentration measured at BTUT was measured in
2010 (21.2 |ig/m3) and is nearly three times greater than the next highest concentration
measured (7.59 |ig/m3, measured in 2008). In all, only 12 concentrations greater than
1 |ig/m3 have been measured at BTUT, all of which were measured prior to 2011.
•	The increases shown for several of the statistical parameters between 2007 and 2010,
particularly the maximum, 95th percentile, and 1-year average concentrations, are a
result of the increased detection rate combined with the higher concentrations
measured.
Figure 26-23. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
BTUT
O 5th Percentile
2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
Maximum	0 95th Percentile
Observations from Figure 26-23 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, 47 in 2012, and 37 in 2013. This explains the increases shown in the 1-year
average concentrations (as well as other statistical parameters) for 2010 through 2013.
26-31

-------
The first year with a median concentration greater than zero is 2012. This indicates
that there were more measured detections than non-detects for the first time since the
onset of sampling.
• The range of concentrations measured in 2013 is considerably larger than the range of
concentrations measured in previous years, as the 1-year average concentration for
2013 is similar to the 95th percentile shown for previous years. All seven of the
1,2-dichloroethane concentrations greater than 0.25 |ig/m3 measured at BTUT were
measured in 2013. Further, measurements collected in 2013 account for more than
one-third of the concentrations greater than 0.1 |ig/m3, with another one-third
measured in 2012, and the final one-third measured between 2009 and 2011.
Figure 26-24. Yearly Statistical Metrics for Dichloromethane Concentrations Measured at
BTUT
1000
900
800
700
^ 600
1
| 500
c
OJ
c
s 400
300
200
100
0
O 5th Percentile	— Minimum	— Median	— Maximum	O 95th Percentile	Average

3 Concentrations in
2010 > 1,000 ng/m3
1 Concentration in
2011 > 1,000 ng/m-
3 Concentrations in
2013 > 1,000 n.g/m3
2008	2009
Year
Observations from Figure 26-24 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, the first concentration greater than 100 |ig/m3 was measured
(203 |ig/m3). In 2009, four concentrations greater than 100 |ig/m3 were measured. In
26-32

-------
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, along with four more greater than
100 |ig/m3. For 2012 only one concentration greater than 100 |ig/m3 was measured.
The maximum dichloromethane concentration was measured at BTUT in 2013
(5,604 |ig/m3) along with two others greater than 1,000 |ig/m3 and eight others 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 32 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. However, the majority of them have been measured during the
second half of any given year (23 of 32). August and September tie for the month
with the greatest number of these higher measurements (6 each), although January
and December tie for second place (5 each).
•	Each of the statistical parameters is at a maximum for 2013. Concentrations measured
in 2013 span four orders of magnitude (0.585 |ig/m3to 5,604 |ig/m3), although 2013 is
not the only year for which this is true.
Figure 26-25. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at BTUT
2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
26-33

-------
Observations from Figure 26-25 for ethylbenzene measurements collected at BTUT
include the following:
•	The maximum concentration of ethylbenzene measured at BTUT was measured in
2013 (5.53 |ig/m3) and is the only concentration greater than 5 |ig/m3 measured at this
site. In total, only seven concentrations greater than 2 |ig/m3 have been measured at
BTUT (of which three were measured in 2013).
•	A steady decreasing trend in the 1-year average concentration is shown from 2004
through 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, 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.
•	The range of ethylbenzene concentrations measured in 2012 is the smallest among the
years of sampling and the 1-year average concentration is at a minimum. Conversely,
the largest range of concentrations was measured in 2013 and the 1-year average
concentration is at its highest since 2005. The range within which the majority of
concentrations fall, as indicated by the 5th and 95th percentiles is also at its largest for
2013, yet the median concentration is at a minimum for this year. Even with the
higher measurements collected, 2013 has the fewest number of measurements greater
than 0.25 |ig/m3.
26-34

-------
Figure 26-26. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at BTUT
Maximum
Concentration for
2004 is 45.4 Hg/m-
2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	- Median	- Maximum	0 95th Percentile	Average
Observations from Figure 26-26 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 12 times
between 2004 and 2007, plus three additional times in 2011.
•	Although the maximum concentration decreased significantly from 2004 to 2005, the
other statistical metrics exhibit increases for 2005. The median increased by nearly
2 |ig/m3 from 2004 to 2005, indicating that concentrations ran higher in 2005 than
2004 (as opposed to being driven by an outlier, as in 2004). 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. In 2008, 95 percent of the concentrations were less than 4 |ig/m3, which is
less than the 1-year average and/or median concentrations for some of the previous
years. After 2008, a steady increasing trend is shown in the 1-year average
formaldehyde concentrations, as well as most other statistical parameters. This trend,
however, levels out for 2012.
26-35

-------
•	Although little change is shown in the 1-year average concentration between 2011
and 2012, the range of concentrations measured is smaller for 2012 and the median
exhibits an increase. The decrease in the concentrations at the upper end of the range
from 2011 to 2012 are balanced out by a higher number of measurements at the mid-
to-upper part of the range. The number of measurements greater than 10 |ig/m3
decreased from nine to one from 2011 to 2012 while the number of measurements
between 5 |ig/m3 and 10 |ig/m3 increased from six to 14 during the same period. 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.
•	For 2013, the 1-year average concentration nearly doubled and the median
concentration increased by 159 percent. Although no formaldehyde concentrations
greater than 15 |ig/m3 were measured in 2013, this year has the highest number of
concentrations greater than 10 |ig/m3 (16) and concentrations greater than 5 |ig/m3
account for more than 75 percent of the measurements in 2013. This is also the only
year for which a formaldehyde concentration less than 2 |ig/m3 was not measured.
Figure 26-27. Yearly Statistical Metrics for Naphthalene Concentrations Measured at BTUT
E
S3 250
2011
Year
O 5th Percentile
O 95th Percentile
26-36

-------
Observations from Figure 26-27 for naphthalene measurements collected at BTUT
include the following:
•	Although PAH sampling began at BTUT in April 2008, complications with the
sampler lead to a 6-month lapse in sampling until mid-October. Thus, Figure 26-27
begins with 2009.
•	The maximum naphthalene concentration (421 ng/m3) was measured in 2009. The
second highest naphthalene concentration (242 ng/m3), measured in 2013, is the only
other naphthalene measurement greater than 200 ng/m3 measured since the onset of
PAH sampling at BTUT.
•	A steady decreasing trend in naphthalene concentrations measured at BTUT is shown
through 2011, with little change shown for 2012.
•	Concentrations increased slightly for 2013, with the 95th percentile for 2013 greater
than the maximum concentrations measured for the two previous years.
•	Concentrations of naphthalene exhibit seasonality. Of the 45 naphthalene
concentrations greater than 100 ng/m3 measured at BTUT since 2009, all but three
were measured during the first or fourth quarters of any given year, with the majority
measured in January (15), November (10), or December (14).
Figure 26-28. Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at BTUT
























I


4-1 T




LqJ

J w		
s-J L-s

2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimurr
Maximum	O 95th Percentile
26-37

-------
Observations from Figure 26-28 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 measured in 2008. No other nickel
concentrations greater than 10 ng/m3 have been measured at BTUT.
•	All 24 non-detects of nickel were measured in 2009 and, with one exception, were
measured on consecutive sample days between June and October.
•	The range of nickel concentrations measured each year is highly variable.
Concentrations measured over a given year have spanned a little as 2.5 ng/m3 (2010)
or up to nearly 30 ng/m3 (2005). This variability is reflected in the undulating pattern
shown in the central tendency statistics, particularly in the years between 2004 and
2011. During this time period, the 1-year average concentrations ranged from
0.75 ng/m3 (2009) to 4.05 ng/m3 (2005). The concentrations measured between 2011
and 2013 exhibit less variability.
Figure 26-29. Yearly Statistical Metrics for Propionaldehyde Concentrations Measured at
BTUT
2004	2005	2006	2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	- Median	— Maximum	O 95th Percentile ••••C"" Average
26-38

-------
Observations from Figure 26-29 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. All but one of the nine concentrations greater than 1.5 |ig/m3 were measured
prior to 2007, with the exception measured in 2013.
•	Even though the maximum concentration decreased from 2004 to 2005, the other
statistical metrics exhibit increases (similar to the formaldehyde concentrations). The
median concentration more than doubled from 2004 to 2005, indicating that
concentrations ran higher in 2005 than 2004 (as opposed to being driven by a few
higher concentrations, as in 2004). The number of concentrations greater than
0.5 |ig/m3 increased nearly four-fold, from nine in 2004 to 33 in 2005, accounting for
more than half of the measurements collected in 2005.
•	After 2005, the propionaldehyde concentrations began to decrease, reaching a
minimum for 2009, when all of the measurements are less than 1 |ig/m3. The
propionaldehyde concentrations measured increased significantly from 2009 to 2010,
with an undulating pattern in the 1-year average concentrations developing afterward.
•	Similar to acetaldehyde and formaldehyde, each of the statistical parameters exhibits
an increase for 2013, with a significant increase shown for the 1-year average
concentration. The most recent year of sampling has the largest number of
propionaldehyde concentrations greater than 1 |ig/m3 (24), nearly double the amount
for the next closest year (13 each in 2005 and 2006). Further, each year of sampling
has a number of concentrations less than the minimum concentration measured in
2013 (0.27 |ig/m3), from as few as six (2006) to as many as 24 (2004).
26.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at the BTUT monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4 for
definitions and explanations regarding the various toxicity factors, time frames, and calculations
associated with these risk-based screenings.
26.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for BTUT and where annual average concentrations could
be calculated, risk was examined by calculating cancer risk and noncancer hazard
approximations. These approximations can be used as risk estimates for cancer and noncancer
effects attributable to the pollutants of interest. Although the use of these approximations is
26-39

-------
limited, they may help identify where policy-makers want to shift their air monitoring priorities.
Refer to Section 3.4.3.3 for an explanation of how cancer risk and noncancer hazard
approximations are calculated and what limitations are associated with them. Annual averages,
cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard approximations are
presented in Table 26-6, where applicable. Cancer risk approximations are presented as
probabilities while the noncancer hazard approximations are ratios and thus, unitless values.
Table 26-6. Risk Approximations for the Utah Monitoring Site
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Bountiful, Utah - BTUT
Acetaldehyde
0.0000022
0.009
55/55
4.18
±0.36
9.21
0.46
Benzene
0.0000078
0.03
53/53
0.94
±0.16
7.31
0.03
1.3 -Butadiene
0.00003
0.002
50/53
0.09
±0.02
2.76
0.05
Carbon Tetrachloride
0.000006
0.1
53/53
0.56
±0.03
3.36
0.01
/?-Dichlorobcnzcnc
0.000011
0.8
19/53
0.05
±0.03
0.51
<0.01
1,2-Dichloroethane
0.000026
2.4
38/53
0.11
±0.03
2.84
<0.01
Dichloromethane
0.000000016
0.6
53/53
225.03
±219.72
3.60
0.38
Ethylbenzene
0.0000025
1
53/53
0.49
±0.24
1.23
<0.01
Formaldehyde
0.000013
0.0098
55/55
8.05
±0.87
104.64
0.82
Propionaldehyde

0.008
55/55
0.89
±0.08

0.11
Arsenic (PMi0)a
0.0043
0.000015
59/59
0.99
±0.40
4.26
0.07
Naphthalene3
0.000034
0.003
56/56
55.48
± 11.39
1.89
0.02
Nickel (PMi,;,)a
0.00048
0.00009
59/59
1.44
±0.24
0.69
0.02
— = A Cancer URE or Noncancer RfC is not available.
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of
viewing.
Observations for BTUT from Table 26-6 include the following:
• The pollutants with the highest annual average concentrations are dichloromethane,
formaldehyde, and acetaldehyde, as discussed in Section 26.4.1.
26-40

-------
•	The pollutants with the highest cancer risk approximations are formaldehyde,
acetaldehyde, and benzene. The cancer risk approximation for formaldehyde for
BTUT (104.64 in-a-million) is the only cancer risk approximation greater than
100 in-a-million calculated across the program. The remaining cancer risk
approximations calculated for BTUT are all less than 10 in-a-million.
•	There were no pollutants of interest with noncancer hazard approximations greater
than 1.0, indicating that no adverse noncancer health effects are expected from these
individual pollutants. The highest noncancer hazard approximation was calculated for
formaldehyde (0.82), which is the highest noncancer hazard approximation calculated
among the site-specific pollutants of interest with noncancer toxicity factors. The
remaining noncancer hazard approximations calculated for BTUT are all less than
0.50.
•	Dichloromethane's relatively high annual average concentration does not translate
into high risk approximations. This is an indication of the toxicity potential of
dichloromethane concentrations in ambient air.
For each of the site-specific pollutants of interest that have cancer risk approximations
greater than 75 in-a-million and/or a noncancer hazard approximation greater than 1.0, a
pollution rose was created 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. Additional information about this analysis is presented
in Section 3.4.3.3. Figure 26-30 is BTUT's pollution rose for formaldehyde.
26-41

-------
Figure 26-30. Pollution Rose for Formaldehyde Concentrations Measured at BTUT
360/0

45
315
	•
225
135
180
o <5 j^g/m3 05-10 J^g/m3 O lu ug m3
Observations from Figure 26-30 include the following:
•	The pollution rose shows that most of the formaldehyde concentrations are shown in
relation to samples days with an average wind direction from the southeast to south or
northwest to north. This matches the wind observations shown on the sample day
wind rose presented in Figure 26-3.
•	The facility map in Figure 26-2 shows that most of the point sources are located to the
south and southwest of BTUT, along the 1-15 corridor and towards Salt Lake City.
•	Formaldehyde concentrations of varying magnitude are shown in relation to the
predominant wind directions, although compared to other NMP sites, even the lowest
concentrations measured at BTUT are higher than half the measurements collected at
other NMP sites.
•	If the formaldehyde concentrations are grouped by average compass direction, the
direction with the most concentrations is northwest, followed by southeast. If the
formaldehyde concentrations are averaged by compass direction, the highest average
concentrations are calculated for west and northeast. However, the westerly direction
only includes two concentrations while the northeasterly direction includes only one.
26-42

-------
Other wind directions, such as northwest, incorporate many concentrations of varying
magnitude.
• The wind data for many of the sample days reflect a lake breeze/valley breeze system,
one in which the wind direction in the morning is different from the
afternoon/evening, switching directions with regularity due to daytime heating and
geographic features such as the Great Salt Lake and the mountains on either side of
the Salt Lake Valley (NHMU, 2015).
26.5.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 26-7 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
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 3.4.3.4. Lastly,
Table 26-7 provides the pollutants with the highest cancer risk approximations (in-a-million) for
BTUT, 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 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 26.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
26-43

-------
Table 26-7. 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)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Bountiful, Utah (Davis County) - BTUT
Benzene
120.37
Benzene
9.39E-04
Formaldehyde
104.64
Formaldehyde
68.30
Formaldehyde
8.88E-04
Acetaldehyde
9.21
Ethylbenzene
67.10
Hexavalent Chromium
6.26E-04
Benzene
7.31
Dichloromethane
46.51
1,3-Butadiene
4.62E-04
Arsenic
4.26
Acetaldehyde
41.70
Naphthalene
2.84E-04
Dichloromethane
3.60
1.3 -Butadiene
15.40
POM, Group 2b
1.79E-04
Carbon Tetrachloride
3.36
Naphthalene
8.35
Ethylbenzene
1.68E-04
1,2-Dichloroethane
2.84
T etrachloroethy lene
6.26
POM, Group 2d
1.23E-04
1,3-Butadiene
2.76
POM, Group 2b
2.03
POM, Group 5a
9.95E-05
Naphthalene
1.89
POM, Group 2d
1.39
Acetaldehyde
9.17E-05
Ethylbenzene
1.23

-------
Table 26-8. 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)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Bountiful, Utah (Davis County) - BTUT
Toluene
539.04
Acrolein
192,602.57
Formaldehyde
0.82
Hexane
370.73
1.3 -Butadiene
7,700.14
Acetaldehyde
0.46
Xylenes
286.60
Formaldehyde
6,969.87
Dichloromethane
0.38
Methanol
205.85
Acetaldehyde
4,633.66
Propionaldehyde
0.11
Ethylene glycol
121.88
Benzene
4,012.18
Arsenic
0.07
Benzene
120.37
Xylenes
2,865.98
1,3-Butadiene
0.05
Formaldehyde
68.30
Naphthalene
2,782.33
Benzene
0.03
Ethylbenzene
67.10
Lead, PM
982.29
Naphthalene
0.02
Methyl isobutyl ketone
51.39
Arsenic, PM
703.33
Nickel
0.02
Dichloromethane
46.51
Hexane
529.62
Carbon Tetrachloride
0.01

-------
Observations from Table 26-7 include the following:
•	Benzene, formaldehyde, ethylbenzene, and dichloromethane are the highest emitted
pollutants with cancer UREs in Davis County.
•	The pollutants with the highest toxicity-weighted emissions (of the pollutants with
cancer UREs) are benzene, formaldehyde, hexavalent chromium, and 1,3-butadiene.
•	Eight of the highest emitted pollutants also have the highest toxicity-weighted
emissions in Davis County.
•	Formaldehyde, which has the highest cancer risk approximation for BTUT, ranks
second for both emissions-based lists behind benzene. Acetaldehyde, 1,3-butadiene,
naphthalene, and ethylbenzene also appear on all three lists in Table 26-7.
Dichloromethane, which has the highest annual average concentration and the fifth
highest cancer risk approximation for BTUT, ranks fourth for emissions in Davis
County but is not among those with the highest toxicity-weighted emissions (it ranks
22nd). Arsenic, carbon tetrachloride, and 1,2-dichloroethane, pollutants that have
some of the highest cancer risk approximations for BTUT, appear on neither
emissions-based list.
•	POM, Group 2b is the ninth highest emitted "pollutant" in Davis County and ranks
sixth for toxicity-weighted emissions. POM, Group 2b includes several PAHs
sampled for at BTUT including acenaphthylene, fluoranthene, and perylene. None of
the PAHs included in POM, Group 2b were identified as pollutants of interest for
BTUT.
•	POM, Group 2d is the 10th highest emitted "pollutant" in Davis County and ranks
eighth 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 2d were identified as pollutants of interest for BTUT.
Observations from Table 26-8 include the following:
•	Toluene, hexane, and xylenes are the highest emitted pollutants with noncancer RfCs
in Davis County.
•	The pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) are acrolein, 1,3-butadiene, and formaldehyde. Although acrolein
was sampled for at BTUT, this pollutant was excluded from the pollutants of interest
designation, and thus subsequent risk-based screening evaluations, due to questions
about the consistency and reliability of the measurements, as discussed in Section 3.2.
•	Four of the highest emitted pollutants also have the highest toxicity-weighted
emissions in Davis County.
•	Formaldehyde, acetaldehyde, and dichloromethane have the highest noncancer hazard
approximations for BTUT (although all are less than 1.0). Formaldehyde and benzene
26-46

-------
are the only listed pollutants of interest to appear on both emissions-based lists.
Acetaldehyde, arsenic, 1,3-butadiene, and naphthalene rank among the pollutants
with the highest toxicity-weighted emissions but do not appear among those with the
highest total emissions. Dichloromethane ranks 10th for its quantity emitted in Davis
County but does not appear among those highest toxicity-weighted emissions.
Propionaldehyde, nickel, and carbon tetrachloride do not appear on either emissions-
based list in Table 26-8.
26.6 Summary of the 2013 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 third year in a row, BTUT has the highest annual average formaldehyde
concentration among NMP sites sampling this pollutant. BTUT also has the highest
annual average concentration of acetaldehyde and second highest annual average
concentration of arsenic among other NMP sites.
~~~ Concentrations of benzene have an overall decreasing trend at BTUT; the 1-year
average concentration for 2013 is the lowest 1-year average concentration of
benzene calculated since the onset of sampling at BTUT. Concentrations of three
carbonyl compounds increased significantly for 2013. The detection rate of
1,2-dichloroethane has been increasing steadily at BTUT over the last few years of
sampling, although this leveled out for 2013.
~~~ Formaldehyde has the highest cancer risk approximation among the pollutants of
interest for BTUT and across the program. This is the only site-specific pollutant of
interest with a cancer risk approximation greater than 100 in-a-million. None of the
pollutants of interest have noncancer hazard approximations greater than an HQ of
1.0.
26-47

-------
27.0	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.
27.1	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 CBSA. The second UATMP site
(RUVT) is located farther south in Rutland, Vermont. Figures 27-1 and 27-2 are the composite
satellite images retrieved from ArcGIS Explorer showing the Burlington monitoring sites and
their immediate surroundings. Figure 27-3 identifies nearby point source emissions locations by
source category, as reported in the 2011 NEI for point sources, version 2. Note that only sources
within 10 miles of the sites are included in the facility counts provided in Figure 27-3. A 10-mile
boundary was chosen to give the reader an indication of which emissions sources and emissions
source categories could potentially have a direct effect on the air quality at the monitoring sites.
Further, this boundary provides both the proximity of emissions sources to the monitoring sites
as well as the quantity of such sources within a given distance of the sites. Sources outside the
10-mile boundaries are still visible on the map for reference, but have been grayed out in order to
emphasize emissions sources within the boundary. Figures 27-4 and 27-5 are the composite
satellite image and emissions sources map for the Rutland site. Table 27-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates for each
site.
27-1

-------
Figure 27-1. Burlington, Vermont (BIJRVT) Monitoring Site

Maih;St
.Adams-Si
,iprucc^
to
-j
K>

-------
Figure 27-2. Under hill, Vermont (UNYT) Monitoring Site

-------
Figure 27-3. NEI Point Sources Located Within 10 Miles of BURVT and UNVT
Franklin
County
lairoilto ' \
County \ \
Chittenden
County
Laws
Washington
i County
NEW YORK
jwrw n
frzwrw tt'WW
Legend
f3"5trw 13WVV
^	n Mtrw rs	n *wv*
re-ioTp* 1V9TN rrtrn-w nrvnrw rrvnrw
No* Out to faenr# d«n«i
-------
Figure 27-4. Rutland, Vermont (RUVT) Monitoring Site

„ >•
tlrVJi
Kill.'! !
,.ir. Mi M
U J
rSUlf ^

£njttT: ***
SoUfC* tlUTGS
c N A*< 4 f, . A *'.
PA£'v>
ruY
to
"-J

-------
Figure 27-5. NEI Point Sources Located Within 10 Miles of RUVT
VUnd&OT
County
NEW
YORK
Source Category Group (No. of Facilities)
•i* Aerospace'Aircraft Manufacturing Facility (3)
f Airport;Airline/Airport Support Operations (6)
Brick. Structural Clay, or Clay Ceramics Plant (1)
* Electricity Generation via Combustion (1)
© Metals ProcessingyFabncation Facility (1)
ffl Pulp and Paper Plant (t)
W Woodwork. Furniture. Millwork & Wood Preserving Facility (2)
Legend
~ RUVT UATMP site
Not* Out ic incut/ itefifttfy «nd> oollocaton ih« total tecilflwt
displayed may not rvprramt all facifews within (he aroa of «itor»st
10 mile radius		] County boundary
27-6

-------
Table 27-1. Geographical Information for the Vermont Monitoring Sites
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Additional Ambient Monitoring Information1
BURVT
50-007-0014
Burlington
Chittenden
Burlington-South
Burlington, VT
44.4762,
-73.2106
Commercial
Urban/City
Center
CO, NO, NO2, NOx, Meteorological parameters,
PM10, PM2 5, PM Coarse, PM25 Speciation.
UNVT
50-007-0007
Underhill
Chittenden
Burlington-South
Burlington, VT
44.52839,
-72.86884
Forest
Rural
Haze, Sulfate TSP, CO, S02, NO, NOy, 03,
Meteorological parameters, PM10, PM Coarse, PM2 5,
PM2 5 Speciation, IMPROVE Speciation.
RUVT
50-021-0002
Rutland
Rutland
Rutland, VT
43.608056,
-72.982778
Commercial
Urban/City
Center
CO, SO2, NO, NO2, NOx, Meteorological parameters,
PM10, PM2 5, PM Coarse, PM2 5 Speciation.
1 Data for additional pollutants are reported to AQS for these sites (EPA, 2014b); however, these data are not generated by ERG and are therefore not included in this report.
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 about one-half 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 27-1. 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 Center in Underhill,
Vermont, which is east of the Burlington area. This research station is part of the University of
Vermont, with research focused on the sugar maple tree and sap collection methods (UVM,
2015). Figure 27-2 shows that the area surrounding the site is rural in nature and heavily
forested. Mount Mansfield, the highest peak in Vermont, lies to the east in Underhill State Park,
less than 3 miles away. This site is intended to serve as a background site for the region for
trends assessment, standards compliance, and long-range transport assessment.
UNVT and BURVT are located approximately 16 miles apart, as shown in Figure 27-3.
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, two heliports at the medical school, two facilities
generating electricity via combustion, and a metals processing/fabrication facility. 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 27-4. 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 27-5 shows that most of the emissions sources within
27-8

-------
10 miles of RUVT are located along Route 7 (Main Street), just south of the monitoring site or
along West Street to the west of the site. The source categories with the greatest number of
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 27-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Vermont monitoring sites. Table 27-2 includes both county-level
population and vehicle registration information. Table 27-2 also contains traffic volume
information for each site as well as the location for which the traffic volume was obtained.
Additionally, Table 27-2 presents the county-level daily VMT for Chittenden and Rutland
Counties.
Table 27-2. Population, Motor Vehicle, and Traffic Information for the Vermont
Monitoring Sites
Site
County
Estimated
County
Population1
County-level
Vehicle
Registration2
Annual
Average
Daily Traffic3
Intersection
Used for Traffic Data
County-
level Daily
VMT4 "
BURVT
Chittenden
159,515
172,203
14,200
Main St, South of Willard St
4,051,781
UNVT
1,100
Pleasant Valley Rd, North of
Harvey Rd
RUVT
Rutland
60,622
79,795
10,400
Bus US-4 (West St) between Pine St
and Evelyn St
1,736,164
bounty-level population estimates reflect 2013 data (Census Bureau, 2014)
2County-level vehicle registrations reflect 2013 data (VT DMV, 2014)
3AADT reflects 2009 data for BURVT and 2011 data for UNVT (CCRPC, 2014) and 2013 data for RUVT (VTrans, 2014a)
4County-level VMT reflects 2013 data (Vtrans, 2014b)
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 27-2 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 south of Willard Street; for UNVT, the data is for Pleasant
27-9

-------
Valley Road, north of Harvey Road; and for RUVT, the data is for US-4 Business
between Pine Street and Evelyn Street.
• The county-level daily VMT for Chittenden County is more than twice the VMT for
Rutland County, with both VMTs in the bottom third compared to other counties with
NMP sites.
27.2 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.
27.2.1	Climate Summary
The city of Burlington is located just to the east of Lake Champlain in northwest
Vermont. Lake Champlain has a moderating effect on the city's temperatures, keeping the city
slightly warmer in winter than it would be given its New England location. The town of
Underhill is located to the east of Burlington but still within the Burlington metro area. The city
of Rutland is located 60 miles south of the Burlington area. Rutland is within the same climatic
division of Vermont as Burlington, but misses the moderating influences of Lake Champlain.
The state of Vermont is affected by many 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 flow parallel to 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, 2015; NO A A, 2015c).
27.2.2	Meteorological Summary
Hourly meteorological data for 2013 were retrieved from NCDC for the weather stations
closest to the Vermont monitoring sites (NCDC, 2013), as described in Section 3.4.2. 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,
27-10

-------
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 27-3. These data were
used to determine how meteorological conditions on sample days vary from conditions
experienced throughout the year.
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 2013. Also included in Table 27-3 is the 95 percent
confidence interval for each parameter. 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 occurred on a l-in-6 day schedule, while sampling at BURVT and
RUVT occurred on a l-in-12 day schedule.
As shown in Table 27-3, meteorological conditions on sample days were representative
of weather conditions experienced throughout the year near these sites. The averages were most
similar for UNVT, where 1 degree or less (or millibar, knot, or percentage) separates the sample
day averages from the full-year averages. The sample day vs. full-year averages for RUVT
exhibit the most variability, although the largest difference was calculated for BURVT's relative
humidity.
Compared to other NMP sites, the Vermont sites experience some of the coldest
temperatures. UNVT and RUVT rank fifth and sixth, respectively for the lowest average
maximum temperature and rank fourth and sixth, respectively, for the lowest average
temperatures. UNVT also has the second lowest average wind speed (behind CELA).
27-11

-------
Table 27-3. Average Meteorological Conditions near the Vermont Monitoring Sites
to
-J
to
Closest Weather
Station
Distance
and

Average
Maximum
Average
Average
Dew Point
Average
Wet Bulb
Average
Relative
Average
Sea Level
Average
Scalar
Wind
(WBAN and
Coordinates)
Direction
from Site
Average
Type1
Temperature
(°F)
Temperature
(°F)
Temperature
(°F)
Temperature
(°F)
Humidity
(%)
Pressure
(mb)
Speed
(kt)
Burlington, Vermont - BURVT
Burlington Intl.
3.1
miles
Sample
Days
56.1
48.4
36.2
42.9
65.3
1015.9
7.0
Airport
(31)
±8.0
±7.3
±7.4
±6.8
±4.2
±2.7
± 1.2
14742
100°
(E)








(44.47, -73.15)

55.6
47.3
36.0
42.5
67.4
1016.5
6.2

2013
±2.2
±2.1
±2.1
± 1.9
± 1.3
±0.8
±0.3



Rutland, Vermont
-RUVT




Rutland State Airport
94737
(43.53, -72.95)
5.4
miles
Sample
Days
(31)
55.6
±7.1
47.2
±6.6
36.3
±6.9
42.2
±6.3
68.3
±3.9
NA
6.5
±0.8
162°
(SSE)

53.7
45.3
35.0
41.0
70.1

6.0

2013
±2.1
± 1.9
±2.1
± 1.8
± 1.3
NA
±0.3
Underhill, Vermont - UNVT
Morrisville-Stowe
12.6
miles
Sample
Days
54.5
44.8
34.9
40.4
71.5
1017.0
3.2
State Airport
(64)
±5.6
±5.1
±5.1
±4.8
±2.8
± 1.7
±0.6
54771









(44.53, -72.61)
OO
(E)

53.5
44.0
35.4
40.0
72.2
1017.1
3.0

2013
±2.2
±2.0
±2.1
± 1.9
± 1.1
±0.8
±0.2
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 Rutland State Airport.

-------
27.2.3 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 3.4.2. A wind rose shows the frequency of wind directions using "petals"
positioned around a 16-point compass, and uses different colors to represent wind speeds.
Figure 27-6 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 27-6 also presents three different
wind roses for the BURVT monitoring site. First, a historical wind rose representing 2003 to
2012	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
2013	is presented. Next, a wind rose representing wind data for days on which samples were
collected in 2013 is presented. These can be used to identify the predominant wind speed and
direction in 2013 and to determine if wind observations on sample days were representative of
conditions experienced over the entire year and historically. Figures 27-7 and 27-8 present the
three wind roses and distance maps for the RUVT and UNVT monitoring sites, respectively.
27-13

-------
Figure 27-6. Wind Roses for the Burlington International Airport Weather Station near
BURVT
Location of BURVT and Weather Station
2003-2012 Historical Wind Rose
NORTH""--,
30%
24%
18%
12%
WIND SPEED
(Knots)
~ =22
Esil 17-21
11 17
I I 7- 11
I 4-7
H 2-4
Calms: 2374%
2013 Wind Rose	Sample Day Wind Rose
NORTH'
NORTH'
WEST
est:
WIND SPEED
(Knots)
~ -22
[ I 17-21
rM 11-17
WIND SPEED
(Knots)
~ >=22
Ml 17-21
|| 11-17
[SOUTH
27-14

-------
Figure 27-7. Wind Roses for the Rutland State Airport Weather Station near RUVT
Location of RUVT and Weather Station	2003-2012 Historical Wind Rose
WIND SPEED
(Knots)
~ >=22
II 17-21
~
: EAS
west:
2013 Wind Rose
NORTH""--
WEST
WIND SPEED
(Kn ots >
17-21
SOUTH
Calms: 16.12%
Sample Day Wind Rose
NORTH --

WIND SPEED
(Knots)
17-21
SOUTH
Caln-s: 13*9%
27-15

-------
Figure 27-8. Wind Roses for the Morrisville-Stowe State Airport Weather Station near
UNVT
Location of UNVT and Weather Station
2003-2012 Historical Wind Rose

12 6 mibr*
yv Waathrn
~	
UNVT




4-
; NORTH"'-.,
WIND SPEED
(Knots)
~ =22
Esil 17-21
11 17
I I 7- 11
I 4-7
H 2-4
Calms: 46.30%
2013 Wind Rose
Sample Day Wind Rose
; NORTH""--,


ft 4%,

2% : ; :
WEST! \ 1

: '• : EAST
WIND SPEED
(Knots)
~	>=22
HI 17-21
11-17
I 1 7- 11
~	4-7
2- 4
Calms: 48 23%
est;
NORTH""--,
10%
8%.
WIND SPEED
(Knots)
~ >=22
11 -17
7- 11
4- 7
2- 4
Calms: 46.04%
~
27-16

-------
Observations from Figure 27-6 for BURVT include the following:
•	The Burlington International Airport weather station is located approximately 3 miles
east of BURVT, which is four times farther away from Lake Champlain than the
monitoring site.
•	The historical wind rose shows that southerly winds are prevalent near BURVT,
accounting for nearly 22 percent of the hourly measurements. Calm winds (those less
than or equal to 2 knots) account for another 21 percent of measurements. Winds
from the northwest quadrant were also commonly observed while winds from the
eastern quadrants were rarely observed.
•	The wind patterns shown on the 2013 wind rose are similar to the historical wind
patterns, indicating that wind conditions observed during 2013 were similar to those
observed over the previous 10 years.
•	The sample day wind rose shows that southerly winds prevailed on sample days, but
account for a higher percentage of observations (nearly 30 percent). The increase in
southerly winds coincides with a decrease in calm winds (down to 13 percent on
sample days). Winds from the north-northwest and north were observed equally on
sample days, which is another difference between the sample day and full-year wind
rose.
Observations from Figure 27-7 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 northwest quadrant, and to a less extent, the
southwest quadrant were also commonly 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 2013 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 2013.
•	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). Westerly winds
were also observed more often on sample days while northwesterly to northerly winds
were observed less frequently.
27-17

-------
Observations from Figure 27-8 for UNVT include the following:
•	The Morrisville-Stowe Airport weather station is located less than 13 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 46 percent of the hourly measurements. Winds from the
northwest to north account for approximately 21 percent of the wind observations
greater than 2 knots. Winds from the south to south-southwest account for another
roughlyl4 percent of observations.
•	The wind patterns shown on the 2013 wind rose are similar to the historical wind
patterns.
•	The sample day wind rose shows that wind conditions on sample days were similar to
those experienced throughout 2013, although number of observations from the north-
northwest is slightly less while the number of observations from the northwest is
slightly higher. A higher percentage of stronger winds from these directions was also
observed on sample days.
27.3 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 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. VOCs were sampled for year-round at BURVT and RUVT, while
hexavalent chromium, PAHs, and metals (PMio) were sampled for in addition to VOCs at
UNVT. Hexavalent chromium sampling at UNVT, however, was discontinued at the end of
June.
27-18

-------
Table 27-4. Risk-Based Screening Results for the Vermont Monitoring Sites
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Burlington, Vermont - BURVT
Benzene
0.13
31
31
100.00
23.13
23.13
Carbon Tetrachloride
0.17
31
31
100.00
23.13
46.27
1.3 -Butadiene
0.03
30
30
100.00
22.39
68.66
1,2-Dichloroethane
0.038
30
30
100.00
22.39
91.04
p-Dichlorobenzene
0.091
5
27
18.52
3.73
94.78
Hexachloro-1,3 -butadiene
0.045
4
4
100.00
2.99
97.76
Ethylbenzene
0.4
2
31
6.45
1.49
99.25
Trichloroethylene
0.2
1
4
25.00
0.75
100.00
Total
134
188
71.28

Rutland, Vermont - RUVT
Benzene
0.13
31
31
100.00
24.60
24.60
Carbon Tetrachloride
0.17
31
31
100.00
24.60
49.21
1.3 -Butadiene
0.03
29
29
100.00
23.02
72.22
1,2-Dichloroethane
0.038
27
27
100.00
21.43
93.65
Ethylbenzene
0.4
5
31
16.13
3.97
97.62
Hexachloro-1,3 -butadiene
0.045
3
4
75.00
2.38
100.00
Total
126
153
82.35

Underhill, Vermont - UNVT
Benzene
0.13
59
60
98.33
26.11
26.11
Carbon Tetrachloride
0.17
58
60
96.67
25.66
51.77
1,2-Dichloroethane
0.038
53
53
100.00
23.45
75.22
Arsenic (PMi0)
0.00023
38
56
67.86
16.81
92.04
1,3-Butadiene
0.03
7
11
63.64
3.10
95.13
Hexachloro-1,3 -butadiene
0.045
5
5
100.00
2.21
97.35
Naphthalene
0.029
4
59
6.78
1.77
99.12
Nickel (PMio)
0.0021
1
60
1.67
0.44
99.56
Trichloroethylene
0.2
1
2
50.00
0.44
100.00
Total
226
366
61.75

Observations from Table 27-4 include the following:
•	Eight pollutants failed at least one screen for BURVT; 71 percent of concentrations
for these eight pollutants were greater than their associated risk screening value (or
failed screens).
•	Six pollutants contributed to 95 percent of failed screens for BURVT and therefore
were identified as pollutants of interest for this site.
27-19

-------
•	Six pollutants failed at least one screen for RUVT; 82 percent of concentrations for
these six pollutants were greater than their associated risk screening value (or failed
screens).
•	Five pollutants contributed to 95 percent of failed screens for RUVT and therefore
were identified as pollutants of interest for this site.
•	Nine pollutants failed at least one screen for UNVT; 62 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 UNVT and therefore
were identified as pollutants of interest for this site. These five include four VOCs
and one PMio metal.
•	The Vermont sites have four pollutants of interest in common: benzene, carbon
tetrachloride, 1,3-butadiene, and 1,2-dichloroethane.
27.4 Concentrations
This section presents various concentration averages used to characterize air toxics
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. Site-specific statistical summaries for all pollutants
sampled for at BURVT, RUVT, and UNVT are provided in Appendices J, M, N, and O.
27.4.1 2013 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
27-20

-------
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 27-5, 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 27-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Vermont Monitoring Sites

# of






Measured
1st
2nd
3rd
4th


Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs. # of
Samples
Average
Oig/m3)
Average
Oig/m3)
Average
(Hg/m3)
Average
(Hg/m3)
Average
(Hg/m3)
Burlington, Vermont - BURVT


0.85
0.54
0.62
0.58
0.65
Benzene
31/31
±0.09
±0.09
±0.16
±0.09
±0.06


0.08
0.06
0.08
0.07
0.07
1.3 -Butadiene
30/31
±0.01
±0.03
±0.03
±0.03
±0.01


0.65
0.62
0.67
0.54
0.62
Carbon Tetrachloride
31/31
±0.05
±0.07
±0.06
±0.08
±0.03


0.07
0.06
0.07
0.04
0.06
/?-Dichlorobcnzcnc
27/31
±0.02
±0.02
±0.02
±0.02
±0.01


0.09
0.10
0.07
0.07
0.08
1,2-Dichloroethane
30/31
±0.03
±0.02
±0.01
±0.01
±0.01


0.02
0.01

0.01
0.01
Hexachloro-1,3 -butadiene
4/31
±0.03
±0.03
0
±0.03
±0.01
Rutland, Vermont - RUVT


1.26
0.64
0.49
0.81
0.81
Benzene
31/31
±0.49
±0.19
±0.10
±0.34
±0.18


0.17
0.08
0.06
0.13
0.11
1.3 -Butadiene
29/31
±0.09
±0.04
±0.02
±0.07
±0.03


0.65
0.66
0.66
0.59
0.63
Carbon Tetrachloride
31/31
±0.08
±0.06
±0.09
±0.03
±0.03


0.10
0.09
0.08
0.04
0.08
1,2-Dichloroethane
27/31
±0.02
±0.02
±0.02
±0.03
±0.01


0.29
0.31
0.31
0.20
0.27
Ethylbenzene
31/31
±0.10
±0.09
±0.13
±0.11
±0.05
a Average concentrations provided for the pollutant below the blue line are presented in ng/m3 for ease of viewing.
27-21

-------
Table 27-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Vermont Monitoring Sites (Continued)

# of






Measured
1st
2nd
3rd
4th


Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs. # of
Samples
Average
(Ug/m3)
Average
(Ug/m3)
Average
(Ug/m3)
Average
(Ug/m3)
Average
(Ug/m3)
Underhill, Vermont - UNVT


0.74
0.23
0.22
0.29
0.37
Benzene
60/60
±0.45
±0.06
±0.02
±0.07
±0.12



<0.01
0.01
0.01
0.01
1.3 -Butadiene
11/60
0
±0.01
±0.01
±0.01
±<0.01


0.59
0.69
0.67
0.56
0.63
Carbon Tetrachloride
60/60
±0.06
±0.04
±0.03
±0.09
±0.03


0.08
0.10
0.05
0.06
0.07
1,2-Dichloroethane
53/60
±0.01
±0.01
±0.01
±0.02
±0.01


0.18
0.37
0.29
0.29
0.28
Arsenic (PMi0)a
56/60
±0.07
±0.08
±0.12
±0.08
±0.05
a Average concentrations provided for the pollutant below the blue line are presented in ng/m3 for ease of viewing.
Observations for BURVT from Table 27-5 include the following:
•	BURVT sampled VOCs on a l-in-12 day schedule, yielding half as many samples as
UNVT.
•	Benzene is the pollutant with the highest annual average concentration for BURVT,
followed by carbon tetrachloride, although their annual averages are similar. All of
the remaining annual average concentrations for the pollutants of interest for BURVT
are less than 0.1 |ig/m3.
•	Concentrations of benzene measured at BURVT range from 0.35 |ig/m3 to
1.02 |ig/m3, which is the only benzene concentration greater than 1 |ig/m3. Six of the
eight concentrations greater than 0.8 |ig/m3 were measured in either January or
February, which explains why the first quarter average concentration is higher than
the other quarterly averages. The difference, however, is not statistically significant.
Similar observations were made in the 2011 and 2012 NMP reports.
•	Concentrations of carbon tetrachloride and p-dichlorobenzene appear lowest during
the fourth quarter of 2013. The two lowest carbon tetrachloride measurements were
collected at BURVT in December. For /;-dichlorobenzene, two of the four non-
detects and the minimum measured detection were measured in December. However,
the differences among the quarterly average concentrations are not statistically
significant.
•	Concentrations of 1,2-dichloroethane measured at BURVT appear higher during the
first half of the year. Concentrations of 1,2-dichloroethane range from 0.04 |ig/m3 to
0.16 |ig/m3, plus a single non-detect. None of the 10 highest concentrations were
measured after May.
27-22

-------
•	For each of the quarterly average concentrations of hexachloro-1,3 -butadiene, the
confidence interval is larger than the average itself, indicating a relatively high level
of variability in the measurements. Hexachloro-l,3-butadiene was detected only four
times at BURVT and was not detected at all during the third quarter, resulting in a
third quarter average concentration of zero.
Observations for RUVT from Table 27-5 include the following:
•	RUVT also sampled VOCs on a l-in-12 day schedule.
•	Benzene, carbon tetrachloride, and ethylbenzene have the highest annual average
concentrations for RUVT, although none of the annual average concentrations are
greater than 1 |ig/m3.
•	The quarterly average concentrations of benzene exhibit considerably variability, with
the first quarter average concentration the highest and the third quarter average the
lowest. Concentrations of benzene measured at RUVT range from 0.30 |ig/m3 to
2.09 |ig/m3. Of the seven benzene concentrations greater than 1 |ig/m3 measured at
RUVT, four were measured during the first quarter (including the maximum,
although a similar concentration was also measured during the fourth quarter), one
was measured during the second quarter, and two were measured during the fourth
quarter. No benzene concentrations greater than 0.65 |ig/m3 were measured during the
third quarter of 2013.
•	Concentrations of 1,3-butadiene measured at RUVT span an order of magnitude,
ranging from 0.033 |ig/m3 to 0.399 |ig/m3, including two non-detects. Concentrations
of 1,3-butadiene appear higher during the first and fourth quarters of 2013 and exhibit
more variability. Nine of the 11 concentrations of 1,3-butadiene greater than
0.1 |ig/m3 were measured at RUVT during the first or fourth quarters.
•	Concentrations of carbon tetrachloride also appear lowest during the fourth quarter of
2013, although the difference for RUVT is less noticeable than the difference for
BURVT.
•	Similar to BURVT, concentrations of 1,2-dichloroethane measured at RUVT appear
higher during the first half of the year. Concentrations of 1,2-dichloroethane range
from 0.05 |ig/m3 to 0.14 |ig/m3, plus four non-detects. All of the concentrations
greater than 0.1 |ig/m3 were measured between January and July and all four non-
detects were measured in October and November.
•	Concentrations of ethylbenzene measured at RUVT also appear lowest during the
fourth quarter of 2013. Concentrations of ethylbenzene measured at RUVT span an
order of magnitude, ranging from 0.065 |ig/m3 to 0.653 |ig/m3, with both the
maximum and minimum concentrations measured during the fourth quarter of 2013.
However, no other ethylbenzene concentrations greater than the median concentration
(0.235 |ig/m3) were measured during the fourth quarter.
27-23

-------
Observations for UNVT from Table 27-5 include the following:
•	Sampling at UNVT occurred on a l-in-6 day schedule.
•	All of the annual average concentrations for the pollutants of interest for UNVT are
less than 1 |ig/m3.
•	Carbon tetrachloride has the highest annual average concentration for UNVT
(0.63 ± 0.03 |ig/m3). The annual average concentrations of this pollutant are similar
across the three Vermont sites, differing by only 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 the Vermont sites as well as all NMP sites sampling benzene.
•	Concentrations of benzene measured during the first quarter of 2013 are considerably
higher than those measured during the rest of the year, based on the quarterly average
concentrations shown in Table 27-5. Concentrations of benzene measured at UNVT
range from 0.109 |ig/m3 to 3.67 |ig/m3. The maximum concentration was measured
on February 21, 2013 and is more than twice the next highest concentration
(1.17 |ig/m3), also measured in February. All other benzene concentrations measured
at UNVT are less than 0.65 |ig/m3. Of the 12 highest benzene measurements collected
at UNVT, all but two were measured during the first quarter of 2013 and only one of
the first quarter benzene concentrations is less than the median concentration for the
year (0.25 |ig/m3).
•	UNVT has the fewest measured detections of 1,3-butadiene (11) among NMP sites
sampling VOCs with Method TO-15, none of which were measured prior to the end
of June.
•	Similar to BURVT and RUVT, concentrations of 1,2-dichloroethane measured at
UNVT appear higher during the first half of the year. Concentrations of
1,2-dichloroethane range from 0.045 |ig/m3 to 0.15 |ig/m3, plus seven non-detects. All
but one of the nine concentrations greater than 0.1 |ig/m3 were measured between
January and May and six of the seven non-detects were measured during the second
half of the year.
•	Arsenic was detected in most of the metals samples collected at UNVT. In addition to
four non-detects, concentrations of arsenic range from 0.05 ng/m3 to 0.80 ng/m3.
Among NMP sites sampling arsenic, UNVT has the lowest annual average
concentration of this pollutant (0.28 ± 0.05 ng/m3).
Tables 4-9 through 4-12 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for the Vermont
monitoring sites from those tables include the following:
27-24

-------
•	BURVT appears twice in Table 4-9 for VOCs. BURVT has the sixth highest annual
average concentration of p-dichlorobenzene and the 10th highest annual average
concentration of 1,2-dichloroethane among NMP sites sampling VOCs.
•	RUVT does not appear in Table 4-9 for VOCs.
•	UNVT does not appear in Tables 4-9 through 4-12 and is often among the sites with
lowest annual average concentrations for the program-level pollutants of interest.
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 pollutants shaded in
gray in Table 27-4 for BURVT, RUVT, and UNVT. Figures 27-9 through 27-16 overlay the
sites' minimum, annual average, and maximum concentrations onto the program-level minimum,
first quartile, median, average, third quartile, and maximum concentrations, as described in
Section 3.4.3.1.
Figure 27-9. Program vs. Site-Specific Average Arsenic (PMio) Concentration
UNVT
0
4
5
Concentration {ng/m3)
6
7
8
9
10
Program: 1st Quartile
2nd Quartile 3rd Quartile 4th Quartile Average
~ ~ ~
Site:
Site Average Site Concentration Range
o
27-25

-------
Figure 27-10. Program vs. Site-Specific Average Benzene Concentrations
1+
Program Max Concentration = 43.5 ^ig/m3
¦4
Program Max Concentration = 43.5 ^ig/m3
r
Program Max Concentration = 43.5 ^ig/m3
0
2 4
6
Concentration {[jg/m3)
8
10

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i

Site: Site Average
o
Site Concentration Range


Figure 27-11. Program vs. Site-Specific Average 1,3-Butadiene Concentrations

Program Max Concentration = 21.5 ^ig/m3

Program Max Concentration = 21.5 ^ig/m3
w
Program Max Concentration = 21.5 ^ig/m3
0
0.3
0.6 0.9
Concentration {[jg/m3)

1.2

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i

Site: Site Average
o
Site Concentration Range


27-26

-------
Figure 27-12. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations
Program Max Concentration = 23.7 ^ig/m3
Program Max Concentration = 23.7 ^ig/m3
Program Max Concentration = 23.7 ^ig/m3
0.75	1	1.25
Concentration (jig/m3)
Program: IstQuartile
2nd Qua rti 1 e 3rd Qua rti 1 e
4th Qua rti le
Average
¦
~
~
~
1
Site: Site Average
Site Concentration Range


o



Figure 27-13. Program vs. Site-Specific Average />-Dichlorobenzene Concentration
0
0.1
0.2
0.3 0.4
Concentration {[jg/m3)
0.5
0.6
0.7

Program:
1st Qua rti le
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i


Site:
Site Average
o
Site Concentration Range



27-27

-------
Figure 27-14. Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations


¦ ,

in
Program Max Concentration = 111 ^ig/m3
I
Program Max Concentration = 111 |xg/m3
0
0.2
0.4 0.6
Concentration (ng/m3)

0.8

Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i

Site: Site Average
o
Site Concentration Range


Figure 27-15. Program vs. Site-Specific Average Ethylbenzene Concentration
Fh
Program Max Concentration = 18.7 ^ig/m3
0
1 2
3
Concentration {[jg/m3)
4
5

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i

Site: Site Average
o
Site Concentration Range


27-28

-------
Figure 27-16. Program vs. Site-Specific Average Hexachloro-l,3-Butadiene Concentration
BURVT
—o
0
0.05
0.1
0.15
Concentration {[jg/m3)
0.2
0.25
0.3
Program: IstQuartile
2nd Quartile 3rdQuartile 4thQuartile Average
~ ~ ~
Site:
Site Average Site Concentration Range
o
Observations from Figures 27-9 through 27-16 include the following:
•	Figure 27-9 presents the box plot for arsenic. UNVT is the only Vermont site that
sampled PMio metals. The maximum arsenic concentration measured at UNVT is
less than the program-level third quartile and the only site-specific maximum
concentration less than 1 ng/m3. UNVT's annual average arsenic (PMio)
concentration is similar to 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.
•	Figure 27-10 for benzene shows all three Vermont sites. Note that the program-
level maximum concentration (43.5 |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 12 |ig/m3. Even though the range of benzene concentrations
is largest for UNVT, this site has the lowest annual average concentration both
among the Vermont sites and across the program. The annual average benzene
concentration for UNVT is just less than the program-level first quartile. RUVT
has the highest annual average concentration of benzene among the Vermont
sites. The annual average concentration for RUVT is similar to the program-level
average concentration. The smallest range of benzene concentrations was
measured at BURVT, whose annual average concentration is less than the
program-level average but greater than the program-level median concentration.
•	Figure 27-11 for 1,3-butadiene also shows all three sites. Note that the program-
level maximum concentration (21.5 |ig/m3) is not shown directly on the box plots
because the scale of the box plots would be too large to readily observe data
points at the lower end of the concentration range. Thus, the scale of the box plots
has been reduced to 1.5 |ig/m3. The range of 1,3-butadiene concentrations is
smallest for UNVT and largest for RUVT. The maximum 1,3-butadiene
concentration measured at RUVT is greater than the program-level average
concentration. The maximum concentration measured at BURVT is less than the
program-level average concentration and the maximum concentration measured at
UNVT is similar to the program-level median concentration. The annual average
27-29

-------
concentration for RUVT is just greater than the program-level third quartile; the
annual average concentration for BURVT is just greater than the program-level
median concentration; and the annual average for UNVT is the lowest annual
average concentration calculated among NMP sites sampling 1,3-butadiene with
Method TO-15 and is one-sixth the program-level average concentration. It
should be noted however, that the program-level average concentration is an order
of magnitude less than the scale of the box plots and is being driven by a few
measurements at the upper end of the concentration range.
Figure 27-12 presents the box plots for carbon tetrachloride for all three sites.
Similar to other VOCs, the program-level maximum concentration (23.7 |ig/m3) is
not shown directly on the box plots 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. The maximum concentrations of carbon tetrachloride are
similar among the Vermont sites. There is more variability in the minimum
concentrations measured at these sites. The annual average concentrations
calculated for these sites are similar to each other and the program-level median
concentration of carbon tetrachloride.
Figure 27-13 is the box plot for /;-dichlorobenzene for BURVT, the only Vermont
site for which this pollutant is a pollutant of interest. Note that the first and second
quartiles are not visible on the box plot because they are zero due to the large
number of non-detects of this pollutant. The maximum concentration measured at
BURVT is about one-sixth of the program-level maximum concentration. The
annual average concentration for BURVT is greater than the program-level
average concentration and just less than the program-level third quartile.
Figure 27-14 presents the box plots for 1,2-dichloroethane for all three sites. Note
that the program-level maximum concentration (111 |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. All of the
measurements of 1,2-dichloroethane measured at the Vermont sites are less than
the program-level average concentration. The program-level average
concentration for this pollutant is being driven by the highest concentrations
measured at a few monitoring sites. The annual average concentrations of
1,2-dichloroethane for the Vermont sites are similar to each and just less than the
median concentration at the program level.
Figure 27-15 is the box plot for ethylbenzene for RUVT, the only Vermont site
for which ethylbenzene is a pollutant of interest. The program-level maximum
concentration (18.7 |ig/m3) is not shown directly on the box plot as the scale has
been reduced to 6 |ig/m3 in order to allow for the observation of data points at the
lower end of the concentration range. The range of ethylbenzene concentrations
measured at RUVT is relatively small. The annual average concentration for
RUVT falls between the program-level median and average concentrations (and
the third quartile).
27-30

-------
• Figure 27-16 presents the box plot for hexachloro-1,3-butadiene for BURVT, the
only Vermont site for which this pollutant is a pollutant of interest. 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 in only four of the 31 valid VOC samples collected at BURVT in 2013.
The maximum concentration of hexachloro-1,3-butadiene measured at BURVT is
less than half the maximum concentration measured across the program. The
annual average concentration for BURVT is just less than the program-level
average concentration.
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 3.4.3.2.
UNVT has sampled PMio metals under the NMP since 2008. In addition, sampling for VOCs
under the NMP began at all three Vermont sites in 2009. Thus, Figures 27-17 through 27-32
present the annual statistical metrics for the pollutants of interest for each of the Vermont sites,
first for BURVT, then for RUVT and UNVT. The statistical metrics presented for assessing
trends include the substitution of zeros for non-detects. If sampling began mid-year, a minimum
of 6 months of sampling is required for inclusion in the trends analysis; in these cases, a 1-year
average concentration is not provided, although the range and percentiles are still presented.
27-31

-------
Figure 27-17. Yearly Statistical Metrics for Benzene Concentrations Measured at BURVT
1.75 -

1.25 -


—5—















0.50 -
0.25 -
0.00 -







	9	

—2—

I


2009 1	2010	2011	2012	2013
0 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
1A 1-year average is not presented due to a low completeness for 2009.
Observations from Figure 27-17 for benzene measurements collected at BURVT include
the following:
•	BURVT began sampling VOCs under the NMP in February 2009. However, a 1-year
average concentration is not provided for 2009 because the late start combined with
low completeness and a l-in-12 sampling schedule did not yield enough valid
samples. However, the range of concentrations measured in 2009 is still provided.
•	The smallest range of benzene concentrations was measured at BURVT in 2009.
Although the range of concentrations widened a little each year through 2011, the
median concentration calculated for each year through 2012 changed little, hovering
on either side of 0.75 |ig/m3. Between 2010 and 2012, the 1-year average
concentration did not change significantly, ranging from 0.77 |ig/m3 (2011) and
0.80 |ig/m3 (2010).
•	Each of the statistical parameters exhibit a decrease for 2013, with each of them at a
minimum for 2013 over the period of sampling shown.
27-32

-------
Figure 27-18. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
BURVT
2011
Year
O 5th Percentile	— Minimum
O 95th Percentile
• Average
1A 1-year average is not presented due to a low completeness for 2009.
Observations from Figure 27-18 for 1,3-butadiene measurements collected at BURVT
include the following:
•	Concentrations of 1,3-butadiene measured at BURVT in 2009 were all less than
0.10 |ig/m3.
•	Each of the statistical parameters increased for 2010, with the exception of the
minimum concentration, which remained at zero due to a single non-detect measured.
For 2010, more than one-third of the concentrations measured were greater than
0.10 |ig/m3.
•	Further increases in the statistical parameters are shown for 2011, including the
minimum concentration as no non-detects were measured. With the exception of the
maximum concentration, all of the statistical parameters for 2011 are at a maximum
for the period of sampling shown.
•	Each of the statistical parameters is shown for 2012 exhibits a decrease from 2011,
except the maximum concentration. This is true for 2013 as well. Excluding 2009,
each of the statistical parameters are at a minimum for 2013.
27-33

-------
Figure 27-19. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at BURVT
0.75
2009 1	2010	2011	2012	2013
O 5th Percentile	— Minimum	~ Median	- Maximum	O 95th Percentile	Average
1A 1-year average is not presented due to a low completeness for 2009.
Observations from Figure 27-19 for carbon tetrachloride measurements collected at
BURVT include the following:
•	The carbon tetrachloride concentrations measured at BURVT since 2009 vary by less
than 0.5 |ig/m3. The maximum concentration (0.845 |ig/m3) was measured in 2009
and the minimum concentration (0.360 |ig/m3) was measured in 2011 and again in
2013.
•	The median concentrations calculated for each year of sampling span less than
0.1 |ig/m3, ranging from 0.61 |ig/m3 (2011) and 0.69 |ig/m3 (2012). This is also true
for the 1-year average concentration, which ranges from 0.60 |ig/m3 (2011) to
0.68 |ig/m3 (2012).
•	The concentrations measured in 2010, 2011, and 2013 are fairly similar. The range of
measurements is considerably tighter for 2012, as little change in the maximum
concentration is shown and nine concentrations measured in 2011 are less than the
minimum concentration measured in 2012. But the increase in the statistical
parameters shown is not just a result of a higher minimum concentration;
concentrations were higher overall in 2012. The number of carbon tetrachloride
concentrations greater than or equal to 0.7 |ig/m3 measured at BURVT increased from
three in 2011 to 14 for 2012, accounting for nearly half of the measurements for
2012. The next highest year has five (both 2009 and 2013).
27-34

-------
Figure 27-20. Yearly Statistical Metrics for o-Dichlorobenzene Concentrations Measured at
BURVT
0.25
0.20
2009
2010
2011
2012
2013
O 5th Percentile	— Minimum	~ Median	- Maximum	O 95th Percentile •••-">••* Average
1A 1-year average is not presented due to a low completeness for 2009.
Observations from Figure 27-20 for p-dichlorobenzene measurements collected at
BURVT include the following:
•	The minimum and 5th percentile for each year shown in Figure 27-20 is zero,
indicating that at least 5 percent of the measurements were non-detects for each year.
The median concentration is also zero for 2010, indicating that at least half of the
measurements for 2010 were non-detects. The percentage of non-detects has varied
from 10 percent (2012) to 63 percent (2010).
•	The maximum and 95th percentile increased each year between 2009 and 2011, when
the highest p-dichlorobenzene concentration was measured (0.14 |ig/m3). These
parameters have a decreasing slight trend in the years that follow.
•	Both the 1-year average and median concentrations increased significantly from 2010
to 2011. Not only did the number of non-detects decrease considerably (from
63 percent to 17 percent), six concentrations greater than the maximum concentration
for 2010 were measured in 2011.
•	Despite the slight decreases shown in the upper end of the concentration range shown
for 2012 and 2013, the 1-year average and median concentrations did not change
significantly.
27-35

-------
Figure 27-21. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured
at BURVT
0.15
2009 1	2010	2011	2012	2013
O 5th Percentile	— Minimum	~ Median	- Maximum	O 95th Percentile	Average
1A 1-year average is not presented due to a low completeness for 2009.
Observations from Figure 27-21 for 1,2-dichloroethane measurements collected at
BURVT include the following:
•	The minimum, 5th percentile, and median concentration for each year shown through
2011 in Figure 27-21 is zero, indicating that at least half of the measurements for each
year through 2011 were non-detects. The percentage of non-detects measured at
BURVT has decreased each year, from a maximum of 91 percent to a minimum of
3 percent. A sharp decrease in the number of non-detects occurred between 2011,
when the percentage of non-detects was at 77 percent, and 2012, when the percentage
of non-detects fell to 6 percent. This change is reflected in the statistical parameters
representing the lower end of the concentration range for 2012 as well as those
representing the central tendency statistics of the dataset.
•	The 95th percentile and maximum concentrations have increased (albeit slightly) each
year of sampling. Thus, magnitude of the concentrations at the upper end of the
concentration range have also increased over the course of sampling.
27-36

-------
Figure 27-22. Yearly Statistical Metrics for Hexachloro-l,3-Butadiene Concentrations
Measured at BURVT
2009 1	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	~ Median	- Maximum	O 95th Percentile	Average
1A 1-year average is not presented due to a low completeness for 2009.
Observations from Figure 27-22 for hexachloro-l,3-budadiene measurements collected at
BURVT include the following:
•	The minimum, 5th percentile, and median concentration for each year shown in
Figure 27-22 is zero, indicating that at least half of the measurements for each year
were non-detects. In fact, all of the measurements were non-detects for 2009 and all
but one were non-detects for 2010.
•	Between 2011 and 2013, the percentage of non-detects ranged from 87 percent (2013)
to 94 percent (2012), still accounting for the majority of measurements. No more than
four measured detections of hexachloro-l,3-budadiene have been measured at
BURVT in any given year.
27-37

-------
Figure 27-23. Yearly Statistical Metrics for Benzene Concentrations Measured at RUVT
2009	2010	2011	2012	2013
O 5th Percentile	— Minimum	— Median	— Maximum	O 95th Percentile	Average
Observations from Figure 27-23 for benzene measurements collected at RUVT include
the following:
•	Sampling for VOCs at RUVT under the NMP also began in February 2009.
•	The maximum benzene concentration was measured at RUVT in 2010 (2.91 |ig/m3).
Six additional benzene concentrations greater than 2 |ig/m3 have been measured at
RUVT, at least one in each year, with the exception of 2009.
•	The maximum concentration doubled from 2009 to 2010 and the 95th percentile
increased by 75 percent. The other statistical parameters also exhibit increases. The
number of benzene concentrations greater than 0.75 |ig/m3 doubled from eight in
2009 to 16 for 2010, accounting for 60 percent of the measurements in 2010.
•	Years with higher benzene concentrations alternate with years with lower
concentrations, giving the box and whisker plots an undulating pattern. The 1-year
average concentrations have varied from 0.72 |ig/m3 (2009) to 1.04 |ig/m3 (2012); the
median concentrations have varied from 0.61 |ig/m3 (2013) to 0.88 |ig/m3 (2012).
27-38

-------
Figure 27-24. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
RUVT
2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	— Maximum	O 95th Percentile	Average
Observations from Figure 27-24 for 1,3-butadiene measurements collected at RUVT
include the following:
•	At least one non-detect of 1,3-butadiene was measured at RUVT each year of
sampling, with the exception of 2009. The number of non-detects has ranged from
zero (2009) to five (2011).
•	The 1-year average concentration for 2010 is greater than the 95th percentile for
2009, indicating that concentrations measured in 2010 were higher than those
measured in 2009. The number of 1,3-butadiene concentrations greater than or equal
to 0.1 |ig/m3 increased from four in 2009 to 10 for 2010. While the range of
1,3-butadiene concentrations measured tripled from 2009 to 2010 and the 1-year
average concentration doubled, the increase in the median concentration is less
dramatic.
•	The box and whisker plot for 1,3-butadiene resembles the box and whisker plot for
benzene, with a similar undulating pattern from year to year. The 1-year average
concentration has ranged from 0.06 |ig/m3 (2009) to 0.13 |ig/m3 (2012) while the
median has ranged from 0.05 |ig/m3 (2009) to 0.09 |ig/m3 (2013).
27-39

-------
Figure 27-25. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at RUVT
20091	2010	2011	2012	2013
O 5th Percentile	— Minimum	~ Median	- Maximum	O 95th Percentile	Average
1A 1-year average is not presented due to low completeness in 2009.
Observations from Figure 27-25 for carbon tetrachloride measurements collected at
RUVT include the following:
•	A few individual carbon tetrachloride concentrations from valid VOC samples were
invalidated in 2009, resulting in a completeness less than 85 percent. As a result, a
1-year average concentration is not presented, although the range of measurements is
still provided.
•	Concentrations of carbon tetrachloride measured at RUVT since 2009 span a
relatively small range, with a minimum concentration of 0.38 |ig/m3 (2010 and 2012)
and a maximum concentration of 0.91 |ig/m3 (2009). Five of the six highest carbon
tetrachloride concentrations (those greater than 0.85 |ig/m3) were measured in 2009.
•	All of the statistical parameters exhibit a decrease from 2009 to 2010. Five
concentrations measured in 2009 are greater than the maximum concentration
measured in 2010 and six concentrations measured in 2010 are less than the minimum
concentration measured in 2009. Yet, the decrease in the median concentration is
relatively small, from 0.69 |ig/m3 for 2009 to 0.66 |ig/m3 for 2010.
•	The majority of concentrations, as indicated by the 5th and 95th percentiles, fell into a
tighter range for 2011, with similar ranges in the years that follow. Most of the
27-40

-------
measurements collected at RUVT between 2011 and 2013 fell between roughly
0.5 |ig/m3 and 0.8 |ig/m3.
Figure 27-26. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured
at RUVT
2009	2010	2011	2012	2013
O 5th Percentile	— Minimum	— Median	— Maximum	O 95th Percentile • ••£¦•~ Average
Observations from Figure 27-26 for 1,2-dichloroethane measurements collected at RUVT
include the following:
•	The box and whisker plot for 1,2-dichloroethane for RUVT resembles the box and
whisker plot for 1,2-dichloroethane for BURVT.
•	The minimum, 5th percentile, and median concentration for each year through 2011
are zero, indicating that at least half of the measurements for each year through 2011
were non-detects. The percentage of non-detects measured at RUVT during the first
3 years of sampling ranged from 87 percent (2011) to 92 percent (2009). A sharp
decrease in the number of non-detects occurred after 2011. Only two non-detects
were measured in 2012 and three non-detects were measured in 2013. This decrease
in non-detects (and thus, zeros substituted in the calculations) is reflected in the
statistical parameters for these years, particularly in the central tendency statistics of
the dataset.
•	The 95th percentile and maximum concentrations have increased each year, with the
exception of 2011, which exhibits no change. Thus, magnitude of the concentrations
27-41

-------
at the upper end of the concentration range have also increased over the course of
sampling.
Figure 27-27. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at
RUVT
I
0
2009
2010

2011
Year
2012
2013
O 5th Percentile
— Minimum
— Median
— Maximum
O 95th Percentile
• ••£¦•• Average
Observations from Figure 27-27 for ethylbenzene measurements collected at RUVT
include the following:
•	The maximum benzene concentration was measured at RUVT in 2012 (0.69 |ig/m3),
although similar concentrations have been in each year of sampling, with the
exception of 2009, when no concentrations greater than 0.5 |ig/m3 were measured.
•	The 1-year average concentration for 2010 is significantly greater than the 1-year
average concentration for 2009. Six ethylbenzene concentrations measured in 2010
are greater than the maximum concentration measured in 2009 while concentrations
at the lower end of the concentration range changed little. Nearly 88 percent of the
ethylbenzene concentrations measured in 2009 are less than 0.25 |ig/m3, which is the
median concentration for 2010.
•	The median concentration has an increasing trend between 2009 and 2012, nearly
doubling from 0.17 |ig/m3 for 2009 to 0.33 |ig/m3 for 2012. The 1-year average
concentration has a similar pattern, with the exception of 2011, for which virtually no
change is shown.
27-42

-------
• For 2013, the measurements at the upper end of the concentration range changed
little, but a decrease is shown at the lower end of the concentration range. Six
concentrations measured in 2013 are less than the minimum concentration measured
in 2012 and the number of ethylbenzene concentrations less than 0.25 |ig/m3 doubled
from 2012 (seven) to 2013 (14), accounting for half of the measurements collected in
2013. This explains the decreases shown in the 1-year average and median
concentrations.
Figure 27-28. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at
UNVT
2010	2011
Year
O 5th Percentile
Median	- Maximum	O 95th Percentile	Average
Observations from Figure 27-28 for arsenic measurements collected at UNVT include the
following:
•	The maximum arsenic concentration was measured at UNVT in 2012 (0.90 ng/m3).
•	With the exception of the 95th percentile, each of the statistical parameters exhibits a
slight decreasing trend between 2008 and 2010. The minimum concentration 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). Between three and six non-detects were measured
each year following 2010.
Overall, a similar range of arsenic concentrations have been measured at UNVT each
year. The 1-year average concentrations of arsenic for UNVT have changed little over
the years of sampling, ranging from 0.21 ng/m3 (2010) to 0.28 ng/m3 (2013).
27-43

-------
Likewise, the median concentration has ranged from 0.17 ng/m3 (2010) to 0.26 ng/m3
(2013).
Figure 27-29. Yearly Statistical Metrics for Benzene Concentrations Measured at UNVT
2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	~ Maximum	O 95th Percentile	Average
Observations from Figure 27-29 for benzene measurements collected at UNVT include
the following:
•	Only two benzene concentrations greater than 1.0 |ig/m3 have been measured at
UNVT, one in 2011 (1.30 |ig/m3) and one in 2013 (1.19 |ig/m3).
•	All of the statistical parameters exhibit increases from 2009 to 2010, with the largest
increases shown for the maximum and 95th percentile. Despite the higher maximum
concentration for 2011, little change is shown in most of the statistical parameters for
2011.
•	All of the statistical parameters exhibit decreases from 2011 to 2012, although the
differences are small for most of them.
•	Even though the second highest benzene concentration was measured in 2013, the
1-year average concentration is at its lowest since 2009 and the median concentration
is at a minimum over the period of sampling. This is due to an increase in the
measurements at the lower end of the concentration range. The number of benzene
concentrations less than 0.3 |ig/m3 more than doubled, from 17 in 2012 to 37 in 2013,
27-44

-------
accounting for nearly 64 percent of the concentrations for 2013. No other year has
more than 30 and most years have fewer than 20.
Figure 27-30. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
UNVT
0.25




0.10 -
0.05 -
o.oo -








I


0 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
Observations from Figure 27-30 for 1,3-butadiene measurements collected at UNVT
include the following:
•	The median 1,3-butadiene concentration (along with the minimum and 5th percentile)
is zero for all years of sampling, indicating that at least half of the measurements were
non-detects. The percentage of non-detects has ranged from 76 percent (2012) to
81 percent (2013).
•	The 1-year average concentration, the 95th percentile, and the maximum
concentration exhibit increases for each year of sampling between 2009 and 2012,
with the largest increase shown for 2012. Prior to 2011, no 1,3-butadiene
concentrations greater than 0.05 |ig/m3 were measured at UNVT; in 2011, two
concentrations greater than 0.05 |ig/m3 were measured. Prior to 2012, no
1,3-butadiene concentrations greater than 0.10 |ig/m3 were measured at UNVT; in
2012, six concentrations greater than 0.10 |ig/m3 were measured. The measurements
collected in 2013 more resemble those measured in 2011, when only one
concentration greater than 0.05 |ig/m3 was measured.
27-45

-------
• The 1-year average concentrations reflect the increases in the magnitude of the
1,3-butadiene measurements collected at UNVT. Yet, measured detections account
for fewer than one-quarter of the measurements collected at UNVT for any given
year.
Figure 27-31. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at UNVT
20091	2010	2011	2012	2013
O 5th Percentile	— Minimum	— Median	— Maximum	O 95th Percentile	Average
1A 1-year average is not presented due to low completeness in 2009
Observations from Figure 27-31 for carbon tetrachloride measurements collected at
UNVT include the following:
•	A few individual carbon tetrachloride concentrations from valid VOC samples were
invalidated in 2009, resulting in a completeness less than 85 percent. As a result, a
1-year average concentration is not presented for 2009, although the range of
measurements is still provided.
•	All of the statistical parameters shown exhibit a decreasing trend through 2011, when
each parameter is at a minimum over the period of sampling.
•	All of the statistical parameters increased for 2012, including a statistically significant
increase in the 1-year average concentration. These changes are the result of an
increase in the number of concentrations at the upper end of the concentration range
as well as fewer concentrations at the lower end of the concentration The number of
carbon tetrachloride concentrations greater than or equal to 0.7 |ig/m3 tripled from
27-46

-------
eight in 2011 to 24 in 2012. At the lower end of the concentration range, the number
of concentrations less than 0.5 |ig/m3 decreased from 10 in 2011 to only one in 2012.
The minimum concentration measured in 2012 is greater than five of the lower
concentrations measured in 2011.
• Although the range of concentrations measured is similar between 2012 and 2013,
slight decreases are shown in the central tendency statistics for 2013. The number of
carbon tetrachloride concentrations greater than 0.7 |ig/m3 decreased by half (down to
12 in 2013 from 24 in 2012).
Figure 27-32. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured
at UNVT
2011
Year
O 5th Percentile
- Maximum	O 95th Percentile	Average
Observations from Figure 27-32 for 1,2-dichloroethane measurements collected at UNVT
include the following:
• The box and whisker plot for 1,2-dichloroethane for UNVT resembles the box and
whisker plots for 1,2-dichloroethane for BURVT and RUVT.
• The minimum, 5th percentile, and median concentration for each year shown through
2011 in Figure 27-32 are zero, indicating that at least half of the measurements for
each year through 2011 were non-detects. The percentage of non-detects measured at
UNVT has decreased each year of sampling, from a maximum of 94 percent in 2009
to a minimum of 12 percent in 2013. A sharp decrease in the number of non-detects
occurred after 2011. This decrease in non-detects (and thus, zeros substituted in the
27-47

-------
calculations) is reflected in the statistical parameters for these years, particularly in
the central tendency statistics of the dataset.
• The 95th percentile has increased for each year of sampling. The maximum
concentration increased from 2009 to 2010, did not change through 2012, then
increased again for 2013. Thus, the magnitude of concentrations at the upper end of
the concentration range have also increased over the course of sampling. However,
the overall concentration range is relatively small.
27.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at the Vermont monitoring sites. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4
for definitions and explanations regarding the various toxicity factors, time frames, and
calculations associated with these risk-based screenings.
27.5.1 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 3.4.3.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 27-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
27-48

-------
Table 27-6. Risk Approximations for the Vermont Monitoring Sites
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Burlington, Vermont - BURVT
Benzene
0.0000078
0.03
31/31
0.65
±0.06
5.06
0.02
1.3 -Butadiene
0.00003
0.002
30/31
0.07
±0.01
2.20
0.04
Carbon Tetrachloride
0.000006
0.1
31/31
0.62
±0.03
3.72
0.01
p-Dichlorobenzene
0.000011
0.8
27/31
0.06
±0.01
0.68
<0.01
1,2-Dichloroethane
0.000026
2.4
30/31
0.08
±0.01
2.17
<0.01
Hexachloro-1,3 -butadiene
0.000022
0.09
4/31
0.01
±0.01
0.26
<0.01
Rutland, Vermont - RUVT
Benzene
0.0000078
0.03
31/31
0.81
±0.18
6.29
0.03
1.3 -Butadiene
0.00003
0.002
29/31
0.11
±0.03
3.33
0.06
Carbon Tetrachloride
0.000006
0.1
31/31
0.63
±0.03
3.80
0.01
1,2-Dichloroethane
0.000026
2.4
27/31
0.08
±0.01
1.99
<0.01
Ethylbenzene
0.0000025
1
31/31
0.27
±0.05
0.67
<0.01
Underhill, Vermont - UNVT
Benzene
0.0000078
0.03
60/60
0.37
±0.12
2.86
0.01
1,3-Butadiene
0.00003
0.002
11/60
0.01
±<0.01
0.19
<0.01
Carbon Tetrachloride
0.000006
0.1
60/60
0.63
±0.03
3.76
0.01
1,2-Dichloroethane
0.000026
2.4
53/60
0.07
±0.01
1.84
<0.01
Arsenic (PMi0)a
0.0043
0.000015
56/60
0.28
±0.05
1.22
0.02
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of
viewing.
Observations from Table 27-6 include the following:
•	For BURVT, benzene and carbon tetrachloride have the highest annual average
concentrations. These two pollutants also have the highest cancer risk approximations
for BURVT (5.06 in-a-million and 3.72 in-a-million, respectively).
•	Benzene and carbon tetrachloride also have the highest annual average concentrations
for RUVT. These pollutants have the highest cancer risk approximations for RUVT
27-49

-------
(6.29 in-a-million and 3.80 in-a-million, respectively), although a similar cancer risk
approximation was calculated for 1,3-butadiene (3.33 in-a-million).
•	Carbon tetrachloride has the highest annual average concentration for UNVT,
followed by benzene. These two pollutants have the highest cancer risk
approximations for UNVT (3.76 in-a-million for carbon tetrachloride and 2.86 in-a-
million for benzene).
•	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 noncancer
health effects are expected from these individual pollutants.
27.5.2 Risk-Based Emissions Assessment
In addition to the risk-based screenings discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 27-7 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
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 3.4.3.4. Lastly,
Table 27-7 provides the pollutants with the highest cancer risk approximations (in-a-million) for
each site, 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.
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 27.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
27-50

-------
Table 27-7. 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)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Burlington, Vermont (Chittenden County) - BURVT
Benzene
103.48
Formaldehyde
8.77E-04
Benzene
5.06
Formaldehyde
67.43
Benzene
8.07E-04
Carbon Tetrachloride
3.72
Acetaldehyde
37.96
1,3-Butadiene
4.06E-04
1,3-Butadiene
2.20
Ethylbenzene
37.92
Arsenic, PM
3.13E-04
1,2-Dichloroethane
2.17
1.3 -Butadiene
13.53
Naphthalene
2.29E-04
/?-Dichlorobcnzcnc
0.68
Naphthalene
6.75
POM, Group 2b
1.52E-04
Hexachloro-1,3 -butadiene
0.26
Dichloromethane
2.55
Hexavalent Chromium
1.19E-04

Tetrachloroethylene
2.22
POM, Group 5a
1.06E-04
POM, Group 2b
1.73
Nickel, PM
9.73E-05
POM, Group 2d
1.02
Ethylbenzene
9.48E-05
Underhill, Vermont (Chittenden County) - UNVT
Benzene
103.48
Formaldehyde
8.77E-04
Carbon Tetrachloride
3.76
Formaldehyde
67.43
Benzene
8.07E-04
Benzene
2.86
Acetaldehyde
37.96
1,3-Butadiene
4.06E-04
1,2-Dichloroethane
1.84
Ethylbenzene
37.92
Arsenic, PM
3.13E-04
Arsenic
1.22
1,3-Butadiene
13.53
Naphthalene
2.29E-04
1,3-Butadiene
0.19
Naphthalene
6.75
POM, Group 2b
1.52E-04

Dichloromethane
2.55
Hexavalent Chromium
1.19E-04
Tetrachloroethylene
2.22
POM, Group 5a
1.06E-04
POM, Group 2b
1.73
Nickel, PM
9.73E-05
POM, Group 2d
1.02
Ethylbenzene
9.48E-05

-------
Table 27-7. 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)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Rutland, Vermont (Rutland County) - RUVT
Benzene
48.90
Benzene
3.81E-04
Benzene
6.29
Formaldehyde
25.16
Formaldehyde
3.27E-04
Carbon Tetrachloride
3.80
Acetaldehyde
17.87
1,3-Butadiene
1.64E-04
1,3-Butadiene
3.33
Ethylbenzene
14.81
Naphthalene
1.13E-04
1,2-Dichloroethane
1.99
1.3 -Butadiene
5.47
POM, Group 2b
7.01E-05
Ethylbenzene
0.67
Naphthalene
3.32
POM, Group 5a
6.22E-05

POM, Group 2b
0.80
Arsenic, PM
4.26E-05
POM, Group 2d
0.45
POM, Group 2d
3.94E-05
T etrachloroethy lene
0.38
Acetaldehyde
3.93E-05
T richloroethy lene
0.30
Ethylbenzene
3.70E-05

-------
Table 27-8. 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)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Burlington, Vermont (Chittenden County) - BURVT
Toluene
250.92
Acrolein
546,915.43
1,3-Butadiene
0.04
Xylenes
174.70
Chlorine
12,098.33
Benzene
0.02
Hexane
106.23
Manganese, PM
9,934.56
Carbon Tetrachloride
0.01
Benzene
103.48
Formaldehyde
6,880.35
Hexachloro-1,3 -butadiene
<0.01
Methanol
90.73
1.3 -Butadiene
6,767.01
p-Dichlorobenzene
<0.01
Formaldehyde
67.43
Arsenic, PM
4,859.91
1,2-Dichloroethane
<0.01
Acetaldehyde
37.96
Acetaldehyde
4,218.14

Ethylbenzene
37.92
Benzene
3,449.29
Hydrochloric acid
35.41
Cadmium, PM
2,474.68
Ethylene glycol
31.16
Nickel, PM
2,252.18
Underhill, Vermont (Chittenden County) - UNVT
Toluene
250.92
Acrolein
546,915.43
Arsenic
0.02
Xylenes
174.70
Chlorine
12,098.33
Benzene
0.01
Hexane
106.23
Manganese, PM
9,934.56
Carbon Tetrachloride
0.01
Benzene
103.48
Formaldehyde
6,880.35
1,3-Butadiene
<0.01
Methanol
90.73
1,3-Butadiene
6,767.01
1,2-Dichloroethane
<0.01
Formaldehyde
67.43
Arsenic, PM
4,859.91

Acetaldehyde
37.96
Acetaldehyde
4,218.14
Ethylbenzene
37.92
Benzene
3,449.29
Hydrochloric acid
35.41
Cadmium, PM
2,474.68
Ethylene glycol
31.16
Nickel, PM
2,252.18

-------
Table 27-8. 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)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Rutland, Vermont (Rutland County) - RUVT
Toluene
107.88
Acrolein
80,780.90
1,3-Butadiene
0.06
Xylenes
60.02
1.3 -Butadiene
2,734.26
Benzene
0.03
Benzene
48.90
Formaldehyde
2,567.26
Carbon Tetrachloride
0.01
Hexane
39.43
Acetaldehyde
1,985.32
Ethylbenzene
<0.01
Methanol
35.40
Benzene
1,630.13
1,2-Dichloroethane
<0.01
Formaldehyde
25.16
Naphthalene
1,107.67

Acetaldehyde
17.87
Arsenic, PM
660.46
Ethylbenzene
14.81
Xylenes
600.20
Ethylene glycol
12.37
Lead, PM
572.30
Methyl isobutyl ketone
5.54
Cadmium PM
518.83

-------
Observations from Table 27-7 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
toxicity-weighted emissions (of the pollutants with cancer UREs) for both counties,
although not necessarily in that order.
•	Six of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Chittenden County while eight of the highest emitted pollutants also
have the highest toxicity-weighted emissions for Rutland County.
•	Benzene is at or near the top of emissions-based lists for both counties as well as at or
near the top of each site's cancer risk approximations. The cancer risk approximation
for carbon tetrachloride is also among the highest for all three sites, but this pollutant
appears on neither emissions-based list for either county. 1,3-Butadiene is another
pollutant for which a cancer risk approximation could be calculated for all three sites
and appears on both emissions-based lists. Ethylbenzene also appears on both
emissions-based lists for Chittenden and Rutland Counties but is only a pollutant of
interest for RUVT.
•	Arsenic has the fourth highest cancer risk approximation for UNVT and ranks fourth
for its toxicity-weighted emissions, but is not one of the highest emitted in Chittenden
County.
•	Naphthalene ranks fifth for its toxicity-weighted emissions and 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.
•	Hexvalent chromium ranks seventh for its toxicity-weighted emissions for Chittenden
County. Although this pollutant was sampled for at UNVT, none of the
concentrations of this pollutant failed screens.
Observations from Table 27-8 include the following:
•	Toluene and xylenes are the highest emitted pollutants with noncancer RfCs in both
Chittenden and Rutland Counties, although the emissions in Chittenden County were
greater than those in Rutland County.
27-55

-------
•	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
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
toxicity-weighted emissions while four 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 among the pollutants with
the highest toxicity-weighted emissions for both counties, is not among the highest
emitted in either county (of the pollutants with noncancer RfCs).
•	Although very low, arsenic has the highest noncancer hazard approximation for
UNVT. While this pollutant ranks fifth 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 none are not among the highest emitted.
27.6 Summary of the 2013 Monitoring Data for the Vermont Monitoring Sites
Results from several of the data treatments described in this section include the
following:
~~~ A total of eight pollutants failed screens for BURVT; six pollutants failed screens for
RUVT; and nine pollutants failed screens for UNVT.
~~~ None of the pollutants of interest for the Vermont sites had annual average
concentrations greater than 1 ng/m3.
~~~ The detection rate of 1,2-dichloroethane has increased significantly at each of the
Vermont sites in the recent years.
~~~ The annual average concentrations for several of UNVT's pollutants of interest were
the lowest annual averages among NMP sites sampling those pollutants.
27-56

-------
28.0	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.
28.1	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 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,
version 2. 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 boundary are still visible on the map for
reference, but have been grayed out in order to emphasize emissions sources within the
boundary. Table 28-1 provides supplemental geographical information such as land use, location
setting, and locational coordinates.
28-1

-------
Figure 28-1. East Highland Park, Virginia (RIVA) Monitoring Site
^o/ia st
. Source: USC*
>UfCt..N* SA NGA. USCS
2 008»M.«r6«ol» COfp
_ q fonifflM or 9
Gre>g^lori

-------
Figure 28-2. NEI Point Sources Located Within 10 Miles of RIVA
TFMWMI
. 9 r
, Richmond •£) ,
J ^
v0e .
i- -t
"tSBTI*
1N> total fsaeilinss
Iki He a*en of ntetst
irwVr
*ofa Oua to fsctty d»nsirf ana collocation
displayed twy not represent all tocrttefc wt
Legend
~ RIVA NATTS site	10 mite radius	[	 County boundary
Source Category Group (No. of Facilities)
T
AirporVAirltne'Airport Support Operations (10)
»
Miscellaneous Commercovindustriai Facility (1)
*
Asphalt Production/Hot Mb Asphalt Pliint (1)
t
Paint and Coaling Manufactunng Facil.ly (2)
0
Bulk Termmate'Bulk Plants (6)
R
Plastic Resin, or Rubber Products Plant <3)
c
Chemical Manufacturing Facility (2)
P
Pnnting/Publishlng/Paper Product Manufacturing Facility (4)
#
Electricity Generation via Combustion (4)
s
Pulp and Paper Plant (1)
F
Food Processing/Agriculture Faollty (1)
X
Rail Yard.'Rail t Ine Operations (4)
0
Institutional (school, hospital, prison etc ) (2)
*
Testing Laboratories (1)
¦
Landfill (1)
M
Tobacco Manufacturing (1)
©
Metals Processmg/F abricatton Facihty (1)
W
Woodwork I urmture, MHIwork & Wood Preservrng Facility (1)
A
Military Base/National Security Facility (1)


28-3

-------
Table 28-1. Geographical Information for the Virginia Monitoring Site
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Additional Ambient Monitoring Information1
RIVA
51-087-0014
East
Highland
Park
Henrico
Richmond, VA
37.55652,
-77.40027
Residential
Suburban
Lead TSP, CO, S02, NOy, NO, N02, NOx, VOCs,
Carbonyl compounds, O3, Meteorological parameters,
PAMS/NMOC, PM10, PM10 Metals, PM Coarse,
PM2.5. PM2.5 Speciation IMPROVE Speciation.
1 Data for additional pollutants are reported to AQS for this site (EPA, 2014b); however, these data are not generated by ERG and are therefore not included in this report.
BOLD ITALICS = EPA-designated NATTS Site
to
00

-------
The RIVA monitoring site is located just northeast of the capital city of Richmond, in
east-central Virginia. The site is located at the MathScience Innovation Center in a residential
area about one-quarter mile from 1-64. The 1-64 interchange with Mechanicsville Turnpike
(US-360) is one-half mile west of the site, as shown in Figure 28-1. 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 28-2 shows, RIVA is located
near several point sources, most of which are located to the south and southwest of the site and
within the city of Richmond. The sources closest to RIVA are a metals processing and
fabrication facility and a heliport at the Medical College of Virginia. The source categories with
the greatest number of emissions sources within 10 miles of RIVA are the airport source
category, which includes airports and related operations as well as small runways and heliports,
such as those associated with hospitals or television stations; bulk terminals and bulk plants;
printing, publishing, and paper product manufacturers; rail yard and rail line operations; and
facilities generating electricity via combustion.
Table 28-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Virginia monitoring site. Table 28-2 includes both county-level
population and vehicle registration information. Table 28-2 also contains traffic volume
information for RIVA as well as the location for which the traffic volume was obtained.
Additionally, Table 28-2 presents the county-level daily VMT for Henrico County.
Table 28-2. Population, Motor Vehicle, and Traffic Information for the Virginia
Monitoring Site
Site
County
Estimated
County
Population1
County-level
Vehicle
Registration2
Annual
Average Daily
Traffic3
Intersection
Used for
Traffic Data
County-level
Daily VMT4
RIVA
Henrico
318,611
350,000
72,000
1-64 at Mechanicsville
Turnpike
8,366,945
bounty-level population estimate reflects 2013 data (Census Bureau, 2014)
2County-level vehicle registration reflects 2013 data (Henrico County, 2014)
3AADT reflects 2012 data (VA DOT, 2012)
4County-level VMT reflects 2013 data (VA DOT, 2014)
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 28-2 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.
28-5

-------
•	The traffic volume experienced near RIVA is in the top third of the range compared
to other NMP monitoring sites, ranking 18th. 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.
28.2 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.
28.2.1	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 can act as a barrier to cold air and the proximity to the Atlantic
Ocean prevents temperatures from dropping too low. Summers are warm and humid, also due to
these influences. Precipitation is well distributed throughout the year, with greater than 3 inches
typical during most months of the year. A northerly wind is most common during the fall and
winter months while southerly winds prevail during the warmest months of the year (Wood,
2004).
28.2.2	Meteorological Summary
Hourly meteorological data for 2013 were retrieved from NCDC for the weather station
closest to the Virginia monitoring site (NCDC, 2013), as described in Section 3.4.2. 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 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 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)
5.7
miles
129°
(SE)
Sample
Days
(63)
69.0
±4.3
59.4
±4.1
47.1
±4.9
53.3
±4.0
66.9
±3.4
1019.5
± 1.6
6.2
±0.6
2013
68.4
+ 1.7
59.0
+ 1.7
47.0
+ 1.9
53.1
+ 1.6
67.6
+ 1.5
1019.0
±0.7
6.2
±0.3
1 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 2013. 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 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 3.4.2. A wind rose shows the frequency of wind directions using "petals"
positioned around a 16-point compass, and uses different colors to represent wind speeds.
Figure 28-3 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 28-3 also presents three different wind roses for the
RIVA monitoring site. First, a historical wind rose representing 2003 to 2012 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 2013 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2013 is
presented. These can be used to identify the predominant wind speed and direction for 2013 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.
28-8

-------
Figure 28-3. Wind Roses for the Richmond International Airport Weather Station near
RIVA
Location of RIVA and Weather Station	2003-2012 Historical Wind Rose
WIND SPEED
(Knots)
~ =22
Esil 17-21
|| 11-17
~
7- 11
4-7
H 2-4
Calms: 15.84%
2013 Wind Rose
NORTH"*--,
20%
16%.
12%
WIND SPEED
(Knots)
~	*22
~	17-21
11-17
I 1 7- 11
~	4-7
Calms: 16.05%
Sample Day Wind Rose
NORTH---.,
20%
16%
12%
WIND SPEED
(Knots)
~ >=22
I H 17-21
IH 11
CO 7- 11
^3 4-7
2- 4
Calms: 1475%
28-9

-------
Observations from Figure 28-3 for RIVA include the following:
•	The Richmond International Airport weather station is located 5.7 miles
east-southeast of RIVA.
•	The historical wind rose shows that the most commonly observed wind direction is
north, although winds from the south are a close second. Winds from the north-
northeast, south-southwest, and southwest were also frequently observed. Winds from
the southeast and northwest quadrants were observed less frequently. Calm winds
(those less than or equal to 2 knots) were observed for approximately 16 percent of
the hourly wind measurements.
•	The 2013 wind rose resembles the historical wind rose in some ways but exhibits
differences as well. Northerly, southerly, and south-southwesterly winds were still
prominent in 2013 but accounted for a higher percentage of observations while fewer
southwesterly to westerly and north-northeasterly to northeasterly winds were
observed. Southerly winds were observed slightly more often than northerly winds in
2013 (while the reverse is true for the historical wind rose).
•	Southerly winds account for the greatest number of wind observations on sample days
near RIVA (approximately 16 percent), followed by south-southwesterly winds
(roughly 10 percent), both of which are greater than the number of northerly wind
observations (9 percent). The calm rate on sample days is just less than 15 percent.
28.3 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 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 each site when reviewing the results of
this analysis. PAHs and hexavalent chromium were sampled for year-round at RIVA. RIVA is
one of two NATTS sites to continue sampling hexavalent chromium beyond the summer of
2013.
28-10

-------
Table 28-4. Risk-Based Screening Results for the Virginia Monitoring Site
Pollutant
Screening
Value
Oig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
East Highland Park, Virginia - RIVA
Naphthalene
0.029
56
58
96.55
100.00
100.00
Total
56
58
96.55

Observations from Table 28-4 include the following:
•	Naphthalene is the only pollutant sampled for at RIVA to fail screens.
•	Naphthalene was detected in all 58 valid PAH samples collected at RIVA.
•	Naphthalene failed greater than 96 percent of its screens, with 56 of 58 measured
detections of naphthalene failing screens.
28.4 Concentrations
This section presents various concentration averages used to characterize air toxics
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
each monitoring site.
•	Annual concentration averages are presented graphically for each site to illustrate
how the site's concentrations compare to the program-level averages, as presented in
Section 4.1.
•	Concentration averages and other statistical metrics are presented from previous years
of sampling in order to characterize concentration trends at each site.
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at RIVA are provided in Appendices M and O.
28.4.1 2013 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
28-11

-------
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 28-5, 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 28-5. 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
Naphthalene
58/58
98.06
± 42.46
75.73
±23.70
90.85
± 19.71
81.95
± 24.09
86.87
± 13.95
Observations for RIVA from Table 28-5 include the following:
•	Concentrations of naphthalene measured at RIVA range from 18.0 ng/m3 to
354 ng/m3.
•	The first quarter average concentration of naphthalene exhibits the most variability, as
indicated by the confidence interval. The maximum concentration was measured at
RIVA on January 10, 2013 (354 ng/m3). Four additional naphthalene concentrations
greater than 100 ng/m3 were also measured during the first quarter of 2013. However,
the 16 concentrations of naphthalene greater than 100 ng/m3 were measured at RIVA
on sample days spread across the year: five were measured during the first quarter,
three were measured during the second quarter, three were measured during the third
quarter, and five were measured during the fourth quarter.
•	Compared to other NMP sites sampling PAHs, RIVA has the eighth highest annual
average concentration of naphthalene, as shown in Table 4-11 of Section 4.
28-12

-------
28.4.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, a box plot was created for the pollutant shaded in
gray in Table 28-4 for RIVA. Figure 28-4 overlays the site's minimum, annual average, and
maximum concentrations onto the program-level minimum, first quartile, median, average, third
quartile, and maximum concentrations, as described in Section 3.4.3.1.
Figure 28-4. Program vs. Site-Specific Average Naphthalene Concentration
-
400	500
Concentration {ng/m3)
Program: 1st Quartile
¦
2nd Quartile 3rd Quartile
~ ~
4th Quartile
~
Average
1
Site: Site Average
o
Site Concentration Range


Observations from Figure 28-4 include the following:
• Figure 28-4 shows that the annual average concentration of naphthalene for RIVA
is just greater than the program-level average concentration (75.26 ng/m3). The
maximum naphthalene concentration measured at RIVA is roughly half the
program-level maximum concentration. There were no non-detects of naphthalene
measured at RIVA or across the program.
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 3.4.3.2.
RIVA began sampling PAHs under the NMP in October 2008. Thus, Figure 28-5 presents the
1-year statistical metrics for the pollutant of interest for RIVA. The statistical metrics presented
for assessing trends include the substitution of zeros for non-detects. If sampling began mid-year,
a minimum of 6 months of sampling is required for inclusion in the trends analysis; in these
cases, a 1-year average concentration is not provided, although the range and percentiles are still
presented.
28-13

-------
Figure 28-5. Yearly Statistical Metrics for Naphthalene Concentrations Measured at RIVA
~ 300
2011
Year
0 5th Percentile
- Minimum
0 95th Percentile
Observations from Figure 28-5 for naphthalene measurements collected at PRRI include
the following:
•	RIVA began sampling PAHs under the NMP in October 2008. Because less than
6 months of data are available for 2008, Figure 28-5 begins with 2009.
•	The three naphthalene concentrations greater than 400 ng/m3 were measured at RIVA
during the fall of 2009. The next highest concentration was measured in 2013
(354 ng/m3).
•	Most of the statistical parameters exhibit a decreasing trend through 2011, with the
most significant change occurring between 2010 and 2011. All of the statistical
parameters are at a minimum in 2011 except the 95th percentile, which is just greater
than the 95th percentile for 2013.
•	With the exception of the maximum concentration, the statistical parameters
calculated for 2013 are similar to those calculated for 2011.
28-14

-------
28.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at the RIVA monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4 for
definitions and explanations regarding the various toxicity factors, time frames, and calculations
associated with these risk-based screenings.
28.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for RIVA and where annual average concentrations could
be calculated, risk was examined by calculating cancer risk and noncancer hazard
approximations. These approximations can be used as risk estimates for cancer and noncancer
effects attributable to the pollutants of interest. Although the use of these approximations is
limited, they may help identify where policy-makers want to shift their air monitoring priorities.
Refer to Section 3.4.3.3 for an explanation of how cancer risk and noncancer hazard
approximations are calculated and what limitations are associated with them. Annual averages,
cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard approximations are
presented in Table 28-6, where applicable. Cancer risk approximations are presented as
probabilities while the noncancer hazard approximations are ratios and thus, unitless values.
Table 28-6. Risk Approximations for the Virginia Monitoring Site
Pollutant
Cancer
URE
(Ug/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections vs.
# of Samples
Annual
Average
(ng/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
East Highland Park, Virginia - RIVA
Naphthalene
0.000034
0.003
58/58
86.87
± 13.95
2.95
0.03
Observations for RIVA from Table 28-6 include the following:
•	The annual average concentration of naphthalene for RIVA is86.87± 13.95 ng/m3.
•	The cancer risk approximation for naphthalene based on RIVA's annual average
concentration is 2.95 in-a-million.
•	The noncancer hazard approximation for naphthalene is considerably less than 1.0
(0.03), indicating that no adverse noncancer health effects are expected from this
individual pollutant.
28-15

-------
28.5.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 28-7 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
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 3.4.3.4. Lastly,
Table 28-7 provides the pollutants with the highest cancer risk approximations (in-a-million) for
RIVA, 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 average concentrations to be
calculated. A more in-depth discussion of this analysis is provided in Section 3.4.3.4. Similar to
the cancer risk and noncancer hazard approximations provided in Section 28.5.1, this analysis
may help policy-makers prioritize their air monitoring activities.
28-16

-------
Table 28-7. 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)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
East Highland Park, Virginia (Henrico County) - RIVA
Benzene
102.27
Formaldehyde
1.12E-03
Naphthalene
2.95
Formaldehyde
86.37
Benzene
7.98E-04

Acetaldehyde
50.16
1,3-Butadiene
5.48E-04
Ethylbenzene
48.29
Naphthalene
2.84E-04
1.3 -Butadiene
18.27
POM, Group 2b
2.22E-04
Tetrachloroethylene
17.17
POM, Group 2d
1.26E-04
Naphthalene
8.35
Ethylbenzene
1.21E-04
POM, Group 2b
2.52
Acetaldehyde
1.10E-04
POM, Group 2d
1.43
POM, Group 5a
8.51E-05
Trichloroethylene
0.85
Arsenic, PM
6.88E-05

-------
Table 28-8. 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)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
East Highland Park, Virginia (Henrico County) - RIVA
Toluene
542.35
Acrolein
276,867.54
Naphthalene
0.03
Hexane
196.44
1.3 -Butadiene
9,132.57

Xylenes
193.08
Formaldehyde
8,812.86
Methanol
181.20
Acetaldehyde
5,572.79
Benzene
102.27
Benzene
3,408.91
Formaldehyde
86.37
Naphthalene
2,783.60
Ethylene glycol
62.63
Xylenes
1,930.83
Acetaldehyde
50.16
Arsenic, PM
1,067.34
Ethylbenzene
48.29
Lead, PM
808.19
Methyl isobutyl ketone
24.42
Propionaldehyde
556.24

-------
Observations from Table 28-7 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, the only pollutant of interest for RIVA, has the seventh highest
emissions and the fourth highest toxicity-weighted emissions for Henrico County.
•	POM, Group 2b is the eighth highest emitted "pollutant" in Henrico County and ranks
fifth for toxicity-weighted emissions. POM, Group 2b includes several PAHs sampled
for at RIVA, including fluorene, peryline, and acenaphthene. POM, Group 2d also
appears on both emissions-based lists for Henrico County and includes anthracene,
phenanthrene, and pyrene. POM, Group 5a includes benzo(a)pyrene and ranked ninth
for toxicity-weighted emissions but is not among the highest emitted. None of the
PAHs sampled for at RIVA included in POM, Groups 2b, 2d, or 5a failed screens.
Observations from Table 28-8 include the following:
•	Toluene, hexane, and xylenes are the highest emitted pollutants with noncancer RfCs
in Henrico County.
•	The pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) are acrolein, 1,3-butadiene, and formaldehyde.
•	Four of the highest emitted pollutants in Henrico County also have the highest
toxicity-weighted emissions.
•	Naphthalene has the sixth highest toxicity-weighted emissions for Henrico County
but is not among the highest emitted pollutants with a noncancer toxicity factor in
Henrico County.
28.6 Summary of the 2013 Monitoring Data for RIVA
Results from several of the data treatments described in this section include the
following:
~~~ Naphthalene was the only pollutant sampledfor at RIVA whose concentrations failed
screens, making naphthalene RIVA's only pollutant of interest.
~~~ RIVA has the eighth highest annual average concentration of naphthalene among
NMP sites sampling PAHs.
28-19

-------
29.0	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.
29.1	Site Characterization
This section characterizes the Washington monitoring site by providing geographical and
physical information about the location of the site and the surrounding area. This information is
provided to give the reader insight regarding factors that may influence the air quality near the
site and assist in the interpretation of the ambient monitoring measurements.
The NATTS site in Washington is located in Seattle. Figure 29-1 is a 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, version 2. 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
boundary are still visible on the map for reference, but have been grayed out in order to
emphasize emissions sources within the boundary. Table 29-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates.
29-1

-------
Figure 29-1. Seattle, Washington (SEWA) Monitoring Site

-------
Figure 29-2. NET Point Sources Located Within 10 Miles of SEWA
iar«om* n^sstrw i»*3otrw	uwmnw tjsnsww iw-iotrw
MlMf Mflpaif
County '
Samrnarniiflfl j t
County
Legend
ZZTUmv	127^9trw	IZT20TTW	tZ2-15TTW	1Z2*1(7irW	«22T(*"W	1ZTOTTW
Note Due (o Polity den&ity and collocation. Gft* total fawbea
Ooola,oe may not rsprasaM at faalltws «v*hin tha area cA Interval
~
SEWA NATTS site
10 mile radius		 County boundary
Source Category Group (No. of Facilities)
Aerospace'Aircraft Manufacturing Facility (2)
T AltportfAirllne/Airport Support Operations (27)
TV AutomoWe.Tiuck Manufacturing Facility (3)
Bnck Structural Clay, or Clay Ceramics Plant (2)
5	Bui* Terminals/BulV Plants (2)
6	Electrical Equipment Manufacturing Facility (t)
F Food Processing/Agncultuw FaciMy (2)
Glass Plant (t)
O Institutional (school, hospital, prison etc > (t)
A	Metal Coating Engraving and Allied Services to Manufacturers (1)
©	Metals Processing/Fabrication Facility (4)
?	Miscellaneous Commercial/Industrial Facility (1)
Q	Pawt and Coating Manufactunng Facility (1)
7	Portland Cement Manufactunng (1)
X	Rail YanfRail Ljne Operations (2)
A	Ship/Boat Manufacturing or Repair Facility (1)
V	Steel Mill (1)
»	Wastewater Treatment Facility (1)
29-3

-------
Table 29-1. Geographical Information for the Washington Monitoring Site
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Additional Ambient Monitoring Information1
SEWA
53-033-0080
Seattle
King
Seattle-Tacoma-
Bellevue, WA
47.568236,
-122.308628
Residential
Urban/City
Center
Haze, CO, S02, NOx, NOy, NO, N02, 03,
Meteorological parameters, PM Coarse, PMio,
PM25, PM2.5 Speciation, IMPROVE Speciation.
1 Data for additional pollutants are reported to AQS for SEWA (EPA, 2014b); however, these data are not generated by ERG and are therefore not included in this report.
BOLD ITALICS = EPA-designated NATTS Site
to
VO

-------
The SEWA monitoring site is located in Seattle, at the southeast corner of the Beacon
Hill Reservoir. With the north reservoir decommissioned and the south reservoir covered, the
entire area is part of Jefferson Park (Seattle, 2015). The reservoir and park are separated from the
Jefferson Park Golf Course to the east by Beacon Avenue, as shown in Figure 29-1. A middle
school and a hospital can be seen to the south of the site in the bottom-center portion of
Figure 29-1. The site is surrounded by residential neighborhoods to the west, north, and east.
Interstate-5, which runs north-south through Seattle, is less than 1 mile to the west of SEWA and
intersects with 1-90 a couple of miles to the north of the site. The area to the west of 1-5 is highly
industrial while the area to the east is primarily residential. Although the emissions sources
within 10 miles of the site are involved in a variety of industries, the airport source category,
which includes airports and related operations as well as small runways and heliports, such as
those associated with hospitals or television stations, has the greatest number of sources. The
closest point sources to SEWA are a metals processing and fabrication facility and a food
processing facility, as shown in Figure 29-2.
Table 29-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Washington monitoring site. Table 29-2 includes both county-
level population and vehicle registration information. Table 29-2 also contains traffic volume
information for SEWA as well as the location for which the traffic volume was obtained.
Additionally, Table 29-2 presents the county-level daily VMT for King County.
Table 29-2. Population, Motor Vehicle, and Traffic Information for the Washington
Monitoring Site


Estimated
County-level
Annual
Intersection
County-


County
Vehicle
Average
Used for
level Daily
Site
County
Population1
Registration2
Daily Traffic3
Traffic Data
VMT4
SEWA
King
2,044,449
1,791,383
176,000
1-5 S at Spokane St Viaduct
23,266,320
bounty-level population estimate reflects 2013 data (Census Bureau, 2014)
2County-level vehicle registration reflects 2013 data (WS DOL, 2013)
3AADT reflects 2013 data (WS DOT, 2013)
4County-level VMT reflects 2013 data (WS DOT, 2013)
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 29-2 include the following:
• King County has the sixth highest county-level population among counties with NMP
sites.
29-5

-------
•	King County has the fifth highest county-level vehicle registration among counties
with NMP sites.
•	The traffic volume experienced near SEWA is the fifth highest compared to other
NMP monitoring sites. The traffic estimate provided is for 1-5 at the Spokane Street
Viaduct.
•	The daily VMT for King County is in the top third compared to other counties with
NMP sites.
29.2 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.
29.2.1	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 south-southwest for much of the year (Wood, 2004; WRCC 2014).
29.2.2	Meteorological Summary
Hourly meteorological data for 2013 were retrieved from NCDC for the weather station
closest to the Washington monitoring site (NCDC, 2013), as described in Section 3.4.2. 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 29-3. These data were used to determine
how meteorological conditions on sample days vary from conditions experienced throughout the
year.
29-6

-------
Table 29-3. Average Meteorological Conditions near the Washington Monitoring Site
Closest









Weather
Distance

Average

Average
Average
Average
Average
Average
Station
and

Maximum
Average
Dew Point
Wet Bulb
Relative
Sea Level
Scalar Wind
(WBAN and
Direction
Average
Temperature
Temperature
Temperature
Temperature
Humidity
Pressure
Speed
Coordinates)
from Site
Type1
(°F)
(°F)
(°F)
(°F)
(%)
(mb)
(kt)
Seattle, Washington - SEWA
Boeing Field/
King County
Intl Airport
2.7
miles
Sample
Davs
(65)
61.1
±3.3
54.4
±2.7
43.7
±2.2
49.0
±2.2
69.8
±3.0
1019.5
± 1.5
3.7
±0.5









24234
(47.53, -122.30)
172°
(S)
2013
61.0
+ 1.3
54.2
+ 1.1
43.4
±0.9
48.7
±0.9
69.3
± 1.2
1019.3
±0.7
4.0
±0.2
Sample day averages are shaded in orange to help differentiate the sample day averages from the full-year averages.
to
vo

-------
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 2013. 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 were representative of average weather conditions experienced
throughout the year. The average sea level pressure for SEWA in 2013 in Table 29-3 is the
highest among all NMP sites.
29.2.3 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 3.4.2. A wind rose shows the frequency of wind directions
using "petals" positioned around a 16-point compass, and uses different colors to represent wind
speeds.
Figure 29-3 presents a map showing the distance between the weather station and SEWA,
which may be useful for identifying topographical influences that can affect the meteorological
patterns experienced at this location. Figure 29-3 also presents three different wind roses for the
SEWA monitoring site. First, a historical wind rose representing 2003 to 2012 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 2013 is presented. Next, a
wind rose representing wind data for days on which samples were collected in 2013 is presented.
These can be used to identify the predominant wind speed and direction for 2013 and to
determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.
29-8

-------
Figure 29-3. Wind Roses for the Boeing Field/King County International Airport Weather
Station near SEWA
Location of SEWA and Weather Station
2003-2012 Historical Wind Rose

NORTH"---^

20%

16%

12%

8%,

» 4% : :
¦westF ¦ * \ M
vS ¦' ' ;EAS

SOUTH
WIND SPEED
(Knots)
~ =22
Esil 17-21
|| 11-17
~
7- 11
4-7
H 2-4
Calms: 24.47%
2013 Wind Rose
NORTH"'-.
WIN C SPEED
(Knots)
17 - 21
11 - 17
SOUTH
Calms: 30.20%
Sample Day Wind Rose
NORTH---.,
est;
WIND SPEED
(Knots)
~ >=22
11 -17
7- 11
4- 7
2- 4
Calms: 31.07%
~
29-9

-------
Observations from Figure 29-3 for SEWA include the following:
•	The Boeing Field/King County Airport weather station is located less than 3 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.
Winds from the northeast quadrant were rarely observed. Calm winds (those greater
than or equal to 2 knots) account for 24 percent of wind observations near SEWA.
•	The wind patterns shown on the 2013 wind rose are similar to the historical wind
patterns, although the percentage of calm winds is higher (30 percent).
•	The wind patterns shown on the sample day wind rose resemble the wind patterns in
2013, albeit with an even higher percentage of calm winds, with nearly one-third of
observations less than 2 knots. This indicates that conditions on sample days were
representative of those experienced over the entire year (and historically).
29.3 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 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 the site when reviewing the results of this analysis. PMio metals,
VOCs, PAHs, carbonyl compounds, and hexavalent chromium were sampled for at SEWA,
although hexavalent chromium sampling was discontinued in June.
Observations from Table 29-4 for SEWA include the following:
•	Fourteen pollutants failed at least one screen for SEWA; 56 percent of concentrations
for these 14 pollutants were greater than their associated risk screening value (or
failed screens).
•	Nine pollutants contributed to 95 percent of failed screens for SEWA and therefore
were identified as pollutants of interest for the site. These nine include two carbonyl
compounds, four VOCs, two PMio metals, and one PAH.
29-10

-------
• Benzene, carbon tetrachloride, and formaldehyde were detected in every valid VOC
or carbonyl compound sample collected at SEWA and failed 100 percent of screens.
1,2-Dichloroethane also failed 100 percent of screens, but was not detected in every
sample collected.
Table 29-4. Risk-Based Screening Results for the Washington Monitoring Site
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Seattle, Washington - SEWA
Arsenic (PMio)
0.00023
57
60
95.00
12.78
12.78
Benzene
0.13
57
57
100.00
12.78
25.56
Carbon Tetrachloride
0.17
57
57
100.00
12.78
38.34
Formaldehyde
0.077
57
57
100.00
12.78
51.12
1,2-Dichloroethane
0.038
51
51
100.00
11.43
62.56
Naphthalene
0.029
48
57
84.21
10.76
73.32
1.3 -Butadiene
0.03
47
49
95.92
10.54
83.86
Acetaldehyde
0.45
39
57
68.42
8.74
92.60
Nickel (PMio)
0.0021
14
60
23.33
3.14
95.74
Ethylbenzene
0.4
6
57
10.53
1.35
97.09
Cadmium (PMio)
0.00056
5
60
8.33
1.12
98.21
Acenaphthene
0.011
4
57
7.02
0.90
99.10
Fluorene
0.011
3
54
5.56
0.67
99.78
Manganese (PMio)
0.03
1
60
1.67
0.22
100.00
Total
446
793
56.24

29.4 Concentrations
This section presents various concentration averages used to characterize air toxics
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
each site.
•	Annual concentration averages are presented graphically for each site to illustrate
how the site's concentrations compare to the program-level averages, as presented in
Section 4.1.
•	Concentration averages and other statistical metrics are presented from previous years
of sampling in order to characterize concentration trends at each site.
29-11

-------
Each analysis is performed where the data meet the applicable criteria specified in the
appropriate sections discussed below. Site-specific statistical summaries for all pollutants
sampled for at SEWA are provided in Appendices J, L, M, N, and O.
29.4.1 2013 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 29-5, 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.
29-12

-------
Table 29-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
the Washington Monitoring Site

# of






Measured
1st
2nd
3rd
4th


Detections
Quarter
Quarter
Quarter
Quarter
Annual
Pollutant
vs. # of
Samples
Average
(Ug/m3)
Average
(Ug/m3)
Average
(Ug/m3)
Average
(Ug/m3)
Average
(Ug/m3)
Seattle, Washington - SEWA


0.46
0.78
0.85
0.76
0.72
Acetaldehyde
57/57
±0.13
±0.29
±0.14
±0.28
±0.11


0.80
0.43
0.35
0.70
0.56
Benzene
57/57
±0.15
±0.09
±0.06
±0.21
±0.08


0.09
0.04
0.05
0.12
0.07
1,3-Butadiene
49/57
±0.03
±0.02
±0.01
±0.06
±0.02


0.64
0.75
0.70
0.65
0.69
Carbon Tetrachloride
57/57
±0.03
±0.07
±0.05
±0.03
±0.03


0.09
0.09
0.03
0.07
0.07
1,2-Dichloroethane
51/57
±0.01
±0.01
±0.01
±0.01
±0.01


0.42
0.54
0.68
0.62
0.57
Formaldehyde
57/57
±0.13
±0.18
±0.08
±0.22
±0.08


0.96
0.61
0.65
0.92
0.79
Arsenic (PMi0)a
60/60
±0.35
±0.16
±0.16
±0.33
±0.13


71.91
61.84
68.08
81.30
70.39
Naphthalene3
57/57
± 26.73
±30.56
± 17.93
± 34.66
± 13.09


1.58
2.39
2.21
0.99
1.78
Nickel (PMi,;,)a
60/60
±0.69
± 1.49
±0.67
±0.27
±0.44
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
Observations from Table 29-5 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 (0.72 ±0.11 |ig/m3), carbon tetrachloride (0.69 ± 0.03 |ig/m3),
formaldehyde (0.57 ± 0.08 |ig/m3), and benzene (0.56 ± 0.08 |ig/m3). These are
similar to the annual average concentrations calculated for 2012.
•	Even though acetaldehyde has the highest annual average concentration among
SEWA's pollutants of interest, this annual average is one of the lowest among NMP
sites sampling carbonyl compounds. SEWA's annual average concentration of
formaldehyde is the lowest among all NMP sites. Few NMP sites have annual
average concentrations of these two pollutants less than 1 |ig/m3. Similar observations
were made in previous NMP reports.
•	Concentrations of acetaldehyde appear lowest during the first quarter of 2013. The
three lowest concentrations of acetaldehyde measured at SEWA in 2013 were
measured in February and March. No acetaldehyde concentrations greater than
1 |ig/m3 were measured at SEWA during the first quarter of 2013; between three and
six were measured in the remaining calendar quarters. A similar observation can be
made for formaldehyde, although the difference is less dramatic. The minimum
concentration of formaldehyde was measured at SEWA on the same day as the
minimum acetaldehyde concentration. The fewest formaldehyde concentrations
29-13

-------
greater than 0.5 |ig/m3 were measured during the first quarter of 2013 (three) while
the number ranges from six to 14 for the remaining calendar quarters.
Concentrations of benzene and 1,3-butadiene appear higher during the colder months
of the year based on the quarterly average concentrations shown in Table 29-5. A
review of the data shows that all five benzene concentrations greater than 1 |ig/m3
were measured in January, November, or December; further all but one of the 20
highest benzene concentrations measured at SEWA were measured in the first (12) or
fourth (7) quarters of 2013. Conversely, the nine lowest concentrations of benzene
were measured at SEWA during the second or third quarters of 2013. A similar
observation can be made for 1,3-butadiene. All but one of the 10 1,3-butadiene
concentrations greater than 0.1 |ig/m3 were measured in January, November, or
December and 14 of the 15 highest concentrations were measured in the first (6) or
fourth (8) quarters of 2013.
Concentrations of 1,2-dichloroethane measured during the third quarter appear
significantly lower than those measured during the rest of the year, based on the
quarterly average concentrations shown in Table 29-5. A review of the data shows
that all six non-detects of this pollutant were measured in either August or September.
In addition, none of the measurements from the third quarter are greater than the
median concentration for the year (0.07 |ig/m3).
The second and third quarter average concentrations of carbon tetrachloride are
greater than the first or four quarter averages. A review of the data shows that the
11 concentrations greater than 0.75 |ig/m3, including one greater than 1 |ig/m3, were
measured between May and August. SEWA is one of only five sites where carbon
tetrachloride concentrations greater than 1 |ig/m3 were measured.
The quarterly average concentrations of naphthalene show that measurements of this
pollutant are highly variable, as indicated by the confidence intervals. Concentrations
measured at SEWA range from 7.73 ng/m3 to 205 ng/m3. Naphthalene concentrations
greater than 100 ng/m3 were measured during each calendar quarter, with two or three
measured each quarter.
Arsenic concentrations measured at SEWA during the first and fourth quarters of
2013 appear higher than those measured during the other calendar quarters and
exhibit considerably more variability. A review of the data shows that arsenic
concentrations measured at SEWA range from 0.13 ng/m3 to 2.42 ng/m3. Of the
15 arsenic concentrations greater than or equal to 1 ng/m3 measured at SEWA, five
were measured during the first quarter, two each were measured during the second or
third quarters, and six were measured during the fourth quarter of 2013. The
maximum arsenic concentrations measured during the first and fourth quarters are
roughly twice the maximum concentrations measured during the second and third
quarters of 2013.
Concentrations of nickel measured at SEWA also appear highly variable, particularly
for the second quarter. Concentrations of nickel measured at SEWA in 2013 range
from 0.17 ng/m3 to 9.75 ng/m3; the maximum nickel concentration measured at this
29-14

-------
site is among the higher nickel concentrations for the program. SEWA is one of the
few NMP sites where multiple nickel concentrations greater than 5 ng/m3 were
measured (ASKY-M and TOOK are the others). Three of the four nickel
concentrations greater than 5 ng/m3 were measured at SEWA between April and
June.
Tables 4-9 through 4-12 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for SEWA from
those tables include the following:
•	SEWA only appears in Table 4-9 for VOCs once; SEWA has the third highest annual
average concentration of carbon tetrachloride among sites sampling VOCs. Note,
however, that with the exceptions of the sites with two highest annual average
concentrations of carbon tetrachloride, the annual averages shown in Table 4-9 span
only 0.03 |ig/m3. A similar observation was made in the 2012 NMP report.
•	SEWA does not appear in Table 4-10 for carbonyl compounds. As indicated above,
SEWA has one of the lowest annual average acetaldehyde concentration and the
lowest annual average concentration of formaldehyde among NMP sites sampling
these pollutants.
•	Table 4-11 for the PAHs shows that SEWA has the ninth highest annual average
concentration of acenaphthene. This pollutant failed screens for SEWA but was not
identified as a site-specific pollutant of interest.
•	As shown in Table 4-12, SEWA has the second highest annual average concentration
of nickel among all sites sampling metals (PMio), behind only ASKY-M. The same
observation was made in the 2012 NMP report. SEWA had the highest annual
average nickel concentration for the 2010 and 2011 NMP reports. SEWA also has the
fourth highest annual average concentration of arsenic among NMP sites sampling
PMio metals.
29.4.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for the pollutants shaded in
gray in Table 29-4 for SEWA. Figures 29-4 through 29-12 overlay the site's minimum, annual
average, and maximum concentrations onto the program-level minimum, first quartile, median,
average, third quartile, and maximum concentrations for each pollutant, as described in
Section 3.4.3.1.
29-15

-------
Figure 29-4. Program vs. Site-Specific Average Acetaldehyde Concentration
0
3
6 9
Concentration {[jg/m3)

12
15

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i


Site: Site Average
o
Site Concentration Range



Figure 29-5. Program vs. Site-Specific Average Arsenic (PMio) Concentration
0
12 3
4 5 6
Concentration {ng/m3)
7
8
9
10

Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
i



Site: Site Average
o
Site Concentration Range




Figure 29-6. Program vs. Site-Specific Average Benzene Concentration
H
Program Max Concentration = 43.5 ^ig/m3
4	6
Concentration {[jg/m3]
Program: 1st Qua rti le
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


29-16

-------
Figure 29-7. Program vs. Site-Specific Average 1,3-Butadiene Concentration




Program Max Concentration = 21.5 ^ig/m3

O
0


0	0.3	0.6	0.9	1.2	1.5
Concentration {[ig/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


Figure 29-8. Program vs. Site-Specific Average Carbon Tetrachloride Concentration
Program Max Concentration = 23.7 ^ig/m3
SEWA
0	0.25	0.5	0.75	1	1.25	1.5	1.75	2
Concentration {[ig/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


Figure 29-9. Program vs. Site-Specific Average 1,2-Dichloroethane Concentration


¦
Program Max Concentration = 111 ^ig/m3


0	0.2	0.4	0.6	0.8	1
Concentration {[ig/m3)
Program: IstQuartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Qua rti le
~
Average
1
Site: Site Average
o
Site Concentration Range


29-17

-------
Figure 29-10. Program vs. Site-Specific Average Formaldehyde Concentration
Bl
3
6
9 12 15
Concentration {[jg/m3)
18
21
Program:
1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site:
Site Average
o
Site Concentration Range


Figure 29-11. Program vs. Site-Specific Average Naphthalene Concentration
I
100
200
300 400 500
Concentration {ng/m3)
600
700
Program:
1st Quartile
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site:
Site Average
o
Site Concentration Range


Figure 29-12. Program vs. Site-Specific Average Nickel (PMio) Concentration
I I I
5
10 15
Concentration {ng/m3)

20
Progra m: 1st Qua rti 1 e
¦
2nd Qua rti 1 e 3rd Qua rti 1 e
~ ~
4th Quartile
~
Average
i
Site: Site Average
o
Site Concentration Range


Observations from Figures 29-4 through 29-12 include the following:
• Figure 29-4 shows that the entire range of acetaldehyde concentrations measured
at SEWA is less than the program-level third quartile. SEWA's annual average
acetaldehyde concentration is considerably less than the program-level average
concentration for acetaldehyde and less than the program-level first quartile (25th
percentile). This site has one of the lowest annual average concentrations of
acetaldehyde among NMP sites sampling carbonyl compounds.
29-18

-------
Figure 29-5 shows that SEWA's annual average arsenic (PMio) concentration
falls between the program-level average concentration and third quartile. 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 29-6 presents the box plot for benzene. Note that the program-level
maximum concentration (43.5 |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 12 |ig/m3. This figure shows that the annual average benzene
concentration for SEWA is less than the program-level average concentration and
similar to the program-level median concentration. The maximum benzene
concentration measured at SEWA is considerably less than the maximum benzene
concentration measured across the program.
Figure 29-7 is the box plot for 1,3-butadiene. Note that the program-level
maximum concentration (21.5 |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 1.5 |ig/m3. This figure shows that the annual average
1,3-butadiene concentration for SEWA is greater than the program-level median
concentration and less than the program-level third quartile. Figure 29-7 also
shows that the maximum 1,3-butadiene concentration measured at SEWA is about
one-fifth the scale of the box plot and considerably less than the maximum
concentration measured across the program. It should be noted however, that the
program-level average concentration is an order of magnitude less than the scale
of the box plot and is being driven by a few measurements at the upper end of the
concentration range.
Figure 29-8 is the box plot for carbon tetrachloride. Similar to other VOCs, the
program-level maximum concentration (23.7 |ig/m3) is not shown directly on the
box plot 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. This figure
shows that the range of carbon tetrachloride concentrations measured at SEWA
spans roughly 0.5 |ig/m3. The annual average concentration of carbon
tetrachloride for SEWA is just greater than the program-level average
concentration and similar to the program-level third quartile, although less than
0.05 |ig/m3 separates these three values.
Figure 29-9 is the box plot for 1,2-dichloroethane. Note that the program-level
maximum concentration (111 |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. All of the concentrations of
1,2-dichloroethane measured at SEWA are less than the program-level average
concentration. The program-level average concentration for this pollutant is being
driven by the highest concentrations measured at a few monitoring sites. The
29-19

-------
annual average concentration of 1,2-dichloroethane for SEWA falls between the
program-level first quartile and second quartile (or median concentration).
•	Figure 29-10 shows that the entire range of formaldehyde concentrations
measured at SEWA is less than the program-level first quartile, indicating that all
of SEWA's formaldehyde concentrations are less than the 25th percentile for the
entire program dataset. This is also true for SEWA's annual average
formaldehyde concentration. As previously discussed, SEWA has the lowest
annual average concentration of formaldehyde among NMP sites sampling
carbonyl compounds, both for 2013 and in previous years.
•	Figure 29-11 shows that the annual average concentration of naphthalene for
SEWA is just less than the program-level average concentration. The maximum
naphthalene concentration measured at SEWA is considerably less than the
program-level maximum concentration.
•	Figure 29-12 is the box plot for nickel. Although the maximum nickel
concentration measured at SEWA is less than half the maximum concentration
measured across the program, it is the fourth highest concentration program-wide.
This site has the second largest range of nickel concentrations measured among
NMP sites sampling PMio metals. SEWA's annual average concentration is
greater than the program-level average concentration and is the second highest
annual average among NMP sites sampling PMio metals.
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 3.4.3.2.
Sampling for PMio metals, VOCs, and carbonyl compounds under the NMP began in 2007 and
sampling for PAHs began in 2008. Thus, Figures 29-13 through 29-21 present the 1-year
statistical metrics for each of the pollutants of interest for SEWA. If sampling began mid-year, a
minimum of 6 months of sampling is required for inclusion in the trends analysis; in these cases,
a 1-year average concentration is not provided, although the range and percentiles are still
presented.
29-20

-------
Figure 29-13. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
SEW A
2007	2008	2009	2010	2011	2012	2013
Year
0 5th Percentile	— Minimum	— Median	- Maximum	0 95th Percentile	Average
Observations from Figure 29-13 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 is 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 (September 2012).
•	The 1-year average concentrations have a slight undulating pattern, with years with
slightly lower concentrations alternating with years with slightly higher
concentrations. The 1-year average acetaldehyde concentration changed little from
2012 to 2013 and is at a minimum for 2013 compared to the other years of sampling.
However, the range is rather small, with the 1-year average concentrations ranging
from 0.72 |ig/m3 (2013) to 0.98 |ig/m3 (2009).
•	The median concentration exhibits a steady increasing trend for the first 5 years of
sampling, ranging from 0.61 |ig/m3 (2007) to 0.85 |ig/m3 (2011). The median then
decreased from 2011 to 2012 (0.68 |ig/m3) and again for 2013 (0.59 |ig/m3), which is
also the minimum for the entire sampling period. These changes, though, are also
relatively small.
29-21

-------
Figure 29-14. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at SEWA
2010
Year
0 5th Percentile
0 95th Percentile
Observations from Figure 29-14 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 ng/m3). The third highest arsenic concentration was measured in 2013 on
January 22, 2013 (2.42 ng/m3). In total, 11 arsenic concentrations greater than
2 ng/m3 have been measured at SEWA, at least one in each year, although 2007 has
the most (three).
•	There have been no non-detects of arsenic measured at SEWA since the onset of
sampling, including 2008, where it appears the minimum concentration is zero. For
2008, the minimum concentration of arsenic is 0.011 ng/m3.
•	The 1-year average concentration fluctuated only slightly between 2007 and 2009,
exhibits a decrease for 2010, after which an increasing trend is shown though the end
of the sampling period, with the 1-year average concentration at a maximum for
2013. However, the 1-year average concentration has only varied by about 0.2 ng/m3,
ranging from 0.58 ng/m3 (2010) to 0.79 ng/m3 (2013). Confidence intervals indicate
that the changes are not statistically significant. The median concentration, also at a
maximum for 2013, has varied by even less, from 0.50 ng/m3 (2011) to 0.63 ng/m3
(2013).
29-22

-------
Figure 29-15. Yearly Statistical Metrics for Benzene Concentrations Measured at SEWA
2010
Year
0 5th Percentile
O 95th Percentile
Average
Observations from Figure 29-15 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.
•	Overall, benzene concentrations have a slight decreasing trend at SEWA, although
this decrease is interrupted by the 2 years that the highest benzene concentrations
were measured. If the maximum concentrations measured in 2009 and 2011 were
removed from the calculations, the 1-year average concentration would have a steady
decreasing trend for the entire period, albeit slight. The 1-year average concentration
of benzene has ranged from 0.56 |ig/m3 (2013) to 0.81 |ig/m3 (2009).
•	The concentrations of benzene appear to have a seasonal trend at SEWA. Of the
66 benzene concentrations greater than 1 |ig/m3, 55 have been measured during the
colder months of the year, either during the first quarter (23) or fourth quarter (32) of
any given year.
29-23

-------
Figure 29-16. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at SEWA
2007	2008	2009	2010	2011	2012	2013
Year
0 5th Percentile	— Minimum	— Median	- Maximum	0 95th Percentile	Average
Observations from Figure 29-16 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, 2011, and 2013, both the minimum and 5th percentile are zero, indicating
that at least 5 percent of the measurements were non-detects. Ten percent of the
measurements were non-detects for 2010, 15 percent were non-detects for 2011, and
14 percent were non-detects for 2013. The percentage of non-detects is 3 percent for
each of the remaining years.
•	The 1-year average concentration has changed little over the years of sampling,
ranging from 0.06 |ig/m3 (2008) to 0.09 |ig/m3 (2011). Interestingly, the year with the
greatest number of non-detects (2011) also has the greatest number of measurements
greater than 0.2 |ig/m3 (seven).
•	The 95th percentile is fairly static between 2011 and 2013, indicating that 95 percent
of the measurements are less than 0.21 |ig/m3 during each of these years.
29-24

-------
Figure 29-17. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
SEW A
2010
Year
O 5th Percentile	— Minimurr
O 95th Percentile
• Average
Observations from Figure 29-17 for carbon tetrachloride measurements collected at
SEWA include the following:
•	Nineteen concentrations of carbon tetrachloride greater than 1.0 |ig/m3 have been
measured since the onset of sampling in 2007. All but two of these were measured in
2008 and 2009, with one each in 2010 and 2013. The maximum carbon tetrachloride
concentration (1.22 |ig/m3) has been measured twice at SEWA, once in 2008 and
once in 2010.
•	All of the statistical metrics increased from 2007 to 2008. Eleven concentrations
measured in 2008 were greater than the maximum concentration measured in 2007. In
addition, the number of carbon tetrachloride concentrations greater than 0.75 |ig/m3
increased from 12 in 2007 to 43 for 2008.
•	Between 2008 and 2011, a steady decreasing trend in the concentrations is shown,
with the 1-year average concentration for 2011 returning to 2007 levels.
•	The range of measurements tightened for 2012 and is the smallest range of
measurements since the onset of sampling. Yet, both the 1-year average and median
concentrations exhibit increases. As the number of concentrations falling into the
0.7 |ig/m3 to 0.8 |ig/m3 range doubled in 2012, from 14 for 2011 to 28 in 2012, the
number of concentrations less than 0.6 |ig/m3 fell from 20 to seven during this time
frame.
29-25

-------
• Despite the increase in the maximum concentration and the 95th percentile for 2013,
both the 1-year average and median concentrations exhibit slight decreases, although
the difference is not statistically significant.
Figure 29-18. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured
at SEW A
.20
.16
.12
.08
.04
.00
2007
2008
2009
2010
2011
2012
2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
Observations from Figure 29-18 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. For 2012, the percentage of measured detections is 93 percent, a considerable
increase from 26 percent in 2011. This percentage leveled off a bit for 2013 (at
88 percent).
•	As the number of measured detections increased, particularly for 2012 (and 2013), the
median and 1-year average concentrations increased correspondingly. The median
concentration is greater than the 1-year average concentration for 2012 and 2013.
This is because there were still several non-detects (or zeros) factoring into the 1-year
average concentration for these years, which can pull an average down in a similar
manner that an outlier can drive an average upward, while the range of measured
detections is rather small.
29-26

-------
• Many of the statistical parameters are at a maximum for 2013, although the maximum
concentration for 2013 is the same as the maximum concentration measured in 2011.
Figure 29-19. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
SEW A
Maximum
Concentration for
2009 is 16.6 ng/m3
2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	— Maximum	O 95th Percentile ••••v—* Average
Observations from Figure 29-19 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. Only nine concentrations greater than
2 |ig/m3 have been measured since the onset of carbonyl compound sampling at
SEWA.
•	The box and whisker plot for formaldehyde bears resemblance to the acetaldehyde
plot. The 1-year average concentrations have an undulating pattern through 2012,
with a "down" year followed by an "up" year. Between 2007 and 2012, the 1-year
average formaldehyde concentrations have ranged from 0.53 |ig/m3 (2012) to
1.04 |ig/m3 (2009). The 1-year average formaldehyde concentration changed little
from 2012 to 2013.
29-27

-------
• The level of variability in the measurements decreased considerably from 2009 to
2010. The difference between the 1-year average and median concentrations is less
than 0.1 |ig/m3 for all years after 2009. Further, the difference between the 5th and
95th percentiles is less than 1 |ig/m3 for 2012 and 2013, as the majority of
measurements fell into a smaller range in these later years.
Figure 29-20. Yearly Statistical Metrics for Naphthalene Concentrations Measured at SEWA
50
0
1
2008
2009
2010
2011
2012
2013
Year
O 5th Percentile	— Minimum	— Median	— Maximum	O 95th Percentile ••••v—* Average
1A 1-year average is not presented because sampling under the NMP did not begin until March 2008.
Observations from Figure 29-20 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 concentration 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 naphthalene measurement greater than 250 ng/m3
measured at this site. Eight additional measurements greater than 200 ng/m3 have
been measured at SEWA and are spread across the years of sampling, except 2008.
•	Each of the statistical parameters shown exhibits an increase from 2008 to 2009.
Although the range of concentrations measured is similar for 2009 and 2010, the 95th
percentile decreased by almost half from one year to the next. The number of
naphthalene concentrations greater than 100 ng/m3 decreased by nearly two-thirds,
29-28

-------
from 19 in 2009 to only seven for 2010. With the exception of the median
concentration, each of the statistical parameters exhibits an increase for 2011, with
the 1-year average concentration nearly returning to 2009 levels. This is partially
driven by the maximum concentration measured this year.
• Little change in the 1-year average concentration is shown between 2011 and 2013.
Figure 29-21. Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at SEWA
2007	2008	2009	2010	2011	2012	2013
Year
O 5th Percentile	— Minimum	— Median	- Maximum	O 95th Percentile	Average
Observations from Figure 29-21 for nickel measurements collected at SEWA include the
following:
•	The two highest concentrations of nickel (14.3 ng/m3 and 11.8 ng/m3) were both
measured at SEWA in 2012, although concentrations greater than 10 ng/m3 were also
measured in 2009 (two) and 2010 (one).
•	The 1-year average concentration exhibits an increasing trend between 2007 and
2009, after which a decrease in shown for 2010, with little change for 2011. An
increase in the 1-year average concentration is shown for 2012, which is followed by
a decrease for 2013. Confidence intervals calculated on the dataset indicate that the
changes shown are not statistically significant as the concentrations measured are
fairly variable from year-to-year. The median concentrations exhibit a similar pattern.
29-29

-------
•	The difference between the 1-year average and median concentrations is greater than
0.50 ng/m3 for all years (and greater than 1.0 ng/m3 for 2012). This indicates that
there is considerable variability in the measurements of nickel.
29.5 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to air toxics at the Washington monitoring site. Refer to Sections 3.2, 3.4.3.3, and 3.4.3.4
for definitions and explanations regarding the various toxicity factors, time frames, and
calculations associated with these risk-based screenings.
29.5.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Washington site and where annual average
concentrations could be calculated, risk was examined by calculating cancer risk and noncancer
hazard approximations. These approximations can be used as risk estimates for cancer and
noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers may want to shift their
air monitoring priorities. Refer to Section 3.4.3.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 29-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Observations from Table 29-6 for SEWA include the following:
•	The pollutants with the highest annual average concentrations for SEWA are
acetaldehyde, carbon tetrachloride, formaldehyde, and benzene.
•	The pollutants with the highest cancer risk approximations are formaldehyde,
benzene, carbon tetrachloride, and arsenic. The cancer risk approximation for
formaldehyde for SEWA is the lowest among this pollutant's site-specific cancer risk
approximations.
•	The noncancer hazard approximations for SEWA are all considerably less than 1.0,
with the highest calculated for acetaldehyde (0.08), indicating that no adverse
noncancer health effects are expected from these individual pollutants.
29-30

-------
Table 29-6. Risk Approximations for the Washington Monitoring Site
Pollutant
Cancer
URE
(Ug/m3)1
Noncancer
RfC
(mg/m3)
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(Ug/m3)
Cancer
Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Seattle, Washington - SEWA
Acetaldehyde
0.0000022
0.009
57/57
0.72
±0.11
1.58
0.08
Benzene
0.0000078
0.03
57/57
0.56
±0.08
4.37
0.02
1.3 -Butadiene
0.00003
0.002
49/57
0.07
±0.02
2.23
0.04
Carbon Tetrachloride
0.000006
0.1
57/57
0.69
±0.03
4.12
0.01
1,2-Dichloroethane
0.000026
2.4
51/57
0.07
±0.01
1.82
<0.01
Formaldehyde
0.000013
0.0098
57/57
0.57
±0.08
7.37
0.06
Arsenic (PMi0)a
0.0043
0.000015
60/60
0.79
±0.13
3.38
0.05
Naphthalene3
0.000034
0.003
57/57
70.39
± 13.09
2.39
0.02
Nickel (PMio)a
0.00048
0.00009
60/60
1.78
±0.44
0.86
0.02
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of
viewing.
29.5.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 29-7 presents the 10 pollutants with the highest emissions from the 2011 NEI (version 2)
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 3.4.3.4. Lastly,
Table 29-7 provides the pollutants with the highest cancer risk approximations (in-a-million) for
SEW A, 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.
29-31

-------
Table 29-7. 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)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Seattle, Washington (King County) - SEWA
Benzene
930.96
Formaldehyde
1.01E-02
Formaldehyde
7.37
Formaldehyde
776.28
Benzene
7.26E-03
Benzene
4.37
Ethylbenzene
460.42
1,3-Butadiene
4.24E-03
Carbon Tetrachloride
4.12
Acetaldehyde
442.08
Naphthalene
2.98E-03
Arsenic
3.38
1.3 -Butadiene
141.43
POM, Group 2b
1.76E-03
Naphthalene
2.39
T etrachloroethy lene
95.67
POM, Group 2d
1.16E-03
1,3-Butadiene
2.23
Naphthalene
87.72
Ethylbenzene
1.15E-03
1,2-Dichloroethane
1.82
POM, Group 2b
19.97
POM, Group 5a
1.11E-03
Acetaldehyde
1.58
POM, Group 2d
13.20
Acetaldehyde
9.73E-04
Nickel
0.86
T richloroethy lene
11.73
Nickel, PM
5.36E-04


-------
Table 29-8. 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)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Seattle, Washington (King County) - SEWA
Toluene
4,999.08
Acrolein
2,910,205.08
Acetaldehyde
0.08
Xylenes
1,895.75
Formaldehyde
79,212.57
Formaldehyde
0.06
Hexane
1,472.55
1,3-Butadiene
70,716.54
Arsenic
0.05
Methanol
1,144.61
Cyanide Compounds, gas
63,595.60
1,3-Butadiene
0.04
Benzene
930.96
Acetaldehyde
49,120.41
Naphthalene
0.02
Formaldehyde
776.28
Benzene
31,032.01
Nickel
0.02
Ethylbenzene
460.42
Naphthalene
29,239.94
Benzene
0.02
Ethylene glycol
455.61
Xylenes
18,957.50
Carbon Tetrachloride
0.01
Acetaldehyde
442.08
Lead, PM
16,900.94
1,2-Dichloroethane
<0.01
Methyl isobutyl ketone
205.29
Nickel, PM
12,405.52


-------
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.3.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 29.5.1, this analysis may help policy-
makers prioritize their air monitoring activities.
Observations from Table 29-7 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 toxicity-weighted emissions (of the pollutants with
cancer UREs) for King County are formaldehyde, benzene, and 1,3-butadiene.
•	Eight of the highest emitted pollutants also have the highest toxicity-weighted
emissions for King County.
•	Formaldehyde 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 and arsenic, which rank third and fourth, respectively, for cancer
risk approximations for SEWA, do not appear on either emissions-based list. This is
also true for 1,2-dichloroethane. Nickel, which appears ninth among the pollutants of
interest for SEWA, has the 10th highest toxicity-weighted emissions for King
County, but is not among the highest emitted (of the pollutants with cancer UREs).
•	POM, Group 2b is the eighth highest emitted "pollutant" in King County and ranks
fifth for toxicity-weighted emissions. POM, Group 2b includes several PAHs sampled
for at SEWA including acenaphthene, fluorene, and perylene. Although
concentrations of acenaphthene and fluorene each failed screens, these pollutants
were not identified as pollutants of interest for SEWA. POM, Group 2d ranks ninth
for total emissions and sixth for its toxicity-weighted emissions. POM, Group 2d
includes several PAHs sampled for at SEWA including anthracene, phenanthrene,
and pyrene. POM, Group 5a also has the eighth highest toxicity-weighted emissions
for King County. Benzo(a)pyrene is part of POM, Group 5a. None of the PAHs
included in POM, Groups 2d and 5a failed screens for SEWA.
29-34

-------
Observations from Table 29-8 for SEWA include the following:
•	Toluene, xylenes, and hexane are the highest emitted pollutants with noncancer RfCs
in King County. The quantity of the emissions of these pollutants are considerably
higher than the emissions for the pollutants topping the emissions-based list in
Table 29-7.
•	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 29-8.
•	Naphthalene, 1,3-butadiene, and nickel are among SEWA's pollutants of interest that
also appear among those with the highest toxicity-weighted emissions, although none
of these appear among the highest emitted (of those with a noncancer RfC).
•	Arsenic, carbon tetrachloride, and 1,2-dichloroethane are pollutants of interest for
SEWA that appear on neither emissions-based list.
29.6 Summary of the 2013 Monitoring Data for SEWA
Results from several of the data treatments described in this section include the
following:
~~~ Fourteen 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 ng/m3.
~~~ The annual average concentration of nickel for SEWA is the second highest among
NMP sites sampling PMw metals. The annual average concentration of carbon
tetrachloride for SEWA is the third highest among NMP sites sampling VOCs.
Conversely, the annual average concentration of formaldehyde for SEWA is the
lowest among NMP sites sampling carbonyl compounds.
~~~ Concentrations of most of the pollutants of interest for SEWA have changed little in
recent years. Concentrations of carbon tetrachloride exhibited a decreasing trend
over much of the sampling period, although this trend did not continue into the later
years of sampling. The number of non-detects of 1,2-dichloroethane has been
decreasing considerably at SEWA, particularly in recent years.
29-35

-------
30.0	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.
30.1	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 monitoring site is located in Horicon, Wisconsin and is the relocated
Mayville NATTS site. The MIWI site is located in Milwaukee. Figure 30-1 is the composite
satellite image retrieved from ArcGIS Explorer showing the HOWI monitoring site and its
immediate surroundings. Figure 30-2 identifies nearby point source emissions locations for this
site by source category, as reported in the 2011 NEI for point sources, version 2. Note that only
sources within 10 miles of the site are included in the facility counts provided in Figure 30-2.
A 10-mile boundary was chosen to give the reader an indication of which emissions sources and
emissions source categories could potentially have a direct effect on the air quality at the
monitoring site. Further, this boundary provides both the proximity of emissions sources to the
monitoring site as well as the quantity of such sources within a given distance of the site. Sources
outside the 10-mile boundary are still visible on the map for reference, but have been grayed out
in order to emphasize emissions sources within the boundary. Figures 30-3 and 30-4 are the
composite satellite image and emissions sources map for MIWI. Table 30-1 provides
supplemental geographical information such as land use, location setting, and locational
coordinates for each site.
30-1

-------
Figure 30-1. Horicon, Wisconsin (HOWI) Monitoring Site
b3 Reservoir
source U5GS J
Source NASA NGA. USG
JOOI MicroioM

-------
Figure 30-2. NET Point Sources Located Within 10 Miles of HOWI
SBKKJY*	«r*M2V*	8T 40l»"W	ar*KTW	aT»irW
r
Fo«<3 du Lac
. Count}
Honcon Mansfi
Dodge
County
on	m vnrw	IT9WW	VStW
Note. Du« 10 faciity density and coltocaton th« total facilities
dispUyad awy rvit	all	*Hhm Bib a'en of t
Legend
HOWI NATTS site	10 mile radius		County boundary
Source Category Group (No. of Facilities)
t AirporVAirllrMWAirport Support Operations (3)
F Food Processing,'Agriculture Facility (1)
• Industrial Machinery or Equipment Plant (3)
» Landfill (1)
® Metals Processing/Fabrication Facility (5)
Mine/Quarry/Mineral Processing Facility (1)
c Printing/Publishing/Paper Product Manufacturing Facility (1)
30-3

-------
Figure 30-3. Milwaukee, Wisconsin (MIWI) Monitoring Site

-------
Figure 30-4. NEI Point Sources Located Within 10 Miles of MIWI

iaks
Micfiyart
Mrtftrautoe •
County |
Legend
Note Due «o (aolfty density and eoltoeallon, ih* total fac*«*»
displayed may not roprovont af tacdibas wrth-n the area of interest
~ MIWI UATMP site	10 mile radius	I | County boundary
Source Category Group (No. ol Facilities)
T
Airpcr1iA»l»rwiA»pon Support Operations {8)
O
Insttutonal (tchool hotoHai pr*cn ate ) |«j
»
Asp^ait ProducnonHor Mix Aspnart P*ant ;?>
Q
Laminar and i aa*w Products Facaty (3)
&
Auwmot>*o.Truck Manjfacturcg Facilty (.21
¦
Metal Can 8o» and Otha* Metal Container Manufacturing (3>
X
Battery Facility f 1)
<•>
Matals Prrxawir^'PaWcalton Facilr, <1*1
T
Brewena&fUts;»«ene&.Wr)enes (11
K
KAno-Quarry\tmarai Pfocesamg acuCyr ID
c
Oemlcal Mapufnctuma Facility (8)
•»
FAace larwojb Cornm*tcialiVidutY Clea^m® Facftty (31
C
Pa r.t and Coaling Manutacfe*>ng Facility 11)
e
Electrical Equoroerrt Manutacturog Facility i2)
R
Plastc Reaai. or RuoOcr Products Plarn (4;
i
Electrk *> G#-flatten via CottlbMtflon (1)
P
PmttngfltyfcHitwtfl/Papar Piodwrt Manufacture Family H1>
E
FlactrcoLatsr-c Placing PolisHIng Anodtzng and Cotanny (5>
6
Pup and Papor Plant < 1 >
F
Food P?oc*vwng
I
Foutdnos Iron and Stool <6S
i
Wastewater Traafmenf Focilcy it)
A
Foundries. Noa-^orrous (T|
4
waiar TreaSmanl FacMy (2>
+
IhctukMal MecNnary or EctuiptnerC Plant (9i
w
Vtoodwrorii, FurriRura Mil** or* A Wood Rresar^ng Facility (3)
30-5

-------
Table 30-1. Geographical Information for the Wisconsin Monitoring Sites
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Additional Ambient Monitoring Information1
HOWI
55-027-0001
Horicon
Dodge
Beaver Dam,
WI
43.466111,
-88.621111
Agricultural
Rural
SVOCs, PCBs, CO, S02, NOy, NO, VOCs, Caibonyl
compounds, O3, Meteorological parameters, PM10,
PM10 Metals, PM Coarse, PM2 5, PM2.5 Speciation,
IMPROVE Speciation.
MIWI
55-079-0026
Milwaukee
Milwaukee
Milwaukee-
Waukesha-West
Allis, WI
43.061258,
-87.913520
Commercial
Urban/City
Center
PAMS/NMOCs, S02, NOy, NO, N02, NOx, Carbonyl
compounds, Hg, O3, Meteorological parameters,
PM10, PM Coarse, PM2 5, PM25 Speciation,
IMPROVE Speciation.
1 Data for additional pollutants are reported to AQS for these sites (EPA, 2014b); however, these data are not generated by ERG and are therefore not included in this report.
BOLD ITALICS = EPA-designated NATTS Site
LtJ
o
On

-------
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 (37 miles to the southeast), Madison (41 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 less than one-half mile south
of the site. The HOWI monitoring site serves as a rural background site for the NATTS program.
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 right-hand side of Figure 30-1, is
the closest major roadway. The Rock River is located just west of the site and can be seen on the
left hand side of Figure 30-1. Figure 30-2 shows that two 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 30-3. Interstate-43 runs north-south less than
one-half mile west of the site. The Milwaukee River runs roughly north-south less than 1 mile
east of the site with the Milwaukee Bay and Lake Michigan approximately 2 miles farther east.
Figure 30-4 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 about 1 mile of MIWI are two electroplating, plating, polishing, anodizing, and coloring
facilities to the south and a pulp and paper plant and a leather and leather products facility to the
east.
30-7

-------
Table 30-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Wisconsin monitoring sites. Table 30-2 includes both county-level
population and vehicle registration information. Table 30-2 also contains traffic volume
information for HOWI and MIWI as well as the location for which each traffic volume was
obtained. Additionally, Table 30-2 presents the county-level daily VMT for Dodge County and
Milwaukee County.
Table 30-2. Population, Motor Vehicle, and Traffic Information for the Wisconsin
Monitoring Sites
Site
County
Estimated
County
Population1
County-level
Vehicle
Registration2
Annual
Average
Daily
Traffic3
Intersection
Used for
Traffic Data
County-level
Daily VMT4
HOWI
Dodge
88,344
99,078
5,100
Hwy 28 (Clason St), north of
Hway 33 in Horicon
2,568,234
MIWI
Milwaukee
956,023
641,582
12,400
N Dr Martin Luther King Jr
Dr, north of W North Ave
16,098,216
bounty-level population estimates reflect 2013 data (Census Bureau, 2014)
2County-level vehicle registrations reflect 2013 data (WI DOT, 2013)
3AADT reflects 2011 data for HOWI and 2013 data for MIWI (WI DOT, 2014a)
4County-level VMT reflects 2013 data (WI DOT, 2014b)
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 30-2 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 rankings 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 Highway 28 (Clason Street) near Highway 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.
30-8

-------
• The daily VMT for Milwaukee County is considerably higher than the VMT for
Dodge County. VMTs for these sites rank 19th and 38th, respectively, compared to
VMTs for other counties with NMP sites.
30.2 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.
30.2.1 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 less
than 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, 2015a and 2015b).
30-9

-------
30.2.2 Meteorological Summary
Hourly meteorological data for 2013 were retrieved from NCDC for the weather stations
closest to the Wisconsin monitoring sites (NCDC, 2013), as described in Section 3.4.2. 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 30-3. These data were used to determine how meteorological
conditions on sample days vary from conditions experienced throughout the year.
Table 30-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
and wind (average scalar wind speed) information for days samples were collected and for all of
2013. Average pressure information is not provided because sea level pressure observations were
not recorded at either weather station. Also included in Table 30-3 is the 95 percent confidence
interval for each parameter. As shown in Table 30-3, average meteorological conditions on
sample days near HOWI appear cooler than conditions experienced throughout the year.
However, sampling under the NMP at HOWI concluded in June, thereby missing the second half
of the year, which also includes the warmest days of the year. Average meteorological conditions
on sample days near MIWI are significantly colder and drier than conditions experienced
throughout the year. Sampling under the NMP at MIWI concluded in March; thus, the sample
day averages incorporate only meteorological data for the first 3 months of the year.
30-10

-------
Table 30-3. 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
5.0
miles
Sample
Davs
(30)
47.3
±7.8
39.8
±7.2
30.0
±6.9
35.7
±6.5
71.7
±6.2
NA
7.1
± 1.1
04898
(43.43, -88.70)
236°
(SW)
2013
52.9
+ 2.3
44.9
±2.2
34.1
±2.0
40.1
± 1.9
69.4
± 1.4
NA
7.0
±0.3
Milwaukee, Wisconsin - MIWI
Lawrence J.
Timmennan
Airport
94869
(43.11, -88.03)
6.8
miles
Sample
Davs
29.9
24.1
17.9
22.3
78.1

8.0
(12)
±5.6
±6.2
±7.7
±6.2
±7.1
NA
± 1.9
299°
(WNW)
2013
53.9
±2.2
45.9
±2.1
35.3
±2.0
41.0
± 1.9
69.7
± 1.2
NA
6.8
±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 either airport.

-------
30.2.3 Wind Rose Comparison
Hourly surface wind data from the weather stations nearest HOWI and MIWI were
uploaded into a wind rose software program to produce customized wind roses, as described in
Section 3.4.2. A wind rose shows the frequency of wind directions using "petals" positioned
around a 16-point compass, and uses different colors to represent wind speeds.
Figure 30-5 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 30-5 also presents three different wind roses for the
HOWI monitoring site. First, a historical wind rose representing 2003 to 2012 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 2013 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2013 is
presented. These can be used to identify the predominant wind speed and direction for 2013 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically. Figure 30-6 presents the distance map and three wind roses
for MIWI.
Observations from Figure 30-5 for HOWI include the following:
•	The Dodge County Airport weather station is located 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. Northerly winds are the only other winds to
individually account for at least 6 percent of the winds near HOWI. The strongest
wind speeds were associated with southerly to west-southwesterly winds. Calm winds
(those less than or equal to 2 knots) were observed for 14 percent of the hourly
measurements.
•	The wind patterns shown on the 2013 wind rose resemble the historical wind patterns,
although the percentage of calm winds was less than 11 percent in 2013.
•	The sample day wind rose shows that winds from the north, north-northeast, and east
accounted for the highest percentage of wind observations on sample days at HOWI,
which is different than the percentages shown on the historical and full-year wind
roses. However, the sample day wind rose includes wind observations for January
through June only, due to the completion of sampling at this site. A full year's worth
of wind observations may look differently.
30-12

-------
Figure 30-5. Wind Roses for the Dodge County Airport Weather Station near HOWI
Location of HOWI and Weather Station
2003-2012 Historical Wind Rose
EST
WIND SPEED
(Kn ots)
SOUTH
2013 Wind Rose
west:
VypctL:
(Knots)
SOUTH
Sample Day Wind Rose
NORTH --
EST
WIND SPEED
(Knots)
I I >-22
i I 17-21
11 -17
7- 11
4- 7
2- 4
Calms: 10.03%
~
30-13

-------
Figure 30-6. Wind Roses for the Lawrence J. Tinimernian Airport Weather Station near
MIWI
Location of MIWI and Weather Station
2006-2012 Historical Wind Rose
E- £
EST
WWD SPEED
(Kn ots}
SOUTH
2013 Wind Rose
Sample Day Wind Rose
WEST
WWD SPEED
[Kn ots >
SOUTH
¦ 2-4
Calms: 16.86%
WEST
WIN D S PE ED
(Kn ots)
SOUTH
Calms: 13.78%
30-14

-------
Observations from Figure 30-6 for MIWI include the following:
•	The Timmerman Airport weather station is located less than 7 miles west-northwest
of MIWI. Note that the airport location is considerably farther from Lake Michigan
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 (10 percent). Winds with a westerly component were observed
more frequently than winds with an easterly component. Calm winds were observed
for approximately 17 percent of the hourly measurements.
•	The wind patterns shown on the 2013 wind rose resemble the historical wind patterns,
indicating that wind conditions in 2013 were similar to those observed historically.
•	While westerly winds were still the most frequently observed wind direction on
sample days near MIWI, this is one of the few similarities the full-year and sample
day wind roses share. However, the sample day wind rose includes wind observations
for January, February, and March only, due to the completion of the monitoring effort
at this site. A full year's worth of wind observations may look differently.
30.3 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 30-4. 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 30-4. 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.
However, sampling was discontinued at MIWI in mid-March and at HOWI in June.
30-15

-------
Table 30-4. Risk-Based Screening Results for the Wisconsin Monitoring Sites
Pollutant
Screening
Value
Oig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Horicon, Wisconsin - HOWI
Hexavalent Chromium
0.000083
0
4
0.00
0.00
0.00
Total
0
4
0.00

Milwaukee, Wisconsin - MIWI
Hexavalent Chromium
0.000083
0
8
0.00
0.00
0.00
Total
0
8
0.00

Observations from Table 30-4 include the following:
•	Thirty hexavalent chromium samples were collected at HOWI prior to the
discontinuation of sampling. This pollutant was detected in only four of the samples
collected at HOWI.
•	Hexavalent chromium did not fail any screens during the 2013 monitoring effort at
HOWI. This was also true for 2011 and 2012.
•	Eleven hexavalent chromium samples were collected at MIWI prior to the
discontinuation of sampling. This pollutant was detected in eight of the samples
collected at MIWI.
•	Hexavalent chromium did not fail any screens during the 2013 portion of the
monitoring effort at MIWI.
30.4 Concentrations
This section typically presents various concentration averages used to characterize
pollution levels at the monitoring sites for each of the site-specific pollutants of interest.
However, because there were no failed screens for HOWI or MIWI, these sites have no
pollutants of interest based on the risk screening process. The short sampling duration at each
site also prevents annual average concentrations (and at least some quarterly average
concentrations) for hexavalent chromium to be calculated. In order to facilitate a review of the
data collected at these sites in 2013, a few statistical calculations are provided in the section that
follows. Site-specific statistical summaries for HOWI and MIWI are also provided in
Appendix O. The concentration comparison and trend analysis were not performed for these
sites.
30-16

-------
30.4.1 2013 Concentration Averages
Quarterly concentration averages were calculated for hexavalent chromium for the
Wisconsin sites, as described above, where applicable. 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, which includes all measured detections and substituted zeros for non-detects for
the entire year of sampling, could not be calculated for these sites as sampling at HOWI and
MIWI was discontinued mid-year. Quarterly average concentrations, where applicable, were
calculated for HOWI and MIWI and are presented in Table 30-5. 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 30-5. 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)
Horicon, Wisconsin - HOWI
Hexavalent Chromium
4/30
0.0013
±0.0018
0.0022
±0.0031
NA
NA
NA
Milwaukee, Wisconsin - MIWI
Hexavalent Chromium
8/11
0.0160
± 0.0093
NA
NA
NA
NA
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
Observations from Table 30-5 include the following:
•	Concentrations of hexavalent chromium measured at HOWI range from 0.0088 ng/m3
to 0.019 ng/m3 (as well as 26 non-detects).
•	For both quarterly average concentrations that could be calculated for HOWI, the
confidence interval is greater than the average itself. This is due to the relatively high
number of non-detects. Only two measured detections, and thus 13 zeros substituted
for non-detects, are incorporated into each quarterly average concentration for HOWI.
•	Concentrations of hexavalent chromium measured at MIWI range from 0.0033 ng/m3
to 0.0405 ng/m3 (as well as three non-detects). These measurements represent a
30-17

-------
decrease in the magnitude of the measurements compared to those measured during
the first 9 months of sampling in 2012.
• The first quarter average concentration for MWIW represents the entire range of
measurements collected at MIWI prior to the conclusion of sampling and thus, is both
a quarterly average and the average for the period of sampling in 2013.
30.5 Additional Risk-Based Screening Evaluations
In order to characterize risk at participating monitoring sites, additional risk-based
screening evaluations were conducted. Because there were no pollutants of interest identified for
the Wisconsin sites and because annual average concentrations could not be calculated for the
pollutant sampled for at these sites, cancer risk and noncancer hazard approximations, as
described in Section 3.4.3.3, were not calculated. The risk-based emissions assessment described
in Section 3.4.3.4 was still conducted, at least in part, as the emissions can be reviewed
independent of concentrations measured.
30.5.1 Risk-Based Emissions Assessment
This section presents an evaluation of county-level emissions based on cancer and
noncancer toxicity, respectively, and is intended to help policy-makers prioritize their air
monitoring activities. Table 30-6 presents the 10 pollutants with the highest emissions from the
2011 NEI (version 2) that have cancer toxicity factors. Table 30-6 also presents the 10 pollutants
with the highest toxicity-weighted emissions, based on the weighting schema described in
Section 3.4.3.4. The emissions and toxicity-weighted emissions are shown in descending order in
Table 30-6. Table 30-7 presents similar information, but is limited to those pollutants with
noncancer toxicity factors. Because not all pollutants have both cancer and noncancer toxicity
factors, the highest emitted pollutants in the cancer table may be different from the noncancer
table, although the actual quantity of emissions is the same. A more in-depth discussion of this
analysis is provided in Section 3.4.3.4.
30-18

-------
Table 30-6. 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)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Horicon, Wisconsin (Dodge County) - HOWI
Benzene
69.92
Formaldehyde
5.95E-04

Formaldehyde
45.75
Benzene
5.45E-04
Acetaldehyde
30.32
1,3-Butadiene
2.50E-04
Ethylbenzene
24.46
Naphthalene
1.78E-04
1.3 -Butadiene
8.32
POM, Group 2b
1.19E-04
Naphthalene
5.22
POM, Group 5a
8.55E-05
POM, Group 2b
1.35
POM, Group 2d
7.54E-05
POM, Group 2d
0.86
Acetaldehyde
6.67E-05
Trichloroethylene
0.85
Ethylbenzene
6.11E-05
POM, Group 6
0.10
Hexavalent Chromium
4.89E-05
Milwaukee, Wisconsin (Milwaukee County) - MIWI
Benzene
235.46
Hexavalent Chromium
2.90E-03

Formaldehyde
183.24
Formaldehyde
2.38E-03
Ethylbenzene
146.27
Benzene
1.84E-03
Acetaldehyde
111.84
1,3-Butadiene
1.13E-03
1,3-Butadiene
37.60
Nickel, PM
1.07E-03
Naphthalene
19.92
Naphthalene
6.77E-04
Dichloromethane
14.89
Arsenic, PM
5.12E-04
POM, Group 2b
4.76
POM, Group 2b
4.19E-04
POM, Group 2d
3.34
Ethylbenzene
3.66E-04
Trichloroethylene
2.86
POM, Group 2d
2.94E-04

-------
Table 30-7. 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)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Horicon, Wisconsin (Dodge County) - HOWI
Toluene
331.94
Acrolein
136,980.59

Xylenes
115.87
Formaldehyde
4,668.80
Hexane
74.18
1.3 -Butadiene
4,159.39
Benzene
69.92
Acetaldehyde
3,368.91
Methanol
50.62
Cyanide Compounds, gas
2,508.26
Formaldehyde
45.75
Benzene
2,330.83
Acetaldehyde
30.32
Naphthalene
1,741.51
Ethylbenzene
24.46
Xylenes
1,158.69
Ethylene glycol
21.53
Lead, PM
1,151.06
Hydrochloric acid
18.25
Hydrochloric acid
912.47
Milwaukee, Wisconsin (Milwaukee County) - MIWI
Toluene
1,012.99
Acrolein
621,919.36

Methanol
644.26
Nickel, PM
24,765.94
Xylenes
582.77
Hydrochloric acid
22,589.06
Hexane
577.45
1,3-Butadiene
18,798.74
Hydrochloric acid
451.78
Formaldehyde
18,698.32
Ethylene glycol
338.20
Acetaldehyde
12,426.36
Benzene
235.46
Manganese, PM
10,644.44
Formaldehyde
183.24
Hydrofluoric acid
9,117.92
Ethylbenzene
146.27
Arsenic, PM
7,940.23
Hydrofluoric acid
127.65
Benzene
7,848.51

-------
Observations from Table 30-6 include the following:
•	Benzene and formaldehyde are the highest emitted pollutants with cancer UREs in
both Dodge and Milwaukee Counties, although the emissions are higher in
Milwaukee County.
•	Formaldehyde is the pollutant with the highest toxicity-weighted emissions (of the
pollutants with cancer UREs), followed by benzene and 1,3-butadiene, for Dodge
County. Hexavalent chromium is the pollutant with the highest toxicity-weighted
emissions for Milwaukee County, followed by formaldehyde and benzene.
•	Eight of the highest emitted pollutants in Dodge County also have the highest
toxicity-weighted emissions. Seven 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 HOWI and MIWI,
has the highest toxicity-weighted emissions for Milwaukee County and the 10th
highest toxicity-weighted emissions for Dodge County, but is not among the highest
emitted for either county. Hexavalent chromium emissions in Dodge County rank
19th and in Milwaukee County rank 16th.
Observations from Table 30-7 include the following:
•	Toluene is the highest emitted pollutant with a noncancer RfC in both counties.
Xylenes and hexane follow toluene for Dodge County while methanol and xylenes
follow toluene for Milwaukee County. The emissions are considerably higher for
Milwaukee County than Dodge County.
•	The pollutant with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) for both counties is acrolein. Formaldehyde and 1,3-butadiene
follow acrolein for Dodge County while nickel and hydrochloric acid follow acrolein
for Milwaukee County.
•	Five of the highest emitted pollutants in Dodge County also have the highest toxicity-
weighted emissions. Four of the highest emitted pollutants in Milwaukee County also
have the highest toxicity-weighted emissions.
•	Hexavalent chromium does not appear among the pollutants with the highest
emissions or toxicity-weighted emissions for either county (among pollutants with
noncancer RfCs).
30-21

-------
30.6 Summary of the 2013 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 sampledfor at HOWI and MIWI,
although sampling was discontinued in March at MIWI and in June at HOWI
~~~ Hexavalent chromium was detected in only four of the 30 valid samples collected at
HOWI; hexavalent chromium was detected in eight of the 11 valid samples collected
at MIWI
~~~ None of the concentrations of hexavalent chromium measured at either site failed
screens.
30-22

-------
31.0	Data Quality
This section discusses the data quality of the ambient air measurements that constitute the
2013 NMP dataset. Each monitoring program under the NMP has its own specific Data Quality
Objectives (DQOs) which have been established and approved by EPA, consistent with the
specific data use needs of the individual monitoring program. Because the DQOs are program-
specific and the ERG laboratory is contracted to perform services for a subset of the overall
program participants, attainment of the individual program DQO(s) is not assessed in this report.
This section establishes data quality through the assessment of Data Quality Indicators (DQI) in
the form of MQOs specific to the program elements conducted by the ERG laboratory. MQOs
are designed to control and evaluate the various phases of the measurement process (sampling,
preparation, analysis, etc.) to ensure that the total measurement quality meets the overall program
data quality needs. In accordance with ERG's EPA-approved QAPP (ERG, 2013), the following
MQOs were assessed: completeness, precision, and accuracy (also called bias).
The quality assessments presented in this section show that the 2013 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 Coefficient of Variation (CV) for
all methods except ASTM D7614 for hexavalent chromium measurement. The analytical
precision for replicate analyses also met the precision MQO of 15 percent CV, with all method
less than 10 percent. Audit samples show that ERG is meeting the accuracy requirements of the
NATTS TAD (EPA, 2009b). These data quality indicators are discussed in further detail in the
following sections.
31.1	Completeness
Completeness refers to the number of valid samples successfully collected and analyzed
compared to the number of total samples scheduled to be collected and analyzed. The MQO for
completeness based on the EPA-approved QAPP specifies that at least 85 percent of samples
collected at a given monitoring site must be analyzed successfully to be considered sufficient for
data trends analysis (ERG, 2013). The MQO of 85 percent completeness was met by all but three
of 143 site-method combinations. Completeness statistics are presented and discussed more
thoroughly in Section 2.4.
31-1

-------
31.2 Method Precision
Precision defines the level of agreement between independent measurements performed
according to identical protocols and procedures. Method precision, which includes sampling and
analytical precision, quantifies random errors associated with collecting ambient air samples and
analyzing the samples in the laboratory. Method precision is evaluated by comparing
concentrations measured in duplicate or collocated samples. A duplicate sample is a sample
collected simultaneously with a primary sample through a common inlet probe such that the
same air parcel is being sampled. This simultaneous collection is typically achieved by teeing the
line from the sampler to two canisters (or other sampling media) and doubling the flow rate
applied to achieve integration over the 24-hour collection period. Collocated samples are
samples collected simultaneously through separate inlet probes, regardless of sampler set-up
(i.e., either two separate sampling systems or a single sampling system with multiple inlets).
Because the samples are not collected using a common inlet, the system is sampling potentially
different air parcels. The overarching difference between the two sample types is whether or not
the potential for non-homogeneity of the air parcel is being considered as part of the precision
calculation. Duplicate samples provide an indication of "intra-system" variability while
collocated samples provide an indication of "inter-system" variability, of which the non-
homogeneity of the air parcels sampled factors into the level of precision measured.
During the 2013 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/Phenols, metals, and hexavalent chromium, only collocated samples may
be collected due to limitations of the sampling media/instrumentation. For each method, these
duplicate or collocated samples were then analyzed in replicate. Replicate measurements are
repeated analyses performed on a duplicate or collocated pair of samples and are discussed in
greater detail in Section 31.3. 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, and is discussed further in Section 31.3.
31-2

-------
Method precision is calculated by comparing the concentrations of the
duplicates/collocates for each pollutant. The CV for duplicate or collocated samples was
calculated for each pollutant and each site. The following approach was employed to estimate
how closely the collected and analyzed samples agree with one another:
Coefficient of Variation (CV) provides a relative measure of data dispersion compared to
the mean. CV can be calculated two ways. The first, which expresses the CV as a ratio of
the standard deviation and the mean, is used for a single variable. The second, which is
provided below, is ideal when comparing paired values, such as a primary concentration
and a duplicate concentration. A coefficient of variation of 1 percent would indicate that
the analytical results could vary slightly due to sampling error, while a variation of
50 percent means that the results are more imprecise.
p = the primary result from a duplicate or collocated pair;
r = the secondary result from a duplicate or collocated pair;
n = the number of valid data pairs (the 2 adjusts for the fact that there are two
values with error).
Coefficients of variation were based on every pair of duplicate or collocated samples
collected during the program year. However, only measurements at or above the MDL were used
in these calculations. Thus, the number of pairs included in the calculations varies significantly
from pollutant to pollutant. 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:
Where:
• 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.
31-3

-------
Table 31-1 presents the 2013 NMP method precision for VOCs, SNMOCs, carbonyl
compounds, PAHs, metals, and hexavalent chromium, presented as the average CV (expressed as
a percentage). With one exception, each analytical method met the program MQO of 15 percent
CV for precision. Only hexavalent chromium results did not meet the MQO of 15 percent. 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 31-1 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. For some methods, such as TO-11A for carbonyl compounds,
the difference is small; for others, such as TO-15 for VOCs, the difference is relatively large.
Table 31-1. Method Precision by Analytical Method
Method/Pollutant
Group
Average
Coefficient of
Variation
(%)
Number of
Pairs Included
in the
Calculation
Total Number
of Pairs Without
the > MDL
exclusion
voc
(TO-15)
10.07
3,953
5,006
SNMOC
9.95
436
536
Carbonyl Compounds
(TO-11 A)
7.12
1,699
1,701
PAHs
(TO-13)
10.28
379
470
Metals Analysis
(Method IO-3.5/FEM)
13.37
1,585
2,072
Hexavalent Chromium
(ASTM D7614)
20.51
27
28
MQO
15.00 percent CV
Tables 31-2 through 31-7 present method precision for VOCs, SNMOCs, carbonyl
compounds, PAHs, metals, and hexavalent chromium, respectively, as the CV per pollutant per
site and the average CV per site, per pollutant, and per method. Also included in these tables is
the number of duplicate and/or collocated pairs included in the CV calculations. For methods
where duplicate or collocated samples are both possible, the type of sample collected at each site
is identified and the average CV based on sample type is also provided. CVs exceeding the
15 percent MQO are bolded in each table. The CVs that exceed the program MQO for precision
are often driven by relatively low concentrations, even though they are greater than the MDL, as
these may result in relatively large CVs.
31-4

-------
31.2.1 VOC Method Precision
Table 31-2 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 ranges from 0 percent (a few pollutants for several sites)
to 82.50 percent (p-dichlorobenzene for TMOK). The CV for/;-dichlorobenzene for TMOK is
based on a single pair of samples greater than the MDL. The number of sites for which a given
pollutant has a CV greater than 15 percent varies, from none (30 pollutants) to greater than 20
(two pollutants). Dichloromethane (20) and methyl isobutyl ketone (21) have the highest number
of sites with average CVs greater than 15 percent.
The pollutant-specific average CV, as shown in orange in Table 31-2, ranges from
0 percent (bromoform, 1,1,2,2-tetrachloroethane, and 1,1,1-trichloroethane) to 35.71 percent
(bromodichloromethane). For the three pollutants with an average CV of 0 percent, the precision
is based on a single pair of measurements greater than the MDL. For bromodichloromethane, the
precision is based on five pairs of measurements collected at a single site (NBIL). The site-
specific average CV, as shown in green in Table 31-2, ranges from 6.33 percent (ROIL) to
13.95 percent (TMOK). None of the sites have a site-specific average CV greater than 15
percent. The overall average method precision for VOCs is 10.07 percent. Note that the results
for acrolein, acetonitrile, acrylonitrile, and carbon disulfide were excluded from the precision
calculations due to the issues described in Section 3.2.
Sites at which duplicate samples were collected are highlighted in blue in Table 31-2
while sites at which collocated samples were collected are highlighted in purple. Collocated
VOC samples were collected at only three of the sites shown in Table 31-2 (BURVT, PXSS, and
TVKY); the remainder collected duplicate VOC samples. The average CV for sites that collected
duplicate samples was calculated and is shown in Table 31-2 in blue while the average CV for
sites collecting collocated samples is shown in purple. The average CV for both precision types
meets the MQO of 15 percent, with the variability associated with collocated samples (11.83
percent) slightly greater than the variability associated with duplicate samples (10.27 percent).
31-5

-------
Table 31-2. VOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant
Pollutant
ADOK
ANAK
BTUT
BURVT
CHNJ
CSNJ
DEMI
ELNJ
Acetylene
4.13
7.46
4.08
7.24
8.69
6.37
8.32
7.58
fcrt-Amyl Methyl Ether
—
—
—
—
—
—
—
—
Benzene
7.69
7.03
18.35
11.93
7.87
3.55
28.12
1.63
Bromochloromethane
—
—
—
—
—
—
—
—
Bromodichloromethane
-
-
-
-
-
-
-
-
Bromofonn
-
-
-
-
-
-
-
-
Bromomethane
-
5.66
5.24
15.65
16.13
20.67
4.88
19.37
1.3 -Butadiene
6.02
5.04
3.40
6.79
36.15
4.10
12.09
4.35
Carbon Tetrachloride
8.37
10.12
3.64
7.49
7.46
9.75
8.91
7.82
Chlorobenzene
-
-
-
-
-
-
-
-
Chloroethane
-
-
-
33.36
17.46
25.38
-
4.56
Chloroform
-
4.36
2.54
8.27
7.60
3.43
48.97
2.51
Chloromethane
0.66
2.24
2.89
7.34
8.08
4.01
6.04
5.14
Chloroprene
-
-
-
-
-
-
-
-
Dibromochloromethane
-
-
-
-
-
-
-
-
1,2-Dibromoethane
-
-
-
-
-
-
-
-
«7-Dichlorobenzene
-
-
-
-
-
-
-
-
o-Dichlorobenzene
-
-
-
-
-
-
-
-
p-Dichlorobenzene
-
2.24
-
-
-
-
-
-
Dichlorodifluoromethane
4.14
2.42
2.48
4.94
8.19
7.59
5.10
3.40
1,1 -Dichloroethane
—
—
—
—
—
—
—
—
1,2-Dichloroethane
7.62
2.23
1.81
12.69
12.14
10.09
7.62
6.16
1,1 -Dichloroethene
—
—
0.00
—
—
—
—
—
cis-1,2-Dichloroethylene
—
—
—
—
—
—
—
—
trans-1,2-Dichloroethylene
—
—
—
9.05
—
—
—
—
Dichloromethane
36.68
25.45
18.27
30.56
7.99
7.90
20.99
32.26
1,2-Dichloropropane
—
—
—
—
—
—
—
—
cis-1,3 -Dichloropropene
—
—
—
—
—
—
—
—
trans-1,3 -Dichloropropene
—
—
—
—
—
—
—
—
Dichlorotetrafluoroethane
6.56
4.46
3.85
6.03
10.14
2.73
4.77
6.43
Ethyl Aery late
—
—
—
—
—
—
—
—
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples.
BOLD ITALICS = EPA-designated NATTS Site
31-6

-------
Table 31-2. VOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
ADOK
ANAK
BTUT
BURVT
CHNJ
CSNJ
DEMI
ELNJ
Ethyl terf-Butyl Ether
—
—
5.25
—
41.49
33.71
—
4.87
Ethylbenzene
10.79
10.70
3.26
14.55
17.79
4.20
8.93
2.76
Hexachloro-1,3 -butadiene
—
—
—
—
—
—
—
—
Methyl Isobutyl Ketone
48.87
32.34
44.27
30.92
18.65
15.05
12.21
6.41
Methyl Methacrylate
—
—
—
—
—
12.53
—
11.85
Methyl tert-Butyl Ether
—
—
11.70
—
18.55
3.66
—
24.72
//-Octane
8.26
8.29
4.88
15.87
9.38
4.45
10.69
6.34
Propylene
38.75
17.69
29.64
19.17
13.14
7.19
7.15
4.62
Styrene
7.26
10.94
4.71
8.31
2.11
10.11
21.59
11.14
1.1,2,2-Tetrachloroethane
—
—
—
—
—
—
—
0.00
Tetrachloroethylene
—
11.88
—
5.12
11.59
7.46
15.01
2.99
Toluene
9.39
9.48
6.84
10.82
2.53
7.76
5.04
1.62
1,2,4-Trichlorobenzene
—
—
—
—
—
—
—
—
1.1.1 -Trichloroethane
—
—
—
—
—
—
—
0.00
1,1,2-Trichloroethane
—
—
—
—
—
—
—
—
Trichloroethylene
—
—
—
9.28
—
1.30
—
22.10
Trichlorofluoromethane
4.63
2.58
2.28
4.95
7.47
6.40
3.32
2.91
Trichlorotrifluoroethane
4.60
5.20
2.67
18.65
7.89
2.15
3.18
3.23
1,2,4-Trimethylbenzene
3.28
6.18
3.78
11.35
4.85
3.04
8.86
4.44
1,3,5-Trimethylbenzene
1.75
6.07
2.04
5.79
7.76
5.34
15.72
6.07
Vinyl chloride
—
—
—
—
—
—
—
—
m,p-Xylene
5.25
10.25
2.50
16.94
14.34
4.83
7.48
2.94
o-Xylene
7.88
10.30
6.45
17.15
19.81
4.00
7.40
3.57
Average by Site
11.08
8.82
7.57
12.97
12.79
8.23
11.77
7.22
# of Pairs Collected per Site
3
6
4
27
6
5
6
6
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples.
BOLD ITALICS = EPA-designated NATTS Site
31-7

-------
Table 31-2. VOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
GLKY
GPCO
KMMS
NBIL
NBNJ
OCOK
PXSS
ROIL
Acetylene
5.98
5.06
4.13
6.45
5.12
7.19
3.80
5.08
tert-Amyl Methyl Ether
—
—
—
—
—
—
—
—
Benzene
6.57
3.75
11.80
6.76
7.11
4.22
6.02
2.24
Bromochloromethane
—
—
—
—
—
—
—
—
Bromodichloromethane
—
—
—
35.71
—
—
—
—
Bromofonn
—
—
—
0.00
—
—
—
—
Bromomethane
12.00
14.81
—
5.66
4.00
37.22
5.00
8.43
1.3 -Butadiene
7.36
10.95
9.63
7.20
2.00
18.48
8.27
7.65
Carbon Tetrachloride
7.22
21.74
6.71
6.26
24.36
3.71
6.96
7.16
Chlorobenzene
—
—
—
—
—
—
—
—
Chloroethane
—
—
—
6.67
—
—
8.96
14.43
Chloroform
—
3.66
7.03
28.70
6.14
7.32
5.12
6.05
Chloromethane
7.36
2.60
4.50
3.04
3.56
19.10
3.07
3.39
Chloroprene
—
—
—
—
—
—
—
—
Dibromochloromethane
—
—
—
19.38
—
—
—
—
1,2-Dibromoethane
—
—
—
—
—
—
—
—
«7-Dichlorobenzene
—
—
—
—
—
—
—
—
o-Dichlorobenzene
—
—
—
—
—
—
—
—
p-Dichlorobenzene
—
—
—
—
—
—
5.08
—
Dichlorodifluoromethane
5.25
2.94
3.27
2.55
2.46
3.79
2.57
3.52
1,1 -Dichloroethane
..
..
..
..
..
..
..
..
1,2-Dichloroethane
5.86
5.92
7.26
0.00
4.24
3.26
4.16
11.33
1,1 -Dichloroethene
..
..
..
—
..
..
..
..
cis-1,2-Dichloroethylene
..
..
..
—
..
..
..
..
trans-1,2-Dichloroethylene
..
..
..
—
..
..
10.10
..
Dichloromethane
58.20
19.97
13.03
15.72
16.84
36.26
38.61
24.73
1,2-Dichloropropane
..
..
..
..
—
..
..
..
cis-1,3 -Dichloropropene
..
..
..
..
—
..
..
..
trans-1,3 -Dichloropropene
..
..
..
..
—
..
..
..
Dichlorotetrafluoroethane
8.04
5.54
3.39
2.82
3.24
3.33
4.80
5.71
Ethyl Aery late
—
—
—
—
—
—
—
—
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples.
BOLD ITALICS = EPA-designated NATTS Site
31-8

-------
Table 31-2. VOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
GLKY
GPCO
KMMS
NBIL
NBNJ
OCOK
PXSS
ROIL
Ethyl tert-Butyl Ether
—
7.58
—
6.50
4.04
—
—
—
Ethylbenzene
16.61
4.65
6.68
4.49
5.84
8.39
8.29
2.89
Hexachloro-1,3 -butadiene
—
—
—
—
—
—
—
—
Methyl Isobutyl Ketone
33.95
27.52
17.81
16.20
23.55
34.11
16.89
8.44
Methyl Methacrylate
—
2.95
—
3.39
—
—
13.83
—
Methyl tert-Butyl Ether
—
48.90
—
—
35.36
—
41.30
—
//-Octane
31.65
15.11
7.62
8.94
6.49
7.17
12.40
4.87
Propylene
18.57
21.87
13.98
7.43
11.18
24.16
12.66
4.19
Styrene
8.84
33.64
11.35
15.77
11.93
5.11
13.16
9.01
1,1,2,2-Tetrachloroethane
—
—
—
—
—
—
—
—
T etrachloroethy lene
—
3.53
2.82
22.57
4.01
4.29
8.33
4.04
Toluene
19.04
3.17
5.13
7.08
5.82
8.79
17.59
2.80
1,2,4-Trichlorobenzene
—
—
—
—
—
—
—
—
1.1.1 -Trichloroethane
—
—
—
—
—
—
—
—
1,1,2-Trichloroethane
—
—
—
—
—
—
—
—
T richloroethy lene
—
—
0.00
—
5.66
—
—
—
T richlorofluoro methane
2.26
3.20
3.03
4.20
2.39
3.84
1.87
7.06
T richlorotrifluoroethane
3.69
3.50
3.40
2.19
2.33
3.48
2.27
2.78
1,2,4-Trimethylbenzene
7.40
6.43
5.20
9.45
6.15
11.32
8.71
4.03
1,3,5 -T rime thy lbenzene
7.07
3.93
10.16
10.30
6.67
4.76
14.28
4.88
Vinyl chloride
—
—
5.24
4.04
—
—
—
—
m,p-Xylene
16.79
4.24
6.72
4.40
4.84
6.04
8.15
2.65
o-Xylene
11.40
4.48
6.11
5.64
4.79
8.78
8.84
0.96
Average by Site
13.69
10.80
7.04
9.02
8.15
11.42
10.38
6.33
# of Pairs Collected per Site
6
7
6
7
6
6
7
5
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples.
BOLD ITALICS = EPA-designated NATTS Site
31-9

-------
Table 31-2. VOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
S4MO
SEWA
SPIL
SSMS
TMOK
TOOK
TROK
TVKY
Acetylene
3.52
4.88
8.16
5.49
8.55
6.22
6.13
6.82
tert-Amyl Methyl Ether
—
—
—
—
—
—
—
—
Benzene
8.06
3.84
6.02
3.77
4.81
4.73
6.48
15.56
Bromochloromethane
—
..
..
..
..
..
..
..
Bromodichloromethane
—
..
..
..
..
..
..
..
Bromofonn
—
..
..
..
..
..
..
..
Bromomethane
5.44
..
16.86
26.44
6.80
7.45
20.40
15.85
1.3 -Butadiene
13.29
5.59
6.98
10.51
9.72
8.24
20.24
8.37
Carbon Tetrachloride
2.29
4.89
5.00
2.91
26.48
4.73
7.38
7.45
Chlorobenzene
—
—
—
—
—
—
—
—
Chloroethane
..
..
..
..
—
6.73
..
14.78
Chloroform
7.23
9.57
9.74
4.04
21.06
4.74
5.26
9.45
Chloromethane
2.74
4.38
3.71
4.50
6.48
17.30
11.51
7.20
Chloroprene
..
..
..
..
..
..
..
..
Dibromochloromethane
..
..
..
..
..
..
..
..
1,2-Dibromoethane
..
..
..
..
..
..
..
..
«7-Dichlorobenzene
..
..
..
..
..
..
..
..
o-Dichlorobenzene
—
—
—
—
—
—
—
—
p-Dichlorobenzene
5.45
..
..
..
82.50
..
..
0.00
Dichlorodifluoromethane
3.24
3.17
4.17
1.85
5.93
5.33
5.59
3.97
1,1 -Dichloroethane
..
..
..
..
..
..
..
4.57
1,2-Dichloroethane
8.75
11.36
7.51
6.10
0.00
3.50
5.94
8.78
1,1 -Dichloroethene
..
..
..
..
—
..
..
6.98
cis-1,2-Dichloroethylene
..
..
..
..
—
..
..
..
trans-1,2-Dichloroethylene
..
..
..
..
—
..
8.27
4.37
Dichloromethane
25.23
65.60
10.35
10.47
17.31
74.85
33.77
24.98
1,2-Dichloropropane
—
..
..
..
..
..
..
—
cis-1,3 -Dichloropropene
—
..
..
..
..
..
..
—
trans-1,3 -Dichloropropene
—
..
..
..
..
..
..
—
Dichlorotetrafluoroethane
7.30
13.38
3.99
3.28
10.77
4.51
5.62
6.41
Ethyl Aery late
—
—
—
—
—
—
—
—
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples.
BOLD ITALICS = EPA-designated NATTS Site
31-10

-------
Table 31-2. VOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
S4MO
SEWA
SPIL
SSMS
TMOK
TOOK
TROK
TVKY
Ethyl tert-Butyl Ether
—
—
13.69
—
—
—
—
—
Ethylbenzene
14.55
5.56
9.65
13.05
9.03
3.91
3.24
15.39
Hexachloro -1,3 -butadiene
—
—
—
—
—
—
—
—
Methyl Isobutyl Ketone
35.03
7.84
36.13
22.00
43.26
41.07
22.60
20.88
Methyl Methacrylate
—
—
—
—
—
6.90
—
—
Methyl tert-Butyl Ether
—
—
35.66
—
—
—
—
—
//-Octane
18.04
9.55
11.31
15.20
5.21
4.86
6.39
14.92
Propylene
16.23
8.90
6.28
12.44
19.42
10.24
10.99
12.74
Styrene
27.78
8.66
1.48
1.10
5.80
5.55
15.13
39.90
1,1,2,2-Tetrachloroethane
—
—
—
—
—
—
—
—
T etrachloroethy lene
5.04
6.39
5.22
3.82
13.34
7.26
4.64
7.78
Toluene
6.47
3.40
8.06
4.69
8.62
5.53
3.39
48.72
1,2,4-Trichlorobenzene
—
—
—
—
—
—
—
—
1,1,1 -Trichloroethane
—
—
—
—
—
—
—
—
1,1,2-Trichloroethane
—
—
—
—
—
—
—
2.89
T richloroethy lene
11.06
—
0.00
40.75
—
—
—
7.44
T richlorofluoro methane
3.33
3.74
2.32
1.88
6.39
5.33
5.23
16.88
T richlorotrifluoroethane
3.25
4.54
4.20
2.57
7.13
4.44
4.90
3.87
1,2,4-Trimethylbenzene
12.92
9.73
10.39
12.64
10.41
5.92
5.07
13.02
1,3,5 -T rime thy lbenzene
10.32
6.19
0.00
0.00
6.03
4.42
5.81
8.51
Vinyl chloride
—
—
—
0.00
—
—
—
4.80
m,p-Xylene
11.23
6.28
9.44
9.55
6.87
4.02
3.02
15.27
o-Xylene
11.98
5.42
10.09
11.05
6.91
4.22
5.44
17.50
Average by Site
10.76
9.26
9.13
8.85
13.95
10.08
9.30
12.38
# of Pairs Collected per Site
6
6
6
6
6
7
7
31
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples.
BOLD ITALICS = EPA-designated NATTS Site
31-11

-------
Table 31-2. VOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
YUOK
# of
Pairs
Average
by
Pollutant
Average
for
Duplicate
Pairs
Average
for
Collocated
Pairs
Acetylene
7.44
190
6.16
6.18
5.95
ten-Amy\ Methyl Ether
—
—
—
—
—
Benzene
5.27
191
7.73
7.26
11.17
Bromochloromethane
..
—
—
—
..
Bromodichloromethane
—
5
35.71
35.71
—
Bromofonn
..
1
0.00
0.00
..
Bromomethane
4.35
67
12.65
12.73
12.16
1.3 -Butadiene
3.32
155
9.43
9.65
7.81
Carbon Tetrachloride
2.26
190
8.44
8.60
7.30
Chlorobenzene
..
—
—
—
..
Chloroethane
..
16
14.70
12.54
19.03
Chloroform
..
119
9.67
10.00
7.61
Chloromethane
4.09
191
5.80
5.79
5.87
Chloroprene
..
—
—
—
..
Dibromochloromethane
..
3
19.38
19.38
..
1,2-Dibromoethane
—
—
—
—
—
«7-Dichlorobenzene
..
—
—
—
..
o-Dichlorobenzene
..
—
—
—
..
p-Dichlorobenzene
..
9
19.06
30.07
2.54
Dichlorodifluoromethane
5.12
191
4.12
4.16
3.83
1,1 -Dichloroethane
—
5
4.57
—
4.57
1,2-Dichloroethane
6.73
134
6.44
6.16
8.54
1,1 -Dichloroethene
..
3
3.49
0.00
6.98
cis-1,2-Dichloroethylene
—
—
—
—
—
trans-1,2-Dichloroethylene
..
7
7.95
8.27
7.84
Dichloromethane
43.59
183
28.38
27.98
31.38
1,2-Dichloropropane
..
—
—
—
..
cis-1,3 -Dichloropropene
..
—
—
—
..
trans-1,3 -Dichloropropene
—
—
—
—
—
Dichlorotetrafluoroethane
5.80
191
5.72
5.71
5.75
Ethyl Aery late
—
--
--
--
—
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV fortius method is
calculated from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting
collocated samples.
BOLD ITALICS = EPA-designated NATTS Site
31-12

-------
Table 31-2. VOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
YUOK
# of
Pairs
Average
by
Pollutant
Average
for
Duplicate
Pairs
Average
for
Collocated
Pairs
Ethyl tert-Butyl Ether
—
26
14.64
14.64
—
Ethylbenzene
9.23
179
8.58
8.01
12.74
Hexachloro-1,3 -butadiene
—
—
—
—
—
Methyl Isobutyl Ketone
19.23
127
25.41
25.75
22.90
Methyl Methacrylate
—
9
8.57
7.52
13.83
Methyl tert-Butyl Ether
—
23
27.48
25.51
41.30
//-Octane
9.43
171
10.29
9.73
14.40
Propylene
7.41
191
14.24
14.16
14.86
Styrene
—
96
12.10
10.91
20.46
1,1,2,2-Tetrachloroethane
—
1
0.00
0.00
—
T etrachloroethy lene
—
75
7.48
7.55
7.08
Toluene
1.06
191
8.51
6.17
25.71
1,2,4-Trichlorobenzene
—
—
—
—
—
1.1.1 -Trichloroethane
—
1
0.00
0.00
—
1,1,2-Trichloroethane
—
2
2.89
—
2.89
T richloroethy lene
—
11
10.84
11.55
8.36
T richlorofluoro methane
5.31
191
4.51
4.05
7.90
T richlorotrifluoroethane
2.99
191
4.36
3.83
8.26
1,2,4-Trimethylbenzene
8.93
158
7.74
7.29
11.03
1,3,5 -T rime thy lbenzene
0.00
83
6.16
5.70
9.53
Vinyl chloride
—
17
3.52
3.09
4.80
m,p-Xylene
6.70
182
7.63
6.83
13.45
o-Xylene
11.84
177
8.48
7.66
14.50
Average by Site
8.10
3,953
10.07
10.27
11.83
# of Pairs Collected per Site
3
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV fortius method is
calculated from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting
collocated samples.
BOLD ITALICS = EPA-designated NATTS Site
31-13

-------
31.2.2 SNMOC Method Precision
The SNMOC method precision for duplicate samples is presented in Table 31-3 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 individual method precision results from duplicate
samples exhibit low- to mid-level variability among the pollutants and sites, ranging from a CV
of 0.24 percent (1,3-butadiene for BTUT) to 45.68 percent (isoprene for NBIL). The CVs for
29 pollutants are less than 15 percent for both sites; conversely, there are only four pollutants
listed where the CV is greater than 15 percent for both sites: 2-methylhexane, 1-pentene,
2,3,4-trimethylpentane, and sum of unknowns.
The pollutant-specific average CV, as shown in orange in Table 31-3, ranges from
0.24 percent (1,3-butadiene) to 35.18 percent (2-methylhexane). The site-specific average CV, as
shown in green in Table 31-3, ranges from 9.94 percent (BTUT) to 10.54 percent (NBIL); these
are the only sites at which duplicate SNMOC samples were collected. No collocated SNMOC
samples were collected during the 2013 program year. The overall average method precision for
SNMOCs is 9.95 percent. Note that the results for TNMOC were not included in the precision
calculations.
Table 31-3. SNMOC Method Precision: Coefficient of Variation
Based on Duplicate Samples by Site and Pollutant
Pollutant
BTUT
NBIL
# of
Pairs
Average
by
Pollutant
Acetylene
5.31
7.84
11
6.57
Benzene
18.74
6.35
8
12.55
1.3 -Butadiene
0.24
—
1
0.24
//-Butane
0.37
4.06
11
2.22
1-Butene
—
—
—
—
67.Y-2-Butcnc
3.16
—
3
3.16
;ra«.v-2-Butcne
18.22
—
4
18.22
Cyclohexane
4.86
5.80
10
5.33
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV fortius
method is calculated from the pollutant-specific averages and is provided in the final column of
the table.
BOLD ITALICS = EPA-designated NATTS Site
31-14

-------
Table 31-3. SNMOC Method Precision: Coefficient of Variation
Based on Duplicate Samples by Site and Pollutant (Continued)
Pollutant
BTUT
NBIL
# of
Pairs
Average
by
Pollutant
Cyclopentane
1.65
16.28
6
8.97
Cyclopentene
—
—
—
—
w-Decane
4.04
11.88
7
7.96
1-Decene
—
—
—
—
w;-Diethylbenzene
—
—
—
—
p-Diethylbenzene
—
—
—
—
2,2-Dimethylbutane
2.92
17.41
8
10.17
2,3 -Dimethylbutane
2.54
4.35
9
3.44
2,3 -Dimethylpentane
3.20
5.16
7
4.18
2,4-Dimethylpentane
4.67
1.97
3
3.32
n-Dodecane
13.95
18.30
4
16.13
1-Dodecene
—
—
—
—
Ethane
0.76
15.56
11
8.16
2-Ethyl-l-butene
—
—
—
—
Ethylbenzene
16.55
10.77
10
13.66
Ethylene
16.38
5.85
11
11.11
«7-Ethyltoluene
8.86
8.23
6
8.54
o-Ethyltoluene
1.73
7.24
4
4.49
p-Ethyltoluene
1.23
15.20
5
8.21
//-Heptane
4.37
6.48
10
5.42
1-Heptene
—
—
—
—
w-Hexane
1.75
5.77
10
3.76
1-Hexene
—
—
—
—
67.v-2-Hc\cnc
—
—
—
—
;ra«.v-2-Hc\cnc
—
—
—
—
Isobutane
1.09
4.85
11
2.97
Isobutene/1 -Butene
—
—
—
—
Isobutylene
—
18.78
1
18.78
Isopentane
16.32
13.45
5
14.88
Isoprene
6.08
45.68
7
25.88
Isopropylbenzene
—
0.46
1
0.46
2-Methyl-1 -butene
1.40
—
1
1.40
3 -Methyl-1 -butene
—
—
—
—
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV fortius
method is calculated from the pollutant-specific averages and is provided in the final column of
the table.
BOLD ITALICS = EPA-designated NATTS Site
31-15

-------
Table 31-3. SNMOC Method Precision: Coefficient of Variation
Based on Duplicate Samples by Site and Pollutant (Continued)
Pollutant
BTUT
NBIL
# of
Pairs
Average
by
Pollutant
2-Methyl-1 -pentene
—
—
—
—
4-Methyl-1 -pentene
—
—
—
—
2 -Me thy 1-2 -butene
0.98
4.58
2
2.78
Methylcyclohexane
1.67
7.37
8
4.52
Methylcyclopentane
5.71
4.41
10
5.06
2-Methylheptane
3.21
—
3
3.21
3-Methylheptane
7.18
2.22
4
4.70
2-Methylhexane
41.31
29.05
11
35.18
3-Methylhexane
3.85
3.02
7
3.43
2-Methylpentane
31.81
7.83
11
19.82
3-Methylpentane
11.96
4.59
10
8.28
w-Nonane
19.86
14.19
9
17.03
1-Nonene
—
8.73
2
8.73
//-Octane
10.44
8.63
8
9.53
1-Octene
14.10
7.86
6
10.98
/7-Pentane
8.45
37.04
11
22.75
1-Pentene
17.58
17.28
6
17.43
67.v-2-Pcntcnc
—
—
—
—
;ra«.v-2-Pcntene
3.21
—
2
3.21
fl-Pincnc
3.06
3.37
6
3.21
/>-Pi nc nc
—
—
—
—
Propane
0.43
4.98
11
2.70
/7-Propylbenzene
7.88
9.14
5
8.51
Propylene
27.00
12.56
11
19.78
Propyne
—
—
—
—
Styrene
—
9.48
1
9.48
Toluene
8.13
3.52
11
5.82
//-Tridccanc
5.99
27.88
3
16.93
1-Tridecene
—
—
—
—
1,2,3 -Trimethylbenzene
30.46
8.98
6
19.72
1,2,4-Trimethylbenzene
17.91
7.46
10
12.69
1,3,5 -T rime thy lbenzene
11.42
5.50
5
8.46
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV fortius
method is calculated from the pollutant-specific averages and is provided in the final column of
the table.
BOLD ITALICS = EPA-designated NATTS Site
31-16

-------
Table 31-3. SNMOC Method Precision: Coefficient of Variation
Based on Duplicate Samples by Site and Pollutant (Continued)
Pollutant
BTUT
NBIL
# of
Pairs
Average
by
Pollutant
2,2,3 -T rime thy lpentane
17.30
7.25
3
12.28
2,2,4 -T rime thy lpentane
10.66
3.90
10
7.28
2,3,4-Trimethy lpentane
29.77
15.07
10
22.42
/7-Undecane
4.71
12.59
7
8.65
1-Undecene
—
—
—
—
«/-Xylene//?-Xylene
5.05
4.93
11
4.99
o-Xylene
8.16
5.03
9
6.60
SNMOC (Sum of Knowns)
20.35
9.03
11
14.69
Sum of Unknowns
26.71
25.21
11
25.96
Average by Site
9.94
10.54
436
9.95
# of Pairs Collected per Site
4
7
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV fortius
method is calculated from the pollutant-specific averages and is provided in the final column of
the table.
BOLD ITALICS = EPA-designated NATTS Site
31.2.3 Carbonyl Compound Method Precision
Table 31-4 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 mid-level variability, ranging from a CV of 0.57 percent
(acetaldehyde for TOOK) to 47.47 percent (2-butanone for SYFL). The number of sites for
which a given pollutant has a CV greater than 15 percent varies from none (five pollutants) to
seven (2-butanone).
The pollutant-specific average CV, as shown in orange in Table 31-4, ranges from
3.50 percent (acetaldehyde) to 10.31 percent (hexaldehyde). The site-specific average CV, as
shown in green in Table 31-4, ranges from 2.53 percent (TOOK) to 13.55 percent (SKFL). None
of the sites collecting duplicate or collocated carbonyl compound samples have average CVs
greater than 15 percent. The overall average method precision is 7.12 percent for carbonyl
compounds.
31-17

-------
Sites at which duplicate samples were collected are highlighted in blue in Table 31-4
while sites at which collocated samples were collected are highlighted in purple. Collocated
carbonyl compound samples were collected at only three of the sites shown in Table 31-4
(DEMI, INDEM, and PXSS); the remainder collected duplicate samples. The average CV for
sites that collected duplicate samples was calculated and is shown in Table 31-4 in blue while the
average CV for sites collecting collocated samples is shown in purple. The average CV for both
precision types meets the MQO of 15 percent, with the variability associated with collocated
samples (9.64 percent) slightly greater than the variability associated with duplicate samples
(6.79 percent).
Table 31-4. Carbonyl Compound Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant
Pollutant
ADOK
AZFL
BTUT
CHNJ
CSNJ
DEMI
ELNJ
GLKY
Acetaldehyde
1.56
4.78
3.83
1.32
4.19
6.84
1.50
1.09
Acetone
1.29
14.59
1.98
6.34
15.18
7.55
26.30
1.82
Benzaldehyde
2.47
10.34
6.59
5.75
4.99
11.56
4.68
8.98
2-Butanone
1.77
16.39
0.83
7.65
15.29
5.49
15.05
2.89
Butyraldehyde
3.92
9.33
5.12
5.04
5.37
13.95
4.39
3.57
Crotonaldehyde
1.48
6.65
4.74
5.09
5.43
10.60
4.20
8.39
2,5-Dimethylbenzaldehyde
--
--
--
--
--
--
--
--
Formaldehyde
1.99
5.10
3.91
2.51
4.27
8.94
2.74
2.33
Hexaldehyde
7.28
6.92
4.65
8.40
3.80
8.78
3.29
14.31
Isovaleraldehyde
—
—
—
—
—
—
—
—
Propionaldehyde
2.59
5.78
4.73
2.57
5.78
19.00
1.47
3.01
Tolualdehydes
4.52
7.13
9.43
8.57
7.19
11.54
5.26
8.55
Valeraldehyde
5.97
7.12
5.67
5.65
6.54
14.88
2.89
9.16
Average by Site
3.17
8.56
4.68
5.35
7.09
10.83
6.52
5.83
# of Pairs Collected per Site
3
6
4
6
6
7
6
6
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV fortius method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples.
BOLD ITALICS = EPA-designated NATTS Site
31-18

-------
Table 31-4. Carbonyl Compound Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
GPCO
INDEM
NBIL
NBNJ
OCOK
ORFL
PXSS
ROIL
Acetaldehyde
10.12
9.45
1.36
2.76
1.42
1.69
6.31
3.65
Acetone
3.44
7.01
2.79
6.12
3.53
20.17
5.36
11.48
Benzaldehyde
23.93
8.20
5.89
6.86
5.60
8.26
4.70
7.26
2-Butanone
7.50
14.19
4.79
6.70
6.28
30.01
16.44
6.30
Butyraldehyde
11.95
10.81
5.37
4.61
1.78
8.40
15.49
6.35
Crotonaldehyde
3.32
7.11
5.74
3.40
2.08
2.36
6.84
5.16
2,5 -Dimethy lbenzaldehy de
—
—
—
—
—
—
—
—
Formaldehyde
11.94
9.46
1.79
3.99
1.41
2.49
9.59
1.50
Hexaldehyde
44.99
6.58
2.51
8.78
8.51
9.52
12.42
6.22
Isovaleraldehyde
—
—
—
—
—
—
—
—
Propionaldehyde
8.11
9.34
2.64
5.27
1.67
8.63
11.53
3.21
Tolualdehydes
19.92
3.94
8.01
3.11
3.67
21.29
10.24
14.30
Valeraldehyde
2.98
6.76
5.00
6.31
1.73
11.02
7.25
14.11
Average by Site
13.47
8.44
4.17
5.27
3.42
11.26
9.65
7.23
# of Pairs Collected per Site
4
11
7
6
6
7
6
6
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV fortius method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples.
BOLD ITALICS = EPA-designated NATTS Site
31-19

-------
Table 31-4. Carbonyl Compound Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
S4MO
SEWA
SKFL
SPIL
SYFL
TMOK
TOOK
TROK
Acetaldehyde
3.13
1.87
3.44
8.36
4.37
0.90
0.57
1.08
Acetone
6.49
1.02
18.36
8.21
11.11
6.56
2.18
1.66
Benzaldehyde
7.42
10.05
10.07
8.21
4.85
2.64
7.19
5.05
2-Butanone
5.91
3.84
24.07
7.92
47.47
3.23
1.14
2.25
Butyraldehyde
3.16
3.11
14.26
9.58
21.16
2.41
1.41
3.26
Crotonaldehyde
3.61
6.26
4.57
6.04
7.76
4.24
1.12
1.71
2,5-Dimethylbenzaldehyde
—
—
—
—
—
—
—
—
Formaldehyde
3.81
5.71
14.93
9.64
3.22
0.71
1.17
2.26
Hexaldehyde
7.48
5.80
21.21
36.21
11.34
4.50
4.53
6.29
Isovaleraldehyde
—
—
—
—
—
—
—
—
Propionaldehyde
5.84
4.96
4.31
10.05
4.27
2.58
1.44
2.06
Tolualdehydes
14.41
7.98
14.48
13.65
6.66
4.22
5.97
5.52
Valeraldehyde
5.73
7.73
19.39
19.81
7.86
2.27
1.11
5.53
Average by Site
6.09
5.30
13.55
12.52
11.82
3.11
2.53
3.33
# of Pairs Collected per Site
6
6
6
5
6
6
6
7
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV fortius method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples.
BOLD ITALICS = EPA-designated NATTS Site
31-20

-------
Table 31-4. Carbonyl Compound Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
WPIN
YUOK
# of
Pairs
Average
by
Pollutant
Average
for
Duplicate
Pairs
Average
for
Collocated
Pairs
Acetaldehyde
4.72
0.75
158
3.50
2.98
7.53
Acetone
13.90
3.85
158
8.01
8.19
6.64
Benzaldehyde
7.97
6.77
158
7.55
7.47
8.16
2-Butanone
10.62
3.73
158
10.30
10.07
12.04
Butyraldehyde
6.99
2.06
158
7.03
6.20
13.42
Crotonaldehyde
8.93
1.77
158
4.95
4.52
8.19
2,5-Dimethylbenzaldehyde
—
—
—
—
—
—
Formaldehyde
5.18
0.68
158
4.66
4.06
9.33
Hexaldehyde
6.90
6.80
156
10.31
10.44
9.26
Isovaleraldehyde
—
—
—
—
—
—
Propionaldehyde
7.18
1.48
158
5.37
4.33
13.29
Tolualdehydes
9.97
9.76
128
9.20
9.29
8.57
Valeraldehyde
7.95
3.84
151
7.47
7.19
9.63
Average by Site
8.21
3.77
1,699
7.12
6.79
9.64
# of Pairs Collected per Site
9
4
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV fortius method is
calculated from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples.
BOLD ITALICS = EPA-designated NATTS Site
31.2.4 PAH Method Precision
The method precision results for collocated PAH samples are shown in Table 31-5 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 2013.
The results from collocated samples exhibit low- to high-level variability, ranging from a CV of
0.78 percent (benzo(a)anthracene for SEW A) to 53.84 percent (chrysene for RUCA, although
coronene has a similar CV for ANAK). The number of sites for which a given pollutant has a CV
greater than 15 percent varies from none (14 pollutants) to two (four pollutants).
31-21

-------
The pollutant-specific average CV, as shown in orange in Table 31-5, ranges from
1.04 percent (dibenz(a,h)anthracene) to 21.28 percent (chrysene). The precision for
dibenz(a,h)anthracene is based on a single pair of measurements greater than the MDL. The site-
specific average CV, as shown in green in Table 31-5, ranges from 5.93 percent (SEWA) to
17.21 percent (ANAK). ANAK is the only site with a site-specific average CV greater than
15 percent. The overall average method precision was 10.28 percent.
Table 31-5. PAH Method Precision: Coefficient of Variation
Based on Collocated Samples by Site and Pollutant
Pollutant
ANAK
DEMI
RUCA
SEWA
SYFL
# of
Pairs
Average
by
Pollutant
Acenaphthene
8.90
8.70
9.31
5.62
4.18
28
7.34
Acenaphthylene
10.17
9.15
16.90
7.71
3.16
19
9.42
Anthracene
9.06
7.03
8.65
5.25
1.96
20
6.39
Benzo(a)anthracene
9.35
12.77
1.18
0.78
--
11
6.02
Benzo(a)pyrene
9.23
11.64
0.97
4.88
--
10
6.68
Benzo(b)fluoranthene
14.70
10.44
9.71
1.06
7.43
20
8.67
Benzo(e)pyrene
13.52
9.68
2.78
3.49
--
14
7.37
Benzo(g,h,i)perylene
29.75
6.29
6.95
5.89
4.96
17
10.77
Benzo(k)fluoranthene
12.75
10.53
4.86
13.23
--
9
10.34
Chrysene
32.27
6.34
53.84
4.22
9.72
28
21.28
Coronene
53.37
3.08
13.88
8.00
--
8
19.58
Cvclopcnta|cd|pvrcnc
36.74
--
2.09
5.87
--
5
14.90
Dibenz(a,h)anthracene
1.04
--
--
--
--
1
1.04
Fluoranthene
18.62
6.69
33.10
6.20
7.86
28
14.50
Fluorene
7.84
6.58
10.50
7.88
4.36
27
7.43
9-Fluorenone
13.58
7.71
12.85
6.56
3.99
28
8.94
Indeno(l,2,3-cd)pyrene
--
--
--
--
--
--
--
Naphthalene
7.23
6.17
10.26
4.53
8.89
28
7.42
Perylene
7.25
--
--
9.42
--
4
8.33
Phenanthrene
12.65
6.52
11.75
4.33
4.25
28
7.90
Pyrene
30.40
6.53
39.95
6.37
8.54
28
18.36
Retene
22.93
4.59
20.95
7.25
10.20
18
13.19
Average by Site
17.21
7.80
14.24
5.93
6.12
379
10.28
# of Pairs Collected per Site
6
7
6
6
3
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
BOLD ITALICS = EPA-designated NATTS Site
31-22

-------
31.2.5 Metals Method Precision
The method precision for all collocated metals samples are presented in Table 31-6 as the
CV per pollutant per site, the average CV per site, the average CV per pollutant, and the overall
average CV across the metals listed. All samples evaluated in this section are collocated samples.
The results from collocated samples exhibit low- to mid-level variability, ranging from a CV of
1.02 percent (antimony for ASKY-M) to 49.11 percent (cobalt for NBIL). The number of sites
for which a given pollutant has a CV greater than 15 percent varies from none (chromium) to
eight (mercury); with several metals exceeding 15 percent for only one site.
The pollutant-specific average CV, as shown in orange in Table 31-6, ranges from
6.12 percent (chromium) to 24.51 percent (mercury), with five of the 11 metals with an average
CV greater than 15 percent. The site-specific average CV, as shown in green in Table 31-6,
ranges from 7.93 percent (ASKY-M) to 20.71 percent (GLKY). GLKY and NBIL have site-
specific average CVs greater than 15 percent. The overall average method precision for metals is
13.37 percent.
31-23

-------
Table 31-6. Metals Method Precision: Coefficient of Variation
Based on Collocated Samples by Site and Pollutant
Pollutant
ASKY-M
BOMA
BTUT
GLKY
NBIL
Antimony
1.02
6.84
2.95
20.71
21.81
Arsenic
6.79
21.37
10.13
13.40
6.34
Beryllium
7.03
20.00
20.00
—
28.28
Cadmium
2.39
19.98
28.53
35.74
4.94
Chromium
—
—
—
—
5.27
Cobalt
24.11
7.76
7.55
12.18
49.11
Lead
2.44
3.87
2.77
18.50
12.30
Manganese
1.51
3.54
6.26
17.66
7.15
Mercury
22.34
16.33
15.71
31.34
38.88
Nickel
4.15
30.19
9.20
26.99
—
Selenium
7.55
13.90
35.69
9.85
9.63
Average by Site
7.93
14.38
13.88
20.71
18.37
# of Pairs Collected per Site
5
28
5
29
9
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV fortius method is
calculated from the pollutant-specific averages and is provided in the final column of the table.
BOLD ITALICS = EPA-designated NATTS Site
31-24

-------
Table 31-6. Metals Method Precision: Coefficient of Variation
Based on Collocated Samples by Site and Pollutant (Continued)
Pollutant
S4MO
TOOK
UNVT
# of
Pairs
Average
by
Pollutant
Antimony
3.77
12.23
38.79
190
13.52
Arsenic
13.99
4.76
5.49
169
10.28
Beryllium
17.32
8.33
—
72
16.83
Cadmium
8.25
19.73
17.00
191
17.07
Chromium
—
6.96
—
30
6.12
Cobalt
15.34
9.60
12.65
175
17.29
Lead
2.92
6.54
1.62
191
6.37
Manganese
5.19
5.82
4.91
191
6.50
Mercury
27.63
15.58
28.28
103
24.51
Nickel
21.97
7.74
15.71
136
16.57
Selenium
7.55
3.96
8.32
137
12.06
Average by Site
12.39
9.20
14.75
1,585
13.37
# of Pairs Collected per Site
55
54
6
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV fortius method is
calculated from the pollutant-specific averages and is provided in the final column of the table.
BOLD ITALICS = EPA-designated NATTS Site
31.2.6 Hexavalent Chromium Method Precision
Table 31-7 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 results from collocated samples exhibit low- to high-
level variability. The site-specific CV ranges from 0.11 percent (SEWA) to 82.40 percent
(SKFL); the CVs for seven of the 14 sites collecting collocated hexavalent chromium samples
are greater than 15 percent. The overall average method precision of hexavalent chromium is
20.51 percent, as shown in orange in Table 31-7. This is the only method exceeding the MQO of
15 percent CV for method precision. Note, however, that most sites sampling hexavalent
chromium discontinued sampling this pollutant at the end of July 2013 and the precision
calculations are based on relatively few pairs; precision for 10 of the 14 sites is based on two or
fewer samples.
31-25

-------
Table 31-7. Hexavalent Chromium Method Precision: Coefficient of Variation
Based on Collocated Samples by Site
Site
CV
(%)
# of
Pairs
BOMA
8.26
2
BTUT
15.50
1
BXNY
11.61
3
CAMS 35
27.55
3
DEMI
7.52
2
GPCO
5.37
1
NBIL
2.10
1
PRRI
32.79
1
PXSS
33.34
3
RIVA
35.53
2
S4MO
17.19
4
SDGA
7.92
1
SEWA
0.11
1
SKFL
82.40
2
Average CV
20.51
# of Pairs
27
Bold = CV greater than 15 percent
Orange shading indicates the average CV
for this method.
BOLD ITALICS = EPA-designated
NATTS Site
31.3 Analytical Precision
Analytical precision is a measurement of random errors associated with the process of
analyzing environmental samples. These errors may result from various factors, including
random "noise" inherent to analytical instruments. Laboratories can evaluate the analytical
precision of ambient air samples by comparing concentrations measured during multiple
analyses of a single sample (i.e., replicate samples). Replicate analyses were run on duplicate or
collocated samples collected during the program year. CVs were calculated for every replicate
analysis run on duplicate or collocated samples collected during the program year. In addition,
replicate analyses were also run on select individual samples to provide an indication of
analytical precision for monitoring sites unable to collect duplicate or collocated samples
(i.e., samplers "unequipped" to collect duplicate or collocated samples). Individual samples with
replicate analyses were also factored into the CV calculations for analytical precision. However,
only results at or above the MDL were used in these calculations, similar to the calculation of
method precision discussed in Section 31.2.
31-26

-------
Table 31-8 presents the 2013 NMP analytical precision for VOCs, SNMOCs, carbonyl
compounds, PAHs, metals, and hexavalent chromium, presented as average CV (expressed as a
percentage). The average CV for each method met the program MQO of 15 percent for
precision. The analytical precision for all methods is less than 9 percent. This table also includes
the number of pairs that were included in the calculation of the analytical precision. The number
of pairs including those with concentrations less than the MDL is also included in Table 31-8 to
provide an indication of the effect that excluding those with concentrations less than the MDL
has on the population of pairs in the dataset.
Table 31-8. Analytical Precision by Analytical Method
Method/Pollutant
Group
Average
Coefficient of
Variation
(%)
Number of
Pairs Included
in the
Calculation
Total Number of
Pairs Without
the > MDL
exclusion
VOCs
(TO-15)
5.81
9,197
11,608
SNMOCs
4.20
1,742
2,089
Carbonyl Compounds
(TO-11A)
2.52
3,626
3,631
PAHs
(TO-13)
3.91
2,073
2,520
PAHs/Phenols
3.16
118
149
Metals Analysis
(Method IO-3.5/FEM)
5.98
3,744
4,815
Hexavalent Chromium
(ASTMD7641)
8.53
58
58
MQO
15.00 percent CV
Tables 31-9 through 31-15 present analytical precision for VOCs, SNMOCs, carbonyl
compounds, PAHs, PAH/Phenols, metals, and hexavalent chromium, respectively, as the CV per
pollutant per site and the average CV per pollutant, per site, and per method. Pollutants
exceeding the 15 percent MQO for CV are bolded in each table. In Tables 31-9 through 31-15,
the number of pairs in comparison to the respective tables listed for duplicate or collocated
analyses in Tables 31-2 through 31-7 is higher, the reason for which is two-fold. One reason is
because each primary and duplicate (or collocated) sample produces a replicate analysis. The
second reason is due to replicates run on individual samples. This is also the reason the number
of sites provided in Tables 31-9 through 31-15 is higher than Tables 31-2 through 31-7. Note that
collocated samples were not collected at KMMS, the one NMP site at which PAHs/Phenols were
31-27

-------
collected, thus, this site/method has no corresponding table in Section 31.2 for method precision.
The replicate analyses of duplicate, collocated, and individual samples indicate that the analytical
precision level is within the program MQOs.
31.3.1 VOC Analytical Precision
Table 31-9 presents analytical precision results from replicate analyses of duplicate,
collocated, and select individual VOC samples as the CV per pollutant per site, the average CV
per site, the average CV per pollutant, and the overall average CV across the VOCs listed. The
analytical precision results from replicate analyses show that, for most of the pollutants, the VOC
analytical precision is within 15 percent. The CV ranged from 0 percent (several pollutants and
several sites) to 62.11 percent (methyl tert-butyl ether for GPCO). The number of sites for which
a given pollutant has a CV greater than 15 percent varies from none (44 pollutants) to six
(methyl tert-butyl ether).
The pollutant-specific average CV, as shown in orange in Table 31-9, ranges from
0 percent (1,1,2,2-tetrachloroethane) to 21.55 percent (methyl tert-butyl ether). The CV for
1,1,2,2-tetrachloroethane is based on a single pair of samples collected at ELNJ, the only site for
which a pair of measurements greater than the MDL were collected. The site-specific average
CV, as shown in green in Table 31-9, ranges from 4.04 percent (CCKY) to 8.85 percent
(UNVT). The overall average analytical precision is 5.81 percent. Note that the results for
acrolein, acetonitrile, acrylonitrile, and carbon disulfide were excluded from the precision
calculations due to the issues described in Section 3.2.
Sites at which duplicate samples were collected are highlighted in blue in Table 31-9,
sites at which collocated samples were collected are highlighted in purple, and sites at which
replicates were run on individual samples are highlighted in brown. Collocated VOC samples
were collected at only three of the sites shown in Table 31-9 (BURVT, PXSS, and TVKY);
replicates were run on individual VOC samples for nine sites, and the remainder of sites
collected duplicate VOC samples. The average CV for sites that collected duplicate samples was
calculated and is shown in Table 31-9 in blue, the average CV for sites collecting collocated
samples is shown in purple, and the average CV for sites where replicates were run on individual
samples is shown in brown. The average CV for all three precision types meets the MQO of
31-28

-------
15 percent, with the variability ranging from 5.18 percent (replicates run on individual samples)
to 6.11 percent (replicates run on duplicate samples).
Table 31-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant
Pollutant
ADOK
ANAK
ASKY
ATKY
BLKY
BTUT
BURVT
CCKY
Acetylene
4.92
4.88
5.10
2.28
4.16
7.58
6.57
2.29
fcrt-Amvl Methyl Ether
—
..
..
..
..
..
..
..
Benzene
7.26
4.10
6.37
4.60
6.83
4.59
5.62
3.41
Bromochloromethane
—
—
—
—
—
—
—
—
Bromodichloromethane
..
..
..
..
..
..
..
..
Bromofonn
..
..
..
..
..
..
..
..
Bromomethane
..
10.50
14.63
7.66
10.50
6.68
9.22
4.71
1.3 -Butadiene
4.18
7.96
7.81
15.99
16.29
2.10
10.68
1.39
Carbon Tetrachloride
5.91
4.13
4.95
3.29
2.10
4.66
4.26
3.93
Chlorobenzene
..
..
..
..
..
..
..
..
Chloroethane
..
..
..
..
..
..
12.26
..
Chloroform
2.67
7.44
9.85
5.41
7.60
5.88
8.25
2.74
Chloromethane
6.26
2.08
3.43
1.87
2.53
3.35
4.10
0.86
Chloroprene
..
..
..
..
..
..
..
..
Dibromochloromethane
..
..
..
..
..
..
..
..
1,2-Dibromoethane
..
..
..
..
..
..
..
..
«7-Dichlorobenzene
..
..
..
..
..
..
..
..
o-Dichlorobenzene
..
..
..
..
..
..
..
..
p-Dichlorobenzene
..
9.39
..
..
..
..
..
..
Dichlorodifluoromethane
5.38
2.47
4.30
1.72
2.53
3.64
3.99
0.79
1,1 -Dichloroethane
..
..
..
..
..
..
..
..
1,2-Dichloroethane
9.97
6.26
8.57
3.74
7.73
4.80
7.34
3.59
1,1 -Dichloroethene
..
..
..
..
..
2.48
..
..
cis-1,2-Dichloroethylene
..
..
..
..
..
..
..
..
trans-1,2-Dichloroethylene
..
..
..
..
..
..
10.80
..
Dichloromethane
4.63
3.35
6.19
6.57
2.22
2.10
4.70
1.49
1,2-Dichloropropane
..
..
..
..
..
..
..
..
cis-1,3 -Dichloropropene
—
—
—
—
—
—
—
—
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples; brown shading identifies sites for which replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-29

-------
Table 31-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant
Pollutant
ADOK
ANAK
ASKY
ATKY
BLKY
BTUT
BURVT
CCKY
trans-1,3 -Dichloropropene
—
—
—
—
—
—
—
—
Dichlorotetrafluoroethane
8.74
7.56
5.94
4.86
5.79
5.63
8.09
2.57
Ethyl Aery late
—
—
—
—
—
—
—
—
Ethyl terf-Butyl Ether
—
—
—
—
—
5.79
—
—
Ethylbenzene
9.26
5.42
9.20
17.81
4.16
3.79
7.63
6.88
Hexachloro-1,3 -butadiene
—
—
—
—
—
—
—
—
Methyl Isobutyl Ketone
19.83
4.03
7.92
8.19
6.01
5.39
9.24
2.60
Methyl Methacrylate
—
—
—
5.66
—
—
—
10.10
Methyl tert-Butyl Ether
—
—
4.29
—
—
6.34
—
—
//-Octane
11.58
8.72
11.18
16.22
7.22
10.80
9.02
7.54
Propylene
3.50
2.04
4.14
3.77
3.84
3.59
4.33
1.00
Styrene
9.62
9.64
10.85
15.48
0.00
3.35
8.69
8.14
1.1,2,2-Tetrachloroethane
—
—
—
—
—
—
—
—
Tetrachloroethylene
—
7.80
7.14
—
—
2.86
6.86
4.88
Toluene
7.59
3.70
5.87
5.45
8.92
8.54
5.02
3.26
1,2,4-Trichlorobenzene
—
—
—
—
—
—
—
—
1.1.1 -Trichloroethane
—
—
—
—
—
—
—
—
1,1,2-Trichloroethane
—
—
—
—
—
—
—
—
Trichloroethylene
—
—
—
—
—
—
7.13
—
Trichlorofluoromethane
3.85
2.24
3.70
1.62
2.36
3.10
4.40
0.91
Trichlorotrifluoroethane
5.13
2.30
4.92
2.85
2.20
3.32
4.46
2.69
1,2,4-Trimethylbenzene
8.01
7.42
6.65
18.64
0.00
4.17
7.78
5.56
1,3,5-Trimethylbenzene
9.42
7.79
14.77
—
0.00
6.22
11.65
9.87
Vinyl chloride
—
—
—
2.10
4.04
—
—
2.44
m,p-Xylene
7.36
4.42
8.43
12.86
10.89
3.49
6.70
4.92
o-Xylene
7.19
5.11
9.65
14.95
0.00
4.68
6.79
6.45
Average by Site
7.37
5.63
7.43
7.65
4.91
4.77
7.24
4.04
# of Pairs Collected per Site
6
12
5
5
5
9
55
5
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples; brown shading identifies sites for which replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-30

-------
Table 31-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
CHNJ
CSNJ
DEMI
ELNJ
GLKY
GPCO
KMMS
LAKY
Acetylene
5.68
5.94
5.25
4.92
5.24
3.95
5.44
5.34
tert-Amyl Methyl Ether
—
—
—
—
—
—
—
—
Benzene
5.67
6.33
4.14
4.43
3.85
6.03
5.61
5.09
Bromochloromethane
—
..
—
..
..
..
..
..
Bromodichloromethane
—
..
—
..
..
..
..
..
Bromofonn
—
..
—
..
..
..
..
..
Bromomethane
10.46
19.72
—
21.48
5.19
7.51
9.40
8.11
1.3 -Butadiene
24.32
6.06
6.51
4.83
6.80
13.30
12.32
7.07
Carbon Tetrachloride
4.55
6.56
5.22
4.86
4.86
4.41
4.93
5.66
Chlorobenzene
—
—
—
—
—
—
—
—
Chloroethane
4.66
2.55
..
4.56
..
..
..
0.00
Chloroform
6.21
5.46
5.73
5.23
7.00
4.59
6.67
6.33
Chloromethane
4.05
4.04
4.49
3.99
5.28
2.23
4.16
2.77
Chloroprene
..
..
..
..
..
..
..
..
Dibromochloromethane
..
..
..
..
..
..
..
..
1,2-Dibromoethane
..
..
..
..
..
..
..
..
«7-Dichlorobenzene
..
..
..
..
..
..
..
..
o-Dichlorobenzene
—
—
—
—
—
—
—
—
p-Dichlorobenzene
..
..
..
..
..
..
..
..
Dichlorodifluoromethane
4.09
3.63
4.51
3.57
5.21
2.56
4.41
3.07
1,1 -Dichloroethane
..
..
—
..
..
..
..
2.24
1,2-Dichloroethane
7.50
11.77
11.57
5.35
7.29
7.55
7.69
12.41
1,1 -Dichloroethene
..
..
..
..
..
..
..
..
cis-1,2-Dichloroethylene
..
..
..
..
..
..
..
..
trans-1,2-Dichloroethylene
..
..
..
..
..
..
..
..
Dichloromethane
7.05
3.38
5.07
4.46
4.58
3.83
6.47
4.32
1,2-Dichloropropane
..
..
..
..
..
..
..
..
cis-1,3 -Dichloropropene
—
—
—
—
—
—
—
—
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples; brown shading identifies sites for which replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-31

-------
Table 31-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
CHNJ
CSNJ
DEMI
ELNJ
GLKY
GPCO
KMMS
LAKY
trans-1,3 -Dichloropropene
—
—
—
—
—
—
—
—
Dichlorotetrafluoroethane
6.29
7.44
7.40
6.22
4.83
8.44
6.03
7.26
Ethyl Aery late
—
—
—
—
—
—
—
—
Ethyl tert-Butyl Ether
5.87
8.23
—
4.61
—
5.52
—
—
Ethylbenzene
15.86
4.71
5.85
4.70
6.92
6.26
8.35
10.60
Hexachloro -1,3 -butadiene
—
—
—
—
—
—
—
—
Methyl Isobutyl Ketone
12.94
5.56
9.58
9.36
5.47
4.56
7.14
9.47
Methyl Methacrylate
—
3.27
—
6.50
—
2.64
—
—
Methyl tert-Butyl Ether
14.25
15.78
—
26.92
10.88
62.11
—
—
//-Octane
7.44
5.69
7.91
3.95
6.27
9.22
13.89
7.36
Propylene
5.37
3.79
4.04
3.69
5.84
2.69
4.16
1.97
Styrene
7.05
4.88
5.36
7.85
10.15
5.12
17.00
6.51
1,1,2,2-Tetrachloroethane
—
—
—
0.00
—
—
—
—
T etrachloroethy lene
10.25
3.71
6.27
5.07
—
6.47
8.31
—
Toluene
8.67
5.88
4.66
3.73
3.79
4.01
5.87
4.74
1,2,4-Trichlorobenzene
—
—
—
—
—
—
—
—
1,1,1 -Trichloroethane
—
—
—
6.61
—
—
—
—
1,1,2-Trichloroethane
—
—
—
—
—
—
—
0.00
T richloroethy lene
—
2.96
—
5.88
—
—
2.70
—
T richlorofluoro methane
3.55
3.63
4.18
3.68
3.34
2.61
3.85
3.20
T richlorotrifluoroethane
3.50
4.19
4.05
4.05
4.73
3.03
4.09
5.33
1,2,4-Trimethylbenzene
8.62
7.22
6.19
4.35
9.10
6.88
8.33
3.47
1,3,5 -T rime thy lbenzene
7.35
7.31
7.30
6.45
9.68
6.32
16.70
2.13
Vinyl chloride
—
—
—
—
—
—
12.31
3.98
m,p-Xylene
12.57
5.03
4.78
3.89
6.20
5.49
6.70
6.42
o-Xylene
15.64
5.78
4.90
3.63
7.44
6.49
6.87
4.85
Average by Site
8.50
6.22
5.87
6.09
6.25
7.55
7.67
5.17
# of Pairs Collected per Site
12
10
12
12
12
14
12
5
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples; brown shading identifies sites for which replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-32

-------
Table 31-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
LEKY
NBIL
NBNJ
OCOK
PXSS
ROIL
RUVT
S4MO
Acetylene
2.95
6.08
6.23
6.85
2.06
5.18
6.80
6.24
tert-Amyl Methyl Ether
—
—
—
—
—
—
—
—
Benzene
1.44
7.22
6.66
3.70
4.27
5.26
8.52
3.59
Bromochloromethane
—
..
..
..
..
..
..
..
Bromodichloromethane
—
6.54
..
..
..
..
..
..
Bromofonn
—
2.06
..
..
..
..
..
..
Bromomethane
4.00
3.02
6.37
1.74
8.74
4.35
2.86
4.94
1.3 -Butadiene
6.57
7.43
4.95
14.93
3.59
10.82
6.88
6.00
Carbon Tetrachloride
1.89
6.44
4.94
5.29
4.76
3.89
8.45
3.97
Chlorobenzene
1.40
—
—
—
—
—
—
—
Chloroethane
..
5.78
6.96
1.27
3.02
3.63
..
7.86
Chloroform
4.94
4.83
6.34
6.33
3.75
3.30
6.03
5.61
Chloromethane
2.83
3.64
3.19
3.97
2.59
2.04
6.28
3.80
Chloroprene
..
..
..
..
..
..
..
..
Dibromochloromethane
..
4.35
..
..
..
..
..
..
1,2-Dibromoethane
..
..
..
..
..
..
..
..
«7-Dichlorobenzene
..
..
..
..
..
..
..
..
o-Dichlorobenzene
—
—
—
—
—
—
—
—
p-Dichlorobenzene
16.32
..
..
..
2.91
..
..
5.41
Dichlorodifluoromethane
2.37
3.37
3.28
3.78
2.22
3.95
7.00
3.96
1,1 -Dichloroethane
..
..
..
..
..
..
..
..
1,2-Dichloroethane
8.20
4.60
6.09
8.58
6.89
6.23
3.30
7.50
1,1 -Dichloroethene
..
..
..
..
..
..
..
..
cis-1,2-Dichloroethylene
..
..
..
..
..
..
..
..
trans-1,2-Dichloroethylene
..
..
..
..
0.00
..
2.57
..
Dichloromethane
2.68
4.48
2.83
5.20
4.87
8.80
8.05
4.16
1,2-Dichloropropane
..
—
..
..
..
..
..
..
cis-1,3 -Dichloropropene
—
—
—
—
—
—
—
—
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples; brown shading identifies sites for which replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-33

-------
Table 31-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
LEKY
NBIL
NBNJ
OCOK
PXSS
ROIL
RUVT
S4MO
trans-1,3 -Dichloropropene
—
—
—
—
—
—
—
—
Dichlorotetrafluoroethane
5.72
5.37
5.50
6.51
5.05
5.22
3.59
5.40
Ethyl Aery late
—
—
—
—
—
—
—
—
Ethyl tert-Butyl Ether
—
7.51
4.60
—
—
—
—
—
Ethylbenzene
4.49
3.66
11.88
8.83
3.54
6.56
3.54
6.21
Hexachloro -1,3 -butadiene
—
—
—
—
—
—
—
—
Methyl Isobutyl Ketone
2.10
7.83
9.19
8.96
5.51
10.43
7.41
9.62
Methyl Methacrylate
—
5.54
—
—
9.23
—
—
—
Methyl tert-Butyl Ether
—
—
31.44
0.00
—
—
—
—
//-Octane
4.70
5.38
11.43
9.69
3.11
8.15
3.18
5.19
Propylene
0.85
4.02
3.92
5.47
2.73
4.83
6.86
3.48
Styrene
7.76
7.34
4.38
8.96
3.39
4.47
4.86
6.74
1,1,2,2-Tetrachloroethane
—
—
—
—
—
—
—
—
T etrachloroethy lene
8.67
5.62
5.72
3.03
5.42
3.93
3.03
4.49
Toluene
3.21
6.24
4.50
4.88
2.93
6.39
4.04
4.04
1,2,4-Trichlorobenzene
—
—
—
—
—
—
—
—
1,1,1 -Trichloroethane
—
—
—
—
—
—
—
—
1,1,2-Trichloroethane
—
—
—
—
—
—
—
—
T richloroethy lene
—
—
6.62
—
—
—
—
8.67
T richlorofluoro methane
1.72
2.99
3.12
3.79
2.00
2.57
6.53
3.55
T richlorotrifluoroethane
1.77
3.74
3.51
4.27
3.33
3.38
7.28
3.43
1,2,4-Trimethylbenzene
4.70
4.98
6.63
8.79
4.53
7.11
4.08
6.00
1,3,5 -T rime thy lbenzene
7.57
5.00
3.48
9.88
6.14
17.27
3.23
3.03
Vinyl chloride
—
2.86
—
—
—
—
—
0.00
m,p-Xylene
3.64
4.91
8.08
6.93
2.90
5.72
3.40
5.48
o-Xylene
4.70
4.00
7.92
7.40
3.93
6.36
2.98
5.89
Average by Site
4.51
5.06
6.78
6.12
4.05
5.99
5.23
5.15
# of Pairs Collected per Site
4
14
12
12
14
10
5
12
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples; brown shading identifies sites for which replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-34

-------
Table 31-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
SEWA
SPAZ
SPIL
SSMS
TMOK
TOOK
TROK
TVKY
Acetylene
3.59
5.17
6.04
3.97
7.66
5.12
3.86
5.64
tert-Amyl Methyl Ether
—
—
—
—
—
—
—
—
Benzene
2.76
3.40
4.64
5.14
7.42
6.38
5.26
5.53
Bromochloromethane
—
..
..
..
..
..
..
..
Bromodichloromethane
—
..
..
..
..
..
..
..
Bromofonn
—
..
..
..
..
..
..
..
Bromomethane
4.04
6.15
16.67
5.81
5.59
7.86
6.05
7.86
1.3 -Butadiene
5.06
7.02
6.43
6.04
5.88
4.58
12.08
6.38
Carbon Tetrachloride
2.94
3.66
4.91
5.57
7.19
14.91
4.73
4.54
Chlorobenzene
—
—
—
—
—
—
—
—
Chloroethane
..
..
..
5.06
..
4.69
..
9.32
Chloroform
4.77
2.56
7.27
5.67
14.31
7.20
7.90
5.55
Chloromethane
3.02
6.05
5.15
3.08
5.72
3.35
3.83
3.89
Chloroprene
..
..
..
..
..
..
..
..
Dibromochloromethane
..
..
..
..
..
..
..
..
1,2-Dibromoethane
..
..
..
..
..
..
..
..
«7-Dichlorobenzene
..
..
..
..
..
..
..
..
o-Dichlorobenzene
—
—
—
—
—
—
—
—
p-Dichlorobenzene
..
4.08
..
..
6.42
..
..
3.50
Dichlorodifluoromethane
2.89
4.73
4.56
3.16
5.69
3.23
3.81
3.45
1,1 -Dichloroethane
..
..
..
..
..
..
..
7.17
1,2-Dichloroethane
7.28
3.68
9.39
6.75
8.84
6.01
5.26
6.52
1,1 -Dichloroethene
..
..
..
..
..
..
..
6.53
cis-1,2-Dichloroethylene
..
..
..
..
..
..
..
..
trans-1,2-Dichloroethylene
..
..
..
..
..
..
10.35
4.24
Dichloromethane
3.27
2.34
5.48
6.07
6.87
4.86
3.59
4.47
1,2-Dichloropropane
..
..
..
..
..
..
..
..
cis-1,3 -Dichloropropene
—
—
—
—
—
—
—
—
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples; brown shading identifies sites for which replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-35

-------
Table 31-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
SEWA
SPAZ
SPIL
SSMS
TMOK
TOOK
TROK
TVKY
trans-1,3 -Dichloropropene
—
—
—
—
—
—
—
—
Dichlorotetrafluoroethane
4.80
4.55
5.50
4.71
7.21
6.14
6.82
7.56
Ethyl Aery late
—
—
—
—
—
—
—
—
Ethyl tert-Butyl Ether
—
—
10.09
—
—
—
—
—
Ethylbenzene
6.82
4.62
6.94
3.68
4.09
4.60
3.65
10.69
Hexachloro -1,3 -butadiene
—
—
—
—
—
—
—
—
Methyl Isobutyl Ketone
7.75
8.46
7.40
6.07
7.88
7.58
6.83
6.56
Methyl Methacrylate
—
0.00
—
—
—
5.02
—
3.29
Methyl tert-Butyl Ether
—
—
50.87
53.64
0.00
3.63
—
—
//-Octane
14.15
6.64
9.51
6.24
6.18
6.83
3.28
10.59
Propylene
2.52
3.54
4.01
2.88
7.44
4.77
4.39
7.53
Styrene
3.33
9.62
9.31
4.73
5.15
5.26
12.47
8.30
1,1,2,2-Tetrachloroethane
—
—
—
—
—
—
—
—
T etrachloroethy lene
3.01
6.02
5.96
2.70
10.32
6.36
4.75
8.31
Toluene
2.29
2.31
5.03
2.95
5.11
5.68
2.77
6.31
1,2,4-Trichlorobenzene
—
—
—
—
—
—
—
—
1,1,1 -Trichloroethane
—
—
—
—
—
—
—
8.84
1,1,2-Trichloroethane
—
—
—
—
—
—
—
6.62
T richloroethy lene
—
—
8.63
4.55
—
—
—
0.00
T richlorofluoro methane
2.97
2.25
3.52
3.00
5.34
3.40
3.61
3.35
T richlorotrifluoroethane
3.85
2.30
3.40
3.48
4.91
3.19
4.12
4.65
1,2,4-Trimethylbenzene
7.06
5.06
8.67
5.19
4.79
4.74
4.50
11.24
1,3,5 -T rime thy lbenzene
3.49
10.20
5.69
6.42
3.91
5.59
5.38
8.09
Vinyl chloride
—
—
—
4.56
—
—
—
4.19
m,p-Xylene
4.92
3.40
7.25
3.73
4.19
4.05
3.45
10.54
o-Xylene
5.99
4.79
8.11
4.90
4.42
4.47
3.62
11.12
Average by Site
4.69
4.72
8.53
6.42
6.25
5.54
5.45
6.54
# of Pairs Collected per Site
12
8
12
12
12
14
14
62
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples; brown shading identifies sites for which replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-36

-------
Table 31-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
UNVT
YUOK
# of
Pairs
Average
by
Pollutant
Average
for
Duplicate
Pairs
Average
for
Collocated
Pairs
Average
for
Unequipped
Acetylene
9.44
5.47
437
5.23
5.46
4.76
4.84
fcrt-Amvl Methyl Ether
—
..
—
—
—
—
—
Benzene
8.66
4.50
438
5.24
5.21
5.14
5.37
Bromochloromethane
..
..
—
—
—
—
—
Bromodichloromethane
—
—
10
6.54
6.54
—
—
Bromofonn
..
..
3
2.06
2.06
—
—
Bromomethane
7.49
9.60
181
8.09
8.35
8.61
7.34
1.3 -Butadiene

9.81
356
8.24
8.29
6.88
8.63
Carbon Tetrachloride
6.35
4.65
437
5.10
5.43
4.52
4.48
Chlorobenzene
..
..
1
1.40
—
—
1.40
Chloroethane
..
..
41
5.12
4.70
8.20
0.00
Chloroform
8.26
8.91
310
6.19
6.33
5.85
5.97
Chloromethane
8.50
L78
438
3.77
3.75
3.53
3.90
Chloroprene
..
..
—
—
—
—
—
Dibromochloromethane
..
..
7
4.35
4.35
—
—
1,2-Dibromoethane
—
—
—
—
—
—
—
«7-Dichlorobenzene
..
..
—
—
—
—
—
o-Dichlorobenzene
..
..
—
—
—
—
—
p-Dichlorobenzene
..
..
27
6.86
7.07
3.20
10.20
Dichlorodifluoromethane
8.61
1.83
438
3.76
3.77
3.22
3.90
1,1 -Dichloroethane
—
—
11
4.71
—
7.17
2.24
1,2-Dichloroethane
7.49
7.79
329
7.16
7.46
6.92
6.52
1,1 -Dichloroethene
..
..
6
4.50
2.48
6.53
—
cis-1,2-Dichloroethylene
—
—
—
—
—
—
—
trans-1,2-Dichloroethylene
..
11.47
15
6.57
10.91
5.01
2.57
Dichloromethane
6.82
3.26
423
4.66
4.72
4.68
4.52
1,2-Dichloropropane
..
..
—
—
—
—
—
cis-1,3 -Dichloropropene
—
—
—
—
—
—
—
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated from
the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated samples;
brown shading identifies sites for which replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-37

-------
Table 31-9. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
UNVT
YUOK
# of
Pairs
Average
by
Pollutant
Average
for
Duplicate
Pairs
Average
for
Collocated
Pairs
Average
for
Unequipped
trans-1,3 -Dichloropropene
—
—
—
—
—
—
—
Dichlorotetrafluoroethane
9.74
5.00
437
6.10
6.22
6.90
5.56
Ethyl Aery late
—
—
—
—
—
—
—
Ethyl tert-Butyl Ether
—
—
55
6.53
6.53
—
—
Ethylbenzene
10.50
5.94
402
6.99
6.54
7.28
7.98
Hexachloro-1,3 -butadiene
—
—
—
—
—
—
—
Methyl Isobutyl Ketone
9.75
9.79
311
7.84
8.33
7.10
6.88
Methyl Methacrylate
—
—
24
5.12
4.59
6.26
5.25
Methyl tert-Butyl Ether
—
—
51
21.55
22.99
—
4.29
//-Octane
8.41
4.72
397
7.98
8.01
7.57
8.05
Propylene
9.39
1.87
438
4.07
4.01
4.86
3.93
Styrene
8.31
10.88
236
7.50
7.41
6.79
7.95
1.1,2,2-Tetrachloroethane
—
—
1
0.00
0.00
—
—
T etrachloroethy lene
—
8.66
171
5.91
5.76
6.86
5.95
Toluene
18.50
2.76
437
5.28
4.96
4.75
6.26
1,2,4-Trichlorobenzene
—
—
—
—
—
—
—
1.1.1-Trichloroethane
—
—
3
7.72
6.61
8.84
—
1.1,2-Trichloroethane
—
—
5
3.31
—
6.62
0.00
T richloroethylene
—
—
24
5.24
5.71
3.57
—
T richlorofluoro methane
8.03
2.04
438
3.35
3.36
3.25
3.37
T richlorotrifluoroethane
7.82
3.34
438
3.90
3.77
4.15
4.13
1,2,4-Trimethylbenzene
6.18
4.74
364
6.51
6.52
7.85
6.04
1,3,5 -Trimethy lbenzene
—
7.75
198
7.35
7.34
8.62
6.82
Vinyl chloride
—
—
49
4.05
4.93
4.19
3.14
m,p-Xylene
7.32
4.36
413
6.01
5.59
6.71
6.81
o-Xylene
10.32
5.00
397
6.36
6.17
7.28
6.52
Average by Site
8.85
5.84
9,197
5.81
6.11
5.99
5.18
# of Pairs Collected per Site
11
7
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated from
the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated samples;
brown shading identifies sites for which replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-38

-------
31.3.2 SNMOC Analytical Precision
Table 31-10 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 ranges from 0 percent (2,4-dimethylpentane for BMCO and 1-hexene for PACO) to
34.79 percent (1-nonene for NBIL). CVs for only seven pollutant-site combinations are greater
than 15 percent.
The pollutant-specific average CV, as shown in orange in Table 31-10, ranges from
0 percent (1-hexene) to 16.42 percent (1-nonene). 1-Nonene is the only SNMOC with an average
CV greater than 15 percent. The site-specific average CV, as shown in green in Table 31-10,
ranges from 2.95 percent (PACO) to 5.20 percent (NBIL). The overall average analytical
precision is 4.20 percent. Note that the results for TNMOC were not included in the precision
calculations.
Sites at which duplicate samples were collected are highlighted in blue in Table 31-10
while sites at which replicates were run on individual samples are highlighted in brown.
Collocated SNMOC samples were not collected at the NMP sites sampling SNMOC. Duplicate
SNMOC samples were collected at only BTUT and NBIL; replicates were run on individual
SNMOC samples collected at the five Garfield County, Colorado sites. The average CV for sites
that collected duplicate samples was calculated and is shown in Table 31-10 in blue while the
average CV for sites where replicates were run on individual samples is shown in brown. The
variability ranges from 3.62 percent (replicates run on individual samples) to 5.24 percent
(replicates run on duplicate samples).
31-39

-------
Table 31-10. SNMOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant
Pollutant
BMCO
BRCO
BTUT
NBIL
PACO
Acetylene
2.84
6.13
4.78
7.88
4.66
Benzene
0.75
3.28
4.48
5.77
4.26
1.3 -Butadiene
—
—
10.77
—
0.25
//-Butane
0.98
1.27
1.52
0.84
0.94
1-Butene
—
—
—
—
—
cv.v-2-Butcnc
—
—
3.70
—
2.99
;ra«.v-2-Butcne
—
—
4.78
—
3.08
Cyclohexane
2.45
0.37
4.02
4.50
0.41
Cyclopentane
3.05
1.70
1.88
17.97
1.58
Cyclopentene
—
—
—
—
—
w-Decane
3.00
5.97
4.63
4.55
3.53
1-Decene
—
—
—
—
—
/w-Diethylbenzene
—
—
—
—
—
p-Diethylbenzene
—
—
—
—
—
2,2-Dimethylbutane
3.81
4.07
7.19
10.81
3.45
2,3 -Dimethy lbutane
2.06
1.07
2.41
3.51
1.85
2,3 -Dimethy lpentane
7.55
4.39
2.30
4.45
2.76
2,4-Dimethylpentane
0.00
5.48
3.79
1.97
8.55
n-Dodecane
—
—
2.74
12.46
3.10
1-Dodecene
—
—
—
—
—
Ethane
0.34
0.58
2.25
0.57
0.50
2-Ethyl-l-butene
—
—
—
—
—
Ethylbenzene
3.82
4.90
6.11
7.01
6.80
Ethylene
3.23
2.61
2.22
1.16
0.74
/w-Ethyltoluene
3.54
5.44
3.29
3.18
1.90
o-Ethyltoluene
2.98
—
10.50
5.01
4.06
p-Ethyltoluene
11.22
6.76
7.69
4.28
3.83
//-Heptane
0.93
3.47
5.30
4.69
2.98
1-Heptene
—
—
—
—
—
n-Hexane
2.57
1.22
2.68
1.74
1.42
1-Hexene
—
—
—
—
0.00
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV fortius method is
calculated from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples and brown shading identifies sites for which
replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-40

-------
Table 31-10. SNMOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
BMCO
BRCO
BTUT
NBIL
PACO
c/.v-2-Hc.\cnc
—
—
—
—
—
;ra«.v-2-Hc\cnc
—
—
—
—
—
Isobutane
1.01
0.87
2.70
1.28
0.29
Isobutene/1 -Butene
—
—
—
—
—
Isobutylene
—
—
—
8.86
—
Isopentane
1.04
1.25
2.67
1.60
1.15
Isoprene
—
—
5.63
2.18
3.91
Isopropylbenzene
—
—
—
2.13
—
2-Methyl-1 -butene
—
—
3.74
—
1.56
3 -Methyl-1 -butene
—
—
—
—
—
2-Methyl-1 -pentene
—
—
—
—
—
4-Methyl-1 -pentene
—
—
—
—
—
2-Methyl-2-butene
—
3.60
2.87
0.73
10.31
Methylcyclohexane
2.19
2.07
3.93
8.90
1.88
Methylcyclopentane
2.21
0.77
2.76
3.89
0.66
2-Methylheptane
2.32
5.62
5.34
—
2.49
3-Methylheptane
3.21
5.91
6.16
2.95
2.75
2-Methylhexane
2.32
2.29
2.82
5.70
2.12
3-Methylhexane
2.16
1.98
3.98
6.06
2.54
2-Methylpentane
2.98
1.33
2.40
2.89
1.14
3-Methylpentane
2.42
1.06
3.14
3.55
1.12
w-Nonane
3.54
8.42
6.48
2.36
3.29
1-Nonene
—
—
—
34.79
7.38
//-Octane
1.26
2.97
4.58
3.63
2.81
1-Octene
—
—
6.16
4.48
4.81
/7-Pentane
1.49
1.49
1.09
1.83
1.40
1-Pentene
—
—
2.55
1.76
4.94
cis-2-Pentene
—
—
—
—
1.58
;ra«.v-2-Pcntene
—
—
8.38
0.93
4.19
fl-Pincnc
3.85
3.55
14.92
4.39
—
/>-Pincnc
--
--
--
--
--
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV fortius method is
calculated from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples and brown shading identifies sites for which
replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-41

-------
Table 31-10. SNMOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
BMCO
BRCO
BTUT
NBIL
PACO
Propane
0.34
0.77
1.20
0.47
0.68
n-Propylbenzene
3.28
—
7.26
4.98
0.65
Propylene
5.53
2.85
2.69
4.83
1.27
Propyne
—
—
—
—
—
Styrene
—
6.14
—
2.77
0.36
Toluene
0.86
3.80
1.91
4.32
3.79
//-Tridccanc
—
—
10.11
19.68
—
1-Tridecene
—
—
—
—
—
1,2,3-Trimethylbenzene
16.97
—
10.24
3.11
3.94
1,2,4-Trimethylbenzene
8.73
5.20
3.46
5.26
2.66
1,3,5-Trimethylbenzene
9.50
7.02
6.26
3.95
3.58
2,2,3 -Trimethylpentane
1.86
—
5.20
7.65
4.11
2,2,4-Trimethylpentane
—
—
3.49
3.62
—
2,3,4-Trimethylpentane
—
1.42
15.86
10.00
15.48
/7-Undecane
0.94
2.53
5.86
4.71
4.23
1-Undecene
—
—
—
—
0.52
«/-Xylene//?-Xylene
3.37
4.83
1.92
3.69
3.60
o-Xylene
5.73
6.74
4.85
4.09
4.04
SNMOC (Sum of Knowns)
0.63
0.78
1.38
1.52
0.50
Sum of Unknowns
2.01
2.91
4.23
4.27
1.43
Average by Site
3.26
3.34
4.78
5.20
2.95
# of Pairs Collected per Site
4
4
9
14
4
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV fortius method is
calculated from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples and brown shading identifies sites for which
replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-42

-------
Table 31-10. SNMOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
RICO
RICO
# of
Pairs
Average
by
Pollutant
Average
for
Duplicate
Pairs
Average
for
Unequipped
Acetylene
1.39
1.12
43
4.12
6.33
3.23
Benzene
4.83
3.11
37
3.78
5.13
3.25
1.3 -Butadiene
9.75
13.86
7
8.66
10.77
7.96
//-Butane
1.86
1.66
43
1.29
1.18
1.34
1-Butene
—
—
—
—
—
—
cv.v-2-Butcnc
2.47
2.68
14
2.96
3.70
2.71
;ra«.v-2-Butcne
3.57
1.46
17
3.22
4.78
2.70
Cyclohexane
3.94
2.17
41
2.55
4.26
1.87
Cyclopentane
5.22
1.75
30
4.73
9.92
2.66
Cyclopentene
—
—
—
—
—
—
w-Decane
12.16
6.38
33
5.75
4.59
6.21
1-Decene
—
—
—
—
—
—
/w-Diethylbenzene
—
—
—
—
—
—
p-Diethylbenzene
—
—
—
—
—
—
2,2-Dimethylbutane
11.19
2.32
32
6.12
9.00
4.97
2,3 -Dimethy lbutane
5.22
1.99
39
2.59
2.96
2.44
2,3 -Dimethy lpentane
2.11
4.03
31
3.94
3.38
4.17
2,4-Dimethylpentane
—
6.03
13
4.30
2.88
5.01
n-Dodecane
—
—
9
6.10
7.60
3.10
1-Dodecene
—
—
—
—
—
—
Ethane
0.94
0.71
43
0.84
1.41
0.61
2-Ethyl-l-butene
—
—
—
—
—
—
Ethylbenzene
2.95
4.61
33
5.17
6.56
4.61
Ethylene
0.79
0.70
43
1.63
1.69
1.61
/w-Ethyltoluene
1.54
5.02
34
3.41
3.23
3.49
o-Ethyltoluene
—
8.15
13
6.14
7.76
5.07
p-Ethyltoluene
10.21
5.71
23
7.10
5.98
7.55
//-Heptane
2.07
2.68
41
3.16
4.99
2.43
1-Heptene
—
—
—
—
—
—
n-Hexane
3.61
2.97
41
2.32
2.21
2.36
1-Hexene
—
—
1
0.00
—
0.00
- = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples and brown shading identifies sites for which replicates
were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-43

-------
Table 31-10. SNMOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
RICO
RICO
# of
Pairs
Average
by
Pollutant
Average
for
Duplicate
Pairs
Average
for
Unequipped
c/.v-2-Hc.\cnc
—
—
—
—
—
—
;ra«.v-2-Hc\cnc
—
—
—
—
—
—
Isobutane
1.28
1.45
43
1.27
1.99
0.98
Isobutene/1 -Butene
—
—
—
—
—
—
Isobutylene
3.94
—
3
6.40
8.86
3.94
Isopentane
1.31
0.79
22
1.40
2.13
1.11
Isoprene
5.45
4.18
19
4.27
3.90
4.51
Isopropylbenzene
—
—
3
2.13
2.13
—
2-Methyl-1 -butene
3.53
2.31
9
2.79
3.74
2.47
3 -Methyl-1 -butene
—
—
—
—
—
—
2-Methyl-1 -pentene
—
—
—
—
—
—
4-Methyl-1 -pentene
—
—
—
—
—
—
2-Methyl-2-butene
2.48
3.67
11
3.94
1.80
5.02
Methylcyclohexane
2.98
2.45
38
3.49
6.42
2.32
Methylcyclopentane
2.61
2.29
41
2.17
3.33
1.71
2-Methylheptane
4.40
6.39
23
4.43
5.34
4.24
3-Methylheptane
3.02
4.86
23
4.12
4.56
3.95
2-Methylhexane
1.04
2.52
43
2.69
4.26
2.06
3-Methylhexane
—
2.41
27
3.19
5.02
2.27
2-Methylpentane
2.58
2.61
43
2.27
2.64
2.13
3-Methylpentane
3.20
1.90
41
2.34
3.35
1.94
w-Nonane
3.60
2.71
36
4.34
4.42
4.31
1-Nonene
7.09
—
6
16.42
34.79
7.23
//-Octane
2.52
3.14
38
2.99
4.10
2.54
1-Octene
3.59
—
21
4.76
5.32
4.20
/7-Pentane
2.05
1.57
43
1.56
1.46
1.60
1-Pentene
3.45
7.97
17
4.13
2.15
5.45
cis-2-Pentene
—
—
1
1.58
—
1.58
;ra«.v-2-Pcntene
—
5.21
10
4.68
4.65
4.70
fl-Pincnc
3.80
8.33
23
6.47
9.66
4.88
/>-Pincnc
—
—
—
—
—
—
- = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples and brown shading identifies sites for which replicates
were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-44

-------
Table 31-10. SNMOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
RICO
RICO
# of
Pairs
Average
by
Pollutant
Average
for
Duplicate
Pairs
Average
for
Unequipped
Propane
1.03
0.91
43
0.77
0.84
0.74
w-Propylbenzene
20.65
12.82
15
8.27
6.12
9.35
Propylene
2.11
1.32
43
2.94
3.76
2.62
Propyne
—
—
—
—
—
—
Styrene
4.83
1.06
7
3.03
2.77
3.09
Toluene
2.51
4.26
43
3.06
3.11
3.04
//-Tridccanc
—
—
6
14.90
14.90
—
1-Tridecene
—
—
—
—
—
—
1,2,3-Trimethylbenzene
8.42
5.68
19
8.06
6.67
8.75
1,2,4-Trimethylbenzene
8.39
4.53
40
5.46
4.36
5.90
1,3,5-Trimethylbenzene
8.91
12.51
24
7.39
5.11
8.31
2,2,3 -Trimethylpentane
—
—
11
4.71
6.43
2.99
2,2,4-Trimethylpentane
5.28
2.57
25
3.74
3.55
3.92
2,3,4-Trimethylpentane
9.19
2.69
29
9.11
12.93
7.19
/7-Undecane
—
3.49
22
3.63
5.28
2.80
1-Undecene
—
—
1
0.52
—
0.52
«/-Xylene//?-Xylene
6.49
4.36
43
4.04
2.80
4.53
o-Xylene
11.35
6.12
40
6.13
4.47
6.80
SNMOC (Sum of Knowns)
1.06
0.84
43
0.96
1.45
0.76
Sum of Unknowns
1.97
1.98
43
2.69
4.25
2.06
Average by Site
4.56
3.84
1742
4.20
5.24
3.62
# of Pairs Collected per Site
4
4
- = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples and brown shading identifies sites for which replicates
were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-45

-------
31.3.3 Carbonyl Compound Analytical Precision
Table 31-11 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 2.52 percent, which is well within the
program MQO of 15 percent CV. The analytical precision results from replicate analyses range
from 0 percent (several pollutants at different sites) to 9.44 percent (hexaldehyde for PACO),
indicating that every pollutant-site combination has a CV less than 15 percent.
The pollutant-specific average CV, as shown in orange in Table 31-11, ranges from
0.58 percent (acetone) to 4.16 percent (tolualdehydes). The site-specific average CV, as shown in
green in Table 31-11, ranges from 0.18 percent (BMCO) to 3.83 percent (RICO). Note that the
site-specific average CV for BMCO is based on a single replicate sample.
Sites at which duplicate samples were collected are highlighted in blue in Table 31-11,
sites at which collocated samples were collected are highlighted in purple, and sites at which
replicates were run on individual samples are highlighted in brown. Collocated carbonyl
compound samples were collected at only three of the sites shown in Table 31-11 (DEMI,
INDEM, and PXSS); replicates were run on individual samples for seven sites, and the
remainder of sites collected duplicate samples. The average CV for sites that collected duplicate
samples was calculated and is shown in Table 31-11 in blue, the average CV for sites collecting
collocated samples is shown in purple, and the average CV for sites where replicates were run on
individual samples is shown in brown. The average CV for all three precision types meets the
MQO of 15 percent, with the variability ranging from 2.31 percent (replicates run on individual
samples) to 2.79 percent (replicates run on collocated samples).
31-46

-------
Table 31-11. Carbonyl Compound Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant
Pollutant
ADOK
ASKY
AZFL
BMCO
BRCO
BTUT
CHNJ
CSNJ
Acetaldehyde
0.76
0.58
1.93
0.61
0.81
0.31
0.85
0.62
Acetone
0.79
0.53
1.24
0.16
0.07
0.32
0.57
0.63
Benzaldehyde
4.39
0.00
4.61
0.00
2.43
3.84
3.00
2.91
2-Butanone
1.74
1.26
3.76
0.00
1.43
1.76
2.37
1.97
Butyraldehyde
2.56
1.27
4.59
0.00
3.72
0.63
2.94
0.92
Crotonaldehyde
1.76
1.21
3.80
0.00
7.16
1.67
3.54
2.43
2,5-Dimethylbenzaldehyde
--
--
--
--
--
--
--
--
Formaldehyde
1.76
0.66
1.80
0.99
1.86
0.46
1.41
1.02
Hexaldehyde
3.96
0.00
4.22
0.00
8.20
4.20
3.70
3.55
Isovaleraldehyde
—
—
—
—
—
—
—
—
Propionaldehyde
1.46
2.08
2.25
0.00
5.04
0.47
1.81
2.01
Tolualdehydes
3.03
4.88
4.39
0.00
5.03
3.23
5.98
4.10
Valeraldehyde
4.36
0.00
3.89
—
5.63
3.52
3.34
3.86
Average by Site
2.42
1.13
3.32
0.18
3.76
1.86
2.68
2.18
# of Pairs Collected per Site
6
1
12
1
5
8
12
12
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated from
the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated samples;
brown shading identifies sites for which replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-47

-------
Table 31-11. Carbonyl Compound Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
DEMI
ELNJ
GLKY
GPCO
INDEM
LEKY
NBIL
NBNJ
Acetaldehyde
1.61
0.54
1.04
0.56
1.41
0.58
1.14
1.02
Acetone
0.45
0.37
1.11
0.22
1.03
0.22
0.62
0.74
Benzaldehyde
4.68
3.50
6.30
1.45
4.71
0.00
4.37
3.81
2-Butanone
2.05
2.52
2.54
0.62
2.09
0.47
2.64
1.58
Butyraldehyde
2.22
2.24
2.85
1.04
4.14
1.90
2.33
2.63
Crotonaldehyde
2.37
2.85
4.51
1.20
3.19
2.13
3.89
0.63
2,5-Dimethylbenzaldehyde
--
--
--
--
--
--
--
--
Formaldehyde
0.72
0.96
1.10
0.45
1.02
0.75
1.07
0.96
Hexaldehyde
3.57
3.68
4.47
2.80
4.72
4.35
3.13
5.46
Isovaleraldehyde
—
—
—
—
—
—
—
—
Propionaldehyde
1.73
1.31
2.54
0.80
3.83
0.00
2.39
1.71
Tolualdehydes
4.91
5.35
3.49
2.90
4.11
5.26
5.03
3.77
Valeraldehyde
4.19
2.95
7.33
3.46
4.34
4.88
5.28
6.80
Average by Site
2.59
2.39
3.39
1.41
3.15
1.87
2.90
2.65
# of Pairs Collected per Site
14
12
12
9
22
2
14
12
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated from
the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated samples;
brown shading identifies sites for which replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-48

-------
Table 31-11. Carbonyl Compound Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
OCOK
OR II
PACO
PXSS
RICO
RICO
ROIL
S4MO
Acetaldehyde
0.64
0.68
0.71
1.99
4.29
0.88
0.88
0.62
Acetone
0.55
0.72
0.10
1.80
0.50
0.46
0.65
0.48
Benzaldehyde
3.99
3.50
3.65
2.65
0.00
6.63
1.89
3.88
2-Butanone
1.61
3.64
1.06
1.47
1.72
2.50
3.51
1.57
Butyraldehyde
2.00
3.66
5.20
2.16
--
4.61
3.40
1.55
Crotonaldehyde
2.11
1.38
3.38
2.75
—
3.06
2.90
1.62
2,5-Dimethylbenzaldehyde
--
--
--
--
--
--
--
--
Formaldehyde
1.02
1.27
1.93
1.69
0.00
2.43
0.69
1.69
Hexaldehyde
4.16
4.91
9.44
2.67
0.00
6.32
2.29
3.65
Isovaleraldehyde
—
—
—
—
—
—
—
—
Propionaldehyde
2.61
2.86
3.37
3.85
0.00
4.20
1.82
2.09
Tolualdehydes
4.24
5.27
6.00
3.85
"
3.07
4.05
4.33
Valeraldehyde
4.02
4.53
4.59
4.07
0.00
8.01
3.10
3.36
Average by Site
2.45
2.95
3.59
2.63
0.81
3.83
2.29
2.26
# of Pairs Collected per Site
12
14
5
13
1
5
12
12
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated from
the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated samples;
brown shading identifies sites for which replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-49

-------
Table 31-11. Carbonyl Compound Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
SEWA
SKFL
SPIL
SYFL
TMOK
TOOK
TROK
WPIN
Acetaldehyde
1.90
0.79
0.89
1.80
0.60
0.99
0.60
1.13
Acetone
0.54
0.60
0.50
0.75
0.66
0.58
0.34
0.55
Benzaldehyde
4.18
4.45
3.59
4.33
2.66
5.39
4.26
4.23
2-Butanone
1.73
2.90
1.37
3.53
1.23
1.70
0.62
2.78
Butyraldehyde
2.85
3.78
1.90
3.49
1.78
2.30
2.81
3.89
Crotonaldehyde
3.85
2.37
1.49
1.72
2.17
2.36
1.67
3.32
2,5-Dimethylbenzaldehyde
--
--
--
--
--
--
--
--
Formaldehyde
2.15
1.63
0.45
1.11
0.80
0.81
0.68
0.79
Hexaldehyde
4.48
5.17
3.06
4.04
3.18
4.40
5.10
4.53
Isovaleraldehyde
—
—
—
—
—
—
—
—
Propionaldehyde
4.12
3.49
1.27
2.91
2.32
1.13
1.61
2.65
Tolualdehydes
3.37
4.96
4.07
3.84
3.13
4.74
4.49
3.63
Valeraldehyde
4.07
4.04
3.69
4.85
4.55
4.55
3.04
4.87
Average by Site
3.02
3.11
2.02
2.94
2.10
2.63
2.29
2.94
# of Pairs Collected per Site
12
12
10
12
12
12
14
18
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated from
the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated samples;
brown shading identifies sites for which replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-50

-------
Table 31-11. Carbonyl Compound Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
YUOK
# of
Pairs
Average
by
Pollutant
Average
for
Duplicate
Pairs
Average
for
Collocated
Pairs
Average
for
Unequipped
Acetaldehyde
0.71
338
1.04
0.91
1.67
1.21
Acetone
0.38
338
0.58
0.60
1.10
0.29
Benzaldehyde
4.23
336
3.44
3.86
4.01
1.82
2-Butanone
2.73
335
1.95
2.19
1.87
1.21
Butyraldehyde
3.34
337
2.65
2.59
2.84
2.79
Crotonaldehyde
1.01
337
2.48
2.36
2.77
2.82
2,5-Dimethylbenzaldehyde
--
--
--
--
--
--
Formaldehyde
0.88
338
1.12
1.09
1.14
1.23
Hexaldehyde
4.20
335
3.99
4.01
3.65
4.04
Isovaleraldehyde
--
--
--
--
--
--
Propionaldehyde
1.26
336
2.15
2.04
3.14
2.10
Tolualdehydes
4.64
274
4.16
4.18
4.29
4.04
Valeraldehyde
3.51
322
4.15
4.22
4.20
3.85
Average by Site
2.45
3,626
2.52
2.55
2.79
2.31
# of Pairs Collected per Site
8
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV fortius method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples; brown shading identifies sites for which replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-51

-------
31.3.4 PAH Analytical Precision
Table 31-12 presents analytical precision results from replicate analyses of collocated and
select individual samples as the CV per pollutant per site, the average CV per site, the average
CV per pollutant, and the overall average CV across the PAHs listed. The CV ranges from
0 percent (benzo(e)pyrene for CHSC and benzo(a)anthracene for WADC) to 28.40 percent
(acenaphthylene for ROCH). CVs for only five pollutant-site combinations are greater than
15 percent.
The pollutant-specific average CV, as shown in orange in Table 31-12, ranges from
1.22 percent (phenanthrene) to 8.03 percent (benzo(k)fluoranthene). The site-specific average
CV, as shown in green in Table 31-12, ranges from 1.98 percent (PXSS) to 5.84 percent
(ANAK). The overall average analytical precision CV is 3.91 percent.
Sites at which collocated PAH samples were collected are highlighted in blue in
Table 31-12 while sites at which replicates were run on individual samples are highlighted in
brown. Collocated PAH samples were collected at only ANAK, DEMI, RUCA, SEW A, and
SYFL; replicates were run on individual PAH samples at the remaining sites. The average CV
for sites that collected collocated PAH samples was calculated and is shown in Table 31-12 in
blue while the average CV for sites where replicates were run on individual samples is shown in
brown. The variability ranges from 3.83 percent (replicates run on individual samples) to
4.15 percent (replicates run on collocated samples).
Table 31-13 presents analytical precision results for the 12 replicate analyses of select
individual samples for KMMS, the only site for which PAH/Phenol samples were collected. The
pollutant-specific average CV ranges from 0.96 percent (benzo(a)anthracene) to 8.93 percent
(anthracene), with a site-specific average CV for KMMS of 3.16 percent.
31-52

-------
Table 31-12. PAH Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant
Pollutant
ANAK
BOMA
BTUT
BXNY
CELA
CHSC
DEMI
Acenaphthene
6.70
3.61
3.42
7.03
4.91
4.73
5.22
Acenaphthylene
5.52
4.39
2.54
4.23
1.71
—
7.52
Anthracene
8.61
2.70
3.39
2.28
2.02
4.38
10.37
Benzo(a)anthracene
4.62
2.85
0.18
2.50
0.84
--
2.63
Benzo(a)pyrene
4.71
3.77
0.36
4.60
9.04
--
5.26
Benzo(b)fluoranthene
5.44
5.05
5.89
2.03
8.70
1.41
3.09
Benzo(e)pyrene
6.14
1.87
2.88
3.95
3.68
0.00
3.44
Benzo(g,h,i)perylene
5.42
6.59
9.12
2.30
4.20
2.01
4.52
Benzo(k)fluoranthene
13.93
17.50
9.37
5.95
3.84
--
4.78
Chrysene
8.90
2.19
9.81
1.91
3.07
9.71
1.36
Coronene
2.16
0.47
4.09
7.03
3.71
--
11.06
Cvc lope nta | cd | pv rcne
7.65
--
1.51
4.54
6.91
--
2.38
Dibenz(a,h)anthracene
3.81
--
--
--
3.04
--
4.60
Fluoranthene
4.98
3.34
1.81
5.49
4.37
3.42
2.55
Fluorene
4.56
2.67
6.41
6.38
3.31
2.85
1.63
9-Fluorenone
4.72
2.42
1.29
6.10
2.46
2.13
2.19
Indeno( 1,2,3 -cd)pyrene
--
--
--
--
--
--
--
Naphthalene
4.95
4.19
5.95
3.87
1.98
1.08
3.25
Perylene
5.49
5.79
5.56
3.55
0.76

9.93
Phenanthrene
4.46
1.18
0.91
0.46
0.95
1.02
1.39
Pyrene
4.43
3.98
1.12
5.48
4.07
3.06
2.73
Retene
5.50
7.62
1.36
7.53
4.98
1.97
5.62
Average by Site
5.84
4.33
3.85
4.36
3.74
2.91
4.55
# of Pairs Collected per Site
18
4
4
4
4
6
15
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV fortius method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting collocated samples and brown shading identifies sites for which replicates
were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-53

-------
Table 31-12. PAH Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
GLKY
GPCO
LBHCA
NBIL
PRRI
PXSS
RIVA
Acenaphthene
4.25
3.39
2.59
6.66
3.03
5.60
1.75
Acenaphthylene
10.46
1.94
—
5.07
5.21
0.77
—
Anthracene
8.61
2.95
6.35
1.22
7.33
2.45
6.69
Benzo(a)anthracene
0.22
0.88
1.97
1.09
6.21
3.92
--
Benzo(a)pyrene
5.97
3.35
0.66
2.51
3.72
1.13
--
Benzo(b)fluoranthene
4.00
4.25
1.14
4.75
2.54
2.21
6.04
Benzo(e)pyrene
5.04
2.75
2.31
0.81
2.26
1.11
--
Benzo(g,h,i)perylene
2.02
4.05
5.66
0.77
3.38
0.16
--
Benzo(k)fluoranthene
11.01
5.84
--
1.51
9.25
1.78
--
Chrysene
4.57
2.10
6.73
5.68
3.22
3.50
12.43
Coronene
--
1.98
--
0.42
5.22
1.02
--
Cvc lope nta | cd | pv rcne
1.41
6.84
--
--
--
2.42
--
Dibenz(a,h)anthracene
--
7.16
3.81
--
2.76
0.96
--
Fluoranthene
2.24
2.96
2.07
5.42
2.90
0.99
2.09
Fluorene
2.00
2.05
1.97
4.23
1.47
0.52
0.84
9-Fluorenone
1.37
2.62
2.30
5.73
1.27
1.22
1.24
Indeno( 1,2,3 -cd)pyrene
--
--
--
--
--
--
--
Naphthalene
2.14
2.88
5.15
3.49
4.35
1.23
1.27
Perylene
--
5.72
4.23
--
3.16
5.40
--
Phenanthrene
0.78
1.15
1.50
1.96
0.92
0.80
1.25
Pyrene
2.53
3.16
3.31
4.61
2.97
1.30
3.29
Retene
5.09
5.38
1.63
—
3.77
3.14
3.31
Average by Site
4.10
3.49
3.14
3.29
3.75
1.98
3.65
# of Pairs Collected per Site
5
8
3
4
6
4
4
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV fortius method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting collocated samples and brown shading identifies sites for which replicates
were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-54

-------
Table 31-12. PAH Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
ROCH
RUCA
S4MO
SEWA
SJJCA
SKFL
SYFL
Acenaphthene
2.29
9.53
1.28
2.54
3.71
4.77
1.37
Acenaphthylene
28.40
3.16
1.56
3.93
4.37
—
5.52
Anthracene
3.22
7.71
10.51
3.98
4.29
5.78
2.14
Benzo(a)anthracene
1.27
1.04
4.80
0.89
1.45
--
--
Benzo(a)pyrene
0.59
2.95
9.30
3.57
3.99
--
--
Benzo(b)fluoranthene
0.86
3.76
5.15
3.08
5.87
2.05
2.37
Benzo(e)pyrene
3.61
3.62
11.02
1.71
2.93
--
4.05
Benzo(g,h,i)perylene
6.09
3.24
8.40
5.67
3.67
6.15
4.38
Benzo(k)fluoranthene
18.57
0.26
6.35
8.42
15.13
--
--
Chrysene
2.37
4.11
3.49
2.69
4.28
0.70
7.15
Coronene
--
4.69
2.20
1.45
3.73
--
--
Cvc lope nta | cd | pv rcne
--
4.60
3.49
3.82
--
--
--
Dibenz(a,h)anthracene
--
8.72
11.92
2.58
--
--
--
Fluoranthene
0.34
2.22
1.40
2.37
1.39
2.97
1.00
Fluorene
1.07
1.75
2.21
3.17
15.84
2.54
1.14
9-Fluorenone
2.51
2.13
2.25
2.86
2.34
3.32
1.94
Indeno( 1,2,3 -cd)pyrene
--
--
--
--
--
--
--
Naphthalene
1.95
5.78
4.47
3.52
4.73
3.49
3.67
Perylene
--
0.41
--
4.48
--
--
--
Phenanthrene
0.91
0.96
2.30
0.98
0.48
0.79
0.86
Pyrene
1.24
2.48
2.73
2.55
1.99
1.87
2.30
Retene
10.48
3.39
4.21
2.92
3.03
2.59
2.97
Average by Site
5.05
3.64
4.95
3.20
4.62
3.09
2.92
# of Pairs Collected per Site
3
13
7
14
4
1
7
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV fortius method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting collocated samples and brown shading identifies sites for which replicates
were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-55

-------
Table 31-12. PAH Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
UNVT
WADC
WPFL
# of
Pairs
Average
by
Pollutant
Average for
Collocated
Pairs
Average
for
Unequipped
Acenaphthene
10.64
1.76
1.92
148
4.28
5.07
4.07
Acenaphthylene
5.50
7.72
1.48
92
5.55
5.13
5.69
Anthracene
3.10
4.62
2.56
119
4.89
6.56
4.45
Benzo(a)anthracene
--
0.00
4.11
57
2.18
2.30
2.15
Benzo(a)pyrene
--
2.69
--
68
3.79
4.12
3.69
B enzo (b )fluoranthene
2.16
4.87
2.96
115
3.74
3.55
3.79
Benzo(e)pyrene
--
1.61
--
81
3.24
3.79
3.06
Benzo(g,h,i)perylene
--
3.41
--
97
4.34
4.65
4.25
B enzo (k)fluoranthene
--
3.01
--
53
8.03
6.85
8.39
Chrysene
4.49
2.50
3.40
146
4.60
4.84
4.53
Coronene
--
1.74
--
50
3.40
4.84
2.87
Cvclopcnta|cd|pvrcnc
--
--
--
29
4.14
4.61
3.87
Dibenz(a,h)anthracene
--
7.13
--
15
5.13
4.93
5.25
Fluoranthene
5.55
4.20
2.61
151
2.86
2.62
2.92
Fluorene
0.17
2.32
1.54
131
3.03
2.45
3.18
9-Fluorenone
4.08
2.60
1.82
151
2.62
2.77
2.58
Indeno(l,2,3-cd)pyrene
--
--
--
--
--
--
--
Naphthalene
2.09
3.06
2.39
151
3.37
4.24
3.15
Perylene
--
--
--
21
4.54
5.08
4.27
Phenanthrene
0.89
0.98
1.37
151
1.22
1.73
1.08
Pyrene
4.05
3.82
1.69
151
2.95
2.90
2.96
Retene
—
3.43
2.33
96
4.19
4.08
4.23
Average by Site
3.88
3.24
2.32
2,073
3.91
4.15
3.83
# of Pairs Collected per Site
5
5
3
- = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated from
the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting collocated samples and brown shading identifies sites for which replicates were
run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-56

-------
Table 31-13. PAH/Phenols Analytical Precision: Coefficient of Variation
Based on Replicate Analyses for KMMS
Pollutant
KMMS
# of
Pairs
Average
by
Pollutant
Acenaphthene
2.41
12
2.41
Anthracene
8.93
7
8.93
Benzo(a)anthracene
0.96
1
0.96
Benzo(a)pyrene
--
--
--
Benzo(b)fluoranthene
2.18
1
2.18
Benzo(g,hi)perylene
—
—
—
Benzo(k)fluoranthene
--
--
--
Chrysene
1.43
1
1.43
«/,£>-Cresols
1.57
12
1.57
o-Cresol
2.58
12
2.58
Dibenz(a,h)anthracene
--
--
--
Fluoranthene
4.94
12
4.94
Fluorene
3.29
12
3.29
Indeno( 1,2,3 -cd)pyrene
—
—
—
Naphthalene
1.48
12
1.48
Phenanthrene
3.41
12
3.41
Phenol
3.61
12
3.61
Pyrene
4.30
12
4.30
Average by Site
3.16
118
3.16
# of Pairs Collected per Site
12
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV.
31.3.5 Metals Analytical Precision
Table 31-14 presents analytical precision results from replicate analyses of collocated and
select individual samples as the CV per pollutant per site, the average CV per site, the average
CV per pollutant, and the overall average CV across the metals listed. The CVs exhibit low- to
mid-level variability, ranging from 0 percent (for several sites and pollutants) to 31.19 percent
(selenium for BTUT).
The pollutant-specific average CV, as shown in orange in Table 31-14, ranges from
1.42 percent (manganese) to 14.07 percent (mercury). The site-specific average CV, as shown in
green in Table 31-14, ranges from 3.09 percent (TOOK) to 9.59 percent (BTUT). The overall
average analytical precision CV is 5.98 percent.
31-57

-------
Sites at which collocated metals samples were collected are highlighted in blue in
Table 31-14 while sites at which replicates were run on individual samples are highlighted in
brown. Collocated metals samples were collected at eight sites; replicates were run on individual
PAH samples at the remaining 10 sites. The average CV for sites that collected collocated metals
samples was calculated and is shown in Table 31-14 in blue while the average CV for sites
where replicates were run on individual samples is shown in brown. The variability ranges from
5.48 percent (replicates run on individual samples) to 6.54 percent (replicates run on collocated
samples).
Table 31-14. Metals Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant
Pollutant
ADOK
ASKY-M
BAKY
BOMA
BTUT
Antimony
1.37
0.98
1.20
1.22
1.44
Arsenic
1.76
6.16
11.01
18.98
7.87
Beryllium
8.89
17.67
17.38
14.14
24.59
Cadmium
6.66
4.15
7.34
8.68
5.95
Chromium
—
—
—
—
—
Cobalt
7.38
1.98
5.25
3.93
2.35
Lead
4.64
0.60
0.55
0.76
0.66
Manganese
2.26
0.51
0.98
1.13
1.74
Mercury
8.72
15.18
30.91
20.75
17.38
Nickel
—
1.04
4.60
2.94
2.70
Selenium
1.70
4.03
11.97
11.75
31.19
Average by Site
4.82
5.23
9.12
8.43
9.59
# of Pairs Collected per Site
2
12
6
57
12
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV fortius method is
calculated from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting collocated samples and brown shading identifies sites for which
replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-58

-------
Table 31-14. Metals Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
CCKY
GLKY
LEKY
NBIL
OCOK
Antimony
2.02
2.00
0.55
5.16
2.60
Arsenic
16.87
18.55
12.40
4.25
3.43
Beryllium
0.00
0.00
—
7.67
4.80
Cadmium
15.36
10.86
8.36
2.61
11.82
Chromium
—
—
—
2.98
—
Cobalt
0.00
8.81
0.00
2.05
4.12
Lead
1.32
1.62
0.34
1.77
2.75
Manganese
0.48
1.45
0.53
1.28
1.89
Mercury
16.33
20.83
0.00
8.33
11.15
Nickel
5.24
15.11
3.53
—
—
Selenium
6.91
9.45
4.81
2.42
1.36
Average by Site
6.45
8.87
3.39
3.85
4.88
# of Pairs Collected per Site
4
60
4
24
4
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV fortius method is
calculated from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting collocated samples and brown shading identifies sites for which
replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-59

-------
Table 31-14. Metals Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
PAFL
PXSS
S4MO
SEWA
SJJCA
TMOK
Antimony
0.59
0.20
1.33
1.40
0.92
5.56
Arsenic
3.42
18.60
13.75
22.12
13.09
5.31
Beryllium
15.71
12.61
7.56
—
0.00
10.41
Cadmium
11.96
2.95
4.35
7.64
13.23
9.59
Chromium
0.88
—
—
—
—
0.31
Cobalt
14.23
1.62
6.64
2.04
0.00
12.85
Lead
1.08
0.61
0.88
1.66
0.34
3.82
Manganese
2.84
1.24
0.95
1.20
0.56
3.67
Mercury
11.07
14.29
20.70
17.89
6.28
9.80
Nickel
—
1.86
4.87
2.72
1.63
0.44
Selenium
3.51
5.34
7.18
12.06
3.79
3.28
Average by Site
6.53
5.93
6.82
7.64
3.98
5.91
# of Pairs Collected per Site
6
4
116
5
5
4
- = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting collocated samples and brown shading identifies sites for which replicates
were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-60

-------
Table 31-14. Metals Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
TOOK
UNVT
# of
Pairs
Average
by
Pollutant
Average
for
Collocated
Pairs
Average
for
Unequipped
Antimony
1.46
1.59
443
1.75
1.90
1.64
Arsenic
2.07
20.88
399
11.14
11.56
10.80
Beryllium
6.77
—
173
9.88
11.20
8.73
Cadmium
3.45
11.98
445
8.16
6.50
9.49
Chromium
1.92
—
78
1.52
2.45
0.59
Cobalt
2.22
12.12
415
4.87
5.01
4.75
Lead
1.78
0.98
445
1.45
1.13
1.71
Manganese
1.71
1.13
445
1.42
1.24
1.56
Mercury
7.33
16.33
253
14.07
15.85
12.64
Nickel
3.08
3.43
318
3.80
4.74
2.86
Selenium
2.23
14.94
330
7.66
10.40
5.47
Average by Site
3.09
9.26
3,744
5.98
6.54
5.48
# of Pairs Collected per Site
108
12
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV fortius method is calculated
from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting collocated samples and brown shading identifies sites for which replicates
were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31.3.6 Hexavalent Chromium Analytical Precision
Table 31-15 presents analytical precision results from replicate analyses of collocated and
select individual samples as the CV per site and the overall average CV for hexavalent
chromium. The site-specific CV ranges from 3.89 percent (SEWA) to 18.13 percent (RIVA).
CVs for only two of the 16 sites collecting hexavalent chromium samples are greater than
15 percent (PRRI and RIVA). The overall average analytical precision of hexavalent chromium
is 8.53 percent, as shown in orange in Table 31-15. Note, however, that most sites sampling
hexavalent chromium discontinued sampling this pollutant at the end of July 2013 and many of
the precision calculations are based on relatively few pairs; precision for eight of the 16 sites is
based on three or fewer samples.
31-61

-------
Sites at which collocated hexavalent chromium samples were collected are highlighted in
blue in Table 31-15 while sites at which replicates were run on individual samples only are
highlighted in brown. Collocated samples were collected at 14 of the 16 sites listed in
Table 31-15; replicates were run only on individual samples at MIWI and GLKY. The average
CV for sites that collected collocated samples was calculated and is shown in Table 31-15 in blue
while the average CV for sites where replicates were run on individual samples is shown in
brown. The variability ranges from 7.98 percent (replicates run on individual samples) to
8.61 percent (replicates run on collocated samples).
Table 31-15. Hexavalent Chromium Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site
Site
CV
(%)
# of
Pairs
BOMA
8.84
4
BTUT
4.09
2
BXNY
8.08
6
CAMS 35
4.67
6
DEMI
8.46
4
GLKY
12.05
2
GPCO
5.29
2
MIWI
3.91
1
NBIL
5.11
2
PRRI
15.58
2
PXSS
13.01
6
RIVA
18.13
4
S4MO
6.10
8
SDGA
8.09
3
SEWA
3.89
2
SKFL
11.19
4
# of Pairs
58
Average CV
8.53
Average CV
8.61
Average CV
7.98
Bold = CV greater than 15 percent
Orange shading indicates the overall average CV for this method.
Blue shading identifies sites collecting collocated samples and brown shading
identifies sites for which replicates were run on individual samples.
BOLD ITALICS = EPA-designated NATTS Site
31-62

-------
31.4 Accuracy
Laboratories typically evaluate their accuracy (or bias) by analyzing audit samples that
are prepared by an external source. The pollutants and the respective concentrations of the audit
samples are unknown to the laboratory. The laboratory analyzes the samples and the external
source compares the measured concentrations to the reference concentrations of those audit
samples and calculates a percent difference. Accuracy, or bias, indicates the extent to which
experimental measurements represent their corresponding "true" or "actual" values.
Laboratories participating in the NATTS program are provided with proficiency test (PT)
audit samples for VOCs, carbonyl compounds, PAHs, metals, and hexavalent chromium, which
are used to quantitatively measure analytical accuracy. Tables 31-16 through 31-20 present
ERG's results for PT audit samples analyzed in 2013. The program MQO for the percentage of
the true value is ± 25 percent recovery, and the values exceeding this criterion are bolded in the
tables. The calculation is as follows:
Percent of True (% Recovery) = —— x 100
^true
Where:
Xiab is the analytical result from the laboratory;
Xtrue is the true concentration of the audit sample.
Note that the "true" value is based on the mean value of the referee laboratory's results.
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 74 results provided in Tables 31-16 through
Table 31-20, only five exceed the ± 25 percent recovery MQO (two for VOCs and three for
PAHs).
31-63

-------
Table 31-16. TO-15 NATTS PT Audit Samples - Percent of True Value
Pollutant
January 2013
July 2013
Acrolein
87.7
91.8
Benzene
94.5
101.9
1.3 -Butadiene
96.1
114.0
Carbon Tetrachloride
108.0
126.6
Chloroform
91.9
112.9
1,2-Dibromoethane
90.1
80.4
cis-1,3 -Dichloropropene
107.0
83.1
trans-1,3 -Dichloropropene
96.6
84.7
1,2-Dichloropropane
96.3
105.9
1,2-Dichloroethane
101.1
108.5
Dichloromethane
94.7
165.0
1.1,2,2-Tetrachloroethane
91.5
87.7
Tetrachloroethylene
91.0
89.4
Trichloroethylene
103.3
102.4
Vinyl chloride
88.4
101.2
Table 31-17. TO-11A NATTS PT Audit Samples - Percent of True Value
Pollutant
January 2013
July 2013
Acetaldehyde
96.4
103.8
Benzaldehyde
100.7
107.7
Formaldehyde
94.7
104.4
Propionaldehyde
86.0
97.1
Table 31-18. TO-13A NATTS PT Audit Sample - Percent of True Value
Pollutant
May 2013
December 2013
Acenaphthene
96.3
121.9
Anthracene
129.5
140.6
Benzo(a)pyrene
93.8
95.6
Fluoranthene
114.1
112.7
Fluorene
109.0
153.8
Naphthalene
124.6
NS
Phenanthrene
87.9
NS
Pyrene
114.5
112.3
NS = Not spiked onto PT audit sample provided to the laboratory
31-64

-------
Table 31-19. Metals NATTS PT Audit Sample - Percent of True Value
Pollutant
May 2013
November 2013
Antimony
82.1
64.9
Arsenic
97.0
NS
Beryllium
98.0
90.1
Cadmium
101.4
98.4
Cobalt
102.4
91.3
Lead
100.1
94.8
Manganese
93.8
100.4
Nickel
98.8
NS
Selenium
84.4
NS
NS = Not spiked onto PT audit sample provided to the laboratory
Table 31-20. Hexavalent Chromium NATTS PT Audit Samples - Percent of True Value
Pollutant
May 2013
November 2013
Hexavalent Chromium
93.5
123.0
ERG's use of the ICP/MS was approved in 2012 as a FEM for the sampling and analysis
of lead for adherence to the National Ambient Air Quality Standards (NAAQS) (EPA 2012a).
This approval requires additional quality assurance steps, including the analysis of quarterly
audit strips. Table 31-21 provides the results of the quarterly NAAQS audit results for lead for
ERG. All results are within the ± 25 percent MQO.
Table 31-21. Metals NAAQS PT Audit Samples - Percent of True Value
Pollutant
Filter #
Analysis #
March 2013
May 2013
August 2013
December 2013


1
84.3
99.3
98.5
98.6

1
2
85.2
98.3
94.6
102.0
Lead

3
89.9
100.5
94.2
102.7


1
87.8
98.6
99.1
98.7

2
2
88.2
97.8
96.1
98.6


3
89.6
93.7
96.8
97.5
31-65

-------
The accuracy of the 2013 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 2013 monitoring effort have
been approved by EPA for accurately measuring ambient levels of various
pollutants - an approval that is based on many years of research into the
development of ambient air monitoring methodologies.
•	When collecting and analyzing ambient air samples, field sampling staff and
laboratory analysts are required to strictly adhere to quality control and quality
assurance guidelines detailed in the respective monitoring methods. This strict
adherence to the well-documented sampling and analytical methods suggests that
the 2013 monitoring data accurately represent ambient air quality.
31-66

-------
32.0	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. NMP data should
ultimately assist a wide range of audiences understand the complex nature of air pollution.
32.1	Summary of Results
Analyses of the 2013 monitoring data identified the following notable results,
observations, trends, and patterns in the program-level and state- and site-specific air monitoring
data.
32.1.1 Program-level Results Summary
•	Number of participating sites. Twenty-five of the 66 monitoring sites are EPA-
designated NATTS sites. An additional 39 UATMP sites participated in the NMP in
2013. Data from two CSATAM sites (ANAK and LBHCA) are also included in the
2013 NMP report.
•	Total number of samples collected and analyzed. Over 9,400 valid samples were
collected by program participants and analyzed at the ERG laboratory, yielding nearly
263,000 valid measurements of air toxics, including duplicate, collocated, and
replicate results.
•	Detects. Of the 175 pollutants monitored, 165 pollutants were detected at least once
over the course of the 2013 monitoring effort. The detection of a given pollutant is
subject to the sensitivity limitation associated with the analytical methods used and
the limitations of the instruments. Simply stated, an MDL is the lowest concentration
of a target pollutant that can be measured and reported with 99 percent confidence
that the pollutant concentration is greater than zero. Approximately 53 percent of the
reported measurements were greater than the associated MDLs. At the method level,
this percentage varies considerably, from 39 percent for hexavalent chromium to 83
percent for carbonyl compounds. Quantification below the MDL is possible and an
acceptable analytical result; therefore, these results are incorporated into the data
analyses. These measurements account for 9 percent of concentrations. Non-detects
account for the remaining 38 percent of results.
•	Program-level Pollutants of Interest. The pollutants of interest at the program-level
are based on the total number of concentrations greater than the risk screening value,
or those "failing the screen". Thirty-eight pollutants failed at least one risk screening
value; of those pollutants, 13 were identified as program-level pollutants of interest.
32-1

-------
•	Mobile Sources. Site-specific hydrocarbon concentrations had virtually no
correlations with county-level motor vehicle ownership data, traffic data, and VMT
data in this year's report. This is a slight change from previous years' reports when
most of the correlations were positive, albeit weak.
•	Seasonal Trends. Formaldehyde concentrations tended to be highest during the
warmer months of the year. Acenaphthene and acetaldehyde concentrations exhibit a
similar pattern. Conversely, benzene and 1,3-butadiene concentrations tended to be
higher during the colder months of the year.
32.1.2 State-level Results Summary
Alaska.
•	The Alaska monitoring site (ANAK) is located in Anchorage and is a CSATAM site.
•	VOCs and PAHs were sampled for at ANAK.
•	Twelve pollutants failed screens for ANAK, of which eight were identified as
pollutants of interest. Benzene and carbon tetrachloride were detected in every valid
sample and failed 100 percent of screens.
•	Of the pollutants of interest for ANAK, benzene has the highest annual average
concentration and is the only pollutant with an annual average greater than 1 |ig/m3,
•	ANAK has the second highest annual average concentrations of benzene and
ethylbenzene among NMP sites sampling this pollutant.
•	Benzene has the highest cancer risk approximation for ANAK, followed by
1,3-butadiene and carbon tetrachloride. None of the noncancer hazard approximations
for ANAK are greater than 1.0 in-a-million.
•	Formaldehyde is the highest emitted pollutant with a cancer toxicity factor in
Anchorage, and also has the highest cancer toxicity-weighted emissions. Toluene is
the highest emitted pollutant with a noncancer toxicity factor in Anchorage, while
acrolein has the highest noncancer toxicity-weighted emissions.
Arizona.
•	The Arizona monitoring sites are located in Phoenix. PXSS is a NATTS site; SPAZ is
a UATMP site.
•	VOCs, carbonyl compounds, PAHs, metals (PMio), and hexavalent chromium were
sampled for at PXSS, although hexavalent chromium sampling was discontinued in
June. VOCs were sampled for at SPAZ.
•	Eighteen pollutants failed screens for PXSS, 10 of which contributed to 95 percent of
failed screens. PXSS failed the second highest number of screens among all NMP
32-2

-------
sites. Six pollutants failed screens for SPAZ, all of which contributed to 95 percent of
failed screens.
•	Of the pollutants of interest for PXSS, formaldehyde has the highest annual average
concentration, followed by acetaldehyde and benzene. These are the only pollutants
of interest with annual average concentrations greater than 1 |ig/m3.
•	Benzene has the highest annual average concentration for SPAZ, and is the only
pollutant with an annual average concentration greater than 1 |ig/m3.
•	SPAZ and PXSS have the highest annual average concentrations of
/;-dichlorobenzene among NMP sites sampling this pollutant.
•	Sampling for the site-specific pollutants of interest has occurred at PXSS and SPAZ
for at least 5 consecutive years; thus, a trends analysis was conducted for each site for
the site-specific pollutants of interest. Benzene concentrations measured at both sites
have decreased over recent years. Arsenic concentrations have also decreased at
PXSS in recent years. The detection rate of 1,2-dichloroethane at both sites has been
steadily increasing over the years, with a significant increase for 2012, which
continued in 2013.
•	Formaldehyde has the highest cancer risk approximation for PXSS and is the only
pollutant of interest with a cancer risk approximation greater than 10 in-a-million for
either site. Benzene has the highest cancer risk approximation for SPAZ. None of the
pollutants of interest for either site have a noncancer hazard approximation greater
than an HQ of 1.0.
•	Benzene is the highest emitted pollutant with a cancer toxicity factor in Maricopa
County, while toluene is the highest emitted pollutant with a noncancer toxicity
factor. Formaldehyde has the highest cancer toxicity-weighted emissions, while
acrolein has the highest noncancer toxicity-weighted emissions for Maricopa County.
California.
•	The 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.
•	PAHs were sampled for at CELA, LBHCA, and RUCA. PAHs and metals (PMio)
were sampled for at SJJCA. Sampling at LBHCA was discontinued in July 2013 at
the end of a 1-year monitoring effort.
•	Naphthalene failed the majority of screens for CELA, LBHCA, and RUCA.
Naphthalene and arsenic contributed almost equally to the total number of failed
screens for SJJCA, together accounting for more than 85 percent of failed screens for
the site.
32-3

-------
•	Naphthalene has the highest annual average concentration for CELA, RUCA, and
SJJCA. Annual average concentrations could not be calculated for LBHCA because
sampling ended in July.
•	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 fluorene and
naphthalene decreased significantly from 2012 levels at CELA. Concentrations of the
pollutants of interest for RUCA have not changed significantly over the last few
years. Concentrations of arsenic have a slight increasing trend at SJJCA over the last
several years, particularly for 2013.
•	Of the pollutants of interest for each site, naphthalene has the highest cancer risk
approximation for all three California sites. The noncancer hazard approximations for
each pollutant of interest are considerably less than an HQ of 1.0 for all three sites.
Cancer risk and noncancer hazard approximations could not be calculated for
LBHCA.
•	Formaldehyde is the highest emitted pollutant with a cancer toxicity factor in Los
Angeles and Riverside Counties, while benzene is the highest emitted pollutant with a
cancer toxicity factor in Santa Clara County. Formaldehyde has the highest cancer
toxicity-weighted emissions for all three counties.
•	Toluene is the highest emitted pollutant with a noncancer toxicity factor in Los
Angeles, Riverside, and Santa Clara Counties, while acrolein has the highest
noncancer toxicity-weighted emissions for all three counties.
Colorado.
•	The NATTS site in Colorado is located in Grand Junction (GPCO). There are also
five UATMP sites located northeast of Grand Junction in Garfield County. The sites
are located in the towns of Battlement Mesa (BMCO), Silt (BRCO), Parachute
(PACO), Carbondale (RFCO), and Rifle (RICO).
•	VOCs, carbonyl compounds, PAHs, and hexavalent chromium were sampled for at
GPCO, although hexavalent chromium sampling was discontinued in June. SNMOCs
and carbonyl compounds were sampled for at the Garfield County sites.
•	Fifteen pollutants failed at least one screen for GPCO, 12 of which contributed to
95 percent of failed screens. Five pollutants failed screens for four of the Garfield
County sites (BMCO, BRCO, PACO, and RICO), while four pollutants failed screens
for RFCO. Benzene, formaldehyde, and acetaldehyde were identified as pollutants of
interest for all five Garfield County sites as well as GPCO.
•	Of the pollutants of interest for GPCO, formaldehyde has the highest annual average
concentration, followed by acetaldehyde.
32-4

-------
•	Benzene has the highest annual average concentration for each of the Garfield County
sites except RFCO, where formaldehyde has the highest annual average
concentration. Annual average concentrations could not be calculated for carbonyl
compounds for RICO due to low sampling completeness.
•	PACO has the highest annual average concentration of benzene among NMP sites.
GPCO has the second highest annual average concentrations of acetaldehyde,
formaldehyde, and naphthalene among NMP sites sampling these pollutants.
•	Sampling for the site-specific pollutants of interest has occurred at GPCO, BRCO,
PACO, and RICO for at least 5 consecutive years; thus, a trends analysis was
conducted for the site-specific pollutants of interest. Benzene concentrations at GPCO
have an overall decreasing trend across the years of sampling. Conversely,
acetaldehyde concentrations at GPCO have been increasing in recent years. 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. Distinct trends
were difficult to identify for the Garfield County sites due to variability in the
measurements and annual averages that could not be calculated for one or more years.
However, benzene concentrations measured at BRCO, PACO, and RICO exhibit a
decreasing trend through 2012 followed by a considerable increase for 2013.
•	Formaldehyde has the highest cancer risk approximation for GPCO (by an order of
magnitude) and is the third highest cancer risk approximation calculated across the
program for 2013. Formaldehyde and benzene have the highest cancer risk
approximation for each of the Colorado sites, where an annual average could be
calculated. All noncancer hazard approximations are less than an HQ of 1.0 for all of
the Colorado sites, where noncancer hazard approximations could be calculated.
•	Benzene is the highest emitted pollutant with a cancer toxicity factor in both Mesa
and Garfield Counties, while formaldehyde has the highest cancer toxicity-weighted
emissions for both counties.
•	While toluene is the highest emitted pollutant with a noncancer toxicity factor for
both Mesa and Garfield Counties, acrolein has the highest noncancer toxicity -
emissions for both counties.
District of Columbia.
•	The Washington, D.C. monitoring site (WADC) is a NATTS site.
•	PAHs and hexavalent chromium were sampled for at WADC, although hexavalent
chromium sampling was discontinued in June.
•	Naphthalene accounted for nearly 97 percent of failed screens for this site and was the
only pollutant identified as a pollutant of interest.
32-5

-------
•	Naphthalene was detected in every valid PAH sample collected at WADC. The
annual average concentration of naphthalene for WADC is the ninth highest annual
average concentration among NMP sites sampling this pollutant.
•	Sampling for the site-specific pollutants of interest has occurred at WADC for at least
5 consecutive years; thus, a trends analysis was conducted for the site-specific
pollutants of interest. Concentrations of naphthalene have decreased since 2009 at
WADC.
•	The cancer risk approximation for naphthalene is 2.83 in-a-million. The noncancer
hazard approximation for naphthalene is considerably less than an HQ of 1.0.
•	Benzene is the highest emitted pollutant with a cancer toxicity factor in the District of
Columbia, while toluene is the highest emitted pollutant with a noncancer toxicity
factor. Formaldehyde has the highest cancer toxicity-weighted emissions, while
acrolein has the highest noncancer toxicity-weighted emissions in the District.
Florida.
•	Three of the Florida monitoring sites are located in the Tampa-St. Petersburg-
Clearwater CBS A (SYFL, AZFL, and SKFL) and two are located in the Orlando-
Kissimmee-Sanford CBSA (ORFL and PAFL). WPFL is located on the western edge
of the Miami-Ft. Lauderdale-West Palm Beach CBSA. SKFL and SYFL are NATTS
sites while the other four are UATMP sites.
•	Carbonyl compounds were sampled for at AZFL and ORFL. Hexavalent chromium
and PAHs were sampled for at SKFL and SYFL in addition to carbonyl compounds.
However, hexavalent chromium sampling was discontinued at SKFL in June and
sampling for both PAHs and hexavalent chromium was discontinued at SYFL in
June. Metals (PMio) were sampled for at PAFL. PAHs were sampled for at WPFL
during a 1-year monitoring effort between March 2013 and March 2014. The 2014
data for WPFL are included in the Florida section but excluded from most program-
level analyses.
•	Acetaldehyde and formaldehyde failed screens for all four Florida sites sampling
carbonyl compounds. Naphthalene failed screens for all three Florida sites that
sampled PAHs. Three additional PAHs failed screens for WPFL. Hexavalent
chromium failed a single screen for SKFL. Arsenic failed 100 percent of screens for
PAFL.
•	Formaldehyde has the highest annual average concentration for the four Florida sites
sampling carbonyl compounds, although the annual average concentrations of
acetaldehyde were just slightly lower. Naphthalene has the highest annual average
concentration for WPFL. PAFL's annual average arsenic concentration ranks sixth
highest among NMP sites sampling metals (PMio).
•	Sampling for the site-specific pollutants of interest has occurred at all of the Florida
sites except WPFL for at least 5 consecutive years; thus, a trends analysis was
32-6

-------
conducted for the site-specific pollutants of interest. The following notable
observations regarding trends include: While acetaldehyde concentrations have been
variable at AZFL, formaldehyde concentrations have changed little over the last few
years. Acetaldehyde concentrations have leveled off after decreasing over the last few
years at SKFL. Formaldehyde concentrations decreased considerably at SKFL from
2012 to 2013. Formaldehyde concentrations measured in 2013 at SYFL exhibit the
least amount of variability over the 10 years of sampling. Formaldehyde
concentrations have changed little at ORFL in the last few years while acetaldehyde
concentrations have exhibited more variability.
•	For the four Florida sites sampling carbonyl compounds, formaldehyde has the
highest cancer risk approximations, ranging from 20 in-a-million to 25 in-a-million.
The cancer risk approximation for arsenic for PAFL is 3.10 in-a-million. Naphthalene
has the highest cancer risk approximation for WPFL, although this was one of the
lowest cancer risk approximations for naphthalene among NMP sites sampling PAHs.
All noncancer hazard approximations for the pollutants of interest for the Florida sites
are less than an HQ of 1.0.
•	Benzene is the highest emitted pollutant with a cancer toxicity factor in Pinellas,
Hillsborough, and Orange Counties, while formaldehyde is the highest emitted
pollutant with a cancer toxicity factor in Palm Beach County. Benzene has the highest
cancer toxicity-weighted emissions for Pinellas County; formaldehyde has the highest
cancer toxicity-weighted emissions for Hillsborough and Palm Beach Counties; and
hexavalent chromium has the highest cancer toxicity-weighted emissions for Orange
County.
•	Toluene is the highest emitted pollutant with a noncancer toxicity factor in all four
Florida counties. Acrolein has the highest noncancer toxicity-weighted emissions for
all four counties.
Georgia.
•	The SDGA monitoring site located in Decatur, east of Atlanta, is a NATTS site.
•	Hexavalent chromium was sampled for at SDGA, although sampling under the NMP
was discontinued in July.
•	Hexavalent chromium was detected in eight of the 30 valid samples collected at
SDGA in 2013, with measured detections of hexavalent chromium ranging from
0.0068 ng/m3 to 0.103 ng/m3.
•	Of eight measured detections, hexavalent chromium failed only one screen.
•	An annual average concentration could not be calculated for hexavalent chromium
since sampling ended in mid-July.
•	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
32-7

-------
pollutants of interest. This analysis shows that the range of concentrations of
hexavalent chromium measured at SDGA have not changed significantly over the last
few years of sampling.
•	Benzene is the highest emitted pollutant with a cancer toxicity factor in DeKalb
County, while formaldehyde has the highest cancer toxicity-weighted emissions.
Toluene is the highest emitted pollutant with a noncancer toxicity factor, while
acrolein has the highest noncancer toxicity-weighted emissions for 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, on the Illinois border near St. Louis.
•	VOCs and carbonyl compounds were sampled for at all three Illinois sites. SNMOCs,
PAHs, hexavalent chromium, and metals (PMio) were also sampled for at NBIL.
NBIL is one of only two NMP sites sampling all six pollutant groups, although
hexavalent chromium sampling was discontinued in June.
•	Twenty pollutants failed screens for NBIL; 12 pollutants failed screens for SPIL; and
11 pollutants failed screens for ROIL.
•	Of the pollutants of interest for each site, acetaldehyde and formaldehyde are the only
pollutants with annual average concentrations greater than 1 |ig/m3. Acetaldehyde has
the highest annual average concentration for NBIL, while formaldehyde has the
highest annual average concentration for SPIL and ROIL.
•	NBIL has highest annual average concentrations of acenaphthene and naphthalene
among NMP sites sampling PAHs. The maximum concentrations of acetaldehyde and
trichloroethylene program-wide were measured at SPIL.
•	Sampling for the site-specific pollutants of interest has occurred at NBIL and SPIL
for at least 5 consecutive years; thus, a trends analysis was conducted for the site-
specific pollutants of interest. Most notably, concentrations of acetaldehyde have
increased significantly 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.
•	Formaldehyde has the highest cancer risk approximation for all three Illinois sites. All
noncancer hazard approximations for the pollutants of interest for the Illinois sites are
less than an HQ of 1.0.
•	Benzene is the highest emitted pollutant with a cancer toxicity factor in Cook County,
while formaldehyde has the highest cancer-toxicity weighted emissions.
Formaldehyde is the highest emitted pollutant with a cancer toxicity factor in
Madison County, while coke oven emissions (PM) have the highest cancer toxicity
emissions.
32-8

-------
• Toluene is the highest emitted pollutant with a noncancer toxicity factor for both
counties, while acrolein has the highest noncancer toxicity-weighted emissions for
both counties.
Indiana.
•	There are two Indiana monitoring sites, one located in Indianapolis (WPIN) and a
second located in Gary, near Chicago (INDEM). Both are UATMP sites.
•	Carbonyl compounds were sampled for at WPIN and INDEM.
•	Formaldehyde and acetaldehyde failed screens for both INDEM and WPIN; all of the
measured detections of formaldehyde failed screens for both sites.
•	Formaldehyde has the highest annual average concentration for both sites, although
concentrations were higher at WPIN than INDEM. WPIN's annual average
concentration of formaldehyde is the sixth highest annual average concentration of
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. Concentrations of acetaldehyde have decreased at
WPIN in the last few years. Concentrations of formaldehyde and acetaldehyde
exhibited a significant decreasing trend at INDEM from 2008 to 2009, which may be
at least partially explained by a sampler change.
•	The cancer risk approximations for formaldehyde are an order of magnitude greater
than the cancer risk approximations for acetaldehyde for both sites. The noncancer
hazard approximations for the pollutants of interest for the Indiana sites are less
than an HQ of 1.0.
•	Benzene is the highest emitted pollutant with a cancer toxicity factor in both Marion
and Lake Counties. Coke oven emissions (PM) have the highest cancer toxicity-
weighted emissions for Lake County while formaldehyde has the highest cancer
toxicity-weighted emissions for Marion County.
•	Toluene is the highest emitted pollutant with a noncancer toxicity factor in both Lake
and Marion Counties while acrolein has the highest noncancer toxicity-weighted
emissions for both counties.
Kentucky.
• Three Kentucky monitoring sites are located in northeast Kentucky, two in Ashland
(ASKY and ASKY-M) and one near Grayson Lake (GLKY). The Grayson Lake
monitoring site is a NATTS site. One monitoring site is located south of Evansville,
Indiana (BAKY). Five monitoring sites are located in or near the Calvert City area
32-9

-------
(ATKY, CCKY, BLKY, LAKY, and TVKY). The final monitoring site is located in
Lexington, in north-central Kentucky (LEKY).
All of the Kentucky monitoring sites sampled for VOCs except ASKY-M and
BAKY. PAHs, carbonyl compounds, PMio metals, and hexavalent chromium were
also sampled for at GLKY, although hexavalent chromium sampling was
discontinued in June. Carbonyl compounds were also sampled for at ASKY and
LEKY and PMio metals were also sampled for at ASKY-M, BAKY, CCKY, and
LEKY.
The number of pollutants failing screens for the Kentucky sites varies from two
(BAKY) to 12 (LEKY, TVKY, and GLKY).
Of the pollutants of interest for each site, formaldehyde has the highest annual
average concentration for all three sites sampling carbonyl compounds (GLKY,
ASKY, and LEKY). Manganese has the highest annual average concentration for
ASKY-M, while arsenic has the highest annual average concentration for BAKY.
Carbon tetrachloride has the highest annual average concentration for ATKY and
CCKY, while 1,2-dichloroethane has the highest annual average concentration for
BLKY, LAKY, and TVKY.
The annual average concentrations of arsenic and nickel calculated for ASKY-M are
the highest annual average concentrations among NMP sites sampling PMio metals.
BAKY, LEKY, and CCKY are also among the NMP sites with highest annual
average concentrations of arsenic. ASKY has the fourth highest annual average
concentration of benzene among NMP sites sampling benzene; in addition, the
maximum concentration of benzene among all sites sampling VOCs was measured at
ASKY.
The Calvert city sites measured some of the highest concentrations of some VOCs,
particularly vinyl chloride, 1,2-dichloroethane, 1,3-butadiene, and carbon
tetrachloride.
The highest cancer risk approximations among the pollutants of interest for the
Kentucky sites were calculated for 1,2-dichloroethane (TVKY, 97.42 in-a-million and
BLKY, 33.15 in-a-million), formaldehyde (LEKY, 37.85 in-a-million), and
1,3-butadiene (TVKY, 30.97 in-a-million). The cancer risk approximation for TVKY
for 1,2-dichloroethane is the second highest cancer risk approximation calculated
among the site-specific pollutants of interest. None of the pollutants of interest for
which noncancer hazard approximations could be calculated were greater than an HQ
of 1.0.
Benzene is the highest emitted pollutant with a cancer toxicity factor in all Kentucky
counties with NMP sites, except Henderson County, where benzene ranks second to
formaldehyde. Coke oven emissions have the highest cancer toxicity-weighted
emissions for Boyd County; formaldehyde has the highest cancer toxicity-weighted
emissions for Carter, Henderson, Livingston, and Fayette Counties; and benzene has
the highest cancer toxicity-weighted emissions for Marshall County.
32-10

-------
•	Toluene is the highest emitted pollutant with a noncancer toxicity factor in Boyd,
Carter, Livingston, and Fayette Counties; carbonyl sulfide is the highest emitted
pollutant with a noncancer toxicity factor in Henderson County; and methanol is the
highest emitted pollutant with a noncancer toxicity factor in Marshall County.
Acrolein has the highest noncancer toxicity-weighted emissions in five of the
Kentucky counties, but ranks second to chlorine in Marshall County.
Massachusetts.
•	The Massachusetts monitoring site (BOMA) is a NATTS site located in Boston.
•	Metals (PMio), PAHs, and hexavalent chromium were sampled for at BOMA,
although hexavalent chromium sampling was discontinued in June.
•	Seven pollutants failed screens for BOMA. Arsenic and naphthalene each accounted
for at least 40 percent of the site's failed screens.
•	Of the pollutants of interest, naphthalene has the highest annual average
concentration.
•	BOMA has 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 the pollutants of interest for BOMA exhibit
little change over recent years of sampling.
•	The pollutants of interest for BOMA with cancer risk approximations greater than
1.0 in-a-million are arsenic and naphthalene. None of the pollutants of interest for
BOMA have noncancer hazard approximations greater than an HQ of 1.0.
•	Formaldehyde is the highest emitted pollutant with a cancer toxicity factor in Suffolk
County and has the highest cancer toxicity-weighted emissions. Toluene is the highest
emitted pollutant with a noncancer toxicity factor in Suffolk County, while acrolein
has the highest noncancer toxicity-weighted emissions.
Michigan.
•	The Michigan monitoring site (DEMI) is a NATTS site located in Dearborn,
southwest of Detroit.
•	VOCs, carbonyl compounds, PAHs, and hexavalent chromium were sampled for at
DEMI, although hexavalent chromium sampling was discontinued in June.
•	Eighteen pollutants failed screens for DEMI, of which 10 were identified as
pollutants of interest.
32-11

-------
•	Formaldehyde and acetaldehyde have the highest annual average concentrations for
DEMI. Compared to other NMP sites sampling PAHs, the annual average
concentration of acenaphthene for DEMI is the third highest while this site's annual
average concentration of naphthalene ranks fifth highest.
•	Sampling for the site-specific pollutants of interest has occurred at DEMI for at least
5 consecutive years; thus, a trends analysis was conducted for the site-specific
pollutants of interest. Benzene concentrations exhibit a steady decreasing trend
although concentrations have leveled out in recent years. In addition, the detection
rate of 1,2-dichloroethane at DEMI has been increasing steadily over the last few
years of sampling.
•	Formaldehyde has the highest cancer risk approximation for DEMI. None of the
pollutants of interest for DEMI had noncancer hazard approximations greater than an
HQ of 1.0.
•	Benzene is the highest emitted pollutant with a cancer toxicity factor in Wayne
County, while coke oven emissions have the highest cancer toxicity-weighted
emissions. Hydrochloric acid is the highest emitted pollutant with a noncancer
toxicity factor in Wayne County, while acrolein has the highest noncancer toxicity -
weighted emissions.
Minnesota.
•	The UATMP site in Minnesota (STMN) is located in St. Cloud.
•	Hexavalent chromium was sampled for at STMN through the end of May 2013 as
part of a 1-year monitoring effort that began in February 2012.
•	Twenty-four samples were collected at STMN in 2013 before sampling concluded;
measured detections of hexavalent chromium range from 0.008 ng/m3 to 0.039 ng/m3,
and includes 16 non-detects.
•	Concentrations of hexavalent chromium did not fail any screens for STMN.
•	Formaldehyde is the highest emitted pollutant with a cancer toxicity factor in Stearns
County and has the highest cancer toxicity-weighted emissions. Toluene is the highest
emitted pollutant with a noncancer toxicity factor in Stearns County, while acrolein
has the highest noncancer toxicity-weighted emissions.
Mississippi.
•	The Mississippi monitoring sites (KMMS and SSMS) are UATMP sites located in
Columbus.
•	Both KMMS and SSMS sampled for VOCs. KMMS also sampled for PAHs between
May and October, with an adjusted methodology that included phenols and cresols.
32-12

-------
•	Fourteen pollutants failed screens for KMMS, with 10 pollutants contributing to
95 percent of the total failed screens. Ten pollutants failed screens for SSMS, with
seven pollutants contributing to 95 percent of the total failed screens. KMMS is the
only NMP site for which xylenes were identified as a pollutant of interest.
•	Xylenes have the highest annual average concentration for KMMS, while carbon
tetrachloride has the highest annual average for SSMS.
•	KMMS has the highest annual average concentration of ethylbenzene among NMP
sites sampling this pollutant. SSMS has the highest annual average concentration of
hexachloro-1,3-butadiene among NMP sites.
•	Ethylbenzene and benzene have the highest cancer risk approximations for KMMS.
Benzene has the highest cancer risk approximation for SSMS. None of the pollutants
of interest for either site have a noncancer hazard approximation greater than an HQ
of 1.0.
•	Ethylbenzene is the highest emitted pollutant with a cancer toxicity factor in Lowndes
County, while formaldehyde has the highest cancer toxicity-weighted emissions.
Methanol is the highest emitted pollutant with a noncancer toxicity factor in Lowndes
County, while acrolein has the highest noncancer toxicity-weighted emissions.
Missouri.
•	The NATTS site in Missouri (S4MO) is located in St. Louis.
•	VOCs, carbonyl compounds, PAHs, metals (PMio), and hexavalent chromium were
sampled for at S4MO.
•	Twenty-one pollutants failed at least one screen for S4MO, 15 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 have the
highest annual average concentrations and are the only pollutants with annual average
concentrations greater than 1 |ig/m3,
•	S4MO has the second highest annual average concentration of hexachloro-1,3-
butadiene, the fourth highest annual average concentration of />dichlorobenzene, and
the fifth highest annual average concentration of arsenic (PMio) among NMP sites
sampling these pollutants.
•	Sampling for the site-specific pollutants of interest has occurred at S4MO for at least
5 consecutive years; thus, a trends analysis was conducted for each of the site-specific
pollutants of interest. Most notably, acetaldehyde concentrations have decreased
significantly since 2010. Some of the lowest concentrations of pollutants such as
arsenic, /?-dichlorobenzene, and naphthalene were measured in 2013. The detection
32-13

-------
rate of 1,2-dichloroethane has been increasing at S4MO over the last few years of
sampling.
•	Formaldehyde has the highest cancer risk approximation for S4MO. None of the
pollutants of interest for S4MO have a noncancer hazard approximation greater
than an HQ of 1.0.
•	Formaldehyde is the highest emitted pollutant with a cancer toxicity factor in
St. Louis (city) and has the highest cancer toxicity-weighted emissions. Toluene is the
highest emitted pollutant with a noncancer toxicity factor, while acrolein has the
highest noncancer toxicity-weighted emissions in St. Louis (city).
New Jersey.
•	Three of the UATMP sites in New Jersey are located in the New York-Newark-Jersey
City CBS A in the towns of Chester (CHNJ), Elizabeth (ELNJ), and North Brunswick
(NBNJ). A fourth UATMP site (CSNJ) is located in the Philadelphia-Camden-
Wilmington CBSA and is sampling under the NMP for the first time.
•	VOCs and carbonyl compounds were sampled for at all four New Jersey sites.
•	Sixteen pollutants failed at least one screen for CSNJ; nine pollutants failed at least
one screen for CHNJ; and 11 pollutants failed at least one screen for both ELNJ and
NBNJ, although the pollutants differed somewhat.
•	Of the site-specific pollutants of interest, formaldehyde and acetaldehyde have the
highest annual average concentrations for each New Jersey site.
•	The annual average concentrations of hexachloro-1,3-butadiene for NBNJ, CSNJ, and
CHNJ rank third, fifth, and sixth, respectively, among NMP sites sampling VOCs.
CSNJ has the third highest annual average concentration of both acetaldehyde and
formaldehyde, while ELNJ has 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 three of the four
New Jersey sites for at least 5 consecutive years; specifically, ELNJ is the longest
running NMP site still participating in the NMP. As such, a trends analysis was
conducted for the site-specific pollutants of interest. Benzene and ethylbenzene
concentrations have decreased significantly at ELNJ since sampling began. In
addition, the detection rates of 1,2-dichloroethane and hexachloro-l,3-butadience
have been increasing steadily over the last few years of sampling at CHNJ, ELNJ, and
NBNJ.
•	Formaldehyde has the highest cancer risk approximation for each of the New Jersey
sites. None of the pollutants of interest for any of the New Jersey sites have
noncancer hazard approximations greater than an HQ of 1.0.
32-14

-------
•	Benzene and formaldehyde are the highest emitted pollutants with cancer UREs in
Camden, Union, Middlesex, and Morris Counties. These two pollutants also have the
highest toxicity-weighted emissions for each county, although the order varied.
•	Toluene is the highest emitted pollutant with a noncancer toxicity factor in Camden,
Union, Middlesex, and Morris Counties. Acrolein has the highest noncancer toxicity-
weighted emissions for each New Jersey county.
New York.
•	The New York monitoring sites are located in New York City (BXNY) and Rochester
(ROCH). Both are NATTS sites.
•	PAHs and hexavalent chromium were sampled for at both BXNY and ROCH,
although hexavalent chromium sampling was discontinued at BXNY in June and at
ROCH in July.
•	Six pollutants failed screens for BXNY and four pollutants failed screens for ROCH.
Naphthalene failed the majority of screens for both sites.
•	Naphthalene has the highest annual average concentration for BXNY and ROCH,
although the annual average concentration for BXNY is twice the annual average
calculated for ROCH.
•	ROCH and BXNY have the second and fifth highest annual average concentrations of
acenaphthene, respectively, among NMP sites sampling PAHs. BXNY also has the
third highest annual average concentration of naphthalene among NMP sites.
•	Even though sampling of PAHs has been conducted at ROCH for greater than
5 consecutive years, a sample collection error resulted in the invalidation of a year
and a half s worth of data. Although a trends analysis was conducted for each of the
site-specific pollutants of interest for ROCH, the gap in data makes definitive trends
hard to identify.
•	Naphthalene has the highest cancer risk approximation among the pollutants of
interest for both ROCH and BXNY. None of the pollutants of interest for either site
have noncancer hazard approximations greater than an HQ of 1.0.
•	Benzene is the highest emitted pollutant with a cancer toxicity factor for Bronx and
Monroe Counties while formaldehyde has the highest cancer toxicity-weighted
emissions for both counties.
•	Toluene is the highest emitted pollutant with a noncancer toxicity factor in both
Bronx and Monroe Counties. Acrolein has the highest noncancer toxicity-weighted
emissions for both counties.
32-15

-------
Oklahoma.
•	There are five UATMP sites in Oklahoma: three located in Tulsa (TOOK, TMOK,
and TROK) and two in Oklahoma City (ADOK and OCOK). The sampling
instrumentation at ADOK was moved mid-year to a site west of Oklahoma City in
Yukon (YUOK).
•	VOCs, carbonyls compounds, and metals (TSP) were sampled for at each of the
Oklahoma sites.
•	Seventeen pollutants failed screens for TOOK; 15 failed screens for TMOK; 14 failed
screens for TROK; 12 failed screens for ADOK; 16 failed screens for OCOK; and 11
failed screens for YUOK.
•	Formaldehyde and acetaldehyde have the highest annual average concentrations for
TOOK, TMOK, TROK, and OCOK. Annual average concentrations could not be
calculated for ADOK and YUOK due to the mid-year relocation of the instruments.
•	TMOK has the third highest annual average concentration of />dichlorobenzene
among NMP sites sampling this pollutant. OCOK has the fourth highest annual
average concentration of hexachloro-1,3-butadiene among NMP sites sampling this
pollutant.
•	Sampling for the site-specific pollutants of interest has occurred at TOOK, TMOK,
and OCOK for at least 5 consecutive years; thus, a trends analysis was conducted for
the site-specific pollutants of interest. After several years of increasing, acetaldehyde,
ethylbenzene, and manganese concentrations decreased at TOOK for 2013. Benzene
concentrations at TOOK and TMOK have decreased in recent years, while
acetaldehyde and formaldehyde concentrations at OCOK have also decreased.
Detection rates of 1,2-dichloroethane and hexachloro 1,3-butadiene have increased at
TOOK, TMOK, and OCOK in recent years.
•	Formaldehyde and benzene have the highest cancer risk approximations for all of the
Oklahoma monitoring sites. None of the pollutants of interest for the Oklahoma sites
have a noncancer hazard approximation greater than an HQ of 1.0.
•	Benzene is the highest emitted pollutant with a cancer toxicity factor in Oklahoma
and Tulsa Counties and has the highest cancer toxicity-weighted emissions for both
counties. Formaldehyde is the highest emitted pollutant with a cancer toxicity factor
in Canadian County and has the highest cancer toxicity-weighted emissions for that
county.
•	Toluene is the highest emitted pollutant with a noncancer toxicity factor in Oklahoma
and Tulsa Counties, while xylenes are the highest emitted pollutant with a noncancer
toxicity factor in Canadian County. Acrolein has the highest noncancer toxicity -
weighted emissions for all three counties.
32-16

-------
Rhode Island.
•	The Rhode Island monitoring site (PRRI) is located in Providence and is a NATTS
site.
•	PAHs and hexavalent chromium were sampled for at PRRI, although sampling for
hexavalent chromium was discontinued at the end of June.
•	Three pollutants failed screens for PRRI, although greater than 95 percent of failed
screens were attributable to naphthalene. As a result, naphthalene is PRRI's only
pollutant of interest.
•	Naphthalene concentrations measured at PRRI span an order of magnitude, ranging
from 17.5 ng/m3 to 187 ng/m3.
•	Sampling for the site-specific pollutants of interest has occurred at PRRI for at least
5 consecutive years; thus, a trends analysis was conducted for the site-specific
pollutants of interest. Although concentrations of naphthalene exhibit little change
over the years of sampling, several of the statistical parameters calculated are at a
minimum for 2013.
•	The cancer risk approximation for naphthalene for PRRI is 2.09 in-a-million. The
noncancer hazard approximation for this pollutant is considerably less than an HQ of
1.0.
•	Benzene is the highest emitted pollutant with a cancer toxicity factor in Providence
County, while formaldehyde has the highest cancer toxicity-weighted emissions.
Toluene is the highest emitted pollutant with a noncancer toxicity factor, while
acrolein has the highest noncancer toxicity-weighted emissions for Providence
County.
South Carolina.
•	The South Carolina monitoring site (CHSC) is located near Chesterfield and is a
NATTS site.
•	Hexavalent chromium and PAHs were sampled for at CHSC, although hexavalent
chromium sampling was discontinued in June.
•	Naphthalene was the only pollutant to fail screens for CHSC. Less than 4 percent of
naphthalene concentrations failed screens for CHSC.
•	Naphthalene concentrations measured at CHSC range from 4.46 ng/m3 to 51.8 ng/m3.
Compared to other NMP sites sampling this pollutant, CHSC has the second lowest
annual average concentration 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
32-17

-------
pollutant of interest. Concentrations of naphthalene measured at CHSC in 2013 are
the lowest since the onset of sampling in 2008.
•	The cancer risk approximation for naphthalene for CHSC is less than 1.0 in-a-million.
The noncancer hazard approximation for this pollutant is considerably less than an
HQ of 1.0.
•	Benzene is the highest emitted pollutant with a cancer toxicity factor in Chesterfield
County while formaldehyde has the highest cancer toxicity-weighted emissions.
Toluene is the highest emitted pollutant with a noncancer toxicity factor, while
acrolein has the highest noncancer toxicity-weighted emissions.
Texas.
•	There are two NATTS sites in Texas: one in Deer Park (CAMS 35) and one in
Karnack (CAMS 85).
•	Hexavalent chromium was sampled for at both CAMS 35 and CAMS 85, although
sampling was discontinued at both sites in June.
•	Hexavalent chromium failed 8 percent of screens for CAMS 35. Because hexavalent
chromium did not fail any screens for CAMS 85, this site has no pollutants of
interest.
•	Concentrations of hexavalent chromium measured at CAMS 35 ranged from
0.0167 ng/m3 to 0.38 ng/m3, including five non-detects. The maximum hexavalent
chromium concentration for the program was measured at CAMS 35. Due to the
discontinuation of sampling, an annual average concentration could not be calculated.
•	Benzene is the highest emitted pollutant with a cancer toxicity factor in Harris
County, while 1,3-butadiene has the highest cancer toxicity-weighted emissions.
Formaldehyde is the highest emitted pollutant with a cancer toxicity factor in
Harrison County and has the highest cancer toxicity-weighted emissions.
•	Toluene is the highest emitted pollutant with a noncancer toxicity factor in both
counties, while acrolein has the highest noncancer toxicity-weighted emissions.
Utah.
•	The NATTS site in Utah (BTUT) is located in Bountiful, north of Salt Lake City.
•	VOCs, carbonyl compounds, SNMOCs, PAHs, metals (PMio), and hexavalent
chromium were sampled for at BTUT. This site is one of only two NMP sites
sampling all six pollutant groups. Hexavalent chromium sampling was discontinued
in June.
32-18

-------
•	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 has the highest annual
average concentration for BTUT, which is consistent with previous years of
sampling. BTUT has the highest annual average concentrations of ethylbenzene,
formaldehyde, and acetaldehyde among NMP sites sampling these pollutants. BTUT
also has the second highest annual average concentration of arsenic among NMP sites
sampling metals (PMio).
•	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 2013 is the lowest 1-year average concentration of benzene calculated since the
onset of sampling at BTUT. Additionally, concentrations of acetaldehyde,
formaldehyde, and propionaldehyde exhibited significant increases in 2013.
•	The pollutant with the highest cancer risk approximation for BTUT is formaldehyde;
this is the highest cancer risk approximation calculated across the program and the
only one greater than 100 in-a-million. None of the pollutants of interest have
noncancer hazard approximations greater than an HQ of 1.0.
•	Benzene is the highest emitted pollutant with a cancer toxicity factor in Davis County
and has the highest cancer toxicity-weighted emissions. Toluene is the highest
emitted pollutant with a noncancer toxicity factor, while acrolein has the highest
noncancer toxicity-weighted emissions for Davis County.
Vermont.
•	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
site, while the remaining sites are UATMP sites.
•	VOCs were sampled for year-round at BURVT and RUVT. VOCs, hexavalent
chromium, PAHs, and metals (PMio) were sampled for at UNVT, although
hexavalent chromium sampling was discontinued in June. Sampled at UNVT
occurred on a l-in-6 day sampling schedule while the other two sites sampled on a
l-in-12 day sampling schedule.
•	Eight pollutants failed screens for BURVT; six pollutants failed screens for RUVT;
and nine pollutants failed screens for UNVT.
•	Benzene has the highest annual average concentrations for BURVT and RUVT, while
carbon tetrachloride has the highest annual average concentration for UNVT.
32-19

-------
•	Annual average concentrations for several of the pollutants of interest for UNVT are
among the lowest compared to other NMP sites sampling the same pollutants.
•	Sampling for several of the site-specific pollutants of interest has occurred at the
Vermont sites for at least 5 consecutive years; thus, a trends analysis was conducted
where applicable. The most notable trend for the Vermont sites is for
1,2-dichloroethane, a pollutant of interest for all three sites. The detection rate of
1,2-dichloroethane has increased significantly over the years at each of the Vermont
sites, particularly in the last 2 years.
•	Benzene and carbon tetrachloride have the highest cancer risk approximations for
each of the Vermont monitoring sites (although not necessarily in that order). None of
the noncancer hazard approximations for these sites are greater than an HQ of 1.0.
•	Benzene and formaldehyde are the highest emitted pollutants with a cancer toxicity
factors in Chittenden and Rutland Counties. Benzene and formaldehyde also have the
highest cancer toxicity-weighted emissions for Rutland County while the order is
reversed for Chittenden County.
•	Toluene is the highest emitted pollutant with a noncancer toxicity factor in both
Vermont counties, while acrolein has the highest noncancer toxicity-weighted
emissions.
Virginia.
•	The NATTS site in Virginia is located near Richmond (RIVA).
•	PAHs and hexavalent chromium were sampled for at RIVA.
•	Naphthalene was the only pollutant to fail screens for RIVA, with greater than
96 percent of naphthalene measurements collected at RIVA failing screens.
•	Naphthalene concentrations measured at RIVA range from 18.0 ng/m3 to 354 ng/m3.
Compared to other NMP sites sampling this pollutant, RIVA has the eighth highest
annual average concentrations of naphthalene.
•	Sampling for PAHs has occurred at RIVA for at least 5 consecutive years; thus, a
trends analysis was conducted for naphthalene. No significant trend in the
concentrations of naphthalene measured at RIVA was noted.
•	The cancer risk approximation for naphthalene at RIVA is 2.95 in-a-million, while
the noncancer hazard approximation is significantly less than a HQ of 1.0.
•	Benzene is the highest emitted pollutant with a cancer toxicity factor in Henrico
County, while formaldehyde has the highest cancer toxicity-weighted emissions.
32-20

-------
•	Toluene is the highest emitted pollutant with a noncancer toxicity factor in Henrico
County, while acrolein has the highest noncancer toxicity-weighted emissions.
Washington.
•	The NATTS site in Washington is located in Seattle (SEWA).
•	VOCs, carbonyl compounds, PAHs, metals (PMio), and hexavalent chromium were
sampled for at SEW A, although hexavalent chromium sampling was discontinued in
June.
•	Fourteen pollutants failed screens for SEW A, of which nine were identified as
pollutants of interest for this site.
•	None of the site-specific pollutants of interest for SEWA have annual average
concentrations greater than 1 |ig/m3, Acetaldehyde and carbon tetrachloride have the
highest annual average concentrations for this site. The annual average concentration
of formaldehyde for SEWA is the lowest among NMP sites sampling this pollutant.
•	SEWA has the second highest annual average concentration of nickel among NMP
sites sampling metals (PMio). This was also true for 2012. 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. Benzene has an overall decreasing trend at SEWA since the
onset of sampling. In addition, the detection rate of 1,2-dichloroethane at SEWA has
been increasing steadily over the last few years of sampling.
•	Formaldehyde has the highest cancer risk approximation for SEWA, although it is the
lowest cancer risk approximation for formaldehyde among NMP sites. All of the
noncancer hazard approximations for the pollutants of interest for SEWA are less
than an HQ of 1.0.
•	Benzene is the highest emitted pollutant with a cancer toxicity factor in King County
while formaldehyde has the highest cancer toxicity-weighted emissions. Toluene is
the highest emitted pollutant with a noncancer toxicity factor in King County, while
acrolein has the highest noncancer toxicity-weighted emissions.
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.
•	Hexavalent chromium was sampled for at both HOWI and MIWI. Hexavalent
chromium sampling was discontinued in March at MIWI at the completion of a
32-21

-------
1-year study beginning in 2012. Hexavalent chromium sampling was also
discontinued at HOWI in June 2013.
•	Thirty hexavalent chromium samples were collected at HOWI prior to the
discontinuation of sampling. This pollutant was detected in only four of the samples
collected at HOWI and did not fail any screens. Concentrations of hexavalent
chromium measured at HOWI in 2013 range from 0.0088 ng/m3 to 0.019 ng/m3 (as
well as 26 non-detects).
•	Eleven hexavalent chromium samples were collected at MIWI prior to the
discontinuation of sampling. This pollutant was detected in eight of the samples
collected at MIWI and did not fail any screens. Concentrations of hexavalent
chromium measured at MIWI in 2013 range from 0.0033 ng/m3 to 0.0405 ng/m3 (as
well as three non-detects). These measurements represent a decrease in the magnitude
of the concentrations compared to those measured during the first 9 months of
sampling in 2012.
•	Benzene is the highest emitted pollutant with a cancer toxicity factor in Dodge
County while formaldehyde has the highest cancer toxicity-weighted emissions.
Benzene is also the highest emitted pollutant with a cancer toxicity factor in
Milwaukee County, while hexavalent chromium has the highest cancer toxicity-
weighted emissions.
•	Toluene is the highest emitted pollutant with a noncancer toxicity factor in Dodge and
Milwaukee Counties, while acrolein has the highest noncancer toxicity-weighted
emissions for each county.
32.1.3 Composite Site-level Results Summary
•	Twenty-seven pollutants were identified as site-specific pollutants of interest, based
on the risk-based screening process. Benzene and 1,3-butadiene were the two most
common pollutants of interest among the monitoring sites. Benzene was identified as
a pollutant of interest for all 39 sites that sampled this pollutant (with Method TO-15
or SNMOC) and 1,3-butadiene was a pollutant of interest for all but one (BRCO is
the exception). Acetaldehyde and formaldehyde were the most common carbonyl
compound pollutants of interest. These two compounds were identified as pollutants
of interest for all 33 sites that sampled carbonyl compounds. Twenty-three of the 25
sites that sampled PAHs had naphthalene as a pollutant of interest (with GLKY and
UNVT as the exceptions). Arsenic was identified as a pollutant of interest for all 20
sites that sampled metals.
•	Hexavalent chromium was identified as a pollutant of interest for only two sites
(SDGA and CAMS 35), although this is the only pollutant sampled for at these sites.
Hexavalent chromium concentrations from several sites (CAMS 85, HOWI, MIWI,
and STMN) did not fail any screens, although this was the only pollutant sampled for
at these sites. EPA dropped the requirement to sample hexavalent chromium under
the NATTS program beginning in July 2013, so all but two of the NATTS sites
stopped sampling this pollutant in either June or July.
32-22

-------
Several pollutants were identified as site-specific pollutants of interest for only one or
two sites. For instance, dichloromethane is a pollutant of interest for only BTUT;
bromomethane is a pollutant of interest for only CSNJ; trichloroethylene is a pollutant
of interest for only SPIL; and xylenes are a pollutant of interest for only KMMS.
Table 32-1 summarizes which pollutants of interest were identified for each site, how
many pollutants of interest were identified for each site, and how many sites for
which each pollutant was identified as a pollutant of interest.
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 24 sites. Naphthalene had the next highest at 11 followed
by benzene with eight.
Formaldehyde tended to have the highest cancer risk approximations on a site-
specific basis. The cancer risk approximation calculated for BTUT (104.6 in-a-
million) from the annual average concentration of formaldehyde is the highest of all
annual average-based cancer risk approximations and the only one greater than
100 in-a-million. This site also had the highest cancer risk approximation in the 2012
report, but the cancer risk approximation for 2012 was half as high. Five other sites
also have cancer risk approximations greater than 50 in-a-million, four for
formaldehyde (GPCO, CSNJ, ELNJ, and PXSS) and one for 1,2-dichloroethane
(TVKY). Benzene and 1,3-butadiene are the only other pollutants for which a cancer
risk approximation greater than 10 in-a-million was calculated.
Carbon tetrachloride often had relatively high cancer risk approximations (based on
annual average concentrations) compared to other pollutants of interest among the
monitoring sites, ranging between 3 in-a-million and 7 in-a-million, but tended to
have relatively low emissions and toxicity-weighted emissions according to the NEI.
This pollutant appears only once in the emissions-based tables for counties with NMP
sites (Marshall County, Kentucky, where the five Calvert City sites are located).
None of the noncancer hazard approximations were greater than an HQ of 1.0. The
noncancer hazard approximation calculated for BTUT's annual average concentration
of formaldehyde (with an HQ of 0.82) is the highest of all annual average-based
noncancer hazard approximations. Formaldehyde tended to have the highest
noncancer hazard approximations on a site-specific basis, followed by 1,3-butadiene
and naphthalene.
Of those pollutants with cancer UREs, benzene, formaldehyde, and ethylbenzene
often had the highest county-level emissions for participating counties.
Formaldehyde, benzene, and 1,3-butadiene typically had the highest toxicity-
weighted emissions (of those with a cancer URE).
32-23

-------
PZ~Z€
to
1
2
I
M
hd
>
6-
0
en
OQ*
1
rt
p.
2
>
H
H
in
p
O
>
~n
r
~n
r
~n
r
~n
r
~n
r
~n
r
a
n
n
o
n
O
n
o
o
o
n
o
n
o
n
>
n
>
n
>
n
>


*
State
NBIL
SDG4
WPFL
SYFL
SKFL
PAFL
ORFL
AZFL
WADC
RICO
RFCO
PACO
GPCO
BRCO
BMCO
SJJCA
RUCA
LBHCA
CELA
SPAZ
PXSS
ANAK
Site
-
-

U>
U>
-
K>
K>
-
Ui
4-^

K>
U>
Ui
u>
-
-

On
o
00
# of Pollutants of Interest
X











X





X



Acenaphthene
X


X
X

X
X

X
X

X
X
X





X

Acet aldehyde
X














X




X

Arsenic
X








X
X
X
X
X
X




X
X
X
Benzene












X





X



Benzo(a)pyrene






















Bromomethane
X








X
X
X
X

X




X
X
X
1,3-Butadiene






















Cadmium
X











X






X
X
X
Carbon Tetrachloride



















X
X
X
p-Dichlorobenzene
X











X






X
X
X
1,2-Dichloroethane






















Dichloromethane









X


X

X




X
X
X
Ethylbenzene
X

X



















Fluoranthene
X

X









X





X



Fluorene
X


X
X







X

X





X

Formaldehyde












X








X
Hexachloro-l,3-butadiene






















Hexavalent Chromium






















Lead






















Manganese
X











X


X




X
X
Naphthalene






















Nickel






















Propionaldehyde






















1,1,2-T richloroethane






















T richloroethylene






















Vinyl chloride






















Xylenes
H
as
2
5T
w
K»
C/5
c
B
5
is
"t
V!
o
«s
S-
r&
C/)
"O
r&
o
s
o
hd
o
p
3
r&
-t
rD
<*2

-------
SZ-Zi
to
1
2
S
M
hj
>
6-
0
en
OQ*
1
rt
p.
2
>
H
H
m
2
«—1
2
<—1
2
m
2
m
| MO
II MN
§
| MA
II KY
II KY
KY
II KY
II KY
II KY
KY
II KY
II KY
II KY
3
3
F
F
State
CSNJ
CHNJ
SSMS
KMMS
S4MO
STMN
DEMI
BOMA
TVKY
LEKY
LAKY
GLKY
CCKY
BLKY
BAKY
ATKY
ASKY-M
ASKY
WPIN
INDEM
SPIL
ROIL
Site
o


o
Ui
o
o
U>
00
00
On

On

K>
Ui
Ui
On
K>
K>
00
00
# of Pollutants of Interest



X
X

X















Acenaphthene
X
X


X

X


X

X





X
X
X
X
X
Acet aldehyde




X


X

X

X
X

X

X





Arsenic
X
X
X
X
X

X

X
X
X
X
X
X

X

X


X
X
Benzene






















Benzo(a)pyrene
X





















Bromomethane
X
X
X
X
X

X

X
X
X
X
X
X

X

X


X
X
1,3-Butadiene




X











X





Cadmium
X
X
X
X
X

X

X
X
X
X
X
X

X

X


X
X
Carbon Tetrachloride


X
X
X



X













p-Dichlorobenzene
X
X
X
X
X

X

X
X
X
X
X
X

X

X


X
X
1,2-Dichloroethane






















Dichloromethane
X

X
X


X














X
Ethylbenzene






















Fluoranthene



X
X

X















Fluorene
X
X


X

X


X







X


X
X
Formaldehyde
X



X



X
X
X

X
X






X
X
Hexachloro-l,3-butadiene






















Hexavalent Chromium




X











X





Lead
















X





Manganese



X
X

X
X














Naphthalene




X











X





Nickel
X





















Propionaldehyde








X




X








1,1,2-T richloroethane




















X

T richloroethylene








X

X











Vinyl chloride



X


















Xylenes
H
as
O"
W
N>
C/5
c
B
B
"t
V!
o
«s
S-
re
GC
¦a
re
o
«
hd
as
S
re
"t
re
W!
n
©
s
s
c
re
a

-------
9Z-Z£
to
1
2
I
M
hd
>
6-
0
en
OQ*
1
rt
p.
2
>
H
H
in
| Total


II WA
$
$
3
$
C
H
>3
><
m
O
s
O
o
O
O
O
O
NY
II NY
2
<—1
2
<—1
State
MIWI
HOWI
SEWA
UNVT
RUVT
BURVT
RIVA
BTUT
CAMS 85
CAMS 35
DSHD
PRRI
YUOK
TROK
TOOK
TMOK
OCOK
ADOK
ROCH
BXNY
NBNJ
ELNJ
Site
U>
Ui
^1
O
O
\D
Ut

On
-
U>
o
-
-
-
00
o
-
o
o
00
*-
U>

00
# of Pollutants of Interest
vo


















X
X


Acenaphthene
U)
U>


X




X




X
X
X
X
X
X


X
X
Acet aldehyde
K>
o


X
X



X




X
X
X
X
X
X




Arsenic
U)
vo


X
X
X
X

X




X
X
X
X
X
X


X
X
Benzene
K>






















Benzo(a)pyrene
-






















Bromomethane
u>
00


X
X
X
X

X




X
X
X
X
X
X


X
X
1,3-Butadiene
K>






















Cadmium
U>


X
X
X
X

X




X
X
X
X
X
X


X
X
Carbon Tetrachloride
K>





X

X







X
X
X




p-Dichlorobenzene
U>


X
X
X
X

X




X
X
X
X
X
X


X
X
1,2-Dichloroethane
-







X














Dichloromethane
00







X





X
X
X
X




X
Ethylbenzene
u>


















X



Fluoranthene
vo


















X
X


Fluorene
U)
U>


X




X




X
X
X
X
X
X


X
X
Formaldehyde
K>





X






X
X
X
X
X




X
Hexachloro-l,3-butadiene
K>






















Hexavalent Chromium
K>






















Lead
K>














X







Manganese
K>
u>


X




X










X



Naphthalene
vo


X




X





X
X







Nickel
K>







X














Propionaldehyde
K>






















1,1,2-T richloroethane
-






















T richloroethylene























Vinyl chloride
-






















Xylenes
H
as
O"
W
N>
C/5
c
B
B
"t
V!
o
«s
S-
re
GC
¦a
re
o
«
hd
as
S
re
"t
re
W!
n
©
s
s
c
re
a

-------
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 twice among the highest
emitted pollutants for counties with NMP sites (Garfield County, Colorado and
Canadian County, Oklahoma). However, due to the high toxicity of this pollutant,
even low emissions translated into high noncancer toxicity-weighted emissions; the
toxicity-weighted value was often several orders of magnitude higher than other
pollutants. Acrolein is a national noncancer risk driver according to NATA. Besides
acrolein, formaldehyde and 1,3-butadiene tended to have the highest toxicity-
weighted emissions among the pollutants with noncancer RfCs.
For the 2013 NMP report, emissions data provided are from version 2 of the 2011
NEI while emissions data for the 2012 NMP report were from version 1 of the 2011
NEI, which explains some changes in the emissions data used to create the point
source emissions maps and the risk-based emissions assessment tables.
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 13 percent detection rate across the program). The five
Calvert City, Kentucky sites together account for more than 67 percent of the
measured detections of vinyl chloride for 2013. Individually, these sites have the
highest number of measured detections among NMP sites sampling VOCs. The
Calvert City sites also account for the 76 highest concentrations of 1,2-dichloroethane
measured across the program. These ambient air measurements agree with
corresponding emissions data in the NEI. These three pollutants appear among the
highest emitted pollutants in Marshall County, Kentucky (among those with a cancer
URE) but are not 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 is a pollutant of interest (34 sites),
and where a trends analysis could be conducted for this pollutant, a dramatic increase
in the number of measured detections is shown over the most recent years of
sampling, particularly for 2012, which was mostly sustained for 2013. 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,
32-27

-------
slowly at first then significantly in 2012. The detection rate of this pollutant is
between 75 percent and 100 percent for most NMP sites for 2013.
32.1.4 Data Quality Results Summary
Completeness, precision, and accuracy were assessed for the 2013 monitoring effort. The
quality assessments presented in this report show that the 2013 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 the MQO for completeness.
Only three out of 143 site- and method-specific data sets failed to comply with the MQO of
85 percent completeness while 63 data sets achieved 100 percent completeness.
Method (sampling and analytical) precision and analytical precision were determined for
the 2013 NMP monitoring efforts using CV calculations based on duplicate, collocated, and
replicate samples. Method precision for each analytical method utilized during the 2013 NMP
was within the MQO of 15 percent CV with the exception of hexavalent chromium. Analytical
precision for each method was determined to be less than 15 percent CV. The precision
calculations presented in this report are based on analytical results greater than or equal to the
sample- and pollutant-specific MDL.
Analytical method accuracy is ensured by using proven methods, as demonstrated by
third-party analysis of proficiency test audit samples, and following strict quality control and
quality assurance guidelines. Most of the pollutants for which audit samples were analyzed met
the MQO for accuracy. Of the 37 pollutants analyzed for via audit samples, five exceeded the
MQO of ± 25 percent recovery.
32.2 Conclusions
Conclusions extrapolated from the data analyses of the data generated from the 2013
NMP monitoring efforts are presented below.
• A large number of concentrations are greater than their respective risk screening
values, particularly for many of the NATTS MQO Core Analytes. For several of the
pollutants, all or nearly all of the measurements fail screens. Examples of frequently
detected pollutants that typically fail all or nearly all of their screens include benzene,
carbon tetrachloride, formaldehyde, acetaldehyde, 1,2-dichloroethane, and
1,3-butadiene. Some of the lesser detected pollutants still fail relatively large numbers
32-28

-------
of screens. For example, even though hexachloro-1,3-butadiene was detected
relatively infrequently, most of the measured detections failed screens. This is also
true for 1,2-dibromoethane and chloroprene.
Over the last few years, the number of concentrations failing screens has increased,
although the percentage of failed screens compared to the number of measured
detections has hovered around 36 percent. Yet, for many of the sites that sampled
year-round in both 2012 and 2013, the number of failed screens was down for 2013
compared to 2012. Four sites (GPCO, PXSS, S4MO, and TOOK) failed more than
100 fewer screens for 2013 compared to 2012. Aside from these four, the difference
in the number of fewer failed screens ranges from one less than in 2012 to 79 fewer
failed screens. Only two sites failed more screens in 2013 than in 2012 and one site's
number of failed screens did not change at all. The decrease for some sites may be
attributable to changes in the risk screening values. The risk screening value for
manganese, which typically failed a majority of its of risk screens, was increased by
an order of magnitude in the last revision by EPA, resulting in a significant difference
between the two reports (from 706 failed screens for 2012 to 61 failed screens for
2013). In addition, the risk screening value for 1,1,2,2-tetrachloroethane was removed
altogether. This pollutant accounted for 112 failed screens in the 2012 report. Other
changes to the risk screening values include the addition of a risk screening value for
coronene (which did not fail any screens) and an update to the dichloromethane risk
screening value (which resulted in a relatively small change to the number of failed
screens for this pollutant).
For those pollutants for which annual average concentrations could be calculated and
that have available cancer UREs, only one of the cancer risk approximations was
greater than 100 in-a-million, which is the first such occurrence since this analysis
was added to the NMP report. In total, 38 site- and pollutant-specific cancer risk
approximations were greater than 10 in-a-million (28 for formaldehyde, four for
benzene, and three each for 1,3-butadiene and 1,2-dichloroethane); and roughly 69
percent were greater than 1.0 in-a-million.
For those pollutants for which annual average concentrations could be calculated and
have available noncancer RfCs, none of the noncancer hazard approximations were
greater than an HQ of 1.0.
When comparing the highest emitted pollutants for a specific county to the pollutants
with the highest toxicity-weighted emissions, the pollutants tended to be more similar
for the pollutants with cancer UREs than for pollutants with noncancer RfCs. This
indicates that pollutants with cancer UREs that are emitted in higher quantities are
often more toxic than pollutants emitted in lower quantities; conversely, the highest
emitted pollutants with noncancer RfCs are not necessarily the most toxic. For
example, toluene is the noncancer pollutant that was emitted in the highest quantities
for many NMP counties, and did not rank less than third for any county with an NMP
site, but was not one of the pollutants with highest toxicity-weighted emissions for
any of these counties. Conversely, while acrolein had the highest noncancer toxicity-
weighted emissions for most NMP counties, and ranked second for only one county,
32-29

-------
it was among the highest emitted pollutants for only two counties with NMP (and
ranked no higher than eighth).
The number of states and sites participating in the NMP changes from year to year.
The number of sites participating in the 2013 NMP increased just slightly, from 64
for 2012 to 66 for 2013. Yet, many of the data analyses utilized in this report require
data from year-round (or nearly year-round) sampling. Of the 66 sites whose data are
included in the 2013 report, nine sites sampled for an abbreviated duration. This can
be due to site initialization and/or site closure/relocation, due to the start or stop date
of special studies, such as those related to CSATAM sites, or due to the removal of a
pollutant from a list of required pollutants in which to sample, as was the case with
hexavalent chromium for the NATTS program. Of the 143 site-method combinations,
30 site-method combinations did not cover the entire year. A majority of these (19)
are a result of the delisting of hexavalent chromium, a pollutant that was infrequently
identified as a site-specific pollutant of interest. Thus, the number of time-period
averages and subsequent risk-based analyses that could not be calculated actually
decreased for 2013 compared to 2012. Fewer data gaps allow for more complete
results and inter-site comparisons.
Of the 66 monitoring sites participating in the 2013 NMP, only two sampled for all
six available pollutant groups under the national program (BTUT and NBIL). Another
four sites (GLKY, PXSS, S4MO, and SEW A) sampled five pollutant groups under
the NMP through the national contract laboratory. The wide range of pollutant groups
sampled for among the sites, which is often the result of different purposes behind the
monitoring at the sites, makes it difficult to draw definitive conclusions regarding air
toxics in ambient air in a global manner.
The data analyses contained in the 2013 NMP report reflect the inclusion of data from
a number of source-oriented monitoring sites. The number of such sites has been
increasing in recent years. Newer source-oriented sites include several of the
Kentucky sites; the two Columbus, Mississippi sites; the Belle Glade, Florida site;
and the new Camden, New Jersey site. Many of these sites are the drivers for certain
pollutant(s) in the 2013 report. This can easily be seen in the graphical site-specific
comparisons to the program-level average concentrations contained in Sections 5
through 31. For many of these pollutants, particularly the VOCs, the highest
concentrations were considerably greater than the majority of measurements, such
that the scale in the figures needed to be greatly reduced.
This report strives to represent data derived from the best laboratory practices and
utilize the best data analysis techniques available. Examples for 2013 include the
improvement of MDLs and the incorporation of updated values for various toxicity
factors. This can lead to adjusting the focus of the report to concentrate on the air
quality issues of highest concern. Thus, the NMP report is dynamic in nature and
scope; yet this approach may prevent the direct comparison of the current report to
past reports. Relatively few major changes were instituted between the 2012 and 2013
NMP reports. One difference between the 2013 report and other reports in recent
years is the removal of the back trajectory analysis, the coefficient of variation
32-30

-------
variability analysis, and the ATSDR MRL screening analysis. Another difference is
the inclusion of additional detail in the QA section (Section 31).
32.3 Recommendations
Based on the conclusions from the 2013 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 2013 report require a full year of data to be most useful and
representative of conditions experienced at each specified location. Therefore, state
and local agencies should be encouraged to implement year-long ambient air
monitoring programs in addition to participating in future monitoring efforts.
•	Monitor for additional pollutant groups based on the results of data analyses in the
annual report. The risk-based analysis where county-level emissions are weighted
based on toxicity identifies those pollutants whose emissions may result in adverse
health effects in a specific area. If a site is not sampling for a pollutant or pollutant
group identified as particularly hazardous for a given area, the agency responsible for
that site should consider sampling for those compounds.
•	Strive to develop standard conventions for interpreting air monitoring data. The lack
of consistent approaches to present and summarize ambient air monitoring data
complicates direct comparisons between different studies. Thought should be given to
the feasibility of establishing standard approaches for analyzing and reporting air
monitoring data for programs with similar objectives.
•	Continue to identify and implement improvements to the sampling and analytical
methods. In 2012, two analytical methods were accepted by governing bodies as
approved methods with which to analyze specific pollutants. ERG's hexavalent
chromium method was approved as an ASTM method and ERG's inorganic method
for both TSP and PMio was accepted as a FEM for lead (NAAQS). These approvals
were obtained after various method enhancements that improve the detection and
recovery of these pollutants. Further research is encouraged to identify other method
improvements that would allow for the characterization of an even wider range of
components in air pollution and enhance the ability of the methods to quantify all
cancer and noncancer pollutants to at least their levels of concern (risk screening
concentrations).
32-31

-------
Revise the pollutants targetedfor sampling based on lessons learned in the field, in
the laboratory, and/or from the annual report. In conjunction with method
improvements, the analytes targeted for monitoring should/need to be reviewed and
revised periodically based on 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. The removal of hexavalent chromium from the target analyte
list for the NATTS program is an example of this. Other pollutants may prove
problematic from a sampling and/or analytical stand point and can be removed from
the target analyte list due to uncertainties associated with its analytical results. In
addition, studies may indicate that one analytical method is better than another at
providing accurate results for a given pollutant. All of these factors should be
considered when determining the pollutants for which to monitor.
Require consistency in sampling and analytical methods. The development of the
NATTS program has shown that there are inconsistencies in collection and analytical
methods that make data comparison difficult across agencies. Requiring agencies to
use specified and accepted measurement methods, consistent with the guidelines
presented in the NATTS TAD, is integral to the identification of trends and
measuring the effectiveness of regulation. At the time of this report, the NATTS TAD
is undergoing revisions by EPA. When completed and released it is expected that the
revised document will enhance method consistency.
Perform case studies based on findings from the annual report. Often, the annual
report identifies an interesting tendency or trend, or highlights an event at a particular
site(s). For example, dichloromethane concentrations have been highest at BTUT and
GPCO for multiple years and trichloroethylene concentrations have been highest at
SPIL for multiple years. Further examination of the data in conjunction with
meteorological phenomena and potential emissions events or incidents, or further site
characterization may help identify state and local agencies pinpoint issues affecting
air quality in their area.
Consider more rigorous study of the effect of automobile emissions on ambient air
quality using multiple years of data. Because many NMP sites have generated years
of continuous data, a real opportunity exists to evaluate the importance and impact of
automobile emissions on ambient air quality. Suggested areas of study include
additional signature compound assessments and parking lot characterizations.
Develop and/or verify HAP and VOC emissions inventories. State/local/tribal
agencies should use the data collected from NMP sites to develop and validate
emissions inventories, or at the very least, identify and/or verify emissions sources of
concern. Ideally, state/local/tribal agencies would compare the ambient monitoring
results with an emissions inventory for source category completeness. The emissions
inventory could then be used to develop modeled concentrations useful to compare
against ambient monitoring data.
Promulgate ambient air standards for HAPs. Concentrations of several pollutants
sampled during the 2013 program year were greater than risk screening values
32-32

-------
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).
32-33

-------
33.0 References
ACE, 2015. U.S. Army Corps of Engineers. Huntington District, Grayson Lake website.
http://www.lrh.usace.armv.mil/Missions/CivilWorks/Recreation/Kentucky/GravsonLake.
aspx Date Last Accessed: 6/5/2014.
AK DMV, 2014. State of Alaska, Department of Administration, Division of Motor Vehicles.
2013 Currently Registered Vehicles, http://doa.alaska.gov/dmv/research/curreg 13.htm
Date Last Accessed: 8/20/2014.
AK DOT, 2012. Alaska Department of Transportation and Public Facilities. Anchorage Traffic
Map 2012. http://www.dot.state.ak.us/stwdplng/mapping/adt.shtml
Date Last Accessed: 8/25/2014.
ASTM, 2012. ASTM, International. ASTM D7614-12 Standard Test Method for Determination
of Total Suspended Particulate (TSP) Hexavalent Chromium in Ambient Air Analyzed
by Ion Chromatography (IC) and Spectrophotometric Measurements.
http://www.astm.org/Standards/D7614.htm Date Last Accessed: 10/27/2014.
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: 10/27/2014.
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: 11/18/15.
AZ DOT, 2014. Arizona Department of Transportation. Transportation Data Management
System website. 2010 and 2011 AADTs.
http://adot.ms2soft.com/tcds/tsearch. asp?loc=Adot&mod=
Date Last Accessed: 8/25/2014.
Bair, 1992. Bair, Frank E. 1992. Weather of U.S. Cities, 4th ed. Detroit, MI: Gale Research Inc.
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. Page 263.
CA DMV, 2013. California Department of Motor Vehicles. Estimated Vehicles Registered by
County for the Period of January 1 through December 31, 2013.
http://www.dmv.ca.gov/about/profile/what dmv does.htm
Date Last Accessed: 8/20/2014.
33-1

-------
CA DOT, 2013a. State of California Department of Transportation, Division of Traffic
Operations. 2013 Traffic Volumes on California State Highways. Sacramento, CA.
http://www.dot.ca.gov/hq/traffops/saferesr/trafdata/ Date Last Accessed: 8/25/2014.
CA DOT, 2013b. California State Transportation Agency. October 2013. 2012 California Public
Road Data, http://www.dot.ca.gov/hq/tsip/hpms/index.php
Date Last Accessed: 8/21/2014.
Calvert City, 2015. Calvert City - Industry website, http://www.calvertcitv.com/industry.htm
Date Last Accessed: 6/5/2015.
Carbondale, 2015. Carbondale, Colorado, The Town website, http://www.carbondale.com/town
Date Last Accessed: 5/4/2015.
CCRPC, 2014. Chittenden County Regional Planning Commission. Chittenden County Traffic
Counts website. 2009 and 2011 Data, http://www.ccrpcvt.org/database/
Date Last Accessed: 8/27/2014.
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: 10/9/2014.
Census Bureau, 2013b. U.S. Census Bureau. Metropolitan and Micropolitan Statistical Areas
website, http://www.census.gov/population/metro/ Date Last Accessed: 10/9/2014.
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/CQ-EST2012-01.html
Date Last Accessed: 8/6/2014.
Census Bureau, 2014. U.S. Census Bureau. Annual Estimates of the Resident Population:
April 1, 2010 to July 1, 2013. Released March 2014.
http://www.census.gov/popest/data/counties/totals/2013/CQ-EST2013-01.html
Date Last Accessed: 8/6/2014.
CO DOR, 2013. Colorado Department of Revenue. December 24, 2013. Colorado Department of
Revenue 2013 Annual Report July 1, 2012-June 30, 2013. Denver, CO.
http://www.colorado.gov/cs/Satellite?c=Page&cid=1213867975035&pagename=Revenu
e-Main%2FXRMLavout Date Last Accessed: 8/20/2014.
CO DOT, 2013a. Colorado Department of Transportation. Online Transportation Information
System (OTIS), Traffic Data Explorer. 2013 data.
http://dtdapps.coloradodot.info/Otis/TrafficData Date Last Accessed: 8/25/2014.
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: 7/20/2015.
DC DOT, 2012. District Department of Transportation. October 2012. 2011 Traffic Volumes
Map. Washington, D.C. http://ddot.dc.gov/node/532932 Date Last Accessed: 8/25/2014.
33-2

-------
Deer Park, 2015. Deer Park Parks and Recreation Department. Sports Complexes website.
http://www.deerparktx.gov/Index.aspx?NID=770 Date Last Accessed: 7/21/2015.
EPA, 1998. U.S. EPA. September 1998. Technical Assistance Document for Sampling and
Analysis of Ozone Precursors. EPA/600-R-98/161. Research Triangle Park, NC.
http://www.epa.gov/ttnamtil/pamsguidance.html Date Last Accessed: 10/15/2014.
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:
10/15/2014.
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: 10/27/2014.
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: 10/27/2014.
EPA, 1999d. U.S. EPA. June 1999. Compendium Method 10-3.5: Determination of Metals in
Ambient Particulate Matter Using Inductively Coupled Plasma/Mass Spectrometry
(ICP/MS). EPA/625/R-96/01 Oa. Cincinnati, OH.
http://www.epa.gov/ttn/amtic/files/ambient/inorganic/mthd-3-5.pdf
Date Last Accessed: 10/27/2014.
EPA, 2007. U.S. EPA. May 2007. 1990-2002 NEI HAP Trends: Success of CAA Air Toxic
Programs in Reducing HAP Emissions and Risk. Paper presented by Anne Pope at the
16th Annual International Emission Inventory Conference. Raleigh, NC.
http://www.epa.g0v/ttn/chief/conference/eil6/session6/a.pope.pdf
Date Last Accessed: 10/30/2015.
EPA, 2009a. U.S. EPA. April 2009. UATMP, NATTS, CSATAM, PAMS, and NMOC Support,
Contract No. EP-D-09-048. 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: 10/9/2014.
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
Date Last Accessed: 10/28/2014.
33-3

-------
EPA, 2010a. U.S. EPA. October 2010. A Preliminary Risk-based Screening Approach for Air
Toxics Monitoring Data Sets, version 2. EPA-904-B-06-001. Atlanta, GA.
http://www.epa.gov/region4/air/airtoxic/Screening 111610 KMEL.pdf
Date Last Accessed: 10/28/2014.
EPA, 2010b. U.S. EPA. December 2010. Data Quality Evaluation Guidelines for Ambient Air
Acrolein Measurements.
http://www.epa.gov/ttn/amtic/files/ambient/airtox/20101217acroleindataqualitveval.pdf
Date Last Accessed: 10/28/2014.
EPA, 201 la. U.S. EPA. June 2011. Technical Report for School: Assessing Outdoor Air Near
Schools: Riggins School, North Birmingham Elementary School, and Lewis Elementary
School (Birmingham, Alabama), http://www.epa.gov/schoolair/NorthBirmi.html
Date Last Accessed: 10/28/2014.
EPA, 201 lb. U.S EPA. March 2011. The 2005 National-Scale Air Toxics Assessment (NATA)
glossary website, http://www.epa.gov/ttn/atw/natamain/gloss 1 .html
Date Last Accessed: 10/28/2014.
EPA, 201 lc. 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. Federal Register Volume 77, Number
106, Page 3,2632. http://www.gpo.gov/fdsvs/pkg/FR-2012-06-01/html/2012-1335Q.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, 2014a. U.S. EPA. Air Toxics - National Air Toxics Trends Stations website.
http://www.epa.gov/ttnamtil/natts.html Date Last Accessed: 10/9/2014.
EPA, 2014b. U.S. EPA. Technology Transfer Network (TTN), Air Quality System (AQS)
website, http://www.epa.gov/ttn/airs/airsaqs/ Date Last Accessed: 10/9/2014.
EPA, 2014c. U.S. EPA. 1986. 40 CFR Part 136, Appendix B. Definition and Procedure for the
Determination of the Method Detection Limit, Revision 1.11. http://www.ecfr.gov/cgi-
bin/text-idx?c=ecfr&tpl=/ecfrbrowse/Title40/40cfrl36 main 02.tpl
Date Last Accessed: 10/15/2014.
EPA, 2014d. U.S. EPA. Technology Transfer Network, About Air Toxics website.
http://www.epa.gov/ttn/atw/allabout.html Date Last Accessed: 10/28/2014.
EPA, 2014e. Email from Ted Palma, U.S. EPA, OAQPS. August 18, 2014. Dose Response
Library spreadsheet.
EPA, 2014f. Email from Donnette Sturdivant, U.S. EPA, Region 4. September 2014. Columbus
Monitoring Sites l-14-14.pdf.
33-4

-------
EPA, 2015a. U.S. EPA. March 4, 2015. 2011 National Emissions Inventory, Version 2.
http://www.epa.gov/ttnchiel/net/201 linventory.html Date Last Accessed: 4/1/2015.
EPA, 2015b. U.S. EPA. Climate Change: Basic Information website.
http://epa.gov/climatechange/basics/ Date Last Accessed: 1/28/2015.
EPA, 2015c. U.S. EPA. Overview of Greenhouse Gases website.
http://epa.gov/climatechange/ghgemissions/gases.html Date Last Accessed: 1/28/2015.
EPA, 2015d. U.S. EPA. The Accelerated Phaseout of Class I Ozone-Depleting Substances
website, http://www.epa.gov/ozone/title6/phaseout/accfact.html
Date Last Accessed: 8/19/2015.
EPA, 2015e. U.S. EPA. Envirofacts Multisystem Search website.
http://www.epa.gov/enviro/facts/multisvstem.html Date Last Accessed: 10/1/2015.
EPA, 2015f. U.S. EPA, Region 4. Kerr-McGee Chemical Corp - Columbus.
http://www.epa.gov/region4/superfund/sites/npl/mississippi/kerrmcgms.html
Date Last Accessed: 7/13/2015.
EPA, 2015g. U.S EPA. June 9, 2015. Longhorn Army Ammunition Plant Site Details.
http://www.epa.gov/region6/6sf/pdffiles/longhorn-tx.pdfDate Last Accessed: 7/21/2015.
ERG, 2013. Eastern Research Group, Inc. 2013. Support for the EPA National Monitoring
Programs (UATMP, NATTS, C SAT AM, PAMS, and NMOC Support), Quality
Assurance Project Plan, Category 1. Contract No. EP-D-09-048. Morrisville, NC.
FAC, 2007. Report of the Federal Advisory Committee on Detection and Quantitation
Approaches and Uses In Clean Water Act Programs, Appendix D: DQ FAC Single
Laboratory Procedure v2.4. December 2007. Submitted to the U.S. EPA.
http://water.epa.gov/scitech/methods/cwa/det/ Date Last Accessed: 10/15/2014.
FCC, 2015. 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: 5/13/2015.
FHWA, 2013a. Federal Highway Administration. March 2013. Highway Statistics 2011.
Glossary website.
http://www.fhwa.dot.gov/policvinformation/statistics/2011/16 glossary.cfm
Date Last Accessed: 11/18/2015.
FHWA, 2013b. Federal Highway Administration. October 2013. Highway Statistics 2012.
Functional System Travel-2012, Annual Vehicle-Miles. Table VM-2.
http://www.fhwa.dot.gov/policvinformation/statistics/2012/vm2.cfm
Date Last Accessed: 8/21/2014.
FHWA, 2014. Federal Highway Administration. January 2014. Highway Statistics 2012. State
Motor Vehicle Registrations-2012. Table MV-1.
http://www.fhwa.dot.gov/policvinformation/statistics/2012/mvl.cfm
Date Last Accessed: 8/20/2014.
33-5

-------
FL DHSMV, 2013. Florida Department of Highway Safety and Motor Vehicles, Division of
Motorist Services. Registered Vehicles by County as of December 1, 2013.
http://flhsmv.gov/html/reports and statistics/reports and statistics.html
Date Last Accessed: 8/20/2014.
FL DOT, 2013a. Florida Department of Transportation. Florida Traffic Online Mapping
Application (2013). http://www2.dot.state.fl.us/FloridaTrafficOnline/viewer.html
Date Last Accessed: 8/25/2014.
FL DOT, 2013b. Florida Department of Transportation, Transportation Statistics Office. Public
Road Mileage and Miles Traveled, 2013.
http://www.dot.state.fl.us/planning/statistics/mileage-rpts/public.shtm
Date Last Accessed: 8/21/2014.
GA DOR, 2013. Georgia Department of Revenue. Motor Vehicle Division website. Registrations
as of May 31, 2013. http: //motor. etax. dor. ga. gov/ stats/renewal s stats. aspx
Date Last Accessed: 5/31/2013.
GA DOT, 2012. Georgia Department of Transportation. Traffic Counts in Georgia mapping
application. 2012 Data.
http://www.dot.ga.gov/informationcenter/statistics/trafflcdata/Pages/default.aspx
Date Last Accessed: 8/25/2014.
GA DOT, 2013. Georgia Department of Transportation, Office of Transportation Data. Mileage
by Route and Road System Report 445 for 2013.
http://www.dot.ga.gov/informationcenter/statistics/RoadData/Pages/default.aspx
Date Last Accessed: 9/9/2014.
GCPH, 2014. Garfield County Public Health Department. June 2014, revised September 2014.
Garfield County 2013 Air Quality Monitoring Report. Rifle, CO. Prepared by Air
Resource Specialists, Inc., Ft. Collins, CO. http://www.garfleld-countv.com/air-
qualitv/air-monitoring-2012.aspx Date Last Accessed: 5/4/2015.
GCRBD, 2014. Email from Rayjean Kramer, Garfield County Road and Bridge Department.
August 25, 2014. 2014 Data.
GSCO, 1998. Georgia State Climate Office, University of Georgia. 1998. Climatology of the
Georgia Piedmont. Athens, Georgia.
http://coastgis.marsci.uga.edu/summit/climatepied.htm Date Last Accessed: 11/12/2015.
HCPID, 2014. Harris County Public Infrastructure Department, Architecture and Engineering
Department. March 31, 2014. Traffic Volumes, Update 2013. Data from December 2004.
Houston, Texas, http://www.eng.hctx.net/traffic/traffic.htm
Date Last Accessed: 8/26/2014.
Henrico County, 2014. Email from Deborah Oxenreider, County of Henrico, Revenue Division.
August 20, 2014. 2013 Data.
33-6

-------
IL DOT, 2012/2013. Illinois Department of Transportation. Getting Around Illinois website.
Average Annual Daily Traffic map viewer. 2012 and 2013 data.
http://www.gettingaroundillinois.com/ Date Last Accessed: 8/25/2014.
IL DOT, 2013. Illinois Department of Transportation, Office of Planning and Programming.
Illinois Travel Statistics 2013. http://www.idot.illinois.gov/transportation-
svstem/Network-Overview/highwav-svstem/illinois-travel-stati sties
Date Last Accessed: 8/21/2014.
IL SCO, 2015. State Climatologist 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: 5/29/2015.
IL SOS, 2014. Illinois Secretary of State, Vehicle Services Department. August 15, 2014.
Vehicle Counts by County, 2014.
http://www.cvberdriveillinois.com/departments/vehicles/statistics/home.html
Date Last Accessed: 8/20/2014.
IN BMV, 2014. Email from Indiana Bureau of Motor Vehicles Response Team.
October 19, 2014.
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: 8/26/2014.
IN DOT, 2013. Indiana Department of Transportation. Mileage and DVMT by County. 2013
Data, http://www.in.gov/indot/2469.htm Date Last Accessed: 8/21/2014.
IPCC, 2014. Intergovernmental Panel on Climate Change, 2014. Climate Change 2014:
Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment
Report of the Intergovernmental Panel on Climate Change [Core Writing Team, R.K.
Pachauri and L.A. Meyer (eds.)]. IPCC, Geneva, Switzerland, Page 151.
http://www.ipcc.ch/report/ar5/index.shtml Date Last Accessed: 11/10/2015.
ISCO, 2002. Indiana State Climate Office, Purdue University. About Indiana Climate website, as
revised by Ken Scheeringa. December 2002.
http://climate.agry.purdue.edu/climate/narrative.asp Date Last Accessed: 6/4/2015.
KY, 2015. Kentucky State Parks. Grayson Lake History website.
http: //parks. ky. gov/parks/recreati onparks/ gray son-1 ake/hi story. aspx
Date Last Accessed: 6/5/2015.
KYTC, 2014a. Email from Justin Carmack, Kentucky Transportation Cabinet, Division of Motor
Vehicle Licensing. August 21, 2014.
KYTC, 2014b. Kentucky Transportation Cabinet. Interactive Statewide Traffic Counts mapping
application. 2011-2013 Data. http://transportation.kv.gov/Planning/Pages/count-
maps.aspx Date Last Accessed: 8/26/2014.
33-7

-------
KYTC, 2014c. Kentucky Transportation Cabinet. May 27, 2014. 2013 DVMT (in thousands) and
Mileage by County and Functional Class.
http://transportation.kv.gov/Planning/Pages/Roadwav-Information-and-Data.aspx
Date Last Accessed: 8/21/2014.
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: 11/18/2015.
Lakes, 2011. Lakes Environmental. June 2011. WRPLOT View 7.0.0.
http://www.weblakes.com/products/wrplot/index.html Date Last Accessed: 11/10/2015.
MA DOT, 2010. Massachusetts Department of Transportation, Highway Division.
Transportation Data Management System. 2010 Traffic Count.
http://www.mhd. state. ma.us/default.asp?pgid=content/traffic01&sid=about
Date Last Accessed: 8/26/2014.
MA DOT, 2014. Telephone conversation with Bob Frey, Massachusetts Department of
Transportation, Office of Transportation Planning. October 2, 2014.
MA RMV, 2014. Email from Elizabeth Rizzuto, Massachusetts Department of Transportation,
Registry of Motor Vehicles. October 14, 2014. Data as of October 5, 2013.
MCC, 2015. Missouri Climate Center, University of Missouri. Wayne L. Decker. Climate of
Missouri website, http://climate.missouri.edu/climate.php
Date Last Accessed: 5/29/2015.
MCWG, 2015. 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: 6/25/2015.
MDS, 2014. Email from Arlene Rothwell, Michigan Department of State. August 20, 2014. Non-
Expired Total Registrations on File as of November 2013.
MI DOT, 2013. Michigan Department of Transportation. Traffic Monitoring Information System
(TMIS) website. Annual Average Daily Traffic Report (2013).
http://mdotnetpublic.state.mi.us/tmispublic/Search.aspx Date Last Accessed: 8/26/2014.
MI DOT, 2014. Email from Don Howe, Michigan Department of Transportation.
August 25, 2014. 2013 AVMT Data.
MN DOT, 2012. Minnesota Department of Transportation, Traffic Forecasting and Analysis. St.
Cloud 2012 Municipality Map. 2009 Traffic Data.
http://www.dot.state.mn.us/traffic/data/maps-muni-2012.html
Date Last Accessed: 8/26/2014.
MNDOT, 2014. Minnesota Department of Transportation. July 24, 2014. Daily (Average) and
Annual (Total) Vehicle Miles - 2013. http://www.dot.state.mn.us/roadway/data/data-
products, html Date Last Accessed: 8/21/2014.
33-8

-------
MN DPS, 2014. Minnesota Department of Public Safety, Driver and Vehicle Services.
February 20, 2014. Motor Vehicle County Summary Report 2013 (PSVH7301).
https://dps.mn.gov/divisions/dvs/forms-documents/Pages/motor-vehicle-forms.aspx
Date Last Accessed: 8/20/2014.
MO DOR, 2014. Missouri Department of Revenue. Total Motor Vehicles by Kind of Vehicle
within County (DM45080-01), as of 12/31/13. Run Date 1/15/14.
http://dor.mo.gov/publicreports/index.php Date Last Accessed: 8/20/2014.
MO DOT, 2013. Missouri Department of Transportation, Transportation Planning. 2013 St.
Louis District Traffic Volume and Commercial Vehicle Count Map. Jefferson City, MO.
http://www.modot.org/safetv/trafficvolumemaps.htm Date Last Accessed: 8/26/2014.
MO DOT, 2014. Email from Tina Schmidtz, Missouri Department of Transportation.
August 21, 2014. Daily VMT by District, County and Functional Class for St. Louis City
and County, Traffic Year 2013.
MPCA, 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: 6/25/2015.
MPCA, 2015. Email from Cassie McMahon, Minnesota Pollution Control Agency.
June 25, 2015. Air Monitoring Summary: Foundry Circle-St. Cloud.
MS DOR, 2014. State of Mississippi, Department of Revenue. December 2013. Annual Report,
Fiscal Year 2013. http://www.dor.ms.gov/info/stats/main.html
Date Last Accessed: 8/20/2014.
MS DOT, 2013. Mississippi Department of Transportation. MDOT Traffic Count Application.
2013 data.
http://sp.rndot.rns.gov/Office%20of%20FIighwavs/Planning/Maps/Pages/Traffic-
Volume-Maps.aspx Date Last Accessed: 8/26/2014.
MS DOT, 2014. Mississippi Department of Transportation. July 18, 2014. 2012 Mississippi
Public Roads Selected Statistics Extent, Travel, and Designation.
http://mdot.ms.gov/portal/planning.aspx?open=Maps Date Last Accessed: 8/21/2014.
MSU, 2015a. Michigan State Climatologist's Office, Michigan State University. Climate of
Dearborn Narrative, http://climate.geo.msu.edu/climate mi/stations/2015/
Date Last Accessed: 6/28/2015.
MSU, 2015b. Michigan State Climatologist's Office, Michigan State University. 29 Year
Summary of Annual Values for Dearborn (2015) 1981-2010.
http://climate.geo.msu.edu/climate mi/stations/2015/ Date Last Accessed: 6/28/2015.
NCDC, 2013. National Climatic Data Center. Quality Controlled Local Climatological Data
(QCLCD). 2013 Data, https://nes.ncdc.noaa.gov/subscriptions.html
Date Last Accessed: 4/1/2015.
33-9

-------
NCDC, 2014. National Climatic Data Center. Quality Controlled Local Climatological Data
(QCLCD). 2014 Data, http://cdo.ncdc.noaa.gov/qclcd/OCLCD?prior=N
Date Last Accessed: 4/1/2015.
NCDC, 2015. National Climatic Data Center. U.S. Climate Normals website. Climate of the
States, Climatography of the United States No. 60. http://www.ncdc.noaa.gov/cgi-
bin/climatenormals/climatenormals.pl?directive=prod select2&prodtvpe=CLIM60&subr
num= Date Last Accessed: 7/14/2015.
NHMU, 2015. Natural History Museum of Utah. Stefan Brems. Salt Lake Valley Weather
Patterns.
https://nhmu.utah.edu/sites/default/files/attachments/SaltLakeVallevWeatherPatterns.pdf
Date Last Accessed: 7/28/2015.
NJ DOT, 2012. New Jersey Department of Transportation. New Jersey's Roadway Mileage and
Daily VMT by Functional Classification Distributed by County. Year 2012.
http://www.state.ni.us/transportation/refdata/roadwav/vmt.shtm
Date Last Accessed: 8/22/2014.
NJ DOT, 2014. New Jersey Department of Transportation. Roadway Information and Traffic
Monitoring System Program mapping application. 2009, 2012 Data.
http://www.state.ni.us/transportation/refdata/roadwav/traffic counts/
Date Last Accessed: 8/26/2014.
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.ni statelib.org:8080/xmlui/bitstream/handle/10929/23355/f4912008d.pdf?seq
uence= 1 &isAllowed=v Date Last Accessed: 8/26/2014.
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: 4/30/2015.
NOAA, 2015a. NOAA, National Weather Service. NWS glossary website.
http://weather.gov/glossary/ Date Last Accessed: 1/28/2015.
NOAA, 2015b. NOAA, Storm Prediction Center. Roger Edwards. The Online Tornado FAQ,
Frequently Asked Questions about Tornadoes website. Last modified 3/5/2015.
http://www.spc.noaa.gov/faq/tornado/ Date Last Accessed: 7/7/2015.
NOAA, 2015c. NOAA, Climate Prediction Center. Climate Divisions with Counties website.
http://www.cpc.ncep.noaa.gov/products/analvsis monitoring/regional monitoring/CLIM
DIVS/states counties climate-divisions.shtml Date Last Accessed: 8/3/2015.
NYS DMV, 2013. New York State Department of Motor Vehicles. Vehicle Registrations in
Force (2013). http://www.dmv.ny.gov/stats.htm Date Last Accessed: 8/20/2014.
33-10

-------
NYS DOT, 2012. New York State Department of Transportation. Traffic Data Viewer mapping
application. 2012 Data.
https://www.dot.nv.gov/divisions/engineering/applications/traffic-data-viewer
Date Last Accessed: 8/26/2014.
OK DEQ, 2013. Email from Randy Ward, Oklahoma Department of Environmental Quality.
August 29, 2013. ADOK site description additions.
OK DOT, 2012. Oklahoma Department of Transportation, Planning and Research Division. 2012
Annual Average Daily Traffic Oklahoma Highway System maps.
http://www.okladot.state.ok.us/maps/aadt/ Date Last Accessed: 8/26/2014.
OK DOT, 2014. Email from Ron Maxwell, Oklahoma Department of Transportation.
October 6, 2014. Spreadsheet with 2013 County-level HPMS data.
OKTC, 2013. Oklahoma Tax Commission, Motor Vehicle Division. Annual Vehicle Registration
Report, Fiscal Year July 1, 2012 - June 30, 2013. Oklahoma City, OK.
http://www.tax.ok.gov/annrpts.html Date Last Accessed: 8/20/2014.
Peltier, 2012. Peltier Technical Services, Inc. Box and Whisker Plot Utility. Downloaded
April 5, 2012. http://peltiertech.com/Utilitv/ Date Last Accessed: 11/10/2015.
Podmolik, 2015. Podmolik, Mary Ellen. 2015. Canadian Pacific seeks new use for Schiller Park
intermodal site. Chicago Tribune. January 20, 2015.
http://www.chicagotribune.com/business/breaking/ct-railwav-terminal-development-
0121-biz-20150120-story.html Date Last Accessed: 5/29/2015.
POLB, 2015. Port of Long Beach. About the Port website.
http://www.polb.com/about/default.asp Date Last Accessed: 4/30/2015.
RI DOT, 2009. Rhode Island Department of Transportation. December 2009. State Highway
Map of Rhode Island, Traffic Flow Map 2009.
http://www.dot.ri.gov/about/maproom/index.php Date Last Accessed: 8/26/2014.
Rutgers, 2015. Office of the New Jersey State Climatologist, Rutgers University. The Climate of
New Jersey website, http://climate.rutgers.edu/stateclim vl/niclimoverview.html
Date Last Accessed: 6/30/2015.
SC DMV, 2013. South Carolina Department of Motor Vehicles. Historical Information and
Statistics website. S.C. Registered Vehicles (12/2013).
http://www.scdmvonline.com/DMVNew/default.aspx?n=historical information
Date Last Accessed: 8/20/2014.
SC DOT, 2013. South Carolina Department of Transportation. Interactive Traffic Count Map,
SCDOT Traffic Counts 2013. http://www.scdot.org/getting/annualtraffic.aspx
Date Last Accessed: 8/26/2014.
SC DOT, 2014. Email from Angela Hance, South Carolina Department of Transportation.
August 22, 2014. DVMT By County as of December 31, 2013.
33-11

-------
SC SCO, 2015. South Carolina Department of Natural Resources, State Climatology Office.
South Carolina Climate website.
http://www.dnr.sc.gov/climate/sco/ClimateData/cli sc climate.php
Date Last Accessed: 7/21/2015.
Seattle, 2015. Seattle Parks and Recreation. Jefferson Park webpage.
http://www.seattle.gov/parks/park detail.asp?ID=l 14 Date Last Accessed: 8/21/2015.
TAMU, 2015. 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: 7/22/2015.
TDWR, 1983. Texas Department of Water Resources. December 1983. Climatic Atlas of Texas.
LP-192. Austin, TX.
https://www.twdb.texas.gov/publications/reports/limited printing/index.asp
Date Last Accessed: 7/22/2015.
TX DMV, 2014. Email from Diane Emrick-Dodson, Texas Department of Motor Vehicles,
Vehicle Titles and Registration Division. August 1, 2013. Vehicle Registration and
License Transactions and Related Collections by County Calendar Year January 1, 2013
to December 31,2013.
TX DOT, 2013a. Texas Department of Transportation. District Traffic Maps - 2012. 2012
Atlanta District Traffic Map. http://www.txdot.gov/inside-txdot/division/transportation-
planning/maps/traffic.html Date Last Accessed: 8/26/2014.
TXDOT, 2013b. Texas Department of Transportation. District and County Statistics (DISCOS)
website. Daily Vehicle Miles (9/1/12 thru 8/31/13) for the Atlanta and Houston Districts.
http://www.txdot.gov/inside-txdot/division/finance/discos.html
Date Last Accessed: 8/22/2014.
UT DOT, 2012. Utah Department of Transportation. AADT Traffic Map Updated Using 2012
Data (via Google Earth kmz file).
http://www.udot.utah.gov/main/f?p=100:pg:0::::V.T:.528 Date Last Accessed: 8/26/2014.
UT DOT, 2014. Utah Department of Transportation. July 10, 2014. 2013 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: 8/22/2014.
UT TC, 2013. Utah Tax Commission. Utah Current Registrations 2013, On-road Registrations
by County and Vehicle Type. Run date 2/11/2013.
http://tax.utah.gov/econstats/mv/registrations-201Q Date Last Accessed: 11/10/2014.
UVM, 2015. The University of Vermont. Proctor Maple Research Center Mission webpage.
http://www.uvm.edu/~pmrc/?Page=mission.htm Date Last Accessed: 8/3/2015.
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: 8/26/2014.
33-12

-------
VA DOT, 2014. Virginia Department of Transportation. August 5, 2014. Daily Vehicle Miles
Traveled (DVMT) by Physical Jurisdiction, with Towns Combined into Counties
(VMT1206). 2013 data, http://www.virginiadot.org/info/2013 traffic data.asp
Date Last Accessed: 9/9/2014.
VT DMV, 2014. Email from Shannon Fassett, Vermont Department of Motor Vehicles.
August 26, 2014. 2013 Data.
Vtrans, 2014a. Vermont Agency of Transportation, Planning, Outreach, and Community Affairs,
Traffic Research Unit. March 2014. 2013 (Route Log) AADTs Federal Aid Urban
Streets, http://vtransplanning.vermont.gov/research/traffic/publications
Date Last Accessed: 8/27/2014.
Vtrans, 2014b. Vermont Agency of Transportation, Highway Research. July 10, 2014. 2013
Miles and Annual Vehicle Miles Traveled (AVMT) by County and Functional Class.
http://vtransplanning.vermont.gov/research/highwav/publications
Date Last Accessed: 8/22/2014.
WI DOT, 2013. Wisconsin Department of Transportation, Division of Motor Vehicles. Facts and
Figures 2013, A Reference Guide. Madison, WI.
http://www.dot.wisconsin.gov/drivers/facts.htm Date Last Accessed: 8/21/2014.
WI DOT, 2014a. Wisconsin Department of Transportation. Roadrunner Interactive Mapping
Application. 2011 data for HOWI, 2013 data for MIWI.
http://www.dot.wisconsin.gov/travel/counts/index.htm Date Last Accessed: 8/27/2014.
WI DOT, 2014b. Wisconsin Department of Transportation. June 24, 2014. 2013 Vehicle Miles
of Travel (VMT) by County, http://www.dot.wisconsin.gov/travel/counts/vmt.htm
Date Last Accessed: 8/22/2014.
WI SCO, 2015a. 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: 8/11/2015.
WI SCO, 2015b. Wisconsin State Climatology Office. Number of days of 1 inch or more of
snow cover per season: Milwaukee Airport (1948/49-2013/14 season) graphic.
http://www.aos.wisc.edu/~sco/clim-history/7cities/milwaukee.html
Date Last Accessed: 8/11/2015.
Wood, 2004. Wood, Richard A. 1992. The Weather Almanac, 11th ed. Farmington Hills, MI:
The Gale Group, Inc.
WRCC, 2014. Western Region Climate Center. Climate Narratives of the States: Climate of
Alaska, Arizona, California, Colorado, Utah, http://www.wrcc.dri.edu/climate-narratives/
Date Last Accessed: 4/30/2015.
WS DOL, 2013. Washington State Department of Licensing. Vehicle/Vessel Fee Collection and
Vehicle Counts Program website. (13) Motor Vehicle Registration by Class and County -
Calendar Year 2013 (revised), http://www.dol.wa.gov/about/vehvesselreports.html
Date Last Accessed: 8/21/2014.
33-13

-------
WS DOT, 2013. Washington State Department of Transportation. 2013 Annual Traffic Report.
http://www.wsdot.wa.gov/mapsdata/travel/annualtrafficreport.htm
Date Last Accessed: 8/27/2014.
33-14

-------
United States	Office of Air Quality Planning and Standards	Publication No. EPA-454/R-15-005a
Environmental Protection	Air Quality Assessment Division	October 2015
Agency	Research Triangle Park, NC

-------