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 a