&EPA
United! S*K
Enviraimerilfll PiutmBmi
Agamy
2011 National Monitoring Programs Annual
Report (UATMP, NATTS, CSATAM)
Volume 1: Main
August 2013
Final Report

-------
                                                            EPA-454/R-13-007a
                                                                  August 2013
2011 National Monitoring Programs Annual Report (UATMP, NATTS, CSATAM)
                             Volume 1: Main
                                  By:
                       Eastern Research Group, Inc.
                          Morrisville, NC 27560
                              Prepared for:
                   Margaret Dougherty and David Shelow
                Office of Air Quality Planning and Standards
                   U.S. Environmental Protection Agency
                     Research Triangle Park, NC 27711
                        Contract No. EP-D-09-048
               Delivery Orders 11, 15, 20, 21, 22, 23, 24, & 25
                   U.S. Environmental Protection Agency
                Office of Air Quality Planning and Standards
                      Air Quality Assessment Division
                     Research Triangle Park, NC 27711

-------
2011 National Monitoring Programs
              Annual Report
      (UATMP, NATTS, CSATAM)
                   Final Report
            EPA Contract No. EP-D-09-048
       Delivery Orders 11,15, 20, 21, 22, 23, 24, & 25
                    Prepared for:

            Margaret Dougherty and David Shelow
           Office of Air Quality Planning and Standards
            U.S. Environmental Protection Agency
              Research Triangle Park, NC 27711
                    Prepared by:

               Eastern Research Group, Inc.
             601 Keystone Park Drive, Suite 700
                 Morrisville, NC 27560
                    August 2013

-------
                                    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	xix
      List of Figures	xx
      List of Tables	xxxiv
      List of Acronyms	xliii
      Abstract	xlv

1.0   Introduction	1-1

      1.1     Background	1-1

      1.2     The Report	1-2

2.0   The 2011 National Monitoring Programs Network	2-1

      2.1     Monitoring Locations	2-1

      2.2     Analytical Methods and Pollutants Targeted for Monitoring	2-12
             2.2.1    VOC and SNMOC Concurrent Sampling and Analytical Methods ... 2-14
             2.2.2    Carbonyl Compound Sampling and Analytical Method	2-18
             2.2.3    PAH Sampling and Analytical Method	2-19
             2.2.4    Metals Sampling and Analytical Method	2-20
             2.2.5    Hexavalent Chromium Sampling and Analytical Method	2-21

      2.3     Sample Collection Schedules	2-22

      2.4     Completeness	2-28

3.0   Summary of the 2011 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     Noncancer Risk-Based Screening Evaluation Using Minimum Risk Levels	3-7

      3.4     Additional Program-Level Analyses of the 2011 National Monitoring
             Programs Dataset	3-8
             3.4.1    The Effect of Mobile Source Emissions on Spatial Variations	3-8
             3.4.2    Variability Analyses	3-9
             3.4.3    Greenhouse Gas Assessment	3-11
                                         in

-------
                          TABLE OF CONTENTS (Continued)
                                                                                    Page
      3.5    Additional Site-Specific Analyses	3-11
             3.5.1    Site Characterization	3-11
             3.5.2    Meteorological Analysis	3-12
                      3.5.2.1  Back Trajectory Analysis	3-13
                      3.5.2.2  Wind Rose Analysis	3-14
             3.5.3    Site-Specific Comparison to Program-level Average
                      Concentrations	3-15
             3.5.4    Site Trends Analysis	3-16
             3.5.5    Preliminary Risk-Based Screening and Pollutants of Interest	3-17
                      3.5.5.1  Emission Tracer Analysis	3-18
                      3.5.5.2  Cancer Risk and Noncancer Hazard Approximations	3-18
                      3.5.5.3  Risk-Based Emissions Assessment	3-19

4.0   Summary of the 2011 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	4-13
             4.1.3    Central Tendency	4-13

      4.2    Preliminary Risk-Based Screening and Pollutants of Interest	4-15
             4.2.1    Concentrations of the Pollutants of Interest	4-21
             4.2.2    Risk-Based Screening Assessment Using MRLs	4-30

      4.3    The Impact of Mobile Sources	4-33
             4.3.1    Mobile Source Emissions	4-33
             4.3.2    Hydrocarbon Concentrations	4-36
             4.3.3    Motor Vehicle Ownership	4-36
             4.3.4    Estimated Traffic Volume	4-38
             4.3.5    Vehicle Miles Traveled	4-39

      4.4    Variability Analysis	4-39
             4.4.1    Coefficient of Variation and Inter-site Variability	4-39
             4.4.2    Quarterly Variability Analysis	4-70

      4.5    Greenhouse Gases	4-104

5.0   Sites in Arizona	5-1

      5.1    Site Characterization	5-1

      5.2    Meteorological Characterization	5-7
             5.2.1    Climate Summary	5-7
             5.2.2    Meteorological Conditions in 2011	5-8

                                           iv

-------
                         TABLE OF CONTENTS (Continued)
                                                                                  Page
             5.2.3    Back Trajectory Analysis	5-10
             5.2.4    Wind Rose Comparison	5-13

      5.3     Pollutants of Interest	5-16

      5.4     Concentrations	5-19
             5.4.1    2011 Concentration Averages	5-19
             5.4.2    Concentration Comparison	5-23
             5.4.3    Concentration Trends	5-28

      5.5     Additional Risk-Based Screening Evaluations	5-32
             5.5.1    Risk-Based  Screening Assessment Using MRLs	5-32
             5.5.2    Cancer Risk and Noncancer Hazard Approximations	5-33
             5.5.3    Risk-Based Emissions Assessment	5-36

      5.6     Summary  of the 2011 Monitoring Data for PXSS and SPAZ	5-40

6.0   Sites in California	6-1

      6.1     Site Characterization	6-1

      6.2     Meteorological Characterization	6-11
             6.2.1    Climate Summary	6-11
             6.2.2    Meteorological Conditions in 2011	6-12
             6.2.3    Back Trajectory Analysis	6-14
             6.2.4    Wind Rose Comparison	6-19

      6.3     Pollutants of Interest	6-24

      6.4     Concentrations	6-26
             6.4.1    2011 Concentration Averages	6-26
             6.4.2    Concentration Comparison	6-28
             6.4.3    Concentration Trends	6-31

      6.5     Additional Risk-Based Screening Evaluations	6-32
             6.5.1    Risk-Based  Screening Assessment Using MRLs	6-32
             6.5.2    Cancer Risk and Noncancer Hazard Approximations	6-32
             6.5.3    Risk-Based Emissions Assessment	6-34

      6.6     Summary  of the 2011 Monitoring Data for CELA, RUCA, and SJJCA	6-40

7.0   Sites in Colorado	7-1

      7.1     Site Characterization	7-1

-------
                         TABLE OF CONTENTS (Continued)
                                                                                  Page
      7.2    Meteorological Characterization	7-12
             7.2.1     Climate Summary	7-12
             7.2.2     Meteorological Conditions in 2011	7-13
             7.2.3     Back Trajectory Analysis	7-16
             7.2.4     Wind Rose Comparison	7-22

      7.3    Pollutants of Interest	7-29

      7.4    Concentrations	7-31
             7.4.1     2011  Concentration Averages	7-32
             7.4.2     Concentration Comparison	7-37
             7.4.3     Concentration Trends	7-42

      7.5    Additional Risk-Based Screening Evaluations	7-47
             7.5.1     Risk-Based  Screening Assessment Using MRLs	7-47
             7.5.2     Cancer Risk and Noncancer Hazard Approximations	7-48
             7.5.3     Risk-Based Emissions Assessment	7-51

      7.6    Summary of the 2011 Monitoring Data for the Colorado Monitoring Sites	7-59

8.0   Site in the District of Columbia                                                8-1

      8.1    Site Characterization	8-1

      8.2    Meteorological Characterization	8-6
             8.2.1     Climate Summary	8-6
             8.2.2     Meteorological Conditions in 2011	8-6
             8.2.3     Back Trajectory Analysis	8-8
             8.2.4     Wind Rose Comparison	8-10

      8.3    Pollutants of Interest	8-12

      8.4    Concentrations	8-13
             8.4.1     2011  Concentration Averages	8-13
             8.4.2     Concentration Comparison	8-15
             8.4.3     Concentration Trends	8-17

      8.5    Additional Risk-Based Screening Evaluations	8-18
             8.5.1     Risk-Based  Screening Assessment Using MRLs	8-18
             8.5.2     Cancer Risk and Noncancer Hazard Approximations	8-19
             8.5.3     Risk-Based Emissions Assessment	8-20

      8.6    Summary of the 2011 Monitoring Data for WADC	8-24
                                          VI

-------
                         TABLE OF CONTENTS (Continued)
                                                                                   Page
9.0   Sites in Florida	9-1

      9.1     Site Characterization	9-1

      9.2     Meteorological Characterization	9-13
             9.2.1     Climate Summary	9-13
             9.2.2     Meteorological Conditions in 2011	9-13
             9.2.3     Back Trajectory Analysis	9-16
             9.2.4     Wind Rose Comparison	9-23

      9.3     Pollutants of Interest	9-31

      9.4     Concentrations	9-33
             9.4.1     2011 Concentration Averages	9-33
             9.4.2     Concentration Comparison	9-36
             9.4.3     Concentration Trends	9-42

      9.5     Additional Risk-Based Screening Evaluations	9-51
             9.5.1     Risk-Based Screening Assessment Using MRLs	9-51
             9.5.2     Cancer Risk and Noncancer Hazard Approximations	9-51
             9.5.3     Risk-Based Emissions Assessment	9-54

      9.6     Summary of the 2011 Monitoring Data for the Florida Monitoring Sites	9-62

10.0  Site in Georgia	10-1

      10.1    Site Characterization	10-1

      10.2    Meteorological Characterization	10-6
             10.2.1    Climate Summary	10-6
             10.2.2    Meteorological Conditions in 2011	10-7
             10.2.3    Back Trajectory Analysis	10-7
             10.2.4    Wind Rose Comparison	10-10

      10.3    Pollutants of Interest	10-12

      10.4    Concentrations	10-13
             10.4.1    2011 Concentration Averages	10-13
             10.4.2    Concentration Comparison	10-15
             10.4.3    Concentration Trends	10-17

      10.5    Additional Risk-Based Screening Evaluations	10-18
             10.5.1    Risk-Based Screening Assessment Using MRLs	10-18
             10.5.2    Cancer Risk and Noncancer Hazard Approximations	10-19
             10.5.3    Risk-Based Emissions Assessment	10-20

                                          vii

-------
                         TABLE OF CONTENTS (Continued)
                                                                                 Page
      10.6   Summary of the 2011 Monitoring Data for SDGA	10-24

11.0  Sites in Illinois	11-1

      11.1   Site Characterization	11-1

      11.2   Meteorological Characterization	11-7
             11.2.1   Climate Summary	11-8
             11.2.2   Meteorological Conditions in 2011	11-8
             11.2.3   Back Trajectory Analysis	11-10
             11.2.4   Wind Rose Comparison	11-13

      11.3   Pollutants of Interest	11-16

      11.4   Concentrations	11-19
             11.4.1   2011 Concentration Averages	11-19
             11.4.2   Concentration Comparison	11-25
             11.4.3   Concentration Trends	11-31

      11.5   Additional Risk-Based Screening Evaluations	11-43
             11.5.1   Risk-Based Screening Assessment Using MRLs	11-43
             11.5.2   Cancer Risk and Noncancer Hazard Approximations	11-43
             11.5.3   Risk-Based Emissions Assessment	11-46

      11.6   Summary of the 2011 Monitoring Data for NBIL and SPIL	11-50

12.0  Sites in Indiana	12-1

      12.1   Site Characterization	12-1

      12.2   Meteorological Characterization	12-8
             12.2.1   Climate Summary	12-8
             12.2.2   Meteorological Conditions in 2011	12-9
             12.2.3   Back Trajectory Analysis	12-11
             12.2.4   Wind Rose Comparison	12-14

      12.3   Pollutants of Interest	12-18

      12.4   Concentrations	12-19
             12.4.1   2011 Concentration Averages	12-19
             12.4.2   Concentration Comparison	12-21
             12.4.3   Concentration Trends	12-22

      12.5   Additional Risk-Based Screening Evaluations	12-25
             12.5.1   Risk-Based Screening Assessment Using MRLs	12-25

                                         viii

-------
                         TABLE OF CONTENTS (Continued)
                                                                                  Page
             12.5.2   Cancer Risk and Noncancer Hazard Approximations	12-26
             12.5.3   Risk-Based Emissions Assessment	12-27

      12.6   Summary of the 2011 Monitoring Data for INDEM and WPIN	12-31

13.0  Site in Kentucky	13-1

      13.1   Site Characterization	13-1

      13.2   Meteorological Characterization	13-6
             13.2.1   Climate Summary	13-6
             13.2.2   Meteorological Conditions in 2011	13-6
             13.2.3   Back Trajectory Analysis	13-8
             13.2.4   Wind Rose Comparison	13-10

      13.3   Pollutants of Interest	13-12

      13.4   Concentrations	13-13
             13.4.1   2011 Concentration Averages	13-13
             13.4.2   Concentration Comparison	13-16
             13.4.3   Concentration Trends	13-19

      13.5   Additional Risk-Based Screening Evaluations	13-19
             13.5.1   Risk-Based Screening Assessment Using MRLs	13-20
             13.5.2   Cancer Risk and Noncancer Hazard Approximations	13-20
             13.5.3   Risk-Based Emissions Assessment	13-22

      13.6   Summary of the 2011 Monitoring Data for GLKY	13-26

14.0  Site in Massachusetts	14-1

      14.1   Site Characterization	14-1

      14.2   Meteorological Characterization	14-6
             14.2.1   Climate Summary	14-6
             14.2.2   Meteorological Conditions in 2011	14-7
             14.2.3   Back Trajectory Analysis	14-7
             14.2.4   Wind Rose Comparison	14-10

      14.3   Pollutants of Interest	14-13

      14.4   Concentrations	14-14
             14.4.1   2011 Concentration Averages	14-14
             14.4.2   Concentration Comparison	14-16
             14.4.3   Concentration Trends	14-20

                                          ix

-------
                         TABLE OF CONTENTS (Continued)
                                                                                  Page
      14.5    Additional Risk-Based Screening Evaluations	14-24
             14.5.1   Risk-Based Screening Assessment Using MRLs	14-24
             14.5.2   Cancer Risk and Noncancer Hazard Approximations	14-24
             14.5.3   Risk-Based Emissions Assessment	14-26

      14.6    Summary of the 2011 Monitoring Data for BOMA	14-30

15.0  Sites in Michigan	15-1

      15.1    Site Characterization	15-1

      15.2    Meteorological Characterization	15-9
             15.2.1   Climate Summary	15-9
             15.2.2   Meteorological Conditions in 2011	15-9
             15.2.3   Back Trajectory Analysis	15-11
             15.2.4   Wind Rose Comparison	15-15

      15.3    Pollutants of Interest	15-19

      15.4    Concentrations	15-21
             15.4.1   2011 Concentration Averages	15-22
             15.4.2   Concentration Comparison	15-26
             15.4.3   Concentration Trends	15-30

      15.5    Additional Risk-Based Screening Evaluations	15-36
             15.5.1   Risk-Based Screening Assessment Using MRLs	15-36
             15.5.2   Cancer Risk and Noncancer Hazard Approximations	15-36
             15.5.3   Risk-Based Emissions Assessment	15-38

      15.6    Summary of the 2011 Monitoring Data for DEMI, RRMI, and SWMI	15-44

16.0  Site in Missouri	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 Conditions in 2011	16-6
             16.2.3   Back Trajectory Analysis	16-8
             16.2.4   Wind Rose Comparison	16-10

      16.3    Pollutants of Interest	16-13

      16.4    Concentrations	16-14
             16.4.1   2011 Concentration Averages	16-15

                                          x

-------
                         TABLE OF CONTENTS (Continued)
                                                                                  Page
             16.4.2   Concentration Comparison	16-19
             16.4.3   Concentration Trends	16-24

      16.5    Additional Risk-Based Screening Evaluations	16-32
             16.5.1   Risk-Based Screening Assessment Using MRLs	16-32
             16.5.2   Cancer Risk and Noncancer Hazard Approximations	16-33
             16.5.3   Risk-Based Emissions Assessment	16-35

      16.6    Summary of the 2011 Monitoring Data for S4MO	16-39

17.0  Sites in New Jersey	17-1

      17.1    Site Characterization	17-1

      17.2    Meteorological Characterization	17-13
             17.2.1   Climate Summary	17-13
             17.2.2   Meteorological Conditions in 2011	17-13
             17.2.3   Back Trajectory Analysis	17-15
             17.2.4   Wind Rose Comparison	17-20

      17.3    Pollutants of Interest	17-26

      17.4    Concentrations	17-30
             17.4.1   2011 Concentration Averages	17-30
             17.4.2   Concentration Comparison	17-36
             17.4.3   Concentration Trends	17-39

      17.5    Additional Risk-Based Screening Evaluations	17-51
             17.5.1   Risk-Based Screening Assessment Using MRLs	17-52
             17.5.2   Cancer Risk and Noncancer Hazard Approximations	17-52
             17.5.3   Risk-Based Emissions Assessment	17-56

      17.6    Summary of the 2011 Monitoring Data for the New Jersey Monitoring
             Sites	17-62

18.0  Sites in New York	18-1

      18.1    Site Characterization	18-1

      18.2    Meteorological Characterization	18-9
             18.2.1   Climate Summary	18-9
             18.2.2   Meteorological Conditions in 2011	18-9
             18.2.3   Back Trajectory Analysis	18-11
             18.2.4   Wind Rose Comparison	18-15
                                          XI

-------
                         TABLE OF CONTENTS (Continued)
                                                                                  Page
      18.3   Pollutants of Interest	18-18

      18.4   Concentrations	18-20
             18.4.1    2011 Concentration Averages	18-20
             18.4.2    Concentration Comparison	18-22
             18.4.3    Concentration Trends	18-25

      18.5   Additional Risk-Based Screening Evaluations	18-25
             18.5.1    Risk-Based Screening Assessment Using MRLs	18-25
             18.5.2    Cancer Risk and Noncancer Hazard Approximations	18-26
             18.5.3    Risk-Based Emissions Assessment	18-27

      18.6   Summary of the 2011 Monitoring Data for MONY and ROCH	18-31

19.0  Sites in Oklahoma                                                           19-1

      19.1   Site Characterization	19-1

      19.2   Meteorological Characterization	19-14
             19.2.1    Climate Summary	19-14
             19.2.2    Meteorological Conditions in 2011	19-14
             19.2.3    Back Trajectory Analysis	19-15
             19.2.4    Wind Rose Comparison	19-23

      19.3   Pollutants of Interest	19-30

      19.4   Concentrations	19-35
             19.4.1    2011 Concentration Averages	19-35
             19.4.2    Concentration Comparison	19-45
             19.4.3    Concentration Trends	19-54

      19.5   Additional Risk-Based Screening Evaluations	19-61
             19.5.1    Risk-Based Screening Assessment Using MRLs	19-61
             19.5.2    Cancer Risk and Noncancer Hazard Approximations	19-61
             19.5.3    Risk-Based Emissions Assessment	19-67

      19.6   Summary of the 2011 Monitoring Data for the Oklahoma Monitoring Sites.. 19-75

20.0  Site in Rhode Island                                                         20-1

      20.1   Site Characterization	20-1

      20.2   Meteorological Characterization	20-6
             20.2.1    Climate Summary	20-6
             20.2.2    Meteorological Conditions in 2011	20-7

                                          xii

-------
                         TABLE OF CONTENTS (Continued)
                                                                                 Page
             20.2.3   Back Trajectory Analysis	20-7
             20.2.4   Wind Rose Comparison	20-10

      20.3    Pollutants of Interest	20-13

      20.4    Concentrations	20-14
             20.4.1   2011 Concentration Averages	20-14
             20.4.2   Concentration Comparison	20-16
             20.4.3   Concentration Trends	20-18

      20.5    Additional Risk-Based Screening Evaluations	20-19
             20.5.1   Risk-Based Screening Assessment Using MRLs	20-19
             20.5.2   Cancer Risk and Noncancer Hazard Approximations	20-19
             20.5.3   Risk-Based Emissions Assessment	20-21

      20.6    Summary of the 2011 Monitoring Data for PRRI	20-25

21.0  Site in South Carolina                                                       21-1

      21.1    Site Characterization	21-1

      21.2    Meteorological Characterization	21-6
             21.2.1   Climate Summary	21-6
             21.2.2   Meteorological Conditions in 2011	21-6
             21.2.3   Back Trajectory Analysis	21-8
             21.2.4   Wind Rose Comparison	21-10

      21.3    Pollutants of Interest	21-12

      21.4    Concentrations	21-13
             21.4.1   2011 Concentration Averages	21-13
             21.4.2   Concentration Comparison	21-15
             21.4.3   Concentration Trends	21-17

      21.5    Additional Risk-Based Screening Evaluations	21-18
             21.5.1   Risk-Based Screening Assessment Using MRLs	21-18
             21.5.2   Cancer Risk and Noncancer Hazard Approximations	21-18
             21.5.3   Risk-Based Emissions Assessment	21-19

      21.6    Summary of the 2011 Monitoring Data for CHSC	21-23

22.0  Sites in South Dakota                                                       22-1

      22.1    Site Characterization	22-1
                                         Xlll

-------
                         TABLE OF CONTENTS (Continued)
                                                                                 Page
      22.2   Meteorological Characterization	22-9
             22.2.1   Climate Summary	22-9
             22.2.2   Meteorological Conditions in 2011	22-9
             22.2.3   Back Trajectory Analysis	22-11
             22.2.4   Wind Rose Comparison	22-14

      22.3   Pollutants of Interest	22-18

      22.4   Concentrations	22-20
             22.4.1   2011 Concentration Averages	22-20
             22.4.2   Concentration Comparison	22-23
             22.4.3   Concentration Trends	22-26

      22.5   Additional Risk-Based Screening Evaluations	22-26
             22.5.1   Risk-Based Screening Assessment Using MRLs	22-27
             22.5.2   Cancer Risk and Noncancer Hazard Approximations	22-27
             22.5.3   Risk-Based Emissions Assessment	22-29

      22.6   Summary of the 2011 Monitoring Data for SSSD and UCSD	22-33

23.0  Sites in Texas	23-1

      23.1   Site Characterization	23-1

      23.2   Meteorological Characterization	23-9
             23.2.1   Climate Summary	23-9
             23.2.2   Meteorological Conditions in 2011	23-9
             23.2.3   Back Trajectory Analysis	23-11
             23.2.4   Wind Rose Comparison	23-14

      23.3   Pollutants of Interest	23-17

      23.4   Concentrations	23-19
             23.4.1   2011 Concentration Averages	23-19
             23.4.2   Concentration Comparison	23-21
             23.4.3   Concentration Trends	23-23

      23.5   Additional Risk-Based Screening Evaluations	23-24
             23.5.1   Risk-Based Screening Assessment Using MRLs	23-24
             23.5.2   Cancer Risk and Noncancer Hazard Approximations	23-24
             23.5.3   Risk-Based Emissions Assessment	23-25

      23.6   Summary of the 2011 Monitoring Data for CAMS 35 and CAMS 85	23-29
                                         xiv

-------
                         TABLE OF CONTENTS (Continued)
                                                                                  Page
24.0  Site in Utah	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 Conditions in 2011	24-7
             24.2.3   Back Trajectory Analysis	24-7
             24.2.4   Wind Rose Comparison	24-10

      24.3    Pollutants of Interest	24-11

      24.4    Concentrations	24-14
             24.4.1   2011 Concentration Averages	24-14
             24.4.2   Concentration Comparison	24-18
             24.4.3   Concentration Trends	24-23

      24.5    Additional Risk-Based Screening Evaluations	24-31
             24.5.1   Risk-Based Screening Assessment Using MRLs	24-31
             24.5.2   Cancer Risk and Noncancer Hazard Approximations	24-32
             24.5.3   Risk-Based Emissions Assessment	24-34

      24.6    Summary of the 2011 Monitoring Data for BTUT	24-38

25.0  Sites in Vermont	25-1

      25.1    Site Characterization	25-1

      25.2    Meteorological Characterization	25-10
             25.2.1   Climate Summary	25-10
             25.2.2   Meteorological Conditions in 2011	25-11
             25.2.3   Back Trajectory Analysis	25-11
             25.2.4   Wind Rose Comparison	25-17

      25.3    Pollutants of Interest	25-21

      25.4    Concentrations	25-24
             25.4.1   2011 Concentration Averages	25-24
             25.4.2   Concentration Comparison	25-29
             25.4.3   Concentration Trends	25-34

      25.5    Additional Risk-Based Screening Evaluations	25-35
             25.5.1   Risk-Based Screening Assessment Using MRLs	25-35
             25.5.2   Cancer Risk and Noncancer Hazard Approximations	25-36
             25.5.3   Risk-Based Emissions Assessment	25-39

                                          xv

-------
                         TABLE OF CONTENTS (Continued)
                                                                                  Page
      25.6   Summary of the 2011 Monitoring Data for the Vermont Monitoring Sites .... 25-45

26.0  Site in Virginia	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 Conditions in 2011	26-6
             26.2.3    Back Trajectory Analysis	26-8
             26.2.4    Wind Rose Comparison	26-10

      26.3   Pollutants of Interest	26-12

      26.4   Concentrations	26-13
             26.4.1    2011  Concentration Averages	26-13
             26.4.2    Concentration Comparison	26-15
             26.43    Concentration Trends	26-16

      26.5   Additional Risk-Based Screening Evaluations	26-17
             26.5.1    Risk-Based  Screening Assessment Using MRLs	26-17
             26.5.2    Cancer Risk and Noncancer Hazard Approximations	26-17
             26.5.3    Risk-Based Emissions Assessment	26-18

      26.6   Summary of the 2011 Monitoring Data for RIVA	26-22

27.0  Site in Washington	27-1

      27.1   Site Characterization	27-1

      27.2   Meteorological Characterization	27-6
             27.2.1    Climate Summary	27-6
             27.2.2    Meteorological Conditions in 2011	27-7
             27.2.3    Back Trajectory Analysis	27-7
             27.2.4    Wind Rose Comparison	27-10

      27.3   Pollutants of Interest	27-12

      27.4   Concentrations	27-14
             27.4.1    2011  Concentration Averages	27-14
             27.4.2    Concentration Comparison	27-18
             27.4.3    Concentration Trends	27-23

      27.5   Additional Risk-Based Screening Evaluations	27-31
             27.5.1    Risk-Based  Screening Assessment Using MRLs	27-31

                                          xvi

-------
                         TABLE OF CONTENTS (Continued)
                                                                                 Page
             27.5.2   Cancer Risk and Noncancer Hazard Approximations	27-31
             27.5.3   Risk-Based Emissions Assessment	27-33

      27.6   Summary of the 2011 Monitoring Data for SEWA	27-37

28.0  Site in Wisconsin	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 Conditions in 2011	28-7
             28.2.3   Back Trajectory Analysis	28-7
             28.2.4   Wind Rose Comparison	28-10

      28.3   Pollutants of Interest	28-12

      28.4   Concentrations	28-13
             28.4.1   2011 Concentration Averages	28-13
             28.4.2   Concentration Comparison	28-14
             28.4.3   Concentration Trends	28-15

      28.5   Additional Risk-Based Screening Evaluations	28-15
             28.5.1   Risk-Based Screening Assessment Using MRLs	28-15
             28.5.2   Cancer Risk and Noncancer Hazard Approximations	28-16
             28.5.3   Risk-Based Emissions Assessment	28-17

      28.6   Summary of the 2011 Monitoring Data for HOWI	28-20

29.0  Data Quality	29-1

      29.1   Completeness	29-1

      29.2   Method Precision	29-2
             29.2.1   VOC Method Precision	29-4
             29.2.2   SNMOC Method Precision	29-12
             29.2.3   Carbonyl Compound Method Precision	29-17
             29.2.4   PAH Method Precision	29-19
             29.2.5   Metals Method Precision	29-20
             29.2.6   Hexavalent Chromium Method Precision	29-21

      29.3   Analytical Precision	29-23
             29.3.1   VOC Analytical Precision	29-24
             29.3.2   SNMOC Analytical Precision	29-32
             29.3.3   Carbonyl Compound Analytical Precision	29-37

                                         xvii

-------
                         TABLE OF CONTENTS (Continued)
                                                                                Page
             29.3.4   PAH Analytical Precision	29-39
             29.3.5   Metals Analytical Precision	29-40
             29.3.6   Hexavalent Chromium Analytical Precision	29-41

      29.4   Accuracy	29-42

30.0  Results, Conclusions, and Recommendations	30-1

      30.1   Summary of Results	30-1
             30.1.1   National-level Results Summary	30-1
             30.1.2   State-level Results Summary	30-2
             30.1.3   Composite Site-level Results Summary	30-19
             30.1.4   Data Quality Results Summary	30-21

      30.2   Conclusions	30-21

      30.3   Recommendations	30-23

31.0  References	31-1
                                        xvin

-------
                        TABLE OF CONTENTS (Continued)
List of Appendices
Appendix A      AQS Site Descriptions for the 2011 NATTS, UATMP, and CSATAM
                Monitoring Sites

Appendix B      Range of Method Detection Limits (MDLs)

Appendix C      2011 VOC Raw Data

Appendix D      2011 SNMOC Raw Data

Appendix E      2011 Carbonyl Compounds Raw Data

Appendix F      2011 PAH Raw Data

Appendix G      2011 Metals Raw Data

Appendix H      2011 Hexavalent Chromium Raw Data

Appendix I       Summary of Invalidated 2011 Samples

Appendix J       2011 Summary Statistics for VOC Monitoring

Appendix K      2011 Summary Statistics for SNMOC Monitoring

Appendix L      2011 Summary Statistics for Carbonyl Compounds Monitoring

Appendix M      2011 Summary Statistics for PAH Monitoring

Appendix N      2011 Summary Statistics for Metals Monitoring

Appendix O      2011 Summary Statistics for Hexavalent Chromium Monitoring

Appendix P      Risk Factors Used Throughout the 2011 NMP Report

Appendix Q      Glossary of Terms
                                       xix

-------
                                 LIST OF FIGURES
2-1    Locations of the 2011 National Monitoring Programs Monitoring Sites	2-3

4-la   Coefficient of Variation Analysis of Acenaphthene Across 23 Sites	4-45
4-lb   Inter-Site Variability for Acenaphthene	4-45
4-2a   Coefficient of Variation Analysis of Acetaldehyde Across 24 Sites	4-46
4-2b   Inter-Site Variability for Acetaldehyde	4-46
4-3a   Coefficient of Variation Analysis of Acrylonitrile Across 22 Sites	4-47
4-3b   Inter-Site Variability for Acrylonitrile	4-47
4-4a   Coefficient of Variation Analysis of Arsenic Across 14 Sites	4-48
4-4b   Inter-Site Variability for Arsenic	4-48
4-5a   Coefficient of Variation Analysis of Benzene Across 23 Sites	4-49
4-5b   Inter-Site Variability for Benzene	4-49
4-6a   Coefficient of Variation Analysis of Benzo(a)pyrene Across 23 Sites	4-50
4-6b   Inter-Site Variability for Benzo(a)pyrene	4-50
4-7a   Coefficient of Variation Analysis of Beryllium Across 14 Sites	4-51
4-7b   Inter-Site Variability for Beryllium	4-51
4-8a   Coefficient of Variation Analysis of 1,3-Butadiene Across 23 Sites	4-52
4-8b   Inter-Site Variability for 1,3-Butadiene	4-52
4-9a   Coefficient of Variation Analysis of Cadmium Across 14 Sites	4-53
4-9b   Inter-Site Variability for Cadmium	4-53
4-10a  Coefficient of Variation Analysis of Carbon Tetrachloride Across 23 Sites	4-54
4-1 Ob  Inter-Site Variability for Carbon Tetrachloride	4-54
4-lla  Coefficient of Variation Analysis of Chloroform Across 23 Sites	4-55
4-llb  Inter-Site Variability for Chloroform	4-55
4-12a  Coefficient of Variation Analysis ofp-Dichlorobenzene Across 23 Sites	4-56
4-12b  Inter-Site Variability for/>-Dichlorobenzene	4-56
4-13a  Coefficient of Variation Analysis of 1,2-Dichloroethane Across 23 Sites	4-57
4-13b  Inter-Site Variability for 1,2-Dichloroethane	4-57
4-14a  Coefficient of Variation Analysis of Ethylbenzene Across 23 Sites	4-58
4-14b  Inter-Site Variability for Ethylbenzene	4-58
4-15a  Coefficient of Variation Analysis of Fluorene Across 23 Sites	4-59
4-15b  Inter-Site Variability for Fluorene	4-59
4-16a  Coefficient of Variation Analysis of Formaldehyde Across 24 Sites	4-60
4-16b  Inter-Site Variability for Formaldehyde	4-60
4-17a  Coefficient of Variation Analysis of Hexachloro-1,3-butadiene Across 23 Sites	4-61
4-17b  Inter-Site Variability for Hexachloro-1,3-butadiene	4-61
4-18a  Coefficient of Variation Analysis of Hexavalent Chromium Across 22 Sites	4-62
4-18b  Inter-Site Variability for Hexavalent Chromium	4-62
4-19a  Coefficient of Variation Analysis of Lead Across 14 Sites	4-63
4-19b  Inter-Site Variability for Lead	4-63
4-20a  Coefficient of Variation Analysis of Manganese Across 14 Sites	4-64
4-20b  Inter-Site Variability for Manganese	4-64
4-21a  Coefficient of Variation Analysis of Naphthalene Across 23 Sites	4-65
4-21b  Inter-Site Variability for Naphthalene	4-65

                                           xx

-------
                         LIST OF FIGURES (Continued)

                                                                                 Page
4-22a  Coefficient of Variation Analysis of Nickel Across 14 Sites	4-66
4-22b  Inter-Site Variability for Nickel	4-66
4-23a  Coefficient of Variation Analysis of Tetrachloroethylene Across 23 Sites	4-67
4-23b  Inter-Site Variability for Tetrachloroethylene	4-67
4-24a  Coefficient of Variation Analysis of Trichloroethylene Across 23 Sites	4-68
4-24b  Inter-Site Variability for Trichloroethylene	4-68
4-25a  Coefficient of Variation Analysis of Vinyl Chloride Across 23 Sites	4-69
4-25b  Inter-Site Variability for Vinyl Chloride	4-69
4-26   Comparison of Average Quarterly Acenaphthene Concentrations	4-73
4-27   Comparison of Average Quarterly Acetaldehyde Concentrations	4-74
4-28   Comparison of Average Quarterly Acrylonitrile Concentrations	4-75
4-29a  Comparison of Average Quarterly Arsenic (PMi0) Concentrations	4-76
4-29b  Comparison of Average Quarterly Arsenic (TSP) Concentrations	4-77
4-30   Comparison of Average Quarterly Benzene Concentrations	4-78
4-31   Comparison of Average Quarterly Benzo(a)pyrene Concentrations	4-79
4-32a  Comparison of Average Quarterly Beryllium (PMio) Concentrations	4-80
4-32b  Comparison of Average Quarterly Beryllium (TSP) Concentrations	4-81
4-33   Comparison of Average Quarterly 1,3-Butadiene Concentrations	4-82
4-34a  Comparison of Average Quarterly Cadmium (PMio) Concentrations	4-83
4-34b  Comparison of Average Quarterly Cadmium (TSP) Concentrations	4-84
4-3 5   Comparison of Average Quarterly Carbon Tetrachloride Concentrations	4-85
4-36   Comparison of Average Quarterly Chloroform Concentrations	4-86
4-37   Comparison of Average Quarterly p-Dich\orobenzene Concentrations	4-87
4-3 8   Comparison of Average Quarterly 1,2-Dichloroethane Concentrations	4-88
4-39   Comparison of Average Quarterly Ethylbenzene Concentrations	4-89
4-40   Comparison of Average Quarterly Fluorene  Concentrations	4-90
4-41   Comparison of Average Quarterly Formaldehyde Concentrations	4-91
4-42   Comparison of Average Quarterly Hexachloro-1,3-butadiene Concentrations	4-92
4-43   Comparison of Average Quarterly Hexavalent Chromium Concentrations	4-93
4-44a  Comparison of Average Quarterly Lead (PMio) Concentrations	4-94
4-44b  Comparison of Average Quarterly Lead (TSP) Concentrations	4-95
4-45a  Comparison of Average Quarterly Manganese (PMio) Concentrations	4-96
4-45b  Comparison of Average Quarterly Manganese (TSP) Concentrations	4-97
4-46   Comparison of Average Quarterly Naphthalene Concentrations	4-98
4-47a  Comparison of Average Quarterly Nickel (PMio) Concentrations	4-99
4-47b  Comparison of Average Quarterly Nickel (TSP) Concentrations	4-100
4-48   Comparison of Average Quarterly Tetrachloroethylene Concentrations	4-101
4-49   Comparison of Average Quarterly Trichloroethylene Concentrations	4-102
4-50   Comparison of Average Quarterly Vinyl Chloride Concentrations	4-103

5-1    Phoenix, Arizona (PXSS) Monitoring Site	5-2
5-2    South Phoenix, Arizona (SPAZ) Monitoring Site	5-3
5-3    NEI Point Sources Located Within 10 Miles of PXSS and SPAZ	5-4
5-4    2011 Composite Back Trajectory Map for PXSS	5-11
5-5    Back Trajectory Cluster Map for PXSS	5-11

                                         xxi

-------
                         LIST OF FIGURES (Continued)

                                                                                 Page
5-6    2011 Composite Back Trajectory Map for SPAZ	5-12
5-7    Back Trajectory Cluster Map for SPAZ	5-12
5-8    Wind Roses for the Phoenix Sky Harbor International Airport Weather Station
       nearPXSS	5-14
5-9    Wind Roses for the Phoenix Sky Harbor International Airport Weather Station
       near SPAZ	5-15
5-10   Program vs. Site-Specific Average Arsenic (PMio) Concentration	5-24
5-11   Program vs. Site-Specific Average Benzene Concentrations	5-24
5-12   Program vs. Site-Specific Average Benzo(a)pyrene Concentration	5-24
5-13   Program vs. Site-Specific Average 1,3-Butadiene Concentrations	5-25
5-14   Program vs. Site-Specific Average Hexavalent Chromium Concentration	5-25
5-15   Program vs. Site-Specific Average Lead (PMio) Concentration	5-25
5-16   Program vs. Site-Specific Average Manganese (PMio) Concentration	5-26
5-17   Program vs. Site-Specific Average Naphthalene Concentration	5-26
5-18   Annual Statistical Metrics for Arsenic (PMio) Concentrations Measured at PXSS	5-29
5-19   Annual Statistical Metrics for Hexavalent Chromium Concentrations Measured
       at PXSS	5-29
5-20   Annual Statistical Metrics for Lead (PMio) Concentrations Measured at PXSS	5-30
5-21   Annual Statistical Metrics for Manganese (PMio) Concentrations Measured
       at PXSS	5-30

6-1    Los Angeles, California (CELA) Monitoring Site	6-2
6-2    NEI Point Sources Located Within 10 Miles of CELA	6-3
6-3    Rubidoux, California (RUCA) Monitoring Site	6-4
6-4    NEI Point Sources Located Within 10 Miles of RUCA	6-5
6-5    San Jose, California (SJJCA) Monitoring Site	6-6
6-6    NEI Point Sources Located Within 10 Miles of SJJCA	6-7
6-7    2011 Composite Back Trajectory Map for CELA	6-15
6-8    Back Trajectory Cluster Map for CELA	6-15
6-9    2011 Composite Back Trajectory Map for RUCA	6-16
6-10   Back Trajectory Cluster Map for RUCA	6-16
6-11   2011 Composite Back Trajectory Map for SJJCA	6-17
6-12   Back Trajectory Cluster Map for SJJCA	6-17
6-13   Wind Roses for the Downtown Los Angeles/USC Campus Weather Station
       near CELA	6-20
6-14   Wind Roses for the Riverside Municipal Airport Weather Station near RUCA	6-21
6-15   Wind Roses for the San Jose International Airport Weather Station near SJJCA	6-22
6-16   Program vs. Site-Specific Average Arsenic (PMio) Concentration	6-29
6-17   Program vs. Site-Specific Average Benzo(a)pyrene Concentrations	6-29
6-18   Program vs. Site-Specific Average Lead (PMio) Concentration	6-29
6-19   Program vs. Site-Specific Average Manganese (PMio) Concentration	6-30
6-20   Program vs. Site-Specific Average Naphthalene Concentrations	6-30

7-1    Grand Junction, Colorado (GPCO) Monitoring Site	7-2
7-2    NEI Point Sources Located Within 10 Miles of GPCO	7-3

                                         xxii

-------
                         LIST OF FIGURES (Continued)

                                                                                 Page
7-3   Battlement Mesa, Colorado (BMCO) Monitoring Site	7-4
7-4   Silt, Colorado (BRCO) Monitoring Site	7-5
7-5   Parachute, Colorado (PACO) Monitoring Site	7-6
7-6   Rifle, Colorado (RICO) Monitoring Site	7-7
7-7   NEI Point Sources Located Within 10 Miles of BMCO, BRCO, PACO, and RICO	7-8
7-8   2011 Composite Back Trajectory Map for GPCO	7-17
7-9   Back Trajectory Cluster Map for GPCO	7-17
7-10  2011 Composite Back Trajectory Map for BMCO	7-18
7-11  Back Trajectory Cluster Map for BMCO	7-18
7-12  2011 Composite Back Trajectory Map for BRCO	7-19
7-13  Back Trajectory Cluster Map for BRCO	7-19
7-14  2011 Composite Back Trajectory Map for PACO	7-20
7-15  Back Trajectory Cluster Map for PACO	7-20
7-16  2011 Composite Back Trajectory Map for RICO	7-21
7-17  Back Trajectory Cluster Map for RICO	7-21
7-18  Wind Roses for the Walker Field Airport Weather Station near GPCO	7-23
7-19  Wind Roses for the Garfield County Regional Airport near BMCO	7-24
7-20  Wind Roses for the Garfield County Regional Airport near BRCO	7-25
7-21  Wind Roses for the Garfield County Regional Airport near PACO	7-26
7-22  Wind Roses for the Garfield County Regional Airport near RICO	7-27
7-23  Program vs. Site-Specific Average Acetaldehyde Concentration	7-37
7-24a Program vs. Site-Specific Average Benzene (Method TO-15) Concentration	7-37
7-24b Program vs. Site-Specific Average Benzene (SNMOC) Concentrations	7-38
7-25  Program vs. Site-Specific Average Benzo(a)pyrene Concentration	7-38
7-26a Program vs. Site-Specific Average 1,3-Butadiene (Method TO-15) Concentration	7-38
7-26b Program vs. Site-Specific Average 1,3-Butadiene (SNMOC) Concentrations	7-39
7-27  Program vs. Site-Specific Average Formaldehyde Concentration	7-39
7-28  Program vs. Site-Specific Average Hexavalent Chromium Concentration	7-39
7-29  Program vs. Site-Specific Average Naphthalene Concentration	7-40
7-30  Annual Statistical Metrics for Acetaldehyde Concentrations Measured at GPCO	7-43
7-31  Annual Statistical Metrics for Benzene Concentrations Measured at GPCO	7-43
7-32  Annual Statistical Metrics for 1,3-Butadiene Concentrations Measured at GPCO	7-44
7-33  Annual Statistical Metrics for Formaldehyde Concentrations Measured at GPCO	7-44
7-34  Annual Statistical Metrics for Hexavalent Chromium Concentrations Measured
      at GPCO	7-45

8-1   Washington, D.C. (WADC) Monitoring Site	8-2
8-2   NEI Point Sources Located Within 10 Miles of WADC	8-3
8-3   2011 Composite Back Trajectory Map for WADC	8-9
8-4   Back Trajectory Cluster Map for WADC	8-9
8-5   Wind Roses for the Ronald Reagan Washington National Airport Weather Station
      near WADC	8-11
8-6   Program vs. Site-Specific Average Benzo(a)pyrene Concentration	8-15
8-7   Program vs. Site-Specific Average Hexavalent Chromium Concentration	8-15
8-8   Program vs. Site-Specific Average Naphthalene Concentration	8-16

                                         xxiii

-------
                         LIST OF FIGURES (Continued)

                                                                                 Page
8-9    Annual Statistical Metrics for Hexavalent Chromium Concentrations Measured at
       WADC	8-17

9-1    St. Petersburg, Florida (AZFL) Monitoring Site	9-2
9-2    Pinellas Park, Florida (SKFL) Monitoring Site	9-3
9-3    Plant City, Florida (SYFL) Monitoring Site	9-4
9-4    NEI Point Sources Located Within 10 Miles of the Tampa/St. Petersburg, Florida
       Monitoring Sites	9-5
9-5    Winter Park, Florida (ORFL) Monitoring Site	9-6
9-6    Orlando, Florida (PAFL) Monitoring Site	9-7
9-7    NEI Point Sources Located Within 10 Miles of ORFL and PAFL	9-8
9-8    2011 Composite Back Trajectory Map for AZFL	9-17
9-9    Back Trajectory Cluster Map for AZFL	9-17
9-10   2011 Composite Back Trajectory Map for SKFL	9-18
9-11   Back Trajectory Cluster Map for SKFL	9-18
9-12   2011 Composite Back Trajectory Map for SYFL	9-19
9-13   Back Trajectory Cluster Map for SYFL	9-19
9-14   2011 Composite Back Trajectory Map for ORFL	9-20
9-15   Back Trajectory Cluster Map for ORFL	9-20
9-16   2011 Composite Back Trajectory Map for PAFL	9-21
9-17   Back Trajectory Cluster Map for PAFL	9-21
9-18   Wind Roses for the St. Petersburg/Whitted Airport Weather Station near AZFL	9-24
9-19   Wind Roses for the St. Petersburg/Clearwater International Airport Weather
       Station near SKFL	9-25
9-20   Wind Roses for the Plant City Municipal Airport Weather Station near SYFL	9-26
9-21   Wind Roses for the Orlando Executive Airport Weather Station near ORFL	9-27
9-22   Wind Roses for the Orlando Executive Airport Weather Station near PAFL	9-28
9-23   Program vs. Site-Specific Average Acetaldehyde Concentrations	9-37
9-24   Program vs. Site-Specific Average Arsenic (PMio) Concentration	9-37
9-25   Program vs. Site-Specific Average Benzo(a)Pyrene Concentrations	9-38
9-26   Program vs. Site-Specific Average Formaldehyde Concentrations	9-38
9-27   Program vs. Site-Specific Average Hexavalent Chromium Concentrations	9-39
9-28   Program vs. Site-Specific Average Lead  (PMio) Concentration	9-39
9-29   Program vs. Site-Specific Average Manganese (PMio) Concentration	9-39
9-30   Program vs. Site-Specific Average Naphthalene Concentrations	9-40
9-31   Annual Statistical Metrics for Acetaldehyde Concentrations Measured at AZFL	9-42
9-32   Annual Statistical Metrics for Formaldehyde Concentrations Measured at AZFL	9-43
9-33   Annual Statistical Metrics for Acetaldehyde Concentrations Measured at ORFL	9-43
9-34   Annual Statistical Metrics for Formaldehyde Concentrations Measured at ORFL	9-44
9-35   Annual Statistical Metrics for Acetaldehyde Concentrations Measured at SKFL	9-44
9-36   Annual Statistical Metrics for Formaldehyde Concentrations Measured at SKFL	9-45
9-37   Annual Statistical Metrics for Acetaldehyde Concentrations Measured at SYFL	9-45
9-38   Annual Statistical Metrics for Formaldehyde Concentrations Measured at SYFL	9-46
9-39   Annual Statistical Metrics for Hexavalent Chromium Concentrations Measured
       at SYFL	9-46

                                         xxiv

-------
                         LIST OF FIGURES (Continued)

                                                                                 Page
10-1  Decatur, Georgia (SDGA) Monitoring Site	10-2
10-2  NEI Point Sources Located Within 10 Miles of SDGA	10-3
10-3  2011 Composite Back Trajectory Map for SDGA	10-9
10-4  Back Trajectory Cluster Map for SDGA	10-9
10-5  Wind Roses for the Hartsfield International Airport Weather Station near SDGA	10-11
10-6  Program vs. Site-Specific Average Benzo(a)pyrene Concentration	10-15
10-7  Program vs. Site-Specific Average Hexavalent Chromium Concentration	10-15
10-8  Program vs. Site-Specific Average Naphthalene Concentration	10-16
10-9  Annual Statistical Metrics for Hexavalent Chromium Concentrations Measured
      at SDGA	10-17

11-1  Northbrook, Illinois (NBIL) Monitoring Site	11-2
11-2  Schiller Park, Illinois (SPIL) Monitoring Site	11-3
11-3  NEI Point Sources Located Within 10 Miles of NBIL and SPIL	11-4
11-4  2011 Composite Back Trajectory Map for NBIL	11-10
11-5  Back Trajectory Cluster Map for NBIL	11-11
11-6  2011 Composite Back Trajectory Map for SPIL	11-11
11-7  Back Trajectory Cluster Map for SPIL	11-12
11-8  Wind Roses for the Palwaukee Municipal Airport Weather Station near NBIL	11-14
11-9  Wind Roses for the O'Hare International Airport Weather Station near SPIL	11-15
11-10 Program vs. Site-Specific Average Acetaldehyde Concentrations	11-25
11-11 Program vs. Site-Specific Average Arsenic (PMio) Concentration	11-26
11-12 Program vs. Site-Specific Average Benzene Concentrations	11-26
11-13 Program vs. Site-Specific Average Benzo(a)pyrene Concentration	11-26
11-14 Program vs. Site-Specific Average 1,3-Butadiene Concentrations	11-27
11-15 Program vs. Site-Specific Average Formaldehyde Concentrations	11-27
11-16 Program vs. Site-Specific Average Hexavalent Chromium Concentration	11-27
11-17 Program vs. Site-Specific Average Lead (PMio) Concentration	11-28
11-18 Program vs. Site-Specific Average Manganese (PMio) Concentration	11-28
11-19 Program vs. Site-Specific Average Naphthalene Concentration	11 -28
11-20 Annual Statistical Metrics for Acetaldehyde Concentrations Measured at NBIL	11-31
11-21 Annual Statistical Metrics for Arsenic (PMio) Concentrations Measured at NBIL.... 11-32
11 -22 Annual Statistical Metrics for Benzene Concentrations Measured at NBIL	11-32
11-23 Annual Statistical Metrics for 1,3-Butadiene Concentrations Measured at NBIL	11-33
11-24 Annual Statistical Metrics for Formaldehyde Concentrations Measured at NBIL	11-33
11-25 Annual Statistical Metrics for Hexavalent Chromium Concentrations Measured
      at NBIL	11-34
11 -26 Annual Statistical Metrics for Lead (PMio) Concentrations Measured at NBIL	11-34
11-27 Annual Statistical Metrics for Manganese (PMio) Concentrations Measured
      at NBIL	11-35
11 -28 Annual Statistical Metrics for Acetaldehyde Concentrations Measured at SPIL	11-35
11-29 Annual Statistical Metrics for Benzene Concentrations Measured at SPIL	11-36
11-30 Annual Statistical Metrics for 1,3-Butadiene Concentrations Measured at SPIL	11-36
11-31 Annual Statistical Metrics for Formaldehyde Concentrations Measured at SPIL	11-37
                                         xxv

-------
                         LIST OF FIGURES (Continued)

                                                                                 Page
12-1  Gary, Indiana (INDEM) Monitoring Site	12-2
12-2  NEI Point Sources Located Within 10 Miles of INDEM	12-3
12-3  Indianapolis, Indiana (WPIN) Monitoring Site	12-4
12-4  NEI Point Sources Located Within 10 Miles of WPIN	12-5
12-5  2011 Composite Back Trajectory Map for INDEM	12-12
12-6  Back Trajectory Cluster Map for INDEM	12-12
12-7  2011 Composite Back Trajectory Map for WPIN	12-13
12-8  Back Trajectory Cluster Map for WPIN	12-13
12-9  Wind Roses for the Lansing Municipal Airport Weather Station near INDEM	12-15
12-10 Wind Roses for the Indianapolis International Airport Weather Station near WPIN . 12-16
12-11 Program vs. Site-Specific Average Acetaldehyde Concentration	12-21
12-12 Program vs. Site-Specific Average Formaldehyde Concentration	12-21
12-13 Annual Statistical Metrics for Acetaldehyde Concentrations Measured at INDEM... 12-23
12-14 Annual Statistical Metrics for Formaldehyde Concentrations Measured at
      INDEM	12-23

13-1  Grayson, Kentucky (GLKY) Monitoring Site	13-2
13-2  NEI Point Sources Located Within 10 Miles of GLKY	13-3
13-3  2011 Composite Back Trajectory Map for GLKY	13-9
13-4  Back Trajectory Cluster Map for GLKY	13-9
13-5  Wind Roses for the Tri-State/MJ. Ferguson Field Airport Weather Station near
      GLKY	13-11
13-6  Program vs. Site-Specific Average Benzene Concentration	13-17
13-7  Program vs. Site-Specific Average Benzo(a)pyrene Concentration	13-17
13-8  Program vs. Site-Specific Average 1,3-Butadiene Concentration	13-17
13-9  Program vs. Site-Specific Average Hexavalent Chromium Concentration	13-18
13-10 Program vs. Site-Specific Average Naphthalene Concentration	13-18

14-1  Boston, Massachusetts (BOMA) Monitoring Site	14-2
14-2  NEI Point Sources Located Within 10 Miles of BOMA	14-3
14-3  2011 Composite Back Trajectory Map for BOMA	14-9
14-4  Back Trajectory Cluster Map for BOMA	14-9
14-5  Wind Roses for the Logan International Airport Weather Station near BOMA	14-12
14-6  Program vs. Site-Specific Average Arsenic (PMio) Concentration	14-17
14-7  Program vs. Site-Specific Average Benzo(a)pyrene Concentration	14-17
14-8  Program vs. Site-Specific Average Hexavalent Chromium Concentration	14-17
14-9  Program vs. Site-Specific Average Lead (PMio) Concentration	14-18
14-10 Program vs. Site-Specific Average Manganese (PMio) Concentration	14-18
14-11 Program vs. Site-Specific Average Naphthalene Concentration	14-18
14-12 Annual Statistical Metrics for Arsenic (PMio) Concentrations Measured at BOMA . 14-20
14-13 Annual Statistical Metrics for Hexavalent Chromium Concentrations Measured
      at BOMA	14-21
14-14 Annual Statistical Metrics for Lead (PMio) Concentrations Measured at BOMA	14-21
14-15 Annual Statistical Metrics for Manganese (PMio) Concentrations Measured
      at BOMA	14-22

                                         xxvi

-------
                         LIST OF FIGURES (Continued)

                                                                                 Page
15-1  Dearborn, Michigan (DEMI) Monitoring Site	15-2
15-2  River Rouge, Michigan (RRMI) Monitoring Site	15-3
15-3  Detroit, Michigan (SWMI) Monitoring Site	15-4
15-4  NEI Point Sources Located Within 10 Miles of DEMI, RRMI, and SWMI	15-5
15-5  2011 Composite Back Trajectory Map for DEMI	15-12
15-6  Back Trajectory Cluster Map for DEMI	15-12
15-7  2011 Composite Back Trajectory Map for RRMI	15-13
15-8  Back Trajectory Cluster Map for RRMI	15-13
15-9  2011 Composite Back Trajectory Map for SWMI	15-14
15-10 Wind Roses for the Detroit City Airport Weather Station near DEMI	15-16
15-11 Wind Roses for the Detroit City Airport Weather Station near RRMI	15-17
15-12 Wind Roses for the Detroit City Airport Weather Station near SWMI	15-18
15-13 Program vs. Site-Specific Average Acetaldehyde Concentrations	15-26
15-14 Program vs. Site-Specific Average Benzene Concentration	15-27
15-15 Program vs. Site-Specific Average Benzo(a)pyrene Concentration	15-27
15-16 Program vs. Site-Specific Average 1,3-Butadiene Concentration	15-27
15-17 Program vs. Site-Specific Average Formaldehyde Concentrations	15-28
15-18 Program vs. Site-Specific Average Hexavalent Chromium Concentration	15-28
15-19 Program vs. Site-Specific Average Naphthalene Concentration	15-28
15-20 Annual Statistical Metrics for Acetaldehyde Concentrations Measured at DEMI	15-31
15-21 Annual Statistical Metrics for Benzene Concentrations Measured at DEMI	15-31
15-22 Annual Statistical Metrics for 1,3-Butadiene Concentrations Measured at DEMI	15-32
15-23 Annual Statistical Metrics for Formaldehyde Concentrations Measured at DEMI	15-32
15-24 Annual Statistical Metrics for Hexavalent Chromium Concentrations Measured
      at DEMI	15-33

16-1  St. Louis, Missouri (S4MO) Monitoring Site	16-2
16-2  NEI Point Sources Located Within 10 Miles of S4MO	16-3
16-3  2011 Composite Back Trajectory Map for S4MO	16-9
16-4  Back Trajectory Cluster Map for S4MO	16-9
16-5  Wind Roses for the St. Louis Downtown Airport Weather Station near S4MO	16-12
16-6  Program vs. Site-Specific Average Acetaldehyde Concentration	16-19
16-7  Program vs. Site-Specific Average Arsenic (PMio) Concentration	16-19
16-8  Program vs. Site-Specific Average Benzene Concentration	16-20
16-9  Program vs. Site-Specific Average Benzo(a)pyrene Concentration	16-20
16-10 Program vs. Site-Specific Average 1,3-Butadiene Concentration	16-20
16-11 Program vs. Site-Specific Average Formaldehyde Concentration	16-21
16-12 Program vs. Site-Specific Average Hexavalent Chromium Concentration	16-21
16-13 Program vs. Site-Specific Average Lead (PMio) Concentration	16-21
16-14 Program vs. Site-Specific Average Manganese (PMio) Concentration	16-22
16-15 Program vs. Site-Specific Average Naphthalene Concentration	16-22
16-16 Annual Statistical Metrics for Acetaldehyde Concentrations Measured at S4MO	16-25
16-17 Annual Statistical Metrics for Arsenic (PMio) Concentrations Measured at S4MO... 16-25
16-18 Annual Statistical Metrics for Benzene Concentrations Measured at S4MO	16-26
16-19 Annual Statistical Metrics for 1,3-Butadiene Concentrations Measured at S4MO	16-26

                                        xxvii

-------
                         LIST OF FIGURES (Continued)

                                                                                 Page
16-20 Annual Statistical Metrics for Formaldehyde Concentrations Measured at S4MO .... 16-27
16-21 Annual Statistical Metrics for Hexavalent Chromium Concentrations Measured
      atS4MO	16-27
16-22 Annual Statistical Metrics for Lead (PMi0) Concentrations Measured at S4MO	16-28
16-23 Annual Statistical Metrics for Manganese (PMio) Concentrations Measured
      atS4MO	16-28

17-1  Chester, New Jersey (CHNJ) Monitoring Site	17-2
17-2  NEI Point Sources Located Within 10 Miles of CHNJ	17-3
17-3  Elizabeth, New Jersey (ELNJ) Monitoring Site	17-4
17-4  New Brunswick, New Jersey (NBNJ) Monitoring Site	17-5
17-5  NEI Point Sources Located Within 10 Miles of ELNJ and NBNJ	17-6
17-6  Paterson, New Jersey (PANJ) Monitoring Site	17-7
17-7  NEI Point Sources Located Within 10 Miles of PANJ	17-8
17-8  2011 Composite Back Trajectory Map for CHNJ	17-16
17-9  Back Trajectory Cluster Map for CHNJ	17-16
17-10 2011 Composite Back Trajectory Map for ELNJ	17-17
17-11 Back Trajectory Cluster Map for ELNJ	17-17
17-12 2011 Composite Back Trajectory Map for NBNJ	17-18
17-13 Back Trajectory Cluster Map for NBNJ	17-18
17-14 2011 Composite Back Trajectory Map for PANJ	17-19
17-15 Wind Roses for the Summerville-Somerset Airport Weather Station near CHNJ	17-21
17-16 Wind Roses for the Newark International Airport Weather Station near ELNJ	17-22
17-17 Wind Roses for the Summerville-Somerset Airport Weather Station near NBNJ	17-23
17-18 Wind Roses for the Essex County Airport Weather Station near PANJ	17-24
17-19 Program vs. Site-Specific Average Acetaldehyde Concentrations	17-36
17-20 Program vs. Site-Specific Average Benzene Concentrations	17-37
17-21 Program vs. Site-Specific Average 1,3-Butadiene Concentrations	17-37
17-22 Program vs. Site-Specific Average Formaldehyde Concentrations	17-38
17-23 Annual Statistical Metrics for Acetaldehyde Concentrations Measured at CHNJ	17-40
17-24 Annual Statistical Metrics for Benzene Concentrations Measured at CHNJ	17-40
17-25 Annual Statistical Metrics for 1,3-Butadiene Concentrations Measured at CHNJ	17-41
17-26 Annual Statistical Metrics for Formaldehyde Concentrations Measured at CHNJ	17-41
17-27 Annual Statistical Metrics for Acetaldehyde Concentrations Measured at ELNJ	17-42
17-28 Annual Statistical Metrics for Benzene Concentrations Measured at ELNJ	17-42
17-29 Annual Statistical Metrics for 1,3-Butadiene Concentrations Measured at ELNJ	17-43
17-30 Annual Statistical Metrics for Formaldehyde Concentrations Measured at ELNJ	17-43
17-31 Annual Statistical Metrics for Acetaldehyde Concentrations Measured at NBNJ	17-44
17-32 Annual Statistical Metrics for Benzene Concentrations Measured at NBNJ	17-44
17-33 Annual Statistical Metrics for 1,3-Butadiene Concentrations Measured at NBNJ	17-45
17-34 Annual Statistical Metrics for Formaldehyde Concentrations Measured at NBNJ	17-45

18-1  New York City, New York (MONY) Monitoring Site	18-2
18-2  NEI Point Sources Located Within 10 Miles of MONY	18-3
18-3  Rochester, New York (ROCH) Monitoring Site	18-4

                                        xxviii

-------
                         LIST OF FIGURES (Continued)

                                                                                 Page
18-4  NEI Point Sources Located Within 10 Miles of ROCH	18-5
18-5  2011 Composite Back Trajectory Map for MONY	18-12
18-6  Back Trajectory Cluster Map for MONY	18-12
18-7  2011 Composite Back Trajectory Map for ROCH	18-13
18-8  Back Trajectory Cluster Map for ROCH	18-13
18-9  Wind Roses for the Central Park Weather Station near MONY	18-16
18-10 Wind Roses for the Greater Rochester International Airport Weather Station
      near ROCH	18-17
18-11 Program vs. Site-Specific Average Benzo(a)pyrene Concentrations	18-23
18-12 Program vs. Site-Specific Average Hexavalent Chromium Concentrations	18-23
18-13 Program vs. Site-Specific Average Naphthalene Concentrations	18-24

19-1  Tulsa, Oklahoma (TOOK) Monitoring Site	19-2
19-2  Tulsa, Oklahoma (TMOK) Monitoring Site	19-3
19-3  NEI Point Sources Located Within 10 Miles of TMOK and TOOK	19-4
19-4  Pryor Creek, Oklahoma (PROK) Monitoring Site	19-5
19-5  NEI Point Sources Located Within 10 Miles of PROK	19-6
19-6  Midwest City, Oklahoma (MWOK) Monitoring Site	19-7
19-7  Oklahoma City, Oklahoma (OCOK) Monitoring Site	19-8
19-8  NEI Point Sources Located Within 10 Miles of MWOK and OCOK	19-9
19-9  2011 Composite Back Trajectory Map for TOOK	19-17
19-10 Back Trajectory Cluster Map for TOOK	19-17
19-11 2011 Composite Back Trajectory Map for TMOK	19-18
19-12 Back Trajectory Cluster Map for TMOK	19-18
19-13 2011 Composite Back Trajectory Map for PROK	19-19
19-14 Back Trajectory Cluster Map for PROK	19-19
19-15 2011 Composite Back Trajectory Map for MWOK	19-20
19-16 Back Trajectory Cluster Map for MWOK	19-20
19-17 2011 Composite Back Trajectory Map for OCOK	19-21
19-18 Back Trajectory Cluster Map for OCOK	19-21
19-19 Wind Roses for the Richard Lloyd Jones Jr. Airport Weather Station near TOOK.... 19-24
19-20 Wind Roses for the Tulsa International Airport Weather Station near TMOK	19-25
19-21 Wind Roses for the Claremore Regional Airport Weather Station near PROK	19-26
19-22 Wind Roses for the Tinker Air Force Base Airport Weather Station near MWOK.... 19-27
19-23 Wind Roses for the Wiley Post Airport Weather Station near OCOK	19-28
19-24 Program vs. Site-Specific Average Acetaldehyde Concentrations	19-46
19-25 Program vs. Site-Specific Average Arsenic (TSP) Concentrations	19-47
19-26 Program vs. Site-Specific Average Benzene Concentrations	19-48
19-27 Program vs. Site-Specific Average 1,3-Butadiene Concentrations	19-49
19-28 Program vs. Site-Specific Average Formaldehyde Concentrations	19-50
19-29 Program vs. Site-Specific Average Lead (TSP) Concentrations	19-51
19-30 Program vs. Site-Specific Average Manganese (TSP) Concentrations	19-52
19-31 Annual  Statistical Metrics for Acetaldehyde Concentrations Measured at TOOK	19-54
19-32 Annual  Statistical Metrics for Arsenic (TSP) Concentrations Measured at TOOK.... 19-55
19-33 Annual  Statistical Metrics for Benzene Concentrations Measured at TOOK	19-55

                                        xxix

-------
                         LIST OF FIGURES (Continued)

                                                                                Page
19-34  Annual Statistical Metrics for 1,3-Butadiene Concentrations Measured at TOOK .... 19-56
19-35  Annual Statistical Metrics for Formaldehyde Concentrations Measured at TOOK.... 19-56
19-36  Annual Statistical Metrics for Lead (TSP) Concentrations Measured at TOOK	19-57
19-37  Annual Statistical Metrics for Manganese (TSP) Concentrations Measured
       at TOOK	19-57

20-1   Providence, Rhode Island (PRRI) Monitoring Site	20-2
20-2   NEI Point Sources Located Within 10 Miles of PRRI	20-3
20-3   2011 Composite Back Trajectory Map for PRRI	20-9
20-4   Back Trajectory Cluster Map for PRRI	20-9
20-5   Wind Roses for the T.F. Green State Airport Weather Station near PRRI	20-12
20-6   Program vs. Site-Specific Average Benzo(a)pyrene Concentration	20-16
20-7   Program vs. Site-Specific Average Hexavalent Chromium Concentration	20-16
20-8   Program vs. Site-Specific Average Naphthalene Concentration	20-17
20-9   Annual Statistical Metrics for Hexavalent Chromium Concentrations Measured
       at PRRI	20-18

21-1   Chesterfield, South Carolina (CHSC) Monitoring Site	21-2
21-2   NEI Point Sources Located Within 10 Miles of CHSC	21-3
21-3   2011 Composite Back Trajectory Map for CHSC	21-9
21-4   Back Trajectory Cluster Map for CHSC	21-9
21-5   Wind Roses for the Monroe Airport Weather Station near CHSC	21-11
21-6   Program vs. Site-Specific Average Benzo(a)pyrene Concentration	21-15
21-7   Program vs. Site-Specific Average Hexavalent Chromium Concentration	21-15
21-8   Program vs. Site-Specific Average Naphthalene Concentration	21-16
21-9   Annual Statistical Metrics for Hexavalent Chromium Concentrations Measured
       at CHSC	21-17

22-1   Sioux Falls, South Dakota (SSSD) Monitoring Site	22-2
22-2   NEI Point Sources Located Within 10 Miles of SSSD	22-3
22-3   Union County, South Dakota (UCSD) Monitoring Site	22-4
22-4   NEI Point Sources Located Within 10 Miles of UCSD	22-5
22-5   2011 Composite Back Trajectory Map for SSSD	22-12
22-6   Back Trajectory Cluster Map for SSSD	22-12
22-7   2011 Composite Back Trajectory Map for UCSD	22-13
22-8   Back Trajectory Cluster Map for UCSD	22-13
22-9   Wind Roses for the Joe Foss Field Airport Weather Station near SSSD	22-16
22-10  Wind Roses for the Sioux Gateway Airport Weather Station near UCSD	22-17
22-11  Program vs. Site-Specific Average Acetaldehyde Concentrations	22-24
22-12  Program vs. Site-Specific Average Benzene Concentrations	22-24
22-13  Program vs. Site-Specific Average 1,3-Butadiene Concentrations	22-25
22-14  Program vs. Site-Specific Average Formaldehyde Concentrations	22-25

23-1   Deer Park,  Texas (CAMS 35) Monitoring Site	23-2
23-2   NEI Point Sources Located Within 10 Miles of CAMS 35	23-3

                                        XXX

-------
                         LIST OF FIGURES (Continued)

                                                                                Page
23-3  Karnack, Texas (CAMS 85) Monitoring Site	23-4
23-4  NEI Point Sources Located Within 10 Miles of CAMS 85	23-5
23-5  2011 Composite Back Trajectory Map for CAMS 35	23-12
23-6  Back Trajectory Cluster Map for CAMS 35	23-12
23-7  2011 Composite Back Trajectory Map for CAMS 85	23-13
23-8  Back Trajectory Cluster Map for CAMS 85	23-13
23-9  Wind Roses for the William P. Hobby Airport Weather Station near CAMS 35	23-15
23-10 Wind Roses for the Shreveport Regional Airport Weather Station near CAMS 85.... 23-16
23-11 Program vs. Site-Specific Average Benzo(a)pyrene Concentration	23-22
23-12 Program vs. Site-Specific Average Hexavalent Chromium Concentrations	23-22
23-13 Program vs. Site-Specific Average Naphthalene Concentration	23-22

24-1  Bountiful, Utah (BTUT) Monitoring Site	24-2
24-2  NEI Point Sources Located Within 10 Miles of BTUT	24-3
24-3  2011 Composite Back Trajectory Map for BTUT	24-9
24-4  Back Trajectory Cluster Map for BTUT	24-9
24-5  Wind Roses for the Salt Lake City International Airport Weather Station near
      BTUT	24-12
24-6  Program vs. Site-Specific Average Acetaldehyde Concentration	24-18
24-7  Program vs. Site-Specific Average Arsenic (PMio) Concentration	24-19
24-8  Program vs. Site-Specific Average Benzene Concentration	24-19
24-9  Program vs. Site-Specific Average Benzo(a)pyrene Concentration	24-19
24-10 Program vs. Site-Specific Average 1,3-Butadiene Concentration	24-20
24-11 Program vs. Site-Specific Average Formaldehyde Concentration	24-20
24-12 Program vs. Site-Specific Average Hexavalent Chromium Concentration	24-20
24-13 Program vs. Site-Specific Average Lead (PMio) Concentration	24-21
24-14 Program vs. Site-Specific Average Manganese (PMio) Concentration	24-21
24-15 Program vs. Site-Specific Average Naphthalene Concentration	24-21
24-16 Annual Statistical Metrics for Acetaldehyde Concentrations Measured at BTUT	24-24
24-17 Annual Statistical Metrics for Arsenic (PMio) Concentrations Measured at BTUT... 24-24
24-18 Annual Statistical Metrics for Benzene Concentrations Measured at BTUT	24-25
24-19 Annual Statistical Metrics for 1,3-Butadiene Concentrations Measured at BTUT	24-25
24-20 Annual Statistical Metrics for Formaldehyde Concentrations Measured at BTUT .... 24-26
24-21 Annual Statistical Metrics for Hexavalent Chromium Concentrations Measured
      at BTUT	24-26
24-22 Annual Statistical Metrics for Lead (PMio) Concentrations Measured at BTUT	24-27
24-23 Annual Statistical Metrics for Manganese (PMio) Concentrations Measured
      at BTUT	24-27

25-1  Burlington, Vermont (BURVT) Monitoring Site	25-2
25-2  Underbill, Vermont (UNVT) Monitoring Site	25-3
25-3  NEI Point Sources Located Within 10 Miles of BURVT and UNVT	25-4
25-4  Rutland, Vermont (RUVT) Monitoring Site	25-5
25-5  NEI Point Sources Located Within 10 Miles of RUVT	25-6
25-6  2011 Composite Back Trajectory Map for BURVT	25-13

                                        xxxi

-------
                         LIST OF FIGURES (Continued)

                                                                                 Page
25-7   Back Trajectory Cluster Map forBURVT	25-13
25-8   2011 Composite Back Trajectory Map for RUVT	25-14
25-9   Back Trajectory Cluster Map for RUVT	25-14
25-10  2011 Composite Back Trajectory Map for UNVT	25-15
25-11  Back Trajectory Cluster Map for UNVT	25-15
25-12  Wind Roses for the Burlington International Airport Weather Station near
       BURVT	25-18
25-13  Wind Roses for the Rutland State Airport Weather Station near RUVT	25-19
25-14  Wind Roses for the Morrisville-Stowe State Airport Weather Station near UNVT ... 25-20
25-15  Program vs. Site-Specific Average Arsenic (PMio) Concentration	25-30
25-16  Program vs. Site-Specific Average Benzene Concentrations	25-30
25-17  Program vs. Site-Specific Average Benzo(a)pyrene Concentration	25-30
25-18  Program vs. Site-Specific Average 1,3-Butadiene Concentrations	25-31
25-19  Program vs. Site-Specific Average Hexavalent Chromium Concentration	25-31
25-20  Program vs. Site-Specific Average Lead (PMio) Concentration	25-31
25-21  Program vs. Site-Specific Average Manganese (PMio) Concentration	25-32
25-22  Program vs. Site-Specific Average Naphthalene Concentration	25-32
25-23  Annual Statistical Metrics for Hexavalent Chromium Concentrations Measured
       at UNVT	25-34

26-1   Richmond, Virginia (RIVA) Monitoring Site	26-2
26-2   NEI Point Sources Located Within 10 Miles of RIVA	26-3
26-3   2011 Composite Back Trajectory Map for RIVA	26-9
26-4   Back Trajectory Cluster Map for RIVA	26-9
26-5   Wind Roses for the Richmond International Airport Weather Station near RIVA	26-11
26-6   Program vs. Site-Specific Average Benzo(a)pyrene Concentration	26-15
26-7   Program vs. Site-Specific Average Hexavalent Chromium Concentration	26-15
26-8   Program vs. Site-Specific Average Naphthalene Concentration	26-16

27-1   Seattle, Washington (SEWA) Monitoring Site	27-2
27-2   NEI Point Sources Located Within 10 Miles of SEWA	27-3
27-3   2011 Composite Back Trajectory Map for SEWA	27-9
27-4   Back Trajectory Cluster Map for SEWA	27-9
27-5   Wind Roses for the Boeing Field/King County International Airport Weather
       Station near SEWA	27-11
27-6   Program vs. Site-Specific Average Acetaldehyde Concentration	27-18
27-7   Program vs. Site-Specific Average Arsenic (PMio) Concentration	27-18
27-8   Program vs. Site-Specific Average Benzene Concentration	27-19
27-9   Program vs. Site-Specific Average Benzo(a)pyrene Concentration	27-19
27-10  Program vs. Site-Specific Average 1,3-Butadiene Concentration	27-19
27-11  Program vs. Site-Specific Average Formaldehyde Concentration	27-20
27-12  Program vs. Site-Specific Average Hexavalent Chromium Concentration	27-20
27-13  Program vs. Site-Specific Average Lead (PMio) Concentration	27-20
27-14  Program vs. Site-Specific Average Manganese (PMio) Concentration	27-21
27-15  Program vs. Site-Specific Average Naphthalene Concentration	27-21

                                        xxxii

-------
                         LIST OF FIGURES (Continued)

                                                                                 Page
27-16  Annual Statistical Metrics for Acetaldehyde Concentrations Measured at SEWA	27-24
27-17  Annual Statistical Metrics for Arsenic (PMio) Concentrations Measured at
       SEWA	27-24
27-18  Annual Statistical Metrics for Benzene Concentrations Measured at SEWA	27-25
27-19  Annual Statistical Metrics for 1,3-Butadiene Concentrations Measured at SEWA.... 27-25
27-20  Annual Statistical Metrics for Formaldehyde Concentrations Measured at SEWA.... 27-26
27-21  Annual Statistical Metrics for Hexavalent Chromium Concentrations Measured
       at SEWA	27-26
27-22  Annual Statistical Metrics for Lead (PMio) Concentrations Measured at SEWA	27-27
27-23  Annual Statistical Metrics for Manganese (PMio) Concentrations Measured
       at SEWA	27-27

28-1   Horicon, Wisconsin (HOWI) Monitoring Site	28-2
28-2   NEI Point Sources Located Within 10 Miles of HOWI	28-3
28-3   2011 Composite Back Trajectory Map for HOWI	28-9
28-4   Back Trajectory Cluster Map for HOWI	28-9
28-5   Wind Roses for the Dodge County Airport Weather Station near HOWI	28-11
28-6   Program vs. Site-Specific Average Hexavalent Chromium Concentration	28-15
                                        xxxin

-------
                                LIST OF TABLES

                                                                                 Page
1-1    Organization of the 2011 National Monitoring Programs Report	1-4

2-1    2011 National Monitoring Programs Sites and Past Program Participation	2-4
2-2    Site Characterizing Information for the 2011 National Monitoring Programs Sites	2-7
2-3    2011 VOC Method Detection Limits	2-16
2-4    2011 SNMOC Method Detection Limits	2-17
2-5    2011 Carbonyl Compound Method Detection Limits	2-19
2-6    2011 PAH Method Detection Limits 	2-20
2-7    2011 Metals Method Detection Limits	2-21
2-8    2011 Hexavalent Chromium Method Detection Limits	2-22
2-9    2011 Sampling Schedules and Completeness Rates	2-24
2-10   Method Completeness Rates for 2011	2-30

3-1    Overview and Lay out of Data Presented	3-1
3-2    NATTS MQO Core Analytes	3-6
3-3    NATTS MQO Core Analytes Selected for Comparative Analysis	3-16
3-4    POM Groups for PAHs	3-21

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	4-9
4-4    Statistical Summaries of the PAH Concentrations	4-10
4-5    Statistical Summaries of the Metals Concentrations	4-11
4-6    Statistical Summary of the Hexavalent Chromium Concentrations	4-12
4-7    Results of the Program-Level Preliminary Risk-Based Screening	4-16
4-8    Site-Specific Risk-Based Screening Comparison	4-19
4-9    Annual Average Concentration Comparison of the VOC Pollutants of Interest	4-22
4-10   Annual Average Concentration Comparison of the Carbonyl Compound Pollutants
       of Interest	4-24
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   Comparison of Maximum Concentrations vs.  ATSDRMRLs	4-31
4-14   Summary of Mobile Source Information by Monitoring Site	4-34
4-15   Greenhouse Gases Measured by Method TO-15	4-105

5-1    Geographical Information for the Arizona Monitoring Sites	5-5
5-2    Population, Motor Vehicle, and Traffic Information for the Arizona Monitoring
       Sites	5-7
5-3    Average Meteorological Conditions near the Arizona Monitoring Sites	5-9
5-4    Risk-Based Screening Results for the Arizona Monitoring Sites	5-17
5-5    Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
       Arizona Monitoring Sites	5-20
5-6    Risk Approximations for the Arizona Monitoring Sites	5-34
                                        xxxiv

-------
                          LIST OF TABLES (Continued)

                                                                                    Page
5-7    Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
       for Pollutants with Cancer UREs for the Arizona Monitoring Sites	5-37
5-8    Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
       Approximations for Pollutants with Noncancer RfCs for the Arizona Monitoring
       Sites	5-38

6-1    Geographical Information for the California Monitoring Sites	6-8
6-2    Population, Motor Vehicle, and Traffic Information for the California Monitoring
       Sites	6-10
6-3    Average Meteorological Conditions near the California Monitoring Sites	6-13
6-4    Risk-Based Screening Results for the California Monitoring Sites	6-25
6-5    Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
       California Monitoring Sites	6-27
6-6    Risk Approximations for the California Monitoring Sites	6-33
6-7    Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
       for Pollutants with Cancer UREs for the California Monitoring Sites	6-3 5
6-8    Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
       Approximations for Pollutants with Noncancer RfCs for the California Monitoring
       Sites	6-37

7-1    Geographical Information for the Colorado Monitoring Sites	7-9
7-2    Population, Motor Vehicle, and Traffic Information for the Colorado Monitoring
       Sites	7-12
7-3    Average Meteorological Conditions near the Colorado Monitoring Sites	7-14
7-4    Risk-Based Screening Results for the Colorado Monitoring Sites	7-29
7-5    Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
       Colorado Monitoring Sites	7-33
7-6    Risk Approximations for the Colorado Monitoring Sites	7-48
7-7    Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
       for Pollutants with Cancer UREs for the Colorado Monitoring Sites	7-52
7-8    Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
       Approximations for Pollutants with Noncancer RfCs for the Colorado Monitoring
       Sites	7-55

8-1    Geographical Information for the Washington, D.C. Monitoring Site	8-4
8-2    Population, Motor Vehicle, and Traffic Information for the Washington, D.C.
       Monitoring Site	8-5
8-3    Average Meteorological Conditions near the Washington, D.C. Monitoring Site	8-7
8-4    Risk-Based Screening Results for the Washington, D.C. Monitoring Site	8-13
8-5    Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
       Washington, D.C. Monitoring Site	8-14
8-6    Risk Approximations for the Washington, D.C. Monitoring Site	8-19
8-7    Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
       for Pollutants with Cancer UREs for the Washington, D.C. Monitoring Site	8-21
                                          xxxv

-------
                          LIST OF TABLES (Continued)

                                                                                    Page
8-8    Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
       Approximations for Pollutants with Noncancer RfCs for the Washington, D.C.
       Monitoring Site	8-22

9-1    Geographical Information for the Florida Monitoring Sites	9-9
9-2    Population, Motor Vehicle, and Traffic Information for the Florida Monitoring
       Sites	9-12
9-3    Average Meteorological Conditions near the Florida Monitoring Sites	9-14
9-4    Risk-Based Screening Results for the Florida Monitoring Sites	9-31
9-5    Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
       Florida Monitoring Sites	9-34
9-6    Risk Approximations for the Florida Monitoring Sites	9-52
9-7    Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
       for Pollutants with Cancer UREs for the Florida Monitoring Sites	9-55
9-8    Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
       Approximations for Pollutants with Noncancer RfCs for the Florida Monitoring
       Sites	9-58

10-1   Geographical Information for the Georgia Monitoring Site	10-4
10-2   Population, Motor Vehicle, and Traffic Information for the Georgia Monitoring
       Site	10-5
10-3   Average Meteorological Conditions near the Georgia Monitoring Site	10-8
10-4   Risk-Based Screening Results for the Georgia Monitoring Site	10-13
10-5   Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
       Georgia Monitoring Site	10-14
10-6   Risk Approximations for the Georgia Monitoring Site	10-19
10-7   Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
       for Pollutants with Cancer UREs for the Georgia Monitoring Site	10-21
10-8   Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
       Approximations for Pollutants with Noncancer RfCs for the Georgia Monitoring
       Site	10-22

11-1   Geographical Information for the Illinois Monitoring Sites	11-5
11-2   Population, Motor Vehicle, and Traffic Information for the Illinois Monitoring
       Sites	11-7
11-3   Average Meteorological Conditions near the Illinois Monitoring Sites	11-9
11-4   Risk-Based Screening Results for the Illinois Monitoring Sites	11-17
11-5   Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
       Illinois Monitoring Sites	11-20
11-6   Risk Approximations for the Illinois Monitoring Sites	11-44
11-7   Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
       for Pollutants with Cancer UREs for the Illinois Monitoring Sites	11 -47
11-8   Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
       Approximations for Pollutants with Noncancer RfCs for the Illinois Monitoring
       Sites	11-48

                                         xxxvi

-------
                          LIST OF TABLES (Continued)

                                                                                   Page
12-1   Geographical Information for the Indiana Monitoring Sites	12-6
12-2   Population, Motor Vehicle, and Traffic Information for the Indiana Monitoring
       Sites	12-8
12-3   Average Meteorological Conditions near the Indiana Monitoring Sites	12-10
12-4   Risk-Based Screening Results for the Indiana Monitoring Sites	12-18
12-5   Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
       Indiana Monitoring Sites	12-20
12-6   Risk Approximations for the Indiana Monitoring Sites	12-26
12-7   Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
       for Pollutants with Cancer UREs for the Indiana Monitoring Sites	12-28
12-8   Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
       Approximations for Pollutants with Noncancer RfCs for the Indiana Monitoring
       Sites	12-29

13-1   Geographical Information for the Kentucky Monitoring Site	13-4
13-2   Population, Motor Vehicle, and Traffic Information for the Kentucky Monitoring
       Site	13-5
13-3   Average Meteorological Conditions near the Kentucky Monitoring Site	13-7
13-4   Risk-Based Screening Results for the Kentucky Monitoring Site	13-13
13-5   Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
       Kentucky Monitoring Site	13-14
13-6   Risk Approximations for the Kentucky Monitoring Site	13-21
13-7   Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
       for Pollutants with Cancer UREs for the Kentucky Monitoring Site	13-23
13-8   Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
       Approximations for Pollutants with Noncancer RfCs for the Kentucky Monitoring
       Site	13-24

14-1   Geographical Information for the Massachusetts Monitoring Site	14-4
14-2   Population, Motor Vehicle, and Traffic Information for the Massachusetts
       Monitoring Site	14-5
14-3   Average Meteorological Conditions near the Massachusetts Monitoring Site	14-8
14-4   Risk-Based Screening Results for the Massachusetts Monitoring Site	14-13
14-5   Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
       Massachusetts Monitoring Site	14-15
14-6   Risk Approximations for the Massachusetts Monitoring Site	14-25
14-7   Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
       for Pollutants with Cancer UREs for the Massachusetts Monitoring Site	14-27
14-8   Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
       Approximations for Pollutants with Noncancer RfCs for the Massachusetts
       Monitoring Site	14-28

15-1   Geographical Information for the Michigan Monitoring Sites	15-6
15-2   Population, Motor Vehicle, and Traffic Information for the Michigan Monitoring
       Sites	15-8

                                         xxxvii

-------
                          LIST OF TABLES (Continued)

                                                                                   Page
15-3  Average Meteorological Conditions near the Michigan Monitoring Sites	15-10
15-4  Risk-Based Screening Results for the Michigan Monitoring Sites	15-20
15-5  Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
      Michigan Monitoring Sites	15-22
15-6  Risk Approximations for the Michigan Monitoring Sites	15-37
15-7  Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
      for Pollutants with Cancer UREs for the Michigan Monitoring Sites	15-39
15-8  Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
      Approximations for Pollutants with Noncancer RfCs for the Michigan Monitoring
      Sites	15-41

16-1  Geographical Information for the Missouri Monitoring Site	16-4
16-2  Population, Motor Vehicle, and Traffic Information for the Missouri Monitoring
      Site	16-5
16-3  Average Meteorological Conditions near the Missouri Monitoring Site	16-7
16-4  Risk-Based Screening Results for the Missouri Monitoring Site	16-14
16-5  Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
      Missouri Monitoring Site	16-15
16-6  Risk Approximations for the Missouri Monitoring Site	16-33
16-7  Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
      for Pollutants with Cancer UREs for the Missouri Monitoring Site	16-36
16-8  Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
      Approximations for Pollutants with Noncancer RfCs for the Missouri Monitoring
      Site	16-37

17-1  Geographical Information for the New Jersey Monitoring Sites	17-9
17-2  Population, Motor Vehicle, and Traffic Information for the New Jersey Monitoring
      Sites	17-12
17-3  Average Meteorological Conditions near the New Jersey Monitoring Sites	17-14
17-4  Risk-Based Screening Results for the New Jersey Monitoring Sites	17-27
17-5  Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
      New Jersey Monitoring Sites	17-31
17-6  Risk Approximations for the New Jersey Monitoring Sites	17-53
17-7  Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
      for Pollutants with Cancer UREs for the New Jersey Monitoring Sites	17-57
17-8  Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
      Approximations for Pollutants with Noncancer RfCs for the New Jersey Monitoring
      Sites	17-59

18-1  Geographical Information for the New York Monitoring Sites	18-6
18-2  Population, Motor Vehicle, and Traffic Information for the New York Monitoring
      Sites	18-8
18-3  Average Meteorological Conditions near the New York Monitoring Sites	18-10
18-4  Risk-Based Screening Results for the New York Monitoring Sites	18-19
                                        xxxvin

-------
                          LIST OF TABLES (Continued)

                                                                                  Page
18-5   Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
       New York Monitoring Sites	18-21
18-6   Risk Approximations for the New York Monitoring Sites	18-26
18-7   Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
       for Pollutants with Cancer UREs for the New York Monitoring Sites	18-28
18-8   Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
       Approximations for Pollutants with Noncancer RfCs for the New York Monitoring
       Sites	18-29

19-1   Geographical Information for the Oklahoma Monitoring Sites	19-10
19-2   Population, Motor Vehicle, and Traffic Information for the Oklahoma Monitoring
       Sites	19-13
19-3   Average Meteorological Conditions near the Oklahoma Monitoring Sites	19-16
19-4   Risk-Based Screening Results for the Oklahoma Monitoring Sites	19-31
19-5   Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
       Oklahoma Monitoring Sites	19-36
19-6   Risk Approximations for the Oklahoma Monitoring Sites	19-62
19-7   Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
       for Pollutants with Cancer UREs for the Oklahoma Monitoring Sites	19-68
19-8   Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
       Approximations for Pollutants with Noncancer RfCs for the Oklahoma Monitoring
       Sites	19-71

20-1   Geographical Information for the Rhode Island Monitoring Site	20-4
20-2   Population, Motor Vehicle, and Traffic Information for the Rhode Island Monitoring
       Site	20-6
20-3   Average Meteorological Conditions near the Rhode Island Monitoring Site	20-8
20-4   Risk-Based Screening Results for the Rhode Island Monitoring Site	20-13
20-5   Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
       Rhode Island Monitoring Site	20-15
20-6   Risk Approximations for the Rhode Island Monitoring Site	20-20
20-7   Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
       for Pollutants with Cancer UREs for the Rhode Island Monitoring Site	20-22
20-8   Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
       Approximations for Pollutants with Noncancer RfCs for the Rhode Island
       Monitoring Site	20-23

21-1   Geographical Information for the South Carolina Monitoring Site	21-4
21-2   Population, Motor Vehicle, and Traffic Information for the South Carolina
       Monitoring Site	21-5
21-3   Average Meteorological Conditions near the South Carolina Monitoring Site	21-7
21-4   Risk-Based Screening Results for the South Carolina Monitoring Site	21-12
21-5   Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
       South Carolina Monitoring Site	21-14
21-6   Risk Approximations for the South Carolina Monitoring Site	21-19

                                         xxxix

-------
                          LIST OF TABLES (Continued)

                                                                                   Page
21-7   Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
       for Pollutants with Cancer UREs for the South Carolina Monitoring Site	21-20
21-8   Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
       Approximations for Pollutants with Noncancer RfCs for the South Carolina
       Monitoring Site	21-21

22-1   Geographical Information for the South Dakota Monitoring Sites	22-6
22-2   Population, Motor Vehicle, and Traffic Information for the South Dakota
       Monitoring Sites	22-8
22-3   Average Meteorological Conditions near the South Dakota Monitoring Sites	22-10
22-4   Risk-Based Screening Results for the South Dakota Monitoring Sites	22-18
22-5   Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
       South Dakota Monitoring Sites	22-21
22-6   Risk Approximations for the South Dakota Monitoring Sites	22-28
22-7   Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
       for Pollutants with Cancer UREs for the South Dakota Monitoring Sites	22-30
22-8   Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
       Approximations for Pollutants with Noncancer RfCs for the South Dakota
       Monitoring Sites	22-31

23-1   Geographical Information for the Texas Monitoring Sites	23-6
23-2   Population, Motor Vehicle, and Traffic Information for the Texas Monitoring Sites... 23-8
23-3   Average Meteorological Conditions near the Texas Monitoring Sites	23-10
23-4   Risk-Based Screening Results for the Texas Monitoring Sites	23-18
23-5   Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
       Texas Monitoring Sites	23-20
23-6   Risk Approximations for the Texas Monitoring Sites	23-25
23-7   Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
       for Pollutants with Cancer UREs for the Texas Monitoring Sites	23-26
23-8   Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
       Approximations for Pollutants with Noncancer RfCs for the Texas Monitoring
       Sites	23-27

24-1   Geographical Information for the Utah Monitoring Site	24-4
24-2   Population, Motor Vehicle, and Traffic Information for the Utah Monitoring Site	24-5
24-3   Average Meteorological Conditions near the Utah Monitoring Site	24-8
24-4   Risk-Based Screening Results for the Utah Monitoring Site	24-13
24-5   Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
       Utah Monitoring Site	24-15
24-6   Risk Approximations for the Utah Monitoring Site	24-32
24-7   Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
       for Pollutants with Cancer UREs for the Utah Monitoring Site	24-35
24-8   Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
       Approximations for Pollutants with Noncancer RfCs for the Utah Monitoring
       Site	24-36

                                           xl

-------
                          LIST OF TABLES (Continued)

                                                                                   Page
25-1   Geographical Information for the Vermont Monitoring Sites	25-7
25-2   Population, Motor Vehicle, and Traffic Information for the Vermont
       Monitoring Sites	25-9
25-3   Average Meteorological Conditions near the Vermont Monitoring Sites	25-12
25-4   Risk-Based Screening Results for the Vermont Monitoring Sites	25-22
25-5   Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
       Vermont Monitoring Sites	25-25
25-6   Risk Approximations for the Vermont Monitoring Sites	25-37
25-7   Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
       for Pollutants with Cancer UREs for the Vermont Monitoring Sites	25-40
25-8   Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
       Approximations for Pollutants with Noncancer RfCs for the Vermont
       Monitoring Sites	25-42

26-1   Geographical Information for the Virginia Monitoring Site	26-4
26-2   Population, Motor Vehicle, and Traffic Information for the Virginia Monitoring
       Site	26-5
26-3   Average Meteorological Conditions near the Virginia Monitoring Site	26-7
26-4   Risk-Based Screening Results for the Virginia Monitoring Site	26-12
26-5   Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
       Virginia Monitoring Site	26-14
26-6   Risk Approximations for the Virginia Monitoring Site	26-18
26-7   Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
       for Pollutants with Cancer UREs for the Virginia Monitoring Site	26-19
26-8   Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
       Approximations for Pollutants with Noncancer RfCs for the Virginia Monitoring
       Site	26-20

27-1   Geographical Information for the Washington Monitoring Site	27-4
27-2   Population, Motor Vehicle, and Traffic Information for the Washington
       Monitoring Site	27-5
27-3   Average Meteorological Conditions near the Washington Monitoring Site	27-8
27-4   Risk-Based Screening Results for the Washington Monitoring Site	27-13
27-5   Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
       Washington Monitoring Site	27-15
27-6   Risk Approximations for the Washington Monitoring Site	27-32
27-7   Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
       for Pollutants with Cancer UREs for the Washington Monitoring Site	27-34
27-8   Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
       Approximations for Pollutants with Noncancer RfCs for the Washington
       Monitoring Site	27-35

28-1   Geographical Information for the Wisconsin Monitoring Site	28-4
28-2   Population, Motor Vehicle, and Traffic Information for the Wisconsin
       Monitoring Site	28-5

                                           xli

-------
                         LIST OF TABLES (Continued)

                                                                                 Page
28-3   Average Meteorological Conditions near the Wisconsin Monitoring Site	28-8
28-4   Risk-Based Screening Results for the Wisconsin Monitoring Site	28-12
28-5   Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
       Wisconsin Monitoring Site	28-14
28-6   Risk Approximations for the Wisconsin Monitoring Site	28-16
28-7   Top 10 Emissions, Toxi city-Weighted Emissions, and Cancer Risk Approximations
       for Pollutants with Cancer UREs for the Wisconsin Monitoring Site	28-18
28-8   Top 10 Emissions, Toxi city-Weighted Emissions, and Noncancer Hazard
       Approximations for Pollutants with Noncancer RfCs for the Wisconsin
       Monitoring Site	28-19

29-1  Method Precision by Analytical Method	29-4
29-2  VOC Method Precision: Coefficient of Variation Based on Duplicate and Collocated
      Samples by Site	29-5
29-3  SNMOC Method Precision: Coefficient of Variation Based on Duplicate and
      Collocated Samples by Site	29-13
29-4  Carbonyl Compound Method Precision: Coefficient of Variation Based on Duplicate
      and Collocated Samples by Site	29-17
29-5  PAH Method Precision: Coefficient of Variation Based on Collocated Samples
      by Site	29-20
29-6  Metals Method Precision: Coefficient of Variation Based on  Collocated Samples
      by Site	29-21
29-7  Hexavalent Chromium Method Precision: Coefficient of Variation Based on
      Collocated Samples by Site	29-22
29-8  Analytical Precision by Analytical Method	29-23
29-9  VOC Analytical Precision: Coefficient of Variation Based on Replicate Analyses
      by Site	29-25
29-10 SNMOC Analytical Precision: Coefficient of Variation Based on Replicate Analyses
      by Site	29-33
29-11 Carbonyl Compound Analytical Precision: Coefficient of Variation Based on
      Replicate Analyses by Site	29-37
29-12 PAH Analytical Precision: Coefficient of Variation Based on Replicate Analyses
      by Site	29-40
29-13 Metals Analytical Precision:  Average Coefficient of Variation Based on Replicate
      Analyses by Site	29-41
29-14 Hexavalent Chromium Analytical Precision: Coefficient of Variation Based on
      Replicate Analyses by Site	29-42
29-15 VOC NATTS PT Audit Samples-Percent Difference from True Value	29-43
29-16 Carbonyl Compound NATTS PT Audit Samples-Percent Difference from
      True Value	29-44
29-17 PAH NATTS PT Audit Samples-Percent Difference from True Value	29-44
29-18 Metals NATTS PT Audit Samples-Percent Difference from True Value	29-44
29-19 Hexavalent Chromium PT Audit Samples-Percent Difference from True Value	29-44
                                         xlii

-------
                            LIST OF ACRONYMS

AADT        Average Annual 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
DQO          Data Quality Objective(s)
EPA          U.S. Environmental Protection Agency
ERG          Eastern Research Group, Inc.
F             Fahrenheit
FHWA        Federal Highway Administration
GC/MS-FID   Gas Chromatography/Mass Spectrometry and Flame lonization Detection
GHG          Greenhouse Gas(es)
GIS           Geographical Information System
GWP          Global Warming Potential
HAP          Hazardous Air Pollutant(s)
HPLC        High-Performance Liquid Chromatography
HQ           Hazard Quotient
HYSPLIT     Hybrid Single-Particle Lagrangian Integrated Trajectory
1C            Ion Chromatography
ICP-MS       Inductively Coupled Plasma/Mass Spectrometry
IPCC          Intergovernmental Panel on Climate Change
kt            Knots
mb           Millibar
MDL          Method Detection Limit
mg/m3        Milligrams per cubic meter
mL           Milliliter
MQO         Method Quality Objective(s)
MRL          Minimal Risk Level
MSA          Metropolitan or Micropolitan Statistical Area(s)
MTBE        Methyl tert-buty\ ether
NAAQS       National Ambient Air Quality Standard
NATA        National Air Toxics Assessment
                                       xliii

-------
                     LIST OF ACRONYMS (Continued)

NATTS       National Air Toxics Trends Stations
ND           Non-detect
NEI          National Emissions Inventory
ng/m3         Nanograms per cubic meter
NMOC        Non-Methane Organic Compound(s)
NMP         National Monitoring Programs
NOAA        National Oceanic and Atmospheric Administration
NOx          Oxides of Nitrogen
NWS         National Weather Service
PAMS        Photochemical Assessment Monitoring Stations
PAH          Polycyclic Aromatic Hydrocarbon(s)
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)
RFG          Reformulated Gasoline
SATMP       School Air Toxics Monitoring Program
SIM          Selected Ion Monitoring
SIP           State Implementation Plan(s)
SNMOC       Speciated Nonmethane Organic Compound(s)
UATMP       Urban Air Toxics Monitoring Program
VOC         Volatile Organic Compound(s)
TAD          Technical Assistance Document
TNMOC       Total Nonmethane Organic Compound(s)
tpy            Tons per year
TRI          Toxics Release Inventory
TSP          Total Suspended Particulate
TSV          Total Spatial Variance
|ig/m3         Micrograms per cubic meter
jiL            Microliter
URE          Unit Risk Estimate(s)
VMT         Vehicle Miles Traveled
WBAN        Weather Bureau/Army/Navy ID
                                       xliv

-------
                                       Abstract

      This report presents the results and conclusions from the ambient air monitoring conducted
as part of the 2011 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 2011 NMP includes data from
samples collected at 51 monitoring sites that collected 24-hour air samples, typically on a l-in-6
or l-in-12 day schedule. Twenty-four sites sampled for 61 volatile organic compounds (VOC); 31
sites sampled for 14 carbonyl compounds; eight sites sampled for 80 speciated nonmethane
organic compounds (SNMOC); 23 sites sampled for 22 polycyclic aromatic hydrocarbons (PAH);
15 sites sampled for 11 metals; and 22 sites sampled for hexavalent chromium. Over 218,900
ambient air concentrations were measured during the 2011 NMP.  This report uses various
graphical, numerical, and statistical analyses to put the vast amount of ambient air monitoring
data collected into perspective. Not surprisingly, the ambient air concentrations measured during
the program varied significantly from city-to-city and from season-to-season.

      The ambient air monitoring data collected during the 2011 NMP serve a wide range of
purposes. Not only do these data characterize the nature and extent of air pollution close to the
51 individual monitoring sites participating in these programs, but they also identify trends and
patterns that may be common to both 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.
                                           xlv

-------
1.0    Introduction
       Air pollution contains many components that originate from a wide range of stationary,
mobile, and natural emissions sources. Because some of these components include air toxics that
are known or suspected to have the potential for negative human health impacts, the
U.S. Environmental Protection Agency (EPA) encourages state, local, and tribal agencies to
understand and appreciate the nature and extent of toxic air pollution in their respective
locations. To achieve this goal, EPA sponsors the National Monitoring Programs (NMP), which
include the Photochemical Assessment Monitoring Stations (PAMS) network, Urban Air Toxics
Monitoring Program (UATMP), National Air Toxics Trends Stations (NATTS) network,
Community-Scale Air Toxics Ambient Monitoring (CSATAM) Program, and monitoring for
other pollutants such as Non-Methane Organic Compounds (NMOCs). This report focuses on the
UATMP, NATTS, and CSATAM programs. These programs have the following program-
specific objectives:
       •   The primary objective of the UATMP is to characterize the composition and
          magnitude of air toxics pollution through ambient air monitoring.
       •   The primary objective of the NATTS network is to obtain a statistically significant
          quantity of high-quality representative air toxics measurements such that long-term
          trends can be  identified.
       •   The primary objective of the CSATAM Program is to conduct local-scale
          investigative air toxics monitoring projects.

1.1    Background
       EPA  began the NMOC program in 1984. Monitoring for selected NMOCs was performed
during the morning hours of the summer ozone season. NMOC data were to be used to better
understand ozone formation and to develop ozone control strategies. The UATMP was initiated
by EPA in 1988 as an extension of the existing NMOC program to meet the increasing need for
information on air toxics. Over the years, the program has grown in both participation and
targeted pollutants (EPA, 2009a). The program has allowed for the identification of compounds
that are prevalent in ambient air and for participating agencies to screen air samples for
concentrations of air toxics that could potentially result in adverse human health effects.

       The NATTS network was created to generate long-term ambient air toxics concentration
data at specific fixed sites across the country. The 10-City Pilot program (LADCO, 2003) was
developed and implemented during 2001 and 2002, leading to the development and initial
                                          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 at fixed sites that remain active over
an extended period of time (EPA, 2009a). The generation of large quantities of high-quality data
over an extended period may allow concentration trends (i.e., any substantial increase or
decrease over a period of time) to be identified. The data generated are also used for validating
modeling results and emissions inventories, assessing current regulatory benchmarks, and
assessing the potential for developing cancerous and noncancerous health effects (EPA, 2012a).
The initial site locations were based on results from preliminary air toxics pilot programs such as
the 1996 National Air Toxics Assessment (NATA), which used air toxics emissions data to
model ambient monitoring concentrations across the nation. Monitoring sites were placed in both
urban and rural locations. Urban areas were chosen to measure population exposure, while rural
areas were chosen to determine background levels of air pollution (EPA, 2009b). Currently, 27
NATTS sites are strategically placed across the country (EPA, 2012b).

       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).

       Many environmental and health agencies have participated in these programs to assess
the sources, effects, and changes in air pollution within their jurisdictions. In past reports,
measurements  from UATMP, NATTS, and CSATAM monitoring sites  have been presented
together and referred to as "UATMP sites." In more recent reports, a distinction is made among
the three programs due to the increasing number of sites covered under  each program. Thus, it is
appropriate to  describe each program; to distinguish among their purposes and scopes; and to
integrate the data, which allows each program's objectives and goals to complement each other.

1.2    The Report
       This report summarizes and interprets the 2011 UATMP,  NATTS, and CSATAM
monitoring efforts of the NMP. Included in this report are data collected at the 51 monitoring

                                          1-2

-------
sites around the country 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. The 51 sites included in this report are located in or near 33 urban or rural
locations in 23 states and the District of Columbia, including 29 metropolitan or micropolitan
statistical areas (MSAs).

       This report provides both a qualitative overview of air toxics pollution at selected urban
and rural locations and a quantitative data analysis of the factors that appear to affect the
behavior of air toxics in urban and rural areas most significantly. This report also focuses on data
characterizations for each of the 51 different air sampling locations, a site-specific approach that
allows for a much more detailed evaluation of the factors  (e.g., emissions sources, natural
sources, meteorological influences) that affect air quality  differently from one location to the
next. Much of the data analysis and interpretation contained in this  report focuses on pollutant-
specific risk potential.

       This report offers participating agencies relevant information and insight into important
air quality issues. For example, participating agencies can use trends and patterns in the
monitoring data to determine whether levels of air pollution present public health concerns, to
identify which emissions sources contribute most to air pollution, or to forecast whether
proposed pollution control initiatives could significantly improve air quality. Monitoring data
may also be compared to modeling results, such as from EPA's NATA.

       Policy-relevant questions that the monitoring data may help answer include the
following:
       •   Which anthropogenic sources substantially affect air quality?
       •   Have pollutant concentrations decreased as a result of regulations (or increased
           despite regulation)?
       •   Which pollutants contribute the greatest health risk on a short-term, intermediate-
           term, and long-term basis?
                                            1-3

-------
       The data analyses contained in this report are applied to each participating UATMP,
NATTS, or 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 2011 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, 2012c).


       This report is organized into 31 sections and 17 appendices. While each state section is

designed to be a stand-alone section to allow those interested in a particular site or state to
understand the associated data analyses without having to read the entire report, it is

recommended that Sections 1  through 4 (Introduction, Monitoring Programs Network Overview,
Data Treatments and Methods, and Summary of Results) and Sections 29 and 30 (Data Quality

and Results, Conclusions, and Recommendations) be read as complements to the individual state
sections. Table  1-1 highlights the contents of each section.
        Table 1-1. Organization of the 2011 National Monitoring Programs Report
 Report
 Section
Section Title
Overview of Contents
           Introduction
                       This section serves as an introduction to the
                       background and scope of the NMP (specifically, the
                       UATMP, NATTS, and CSATAM).
           The 2011 National Monitoring
           Programs Network
                       This section provides information on the 2011 NMP
                       and network:
                       • Monitoring locations
                       • Pollutants selected for monitoring
                       • Sampling and analytical methods
                       • Sampling schedules
                       • Completeness of the air monitoring programs.
           Summary of the 2011 National
           Monitoring Programs Data
           Treatments and Methods
                       This section presents and discusses the data treatments
                       used on the 2011 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 risk.	
                                           1-4

-------
Table 1-1. Organization of the 2011 National Monitoring Programs Report (Continued)
Report
Section
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Section Title
Summary of the 201 1 National
Monitoring Programs Data
Sites in Arizona
Sites in California
Sites in Colorado
Site in the District of Columbia
Sites in Florida
Site in Georgia
Sites in Illinois
Sites in Indiana
Site in Kentucky
Site in Massachusetts
Sites in Michigan
Site in Missouri
Sites in New Jersey
Sites in New York
Sites in Oklahoma
Overview of Contents
This section presents and discusses the results of the
data treatments from the 201 1 NMP data.
Monitoring results for the sites in the Phoenix-Mesa-
Glendale, AZ MSA (PXSS and SPAZ)
Monitoring results for the sites in the Los Angeles-
Long Beach-Santa Ana, CA MSA (CELA), the
Riverside-San Bernardino-Ontario, CA MSA (RUCA),
and the San Jose-Sunnyvale-Santa Clara, CA MSA
(SJJCA)
Monitoring results for the sites in the Grand Junction,
CO MSA (GPCO) and Garfield County (BMCO,
BRCO, PACO, and RICO)
Monitoring results for the site in the Washington-
Arlington-Alexandria, DC-VA-MD-WV MSA
(WADC)
Monitoring results for the sites in the Orlando-
Kissimmee-Sanford, FL MSA (ORFL and PAFL) and
the Tampa-St. Petersburg-Clearwater, FL MSA
(AZFL, SKFL, and SYFL)
Monitoring results for the site in the Atlanta-Sandy
Springs-Marietta, GA MSA (SDGA)
Monitoring results for the sites in the Chicago- Joliet-
Naperville, IL-IN-WI MSA (NBIL and SPIL)
Monitoring results for the sites in the Chicago-Joliet-
Naperville, IL-IN-WI MSA (INDEM) and the
Indianapolis-Carmel, IN MSA (WPIN)
Monitoring results for the site in Gray son, KY
(GLKY)
Monitoring results for the site in the Boston-
Cambridge-Quincy, MA-NH MSA (BOMA)
Monitoring results for the sites in the Detroit- Warren-
Livonia, MI MSA (DEMI, RRMI, and SWMI)
Monitoring results for the site in the St. Louis, MO-IL
MSA (S4MO)
Monitoring results for the sites in the New York-
Northern New Jersey-Long Island, NY-NJ-PA MSA
(CHNJ, ELNJ, NBNJ, and PANJ)
Monitoring results for the site in the New York-
Northern New Jersey-Long Island, NY-NJ-PA MSA
(MONY) and the Rochester, NY MSA (ROCH)
Monitoring results for the sites in the Tulsa, OK MSA
(TOOK and TMOK), the Oklahoma City, OK MSA
(MWOK and OCOK), and Pryor Creek, OK (PROK)
                                     1-5

-------
Table 1-1. Organization of the 2011 National Monitoring Programs Report (Continued)
Report
Section
20
21
22
23
24
25
26
27
28
29
30
31
Section Title
Site in Rhode Island
Site in South Carolina
Sites in South Dakota
Sites in Texas
Site in Utah
Sites in Vermont
Site in Virginia
Site in Washington
Site in Wisconsin
Data Quality
Results, Conclusions, and
Recommendations
References
Overview of Contents
Monitoring results for the site in the Providence-New
Bedford-Fall River, RI-MA MSA (PRRI)
Monitoring results for the site in Chesterfield, SC
(CHSC)
Monitoring results for the sites in the Sioux City, IA-
NE-SD MSA (UCSD) and the Sioux Falls, SD MSA
(SSSD)
Monitoring results for the sites in the Houston-Sugar
Land-Baytown, TX MSA (CAMS 35) and the
Marshall, TX MSA (CAMS 85)
Monitoring results for the site in the Ogden-Clearfield,
UT MSA (BTUT)
Monitoring results for the sites in the Burlington-South
Burlington, VT MSA (BURVT and UNVT) and the
Rutland, VT MSA (RUVT)
Monitoring results for the site in the Richmond, VA
MSA (RIVA)
Monitoring results for the site in the Seattle-Tacoma-
Bellevue, WA MSA (SEWA)
Monitoring results for the site in the Beaver Dam, WI
MSA (HOWI)
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 201 1 NMP ambient air
monitoring data.
This section summarizes the most significant findings
of the report and makes several recommendations for
future projects that involve ambient air monitoring.
This section lists the references cited throughout the
report.
                                     1-6

-------
2.0    The 2011 National Monitoring Programs Network
       Agencies operating UATMP, NATTS, or CSATAM sites may choose to have their
samples analyzed by EPA's contract laboratory, ERG, in Morrisville, NC. Data from 51
monitoring sites that collected 24-hour integrated ambient air samples for up to 12 months, at
l-in-6 or l-in-12 day sampling intervals, and sent them to ERG for analysis are included in this
report.  Samples were analyzed for concentrations of selected hydrocarbons, halogenated
hydrocarbons, and polar compounds from canister samples (Speciated Nonmethane Organic
Compounds (SNMOCs) and/or Method TO-15),
carbonyl compounds from sorbent cartridge samples
(Method TO-11A), polycyclic aromatic hydrocarbons
(PAHs) from polyurethane foam (PUF) and XAD-2ฎ
resin samples (Method TO-13 A), trace metals from
filters (Method IO-3.5), and hexavalent chromium from
sodium bicarbonate-coated filters (EPA-approved
method). Section 2.2 provides further details on each of
the sampling methodologies used to collect and analyze
samples.
Agencies operating these sites 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 laboratories. In
these cases, data are generated by
sources other than ERG and are
not included in this report.
       The following sections review the monitoring locations, pollutants selected for
monitoring, sampling and analytical methods, collection schedules, and completeness of the
2011NMPdataset.

2.1    Monitoring Locations
       For the NATTS network, monitor siting is based on the need to assess population
exposure and background-level concentrations. For the UATMP and CSATAM programs,
representatives from the state, local, and tribal agencies that voluntarily participate in the
programs select the monitoring locations based on specific siting criteria and study needs.
Among these programs, monitors were placed in urban areas near the centers of heavily
populated cities (e.g., Chicago, IL and Phoenix, AZ), while others were placed in moderately
populated rural areas (e.g., Horicon, WI and Chesterfield, SC). Figure 2-1 shows the locations of
the 51 monitoring sites participating in the 2011 programs, which encompass 33  different urban
and rural areas. Outlined in Figure 2-1 are the associated core-based statistical areas (CBSA), as
                                          2-1

-------
designated by the U.S. Census Bureau, where each site is located (Census Bureau, 2010). A
CBS A 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, 2012a).

       Table 2-1  lists the respective monitoring program and the years of program participation
for the 51 monitoring sites. All 51 monitoring sites have been included in previous annual
reports, although two sites began sampling again under the NMP in 2011 after nine years. These
two sites are highlighted in Table 2-1.

       As Figure 2-1 and Table 2-1  show, the 2011 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 two programs. Table 2-2 shows that the locations of the monitoring
sites vary significantly from site to site. These sites are located in areas of differing elevation,
population, land use, climatology, and topography. A more detailed look at each monitoring
site's surroundings is provided in the individual state sections.
       For record-keeping and reporting purposes, each site was assigned the following:
       •  A unique four- or five-letter site code used to track samples from the monitoring site
          to the ERG laboratory.
       •  A unique nine-digit AQS site code used to index monitoring results in the AQS
          database.
       This report cites the four- or five-letter site code when presenting selected monitoring
results. For reference, each site's AQS site code is provided in Table 2-2.
                                           2-2

-------
             Figure 2-1. Locations of the 2011 National Monitoring Programs Monitoring Sites
                                         Underbill,
                                              A^
                                              Rutland, VT
                 	         Rochester, NY  \ LJ^Boston, MA
     Horicon.W."'  '    '~  '    '
       ,    Q I   \ Dearborn, Ml
       V, _   .^Detroit, MIV|E^  v^-Paterson, NJ
 North brook, It-rrHgl/Riyer Rouge, Ml  Chester, NJ
Schiller Park, lirฃ-fGary, i_N
            Indianapolis, IN
                                            Sioux Falls, SD
                                           Union CountyTSD
                                                                                                                 Providence, Rl
                                                                                                                 New York, NY
                                                                                                                 Elizabeth, NJ
                                                                                                             New Brunswick, NJ
 San Jose, CA       /    Parachute, CO  Rif|e co
             Battlement Mesa, CO       silt, CO
                                                                                                Washington,-DC
                    Grand Junction, CO
                                                              P/yor Creek, OK
                                                                     OK
                                          Oklahoma City, OK
                                                                                                  Chesterfield, SC
                                                    Karnack, TX

                                                 Deer Park, TX
                                                                                                    Winter Park, FL
                                                                                                     Orlando, FL
                                                                              Pinellas Park, FL
                                                                              St. Petersburg, FL
Legend
Program
•   CSATAM
o   NATTS
•   UATMP
    Metropolitan/Micropolitan Statistical Area

-------
                         Table 2-1. 2011 National Monitoring Programs Sites and Past Program Participation
Monitoring Location
and Site
Battlement Mesa, CO (BMCO)
Boston, MA (BOMA)
Bountiful, UT (BTUT)
Burlington, VT (BURVT)
Chester, NJ (CHNJ)
Chesterfield, SC (CHSC)
Dearborn, MI (DEMI)
Decatur, GA (SDGA)
Deer Park, TX (CAMS 35)
Detroit, MI (SWMI)
Elizabeth, NJ (ELNJ)
Gary, IN (INDEM)
Grand Junction, CO (GPCO)
Grayson, KY (GLKY)
Horicon, WI (HOW)
Indianapolis, IN (WPIN)
Karnack, TX (CAMS 85)
Program
UATMP
NATTS
NATTS
UATMP
UATMP
NATTS
NATTS
NATTS
NATTS
UATMP
UATMP
UATMP
NATTS
NATTS
NATTS
UATMP
NATTS
2001 and Earlier




2001

2001


2001
1999-2001






2002





-------
                   Table 2-1. 2011 National Monitoring Programs Sites and Past Program Participation (Continued)
Monitoring Location
and Site
Los Angeles, CA (CELA)
Midwest City, OK (MWOK)
New Brunswick, NJ (NBNJ)
New York, NY (MONY)
Northbrook, IL (NBIL)
Oklahoma City, OK (OCOK)
Orlando, FL (PAFL)
Parachute, CO (PACO)
Paterson, NJ (PANJ)
Phoenix, AZ (PXSS)
Phoenix, AZ (SPAZ)
Pinellas Park, FL (SKFL)
Plant City, FL (SYFL)
Providence, RI (PRRI)
Pryor Creek, OK (PROK)
Richmond, VA (RIVA)
Rifle, CO (RICO)
Program
NATTS
UATMP
UATMP
NATTS
NATTS
UATMP
UATMP
UATMP
CSATAM
NATTS
UATMP
NATTS
NATTS
NATTS
UATMP
NATTS
UATMP
2001 and Earlier


2001






2001
2001






2002


•/






•/







2003


•/


-------
                   Table 2-1. 2011 National Monitoring Programs Sites and Past Program Participation (Continued)
Monitoring Location
and Site
River Rouge, MI (RRMI)
Rochester, NY (ROCH)
Rubidoux, CA (RUCA)
Rutland, VT (RUVT)
San Jose, CA (SJJCA)
Schiller Park, IL (SPIL)
Seattle, WA (SEW A)
Silt, CO (BRCO)
Sioux Falls, SD (SSSD)
St. Louis, MO (S4MO)
St. Petersburg, FL (AZFL)
Tulsa, OK (TMOK)
Tulsa, OK (TOOK)
Underbill, VT (UNVT)
Union County, SD (UCSD)
Washington, D.C. (WADC)
Winter Park, FL (ORFL)
Program
UATMP
NATTS
NATTS
UATMP
NATTS
UATMP
NATTS
UATMP
UATMP
NATTS
UATMP
UATMP
UATMP
NATTS
UATMP
NATTS
UATMP
2001 and Earlier
2001


1995-1999






1991-1992, 2001





1990-1991
2002



S





•/
•/


•/



2003





•/



•/
•/






-------
                          Table 2-2. Site Characterizing Information for the 2011 National Monitoring Programs Sites
Site
Code
AZFL
BMCO
BOMA
BRCO
BTUT
BURVT
CAMS 35
CAMS 85
CELA
CHNJ
CHSC
DEMI
AQS
Code
12-103-0018
NA
25-025-0042
08-045-0009
49-011-0004
50-007-0014
48-201-1039
48-203-0002
06-037-1103
34-027-3001
45-025-0001
26-163-0033
Location
St. Petersburg, FL
Battlement Mesa, CO
Boston, MA
Silt, CO
Bountiful, UT
Burlington, VT
Deer Park, TX
Karnack TX
Los Angeles, CA
Chester NT
Chesterfield SC
Dearborn, MI
Land Use
Residential
Residential
Commercial
Agricultural
Residential
Commercial
Residential
Agricultural
Residential
Agricultural
Forest
Industrial
Location
Setting
Suburban
Rural
Urban/City
Center
Rural
Suburban
Urban/City
Center
Suburban
Rural
Urban/City
Center
Rural
Rural
Suburban
County-level
Population"
917,398
56,270
730,932
56,270
311,811
157,491
4,180,894
66,296
9,889,056
494,976
46,557
1,802,096
County-level
Vehicle
Registration,
# of Vehicles'1
(Year)
877,075
(2011)
72,957
(2010)
481,199
(2011)
72,957
(2010)
239,582
(2011)
169,767
(2012)
3,164,173
(2011)
70,858
fioin
7,360,573
(2011)
389,359
(2010)d
40,492
(2011)
1,334,752
(2011)
Estimated
Daily Traffic,
AADTb
(Year)
40,500
(2011)
2,527
(2002)
31,400
(2007)
150
(2002)
113,955
(2010)
14,000
(2007)
31,043
(2004)
1,250
(2010)
230,000
(2011)
12,917
(2010)
550
(2011)
92,800
(2011)
County-level
Stationary
Source HAP
Emissions
from the 2008
NEIC
(tpy)
1,381.26
1,364.26
572.38
1,364.26
391.90
347.53
9,322.29
593.11
14,794.19
198.46
97.19
7,384.27
County-level
Mobile Source
HAP
Emissions
from the 2008
NEIC
(tpy)
3,808.72
353.08
1,156.01
353.08
1,198.09
623.35
11,313.66
413.72
14,628.66
1,907.47
209.23
7,014.06
to
     BOLD ITALICS = EPA-designaled NATTS site
     ""Reference: Census Bureau, 2012b
     blndividual references provided in each state section.
     c Reference: EPA, 2012d
     dThe proportion of county-level population to the state-level population was applied to state-level vehicle registration figure and used as a surrogate when county-
     level vehicle registration counts were not available.
     eGPCO's hexavalent chromium monitor is at a separate, but adjacent, location; thus, this site has two AQS codes.
     NA = Data not loaded into AQS per agency request

-------
                   Table 2-2. Site Characterizing Information for the 2011 National Monitoring Programs Sites (Continued)
Site
Code
ELNJ
GLKY
GPCOe
HOW
INDEM
MONY
MWOK
NBIL
NBNJ
OCOK
ORFL
PACO
AQS
Code
34-039-0004
21-043-0500
08-077-0017
08-077-0018
55-027-0001
18-089-0022
36-005-0080
40-109-0041
17-031-4201
34-023-0006
40-109-1037
12-095-2002
08-045-0005
Location
Elizabeth, NJ
Grayson, KY
Grand Junction, CO
Horicon, WI
Gary, IN
New York, NY
Midwest City, OK
Northbrook, IL
New Brunswick NT
Oklahoma City, OK
Winter Park, FL
Parachute, CO
Land Use
Industrial
Residential
Commercial
Agricultural
Industrial
Residential
Commercial
Residential
Agricultural
Residential
Commercial
Residential
Location
Setting
Suburban
Rural
Urban/City
Center
Rural
Urban/City
Center
Urban/City
Center
Urban/City
Center
Suburban
Rural
Suburban
Urban/City
Center
Urban/City
Center
County-level
Population"
539,494
27,586
147,083
88,661
495,558
1,392,002
732,371
5,217,080
814,217
732,371
1,169,107
56,270
County-level
Vehicle
Registration,
# of Vehicles'1
(Year)
424,894
(2010)d
32,398
(2011)
178,425
(2010)
100,176
(2011)
419,431
(2011)
246,748
(2011)
832,160
(2011)
2,072,399
(2011)
640,893
(2010)d
832,160
(2011)
1,056,627
(2011)
72,957
(2010)
Estimated
Daily Traffic,
AADTb
(Year)
250,000
(2006)
428
(2009)
11,000
(2011)
5,000
(2008)
34,240
(2010)
91,465
(2010)
40,900
(2011)
34,600
(2011)
114,322
(2010)
40,900
(2011)
32,500
(2011)
16,000
(2011)
County-level
Stationary
Source HAP
Emissions
from the 2008
NEIC
(tpy)
360.61
55.08
532.80
531.88
1,486.55
2,171.17
1,242.77
15,376.26
475.76
1,242.77
1,791.25
1,364.26
County-level
Mobile Source
HAP
Emissions
from the 2008
NEIC
(tpy)
1,342.05
179.45
573.11
467.91
1,857.03
1,217.06
3,717.21
11,796.13
2,290.35
3,717.21
4,785.53
353.08
to
oo
     BOLD ITALICS = EPA-designaled NATTS site
     ""Reference: Census Bureau, 2012b
     blndividual references provided in each state section.
     c Reference: EPA, 2012d
     dThe proportion of county-level population to the state-level population was applied to state-level vehicle registration figure and used as a surrogate when county-
     level vehicle registration counts were not available.
     eGPCO's hexavalent chromium monitor is at a separate, but adjacent, location; thus, this site has two AQS codes.
     NA = Data not loaded into AQS per agency request

-------
                   Table 2-2. Site Characterizing Information for the 2011 National Monitoring Programs Sites (Continued)
Site
Code
PAFL
PANJ
PROK
PRRI
PXSS
RICO
RIVA
ROCH
RRMI
RUCA
RUVT
S4MO
AQS
Code
12-095-1004
34-031-0005
40-097-0187
44-007-0022
04-013-9997
08-045-0007
51-087-0014
36-055-1007
26-163-0005
06-065-8001
50-021-0002
29-510-0085
Location
Orlando, FL
Paterson, NJ
Pryor Creek, OK
Providence, RI
Phoenix, AZ
Rifle, CO
Richmond, VA
Rochester, NY
River Rouge, MI
Rubidoux, CA
Rutland, VT
St. Louis, MO
Land Use
Commercial
Commercial
Industrial
Residential
Residential
Commercial
Residential
Residential
Industrial
Residential
Commercial
Residential
Location
Setting
Suburban
Urban/City
Center
Suburban
Urban/City
Center
Urban/City
Center
Urban/City
Center
Suburban
Urban/City
Center
Suburban
Suburban
Urban/City
Center
Urban/City
Center
County-level
Population"
1,169,107
502,007
41,389
626,709
3,880,244
56,270
310,445
745,625
1,802,096
2,239,620
61,289
1,316,761
County-level
Vehicle
Registration,
# of Vehicles'1
(Year)
1,056,627
(2011)
396,602
(2010)d
39,968
(2011)
485,837
(2010)d
3,776,819
(2011)
72,957
(2010)
354,721
(2011)
550,992
(2011)
1,334,752
(2011)
1,711,492
(2011)
70,900
(2012)
1,114,812
(2011)
Estimated
Daily Traffic,
AADTb
(Year)
46,000
(2011)
22,272
(2010)
15,100
(2011)
136,800
(2009)
184,000
(2010)
17,000
(2011)
73,000
(2011)
86,198
(2010)
98,500
(2011)
145,000
(2011)
7,200
(2010)
79,558
(2011)
County-level
Stationary
Source HAP
Emissions
from the 2008
NEIC
(tpy)
1,791.25
162.17
329.16
906.46
1,618.22
1,364.26
740.28
1,809.55
7,384.27
2,552.70
135.82
1,054.65
County-level
Mobile Source
HAP
Emissions
from the 2008
NEIC
(tpy)
4,785.53
1,064.24
256.05
1,485.96
11,681.75
353.08
1,020.76
2,250.12
7,014.06
3,490.17
308.74
1,157.32
to
     BOLD ITALICS = EPA-designaled NATTS site
     ""Reference: Census Bureau, 2012b
     blndividual references provided in each state section.
     c Reference: EPA, 2012d
     dThe proportion of county-level population to the state-level population was applied to state-level vehicle registration figure and used as a surrogate when county-
     level vehicle registration counts were not available.
     eGPCO's hexavalent chromium monitor is at a separate, but adjacent, location; thus, this site has two AQS codes.
     NA = Data not loaded into AQS per agency request

-------
              Table 2-2. Site Characterizing Information for the 2011 National Monitoring Programs Sites (Continued)
Site
Code
SDGA
SEWA
SJJCA
SKFL
SPAZ
SPIL
SSSD
SWMI
SYFL
TMOK
TOOK
UCSD
AQS
Code
13-089-0002
53-033-0080
06-085-0005
12-103-0026
04-013-4003
17-031-3103
46-099-0008
26-163-0015
12-057-3002
40-143-1127
40-143-0235
46-127-0001
Location
Decatur, GA
Seattle, WA
San Jose, CA
Pinellas Park, FL
Phoenix, AZ
Schiller Park, IL
Sioux Falls, SD
Detroit, MI
Plant City, FL
Tulsa, OK
Tulsa, OK
TTninn County SD
Land Use
Residential
Industrial
Commercial
Residential
Residential
Mobile
Commercial
Commercial
Residential
Residential
Industrial
Agricultural
Location
Setting
Suburban
Suburban
Urban/City
Center
Suburban
Urban/City
Center
Suburban
Urban/City
Center
Urban/City
Center
Rural
Urban/City
Center
Urban/City
Center
Rural
County-level
Population"
699,893
1,969,722
1,809,378
917,398
3,880,244
5,217,080
171,752
1,802,096
1,267,775
610,599
610,599
14,651
County-level
Vehicle
Registration,
# of Vehicles'1
(Year)
472,535
(2011)
1,783,335
(2011)
1,517,190
(2011)
877,075
(2011)
3,776,819
(2011)
2,072,399
(2011)
210,914
(2011)
1,334,752
(2011)
1,135,945
(2011)
603,926
(2011)
603,926
(2011)
25,419
(2011)
Estimated
Daily Traffic,
AADTb
(Year)
140,820
(2011)
226,000
(2011)
104,000
(2011)
47,000
(2011)
128,000
(2010)
190,000
(2010)
18,700
(2011)
93,000
(2011)
10,600
(2011)
12,600
(2011)
63,000
(2011)
156
(2007)
County-level
Stationary
Source HAP
Emissions
from the 2008
NEIC
(tpy)
779.22
3,191.49
3,325.51
1,381.26
1,618.22
15,376.26
382.22
7,384.27
2,633.02
1,219.02
1,219.02
62.28
County-level
Mobile Source
HAP
Emissions
from the 2008
NEIC
(tpy)
3,044.68
9,694.40
2,772.68
3,808.72
11,681.75
11,796.13
600.33
7,014.06
4,579.82
3,065.07
3,065.07
122.79
BOLD ITALICS = EPA-designaled NATTS site
""Reference: Census Bureau, 2012b
blndividual references provided in each state section.
c Reference: EPA, 2012d
dThe proportion of county-level population to the state-level population was applied to state-level vehicle registration figure and used as a surrogate when county-
level vehicle registration counts were not available.
eGPCO's hexavalent chromium monitor is at a separate, but adjacent, location; thus, this site has two AQS codes.
NA = Data not loaded into AQS per agency request

-------
              Table 2-2. Site Characterizing Information for the 2011 National Monitoring Programs Sites (Continued)





Site
Code

UNVT

WADC

WPIN





AQS
Code

50-007-0007

11-001-0043

18-097-0078






Location

Underbill, VT

Washington, D.C.

Indianapolis, IN






Land Use

Forest

Commercial

Residential





Location
Setting

Rural
Urban/City
Center

Suburban





County-level
Population"

157,491

617,996

911,296


County-level
Vehicle
Registration,
# of Vehicles'1
(Year)
169,767
(2012)
213,232
(2010)
820,767
(2011)



Estimated
Daily Traffic,
AADTb
(Year)
1,100
(2011)
7,700
(2009)
143,970
(2010)
County-level
Stationary
Source HAP
Emissions
from the 2008
NEIC
(tpy)

347.53

632.23

2,965.43
County-level
Mobile Source
HAP
Emissions
from the 2008
NEIC
(tpy)

623.35

1,257.69

3,380.45
BOLD ITALICS = EPA-designated NATTS site
""Reference: Census Bureau, 2012b
blndividual references provided in each state section.
c Reference: EPA, 2012d
dThe proportion of county-level population to the state-level population was applied to state-level vehicle registration figure and used as a surrogate when county-
level vehicle registration counts were not available.
eGPCO's hexavalent chromium monitor is at a separate, but adjacent, location; thus, this site has two AQS codes.
NA = Data not loaded into AQS per agency request

-------
       The proximity of the monitoring sites to different emissions sources, especially industrial

facilities and heavily traveled roadways, often explains the observed spatial variations in ambient

air quality. To provide a first approximation of the potential contributions of stationary and

mobile source emissions on ambient air quality at each site, Table 2-2 also lists the following:

       •  The number of people living within each monitoring site's respective county.

       •  The county-level number of motor vehicles registered in each site's respective
          county, based on total vehicle registrations.

       •  The number of vehicles passing the nearest available roadway to the monitoring site,
          generally expressed as average annual daily traffic (AADT).

       •  Stationary and mobile source HAP emissions for the monitoring site's residing
          county, according to the 2008 National Emissions Inventory (NEI).


This information is discussed in further detail in the individual state sections.


2.2    Analytical Methods and Pollutants  Targeted for Monitoring

       Air pollution typically contains hundreds of components, including, but not limited to,

volatile organic compounds (VOCs), metals,  and paniculate matter (PM). Because the sampling

and analysis required to monitor for every component of air pollution has been prohibitively

expensive, the NMP focuses on specific pollutants that are analyzed using specific methods, as

listed below. The target pollutants varied significantly from monitoring site to monitoring site.

       •  Compendium Method TO-15 was used to measure ambient air concentrations of
          61 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
          14 carbonyl compounds.

       •  Compendium Method TO-13A was used to measure ambient air concentrations of
          22 PAHs.

       •  Compendium MethodIO-3.5 was used to measure ambient air concentrations of
          11 metals.

       •  EPA-approved hexavalent chromium method was used to measure ambient air
          concentrations of hexavalent chromium.
                                         2-12

-------
       At each monitoring site, the sample collection equipment 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 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, 2012e) in accordance with the specifications presented in the
NATTS Technical Assistance Document (TAD) (EPA, 2009b). This procedure involves
analyzing at least seven replicate  standards prepared on/in the appropriate sampling media (per
analytical method). Instrument-specific detection limits (replicate analysis of standards only) are
not determined because sample contamination and preparation variability would not be
considered.

       For the metals, however, the MDL procedure described by "Appendix D: DQ FAC
Single Laboratory Procedure v2.4" (FAC, 2007) was used to determine MDLs for chromium for
both quartz and Teflonฎ  filter types, as well as manganese, cobalt, nickel, cadmium, and lead for
quartz filters. The method involves analyzing at least seven replicate samples extracted from
                                          2-13

-------
blank sampling and calculating the MDLs from the results. For all other metals analytes, the
MDL procedure described in 40 CFR was used.

       Tables 2-3 through 2-8 identify the specific target pollutants for each analytical method
and their corresponding MDLs. For the VOC and SNMOC analyses, the experimentally-
determined MDLs do not change within a given year unless the sample was diluted. The 2011
VOC and SNMOC MDLs are presented in Tables 2-3 and 2-4, respectively. For the rest of the
analyses, the MDLs vary due to the actual volume pulled through the sample or if the sample
was diluted. For these analyses, the range and average MDL is presented for each pollutant in
Tables 2-5 through 2-8, based on valid samples. 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,
2006).

2.2.1   VOC and SNMOC Concurrent Sampling and Analytical Methods
       VOC and SNMOC sampling and analysis can be  performed concurrently in accordance
with a combination of EPA Compendium Method TO-15 (EPA, 1999a) and the procedure
presented in EPA's "Technical Assistance Document for Sampling and Analysis of Ozone
Precursors" (EPA, 1998). When referring to SNMOC, this report may refer to this method as the
"concurrent SNMOC method" or "concurrent SNMOC analysis" because both methods were
often employed at the same time to analyze the same sample. Ambient air samples for VOC
and/or SNMOC analysis were collected in passivated stainless steel canisters. The ERG
laboratory distributed the prepared canisters (i.e., cleaned and evacuated) to the monitoring sites
before each scheduled sample collection event, and site operators connected the canisters to air
sampling equipment prior to each sample day. Prior to field sampling, the passivated canisters
had internal pressures much lower than atmospheric pressure. Using this pressure differential,
ambient air flowed into the canisters automatically  once an associated system solenoid valve was
opened. A mass flow controller on the  sampling device inlet ensured that ambient air entered the
canister at an integrated constant rate across the collection period. At the end of the 24-hour

                                         2-14

-------
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 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 61 VOCs and/or 80
SNMOCs, and calculated the total nonmethane organic compounds (TNMOC) concentration.
TNMOC is the sum of all hydrocarbon concentrations within the sample. Because isobutene and
1-butene elute from the GC column at the same time, the SNMOC analytical method reports
only the sum concentration for these two compounds, and not the separate concentration for each
compound. The  same approach applies to m-xylene and/>-xylene for both the VOC and
concurrent SNMOC methods. Raw data for both methods are presented in Appendices  C  and D.

       Table 2-3 presents the MDLs for the laboratory analysis of VOC samples with Method
TO-15 and Table 2-4 presents the MDLs for the analysis of SNMOC samples. The MDL  for
every VOC is lower than 0.16 parts per billion by volume (ppbv). SNMOC detection limits are
expressed in parts per billion Carbon (ppbC). All of the SNMOC MDLs are less than 0.45 ppbC.
                                         2-15

-------
                        Table 2-3. 2011 VOC Method Detection Limits
Pollutant
Acetonitrile
Acetylene
Acrolein
Acrylonitrile
tert-Amyl Methyl Ether
Benzene
Bromochloromethane
Bromodichloromethane
Bromoform
Bromomethane
1,3 -Butadiene
Carbon Bisulfide
Carbon Tetrachloride
Chlorobenzene
Chloroethane
Chloroform
Chloromethane
Chloromethylbenzene
Chloroprene
Dibromochloromethane
2011
MDL
(ppbv)
0.0235
0.0163
0.0501
0.0116
0.0165
0.0288
0.0083
0.0247
0.0250
0.0091
0.0072
0.0331
0.0237
0.0212
0.0087
0.0088
0.0119
0.0389
0.0076
0.0207
Pollutant
1 ,2-Dibromoethane
/w-Dichlorobenzene
o-Dichlorobenzene
ฃ>-Dichlorobenzene
Dichlorodifluoromethane
1 , 1 -Dichloroethane
1 ,2-Dichloroethane
1, 1-Dichloroethene
c/'s-l,2-Dichloroethylene
/raซ5-l,2-Dichloroethylene
Dichloromethane
1 ,2-Dichloropropane
cis- 1 , 3 -Dichloropropene
trans- 1 , 3 -Dichloropropene
Dichlorotetrafluoroethane
Ethyl Aery late
Ethyl tert-Butyl Ether
Ethylbenzene
Hexachloro- 1 ,3 -Butadiene
Methyl Ethyl Ketone
2011
MDL
(ppbv)
0.0183
0.0342
0.0368
0.0348
0.0114
0.0084
0.0086
0.0088
0.0089
0.0081
0.0100
0.0227
0.0222
0.0246
0.0088
0.0200
0.0076
0.0169
0.0369
0.1570
Pollutant
Methyl Isobutyl Ketone
Methyl Methacrylate
Methyl tert-Butyl Ether
w-Octane
Propylene
Styrene
1 , 1 ,2,2-Tetrachloroethane
Tetrachloroethylene
Toluene
1 ,2,4-Trichlorobenzene
1,1,1 -Trichloroethane
1,1,2-Trichloroethane
Trichloroethylene
Trichlorofluoromethane
Trichlorotrifluoroethane
1 ,2,4-Trimethylbenzene
1 ,3 ,5-Trimethylbenzene
Vinyl Chloride
/w.^-Xylene1
o-Xylene
2011
MDL
(ppbv)
0.0223
0.0223
0.0083
0.0149
0.0393
0.0223
0.0244
0.0178
0.0162
0.0365
0.0222
0.0254
0.0254
0.0096
0.0109
0.0285
0.0262
0.0079
0.0336
0.0180
sum of/w-xylene and^-xylene concentrations and not concentrations of the individual isomers.
                                              2-16

-------
                      Table 2-4. 2011 SNMOC Method Detection Limits1
Pollutant
Acetylene
Benzene
1,3 -Butadiene
w-Butane
c/s-2-Butene
/raws-2-Butene
Cyclohexane
Cyclopentane
Cyclopentene
w-Decane
1-Decene
/w-Diethylbenzene
ฃ>-Diethylbenzene
2,2-Dimethylbutane
2,3 -Dimethy Ibutane
2,3 -Dimethy Ipentane
2,4-Dimethylpentane
w-Dodecane
1-Dodecene
Ethane
2-Ethyl-l-butene
Ethylbenzene
Ethylene
/w-Ethyltoluene
o-Ethyltoluene
ฃ>-Ethyltoluene
w-Heptane
2011
MDL
(ppbC)1
0.080
0.220
0.190
0.190
0.170
0.141
0.189
0.124
0.190
0.200
0.200
0.200
0.120
0.140
0.210
0.440
0.230
0.430
0.430
0.090
0.330
0.180
0.290
0.170
0.190
0.250
0.190
Pollutant
1-Heptene
w-Hexane
1-Hexene
c/s-2-Hexene
/raซs-2-Hexene
Isobutane
Isobutene/ 1 -Butene2
Isopentane
Isoprene
Isopropylbenzene
2-Methy 1-1 -Butene
3 -Methy 1-1 -Butene
2-Methy 1- 1 -Pentene
4-Methyl- 1 -Pentene
2-Methyl-2-Butene
Methylcyclohexane
Methylcyclopentane
2-Methy Iheptane
3-Methylheptane
2-Methy Ihexane
3-Methylhexane
2-Methy Ipentane
3 -Methy Ipentane
w-Nonane
1-Nonene
w-Octane
1-Octene
2011
MDL
(ppbC)1
0.440
0.180
0.330
0.330
0.330
0.160
0.130
0.180
0.190
0.200
0.190
0.190
0.330
0.330
0.190
0.180
0.160
0.130
0.160
0.220
0.140
0.150
0.170
0.210
0.250
0.230
0.260
Pollutant
w-Pentane
1 -Pentene
c/s-2-Pentene
/raซs-2-Pentene
a-Pinene
6-Pinene
Propane
w-Propylbenzene
Propylene
Propyne
Styrene
Toluene
w-Tridecane
1-Tridecene
1,2,3-Trimethylbenzene
1,2,4-Trimethylbenzene
1 , 3 ,5 -Trimethy Ibenzene
2,2,3 -Trimethy Ipentane
2,2,4-Trimethy Ipentane
2,3,4-Trimethy Ipentane
w-Undecane
1-Undecene
7w-Xylene//?-Xylene2
o-Xylene
Sum of Knowns
Sum of Unknowns
TNMOC
2011
MDL
(ppbC)1
0.110
0.130
0.180
0.140
0.200
0.200
0.100
0.190
0.090
0.100
0.260
0.280
0.430
0.430
0.150
0.180
0.190
0.260
0.160
0.150
0.200
0.200
0.240
0.170
NA
NA
NA
1 Concentration in ppbC = concentration in ppbv * number of carbon atoms in the compound.
2 Because isobutene and 1-butene elute from the GC column at the same time, the SNMOC analytical method reports
 the sum concentration for these two compounds and not concentrations of the individual compounds. For the same
 reason, the /w-xylene and^-xylene concentrations are reported as a sum concentration.
NA = Not applicable
                                               2-17

-------
2.2.2   Carbonyl Compound Sampling and Analytical Method
       Following the specifications of EPA Compendium Method TO-11A (EPA, 1999b),
ambient air samples for carbonyl compound analysis were collected by passing ambient air
through an ozone scrubber and then through cartridges containing silica gel coated with
2,4-dinitrophenylhydrazine (DNPH), a compound known to react selectively and reversibly with
many aldehydes and ketones. Carbonyl compounds in ambient air are retained in the sampling
cartridge, while other compounds pass through the cartridge without reacting with the DNPH-
coated matrix. The ERG laboratory distributed the DNPH cartridges to the monitoring sites prior
to each scheduled sample collection event and site operators connected the cartridges to the air
sampling equipment. After each 24-hour sampling period, site operators recovered and returned
the cartridges, along with the Chain of Custody forms and all associated documentation, to the
ERG laboratory for analysis.

       To quantify concentrations of carbonyl compounds in the  sampled ambient air, laboratory
analysts extracted the exposed DNPH cartridges with acetonitrile. High-performance liquid
chromatography (HPLC) analysis and ultraviolet 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 14 carbonyl compounds. 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 for every pollutant is less than 0.01 ppbv.
                                         2-18

-------
                 Table 2-5. 2011 Carbonyl Compound Method Detection Limits
Pollutant
Acetaldehyde
Acetone
Benzaldehyde
Butyraldehyde
Crotonaldehyde
2,5-Dimethylbenzaldehyde
Formaldehyde
Hexaldehyde
Isovaleraldehyde
Propionaldehyde
Tolualdehydes1
Valeraldehyde
Minimum
MDL
(ppbv)
0.0040
0.0040
0.0010
0.0020
0.0020
0.0010
0.0060
0.0010
0.0009
0.0020
0.0020
0.0010
Maximum
MDL
(ppbv)
0.0300
0.0300
0.0100
0.0140
0.0140
0.0100
0.0840
0.0090
0.0070
0.0110
0.0140
0.0110
Average
MDL
(ppbv)
0.0068
0.0068
0.0024
0.0032
0.0032
0.0024
0.0099
0.0020
0.0014
0.0026
0.0033
0.0025
            1 The three tolualdehyde isomers elute from the HPLC column at the same time; thus,
             the analytical method reports only the sum concentration for these three isomers and
             not the individual concentrations.

2.2.3   PAH Sampling and Analytical Method
       PAH sampling and analysis was performed in accordance with EPA Compendium
Method TO-13A (EPA, 1999c) and ASTM D6209 (ASTM, 2013). The ERG laboratory prepared
sampling media and supplied them to the sites before each scheduled sample collection event.
The clean sampling PUF/XAD-2ฎ cartridge and glass fiber filter are installed in a high volume
sampler by the site operators and allowed to sample for 24 hours. Sample collection modules and
Chain of Custody forms and all associated documentation were returned to the ERG laboratory
after sample collection. Within 14 days of sampling, the filter and cartridge are extracted
together using a toluene in hexane solution using the Dionex Accelerated Solvent Extractor
(ASE) 350 or ASE 300. The sample extract is concentrated to a final volume of 1.0 milliliter
(mL). A volume of 1 microliter (uL) is injected into the GC/MS operating in the SIM mode to
analyze 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 ranged from
0.032 ng/m3 (pyrene) to 0.145 ng/m3 (naphthalene).
                                         2-19

-------
                      Table 2-6. 2011 PAH Method Detection Limits
Pollutant
Acenaphthene
Acenaphthylene
Anthracene
Benzo(a)anthracene
Benzo(a)pyrene
Benzo(b)fluoranthene
Benzo(e)pyrene
Benzo(g,h,i)perylene
Benzo(k)fluoranthene
Chrysene
Coronene
Cyclopenta[cd]pyrene
Dibenz(a,h)anthracene
Fluoranthene
Fluorene
9-Fluorenone
Indeno(l,2,3-cd)pyrene
Naphthalene
Perylene
Phenanthrene
Pyrene
Retene
Minimum
MDL
(ng/m3)
0.029
0.025
0.023
0.027
0.023
0.025
0.024
0.016
0.025
0.021
0.013
0.033
0.018
0.022
0.030
0.026
0.018
0.067
0.021
0.022
0.019
0.026
Maximum
MDL
(ng/m3)
0.545
0.483
0.625
0.904
0.887
0.465
0.451
0.780
0.640
0.416
0.863
0.718
0.800
0.406
0.561
0.484
0.754
1.570
0.865
0.409
0.349
0.481
Average
MDL
(ng/m3)
0.051
0.045
0.057
0.083
0.081
0.043
0.042
0.071
0.059
0.038
0.079
0.066
0.073
0.038
0.052
0.045
0.069
0.145
0.079
0.038
0.032
0.048
2.2.4   Metals Sampling and Analytical Method
       Sampling for the determination of metals in or on particulate matter was performed by
the sites in accordance with EPA Compendium Method IO-3.5  (EPA, 1999d). Ambient air
samples for metals analysis were collected by passing ambient air through either 47mm Teflonฎ
filters or 8" x 10" quartz filters, depending on the separate and distinct sampling apparatus used
to collect the sample; the 47mm Teflonฎ filter is used for low-volume samplers, whereas the
8" x 10" quartz filter is used for high-volume samplers. EPA provides the filters to the
monitoring sites. Sites sampled for either parti culate matter less than 10 microns (PMio) or total
suspended particulate (TSP). Particulates in ambient air were collected on the filters and, after a
24-hour sampling period, site operators recovered and returned the filters, along with the Chain
of Custody forms and all associated documentation, to the ERG laboratory for analysis.

       Upon receipt at the laboratory, the whole filters (47mm Teflonฎ) or filter strips
(8" x 10" quartz) were digested using a dilute nitric acid, hydrochloric acid, and/or hydrofluoric
                                          2-20

-------
acid (Teflon  only) solution. The digestate was then quantified using inductively coupled plasma
mass spectrometry (ICP-MS) to determine the concentration of individual metals present in the
original air sample. Raw data for speciated metals are presented in Appendix G.

       Table 2-7 lists the MDLs for the analysis of the metals samples. Due to the difference in
sample volume/filter collection media, there are two sets of MDLs listed in Table 2-7. Although
the sensitivity varies from pollutant-to-pollutant and from site-to-site due to the different
volumes pulled through the samples, the average MDL for valid samples ranged from
0.007 ng/m3 (mercury) to 2.07 ng/m3 (chromium) for the quartz filters and from 0.010 ng/m3
(cadmium) to 8.14 ng/m3 (chromium) for the Teflonฎ filters.
                     Table 2-7. 2011 Metals Method Detection Limits
Pollutant
Minimum
MDL
(ng/m3)
Maximum
MDL
(ng/m3)
Average
MDL
(ng/m3)
8 X 10" Quartz Filters
Antimony
Arsenic
Beryllium
Cadmium
Chromium
Cobalt
Lead
Manganese
Mercury
Nickel
Selenium
0.007
0.033
0.001
0.004
1.430
0.015
0.060
0.074
0.005
0.230
0.009
0.055
0.389
0.059
0.079
3.530
0.067
19.400
23.000
0.565
3.940
0.513
0.013
0.049
0.008
0.010
2.072
0.025
0.132
0.163
0.007
0.361
0.021
Pollutant
Minimum
MDL
(ng/m3)
Maximum
MDL
(ng/m3)
Average
MDL
(ng/m3)
47mm Teflonฎ Filters
Antimony
Arsenic
Beryllium
Cadmium
Chromium
Cobalt
Lead
Manganese
Mercury
Nickel
Selenium
0.020
0.100
0.010
0.010
6.080
0.020
0.020
0.200
0.020
0.290
0.060
0.060
0.220
0.020
0.020
10.800
0.030
0.080
0.270
0.030
1.500
0.180
0.023
0.160
0.011
0.010
8.138
0.021
0.024
0.204
0.022
1.096
0.069
2.2.5   Hexavalent Chromium Sampling and Analytical Method
       Hexavalent chromium was measured using an EPA-approved approach. For a detailed
description of the method, refer to the "Standard Operating Procedure for the Determination of
Hexavalent Chromium in Ambient Air Analyzed by Ion Chromatography (1C)" (EPA, 2006).
Ambient air samples for hexavalent chromium analysis were collected by passing ambient air
through sodium bicarbonate impregnated acid-washed cellulose filters. ERG prepared and
distributed filters secured in Teflonฎ cartridges to the monitoring sites prior to each scheduled
sample collection event and site operators connected the cartridges to the air sampling
equipment. After a 24-hour sampling period, site operators recovered the cartridges and Chain of
                                         2-21

-------
Custody forms and returned them to the ERG laboratory for analysis. Upon receipt at the
laboratory, the filters were extracted using a sodium bicarbonate solution. Ion chromatography
(1C) analysis and ultraviolet-visible detection of these extracts determined the amount of
hexavalent chromium present in each sample.

       Although the sensitivity varies from site-to-site due to the different volumes pulled
through the samples, the average detection limit for valid samples reported by the ERG
laboratory across the program, which is presented in Table 2-8, is 0.0040 ng/m3. Raw data for
the hexavalent chromium method are presented in Appendix H.
             Table 2-8. 2011 Hexavalent Chromium Method Detection Limits
Pollutant
Hexavalent Chromium
Minimum
MDL
(ng/m3)
0.0024
Maximum
MDL
(ng/m3)
0.0081
Average
MDL
(ng/m3)
0.0040
2.3    Sample Collection Schedules
       Table 2-9 presents the first and last date on which sample collection occurred for each
monitoring site sampling under the NMP in 2011. The first sample date for each site is generally
at the beginning of January 2011 and sampling continued through the end of December 2011,
although there were a few exceptions:
       •  The San Jose, CA site (SJJCA) samples PMio metals under the NMP. However, in
          December 2010, this site changed samplers. This site began sampling PMio metals
          using a low-volume sampler with Teflonฎ filters on December 16, 2010 (rather than
          the previous high-volume sampler with quartz filters). As a result, data from the three
          December 2010 samples collected with the low-volume sampler have been included
          with the 2011 data for this site.
       •  The River Rouge and Detroit, MI sites (RRMI and SWMI) began sampling carbonyl
          compounds under the NMP at the end of January.
       •  In May, the Grayson, KY site (GLKY)  began sampling PMio metals under the NMP,
          in addition to VOCs, hexavalent chromium, and PAHs. This site also began sampling
          carbonyl compounds under the NMP in August.
       •  The Silt, CO site (BRCO) experienced carbonyl compound sampler problems,
          delaying sampling until September 2011.
       •  The Paterson, NJ site (PANJ) stopped sampling under the NMP in May.
                                        2-22

-------
       •  The Midwest City, OK site (MWOK) stopped sampling in November 2011 and the
          instrumentation was moved to a different location in Oklahoma City. Because less
          than one month of data was available for 2011 for this new site, data from the new
          location are not included in the 2011 NMP annual report and will be included in the
          2012 NMP report.


       According to the NMP schedule, 24-hour integrated samples were to be collected at each
monitoring site every l-in-6 days (or l-in-12 days, dependent upon location and monitoring

objectives) and each sample collection began and ended at midnight, local standard time.

However, there were some exceptions:

       •  The Garfield County, CO sites (BMCO, BRCO, PACO, RICO) collected samples by
          initiating the samplers manually. For these sites, samples were generally collected
          from mid-morning of one day to mid-morning of the next. In addition, SNMOC
          samples were collected on a l-in-6 day schedule while carbonyl compounds were
          collected on a l-in-12  day schedule.

       •  The South Phoenix, AZ site (SPAZ) collected VOC samples on a l-in-12 day
          schedule.

       •  The Paterson, NJ  site (PANJ) collected VOC samples on a l-in-12 day schedule.

       •  The Orlando, FL site (PAFL) collected metals samples on a l-in-12  day schedule.

       •  The Burlington, VT and Rutland, VT sites (BURVT and RUVT) collected VOC
          samples on a l-in-12 day  schedule.
                                        2-23

-------
                                                  Table 2-9. 2011 Sampling Schedules and Completeness Rates
Site
AZFL
BMCO
BOMA
BRCO
BTUT
BURVT2
CAMS 35
CAMS 85
CELA
CHNJ
CHSC
DEMI
ELNJ
GLKY
Monitoring Period1
First
Sample
1/3/11
1/3/11
1/3/11
1/3/11
1/3/11
1/9/11
1/3/11
1/3/11
1/3/11
1/3/11
1/3/11
1/3/11
1/3/11
1/3/11
Last
Sample
12/31/11
12/29/11
12/29/11
12/29/11
12/29/11
12/23/11
12/29/11
12/29/11
12/29/11
12/29/11
12/29/11
12/29/11
12/29/11
12/29/11
Carbonyl
Compounds
A
62
20
_
6
60
	
_
	
_
61
„
62
61
26
B
61
30
„
9
61
	
„
	
_
61
„
61
61
26
C
>100
672
„
672
98
	
„
	
„
100
„
>100
100
100
voc
A
„
	
_
	
60
30
„
	
„
61
„
61
61
61
B
„
	
„
	
61
30
_
	
_
61
„
61
61
61
C
„
	
„
	
98
100
_
	
_
100
„
100
100
100
Hexavalent
Chromium
A
„
	
61
	
61
	
60
61
„
	
60
61
„
61
B
„
	
61
	
61
	
61
61
_
	
61
61
„
61
C
„
	
100
	
100
	
98
100
„
	
98
100
„
100
Metals
A
_
	
60
	
60
	
„
	
„
	
„
	
„
41
B
„
	
61
	
61
	
„
	
„
	
„
	
_
41
C
„
	
98
	
98
	
„
	
„
	
„
	
„
100
SNMOC
A
„
51
„
54
60
	
„
	
„
	
_
	
„
	
B
„
61
„
61
61
	
„
	
_
	
„
	
„
	
C
_
84
„
89
98
	
„
	
„
	
„
	
„
	
PAH
A
„
	
61
	
62
	
58
	
59
	
60
60
„
61
B
„
	
61
	
61
	
61
	
61
	
61
61
_
61
C
„
	
100
	
>100
	
95
	
97
	
98
98
„
100
to
to
   A = Number of valid samples collected.
   B = Number of valid samples that should be collected in 2011 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.
     Sampling schedule was a l-in-12 day schedule rather than a l-in-6 schedule.
     Includes three samples from December 2010.
   BOLD ITALICS = EPA-designated NATTS site.
   Shading indicates that completeness is below the MQO of 85%.

-------
                                          Table 2-9. 2011 Sampling Schedules and Completeness Rates (Continued)
Site
GPCO
HOW
INDEM
MONY
MWOK
NBIL
NBNJ
OCOK
ORFL
PACO
PAFL2
PANJ2
PROK
PRRI
PXSS
Monitoring Period1
First
Sample
1/3/11
1/3/11
1/3/11
1/3/11
1/3/11
1/3/11
1/3/11
1/3/11
1/3/11
1/3/11
1/3/11
1/3/11
1/3/11
1/3/11
1/3/11
Last
Sample
12/29/11
12/29/11
12/29/11
12/29/11
11/29/11
12/29/11
12/29/11
12/29/11
12/29/11
12/29/11
12/29/11
5/15/11
12/29/11
12/29/11
12/29/11
Carbonyl
Compounds
A
60
	
57
	
56
62
62
60
60
21
_
_
58
_
48
B
61
	
61
	
56
61
61
61
61
30
_
_
61
„
61
C
98
	
93
	
100
>100
>100
98
98
702
„
„
95
„
79
voc
A
60
	
„
	
56
55
58
61
_
	
„
12
56
„
61
B
61
	
„
	
56
61
61
61
„
	
„
12
61
_
61
C
98
	
„
	
100
90
95
100
„
	
_
100
92
„
100
Hexavalent
Chromium
A
59
61
„
61
„
61
„
	
„
	
„
„
„
57
62
B
61
61
„
61
„
61
„
	
„
	
„
„
„
61
61
C
97
100
„
100
„
100
„
	
„
	
_
„
„
93
>100
Metals
A
_
	
_
	
56
53
„
61
„
	
31
_
56
„
61
B
„
	
„
	
56
61
_
61
„
	
31
„
61
„
61
C
„
	
„
	
100
87
„
100
„
	
100
„
92
„
100
SNMOC
A
„
	
„
	
„
54
„
	
„
53
„
„
„
_
	
B
„
	
„
	
„
61
„
	
„
61
„
„
„
„
	
C
„
	
_
	
„
89
„
	
„
87
_
_
„
„
	
PAH
A
61
	
„
60
„
61
„
	
„
	
„
„
„
57
57
B
61
	
_
61
„
61
_
	
_
	
_
_
_
60
61
C
100
	
„
98
„
100
„
	
„
	
„
„
„
95
93
to
to
   A = Number of valid samples collected.
   B = Number of valid samples that should be collected 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.
     Sampling schedule was a l-in-12 day schedule rather than a l-in-6 schedule.
     Includes three samples from December 2010.
   BOLD ITALICS = EPA-designated NATTS site.
   Shading indicates that completeness is below the MQO of 85%.

-------
                                          Table 2-9. 2011 Sampling Schedules and Completeness Rates (Continued)
Site
RICO
RIVA
ROCH
RRMI
RUCA
RUVT2
S4MO
SDGA
SEWA
SJJCA3
SKFL
SPAZ2
SPIL
SSSD
SWMI2
Monitoring Period1
First
Sample
1/3/11
1/3/11
1/3/11
1/21/11
1/3/11
1/9/11
1/3/11
1/3/11
1/3/11
12/16/10
1/3/11
1/6/11
1/3/11
1/3/11
1/27/11
Last
Sample
12/29/11
12/29/11
12/29/11
12/29/11
12/29/11
12/23/11
12/29/11
12/31/11
12/29/11
12/29/11
12/29/11
12/29/11
12/29/11
12/29/11
12/29/11
Carbonyl
Compounds
A
17
	
_
57
„
	
59
	
60
	
61
	
62
59
28
B
30
	
_
58
„
	
61
	
61
	
61
	
61
61
29
C
572
	
„
98
„
	
97
	
98
	
100
	
>100
97
97
voc
A
_
	
„
	
„
30
57
	
61
	
„
31
56
60
	
B
„
	
„
	
„
30
61
	
61
	
„
31
61
61
	
C
„
	
_
	
„
100
93
	
100
	
„
100
92
98
	
Hexavalent
Chromium
A
„
61
56
	
„
	
61
61
61
	
62
	
„
	
	
B
_
61
61
	
„
	
61
61
61
	
61
	
„
	
	
C
„
100
92
	
_
	
100
100
100
	
>100
	
„
	
	
Metals
A
„
	
_
	
„
	
59
	
61
64
_
	
„
	
	
B
„
	
„
	
„
	
61
	
61
64
„
	
„
	
	
C
_
	
„
	
_
	
97
	
100
100
„
	
„
	
	
SNMOC
A
53
	
„
	
_
	
„
	
„
	
„
	
„
60
	
B
61
	
„
	
„
	
„
	
„
	
„
	
„
61
	
C
87
	
„
	
„
	
„
	
„
	
„
	
_
98
	
PAH
A
„
61
58
	
61
	
61
61
60
61
61
	
„
	
	
B
_
61
61
	
61
	
61
61
61
61
61
	
_
	
	
C
„
100
95
	
100
	
100
100
98
100
100
	
„
	
	
to
to
   A = Number of valid samples collected.
   B = Number of valid samples that should be collected 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.
     Sampling schedule was a l-in-12 day schedule rather than a l-in-6 schedule.
     Includes three samples from December 2010.
   BOLD ITALICS = EPA-designated NATTS site.
   Shading indicates that completeness is below the MQO of 85%.

-------
                                          Table 2-9. 2011 Sampling Schedules and Completeness Rates (Continued)
Site
SYFL
TMOK
TOOK
UCSD
UNVT
WADC
WPIN
Monitoring Period1
First
Sample
1/3/11
1/3/11
1/3/11
1/6/11
1/3/11
1/3/11
1/3/11
Last
Sample
12/29/11
12/29/11
12/29/11
12/29/11
12/29/11
12/29/11
12/29/11
Carbonyl
Compounds
A
60
61
56
61
_
_
51
B
61
61
61
61
„
_
61
C
98
100
92
100
„
„
84
voc
A
„
60
57
56
60
„
—
B
_
61
61
61
61
_
—
C
_
98
93
92
98
_
—
Hexavalent
Chromium
A
59
	
„
	
60
61
—
B
61
	
„
	
61
61
—
C
97
	
„
	
98
100
—
Metals
A
„
58
56
	
57
_
—
B
„
61
61
	
61
„
—
C
„
95
92
	
93
„
—
SNMOC
A
„
	
„
56
„
„
—
B
„
	
_
61
„
„
—
C
„
	
„
92
„
„
—
PAH
A
60
	
„
	
60
61
—
B
61
	
„
	
61
61
—
C
98
	
_
	
98
100
—
   A = Number of valid samples collected.
to B = Number of valid samples that should be collected based on sample schedule and start/end date of sampling.
K> 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.
     Sampling schedule was a l-in-12 day schedule rather than a l-in-6 schedule.
     Includes three samples from December 2010.
   BOLD ITALICS = EPA-designated NATTS site.
   Shading indicates that completeness is below the MQO of 85%.

-------
       Table 2-9 shows the following:
       •  24 sites collected VOC samples and 31 sites collected carbonyl compound samples;
          VOC and carbonyl compound samples were collected concurrently at 19 sites.
       •  8 sites collected SNMOC samples.
       •  23 sites collected PAH samples.
       •  15 sites collected metals samples.
       •  22 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 for a given scheduled sample day, site operators were
instructed to reschedule or "make up" samples on other days. This practice explains why some
monitoring locations periodically strayed from the l-in-6 or l-in-12 day sampling schedule.

       The l-in-6 or  l-in-12 day sampling schedule provides cost-effective approaches to data
collection for trends characterization of toxic pollutants in ambient air and ensures that sample
days are evenly distributed among the seven days of the week to allow weekday/weekend
comparison of air quality. Because the l-in-6 day schedule yields twice the number of
measurements than the l-in-12 day schedule, data characterization based on this schedule tends
to be more representative.

2.4    Completeness
       Completeness refers to the number of valid samples collected and analyzed compared to
the number of total samples expected based on a l-in-6 or l-in-12 day sample schedule.
Monitoring programs that consistently generate valid samples have higher completeness than
programs that consistently have invalid samples. The completeness of an air monitoring
program, therefore, can be a qualitative measure of the reliability of air sampling and laboratory
                                          2-28

-------
analytical equipment as well as a measure of the efficiency with which the program is managed.
The completeness for each monitoring site and method sampled is presented in Table 2-9.

       The measurement quality objective (MQO) for completeness based on the EPA-approved
Quality Assurance Project Plan (QAPP) specifies that at least 85 percent of samples from a given
monitoring site must be collected and analyzed successfully to be considered sufficient for data
trends analysis (ERG, 2011). The data in Table 2-9 show that seven datasets from a total of 123
datasets from the 2011 NMP monitoring sites did not meet this MQO (shaded cells in Table 2-9):
       •   Four of the seven site-method combinations for which completeness was less than
          85 percent were for Garfield County carbonyl compound sites (BRCO, BMCO,
          PACO, and RICO). These sites tended to experience issues with their carbonyl
          compound samplers.  Prolonged issues resulted in BRCO sampling carbonyl
          compounds for only four months out of the year.
       •   Canister issues and several missed sample days led to a completeness less than
          85 percent for BMCO SNMOC.
       •   Maintenance of the primary carbonyl compound sampler at PXSS in 2010 led to a
          problem with the ozone denuder. As a result, the primary sampling results from mid-
          February 2010 through March 2011 were invalidated.
       •   Intermittent sampler issues throughout 2011 resulted in a carbonyl compound
          completeness less than 85  percent for WPIN.

       Although the completeness for S4MO's VOC is 93 percent, it should be noted that the
Missouri Department of Natural Resources discovered a sampler contamination issue and
invalidated all of its acrylonitrile results for this site. This is discussed in more detail in the
Missouri state section (Section 16).

       Appendix I identifies samples that were invalidated and lists the reason for invalidation,
based on the applied AQS null code.

       Table 2-10 presents method-specific completeness. Method-specific completeness was
greater than 90 percent for all six methods performed under the 2011 NMP and ranged from
90.37 percent for SNMOC to 98.96 percent for hexavalent chromium.
                                         2-29

-------
Table 2-10. Method Completeness Rates for 2011
Method
voc
SNMOC
Carbonyl Compounds
PAH
Metals Analysis
Hexavalent Chromium
# of Valid
Samples
1,281
441
1,594
1,382
834
1,328
#of
Samples
Scheduled
1,318
488
1,671
1,402
863
1,342
Method
Completeness
(%)
97.19
90.37
95.39
98.57
96.64
98.96
Minimum
Site-Specific
Completeness
(%)
90.16
(NBNJ)
83.60
(BMCO)
56.67
(RICO)
93.44
(PXSS)
86.89
(NBIL)
91.80
(ROCH)
Maximum
Site-Specific
Completeness
(%)
100.00
(12 sites)
98.36
(BTUT)
>100
(5 sites)
>100
(BTUT)
100.00
(7 sites)
>100
(2 sites)
                   2-30

-------
3.0
Summary of the 2011 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 2011
NMP sampling year. These data were analyzed on
a program-wide basis as well as a site-specific
basis.
                                           Results from the program-wide data
                                           analyses are presented in Section 4
                                           and results from the site-specific data
                                           analyses are presented in the
                                           individual state sections, Sections 5
                                           through 28.
       A total of 218,948 valid air toxics concentrations (including non-detects, duplicate
analyses, replicate analyses, and analyses for collocated samples) were produced from 8,835
valid samples collected at 51 monitoring sites during the 2011 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 here.
                    Table 3-1. Overview and Layout of Data Presented
Pollutant Group
VOCs
SNMOCs
Carbonyl Compounds
PAHs
Metals
Hexavalent Chromium
Number of Sites
24
8
31
23
15
22
Appendix
Raw Data
C
D
E
F
G
H
Statistical Summary
J
K
L
M
N
O
3.1    Approach to Data Treatment
       This section examines the various statistical tools employed to characterize the data
collected during the 2011 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.
                                           5-1

-------
       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 thepreprocessed daily
measurement.

       Concentrations of m,/>-xylene and o-xylene were summed together and are henceforth
referred to as "total xylenes," "xylenes (total)," or simply "xylenes" throughout the remainder of
this report, with a few exceptions. One exception is Section 4.1, which examines the results of
basic statistical calculations performed on the dataset. Table 4-1 and Table 4-2,  which are the
method-specific statistics for VOCs and SNMOCs, respectively, present the xylenes results
retained as m,/>-xylene and o-xylene species. This is also true of the Data Quality
section (Section 29).

       For the 2011 NMP, where statistical parameters are calculated based on  the preprocessed
daily measurements, zeros have been substituted for non-detect results. In past reports, the
substitution of zeros was applied  only to risk-related analyses; however, in the 2010 and 2011
NMP reports, the substitution of zeros was applied to all analyses.  This approach is consistent
with how data are loaded into AQS per the NATTS TAD (EPA, 2009b) as well  as other EPA air
toxics monitoring programs such as the School Air Toxics Monitoring Program (SATMP)
(EPA, 201 la) and the NATTS Network Assessment (EPA, 2012f). The substitution of zeros for
non-detects results in lower average concentrations of pollutants that are rarely measured at or
above the associated method detection limit and/or have a relatively  high MDL.

       In order to compare concentrations across multiple sampling methods, all concentrations
have been converted to a common unit of measure: microgram per cubic meter (ug/m3).
However, whenever a particular sampling method is isolated from  others,  such as in Tables 4-1
through 4-6, the statistical parameters are presented in the units of measure associated with the
particular sampling method. As such, 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.
                                           5-2

-------
       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 concentrations from January, February, and March; the
second quarter includes April, May, and June; the third quarter includes July, August, and
September; and the fourth quarter includes October, November, and December. A site must have
a minimum of 75 percent valid samples of the total number of samples possible within a given
quarter to have a quarterly average. 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 (2011). Annual average concentrations were calculated for monitoring
sites where three quarterly averages could be calculated and where method completeness was
greater than or equal to 85  percent. Sites that do not meet these requirements do not have an
annual average concentration presented.

       The concentration averages presented in this report are often provided with their
associated 95 percent confidence intervals. Confidence intervals represent the interval within
which the true average concentration falls 95 percent of the time. The confidence interval
includes an equal amount of quantities above and below the concentration average. For example,
an average concentration may be written as 1.25 ฑ 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. In NMP annual  reports prior to

                                           3-3

-------
2003, this subset was based on the frequency and magnitude of concentrations (previously called
"prevalent compounds"). Since the 2003 NMP annual report, health risk-based calculations have
been used to identify "pollutants of interest." For the 2011 NMP report, the pollutants of interest
are also based on risk potential. The following paragraphs provide an overview of health risk
terms and concepts and outline how the pollutants of interest are determined and then used
throughout the remainder of the report.

       EPA defines risk as "the probability that damage to life, health, or the environment will
occur as a result of a given hazard (such as exposure to a toxic chemical)" (EPA, 201 Ib). Human
health risk can be defined in terms of time. Chronic effects develop from repeated exposure over
long periods of time; acute effects develop from a single exposure or from exposures over short
periods of time (EPA, 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, 2012g).

       Health risks are typically divided into cancer and noncancer effects when referring to
human health risk. Cancer risk is defined as the likelihood of developing cancer as a result of
exposure to a given concentration over a 70-year period, and is presented as the number of
people at risk for cancer per million people. Noncancer health effects include conditions such as
asthma; noncancer health risks are presented as a hazard quotient, the value below which no
adverse health effects are expected (EPA, 201 Ib). Cancer risk is presented as a probability while
the hazard quotient is a ratio and thus, a unitless value.

       In order to assess health risk, EPA and other agencies develop toxicity factors, such as
cancer unit risk estimates (UREs) and noncancer reference concentrations (RfCs), to estimate
cancer and noncancer risks and to identify (or screen) where air toxics concentrations may
present a human health risk. EPA has published a guidance document outlining a risk-based
screening approach that utilizes a risk-based methodology for performing an initial screen of
ambient air toxics monitoring datasets (EPA, 2010a). This preliminary risk-based screening
process provides a risk-based methodology for analysts and interested parties to identify which

                                           3-4

-------
pollutants may pose a risk in their area. Cancer UREs and noncancer RfCs are used as screening
values. Not all pollutants analyzed under the NMP have screening values; of the 177 pollutants
sampled under the NMP, 81 pollutants have screening values in the guidance document. The
screening values used in this analysis are presented in Appendix Q1.

       The preprocessed daily measurements of the target pollutants were compared to these
chronic risk screening values in order to identify pollutants of interest across the program. The
following risk-based screening process was used to identify pollutants of interest:
       1.  The TO-15 and SNMOC methods have 12 pollutants in common. If a pollutant was
          measured by both the TO-15  and  SNMOC methods at the same site, the TO-15
          results were used. The purpose of this data treatment is to have one concentration per
          pollutant for each sample day.
       2.  Each preprocessed daily measurement was compared to the risk screening value.
          Concentrations that are greater than the risk  screening value are described as "failing
          the screen."
       3.  The number of failed screens was summed for each applicable pollutant.
       4.  The percent contribution of the number of failed screens to the total number of failed
          screens program-wide was  calculated for each applicable pollutant.
       5.  The pollutants contributing to the top 95 percent of the total failed screens were
          identified as pollutants of interest.

       In regards to Step 5 above, the actual cumulative contribution may exceed 95 percent in
order to include all pollutants contributing to the minimum 95 percent criteria (refer to
hexachloro-1,3-butadiene 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 values may be artificially high (or
non-existent) due to site conditions and potential cross-contamination with concurrent sampling
of carbonyl compounds using Method TO-11 A. The inclusion of acetonitrile in data analysis
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, 2012h).
                                           3-5

-------
calculations must be determined on a site-specific basis by the agency responsible for the site.
Thus, acetonitrile results are excluded from certain program-wide and site-specific data analyses,
particularly those related to risk.

       In addition to the preliminary risk-based screening process described above, the
pollutants of interest designation was further refined based on the NATTS TAD (EPA, 2009b).
This document 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).
                         Table 3-2. NATTS MQO Core Analytes
Pollutant
Acrolein
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
Acetaldehyde
Formaldehyde
Naphthalene
Benzo(a)pyrene
Arsenic
Beryllium
Cadmium
Manganese
Lead
Nickel
Hexavalent chromium
Class/Method
VOCs/TO-15
Carbonyl Compounds/
TO-11A
PAHs/TO-13A
Metals/IO-3.5
Metals/EPA

-------
       With the exception of acrolein, all of the pollutants listed in Table 3-2 are inherently
considered pollutants of interest due to their designation as NATTS MQO Core Analytes. If a
pollutant listed in Table 3-2 did not meet the pollutant of interest criteria based on the
preliminary risk-based screening approach outlined above, that pollutant was added to the list of
program-wide pollutants of interest.

       Although it is a NATTS MQO Core Analyte, acrolein was excluded from the preliminary
risk-based screening process due to questions about the consistency and reliability of the
measurements (EPA, 201 Ob). 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 (as
discussed above).

       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 EPA's national contract laboratory or may measure analytes other than
those targeted under the NMP. In these cases, data  are generated by sources other than ERG and
are not included in the preliminary risk-based screening process or any other data analysis
contained in this report.

3.3    Noncancer Risk-Based Screening Evaluation Using Minimum Risk Levels
       In addition to the preliminary risk-based screening process  described above, a second
risk-based screening was conducted using the Agency for Toxic Substances and Disease
Registry (ATSDR) Minimal Risk Level (MRL) health benchmarks (ATSDR, 2012a). An MRL is
a concentration of a hazardous substance that is "likely to be without appreciable risk of adverse
noncancer health effects over a specified duration of exposure," similar to EPA's RfCs (ATSDR,
2012b). MRLs are intended to be used as screening tools, similar to the preliminary risk-based
screening process discussed above, where "exposure to a level above the MRL does not mean
that adverse health effects will occur" (ATSDR, 2012b). ATSDR defines MRLs for three
durations of exposure: acute, intermediate, and chronic exposure. Acute risk results from
exposures of 1 to  14 days;  intermediate risk results from exposures of 15 to 364 days; and
chronic risk results from exposures of 1 year or greater (ATSDR, 2012b). MRLs, as published by

                                           3-7

-------
ATSDR, are presented in parts per million (ppm) for gases and milligrams per cubic meter
(mg/m3) for particulates. The MRLs used in this report have been converted to |ig/m3, have one
significant figure, and are presented in Appendix Q.

       For this risk-based screening evaluation, the preprocessed daily measurements were
compared to acute MRLs; quarterly averages were compared to intermediate MRLs; and annual
averages were compared to chronic MRLs. Section 4.2.2 presents the number of preprocessed
daily measurements, quarterly averages, and/or annual averages that are greater than their
respective MRL for each pollutant, summed to the program level. The number of site-specific
concentrations and/or time period averages that are greater than their respective MRLs is
expanded upon in the individual state sections.

3.4    Additional Program-Level Analyses of the 2011 National Monitoring Programs
       Dataset
       This section summarizes additional analyses performed on the 2011 NMP dataset at the
program level. Additional program-level analyses include an examination of the potential effect
of motor vehicles and a review of how concentrations vary among the  sites themselves and from
quarter-to-quarter. The results of these analyses are presented in Sections 4.3 and 4.4.

3.4.1   The Effect of Mobile Source Emissions on Spatial Variations
       Mobile source emissions from motor vehicles contribute significantly to air pollution.
"Mobile sources" refer to emitters of air pollutants that move, or can be moved, from place to
place and include both on-road and non-road emissions (EPA, 2012i). Pollutants found in motor
vehicle exhaust generally result from incomplete combustion of vehicle fuels. Although modern
vehicles and, more recently, vehicle fuels have been engineered to minimize air emissions, all
motor vehicles with internal combustion engines emit a wide range of pollutants. The magnitude
of these emissions primarily depends on the volume of traffic, while the chemical profile of these
emissions depends more on vehicle design and fuel formulation. This report uses a variety of
parameters  to quantify and evaluate the effect of motor vehicle emissions on ambient air quality,
which are discussed further in Section 4.3:
       •  Emissions data from the NEI
       •  Total  hydrocarbon concentrations

-------
       •   Motor vehicle ownership data
       •   Estimated daily traffic volume
       •   Vehicle miles traveled (VMT).

       This report uses Pearson correlation coefficients to measure the degree of correlation
between two variables, such as the ones listed above. By definition, Pearson correlation
coefficients always lie between -1 and +1. Three qualification statements apply:
       •   A correlation coefficient of-1 indicates a perfectly "negative" relationship, indicating
           that increases in the magnitude of one variable are associated with proportionate
           decreases in the magnitude of the other variable, and vice versa.
       •   A correlation coefficient of+1 indicates a perfectly "positive" relationship, indicating
           that the magnitudes of two variables both increase and both decrease proportionately.
       •   Data that are completely uncorrelated have Pearson correlation coefficients of 0.

Therefore, the sign (positive or negative) and magnitude of the Pearson correlation coefficient
indicate the direction and strength, respectively, of data correlations. In this report, correlation
coefficients greater than or equal to 0.50 and less than or equal to -0.50 are classified as strong,
while correlation coefficients less than 0.50 and greater than -0.50 are classified as weak.

       The number of observations used in a calculation is an important factor to consider when
analyzing the correlations. A correlation using relatively few observations may skew the
correlation, making the  degree of correlation appear higher (or lower) than it may actually be.
Thus, in this report, a minimum of five data points must be available to present a correlation.

3.4.2  Variability Analyses
       Variability refers to the degree of difference among values in a dataset. Three types of
variability are analyzed for this report. The first type examines the coefficient of variation (CV)
for each of the program-level pollutants of interest across the program sites.  The CV provides a
relative measure of variability by expressing the standard deviation to the magnitude of the
arithmetic mean for each of the program-level pollutants of interest, as identified in Section 4.2.
It is particularly useful when comparing different sets of data because it is unitless (Pagano, P.
and Gauvreau, K., 2000). In this report, variability across data distributions for different sites and

                                            3-9

-------
different pollutants are compared. The CVs are shown in the form of scatter plots, where data
points represent the CV and a trend line is plotted to show linearity. In addition, the "R2" value is
also shown on each scatter plot. R2 is the coefficient of determination and is an indicator of how
dependant one variable is on the other. If R2 is equal to 1.0, the data exhibit perfect linearity; the
lower R2, the less dependent the variables are each other (Pagano, P. and Gauvreau, K., 2000).
Pollutants of interest whose data points are clustered together indicate uniformity in how the
concentrations are dispersed among the sites.  This suggests that concentrations are affected by
typical and consistent sources (e.g., mobile sources). Data points that are not clustered suggest
the likelihood of a stationary source not typically found in most urban areas (e.g., coke
manufacturing facility).

       The second type of variability assessed in this report is inter-site variability and is paired
with the CV analysis in Section 4.4. The annual average concentration for each site is plotted in
the form of a bar graph for each program-wide pollutant of interest. The criteria for calculating
an annual average is discussed in Section 3.1  and sites that do not meet these requirements do
not have an annual average concentration presented. This assessment allows the reader to
visualize how concentrations varied across the sites for a particular pollutant of interest. In order
to further this analysis, the program-level average concentrations, as presented in Tables 4-1
through 4-6 in Section 4.1, are plotted against the site-specific annual averages. This allows the
reader to see how the site-specific annual averages compared to the program-level average for
each pollutant. Note that the average concentrations shown for VOCs, SNMOCs, and carbonyl
compounds in Tables 4-1 through 4-3 are presented in method-specific units, but have been
converted to a common unit of measurement (|ig/m3) for the purposes of this analysis.

       Quarterly variability is the third type of variability assessed in this report. The
concentration data for each site were divided into the four quarters of the year, as described in
Section 3.1. The completeness criteria, also described in Section 3.1, are maintained here as well.
The site-specific quarterly averages are illustrated by bar graphs for each program-level pollutant
of interest. This analysis allows for a determination of a quarterly (or seasonal) correlation with
the magnitude of concentrations for a specific pollutant.
                                           3-10

-------
3.4.3   Greenhouse Gas Assessment
       Currently, there is considerable discussion about climate change among atmospheric and
environmental scientists. Climate change refers to an extended period of change in
meteorological variables used to determine climate, such as temperature and precipitation.
Researchers are typically concerned with greenhouse gases (GHGs), which are those that cause
heat to be retained in the atmosphere (EPA, 2012J).

       Agencies researching the effects of greenhouse gases tend to concentrate primarily on
tropospheric levels of these gases. The troposphere is the lowest level of the atmosphere, whose
height varies depending on season and latitude. This is also the layer in which weather
phenomenon occur (NOAA, 2013a). 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 Global Warming Potential (GWP), and the average
concentration across the NMP program. GWP is a way to determine a pollutant's ability to retain
heat relative to carbon dioxide, which is one of the predominant anthropogenic GHGs in the
atmosphere; higher GWPs indicate a higher potential contribution to global warming (EPA,
2012J). In the future,  additional GHG pollutants may be  added to the NMP Method TO-15 target
pollutant list in order to assess their surface-level ambient concentrations.

3.5    Additional Site-Specific Analyses
       In addition to many of the analyses described in the preceding sections, the state-specific
sections contain additional analyses that are applicable only at the local level. This section
provides an overview of these analyses but does not discuss their results. Results of these site-
specific analyses are presented in the individual state-specific sections (Sections 5 through 28).

3.5.1   Site Characterization
       For each site participating in the 2011 NMP, a site characterization was performed. This
analysis 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.
                                          3-11

-------
       Regarding the plotting of emissions sources: for the 2011 (and 2010) NMP report, the
locations of point sources located near the monitoring sites were obtained from Version 2 of the
2008 NEI (EPA, 2012d). The 2008 NEI was compiled using a more streamlined approach by
EPA from state, local, and tribal agencies, as well as limited emission inventory data from other
federal programs, such as EPA's Toxics Release Inventory (TRI). By comparison, the 2008-2009
NMP report used Version 3  of the 2005 NEI, which included additional datasets not available for
the 2008 NEI. Thus, the total number of emission sources surrounding the monitoring sites is
generally lower in the 2008 NEI vs. the 2005 NEI. When comparing facility maps and emission
estimates presented in the 2011 (and 2010) NMP report to those presented in the 2008-2009
NMP report, it should be noted that the emissions inventory used in each report was for different
base years and was compiled differently.

3.5.2   Meteorological Analysis
       Several site-specific meteorological analyses were performed in order to help readers to
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 2011 and one average for sample days only. These two averages allow for the determination
of how meteorological conditions on sample days varied from typical conditions 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, 2010 and 2011). 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.
                                          3-12

-------
Although there are limitations to using NWS data, the data used is standardized and quality-
assured per NWS protocol.

       In addition to the climate summary and the statistical calculations performed on
meteorological observations collected near each monitoring site, the following sections describe
additional meteorological analyses that were performed for each monitoring site. These analyses
were performed to further characterize the meteorology at or near each monitoring site and to
determine if the meteorological conditions on days samples were collected were representative
of conditions typically experienced near each site.

3.5.2.1 Back Trajectory Analysis
       For all  sites sampling under the NMP for 2011, a back trajectory analysis was conducted.
A back trajectory traces the origin of an air parcel in relation to the location where it is currently
being measured.  The method of constructing a back trajectory uses the Lagrangian frame of
reference. In simplest terms, an air parcel can be traced back 1 hour to a new point of reference
based on the current measured wind speed and direction. At this new point of reference (that is
now 1 hour prior to the current observation), the wind speed and direction are used again to
determine where the air was 1 hour before. Back trajectory calculations are also governed by
other meteorological parameters,  such as pressure and temperature. Each time segment is
referred to as a "time step."

       Gridded meteorological data and the model used for back trajectory analyses were
prepared and developed by the National Oceanic and Atmospheric Administration (NOAA)
using data from the NWS and other cooperative agencies. The model used is the Hybrid Single-
Particle Lagrangian Integrated Trajectory (HYSPLIT) model (Draxler, R.R. and Rolph, G.D.,
1997 and 1998; Draxler, R.R., 1999). Back trajectories were computed using the HYPLIT model
to represent four times for each sample day, one at OOZ, 06Z, 12Z, and 18Z. "Z" time is "Zulu
Time" and the same time as UTC (Universal Time Coordinated) or GMT (Greenwich Mean
Time), or the local  time at the prime meridian (NOAA, 2013a). Although back trajectories can
be modeled for extended periods of time, trajectories were constructed for durations of 24 hours
to match the 24-hour sampling duration. Trajectories are modeled with an initial height of
50 meters above ground level (AGL), and  each sample day's back trajectories are plotted to

                                          3-13

-------
create a composite back trajectory map. A composite back trajectory map was constructed for
each monitoring site using Geographical Information System (GIS) software. The composite
back trajectory map can be used in the estimation of a 24-hour air shed domain for each site. An
air shed domain is the geographical area surrounding a site from which an air parcel may
typically travel within the 24-hour time frame. Information about the maximum and average
trajectory length may also be provided in reference to the composite back trajectory maps. Note
that the distances provided are straight-line distances, or the length from the site to end point, not
necessarily the length of the actual trajectory.  Agencies can use the air shed domain to evaluate
regions where long-range transport may affect their monitoring site.

       In addition to the composite back trajectory map, the HYSPLIT model was used to
perform trajectory cluster analysis. This analysis is a grouping technique that allows the model to
create a subset of trajectories or "clusters" that represent trajectories originating from similar
locations. For each monitoring site, data from each sample day's back trajectories were used as
input for the cluster analysis program. The model compares the end points between each
trajectory and calculates a  spatial variance. Trajectories that are similar to each other have lower
spatial variances while trajectories that are dissimilar have larger spatial variances. The model
then provides the user with information about total spatial variance (TSV) among the
trajectories, which allows the user to determine how many clusters best represent a given group
of trajectories (Draxler, R.R., et. al., 2009). Similar to the composite map, once the cluster
trajectories for each site were computed, a cluster map was constructed for each monitoring site
using GIS software. Both the direction and the distance from monitoring site are considered in
the clustering process. A minimum of 30 trajectories must be available for the model to run the
cluster analysis. Since four trajectories were computed for each sample day, a minimum of 30
sample days was the criteria used to perform the cluster analysis for this report. The cluster
analysis is useful  for scientifically and quantitatively determining where air most often originates
for a given location.

3.5.2.2 Wind Rose Analysis
       Wind roses were constructed for each site to help identify the predominant direction from
which the wind blows. A wind rose shows the frequency of wind directions as petals positioned
around a 16-point compass, and uses color or shading to represent wind speeds. Wind roses are

                                          3-14

-------
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 historical wind rose
for up to 10 years prior to sampling; second, 2011 data were used to construct a wind rose
presenting wind data for the entire calendar year; and lastly, a wind rose was constructed to
present wind data for sample days only. In addition to the wind roses, a map showing the
distance between the NWS station used and the monitoring site is presented. This allows for
topographical influences on the wind patterns to potentially be identified.

       A wind rose is often used in determining where to install an ambient monitoring site
when trying to capture emissions from an upwind source. A wind rose may also be useful in
determining whether high concentrations correlate with a specific wind direction. While the
composite back trajectory map shows where air parcels originated on a number of days, the wind
rose shows the frequency at which wind speed and direction are measured near the monitoring
site. Thus, the back trajectory analysis focuses on long range transport, while the wind rose
captures day-to-day fluctuations at the surface. Both are used to identify potential meteorological
influences on a monitoring site.

3.5.3   Site-Specific Comparison to Program-level Average Concentrations
       To better understand how a site's concentrations compare to the program-level
concentrations, as presented in Tables 4-1 through 4-6 of Section 4.1, the site-specific and
program-level  concentrations are presented together graphically for the selected NATTS MQO
Core Analytes listed in Table 3.3. This analysis is an extension of the analysis discussed in
Section 3.4.2 and utilizes box and whisker plots, or simply box plots,  to visually show this
comparison. These box plots were created in Microsoft Excel, using the Peltier Box and Whisker
Plot Utility (Peltier, 2012). Note that for sites that sampled SNMOCs, benzene and 1,3-butadiene
are shown only in comparison to  those sites sampling SNMOCs as opposed to sites sampling
these pollutants with Method TO-15, to match the program-level averages presented in
Tables 4-1 and 4-2 in Section 4.1.

       The box plots used in this analysis overlay the site-specific minimum, annual average,
and maximum concentrations over several program-level statistical metrics. For the program-

                                          3-15

-------
level, the first, second (median), third, and fourth (maximum) quartiles are shown as colored
segments on a "bar" where the color changes indicate the exact numerical value of the quartile.
The thin vertical line represents the program-level average concentration. The site-specific
annual average is shown as a white circle plotted on top of the bar and the horizontal lines
represent the minimum and maximum concentration measured at the site. An example of this
figure is shown in Figure 5-10. Note that the program-level average concentrations shown for
VOCs, SNMOCs, and carbonyl compounds in Tables 4-1 through 4-3 are presented in method-
specific units, but have been converted to a common unit of measurement (|ig/m3) for the
purposes of this analysis. These graphs are presented in Sections 5 through 28, and are grouped
by pollutant within each state section.  This allows for both a "site vs. program" comparison, and
an inter-site comparison for sites within a given state.
           Table 3-3. NATTS MQO Core Analytes Selected for Comparative Analysis
Pollutant
Benzene
1,3 -Butadiene
Acetaldehyde
Formaldehyde
Benzo(a)pyrene
Naphthalene
Arsenic
Manganese
Lead
Hexavalent Chromium
Class/Method
VOCs/TO-15
Carbonyl
Compounds/TO-llA
PAHs/TO-13A
MetaMO-3.5
Metals/EPA
3.5.4   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 the
selected NATTS MQO Core Analytes shown in Table 3-3. Twenty-nine of the 51 sites have
sampled at least one pollutant group long enough for the trends analysis to be conducted. The
approach to this trends analysis is described below and the results are presented in the individual
state sections (Sections 5 through 28).
                                         3-16

-------
       The trends figures and analyses are presented as 1-year statistical metrics, which is a
change from previous reports (where 3-year rolling average concentrations were calculated).  The
following criteria were used to calculate valid statistical metrics:
       •  Analysis must have been performed under the NMP.
       •  There must be a minimum of at least 5 years of consecutive data.
       •  There must be a minimum of 85 percent completeness for each sampling year, 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).

       Five individual statistical metrics were used in this analysis and are presented as box and
whisker plots, an example of which can be seen in Figure 5-18. 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, 50thpercentile (or
median), and 95th percentile (as shown by  the y-values corresponding with the bottom, gray line,
or top of the box, respectively); and the average concentration (as denoted by the orange
diamond). Each of the five metrics represents all measurements from that 1-year period.

       Data used in this analysis were downloaded from EPA's AQS database (EPA, 2012c),
where non-detects are uploaded into AQS  as zeros (EPA, 2009b). Similar to other analyses
presented in this report, zeros representing these non-detects were incorporated into the
statistical calculations. The results from sample days  with precision data (duplicates, collocates,
and/or replicates) were averaged together to allow for the determination of a single concentration
per pollutant for each site, reflecting the data treatment described in Section 3.1.

3.5.5   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 risk-
related analyses at the site-specific level, as described below:
       •  Comparison to ATSDR MRLs, as described in Section 3.3, including the emission
          tracer analysis described below
                                          3-17

-------
       •  The calculation of cancer risk and noncancer hazard approximations in relation to
          cancer and noncancer health effects
       •  Risk-based emissions assessment.

3.5.5.1 Emission Tracer Analysis
       The preprocessed daily measurements and time-period average concentrations for each
site-specific pollutant of interest were compared to the ATSDR MRL noncancer health
benchmarks in the same fashion described in Section 3.3.  To further this analysis, pollution roses
were created for each of the site-specific pollutants of interest that have preprocessed daily
measurements greater than their respective ATSDR acute MRL health benchmark (where
applicable).  This analysis is performed to help identify the geographical area where the
emissions sources of these pollutants may have  originated. A pollution rose is a plot of the
ambient concentration versus the wind speed and direction; high concentrations may be shown in
relation to the direction of potential emissions sources.

3.5.5.2 Cancer Risk and Noncancer Hazard Approximations
       Risk was further examined by calculating cancer risk and noncancer hazard
approximations for each of the site-specific pollutants of interest. The cancer risk
approximations presented in this report estimate the cancer risk due to exposure at the annual
average concentration over a 70-year period (not the risk resulting from exposure over the time
period covered in this report). A cancer risk approximation less than 1 in-a-million is considered
negligible; a cancer risk greater than 1 in-a-million but less than 100 in-a-million is generally
considered acceptable; and a cancer risk greater than  100  in-a-million is considered significant
(EPA, 2009c). The noncancer hazard approximation is presented as the Noncancer Hazard
Quotient (HQ), which is a unitless value. According to EPA, "If the HQ is calculated to be equal
to or less than 1.0, then no adverse health effects are expected as a result of exposure. If the HQ
is greater than 1.0, then adverse health effects are possible" (EPA, 201 Ib).

       The toxicity factors applied  to calculate  the cancer risk and noncancer hazard
approximations are typically UREs  (for cancer) or RfCs (for noncancer), which are developed by
EPA. However, UREs and RfCs are not available for all pollutants. In the absence of EPA
values, toxicity factors developed by agencies with credible methods and that are similar in
                                          3-18

-------
scope and definition were used (EPA, 2012h). Cancer URE and noncancer RfC toxicity factors
can be applied to the annual averages to approximate risk based on ambient monitoring data.
While the cancer risk and noncancer hazard approximations do not incorporate human activity
patterns and therefore do not reflect true human inhalation exposure, they may allow analysts to
further refine their focus by identifying concentrations of specific pollutants that may present
health risks. Cancer UREs and/or noncancer RfCs, site-specific annual averages, and
corresponding annual average-based cancer risk and noncancer hazard approximations are
presented in each state section (Sections 5 through 28).

3.5.5.3 Risk-Based Emissions Assessment
       A pollutant emitted in high quantities does not necessarily present a higher risk to human
health than a pollutant emitted in very low quantities. The more toxic the pollutant, the more risk
associated with its emissions in ambient air. The development of various health-based toxicity
factors has allowed analysts to apply weight to the emissions  of pollutants based on toxicity
rather than mass emissions. This approach considers both a pollutant's toxicity potential and the
quantity emitted.

       This assessment  compares county-level emissions to toxicity-weighted emissions based
on the EPA-approved approach described below (EPA, 2007). The 10 pollutants with the highest
total mass emissions and the 10 pollutants with the highest associated  toxicity-weighted
emissions for pollutants  with cancer and noncancer toxicity factors are presented in each state
section. While the absolute magnitude of the pollutant-specific toxicity-weighted emissions is
not meaningful, the relative magnitude  of toxicity-weighted emissions is useful in identifying the
order of potential priority for air quality managers. Higher values suggest greater priority;
however, even the highest values may not reflect potential cancer effects greater than the level of
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.
                                           3-19

-------
       The toxicity-weighted emissions approach consists of the following steps:
       1.  Obtain HAP emissions data for all anthropogenic sectors from the NEI. For point
          sources, sum the process-level emissions to the county-level.
       2.  Apply the mass extraction speciation profiles to extract metal and cyanide mass. The
          only exception is for two chromium species: chromium and chromium compounds.
       3.  For chromium and chromium compounds, trivalent chromium (non-toxic) must be
          separated from hexavalent chromium (toxic). To do this, apply the chromium
          speciation profile to extract the hexavalent chromium mass by industry group.
       4.  Apply weight to the emissions derived from the steps above based on their toxicity.
          The results of the toxicity-weighting process are unitless.
          a.  To apply weight based on cancer toxicity, multiply the emissions of each
             pollutant by its cancer URE.
          b.  To apply weight based on noncancer toxicity, divide the emissions of each
             pollutant by its noncancer RfC.

       The PAHs measured using Method TO-13A are a sub-group of Poly cyclic Organic
Matter (POM). Because these compounds are often not speciated into individual compounds in
the NEI, the PAHs are grouped into POM Groups in order to assess risk attributable to these
pollutants (EPA, 201 Ic).  Thus, emissions data and toxicity-weighted emissions for PAHs are
presented by POM Groups for this analysis. Table 3-4 presents the 22 PAHs measured by
Method TO-13A and their associated POM Groups. The POM groups are sub-grouped in
Table 3-4 because toxicity research has led to the refining of UREs for certain PAHs (EPA,
2012h). Note that naphthalene emissions are reported to the NEI individually; therefore,
naphthalene is not included in one of the POM Groups. Also note that four pollutants analyzed
by Method TO-13A and listed in Table 3-4 do not have assigned POM Groups.
                                         3-20

-------
         Table 3-4. POM Groups for PAHs
Pollutant
Acenaphthene
Acenaphthylene
Anthracene
Benzo(a)anthracene
Benzo(a)pyrene
Benzo(b)fluoranthene
Benzo(e)pyrene
Benzo(g,h,i)perylene
Benzo(k)fluoranthene
Chrysene
Coronene
Cyclopenta[cd]pyrene
Dibenz(a,h)anthracene
Fluoranthene
Fluorene
9-Fluorenone
Indeno( 1 ,2,3 -cd)pyrene
Naphthalene*
Perylene
Phenanthrene
Pyrene
Retene
POM Group
Group 2
Group 2
Group 2
POM
Subgroup
Group 2b
Group 2b
Group 2d
Group 6
Group 5
GroupSa
Group 6
Group 2
Group 2
Group 2b
Group 2b
Group 6
Group 7
NA
NA
Group 5
Group 2
Group 2
GroupSb
Group 2b
Group 2b
NA
Group 6
NA
Group 2
Group 2
Group 2
Group 2b
Group 2d
Group 2d
NA
* Naphthalene emissions are reported to the NEI individually;
therefore, naphthalene is not included in one of the POM Groups.
NA = no POM Group assigned.
                         3-21

-------
4.0    Summary of the 2011 National Monitoring Programs Data
       This section summarizes the results of the data analyses performed on the NMP dataset,
as described in Section 3.

4.1    Statistical Results
       This section examines the following statistical parameters for the target pollutants of each
analytical method:  1) detection rates, 2) concentration ranges and data distribution, and 3) central
tendency statistics. Tables 4-1 through 4-6 present statistical summaries for the target pollutants
and Sections 4.1.1 through 4.1.3 review the basic findings of these statistical calculations.

4.1.1   Target Pollutant Detection Rates
       There is an experimentally determined MDL for every target pollutant, as described in
Section 2.2. Quantification below the MDL is possible, although the measurement's reliability is
lower. If a concentration does not exceed the MDL, it does not mean that the pollutant is not
present in the air. If the instrument does not generate a numerical concentration, the
measurement is marked as "ND," or "non-detect." As explained in Section 2.2, data analysts
should exercise caution when interpreting monitoring data with a high percentage of reported
concentrations at levels near or below the corresponding MDLs. A thorough review of the
number of measured detections, the number of non-detects, and the total number of samples is
beneficial to understanding the representativeness of the interpretations made.

       Tables 4-1 through 4-6 summarize the number of times the target pollutants were
detected out of the number of valid samples collected and analyzed. Approximately 54 percent of
the reported measurements (based on the preprocessed daily measurements) were above the
MDLs across the program. The following list provides the percentage of measurements that were
above the MDLs for each method:
       •   40.7 percent for VOCs
       •   51.1 percent for SNMOCs
       •   81.5 percent for carbonyl compounds
       •   63.0 percent for PAHs
                                           4-1

-------
       •  82.6 percent for metals
       •  79.4 percent for hexavalent chromium samples.
       Some pollutants were always detected while others were infrequently detected or not
detected at all. Similar to previous years' reports, acetaldehyde, formaldehyde, and acetone had
the greatest number of measured detections (1,594), using the preprocessed daily measurements.
These pollutants were reported in every valid carbonyl compound sample collected (1,594).
Acetylene, benzene, chloromethane, dichlorodifluoromethane, propylene, toluene,
trichlorofluoromethane, and trichlorotrifluoroethane were detected in every valid VOC sample
collected (1,281). Eight pollutants, including acetylene, ethylene, ethane, and toluene, were
detected in every valid SNMOC sample  collected (441). Naphthalene, phenanthrene,
fluoranthene, and pyrene were detected in every valid PAH sample collected (1,382). Cadmium,
lead, manganese, and nickel were detected in every valid metal sample collected (834).
Hexavalent chromium was detected in 1,068 samples (out of 1,328 samples).

       Although BTUT and NBIL have the greatest number of measured detections (7,321 for
BTUT and 6,995 for NBIL), they were also the only two sites that collected samples for all six
analytical methods/pollutant groups. However, the detection rates for these sites (66 and
68 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 99 percent. Conversely,
VOCs had the lowest percentage of concentrations greater than the MDLs (40.7 percent). A site
measuring only VOCs would be expected to have lower detection rates, such as SPAZ
(48.3 percent).
                                          4-2

-------
                                           Table 4-1. Statistical Summaries of the VOC Concentrations
Pollutant
Acetonitrile
Acetylene
Acrolein
Acrylonitrile3
tert-Amyl Methyl Ether
Benzene
Bromochloromethane
Bromodichloromethane
Bromoform
Bromomethane
1,3 -Butadiene
Carbon Bisulfide
Carbon Tetrachloride
Chlorobenzene
Chloroethane
Chloroform
Chloromethane
Chloromethylbenzene
Chloroprene
Dibromochloromethane
1 ,2-Dibromoethane
m -Dichlorobenzene
o-Dichlorobenzene
^-Dichlorobenzene
Dichlorodifluoromethane
1 , 1 -Dichloroethane
#of
Measured
Detections1
1,275
1,281
1,254
300
33
1,281
o
6
61
69
737
966
1,219
1,279
20
203
756
1,281
9
1
299
31
88
108
736
1,281
5
#
of Non-
Detects1
6
0
27
924
1,248
0
1,278
1,220
1,212
544
315
62
2
1,261
1,078
525
0
1,272
1,280
982
1,250
1,193
1,173
545
0
1,276
Minimum2
(ppbv)
0.045
0.082
0.078
0.013
0.004
0.053
0.009
0.012
0.004
0.007
0.007
0.004
0.006
0.007
0.006
0.003
0.133
0.008
Maximum
(ppbv)
1,030
101
11.0
0.778
0.010
7.42
0.049
3.05
0.062
0.437
4.29
22.8
0.260
0.084
8.18
6.72
2.53
0.124
0.016
0.002
0.005
0.004
0.002
0.003
0.114
0.008
1.13
0.022
0.329
0.083
0.469
1.23
0.021
Arithmetic
Mean
(ppbv)
14.7
0.935
0.489
0.039
0.000
0.305
0.0001
0.021
0.001
0.010
0.043
0.925
0.096
0.000
0.011
0.079
0.611
0.000
Median
(ppbv)
0.392
0.598
0.370
0
0
0.218
0
0
0
0.010
0.028
0.033
0.098
0
0
0.018
0.600
0
Mode
(ppbv)
0.206
1.09
0
0
0
0.149
0
0
0
0
0
0
0.098
0
0
0
0.610
0
First
Quartile
(ppbv)
0.200
0.387
0.215
0
0
0.158
0
0
0
0
0.008
0.013
0.085
0
0
0
0.543
0
Third
Quartile
(ppbv)
1.91
0.970
0.597
0
0
0.325
0
0
0
0.014
0.050
0.425
0.109
0
0
0.030
0.663
0
Standard
Deviation
(ppbv)
67.0
3.31
0.526
0.103
0.001
0.359
0.001
0.163
0.004
0.019
0.133
2.60
0.022
0.003
0.229
0.427
0.114
0.005
Coefficient
of
Variation
4.55
3.54
1.08
2.62
6.36
1.18
26.4
7.90
5.34
1.99
3.07
2.81
0.234
10.4
21.0
5.38
0.186
18.6
NA
0.008
0.000
0.001
0.001
0.013
0.554
0.0001
0
0
0
0
0.008
0.548
0
0
0
0
0
0
0.544
0
0
0
0
0
0
0.526
0
0
0
0
0
0.017
0.578
0
0.055
0.002
0.012
0.003
0.024
0.059
0.001
6.52
6.82
9.12
4.64
1.86
0.106
17.0
2 Excludes zeros for non-detects
3 Because S4MO invalidated all acrylonitrile data for 2011, the number of measured detections plus the number of non-detects does not equal the total number of VOC samples
collected (1,281).
NA = Not applicable for these parameters

-------
                                    Table 4-1. Statistical Summaries of the VOC Concentrations (Continued)
Pollutant
1 ,2-Dichloroethane
1 , 1 -Dichloroethene
c/'s-l,2-Dichloroethylene
trans- 1 ,2-Dichloroethylene
Dichloromethane
1,2-Dichloropropane
cis- 1 , 3 -Dichloropropene
trans- 1 ,3 -Dichloropropene
Dichlorotetrafluoroethane
Ethyl Aery late
Ethyl fert-Butyl Ether
Ethylbenzene
Hexachloro- 1 ,3 -butadiene
Methyl Ethyl Ketone
Methyl Isobutyl Ketone
Methyl Methacrylate
Methyl fer/-Butyl Ether
w-Octane
Propylene
Styrene
1 , 1 ,2,2-Tetrachloroethane
Tetrachloroethylene
Toluene
1,2,4-Trichlorobenzene
1,1,1 -Trichloroethane
1, 1,2-Trichloroethane
#of
Measured
Detections1
320
22
0
14
1,264
1
11
6
1,241
4
8
1,277
113
1,280
1,202
54
37
1,241
1,281
1,219
55
958
1,281
48
1,096
8
#
of Non-
Detects1
961
1,259
1,281
1,267
17
1,280
1,270
1,275
40
1,277
1,273
4
1,168
1
79
1,227
1,244
40
0
62
1,226
323
0
1,233
185
1,273
Minimum2
(ppbv)
0.011
0.005
Maximum
(ppbv)
0.174
0.018
Arithmetic
Mean
(ppbv)
0.006
0.0001
Median
(ppbv)
0
0
Mode
(ppbv)
0
0
First
Quartile
(ppbv)
0
0
Third
Quartile
(ppbv)
0
0
Standard
Deviation
(ppbv)
0.012
0.001
Coefficient
of
Variation
2.04
8.14
Not Detected
0.006
0.040
0.050
349
0.012
0.007
0.011
0.004
0.020
0.005
0.004
0.003
0.059
0.006
0.009
0.005
0.008
0.057
0.006
0.004
0.004
0.023
0.004
0.002
0.008
0.088
0.057
0.078
0.078
0.016
2.65
0.064
34.3
2.43
0.34
0.274
2.62
34.1
3.06
0.048
1.27
14.9
0.083
0.081
0.017
0.0002
1.01
0
0.128
0
0
0
0.087
0
0.230
0.002
12.6
11.8
12.5
NA
0.0002
0.0001
0.019
0.0001
0.0001
0.093
0.001
0.541
0.043
0.002
0.001
0.061
0.704
0.059
0.0005
0.023
0.586
0.001
0.010
0.000
0
0
0.018
0
0
0.064
0
0.430
0.034
0
0
0.041
0.374
0.034
0
0.014
0.339
0
0.010
0
0
0
0.017
0
0
0.038
0
0.474
0
0
0
0
0.348
0
0
0
0.288
0
0
0
0
0
0.016
0
0
0.041
0
0.285
0.022
0
0
0.027
0.265
0.023
0
0
0.187
0
0.008
0
0
0
0.020
0
0
0.101
0
0.643
0.052
0
0
0.065
0.558
0.049
0
0.027
0.645
0
0.012
0
0.003
0.002
0.008
0.003
0.001
0.123
0.004
1.00
0.074
0.014
0.012
0.100
1.85
0.156
0.003
0.048
0.959
0.003
0.006
0.001
14.6
17.9
0.440
20.7
13.2
1.32
4.12
1.85
1.72
7.34
11.4
1.64
2.63
2.65
5.50
2.06
1.64
6.82
0.570
13.0
2 Excludes zeros for non-detects
3 Because S4MO invalidated all acrylonitrile data for 2011, the number of measured detections plus the number of non-detects does not equal the total number of VOC samples
collected (1,281).
NA = Not applicable for these parameters

-------
Table 4-1. Statistical Summaries of the VOC Concentrations (Continued)
Pollutant
Trichloroethylene
Trichlorofluoromethane
Trichlorotrifluoroethane
1 ,2,4-Trimethylbenzene
1 , 3 , 5-Trimethylbenzene
Vinyl chloride
m,p-Xylene
o-Xylene
#of
Measured
Detections1
251
1,281
1,281
1,275
1,226
82
1,273
1,267
#
of Non-
Detects1
1,030
0
0
6
55
1,199
8
14
Minimum2
(ppbv)
0.004
0.051
0.023
0.005
0.004
0.003
0.009
0.005
Maximum
(ppbv)
1.56
1.41
0.180
1.05
0.386
0.033
6.08
1.99
Arithmetic
Mean
(ppbv)
0.009
0.280
0.097
0.088
0.032
0.001
0.243
0.095
Median
(ppbv)
0
0.270
0.096
0.061
0.024
0
0.148
0.062
Mode
(ppbv)
0
0.262
0.096
0.033
0
0
0.033
0.023
First
Quartile
(ppbv)
0
0.255
0.089
0.039
0.016
0
0.078
0.036
Third
Quartile
(ppbv)
0
0.290
0.104
0.103
0.037
0
0.267
0.105
Standard
Deviation
(ppbv)
0.067
0.061
0.012
0.095
0.031
0.002
0.386
0.127
Coefficient
of
Variation
7.57
0.219
0.129
1.07
0.937
6.21
1.11
0.997
1 Out of 1,281 valid samples
2 Excludes zeros for non-detects
3 Because S4MO invalidated all acrylonitrile data for 20 1 1,
collected (1,281).
NA = Not applicable for these parameters
             the number of measured detections plus the number of non-detects does not equal the total number of VOC samples

-------
                                      Table 4-2. Statistical Summaries of the SNMOC Concentrations
Pollutant
Acetylene
Benzene
1,3 -Butadiene
w-Butane
c/s-2-Butene
/raซs-2-Butene
Cyclohexane
Cyclopentane
Cyclopentene
w-Decane
1-Decene
Tw-Diethylbenzene
ฃ>-Diethylbenzene
2,2-Dimethylbutane
2,3 -Dimethylbutane
2,3 -Dimethylpentane
2,4-Dimethylpentane
w-Dodecane
1-Dodecene
Ethane
2-Ethyl-l-butene
Ethylbenzene
Ethylene
/w-Ethyltoluene
o-Ethyltoluene
ฃ>-Ethyltoluene
w-Heptane
#of
Measured
Detections1
441
436
151
361
325
337
425
414
117
405
5
401
357
424
351
423
423
427
355
441
2
433
441
403
319
403
435
#
of Non-
Detects1
0
5
290
80
116
104
16
27
324
36
436
40
84
17
90
18
18
14
86
0
439
8
0
38
122
38
6
Minimum2
(ppbC)
0.271
0.241
0.041
0.339
0.048
0.066
0.101
0.076
0.077
0.073
0.096
0.075
0.077
0.088
0.101
0.107
0.080
0.073
0.071
1.59
0.241
0.086
0.717
0.082
0.086
0.062
0.079
Maximum
(ppbC)
237
12.9
4.24
81.9
2.36
5.54
16.9
21.0
8.21
9.12
4.27
2.38
1.82
2.93
3.86
4.07
2.28
6.60
8.57
1,390
0.557
4.05
674
2.68
2.71
1.99
11.9
Arithmetic
Mean
(ppbC)
2.25
1.65
0.120
11.7
0.197
0.276
1.92
0.615
0.134
0.571
0.013
0.404
0.269
0.555
0.730
0.611
0.406
0.519
0.302
54.9
0.002
0.531
4.81
0.328
0.222
0.246
1.65
Median
(ppbC)
1.12
1.13
0
5.70
0.146
0.177
0.954
0.402
0
0.383
0
0.350
0.214
0.421
0.455
0.478
0.324
0.437
0.201
26.2
0
0.341
2.09
0.266
0.182
0.214
0.924
Mode
(ppbC)
1.10
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
100
0
0
1.83
0
0
0
0
First
Quartile
(ppbC)
0.733
0.704
0
1.53
0
0.092
0.240
0.241
0
0.170
0
0.239
0.118
0.262
0.165
0.303
0.197
0.322
0.112
5.89
0
0.232
1.57
0.162
0
0.151
0.261
Third
Quartile
(ppbC)
1.70
2.25
0.184
16.0
0.230
0.297
2.81
0.723
0.086
0.668
0
0.522
0.356
0.709
1.04
0.803
0.539
0.628
0.363
69.4
0
0.636
3
0.435
0.320
0.310
2.44
Standard
Deviation
(ppbC)
12.4
1.42
0.285
15.1
0.268
0.423
2.41
1.14
0.488
0.801
0.207
0.290
0.272
0.453
0.802
0.482
0.311
0.464
0.511
97.1
0.029
0.509
35.3
0.275
0.272
0.187
1.88
Coefficient
of
Variation
5.51
0.860
2.38
.29
.36
.53
.25
.85
3.64
1.40
16.0
0.717
1.01
0.815
1.10
0.788
0.765
0.895
1.69
1.77
15.9
0.959
7.33
0.841
1.23
0.762
1.14
1 Out of 441 valid samples
2 Excludes zeros for non-detects
NA = Not applicable for these parameters

-------
                                Table 4-2. Statistical Summaries of the SNMOC Concentrations (Continued)
Pollutant
1-Heptene
w-Hexane
1-Hexene
c/s-2-Hexene
trans-2-Rexene
Isobutane
Isobutene/ 1 -Butene
Isopentane
Isoprene
Isopropylbenzene
2-Methyl-l-butene
3 -Methyl- 1 -butene
2-Methy 1- 1 -pentene
4-Methy 1- 1 -pentene
2-Methyl-2 -butene
Methylcyclohexane
Methylcyclopentane
2-Methy Iheptane
3-Methylheptane
2-Methy Ihexane
3 -Methy Ihexane
2-Methy Ipentane
3-Methylpentane
w-Nonane
1-Nonene
w-Octane
1-Octene
#of
Measured
Detections1
308
439
357
54
25
440
181
418
370
273
249
6
12
14
287
418
438
349
357
427
411
441
441
403
135
423
164
#
of Non-
Detects1
133
2
84
387
416
1
260
23
71
168
192
435
429
427
154
23
3
92
84
14
30
0
0
38
306
18
277
Minimum2
(ppbC)
0.050
0.137
0.086
0.076
0.064
0.208
0.107
0.729
0.076
0.081
0.096
0.161
0.062
0.116
0.087
0.101
0.104
0.059
0.054
0.168
0.135
0.243
0.099
0.066
0.054
0.090
0.057
Maximum
(ppbC)
3.87
67.9
0.566
0.404
0.543
129
13.5
110
7.20
0.400
2.84
2.25
0.382
0.358
7.81
33.1
21.5
3.15
2.43
6.76
6.66
28.3
22.1
11.0
3.60
8.94
0.727
Arithmetic
Mean
(ppbC)
0.436
3.48
0.190
0.020
0.010
11.8
0.889
14.0
0.406
0.094
0.220
0.010
0.005
0.006
0.262
3.58
1.85
0.482
0.386
1.31
1.19
3.46
1.98
0.668
0.090
1.15
0.078
Median
(ppbC)
0.202
1.72
0.211
0
0
5.73
0
10.2
0.195
0.116
0.163
0
0
0
0.191
1.44
1.04
0.290
0.251
0.930
0.839
2.24
1.19
0.401
0
0.643
0
Mode
(ppbC)
0
1.01
0
0
0
1.14
0
0
0
0
0
0
0
0
0
0
2.76
0
0
0
0
1.16
2.49
0
0
0
0
First
Quartile
(ppbC)
0
0.599
0.137
0
0
1.50
0
4.36
0.127
0
0
0
0
0
0
0.266
0.384
0.096
0.108
0.549
0.390
0.967
0.529
0.171
0
0.241
0
Third
Quartile
(ppbC)
0.671
4.29
0.263
0
0
15.7
1.52
18.9
0.362
0.155
0.312
0
0
0
0.326
5.42
2.47
0.761
0.576
1.82
1.71
4.48
2.58
0.827
0.139
1.60
0.143
Standard
Deviation
(ppbC)
0.562
5.44
0.115
0.061
0.051
15.5
1.57
14.4
0.671
0.081
0.318
0.121
0.033
0.035
0.505
4.76
2.33
0.538
0.406
1.09
1.11
3.70
2.34
0.938
0.248
1.30
0.121
Coefficient
of
Variation
1.29
1.56
0.604
3.01
5.02
1.31
1.77
1.03
1.65
0.863
1.45
11.7
7.02
5.92
1.93
1.33
1.26
1.11
1.05
0.834
0.933
1.07
1.18
1.40
2.75
1.13
1.56
1 Out of 441 valid samples
2 Excludes zeros for non-detects
NA = Not applicable for these parameters

-------
                                Table 4-2. Statistical Summaries of the SNMOC Concentrations (Continued)
Pollutant
w-Pentane
1-Pentene
c/s-2-Pentene
/raซs-2-Pentene
a-Pinene
6-Pinene
Propane
w-Propylbenzene
Propylene
Propyne
Styrene
Toluene
w-Tridecane
1-Tridecene
1 ,2,3 -Trimethylbenzene
1 ,2,4-Trimethylbenzene
1 ,3 ,5-Trimethylbenzene
2,2,3 -Trimethylpentane
2,2,4-Trimethylpentane
2,3 ,4-Trimethylpentane
w-Undecane
1-Undecene
7w-Xylene/ฃ>-Xylene
o-Xylene
SNMOC (Sum of Knowns)
Sum of Unknowns
TNMOC
#of
Measured
Detections1
441
422
266
363
122
46
441
347
438
9
83
441
41
14
359
430
254
108
332
383
401
57
246
426
441
441
441
#
of Non-
Detects1
0
19
175
78
319
395
0
94
o
6
432
358
0
400
427
82
11
187
333
109
58
40
384
195
15
NA
NA
NA
Minimum2
(ppbC)
0.160
0.075
0.062
0.064
0.071
0.094
1.41
0.075
0.294
0.101
0.052
0.248
0.061
0.073
0.074
0.108
0.059
0.074
0.084
0.056
0.078
0.053
0.076
0.097
15.9
11.2
30.1
Maximum
(ppbC)
199
17.2
121
2.63
1.70
2.00
376
1.24
46.4
0.225
5.06
174
2.45
0.377
1.33
10.2
2.01
0.707
4.26
1.48
13.1
2.72
13.7
4.97
2,740
5,510
5,660
Arithmetic
Mean
(ppbC)
7.88
0.425
0.735
0.209
0.106
0.067
34.5
0.155
1.19
0.003
0.104
4.60
0.024
0.005
0.166
0.557
0.210
0.062
0.404
0.212
0.429
0.055
1.13
0.539
192
88.6
280
Median
(ppbC)
3.76
0.233
0.103
0.158
0
0
18.0
0.148
0.836
0
0
2.29
0
0
0.155
0.417
0.134
0
0.235
0.174
0.287
0
0.360
0.396
116
47.0
195
Mode
(ppbC)
1.94
0
0
0
0
0
126
0
1.03
0
0
2.72
0
0
0
0
0
0
0
0
0
0
0
0
250
29.5
245
First
Quartile
(ppbC)
1.5
0.179
0
0.107
0
0
6.15
0.103
0.640
0
0
1.14
0
0
0.108
0.278
0
0
0.086
0.120
0.150
0
0
0.231
49.1
28.5
101
Third
Quartile
(ppbC)
8.85
0.339
0.154
0.253
0.142
0
43.6
0.207
1.16
0
0
4.24
0
0
0.215
0.661
0.324
0
0.503
0.249
0.493
0
1.75
0.649
250
76.4
328
Standard
Deviation
(ppbC)
14.1
1.12
7.95
0.255
0.220
0.237
43.2
0.133
2.51
0.024
0.424
14.2
0.141
0.028
0.140
0.653
0.267
0.124
0.549
0.192
0.765
0.258
1.78
0.555
234
281
366
Coefficient
of
Variation
1.79
2.62
10.8
1.22
2.08
3.55
1.25
0.863
2.11
7.18
4.07
3.08
5.85
6.24
0.844
1.17
1.27
2.00
1.36
0.909
1.78
4.70
1.57
1.03
1.22
3.18
1.31
1 Out of 441 valid samples
2 Excludes zeros for non-detects
NA = Not applicable for these parameters

-------
                               Table 4-3. Statistical Summaries of the Carbonyl Compound Concentrations
Pollutant
Acetaldehyde
Acetone
Benzaldehyde
Butyraldehyde
Crotonaldehyde
2,5-Dimethylbenzaldehyde
Formaldehyde
Hexaldehyde
Isovaleraldehyde
Propionaldehyde
Tolualdehydes
Valeraldehyde
#of
Measured
Detections1
1,594
1,594
1,572
1,591
1,570
0
1,594
1,562
3
1,588
1,364
1,563
#
of Non-
Detects1
0
0
22
3
24
1,594
0
32
1,591
6
230
31
Minimum2
(ppbv)
0.023
0.031
0.004
0.008
0.005
Maximum
(ppbv)
8.05
6.48
0.359
0.924
2.19
Arithmetic
Mean
(ppbv)
1.11
1.08
0.032
0.103
0.108
Median
(ppbv)
0.928
0.881
0.024
0.084
0.040
Mode
(ppbv)
1.07
1.21
0.017
0.065
0.016
First
Quartile
(ppbv)
0.636
0.580
0.016
0.058
0.022
Third
Quartile
(ppbv)
1.38
1.37
0.035
0.121
0.123
Standard
Deviation
(ppbv)
0.719
0.771
0.030
0.079
0.163
Coefficient
of
Variation
0.647
0.716
0.956
0.770
1.51
Not Detected
0.026
0.004
0.014
0.008
0.005
0.005
22.5
1.29
0.028
0.604
0.256
0.550
2.33
0.039
0.00004
0.132
0.030
0.030
1.82
0.024
0
0.110
0.024
0.022
1.79
0.012
0
0.081
0
0.011
1.22
0.015
0
0.078
0.014
0.014
2.76
0.039
0
0.165
0.038
0.035
1.89
0.086
0.001
0.084
0.029
0.041
0.813
2.20
23.9
0.633
0.947
1.35
2 Excludes zeros for non-detects

-------
                                      Table 4-4. Statistical Summaries of the PAH Concentrations
Pollutant
Acenaphthene
Acenaphthylene
Anthracene
Benzo(a)anthracene
Benzo(a)pyrene
Benzo(b)fluoranthene
Benzo(e)pyrene
Benzo(g,h,i)perylene
Benzo(k)fluoranthene
Chrysene
Coronene
Cyclopenta[cd]pyrene
Dibenz(a,h)anthracene
Fluoranthene
Fluorene
9-Fluorenone
Indeno( 1 ,2,3 -cd)pyrene
Naphthalene
Perylene
Phenanthrene
Pyrene
Retene
#of
Measured
Detections1
1,366
634
1,015
968
844
1,275
1,131
1,225
817
1,321
852
250
287
1,382
1,376
1,379
1,074
1,382
423
1,382
1,382
1,363
#
of Non-
Detects1
16
748
367
414
538
107
251
157
565
61
530
1,132
1,095
0
6
3
308
0
959
0
0
19
Minimum2
(ng/m3)
0.0793
0.0428
0.0229
0.015
0.0175
0.0194
0.0147
0.0115
0.00906
0.0204
0.0127
0.0147
0.0158
0.0925
0.365
0.101
0.0175
2.27
0.00889
0.387
0.0408
0.0204
Maximum
(ng/m3)
111
17.6
17.1
2.4
1.99
2.80
1.87
1.55
0.861
3.04
0.705
1.13
0.276
39.5
70.5
15.4
1.34
779
0.398
178
19.0
10.6
Arithmetic
Mean
(ng/m3)
4.64
0.666
0.369
0.088
0.084
0.219
0.112
0.126
0.056
0.218
0.051
0.023
0.009
2.42
4.98
1.46
0.110
81.7
0.016
9.92
1.40
0.410
Median
(ng/m3)
2.16
0
0.171
0.043
0.038
0.117
0.064
0.069
0.030
0.134
0.030
0
0
1.27
3.03
1.03
0.061
62.8
0
5.43
0.770
0.203
Mode
(ng/m3)
0
0
0
0
0
0
0
0
0
0
0
0
0
1.51
1.91
1.04
0
102
0
11
1.29
0
First
Quartile
(ng/m3)
0.986
0
0
0
0
0.054
0.030
0.034
0
0.070
0
0
0
0.690
1.69
0.589
0.025
32.8
0
2.87
0.419
0.117
Third
Quartile
(ng/m3)
4.55
0.587
0.411
0.111
0.109
0.276
0.145
0.150
0.076
0.293
0.061
0
0
2.51
5.16
1.77
0.138
108
0.029
10.1
1.47
0.410
Standard
Deviation
(ng/m3)
8.09
1.54
0.826
0.158
0.149
0.284
0.145
0.168
0.086
0.252
0.079
0.080
0.023
3.76
6.98
1.49
0.150
73.3
0.033
15.8
1.95
0.662
Coefficient
of
Variation
1.74
2.31
2.24
1.80
1.76
1.30
1.29
1.33
1.53
1.15
1.55
3.50
2.47
1.55
1.40
1.02
1.37
0.898
2.01
1.59
1.39
1.61
1 Out of 1,382 valid samples
2 Excludes zeros for non-detects

-------
                                    Table 4-5. Statistical Summaries of the Metals Concentrations
Pollutant
Antimony (PM10)
Arsenic (PM10)
Beryllium (PM10)
Cadmium (PM10)
Chromium (PM10)
Cobalt (PM10)
Lead (PM10)
Manganese (PM10)
Mercury (PM10)
Nickel (PM10)
Selenium (PM10)
Antimony (TSP)
Arsenic (TSP)
Beryllium (TSP)
Cadmium (TSP)
Chromium (TSP)
Cobalt (TSP)
Lead (TSP)
Manganese (TSP)
Mercury (TSP)
Nickel (TSP)
Selenium (TSP)
#of
Measured
Detections1'2
546
541
517
547
526
536
547
547
516
547
490
287
287
287
287
287
287
287
287
287
287
287
#
of Non-
Detects1'2
1
6
30
0
21
11
0
0
31
0
57
0
0
0
0
0
0
0
0
0
0
0
Minimum3
(ng/m3)
0.020
0.004
0.0003
0.006
0.004
0.002
0.320
0.403
0.0005
0.073
0.0001
0.063
0.099
0.003
0.025
0.850
0.065
0.571
2.07
0.005
0.271
0.052
Maximum
(ng/m3)
8.87
4.27
0.27
2.23
11
2.96
30.9
395
0.526
7.97
3.24
4.12
1.76
0.108
1.50
9.06
12.3
19.0
104
0.089
5.44
2.57
Arithmetic
Mean
(ng/m3)
1.27
0.587
0.012
0.155
2.68
0.179
3.82
8.81
0.019
1.27
0.440
0.565
0.559
0.022
0.185
1.85
0.708
3.70
17.8
0.013
1.20
0.715
Median
(ng/m3)
0.830
0.436
0.006
0.083
2.49
0.095
2.50
4.48
0.010
0.980
0.300
0.432
0.519
0.018
0.145
1.70
0.370
3.12
14.4
0.012
1.01
0.690
Mode
(ng/m3)
0.190
0.090
0.002
0.040
0
0.040
2.47
13.8
0.01
1.18
0
0.395
0.247
0.009
0.131
1.45
0.368
2.30
21.8
0.012
1.28
1.01
First
Quartile
(ng/m3)
0.478
0.258
0.002
0.050
1.89
0.050
1.57
2.36
0.007
0.629
0.140
0.289
0.349
0.010
0.090
1.43
0.227
2.25
9.32
0.009
0.690
0.399
Third
Quartile
(ng/m3)
1.43
0.761
0.010
0.153
3.25
0.180
4.42
9.86
0.020
1.43
0.540
0.696
0.704
0.029
0.218
2.09
0.663
4.36
23.1
0.016
1.50
0.951
Standard
Deviation
(ng/m3)
1.44
0.525
0.028
0.226
1.44
0.303
4.00
20.6
0.032
1.13
0.488
0.489
0.288
0.016
0.175
0.719
1.30
2.41
12.8
0.007
0.706
0.396
Coefficient
of
Variation
1.13
0.894
2.25
1.46
0.536
1.69
1.05
2.34
1.70
0.890
1.11
0.865
0.515
0.723
0.944
0.389
1.83
0.652
0.719
0.530
0.589
0.554
1 For PM10, out of 548 valid samples
2 For TSP, out of 287 valid samples
3 Excludes zeros for non-detects

-------
                                  Table 4-6. Statistical Summary of the Hexavalent Chromium Concentrations
Pollutant
Hexavalent Chromium
#of
Measured
Detections1
1,068
#
of Non-
Detects1
260
Minimum2
(ng/m3)
0.001
Maximum
(ng/m3)
0.268
Arithmetic
Mean
(ng/m3)
0.024
Median
(ng/m3)
0.018
Mode
(ng/m3)
0
First
Quartile
(ng/m3)
0.009
Third
Quartile
(ng/m3)
0.031
Standard
Deviation
(ng/m3)
0.026
Coefficient
of
Variation
1.08
   1 Out of 1,328 valid samples
   2 Excludes zeros for non-detects
to

-------
4.1.2   Concentration Range and Data Distribution
       The concentrations measured during the 2011 NMP show a wide range of variability. The
minimum and maximum concentration measured (excluding zeros substituted for non-detects)
for each target pollutant are presented in Tables 4-1 through 4-6 (in respective pollutant group
units).  Some pollutants, such as acetonitrile, had a wide range of concentrations measured, while
other pollutants, such as dichlorotetrafluoroethane, did not, even though they were both detected
frequently. The pollutant for each method-specific pollutant group with the largest range in
measured concentrations is as follows:
       •  For VOCs, acetonitrile (0.045 ppbv to 1,030 ppbv)
       •  For SNMOCs, ethane (1.59 ppbC to 1,390 ppbC)
       •  For carbonyl compounds, formaldehyde (0.026 ppbv to 22.5 ppbv)
       •  For PAHs, naphthalene (2.27 ng/m3 to 779 ng/m3)
       •  For metals, both size fractions, manganese (0.403 ng/m3 to 395 ng/m3 for PMio and
          2.07 ng/m3 to 104 ng/m3 for TSP)
       •  For hexavalent chromium, 0.001 ng/m3 to 0.268 ng/m3.

4.1.3   Central Tendency
       In addition to the number of measured detections and the concentration ranges,
Tables 4-1 through 4-6 also present a number of central tendency and data distribution statistics
(arithmetic mean, median, mode, first and third quartiles, standard deviation, and coefficient of
variation) for each of the pollutants sampled during the 2011  NMP in respective pollutant group
units. A multitude of observations can be made from these tables. The pollutants with the three
highest average concentrations, by mass, for each pollutant group are provided below, with
respective confidence intervals, although the 95 percent confidence interval is not provided in
the table.

       The top three VOCs by average mass concentration, as presented in Table 4-1, are:
       •  acetonitrile (14.7 ฑ 3.67 ppbv)
       •  dichloromethane (1.01 ฑ 0.693 ppbv)
       •  acetylene (0.934 ฑ0.182 ppbv).

                                          4-13

-------
       The top three SNMOCs by average mass concentration, as presented in Table 4-2, are:
       •  ethane (54.9 ฑ 9.09 ppbC)
       •  propane (34.5 ฑ4.04 ppbC)
       •  isopentane (14.0 ฑ 1.34 ppbC).

       The top three carbonyl compounds by average mass concentration, as presented in
Table 4-3, are:
       •  formaldehyde (2.33 ฑ 0.09 ppbv)
       •  acetaldehyde (1.11 ฑ 0.04 ppbv)
       •  acetone (1.08 ฑ 0.04 ppbv).

       The top three PAHs by average mass concentration, as presented in Tables 4-4, are:
       •  naphthalene (81.7 ฑ3.87 ng/m3)
       •  phenanthrene (9.92 ฑ 0.83 ng/m3)
       •  fluorene (4.98 ฑ 0.37 ng/m3).

       The top three metals by average mass concentration for both PMio and TSP fractions, as
presented in Table 4-5, are;
       •  manganese (PMio = 8.81 ฑ 1.73 ng/m3, TSP = 17.80 ฑ 1.49 ng/m3)
       •  lead (PMio = 3.82 ฑ 0.34 ng/m3, TSP = 3.70 ฑ 0.28 ng/m3)
       •  total chromium (PMW = 2.68 ฑ 0.12 ng/m3, TSP = 1.85 ฑ 0.08 ng/m3).

       The average mass concentration of hexavalent chromium, as presented in Table 4-6, is
0.024 ฑ0.001 ng/m3.

       Appendices J through O present similar statistical calculations on a site-specific basis.
                                         4-14

-------
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 formaldehyde failed the highest number of screens (1,592), although
acetaldehyde and benzene were not far behind (1,570 and  1,488, respectively). These three
pollutants were also among those with the highest number of measured detections. Conversely,
four pollutants (chloroprene, c/s-l,3-dichloropropene, tetrachloroethylene, and
frvms-l^-dichloropropene) failed  only one screen each. The number of measured detections for
these four pollutants varied significantly. Tetrachloroethylene was detected in 958 samples while
chloroprene was detected only once (both out of 1,281  valid samples).

       While seven pollutants exhibited a failure rate of 100 percent, most of them were
infrequently detected. Of these seven, benzene was detected in all 1,488 samples (both SNMOC
and Method TO-15), while the other pollutants were detected less frequently. Thus, the number
of failed screens, the number of measured detections, and the failure rate must all be considered
when reviewing the results of the preliminary risk-based screening process. Other pollutants with
relatively high failure rates include acetaldehyde, acrylonitrile, carbon tetrachloride,
1,2-dichloroethane, 1,2-dibromoethane, formaldehyde, and 1,1,2,2-tetrachloroethane. While each
of these pollutants failed more than 98 percent of screens,  four of them (acrylonitrile,
1,2-dichloroethane, 1,2-dibromoethane, and 1,1,2,2-tetrachloroethane) were detected in fewer
than 25 percent of samples collected.
                                           4-15

-------
           Table 4-7. Results of the Program-Level Preliminary Risk-Based Screening
Pollutant
Formaldehyde
Acet aldehyde
Benzene
Carbon Tetrachloride
Naphthalene
1,3-Butadiene
Arsenic
Manganese
Ethylbenzene
p-Dichlorobenzene
1 ,2-Dichloroethane
Acrylonitrile
Acenaphthene
Fluorene
Hexachloro- 1 , 3 -butadiene
Nickel
1 , 1 ,2,2-Tetrachloroethane
Propionaldehyde
Fluoranthene
Hexavalent Chromium
Cadmium
Trichloroethylene
Dichloromethane
1,2-Dibromoethane
Benzo(a)pyrene
Lead
Chloroform
Xylenes
Chloromethylbenzene
Acenaphthylene
Bromomethane
1,1,2-Trichloroethane
Carbon Bisulfide
Cobalt
Chloroprene
cis- 1 , 3 -Dichloropropene
trans- 1 ,3 -Dichloropropene
Tetrachloroethylene
Screening
Value
(Hg/m3)
0.077
0.45
0.13
0.17
0.029
0.03
0.00023
0.005
0.4
0.091
0.038
0.015
0.011
0.011
0.045
0.0021
0.017
0.8
0.011
0.000083
0.00056
0.2
7.7
0.0017
0.00057
0.015
9.8
10
0.02
0.011
0.5
0.0625
70
0.01
0.0021
0.25
0.25
3.8
Total
#of
Failed
Screens
1,592
1,570
1,488
1,261
1,074
995
694
518
401
351
320
300
132
117
105
97
55
49
43
40
36
35
34
31
19
17
14
12
9
4
4
4
2
2
1
1
1
1
11,429
#of
Measured
Detections
1,594
1,594
1,488
1,279
1,382
1,045
828
834
1,487
736
320
300
1,366
1,376
113
834
55
1,588
1,382
1,068
834
251
1,264
31
844
834
756
1,484
9
634
737
8
1,219
823
1
11
6
958
31,373
%of
Failed
Screens
99.87
98.49
100.00
98.59
77.71
95.22
83.82
62.11
26.97
47.69
100.00
100.00
9.66
8.50
92.92
11.63
100.00
3.09
3.11
3.75
4.32
13.94
2.69
100.00
2.25
2.04
1.85
0.81
100.00
0.63
0.54
50.00
0.16
0.24
100.00
9.09
16.67
0.10
36.43
%of
Total
Failures
13.93
13.74
13.02
11.03
9.40
8.71
6.07
4.53
3.51
3.07
2.80
2.62
1.15
1.02
0.92
0.85
0.48
0.43
0.38
0.35
0.31
0.31
0.30
0.27
0.17
0.15
0.12
0.10
0.08
0.03
0.03
0.03
0.02
0.02
0.01
0.01
0.01
0.01
Cumulative
%
Contribution
13.93
27.67
40.69
51.72
61.12
69.82
75.89
80.43
83.94
87.01
89.81
92.43
93.59
94.61
95.53
96.38
96.86
97.29
97.66
98.01
98.33
98.64
98.93
99.20
99.37
99.52
99.64
99.75
99.83
99.86
99.90
99.93
99.95
99.97
99.97
99.98
99.99
100.00

BOLD = EPA MQO NATTS Core Analyte
                                       4-16

-------
       The 18 NATTS MQO Core Analytes (excluding acrolein) are bolded in Table 4-7.
Several NATTS MQO Core Analytes failed screens, but did not contribute to the top 95 percent
of failed screens (such as hexavalent chromium). However, as described in Section 3.2, all
NATTS MQO Core Analytes are inherently designated as program-wide pollutants of interest.
Two pollutants, beryllium and vinyl chloride, were added as pollutants of interest because they
are NATTS MQO Core Analytes, even though they did not fail any screens. These two
pollutants are not shown in Table 4-7. Note that seven of the 15 pollutants contributing to the top
95 percent of failed screens (ethylbenzene,/>-dichlorobenzene, 1,2-dichloroethane, acrylonitrile,
acenaphthene, fluorene, and hexachloro-1,3-butadiene) are not NATTS MQO Core Analytes.
       The program-level pollutants of interest, as indicated by the shading and/or bolding in
Table 4-7, were identified as follows:
       •  Acenaphthene                               •   1,2-Dichloroethane
       •  Acetaldehyde
       •  Acrylonitrile
       •  Arsenic
       •  Benzene
       •  Benzo(a)pyrene
       •  Beryllium
       •  1,3-Butadiene
       •  Cadmium
       •  Carbon Tetrachloride
       •  Chloroform
       •  />-Dichlorobenzene
•  Ethylbenzene
•  Fluorene
•  Formaldehyde
•  Hexachloro-1,3-butadiene
•  Hexavalent Chromium
•  Lead
•  Manganese
•  Naphthalene
•  Nickel
•  Tetrachloroethylene
•  Trichloroethylene
•  Vinyl Chloride
                                          4-17

-------
       The list of pollutants of interest identified via the preliminary risk-based screening
approach for 2011 is similar to the list of pollutants of interest for 2010. There is, however, one
exception. Hexachloro-1,3-butadiene was not identified as a pollutant of interest for 2010,
although it did fail screens.  This pollutant was detected in 10 times as many samples in 2011
compared to 2010 and failed roughly 10 times more screens in 2011 than in 2010.

       Of the 81 pollutants sampled for under the NMP that have corresponding screening
values, concentrations of 38 pollutants failed at least one screen (or roughly 48 percent of
pollutants). Of these, a total of 11,429 out of 31,373 concentrations (or 36.4 percent) failed
screens. If the measured detections for vinyl chloride and beryllium (the two NATTS MQO Core
Analytes that did not fail any screens) are included in the total number of concentrations
(32,259), the percentage of failed screens is 35.4 percent. 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 (11,429 of 53,222, or 21.5 percent).

       Table 4-8 presents the total number of failed screens per site, in descending order, as a
means of comparing the results of the preliminary risk-based screening process across the sites.
As shown, S4MO has the largest number of failed screens (602), followed by PXSS (589) and
NBIL (556). In addition to the number of failed screens, Table 4-8 also provides the total number
of screens conducted (one screen per valid preprocessed daily measurement for each site for all
pollutants with screening values). The failure rate, as a percentage, was determined from the
number of failed screens and the total number of screens conducted (based on applicable
measured detections) and is also provided in  Table 4-8.
                                          4-18

-------
Table 4-8. Site-Specific Risk-Based Screening Comparison
Site
S4MO
PXSS
NBIL
TOOK
TMOK
BTUT
GPCO
SEWA
DEMI
OCOK
PROK
MWOK
ELNJ
SPIL
NBNJ
CHNJ
SSSD
GLKY
UCSD
UNVT
SKFL
SYFL
SPAZ
SJJCA
RICO
PACO
AZFL
BURVT
BOMA
ORFL
INDEM
MONY
RRMI
BMCO
WPIN
ROCH
RUVT
CAMS 35
BRCO
PRRI
CELA
#of
Failed
Screens
602
589
556
521
511
510
498
482
468
457
422
417
406
405
364
345
319
303
300
179
176
156
154
143
142
125
123
123
122
120
115
115
110
102
102
97
96
75
70
66
64
Total # of
Measured
Detections1
2,724
2,584
2,662
1,701
1,769
2,473
2,076
2,531
2,112
1,776
1,640
1,654
1,319
1,236
1,266
1,244
1,251
2,142
1,143
1,884
964
814
549
1,243
414
395
186
533
1,471
180
171
893
169
367
153
735
472
629
331
798
691
%of
Failed
Screens
22.10
22.79
20.89
30.63
28.89
20.62
23.99
19.04
22.16
25.73
25.73
25.21
30.78
32.77
28.75
27.73
25.50
14.15
26.25
9.50
18.26
19.16
28.05
11.50
34.30
31.65
66.13
23.08
8.29
66.67
67.25
12.88
65.09
27.79
66.67
13.20
20.34
11.92
21.15
8.27
9.26
#of
Pollutant
Groups
Analyzed
5
5
6
3
3
6
4
5
4
3
3
3
2
2
2
2
3
5
3
4
3
3
1
2
2
2
1
1
3
1
1
2
1
2
1
2
1
2
2
2
1
     1 Total number of measured detections for all pollutants with
     screening values, not just those failing screens.
     BOLD ITALICS = EPA-designaled NATTS Site
                            4-19

-------
          Table 4-8. Site-Specific Risk-Based Screening Comparison (Continued)
Site
PANJ
WADC
RIVA
SWMI
SDGA
RUCA
PAFL
CHSC
CAMS 85
HOW
#of
Failed
Screens
63
61
57
57
56
54
28
3
0
0
Total # of
Measured
Detections1
237
731
681
84
668
625
308
452
50
41
%of
Failed
Screens
26.58
8.34
8.37
67.86
8.38
8.64
9.09
0.66
0.00
0.00
#of
Pollutant
Groups
Analyzed
1
2
2
1
2
1
1
2
1
1
                    1 Total number of measured detections for all pollutants with
                    screening values, not just those failing screens.
                    BOLD ITALICS = EPA-designaled NATTS Site
       The total number of screens and the number of pollutant groups measured by each site
must also be considered when interpreting the results in Table 4-8. For example, sites sampling
three, four, or five pollutant groups tended to have a higher number of failed screens. Although
WPIN, ORFL, INDEM, and SWMI have the highest failure rates (67-68 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,
BTUT, and SEWA. These sites each sampled five  or six pollutant groups and have a failure rate
around 20 percent. For this reason, the number of pollutant groups for which sampling was
conducted is also presented in Table 4-8. Note that measurements for two sites, HOWI and
CAMS 85, did not fail any screens (both of these sites sampled only hexavalent chromium).

       The following sections from this point forward focus only on those  pollutants designated
as program-level pollutants of interest.
                                          4-20

-------
4.2.1   Concentrations of the Pollutants of Interest
       Concentrations of the program-level pollutants of interest vary significantly, among the
pollutants and among the sites. Tables 4-9 through 4-12 present the top 10 annual average
concentrations and 95 percent confidence intervals by site for each of the program-level
pollutants of interest (for VOCs, carbonyl compounds, PAHs, and metals respectively). As
described in Section 3.1, an annual average is the average concentration of all measured
detections and zeros substituted for non-detects for a given year. Further, an annual average is
only considered valid where there are at least three quarterly averages and where the site-specific
method completeness is at least 85 percent. The annual average concentrations for PAHs in
Table 4-11 and metals in Table 4-12 are reported in ng/m3 for ease of viewing, while annual
average concentrations in Tables 4-9 and 4-10, for VOCs and carbonyl compounds, respectively,
are reported in ug/m3. Note that not all sites sampled each pollutant; thus, the list of possible
sites presented in Tables 4-9 through 4-12 is limited to those sites  sampling each pollutant. For
example, only five sites sampled TSP metals; thus, all five sites appear in Table 4-12 for each
metal (TSP) pollutant of interest shown.
                                          4-21

-------
                      Table 4-9. Annual Average Concentration Comparison of the VOC Pollutants of Interest
to
to
Rank
1
2
3
4
5
6
7
8
9
10
Acrylonitrile
(Hg/m3)
SPIL
0.77 ฑ0.11
UCSD
0.17 ฑ0.04
GLKY
0.16 ฑ0.07
NBIL
0.13 ฑ0.03
CHNJ
0.10 ฑ0.05
NBNJ
0.09 ฑ0.05
TMOK
0.08 ฑ 0.07
OCOK
0.07 ฑ0.05
GPCO
0.06 ฑ0.03
ELNJ
0.04 ฑ0.03
Benzene
(jig/m )
TOOK
3.59 ฑ0.98
SPAZ
1.65 ฑ0.45
PACO
1.45 ฑ0.21
TMOK
1.35ฑ0.18
PXSS
1.34 ฑ0.22
GPCO
1.34 ฑ0.35
RICO
1.27 ฑ0.18
BTUT
1.14ฑ0.15
ELNJ
1.03 ฑ0.13
RUVT
0.89 ฑ0.16
1,3-Butadiene
(Hg/m3)
SPAZ
0.29 ฑ0.10
PXSS
0.23 ฑ 0.06
OCOK
0.20 ฑ0.31
RICO
0.18 ฑ0.04
SPIL
0.16 ฑ0.03
ELNJ
0.14 ฑ0.02
GPCO
0.13 ฑ0.02
NBIL
0.12 ฑ0.10
TMOK
0.10 ฑ0.02
BTUT
0.10 ฑ0.02
Carbon
Tetrachloride
(Hg/m3)
DEMI
0.65 ฑ0.02
SEWA
0.65 ฑ0.03
NBIL
0.64 ฑ0.03
PXSS
0.63 ฑ0.03
SPAZ
0.63 ฑ0.07
TOOK
0.63 ฑ0.05
GLKY
0.63 ฑ0.03
CHNJ
0.62 ฑ0.03
PROK
0.62 ฑ0.03
S4MO
0.61 ฑ0.04
Chloroform
(Hg/m3)
NBIL
6.06 ฑ2.17
DEMI
0.82 ฑ0.14
PXSS
0.37 ฑ0.07
S4MO
0.35 ฑ0.40
SPAZ
0.18 ฑ0.04
ELNJ
0.14 ฑ0.03
NBNJ
0.13 ฑ0.02
SEWA
0.12 ฑ0.01
CHNJ
0.10 ฑ0.03
SPIL
0.09 ฑ0.02
/7-Dichlorobenzene
(Hg/m3)
SPAZ
0.26 ฑ 0.07
S4MO
0.21ฑ0.11
PROK
0.20 ฑ 0.07
PXSS
0.20 ฑ 0.04
TOOK
0.15 ฑ0.03
ELNJ
0.10 ฑ0.02
MWOK
0.09 ฑ0.01
TMOK
0.08 ฑ0.02
SPIL
0.07 ฑ0.03
BURVT
0.07 ฑ0.01
            BOLD ITALICS = EPA-designaled NATTS Site

-------
                Table 4-9. Annual Average Concentration Comparison of the VOC Pollutants of Interest (Continued)
Rank
1
2
3
4
5
6
7
8
9
10
1,2-Dichloro ethane
(Hg/m3)
CHNJ
0.034 ฑ0.019
S4MO
0.032 ฑ0.013
OCOK
0.031 ฑ0.024
NBNJ
0.030 ฑ0.012
ELNJ
0.028 ฑ0.014
PROK
0.028 ฑ0.012
GLKY
0.028 ฑ0.01
BTUT
0.028 ฑ0.013
GPCO
0.027 ฑ0.012
UCSD
0.026 ฑ0.011
Ethylbenzene
(Hg/m3)
SPAZ
1.06 ฑ0.28
PXSS
0.82 ฑ0.13
TOOK
0.68 ฑ0.13
GPCO
0.62 ฑ0.11
DEMI
0.61 ฑ0.13
TMOK
0.55 ฑ0.09
ELNJ
0.51 ฑ0.16
NBNJ
0.49 ฑ0.39
BTUT
0.47 ฑ0.13
S4MO
0.38 ฑ0.04
Hexachloro-1,3-
Butadiene
(Hg/m3)
S4MO
0.023 ฑ0.014
BURVT
0.019 ฑ0.015
MWOK
0.017 ฑ0.023
SPIL
0.016 ฑ0.025
NBNJ
0.015 ฑ0.01
GPCO
0.014 ฑ0.01
GLKY
0.014 ฑ0.009
UCSD
0.013 ฑ0.011
CHNJ
0.011 ฑ0.009
BTUT
0.010 ฑ0.008
Tetrachloroethylene
(Hg/m3)
PXSS
0.55 ฑ0.28
SPAZ
0.41 ฑ0.16
NBIL
0.35 ฑ0.13
SPIL
0.31 ฑ0.08
GPCO
0.26 ฑ0.05
ELNJ
0.20 ฑ0.05
DEMI
0.18 ฑ0.03
S4MO
0.18 ฑ0.04
MWOK
0.15 ฑ0.06
NBNJ
0.13 ฑ0.02
Trichloroethylene
(Hg/m3)
SPIL
0.64 ฑ0.41
ELNJ
0.09 ฑ0.09
SPAZ
0.05 ฑ0.03
CHNJ
0.05 ฑ0.09
NBIL
0.04 ฑ 0.02
GPCO
0.04 ฑ 0.02
S4MO
0.03 ฑ0.01
NBNJ
0.02 ฑ0.01
PXSS
0.02 ฑ0.01
UCSD
0.02 ฑ 0.02
Vinyl Chloride
(Hg/m3)
NBIL
0.003 ฑ 0.002
NBNJ
0.003 ฑ 0.002
SPIL
0 003 ฑ 0 003
S4MO
0.003 ฑ 0.002
DEMI
0.003 ฑ 0.002
CHNJ
0.002 ฑ 0.002
GPCO
0.002 ฑ 0.002
OCOK
0.001 ฑ0.002
PXSS
0.001 ฑ0.001
UCSD
0.001 ฑ0.001
J^.
K>
           BOLD ITALICS = EPA-designated NATTS Site

-------
              Table 4-10. Annual Average Concentration Comparison of the Carbonyl Compound Pollutants of Interest
J^.
K>
Rank
1
2
3
4
5
6
7
8
9
10
Acet aldehyde
(Ug/m3)
ELNJ
3.24 ฑ0.39
SPIL
2.94 ฑ0.61
S4MO
2.75 ฑ0.43
TOOK
2.75 ฑ0.41
NBNJ
2.49 ฑ0.27
GPCO
2.43 ฑ0.29
OCOK
2.41 ฑ0.33
TMOK
2.40 ฑ0.34
UCSD
2.33 ฑ0.40
BTUT
2.19 ฑ0.35
Formaldehyde
(Ug/m3)
BTUT
4.49 ฑ1.15
S4MO
4.25 ฑ 0.92
OCOK
4 06 ฑ 0 80
MWOK
4.05 ฑ0.53
TMOK
3.93 ฑ0.62
PROK
3.84 ฑ0.67
TOOK
3.74 ฑ0.57
ELNJ
3.45 ฑ0.44
SPIL
3.29 ฑ0.46
RRMI
3.23 ฑ0.47
                                             BOLD ITALICS = EPA-designated NATTS Site

-------
                       Table 4-11. Annual Average Concentration Comparison of the PAH Pollutants of Interest
to
Rank
1
2
3
4
5
6
7
8
9
10
Acenaphthene
(ng/m3)
DEMI
13.42 ฑ5. 17
NBIL
13.14ฑ4.14
ROCH
12.29 ฑ3.55
GPCO
10.54 ฑ2.17
MONY
8.99 ฑ1.80
CAMS 35
6.58 ฑ2.30
S4MO
5.41 ฑ1.45
CELA
4.92 ฑ0.76
PRRI
3 65 ฑ083
RIVA
3. 13 ฑ0.64
Benzo(a)pyrene
(ng/m3)
GPCO
0.20 ฑ0.08
MONY
0.20 ฑ 0.04
DEMI
0.19 ฑ0.07
PRRI
0.16 ฑ0.03
NBIL
0.15 ฑ0.03
S4MO
0.13 ฑ0.03
BOMA
0.11 ฑ0.02
PXSS
0.10 ฑ0.04
ROCH
0.09 ฑ0.03
SEWA
0.08 ฑ0.05
Fluorene
(ng/m3)
NBIL
13.46 ฑ4.25
DEMI
11.19ฑ3.84
ROCH
9.92 ฑ2.81
MONY
9.47 ฑ1.55
CAMS 35
7.84 ฑ2.85
GPCO
7.68 ฑ0.98
S4MO
6.08 ฑ 1.40
CELA
5.62 ฑ0.70
PRRI
4.97 ฑ0.96
RUCA
4.09 ฑ0.56
Naphthalene
(ng/m3)
GPCO
155. 52 ฑ29.71
DEMI
143. 35 ฑ23.07
MONY
135.66 ฑ15.50
CELA
131. 96 ฑ21.23
WADC
102.71 ฑ20.46
NBIL
99.39 ฑ29.75
CAMS 35
94. 14 ฑ18.02
PRRI
91.41 ฑ14.27
RUCA
91. 18 ฑ17.14
SDGA
90.82 ฑ15.61
                              BOLD ITALICS = EPA-designated NATTS Site

-------
                      Table 4-12. Annual Average Concentration Comparison of the Metals Pollutants of Interest
to
Rank
1
2
o
J
4
5
6
7
8
9
10
Arsenic
(PM10)
(ng/m3)
S4MO
0.87 ฑ0.14
PXSS
0.77 ฑ0.16
NBIL
0.73 ฑ0.15
Sฃ^
0.66 ฑ0.12
PAFL
0.62 ฑ0.17
BTUT
0.59 ฑ0.19
BOM4
0.43 ฑ 0.07
SJJCA
0.39 ฑ0.07
UNVT
0.25 ฑ 0.06

Arsenic
(TSP)
(ng/m3)
TOOK
0.76 ฑ0.08
TMOK
0.63 ฑ 0.08
PROK
0.54 ฑ0.07
OCOK
0.47 ฑ 0.07
MWOK
0.40 ฑ 0.04





Beryllium
(PM10)
(ng/m3)
PXSS
0.047 ฑ0.016
BTUT
0.013 ฑ0.004
BOMA
0.010 ฑ0.007
S4MO
0.008 ฑ 0.002
NBIL
0 007 ฑ 0 004
SJJCA
0.007 ฑ0.001
SEWA
0.006 ฑ 0.005
UNVT
0.006 ฑ 0.002
PAFL
0.004 ฑ0.001

Beryllium
(TSP)
(ng/m3)
PROK
0.029 ฑ 0.006
TOOK
0.026 ฑ 0.004
MWOK
0.019 ฑ0.003
OCOK
0.018 ฑ0.003
TMOK
0.018 ฑ0.003





Cadmium
(PM10)
(ng/m3)
S4MO
0.56 ฑ0.12
NBIL
0 1 4 ฑ 0 03
BTUT
0.14 ฑ0.05
PXSS
0.14 ฑ0.03
BOMA
0.13 ฑ0.01
SEWA
0.10 ฑ0.03
SJJCA
0.07 ฑ 0.02
UNVT
0.07 ฑ0.01
PAFL
0.06 ฑ0.01

Cadmium
(TSP)
(ng/m3)
TOOK
0.31 ฑ0.08
TMOK
0.22 ฑ0.03
PROK
0.16 ฑ0.02
OCOK
0.13 ฑ0.03
MWOK
0.12 ฑ0.02





Hexavalent
Chromium
(ng/m3)
PXSS
0.065 ฑ0.012
CAMS 35
0.050 ฑ 0.007
DEMI
0.047 ฑ 0.009
MONY
0.041 ฑ0.005
SEWA
0.033 ฑ 0.007
S4MO
0.033 ฑ 0.006
BOMA
0.026 ฑ 0.006
SKFL
0 023 ฑ 0 005
PRRI
0.022 ฑ 0.006
CAMS 85
0.022 ฑ 0.003
                BOLD ITALICS = EPA-designated NATTS Site

-------
                Table 4-12. Annual Average Concentration Comparison of the Metals Pollutants of Interest (Continued)
to
Rank
1
2
o
J
4
5
6
7
8
9
10
Lead
(PM10)
(ng/m3)
S4MO
10.42 ฑ 1.72
NBIL
4.16 ฑ1.07
PXSS
4.06 ฑ 0.64
BTUT
3.51 ฑ0.98
BOMA
3.07 ฑ0.44
SJJCA
3.02 ฑ0.56
SEWA
2.89 ฑ0.44
PAFL
2.13 ฑ0.47
UNVT
1.46 ฑ0.29

Lead
(TSP1
(ng/m5)
TOOK
5.87 ฑ0.87
TMOK
4.91 ฑ0.64
PROK
2.69 ฑ0.3
OCOK
2.61 ฑ0.26
MWOK
2.48 ฑ0.22





Manganese
(PM10)
(ng/m3)
PXSS
22.82 ฑ6.79
S4MO
18.42 ฑ13. 18
NBIL
8.30 ฑ2.29
SEWA
8.17 ฑ2.57
BTUT
7.58 ฑ1.27
SJJCA
6.62 ฑ1.14
BOMA
3.48 ฑ0.44
PAFL
2.23 ฑ0.35
UNVT
1.84 ฑ0.32

Manganese
(TSP1
(ng/nr5)
TOOK
30.09 ฑ4.58
TMOK
20.52 ฑ3.06
OCOK
14.04 ฑ2.01
MWOK 13.26
ฑ1.94
PROK
11. 31 ฑ1.88





Nickel
(PM10)
(ng/m3)
SEWA
1.90 ฑ0.46
PXSS
1.74 ฑ0.33
BTUT
1.73 ฑ0.35
BOMA
1.38 ฑ0.17
NBIL
1.27 ฑ0.21
SJJCA
1.27 ฑ0.17
S4MO
1.20 ฑ0.26
PAFL
0.72 ฑ0.10
UNVT
0.53 ฑ0.10

Nickel
(TSP1
(ng/m5)
TOOK
1.75 ฑ0.19
TMOK
1.42 ฑ0.19
MWOK
1.20 ฑ0.21
OCOK
0.84 ฑ0.08
PROK
0.79 ฑ0.09





                        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 chloroform for NBIL (6.06 ฑ2.17 |ig/m3). As shown in
   Table 4-9, this annual average concentration is more than seven times the next
   highest concentration of chloroform (DEMI, 0.82 ฑ 0.14 |ig/m3). The relatively large
   confidence interval associated with this annual average indicates a wide range of
   concentrations are factored into this average. Twenty-five of the 26 highest
   chloroform concentrations measured program-wide, ranging from 2.61 |ig/m3to
   32.9 |ig/m3, were measured at NBIL. NBIL also had the highest annual average
   chloroform concentration among sites sampling this pollutant in 2010.

•  Behind NBIL's chloroform concentration, the next four highest annual average
   concentrations were calculated for formaldehyde, ranging from 4.05 ฑ 2.17 |ig/m3
   (MWOK) to 4.49 ฑ2.17 |ig/m3 (BTUT). All other annual average concentrations
   were less than 4.0 |ig/m3. Formaldehyde and acetaldehyde account  for 30 of the 32
   annual average concentrations greater than 2.0 |ig/m3 in Tables 4-9 through 4-12 (the
   two exceptions being for NBIL's chloroform and TOOK's benzene).

•  Among the VOCs, the annual average concentrations of benzene are the only annual
   averages consistently greater than 1 |ig/m3.  TOOK's annual average benzene
   concentration (3.59 ฑ 0.98 |ig/m3) is  significantly higher than the next highest annual
   average benzene concentration (1.65 ฑ 0.45 |ig/m3 for SPAZ), which is similar to the
   2010 NMP report. Across the program, four of the five benzene measurements
   greater than 10 |ig/m3 (and 11 of the  15 benzene concentrations greater than 5 |ig/m3)
   were measured at TOOK. Three of the five  Colorado sites have one of the top 10
   annual average benzene concentrations and both Tulsa, Oklahoma sites appear in
   Table 4-9 for benzene.

•  The difference between the highest and tenth highest annual average concentration of
   carbon tetrachloride is only 0.04 |ig/m3. The difference between the highest and
   lowest annual average concentration of this pollutant among all NMP sites is
   0.12 |ig/m3, indicating the relative uniformity in concentrations of this pollutant in
   ambient air.

•  The annual average concentration of acrylonitrile for SPIL  (0.77 ฑ0.11 |ig/m3) is
   more than four times higher than the next highest annual average concentration of
   this pollutant (0.17 ฑ 0.04 |ig/m3 for UCSD). Thirty-one of the 38 highest
   concentrations of acrylonitrile (those greater than 0.75  |ig/m3) were measured at SPIL
   and of the 24 measurements greater than 1.0 |ig/m3 program-wide,  19 were measured
   at SPIL. However, the highest concentration of acrylonitrile (1.69 |ig/m3) was not
   measured at SPIL; it was  measured at TMOK.

•  The only other VOC for which an annual average concentration was greater than
   1.0 |ig/m3 was calculated for ethylbenzene (SPAZ,  1.06 ฑ 0.28 |ig/m3). SPAZ had the
   second highest number of ethylbenzene concentrations greater than 1 |ig/m3 (12),
   behind only PXSS (20), which had the second highest annual average concentration
   of ethylbenzene (0.82 ฑ 0.13 |ig/m3).

                                   4-28

-------
•  The annual average concentration of trichloroethylene for SPIL (0.64 ฑ 0.41 |ig/m3)
   is seven times higher than the next highest annual average
   (ELNJ, 0.09 ฑ 0.09 |ig/m3). Of the 13 concentrations of trichloroethylene measured
   across the program that are greater than 0.75 |ig/m3, 10 were measured at SPIL (and
   two were measured at ELNJ). Similar trends in SPIL's trichloroethylene
   concentrations were seen in the 2008-2009 and 2010 NMP reports. The confidence
   intervals shown for both sites indicate a relatively high-level of variability in the
   measurements. Note that trichloroethylene was detected in 75 percent of samples at
   SPIL and 40 percent of samples at ELNJ.

•  Although BTUT has the highest annual average concentration of formaldehyde
   (4.49 ฑ1.15 |ig/m3), the maximum concentration was not measured at this site.
   However,  nine of the 26 concentrations greater than 10 |ig/m3 were measured at
   BTUT.  While just greater than 0.50 |ig/m3 separates the annual averages for the top
   five sites, the confidence interval for BTUT's annual average suggests a slightly
   higher level of variability  in the measurements than the other sites. The site with the
   maximum formaldehyde measurement (2.71 ฑ1.10 |ig/m3) is NBNJ, who's annual
   average formaldehyde concentration ranked 13th.

•  Although ELNJ has the highest annual average concentration of acetaldehyde
   (3.24 ฑ 0.39 |ig/m3), the maximum concentration was measured at SPIL (14.5 |ig/m3),
   which has the second highest annual average concentration (2.94 ฑ 0.61 |ig/m3)
   among sites sampling carbonyl compounds.  These two sites account  for 18 of the 51
   acetaldehyde measurements greater than 5 |ig/m3 (10 for ELNJ and eight for SPIL).

•  Although GPCO has the highest annual average concentration of naphthalene
   (155.52 ฑ  29.71 ng/m3), the highest measurement of this pollutant was not measured
   at this site. NBIL has the maximum measurement among all  sites sampling
   naphthalene (799 ng/m3), although its annual average concentration ranked sixth
   (99.39 ฑ 29.75 ng/m3). GPCO has the highest number of naphthalene measurements
   greater than 300 ng/m3 (five)  among sites sampling this pollutant.

•  S4MO has the highest annual average concentration of three of the six PMio metals:
   arsenic, cadmium, and lead. In addition, S4MO's annual average manganese
   concentration ranks second highest. Several of S4MO's annual averages, such as
   lead, are significantly higher than the other annual averages listed. The two maximum
   concentrations of lead  across the program were measured at S4MO (30.9 ng/m3 and
   27.5 ng/m3). Additionally, 22 of the 32 highest lead concentrations (those greater than
   10 ng/m3) were measured  at S4MO.  The maximum manganese concentration among
   all sites sampling PMio metals was measured at S4MO (395 ng/m3) and is more than
   twice the next highest manganese concentration (130 ng/m3, measured at PXSS). The
   maximum concentration of cadmium was also measured at S4MO, while the second
   highest concentration of arsenic was measured at S4MO (behind only BTUT).

•  TOOK has the highest annual average concentration of five of the six TSP metals:
   arsenic, cadmium, lead, manganese, nickel. TOOK's annual average beryllium
   concentration ranked second, behind PROK. Note that only Oklahoma sites  sampled
   TSP metals.

                                  4-29

-------
       •  S4MO was on the top 10 list for 21 of the 25 program-level pollutants of interest;
          PXSS and NBIL were both on the top 10 list for 17 of the 25 program-level pollutants
          of interest. GPCO appears in Tables 4-9 through 4-12 a total of 14 times. Conversely,
          12 sites do not appear in Table 4-9 through 4-12 at all. Note, however, that some sites
          did not meet the criteria for annual averages to be calculated.

4.2.2   Risk-Based Screening Assessment Using MRLs
       Table 4-13 presents the pollutants analyzed under the NMP that have associated ATSDR
MRLs. Note that some pollutants do not have MRLs for one or more of the designated time
frames (acute, intermediate, or chronic).  None of the preprocessed daily measurements are
greater than associated acute MRL; none of the quarterly  average concentrations, where they
could be calculated, are greater than the associated intermediate MRL; and none of the annual
average concentrations, where they could be calculated, are greater than the associated chronic
MRL.  Thus, Table 4-13 also presents the maximum preprocessed daily measurement, quarterly
average, and annual average concentration associated with each pollutant. This allows the reader
to see how close (or how far) from the MRL(s) some concentrations were. For example, the
acute MRL for benzene is 30 |ig/m3 and the maximum concentration measured was nearly
24 |ig/m3.  Conversely, the  acute MRL for acetone is 60,000 |ig/m3 while the maximum
concentration measured was 15.4 |ig/m3.

       The pollutant with the concentration closest to the acute MRL is benzene (the  acute MRL
is 30 |ig/m3 and the maximum benzene measurement is 23.8 |ig/m3).  The pollutant with the
quarterly average concentration closest to the intermediate MRL is also benzene (the
intermediate MRL is 20 |ig/m3 and the maximum quarterly average is 4.39 |ig/m3).  The pollutant
with the annual  average concentration closest to the chronic MRL is manganese (the chronic
MRL  is 0.04 |ig/m3 and the maximum annual average is 0.03 |ig/m3).

       Because none of the preprocessed daily measurements are greater than associated acute
MRLs, the emission tracer analysis described in Section 3.5.5.1 was not performed.
                                         4-30

-------
                            Table 4-13. Comparison of Maximum Concentrations vs. ATSDR MRLs
Pollutant
Acetone
Acrylonitrile
Benzene
Bromomethane
1,3 -Butadiene
Cadmium
Carbon Bisulfide
Carbon Tetrachloride
Chloroethane
Chloroform
Chloromethane
Chromium
Cobalt
ฃ>-Dichlorobenzene
1,2-Dichloroethane
1, 1-Dichloroethene
trans- 1 ,2-Dichloroethylene
Dichloromethane
1 ,2-Dichloropropane
cis- 1 , 3 -Dichloropropene
trans- 1 ,3 -Dichloropropene
Ethylbenzene
Formaldehyde
w-Hexane
Hexavalent Chromium
Manganese
Mercury
Methyl tert-Butyl Ether
ATSDR Acute
MRL1
(Hg/m3)
60,000
200
30
200
200
0.03
—
—
40,000
500
1,000
—
—
10,000
—
—
800
2,000
200
—
—
20,000
50
—
—
—
—
7,000
Maximum
Prep recessed Daily
Measurement
(Hg/m3)
15.40
1.69
23.80
1.70
9.51
0.01
—
—
21.60
32.90
5.24
—
—
2.83
—
—
0.20
1,214.00
0.06
—
—
11.50
27.70
—
—
—
—
0.99
ATSDR
Intermediate
MRL1
(Hg/m3)
30,000
—
20
200
—
—
—
200
—
200
400
0.3
—
1,000
—
80
800
1,000
30
40
40
9,000
40
—
0.3
—
—
3,000
Maximum
Quarterly Average
Concentration
(Hg/m3)
9.67
—
4.39
0.20
—
—
—
0.72
—
12.41
1.55
<0.01
—
0.54
—
<0.01
0.03
156.90
<0.01
0.03
0.02
1.65
8.37
—
0.01
—
—
0.08
ATSDR Chronic
MRL1
(Hg/m3)
30,000
—
10
20
—
0.01
900
200
—
100
100
—
0.1
60
2,000
—
—
1,000
—
30
30
300
10
2,000

0.04
0.2
3,000
Maximum Annual
Average
Concentration
(Hg/m3)
4.92
—
3.59
0.07
—
0.01
29.46
0.65
—
6.06
1.42
—
0.01
0.26
0.03
—
—
53.90

0.01
0.01
1.06
4.49
4.41
—
0.03
O.01
0.02
Reflects the use of one significant digit for MRLs

-------
                          Table 4-13. Comparison of Maximum Concentrations vs. ATSDR MRLs (Continued)
Pollutant
Naphthalene
Nickel
Styrene
Tetrachloroethylene
Toluene
1,1,1 -Trichloroethane
Trichloroethylene
Vinyl Chloride
Xylenes
ATSDR Acute
MRL1
(Hg/m3)
—
—
20,000
1,000
4,000
10,000
10,000
1,000
9,000
Maximum
Prep recessed Daily
Measurement
(Hg/m3)
—
—
13.10
8.63
93.90
0.44
8.40
0.08
34.76
ATSDR
Intermediate
MRL1
(Hg/m3)
—
0.2
—
—
—
4,000
500
80
3,000
Maximum
Quarterly Average
Concentration
(Hg/m3)
—
0.01
—
—
—
0.10
1.28
0.01
6.45
ATSDR Chronic
MRL1
(Hg/m3)
4
0.09
900
300
300
—
—
—
200
Maximum Annual
Average
Concentration
(Hg/m3)
0.16
0.01
1.45
0.55
6.41
—
—
—
4.06
    Reflects the use of one significant digit for MRLs
-^
to

-------
4.3    The Impact of Mobile Sources
       Ambient air is significantly impacted by mobile sources, as discussed in Section 3.4.1.
Table 4-14 contains several parameters that are used to assess mobile source impacts on 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 account for nearly half of air pollution in the United
States. Mobile source emissions can be broken into two categories: on-road and non-road. On-
road emissions come from mobile sources such as automobiles, buses, and construction vehicles
that use roadways; non-road emissions come from the remaining mobile sources such as
locomotives, lawn mowers, and boats (EPA, 20121). Table 4-14 contains county-level on-road
and non-road HAP emissions from the 2008 NEI.

       Mobile source emissions tend to be highest in large urban areas and lowest in rural  areas.
Estimated on-road county emissions were highest in Los Angeles County, CA (where CELA is
located), followed by Harris County, TX (where CAMS 35 is located), and Maricopa County,
AZ (where PXSS and SPAZ are located) while estimated on-road emissions were lowest in
Chesterfield County, SC and Union County,  SD (where CHSC  and UCSD are located,
respectively). Estimated non-road county emissions were also highest in Los Angeles County,
CA, followed by Cook County, IL (where NBIL and SPIL are located) and Maricopa County,
AZ. Estimated non-road county emissions were lowest  in Union County, SD and Carter County,
KY (where GLKY is located).
                                         4-33

-------
                        Table 4-14. Summary of Mobile Source Information by Monitoring Site
Site
AZFL
BMCO
BOMA
BRCO
BTUT
BURVT
CAMS 35
CAMS 85
CELA
CHNJ
CHSC
DEMI
ELNJ
GLKY
GPCO
HOW
INDEM
MONY
MWOK
NBIL
NBNJ
OCOK
ORFL
PACO
PAFL
County-level
Motor Vehicle
Registration1
(# of Vehicles)
877,075
72,957
481,199
72,957
239,582
169,767
3,164,173
70,585
7,360,573
389,359
40,792
1,334,752
424,894
32,398
178,425
100,176
419,431
246,748
832,160
2,072,399
640,893
832,160
1,056,627
72,957
1,056,627
Estimated
10-Mile Vehicle
Ownership1
(# of Vehicles)
555,080
7,703
1,100,560
33,018
206,495
130,202
528,398
2,410
2,647,604
187,012
4,852
775,162
1,724,607
17,159
145,607
24,846
348,652
1,007,684
427,423
353,553
637,915
429,679
907,230
7,703
795,455
Annual
Average Daily
Traffic1
(# of Vehicles)
40,500
2,527
31,400
150
113,955
14,000
31,043
1,250
230,000
12,917
550
92,800
250,000
428
11,000
5,000
34,240
91,465
40,900
34,600
114,322
40,900
32,500
16,000
46,000
County-level
Daily VMT1
21,395,381
1,901,434
10,695,874
1,901,434
6,866,779
4,027,945
56,650,489
2,578,700
214,458,140
14,256,044
1,276,517
42,804,737
12,485,902
1,084,000
2,031,327
2,626,054
16,226,000
9,698,000
27,190,328
86,863,779
20,415,685
27,190,328
33,325,315
1,901,434
33,325,315
County-Level
On-road
Emissions2
(tpy)
2,650.97
260.03
715.05
260.03
861.85
371.91
8,521.88
284.17
9,556.40
1,202.20
128.86
5,900.70
951.86
163.87
392.28
247.41
1,222.76
825.34
2,900.47
7,721.47
1,617.22
2,900.47
3,198.03
260.03
3,198.03
County-Level
Non-road
Emissions2
(tpy)
1,157.75
93.05
440.97
93.05
336.23
251.44
2,791.78
129.54
5,072.27
705.27
80.37
1,113.36
390.19
15.58
180.83
220.50
634.27
391.72
816.74
4,074.66
673.13
816.74
1,587.50
93.05
1,587.50
Hydrocarbon
Average3
(ppbv)
NA
NA
NA
NA
3.75
2.42
NA
NA
NA
1.81
NA
3.27
4.48
1.30
4.67
NA
NA
NA
1.90
5.71
2.66
2.42
NA
NA
NA
Reference: EPA, 2012d
3This parameter is only available for monitoring sites sampling VOCs.
BOLD ITALICS = EPA-designated NATTS Site
NA = Data not available

-------
                  Table 4-14. Summary of Mobile Source Information by Monitoring Site (Continued)
Site
PANJ
PROK
PRRI
PXSS
RICO
RIVA
ROCH
RRMI
RUCA
RUVT
S4MO
SDGA
SEWA
SJJCA
SKFL
SPAZ
SPIL
SSSD
SWMI
SYFL
TMOK
TOOK
UCSD
UNVT
WADC
WPIN
County-level
Motor Vehicle
Registration1
(# of Vehicles)
396,602
39,968
485,837
3,776,819
72,957
354,721
550,992
1,334,752
1,711,492
70,900
1,114,812
472,535
1,783,335
1,517,190
877,075
3,776,819
2,072,399
210,914
1,334,752
1,135,945
603,926
603,926
25,419
169,767
213,232
820,767
Estimated
10-Mile Vehicle
Ownership1
(# of Vehicles)
1,071,818
29,285
509,773
1,314,732
23,530
544,138
480,772
572,987
784,472
40,098
673,974
492,952
890,137
1,242,743
674,583
826,196
805,601
229,582
721,842
288,235
323,647
448,957
9,676
36,243
666,556
718,087
Annual
Average Daily
Traffic1
(# of Vehicles)
22,272
15,100
136,800
184,000
17,000
73,000
86,198
98,500
145,000
7,200
79,558
140,820
226,000
104,000
47,000
128,000
190,000
18,700
93,000
10,600
12,600
63,000
156
1,100
7,700
143,970
County-level
Daily VMT1
8,178,167
1,656,458
NA
89,448,000
1,901,434
8,246,774
17,772,000
42,804,737
55,717,760
1,766,027
19,896,584
20,187,000
23,282,703
41,250,490
21,395,381
89,448,000
86,863,779
3,751,886
42,804,737
34,351,899
20,348,926
20,348,926
808,049
4,027,945
9,775,000
32,005,000
County-Level
On-road
Emissions2
(tpy)
616.98
172.07
1,104.51
7,862.48
260.03
831.85
1,566.25
5,900.70
2,486.42
158.14
974.72
2,272.55
6,932.11
1,960.08
2,650.97
7,862.48
7,721.47
467.40
5,900.70
3,252.93
2,197.21
2,197.21
91.81
371.91
929.71
2,664.97
County-Level
Non-road
Emissions2
(tpy)
447.26
83.98
381.46
3,819.27
93.05
188.91
683.88
1,113.36
1,003.76
150.60
182.60
772.13
2,762.29
812.60
1,157.75
3,819.27
4,074.66
132.93
1,113.36
1,326.89
867.85
867.85
30.98
251.44
327.98
715.48
Hydrocarbon
Average3
(ppbv)
4.57
2.02
NA
5.16
NA
NA
NA
NA
NA
2.47
2.65
NA
2.37
NA
NA
6.74
2.86
1.75
NA
NA
3.89
9.15
0.98
0.94
NA
NA
Reference: EPA, 2012d
3This parameter is only available for monitoring sites sampling VOCs.
BOLD ITALICS = EPA-designated NATTS Site
NA = Data not available

-------
4.3.2   Hydrocarbon Concentrations
       Hydrocarbons are organic compounds that contain only carbon and hydrogen.
Hydrocarbons are derived mostly from crude petroleum sources and are classified according to
their arrangement of atoms as alicyclic, aliphatic, and aromatic. Hydrocarbons are of prime
economic importance because they encompass the constituents of the major fossil fuels,
petroleum and natural gas, as well as plastics, waxes, and oils. Hydrocarbons in the atmosphere
originate from natural sources and from various anthropogenic sources, such as the combustion
of fuel and biomass, petroleum refining, petrochemical manufacturing, solvent use, and gas and
oil production and use. In urban air pollution, these components, along with oxides of nitrogen
(NOX) and sunlight,  contribute to the formation of tropospheric ozone. Thus, the concentration of
hydrocarbons in ambient air may act as an indicator of mobile source activity levels.  Several
hydrocarbons are  sampled with Method TO-15,  including benzene, ethylbenzene, and toluene.

       Table 4-14 presents the average of the sum of hydrocarbon concentrations for each site
sampling VOCs. Note that only sites sampling VOCs have data in this column. Table 4-14 shows
that TOOK,  SPAZ, and NBIL have the highest hydrocarbon averages among the sites monitoring
VOCs. Each of these sites is located in a highly  populated urban area and in relatively close
proximity to heavily traveled roadways. For example, TOOK is located near Exit 3 A of 1-244 in
Tulsa, Oklahoma. The sites with the lowest hydrocarbon averages are GLKY, UCSD, and
UNVT. All three of these sites are located in rural areas. The average sum of hydrocarbon
concentrations can be compared to other indicators of mobile source activity, such as the ones
discussed below, to  determine if correlations exist.

4.3.3   Motor Vehicle Ownership
       Another indicator of motor vehicle activity near the monitoring sites is the total number
of vehicles owned by residents in the county where each monitoring site is located, which
includes passenger vehicles, trucks, and commercial vehicles, as well as vehicles that can be
regional in use such as boats or snowmobiles. Actual county-level vehicle registration data were
obtained from each applicable state or local agency, where possible. If data were not available,
vehicle registration data are available at the state-level (FHWA, 2011). The county proportion of
the state population  was then applied to the state registration count.
                                          4-36

-------
       The county-level motor vehicle ownership data and the average summed hydrocarbon
concentrations are presented in Table 4-14. As previously discussed, TOOK, SPAZ, and NBIL
have the highest average summed hydrocarbon concentrations, respectively, while GLKY and
UCSD have the lowest. Table 4-14 also shows that SPAZ, PXSS, NBIL, and SPIL have the
highest county-level vehicle ownership of the sites sampling VOCs, while PROK, GLKY, and
UCSD have the lowest. The Pearson correlation coefficient calculated between these two
datasets is 0.46. Although this correlation falls just below the "strong" classification, it does
indicate a positive correlation between hydrocarbon concentrations and vehicle registration.
CELA, which has the highest county-level vehicle ownership of all NMP sites, did not sample
VOCs under the NMP.

       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. Therefore, a vehicle
registration-to-population ratio was developed for each county with a monitoring site. Each ratio
was then applied to the 10-mile population surrounding the sites (provided in the individual state
sections) to estimate a 10-mile vehicle ownership, which is also presented in Table 4-14. Table
4-14 shows that ELNJ, PXSS, and PANJ have the highest 10-mile estimated vehicle ownership
of the sites sampling VOCs, while PROK, GLKY, and UCSD have the lowest. The Pearson
correlation coefficient calculated between the average summed hydrocarbon calculations and the
10-mile vehicle registration estimates  is 0.41. While this  correlation also falls below the "strong"
classification, it does indicate a positive correlation between hydrocarbon concentrations and the
estimated vehicle registration within a 10-mile radius. CELA, which had the highest 10-mile
estimated vehicle ownership of all NMP sites, did not sample VOCs under the NMP.

       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.
                                          4-37

-------
       •  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 total volume of traffic on a highway segment for 1 year, divided by
the number of days in the year," and incorporates both directions of traffic (FL DOT, 2007).
AADT counts obtained were based on data from 2002 to 2011, primarily 2009 forward. The
updated traffic values are presented in Table 4-14. The traffic data presented in Table 4-14
represent the most recently available data applicable to the monitoring sites.

       There are several limitations to obtaining the AADT near each monitoring site. AADT
statistics are developed for roadways, such as interstates, state highways, or local roadways,
which are managed by different municipalities or government agencies. AADT is not always
available in 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 28).

       Table 4-14 shows that ELNJ, SEW A, and SPIL have the highest daily traffic volumes of
the sites sampling VOCs, while IHSTVT, GLKY, and UCSD have the lowest. For all monitoring
sites (not just those sampling VOCs), the highest daily traffic volume occurs near ELNJ, CELA,
and SEWA. ELNJ is located near Exit 13 on 1-95; CELA is located  in downtown Los Angeles;
and SEWA is located in Seattle near the intersection of 1-5 and 1-9. ELNJ has the highest traffic
volume and the seventh highest hydrocarbon average, but SEWA, SPIL, and PXSS, which have
the second, third, and fourth highest traffic volumes, have the  17th, 11th, and 4th highest
hydrocarbon averages, respectively. Again, CELA did not measure  VOCs  under the NMP. A
Pearson correlation coefficient calculated between the average hydrocarbon calculations and the
traffic counts is 0.28. While this correlation is not a "strong" correlation, it does indicate a
positive correlation between hydrocarbon concentrations and traffic volumes.
                                          4-38

-------
4.3.5   Vehicle Miles Traveled
       Another approach to determine how mobile sources affect urban air quality is to review
VMT. VMT is "the sum of distances traveled by all motor vehicles in a specified system of
highways for a given period of time" (OR DOT, 2012). Thus, VMT values tend to be large (in
the millions). In past NMP reports, daily VMT data from the Federal Highway Administration
(FHWA) were obtained. However, VMT was only available by urban area; thus, no VMT was
available for sites located in rural areas. Beginning with the 2010 NMP, county-level VMT was
obtained from state organizations, primarily departments of transportation. However, these data
are not readily available for all states. In addition, not all states provide this information on the
same level. For example, many states provide VMT for all public roads, while the state of
Colorado provided this information for state highways only. County-level VMT are presented in
Table 4-14, where available.

       The sites with the highest county-level VMT, where available, are CELA (Los  Angeles
County, CA), PXSS and SPAZ (Maricopa County, AZ), and SPIL and NBIL (Cook County, IL).
The sites with the lowest county-level VMT, where available, are CHSC (Chesterfield County,
SC), GLKY (Carter County, KY), and UCSD (Union County,  SD). A Pearson correlation
coefficient calculated between the average summed hydrocarbon concentrations and VMT,
where available, is 0.46, indicating a positive 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, RUCA, CAMS 35, and  SJJCA). In addition,
county-level VMT were not readily available for Rhode Island.

4.4    Variability Analysis
       This section presents the results of the three variability analyses described in
Section 3.4.2.

4.4.1   Coefficient of Variation and Inter-site Variability
       The site-specific CVs and the inter-site comparison analyses are  discussed together in
this section. Figures 4-la through 4-25a are graphical displays of site-specific CVs (standard
deviation vs. annual average concentration) for the program-level pollutants of interest.
Figures 4-lb through 4-25b are bar graphs depicting the site-specific annual averages overlain on

                                          4-39

-------
the program-level averages, as discussed in Section 4.1. For each program-level pollutant of
interest, the CV graph is shown first, followed by the inter-site variability graph. The figures are
aligned this way because they tend to complement each other; the data point with the highest
annual average concentration and/or standard deviation in the CV graph is easily identifiable in
the inter-site variability graph.  Further, the inter-site variability graphs allow the reader to see
how the individual site-specific annual averages feed into the program-level averages (i.e., if a
specific site(s) is driving the program average). In addition to the standard deviations on the CV
graphs, the confidence intervals provided on the inter-site variability graphs are a further
indication of the amount of variability contained within the site-specific annual averages.

       Several items to note about these figures:  Some sites do not have annual averages
presented on the inter-site variability graphs because they did not meet the criteria specified in
Section 3.1. These same sites without annual averages on the inter-site variability graphs are not
represented by a data point on the corresponding CV graphs. For the sites sampling metals, the
program-level average for sites collecting PMio samples is presented in green while the program-
level average for sites collecting TSP samples is presented in pink. The annual averages for the
sites sampling only SNMOCs are not included in the graphs for benzene, 1,3-butadiene, or
ethylbenzene.

       The CV figures show that few of the pollutants appear to exhibit the "clustering"
discussed in Section 3.4.2. Figure 4-10a for carbon tetrachloride exhibits clustering, or
uniformity in concentrations. Carbon tetrachloride is a pollutant that was used worldwide as a
refrigerant. However, it was  identified as an ozone-depleting substance in the stratosphere and its
use was banned at the Kyoto Protocol. This pollutant has a long lifetime in the atmosphere, but
slowly degrades over time. Today, its concentration  in ambient air is fairly ubiquitous regardless
of where it is measured. The CVs  shown in Figure 4-10a not only support the expected
uniformity (i.e., lack of variability) in "background" concentrations of carbon tetrachloride, but
are also a testament to the representativeness of the data generated under the NMP. Figure 4-1 Ob
supports what is shown in Figure 4-10a. The inter-site variability is relatively low, with the
annual average concentrations  of carbon tetrachloride ranging from 0.53 |ig/m3 for GPCO to
0.65 |ig/m3 for SEWA and DEMI. Further, the confidence intervals for all sites shown are less
than ฑ0.07 |ig/m3.

                                           4-40

-------
       Figure 4-13a shows that 1,2-dichloroethane also appears to exhibit clustering. However,
it should be noted that the y-axis scale for the standard deviation is three times larger than the
x-axis scale for the annual average concentration. This indicates that there is a relatively high-
level of variability in the individual concentrations, as supported in Figure 4-13b by the
relatively large confidence intervals shown for every site. In addition, this pollutant was detected
infrequently (only 25 percent of samples) and therefore has many zero substitutions factored into
each annual average, which contributes to both the low range of annual average concentrations
(all less than 0.035 |ig/m3) and the large confidence intervals.

       Other pollutants, such as hexchloro-1,3-butadiene and vinyl chloride, exhibit the same
large confidence intervals on the inter-site variability graphs as 1,2-dichloroethane due to a large
number of non-detects and the associate zero substitutions. Hexchloro-1,3-butadiene was
detected in fewer than 10 percent of samples, resulting in a large number of zero substitutions.
With the exception of the sites that did not detect this pollutant, the standard deviations are all
greater than the annual average for each site in Figure 4-17a, as shown by the values along the
x- and y-axes. Further, nearly every site has a relatively large confidence interval shown in
Figure 4-17b.  The largest confidence intervals were calculated for SPIL and MWOK; although
these two sites measured the two greatest concentrations of hexachloro-l,3-butadiene
(0.684 |ig/m3 for SPIL and 0.631 |ig/m3 for MWOK), this pollutant was detected only three times
at SPIL and four times at MWOK.

       Similarly, vinyl chloride is another infrequently detected pollutant (less than a 7 percent
detection rate) for which the annual averages have large standard  deviations in Figure 4-25a
(note how many of the standard deviations are more than twice the corresponding annual
averages) and large confidence intervals in Figure 4-25b. Although the concentrations of vinyl
chloride were at least an order of magnitude less than the concentrations of
hexchloro-1,3-butadiene, nearly half of the vinyl chloride measurements were greater than the
MDL while only two (the high concentrations for MWOK and SPIL) were greater than the MDL
for hexachloro-1,3 -butadiene.
                                           4-41

-------
       Several CVs for the program-level pollutants of interest follow a linear trend line.
Examples of pollutants whose annual average concentrations exhibit this trend include
acenaphthalene, acetaldehyde, benzene, fluorene, hexavalent chromium, lead, naphthalene, and
tetrachloroethylene. This means that as the annual averages increase, so do the standard
deviations, indicating increasing variability. This increased variability  is often a result of an
increased range of individual measurements that are used to calculate the annual average. This is
supported by the corresponding inter-site variability graphs for each pollutant. The site-specific
annual averages that extend well above the program-level average concentration for each
pollutant tend to have a wider confidence interval associated with them, indicating a wider range
of measurements and the possible influence of outliers. The annual averages considerably less
than the program-level average concentration tend to have much smaller confidence intervals.
Figures 4-la and 4-lb for acenaphthalene and Figures 4-15a and 4-15b for fluorene are good
examples of this trend.  The higher annual averages for sites such as DEMI, NBIL and ROCH
have large confidence intervals associated with them while sites such as CHSC, GLKY, and
UNVT have significantly lower annual averages as  well as very small confidence intervals.  To
illustrate this point, the range of measured detections of acenaphthene for DEMI was
0.698 ng/m3 to 111 ng/m3 while the range of measurements for GLKY was 0.0793 ng/m3 to
1.02 ng/m3.  DEMF sill ng/m3 measureme
acenaphthene measured across the program.
1.02 ng/m3.  DEMF sill ng/m3 measurement of acenaphthene was the highest measurement of
       Some of the pollutants' annual averages follow a linear pattern, but one of the annual
average concentrations is significantly higher than other annual average concentrations of the
other sites, one of the standard deviations is significantly higher than other sites, or both (such as
benzene, beryllium, and tetrachloroethylene). Figures 4-5a and 4-5b show that the annual
average benzene concentration for TOOK is more than twice the next highest annual average
concentration for this pollutant. A review of TOOK's benzene data shows that all but seven of
TOOK's preprocessed daily measurements (out of 57) were greater than the program-level
average concentration of 0.98 |ig/m3. Thus, concentrations of benzene at TOOK tend to run
higher than at other sites. Figure 4-23a shows that the magnitude of the standard deviation axis is
more than twice the annual average axis for tetrachloroethylene. Figure 4-23b shows that this is
the result of the PXSS data. The maximum concentration measured at PXSS (8.63 |ig/m3) is
more than three times higher than the next highest tetrachloroethylene measurement (2.42 |ig/m3,

                                         4-42

-------
measured at NBIL) and nearly five times higher than the next highest tetrachloroethylene
measurement at PXSS (1.74 |ig/m3). Thus, the annual average concentration of
tetrachloroethylene for PXSS is reflecting the influence of the outlying concentration.

       Although many of the other pollutants of interest do not exhibit easily classifiable
clustering or appear to follow a linear pattern, some of them are influenced by one or more data
points that do not fall in line with the others. For example, the larger standard deviation
(1.20 |ig/m3) exhibited for 1,3-butadiene in Figure 4-8a indicates that this particular annual
average is likely influenced by outlier(s). Figure 4-8b shows that this data point represents
OCOK's annual average.  Excluding this data point would allow the rest to follow a more linear
trend line. Although this site did not have the highest annual average concentration of
1,3-butadiene, the highest individual measurement  of this pollutant across the program was
measured at OCOK (9.51 jig/m3).  The next highest measurement of 1,3-butadiene at OCOK was
considerably less (approximately 0.25 |ig/m3).  The next highest measurement at the program
level was 2.68 |ig/m3, measured at NBIL.

       If the data point that represents SPIL's annual average and standard deviation was
removed and the scales adjusted, the trichloroethylene concentrations would appear to exhibit a
more linear trend, although some clustering would  still be shown among the sites with the lowest
annual averages. The annual average trichloroethylene concentration for SPIL
(0.64 ฑ 0.41 |ig/m3) is more than seven times the annual average trichloroethylene concentration
for any other site, as shown in both Figures 4-23 a and  4-23b. If the two highest concentrations
measured at SPIL (those greater than 3 |ig/m3) were removed from the calculation,  SPIL's
annual average would still be more than four times the next highest annual average (as calculated
for ELNJ).  SPIL's annual average trichloroethylene concentration is discussed in more detail in
Section 4.2.1.

       Acrylonitrile appears to exhibit clustering in Figure 4-3 a with the exception of the annual
average concentration for SPIL. The annual average concentration of acrylonitrile for SPIL
(0.77 |ig/m3) is nearly five times greater than the next highest concentration, as shown in
Table 4-9 and Figure 4-3b. Without this data point, the annual average concentrations of
acrylonitrile range from 0.01  |ig/m3 to 0.17 |ig/m3. However, this pollutant was detected  in fewer

                                          4-43

-------
than 25 percent of samples, resulting in a large number of zero substitutions; thus the standard
deviations for this pollutant tended to be rather large, as compared to the annual averages
themselves. This is evident by the number of sites whose standard deviations are larger than their
corresponding annual averages in Figure 4-3a.

       Chloroform appears to exhibit clustering in Figure 4-1 la. However, chloroform was
detected in fewer than 60 percent of samples. This can yield relatively low annual averages and
standard deviations, due in part to the substitution of zeros for many non-detects. If the data
point that represents NBIL's annual average, which is more than seven times greater than the
next highest annual average, was removed from Figure 4-1 la and the scales adjusted, most of the
chloroform concentrations still exhibit clustering. However, the three sites (DEMI, PXSS, and
S4MO) with annual averages just above and just below the program-level average concentration
of chloroform (0.39 |ig/m3) stand out more, as they do in Figure 4-1 Ib. While S4MO's annual
average concentration is less than the annual averages for DEMI and PXSS, S4MO's standard
deviation (1.50 |ig/m3) is relatively large compared to its annual average concentration
(0.35 ฑ 0.40 |ig/m3). This is due  to one particularly high measurement (11.5 |ig/m3), skewing the
data. This is illustrated not only by the confidence intervals shown in Figure 4-1 Ib but also by
the  site-specific annual average concentration comparison for chloroform shown in Table 4-9.
                                          4-44

-------
Figure 4-1 a. Coefficient of Variation Analysis of Acenaphthene Across 23 Sites
I15
 B
.a



I

1
•3
y=1.2448x-1.1506


   R1 = 0.9163
                                     6          8          10


                                     Annual Average Concentration (ng/m3)
               Figure 4-lb. Inter-Site Variability for Acenaphthene
 .a

 1 10

 s
 3
         n
                                                   il
     BOMA BTUT CAMS CELA  CH5C DEMI GLKY GPCO MONY NBIL PRRI PX5S RIVA ROCH RUCA S4MO SDGA 5EWA 5JJCA 5KFL 5YFL UNVT WADC


              35


                                            Monitoring Site
                        Program Average
                                                      nSite-SpecificAnnual Average
                                           4-45

-------
Figure 4-2a. Coefficient of Variation Analysis of Acetaldehyde Across 24 Sites
                                                             y=0.6977x-0.3503
                                                               ff- 0.7359
                                            n  o
                           1          1.5          2
                                  Annual Average Concentration (
              Figure 4-2b. Inter-Site Variability for Acetaldehyde
                                         Monrtoring Site
                      Program Average
                                                   DSite-Specific Annual Average
                                         4-46

-------
Figure 4-3a. Coefficient of Variation Analysis of Acrylonitrile Across 22 Sites
                                                      y=0.5078x +0.0733
                                                         R! = 0.6244
       oo
                                0.3       0.4        0.5
                                    Annual Average Concentration (
               Figure 4-3b. Inter-Site Variability for Acrylonitrile
"E
JO.6
E
a
ซ 0.5
s
3
&0.4
     Jฑt
                                     ?
                                           Monitoring Site

                       Program Average                    D Site-Specific Annual Average
                                          4-47

-------
   Figure 4-4a. Coefficient of Variation Analysis of Arsenic Across 14 Sites
                                                                     y=0.6659x + 0.0602

                                                                        ซ.'- 0.5519
I 0.3
                                                                y=0.2846x + 0.0947

                                                                   R2 = 0.5903
              0.1        0.2       0.3        0.4        0.5       0.6        0.7        0.8


                                        Annual Average Concentration (ng/m3)




                            O  PM10   O  TSP  	Linear(PMlO)  	Linear(TSP)
                    Figure 4-4b. Inter-Site Variability for Arsenic
S
E
a

•E 0.6
I
3
      BOMA   BTUT   GLKY    NBIL    PAFL   PXS5   S4MO  SEWA   SJJCA   UNVT   MWOK  OCOK   PROK   TMOK   TOOK


                                                Monitoring Site
              D Program PMlOAverage
                                          D Program TSP Average
                                                                    D Site-Specific Annual Average
                                              4-48

-------
   Figure 4-5a. Coefficient of Variation Analysis of Benzene Across 23 Sites
                                                             y= 1.0979X-0.4515
                                                               R! = 0.937
"I"
                                O
                                o
                                   1.5         2         2.5
                                   Annual Average Concentration (
                 Figure 4-5b. Inter-Site Variability for Benzene
               Hh

                                          Monitoring Site

                      Program Average                   D Site-Specific Annual Average
                                         4-49

-------
Figure 4-6a. Coefficient of Variation Analysis of Benzo(a)pyrene Across 23 Sites
I-
|
I
y=0.945Bxt 0.034
  R* = 0.5462
                                       0.1               0.15

                                     Annual Average Concentration (ng/m3)
               Figure 4-6b. Inter-Site Variability for Benzo(a)pyrene
  _ 0.2

  |

  a
  E
  •g 0.15
  I
  3
       BOMA BTUT CAMS CELA CHSC DEMI GLKY GPCO MONY NBIL  PRRI PXSS RIVA ROCH RUCA S4MO SDGA SEWA SJJCA SKFL  SYFL UNVT WADC
                35

                                             Monitoring Site
                         Program Average
                                                       nsite-SpecificAnnual Average
                                            4-50

-------
  Figure 4-7a. Coefficient of Variation Analysis of Beryllium Across 14  Sites
                                                                    y=1.2502x + 0.0024

                                                                       R! = 0.8486
I:
e
.a
                                       op--
                                                         V=0.8987jt-0.0052
     0       0.005      0.01      0.015       0.02      0.025      0.03      0.035       0.04      0.045      0.05

                                        Annual Average Concentration (ng/m5)



                             O  PM10    O  TSP  	Linear (PM10)  	Linear(TSP)
                   Figure 4-7b. Inter-Site Variability for Beryllium
I
= •
\
I

r
       BOMA   BTUT   GLKY    NBIL    PAFL   PXSS   S4MO   SEWA   SJJCA   UNVT   MWOK  OCOK   PROK   TMOK   TOOK

                                                Monitoring Site
              D Program PM10 Average
                                          D Program TSP Average
                                                                     D Site-Specific Annual Average
                                               4-51

-------
Figure 4-8a. Coefficient of Variation Analysis of 1,3-Butadiene Across 23 Sites
                                                    V=l.S921x- 0.0484
                                                       ff= 0.234
                                    0.15         0.2
                                 Annual Average Concentration (n
              Figure 4-8b. Inter-Site Variability for 1,3-Butadiene



"E
1
E
.a

s
3
S
0.1 -




























I
1

* d




i
1

/






[ฑ
c/0<





+

?








-------
  Figure 4-9a. Coefficient of Variation Analysis of Cadmium Across 14 Sites
                                                            y=0.7S72x+0.0072


                                                               R; = 0.9061
                                 ,-' v=10236x-0.052


                                      R= = 0.7595
                                   0.2             0.3             0.4



                                       Annual Average Concentration (ng/m3)
                            O  PM10    O TSP  	Linear(PMlO)   	Linear(TSP)
                  Figure 4-9b. Inter-Site Variability for Cadmium
T
E
a


=• 0.4

s
3


} 0.3
      BOMA   BTUT   GLKY   NBIL   PAFL    PXS5   S4MO   SEWA   SJJCA   UNVT   MWOK  OCOK  PROK  TMOK   TOOK



                                               Monitoring Site
              D Program PM10 Average
                                         D Program TSP Average
                                                                   D Site-Specific Annual Average
                                              4-53

-------
Figure 4-10a. Coefficient of Variation Analysis of Carbon Tetrachloride Across 23 Sites
                                0.2         0.3         0.4

                                       Annual Average Concentration (\.
                                                                            \ O
                                                                Y=-0.3B7Sx + 0.3687
                                                                   R! = 0.1949     Q  V!

                                                                              ^>V
                                                                              O  o
                                                                           O    O   O
              Figure 4-1 Ob. Inter-Site Variability for Carbon Tetrachloride
^
      .s

      •S 04
      I
      3
      &
rfl
                                                  -h
                                              Monitoring S te
                           Program Average
                                                       DSite-SpecificAnnual Average
                                             4-54

-------
Figure 4-lla. Coefficient of Variation Analysis of Chloroform Across 23 Sites
                                               y= 1.3197X-0.0043
                                                 Rz = 0.9761
                           2345
                                  Annual Average Concentration (ujj/m3)
              Figure 4-1 Ib. Inter-Site Variability for Chloroform
    •^  ^
   ^  ซ?•
                                         Monitoring Site
                      Program Average
                                                   D Site -Specific Annual Average
                                         4-55

-------
Figure 4-12a. Coefficient of Variation Analysis of/J-Dichlorobenzene Across 23 Sites
    &
    •E 0.2
                                       Y=0.9855x +0.0091

                                         R= = 0.6357
                                  0.1           0.15

                                      Annual Average Concentration (
              Figure 4-12b. Inter-Site Variability for/7-Dichlorobenzene
             rfi
               
-------
Figure 4-13a. Coefficient of Variation Analysis of 1,2-Dichloroethane Across 23 Sites
    i
                                                               y = 1.4603x +0.0109


                                                                  RJ = 0.3805
                                      0.015       0.02       0.025


                                       Annual Average Concentration (i^g/m3)
               Figure 4-13b. Inter-Site Variability for 1,2-Dichloroethane
    _ 0.04

    i

    1
    c
    a

                                              Monitoring Site
                          Program Average
                                                       DSite-SpecificAnnual Average
                                             4-57

-------
Figure 4-14a. Coefficient of Variation Analysis of Ethylbenzene Across 23 Sites
.2 0.8

•e
i
5 0.6
                                                V=0.7195x
                                                   R' = 0.2712
                      O    O
                               0.4            0.6            0.8

                                   Annual Average Concentration ((^g/m5)
              Figure 4-14b. Inter-Site Variability for Ethylbenzene
 1-
 E
 a
 •ฃ o.s
 s
 s
                                        Hh
                                                    rh

                                          Monitoring Site
                       Program Average
                                                   DSite-SpecificAnnual Average
                                         4-58

-------
  Figure 4-15a. Coefficient of Variation Analysis of Fluorene Across 23 Sites
                                                        y=1.259x-2.0842

                                                          R! = 0.8677
I
&
                                     6           S           10

                                     Annual Average Concentration (ng/m3)
                 Figure 4-15b. Inter-Site Variability for Fluorene
.S
e
•S 10
                            n
                                                                  rh
^^
                                                                                          rir
    BOMA BTUT CAMS CELA CHSC DEMI GLKY GPCO MONY NBIL  PRRI PXSS RIVA ROCH RUCA S4MO SDGA SEWA SJJCA SKFL SYFL  UNVT WADC

             35

                                            Monitoring Site
                        Program Average
                                                      D Site-Specific Annual Average
                                            4-59

-------
Figure 4-16a. Coefficient of Variation Analysis of Formaldehyde Across 24 Sites
                                                      ฅ= 0.854Ix-0.5856

                                                         R! = 0.5261
                             1.5       2       2.5       3

                                    Annual Average Concentration (|Jg/n
              Figure 4-16b. Inter-Site Variability for Formaldehyde
 _4

 ^E

 1
  c
 .e
                                          Monitoring Site
                       Program Average
                                                    DSite-SpecificAnnual Average
                                          4-60

-------
Figure 4-17a. Coefficient of Variation Analysis of Hexachloro-l,3-butadiene Across 23 Sites
                                           0.01             0.015

                                         Annual Average Concentration (ug/m5)
              Figure 4-17b. Inter-Site Variability for Hexachloro-l,3-butadiene
        _ 0.03
        i
        1
        g 0.025
        C

        I
        3 0.02
                             Program Average
                                                -  ^  ,*   ฃ
                                                 
-------
Figure 4-18a. Coefficient of Variation Analysis of Hexavalent Chromium Across 22 Sites
      J
      B
      .2
      I ฐ
      &
      •e
      i
      K
                                                               V=0.5671x +0.0051

                                                                  R! = 0.8347
                                              0.03         0.04

                                          Annual Average Concentration (ng/m3)
               Figure 4-18b. Inter-Site Variability for Hexavalent Chromium
      _ 0.06


      I
      3 0.04
                                                                 il
            BOMA BTUT CAMS CAMS CHSC DEMI GLKY GPCO HOWI MONY NBIL  PRRI  PXSS RIVA ROCH S4MO SDGA SEWA SKFL SYFL  UNVT WADC

                    35  85

                                                 Monitoring Site
                             Program Average
                                                          nsite-SpecificAnnual Average
                                                4-62

-------
    Figure 4-19a. Coefficient of Variation Analysis of Lead Across 14 Sites
r
                                                 y=0.6267x + 0.3349
                                                    R! = 0.8648
                                  o    ,,-6
                         00
                                 y=0.6718x- 0.7655
                                    R! = 0.995
                                  468
                                     Annual Average Concentration (ng/m3)

                           O  PM10    O TSP   	Linear(PMlO)  	Linear(TSP)
                    Figure 4-19b. Inter-Site Variability for Lead
\
= s
I
1
I'
                                                                         fn  rf
      BOMA   BTUT   GLKY   NBIL   PAFL   PXSS   S4MO   5EWA   SJJCA   UNVT   MWOK  OCOK  PROK   TMOK  TOOK
                                             Monitoring Site
             D Program PMlOAverage
                                        D Program TSP Average
                                                                  D Site-Specific Annual Average
                                             4-63

-------
Figure 4-20a. Coefficient of Variation Analysis of Manganese Across 14 Sites
!
                             y= 1.9349X-4.9954
                                R' = 0.7189
                                                                         y=0.5592x
                                                                            R== 0.994
                                 0--"CT
                                          15           20
                                      Annual Average Concentration (ng/m3)
                           O  PM10    O TSP   	Unear(PMlO)  	Linear(TSP)
                Figure 4-20b. Inter-Site Variability for Manganese
r
E
.2
E 20
I
3

I15
      BOMA   BTUT   GLKY   NBIL   PAFL   PX5S   S4MO   SEW A   SJJCA   UNVT   MWOK  OCOK   PROK   TMOK   TOOK
                                             Monitoring Site
             D Program PMlOAverage
                                        D Program TSP Average
                                                                 D Site-Specific Annual Average
                                             4-64

-------
Figure 4-21a. Coefficient of Variation Analysis of Naphthalene Across 23 Sites
"i,.
 B
a
i
•3  60
ฃ
                                                 y=0.6549x +2.2998
                             60       BO       100      120

                                Annual Average Concentration (ng/m5)
                                                              140       160
             Figure 4-21b. Inter-Site Variability for Naphthalene
Concentration (ng/m3)
H1 I-1 H1 H1 I-1 W
S S ฃ S S g
| so
ฅ
60 -
40
20
0













T
1


BO MA




i




































n





















n

















































































































































l—r—
T



T
ill






































































































BTUT CAMS CELA CHSC DEMI GLKY GPCO MONY NBIL PRRI PXSS RIVA ROCH RUC-A S4MO SDGA SEWA SJJCA SKFL




i

SYFL





fl























UNVT WADC
35
Monitoring Site



Pro|
ram Average
DSite-SpecificAnnual Average
                                      4-65

-------
Figure 4-22a. Coefficient of Variation Analysis of Nickel Across 14 Sites
1"
e

I  i
S
|
1 o.s
                                                       y=0.9749x-0.3606


                                                          R! = 0.8149
                                                                   o

                                                                 o-'''
                                                          y=0.4865x-0.0142


                                                             R; = 0.6739
          0.2       0.4       0.6        0.8        1        1.2       1.4        1.6        l.S


                                    Annual Average Concentration (ng/m3)




                         O PM10   O  TSP   	Linear(PMlO)  	Linear(TSP)
                Figure 4-22b. Inter-Site Variability for Nickel
   BOMA   BTUT    GLKY   NBIL    PAFL   PXS5   S4MO  SEWA   SJJCA   UNVT   MWOK  OCOK   PROK  TMOK   TOOK


                                            Monitoring Site
          D Program PM10 Average
                                      D Program TSP Average
                                                                D Site-Specific Annual Average
                                           4-66

-------
Figure 4-23a. Coefficient of Variation Analysis of Tetrachloroethylene Across 23 Sites
     1
                                           y=1.6099x-0.071
                                             R! = 0.8417
                                   0.2            0.3

                                       Annual Average Concentration (
              Figure 4-23b. Inter-Site Variability for Tetrachloroethylene
     ,-.0.6
     i
     1
     I 0.5


     i
     50.4

                                              Monitoring Site
                           Program Average
                                                       DSite-SpecificAnnual Average
                                             4-67

-------
Figure 4-24a. Coefficient of Variation Analysis of Trichloroethylene Across 23 Sites
   1,
   "E 0.8
   1
V= 2.3963X +0.0118
   R'- 0.9665
                               0.2          0.3          0.4          0.5
                                      Annual Average Concentration (ug/m5)
               Figure 4-24b. Inter-Site Variability for Trichloroethylene
    _0.8
    1
                                                                                   JฑL
                          Program Average

                                             Monitoring Site
                                                       D Site-Specific Annual Average
                                             4-68

-------
Figure 4-25a. Coefficient of Variation Analysis of Vinyl Chloride Across 23 Sites
Jo.oos

e
a



I
                                                           y=2.6421x


                                                              R2 = 0.9021
    0 i J


      0
                            0.001         0.0015        0.002         0.0025


                                     Annual Average Concentration (ug/m3)
               Figure 4-25b. Inter-Site Variability for Vinyl Chloride
  1
  7 0.004
  a



  I

  3 o.oos
                                             Monitoring Site
                         Program Average
                                                      D Site -Specific Annual Average
                                            4-69

-------
4.4.2   Quarterly Variability Analysis
       Figures 4-26 through 4-50 provide a graphical display of the site-specific quarterly
average concentrations for each of the program-level pollutants of interest. Quarterly averages
are calculated based on the criteria specified in Section 3.1. If the pollutant of interest has a
corresponding ATSDR Intermediate MRL, as defined in Section 3.3, then this value is indicated
on the graph and is plotted where applicable. The National Ambient Air Quality Standard
(NAAQS) for lead (TSP) is a 3-month standard. Because this time period aligns well with the
quarterly averages presented in this section, the NAAQS for lead (TSP) is also provided in
Figure 4-44a and 4-44b. Note that the scales on the PMio and TSP graphs are the same for a
given speciated metal.

       Data gaps, or missing quarterly averages, in the figures for the pollutants of interest can
be attributed to two reasons. First, some of the program-wide pollutants of interest were
infrequently detected in some quarters and thus have a quarterly average concentration of zero as
a result of the substitution of zeros for non-detects. One example of this is Figure 4-28 for
acrylonitrile.  This pollutant was infrequently detected (300 measured detections out of 1,281
valid samples); of the 85 possible quarterly averages of this pollutant, 22 of them are zero. Thus,
few quarterly averages appear in Figure 4-28. Further, most of the remaining quarterly averages
have relatively few measurements and include many zero substitutions for non-detects, resulting
in relatively low quarterly averages.  (Although this pollutant was detected in only 23 percent of
VOC  samples collected, its risk screening value is relatively low; thus, all 300 measured
detections of this pollutant failed screens.)

       Another reason for data gaps in the figures is due to the sampling duration of each site.
Some sites started late or ended early, which may result in a lack of quarterly averages. For
example, benzene is almost always detected in VOC samples, thus the gaps in Figure 4-30 are
primarily due to sampling duration. PANJ stopped sampling VOCs in May 2011; thus, the third
and fourth quarterly averages are blank. Because the criteria in Section 3.1 require a site to have
75 percent of the possible samples within a quarter (12 for a site sampling on a l-in-6 day
schedule), PAJN could not get a quarterly average for the second quarter because it did not
sample long enough within that quarter. Therefore, the only quarterly average that could be
calculated for PANJ was for the first quarter.

                                           4-70

-------
       Some pollutants of interest, such as acetaldehyde, benzene, carbon tetrachloride,
ethylbenzene, formaldehyde, and naphthalene, were detected year-round. Comparing the
quarterly averages for sites with four valid quarterly averages in a year may reveal a trend for
these pollutants. For example, formaldehyde averages tended to be highest in the third quarter, as
shown in Figure 4-41, with 18 of the 30 sites sampling formaldehyde (and have quarterly
averages) exhibiting the highest quarterly average during July through September. Thus, it
appears that formaldehyde concentrations tend to be highest during the summer months.
Conversely, benzene averages tended to be higher during the first quarter followed by the fourth
quarter, or the colder months, as shown in Figure 4-30. The seasonal behavior of benzene and
formaldehyde suggests the influence of reformulated gasoline (RFG) because the benzene
content is typically lowered during warmer periods (i.e., summer and spring). Refineries
typically  begin production of RFG during the spring and end in the autumn. Additionally,
methyl tert-buty\ ether (MTBE) is often used as an RFG additive in fuels to replace the lowered
benzene content. Research has shown that the combustion of fuels containing MTBE leads to the
secondary production of formaldehyde. Thus, while benzene concentrations decrease during the
summer months, formaldehyde concentrations may increase if MTBE is used in the gasoline
blend.

       Other notable trends include benzo(a)pyrene and 1,3-butaidene with higher
concentrations in the first and fourth quarters; acenaphthene and fluorene with higher
concentrations in the third quarter; and 1,2-dichloroethane with higher concentrations in the
second and fourth quarters.  Arsenic tended to be highest during the fourth quarter for eight of
the nine sites sampling PMio metals year-round  and four of the  five sites sampling TSP metals (a
fourth quarter average could not be calculated for MWOK because it did not meet the
completeness criteria).

       Other notable trends may also be revealed in these graphs. Figure 4-38 for
1,2-dichlorethane shows that most of the measured detections of this pollutant were measured
during the second and fourth quarter averages of 2011, as indicated by the red (second quarter)
and purple (fourth quarter) bars. Over  50 percent of the measured detections were measured
during the fourth quarter and another 28 percent were measured during the second quarter.
Figure 4-50 for vinyl chloride shows that this pollutant was infrequently detected,  as many sites

                                          4-71

-------
have fewer than four quarterly averages shown, even though they sampled year-round and met
the completeness criteria. Note that BTUT, RUVT, and SPAZ did not detect this pollutant at all.

       The quarterly average comparison also allows for the identification of sites with
unusually high concentrations of the pollutants of interest compared to other sites and when
those high concentrations were measured. This is evident in Figures 4-28, 4-30, 4-32a, 4-33,
4-36, and 4-49 for acrylonitrile, benzene, beryllium, 1,3-butadiene, chloroform, and
trichloroethylene, respectively, to name a few. For example, Figure 4-36 shows that the quarterly
averages of chloroform for NBIL are significantly higher than for other sites sampling VOCs, as
most of the other bars are barely visible on the graph. Figure 4-49 shows that the quarterly
averages of trichloroethylene for SPIL are significantly higher than for other sites sampling
VOCs. Figure 4-32a and 4-32b show that PXSS's third quarter average concentration of
beryllium is significantly higher than this site's other the quarterly averages as well as all other
sites sampling speciated metals (both PMi0 and TSP). Conversely, these graphs may also reveal
when there is very little variability in the quarterly averages across  other sites. Figure 4-35 for
carbon tetrachloride shows that the quarterly averages of this pollutant did not vary significantly
across the sites. Other pollutants may not exhibit such trends.

       These graphs also show that only 10 of the 25 program-level pollutants of interest have
ATSDR Intermediate MRLs. For the 10 that do, the quarterly average concentrations are
significantly less than their respective ATSDR Intermediate MRLs, generally by an order of
magnitude or more, which is also discussed in Section 4.2.2. In all 10 cases, the scale on the
graph is well below the ATSDR Intermediate MRL.
                                          4-72

-------
             Figure 4-26. Comparison of Average Quarterly Acenaphthene Concentrations
BOMA BTUT CAMS CELA CHSC DEMI  GLKY GPCO MONY NBIL  PRRI PXSS RIVA  ROCH RUCA S4MO SDGA SEWA SJJCA SKFL SYFL  UNVT WADC
                 11st Quarter
           Monitoring Site




• 2nd Quarter              • 3rd Quarter               14th Quarter

-------
Figure 4-27. Comparison of Average Quarterly Acetaldehyde Concentrations
   11st Quarter
          Monitoring Site




1 2nd Quarter              • 3rd Quarter
4th Quarter

-------
                    Figure 4-28. Comparison of Average Quarterly Acrylonitrile Concentrations
  0.9
  0.8
  0.7
•5 0.6
2
Q 0.5

&
2

| 0.4
&
v

  0.3
                       11st Quarter
           Monitoring Site

12nd Quarter                3rd Quarter              B4th Quarter

-------
        Figure 4-29a. Comparison of Average Quarterly Arsenic (PMio) Concentrations
BOMA        BTUT        GLKY        NBIL        PAFL        PXSS        S4MO



                                               Monitoring Site




            • IstQuarter              B2nd Quarter              B3rd Quarter
                                                                              SEWA
                                                                                         SJJCA
14th Quarter
                                                                                                    UNVT

-------
   Figure 4-29b. Comparison of Average Quarterly Arsenic (TSP) Concentrations
MWOK
                      OCOK
                                           PROK



                                        Monitoring Site
                                                                TMOK
                                                                                      TOOK
       11st Quarter
12nd Quarter
3rd Quarter
14th Quarter

-------
                             Figure 4-30. Comparison of Average Quarterly Benzene Concentrations
J^.
I


oo
               ATSDRIntermediateMRL=20
                              11st Quarter
         Monitoring Site




12nd Quarter              • 3rd Quarter              B4th Quarter

-------
            Figure 4-31. Comparison of Average Quarterly Benzo(a)pyrene Concentrations
BOMA BTUT CAMS CELA CHSC DEMI GLKY  GPCO MONY NBIL  PRRI PXSS RIVA  ROCH RUCA S4MO SDGA SEWA SJJCA SKFL SYFL UNVT WADC
           35
                11st Quarter
          Monitoring Site

12nd Quarter                3rd Quarter              B4th Quarter

-------
                        Figure 4-32a. Comparison of Average Quarterly Beryllium (PMio) Concentrations
          0.12
oo
o
                  BOMA
                              BTUT
                              11st Quarter
                                         GLKY
                                                    NBIL
                                                                                     S4MO
        PAFL       PXS5


          Monitoring Site



2nd Quarter              B3rd Quarter
                                                                                                SEWA
                                                                                                           5JJCA
14th Quarter
                                                                                                                      UNVT

-------
                        Figure 4-32b. Comparison of Average Quarterly Beryllium (TSP) Concentrations
oo
          0.12
           0.1
        Is0-08
        .a
        +ri

        2
        4-1
        c


        i

        3 3.06
        I
          0.04
          0.02
                       MWOK
                                              OCOK
                                                                    PROK



                                                                Monitoring Site
                                                                                         TMOK
                                                                                                               TOOK
                              11st Quarter
12nd Quarter
3rd Quarter
14th Quarter

-------
                          Figure 4-33. Comparison of Average Quarterly 1,3-Butadiene Concentrations
oo
to
                              11st Quarter
          Monitoring Site



12nd Quarter               3rd Quarter
4th Quarter

-------
                        Figure 4-34a. Comparison of Average Quarterly Cadmium (PMio) Concentrations
oo
                 BOMA        BTUT        GLKY        NBIL
IstQuarter
                                                             PAFL       PXSS



                                                               Monitoring Site




                                                      2nd Quarter              B3rd Quarter
                                                                                   S4MO       SEWA       SJJCA       UNVT
14th Quarter

-------
                        Figure 4-34b. Comparison of Average Quarterly Cadmium (TSP) Concentrations
oo
        .a
        s
        2

               0.1
             6E-16
              -0.1
                              11st Quarter
          Monitoring Site


12nd Quarter              B3rd Quarter              B4th Quarter

-------
                      Figure 4-35. Comparison of Average Quarterly Carbon Tetrachloride Concentrations
oo
                ATSDRIntermediate MRL= 200
                             11st Quarter
          Monitoring Site




1 2nd Quarter              • 3rd Quarter
4th Quarter

-------
                           Figure 4-36. Comparison of Average Quarterly Chloroform Concentrations
oo
          14
          12
        JTlO
        I  8

        I
        I

        ฃ
              i ATSDRIntermediateMRL=200 ng/m3
              !	                                ...j
                               1st Quarter
          Monitoring Site


2nd Quarter               3rd Quarter
4th Quarter

-------
                        Figure 4-37. Comparison of Average Quarterly /7-Dichlorobenzene Concentrations
oo
                                                                                       : ATSDRIntermediate MRL=1,000 |ag/m3
                              11st Quarter
          Monitoring Site




12nd Quarter               3rd Quarter              B4th Quarter

-------
                       Figure 4-38. Comparison of Average Quarterly 1,2-Dichloroethane Concentrations
oo
oo
                          11st Quarter
            Monitoring Site


12nd Quarter               B3rd Quarter
4th Quarter

-------
                          Figure 4-39. Comparison of Average Quarterly Ethylbenzene Concentrations
oo
VO
                ATSDR Intermediate MRL= 9,000 ug/m3
                              11st Quarter
          Monitoring Site


12nd Quarter                3rd Quarter              B4th Quarter

-------
                              Figure 4-40. Comparison of Average Quarterly Fluorene Concentrations
J^.
I
o
              BOMA BTUT CAMS CELA CHSC DEMI GLKY GPCO MONY NBIL  PRRI  PXSS RIVA  ROCH RUCA S4MO SDGA SEWA SJJCA  SKFL  SYFL  UNVT WADC
                         35
                               11st Quarter
           Monitoring Site

12nd Quarter                3rd Quarter               B4th Quarter

-------
Figure 4-41. Comparison of Average Quarterly Formaldehyde Concentrations
                                                              ATSDRIntermediateMRL =
    11st Quarter
          Monitoring Site




1 2nd Quarter              • 3rd Quarter
4th Quarter

-------
                    Figure 4-42. Comparison of Average Quarterly Hexachloro-l,3-butadiene Concentrations
          0.06
VO
to
                             11st Quarter
          Monitoring Site


12nd Quarter               3rd Quarter
4th Quarter

-------
            Figure 4-43. Comparison of Average Quarterly Hexavalent Chromium Concentrations
0.09
0.08
       ATSDR Intermediate MRL= 300 ng/m3
o.oi
     BOMA  BTUT CAMS CAMS CHSC DEMI GLKY GPCO HOWI MONY  NBIL  PRRI  PXSS  RIVA ROCH S4MO SDGA SEWA  SKFL  SYFL UNVT WADC
                35   85
                    11st Quarter
          Monitoring Site

12nd Quarter              • 3rd Quarter              • 4th Quarter

-------
                Figure 4-44a. Comparison of Average Quarterly Lead (PMio) Concentrations
12
    ;-"
    i  NAAQS for Lead = 150 ng/m3
       BOMA
                  BTUT
                             GLKY
                    11st Quarter
NBIL        PAFL        PXSS        S4MO

             Monitoring Site


  • 2nd Quarter              B3rd Quarter
                                                                                     SEWA
                                                                                                SJJCA
14th Quarter
                                                                                                           UNVT

-------
                  Figure 4-44b. Comparison of Average Quarterly Lead (TSP) Concentrations
  12
  10
      : NAAQS for Lead = 150 ng/m3
.a
4-1

2
              MWOK
                                    OCOK
                                                         PROK



                                                      Monitoring Site
                                                                              TMOK
                                                                                                    TOOK
                     11st Quarter
12nd Quarter
3rd Quarter
14th Quarter

-------
       Figure 4-45a. Comparison of Average Quarterly Manganese (PMio) Concentrations
BOMA
            BTUT
                       GLKY
             11st Quarter
NBIL        PAFL        PXSS        S4MO



             Monitoring Site




  • 2nd Quarter              B3rd Quarter
                                                                              SEWA
                                                                                         SJJCA
14th Quarter
                                                                                                    UNVT

-------
  Figure 4-45b. Comparison of Average Quarterly Manganese (TSP) Concentrations
MWOK
                      OCOK
                                           PROK



                                        Monitoring Site
                                                                TMOK
                                                                                      TOOK
        11st Quarter
12nd Quarter
3rd Quarter
14th Quarter

-------
                            Figure 4-46. Comparison of Average Quarterly Naphthalene Concentrations
          250
VO
oo
               BOMA BTUT CAMS CELA  CHSC  DEMI GLKY GPCO MONY  NBIL PRRI  PXSS RIVA ROCH RUCA S4MO SDGA SEWA SJJCA SKFL  SYFL  UIWT WADC

                          35
                               11st Quarter
           Monitoring Site



12nd Quarter               • 3rd Quarter              • 4th Quarter

-------
                          Figure 4-41 a. Comparison of Average Quarterly Nickel (PMio) Concentrations
          3.5
VO
          0.5
                ATSDRIntermediate MRL= 200 ng/m3
                  BOMA       BTUT       GLKY        NBIL        PAFL        PXSS       S4MO



                                                                 Monitoring Site




                              • IstQuarter               B2nd Quarter              B3rd Quarter
                                                                                                5EWA
                                                                                                            SJJCA
14th Quarter
                                                                                                                       UNVT

-------
                          Figure 4-47b. Comparison of Average Quarterly Nickel (TSP) Concentrations
          3.5
O
O
                ATSDRIntermediate MRL= 200 ng/mE
                      MWOK
                                            OCOK
                                                                  PROK


                                                               Monitoring Site
                                                                                       TMOK
                                                                                                             TOOK
                             11st Quarter
12nd Quarter
3rd Quarter
14th Quarter

-------
Figure 4-48. Comparison of Average Quarterly Tetrachloroethylene Concentrations
       11st Quarter
          Monitoring Site




12nd Quarter               3rd Quarter              B4th Quarter

-------
                        Figure 4-49. Comparison of Average Quarterly Trichloroethylene Concentrations
o
to
                ATSDR Intermediate MRL = 500
                              11st Quarter
          Monitoring Site



12nd Quarter              B3rd Quarter
4th Quarter

-------
                          Figure 4-50. Comparison of Average Quarterly Vinyl Chloride Concentrations
J^.
I
o
          0.009
          O.OOS
                 ATSDRIntermediate MRL= 80
                              11st Quarter
          Monitoring Site

12nd Quarter                3rd Quarter
4th Quarter

-------
4.5   Greenhouse Gases
       Table 4-15 presents the program-level average concentrations for the 10 GHGs measured
using Method TO-15, in descending order by GWP. As shown, most of the GHGs were detected
in nearly every sample collected (a total 1,281 valid VOC samples). Chloroform, bromomethane,
and 1,1,1-trichloroethane were the only pollutants detected in less than 95 percent of VOC
samples collected, although even these were detected in greater than 50 percent of samples.
Dichlorodifluoromethane  and dichlorotetrafluoroethane have the highest GWPs of the GHGs
measured by Method TO-15 (10,900 and  10,000 respectively),  while bromomethane and
dichloromethane have the lowest GWPs (5 and 8.7, respectively). Dichloromethane has the
highest program-level average concentration among the GHGs measured, although the
associated confidence interval indicates that this concentration  is likely influenced by outliers. A
review of the data shows that a single site contributed to this high average concentration. Three
concentrations of this pollutant greater than 100 |ig/m3 were measured at BTUT (ranging from
175 |ig/m3 to 349 |ig/m3).  An additional three concentrations greater than 20 |ig/m3 were
measured at BTUT (ranging from 22.4 |ig/m3 to 70 |ig/m3). Besides dichloromethane, only three
additional  GHGs shown in Table 4-15 have program-level average concentrations greater than
1 |ig/m3: dichlorodifluoromethane, trichlorofluoromethane, and chloromethane.
                                         4-104

-------
    Table 4-15. Greenhouse Gases Measured by Method TO-15
Pollutant
Dichlorodifluoromethane
Dichlorotetrafluoroethane
Trichlorotrifluoroethane
Trichlorofluoromethane
Carbon Tetrachloride
1,1,1 -Trichloroethane
Chloroform
Chloromethane
Dichloromethane
Bromomethane
Global
Warming
Potential1
(100 yrs)
10,900
10,000
6,130
4,750
1,400
146
31
13
8.7
5
Total # of
Measured
Detections
1,281
1,241
1,281
1,281
1,279
1,096
756
1,281
1,264
737
2011
Program
Average
(Hg/m3)
2.75
ฑ0.02
0.13
ฑ<0.01
0.74
ฑ0.01
1.58
ฑ0.02
0.60
ฑ0.01
0.05
ฑ<0.01
0.39
ฑ0.11
1.26
ฑ0.01
3.51
ฑ2.41
0.04
ฑ<0.01
:GWPs presented here are from the Intergovernmental Panel on Climate Change
(IPCC) Fourth Assessment Report (AR4) (IPCC, 2012).
                              4-105

-------
5.0    Sites in Arizona
       This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS and UATMP sites in Arizona, and integrates these
concentrations with emissions, meteorological, and risk information. Data generated by sources
other than ERG are not included in the data analyses contained in this report. Readers are
encouraged to refer to Sections  1 through 4 and the glossary (Appendix P) for detailed
discussions and definitions regarding the various data analyses presented below.

5.1    Site Characterization
       This section characterizes the Arizona monitoring sites by providing geographical and
physical information about the location of the sites and the surrounding areas. This information
is provided to give the reader insight regarding factors that may influence the air quality near the
sites and assist in the interpretation of the ambient monitoring measurements.

       The Arizona monitoring sites are located in Phoenix, Arizona. Figures 5-1  and 5-2 are
composite satellite images retrieved from ArcGIS Explorer showing the monitoring sites in their
urban locations. Figure 5-3 identifies nearby point source emissions locations by source
category, as reported in the 2008 NEI for point sources. Note that only sources within 10 miles
of the sites are included in the facility counts provided in Figure 5-3. A 10-mile boundary was
chosen to give the reader an indication of which emissions sources and emissions source
categories could potentially have a direct effect on the air quality at the monitoring sites. Further,
this boundary provides both the proximity of emissions sources to the monitoring sites as well as
the quantity of such sources within a given distance of the sites. Sources outside the 10-mile radii
are still visible on the map, but have been grayed out in order to show emissions sources just
outside the boundary. Table 5-1 provides supplemental geographical information such as land
use, location setting, and locational coordinates.
                                           5-1

-------
                                       Figure 5-1. Phoenix, Arizona (PXSS) Monitoring Site
to

-------
Figure 5-2. South Phoenix, Arizona (SPAZ) Monitoring Site

-------
Figure 5-3. NEI Point Sources Located Within 10 Miles of PXSS and SPAZ
          112'25'CTW       112'2Q'0*W
                                                112-10'0-W       112'5'Q-W
                 Legend
                       PXSS NATTS site
                                             Note: Due to facili'.v density and collocation, the total facilities
                                             displayed may not represent a\\ facilities within the area of interest.
                                             10 mile radius
                  •jjf   SPAZ UATMP site       County boundary

                  Source Category Group (No. of Facilities)
                   •fi   Aerospace/Aircraft Manufacturing (1)
                   4>   Aircraft Operations (39)
                   c   Chemical Manufacturing (1)
                   *   Electricity Generation via Combustion (4)
                   -&   Flexible Polyurethane Foam Production (1 )
                   R   Furniture Plant (3)
                   ?   Miscellaneous Commercial/Industrial (4)
                   M   Miscellaneous Manufacturing (1)
                   R   Rubber and Miscellaneous Plastics Products (1 )
                                             5-4

-------
                                  Table 5-1. Geographical Information for the Arizona Monitoring Sites
Site
Code

PXSS
SPAZ
AQS Code

04-013-9997
04-013-4003
Location

Phoenix
Phoenix
County

Maricopa
Maricopa
Micro- or
Metropolitan
Statistical Area
Phoenix-Mesa-
Glendale, AZ
MSA
Phoenix-Mesa-
Glendale, AZ
MSA
Latitude
and
Longitude

33.503731,
-112.095809
33.40316,
-112.07533
Land Use

Residential
Residential
Location
Setting

Urban/City
Center
Urban/City
Center
Additional Ambient Monitoring Information1
Haze, CO, SO2, NO, NO2, NOX, PAMS, O3,
Meteorological parameters, PM10, PM2 5, PM Coarse,
PM2 5 Speciation.
CO, PAMS, O3, Meteorological parameters, PM2 5,
PM10,PM Coarse.
:Data for additional pollutants are reported to AQS for these sites (EPA, 2012c); however, these data are not generated by ERG and are therefore not included in this report.
 BOLD ITALICS = EPA-designated NATTS Site

-------
       PXSS is located in central Phoenix. Figure 5-1 shows that PXSS is located in a highly
residential area on North 17th Avenue. The Grand Canal is shown along the bottom of
Figure 5-1. The monitoring site is approximately three-quarters of a mile east of 1-17 and 2 miles
north of I-10. Figure 5-2 shows that SPAZ is located near the intersection of West Tamarisk
Avenue and South Central Avenue in South Phoenix. SPAZ is bounded on the west side by
residential properties and commercial properties on the east side. SPAZ is located approximately
1 mile south of 1-17.

       SPAZ and PXSS are located approximately 7 miles apart. The majority of emissions
sources are located between the sites, to the south of PXSS and north of SPAZ, as shown in
Figure 5-3. The source category with the greatest number of emissions sources near these
monitoring sites is the aircraft operations source category, which includes airports as well as
small runways, heliports, or landing pads. The emissions source nearest PXSS is a landing strip
at a hospital while the source nearest SPAZ is a landing strip at a police station.

       Table 5-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Arizona monitoring sites. Table 5-2 includes county-level
population and vehicle registration information. Table 5-2 also includes a county-level vehicle
registration-to-population ratio, which was calculated to represent the number of vehicles per
person within each monitoring site's residing county. In addition, the population within 10 miles
of each site is presented, based on postal code population data estimates. An estimate of 10-mile
vehicle ownership was determined by applying the county-level vehicle registration-to-
population ratio to the 10-mile population surrounding each monitoring site. Table 5-2 also
contains traffic volume information for each site. Finally, Table 5-2 presents the county-level
daily VMT for Maricopa County.
                                           5-6

-------
Table 5-2. Population, Motor Vehicle, and Traffic Information for the Arizona Monitoring
                                          Sites
Site
PXSS
SPAZ
Estimated
County
Population1
3,880,244
County-level
Vehicle
Registration2
3,776,819
Vehicles per
Person
(Registration:
Population)
0.97
Population
within 10
miles3
1,350,735
848,821
Estimated
10-mile
Vehicle
Ownership
1,314,732
826,196
Annual
Average
Daily
Traffic4
184,000
128,000
County-
level Daily
VMT5
89,448,000
1 County-level population estimate reflects 2011 data from the U.S. Census Bureau (Census Bureau, 2012b)
2 County-level vehicle registration reflects 2011 data from the Arizona DOT (AZ DOT, 201 la)
3 10-mile population estimates reflect 2011 data from Xionetic (Xionetic, 2011)
4 AADT reflects 2010 data from the Arizona DOT (AZ DOT, 2010)
5 County-level VMT reflects 2010 data for all public roads from the Arizona DOT (AZ DOT, 201 Ib)
BOLD ITALICS = EPA-designated NATTS Site
       Observations from Table 5-2 include the following:

       •  Maricopa County has the fourth highest county-level population and second highest
          county-level vehicle registration compared to other counties with NMP sites.

       •  The vehicle-per-person ratio is just less than one vehicle per person. This ratio falls
          within the top third compared to other NMP sites.

       •  The 10-mile population and estimated vehicle ownership are higher near PXSS than
          SPAZ.

       •  PXSS experiences a higher traffic volume compared to SPAZ, based on locations
          along 1-17 (between exits 195B  and  196 for PXSS and between exits 202 and 203 for
          SPAZ). The traffic volume near PXSS is the fifth highest compared to traffic volumes
          near other NMP sites, with the traffic volume near SPAZ ranking tenth.

       •  The daily VMT for Maricopa County is the second highest compared to other
          counties with NMP sites (where VMT data were available).


5.2    Meteorological Characterization

       The following sections characterize the meteorological conditions near the monitoring

sites in Arizona on sample days, as well as over the course of the year.
5.2.1   Climate Summary

       The Phoenix area 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 50ฐF. A

summer "monsoon" period brings precipitation to the area for part of the summer, while storms
                                           5-7

-------
originating off the Pacific Ocean bring rain in the winter and early spring. Winds are generally
light (Bair, 1992, and WRCC, 2013).

5.2.2   Meteorological Conditions in 2011
       Hourly meteorological data from the NWS weather station nearest these sites were
retrieved for 2011 (NCDC, 2011). The closest weather station to PXSS and SPAZ is located at
Phoenix Sky Harbor International Airport (WBAN 23183). Additional information about the Sky
Harbor weather station, such as the distance between the sites and the weather station, is
provided in Table 5-3. These data were used to determine how meteorological conditions on
sample days vary from conditions experienced throughout the year.

       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 the entire year for 2011. Also included in Table 5-3 is the
95 percent confidence interval for each parameter. As shown in Table 5-3, average
meteorological conditions on sample days were representative of average weather conditions
throughout the year. The greatest difference between the full-year averages and  sample day
averages is for the maximum and average temperatures for PXSS. This may be due to make-up
samples; thirteen of the nineteen make-up samples were collected between May and October
2011, during the warmer months of the year. Table 5-3 also shows that these sites experienced
the lowest relative humidity levels among NMP sites.

-------
Table 5-3. Average Meteorological Conditions near the Arizona Monitoring Sites
Closest NWS
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
Average
Temperature
Average
Dew Point
Temperature
Average
Wet Bulb
Temperature
Average
Relative
Humidity
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Phoenix, Arizona - PXSS
Phoenix Sky
Harbor Intl.
Airport
23183
(33.44, -111.99)
7.19
miles
136ฐ
(SE)
Sample
Day
2011
87.5
ฑ4.0
86.2
ฑ1.7
76.6
ฑ3.8
75.3
ฑ 1.7
35.6
ฑ3.1
34.8
ฑ1.5
55.5
ฑ2.3
54.7
ฑ1.0
28.6
ฑ3.7
28.8
ฑ1.6
1010.8
ฑ1.2
1011.3
ฑ0.6
5.5
ฑ0.5
5.3
ฑ0.2
South Phoenix, Arizona - SPAZ
Phoenix Sky
Harbor Intl.
Airport
23183
(33.44, -111.99)
5.46
miles
70ฐ
(ENE)
Sample
Day
2011
86.2
ฑ6.0
86.2
ฑ1.7
75.2
ฑ5.8
75.3
ฑ1.7
35.1
ฑ4.7
34.8
ฑ 1.5
54.8
ฑ3.4
54.7
ฑ1.0
30.0
ฑ6.1
28.8
ฑ 1.6
1011.4
ฑ1.9
1011.3
ฑ0.6
5.6
ฑ0.7
5.3
ฑ0.2
Sample day averages are highlighted to help differentiate the sample day averages from the full-year averages.

-------
5.2.3   Back Trajectory Analysis

       Figure 5-4 is the composite back trajectory map for days on which samples were

collected at the PXSS monitoring site in 2011. Included in Figure 5-4 are four back trajectories

per sample day. Figure 5-5 is the corresponding cluster analysis. Similarly, Figures 5-6 and 5-7

are the composite back trajectory map and corresponding cluster analysis for days on which

samples were collected at SPAZ. An in-depth description of these maps and how they were

generated is presented in Section 3.5.2.1. For the composite maps, each line represents the

24-hour trajectory along which a parcel of air traveled toward the monitoring site on a given

sample day and time, based on an initial height of 50 meters AGL. For the cluster analyses, each

line corresponds to a trajectory representative of a given cluster of back trajectories. Each

concentric circle around the sites in Figures 5-4 through 5-7 represents 100 miles.


       Observations from Figures 5-4 and 5-5 for PXSS include the following:

       •  The 24-hour air shed domain for PXSS is the smallest in size, based on average back
          trajectory length, compared to other NMP sites. The farthest away a back trajectory
          originated from PXSS was over northern Utah, or just greater than 500 miles away.
          However, most trajectories (89 percent) originated less than 300 miles from PXSS
          and the average trajectory length was approximately 166 miles.

       •  Back trajectories originated from a variety of directions at PXSS, although many
          trajectories originated from the southwest and west. A secondary group of trajectories
          originated from the north to northeast of the site. Back trajectories also originated
          from the east of the site.

       •  The cluster analysis map supports the observations above regarding the direction of
          trajectory origin as well  as the observations about trajectory distances. Nearly half
          (43 percent) of back trajectories originated to the southwest and west of PXSS, over
          southwest Arizona, southern California, and Baja California, Mexico. The short
          cluster trajectory (40 percent) represents back trajectories originating from nearly all
          directions, but generally less than 200  miles in length.


       Observations from Figures 5-6 and 5-7 for SPAZ include the following:

       •  Samples were collected every 12 days at SPAZ, which is half the frequency of sample
          collection at PXSS. As a result, fewer trajectories are shown in Figure 5-6 than
          Figure 5-4.

       •  The composite trajectory map for SPAZ has a trajectory distribution pattern similar to
          PXSS. The cluster analysis maps are also similar to each other. One difference,
          however, is that for SPAZ, the shorter trajectories of varying directions are included
          with the back trajectories originating from the north and east (as represented by the 42
          percent cluster),  while they are represented by a separate cluster trajectory for PXSS.

                                           5-10

-------
     Figure 5-4. 2011 Composite Back Trajectory Map for PXSS
       Figure 5-5. Back Trajectory Cluster Map for PXSS
\    \ป   \
                           5-11

-------
Figure 5-6. 2011 Composite Back Trajectory Map for SPAZ
    Figure 5-7. Back Trajectory Cluster Map for SPAZ
                        5-12

-------
5.2.4   Wind Rose Comparison
       Hourly surface wind data from the NWS weather station at Phoenix Sky Harbor
International Airport were uploaded into a wind rose software program to produce customized
wind roses, as described in Section 3.5.2.2. A wind rose shows the frequency of wind directions
using "petals" positioned around a 16-point compass, and uses different colors to represent wind
speeds.

       Figure 5-8 presents a map showing the distance between the NWS station and PXSS,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 5-8 also presents three different wind roses for the
PXSS monitoring site. First, a historical wind rose representing 2001 to 2010 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 2011 is presented.
Next, a wind rose representing wind observations for days on which samples were collected in
2011 is presented. These can be  used to identify the predominant wind speed and direction for
2011 and determine if wind observations on sample days were representative of conditions
experienced over the entire year and historically. Figure 5-9 presents the distance map and three
wind roses for SPAZ.
                                          5-13

-------
Figure 5-8. Wind Roses for the Phoenix Sky Harbor International Airport Weather Station
                                     near PXSS
   Distance between PXSS and NWS Station
2001-2010 Historical Wind Rose
                                                                              WIND SPEED
                                                                              (Knots)
                                                                              o *=
                                                                              ^| 17 • 21
                                                                              ^| 11 - 17
                                                                              ^| 7- 11
                                                                              CH 4-7
                                                                              !• 2- 4
                                                                              Calms: 15.34%
            2011 Wind Rose
    Sample Day Wind Rose
                                        5-14

-------
Figure 5-9. Wind Roses for the Phoenix Sky Harbor International Airport Weather Station
                                      near SPAZ
   Distance between SPAZ and NWS Station
2001-2010 Historical Wind Rose
   :   !          V
            2011 Wind Rose
                                                        Sample Day Wind Rose
                                 Calms: 13.92%
                           WIND SPEED
                           (Knots)

                           n ~*
                           ^| 17 • 21
                           JH 11 - 17
                           ^| 7- 11
                           n 4^7
                           !• 2- 4
                           Calms; 18.56^
                                         5-15

-------
       Observations from Figures 5-8 and 5-9 for the Arizona monitoring sites include the
following:
       •  The NWS 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 approximately 7.2 miles southeast of PXSS and 5.5 miles east-northeast of
          SPAZ.
       •  Because the Phoenix Sky Harbor weather station is the closest weather station to both
          sites, the historical and 2011 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 20 percent of observations), followed by
          westerly and east-southeasterly winds. Winds from the northwest, north, and
          northeast were infrequently observed, as were winds from the south. Calm winds
          (< 2 knots) account for 15 percent of the hourly wind measurements from 2001 to
          2010.
       •  The 2011 wind patterns are similar to the historical wind patterns, with just slightly
          more calm wind observations (nearly 19 percent). Further, the sample day wind
          patterns for each site resemble the historical and 2011 wind patterns, indicating that
          wind conditions  on sample days were representative  of those experienced over the
          entire year and historically.

5.3    Pollutants of Interest
       Site-specific "pollutants of interest" were determined for the Arizona monitoring sites in
order to allow 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. In
addition, if any of the NATTS MQO Core Analytes measured by each monitoring site did not
meet the pollutant of interest criteria based on the preliminary risk-based screening, that pollutant
was added to the list of site-specific pollutants of interest.  A more in-depth description of the
risk-based screening process is presented in Section 3.2.

       Table 5-4 presents the results of the preliminary risk-based screening process for PXSS
and SPAZ. The pollutants that failed at least one screen and contributed to 95 percent of the total
failed screens for each monitoring site are shaded. NATTS MQO Core Analytes are bolded.
                                          5-16

-------
Thus, pollutants of interest are shaded and/or bolded. PXSS sampled for VOCs, carbonyl
compounds, PAHs, metals (PMio), and hexavalent chromium; SPAZ sampled for VOCs only.
        Table 5-4. Risk-Based Screening Results for the Arizona Monitoring Sites
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Phoenix, Arizona - PXSS
Benzene
Carbon Tetrachloride
1,3-Butadiene
Manganese (PM10)
Arsenic (PM10)
Naphthalene
Acetaldehyde
ฃ>-Dichlorobenzene
Formaldehyde
Ethylbenzene
Hexavalent Chromium
1 ,2-Dichloroethane
Nickel (PM10)
Acrylonitrile
Hexachloro- 1 ,3 -butadiene
1 , 1 ,2,2-Tetrachloroethane
Chloromethylbenzene
Benzo(a)pyrene
Cadmium (PM10)
1 ,2-Dibromoethane
Dichloromethane
Propionaldehyde
Tetrachloroethylene
0.13
0.17
0.03
0.005
0.00023
0.029
0.45
0.091
0.077
0.4
0.000083
0.038
0.0021
0.015
0.045
0.017
0.02
0.00057
0.00056
0.0017
7.7
0.8
3.8
Total
61
61
56
55
53
50
48
48
48
44
14
12
12
11
5
3
2
1
1
1
1
1
1
589
61
61
56
61
61
57
48
55
48
61
62
12
61
11
6
o
J
2
30
61
1
61
48
56
983
100.00
100.00
100.00
90.16
86.89
87.72
100.00
87.27
100.00
72.13
22.58
100.00
19.67
100.00
83.33
100.00
100.00
3.33
1.64
100.00
1.64
2.08
1.79
59.92
10.36
10.36
9.51
9.34
9.00
8.49
8.15
8.15
8.15
7.47
2.38
2.04
2.04
1.87
0.85
0.51
0.34
0.17
0.17
0.17
0.17
0.17
0.17
10.36
20.71
30.22
39.56
48.56
57.05
65.20
73.34
81.49
88.96
91.34
93.38
95.42
97.28
98.13
98.64
98.98
99.15
99.32
99.49
99.66
99.83
100.00

South Phoenix, Arizona - SPAZ
Benzene
1,3-Butadiene
Carbon Tetrachloride
ฃ>-Dichlorobenzene
Ethylbenzene
1,2-Dichloroethane
Trichloroethylene
Xylenes
Acrylonitrile
Hexachloro- 1 ,3 -butadiene
0.13
0.03
0.17
0.091
0.4
0.038
0.2
10
0.015
0.045
Total
31
30
30
24
24
7
3
3
1
1
154
31
30
30
29
31
7
12
31
1
1
203
100.00
100.00
100.00
82.76
77.42
100.00
25.00
9.68
100.00
100.00
75.86
20.13
19.48
19.48
15.58
15.58
4.55
1.95
1.95
0.65
0.65
20.13
39.61
59.09
74.68
90.26
94.81
96.75
98.70
99.35
100.00

                                       5-17

-------
Observations from Table 5-4 include the following:

•  The number of pollutants failing screens varied significantly between the two
   monitoring sites; this is expected given the different pollutants measured at each site.

•  Twenty-three pollutants failed at least one screen for PXSS, of which 13 are NATTS
   MQO Core Analytes.

•  Thirteen pollutants, of which  10 are NATTS MQO Core Analytes, were initially
   identified as pollutants of interest for PXSS. Benzo(a)pyrene, cadmium, and
   tetrachloroethylene were added to the pollutants of interest for PXSS because they are
   NATTS MQO Core Analytes, even though they did not contribute to 95 percent of
   the total failed screens. Five additional NATTS MQO Core Analytes were added to
   the pollutants of interest for PXSS, even though their concentrations did not fail any
   screens: beryllium, chloroform, lead, trichloroethylene, and vinyl chloride. These five
   pollutants are not shown in Table 5-4 but are shown in subsequent tables in the
   sections that follow.

•  For PXSS, approximately 60 percent of the measured detections failed screens (of
   those pollutants failing at least one screen).

•  PXSS failed the second highest number of screens (589) among all NMP sites, behind
   only S4MO with 602 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 at this site, as discussed in Section 4.2.

•  Ten pollutants failed screens for SPAZ, of which four are NATTS MQO Core
   Analytes. Eight pollutants were initially identified as pollutants of interest for SPAZ.
   Chloroform and tetrachloroethylene were added to the pollutants of interest for SPAZ
   because they are NATTS MQO Core Analytes, even though their concentrations did
   not fail any screens. These two pollutants are not shown in Table 5-4 but are shown in
   subsequent tables in the sections that follow. While vinyl chloride is also a NATTS
   MQO Core Analyte, it was not detected at SPAZ, and therefore not added to the list
   of pollutants of interest.

•  For SPAZ, nearly 76 percent of the measured detections failed screens (of the
   pollutants failing at least one screen).

•  Of the VOCs, which were measured by Method  TO-15 at both sites, the following
   pollutants of interest failed  100 percent of screens for both sites: acrylonitrile,
   benzene, 1,3-butadiene, carbon tetrachloride, and 1,2-dichloroethane.
   Hexachloro-1,3-butadiene also failed 100 percent of screens for SPAZ.  Acetaldehyde,
   chloromethylbenzene, 1,2-dibromoethane, formaldehyde, and
   1,1,2,2-tetrachloroethane failed 100 percent of screens at PXSS. However, many of
   these pollutants were detected infrequently.
                                   5-18

-------
5.4    Concentrations
       This section presents various concentration averages used to characterize pollution levels
at the Arizona monitoring sites. Where applicable, the following calculations and data analyses
were performed: Time period-based concentration averages (quarterly and annual) are provided
for the pollutants of interest for the Arizona monitoring sites, where the data meet the applicable
criteria. Concentration averages for select pollutants are also presented graphically for the sites to
illustrate how the sites' concentrations compare to the program-level averages, as presented in
Section 4.1. In addition, concentration averages for select pollutants are presented from previous
years of sampling in order to characterize concentration trends at the sites.  Additional site-
specific statistical summaries for PXSS and SPAZ are provided in Appendices J, L, M, N, and O.

5.4.1   2011 Concentration Averages
       Quarterly and annual concentration averages were calculated for the pollutants of interest
for each Arizona 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 of the total number of samples
possible within a given quarter for a quarterly average to be calculated. An annual average
includes all measured detections and substituted zeros for non-detects for the entire year of
sampling. Annual averages were calculated for pollutants where three valid quarterly averages
could be calculated and where method completeness was greater than or equal to 85 percent, as
presented in Section 2.4.  Quarterly and annual average concentrations for the Arizona
monitoring sites are presented in Table 5-5, where applicable. Note that concentrations of the
PAHs, metals, and hexavalent chromium for PXSS are presented in ng/m3 for ease of viewing.
Also note that if a pollutant was not detected in a given calendar quarter, the quarterly average
simply reflects "0" because only zeros substituted for non-detects were factored into the
quarterly average concentration.
                                           5-19

-------
 Table 5-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
                                 the Arizona Monitoring Sites
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Hg/m3)
2nd
Quarter
Average
(Hg/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Hg/m3)
Phoenix, Arizona - PXSS
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
ฃ>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
Arsenic (PM10)a
Benzo(a)pyrene a
Cadmium (PM10) a
Bery Ilium (PM10)a
Hexavalent Chromium3
Lead(PM10)a
Manganese (PM10) a
Naphthalene a
Nickel (PM10)a
48/48
61/61
56/61
61/61
57/61
55/61
12/61
61/61
48/48
56/61
16/61
5/61
61/61
30/57
61/61
61/61
62/62
61/61
61/61
57/57
61/61
NA
1.70
ฑ0.38
0.32
ฑ0.10
0.58
ฑ0.02
0.28
ฑ0.06
0.27
ฑ0.06
0.01
ฑ0.02
0.90
ฑ0.23
NA
0.41
ฑ0.10
0.04
ฑ0.02
O.01
ฑ<0.01
0.68
ฑ0.33
0.16
ฑ0.09
0.15
ฑ0.07
0.02
ฑ0.01
0.08
ฑ0.02
4.02
ฑ1.30
11.21
ฑ2.96
107.68
ฑ 27.68
0.92
ฑ0.19
2.26
ฑ0.60
0.73
ฑ0.16
0.08
ฑ0.03
0.55
ฑ0.07
0.31
ฑ0.16
0.10
ฑ0.04
0.05
ฑ0.04
0.51
ฑ0.17
3.58
ฑ0.73
0.83
ฑ1.35
0.02
ฑ0.02
O.01
ฑ<0.01
0.46
ฑ0.21
O.01
ฑ0.01
0.08
ฑ0.02
0.03
ฑ0.01
0.07
ฑ0.03
2.92
ฑ1.14
14.89
ฑ3.21
61.28
ฑ38.49
1.42
ฑ0.29
2.66
ฑ0.53
0.83
ฑ0.20
0.09
ฑ0.04
0.66
ฑ0.05
0.43
ฑ0.17
0.13
ฑ0.05
0
0.59
ฑ0.13
4.29
ฑ0.41
0.21
ฑ0.07
0.01
ฑ0.01
O.01
ฑ<0.01
0.95
ฑ0.37
O.01
ฑ0.01
0.15
ฑ0.07
0.11
ฑ0.05
0.05
ฑ0.02
4.09
ฑ1.00
48.14
ฑ21.67
50.61
ฑ 16.97
2.59
ฑ0.84
3.52
ฑ0.80
1.98
ฑ0.49
0.39
ฑ0.13
0.72
ฑ0.05
0.43
ฑ0.13
0.28
ฑ0.08
0.03
ฑ0.02
1.20
ฑ0.29
4.41
ฑ0.77
0.78
ฑ0.26
0.02
ฑ0.02
0
1.00
ฑ0.30
0.21
ฑ0.10
0.18
ฑ0.06
0.03
ฑ0.01
0.06
ฑ0.02
5.19
ฑ1.64
15.37
ฑ4.79
125.83
ฑ29.09
1.97
ฑ0.75
NA
1.34
ฑ0.22
0.23
ฑ0.06
0.63
ฑ0.03
0.37
ฑ0.07
0.20
ฑ0.04
0.02
ฑ0.01
0.82
ฑ0.13
NA
0.55
ฑ0.28
0.02
ฑ0.01
O.01
ฑO.01
0.77
ฑ0.16
0.10
ฑ0.04
0.14
ฑ0.03
0.05
ฑ0.02
0.06
ฑ0.01
4.06
ฑ0.64
22.82
ฑ6.79
89.36
ฑ 16.23
1.74
ฑ0.33
NA = Not available due to the criteria for calculating a quarterly and/or annual average
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
                                              5-20

-------
 Table 5-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
                        the Arizona Monitoring Sites (Continued)
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Hg/m3)
2nd
Quarter
Average
(Hg/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Hg/m3)
South Phoenix, Arizona - SPAZ
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
1 ,2-Dichloroethane
Ethylbenzene
ฃ>-Dichlorobenzene
Tetrachloroethylene
Trichloroethylene
Xylenes
31/31
30/31
30/31
26/31
7/31
31/31
29/31
28/31
12/31
31/31
2.33
ฑ1.26
0.46
ฑ0.32
0.64
ฑ0.14
0.22
ฑ0.06
0.01
ฑ0.03
1.22
ฑ0.69
0.37
ฑ0.16
0.37
ฑ0.16
0.11
ฑ0.08
4.76
ฑ2.72
0.76
ฑ0.30
0.12
ฑ0.06
0.49
ฑ0.24
0.18
ฑ0.06
0.03
ฑ0.04
0.51
ฑ0.23
0.12
ฑ0.07
0.16
ฑ0.06
0.05
ฑ0.08
1.86
ฑ0.89
0.84
ฑ0.29
0.10
ฑ0.05
0.64
ฑ0.09
0.17
ฑ0.13
0.01
ฑ0.02
0.66
ฑ0.23
0.23
ฑ0.18
0.09
ฑ0.07
0.01
ฑ0.02
2.33
ฑ0.94
2.36
ฑ0.68
0.41
ฑ0.13
0.71
ฑ0.10
0.16
ฑ0.08
0.04
ฑ0.04
1.65
ฑ0.57
0.03
ฑ0.09
0.89
ฑ0.38
0.03
ฑ0.04
6.45
ฑ2.38
1.65
ฑ0.45
0.29
ฑ0.10
0.63
ฑ0.07
0.18
ฑ0.04
0.02
ฑ0.01
1.06
ฑ0.28
0.26
ฑ0.07
0.41
ฑ0.16
0.05
ฑ0.03
4.06
ฑ1.14
NA = Not available due to the criteria for calculating a quarterly and/or annual average
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.

       Observations for PXSS from Table 5-5 include the following:

       •  The pollutant with the highest annual average concentration by mass is benzene. This
          is the only pollutant with an annual average concentration greater than 1 |ig/m3
          (1.34 ฑ 0.22 |ig/m3). The quarterly average concentrations of benzene exhibit a
          seasonal trend, with higher quarterly averages for the colder months of the year.

       •  Similarly, concentrations of benzo(a)pyrene,  1,3-butadiene, and/>-dichlorobenzene
          are also higher during the colder months of the year. Ethylbenzene and naphthalene
          also appear to exhibit this trend, but the difference in the quarterly  average
          concentrations is not statistically significant for these pollutants.

       •  The confidence interval for the second quarter average concentration of
          tetrachloroethylene is higher than the average itself, indicating the  likely influence of
          outliers. The highest concentration of this pollutant was measured at PXSS on
          May 9, 2011 (8.63 |ig/m3) and is nearly five times greater than the  next highest
          concentration of this pollutant measured at PXSS (1.74 |ig/m3) and more than three
          times the next highest concentration measured among NMP sites sampling VOCs
          (2.42 |ig/m3, measured at NBIL).
                                           5-21

-------
•  The third quarter average concentration of manganese is three times higher than the
   other quarterly averages and has a relatively large confidence interval associated with
   it. Three of the five concentrations of manganese greater than 100 ng/m3 measured
   across the program were measured at PXSS. All three of these measurements were
   measured during the third quarter of 2011, one in each month and ranged from
   111 ng/m3to 130 ng/m3.

•  Regarding hexavalent chromium measurements at PXSS: stainless steel filter holders
   used in the hexavalent chromium sampler at PXSS may have contaminated the
   samples collected at this site. The filter holder was exchanged with a Teflonฎ filter
   holder at the end of February 2011. Although the first quarter average is the highest
   quarterly average for 2011, the difference is not statistically significant and the
   maximum concentration of hexavalent chromium was not measured during the first
   quarter. Of the nine concentrations greater than 0.1 ng/m3, four were measured during
   the first quarter,  two in the second, one in the third, and two in the fourth.

•  Note that neither acetaldehyde nor formaldehyde have first quarter or annual average
   concentrations presented in Table  5-5.  This is because maintenance of the primary
   carbonyl compound sampler at PXSS led to a problem with the ozone denuder,
   resulting in the invalidation of the sampling results through the end of March 2011.
   However, Appendix L provides the pollutant-specific average concentrations for all
   valid  samples collected over the entire sample period for each site.

Observations for SPAZ from Table 5-5 include the following:

•  The pollutant with the highest annual average concentration by mass for SPAZ is
   xylenes (4.06 ฑ 1.14 |ig/m3), which is more than twice the next highest annual
   average for benzene (1.65 ฑ 0.45 |ig/m3). Ethylbenzene is the only other pollutant of
   interest with an annual average greater than  1 |ig/m3 (1.06 ฑ 0.28 |ig/m3).

•  The first and fourth quarterly averages of xylene are more than twice the quarterly
   averages for the  second and third quarters and have relatively large confidence
   intervals associated with them. Xylenes were detected in all 31 valid VOC samples
   collected at SPAZ  and ranged from 0.879 |ig/m3to 11.93 |ig/m3. Of these  31, the
   eight  highest concentrations  (those greater than 5 |ig/m3) were all measured in the
   first and fourth quarters; conversely, the four lowest concentrations (those less than
   1 |ig/m3) were measured during the second and third quarters. SPAZ is the only site
   for which xylenes is a pollutant of interest.

•  Similar to PXSS, the quarterly average concentrations of several of the VOCs exhibit
   a seasonal trend, with the averages being higher during the colder months  of year.
   However, the confidence intervals indicate that the differences in the quarterly
   averages are not statistically  significant for SPAZ.
                                   5-22

-------
       Tables 4-9 through 4-12 present the sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for PXSS and
SPAZ from those tables include the following:
       •  PXSS and SPAZ appear in Tables 4-9 through 4-12 a total of 25 times.
       •  SPAZ has the highest annual average concentration of 1,3-butadiene,
          />-dichlorobenzene, and ethylbenzene among all NMP sites sampling VOCs, and the
          second highest annual averages of benzene and tetrachloroethylene. PXSS has the
          highest annual average concentration of tetrachloroethylene and the second highest
          annual average concentrations of 1,3-butadiene and ethylbenzene among sites
          sampling VOCs.
       •  PXSS has the highest annual average concentration of hexavalent chromium among
          all NMP sites sampling this pollutant. However, the annual average for 2011 is half of
          what it was in 2010 (0.13 ฑ 0.03 ng/m3).
       •  PXSS also has the highest annual average concentrations of beryllium and
          manganese, the second highest annual averages of arsenic and nickel, and the third
          highest annual average concentration of lead, among NMP sites sampling PMi0
          metals.

5.4.2   Concentration Comparison
       In order to better illustrate how a site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for the selected NATTS MQO Core
Analytes listed in Section 3.5.3, where applicable. Thus, box plots for benzene and 1,3-butadiene
were created for both PXSS and SPAZ. Box plots were also created for arsenic, benzo(a)pyrene,
hexavalent chromium, lead, manganese, and naphthalene for PXSS. Figures 5-10 through 5-17
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 3.5.3.
                                         5-23

-------
       Figure 5-10. Program vs. Site-Specific Average Arsenic (PMi0) Concentration
                                              2         2.5
                                              Concentration (ng/m3)
                 Program:   IstQuartile      2ndQuartile      SrdQuartile     4thQuartile     Average
                 Site:
                          Site Average     Site Minimum/Maximum

          Figure 5-11. Program vs. Site-Specific Average Benzene Concentrations
SFiZ

i Program Max Concentration = 23. 8 UR/m3


1

™ '


|



:
j

                                          4         5
                                              Concentration
                 Program:   IstQuartile      2ndQuartile      SrdQuartile     4thQuartile     Average
                 Site:      Site Average     Site Minimum/Maximum

      Figure 5-12. Program vs. Site-Specific Average Benzo(a)pyrene Concentration
                                              Concentration (ng/mi)
                 Program:   IstQuartile      2ndQuartile      SrdQuartile     4thQuartile     Average
                 Site:
                          Site Average     Site Minimum/Maximum

                                                5-24

-------
     Figure 5-13. Program vs. Site-Specific Average 1,3-Butadiene Concentrations
                                                                                                j
                                                                i Program Max Concentration = 9.51 |ag/m3

         -o-
                                                                 Program Max Concentration = 9.51 ug/m3
                                                1.5
                                           Concentration (
              Program:   IstQuartile     2ndQuartile      SrdQuartile     4thQuartile      Average
              Site:      Site Average     Site Minimum/Maximum

Figure 5-14. Program vs. Site-Specific Average Hexavalent Chromium Concentration
                0.05
                                0.1
                                               0.15
                                           Concentration (ng/m3)
                                                               3.2
                                                                               0.25
                                                                                               D.3
              Program:   IstQuartile     2ndQuartile      SrdQuartile     4thQuartile      Average
              Site:      Site Average     Site Minimum/Maximum

      Figure 5-15. Program vs. Site-Specific Average Lead (PMi0) Concentration
                                         15            20
                                           Concentration (ng/m3)
Program:
Site:
IstQuartile
Site Average
0
2ndQuartile
SrdQuartile
n
4thQuartile Average
n
^B 1 	 1
Site Minimum/Maximum
                                             5-25

-------
Figure 5-16. Program vs. Site-Specific Average Manganese (PMi0) Concentration





1
,: Program Max Concentration =



395 ng/m3


                             75         100         125
                                   Concentration (ng/m3)
           Program:   IstQuartile    2ndQuartile     SrdQuartile     4thQuartile     Average
          Site:     Site Average    Site Minimum/Maximum

   Figure 5-17. Program vs. Site-Specific Average Naphthalene Concentration



,
1



*^L r
Program Max Concentration -779 ng/m



, !
i

                               ::•:•      250       3:0
                                   Concentration (nฃ/m3)
           Program:   IstQuartile    2ndQuartile     SrdQuartile     4thQuartile     Average
          Site:     Site Average    Site Minimum/Maximum

  Observations from Figures 5-10 through 5-17 include the following:

     •   Figure 5-10 shows that the annual average arsenic (PMio) concentration for PXSS
         is greater than the program-level average for arsenic (PMio). Although the
         maximum concentration of arsenic measured across the program was not
         measured at PXSS, the maximum concentration measured at PXSS (3.05 ng/m3)
         was among the higher arsenic measurements. There were no non-detects of
         arsenic measured at PXSS.

     •   Figure 5-11 for benzene shows both sites, as both SPAZ and PXSS sampled
         VOCs. Note that the program-level maximum concentration (23.8 |ig/m3) is not
         shown directly on the box plots because the scale of the box plots would be too
         large to readily observe data points at the lower end of the concentration range.
         Thus, the scale of the box plots has been reduced to 10 |ig/m3. While neither
         Arizona site measured the maximum benzene concentration measured across the
         program, both annual averages are greater than the  program-level average
         concentration. The  annual average benzene concentration for SPAZ is slightly
         higher than the annual average concentration for PXSS. SPAZ has the second
         highest annual average concentration of benzene among NMP sites sampling this
         pollutant, as discussed above. There were no non-detects of benzene measured at
         either site (or among sites sampling VOCs).
                                     5-26

-------
•  Figure 5-12 is the box plot for benzo(a)pyrene for PXSS. Note that the program-
   level first quartile for this pollutant is zero and is not visible on this box plot. The
   box plot shows that the annual average concentration for PXSS is just greater than
   the program-level average concentration and just less than the program-level third
   quartile. Figure 5-12 also shows that the maximum concentration measured at
   PXSS is considerably less than the maximum concentration measured across the
   program. Several non-detects of benzo(a)pyrene were measured at PXSS.

•  Figure 5-13 for 1,3-butadiene also shows both sites. Similar to the box plots for
   benzene, the program-level maximum 1,3-butadiene concentration (9.51 |ig/m3) is
   not shown directly on the box plots as the scale has been reduced to 3 |ig/m3 in
   order to allow for the observation of data points at the lower end of the
   concentration range. The annual average concentrations for both sites are more
   than twice the program-level average concentration. Further, these two sites have
   the highest annual average concentrations of this pollutant across the program, as
   mentioned above, with the 1,3-butadiene annual average concentration for  SPAZ
   slightly higher than the annual average concentration for PXSS. There were five
   non-detects of 1,3-butadiene measured at PXSS and one non-detect measured at
   SPAZ.

•  Figure 5-14 is the box plot for hexavalent chromium for PXSS. Figure 5-14 shows
   that the maximum concentration of this pollutant across the program
   (0.268 ng/m3) was measured at PXSS. The box plot also shows that this site's
   annual average concentration of hexavalent chromium (0.065 ng/m3) is nearly
   three times greater than the program-level average concentration (0.024 ng/m3).
   As discussed above, PXSS has the highest annual average concentration of
   hexavalent chromium among NMP sites sampling hexavalent chromium. There
   were no non-detects of hexavalent chromium measured at PXSS.

•  Figure 5-15 shows that the annual average concentration of lead (PMio) for PXSS
   is just greater than the program-level average concentration. While the maximum
   concentration measured at PXSS is considerably less than the program maximum
   concentration, PXSS has the third highest annual average concentration of  lead
   among the NMP sites, behind only S4MO and NBIL. There were no non-detects
   of lead measured at PXSS (or among sites sampling PMio metals).

•  Figure 5-16 is the box plot for manganese (PMio) for PXSS. Note that the
   program-level maximum concentration  (395 ng/m3) is not shown directly on the
   box plot as the scale has been reduced to 200 ng/m3 in order to allow for the
   observation of data points at the lower end of the concentration range. Figure 5-16
   shows the annual average concentration of manganese for PXSS (22.82 ng/m3) is
   greater than the program-level average concentration (8.81  ng/m3), nearly two and
   a half times as high). PXSS has the highest annual average concentration of
   manganese among NMP sites sampling PMio metals, as discussed above. While
   the maximum concentration measured at PXSS (130 ng/m3) is considerably less
   than the program-level maximum concentration, this is the  second highest
   measurement of manganese measured among the NMP sites sampling
                               5-27

-------
             metals. There were no non-detects of manganese measured at PXSS (or among
             sites sampling PMi0 metals).
          •  Figure 5-17 is the box plot for naphthalene. Note that the program-level
             maximum concentration (779 ng/m3) is not shown directly on the box plot as the
             scale has been reduced to 500 ng/m3 in order to allow for the observation of data
             points at the lower end of the concentration range. Figure 5-17 shows that the
             annual average naphthalene concentration for PXSS (89.36 ng/m3) is greater than
             the program-level average concentration (81.68 ng/m3). The maximum
             naphthalene concentration measured at PXSS (293 ng/m3) is considerably less
             than the program-level maximum concentration. There were no non-detects of
             naphthalene measured at PXSS (or among sites sampling PAHs).
          •  Recall that annual averages could not be calculated for formaldehyde and
             acetaldehyde, as discussed in Section 5.4.1.

5.4.3   Concentration Trends
       A site-specific trends evaluation was completed for sites that have sampled one or more
of the selected NATTS MQO Core Analytes for 5 consecutive years or longer, as described in
Section 3.5.4. PXSS has sampled PMi0 metals and hexavalent chromium under the NMP since
2006; thus, Figures 5-18 through 5-21 present the annual statistical metrics for arsenic,
hexavalent chromium, lead, and manganese, respectively. The statistical metrics presented for
assessing trends include the substitution of zeros for non-detects.

       SPAZ began sampling VOCs and PXSS began sampling VOCs, carbonyl compounds,
and PAHs under the NMP in 2007. However, they did not begin sampling until halfway through
the year; therefore,  annual statistical metrics could not be calculated for 2007. Thus, the trends
analysis was not conducted for the pollutants for these methods for PXSS and SPAZ.
                                         5-28

-------
Figure 5-18. Annual Statistical Metrics for Arsenic (PMi0) Concentrations Measured at
                                    PXSS



tf\
[ration (ng/r
rage Concen









r
— •




T
1 r— ^
L 6
5 	 1

	 ซ, 	 ^ ...^
1 — f— —f— 	 — t— — ป—
2006 2007 2008 2009 2010 2011
Year
• 5th Percentile — Minimum — Median - Maximum • 95th Percentile ..^.. Average
  Figure 5-19. Annual Statistical Metrics for Hexavalent Chromium Concentrations
                              Measured at PXSS



t rat ion (ng/m3)
S 1
ti I
1
3
1
"c,



"


























ซ
— '
20
0




••


t f
._ *
J... | 	 |.-ป-..|... r^
•^ ^? Lzl L5 L_jJ El^i
06 2007 2008 2009 2010 2011
Year
5th Percentile — Minimum — Median - Maximum • 95th Percentile ...+*. Average
                                     5-29

-------
Figure 5-20. Annual Statistical Metrics for Lead (PMi0) Concentrations Measured at PXSS
                                          2008          2009
                                                 Year
                 •  5th Percentile    —  Minimum     — Median     -  Maximum    •  95th Percentile
                                                                                > Average
Figure 5-21. Annual Statistical Metrics for Manganese (PMi0) Concentrations Measured at
                                             PXSS
         80
                                          200S          2009
                                                 Year
                 *  5th Percentile    —  Minimum     — Median     -  Maximurr
                                                                                .-*
                                                                   95th Percentile   ..^.. Average
                                              5-30

-------
Observations from Figure 5-18 for arsenic measurements at PXSS include the following:

       •   PXSS began sampling arsenic under the NMP in January 2006.

       •   The maximum arsenic concentration (6.73 ng/m3) was measured on
          December 26, 2007 and is more than twice the next highest concentration
          (3.05 ng/m3), measured on August 19, 2011.

       •   The average concentration increased from 2010 to 2011 after several years of slight
          decreasing,  although the changes across the years of sampling are not statistically
          significant.  The averages range from 0.51 ng/m3 in 2010 to 0.77 ng/m3 in 2011.


       Observations from Figure 5-19 for hexavalent chromium measurements at PXSS include

the following:

       •   PXSS began sampling hexavalent chromium in January 2006.

       •   The maximum hexavalent chromium concentration  shown was measured on
          July 10, 2006 (0.989 ng/m3). The second highest concentration was also measured in
          2006 (0.737 ng/m3).

       •   There is little statistical difference in the average concentrations, which range from
          0.06 ng/m3 in 2011 to 0.13 ng/m3 in 2006 and 2010.

       •   For 2011, the difference between the average and the median concentrations is at a
          minimum, indicating decreasing variability in the concentrations measured.


       Observations from Figure 5-20 for lead measurements at PXSS include the following:

       •   The maximum lead concentration (23.1 ng/m3) was  measured on January 1, 2009.
          The second highest concentration (17.4 ng/m3) was  measured on December 31, 2006.

       •   Similar to arsenic, the average lead concentrations exhibit an increase for 2011 after a
          decreasing trend between 2008 and 2010.

       •   There have  been no non-detects reported for lead since the onset of lead sampling at
          PXSS. The  minimum concentrations for 2007 and 2008 are considerably less than the
          others, but are still numerical measurements.


       Observations from Figure 5-21 for manganese measurements at PXSS include the

following:

       •   Three manganese concentrations greater than 100 ng/m3 have been measured at PXSS
          since 2006; all three were measured in 2011. Of the nine concentrations greater than
          50 ng/m3, five were measured during 2011.
                                         5-31

-------
       •  The annual average concentration of manganese for 2011 is twice the average for
          2010 and represents a significant increase between the two years. Over the course of
          sampling, the measurements from 2011 exhibit the highest variability while the
          measurements from 2010 exhibit the least. Regardless, this site still has some of the
          highest levels of manganese among sites sampling PMi0 metals; PXSS had the second
          highest annual average concentration of manganese for 2010 and the highest annual
          average across the program for 2011.
       •  The 95th percentile for 2011 is significantly higher than for the previous years of
          sampling, indicating that a greater number of concentrations were on the higher side
          for 2011.
       •  The widening difference between the median and the annual average concentration
          for 2011 is  a further indication of the increasing variability of the 2011 manganese
          measurements, whereas for 2010, the median and average concentrations were less
          than 0.30 ng/m3 apart.

5.5    Additional Risk-Based Screening Evaluations
       The following risk-based screening evaluations were conducted to characterize risk at
each Arizona monitoring site. Refer to Sections 3.3 and  3.5.5 for definitions and  explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.
5.5.1   Risk-Based Screening Assessment Using MRLs
       A risk-based screening was conducted by comparing the concentration data from the
Arizona monitoring sites to the ATSDR MRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 to 14 days); intermediate (exposures of 15 to 364 days); and chronic (exposures
of 1 year or greater). The preprocessed daily measurements of the pollutants of interest were
compared  to the acute MRLs; the quarterly averages were compared to the intermediate MRLs;
and the annual averages were compared to the chronic MRLs.

       As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2011.
                                          5-32

-------
5.5.2   Cancer Risk and Noncancer Hazard Approximations

       For the pollutants of interest for the Arizona monitoring sites and where annual average

concentrations could be calculated, risk was examined by calculating cancer risk and noncancer

hazard approximations. These approximations can be used as risk estimates for cancer and

noncancer effects attributable to the pollutants of interest. Although the use of these

approximations is limited, they may help identify where policy-makers may want to shift or
confirm their air-monitoring priorities. Refer to Section 3.5.5.2 for an explanation of how cancer

risk and noncancer hazard approximations are calculated and what limitations are associated with
them. Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer

hazard approximations are presented in Table 5-6, where applicable. Cancer risk approximations

are presented as probabilities while the noncancer hazard approximations are ratios and thus,

unitless values.


       Observations for PXSS from Table 5-6 include the following:

       •  The pollutants with the highest annual average concentrations by mass are benzene,
          ethylbenzene, and carbon tetrachloride, with only benzene having an annual average
          concentration greater than 1 |ig/m3.

       •  Based on the annual averages and  cancer UREs, benzene, 1,3-butadiene, and carbon
          tetrachloride have the three highest cancer risk approximations. An additional four
          pollutants have cancer risk approximations greater than 1.0 in-a-million.

       •  None of the pollutants of interest for PXSS have noncancer hazard approximations
          greater than 1.0, indicating that no adverse health effects are expected from these
          individual pollutants. The pollutant with the highest noncancer hazard approximation
          for PXSS is manganese (0.46).

       •  The noncancer hazard approximation for PXSS for manganese is the third  highest
          noncancer hazard approximation calculated across the program and the second
          highest for this pollutant (second only to TOOK).

       •  Annual averages (and therefore cancer risk and noncancer hazard approximations)
          could not be calculated for acetaldehyde and formaldehyde, as discussed in
          Section 5.4.1.
                                          5-33

-------
                  Table 5-6. Risk Approximations for the Arizona Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
# of Measured
Detections vs.
# of Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Phoenix, Arizona - PXSS
Acetaldehyde
Arsenic (PM10)a
Benzene
Benzo(a)pyrene a
Bery Ilium (PM10)a
1,3 -Butadiene
Cadmium (PM10)a
Carbon Tetrachloride
Chloroform
ฃ>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
Hexavalent Chromium a
Lead(PM10)a
Manganese (PM10) a
Naphthalene a
Nickel (PM10)a
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
0.0000022
0.0043
0.0000078
0.00176
0.0024
0.00003
0.0018
0.000006

0.000011
0.000026
0.0000025
0.000013
0.012


0.000034
0.00048
0.00000026
0.0000048
0.0000088
0.009
0.000015
0.03

0.00002
0.002
0.00001
0.1
0.098
0.8
2.4
1
0.0098
0.0001
0.00015
0.00005
0.003
0.00009
0.04
0.002
0.1
48/48
61/61
61/61
30/57
61/61
56/61
61/61
61/61
57/61
55/61
12/61
61/61
48/48
62/62
61/61
61/61
57/57
61/61
56/61
16/61
5/61
NA
0.01
ฑ0.01
1.34
ฑ0.22
0.01
ฑ0.01
0.01
ฑ0.01
0.23
ฑ0.06
0.01
ฑ0.01
0.63
ฑ0.03
0.37
ฑ0.07
0.20
ฑ0.04
0.02
ฑ0.01
0.82
ฑ0.13
NA
O.01
ฑO.01
0.01
ฑ 0.01
0.02
ฑ0.01
0.09
ฑ0.02
O.01
ฑO.01
0.55
ฑ0.28
0.02
ฑ0.01
0.01
ฑ0.01
NA
3.33
10.48
0.18
0.11
6.89
0.25
3.80

2.19
0.57
2.05
NA
0.78


3.04
0.83
0.14
0.10
0.01
NA
0.05
0.04

0.01
0.11
0.01
0.01
0.01
O.01
0.01
O.01
NA
O.01
0.03
0.46
0.03
0.02
0.01
0.01
0.01
— = a Cancer URE or Noncancer RfC is not available.
NA = Not available due to the criteria for calculating an annual average.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 5-5.
                                                   5-34

-------
          Table 5-6. Risk Approximations for the Arizona Monitoring Sites (Continued)
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
# of Measured
Detections vs.
# of Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
South Phoenix, Arizona - SPAZ
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
1 ,2-Dichloroethane
Ethylbenzene
ฃ>-Dichlorobenzene
Tetrachloroethylene
Trichloroethylene
Xylenes
0.0000078
0.00003
0.000006

0.000026
0.0000025
0.000011
0.00000026
0.0000048
	
0.03
0.002
0.1
0.098
2.4
1
0.8
0.04
0.002
0.1
31/31
30/31
30/31
26/31
7/31
31/31
29/31
28/31
12/31
31/31
1.65
ฑ0.45
0.29
ฑ0.10
0.63
ฑ0.07
0.18
ฑ0.04
0.02
ฑ0.01
1.06
ฑ0.28
0.26
ฑ0.07
0.41
ฑ0.16
0.05
ฑ0.03
4.06
ฑ1.14
12.88
8.66
3.79

0.56
2.66
2.91
0.11
0.24
	
0.06
0.14
0.01
0.01
<0.01
0.01
0.01
0.01
0.02
0.04
— = a Cancer URE or Noncancer RfC is not available.
NA = Not available due to the criteria for calculating an annual average.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 5-5.
          Observations for SPAZ from Table 5-6 include the following:

          •   The pollutants with the highest annual average concentrations by mass are xylenes,
              benzene, and ethylbenzene. These three pollutants are the only ones with an annual
              average concentration greater than 1 |ig/m3.

          •   Based on the annual averages and cancer UREs, benzene, 1,3-butadiene, and carbon
              tetrachloride have the three highest cancer risk approximations. Two additional
              pollutants have cancer risk approximations greater than 1.0 in-a-million.

          •   The cancer risk approximation for SPAZ for benzene (12.88 in-a-million) is the
              second highest cancer risk approximation calculated for this pollutant across the
              program (second only to TOOK). The cancer risk approximation for PXSS for
              benzene (10.48 in-a-million) ranks fifth.

          •   None of the pollutants of interest for SPAZ have noncancer hazard approximations
              greater than 1.0, indicating no adverse health effects are expected from these
              individual pollutants. The pollutant with the highest noncancer hazard approximation
              for SPAZ is 1,3-butadiene (0.14).
                                              5-35

-------
5.5.3   Risk-Based Emissions Assessment
       In addition to the risk-based screenings discussed above, Tables 5-7 and 5-8 present 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 2008 NEI, the 10
pollutants with the highest toxicity-weighted emissions, and the 10 pollutants with the highest
cancer risk approximations (in-a-million), as calculated from the annual averages provided in
Table 5-6. Table 5-8 presents similar information, but identifies the 10 pollutants with the
highest noncancer hazard approximations (HQ), also calculated from annual averages provided
in Table 5-6.

       The pollutants listed in  Table 5-7 and 5-8 are limited to those that have cancer and
noncancer toxicity factors, respectively. As a result, although the actual value of the emissions is
the same, the highest emitted pollutants in the cancer table may be different from the noncancer
table. The cancer risk and noncancer hazard approximations based on each site's annual averages
are limited to those pollutants for which each respective site sampled. As discussed in
Section 5.3, PXSS sampled for VOCs, carbonyl compounds, PAHs, metals (PMio), and
hexavalent chromium; SPAZ sampled for VOCs only. 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, cancer risk and noncancer hazard approximations were
not calculated for formaldehyde and acetaldehyde. A more in-depth discussion of this analysis is
provided in Section 3.5.5.3. Similar to the cancer risk and noncancer hazard approximations, this
analysis may help policy-makers prioritize their air monitoring activities.
                                          5-36

-------
Table 5-7. 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
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity- Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Phoenix, Arizona (Maricopa County) - PXSS
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
Dichloromethane
POM, Group 2b
POM, Group la
Propylene oxide
1,256.37
969.28
766.10
487.50
179.20
90.54
28.45
14.68
2.61
1.66
Formaldehyde
Benzene
1,3 -Butadiene
Naphthalene
Ethylbenzene
Hexavalent Chromium, PM
POM, Group 2b
Acetaldehyde
Arsenic, PM
POM, Group 5a
1.26E-02
9.80E-03
5.38E-03
3.08E-03
1.92E-03
1.70E-03
1.29E-03
1.07E-03
7.14E-04
3.83E-04
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Arsenic
Naphthalene
ฃ>-Dichlorobenzene
Ethylbenzene
Nickel
Hexavalent Chromium
1 ,2-Dichloroethane
10.48
6.89
3.80
3.33
3.04
2.19
2.05
0.83
0.78
0.57
South Phoenix, Arizona (Maricopa County) - SPAZ
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
Dichloromethane
POM, Group 2b
POM, Group la
Propylene oxide
1,256.37
969.28
766.10
487.50
179.20
90.54
28.45
14.68
2.61
1.66
Formaldehyde
Benzene
1,3 -Butadiene
Naphthalene
Ethylbenzene
Hexavalent Chromium, PM
POM, Group 2b
Acetaldehyde
Arsenic, PM
POM, Group 5a
1.26E-02
9.80E-03
5.38E-03
3.08E-03
1.92E-03
1.70E-03
1.29E-03
1.07E-03
7.14E-04
3.83E-04
Benzene
1,3 -Butadiene
Carbon Tetrachloride
ฃ>-Dichlorobenzene
Ethylbenzene
1 ,2-Dichloroethane
Trichloroethylene
Tetrachloroethylene
12.88
8.66
3.79
2.91
2.66
0.56
0.24
0.11


-------
        Table 5-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with
                                       Noncancer RfCs for the Arizona Monitoring Sites
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Phoenix, Arizona (Marico]
Toluene
Xylenes
Benzene
Formaldehyde
Hexane
Ethylbenzene
Acetaldehyde
Ethylene glycol
1,3 -Butadiene
Glycol ethers, gas
3,931.16
2,893.73
1,256.37
969.28
963.82
766.10
487.50
240.60
179.20
98.40
Acrolein
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Lead, PM
Benzene
Naphthalene
Xylenes
Arsenic, PM
Propionaldehyde
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer Hazard
Approximation
(HQ)
pa County) - PXSS
2,430,403.47
98,905.78
89,599.85
54,166.44
42,964.38
41,879.01
30,178.52
28,937.26
11,077.09
8,611.73
Manganese
1,3 -Butadiene
Arsenic
Benzene
Naphthalene
Lead
Nickel
Cadmium
Tetrachloroethylene
Trichloroethylene
0.46
0.11
0.05
0.04
0.03
0.03
0.02
0.01
0.01
0.01
South Phoenix, Arizona (Maricopa County) - SPAZ
Toluene
Xylenes
Benzene
Formaldehyde
Hexane
Ethylbenzene
Acetaldehyde
Ethylene glycol
1,3 -Butadiene
Glycol ethers, gas
3,931.16
2,893.73
1,256.37
969.28
963.82
766.10
487.50
240.60
179.20
98.40
Acrolein
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Lead, PM
Benzene
Naphthalene
Xylenes
Arsenic, PM
Propionaldehyde
2,430,403.47
98,905.78
89,599.85
54,166.44
42,964.38
41,879.01
30,178.52
28,937.26
11,077.09
8,611.73
1,3 -Butadiene
Benzene
Xylenes
Trichloroethylene
Tetrachloroethylene
Carbon Tetrachloride
Chloroform
Ethylbenzene
ฃ>-Dichlorobenzene
1,2-Dichloroethane
0.14
0.06
0.04
0.02
0.01
0.01
0.01
0.01
0.01
0.01
fj\


-------
Observations from Table 5-7 include the following:

•  Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
   cancer UREs in Maricopa County.

•  The pollutants with the highest toxi city-weighted emissions (of the pollutants with
   cancer UREs) are formaldehyde, benzene, and 1,3-butadiene.

•  Seven of the highest emitted pollutants in Maricopa County also have the highest
   toxicity-weighted emissions.

•  Benzene, 1,3-butadiene, and carbon tetrachloride have highest cancer risk
   approximations for both PXSS and SPAZ. While benzene and 1,3-butadiene both
   appear among the pollutants with the 10 highest emissions and 10 highest toxicity-
   weighted emissions for Maricopa County, carbon tetrachloride does not appear on
   either list. In addition to benzene and 1,3-butadiene, ethylbenzene also appears on all
   three lists for both sites. Naphthalene also appears on all three lists for PXSS.

•  POM, Group 2b is the eighth highest emitted "pollutant" in Maricopa County and
   ranks seventh for toxicity-weighted emissions. POM, Group 2b includes several
   PAHs sampled for at PXSS including acenaphthene, benzo(e)pyrene, fluoranthene,
   and perylene. None of the PAHs included in POM, Group 2b were identified as
   pollutants of interest for PXSS.

•  POM, Group 5a is the pollutant with the tenth highest toxicity-weighted emissions for
   Maricopa County.  Benzo(a)pyrene is part of POM, Group 5a.


Observations from Table 5-8 include the following:

•  Toluene, xylenes, and benzene are the highest emitted pollutants with noncancer
   RfCs in Maricopa County.

•  The pollutants with the highest toxicity-weighted emissions (of the pollutants with
   noncancer RfCs) are acrolein, formaldehyde, and 1,3-butadiene.

•  Five of the highest emitted pollutants also have the highest toxicity-weighted
   emissions.

•  Acrolein has the highest toxicity-weighted emissions (by two orders of magnitude)
   for Maricopa County. Although acrolein was sampled for at both sites, this pollutant
   was excluded from the pollutants of interest designation, and thus subsequent risk-
   based screening evaluations, due to questions about the consistency and reliability of
   the measurements, as discussed in Section 3.2.

•  Manganese has the highest noncancer hazard approximation for PXSS (although
   considerably less than an HQ of 1.0), followed by 1,3-butadiene. 1,3-Butadiene has
   the highest noncancer hazard approximation for SPAZ. In addition to 1,3-butadiene,
   benzene and xylenes all appear on all three lists for both sites.

                                   5-39

-------
       •   Five metals appear among the 10 pollutants with highest noncancer hazard
          approximations for PXSS. Two of these (arsenic and lead) are among those with the
          highest toxicity-weighted emissions while none appear among the highest emitted
          pollutants in Maricopa County.


5.6    Summary of the 2011 Monitoring Data for PXSS and SPAZ

       Results from several of the data treatments described in this section include the
following:

       ปซป  Twenty-three pollutants failed screens for PXSS; 14 of these are NA TTSMQO Core
          Analytes. Ten pollutants failed screens for SPAZ, of which four are NATTSMQO
          Core Analytes.

       *ป*  Of the site-specific pollutants of interest for PXSS, benzene had the highest annual
          average concentration and was the only pollutant with an annual average greater
          than 1 jug/m3. For SPAZ,  xylenes had the highest annual average concentration
          among this site's pollutants of interest.

       ปซป  Concentrations of several VOCs, including benzene and 1,3-butadiene, tended to be
          slightly higher during the colder months of the year.

       ปซป  PXSS has the highest annual average concentration oftetrachloroethylene,
          hexavalent chromium,  beryllium (PMw), and manganese (PMw) among all NMP sites
          sampling these pollutants; SPAZ has the highest annual average concentration of
          1,3-butadiene, p-dichlorobenzene, and ethylbenzene among all NMP sites sampling
          VOCs.
                                         5-40

-------
6.0    Sites in California
       This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at three NATTS sites in California, and integrates these concentrations
with emissions, meteorological, and risk information. Data generated by sources other than ERG
are not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) for detailed discussions and definitions
regarding the various data analyses presented below.

6.1    Site Characterization
        This section characterizes the California monitoring sites by providing geographical and
physical information about the locations of the sites and the surrounding areas. This information
is provided to give the reader insight regarding factors that may influence the air quality near the
sites and assist in the interpretation of the ambient monitoring measurements.

       The California monitoring sites are located in Los Angeles, Rubidoux, and San Jose.
Figures 6-1, 6-3, and 6-5 are composite satellite images retrieved from ArcGIS Explorer showing
the monitoring sites in their urban locations. Figures 6-2, 6-4, and 6-6 identify nearby point
source emissions locations by source category, as reported  in the 2008 NEI for point sources.
Note that only sources within 10 miles of the sites are included in the facility counts provided in
Figures 6-2, 6-4, and 6-6. A 10-mile boundary was chosen  to give the reader an indication of
which emissions sources and emissions source categories could potentially have a direct effect
on the air quality at the monitoring sites. Further, this boundary provides  both the proximity of
emissions sources to the monitoring sites as well as the quantity of such sources within a given
distance of the sites. Sources outside each 10-mile radius are still visible on the maps, but have
been grayed out in order to show emissions sources just outside the boundary. Table 6-1 provides
supplemental geographical information such as land use, location setting, and locational
coordinates.
                                           6-1

-------
to
                                     Figure 6-1. Los Angeles, California (CELA) Monitoring Site
                                     __^  " V, .'- ป                      i    '      ?-!_.•ซ-

-------
            Figure 6-2. NEI Point Sources Located Within  10  Miles of CELA
                                           2
                                      0         ,?        +       *""
                                                    U8"15"0"W              115 10'0~W              118'5H1"W               118'OXTW
                                                                 Note: Due to facilily density and collocation, the total facilities
                                                                 displayed may not represent all facilities within the area of interest
Legend
  •&   CELA NATTS site          10 mile radius
                                                                                         County boundary
Source Category Group (No. of Facilities)
Ip Aerospace/Aircraft Manufacturing (2)
0 Air-comJitioning/Refrigeratwn (1)
-f" Aircraft Operations (62)
 I  Asphgtt Processing/Roofing Manufacturing (2)
 <>  Auto Body Shop/Pamlers (12)
S Automobile/Tnjck Manufacturing (4)
0  Bakery (4)
X Battery Manufacturing (2)
tf Burning Construction (4)
 B  Bulk Terminals/Bulk Plants (2)
C  Cltemtcal Manufacturing (8)
•  Concrete  Batch Plant (1)
XJ Crematory - Animal/Human (2)
6 Etectncat  Equipment (2)
 ฃ  EteclrMy Generation via Combustion (5)
 E  Electroplating, Plating. Polishing. Anodizing. & Coloring (17)
ฉ Fabricated Metal Products (11)
                          fe9 Flexible PolyureUiane Foam Production (2)
                           F Food Processing/Agriculture (13)
                          [" ! Furniture Plant {20)
                          1* Glass Manufacturing (2)
                          A Grain Handling (2)
                           (V Heating Equipment Manufacturing (1)
                          [3 Hospital (4)
                          $ Hoi MiK Asphalt Plant (1)
                          T Industrial Machinery and Equipment (5)
                          Ip Institutional • prison (1)
                          ^ Institutional - school (1}
                           I Iron and Steel Foundry (7}
                          A Landfill (1)
                          V Leather and Leather Products (1}
                          V Mineral Products (1)
                          ? Miscellaneous Commercial Industrial (31)
M Miscellaneous Manufacturing (6)
0 Municipal V^iste Combustor(1 >
• Oil artd/or Gas Production (11)
jt Pel roleum Refinery (1)
^—- PnarmaceuttcalManufaclunng (1)
 1 Primary Metal Production (8)
P Printing/Publishing (26)
IB Pulp and Paper Pian(/VWปd Products (9)
R Rubber and Miscellaneous Plastics Products (6)
2 Secondary  Melal Processing (4)
< Site Remediation Activity {1)
V S1eelMill(2)
S Surface Coaling (10)
TT TeleoonvntmBattons (2)
T Textile Mill (8)
*v Transportation Equipmenl (2)
l V^atewater Treatment (6)
W V*odworh, Furniture, Miltoorfc & V\food Preserving (2)
                                                                 6-3

-------
Figure 6-3. Rubidoux, California (RUCA) Monitoring Site

-------
            Figure 6-4. NEI Point Sources Located Within  10 Miles of RUCA
                                     117'30'CTW              117'25'frW             117'20^"W             117"15TJPW
                                                                Note: Due to facility density and collocation, the total facilities
                                                                displayed may not represent all facilities within the area of interest.
                            Legend

                             @   RUCA NATTS site          10 mile radius Q   ] County boundary
Source Category Group (No. of Facilities)
i$* Ae rospacefAircraft Manufacturing (1)
41 Aircraft Operations (12)
 I  Asphalt Processing/Roofing Manufacturing (3)
 0  Auto Body Shop/Painters (6)
H AutomobileiTruck Manufacturing (7)
$  Bakery (3)
ฑ  BoatManufactunngd)
    Brick Manufacturing & Structural Clay (2)
ft  Building Construction (3)
B  Bulk Terminals^Bulk Plants (2)
C  Chemical Manufacturing (1)
•  Concrete Batch Plant (5)
8  Electrical Equipment (5)
 f  Electricity Generation via Combustion (4)
ฉ Fabricated Metal Products (10)
Flexible Polyurethane Foam Production (1)
Food Processing/Agriculture (18)
Furniture Plant (1)
GasolmeiDiesel Service Station (1)
Hospital (1)
Hot Mix Asphalt Plant(1)
Industrial Machinery and Equipment (2)
Institutional • prison (1)
Institutional - school (5)
Landfill (4)
Lumber/sawmill (1)
Mine/Quarry (3)
Mineral Products (2)
Miscellaneous Commercial/Industrial (10)
Miscellaneous Manufacturing (2)
• Oil and/or Gas Production (3)
* Other Solid Vvaste Incineration (1)
7 Portland Cement Manufactunng (2)
 1 Primary Metal Production (4)
P PrrtngCPublishing(IO)
H Pulp and Paper PlanuWsod Products (4)
R Rubber and Miscellaneous Plastics Products (6)
2 Secondary Metal Processing (2)
< SiteRenwdiafion Activity (1)
> Solid Waste Disposal - Comrnerciairinstitutional (1)
V Steel Mill (4)
S Surface Coaling (4)
** Transportation Equipment (3)
$ฃ Transportation and Marketing of Petroleum Products (1)
• Mfeslewater Treatment (9)
W Woodwork, Furniture. Milrwork 4 Wsod Preserving (2)
                                                                6-5

-------
                                       Figure 6-5. San Jose, California (SJJCA) Monitoring Site
Oi

-------
           Figure 6-6. NEI Point Sources  Located Within  10  Miles  of SJJCA
                    Legend
                                                      1ZV55'CrW            tZrSO'FW            121*4510"W           121 4QXJ-W
                                                                  Note: Due to facility density and collocation, the total facilities
                                                                  displayed may not represent all facilities within the area of interest.
                            SJJCA NATTS site    Q    10 mile radius
County boundary
Source Category Group (No. of Facilities)
 :\h  An (2)
 HH Aircraft Operations (19)
 I  Asphalt Processing/Roofing Manufacturing (4)
 0  Auto Body Shop/Painters (235)
 H Automobile/Truck Manufacturing (17)
 fe  BaKery(2)
 X  Battery ManufBdunng (1)
    Brick Manufacturing & Structural Clay (3)
 ft  Budding ConsUuclion (15)
 5  Bulk Terminals/Bulk Plants (5)
 C  Chemical Manufactunr^<5)
 O  Clay Ceramics Manufacturing (1)
 •  Concrete  Batch Plant (10)
XCrematory - Aiwnal/Human (3)
 ฎ  Dry Cleaning Facdtty (161)
 &  Electncal  Equipment (250)
 f  Electricity Generation via Combuston (2)
 E  Etectroplaling, Plalmg, Pobshmg. ArwdJang, SCotonng (14)
 4  Ermine Testing (1)
ฉ Fabricated Metal Producls (35)
^ Flexible Potyurethane Foam Produclidn (2)
 F  Food Process ing'AgricurtuTe (43)
I   , Furniture Plant (37)
V  Gasoline/Diesel Service Station (9)
^t Glass Manufacturing (1)
A Grain Handling (2)
 //  Healing Equipment Manufacturing (7)
[3 Hospital (16)
ฃ  Hoi Mix Asphalt Plant (1)
^fr Industrial Machinery and E(^iiprnenl (19)
l|r Institutional - school (27)
 I  Iron and Steel Foundry (1)
^ Laboratory (7)
W Leather and Leather  Products {1)
 /  Lumber/sawmill (1)
 A  Mflftary Base/Nalional Secitnty Facility (7)
 X  Mine/Quarry (5)
    4i Mineral Wool Manufacturing (1)
    ? Miscellaneous Commercial/Industrial (212)
    M Miscellaneous Manufacturing (69)
    • 0ซ1 and/or C3as Production (3)
   ••—I Pharmaceutical Manufacturing (9)
    1  Primary Melal Production (2)
    P Printing/Publishing (30)
    B Pulp and Paper Plant/Wood Products (W)
    R Rubber and Miscellaneous Plastics Products (3)
    2  Secondary Melal Processing (3)
    <  Erie Remedialion Activity (13)
    > Solid Waste Disposal - Commercial,'mstitutซ>na] (26)
    V Steel Mill (1}
    S Surface Coating (32)
    IT Tetecommunica1.K)ns{102)
    T Textile Mill (5)
   ซ*• Transportation Equipment (5)
    •ifc Transportation and Marketing of Petroleum Products (3)
    1  Wfcstewater Trea tmenl (IS)
   W Woodwork. Furrature. Miflwork & Wood Preserving (2)
                                                                  6-7

-------
                               Table 6-1. Geographical Information for the California Monitoring Sites
Site
Code
CELA
RUCA
SJJCA
AQS Code
06-037-1103
06-065-8001
06-085-0005
Location
Los
Angeles
Rubidoux
San Jose
County
Los
Angeles
Riverside
Santa
Clara
Micro- or
Metropolitan
Statistical Area
Los Angeles-Long
Beach-Santa Ana,
CAMSA
Riverside-San
Bernardino-
Ontario, CA MSA
San Jose-
Sunnyvale-Santa
Clara, CAMSA
Latitude
and
Longitude
34.06659,
-118.22688
33.99958,
-117.41601
37.3485,
-121.895
Land Use
Residential
Residential
Commercial
Location
Setting
Urban/City
Center
Suburban
Urban/City
Center
Additional Ambient Monitoring Information1
TSP, TSP Speciation, Hexavalent chromium, CO,
SO2, NO, NO2, NOX, NOy, PAMS, Carbonyl
compounds, VOCs, O3, Meteorological parameters,
PM10, PM10 Speciation, PM2 5, PM25 Speciation.
Haze, TSP Speciation, Hexavalent chromium, CO,
SO2, NO, NO2, NOX, PAMS, VOCs, Carbonyl
compounds, O3, Meteorological parameters, PM10,
PM10 Speciation, PM coarse, PM25,
PM25 Speciation.
TSP Speciation, Hexavalent chromium, CO, SO2,
NO, NO2, NOX, VOCs, Carbonyl compounds, O3,
NMOC, Meteorological parameters, PM10, PM10
Speciation, Black carbon, PM2 5, PM2 5 Speciation.
Data for additional pollutants are reported to AQS for these sites (EPA, 2012c); however, these data are not generated by ERG and are therefore not included in this report
BOLD ITALICS = EPA-designated NATTS Site
oo

-------
       CELA is located on the rooftop of a two-story building northeast of downtown Los
Angeles, just southeast of Dodgers' Stadium. Figure 6-1 shows that CELA is surrounded by
major freeways, including 1-5 and Route 110. Highway 101 is located farther south. Although
the area is classified as residential, a freight yard is located to the south of the site. The Los
Angeles River runs north-south just east of the site. This monitoring site was originally set up as
an emergency response monitor. As Figure 6-2 shows, CELA is situated among numerous point
sources. There is a cluster of emissions sources located just to the southwest of CELA. A large
number of emissions sources within 10 miles of CELA are involved in aircraft operations, which
include airports as well as small runways, heliports, or landing pads; furniture products;
electroplating, plating, polishing, anodizing, and coloring; and printing or publishing.

       RUCA is located just outside of Riverside, in a residential area of the suburban town of
Rubidoux. 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. Figure 6-3 shows that RUCA
is adjacent to a power substation near the intersection of Mission Boulevard and Riverview
Drive. RUCA and CELA are located less than 45 miles apart. Figure 6-4 shows that fewer
emissions sources surround RUCA than CELA.  Most of the emissions sources are located to the
northeast and northwest of the site. The point source located closest to RUCA is Flabob Airport.
The emissions source categories with the highest number of sources near RUCA include food
processing, aircraft operations, printing and publishing, and fabricated metals products.

       SJJCA is located in central San Jose. Figure 6-5 shows that SJJCA is located in a
commercial area surrounded by residential areas. A railroad is  shown just  east of the monitoring
site, running north-south in Figure 6-5. Guadalupe Parkway, which can be seen on the bottom
left of Figure 6-5, 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.
Figure 6-6 shows that the density of point sources is significantly higher near SJJCA than CELA
and RUCA. The emissions source categories with the highest number of sources are electrical
equipment; auto body/paint shops; dry cleaning; and telecommunications.
                                          6-9

-------
       Table 6-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the California monitoring sites.  Table 6-2 includes county-level
population and vehicle registration information.  Table 6-2 also includes a county-level vehicle
registration-to-population ratio, which was calculated to represent the number of vehicles per
person within each monitoring site's residing county. In addition, the population within 10 miles
of each site is presented, based on postal code population data estimates. An estimate of 10-mile
vehicle ownership was determined by applying the county-level vehicle registration-to-
population ratio to the 10-mile population surrounding each monitoring site. Table 6-2 also
contains traffic volume information for each site. Finally, Table 6-2 presents the county-level
daily VMT for Los Angeles, Riverside, and Santa Clara Counties.
     Table 6-2. Population, Motor Vehicle, and Traffic Information for the California
                                     Monitoring Sites
Site
CELA
RUCA
SJJCA
Estimated
County
Population1
9,889,056
2,239,620
1,809,378
County-level
Vehicle
Registration2
7,360,573
1,711,492
1,517,190
Vehicles per
Person
(Registration:
Population)
0.74
0.76
0.84
Population
within 10
miles3
3,557,102
1,026,542
1,482,077
Estimated
10-mile
Vehicle
Ownership
2,647,604
784,472
1,242,743
Annual
Average
Daily
Traffic4
230,000
145,000
104,000
County-
level Daily
VMT5
214,458,140
55,717,760
41,250,490
  Bounty-level population estimates reflect 2011 data from the U.S. Census Bureau (Census Bureau, 2012b)
  2County-level vehicle registration reflects 2011 data from the California DMV (CA DMV, 2011)
  310-mile population estimates reflect 2011 data from Xionetic (Xionetic, 2011)
  4AADT reflects 2011 data from the California DOT (CA DOT, 2011)
  5County-level VMT reflects 2011 data from the California DOT (CA DOT, 2012)
  BOLD ITALICS = EPA-designated NATTS Site

       Observations from Table 6-2 include the following:

       •  Los Angeles County (CELA) has the highest county-level population and vehicle
          registration compared to all counties with NMP sites. CELA also has the highest
          10-mile estimated vehicle ownership. However, the 10-mile population near this site
          ranks second behind MONY, which is located in Bronx County and part of New York
          City.

       •  Riverside and Santa Clara Counties are also in the top 10 for county-level population
          and vehicle registration among counties with NMP sites.

       •  Among the California sites, the vehicle-per-person ratio is lowest for the most
          populous area (CELA) and highest for the least populated area (SJJCA), based on
          county population, although all three sites are located in highly populated areas. In
          general, this trend is also true among all NMP sites.
                                           6-10

-------
       •  CELA experiences the second highest annual average daily traffic among NMP sites,
          and has a substantially higher traffic volume than both RUCA and SJJCA (although
          all three rank in the top third among NMP sites). The traffic count for CELA is based
          on data from 1-5 near Exit 136A at Main Street. The traffic count for RUCA is based
          on data from Mission Boulevard at Rubidoux Boulevard. The traffic count for SJJCA
          is based on the intersection of Guadalupe Parkway at West Taylor Street.
       •  Los Angeles County's daily VMT is the highest among all counties with NMP sites,
          where VMT was available. This VMT is an order of magnitude higher than the next
          highest county-level VMT (Maricopa County, AZ).  The VMT for Riverside and
          Santa Clara Counties are also in the top 10 for VMT among counties with NMP sites
          (where VMT data were available).

6.2    Meteorological Characterization
       The following sections characterize the meteorological conditions near the monitoring
sites in California on sample days, as well as over the course of the year.
6.2.1   Climate Summary
       The climate of Los Angeles and the surrounding areas is generally mild. While the
proximity to the Pacific Ocean acts as a moderating influence on the Los Angeles area, the
elevation changes between the mountains and valleys allow the distance from the ocean to create
substantial differences in temperature, rainfall, and wind over a relatively short distance.
Precipitation falls primarily in winter months, while summers tend to be dry. 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 (Bair, 1992 and WRCC, 2013).

       San Jose is located to the southeast of San Francisco, near the base of the San Francisco
Bay. The city is situated in the Santa Clara Valley, between the Santa Cruz Mountains to the
south and west and the Diablo Range to the east. San Jose experiences a Mediterranean  climate,
with distinct wet-dry seasons.  The period from November through March represents the wet
season, with cool but mild conditions prevailing. Little rainfall occurs the rest of the year and
conditions tend to be warm and sunny. San Jose is not outside the marine influences of the cold
ocean currents typically affecting the San Francisco area (Bair,  1992 and NOAA, 1999).
                                          6-11

-------
6.2.2   Meteorological Conditions in 2011
       Hourly meteorological data from the NWS weather stations nearest these sites were
retrieved for 2010 and 2011 (NCDC, 2010 and 2011). The weather station nearest CELA is
located at Downtown Los Angeles/USC Campus; the nearest NWS weather station to RUCA is
located at Riverside Municipal Airport; and the nearest NWS station to SJJCA is located at San
Jose International (WBAN 93134, 03171, and 23293, respectively). Additional information
about these weather stations, such as the distance between the sites and the weather stations, is
provided in Table 6-3. These data were used to determine how meteorological conditions on
sample days vary from conditions experienced throughout the year.

       Table 6-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for the entire year for 2011. For SJJCA, meteorological data for the
additional days from 2010, as discussed in Section 2.3, have been included in the averages
presented in Table 6-3. Also included in Table 6-3 is the 95 percent confidence interval for each
parameter. As shown in Table 6-3, average meteorological conditions on sample days near these
sites were representative of average weather conditions throughout the year. The most significant
difference in the table is for average relative humidity for RUCA. Table 6-3 shows a marked
wind speed difference between CELA and RUCA (which are located 45 miles apart), as alluded
to in Section 6.2.1, although wind speeds for both sites are very light. As expected, conditions
tended to be cooler near SJJCA than near CELA and RUCA.
                                         6-12

-------
                  Table 6-3. Average Meteorological Conditions near the California Monitoring Sites
Closest NWS Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(ฐF)
Average
Temperature
(ฐF)
Average
Dew Point
Temperature
(ฐF)
Average
Wet Bulb
Temperature
(ฐF)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar
Wind
Speed
(kt)
Los Angeles, California - CELA
Downtown
L.A./USC Campus
Airport
93134
(34.03, -118.30)
4.57
miles
248ฐ
(WSW)
Sample
Day
2011
72.8
ฑ2.4
72.2
ฑ0.9
62.7
ฑ1.8
62.6
ฑ0.7
47.8
ฑ2.5
48.2
ฑ1.0
55.0
ฑ1.7
55.1
ฑ0.7
62.7
ฑ3.6
63.7
ฑ1.4
1015.0
ฑ1.0
1015.1
ฑ0.4
1.0
ฑ0.2
1.2
ฑ0.1
Rubidoux, California - RUCA
Riverside Municipal
Airport
03171
(33.95, -117.44)
3.49
miles
214ฐ
(SW)
Sample
Day
2011
78.3
ฑ3.5
77.9
ฑ1.3
64.8
ฑ2.6
64.3
ฑ1.0
42.3
ฑ3.6
43.8
ฑ1.3
53.6
ฑ2.1
53.8
ฑ0.8
51.5
ฑ4.4
54.2
ฑ1.7
1014.0
ฑ1.1
1014.1
ฑ0.4
3.8
ฑ0.4
3.8
ฑ0.2
San Jose, California - SJJCA
San Jose Intl.
Airport
23293
(37.36, -121.93)
1.90
miles
316ฐ
(NW)
Sample
Day
Dec
2010-
2011
69.4
ฑ2.3
68.7
ฑ1.0
58.6
ฑ1.7
58.2
ฑ0.8
46.5
ฑ2.1
46.5
ฑ0.9
52.4
ฑ1.6
52.2
ฑ0.7
67.9
ฑ2.9
68.6
ฑ1.2
1016.8
ฑ1.1
1016.8
ฑ0.5
4.9
ฑ0.5
5.3
ฑ0.2
Sample day averages are highlighted to help differentiate the sample day averages from the full-year averages.

-------
6.2.3   Back Trajectory Analysis

       Figure 6-7 is the composite back trajectory map for days on which samples were

collected at the CELA monitoring site in 2011. Included in Figure 6-7 are four back trajectories

per sample day. Figure 6-8 is the corresponding cluster analysis. Similarly, Figures 6-9 and 6-11

are the composite back trajectory maps for days on which samples were collected at RUCA and

SJJCA and Figures 6-10 and 6-12 are the corresponding cluster analyses, respectively. An in-

depth description of these maps and how they were generated is presented in Section 3.5.2.1. For

the composite maps, each line represents the 24-hour trajectory along which a parcel of air

traveled toward the monitoring site on a given sample day and time, based on an initial height of

50 meters AGL. For the cluster analyses, each line corresponds to a trajectory representative of a

given cluster of back trajectories. Each concentric circle around the sites in Figures 6-7 through

6-12 represents 100 miles.


       Observations from Figures 6-7 and 6-8 for CELA include the following:

       •  The 24-hour air shed domain for CELA is among the smaller ones compared to other
          NMP monitoring sites, based on the average length of trajectories. Although the
          farthest away a trajectory originated was over southwest Montana, or greater than
          800 miles away, most trajectories (83 percent) originated within 300 miles of CELA
          (and only eight back trajectories originated  greater than 500 miles away).

       •  Back trajectories originated from  a variety of directions at CELA. However, a large
          number of trajectories originated from the northwest over the Pacific Ocean and along
          the California coastline. Another cluster originated from the east-northeast. Few
          trajectories originated from the east, southeast, south, or southwest. The longest back
          trajectories originated to the northeast of the site.

       •  The cluster analysis shows that roughly 70 percent of trajectories originated from the
          northwest and/or offshore, although of varying distances.  The shorter cluster
          trajectory (19 percent) includes back trajectories originating to the northwest of Los
          Angeles as well as shorter trajectories originating just offshore. The cluster analysis
          also shows that 25 percent of trajectories originated from the northeast quadrant
          (including northerly and easterly directions). The long cluster  (3 percent) originating
          over Idaho represents the longer trajectories originating over northeast Utah, Idaho,
          and Montana.
                                           6-14

-------
Figure 6-7. 2011 Composite Back Trajectory Map for CELA
    Figure 6-8. Back Trajectory Cluster Map for CELA
                         6-15

-------
Figure 6-9. 2011 Composite Back Trajectory Map for RUCA
   Figure 6-10. Back Trajectory Cluster Map for RUCA
                          6-16

-------
Figure 6-11. 2011 Composite Back Trajectory Map for SJJCA
    Figure 6-12. Back Trajectory Cluster Map for SJJCA
                          6-17

-------
Observations from Figures 6-9 and 6-10 for RUCA include the following:

•  The composite back trajectory map for RUCA resembles the one for CELA, which is
   not surprising given their relatively close proximity. The 24-hour air shed domain for
   RUCA is similar in size to CELA. The farthest away a trajectory originated was also
   over western Montana, or greater than 850 miles away. Most trajectories (86 percent)
   originated within 300 miles of RUCA and only eight trajectories originated farther
   than 500 miles away.

•  Back trajectories originated from a variety of directions at RUCA. A large cluster of
   trajectories originated from the northwest of the site and a secondary cluster
   originated from the northeast. Few trajectories originated from the east, southeast, or
   south. The longest back trajectories originated to the northeast of the site.

•  The cluster analysis for RUCA is similar to the cluster analysis for CELA.
   Figure 6-10 shows that nearly 70 percent of trajectories originated from the northwest
   of the site. However, the shorter trajectory (31 percent) includes some back
   trajectories originating from the northwest along the coastline and offshore as well as
   shorter trajectories originating from the west (either offshore or inland, as the
   clustering program uses both direction and distance to determine clusters). Another
   cluster (29 percent) represents those trajectories originating from the north, northeast,
   and east over southern California, southern Nevada, northwest Arizona, and
   southwest Utah. The long cluster trajectory marked with 3 percent represents the
   eight back trajectories originating over the inter-mountain west.


Observations from Figures 6-11 and 6-12 for SJJCA include the following:

•  Based on the average length of the back trajectories, the 24-hour air shed domain for
   SJJCA is somewhat larger than the air shed domains for the other two California
   sites. The farthest away a back trajectory originated was also over western Montana
   or just greater than 700 miles away, which is shorter than the longest back trajectories
   for CELA and RUCA. Only 71 percent of back trajectories originated within
   300 miles of SJJCA, while greater than 80 percent originated within 300 miles of
   CELA and RUCA. Eight back trajectories originated farther than 500 miles away
   from the site. Recall from Section 2.3, that three additional sample days from
   December 2010 are included in the composite back trajectory map for SJJCA.

•  Back trajectories originated from a variety of directions at SJJCA, seemingly more so
   than for the  other two California sites. However, the composite map still shows a
   large number of trajectories originated from the northwest and along the coast. Fewer
   back trajectories originated from the southeast quadrant.

•  The cluster analysis shows that 50 percent of trajectories originated from the
   northwest to northeast, with the cluster program grouping those from  the northwest,
   north, and northeast together. Back trajectories originating from the northeast to east
   to southeast are represented by  the cluster trajectory originating to the east
   (14 percent). The short cluster (18 percent) originating just offshore includes back
   trajectories less than 200 miles  in length from a variety of directions as well as those
   longer trajectories originating from the south. Fifteen percent of back trajectories

                                    6-18

-------
          originated farther offshore. The longest trajectory (3 percent) includes the eleven
          trajectories originating over the Pacific Northwest and Idaho.

6.2.4   Wind Rose Comparison
       Hourly surface wind data from the NWS weather stations at the Downtown Los
Angeles/USC Campus (for CELA), Riverside Municipal Airport (for RUCA), and San Jose
International Airport (for SJJCA) were uploaded into a wind rose software program to produce
customized wind roses, as described in Section 3.5.2.2. A wind rose shows the frequency of wind
directions using "petals" positioned around a 16-point compass, and uses different colors to
represent wind speeds.

       Figure 6-13 presents a map showing the distance between the NWS station and CELA,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 6-13 also presents three different wind roses for the
CELA monitoring site. First, a historical wind rose representing 2001 to 2010 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 2011 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2011 is
presented. These can be used to identify the predominant wind speed and direction for 2011 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically. Figures 6-14 and 6-15 present the distance maps and wind
roses for RUCA and SJJCA, respectively.
                                          6-19

-------
Figure 6-13. Wind Roses for the Downtown Los Angeles/USC Campus Weather Station
                                   near CELA
 Distance between CELA and NWS Station
2001-2010 Historical Wind Rose
                                                                           WIND SPEED
                                                                           (Knots)

                                                                           n ซ=
                                                                           ^| 17 • 21
                                                                           JH 11 - 17

                                                                            H I-7
                                                                           !• 2- 4
                                                                           Calms: 33.72%
          2011 Wind Rose
    Sample Day Wind Rose
	 _

,,."-*""""""




^^^^p^^
WEST •; 	 -T 	

v",%


NORTH'---.,

""S 10%
*"*% 6%-%
4%,
f 2% '; ! !
—•CJ---A 	 t 	 > 	
^>T ; I IEAST







WEST •— ^— *^~*
,"•"" '' ' WIND SPEED \ \ ""---....
(Knots) '\ %v-,
..-'' ,'' I I ซ22 *-, "'-- 	
-"'' •
SOUTH---'" --



•
17-21 *~-v>
11 - 17 ~---------
NORTH''---,

10%
""-^ 8%,
""--,^ 6%-s
4*,
2% ': ';
^T"; : : EAST
^•'' / / WND SPEED
(Knots)
.,-•""' ,-'' n -22
^| 17-21
SOUTH---"' • 11'17
7- 11 ^^
D 4-7
4- 7 ^_
• 2- 4
2- 4
Calm;: ?2^:i'S.
                              Calms: 89.39%
                                      6-20

-------
Figure 6-14. Wind Roses for the Riverside Municipal Airport Weather Station near RUCA
  Distance between RUCA and NWS Station
2001-2010 Historical Wind Rose

                                                                                     WIND SPEED
                                                                                     (Knots)
                                                                                     || >= 22
                                                                                     ^| 17 • 21
                                                                                     ^| 11 • 17



                                                                                     Calms: 30.74%
             2011 Wind Rose
                                   ™ND SPEED
                                   (Knots)
                                   cn 4-7
                                   Calms: 35.16%
                                                            Sample Day Wind Rose
                              WIND SPEED
                              (Knots)
                              |1 >= 22
                              ^| 17 - 21
                              ^| 11 • 17
                              IB 7- 11
                              |  | 4- 7
                              IB 2- 4
                              Calms: 33.88%
                                            6-21

-------
Figure 6-15. Wind Roses for the San Jose International Airport Weather Station near
                                    SJJCA
Distance between SJJCA and NWS Station
2001-2010 Historical Wind Rose
          2011 Wind Rose
                                                   Sample Day Wind Rose
                                     6-22

-------
Observations from Figure 6-13 for CELA include the following:

•  The NWS weather station at the Downtown Los Angeles/USC Campus is located
   approximately 4.6 miles west-southwest of CELA.

•  Historically, winds were generally light near this site, with calm winds (< 2 knots)
   observed for nearly 60 percent of the wind observations. For wind speeds greater than
   2 knots, westerly winds were most common, followed by west-southwesterly and
   easterly winds. Wind speeds greater than 11 knots were not measured at this weather
   station.

•  The 2011 full-year and sample day wind roses are similar to the historical wind rose
   in that calm winds make up the bulk of the observations and that westerly winds were
   prominent. The wind patterns shown on the full-year and sample day wind roses
   generally resemble the historical wind patterns, indicating that conditions in 2011 and
   on sample days were representative of those experienced historically. However, west-
   southwesterly winds were rarely observed in 2011.


Observations from Figure 6-14 for RUCA include the following:

•  The NWS weather station at the Riverside Municipal Airport is located across the
   Santa Ana River and Wildlife Area, approximately 3.5 miles southwest of RUCA.

•  Although calm winds were observed for approximately 31 percent of the wind
   observations near RUCA, westerly and west-northwesterly winds were also
   frequently observed, based on the historical wind rose.

•  The 2011 wind rose exhibits a slightly  higher percentage of calm winds compared to
   the historical wind rose. Westerly winds make up almost the same percentage of wind
   observations in 2011 as both westerly and west-northwesterly winds on the historical
   wind rose, as west-northwesterly winds were observed less frequently in 2011 than
   historically.

•  The wind patterns shown on the sample day wind rose resemble the wind patterns
   shown on the full-year wind rose, indicating that conditions on sample days were
   representative of those experienced over the entire year.


Observations from Figure 6-15 for SJJCA include the following:

•  The NWS weather station at the San Jose International Airport is located
   approximately 2 miles northwest of SJJCA, across 1-880, the Guadalupe Parkway,
   and the Guadalupe River.

•  Between 2001 and 2010, approximately 47 percent of winds were from the west-
   northwest to north. Another 20 percent of winds were from the east-southeast to
   south. Winds from the northeastern and southwestern quadrants were rarely observed.
   Approximately one-fifth of the winds were calm.
                                   6-23

-------
       •  The additional days from December 2010 have been included in the 2011 full-year
          and sample day wind roses, as discussed in Section 2.3.
       •  The wind patterns on the full-year and sample day wind roses exhibit a shift in
          primary wind direction, from northwest to north on the historical wind rose to west to
          northwest on the 2011 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 2011 wind roses. This shift was also shown on the 2009 and 2010 wind roses in
          the 2008-2009 and 2010 NMP reports.
       •  The wind patterns shown on the sample day wind rose resemble the wind patterns
          shown on the full-year wind rose, indicating that conditions on sample days were
          representative of those experienced over the entire year.

6.3    Pollutants of Interest
       Site-specific "pollutants of interest" were determined for the California monitoring sites
in order to allow 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. In addition, if any of the NATTS MQO Core Analytes measured by each monitoring site
did not meet the pollutant of interest criteria based on the preliminary risk-based screening, that
pollutant was added to the list of site-specific pollutants of interest. A more in-depth description
of the risk-based screening process is presented in Section 3.2.
       Table 6-4 presents the results of the preliminary risk-based screening process for CELA,
RUCA, and SJJCA. The pollutants that failed at least one screen and contributed to 95 percent of
the total failed screens for each monitoring site are shaded. NATTS MQO Core Analytes are
bolded. Thus, pollutants of interest are shaded and/or bolded. CELA and RUCA sampled for
PAHs only, while SJJCA sampled for metals (PMio) and PAHs.
                                          6-24

-------
 Table 6-4. Risk-Based Screening Results for the California Monitoring Sites
Pollutant
Screening
Value
(Hg/m3)
# of Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
Los Angeles, California - CELA
Naphthalene
Acenaphthene
Fluorene
0.029
0.011
0.011
Total
58
3
3
64
59
59
59
177
98.31
5.08
5.08
36.16
90.63
4.69
4.69
90.63
95.31
100.00

Rubidoux, California - RUCA
Naphthalene
0.029
Total
54
54
61
61
88.52
88.52
100.00
100.00

San Jose, California - SJJCA
Naphthalene
Arsenic (PM10)
Manganese (PM10)
Nickel (PM10)
Benzo(a)pyrene
Acenaphthylene
0.029
0.00023
0.005
0.0021
0.00057
0.011
Total
52
43
35
8
3
2
143
61
63
64
64
19
27
298
85.25
68.25
54.69
12.50
15.79
7.41
47.99
36.36
30.07
24.48
5.59
2.10
1.40
36.36
66.43
90.91
96.50
98.60
100.00

Observations from Table 6-4 include the following:

•  Naphthalene failed the majority of screens for all three California monitoring sites,
   with its site-specific contribution to the total failed screens ranging from 36 percent
   (SJJCA) to 100 percent (RUCA).

•  Three pollutants failed screens for CELA, including one NATTS MQO Core Analyte
   (naphthalene). Benzo(a)pyrene was added to the pollutants of interest for CELA
   because it is a NATTS MQO Core Analyte, even though it did not fail any screens.
   Benzo(a)pyrene is not shown in Table 6-4 but is shown in subsequent tables in the
   sections that follow.

•  Naphthalene was the only pollutant to fail screens for RUCA. Similar to CELA,
   benzo(a)pyrene was added to RUCA's pollutants of interest because it is a NATTS
   MQO Core Analyte, even though it did not fail any screens. Benzo(a)pyrene is not
   shown in Table 6-4 but is shown in subsequent tables in the sections that follow.

•  Six pollutants failed screens for SJJCA, of which five are NATTS MQO Core
   Analytes. Four of these were initially identified as pollutants of interest for SJJCA.
   Benzo(a)pyrene was added as a pollutant of interest, even though it did not contribute
   to 95 percent of SJJCA's total failed screens, because it is a NATTS MQO Core
   Analyte. Three additional NATTS MQO  Core Analytes were added to the pollutants
   of interest for SJJCA, even though their concentrations did not fail any screens:
   beryllium, cadmium, and lead. These three pollutants are not shown in Table 6-4, but
   are shown in subsequent tables in the sections that follow.
                                   6-25

-------
       •  The instrumentation at SJJCA for sampling metals was changed in mid-December
          2010 and the site began sampling metals on Teflonฎ filters rather than quartz filters.
          In light of this change, the three samples from 2010 on the new instrumentation were
          included in all data analyses contained in this report.

6.4    Concentrations
       This section presents various concentration averages used to characterize pollution levels
at the California monitoring sites. Where applicable, the following calculations and data analyses
were performed: Time period-based concentration averages (quarterly and annual) are provided
for the pollutants of interest for the California monitoring sites, where the data meet the
applicable criteria. Concentration averages for select pollutants are also presented graphically for
the sites to illustrate how the sites' concentrations compare to the program-level averages, as
presented in Section 4.1. In addition, concentration averages for select pollutants are presented
from previous years of sampling in order to characterize concentration trends at the sites.
Additional site-specific statistical summaries for CELA, RUCA, and SJJCA are provided in
Appendices M and N.

6.4.1   2011 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 of the total number of
samples possible within a given quarter for a quarterly average to be calculated. An annual
average includes all measured detections and substituted zeros for non-detects  for the entire year
of sampling. Annual averages were calculated for pollutants where three valid  quarterly averages
could be calculated and where method completeness was greater than or equal  to 85 percent, as
presented in Section 2.4. Quarterly and annual average concentrations for the California
monitoring sites are presented in Table 6-5, where applicable. Note that if a pollutant was not
detected in a given calendar quarter, the quarterly average simply reflects "0" because only zeros
substituted for non-detects were factored into the quarterly average concentration. Note also that
the three December 2010 metals samples for SJJCA have been included in the  first quarter
averages shown in Table 6-5.
                                           6-26

-------
Table 6-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest
                         for the California Monitoring Sites
Pollutant
#of
Measured
Detections vs.
# of Samples
1st
Quarter
Average
(ng/m3)
2nd
Quarter
Average
(ng/m3)
3rd
Quarter
Average
(ng/m3)
4th
Quarter
Average
(ng/m3)
Annual
Average
(ng/m3)
Los Angeles, California - CELA
Acenaphthene
Benzo(a)pyrene
Fluorene
Naphthalene
59/59
41/59
59/59
59/59
3.96
ฑ1.25
0.06
ฑ0.02
4.92
ฑ1.25
160.49
ฑ50.18
3.89
ฑ1.11
0.02
ฑ0.01
4.69
ฑ1.22
78.35
ฑ 22.87
6.06
ฑ1.39
0.03
ฑ0.01
6.54
ฑ1.42
118.30
ฑ 30.47
5.83
ฑ2.18
0.12
ฑ0.05
6.37
ฑ1.78
173.46
ฑ51.51
4.92
ฑ0.76
0.06
ฑ0.02
5.62
ฑ0.70
131.96
ฑ21.23
Rubidoux, California - RUCA
Benzo(a)pyrene
Naphthalene
25/61
61/61
0.02
ฑ0.01
80.99
ฑ26.99
0.01
ฑ0.01
55.39
ฑ 17.96
0.01
ฑ0.01
85.68
ฑ 24.06
0.08
ฑ0.06
143.05
ฑ50.61
0.03
ฑ0.02
91.18
ฑ17.14
San Jose, California - SJJCA
Arsenic (PM10)
Benzo(a)pyrene
Bery Ilium (PM10)
Cadmium (PM10)
Lead (PM10)
Manganese (PM10)
Naphthalene
Nickel (PM10)
63/64
19/61
55/64
64/64
64/64
64/64
61/61
64/64
0.41
ฑ0.17
0.09
ฑ0.05
0.01
ฑ<0.01
0.07
ฑ0.03
2.78
ฑ1.19
5.78
ฑ2.19
103.53
ฑ44.91
1.09
ฑ0.36
0.24
ฑ0.07
0
0.01
ฑ<0.01
0.04
ฑ0.01
1.73
ฑ0.43
4.00
ฑ0.95
34.63
ฑ8.01
1.05
ฑ0.27
0.32
ฑ0.09
0.01
ฑ0.01
0.01
ฑ<0.01
0.04
ฑ0.01
2.70
ฑ1.09
6.70
ฑ1.55
45.44
ฑ 10.88
1.35
ฑ0.29
0.61
ฑ0.14
0.20
ฑ0.17
0.01
ฑ<0.01
0.13
ฑ0.05
4.92
ฑ1.07
10.19
ฑ3.12
138.94
ฑ42.95
1.61
ฑ0.38
0.39
ฑ0.07
0.07
ฑ0.05
0.01
ฑ<0.01
0.07
ฑ0.02
3.02
ฑ0.56
6.62
ฑ1.14
80.06
ฑ 18.30
1.27
ฑ0.17
 Observations for the California monitoring sites from Table 6-5 include the following:

 •   Naphthalene and benzo(a)pyrene are pollutants of interest for each site.

 •   The pollutant with the maximum annual average concentration for each site is
     naphthalene, which ranged from 80.06 ฑ 18.30 ng/m3 for  SJJCA to
     131.96 ฑ 21.23 ng/m3 for CELA. The naphthalene concentrations for SJJCA exhibit a
     seasonal trend, with higher concentrations measured during the colder months of the
     year (first and fourth  quarters). Naphthalene concentrations for CELA appear to
     exhibit this trend too, although the difference in the quarterly averages is not
     statistically significant. Naphthalene concentrations for RUCA appear highest during
     the fourth quarter, but the relatively high confidence interval associated with this
     average indicates the possible influence of outliers.
                                     6-27

-------
       •  Benzo(a)pyrene concentrations also appear to be higher during the colder months of
          the year, particularly for SJJCA. However, this pollutant was detected in just over
          one-quarter of the samples collected at SJJCA. Benzo(a)pyrene was not detected at all
          during the second quarter at SJJCA and was detected only twice during the third
          quarter of 2011.

       •  Manganese has the highest annual average concentration (6.62 ฑ1.14 ng/m3) of the
          PMio metals measured at SJJCA. The quarterly averages for manganese show that the
          highest concentrations were measured during the fourth quarter of 2011.  A review of
          the data shows that of the 11 concentrations of manganese greater than 10 ng/m3
          measured at this site, seven were measured during the fourth quarter of 2011,
          including the two highest concentrations (25.4 ng/m3 measured on November 2, 2011
          and 18.1 ng/m3 measured on October 27, 2011).

       •  The highest quarterly average concentration was calculated for the fourth quarter  of
          2011 for most of the PMio metals.


       Tables 4-9 through 4-12 present the sites with the 10 highest annual average

concentrations for each of the program-level pollutants of interest. Observations for the

California sites from those tables include the following:

       •  CELA appears in Table 4-11 for PAHs three times. CELA has the fourth highest
          annual average concentration naphthalene among NMP sites sampling PAHs, behind
          GPCO, DEMI, and MONY. This site also ranks eighth for acenaphthene and
          fluorene, but does not appear in Table 4-11 for benzo(a)pyrene (it ranks 14th). RUCA
          has the ninth highest annual average concentration of naphthalene and tenth highest
          annual average concentration of fluorene. SJJCA does not appear in Table 4-11.

       •  Because only nine sites sampled PMio metals, SJJCA appears in Table 4-12 for every
          program-level metal pollutant of interest. However, this site was not in the top five
          for any of the speciated metals.


6.4.2   Concentration Comparison

       In order to better illustrate how a site's annual average concentrations compare to the

program-level averages, a site-specific box plot was created for the selected NATTS MQO Core

Analytes listed in Section 3.5.3, where applicable. Thus, box plots for benzo(a)pyrene and

naphthalene were created for all three California sites. Box plots were also created for arsenic,

lead, and manganese for SJJCA. Figures 6-16 through 6-20 overlay the sites' minimum, annual

average, and maximum concentrations onto the program-level minimum, first quartile, median,

average, third quartile, and maximum concentrations, as described in Section 3.5.3.
                                          6-28

-------
      Figure 6-16. Program vs. Site-Specific Average Arsenic (PMi0) Concentration
SJJCA
             DL5
                                 1.5
                                           2         2.5
                                           Concentration (ng/mi)
                                                                         3.5
Program
Site:
: IstQuartile
Site Average
o
2ndQuartile 3rdQuartile 4thQuartile Avc
• n
Site Minimum/Maximum
'rage

                                                                                             4.5
     Figure 6-17. Program vs. Site-Specific Average Benzo(a)pyrene Concentrations
CELA
i
                                     0.75          1          1.Z5
                                            Concentration (ng/m3)
                Program:   IstQuartile      2ndQuartile     3rdQuartile     4thQuartile     Average
                Site:      Site Average    Site Minimum/Maximum

        Figure 6-18. Program vs. Site-Specific Average Lead (PMio) Concentration
                                          15            20
                                           Concentration (ng/m3)
Program
Site:

: IstQuartile
Site Average
0
2ndQuartile SrdQuartile 4thQuartile Ave
• n
Site Minimum/Maximum


rage


                                             6-29

-------
    Figure 6-19. Program vs. Site-Specific Average Manganese (PMi0) Concentration
SJJCA
                                                         Program Max Concentration = 395 ng/m3
                                   75         100         125
                                         Concentration (ng/m3)
                                                                  ISO
                                                                             175
                                                                                       200
               Program:   IstQuartile     2ndQuartile     SrdQuartile     4thQuartile     Average
               Site:      Site Average    Site Minimum/Maximum
                          o
       Figure 6-20. Program vs. Site-Specific Average Naphthalene Concentrations
CELA
                                                            [  Program Max Concentration =779 ng/m3
                                                            : P rogra m M a x Co n ce ntratio n = 779 ng/m3
                                                          1 i Program Max Concentration = 779 ng/m3
                                     200       250       300
                                         Concentration (ng/m3)
Program
Site:
: IstQuartile
Site Average
O
2ndQuartile SrdQuartile 4thQuartile Ave
Site Minimum/Maximum
rage

      Observations from Figures 6-16 through 6-20 include the following:

          •   Figure 6-16 shows that the annual average arsenic (PMio) concentration for
             SJJCA is less than the program-level average and median concentrations of
             arsenic (PMio). A single non-detect of arsenic was measured at SJJCA.

          •   Figure 6-17 for benzo(a)pyrene shows all three sites. Note that the program-level
             first quartile for this pollutant is zero and is not visible on the box plots. Each of
             the annual average concentrations of benzo(a)pyrene for the California sites is
             less than the program-level average concentration. Figure 6-17 also shows that
             while the maximum benzo(a)pyrene concentration measured at each California
             site is less than the maximum concentration measured across the program, the
             maximum concentration measured at SJJCA is more than twice the maximum
             concentration measured at RUCA and three times the maximum concentration
                                           6-30

-------
             measured at CELA. Non-detects of benzo(a)pyrene were reported for each of
             these three sites.

          •  Figure 6-18 shows that the annual average concentration of lead (PMio) for
             SJJCA is less than the program-level average and but greater than the program-
             level median concentration. The maximum lead concentration measured at SJJCA
             is less than the maximum concentration measured across the program. There were
             no non-detects of lead measured at SJJCA or across the program.

          •  Figure 6-19 is the box plot for manganese (PMio). Note that the program-level
             maximum concentration (395 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
             200 ng/m3. Figure 6-19  shows that the annual average concentration of
             manganese (PMio) for SJJCA is also less than the program-level average
             concentration. The maximum manganese concentration measured at SJJCA is
             considerably less than the maximum concentration measured across the program.
             There were no non-detects of manganese measured at SJJCA or across the
             program.

          •  Figure 6-20 for naphthalene also shows all three sites. Note that the program-level
             maximum concentration (779 ng/m3) is not shown directly on the box plot as the
             scale has been reduced to 500 ng/m3 to allow for the observation of data points at
             the lower end of the concentration range. Figure 6-20 shows that the annual
             average concentrations for CELA and RUCA are greater than the program-level
             average concentration, while the annual average concentration for SJJCA  is just
             less than the program-level average. Figure 6-20 also shows that the maximum
             concentrations for all three sites are less than the maximum concentration
             measured across the program. However, CELA's maximum concentration
             (430 ng/m3) is the fifth highest concentration measured among sites sampling
             PAHs. There were no non-detects  of naphthalene measured at CELA, RUCA, or
             SJJCA or across the program.


6.4.3   Concentration Trends

       A site-specific trends evaluation was completed for  sites that have sampled one or more

of the selected NATTS MQO Core Analytes for 5 consecutive years or longer, as described in
Section 3.5.4. Although both CELA and RUCA began sampling PAHs in 2007, they did  not

begin until April and May, respectively, and therefore average concentrations could not be
calculated for 2007. SJJCA did not begin sampling under the NMP until 2008. Therefore, the

trends analysis was not conducted for the California sites.
                                         6-31

-------
6.5    Additional Risk-Based Screening Evaluations
       The following risk-based screening evaluations were conducted to characterize risk at
each California monitoring site. Refer to Sections 3.3 and 3.5.5 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.

6.5.1   Risk-Based Screening Assessment Using MRLs
       A risk-based screening was conducted by comparing the concentration data from the
California monitoring sites to the ATSDR MRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 to 14 days); intermediate (exposures of 15 to 364 days); and chronic (exposures
of 1 year or greater). The preprocessed daily measurements of the pollutants of interest were
compared to the acute MRLs; the quarterly averages were compared to the intermediate MRLs;
and the annual averages were compared to the chronic MRLs.

       As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2011.

6.5.2   Cancer Risk and Noncancer Hazard Approximations
       For the pollutants of interest for the California 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 may want to shift or
confirm their air-monitoring priorities.  Refer to Section 3.5.5.2 for an explanation of how cancer
risk and noncancer hazard approximations are calculated and what limitations are associated with
them. Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer
hazard approximations are presented in Table 6-6, where applicable. Cancer risk approximations
are presented as probabilities while the noncancer hazard approximations are ratios and thus,
unitless values.
                                          6-32

-------
            Table 6-6. Risk Approximations for the California Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
# of Measured
Detections vs.
# Samples
Annual
Average
(ng/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Los Angeles, California - CELA
Acenaphthene
Benzo(a)pyrene
Fluorene
Naphthalene
0.000088
0.00176
0.000088
0.000034



0.003
59/59
41/59
59/59
59/59
4.92
ฑ0.76
0.06
ฑ0.02
5.62
ฑ0.70
131.96
ฑ21.23
0.43
0.10
0.49
4.49



0.04
Rubidoux, California - RUCA
Benzo(a)pyrene
Naphthalene
0.00176
0.000034

0.003
25/61
61/61
0.03
ฑ0.02
91.18
ฑ17.14
0.05
3.10

0.03
San Jose, California - SJJCA
Arsenic (PM10)
Benzo(a)pyrene
Beryllium (PM10)
Cadmium (PM10)
Lead (PM10)
Manganese (PM10)
Naphthalene
Nickel (PM10)
0.0043
0.00176
0.0024
0.0018


0.000034
0.00048
0.000015

0.00002
0.00001
0.00015
0.00005
0.003
0.00009
63/64
19/61
55/64
64/64
64/64
64/64
61/61
64/64
0.39
ฑ0.07
0.07
ฑ0.05
0.01
ฑ0.01
0.07
ฑ0.02
3.02
ฑ0.56
6.62
ฑ1.14
80.06
ฑ 18.30
1.27
ฑ0.17
1.69
0.13
0.02
0.13


2.72
0.61
0.03

0
0.01
0.02
0.13
0.03
0.01
— = a Cancer URE or Noncancer RfC is not available.
       Observations for the California sites from Table 6-6 include the following:

       •  Naphthalene has the highest annual average concentration for each of the California
          sites, as discussed in the previous section.

       •  Naphthalene also has the highest cancer risk approximation among the pollutants of
          interest for all three California monitoring sites. The cancer risk approximations range
          from 2.72 in-a-million for SJJCA to 4.49 in-a-million for CELA.

       •  Of the metals sampled at SJJCA, arsenic has the highest cancer risk approximation
          and is the only metal for which a cancer risk approximation is greater than
          1.0 in-a-million (1.69 in-a-million).
                                          6-33

-------
       •  All of the noncancer hazard approximations for the pollutants of interest for the
          California monitoring sites are less than 1.0, indicating that no adverse health effects
          are expected from these individual pollutants.

6.5.3   Risk-Based Emissions Assessment
       In addition to the risk-based screenings discussed above, Tables 6-7 and 6-8 present 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 2008 NEI, the 10
pollutants with the highest toxicity-weighted emissions, and the  10 pollutants with the highest
cancer risk approximations (in-a-million), as calculated from the annual averages and provided
in Table 6-6. Table 6-8 presents similar information, but identifies the 10 pollutants with the
highest noncancer hazard approximations (HQ), also calculated from annual averages provided
in Table 6-6.

       The pollutants listed in Tables 6-7 and 6-8 are limited to  those that have cancer and
noncancer toxicity factors, respectively. As a result, although the actual value of the emissions is
the same, the highest emitted pollutants in the cancer table may be different from the noncancer
table. Further, the cancer risk and noncancer hazard approximations based on each site's annual
averages are limited to those pollutants for which each respective site sampled. As discussed in
Section 6.3,  all three California monitoring sites sampled for PAHs and SJJCA also sampled
PMio metals. In addition, the cancer risk and noncancer hazard approximations are limited to
those pollutants with enough data to meet the criteria for annual  averages to be calculated. A
more in-depth discussion of this analysis is provided in Section 3.5.5.3. Similar to the cancer risk
and noncancer hazard approximations, this analysis may help policy-makers prioritize their air
monitoring activities.
                                           6-34

-------
Table 6-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
                                             the California Monitoring Sites
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
Los Angeles, California (Los Angeles County) - CELA
Formaldehyde
Benzene
Dichloromethane
Acetaldehyde
Ethylbenzene
1,3 -Butadiene
Naphthalene
ฃ>-Dichlorobenzene
Tetrachloroethylene
Trichloroethylene
3,019.71
1,847.23
1,347.58
1,236.99
959.04
358.22
207.88
144.14
105.99
50.40
Formaldehyde
Hexavalent Chromium, PM
Benzene
1,3 -Butadiene
Naphthalene
Arsenic, PM
Nickel, PM
Acetaldehyde
Ethylbenzene
POM, Group la
3.93E-02
2.57E-02
1.44E-02
1.07E-02
7.07E-03
5.28E-03
5.01E-03
2.72E-03
2.40E-03
1.99E-03
Naphthalene 4.49
Fluorene 0.49
Acenaphthene 0.43
Benzo(a)pyrene 0.10

Rubidoux, California (Riverside County) - RUCA
Formaldehyde
Benzene
Acetaldehyde
Ethylbenzene
Tetrachloroethylene
Dichloromethane
1,3 -Butadiene
Naphthalene
1 ,3 -Dichloropropene
ฃ>-Dichlorobenzene
793.92
409.04
353.08
207.95
163.53
150.35
81.30
55.49
37.72
28.62
Hexavalent Chromium, PM
Formaldehyde
Benzene
1,3 -Butadiene
Naphthalene
Arsenic, PM
Acetaldehyde
POM, Group la
Ethylbenzene
Nickel, PM
2.13E-02
1.03E-02
3.19E-03
2.44E-03
1.89E-03
1.82E-03
7.77E-04
5.85E-04
5.20E-04
4.85E-04
Naphthalene 3.10
Benzo(a)pyrene 0.05


-------
   Table 6-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
                                           the California Monitoring Sites (Continued)
Oi
OJ
Oi
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
San Jose, California (Santa Clara County) - SJJCA
Formaldehyde
Benzene
Acetaldehyde
Ethylbenzene
Dichloromethane
Tetrachloroethylene
1,3 -Butadiene
Naphthalene
ฃ>-Dichlorobenzene
POM, Group la
577.88
353.73
254.78
201.33
102.75
97.93
74.71
39.30
25.48
19.28
Formaldehyde
Hexavalent Chromium, PM
Benzene
1,3 -Butadiene
POM, Group la
Naphthalene
Arsenic, PM
Acetaldehyde
Ethylbenzene
Nickel, PM
7.51E-03
7.13E-03
2.76E-03
2.24E-03
1.70E-03
1.34E-03
9.76E-04
5.61E-04
5.03E-04
2.82E-04
Naphthalene
Arsenic
Nickel
Benzo(a)pyrene
Cadmium
Beryllium
2.72
1.69
0.61
0.13
0.13
0.02


-------
Table 6-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
                                         RfCs for the California Monitoring Sites
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity- Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Noncancer Hazard
Approximation
Pollutant (HQ)
Los Angeles, California (Los Angeles County) - CELA
1,1,1 -Trichloroethane
Toluene
Formaldehyde
Benzene
Dichloromethane
Hexane
Acetaldehyde
Ethylbenzene
Xylenes
Ethylene glycol
7,345.47
5,503.09
3,019.71
1,847.23
1,347.58
1,286.59
1,236.99
959.04
873.15
655.71
Acrolein
Chlorine
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Nickel, PM
Arsenic, PM
Hexamethylene- 1 ,6-diisocyanate, gas
Naphthalene
Benzene
8,698,970.49
367,827.04
308,133.92
179,110.50
137,443.72
115,985.91
81,886.64
78,379.05
69,291.67
61,574.38
Naphthalene 0.04

Rubidoux, California (Riverside County) - RUCA
Toluene
1,1,1 -Trichloroethane
Formaldehyde
Benzene
Acetaldehyde
Hexane
Ethylbenzene
Xylenes
Tetrachloroethylene
Dichloromethane
1,289.54
807.28
793.92
409.04
353.08
300.78
207.95
178.62
163.53
150.35
Acrolein
Chlorine
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Arsenic, PM
Naphthalene
Hexavalent Chromium, PM
Bromomethane
Benzene
1,964,801.08
87,536.25
81,012.73
40,650.19
39,230.78
28,239.83
18,496.03
17,714.98
16,899.96
13,634.82
Naphthalene 0.03


-------
   Table 6-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
                                      RfCs for the California Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity- Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer Hazard
Approximation
(HQ)
San Jose, California (Santa Clara County) - SJJCA
Toluene
1,1,1 -Trichloroethane
Formaldehyde
Benzene
Hexane
Acetaldehyde
Ethylbenzene
Xylenes
Ethylene glycol
Propionaldehyde
1,394.35
1,290.78
577.88
353.73
265.48
254.78
201.33
176.68
120.59
113.19
Acrolein
Chlorine
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Arsenic, PM
Propionaldehyde
Naphthalene
Benzene
Lead, PM
1,999,590.36
109,927.46
58,967.45
37,354.73
28,308.52
15,130.81
14,149.25
13,100.12
11,790.92
10,622.90
Manganese
Arsenic
Naphthalene
Lead
Nickel
Cadmium
Beryllium
0.13
0.03
0.03
0.02
0.01
0.01
0.01

oo

-------
Observations from Table 6-7 include the following:

•  Formaldehyde and benzene are the two highest emitted pollutants with cancer UREs
   in all three California counties. The quantity emitted is greater for Los Angeles
   County than Riverside and Santa Clara Counties.

•  Formaldehyde and hexavalent chromium are the pollutants with the highest toxicity-
   weighted emissions (of the pollutants with cancer UREs) for Los Angeles and Santa
   Clara Counties, while the order is reversed for Riverside County.

•  Six of the highest emitted pollutants also have the highest toxi city-weighted
   emissions for Los Angeles and Riverside Counties, while there are seven in common
   for Santa Clara County. While hexavalent chromium is at or near the top in toxicity-
   weighted emissions for all three counties, this pollutant is not among the 10 highest
   emitted pollutants in these counties. Hexavalent chromium emissions rank between
   13th highest for RUCA to 18th highest for SJJCA.

•  Naphthalene has the highest cancer risk approximation for all three sites. Naphthalene
   is the only pollutant to appear on all three lists for all three counties.

•  Arsenic and nickel, which have the second and third highest cancer risk
   approximations for SJJCA, respectively, have the seventh and tenth highest toxicity-
   weighted emissions for Santa Clara County, but are not among the 10 highest emitted
   pollutants for this county. These are the only pollutants sampled by SJJCA, other than
   naphthalene, to appear on either emissions-based list.


Observations from Table 6-8 include the following:

•  Toluene,  1,1,1-trichloroethane, formaldehyde, and benzene are the highest emitted
   pollutants with noncancer RfCs in all three California counties (although not
   necessarily in that order). Consistent with pollutants having cancer UREs, emissions
   are greater in Los Angeles County than Riverside and Santa Clara Counties.

•  Acrolein, chlorine, and formaldehyde are the pollutants with the highest toxicity-
   weighted emissions (of the pollutants with noncancer RfCs) for all three counties.
   While acrolein and chlorine rank highest for toxicity-weighted emissions for each
   county, neither pollutant appears among the highest emitted. Conversely,
   formaldehyde has the third highest emissions for each county.

•  Three of the highest emitted pollutants also have the highest toxicity-weighted
   emissions for Los Angeles and Riverside Counties, while four of the highest  emitted
   pollutants also have the highest toxicity-weighted emissions for Santa Clara County.

•  Naphthalene, the only pollutant for which a noncancer hazard approximation could be
   calculated for CELA and RUCA, has one of the 10 highest toxicity-weighted
   emissions for these sites' respective counties, but does not appear on the list of the 10
   highest total emissions for either county (of pollutants with noncancer RfCs). This is
   also true for Santa Clara County.


                                   6-39

-------
       •  Besides naphthalene, arsenic and lead are the only two pollutants for which
          noncancer hazard approximations could be calculated for SJJCA and that also appear
          on the list of 10 highest toxi city-weighted emissions. Manganese, which has the
          highest noncancer hazard approximation for SJJCA, appears on neither emissions-
          based list. None of the metals appear among the pollutants with the highest total
          emissions.


6.6    Summary of the 2011 Monitoring Data for CELA, RUCA, and SJJCA

       Results from several of the data treatments described in this section include the
following:

       ปซป  Three PAHs, including naphthalene, failed screens for CELA, while only naphthalene
          failed screens for RUCA. Three PAHs and three PMw metals failed screens for
          SJJCA.

       ปซป  Naphthalene had the highest annual average concentration among all the pollutants
          of interest for the California sites. Among the metals sampled at SJJCA, manganese
          had the highest annual average concentration.

       ปซป  Concentrations of naphthalene and benzo(a)pyrene exhibit a seasonal trend at
          SJJCA, with higher concentrations during the first and fourth quarters (or the colder
          months) of 2011.
                                          6-40

-------
7.0    Sites in Colorado
       This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS and UATMP sites in Colorado, and integrates these
concentrations with emissions, meteorological, and risk information. Data generated by sources
other than ERG are not included in the data analyses contained in this report. Readers are
encouraged to refer to Sections 1 through 4 and the glossary (Appendix P) for detailed
discussions and definitions regarding the various data analyses presented below.

7.1    Site Characterization
       This section characterizes the Colorado monitoring sites by providing geographical and
physical information  about the location of the sites and the surrounding areas. This information
is provided to give the reader insight regarding factors that may influence the air quality near the
sites and assist in the interpretation of the ambient monitoring measurements.

       The NATTS site is located in Grand Junction (GPCO) and the other four sites are located
in Garfield County, between 35 and 55 miles northeast of Grand Junction, in the towns of
Battlement Mesa (BMCO), Silt (BRCO), Parachute (PACO), and Rifle (RICO). Figure 7-1 for
GPCO and Figures 7-3 through 7-6 for the Garfield County sites are composite satellite images
retrieved from ArcGIS Explorer showing the monitoring sites in their urban and rural locations.
Figures 7-2 and 7-7 identify nearby point source  emissions locations by source category, as
reported in the 2008 NEI for point sources.  Note  that only sources within 10 miles of each site
are included in the facility counts provided  in Figures 7-2 and 7-7. A 10-mile boundary  was
chosen to give the reader an indication of which emissions sources and emissions source
categories could potentially have a direct effect on the air quality at the monitoring sites. Further,
this boundary provides both the proximity of emissions sources to the monitoring sites as well as
the quantity of such sources within a given  distance of the sites. Sources outside the 10-mile radii
are still visible on the maps, but have been grayed out in order to show emissions sources just
outside the boundary. Table 7-1 provides supplemental geographical information such as land
use, location setting,  and locational coordinates.
                                           7-1

-------
                                  Figure 7-1. Grand Junction, Colorado (GPCO) Monitoring Site
to

-------
         Figure 7-2. NEI Point Sources Located Within 10 Miles of GPCO
          0      2.5       5              10
          I	1	1	1	1	1	1	1	1
                        MM
Legend
       GPCO NATTS site
10 mile radius
                                                 Note: Due to facility density and collocation, the total facilities
                                                 displayed may not represent all facilities within the area of interest.
County boundary
Source Category Group (No. of Facilities)
 41  Aircraft Operations (9)
 0  Auto Body Shop/Painters (4)
 SS  Automobile/Truck Manufacturing (2)
    Brick Manufacturing & Structural Clay (1)
 B  Bulk Terminals/Bulk Plants (4)
 C  Chemical Manufacturing (1)
 •  Concrete Batch Plant (6)
X Crematory-Animal/Human (4)
 (1)  Dry Cleaning (4)
 e  Electrical Equipment (2)
 E  Electroplating. Plating. Polishing, Anodizing, and Coloring (2)
 ฉ  Fabricated Metal Products (1)
 E!  Furniture Plant (1)
                              Gasoline/Diesel Service Station (47)
                              Hot Mix Asphalt Plant (2)
                              Industrial Machinery and Equipment (1)
                              Institutional - school (4}
                              Landfill (1)
                              Mine/Quarry (23)
                              Miscellaneous Commercial/Industrial (3)
                              Oil and/or Gas Production (4)
                              Rubber and Miscellaneous Plastics Products (3)
                              Secondary Metal Processing (1)
                              Site Remediation Activity (2)
                              Surface Coating (2)
                              Textile Mill (1)
                              Wastewater Treatment (1)
                                                 7-3

-------
Figure 7-3. Battlement Mesa, Colorado (BMCO) Monitoring Site

-------
Figure 7-4. Silt, Colorado (BRCO) Monitoring Site

-------
Figure 7-5. Parachute, Colorado (PACO) Monitoring Site


-------
Figure 7-6. Rifle, Colorado (RICO) Monitoring Site

-------
Figure 7-7. NEI Point Sources Located Within 10 Miles of BMCO, BRCO, PACO, and
                                                 RICO
                   i08-i5T>-w   1Q8-10WV   ios'5T)-w
                tOB'IO'O'W   K>B 5'0"W
                                        107'55'G-W  107-50'CTW  t07r'45'0'W  107^0BlrV   1Q7"35'G'W   107'30'0'W   107'25'0"W
                                                     Note Due to facility density and collocation, Ihe total facilities
                                                     displayed may not represent all facilities within the area of interest.
                  Legend
                   ฉ   BMCO UATMP site   jฃ   PACO UATMP site
                   ฎ   BRCO UATMP site   5   RICO UATMP site
               10 mile radius
             J County boundary
        Source Category Group (NO. Of Facilities)   *
          -f   Aircraft Operations (7)
          *   Building Construction (1)
          B   Bulk Terminals/Bulk Plants (1)
          •   Concrete Batch Plant (1)
          ffi   Dry Cleaning (1)
          *   Electricity Generation via Combustion (1)
'A
•
Gasoline/Diesel Service Station (17)
Hot Mix Asphalt Plant (1)
Landfill (1)
Mine/Quarry (12)
Miscellaneous Commercial/Industrial (2)
Oil and/or Gas Production (920)
Pipeline Compressor Station (8)
                                                   7-8

-------
                                Table 7-1. Geographical Information for the Colorado Monitoring Sites
Site
Code
GPCO
BMCO
BRCO
PACO
RICO
AQS Code
08-077-0017
08-077-0018
None
08-045-0009
08-045-0005
08-045-0007
Location
Grand
Junction
Battlement
Mesa
Silt
Parachute
Rifle
County
Mesa
Garfield
Garfield
Garfield
Garfield
Micro- or
Metropolitan
Statistical Area
Grand Junction,
CO MSA
Not in an MSA
Not in an MSA
Not in an MSA
Not in an MSA
Latitude
and
Longitude
39.064289,
-108.56155
39.4399898,
-108.029769
39.487755,
-107.659685
39.453654,
-108.053259
39.531813,
-107.782298
Land Use
Commercial
Residential
Agricultural
Residential
Commercial
Location
Setting
Urban/City
Center
Rural
Rural
Urban/City
Center
Urban/City
Center
Additional Ambient Monitoring Information1
Meteorological parameters, CO, PM10, PM10
Speciation, PM2 5, and PM2 5 Speciation.
No AQS entry.
None.
PMio, PM10 Speciation.
PMio, PM10 Speciation.
:Data for additional pollutants are reported to AQS for these sites (EPA,
BOLD ITALICS = EPA-designaled NATTS Site
2012c); however, these data are not generated by ERG and are therefore not included in this report.

-------
       The GPCO monitoring site is comprised of two locations. The first location is a small
1-story shelter that houses the VOC and carbonyl compound samplers, with the PAH sampler
located just outside the shelter. The second location, which is on the roof of an adjacent 2-story
building, is comprised of the hexavalent chromium samplers. As a result, two AQS codes are
provided in Table 7-1. Figure 7-1 shows that the area surrounding GPCO is of mixed usage, with
commercial businesses to the west, northwest, and north; residential areas to the northeast and
east; and industrial areas to the southeast, south, and southwest. This site's location is next to one
of the major east-west roads in Grand Junction (1-70 Business). A railroad runs east-west to the
south of the GPCO monitoring site, and merges with another railroad to the southwest of the site.
As Figure 7-2 shows, GPCO is located within 10 miles of numerous emissions sources. Many of
the sources are located along a diagonal line running roughly northwest to southeast along
Highways 6 and 50 and Business-70. Many of the point sources near GPCO fall into the
gasoline/diesel service station and mine/quarry source categories. The sources closest to GPCO
are a bulk terminal/bulk plant, an automobile/truck manufacturer, and a gasoline/diesel service
station.

       The BMCO monitoring site is located in Battlement Mesa, a rural community located to
the southeast of Parachute. The monitoring site is located on the roof of the Grand Valley Fire
Protection District facility, near the intersection of Stone Quarry Road and West Battlement
Parkway, as shown in Figure 7-3. The site is surrounded primarily by residential subdivisions. A
cemetery is located to the south of the site and a church is located to the east.

       The BRCO monitoring site is located on Bell/Melton Ranch, off Owens Drive,
approximately 4 miles south of the town of Silt. The site is  both rural and agricultural in nature.
As shown in Figure 7-4, the closest major roadway is County Road 331, Dry Hollow Road.

       PACO is located on the roof of the old Parachute High School building, which is
presently operating as a day care facility. This location is in the center of the town of Parachute,
as shown in Figure 7-5. The surrounding area is considered residential. Interstate-70 is less  than
a quarter of a mile from the monitoring site.
                                          7-10

-------
       RICO is located on the roof of the Henry Annex Building in downtown Rifle. This
location is near the crossroads of several major roadways through town, as shown in Figure 7-6.
Highway 13 and US-6 intersect just south of the site and 1-70 is just over a half-mile south of the
monitoring site across the Colorado River. The surrounding area is considered commercial.

       The four  Garfield County sites are located along a line running roughly east-west and
spanning approximately 20 miles; hence, they are shown together in Figure 7-7. There are more
than 900 petroleum or natural gas wells (collectively shown as the oil and/or gas production
source category) within 10 miles of these sites. One reason Garfield County is conducting air
monitoring is to characterize the effects these wells may have on the surrounding areas (GCPH,
2010).

       Table 7-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Colorado monitoring sites. Table 7-2 includes county-level
population and vehicle registration information.  Table 7-2 also includes a county-level vehicle
registration-to-population ratio, which was calculated to represent the number of vehicles per
person within each monitoring site's residing county. In addition, the population within 10 miles
of each site is presented, based on postal code population data estimates. An estimate of 10-mile
vehicle ownership was then determined by applying the county-level vehicle registration-to-
population ratio to the 10-mile population surrounding each monitoring site. Table 7-2 also
contains traffic volume information for each site. Finally, Table 7-2 presents the county-level
daily VMT for Mesa and Garfield Counties. Note that the VMT presented is for state highways
only, which differs from the VMT presented in this table in other state sections.

       Observations from  Table 7-2 include the following:
       •  Mesa County's population and vehicle ownership are considerably higher than those
          for Garfield County. This is also true for 10-mile population and vehicle ownership.
          However, both  counties  rank in the bottom-third compared to other counties with
          NMP sites.
       •  The vehicle-per-person ratios for the Colorado sites are among the highest for all
          NMP sites, with the Garfield County  sites'  ranking second behind only UCSD, and
          GPCO ranking  fourth.
       •  The traffic volumes near PACO, RICO, and GPCO are considerably higher than the
          traffic volumes near the  other Colorado monitoring sites. The lowest traffic volume
                                           7-11

-------
           among all NMP sites is for BRCO. The traffic estimate provided for GPCO is based
           on Business-70 near 7th Street; from South Battlement Parkway for BMCO; from the
           junction of County Roads 331 and 326 for BRCO; from 1-70 near exit 75 for PACO;
           and from Highway 13 between Highway 6 and 1-70 for RICO.
           While the Mesa and Garfield County VMTs are fairly similar to each other, they are
           also among the lowest for counties with NMP sites, where VMT data were available.
           However, the county-level VMT available from the Colorado DOT is for state
           highways only.
      Table 7-2. Population, Motor Vehicle, and Traffic Information for the Colorado
                                     Monitoring Sites
Site
GPCO
BMCO
BRCO
PACO
RICO
Estimated
County
Population1
147,083
56,270
County-level
Vehicle
Registration2
178,425
72,957
Vehicles per
Person
(Registration:
Population)
1.21
1.30
Population
within 10
miles3
120,030
5,941
25,466
5,941
18,148
Estimated
10-mile
Vehicle
Ownership
145,607
7,703
33,018
7,703
23,530
Annual
Average
Daily
Traffic4
11,000
2,527
150
16,000
17,000
County-level
Daily VMT5
2,031,327
1,901,434
Bounty-level population estimates reflect 2011 data from the U.S. Census Bureau (Census Bureau, 2012b)
2County-level vehicle registration reflects 2010 data from the Colorado DOR (CO DOR, 2011)
310-mile population estimates reflect 2011 data from Xionetic (Xionetic, 2011)
4AADT reflects 2002 data for BMCO and BRCO from Garfield County (GCRBD, 2002) and 2011 data for GPCO,
 PACO, and RICO from the Colorado DOT (CO DOT, 201 la)
5County-level VMT reflects 2011 data for state highways only from the Colorado DOT (CO DOT, 201 Ib)
BOLD ITALICS = EPA-designaled NATTS Site
 7.2    Meteorological Characterization
       The following sections characterize the meteorological conditions near the monitoring
 sites in Colorado on sample days, as well as over the course of the year.

 7.2.1  Climate Summary
       Grand Junction is located in a mountain valley on the west side of the Rockies. The
 valley location of the city helps protect it from dramatic weather changes. The area tends to be
 fairly dry and winds tend to flow out of the east-southeast on average, due to the valley breeze
 effect (Bair, 1992). 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
                                           7-12

-------
a river valley running north of the Grand Mesa. Similar to Grand Junction, these towns are
shielded from drastic changes in weather by the surrounding terrain and tend to experience fairly
dry conditions for most of the year. Wind patterns in these towns are affected by the high
canyons, the Colorado River, and valley breezes (GCPH, 2010 and WRCC, 2013).

7.2.2   Meteorological Conditions in 2011
       Hourly meteorological data from the NWS weather stations nearest these sites were
retrieved for 2011 (NCDC, 2011). The weather station nearest GPCO is located at Walker Field
Airport (WBAN 23066); the closest weather station to the four Garfield County sites is located at
Garfield County Regional Airport (WBAN 03016).  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.

       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 the entire year for 2011. 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 GPCO were representative of average weather
conditions throughout the year.  Temperatures on sample days near the Garfield County sites
appear slightly cooler than over the course of the year. This may be attributable to a few missed
sample days in May and  September for each site.
                                         7-13

-------
                 Table 7-3. Average Meteorological Conditions near the Colorado Monitoring Sites
Closest NWS
Station (WBAN
and Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(ฐF)
Average
Temperature
<ฐF)
Average
Dew Point
Temperature
<ฐF)
Average
Wet Bulb
Temperature
(ฐF)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Grand Junction, Colorado - GPCO
Walker Field
Airport
23066
(39.13, -108.54)
4.96
miles
22ฐ
(NNE)
Sample
Day
2011
63.8
ฑ5.2
64.5
ฑ2.2
51.5
ฑ4.8
52.3
ฑ2.0
28.2
ฑ3.5
28.5
ฑ1.4
40.5
ฑ3.4
40.9
ฑ1.4
48.9
ฑ4.6
47.7
ฑ1.9
1014.7
ฑ2.2
1014.9
ฑ0.9
6.3
ฑ0.7
6.2
ฑ0.3
Battlement Mesa, Colorado - BMCO
Garfield Co.
Regional Airport
03016
(39.53, -107.73)
16.41
miles
76ฐ
(ENE)
Sample
Day
2011
59.4
ฑ5.6
61.4
ฑ2.2
45.5
ฑ5.1
47.2
ฑ1.9
27.3
ฑ4.2
28.1
ฑ1.4
37.2
ฑ4.0
38.2
ฑ1.5
56.6
ฑ4.3
55.2
ฑ1.7
1016.4
ฑ2.6
1016.7
ฑ0.9
5.1
ฑ0.8
4.5
ฑ0.3
Silt, Colorado - BRCO
Garfield Co.
Regional Airport
03016
(39.53, -107.73)
4.23
miles
316ฐ
(NW)
Sample
Day
2011
59.5
ฑ5.8
61.4
ฑ2.2
45.3
ฑ5.3
47.2
ฑ1.9
26.6
ฑ4.2
28.1
ฑ1.4
36.7
ฑ4.2
38.2
ฑ1.5
55.9
ฑ4.3
55.2
ฑ 1.7
1016.2
ฑ2.7
1016.7
ฑ0.9
5.1
ฑ0.8
4.5
ฑ0.3
Parachute, Colorado - PACO
Garfield Co.
Regional Airport
03016
(39.53, -107.73)
17.22
miles
81ฐ
(E)
Sample
Day
2011
59.4
ฑ5.6
61.4
ฑ2.2
45.5
ฑ5.1
47.2
ฑ1.9
27.3
ฑ4.2
28.1
ฑ1.4
37.1
ฑ4.0
38.2
ฑ 1.5
56.8
ฑ4.3
55.2
ฑ1.7
1016.5
ฑ2.6
1016.7
ฑ0.9
5.0
ฑ0.8
4.5
ฑ0.3
Sample day averages are highlighted to help differentiate the sample day averages from the full-year averages.

-------
          Table 7-3. Average Meteorological Conditions near the Colorado Monitoring Sites (Continued)
Closest NWS
Station (WBAN
and Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(ฐF)
Average
Temperature
<ฐF)
Average
Dew Point
Temperature
<ฐF)
Average
Wet Bulb
Temperature
(ฐF)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Rifle, Colorado - RICO
Garfield Co.
Regional Airport
03016
(39.53, -107.73)
2.89
miles
105ฐ
(ESE)
Sample
Day
2011
59.8
ฑ5.7
61.4
ฑ2.2
45.7
ฑ5.2
47.2
ฑ1.9
26.9
ฑ4.0
28.1
ฑ1.4
37.1
ฑ4.0
38.2
ฑ1.5
55.7
ฑ4.3
55.2
ฑ1.7
1016.3
ฑ2.6
1016.7
ฑ0.9
5.2
ฑ0.8
4.5
ฑ0.3
Sample day averages are highlighted to help differentiate the sample day averages from the full-year averages.

-------
7.2.3   Back Trajectory Analysis

       Figure 7-8 is the composite back trajectory map for days on which samples were

collected at the GPCO monitoring site in 2011. Included in Figure 7-8 are four back trajectories

per sample day. Figure 7-9 is the corresponding cluster analysis. Similarly, Figures 7-10 through

7-17 are the composite back trajectory maps and corresponding cluster analyses for the Garfield

County monitoring sites. An in-depth description of these maps and how they were generated is
presented in Section 3.5.2.1. For the composite maps, each line represents the 24-hour trajectory

along which a parcel of air traveled toward the monitoring site on a given sample day and time,

based on an  initial height of 50 meters AGL. For the cluster analyses, each line corresponds to a

trajectory representative of a given cluster of back trajectories. Each concentric circle around the

sites in Figures 7-8 through 7-17 represents 100 miles.


       Observations for GPCO from Figures 7-8 and 7-9 include the following:

       •  The 24-hour air shed domain for GPCO is one of the smallest in size, based on
          average back trajectory length, compared to other NMP monitoring sites. The farthest
          away a back trajectory originated was over southwest Arizona, or just greater than
          450 miles away. However, most trajectories (88 percent) originated within 300 miles
          of GPCO and the average trajectory length was approximately 169 miles.

       •  Back trajectories originated from a variety of directions at GPCO, although a large
          cluster of trajectories originated to the southwest of GPCO and the majority of the
          trajectories had a westerly component.

       •  The cluster analysis shows that back trajectories frequently originated from the
          southwest, accounting for roughly 37 percent of back trajectories. Another 27 percent
          of trajectories originated from the northwest and west of GPCO. Shorter back
          trajectories originating over southeast Utah and southwest Colorado are represented
          by the short cluster labeled 24 percent and originating to the south of GPCO. The
          cluster originating  to the east of GPCO represents back trajectories  originating from
          the northeast, east, and southeast of the site (12 percent). Thus, air moving towards
          GPCO is generally originating in Colorado, Utah, and Arizona.
                                          7-16

-------
Figure 7-8. 2011 Composite Back Trajectory Map for GPCO
    Figure 7-9. Back Trajectory Cluster Map for GPCO
                         7-17

-------
Figure 7-10. 2011 Composite Back Trajectory Map for BMCO
    Figure 7-11. Back Trajectory Cluster Map for BMCO
                          7-18

-------
Figure 7-12. 2011 Composite Back Trajectory Map for BRCO
    Figure 7-13. Back Trajectory Cluster Map for BRCO
                          7-19

-------
Figure 7-14. 2011 Composite Back Trajectory Map for PACO
    Figure 7-15. Back Trajectory Cluster Map for PACO
                          7-20

-------
Figure 7-16. 2011 Composite Back Trajectory Map for RICO
    Figure 7-17. Back Trajectory Cluster Map for RICO
                          7-21

-------
       Observations from Figures 7-10 through 7-17 for the Garfield County sites include the
following:
       •  The composite back trajectory maps for the Garfield County sites resemble the one
          for GPCO. This is expected, given the sites' close proximity to GPCO (and to each
          other).
       •  The 24-hour air shed domains for the Garfield County sites were similar in size to
          GPCO, with the average trajectory length ranging from 171 miles for PACO to
          181 miles for BRCO. The longest trajectories for these sites originated over central
          Arizona. For each Garfield County site, approximately 85 percent of back trajectories
          originated within 300 miles of the site.
       •  The cluster maps for the Garfield County sites resemble the cluster map for GPCO,
          with most of the trajectories having a southwesterly or northwesterly component.

7.2.4   Wind Rose Comparison
       Hourly surface wind data from the NWS weather stations at the Walker  Field Airport (for
GPCO) and Garfield County Regional Airport (for BMCO, BRCO, PACO, and RICO) were
uploaded into a wind rose software program to produce customized wind roses,  as described in
Section 3.5.2.2. A wind rose shows the frequency of wind directions using "petals" positioned
around a 16-point compass, and uses different colors to represent wind speeds.

       Figure 7-18 presents a map showing the distance between the NWS station and GPCO,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 7-18 also presents three  different wind roses for the
GPCO monitoring site. First, a historical wind rose representing 2001 to 2010 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 2011 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2011 is
presented. These can be used to identify the predominant wind speed and direction for 2011 and
determine if wind observations  on sample days were representative of conditions experienced
over the entire year and historically. Figures 7-19 through 7-22 present the distance maps and
wind roses for the four Garfield County sites.
                                          7-22

-------
 Figure 7-18. Wind Roses for the Walker Field Airport Weather Station near GPCO
Distance between GPCO and NWS Station
2001-2010 Historical Wind Rose
          2011 Wind Rose
                                                   Sample Day Wind Rose
                             Calms: 17.39%
                                     7-23

-------
   Figure 7-19. Wind Roses for the Garfield County Regional Airport near BMCO
Distance between BMCO and NWS Station
2001-2010 Historical Wind Rose

                                            WEST;
          2011 Wind Rose
    Sample Day Wind Rose
                                     7-24

-------
    Figure 7-20. Wind Roses for the Garfield County Regional Airport near BRCO
 Distance between BRCO and NWS Station
              2001-2010 Historical Wind Rose

r
 N




+
                                             WEST;
          2011 Wind Rose
                                                   Sample Day Wind Rose
                                     7-25

-------
   Figure 7-21. Wind Roses for the Garfield County Regional Airport near PACO
Distance between PACO and NWS Station
2001-2010 Historical Wind Rose
                                           WEST
         2011 Wind Rose
                                                  Sample Day Wind Rose
                                    7-26

-------
   Figure 7-22. Wind Roses for the Garfield County Regional Airport near RICO
Distance between RICO and NWS Station
2001-2010 Historical Wind Rose
                                  N
                                 +
                                             WEST;
         2011 Wind Rose
                                                   Sample Day Wind Rose
                             Calms: 36.76%
                                     7-27

-------
       Observations from Figure 7-18 for GPCO include the following:

       •  The Walker Field Airport weather station is located approximately 5 miles north-
          northeast of GPCO. Most of the city of Grand Junction lies between the site and the
          weather station.

       •  The historical wind rose shows that easterly, east-southeasterly, and southeasterly
          winds were prevalent near GPCO. Calm winds (< 2 knots) were observed for less
          than 15 percent of the hourly wind  measurements.

       •  The 2011 wind rose exhibits similar wind patterns as the historical wind rose. Further,
          the sample day wind patterns also resemble the historical and full-year wind patterns,
          indicating that wind conditions on sample days were representative of those
          experienced over the entire year and historically.


       Observations from Figures 7-19 through 7-22 for the Garfield County sites include the
following:

       •  The NWS weather station at Garfield County Regional Airport is the closest weather
          station to all four monitoring sites in Garfield County. The weather station is located
          just east of Rifle. The distances from the weather station to the sites varies from less
          than 3 miles (RICO) to just over 17 miles (PACO).

       •  The historical and 2011 wind roses for the Garfield County sites are identical to each
          other. This is because the wind observations come from the same NWS weather
          station for all four sites.

       •  The historical wind roses show that calm winds were prevalent near the monitoring
          sites, representing one-third of wind observations. Westerly and southerly winds were
          also common.

       •  Calm winds were observed for greater than one third of the wind observations in
          2011. The 2011 wind roses exhibit  a higher percentage of westerly winds and a
          significantly lower percentage of southerly and south-southwesterly winds than the
          historical wind rose.

       •  The sample day wind patterns for each site resemble the full-year wind patterns, but
          with a higher percentage of northwesterly winds. This resemblance indicates that
          conditions on sample days were representative of those experienced over the entire
          year.
                                          7-28

-------
7.3    Pollutants of Interest
       Site-specific "pollutants of interest" were determined for the Colorado monitoring sites in
order to allow 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. In
addition, if any of the NATTS MQO Core Analytes measured by each monitoring site did not
meet the pollutant of interest criteria based on the preliminary risk-based screening, that pollutant
was added to the list of site-specific pollutants of interest. A more in-depth description of the
risk-based screening process is presented in Section 3.2.

       Table 7-4 presents the results of the preliminary risk-based screening process for each
Colorado monitoring site. The pollutants that failed at least one screen and contributed to
95 percent of the total failed screens for each monitoring site are shaded. NATTS MQO Core
Analytes are bolded. Thus, pollutants of interest are shaded and/or bolded. GPCO  sampled for
VOCs, carbonyl  compounds, PAHs, and hexavalent chromium; the Garfield County sites
sampled for SNMOCs and carbonyl compounds only.
        Table 7-4. Risk-Based Screening Results for the Colorado Monitoring Sites
Pollutant
Screening
Value
(Ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
Grand Junction, Colorado - GPCO
Naphthalene
Acet aldehyde
Benzene
Formaldehyde
Carbon Tetrachloride
1,3-Butadiene
Ethylbenzene
Acenaphthene
Acrylonitrile
1 ,2-Dichloroethane
Fluorene
ฃ>-Dichlorobenzene
Hexachloro- 1 , 3 -butadiene
Benzo(a)pyrene
0.029
0.45
0.13
0.077
0.17
0.03
0.4
0.011
0.015
0.038
0.011
0.091
0.045
0.00057
61
60
60
60
58
53
36
21
17
16
11
9
8
7
61
60
60
60
60
53
60
61
17
16
61
32
8
44
100.00
100.00
100.00
100.00
96.67
100.00
60.00
34.43
100.00
100.00
18.03
28.13
100.00
15.91
12.25
12.05
12.05
12.05
11.65
10.64
7.23
4.22
3.41
3.21
2.21
1.81
1.61
1.41
12.25
24.30
36.35
48.39
60.04
70.68
77.91
82.13
85.54
88.76
90.96
92.77
94.38
95.78
                                           7-29

-------
Table 7-4. Risk-Based Screening Results for the Colorado Monitoring Sites (Continued)
Pollutant
1 , 1 ,2,2-Tetrachloroethane
Trichloroethylene
1 ,2-Dibromoethane
Dichloromethane
Acenaphthylene
Hexavalent Chromium
Propionaldehyde
1 , 1 ,2-Trichloroethane
Xylenes
Screening
Value
(Ug/m3)
0.017
0.2
0.0017
7.7
0.011
0.000083
0.8
0.0625
10
Total
#of
Failed
Screens
7
4
3
2
1
1
1
1
1
498
#of
Measured
Detections
7
22
3
60
41
42
60
1
60
949
%of
Screens
Failed
100.00
18.18
100.00
3.33
2.44
2.38
1.67
100.00
1.67
52.48
% of Total
Failures
1.41
0.80
0.60
0.40
0.20
0.20
0.20
0.20
0.20
Cumulative
%
Contribution
97.19
97.99
98.59
99.00
99.20
99.40
99.60
99.80
100.00

Battlement Mesa, Colorado - BMCO
Benzene
Formaldehyde
Acet aldehyde
Ethylbenzene
1,3-Butadiene
0.13
0.077
0.45
0.4
0.03
Total
51
20
14
9
8
102
51
20
20
51
8
150
100.00
100.00
70.00
17.65
100.00
68.00
50.00
19.61
13.73
8.82
7.84
50.00
69.61
83.33
92.16
100.00

Silt, Colorado - BRCO
Benzene
1,3-Butadiene
Formaldehyde
Acet aldehyde
0.13
0.03
0.077
0.45
Total
53
6
6
5
70
53
7
6
6
72
100.00
85.71
100.00
83.33
97.22
75.71
8.57
8.57
7.14
75.71
84.29
92.86
100.00

Parachute, Colorado - PACO
Benzene
1,3-Butadiene
Formaldehyde
Acet aldehyde
Ethylbenzene
0.13
0.03
0.077
0.45
0.4
Total
52
21
21
19
12
125
52
22
21
21
53
169
100.00
95.45
100.00
90.48
22.64
73.96
41.60
16.80
16.80
15.20
9.60
41.60
58.40
75.20
90.40
100.00

Rifle, Colorado - RICO
Benzene
1,3-Butadiene
Acet aldehyde
Formaldehyde
Ethylbenzene
0.13
0.03
0.45
0.077
0.4
Total
51
42
17
17
15
142
51
42
17
17
53
180
100.00
100.00
100.00
100.00
28.30
78.89
35.92
29.58
11.97
11.97
10.56
35.92
65.49
77.46
89.44
100.00

                                      7-30

-------
       Observations from Table 7-4 include the following:

       •  Twenty-three pollutants failed at least one screen for GPCO, of which eight are
          NATTS MQO Core Analytes.

       •  Fourteen pollutants were initially identified as pollutants of interest for GPCO based
          on the risk-based screening process, of which seven are NATTS MQO Core Analytes.
          Trichloroethylene and hexavalent chromium were added to GPCO's pollutants of
          interest because they are NATTS MQO Core Analytes, even though they did not
          contribute to 95 percent of GPCO's total failed screens. Three additional NATTS
          MQO Core Analytes were also added to GPCO's pollutants of interest even though
          their concentrations did not fail any screens: chloroform, tetrachloroethylene, and
          vinyl chloride. These three pollutants are not shown in Table 7-4 but are shown in
          subsequent tables in the sections that follow.

       •  The number of pollutants failing screens for the Garfield County sites range from four
          to five. Four pollutants (benzene, 1,3-butadiene,  formaldehyde, and acetaldehyde)
          failed screens for each Garfield County site. These four pollutants were identified as
          pollutants of interest for all four sites. Ethylbenzene also failed screens for three of
          the four Garfield County sites (BRCO being the  exception), and was identified as a
          pollutant of interest for BMCO, PACO, and RICO.

       •  Note  that carbonyl compound samples were collected on a l-in-12 day sampling
          schedule at the Garfield County sites, while SNMOCs were collected on a l-in-6 day
          sampling schedule; thus, the number of carbonyl compounds samples collected at
          these sites were often less than half the number of SNMOC samples.

       •  Benzene and formaldehyde failed 100 percent of screens for all five Colorado  sites.


7.4    Concentrations
       This section presents various concentration averages used to characterize pollution levels

at the Colorado monitoring sites. Where applicable, the following calculations and data analyses

were performed: Time period-based concentration averages (quarterly and annual) are provided

for the pollutants of interest for the Colorado monitoring sites, where the data meet the

applicable criteria. Concentration averages for select pollutants are also presented graphically  for

the sites to illustrate how the sites' concentrations compare to the program-level averages, as

presented in Section 4.1. In addition, concentration averages for select pollutants are presented

from previous  years of sampling in order to characterize concentration trends at the sites.

Additional site-specific statistical summaries for the five Colorado sites are provided in

Appendices J through M and Appendix O.
                                          7-31

-------
7.4.1   2011 Concentration Averages
       Quarterly and annual concentration averages were calculated for the pollutants of interest
for each Colorado 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 of the total number of
samples possible within  a given quarter for a quarterly average to be calculated. An annual
average includes all measured detections and substituted zeros for non-detects for the entire year
of sampling. Annual averages were calculated for pollutants where three valid quarterly averages
could be calculated  and where method completeness was greater than  or equal to 85 percent, as
presented in Section 2.4. Quarterly and annual average concentrations for the Colorado
monitoring sites are presented in Table 7-5, where applicable. Note that concentrations of the
PAHs and hexavalent chromium for GPCO are presented in ng/m3  for ease of viewing. Also note
that if a pollutant was not detected in a given calendar quarter, the quarterly average simply
reflects "0" because only zeros substituted for non-detects were factored into the quarterly
average concentration.
                                           7-32

-------
     Table 7-5. Quarterly and Annual Average Concentrations of the Pollutants of
                   Interest for the Colorado Monitoring Sites
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Ug/m3)
2nd
Quarter
Average
(Ug/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Ug/m3)
Grand Junction, Colorado - GPCO
Acetaldehyde
Acrylonitrile
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
ฃ>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
Hexachloro- 1 ,3 -butadiene
1 , 1 ,2,2-Tetrachloroethane
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
Acenaphthene3
Benzo(a)pyrenea
Fluorene3
Hexavalent Chromium3
Naphthalene3
60/60
17/60
60/60
53/60
60/60
42/60
32/60
16/60
60/60
60/60
8/60
7/60
53/60
22/60
6/60
61/61
44/61
61/61
42/59
61/61
1.81
ฑ0.34
0.01
ฑ0.03
1.32
ฑ0.29
0.15
ฑ0.04
0.44
ฑ0.08
0.07
ฑ0.02
0.02
ฑ0.02
0.01
ฑ0.02
0.46
ฑ0.13
2.84
ฑ0.37
0.02
ฑ0.02
0.01
ฑ0.01
0.26
ฑ0.07
0.04
ฑ0.02
<0.01
ฑ<0.01
15.27
ฑ6.69
0.23
ฑ0.11
7.65
ฑ2.43
0.01
ฑ0.01
240.29
ฑ 96.22
1.47
ฑ0.24
0.15
ฑ0.11
1.49
ฑ1.35
0.07
ฑ0.03
0.50
ฑ0.08
0.08
ฑ0.03
0.01
ฑ0.02
0.04
ฑ0.03
0.34
ฑ0.06
2.53
ฑ0.27
0.01
ฑ0.02
0.01
ฑ0.01
0.18
ฑ0.08
0.02
ฑ0.03
<0.01
ฑ<0.01
9.75
ฑ4.24
0.03
ฑ0.02
7.65
ฑ2.26
0.01
ฑ<0.01
106.40
ฑ 36.96
3.12
ฑ0.45
0.06
ฑ0.04
0.90
ฑ0.14
0.09
ฑ0.02
0.62
ฑ0.08
0.07
ฑ0.04
0.06
ฑ0.02
0.02
ฑ0.02
0.69
ฑ0.18
2.84
ฑ0.37
0.02
ฑ0.02
0.02
ฑ0.02
0.15
ฑ0.05
0.04
ฑ0.04
<0.01
ฑ<0.01
12.07
ฑ2.27
0.06
ฑ0.07
9.39
ฑ1.47
0.01
ฑ<0.01
121.08
ฑ 19.58
3.36
ฑ0.62
0.01
ฑ0.02
1.69
ฑ0.40
0.23
ฑ0.06
0.57
ฑ0.06
0.11
ฑ0.03
0.07
ฑ0.02
0.04
ฑ0.03
1.00
ฑ0.33
2.74
ฑ0.41
0.01
ฑ0.01
O.01
ฑ0.01
0.45
ฑ0.15
0.07
ฑ0.06
O.01
ฑO.01
4.95
ฑ1.72
0.49
ฑ0.26
5.90
ฑ1.66
0.03
ฑ0.02
156.63
ฑ 47.25
2.43
ฑ0.29
0.06
ฑ0.03
1.34
ฑ0.35
0.13
ฑ0.02
0.53
ฑ0.04
0.08
ฑ0.02
0.04
ฑ0.01
0.03
ฑ0.01
0.62
ฑ0.11
2.74
ฑ0.17
0.01
ฑ0.01
0.01
ฑ0.01
0.26
ฑ0.05
0.04
ฑ0.02
O.01
ฑO.01
10.54
ฑ2.17
0.20
ฑ0.08
7.68
ฑ0.98
0.02
ฑ0.01
155.52
ฑ29.71
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
a Average concentrations provided for the pollutants below the blue line for GPCO are presented in ng/m3
 for ease of viewing.
                                       7-33

-------
     Table 7-5. Quarterly and Annual Average Concentrations of the Pollutants of
            Interest for the Colorado Monitoring Sites (Continued)
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Ug/m3)
2nd
Quarter
Average
(Ug/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Ug/m3)
Battlement Mesa, Colorado - BMCO
Acetaldehyde
Benzene
1,3 -Butadiene
Ethylbenzene
Formaldehyde
20/20
51/51
8/51
51/51
20/20
0.71
ฑ0.29
1.37
ฑ0.36
0.02
ฑ0.03
0.24
ฑ0.04
1.42
ฑ0.57
NA
NA
NA
NA
NA
NA
2.65
ฑ1.04
0
0.65
ฑ0.36
NA
NA
1.22
ฑ0.36
0.05
ฑ0.03
0.28
ฑ0.14
NA
NA
NA
NA
NA
NA
Silt, Colorado - BRCO
Acetaldehyde
Benzene
1,3 -Butadiene
Formaldehyde
6/6
53/54
7/54
6/6
NA
1.06
ฑ0.29
0.02
ฑ0.02
NA
NA
0.48
ฑ0.12
0
NA
NA
NA
NA
NA
NA
0.85
ฑ0.15
0.02
ฑ0.02
NA
NA
0.86
ฑ0.13
0.01
ฑ0.01
NA
Parachute, Colorado - PACO
Acetaldehyde
Benzene
1,3 -Butadiene
Ethylbenzene
Formaldehyde
21/21
52/53
22/53
53/53
21/21
0.85
ฑ0.33
1.61
ฑ0.47
0.13
ฑ0.06
0.25
ฑ0.07
1.41
ฑ0.33
NA
NA
NA
NA
NA
NA
1.61
ฑ0.46
0
0.59
ฑ0.27
NA
1.07
ฑ0.29
1.56
ฑ0.35
0.11
ฑ0.05
0.27
ฑ0.06
1.83
ฑ0.39
NA
1.45
ฑ0.21
0.07
ฑ0.03
0.32
ฑ0.07
NA
Rifle, Colorado - RICO
Acetaldehyde
Benzene
1,3 -Butadiene
Ethylbenzene
Formaldehyde
17/17
51/53
42/53
53/53
17/17
1.47
ฑ0.58
1.48
ฑ0.36
0.27
ฑ0.09
0.34
ฑ0.06
1.99
ฑ0.84
NA
0.64
ฑ0.14
0.09
ฑ0.04
0.22
ฑ0.03
NA
NA
NA
NA
NA
NA
NA
1.75
ฑ0.32
0.28
ฑ0.08
0.47
ฑ0.11
NA
NA
1.27
ฑ0.18
0.18
ฑ0.04
0.35
ฑ0.04
NA
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
a Average concentrations provided for the pollutants below the blue line for GPCO are presented in ng/m3
 for ease of viewing.
                                      7-34

-------
Observations for GPCO from Table 7-5 include the following:

•  The pollutants with the highest annual average concentrations by mass are
   formaldehyde (2.74 ฑ 0.17 |ig/m3), acetaldehyde (2.43 ฑ 0.29 |ig/m3), and benzene
   (1.34 ฑ 0.35 |ig/m3). These are also the only pollutants with annual average
   concentrations greater than 1 |ig/m3.

•  The confidence intervals associated with the quarterly average concentrations of
   acrylonitrile are relatively high and in some case greater than the quarterly averages
   themselves. This pollutant was detected 17 times in 2011 and its measurements
   ranged from 0.0739 |ig/m3to 0.757 |ig/m3. Thus, a large number of zeros (43) were
   substituted into the calculations, leading to the appearance of a higher level of
   variability within the average concentrations, which is reflected in the confidence
   intervals. Nine of the 17 measured detections were measured during the second
   quarter. The minimum and maximum acrylonitrile concentrations were measured
   during this quarter.

•  Acetaldehyde concentrations were highest during third and fourth quarters of the
   year.  This trend is not reflected in the formaldehyde concentrations.

•  Benzene's second quarter confidence interval is nearly as great as the quarterly
   average concentration, indicating the likely influence of outliers. The maximum
   concentration of benzene was measured at GPCO on June 8, 2011  (10.6 |ig/m3). This
   concentration is more than twice the next highest measurement (3.55 |ig/m3 measured
   on November 29, 2011) and more than seven times the next highest concentration
   measured in that quarter (1.49 |ig/m3 measured on April 27, 2011). The June 8th
   concentration is also the fourth highest benzene concentration measured across the
   program.

•  For many VOCs, the fourth quarter averages were greater than the other quarterly
   averages, particularly for 1,3-butadiene, ethylbenzene, and tetrachloroethylene.
   However, the differences across the averages are not statistically significant.

•  The first quarter average concentration of naphthalene is the highest of the quarterly
   averages and has a relatively large confidence interval associated with  it. A review of
   the data shows that concentrations of naphthalene range from 38.1 ng/m3 to
   650 ng/m3, with three of the highest measurements of naphthalene (those greater than
   400 ng/m3) measured in January and February of 2011. GPCO has three of six highest
   naphthalene concentrations measured across the program

•  Average benzo(a)pyrene concentrations for the first and fourth quarters of 2011 are
   significantly higher than the average concentrations for the other two quarters. A
   review of the data shows that 22 of the 25 concentrations of this pollutant greater than
   0.01 ng/m3 were measured during the first and fourth quarters of 2011. Conversely,
   14 of the 16 non-detects were measured during the second and third quarters of 2011.
   Three measurements of benzo(a)pyrene greater than 1 ng/m3 were measured during a
   1-month period between November and December 2011; these measurements account
   for three of the seven measurements of benzo(a)pyrene greater than 1 ng/m3 measured
   across the program for all NMP sites sampling PAHs.

                                   7-35

-------
       Observations for the Garfield County sites from Table 7-5 include the following:

       •   Of the SNMOCs, benzene has the highest annual average concentration by mass for
          each of the Garfield County sites. Annual average concentrations of benzene ranged
          from 0.86 ฑ 0.13 |ig/m3 for BRCO to 1.45 ฑ 0.21 |ig/m3 for PACO. While PACO's
          benzene concentrations were steady across 2011, the quarterly average concentrations
          of benzene for the other sites exhibit more variability.

       •   The annual average concentrations of 1,3-butadiene varied significantly across the
          Garfield County sites, ranging from 0.01 ฑ 0.0.1 |ig/m3 for BRCO to
          0.18 ฑ 0.04 |ig/m3 for RICO. A comparison of the Garfield County sites shows that
          the 10 highest concentrations of 1,3-butadiene were measured at RICO, all of which
          were measured during the first and fourth quarters of 2011. Note that this pollutant
          was not detected during the second quarter at BRCO nor at BMCO and PACO during
          the third quarter of 2011. However, the lack of quarterly averages across all  sites and
          all quarters makes a seasonal trend difficult to determine.

       •   Annual average concentrations for the carbonyl compound pollutants of interest could
          not be calculated for any of the Garfield  County sites because these sites did not meet
          the quarterly completeness criteria specified above. However, Appendix L provides
          the pollutant-specific average concentrations for all valid samples collected over the
          entire sample period for each site.

       •   Annual average concentrations for the SNMOC pollutants of interest could not be
          calculated for BMCO because the overall method completeness criterion was not met.
          Appendix K also provides the pollutant-specific average concentrations for all valid
          SNMOC samples collected over the entire sample period.


       Tables 4-9 through 4-12 present the sites with the 10 highest annual average

concentrations for each of the program-level pollutants of interest. Observations for the  Colorado

sites from those tables include the following:

       •   Annual average concentrations for GPCO appear in Tables 4-9 through 4-12 a total of
          14 times.

       •   GPCO appears frequently in Table 4-9 for VOCs, although its highest ranking is
          fourth (ethylbenzene). PACO's annual average benzene concentration ranked third
          among all sites sampling this pollutant, with GPCO and RICO ranking sixth  and
          seventh, respectively. RICO and GPCO rank fourth and seventh for 1,3-butadiene,
          respectively.

       •   GPCO's annual average acetaldehyde concentration ranked sixth highest among
          NMP sites sampling carbonyl compounds, as shown in Table 4-10. GPCO's
          formaldehyde concentration does not appear in this table (it ranked 12th).

       •   As shown in Table 4-11 for the PAHs, GPCO has the highest annual concentration of
          naphthalene and benzo(a)pyrene among  all NMP sites sampling PAHs. This site also


                                         7-36

-------
          has the fourth highest annual average of acenaphthene and the sixth highest annual
          average of fluorene.

7.4.2   Concentration Comparison
       In order to better illustrate how a site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for the selected NATTS MQO Core
Analytes listed in Section 3.5.3, where applicable. Thus, box plots for benzene and 1,3-butadiene
were created for each of the Colorado sites. Note that these box plots are 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). Box plots for acetaldehyde and
formaldehyde were created for GPCO (the only site for which annual averages for these
pollutants could be calculated). Box plots  were also created for benzo(a)pyrene, hexavalent
chromium,  and naphthalene for GPCO. Figures 7-23 through 7-29 overlay the sites' minimum,
annual average, and maximum concentrations onto the program-level minimum, first quartile,
median, average, third quartile, and maximum concentrations, as described in Section 3.5.3.
        Figure 7-23. Program vs. Site-Specific Average Acetaldehyde Concentration
 6PCC
                                         Concentration (
                Program:   1st Quartile    2nd Quartile    3rd Quartile     4th Quartile     Average
                Site:     Site Average    Site Minimum/Maximum

 Figure 7-24a. Program vs. Site-Specific Average Benzene (Method TO-15) Concentration
 GPCO
                                                           Program Max Concentration = 23.8 ng/m3
                                     4       5
                                         Concentration (
Program:
Site:
1st Quartile
Site Average
o
2nd Quartile
3rd Quartile
n
4th Quartile Average
n
^ 1 — 1
Site Minimum/Maximum
                                          7-37

-------
   Figure 7-24b. Program vs. Site-Specific Average Benzene (SNMOC) Concentrations
BRCC
                                          3            4
                                           Concentration (mj/mj)
                Program:   IstQuartile      2ndQuartile     SrdQuartile      4thQuartile      Average
                Site:     Site Average     Site Minimum/Maximum

      Figure 7-25. Program vs. Site-Specific Average Benzo(a)pyrene Concentration

            -o-
                                     0.75          1          1.Z5
                                            Concentration (ng/m3)
                Program:   IstQuartile      2ndQuartile     3rdQuartile      4thQuartile      Average
                Site:      Site Average     Site Minimum/Maximum

     Figure 7-26a. Program vs. Site-Specific Average 1,3-Butadiene (Method TO-15)
                                        Concentration
                                                                Program Max Concentration = 9.51
Program:
Site:

IstQuartile
Site Average
0
2ndQuartile
SrdQuartile
n
4thQuartile Average
n
^m i 	 i
Site Minimum/Maximum




                                             7-38

-------
Figure 7-26b. Program vs. Site-Specific Average 1,3-Butadiene (SNMOC) Concentrations
BRCC
                                                                     Program Max Concentration = 2.35 ug/rna
                                                                     Program Max Concentration = 2.35 ug/ma
                    -O
                                                                     Program Max Concentration = 2.35
                                          0.4        3.5
                                              Concentration (
                 Program:    IstQuartile     2ndQuartile     SrdQuartile      4thQuartile      Average
                 Site:      Site Average     Site Minimum/Maximum

       Figure 7-27. Program vs. Site-Specific Average Formaldehyde Concentration

                                                    IS
                                              Concentration
                 Program:   IstQuartile      2ndQuartile     SrdQuartile      4thQuartile      Average
                 Site:      Site Average     Site Minimum/Maximum

  Figure 7-28. Program vs. Site-Specific Average Hexavalent Chromium Concentration
                                                   0.15
                                              Concentration (ng/m3)
                 Program:   IstQuartile      2ndQuartile     SrdQuartile      4thQuartile      Average
                 Site:      Site Average     Site Minimum/Maximum

                                                7-39

-------
       Figure 7-29. Program vs. Site-Specific Average Naphthalene Concentration
GFCC

^m
^^*
3 50

i Program Max Concentration

100 150 200 250 300 350 400
Concentration (ng/mi)
J
= 779 ng/m3

450 5C
Program
Site:
: IstQuartile
Site Average
O
2ndQuartile 3rd Quartile 4thQuartile Ave
Site Minimum/Maximum
rage

      Observations from Figures 7-23 through 7-29 include the following:

         •  Figure 7-23 shows that GPCO's annual average acetaldehyde concentration is
            greater than the program-level average concentration and just less than the
            program-level third quartile (or 75th percentile). The maximum acetaldehyde
            concentration measured at GPCO is less than the maximum concentration
            measured across the program.

         •  Figure 7-24a presents the benzene concentrations for GPCO compared to the
            benzene concentrations measured across the program for NMP sites sampling
            VOCs with Method TO-15; Figure 7-24b presents the annual average benzene
            concentrations for the Garfield County sites compared to the benzene
            concentrations measured across the program for NMP sites sampling SNMOCs.
            The box plots are presented this way to correspond with Tables 4-1 and 4-2 in
            Section 4.1, as discussed in Section 3.5.3.

         •  Figure 7-24a is the box plot for benzene for GPCO. The program-level maximum
            concentration (23.8 |ig/m3) is not shown directly on the box plot because the scale
            of the box plot would be too large to readily observe data points at the lower end
            of the concentration range.  Thus, the scale has been reduced to 10 |ig/m3. The box
            plot shows that the annual average  benzene concentration for GPCO is greater
            than the program-level average concentration as well as the third quartile for the
            program. While the maximum benzene  concentration measured at GPCO is less
            than the maximum measured across the program, it is greater than the scale on the
            graph (10.6 |ig/m3).

         •  The annual average benzene concentrations for PACO and RICO are greater than
            the program-level average concentration while the annual average concentration
            for BRCO is just less than the program-level average, as shown in Figure 7-24b.
            The annual averages for PACO and RICO are also greater than the program-level
            third quartile. The maximum benzene concentration measured by sites sampling
            SNMOCs was not measured at the  Garfield County sites.

         •  Figure 7-25 is the box plot  for benzo(a)pyrene for GPCO. Note that the program-
            level first quartile for this pollutant is zero and is not visible on this box plot. The
            box plot shows that the annual average concentration for GPCO is greater than the
                                        7-40

-------
   program-level average concentration as well as the program-level third quartile.
   Recall from the previous section that GPCO has the highest annual average
   benzo(a)pyrene concentration among all sites sampling PAHs. Figure 7-25 also
   shows that while the maximum benzo(a)pyrene concentration measured at GPCO
   (1.54 ng/m3) is not the maximum concentration measured across the program
   (1.99 ng/m3), it was the second highest.  The third highest concentration measured
   across the program was also measured at GPCO. Several non-detects of
   benzo(a)pyrene were measured at GPCO.

•  Similar to the box plots for benzene, Figure 7-26a presents the annual average
   concentration of 1,3-butadiene for GPCO compared to the  1,3-butadiene
   concentrations measured across the program for NMP sites sampling VOCs with
   Method TO-15; Figure 7-26b presents the annual average 1,3-butadiene
   concentrations for the Garfield County sites compared to the 1,3-butadiene
   concentrations measured across the program for NMP sites sampling SNMOCs.
   Note that the program-level maximum concentrations are not shown directly on
   either box plot as the scale has been reduced to allow for the observation of data
   points at the lower end of the concentration range.

•  Figure 7-26a shows that, similar to benzene, GPCO's annual average
   1,3-butadiene concentration is greater than the program-level average
   concentration and greater than the third quartile for the program. However, the
   maximum concentration measured at GPCO is considerably less than the
   maximum concentration measured across the program.  A few non-detects of
   1,3-butadiene were measured at GPCO.

•  The first and second quartiles (and median concentration) at the program-level are
   zero, and thus, not shown in Figure 7-26b,  indicating that at least half of the
   1,3-butadiene concentrations measured by  sites sampling SNMOCs were
   non-detects. Figure 7-26b shows that, of the Garfield County sites for which
   annual average concentrations could be calculated, RICO's annual average
   1,3-butadiene concentration is the highest and BRCO's annual average the lowest.
   This figure also shows that the annual average concentration for BRCO is less
   than the program-level average, the annual average concentration for PACO is
   similar to the program-level average, and the annual average concentration for
   RICO is more than twice the program-level average concentration.

•  Figure 7-27 shows that GPCO's  annual average formaldehyde concentration is
   just less than the program-level average concentration.  The range of
   formaldehyde concentrations measured at GPCO appears rather small, ranging
   from 1.65 |ig/m3 to 4.28 |ig/m3. Note that the minimum formaldehyde
   concentration measured at GPCO is greater than the program-level first quartile.
                               7-41

-------
          •  Figure 7-28 is the box plot for hexavalent chromium for GPCO. The figure shows
             that the annual average concentration for GPCO is less than both the program-
             level average and median concentrations. Although the maximum concentration
             measured at GPCO is less than the maximum concentration measured across the
             program, GPCO's maximum concentration is the seventh highest measurement of
             this pollutant among sites sampling hexavalent chromium. Several non-detects of
             hexavalent chromium were measured at GPCO.
          •  Figure 7-29 is the box plot for naphthalene. Note that the program-level
             maximum concentration (779 ng/m3) is not shown directly on the box plot as the
             scale has been reduced to 500 ng/m3 to allow for the observation of data points at
             the lower end of the concentration range. Figure 7-29 shows that the annual
             average naphthalene concentration for GPCO is greater than the program-level
             average concentration as well as the program-level third quartile. Recall from the
             previous section that GPCO has the highest annual average naphthalene
             concentration among all sites sampling PAHs. The maximum naphthalene
             concentration measured at GPCO (650 ng/m3) is the second highest measurement
             across the  program. The minimum concentration of naphthalene measured at
             GPCO (38.1  ng/m3) is greater than the program-level first quartile (32.8 ng/m3).

7.4.3   Concentration Trends
       A site-specific trends evaluation was completed for sites that have sampled one or more
of the selected NATTS MQO  Core Analytes for 5 consecutive years or longer, as described in
Section 3.5.4. GPCO has  sampled carbonyl compounds and VOCs since 2004 and hexavalent
chromium since 2005. Thus, Figures 7-30 through 7-35 present the annual statistical metrics for
acetaldehyde, benzene, 1,3-butadiene, formaldehyde, and hexavalent chromium for GPCO,
respectively. The statistical metrics presented for assessing trends include the substitution of
zeros for non-detects.
       GPCO began sampling PAHs and BRCO, PACO, and RICO began sampling SNMOCs
and carbonyl compounds under the NMP in 2008; because this is fewer than 5 consecutive years
of sampling, the trends analysis was not conducted for the listed pollutants for these methods.
BMCO began sampling SNMOCs and carbonyl compounds under the NMP at the end of 2010;
thus, the trends analysis was not conducted for this site either.
                                         7-42

-------
Figure 7-30. Annual Statistical Metrics for Acetaldehyde Concentrations
                          Measured at GPCO


1
TS ,= .
ge Concentration fi'
j I
C ™
8
=t












The maximum
concentrationfor |
2004 is 93.0 Hg/m3 |




ซ.
^^m
— f—


I
2004





T T T
p— i Li
E , t i I

|_l tf^
^ -^ --- * -t- '-•-'
2005 2006 2007 2008 2009 2010 2011
Year
• 5th Percentile — Minimum — Median — Maximum • 95th Percentile "^--Average
   Figure 7-31. Annual Statistical Metrics for Benzene Concentrations
                          Measured at GPCO
      2004       2005       2006       2007       200B      2009      2010
         5th Percentile    — Minimum     — Median     —  Maximum     *  95th Percentile   * + ^.. Average
                                  7-43

-------
Figure 7-32. Annual Statistical Metrics for 1,3-Butadiene Concentrations
                           Measured at GPCO
       2004       2005      2006      2007      200S       2009       2010       2011
                                      Year
          5th Percentile     —  Minimum    —  Median    —  Maximum    •  95th Percentile   "^--Average
Figure 7-33. Annual Statistical Metrics for Formaldehyde Concentrations
                           Measured at GPCO


E
B
.&
|
1
a
& s
e 8
1






4
.•

The maximum
concentration for
2004 is 40.5 u.g/m.3



rn Jn ^ rh ฑ
I
_ 	 "*" *^ -•-..
• I f
U.J 1 L^
r
k L
2004 2005 2006 2007 200B 2009 2010 2011
Year
• 5th Percentile — Minimum — Median — Max mum • 95th Percentile * + ^.. Average
                                   7-44

-------
     Figure 7-34. Annual Statistical Metrics for Hexavalent Chromium Concentrations
                                  Measured at GPCO
        r
        E
        .2
        E
        •ฃ 0.4
        I
        3
        a,
        2
                - Minimum
                                    -  Maximum
                                                9SthPercentile
                                                           • SthPercentile
                                                                        • Average
       Observations from Figure 7-30 for acetaldehyde measurements at GPCO include the
following:

       •  The maximum acetaldehyde concentration was measured during 2004. The maximum
          concentrations measured in subsequent time periods were significantly lower. The
          two highest acetaldehyde concentrations (93.0 |ig/m3 and 54.9 |ig/m3) were measured
          in 2004 and the six highest acetaldehyde concentrations (ranging from 6.35 |ig/m3to
          93.0 |ig/m3) were all measured in 2004 and 2005.

       •  After the first two years of sampling, the median and average concentrations fluctuate
          only slightly from year to year. The average concentration ranged from 2.00 |ig/m3
          for 2010 to 2.90 |ig/m3 for 2009  for the period from 2006 to 2011.

       •  Although difficult to discern in Figure 7-30, the average and median concentrations
          differ less than 0.15 |ig/m3 for each year after 2005, indicating relatively little
          variability in the central tendency of acetaldehyde concentrations measured over the
          periods shown.
                                          7-45

-------
       Observations from Figure 7-31 for benzene measurements 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.

       •   Even with the maximum concentration, most of the statistical metrics for 2011
          decreased from 2010 to 2011.

       •   Although there have been fluctuations, both the average  and median concentrations of
          benzene have decreased slightly over time.

       •   There have been no non-detects of benzene reported over the period of sampling.


       Observations from Figure 7-32 for 1,3-butadiene measurements at GPCO include the

following:

       •   The maximum 1,3-butadiene concentration was measured on December 11, 2004 and
          is the only 1,3-butadiene concentration greater than 1 |ig/m3 measured at GPCO.

       •   The average concentrations appear to have a slight decreasing trend; however,
          confidence intervals calculated from the individual concentrations show that this
          decrease is not statistically significant.

       •   The difference between the average and the median concentrations is at a minimum
          for 2011, which can be an indication of decreasing variability in the measurements.

       •   The number of non-detects, and subsequently zeros substituted for non-detects,
          decreased from approximately 30 percent in 2004 and 2005, to eight percent in 2006,
          to none in 2007, 2008, and 2009. The number of non-detects began to increase
          slightly after 2009, up to three percent in 2010 and nearly 12 percent in 2011.


       Observations from Figure 7-33 for formaldehyde measurements at GPCO include the

following:

       •   The trends graph for formaldehyde resembles the trends  graph for acetaldehyde in
          that the maximum formaldehyde concentration was measured in 2004 and is
          significantly higher than the maximum concentrations measured in subsequent years.
          The three highest concentrations of formaldehyde were measured on the same days as
          the three highest acetaldehyde concentrations.

       •   Even with decreasing maximum concentrations, the average formaldehyde
          concentrations (as well as several other statistical parameters) have a slight increasing
          trend through 2006. Between 2006 and 2009, the average concentration was
          approximately 4 |ig/m3. A significant decrease in all of the statistical metrics is shown
          for 2010, with little change for 2011.

                                         7-46

-------
       •  The difference between the 5th and 95th percentiles has been decreasing since 2007,
          and is at a minimum for 2011, which indicates a smaller range in the majority of the
          concentrations measured. The decreasing difference between the median and average
          concentrations is a further indicator of the decreasing variability in the formaldehyde
          measurements at GPCO.


       Observations from Figure 7-34 for hexavalent chromium measurements at GPCO include
the following:

       •  The maximum hexavalent chromium concentration was measured on July 5, 2008
          (0.685 ng/m3). Only two additional hexavalent chromium concentrations measured at
          GPCO are greater than 0.1 ng/m3, one  measured on December 29, 2011 (0.154 ng/m3)
          and the other on August 9, 2006 (0.113 ng/m3).

       •  The average concentration of hexavalent chromium for 2008 is being driven by the
          outlier measured that year. If that measurement was removed from the calculation,
          concentrations of hexavalent chromium would exhibit a steady decreasing trend
          between 2006 and 2009, followed by an increasing trend for 2010 and 2011.

       •  Both the minimum concentration and 5th percentile for all years of sampling are zero,
          indicating the presence of non-detects. For 2009, the median (or 50th percentile) is
          also zero indicating that at least half of the measurements were non-detects. The
          percentage of non-detects has ranged from 18 percent in 2006 to 60 percent in 2009.


7.5    Additional Risk-Based Screening Evaluations

       The following risk-based screening evaluations were conducted to  characterize risk at

each Colorado monitoring site. Refer to Sections 3.3 and 3.5.5 for definitions and explanations

regarding the various toxicity factors, time frames, and calculations associated with these risk-

based screenings.
7.5.1   Risk-Based Screening Assessment Using MRLs
       A risk-based screening was conducted by comparing the concentration data from the

Colorado monitoring sites to the ATSDR MRLs, where available. As described in Section 3.3,

MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute

(exposures of 1 to 14 days); intermediate (exposures of 15 to 364 days); and chronic (exposures

of 1 year or greater). The preprocessed daily measurements of the pollutants of interest were

compared to the acute MRLs; the quarterly averages were compared to the intermediate MRLs;

and the annual averages were compared to the chronic MRLs.
                                         7-47

-------
         As discussed in Section 4.2.2, none of the measured detections or time-period average
   concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
   noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2011.

   7.5.2  Cancer Risk and Noncancer Hazard Approximations
         For the pollutants of interest for the Colorado 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 may want to shift or
   confirm their air-monitoring priorities.  Refer to Section 3.5.5.2 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.

               Table 7-6. Risk Approximations for the Colorado Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs.
# of Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Grand Junction, Colorado - GPCO
Acenaphthene3
Acetaldehyde
Acrylonitrile
Benzene
Benzo(a)pyrenea
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
0.000088
0.0000022
0.000068
0.0000078
0.00176
0.00003
0.000006
	

0.009
0.002
0.03
_
0.002
0.1
0.098
61/61
60/60
17/60
60/60
44/61
53/60
60/60
42/60
0.01
ฑ<0.01
2.43
ฑ0.29
0.06
ฑ0.03
1.34
ฑ0.35
<0.01
ฑ0.01
0.13
ฑ0.02
0.53
ฑ0.04
0.08
ฑ0.02
0.93
5.35
4.07
10.42
0.35
4.01
3.20
	

0.27
0.03
0.04
_
0.07
0.01
<0.01
— = a Cancer URE or Noncancer RfC is not available.
NA = Not available due to the criteria for calculating an annual average.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 7-5.
                                             7-48

-------
         Table 7-6. Risk Approximations for the Colorado Monitoring Sites (Continued)
| Pollutant
ฃ>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Fluorene3
Formaldehyde
Hexachloro- 1 , 3 -butadiene
Hexavalent Chromium3
Naphthalene3
1 , 1 ,2,2-Tetrachloroethane
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
Cancer
URE
(Hg/m3)1
0.000011
0.000026
0.0000025
0.000088
0.000013
0.000022
0.012
0.000034
0.000058
0.00000026
0.0000048
0.0000088
Noncancer
RfC
(mg/m3)
0.8
2.4
1

0.0098
0.09
0.0001
0.003

0.04
0.002
0.1
#of
Measured
Detections
vs.
# of Samples
32/60
16/60
60/60
61/61
60/60
8/60
42/59
61/61
7/60
53/60
22/60
6/60
Annual
Average
(Hg/m3)
0.04
ฑ0.01
0.03
ฑ0.01
0.62
ฑ0.11
0.01
ฑ<0.01
2.74
ฑ0.17
0.01
ฑ0.01
0.01
ฑ0.01
0.16
ฑ0.03
0.01
ฑ0.01
0.26
ฑ0.05
0.04
ฑ0.02
O.01
ฑO.01
Cancer Risk
Approximation
(in-a-million)
0.44
0.69
1.54
0.68
35.60
0.32
0.19
5.29
0.56
0.07
0.21
0.02
Noncancer
Hazard
Approximation
(HQ)
O.01
0.01
O.01

0.28
O.01
0.01
0.05

0.01
0.02
O.01
Battlement Mesa, Colorado - BMCO
Acetaldehyde
Benzene
1,3 -Butadiene
Ethylbenzene
Formaldehyde
0.0000022
0.0000078
0.00003
0.0000025
0.000013
0.009
0.03
0.002
1
0.0098
20/20
51/51
8/51
51/51
20/20
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Silt, Colorado - BRCO
Acetaldehyde
Benzene
1,3 -Butadiene
Formaldehyde
0.0000022
0.0000078
0.00003
0.000013
0.009
0.03
0.002
0.0098
6/6
53/54
7/54
6/6
NA
0.86
ฑ0.13
0.01
ฑ0.01
NA
NA
6.71
0.34
NA
NA
0.03
0.01
NA
— = a Cancer URE or Noncancer RfC is not available.
NA = Not available due to the criteria for calculating an annual average.
3 For the annual average concentration of this pollutant in ng/m3, refer to Table 7-5.
                                                 7-49

-------
         Table 7-6. Risk Approximations for the Colorado Monitoring Sites (Continued)
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs.
# of Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Parachute, Colorado - PACO
Acetaldehyde
Benzene
1,3 -Butadiene
Ethylbenzene
Formaldehyde
0.0000022
0.0000078
0.00003
0.0000025
0.000013
0.009
0.03
0.002
1
0.0098
21/21
52/53
22/53
53/53
21/21
NA
1.45
ฑ0.21
0.07
ฑ0.03
0.32
ฑ0.07
NA
NA
11.27
2.10
0.79
NA
NA
0.05
0.04
0.01
NA
Rifle, Colorado - RICO
Acetaldehyde
Benzene
1,3 -Butadiene
Ethylbenzene
Formaldehyde
0.0000022
0.0000078
0.00003
0.0000025
0.000013
0.009
0.03
0.002
1
0.0098
17/17
51/53
42/53
53/53
17/17
NA
1.27
ฑ0.18
0.18
ฑ0.04
0.35
ฑ0.04
NA
NA
9.93
5.53
0.87
NA
NA
0.04
0.09
0.01
NA
— = a Cancer URE or Noncancer RfC is not available.
NA = Not available due to the criteria for calculating an annual average.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 7-5.
         Observations for GPCO from Table 7-6 include the following:

         •   Formaldehyde, acetaldehyde, and benzene have the highest annual average
             concentrations for GPCO.

         •   Formaldehyde also has the highest cancer risk approximation (35.60 in-a-million) for
             this site. Benzene has the second highest cancer risk approximation
             (10.42 in-a-million) and acetaldehyde has the third highest cancer risk approximation
             (5.35 in-a-million), with naphthalene a close fourth (5.29 in-a-million).

         •   None  of the pollutants of interest for GPCO have noncancer hazard approximations
             greater than  1.0, indicating no adverse health effects are expected from these
             individual pollutants. Formaldehyde has the highest noncancer hazard approximation
             (0.28) among the pollutants of interest for GPCO.
                                             7-50

-------
       Observations for the Garfield County sites from Table 7-6 include the following:

       •  Benzene's cancer risk approximation is the highest among each site's pollutants of
          interest, where risk approximations could be calculated. Benzene's cancer risk
          approximations range from 6.71 in-a-million for BRCO to 11.27 in-a-million for
          PACO. PACO's benzene cancer risk approximation is the third highest benzene
          cancer risk approximation  compared to other NMP sites.

       •  None of the noncancer hazard approximations calculated  for the Garfield County sites
          are greater than 1.0, indicating no adverse health effects are expected from these
          individual pollutants. The highest noncancer hazard approximation was calculated for
          1,3-butadiene for RICO (0.09).

       •  Annual averages, and thus cancer risk and noncancer hazard approximations, could
          not be calculated for acetaldehyde and formaldehyde for any of the Garfield County
          sites because the completeness criteria were not met, as discussed in Section 7.4.1.

       •  Annual averages, and thus cancer risk and noncancer hazard approximations, could
          not be calculated for the SNMOCs for BMCO due to completeness issues, as
          discussed in Section 7.4.1.


7.5.3   Risk-Based Emissions Assessment

       In addition to the risk-based screenings discussed above, Tables 7-7 and 7-8 present 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 2008 NEI, the 10

pollutants with the highest toxicity-weighted emissions, and the 10 pollutants with the highest

cancer risk approximations (in-a-million), as calculated from the annual averages provided in

Table 7-6. Table 7-8 presents similar information, but identifies the 10 pollutants with the

highest noncancer hazard approximations (HQ), also calculated from annual averages provided

in Table 7-6.
                                          7-51

-------
   Table 7-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
                                                the Colorado Monitoring Sites
to
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Grand Junction, Colorado (Mesa County) - GPCO
Benzene
Formaldehyde
Acetaldehyde
Ethylbenzene
1,3 -Butadiene
Naphthalene
Dichloromethane
POM, Group 2b
Tetrachloroethylene
POM, Group 6
140.94
112.12
45.52
41.56
14.18
7.63
6.38
2.02
1.20
0.20
Formaldehyde
Benzene
1,3 -Butadiene
POM, Group 3
Naphthalene
POM, Group 2b
Arsenic, PM
Hexavalent Chromium, PM
Ethylbenzene
Acetaldehyde
1.46E-03
1.10E-03
4.25E-04
3.93E-04
2.59E-04
1.78E-04
1.58E-04
1.37E-04
1.04E-04
l.OOE-04
Formaldehyde
Benzene
Acetaldehyde
Naphthalene
Acrylonitrile
1,3 -Butadiene
Carbon Tetrachloride
Ethylbenzene
Acenaphthene
1 ,2-Dichloroethane
35.60
10.42
5.35
5.29
4.07
4.01
3.20
1.54
0.93
0.69
Battlement Mesa, Colorado (Garfield County) - BMCO
Formaldehyde
Benzene
Acetaldehyde
Ethylbenzene
1,3 -Butadiene
Naphthalene
POM, Group 2b
Tetrachloroethylene
Dichloromethane
POM, Group 6
324.22
269.17
79.93
27.66
16.04
6.02
2.69
1.34
0.74
0.32
Formaldehyde
Benzene
POM, Group 3
1,3 -Butadiene
POM, Group 2b
Naphthalene
Acetaldehyde
POM, Group 5a
Ethylbenzene
POM, Group 6
4.21E-03
2.10E-03
8.97E-04
4.81E-04
2.37E-04
2.05E-04
1.76E-04
l.OOE-04
6.92E-05
5.61E-05


-------
Table 7-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
                                       the Colorado Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity- Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations
Based on Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
Silt, Colorado (Garfield County) - BRCO
Formaldehyde
Benzene
Acetaldehyde
Ethylbenzene
1,3 -Butadiene
Naphthalene
POM, Group 2b
Tetrachloroethylene
Dichloromethane
POM, Group 6
324.22
269.17
79.93
27.66
16.04
6.02
2.69
1.34
0.74
0.32
Formaldehyde
Benzene
POM, Group 3
1,3 -Butadiene
POM, Group 2b
Naphthalene
Acetaldehyde
POM, Group 5a
Ethylbenzene
POM, Group 6
4.21E-03
2.10E-03
8.97E-04
4.81E-04
2.37E-04
2.05E-04
1.76E-04
l.OOE-04
6.92E-05
5.61E-05
Benzene 6.71
1,3-Butadiene 0.34

Parachute, Colorado (Garfield County) - PACO
Formaldehyde
Benzene
Acetaldehyde
Ethylbenzene
1,3 -Butadiene
Naphthalene
POM, Group 2b
Tetrachloroethylene
Dichloromethane
POM, Group 6
324.22
269.17
79.93
27.66
16.04
6.02
2.69
1.34
0.74
0.32
Formaldehyde
Benzene
POM, Group 3
1,3 -Butadiene
POM, Group 2b
Naphthalene
Acetaldehyde
POM, Group 5a
Ethylbenzene
POM, Group 6
4.21E-03
2.10E-03
8.97E-04
4.81E-04
2.37E-04
2.05E-04
1.76E-04
l.OOE-04
6.92E-05
5.61E-05
Benzene 11.27
1,3-Butadiene 2.10
Ethylbenzene 0.79


-------
Table 7-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
                                       the Colorado Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity- Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Rifle, Colorado (Garfield County) - RICO
Formaldehyde
Benzene
Acetaldehyde
Ethylbenzene
1,3 -Butadiene
Naphthalene
POM, Group 2b
Tetrachloroethylene
Dichloromethane
POM, Group 6
324.22
269.17
79.93
27.66
16.04
6.02
2.69
1.34
0.74
0.32
Formaldehyde
Benzene
POM, Group 3
1,3 -Butadiene
POM, Group 2b
Naphthalene
Acetaldehyde
POM, Group 5a
Ethylbenzene
POM, Group 6
4.21E-03
2.10E-03
8.97E-04
4.81E-04
2.37E-04
2.05E-04
1.76E-04
l.OOE-04
6.92E-05
5.61E-05
Benzene
1,3 -Butadiene
Ethylbenzene
9.93
5.53
0.87


-------
Table 7-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
                                         RfCs for the Colorado Monitoring Sites
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity- Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Grand Junction, Colorado (Mesa County) - GPCO
Toluene
Xylenes
Benzene
Formaldehyde
Methanol
Hexane
Acetaldehyde
Ethylbenzene
Hydrochloric acid
1,3 -Butadiene
240.58
188.54
140.94
112.12
88.05
50.96
45.52
41.56
27.54
14.18
Acrolein
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Benzene
Naphthalene
Arsenic, PM
Xylenes
Hydrochloric acid
Manganese, PM
553,576.79
11,440.45
7,087.62
5,058.01
4,698.01
2,543.73
2,443.88
1,885.37
1,376.84
1,158.03
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Naphthalene
Benzene
Acrylonitrile
Trichloroethylene
Tetrachloroethylene
Carbon Tetrachloride
Chloroform
0.28
0.27
0.07
0.05
0.04
0.03
0.02
0.01
0.01
0.01
Battlement Mesa, Colorado (Garfield County) - BMCO
Toluene
Formaldehyde
Xylenes
Benzene
Hexane
Acetaldehyde
Methanol
Acrolein
Ethylbenzene
1,3 -Butadiene
419.91
324.22
321.33
269.17
82.14
79.93
60.71
40.42
27.66
16.04
Acrolein
Formaldehyde
Benzene
Acetaldehyde
1,3 -Butadiene
Xylenes
Naphthalene
Cyanide Compounds, gas
Lead, PM
Arsenic, PM
2,021,060.94
33,083.40
8,972.23
8,880.90
8,021.37
3,213.31
2,008.12
988.69
430.70
233.88


-------
Table 7-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
                                   RfCs for the Colorado Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity- Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Noncancer
Hazard
Approximation
Pollutant (HQ)
Silt, Colorado (Garfield County) - BRCO
Toluene
Formaldehyde
Xylenes
Benzene
Hexane
Acetaldehyde
Methanol
Acrolein
Ethylbenzene
1,3 -Butadiene
419.91
324.22
321.33
269.17
82.14
79.93
60.71
40.42
27.66
16.04
Acrolein
Formaldehyde
Benzene
Acetaldehyde
1,3 -Butadiene
Xylenes
Naphthalene
Cyanide Compounds, gas
Lead, PM
Arsenic, PM
2,021,060.94
33,083.40
8,972.23
8,880.90
8,021.37
3,213.31
2,008.12
988.69
430.70
233.88
Benzene 0.03
1,3-Butadiene 0.01

Parachute, Colorado (Garfield County) - PACO
Toluene
Formaldehyde
Xylenes
Benzene
Hexane
Acetaldehyde
Methanol
Acrolein
Ethylbenzene
1,3 -Butadiene
419.91
324.22
321.33
269.17
82.14
79.93
60.71
40.42
27.66
16.04
Acrolein
Formaldehyde
Benzene
Acetaldehyde
1,3 -Butadiene
Xylenes
Naphthalene
Cyanide Compounds, gas
Lead, PM
Arsenic, PM
2,021,060.94
33,083.40
8,972.23
8,880.90
8,021.37
3,213.31
2,008.12
988.69
430.70
233.88
Benzene 0.05
1,3-Butadiene 0.04
Ethylbenzene <0.01


-------
Table 7-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
                                   RfCs for the Colorado Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity- Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Rifle, Colorado (Garfield County) - RICO
Toluene
Formaldehyde
Xylenes
Benzene
Hexane
Acetaldehyde
Methanol
Acrolein
Ethylbenzene
1,3 -Butadiene
419.91
324.22
321.33
269.17
82.14
79.93
60.71
40.42
27.66
16.04
Acrolein
Formaldehyde
Benzene
Acetaldehyde
1,3 -Butadiene
Xylenes
Naphthalene
Cyanide Compounds, gas
Lead, PM
Arsenic, PM
2,021,060.94
33,083.40
8,972.23
8,880.90
8,021.37
3,213.31
2,008.12
988.69
430.70
233.88
1,3 -Butadiene
Benzene
Ethylbenzene
0.09
0.04
0.01


-------
       The pollutants listed in Tables 7-7 and 7-8 are limited to those that have cancer and

noncancer toxicity factors, respectively. As a result, although the actual value of the emissions is

the same, the highest emitted pollutants in the cancer table may be different from the noncancer

table. Further, the cancer risk and noncancer hazard approximations based on each site's annual

averages are limited to those pollutants for which each respective monitoring site sampled. As

discussed in Section 7.3, GPCO sampled for VOCs, carbonyl compounds, PAHs, and hexavalent

chromium; the Garfield County sites sampled for SNMOCs and carbonyl compounds only. In

addition, the cancer risk and noncancer hazard approximations are limited to those pollutants

with enough data to meet the criteria for annual averages to be calculated. A more in-depth
discussion of this analysis is provided in Section 3.5.5.3. Similar to the cancer risk and

noncancer hazard  approximations, this analysis may help policy-makers prioritize their air

monitoring activities.


       Observations from Table 7-7 include the following:

       •  Benzene, formaldehyde, and acetaldehyde are the highest emitted pollutants with
          cancer UREs in Mesa County. These same pollutants also topped the list for Garfield
          County, although not in the  same order. Note that the quantity emitted for each
          pollutant was roughly twice as high in Garfield County than Mesa County.

       •  The two pollutants with the  highest toxicity-weighted emissions (of the pollutants
          with cancer UREs) are formaldehyde and benzene for both Mesa and Garfield
          Counties. These two counties have eight pollutants in common among the pollutants
          with the highest toxicity-weighted emissions.

       •  Seven of the highest emitted pollutants also have the highest toxicity-weighted
          emissions in Mesa County while eight of the highest emitted pollutants also have the
          highest toxicity-weighted emissions in Mesa County.

       •  For GPCO, six of the 10 pollutants with the highest cancer risk approximations also
          appear on both emissions-based lists for Mesa County. In addition, POM, Group 2b is
          the eighth highest emitted "pollutant" in Mesa County and ranks sixth for toxicity-
          weighted emissions. POM, Group 2b includes several PAHs sampled for at GPCO
          including acenaphthene, which has the ninth highest cancer risk approximation for
          GPCO.

       •  Where cancer risk approximations could be calculated for the Garfield County sites,
          all of the pollutants of interest listed appear on both emissions-based lists.

       •  POM, Groups 2b, 3, 5a, and 6 appear on Garfield County's emissions-based lists.
          PAHs were not sampled at the Garfield County sites.
                                          7-58

-------
       Observations from Table 7-8 include the following:

       •  Toluene is the highest emitted pollutant with a noncancer RfC in both Mesa and
          Garfield Counties, although the emissions are higher in Garfield County. These two
          counties have an additional eight pollutants in common on their lists of highest
          emitted pollutants with noncancer RfCs.

       •  The two pollutants with the highest toxi city-weighted emissions (of the pollutants
          with noncancer RfCs) for both counties are acrolein and formaldehyde. 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.

       •  Six of the 10 highest emitted pollutants in Mesa County also have the highest
          toxi city-weighted emissions. Six of the 10 highest emitted pollutants in Garfield
          County (including acrolein) also have the highest toxicity-weighted emissions. This is
          the only county for which acrolein appears among the highest emitted pollutants.

       •  Formaldehyde, acetaldehyde, benzene,  and  1,3-butadiene appear on all three lists for
          GPCO. Additionally, 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.

       •  Benzene and 1,3-butadiene appear on all three lists for the Garfield County sites.
          Ethylbenzene, a pollutant of interest for PACO and RICO but not BRCO, is one of
          the highest emitted pollutants in Garfield County, but is not among the most toxic.


7.6    Summary of the 2011 Monitoring Data for the Colorado Monitoring Sites

       Results from several of the data treatments described in this section include the
following:

       ปซป  Twenty-three pollutants failed screens for GPCO. The number of pollutants failing
          screens for the Garfield County sites ranged from four to five.

       ปซป  The pollutants with the highest annual average concentrations for GPCO were
          formaldehyde, acetaldehyde, and benzene. These were also the only pollutants with
          annual average concentrations greater than 1 jug/m3. Benzene had the highest annual
          average concentration for each of the Garfield County sites.

       *ป*  GPCO has the highest annual average  concentrations of naphthalene and
          benzo(a)pyrene among allNMP sites sampling PAHs.

       ปซป  Benzene concentrations at GPCO have an overall decreasing trend. In recent years,
          concentrations ofhexavalent chromium have increased at GPCO while
          concentrations of formaldehyde have decreased.
                                          7-59

-------
8.0    Site in the District of Columbia
       This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS site in Washington, D.C., and integrates these
concentrations with emissions, meteorological, and risk information. Data generated by sources
other than ERG are not included in the data analyses contained in this report. Readers are
encouraged to refer to Sections  1 through 4 and the glossary (Appendix P) for detailed
discussions and definitions regarding the various data analyses presented below.

8.1    Site Characterization
       This section characterizes the Washington, D.C. monitoring site by providing
geographical and physical information about the location of the site and the surrounding area.
This information is provided to  give the reader insight regarding factors that may influence the
air quality near the site and assist in the interpretation of the ambient monitoring measurements.

       Figure 8-1 is a composite satellite image retrieved from ArcGIS Explorer showing the
monitoring site in its urban location. Figure 8-2 identifies nearby point source emissions
locations by source category, as reported in the 2008 NEI for point sources. Note that only
sources within 10 miles of the site are included in the facility counts provided in Figure 8-2. A
10-mile boundary was chosen to give the reader an indication of which emissions sources and
emissions source categories could potentially have a direct effect on the air quality at the
monitoring site. Further, this boundary provides both the proximity of emissions sources to the
monitoring site as well as the quantity of such sources within a given distance of the site. Sources
outside the 10-mile radius are still visible on the map, but have been grayed out in order to show
emissions sources just outside the boundary. Table  8-1 provides  supplemental geographical
information such as land use, location setting, and locational coordinates.
                                            3-1

-------
                                        Figure 8-1. Washington, B.C. (WADC) Monitoring Site
oo
to

-------
 Figure 8-2. NEI Point Sources Located Within  10 Miles of WADC
                                          ^.Montgomery    /     Prince George's  \
                                                CountV     /        County     ,
77-20'drW         77'15'0'W          77'1CrO*Vtf          TT'S'ITW           77-fl'trW          76"55T3*W
                                           Note: Due to facility density and collocation, the total facilities
         •         i                         displayed may not represent all facilities within the area of interest.

          @  WADC NATTS site        10 rnile radius |     j  County boundary

         Source Category Group (No. of Facilities)
          ts   Air-conditioning/Refrigeration (2)
          -f   Aircraft Operations (27)
           I   Asphalt Processing/Roofing Manufacturing (3)
           *   Electricity Generation via Combustion (3)
           F   Food Processing/Agriculture (1)
          13   Hospital (5)
          ^   Institutional - school (10)
           A   Military Base/National Security (5)
          ?   Miscellaneous Commercial/Industrial (18)
           P   Printing/Publishing (5)
          *   Transportation and Marketing of Petroleum Products  (1)
                                           8-3

-------
                            Table 8-1. Geographical Information for the Washington, D.C. Monitoring Site

Site
Code
WADC

AQS Code
11-001-0043

Location
Washington

County
District
Of
Columbia
Micro- or
Metropolitan
Statistical Area
Washington-
Arlington-
Alexandria, DC-
VA-MD-WVMSA
Latitude
and
Longitude
38.921847,
-77.013178

Land Use
Commercial

Location
Setting
Urban/City
Center

Additional Ambient Monitoring Information1
Arsenic, CO, VOCs, SO2, NOy, NO, NO2, NOX,
PAMS, Carbonyl compounds, O3, Meteorological
parameters, PM10, PM25, PM10 Speciation, Black
carbon, PM Coarse, PM2 5 Speciation.
  BOLD ITALICS = EPA-designated NATTS Site
oo

-------
       Figure 8-1 shows that the WADC monitoring site is located in an open field at the
southeast of end of the McMillan Water Reservoir in Washington, D.C. It is also located near
several heavily traveled roadways. The site is located in a commercial area, and is surrounded by
a hospital, a cemetery, and a university. As Figure 8-2 shows, WADC is surrounded
predominantly by sources in the aircraft operations source category. This category includes
airports as well as small runways, heliports, or landing pads.  Aside from aircraft operations, the
source category with the highest number of sources within 10 miles of the WADC monitoring
site is schools. The closest sources to WADC are hospitals and heliports at hospitals.

       Table 8-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Washington D.C. monitoring site.  Table 8-2 includes county-
level population and vehicle registration information. Table  8-2 also includes a county-level
vehicle registration-to-population ratio, which was calculated to represent the number of vehicles
per person within the District of Columbia. In addition, the population within 10 miles of the site
is presented, based on postal code population data estimates.  An estimate of 10-mile vehicle
ownership was then determined by applying the county-level vehicle registration-to-population
ratio to the 10-mile population surrounding the monitoring site. Table 8-2 also contains traffic
volume information for WADC. Finally, Table 8-2 presents the county-level daily VMT for the
District of Columbia.
  Table 8-2. Population, Motor Vehicle, and Traffic Information for the Washington, D.C.
                                      Monitoring Site



Site
WADC

Estimated
County
Population1
617,996

County-level
Vehicle
Registration2
213,232
Vehicles per
Person
(Registration:
Population)
0.35

Population
within 10
miles3
1,931,834
Estimated
10 mile
Vehicle
Ownership
666,556
Annual
Average
Daily
Traffic4
7,700
County-
level
Daily
VMT5
9,775,000
1 County-level population estimate reflects 2011 data from the U.S. Census Bureau (Census Bureau, 2012b)
2 County-level vehicle registration reflects 2010 data from the Federal Highway Administration (FHWA, 2011)
3 10-mile population estimate reflects 2011 data from Xionetic (Xionetic, 2011)
4 AADT reflects 2009 data from the District DOT (DC DOT, 2011)
5 County-level VMT reflects 2011 data from the District DOT (DC DOT, 2012)
BOLD ITALICS = EPA-designaled NATTS Site
                                            8-5

-------
       Observations from Table 8-2 include the following:
       •  The District's population is in the middle of the range compared to other counties
          with NMP sites. Its 10-mile population however, ranks fifth highest and is three times
          higher than its county-level population.
       •  The District-level vehicle registration is in the bottom third compared to other
          counties with NMP sites, while its 10-mile ownership is in the middle of the range.
       •  The vehicle-per-person ratio is among the lowest compared to other NMP sites.
       •  The traffic volume experienced near WADC is in the bottom third compared to other
          NMP monitoring sites. The traffic volume provided is for the intersection of Bryant
          Street and First Street.
       •  The district-level VMT is in the middle-third compared to other county-level VMT,
          where VMT is available.

8.2    Meteorological Characterization
       The following sections characterize the meteorological conditions near the monitoring
site in Washington, D.C. on sample days, as well as over the course of the year.

8.2.1   Climate Summary
       Located on the Potomac River that divides Virginia and Maryland, the capital enjoys all
four seasons, although its weather is somewhat variable. Summers are warm and often humid,
where southerly winds prevail, which 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. Precipitation is evenly distributed across the seasons
(Bair, 1992).

8.2.2   Meteorological Conditions in 2011
       Hourly meteorological data from the NWS weather station nearest this site were retrieved
for 2011 (NCDC, 2011). The closest weather station to WADC is located at Ronald Reagan
Washington National Airport (WBAN 13743). Additional information about the National
Airport weather station, such as the distance between the site and the weather station, is provided
in Table 8-3. These data were used to determine how meteorological conditions on sample days
vary from conditions experienced throughout the year.
                                           8-6

-------
                      Table 8-3. Average Meteorological Conditions near the Washington, D.C. Monitoring Site
Closest NWS
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(ฐF)
Average
Temperature
(ฐF)
Average
Dew Point
Temperature
(ฐF)
Average
Wet Bulb
Temperature
(ฐF)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Washington, D.C. - WADC
Ronald Reagan
Washington
National Airport
13743
(38.87, -77.03)
4.07
miles
183ฐ
(S)
Sample
Day
2011
66.4
ฑ4.3
67.7
ฑ1.9
58.7
ฑ4.0
59.6
ฑ1.8
44.9
ฑ4.5
46.4
ฑ1.9
51.9
ฑ3.7
52.9
ฑ1.6
63.5
ฑ3.7
64.9
ฑ1.5
1017.0
ฑ1.7
1016.5
ฑ0.7
7.5
ฑ0.8
7.0
ฑ0.3
        Sample day averages are highlighted to help differentiate the sample day averages from the full-year averages.
oo

-------
       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 the entire year for 2011. Also included in Table 8-3 is the
95 percent confidence interval for each parameter. As shown in Table 8-3, average
meteorological conditions on sample days were generally representative of average weather
conditions throughout the year. The largest differences were calculated for dew point
temperature and relative humidity, although the differences were slight. These differences may
result from make-up samples collected in November and December 2011.

8.2.3   Back Trajectory Analysis
       Figure 8-3 is the composite back trajectory map for days on which samples were
collected at the WADC monitoring site in 2011. Included in Figure 8-3 are  four back trajectories
per sample day. Figure 8-4 is the corresponding cluster analysis. An in-depth description of these
maps  and how they were generated is presented in Section 3.5.2.1. For the composite map, each
line represents the 24-hour trajectory along which a parcel of air traveled toward the monitoring
site on a given sample day and time, based on an initial height of 50 meters AGL. For the cluster
analysis, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric circle around the site in Figures 8-3 and 8-4 represents  100 miles.

-------
   Figure 8-3. 2011 Composite Back Trajectory Map for WADC
y
       Figure 8-4. Back Trajectory Cluster Map for WADC
                              8-9

-------
       Observations from Figures 8-3 and 8-4 include the following:
       •  Back trajectories originated from a variety of directions at WADC. The most
          common direction of trajectory origin appears to be from the northwest.
       •  The 24-hour air shed domain for WADC was comparable in size to many other NMP
          monitoring sites. The farthest away a back trajectory originated was over the Upper
          Peninsula of Michigan, or nearly than 700 miles away. However, the average
          trajectory length was 239 miles and 88 percent of back trajectories originated within
          400 miles of the site.
       •  The cluster analysis  confirms that back trajectories originating from the northwest
          were  most common (26 percent). The cluster trajectory originating from the west of
          WADC (16 percent) represents back trajectories originating over Virginia and West
          Virginia and generally less than 200 miles in length. Twelve percent of trajectories
          originated to the north of WADC, another 10 percent originated to the south over
          North Carolina, and  another 7 percent originated from the east and offshore. The
          short cluster (28 percent) represents trajectories originating within 200 miles of the
          site and of varying direction, but most often to the south of the site as well as longer
          trajectories originating off the coast of North Carolina and Virginia.

8.2.4  Wind Rose Comparison
       Hourly surface wind data from the NWS weather station at Ronald Reagan Washington
National Airport were uploaded into a wind rose software program to produce customized wind
roses, as described in Section 3.5.2.2. A wind rose shows the frequency of wind directions using
"petals" positioned around a 16-point compass, and uses different colors to represent wind
speeds.
       Figure 8-5 presents a map showing the distance between the NWS station and WADC,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 8-5 also presents three different wind roses for the
WADC monitoring site. First, a historical wind rose representing 2001 to 2010 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 2011 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2011 is
presented. These can be used to identify the predominant wind speed and direction for 2011 and
determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.
                                           8-10

-------
Figure 8-5. Wind Roses for the Ronald Reagan Washington National Airport Weather
                              Station near WADC
Distance between WADC and NWS Station
2001-2010 Historical Wind Rose
                                 4-
                       ......ป.ซ.., •/       I
                                 *  1
          2011 Wind Rose
    Sample Day Wind Rose
                                      8-11

-------
       Observations from Figure 8-5 for WADC include the following:
       •  The NWS weather station at Washington National Airport is located approximately
          4.1 miles to the south of WADC. Note that between WADC and Washington
          National is the city of Washington and the Potomac River.
       •  Historically, southerly to south-south westerly winds account for approximately
          25 percent of wind observations near WADC, while northwesterly to northerly winds
          account for  another 25 percent of observations. Calm winds (< 2 knots) were
          observed for less than 10 percent of the hourly measurements.
       •  The wind patterns on both the full-year and sample day wind roses are similar to the
          wind patterns  shown on the historical wind rose, although southerly winds accounted
          for a  slightly higher percentage of wind observations in 2011. The similarities in the
          wind roses indicate that wind patterns in 2011 were similar to what is expected
          climatologically near this site.

8.3    Pollutants of Interest
       Site-specific "pollutants of interest" were determined for the Washington, D.C.
monitoring site in order to allow 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." 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. In addition, if any of the NATTS MQO Core Analytes measured by the monitoring site
did not meet the pollutant of interest criteria based on the preliminary risk-based screening, that
pollutant was added to  the list of site-specific pollutants of interest. A more in-depth description
of the risk-based screening process is presented in Section 3.2.

       Table 8-4 presents the results of the preliminary risk-based screening process. The
pollutants that failed at least one screen and contributed to 95 percent of the total failed screens
for the WADC monitoring site are shaded. NATTS MQO Core Analytes are bolded. Thus,
pollutants of interest are shaded and/or bolded. WADC sampled for hexavalent chromium and
PAHs.
                                           8-12

-------
    Table 8-4. Risk-Based Screening Results for the Washington, D.C. Monitoring Site
Pollutant
Screening
Value
(Ug/m3)
# of Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
Washington, D.C. - WADC
Naphthalene
0.029
Total
61
61
61
61
100.00
100.00
100.00
100.00

       Observations from Table 8-4 include the following:
       •  Naphthalene is the only pollutant to failed screens for WADC. Naphthalene failed
          100 percent of its screens.
       •  Benzo(a)pyrene and hexavalent chromium were added as pollutants of interest for
          WADC because they are the other NATTS MQO Core Analytes measured by this
          site. These two pollutants are not shown in Table 8-4 but are shown in subsequent
          tables in the sections that follow.
8.4    Concentrations
       This section presents various concentration averages used to characterize pollution levels
at the Washington D.C. monitoring site. Where applicable, the following calculations and data
analyses were performed: Time period-based concentration averages (quarterly and annual) are
provided for the pollutants of interest, where the data meet the applicable criteria. Concentration
averages for select pollutants are also presented graphically to illustrate how the site's
concentrations compare to the program-level averages, as presented in Section 4.1. In addition,
concentration averages for select pollutants are presented from previous years of sampling in
order to characterize concentration trends at the site. Additional site-specific statistical
summaries for WADC are provided in Appendices M and O.

8.4.1   2011 Concentration Averages
       Quarterly and annual concentration averages 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 of 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
                                           8-13

-------
could be calculated and where method completeness was greater than or equal to 85 percent, as

presented in Section 2.4. Quarterly and annual average concentrations for WADC are presented

in Table 8-5, where applicable. Note that if a pollutant was not detected in a given calendar

quarter, the quarterly average simply reflects "0" because only zeros substituted for non-detects

were factored into the quarterly average concentration.
     Table 8-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest
                           for the Washington, D.C. Monitoring Site



Pollutant
#of
Measured
Detections
vs. # of
Samples

1st
Quarter
Average
(ng/m3)

2nd
Quarter
Average
(ng/m3)

3rd
Quarter
Average
(ng/m3)

4th
Quarter
Average
(ng/m3)


Annual
Average
(ng/m3)
Washington, D.C. - WADC

Benzo(a)pyrene

Hexavalent Chromium

Naphthalene

37/61

49/61

61/61
0.07
ฑ0.04
0.02
ฑ0.01
67.83
ฑ17.12
0.02
ฑ0.02
0.02
ฑ0.01
65.20
ฑ 12.60
0.03
ฑ0.02
0.02
ฑ0.01
123.41
ฑ 50.74
0.09
ฑ0.04
0.02
ฑ0.01
137.87
ฑ 46.65
0.06
ฑ0.02
0.02
ฑ0.01
102.71
ฑ 20.46
       Observations for WADC from Table 8-5 include the following:

       •  Naphthalene was detected in every PAH sample collected at WADC while
          benzo(a)pyrene was detected in approximately 60 percent of the PAH samples
          collected. Hexavalent chromium was detected in 80 percent of the samples collected.

       •  The annual average concentration of naphthalene is significantly higher than the
          annual average concentrations of benzo(a)pyrene and hexavalent chromium.

       •  Benzo(a)pyrene concentrations appear higher during the colder months of the year, as
          indicated by the quarterly averages. Of the 14 measurements of benzo(a)pyrene
          greater than or equal to 0.1 ng/m3, eight were measured during the fourth quarter and
          four during the first quarter (and one each in the second and third quarters). In all,
          there were eight measured detections of this pollutant during the first quarter of 2011,
          five during the second quarter,  eight during the third quarter, and  16 during the fourth
          quarter. The number of non-detects was highest during the warmer quarters of the
          year (five, nine, seven, and three, respectively).

       •  Concentrations of hexavalent chromium were relatively consistent across the year.
          The third and fourth quarterly average concentrations of naphthalene are twice as
          high as the first and second quarterly average concentrations. The maximum
          naphthalene concentration was measured at WADC on August 31, 2011 (416 ng/m3).
          Of the 19 naphthalene concentrations greater than 100 ng/m3 measured at WADC, 15
          were measured during the third and fourth quarters; conversely, of the 15 naphthalene
                                          8-14

-------
           concentrations less than 50 ng/m3 measured at WADC, 10 were measured during the
           first and second quarters. However, both relatively high and low naphthalene
           concentrations were measured in each quarter of 2011.

8.4.2  Concentration Comparison
       In order to better illustrate how a site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for the selected NATTS MQO Core
Analytes listed in Section 3.5.3, where applicable. Thus, box  plots for benzo(a)pyrene,
hexavalent chromium, and naphthalene were created for WADC. Figures 8-6 through 8-8
overlay the site's minimum, annual average, and maximum concentrations onto the program-
level minimum, first quartile, median, average, third quartile, and maximum concentrations, as
described in Section 3.5.3.
       Figure 8-6. Program vs. Site-Specific Average Benzo(a)pyrene Concentration
     i
                                          Concentration (ng/m3)
                Program:  IstQuartile
2nd Quartile
3rd Quartile
  D
4th Quartile
                                    Average
                Site:
                        Site Average
Site Minimum/Maximum
                           o
    Figure 8-7. Program vs. Site-Specific Average Hexavalent Chromium Concentration
     I
                                              0.15
                                          Concentration (ng/m3)
                Program:   IstQuartile     2ndQuartile    SrdQuartile     4thQuartile     Average
                Site:      Site Average    Site Minimum/Maximum

                                            8-15

-------
       Figure 8-8. Program vs. Site-Specific Average Naphthalene Concentration
WADC
                                                           Program Max Concentration =779 ng/m3
            50
                   100
                                    200       250       300
                                        Concentration (ng/m3)
                                                            35:
                                                                            45:
                                                                                    5::
              Program:  IstQuartile     2ndQuartile     3rd Quartile     4thQuartile     Average
              Site:
                      Site Average    Site Minimum/Maximum

      Observations from Figures 8-6 through 8-8 include the following:

         •   Figure 8-6 is the box plot for benzo(a)pyrene. Note that the program-level first
             quartile for this pollutant is zero and is not visible on this box plot. The box plot
             shows that the annual average concentration for WADC is less than the program-
             level average concentration but greater than the program-level median
             concentration. Figure 8-6 also shows that the maximum concentration measured
             at WADC is considerably less than the maximum concentration measured across
             the program. There were several non-detects of benzo(a)pyrene measured at
             WADC.

         •   Figure 8-7 is the box plot for hexavalent chromium. Figure 8-7 shows that
             WADC's annual average concentration (0.0167 ng/m3) is less than the program-
             level average (0.0237 ng/m3) and just less than the program-level median
             concentration (0.0178 ng/m3). The maximum concentration measured at WADC
             is less than the program-level maximum concentration. There were several non-
             detects of hexavalent chromium measured at WADC.

         •   Figure 8-8 is the box plot for naphthalene.  Note that the program-level maximum
             concentration (779 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 500 ng/m3.
             Figure 8-8 shows that the annual naphthalene average for WADC is greater than
             the program-level  average concentration and just less than the program-level third
             quartile. The annual average concentration of naphthalene for WADC ranked  12th
             compared to other NMP sites sampling PAHs. The maximum naphthalene
             concentration measured at WADC is less than the program-level maximum
             concentration. The minimum concentration measured at WADC is just less than
             the program-level  first quartile. There  were no non-detects of naphthalene
             measured at WADC or across the  program.
                                         8-16

-------
8.4.3   Concentration Trends
       A site-specific trends evaluation was completed for sites that have sampled one or more
of the selected NATTS MQO Core Analytes for 5 consecutive years or longer, as described in
Section 3.5.4. WADC has sampled hexavalent chromium under the NMP since 2005. Thus,
Figure 8-9 presents the annual statistical metrics for hexavalent chromium for WADC. The
statistical metrics presented for assessing trends include the substitution of zeros for non-detects.
WADC did not begin sampling PAHs until 2008; thus, the trends analysis was not conducted for
the pollutants for these methods.

     Figure 8-9. Annual Statistical Metrics for Hexavalent Chromium Concentrations
                                  Measured at WADC
                    The maximum
                    concentration for
                    2006 is 0.645 ng/m3
                  5th Percentile    - Minimum     —  Median    -  Maximum    • 95th Percentile   • ••*'• Average
       Observations from Figure 8-9 for hexavalent chromium measurements at WADC include
the following:
       •  Sampling for hexavalent chromium began in March 2005 but because fewer than
          85 percent of possible samples were collected, Figure 8-9 excludes 2005 data and
          begins with 2006.
       •  The maximum hexavalent chromium concentration shown was measured on
          July 4, 2006 (0.645 ng/m3).
                                          8-17

-------
       •  The average hexavalent chromium concentration decreased significantly from 2006 to
          2007 and remained steady through 2009. During this time, the median decreased to
          zero for 2008 and 2009, indicating that at least 50 percent of the measurements were
          non-detects. The percentage of non-detects increased from 32 percent in 2006 to a
          maximum of 72 percent in 2009. The substitution of zeros for these non-detects is the
          likely reason for this decrease in these statistical parameters. The number of non-
          detects decreased in 2010 (33 percent) and again in 2011 (20 percent); accordingly,
          with fewer zero substitutions for non-detects, the median and average concentrations
          increased.
       •  The maximum concentrations are roughly the same across the last three years of
          sampling (approximately 0.08 ng/m3). However, the 95th percentile exhibits mon
          fluctuations, indicating more variability among the majority of measurements.
8.5    Additional Risk-Based Screening Evaluations
       The following risk-based screening evaluations were conducted to characterize risk at the
WADC monitoring site. Refer to Sections 3.3 and 3.5.5 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.

8.5.1   Risk-Based Screening Assessment Using MRLs
       A risk-based screening was conducted by comparing the concentration data from the
Washington D.C. monitoring site to the ATSDR MRLs, where available. As described in
Section 3.3, MRLs are noncancer health risk benchmarks and are defined for three exposure
periods: acute (exposures of 1 to 14 days); intermediate (exposures of 15 to 364 days); and
chronic (exposures of 1 year or greater). The preprocessed daily measurements of the pollutants
of interest were compared to the acute MRLs; the quarterly averages were compared to the
intermediate MRLs; and the annual averages were compared to the chronic MRLs.

       As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2011.
                                          8-18

-------
8.5.2   Cancer Risk and Noncancer Hazard Approximations
       For the pollutants of interest for WADC and where annual average concentrations could
be calculated, risk was examined by calculating cancer risk and noncancer hazard
approximations. These approximations can be used as risk estimates for cancer and noncancer
effects attributable to the pollutants of interest. Although the use of these approximations is
limited, they may help identify where policy-makers may want to shift or confirm their air-
monitoring priorities. Refer to Section 3.5.5.2 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 8-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.

         Table 8-6. Risk Approximations for the Washington, D.C. Monitoring Site



Pollutant


Cancer
URE
(Hg/m3)1


Noncancer
RfC
(mg/m3)
Washin

Benzo(a)pyrene

Hexavalent Chromium

Naphthalene

0.00176

0.012

0.000034

—

0.0001

0.003
#of
Measured
Detections
vs. # of
Samples


Annual
Average
(ng/m3)


Cancer Risk
Approximation
(in-a-million)

Noncancer
Hazard
Approximation
(HQ)
gton, D.C. - WADC

37/61

49/61

61/61
0.06
ฑ0.02
0.02
ฑ<0.01
102.71
ฑ 20.46

0.10

0.20

3.49

—

<0.01

0.03
— = a Cancer URE or Noncancer RfC is not available.
       Observations for WADC from Table 8-6 include the following:
          •   As discussed in Section 8.4.1, naphthalene's annual average concentration is four
              orders of magnitude higher than the annual average concentrations for the other
              two pollutants of interest.
          •   Naphthalene's cancer risk approximation is greater than 1.0 in-a-million
              (3.49 in-a-million). Its noncancer hazard approximation is significantly less than
              1.0 (0.03), indicating no adverse health effects are expected from this pollutant.
          •   The cancer risk approximation for benzo(a)pyrene is considerably less than the
              cancer risk approximation for naphthalene (0.10 in-a-million). A noncancer RfC
              is not available for benzo(a)pyrene; thus, a noncancer hazard approximation could
              not be calculated.
                                          8-19

-------
          •   The cancer surrogate risk approximation based on hexavalent chromium's annual
              average concentration is less than 1.0 in-a-million (0.20 in-a-million). The
              noncancer hazard approximation is also low (<0.01).

8.5.3   Risk-Based Emissions Assessment
       In addition to the risk-based screenings discussed above, Tables 8-7 and 8-8 present an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 8-7 presents the 10 pollutants with the highest emissions from the 2008 NEI, the 10
pollutants with the highest toxicity-weighted emissions, and the 10 pollutants with the highest
cancer risk approximations (in-a-million), as calculated from the annual averages provided in
Table 8-6. Table 8-8 presents similar information, but identifies the 10 pollutants with the
highest noncancer hazard approximations (HQ), also calculated from annual averages provided
in Table 8-6.

       The pollutants listed in Tables 8-7 and 8-8 are limited to those that have cancer and
noncancer toxicity factors, respectively. As a result, although the actual value of the emissions is
the same, the highest emitted pollutants in the cancer table may be different from the noncancer
table. Further, cancer risk and noncancer hazard approximations based on each site's annual
averages are limited to those pollutants for which each respective site sampled. As discussed in
Section 8.3, WADC sampled for PAHs and hexavalent chromium. In addition,  the cancer risk
and noncancer hazard approximations are limited to those pollutants with enough data to meet
the criteria for annual averages to be calculated. A more in-depth discussion of this analysis is
provided in Section 3.5.5.3. Similar to the cancer risk and noncancer hazard approximations, this
analysis may help policy-makers prioritize their air monitoring activities.
                                           8-20

-------
   Table 8-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs
                                             the Washington, D.C. Monitoring Site
for
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity- Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Washington, D.C. - WADC
Benzene
Formaldehyde
Acetaldehyde
Ethylbenzene
1,3 -Butadiene
Naphthalene
Dichloromethane
POM, Group 2b
POM, Group la
POM, Group 6
149.75
139.99
83.08
76.80
24.93
14.40
4.38
3.56
0.39
0.35
Formaldehyde
Benzene
1,3 -Butadiene
POM, Group 3
Naphthalene
POM, Group 2b
Hexavalent Chromium, PM
Ethylbenzene
Acetaldehyde
POM, Group 5a
1.82E-03
1.17E-03
7.48E-04
6.38E-04
4.90E-04
3.13E-04
2.01E-04
1.92E-04
1.83E-04
1.32E-04
Naphthalene
Hexavalent Chromium
Benzo(a)pyrene
3.49
0.20
0.10

oo
to

-------
   Table 8-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer

                                         RfCs for the Washington, D.C. Monitoring Site
oo

to
to
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity- Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Noncancer
Hazard
Approximation
Pollutant (HQ)
Washington, D.C. - WADC
Toluene
Methanol
Xylenes
Benzene
Formaldehyde
Hexane
Acetaldehyde
Ethylbenzene
Ethylene glycol
1,3 -Butadiene
433.60
342.72
309.21
149.75
139.99
93.23
83.08
76.80
36.01
24.93
Acrolein
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Benzene
Naphthalene
Xylenes
Nickel, PM
Arsenic, PM
Propionaldehyde
361,586.14
14,284.26
12,467.17
9,231.05
4,991.64
4,799.27
3,092.12
1,910.51
1,536.26
966.01
Naphthalene 0.03
Hexavalent Chromium O.01


-------
Observations from Table 8-7 include the following:

•  Benzene and formaldehyde are the highest emitted pollutants with cancer UREs in the
   District of Columbia. Formaldehyde and benzene are the pollutants with the highest
   toxi city-weighted emissions (of the pollutants with cancer UREs).

•  Seven of the highest emitted pollutants also have the highest toxi city-weighted
   emissions.

•  Naphthalene is the only pollutant sampled for at WADC that appears on both
   emissions-based lists. Naphthalene is the sixth highest emitted pollutant with a cancer
   URE in the District of Columbia and has the fifth highest  toxicity-weighted emissions
   (of the pollutants with cancer UREs).

•  While hexavalent chromium is not one of the 10 highest emitted pollutants in the
   District, its toxi city-weighted emissions ranked seventh highest (of the pollutants with
   cancer UREs).

•  Several POM Groups are among the highest emitted "pollutants" in the District
   and/or rank  among the highest toxicity-weighted emissions. POM, Group 5a, which
   includes benzo(a)pyrene, appears on both emissions-based lists for the District. POM,
   Group 2b includes several PAHs sampled for at WADC including acenaphthylene,
   fluoranthene, fluorene, and perylene. POM, Group 6 includes benzo(a)anthracene and
   indeno(l,2,3-cd)pyrene. None of the PAHs included in POM, Groups 2b or 6 were
   identified as pollutants of interest for WADC.


Observations from Table 8-8 include the following:

•  Toluene, methanol, and xylenes are the highest emitted pollutants with noncancer
   RfCs in the District of Columbia.

•  The pollutants with the highest toxicity-weighted emissions (of the pollutants with
   noncancer RfCs) are acrolein, formaldehyde, and 1,3-butadiene.

•  Five of the highest emitted pollutants in the District of Columbia also have the
   highest toxicity-weighted emissions.

•  Naphthalene has the highest noncancer hazard approximation for WADC (albeit low).
   Naphthalene has the sixth highest toxicity-weighted emissions but is not one of the 10
   highest emitted pollutants (of the pollutants with noncancer RfCs).

•  Hexavalent  chromium, the only other pollutant of interest for which a noncancer RfC
   is available, does not appear on either emissions-based list.
                                   8-23

-------
8.6    Summary of the 2011 Monitoring Data for WADC

       Results from several of the data treatments described in this section include the
following:

       ปซป  Naphthalene was the only pollutant to fail screens for WADC. While naphthalene was
          the only pollutant of interest identified via the risk screening process, benzo(a)pyrene
          and hexavalent chromium were added to WADC"s pollutants  of interest because they
          are NATTSMQO Core Analytes.

       *ป*  Of the site-specific pollutants of the interest, naphthalene had the highest annual
          average concentration for WADC. Naphthalene concentrations were highest during
          the secondhalfoftheyear.

       ปซป  The number ofnon-detects of hexavalent chromium measured at WADC has been
          decreasing over recent years of sampling.
                                          8-24

-------
9.0    Sites in Florida
       This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS and UATMP sites in Florida, and integrates these
concentrations with emissions, meteorological, and risk information. Data generated by sources
other than ERG are not included in the data analyses contained in this report. Readers are
encouraged to refer to Sections 1 through 4 and the glossary (Appendix P) for detailed
discussions and definitions regarding the various data analyses presented below.

9.1    Site Characterization
       This section characterizes the Florida monitoring sites by providing geographical and
physical information about the location of the sites and the surrounding areas. This information
is provided to give the reader insight regarding factors that may influence the air quality near the
sites and assist in the interpretation of the ambient monitoring measurements.

       The five Florida sites are located in two different urban areas. Three sites (AZFL,  SKFL,
and SYFL) are located in the Tampa-St. Petersburg-Clearwater, FL MSA. ORFL and PAFL are
located in the Orlando-Kissimmee, FL MSA. Figures 9-1 through 9-3 are composite satellite
images retrieved from ArcGIS Explorer showing the Tampa/St. Petersburg monitoring sites in
their urban and rural locations. Figure 9-4 identifies nearby point source emissions locations that
surround these three sites by source category, as reported in the 2008 NEI for point sources. Note
that only sources within 10 miles of the sites are included in the facility counts provided in
Figure 9-4. A 10-mile boundary was chosen to give the reader an indication of which emissions
sources and emissions source categories could potentially have a direct effect on the air quality at
the monitoring sites. Further, this boundary provides both the proximity of emissions sources to
the monitoring sites as well as the quantity of such sources within a given distance of the sites.
Sources outside the 10-mile radii are still visible on the map, but have been grayed out in order
to  show emissions  sources just outside the boundary. Figures 9-5 and 9-6 are the composite
satellite images for the two sites in the Orlando area and Figure 9-7 is the emissions sources map
for these sites. Table 9-1 provides supplemental geographical information such as land use,
location setting, and locational coordinates.
                                           9-1

-------
                                     Figure 9-1. St. Petersburg, Florida (AZFL) Monitoring Site
to

-------
Figure 9-2. Pinellas Park, Florida (SKFL) Monitoring Site

-------
Figure 9-3. Plant City, Florida (SYFL) Monitoring Site

-------
        Figure  9-4. NEI Point Sources Located Within  10 Miles of the
                  Tampa/St.  Petersburg, Florida Monitoring  Sites
            82'45'trW   82'40'CrW   82'35'0'W   82'30tTW    B2'25'0'W
      82 55XTW   82'50'0-W   B2'45T)-W
Legend
      AZFLUATMPsite
                                              82 35XJ-W   a2'3(K™   82'25' Air-conditioning/Refrigeration (1)
 4" Aircraft Operations (16)
 I  Asphalt Processing/Roofing Manufacturing (1)
 0  Auto Body Shop/Painters (2)
 K Automobile/Truck Manufacturing (2)
 6  Bakery (1)
 i Boat Manufacturing (4)
 ~ Brick Manufacturings Structural Clay (1)
 •  Building Construction (1)
 C  Chemical Manufacturing (2)
 •  Concrete Batch Plant (1)
 f  Electricity Generation via Combustion (2)
 E  Electroplating, Plating, Polishing, Anodizing, and Coloring (1)
 ฉ Fabricated Metal Products (10)
 F  Food Processing/Agriculture (6)
 •  Gasoline/Diesel Service  Station (1)
Hospital (1)
Hot Mix A$phalt Plant (2)
Industrial Machinery and Equipment (1)
Institutional - school (2)
Miscellaneous Commerciain'ndustrial (3)
Miscellaneous Manufacturing (3)
Municipal Waste Cornbustor (2)
Oil and/or Gas Production (1)
Pharmaceutical  Manufacturing (1)
Printing/Publishing (12)
Pulp and Paper  Plant/Wood Products (2)
Rubber and Miscellaneous Plaslics Products (3)
Secondary Melal Processing (3)
Solid Waste Disposal - Commercial/Institutional (1)
Surface Coating (4)
Telecommunications (2)
Transportation Equipment (1}
Wastewater Treatment (2)
                                                    9-5

-------
Figure 9-5. Winter Park, Florida (ORFL) Monitoring Site

-------
Figure 9-6. Orlando, Florida (PAFL) Monitoring Site

-------
Figure 9-7. NEI Point Sources Located Within 10 Miles of ORFL and PAFL
Legend
                                                Note: Due to facility density and collocation, the total facilities
                                                displayed may not represent ail facilities within the area of interest
      ORFL UATMP site  ฎ  PAFL UATMP site        1 0 mile radius f   1 County boundary
Source Category Group (No. of Facilities)
 -f  Aircraft Operations (23)
  I   Asphalt Processing/Roofing Manufacturing (1)
  0   Auto Body Shop/Painters (1)
 ft  Automobile/Truck Manufacturing (1)
  6   Bakery (3)
 ฑ  Boat Manufacturing (1)
     Brick Manufacturing & Structural Clay (1)
 A  Cement Kiln/Dryer (1)
  C   Chemical Manufacturing (1)
 fl>  Dry Cleaning (1)
  6   Electrical Equipment (2)
  *   Electricity Generation via Combustion (1)
  E   Electroplating, Plating, Polishing, Anodizing, and Coloring (1)
 ฉ  Fabricated Metal Products (2)
  F   Food Processing/Agriculture (2)
                                                          B  Furniture Plant (2)
                                                          •   Gasoline/Diesel Service Station (1)
                                                           :   Hospital (2)
                                                          $  Hot Mix Asphalt Plant (4)
                                                          -$•  Industrial Machinery and Equipment (2)
                                                          ^t  Institutional - school (3)
                                                          V  Mineral Products (1)
                                                          ?  Miscellaneous Commercial/Industrial (5)
                                                          M  Miscellaneous Manufacturing (2)
                                                          •  Oil and/or Gas Production (1}
                                                          P  Printing/Publishing (2)
                                                          H  Pulp and Paper Plant/Wood Products (3)
                                                          R  Rubber and Miscellaneous Plastics Products (3)
                                                          S  Surface Coating (5)
                                                          rr  Telecommunications (1)
                                                 9-8

-------
                                Table 9-1. Geographical Information for the Florida Monitoring Sites
Site
Code
AZFL
SKFL
SYFL
ORFL
PAFL
AQS Code
12-103-0018
12-103-0026
12-057-3002
12-095-2002
12-095-1004
Location
St.
Petersburg
Pinellas
Park
Plant City
Winter
Park
Orlando
County
Pinellas
Pinellas
Hillsborough
Orange
Orange
Micro- or
Metropolitan
Statistical Area
Tampa-St.
Petersburg-
Clearwater, FL
Tampa-St.
Petersburg-
Clearwater, FL
Tampa-St.
Petersburg-
Clearwater, FL
Orlando-
Kissimmee, FL
Orlando-
Kissimmee, FL
Latitude
and
Longitude
27.785556,
-82.74
27.850041,
-82.714590
27.96565,
-82.2304
28.596444,
-81.362444
28.550833,
-81.345556
Land Use
Residential
Residential
Residential
Commercial
Commercial
Location
Setting
Suburban
Suburban
Rural
Urban/City
Center
Suburban
Additional Ambient Monitoring Information1
NO, NO2, NOX, VOCs, O3, Meteorological
parameters, PM10, PM10 Speciation, PM25.
VOCs, Meteorological parameters, PM10 Speciation,
Black carbon, PM25 Speciation, PM25.
CO, SO2, NOy, NO, NO2, NOX, VOCs, O3,
Meteorological parameters, PM10, PM10 Speciation,
PM2 5, PM2 5 Speciation, PM Coarse.
CO, SO2, NO, NO2, NOX, VOCs, O3, Meteorological
parameters, PM10, PM25.
Meteorological parameters, PM10, PM25.

VO
  BOLD ITALICS = EPA-designated NATTS Site

-------
       AZFL is located at Azalea Park in St. Petersburg. Figure 9-1 shows that the area
surrounding AZFL consists of mixed land use, including residential, commercial, and industrial
properties. Heavily traveled roadways are located less than 1 mile from the monitoring site.
AZFL is located just over 1 mile east of Boca Ciega Bay.

       SKFL is located in Pinellas Park,  north of St. Petersburg. This site is on the property of
Skyview Elementary School near 86th Avenue North. Figure 9-2 shows that SKFL is located in a
primarily residential area. However, a railroad intersects the Pinellas Park Ditch near a
construction company in the bottom left corner of Figure 9-2. Population exposure is the purpose
behind monitoring at this location. This site is the Pinellas County NATTS site.

       SYFL is located in Plant City, which is also part of the Tampa-St. Petersburg-Clearwater,
FL MSA, although it is on the eastern outskirts of the area. Unlike the other Florida sites, the
SYFL monitoring site is in a rural area, although, as Figure 9-3 shows, a residential community
and country club lie just to the west of the site. Located to the south of the site is a tank that is
part of the local water treatment facility. This site serves as a background site, although the effect
of increased development in the area is likely being  captured by the monitoring site. This site is
the Tampa NATTS site.

       Figure 9-4 shows the location of the Tampa/St. Petersburg sites in relation to each other.
SYFL is located the farthest east and AZFL is the farthest west, although SKFL is located within
a few miles of AZFL.  A large cluster of point sources is located just north of SKFL. Another
cluster of emissions sources is located about halfway between SYFL and the other two sites,
although grayed out and not included in the facility counts in Figure 9-4. Aircraft operations,
which include airports as well as small runways, heliports, or landing pads; printing and
publishing facilities; and fabricated metal product facilities are the source categories with the
highest number of emissions sources in the Tampa/St. Petersburg area (based on the areas
covered by the 10-mile radii).

       ORFL is located in Winter Park, north of Orlando. Figure 9-5 shows that ORFL is
located near Lake Mendsen, east of Lake Killarney and south of Winter Park Village. This site
lies in a commercial area and serves as a population exposure monitor.
                                          9-10

-------
       PAFL is located in northern Orlando, on the northwestern edge of the Orlando Executive
Airport property, as shown in Figure 9-6. The area is considered commercial and experiences
heavy traffic. The airport is bordered by Colonial Drive to the north and the East-West
Expressway (Toll Road 408) to the south (although not shown in Figure 9-6). A large shopping
complex is located to the northeast of the site, just north of the airport, between Colonial Drive
and Maguire Boulevard. Interstate-4 runs north-south less than 2 miles to the west of the
monitoring site.

       Figure 9-7 shows that ORFL is located a few miles north of PAFL. Most of the point
sources are located on the western side of the 10-mile radii. Although the emissions sources
surrounding ORFL and PAFL are involved in a variety of industries and processes, the aircraft
operations source category has the highest number of emissions sources within 10 miles of these
sites.

       Table 9-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Florida monitoring sites.  Table 9-2 includes county-level
population and vehicle registration information.  Table 9-2 also includes a county-level vehicle
registration-to-population ratio, which was calculated to represent the number of vehicles per
person within each monitoring site's residing county. In addition, the population within 10 miles
of each site is presented, based on postal code population data estimates. An estimate of 10-mile
vehicle ownership was determined by applying the county-level vehicle registration-to-
population ratio to the 10-mile population surrounding each monitoring site. Table 9-2 also
contains traffic volume information for each site.  Finally, Table 9-2 presents the county-level
daily VMT for Pinellas, Hillsborough, and Orange Counties.
                                          9-11

-------
Table 9-2. Population, Motor Vehicle, and Traffic Information for the Florida Monitoring
                                         Sites
Site
AZFL
SKFL
SYFL
ORFL
PAFL
Estimated
County
Population1
917,398
1,267,775
1,169,107
County-level
Vehicle
Registration2
877,075
1,135,945
1,056,627
Vehicles per
Person
(Registration:
Population)
0.96
0.90
0.90
Population
within 10
miles3
580,599
705,597
321,686
1,003,806
880,133
Estimated
10-mile
Vehicle
Ownership
555,080
674,583
288,235
907,230
795,455
Annual
Average
Daily
Traffic4
40,500
47,000
10,600
32,500
46,000
County-
level
Daily VMT5
21,395,381
34,351,899
33,325,315
  Bounty-level population estimates reflect 2011 data from the U.S. Census Bureau (Census Bureau, 2012b)
  2County-level vehicle registration reflects 2011 data from the Florida Department of Highway Safety & Motor
   Vehicles (FL DHSMV, 2011)
  310-mile population estimates reflect 2011 data from Xionetic (Xionetic, 2011)
  4AADT reflects 2011 data from the Florida DOT (FL DOT, 201 la)
  5County-level VMT reflects 2011 data for all public roads from the Florida DOT (FL DOT, 201 Ib)
  BOLD ITALICS = EPA-designated NATTS Site
       Observations from Table 9-2 include the following:

      •  Hillsborough County, where SYFL is located, is the most populous of the Florida
         counties with monitoring sites, although Orange County also has over 1 million
         people. Broward County ranks just less than Orange County compared to other
         counties with NMP sites covered in this report.

      •  Of the five Florida monitoring sites, ORFL has the highest population within 10 miles
         of all the Florida sites. ORFL's 10-mile population ranks 12th highest among NMP
         sites. Note the difference between SYFL's 10-mile and county-level populations. This
         is an example of a site located within a populous county that is not near the
         population center.

      •  The vehicle registration counts for two of the three Florida counties are over
         1 million, with Hillsborough County having the most and Pinellas County having the
         least. The 10-mile ownership estimates are more variable, with SYFL having the least
         number of vehicles  and ORFL having the most.

      •  The vehicle-per-person ratios range from 0.90 (both Orange and Hillsborough
         Counties) to 0.96 (Pinellas County).

      •  The traffic volume near SYFL is the lowest among the Florida sites and highest near
         SKFL. Traffic volumes near most of the Florida monitoring sites are in the middle of
         the range compared to other NMP sites, with traffic near SYFL being in the bottom
         third compared to other NMP sites. The following list provides the roadways or
         intersections from which the traffic data were obtained: AZFL - 66th StreetNorth,
         north of 9th Street; ORFL - Orlando Avenue, north of Morse Boulevard; PAFL - East
         Colonial Drive, between Primrose Road and Bumby Avenue;  SKFL - Park
         Boulevard, east of 66th Street North; and SYFL - Martin Luther King Jr. Boulevard
         (574), east of Mclntosh Road.
                                         9-12

-------
       •  VMT is highest for Hillsborough County and lowest for Pinellas County (among the
          Florida sites). The Hillsborough, Orange, and Pinellas County VMTs ranked eighth,
          ninth, and 14th highest among counties with NMP sites, respectively.

9.2    Meteorological Characterization
       The following sections characterize the meteorological conditions near the monitoring
sites in Florida on sample days, as well as over the course of the year.

9.2.1  Climate Summary
       The Tampa and Orlando areas experience very mild winters and warm, humid summers.
Precipitation tends to be concentrated during the summer, as afternoon thunderstorms occur
frequently. Semi-permanent high pressure offshore over the Atlantic Ocean extends westward
towards Florida in the winter, resulting in reduced precipitation amounts. Land and sea breezes
affect coastal locations and the proximity to the Atlantic Ocean or Gulf of Mexico can have a
marked affect on the local meteorological conditions.  Florida's orientation and location between
the warm waters of the Gulf of Mexico, the Atlantic Ocean, and Caribbean Sea make it
susceptible to tropical systems (Bair, 1992 and FCC, 2013).

9.2.2  Meteorological Conditions in 2011
       Hourly meteorological data from NWS weather stations nearest these sites were retrieved
for 2011 (NCDC, 2011). These data were used to determine  how meteorological conditions on
sample days vary from conditions experienced throughout the year. The weather station closest
to the AZFL monitoring site is located at St. Petersburg/Whitted Airport (WBAN 92806); closest
to SYFL is at Plant City Municipal Airport (WBAN 92824); closest to SKFL is at
St. Petersburg/Clearwater International Airport (WBAN 12873); and closest to ORFL and PAFL
is at Orlando Executive Airport (WBAN 12841).  Additional information about each of these
weather stations, such as the distance between the sites and the weather stations, is provided in
Table 9-3.
                                          9-13

-------
                               Table 9-3. Average Meteorological Conditions near the Florida Monitoring Sites
Closest NWS Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(ฐF)
Average
Temperature
(ฐF)
Average
Dew Point
Temperature
(ฐF)
Average
Wet Bulb
Temperature
(ฐF)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar
Wind Speed
(kt)
St. Petersburg, Florida - AZFL
St. Petersburg/
Whirled Airport
92806
(27.77, -82.63)
6.77
miles
94ฐ
(E)
Sample
Day
2011
79.6
ฑ2.1
80.4
ฑ0.9
73.6
ฑ2.3
74.2
ฑ0.9
65.1
ฑ2.6
65.4
ฑ1.0
68.3
ฑ2.3
68.7
ฑ0.9
76.1
ฑ2.5
75.4
ฑ1.0
1016.7
ฑ 1.0
1016.5
ฑ0.4
7.6
ฑ0.8
7.3
ฑ0.3
Pinellas Park, Florida - SKFL
St Petersburg-
Clearwater Intl.
Airport
12873
(27.91, -82.69)
4.46
miles
12ฐ
(NNE)
Sample
Day
2011
80.8
ฑ2.2
81.6
ฑ0.9
72.9
ฑ2.3
73.4
ฑ0.9
63.0
ฑ2.6
63.3
ฑ1.0
66.8
ฑ2.2
67.1
ฑ0.9
72.8
ฑ2.2
72.4
ฑ0.9
1017.0
ฑ1.0
1017.0
ฑ0.4
6.6
ฑ0.6
6.5
ฑ0.2
Plant City, Florida - SYFL
Plant City
Municipal Airport
92824
(28.00, -82.16)
4.56
miles
50ฐ
(NE)
Sample
Day
2011
83.2
ฑ2.5
84.7
ฑ0.9
73.0
ฑ2.5
73.9
ฑ0.9
62.9
ฑ2.8
63.5
ฑ1.0
66.8
ฑ2.4
67.4
ฑ0.9
73.9
ฑ2.4
73.5
ฑ1.0
NA
NA
4.4
ฑ0.4
4.2
ฑ0.2
Winter Park, Florida - ORFL
Orlando Executive
Airport
12841
(28.55, -81.33)
3.95
miles
145ฐ
(SE)
Sample
Day
2011
82.2
ฑ2.4
82.8
ฑ0.9
72.9
ฑ2.3
73.0
ฑ0.9
62.4
ฑ2.5
62.1
ฑ1.0
66.4
ฑ2.2
66.3
ฑ0.9
72.1
ฑ2.2
71.3
ฑ1.0
1017.1
ฑ1.1
1017.4
ฑ0.4
6.1
ฑ0.6
5.8
ฑ0.2
VO
           Sample day averages are highli;
           NA= Sea level pressure was not
jhted to help differentiate the sample day averages from the full-year averages.
recorded at the Plant City Municipal Airport.

-------
             Table 9-3. Average Meteorological Conditions near the Florida Monitoring Sites (Continued)
Closest NWS Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(ฐF)
Average
Temperature
(ฐF)
Average
Dew Point
Temperature
(ฐF)
Average
Wet Bulb
Temperature
(ฐF)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Orlando, Florida - PAFL
Orlando Executive
Airport
12841
(28.55, -81.33)
0.84
miles
111ฐ
(ESE)
Sample
Day
2011
81.1
ฑ3.7
82.8
ฑ0.9
71.9
ฑ3.6
73.0
ฑ0.9
61.2
ฑ3.9
62.1
ฑ1.0
65.4
ฑ3.5
66.3
ฑ0.9
71.4
ฑ3.1
71.3
ฑ1.0
1017.9
ฑ1.5
1017.4
ฑ0.4
5.8
ฑ0.9
5.8
ฑ0.2
Sample day averages are highlighted to help differentiate the sample day averages from the full-year averages.
NA= Sea level pressure was not recorded at the Plant City Municipal Airport

-------
       Table 9-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for the entire year for 2011. Also included in Table 9-3 is the
95 percent confidence interval for each parameter. As shown in Table 9-3, average
meteorological conditions on sample days in 2011 at the Florida monitoring sites were
representative of average weather conditions experienced throughout the entire year. The largest
difference is shown for PAFL and the temperature parameters. Sampling at PAFL took place on
a l-in-12 day schedule, yielding roughly half the sample days as the other Florida monitoring
sites. This may result in more variability in the sample day averages. Temperatures on sample
days at SYFL also appear slightly cooler than those for the entire year.  SYFL did not deviate
from the l-in-6 day sample schedule until then end of the year, where make-up samples were
collected at SYFL in December, which may explain the differences shown in Table 9-3.

9.2.3   Back Trajectory Analysis
       Figure 9-8 is the composite back trajectory map for days on which samples were
collected at the AZFL monitoring site in 2011. Included in Figure 9-8 are four back trajectories
per sample day. Figure 9-9 is the corresponding cluster analysis. Similarly, Figures 9-10 through
9-17 are the composite back trajectory maps and corresponding cluster analyses for the
remaining Florida monitoring sites. An in-depth description  of these maps and how they
generated is presented in Section 3.5.2.1. For the composite  maps, each line represents the
24-hour trajectory along which a parcel of air  traveled toward the monitoring site on a given
sample day and time, based on an initial height of 50 meters AGL.  For the cluster analyses, each
line corresponds to a trajectory representative  of a given cluster of back trajectories. Each
concentric circle  around the sites in Figures 9-8 through 9-17 represents 100 miles.
r were
                                          9-16

-------
Figure 9-8. 2011 Composite Back Trajectory Map for AZFL
     Figure 9-9. Back Trajectory Cluster Map for AZFL
                          9-17

-------
Figure 9-10. 2011 Composite Back Trajectory Map for SKFL
    Figure 9-11. Back Trajectory Cluster Map for SKFL
                         9-18

-------
Figure 9-12. 2011 Composite Back Trajectory Map for SYFL
    Figure 9-13. Back Trajectory Cluster Map for SYFL
                         9-19

-------
Figure 9-14. 2011 Composite Back Trajectory Map for ORFL
    Figure 9-15. Back Trajectory Cluster Map for ORFL
                         9-20

-------
Figure 9-16. 2011 Composite Back Trajectory Map for PAFL
    Figure 9-17. Back Trajectory Cluster Map for PAFL
                                    '' ~A /
                                           '    - \
                                        ?     /  s -*
                         9-21

-------
       Observations from Figures 9-8 through 9-13 for the Tampa/St. Petersburg sites include
the following:

       •  The composite back trajectory maps for the Tampa/St. Petersburg sites are generally
          similar to each other in trajectory distribution, which is not unexpected given their
          close proximity to each other. Back trajectories originated from a variety of directions
          at the Tampa/St. Petersburg sites.

       •  The 24-hour air shed domains for these sites were comparable in size to other NMP
          monitoring sites, with the average trajectory length ranging from 215 miles for AZFL
          to 222 miles for SYFL. For AZFL and SKFL, the farthest away a back trajectory
          originated was just greater than 500 miles away, originating eastward over the
          Atlantic Ocean. For SYFL, the longest back trajectory originated over central North
          Carolina, nearly 550 miles away. However, most trajectories (roughly 88 percent for
          each site) originated within 400 miles of the Tampa/St. Petersburg monitoring sites.

       •  The cluster maps for AZFL, SKFL, and SYFL are similar to each other in
          geographical breakup, although the percentages differ somewhat.  The  cluster maps
          for all three sites show that the approximately one-third or more of the back
          trajectories are represented by the short cluster originating just west of the Tampa/St.
          Petersburg area and over the Gulf of Mexico. This cluster includes back trajectories
          of varying lengths originating to the west of the sites over the Gulf of Mexico as well
          as shorter trajectories originating from a variety of directions around the sites.

       •  The cluster maps group the remaining back trajectories into four directions:
          northwestward over the Florida Panhandle, northeastward off the Southeast Coast,
          eastward over the Atlantic Ocean, and  southward over south Florida and the Straights
          of Florida.


        Observations from Figures 9-14 through 9-17 for ORFL and PAFL include the

following:

       •  The composite back trajectory map for PAFL has fewer trajectories compared to the
          map for ORFL. This is because sampling at PAFL occurred on a l-in-12 day
          schedule, yielding approximately half the sample days as ORFL.

       •  The 24-hour air shed domains were similar in size compared to the Tampa/
          St. Petersburg monitoring sites. The longest trajectory originated northeastward over
          the Atlantic Ocean for PAFL, or 560 miles away, with a few additional back
          trajectories originating from a similar location. The longest back trajectory for ORFL
          originated in southeast Virginia, nearly 650 miles away.  PAFL does not have a
          similar back trajectory because this site did not sample on this date.

       •  Nearly 90 percent of back trajectories originated with 400 miles of ORFL and PAFL.

       •  The cluster map for ORFL is similar to the cluster maps  for the Tampa/St. Petersburg
          sites in trajectory distribution. Nearly half of all back trajectories are represented by
          the short cluster originating to the southwest of the site (44 percent). This cluster

                                          9-22

-------
          includes back trajectories originating from a variety of directions, although primarily
          from the western quadrants, and less than 200 miles away. The cluster map groups
          the remaining back trajectories into four directions: those originating northwestward
          over the Florida Panhandle, Georgia, and Alabama; those originating northeastward
          off the Southeast Coast; those originating eastward over the Atlantic Ocean and
          northern Bahamas; and those originating southward over south Florida and the
          surrounding waters.
       •  The cluster map for PAFL has a short cluster (53 percent) similar to the one for
          ORFL, but this cluster also includes those back trajectories originating from south
          Florida (while the cluster map for ORFL separates them into a separate cluster).
       •  The cluster map for PAFL groups the remaining back trajectories as follows:
          northwestward over the Florida Panhandle, Georgia, and Alabama; northeastward off
          the Southeast Coast, and east-southeastward over the Atlantic Ocean.

9.2.4   Wind Rose Comparison
       Hourly surface wind data from the NWS weather stations nearest the Florida sites, as
presented in Section 9.2.2, were  uploaded into a wind rose software program to produce
customized wind roses, as described in Section 3.5.2.2. A wind rose shows the frequency of wind
directions using "petals" positioned around a 16-point compass, and uses different colors to
represent wind speeds.

       Figure 9-18 presents a map showing the distance between the NWS station and AZFL,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 9-18 also presents three different wind roses for the
AZFL monitoring site. First, a historical wind rose representing 2001 to 2010 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 2011 is presented.
Next, a wind rose representing wind data for days on which  samples were collected in 2011 is
presented. These can be used to identify the predominant wind speed and direction for 2011 and
determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically. Figures 9-19 through 9-22 present the three wind roses and
distance maps for SKFL, SYFL, ORFL, and PAFL, respectively.
                                          9-23

-------
Figure 9-18. Wind Roses for the St. Petersburg/Whitted Airport Weather Station near
                                     AZFL
Distance between AZFL and NWS Station
2001-2010 Historical Wind Rose
                 .,.„*..-  ;
   ซ•••-•   s  ;; :"*,","! . :
 "     ''  19* ปv.ป- f * ป
 #      ' "	! •
                                  4-
          2011 Wind Rose
    Sample Day Wind Rose
                                      9-24

-------
Figure 9-19. Wind Roses for the St. Petersburg/Clearwater International Airport Weather
                                Station near SKFL
  Distance between SKFL and NWS Station
2001-2010 Historical Wind Rose
           2011 Wind Rose
    Sample Day Wind Rose
                                      9-25

-------
Figure 9-20. Wind Roses for the Plant City Municipal Airport Weather Station near SYFL
  Distance between SYFL and NWS Station
2008-2010 Historical Wind Rose
           2011 Wind Rose
    Sample Day Wind Rose
                                       9-26

-------
Figure 9-21. Wind Roses for the Orlando Executive Airport Weather Station near ORFL
 Distance between ORFL and NWS Station
2001-2010 Historical Wind Rose
  r ป^ To1 !  tUMIIIOt
       t  cc-sr,1
       ฃ" Mivcinnst
           2011 Wind Rose
    Sample Day Wind Rose
                                                WEST
                                                                               WND SPEED
                                                                               (Knots)
                                                                               n -22
                                                                               ^| 17 - 21
                                                                               ^| 11 • 17
                                                                               IH 7- 11
                                                                               O 4-7
                                                                               IB 2- 4
                                                                               Calms: 14.89%
                                         9-27

-------
Figure 9-22. Wind Roses for the Orlando Executive Airport Weather Station near PAFL
 Distance between PAFL and NWS Station
2001-2010 Historical Wind Rose
          2011 Wind Rose
    Sample Day Wind Rose
                                     9-28

-------
Observations from Figure 9-18 for AZFL include the following:

•  The NWS weather station at St. Petersburg/Whitted Airport is located approximately
   6.8 miles east of AZFL. Between them is most of the city of St. Petersburg. Note that
   the Whitted Airport is 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 calm winds (< 2 knots) accounted for less than
   10 percent of the hourly wind measurements. Winds from the north, northeasterly
   quadrant, and east were the most commonly observed wind directions near AZFL
   while winds from the western quadrants were observed less frequently.

•  The full-year wind rose shows that while winds from all  directions were observed
   near AZFL, winds from the north and east were the predominant wind directions,
   similar to the historical wind rose.

•  The sample day wind patterns favor the full-year wind patterns, indicating that
   conditions on sample days were representative of wind conditions experienced in
   2011.

•  Both the full-year and sample day wind roses resemble the historical wind rose,
   indicating that conditions on sample days and over the entire year were similar to
   wind conditions experienced historically.


Observations from Figure 9-19 for SKFL include the following:

•  The NWS weather station at St. Petersburg/Clearwater Airport is  located just less
   than 4.5 miles north-northeast of SKFL. Note that the St. Petersburg/Clearwater
   Airport is located on Old Tampa Bay while SKFL is farther inland.

•  The historical wind rose shows that winds from a variety of directions were observed
   near SKFL, although winds from the north, northeast quadrant, east, and
   east-southeast were the most commonly observed wind directions. Calm winds
   accounted for approximately 10 percent of the hourly wind measurements.

•  The 2011 and sample day wind roses resemble the historical wind rose, indicating
   that conditions on sample days and over the entire year were similar to wind
   conditions experienced historically.

Observations from Figure 9-20 for SYFL include the following:

•  The NWS weather station at Plant City Municipal Airport is located 4.6 miles
   northeast of SYFL. Note that this weather station has less historical data than the
   other sites. This station did not begin operating until 2006 and data availability is
   lacking until mid-2007; thus, the historical wind rose includes data from the first full-
   year of data (2008) through 2010.
                                   9-29

-------
•  The historical wind rose shows that calm winds (< 2 knots) account for approximately
   25 percent of the hourly wind measurements between 2008 and 2010. Winds from the
   eastern quadrants were observed more often than the western quadrants, although
   winds from all directions were observed near SYFL.

•  Both the full-year and sample day wind patterns are similar to the historical wind
   patterns, indicating that conditions on sample days were representative of wind
   conditions experienced throughout the year and historically.


Observations  from Figures 9-21 and 9-22 for ORFL and PAFL include the following:

•  The closest NWS weather station to both ORFL and PAFL is the Orlando Executive
   Airport. The weather station is located just less than 4 miles southeast of ORFL and
   less than 1 mile east-southeast of PAFL, as PAFL is located on the edge of the
   Orlando Executive Airport property. Thus, the historical and full-year wind roses for
   these sites are identical.

•  The historical wind roses show that winds from all directions were observed near
   these sites, with easterly winds being observed the most, followed by winds from due
   north and  due south.  Winds with an easterly component were observed more often
   than winds with a westerly component. Calm winds were observed for less than 15
   percent of the wind observations.

•  The wind  patterns shown on the full-year wind roses resemble the wind patterns on
   the historical wind roses.

•  The 2011  sample day wind rose for ORFL exhibits the same prominence of easterly,
   northerly,  and southerly winds, but winds from the southwest quadrant (and
   northwest quadrant to some extent) account for a higher percentage of wind
   observations than they do for the historical and full-year wind roses.

•  The 2011  sample day wind rose for PAFL shares the easterly prominence with the
   2011 wind rose, but that is where the similarities in wind direction end. Winds from
   the south and north account for fewer observations and winds from the east-northeast
   account for more. Further, winds from the western quadrants account for more
   observations than those from the eastern quadrants (with the exception of winds from
   the east-northeast and east). Note, however, that PAFL samples on a l-in-12 day
   sampling schedule, leading to roughly half the sample days included in the sample
   day wind rose as ORFL.
                                   9-30

-------
9.3    Pollutants of Interest
       Site-specific "pollutants of interest" were determined for the Florida monitoring sites in
order to allow 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. In
addition, if any of the NATTS MQO Core Analytes measured by each monitoring site did not
meet the pollutant of interest criteria based on the preliminary risk-based screening, that pollutant
was added to the list of site-specific pollutants of interest. A more in-depth description of the
risk-based screening process is presented in Section 3.2.

       Table 9-4 presents the results of the preliminary risk-based screening process for each
Florida monitoring site. The pollutants that failed at least one screen and contributed to 95
percent of the total failed screens for each monitoring site are shaded. NATTS MQO Core
Analytes are bolded. Thus, pollutants of interest are shaded and/or bolded. AZFL and ORFL
sampled for carbonyl compounds only. SKFL and SYFL sampled hexavalent chromium and
PAHs in addition to carbonyl compounds. PAFL sampled only PMio metals.
         Table 9-4. Risk-Based Screening Results for the Florida Monitoring Sites
Pollutant
Screening
Value
(Ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
St. Petersburg, Florida - AZFL
Formaldehyde
Acetaldehyde
0.077
0.45
Total
62
61
123
62
62
124
100.00
98.39
99.19
50.41
49.59
50.41
100.00

Pinellas Park, Florida - SKFL
Acetaldehyde
Formaldehyde
Naphthalene
Hexavalent Chromium
Propionaldehyde
0.45
0.077
0.029
0.000083
0.8
Total
61
61
52
1
1
176
61
61
61
55
61
299
100.00
100.00
85.25
1.82
1.64
58.86
34.66
34.66
29.55
0.57
0.57
34.66
69.32
98.86
99.43
100.00

                                          9-31

-------
Table 9-4. Risk-Based Screening Results for the Florida Monitoring Sites (Continued)
Pollutant
Screening
Value
(Ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
Plant City, Florida - SYFL
Formaldehyde
Acetaldehyde
Naphthalene
Acenaphthene
0.077
0.45
0.029
0.011
Total
60
57
37
2
156
60
60
60
60
240
100.00
95.00
61.67
3.33
65.00
38.46
36.54
23.72
1.28
38.46
75.00
98.72
100.00

Winter Park, Florida - ORFL
Acetaldehyde
Formaldehyde
0.45
0.077
Total
60
60
120
60
60
120
100.00
100.00
100.00
50.00
50.00
50.00
100.00

Orlando, Florida - PAFL
Arsenic (PM10)
0.00023
Total
28
28
31
31
90.32
90.32
100.00
100.00

    Observations from Table 9-4 include the following:

    •  Acetaldehyde and formaldehyde were the only two pollutants to fail screens for
       AZFL and ORFL. Although these two pollutants contributed equally to the total
       number of failed screens for ORFL, there was one more failed screen for
       formaldehyde than acetaldehyde for AZFL. These two sites sampled only carbonyl
       compounds; among the carbonyl compounds, only acetaldehyde, formaldehyde, and
       propionaldehyde have risk screening values. Propionaldehyde did not fail any screens
       for these two sites.

    •  Five pollutants, of which four are NATTS MQO Core Analytes, failed screens for
       SKFL. Acetaldehyde,  formaldehyde, and naphthalene were identified as pollutants of
       interest via the risk-based screening process. Hexavalent chromium was added to
       SKFL's pollutants of interest because it is a NATTS MQO Core Analyte, even
       though it did not contribute to 95 percent of SKFL's failed screens. Benzo(a)pyrene
       was also added as a pollutant of interest for SKFL, even though it did not fail any
       screens, because it is a NATTS MQO Core Analyte. Benzo(a)pyrene is not shown in
       Table 9-4 but is shown in subsequent tables in the sections that follow.

    •  Four pollutants, of which three are NATTS MQO Core Analytes, failed screens for
       SYFL. Acetaldehyde, formaldehyde, and naphthalene were identified as pollutants of
       interest via the risk-based screening process. Benzo(a)pyrene and hexavalent
       chromium were added to SYFL's pollutants of interest because they are NATTS
       MQO Core Analytes, even though their concentrations did not fail any screens. These
       pollutants are not shown in Table 9-4 but are  shown in subsequent tables in the
       sections that follow.
       Formaldehyde failed 100 percent of screens for all four sites sampling carbonyl
       compounds.
                                      9-32

-------
       •  Arsenic was the only speciated metal to fail screens for PAFL. Ninety percent of the
          measurements of arsenic failed screens. Five additional pollutants were added to
          PAFL's pollutants of interest because they are NATTS MQO Core Analytes, even
          though they did not fail any screens (beryllium, cadmium, lead, manganese, and
          nickel). These pollutants are not shown in Table 9-4 but are shown in subsequent
          tables in the sections that follow.

9.4  Concentrations
      This section presents various concentration  averages used to characterize pollution levels
at the Florida monitoring sites. Where applicable, the following calculations and data analyses
were performed: Time period-based concentration averages (quarterly and annual) are provided
for the pollutants of interest for the Florida monitoring sites, where the data meet the applicable
criteria. Concentration averages for select pollutants are also presented graphically for the sites to
illustrate how the sites' concentrations compare to the program-level averages, as presented in
Section 4.1. In addition, concentration averages for select pollutants are presented from previous
years of sampling in order to characterize concentration trends at the sites. Additional site-
specific statistical summaries for the Florida monitoring sites are provided in Appendices L, M,
N,and O.

9.4.1   2011 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 of the total number of samples
possible within a given quarter for a quarterly average to be calculated. An annual average
includes all  measured detections and substituted zeros for non-detects for the entire year of
sampling. Annual averages were calculated for pollutants where three valid quarterly averages
could be calculated and where method completeness was greater than or equal to 85 percent, as
presented in Section 2.4. Quarterly and annual average concentrations for the Florida monitoring
sites are presented in Table 9-5, where applicable.  Note that concentrations of the PAHs, metals,
and hexavalent chromium are presented in ng/m3 for ease of viewing. Also note that if a
pollutant was not detected in a given calendar quarter, the quarterly average simply reflects "0"
                                           9-33

-------
because only zeros substituted for non-detects were factored into the quarterly average

concentration.
   Table 9-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest
                              for the Florida Monitoring Sites
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Hg/m3)
2nd
Quarter
Average
(Hg/m3)
3rd
Quarter
Average
(jig/m3)
4th
Quarter
Average
(jig/m3)
Annual
Average
(jig/m3)
St. Petersburg, Florida - AZFL
Acetaldehyde
Formaldehyde
62/62
62/62
2.98
ฑ 0.59
1.69
ฑ 0.28
2.46
ฑ 0.67
1.86
ฑ 0.30
2.25
ฑ 0.49
1.81
ฑ 0.33
1.07
ฑ 0.16
2.03
ฑ 0.20
2.13
ฑ 0.30
1.86
ฑ 0.13
Pinellas Park, Florida - SKFL
Acetaldehyde
Formaldehyde
Benzo(a)pyrenea
Hexavalent Chromium3
Naphthalene3
61/61
61/61
55/61
55/62
61/61
3.15
ฑ 1.02
1.16
ฑ 0.31
0.05
ฑ 0.02
0.02
ฑ 0.01
83.83
ฑ 35.05
1.76
ฑ 0.24
2.72
ฑ 0.36
0.03
ฑ 0.01
0.03
ฑ 0.01
79.78
ฑ 25.27
1.58
ฑ 0.28
2.60
ฑ 0.31
0.04
ฑ 0.01
0.03
ฑ 0.01
99.97
ฑ 31.68
1.33
ฑ 0.19
2.13
ฑ 0.26
0.07
ฑ 0.05
0.01
ฑ 0.01
66.37
ฑ 24.19
1.95
ฑ 0.31
2.16
ฑ 0.21
0.05
ฑ 0.01
0.02
ฑ <0.01
82.22
ฑ 14.07
Plant City, Florida - SYFL
Acetaldehyde
Formaldehyde
Benzo(a)pyrene3
Hexavalent Chromium3
Naphthalene3
60/60
60/60
25/60
42/59
60/60
1.04
ฑ 0.18
1.62
ฑ 0.27
0.02
ฑ 0.01
0.01
ฑ 0.01
47.12
ฑ 15.91
1.34
ฑ 0.40
3.36
ฑ 0.97
0.02
ฑ 0.02
0.02
ฑ 0.01
46.58
ฑ 16.20
1.00
ฑ 0.25
2.75
ฑ 0.65
0.01
ฑ 0.01
0.02
ฑ 0.01
36.77
ฑ 9.75
0.75
ฑ 0.09
1.49
ฑ 0.20
0.05
ฑ 0.03
0.01
ฑ <0.01
38.20
ฑ 13.96
1.03
ฑ 0.13
2.30
ฑ 0.34
0.02
ฑ 0.01
0.01
ฑ <0.01
42.00
ฑ 6.68
Winter Park, Florida - ORFL
Acetaldehyde
Formaldehyde
60/60
60/60
1.82
ฑ 0.37
1.42
ฑ 0.28
2.08
ฑ 0.24
2.12
ฑ 0.30
1.81
ฑ 0.28
2.48
ฑ 0.36
1.67
ฑ 0.26
1.49
ฑ 0.25
1.85
ฑ 0.14
1.89
ฑ 0.18
      3 Average concentrations provided below the blue line for this site and/or pollutant are presented in
      ng/m3 for ease of viewing.
                                            9-34

-------
Table 9-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest
                   for the Florida Monitoring Sites (Continued)
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Ug/m3)
2nd
Quarter
Average
(Ug/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Ug/m3)
Orlando, Florida - PAFLa
Arsenic (PM10)a
Beryllium (PM10)a
Cadmium (PM10)a
Lead (PM10)a
Manganese (PM10)a
Nickel (PM10)a
31/31
31/31
31/31
31/31
31/31
31/31
0.63
ฑ 0.31
0.01
ฑ <0.01
0.07
ฑ 0.02
2.25
ฑ 1.04
1.83
ฑ0.47
0.63
ฑ 0.15
0.31
ฑ 0.07
0.01
ฑ <0.01
0.05
ฑ 0.01
2.07
ฑ 0.86
2.39
ฑ0.82
0.98
ฑ 0.32
0.79
ฑ 0.36
0.01
ฑ 0.01
0.05
ฑ 0.01
1.86
ฑ 0.74
2.74
ฑ0.78
0.77
ฑ 0.10
0.70
ฑ 0.47
0.01
ฑ 0.01
0.05
ฑ 0.03
2.35
ฑ 1.40
1.96
ฑ0.81
0.53
ฑ 0.07
0.62
ฑ 0.17
0.01
ฑ 0.01
0.06
ฑ 0.01
2.13
ฑ 0.47
2.23
ฑ0.35
0.72
ฑ 0.10
   a Average concentrations provided below the blue line for this site and/or pollutant are presented in
   ng/m3 for ease of viewing.


    Observations from Table 9-5 include the following:

    •  SYFL's annual average concentration of formaldehyde (2.30 ฑ 0.34 |ig/m3) is the
       highest annual average concentration among the Florida sites. The annual average
       concentration of formaldehyde is higher than the annual average acetaldehyde
       concentration for both SKFL and SYFL; the annual average of acetaldehyde is higher
       than formaldehyde for AZFL; and the annual averages for the two carbonyl
       compounds were similar to each for ORFL.

    •  The annual average concentrations of formaldehyde range from 1.89 ฑ 0.18 |ig/m3
       (ORFL) to 2.30 ฑ 0.34 |ig/m3 (SYFL). The annual average concentrations of
       acetaldehyde range from 1.03 ฑ 0.13 |ig/m3 (SYFL) to 2.13 ฑ 0.30 |ig/m3 (AZFL).

    •  The first quarter acetaldehyde average for SKFL is greater than the other quarterly
       average concentrations and has a relatively large confidence interval associated with
       it. A review of the data shows that the maximum concentration of acetaldehyde
       measured at SKFL was measured on February 14, 2011 (8.94 |ig/m3) and is more
       than twice the next highest concentration measured at SKFL  (4.33 |ig/m3 measured
       on January 3, 2011). The nine highest concentrations of acetaldehyde were measured
       at SKFL during the first quarter of 2011.

    •  Concentrations of formaldehyde appear to be higher during the warmer months of the
       year, as the second and third quarter formaldehyde averages for SYFL and ORFL are
       greater than the other quarterly average concentrations. This trend appears to
       continue at SKFL but the confidence intervals calculated for  these averages indicate
       that the difference among the quarterly averages is not statistically significant.  This
       trend is not shown for AZFL.
                                       9-35

-------
       •   As previously discussed, SKFL and SYFL both sampled hexavalent chromium and
          PAHs in addition to carbonyl compounds. Hexavalent chromium, naphthalene, and
          benzo(a)pyrene are all pollutants of interest for these two sites. The annual average
          concentrations of these three pollutants are higher for SKFL than SYFL, particularly
          for naphthalene.

       •   For PAFL, manganese and lead have the highest annual average concentrations
          among the PMio metals. These are the only two metals with annual average
          concentrations greater than 1 ng/m3.

       •   For PAFL, the first and fourth quarter lead averages have relatively large confidence
          intervals associated with them. A review of the data shows that two highest
          concentrations of lead were measured on December 29, 2011  (6.53 ng/m3) and
          March 16, 2011 (5.16 ng/m3). All other measurements of lead are less than 4 ng/m3.


       Tables 4-9 through 4-12 present the sites with the 10 highest annual average

concentrations for each of the program-level pollutants of interest. Observations for the Florida

sites from those tables include the following:

       •   None of the Florida monitoring sites appear in Table 4-10 for carbonyl compounds or
          Table 4-11 for PAHs.

       •   SKFL has the eighth highest annual average concentration of hexavalent chromium
          among NMP sites sampling this pollutant, as shown in Table 4-12.

       •   Because only nine NMP sites sampled PMi0 metals, all nine sites appear in
          Table 4-12. The annual average concentrations of the PMio metals for PAFL ranked
          eighth or ninth with the exception of arsenic, which ranked fifth.


9.4.2   Concentration Comparison

       In order to better illustrate how a site's annual concentration averages compare to the

program-level averages, a site-specific box plot was created for the selected NATTS MQO Core

Analytes listed in Section 3.5.3, where applicable. Thus, box plots for acetaldehyde and

formaldehyde were created for AZFL, SKFL, SYFL, and ORFL. Box plots were also created for

benzo(a)pyrene, hexavalent chromium, and naphthalene for SKFL and SYFL and for arsenic,

lead, and manganese for PAFL. Figures 9-23 through 9-30 overlay the sites' minimum, annual

average, and maximum concentrations onto the program-level  minimum, first quartile, median,

average, third quartile, and maximum concentrations, as described in Section 3.5.3.
                                         9-36

-------
       Figure 9-23. Program vs. Site-Specific Average Acetaldehyde Concentrations
AZFL
5YFL
                                        6           8
                                               Concentration (
                                                                10
                 Program:   IstQuartile     2ndQuartile     SrdQuartile     4thQuartile     Average
                 Site:      Site Average     Site Minimum/Maximum

                                                                                                     16
       Figure 9-24. Program vs. Site-Specific Average Arsenic (PMio) Concentration
    m
PiFL
              3.5
                         1         1.5          2          2.5          3         3.5          4          4.5
                                               Concentration (ng/m3)
                 Program:   IstQuartile     2ndQuartile     3rdQuartile      4thQuartile      Average
                 Site:       Site Average     Site Minimum/Maximum

                                                 9-37

-------
     Figure 9-25. Program vs. Site-Specific Average Benzo(a)pyrene Concentrations
   E
   E
                                    0.75          1          1.25
                                           Concentration (ng/m3)
                Program:   IstQuartile     2ndQuartile     3rdQuartile     4thQuartile     Average
                Site:      Site Average     Site Minimum/Maximum

      Figure 9-26. Program vs. Site-Specific Average Formaldehyde Concentrations
    •H-
ORFL
                                 10
                                               15
                                          Concentration (|
                                                              20
                                                                             25
               Program:   IstQuartile     2ndQuartile     SrdQuartile     4thQuartile     Average
               Site:      Site Average     Site Minimum/Maximum
                           o
                                                                                           3D
                                            9-38

-------
  Figure 9-27. Program vs. Site-Specific Average Hexavalent Chromium Concentrations
SKFL
                                                  :us
                                             Concentration (ng/m3)
                Program:   IstQuartile     2ndQuartile      SrdQuartile     4thQuartile      Average
                Site:      Site Average     Site Minimum/Maximum

        Figure 9-28. Program vs. Site-Specific Average Lead (PMio) Concentration
                                            15            20
                                             Concentration (ng/m3)
                Program:   IstQuartile     2ndQuartile      SrdQuartile     4thQuartile      Average
                Site:      Site Average     Site Minimum/Maximum

     Figure 9-29. Program vs. Site-Specific Average Manganese (PMio) Concentration
   li
; Program Max Concentration = 395 ng/m3
I
                                      75          100         1Z5
                                             Concentration (ng/m3J
                Program:   IstQuartile     2ndQuartile      SrdQuartile     4thQuartile      Average
                Site:      Site Average     Site Minimum/Maximum

                                               9-39

-------
      Figure 9-30. Program vs. Site-Specific Average Naphthalene Concentrations
                                                          Program Max Concentration =773 ng/ms
5YFL
                                                         i Program Max Concentration =773 ng/m3
           50
                   100
                           153
                                   200      250       300
                                       Concentration (ng/m3)
                                                                    4;:
                                                                            450
                                                                                    5DD
              Program:   IstQuartile     2ndQuartile     SrdQuartile    4thQuartile     Average
              Site:     Site Average     Site Minimum/Maximum

  Observations from Figures 9-23 through 9-30 include the following:

  •   Figure 9-23 for acetaldehyde shows that the annual average concentration for AZFL is
      the only one that is greater than the program-level average concentration among the
      Florida sites, although the annual average for SKFL (1.95 |ig/m3) is just less than the
      program-level average (2.00 |ig/m3). SKFL's maximum concentration is the fourth
      highest acetaldehyde concentration measured among NMP sites sampling this pollutant.
      There were no non-detects of acetaldehyde measured at the Florida sites or across the
      program.

  •   Figure 9-24 for arsenic shows that PAFL's annual average concentration is just greater
      than the program-level average concentration. The maximum arsenic concentration
      measured at PAFL is less than the maximum concentration measured among sites
      sampling PMio metals. There were no non-detects of arsenic measured at PAFL, although
      there were a few reported  across the program.

  •   Figure 9-25 presents the box plots for benzo(a)pyrene. Note that the program-level first
      quartile for this pollutant is zero and is not visible on the box plots. The box plots show
      that the annual  average concentration for SKFL is slightly higher than the annual  average
      concentration for SYFL and that both annual average concentrations are less than the
      program-level average concentration. The annual average for SYFL is also less than the
      program-level median concentration. Figure 9-25 also shows that the maximum
      concentrations  measured at these sites are considerably less than the maximum
      concentration measured across the program.

  •   Figure 9-26 for formaldehyde shows that the annual average concentrations of
      formaldehyde for AZFL, SKFL, SYFL, and ORFL are all less than the program-level
      average  concentration. There is little difference among the site-specific annual averages
      of formaldehyde across the Florida sites (less than 0.45 |ig/m3 separates them). Note that
      the range of formaldehyde concentrations measured at SYFL is roughly twice the range
      measured at AZFL, SKFL, and ORFL.
                                         9-40

-------
•  Figure 9-27 presents the box plots for hexavalent chromium, which was measured at
   SKFL and SYFL. The annual average concentration for SKFL is similar to the program-
   level average while the annual average for SYFL is less than the program-level average.
   SYFL's annual average concentration is also less than the program-level median
   concentration. The maximum concentrations measured at SKFL and SYFL are both less
   than the maximum concentration measured among NMP sites sampling this pollutant.

•  Figure 9-28 for lead shows that PAFL's annual average concentration is less than both
   the program-level average and median concentrations. The maximum lead concentration
   measured at PAFL is considerably less than the maximum concentration measured
   among NMP sites sampling PMio metals.

•  Figure 9-29 presents the box plot for manganese. Note that the program-level maximum
   concentration (395 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 200 |ig/m3. Figure 9-29 shows
   that PAFL's annual average concentration is less than the program-level average, median,
   and first quartile concentrations. The maximum manganese concentration measured at
   PAFL is roughly equivalent to the program-level median concentration. This site has the
   smallest range of manganese measurements among NMP sites sampling PMio metals.

•  Figure 9-30 presents the box plots for naphthalene. Note that the program-level
   maximum concentration (779 ng/m3) is not shown directly on the box plots as the  scale
   has been reduced to 500 ng/m3 in order to allow for the observation of data points  at the
   lower end of the concentration range.  The box plots show that the annual average
   concentration for SKFL is nearly twice the  annual average concentration for SYFL.
   SKFL's annual naphthalene concentration is just greater than the program-level  average
   concentration, although difficult to discern  in Figure 9-30. The maximum concentration
   measured at SKFL is more than 100 ng/m3  greater than the maximum concentration
   measured at SYFL.
                                     9-41

-------
9.4.3   Concentration Trends
       A site-specific trends evaluation was completed for sites that have sampled one or more
of the selected NATTS MQO Core Analytes for 5 consecutive years or longer, as described in
Section 3.5.4. AZFL, ORFL, SKFL, and SYFL have sampled carbonyl compounds as part of the
NMP for at least 5 consecutive years. Thus, Figures 9-31 through 9-38 present the annual
statistical metrics for acetaldehyde and formaldehyde for each of these sites. In addition, SYFL
has sampled hexavalent chromium since 2005; thus, Figure 9-39 presents the annual statistical
metrics for hexavalent chromium for SYFL. The statistical metrics presented for assessing trends
include the substitution of zeros for non-detects. Sampling for PAHs at SKFL and SYFL,
hexavalent chromium at SKFL, and PMio metals at PAFL began in 2008, which is less than the
5 consecutive year criteria; thus, the trends analysis was not conducted for the pollutants for
these methods.
         Figure 9-31. Annual Statistical Metrics for Acetaldehyde Concentrations
                                   Measured at AZFL
        I5
        I
















I




I
















T
-i-
W
-s-
















— i


..<

1





3—


If

J
















— t

•
'<

:

.



3 —

•
>.



[















-I


•.ป
T
1

















I













'"•-...








—t

j

-j






t— |

>.
•

i-



















•d

-3










r














— <



^


-;



>—



ฅ'


r














— c



~<
^
q





t-



>•
•
(^

















rh





T











             2001    2002    1003     1004    2005
                                                   1337    1008    2009    2010    2011
               O 5th Percentile     —  Minimum    — Median     —  Maximum    0  95th Percentile
                                                                         ' Average
                                          9-42

-------
 Figure 9-32. Annual Statistical Metrics for Formaldehyde Concentrations
                            Measured at AZFL




i
E -
.a 10
I
1
B
!



























— (
<




IS







1 i
n r
> 	 ^ fy 1 rtn 1 JU
E JL
r-t-i _ 	 •*! **•-.. -^ r-E-i
,_ Hr1 ^.. 	 • ^_ --..^ r

LjjJ L-P-1 L-2-1 ^f LaJ L^ i-a-i
01 2002 2003 2004 2DD5 2006 2007 2008 2009 2010 2011
Year
O 5th Percentile - Minimum — Median — Maximum 0 95th Percentile •••*•• Average
  Figure 9-33. Annual Statistical Metrics for Acetaldehyde Concentrations
                            Measured at ORFL
I
e
.9
S
                                 2007      2008
                                      Year
        * 5th Percentile     -  Minimum     —  Median    -  Maximum    •  95th Percentile
                                                                   • Average
                                   9-43

-------
 Figure 9-34. Annual Statistical Metrics for Formaldehyde Concentrations
                           Measured at ORFL




"1
a Iu
E
1
3 3
B
e
ฃ
< 6
























[

^ta
-r
1004
*


	





•



>••
•
•^
L
2005
5th Percentile


	


[
MM
-ฃ-^
2006




— Minimum







"•••
^^m
— •—

2D07



Year
Median
••1
^





1 	 * — 1
^ r*n J_

™ "^
r , ^

200S 2009 2010 2011
Maximum * 95th Percentile ..^.. Average
  Figure 9-35. Annual Statistical Metrics for Acetaldehyde Concentrations
                            Measured at SKFL
I
e
.9
S
                                      2008
                                      Ysar
        * 5th Percentile     -  Minimum    — Median    - Maximum
                                                      95th Percentile
                                                                   • Averปi e
                                   9-44

-------
Figure 9-36. Annual Statistical Metrics for Formaldehyde Concentrations
                          Measured at SKFL




ฃ
E
a
i
a
s, 4 .
E 4
S










!


The max mum
2005 is 91.7 |ag/m3
j



T T
I T
I
•• ' 1— "— 1

— '"ita^ ~^ T ^



T ^^^^^ ' ^^^^m ^^^^ "
1 — ^ • 	 — ^
J


1 	 +—* ซ
2005 2006 2007 2008 2009 2010 2011
Year
• 5th Percentile — Minimum — Median — Maximum • 95th Percentile "^"Average
Figure 9-37. Annual Statistical Metrics for Acetaldehyde Concentrations
                          Measured at SYFL
                                2007       2008
                                     Ysar
       *  5th Percentile    -  Minimum    — Median    - Maximum     • 95th Percentile
                                                                 • Average
                                  9-45

-------
    Figure 9-38. Annual Statistical Metrics for Formaldehyde Concentrations
                            Measured at SYFL



"1
Concentration
ft
C
ฃ
<








[

—4—1 r— '



I 1
i T


r ,— 5-, |— "-i _,_,
1 	 + 	 -f 	

t ! — •— ! ' — a—1 — *~~^ L
2004 2005 2006 2007 2008 2009 2010 2011
Year
• 5th Percentile — Minimum — Median — Maximum • 95th Percentile "^"Average
Figure 9-39. Annual Statistical Metrics for Hexavalent Chromium Concentrations
                           Measured at SYFL


_
ge Concentration (n
3 c
i <
4









•


*••



1005
- Minimum




r r
r L
I

c

	 ^"

2006 2007 2008 2009 2010 2011
Year
- Median - Maximum • 95th Percentile • SthPercentile •••*•• Average
                                  9-46

-------
       Observations from Figure 9-31 for acetaldehyde measurements at AZFL include the
following:

       •   Carbonyl compounds have been measured at AZFL since 2001, making this site one
          of the longest running UATMP sites.

       •   The maximum acetaldehyde concentration was measured in 2010 (8.09 |ig/m3),
          although a similar concentration was also measured in 2003 (8.00 |ig/m3).

       •   The average and median acetaldehyde concentrations increased through 2004 then
          began to decrease significantly. The average began to increase again in 2009. This
          increase cannot be attributed to an outlier here or there because the trend continued
          into 2010 and the all statistical metrics exhibited this increase. The 95th percentile
          more than doubled from 2008 to 2009, as did the average concentration.  Although a
          decrease is shown for 2011, additional years of sampling are required to determine if
          this decrease will continue.

       •   With the exception of 2001, the minimum concentration for each year is greater than
          zero. Only two non-detects of acetaldehyde have been reported since the onset of
          carbonyl compound sampling (both in 2001).


       Observations from Figure 9-32 for formaldehyde measurements at AZFL include the

following:

       •   The maximum formaldehyde concentration was measured in 2001, after which the
          highest concentration measured decreased by nearly half. The three highest
          concentrations of formaldehyde (ranging from 9.30 |ig/m3 to  16.1 |ig/m3) were all
          measured in 2001.

       •   The average formaldehyde concentration 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 for 2009.

       •   The trends shown for formaldehyde in Figure 9-32 are almost the opposite of the
          trends shown for acetaldehyde in Figure 9-30.

       •   The range of formaldehyde concentrations measured at AZFL is at a minimum for
          2011.

       •   The minimum concentration for each period is greater than zero. No non-detects of
          formaldehyde have been reported since the onset of carbonyl  compound sampling in
          2001.
                                         9-47

-------
      Observations from Figure 9-33 for acetaldehyde measurements at ORFL include the
following:
       •   Sampling for carbonyl compounds began at ORFL in April 2003. Because fewer than
          85 percent of possible samples were collected in 2003, Figure 9-33 excludes data
          from 2003.

       •   The maximum acetaldehyde concentration was measured in 2006 (9.55 |ig/m3).

       •   The average concentration was at a maximum in 2006. After 2006, the average
          concentrations have varied from 1.45 |ig/m3 in 2010 to 1.85 |ig/m3 in 2011.

       •   Even though the range of acetaldehyde concentrations measured at ORFL is at a
          minimum for 2011, the average concentration is at a maximum for the period from
          2007 to 2011 (as is the median concentration).

       •   The minimum concentration for each period is greater than zero, indicating that no
          non-detects of acetaldehyde have been reported for this pollutant.


       Observations from Figure 9-34 for formaldehyde measurements at ORFL include the

following:

       •   The maximum formaldehyde concentration was measured in 2007 (16.1 |ig/m3),
          although concentrations greater than 10 |ig/m3 were also measured in 2005  and 2008.

       •   Even with the relatively high concentrations measured in the middle years of
          sampling, the average concentrations exhibit a steady decreasing trend. 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 2011, similar to
          acetaldehyde.

       •   The minimum concentration for each year is greater than zero, indicating that no non-
          detects of formaldehyde have been reported for this pollutant.


       Observations from Figure 9-35 for acetaldehyde measurements at SKFL include the
following:

       •   Sampling for carbonyl compounds began at SKFL in July 2004. Because fewer than
          85 percent of possible samples were collected in 2004, Figure 9-35 excludes data
          from 2004.

       •   The maximum acetaldehyde concentration shown was measured in
          2010 (10.3 |ig/m3). The four highest concentrations of acetaldehyde were all
          measured in 2010 and 2011.
                                         9-48

-------
       •  The average acetaldehyde concentration increased steadily between 2007 and 2010,
          after which a significant decrease is shown for 2011. Additional sampling is needed
          to see if this decrease continues.

       •  The range of concentrations measured, as indicated by the minimum and maximum
          concentrations as well as the 5th and 95th percentiles, increased from 2007 through
          2010. Although the second highest concentration was measured at SKFL in 2011, the
          95th percentile is significantly lower for 2011.

       •  The minimum concentration for each year is greater than zero, indicating that no non-
          detects of acetaldehyde  have been reported for this pollutant.


       Observations from Figure 9-36 for formaldehyde measurements at SKFL include the

following:

       •  The maximum formaldehyde concentration was measured at SKFL on July 9, 2005
          (91.7 |ig/m3). Note that  for 2005, the average concentration is greater than the 95th
          percentile, reflecting the effects that an outlier can have on statistical measurements.
          All other concentrations measured at this site were less than 6 |ig/m3 for the years
          shown.

       •  The average and median concentrations exhibit a steady decreasing trend over the
          years shown, with the exception of 2011.

       •  The range of concentrations measured, as indicated by the minimum and maximum
          concentrations as well as the 5th and 95th percentiles, decreased significantly for 2009
          and 2010, then increased for 2011.  Note that the median is  greater than the average
          concentration for 2011,  which is unusual.  This means that there is more variability in
          the concentrations measured in 2011 than in 2010. For instance, the median
          concentration for 2011 is greater than the maximum concentration measured for 2010,
          which means that 50 percent of the concentrations measured in 2011 are greater than
          the maximum concentration measured in 2010.

       •  The minimum concentration for each period is greater than  zero, indicating that no
          non-detects of formaldehyde have been reported for this pollutant.


       Observations from Figure 9-37 for acetaldehyde measurements at SYFL include the

following:

       •  Carbonyl compounds have been measured at SYFL  since January 2004.

       •  The maximum acetaldehyde concentration was measured on January 18, 2007
          (15.3 |ig/m3). The next highest concentration, also measured in 2007, is roughly half
          as high (7.55 |ig/m3).
                                         9-49

-------
       •   With the exception of 2007, the average concentrations have fluctuated between
          1.03 |ig/m3 (2011) and  1.60 |ig/m3 (2004).

       •   All of the statistical parameters increased for 2007. Aside from the two
          measurements of acetaldehyde discussed above, 2007 had the greatest number of
          acetaldehyde concentrations greater than 3 |ig/m3 (16), while every other year of
          sampling had three or less. Thus, it is not just the outliers driving this average
          concentration.

       •   Only one non-detect of acetaldehyde has been reported since the onset of carbonyl
          compound sampling at  SYFL.


       Observations from Figure 9-38 for formaldehyde measurements at SYFL include the

following:

       •   The maximum formaldehyde concentration measured at SYFL was measured in 2005
          (32.5 |ig/m3) and was nearly twice the next highest concentration measured in 2008
          (17.1 |ig/m3), although  several measurements similar in magnitude to this one were
          also measured in 2007.  In all,  eight formaldehyde concentrations greater than
          10 |ig/m3 have been measured at SYFL.

       •   The average formaldehyde concentration has fluctuated over the years, ranging from
          1.58 |ig/m3 in 2006 to 3.19 |ig/m3 in 2007.

       •   Similar to acetaldehyde, all of the statistical parameters exhibit an  increase for 2007,
          even though the maximum concentrations were not measured during this year. The
          difference between the  5th and 95th percentiles for 2007 compared to the difference for
          the remaining years show that the measurements were higher in 2007. The number of
          formaldehyde concentrations greater than 5 |ig/m3 is highest for 2007 (seven), while
          every other year of sampling had two or less.

       •   The minimum concentration for each year is greater than zero. No non-detects of
          formaldehyde have been reported since the  onset of carbonyl compound sampling in
          2004.


       Observations from Figure 9-39 for hexavalent chromium measurements at SYFL include
the following:

       •   Hexavalent chromium sampling at SYFL began in January 2005.

       •   The maximum hexavalent chromium concentration measured at SYFL (0.134 ng/m3)
          was measured on July 3, 2005 and is similar in magnitude to the next highest
          concentrations, measured on July 4, 2006 (0.120 ng/m3) and March 17, 2005
          (0.119 ng/m3).  These are the only  three measurements greater than 0.1 ng/m3.
                                         9-50

-------
       •  The average hexavalent chromium concentration has decreased significantly over the
          years since the onset of sampling, reaching a minimum in 2009.  This is the year with
          the smallest range of concentrations measured. With the exception of the minimum
          and fifth percentiles, all statistical parameters increased in 2010 and again in 2011.
       •  For 2008 and 2009, the median concentration decreased to zero, indicating that at
          least 50 percent of the measurements were non-detects. The percentage of non-detects
          increased from 28 percent in 2006 to a maximum of 70 percent in 2009. The number
          of non-detects decreased in 2010 (41 percent) and again in 2011 (17 percent). The
          changes in the statistical parameters are at least partially attributable to the number of
          non-detects (and thus, zeros) factored into the calculations.

9.5    Additional Risk-Based Screening Evaluations
       The following risk-based screening evaluations were conducted to characterize risk at
each Florida monitoring site. Refer to Sections 3.3 and 3.5.5 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.

9.5.1   Risk-Based Screening Assessment Using MRLs
       A risk-based screening was conducted by comparing the concentration data from the
Florida monitoring sites to the ATSDR MRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 to 14 days); intermediate (exposures of 15 to 364  days); and chronic (exposures
of 1 year or greater). The preprocessed daily measurements of the pollutants of interest were
compared to the acute MRLs; the quarterly averages were compared to the intermediate MRLs;
and the annual averages were compared to the chronic MRLs.

       As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2011.

9.5.2   Cancer Risk and Noncancer Hazard Approximations
       For the pollutants of interest for the Florida sites and where annual average
concentrations could be calculated, risk was examined by calculating cancer risk and noncancer
hazard approximations. These approximations can be used as risk estimates for cancer and
noncancer effects attributable to the pollutants of interest. Although the use of these

                                          9-51

-------
  approximations is limited, they may help identify where policy-makers may want to shift or
  confirm their air-monitoring priorities. Refer to Section 3.5.5.2 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 Florida Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(jig/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
St. Petersburg, Florida - AZFL
Acetaldehyde
Formaldehyde
0.0000022
0.000013
0.009
0.0098
62/62
62/62
2.13
ฑ 0.3
1.86
ฑ 0.13
4.70
24.15
0.24
0.19
Pinellas Park, Florida - SKFL
Acetaldehyde
Benzo(a)pyrene 3
Formaldehyde
Hexavalent Chromium3
Naphthalene3
0.0000022
0.00176
0.000013
0.012
0.000034
0.009

0.0098
0.0001
0.003
61/61
55/61
61/61
55/62
61/61
1.95
ฑ 0.31
<0.01
ฑ <0.01
2.16
ฑ 0.21
<0.01
ฑ 0.01
0.08
ฑ 0.01
4.29
0.08
28.09
0.28
2.80
0.22

0.22
0.01
0.03
Plant City, Florida - SYFL
Acetaldehyde
Benzo(a)pyrene 3
Formaldehyde
Hexavalent Chromium3
Naphthalene 3
0.0000022
0.00176
0.000013
0.012
0.000034
0.009

0.0098
0.0001
0.003
60/60
25/60
60/60
42/59
60/60
1.03
ฑ 0.13
O.01
ฑ O.01
2.30
ฑ 0.34
0.01
ฑ O.01
0.04
ฑ 0.01
2.26
0.04
29.84
0.16
1.43
0.11

0.23
O.01
0.01
Winter Park, Florida - ORFL
Acetaldehyde
Formaldehyde
0.0000022
0.000013
0.009
0.0098
60/60
60/60
1.85
ฑ 0.14
1.89
ฑ 0.18
4.06
24.60
0.21
0.19
— = a Cancer URE or Noncancer RfC is not available.
3 For the annual average concentration of this pollutant in ng/m3, refer to Table 9-5.
                                             9-52

-------
        Table 9-6. Risk Approximations for the Florida Monitoring Sites (Continued)
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Orlando, Florida - PAFL
Arsenic (PM10)a
Bery Ilium (PM10)a
Cadmium (PM10)a
Lead(PM10)a
Manganese (PM10)a
Nickel (PM10)a
0.0043
0.0024
0.0018


0.00048
0.000015
0.00002
0.00001
0.00015
0.00005
0.00009
31/31
31/31
31/31
31/31
31/31
31/31
<0.01
ฑ <0.01
<0.01
ฑ <0.01
<0.01
ฑ <0.01
<0.01
ฑ <0.01
<0.01
ฑ<0.01
<0.01
ฑ <0.01
2.66
0.01
0.10


0.35
0.04
O.01
0.01
0.01
0.04
0.01
— = a Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 9-5.
        Observations for the Florida sites from Table 9-6 include the following:

        •   Formaldehyde has the highest cancer risk approximations among the sites sampling
            carbonyl compounds, ranging from 24.15 in-a-million (AZFL) to 29.84 in-a-million
            (SYFL).

        •   The cancer risk approximations for acetaldehyde are an order of magnitude less than
            the cancer risk approximations for formaldehyde, ranging from 2.26 in-a-million
            (SYFL) to 4.70 in-a-million (AZFL).

        •   For the two sites sampling PAHs and hexavalent chromium in addition to carbonyl
            compounds, naphthalene has the third highest cancer risk approximations for each
            site, behind formaldehyde and acetaldehyde. Cancer risk approximations for
            hexavalent chromium and benzo(a)pyrene are less than 1.0 in-a-million for both sites.

        •   For PAFL, arsenic has the highest cancer risk approximation (2.66 in-a-million). The
            cancer risk approximations are less than 1.0 in-a-million for the remaining metals,
            where a cancer URE is available.

        •   All of the noncancer hazard approximations for the site-specific pollutants of interest
            are less than 1.0, indicating that no adverse health  effects are expected from these
            individual pollutants.
                                            9-53

-------
9.5.3   Risk-Based Emissions Assessment
       In addition to the risk-based screenings discussed above, Tables 9-7 and 9-8 present an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 9-7 presents the 10 pollutants with the highest emissions from the 2008 NEI, the 10
pollutants with the highest toxicity-weighted emissions, and the  10 pollutants with the highest
cancer risk approximations (in-a-million), as calculated from the annual averages provided in
Table 9-6. Table 9-8 presents similar information, but identifies  the 10 pollutants with the
highest noncancer hazard approximations (HQ), also calculated from annual averages provided
in Table 9-6.

       The pollutants listed in Tables 9-7 and 9-8 are limited to  those that have cancer and
noncancer toxicity factors, respectively. As a result, although the actual value of the emissions is
the same, the highest emitted pollutants in the cancer table may be different from the noncancer
table. Further, the cancer risk and noncancer hazard approximations based on each site's annual
averages are limited to those pollutants for which each respective site  sampled. As discussed in
Section 9.3, AZFL and ORFL sampled for carbonyl compounds only; SKFL and SYFL sampled
hexavalent chromium and PAHs in addition to carbonyl compounds; and PAFL sampled only
PMio metals. In addition, the cancer risk and noncancer hazard approximations are limited to
those pollutants with enough data to meet the criteria for annual  averages to be calculated. A
more in-depth discussion of this analysis is provided in Section 3.5.5.3. Similar to the cancer risk
and noncancer hazard approximations, this analysis may help policy-makers prioritize their air
monitoring activities.
                                          9-54

-------
Table 9-7. 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)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity- Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
St. Petersburg, Florida (Pinellas County) - AZFL
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Naphthalene
Dichloromethane
POM, Group 2b
Nickel, PM
Tetrachloroethylene
430.92
261.69
214.09
118.05
59.55
26.34
10.63
3.58
2.82
0.84
Benzene
Formaldehyde
1,3 -Butadiene
Nickel, PM
Hexavalent Chromium, PM
Naphthalene
Ethylbenzene
Arsenic, PM
POM, Group 2b
Acetaldehyde
3.36E-03
2.78E-03
1.79E-03
1.35E-03
9.45E-04
8.96E-04
6.54E-04
4.13E-04
3.15E-04
2.60E-04
Formaldehyde
Acetaldehyde
24.15
4.70

Pinellas Park, Florida (Pinellas County) - SKFL
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Naphthalene
Dichloromethane
POM, Group 2b
Nickel, PM
Tetrachloroethylene
430.92
261.69
214.09
118.05
59.55
26.34
10.63
3.58
2.82
0.84
Benzene
Formaldehyde
1,3 -Butadiene
Nickel, PM
Hexavalent Chromium, PM
Naphthalene
Ethylbenzene
Arsenic, PM
POM, Group 2b
Acetaldehyde
3.36E-03
2.78E-03
1.79E-03
1.35E-03
9.45E-04
8.96E-04
6.54E-04
4.13E-04
3.15E-04
2.60E-04
Formaldehyde
Acetaldehyde
Naphthalene
Hexavalent Chromium
Benzo(a)pyrene
28.09
4.29
2.80
0.28
0.08


-------
Table 9-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
                                         the Florida Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants
with Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity- Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Plant City, Florida (Hillsborough County) - SYFL
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Naphthalene
Dichloromethane
Methyl tert butyl ether
POM, Group 2b
Nickel, PM
545.24
339.12
333.41
174.74
77.10
37.70
11.51
8.62
5.02
2.88
Formaldehyde
Benzene
1,3 -Butadiene
Hexavalent Chromium, PM
Nickel, PM
Naphthalene
Arsenic, PM
Ethylbenzene
POM, Group 2b
Acetaldehyde
4.33E-03
4.25E-03
2.31E-03
1.44E-03
1.38E-03
1.28E-03
1.01E-03
8.48E-04
4.42E-04
3.84E-04
Formaldehyde
Acetaldehyde
Naphthalene
Hexavalent Chromium
Benzo(a)pyrene
29.84
2.26
1.43
0.16
0.04

Winter Park, Florida (Orange County) - ORFL
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
Dichloromethane
POM, Group 2b
Tetrachloroethylene
Propylene oxide
581.25
350.51
334.02
169.72
80.64
34.81
10.25
6.17
2.34
1.17
Formaldehyde
Benzene
1,3 -Butadiene
Naphthalene
Ethylbenzene
Hexavalent Chromium, PM
POM, Group 2b
Acetaldehyde
POM, Group 3
Arsenic, PM
4.56E-03
4.53E-03
2.42E-03
1.18E-03
8.35E-04
6.86E-04
5.43E-04
3.73E-04
3.57E-04
2.84E-04
Formaldehyde
Acetaldehyde
24.60
4.06


-------
Table 9-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
                                         the Florida Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants
with Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity- Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Orlando, Florida (Orange County) - PAFL
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
Dichloromethane
POM, Group 2b
Tetrachloroethylene
Propylene oxide
581.25
350.51
334.02
169.72
80.64
34.81
10.25
6.17
2.34
1.17
Formaldehyde
Benzene
1,3 -Butadiene
Naphthalene
Ethylbenzene
Hexavalent Chromium, PM
POM, Group 2b
Acetaldehyde
POM, Group 3
Arsenic, PM
4.56E-03
4.53E-03
2.42E-03
1.18E-03
8.35E-04
6.86E-04
5.43E-04
3.73E-04
3.57E-04
2.84E-04
Arsenic
Nickel
Cadmium
Beryllium
2.66
0.35
0.10
0.01


-------
   Table 9-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
                                             RfCs for the Florida Monitoring Sites
oo
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity- Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Noncancer Hazard
Approximation
Pollutant (HQ)
St. Petersburg, Florida (Pinellas County) - AZFL
Toluene
Xylenes
Methanol
Benzene
Hexane
Ethylbenzene
Formaldehyde
Acetaldehyde
Styrene
Ethylene glycol
1,374.54
1,035.91
540.89
430.92
353.10
261.69
214.09
118.05
113.79
73.16
Acrolein
Nickel, PM
1,3 -Butadiene
Formaldehyde
Benzene
Acetaldehyde
Xylenes
Manganese, PM
Naphthalene
Lead, PM
625,802.58
31,311.86
29,777.05
21,845.80
14,364.15
13,116.58
10,359.11
10,118.05
8,780.20
6,986.71
Acetaldehyde 0.24
Formaldehyde 0.19

Pinellas Park, Florida (Pinellas County) - SKFL
Toluene
Xylenes
Methanol
Benzene
Hexane
Ethylbenzene
Formaldehyde
Acetaldehyde
Styrene
Ethylene glycol
1,374.54
1,035.91
540.89
430.92
353.10
261.69
214.09
118.05
113.79
73.16
Acrolein
Nickel, PM
1,3 -Butadiene
Formaldehyde
Benzene
Acetaldehyde
Xylenes
Manganese, PM
Naphthalene
Lead, PM
625,802.58
31,311.86
29,777.05
21,845.80
14,364.15
13,116.58
10,359.11
10,118.05
8,780.20
6,986.71
Formaldehyde 0.22
Acetaldehyde 0.22
Naphthalene 0.03
Hexavalent Chromium <0.01


-------
   Table 9-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
                                       RfCs for the Florida Monitoring Sites (Continued)
VO
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity- Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Plant City, Florida (Hillsborough County) - SYFL
Toluene
Xylenes
Methanol
Benzene
Hexane
Cyanide Compounds, gas
Hydrochloric acid
Ethylbenzene
Formaldehyde
Acetaldehyde
1,643.63
1,310.35
688.28
545.24
438.64
428.19
350.01
339.12
333.41
174.74
Acrolein
Cyanide Compounds, gas
1,3 -Butadiene
Formaldehyde
Nickel, PM
Manganese, PM
Acetaldehyde
Benzene
Hydrochloric acid
Arsenic, PM
851,462.75
535,232.13
38,551.07
34,021.47
31,976.42
23,688.87
19,415.44
18,174.77
17,500.47
15,600.20
Formaldehyde
Acetaldehyde
Naphthalene
Hexavalent Chromium

Noncancer Hazard
Approximation
(HQ)

0.23
0.11
0.01
<0.01

Winter Park, Florida (Orange County) - ORFL
Toluene
Xylenes
Methanol
Benzene
Hexane
Formaldehyde
Ethylbenzene
Acetaldehyde
Styrene
Ethylene glycol
1,693.63
1,316.88
629.52
581.25
434.13
350.51
334.02
169.72
164.64
81.80
Acrolein
1,3 -Butadiene
Formaldehyde
Benzene
Acetaldehyde
Xylenes
Cyanide Compounds, gas
Naphthalene
Arsenic, PM
Lead, PM
870,786.72
40,319.80
35,765.86
19,374.99
18,857.75
13,168.77
12,178.13
11,604.18
4,404.23
3,678.02
Acetaldehyde
Formaldehyde

0.21
0.19


-------
Table 9-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
                                    RfCs for the Florida Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity- Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer Hazard
Approximation
(HQ)
Orlando, Florida (Orange County) - PAFL
Toluene
Xylenes
Methanol
Benzene
Hexane
Formaldehyde
Ethylbenzene
Acetaldehyde
Styrene
Ethylene glycol
1,693.63
1,316.88
629.52
581.25
434.13
350.51
334.02
169.72
164.64
81.80
Acrolein
1,3 -Butadiene
Formaldehyde
Benzene
Acetaldehyde
Xylenes
Cyanide Compounds, gas
Naphthalene
Arsenic, PM
Lead, PM
870,786.72
40,319.80
35,765.86
19,374.99
18,857.75
13,168.77
12,178.13
11,604.18
4,404.23
3,678.02
Manganese
Arsenic
Lead
Nickel
Cadmium
Beryllium
0.04
0.04
0.01
0.01
0.01
<0.01


-------
Observations from Table 9-7 include the following:

•  Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
   cancer UREs in Pinellas, Hillsborough, and Orange Counties, although not
   necessarily in that order.

•  Benzene, formaldehyde, and 1,3-butadiene have the highest toxi city-weighted
   emissions for all three counties, although the order varies.

•  Seven of the highest emitted pollutants in Hillsborough and Orange Counties also
   have the highest toxi city-weighted emissions. Eight of the highest emitted pollutants
   in Pinellas 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.

•  PAFL sampled only PMio metals; arsenic has the highest cancer risk approximation
   for this site. Arsenic ranks 10th among the toxicity-weighted emissions for Orange
   County, but is not among the highest emitted pollutants, indicating the relative
   toxicity of a low quantity of emissions.

•  Hexavalent chromium, which was sampled for at SKFL and SYFL, ranks between
   fourth and sixth among each county's toxicity-weighted emissions but is not among
   the highest emitted for any of the Florida counties.

•  POM, Group 2b is one of the highest emitted "pollutants" in all three counties and
   appears among the pollutants with the highest toxicity-weighted emissions. POM,
   Group 2b includes several PAHs  sampled for at SKFL and SYFL including
   acenaphthene, benzo(e)pyrene, and fluorene. None of these pollutants failed screens
   for SKFL or SYFL.
Observations from Table 9-8 include the following:

•  Toluene, xylenes, and methanol are the highest emitted pollutants with noncancer
   RfCs in all three 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 three
   Florida counties.

•  Four of the highest emitted pollutants in Pinellas and Orange Counties also have the
   highest toxicity-weighted emissions. Five of the highest emitted pollutants in
   Hillsborough County also have the highest toxicity-weighted emissions.
                                   9-61

-------
       •   Formaldehyde and acetaldehyde, which have the highest noncancer hazard
          approximations for the sites sampling carbonyl compounds, appear on both
          emissions-based lists for each site/county. Naphthalene is among the pollutants with
          the highest toxicity-weighted emissions for Pinellas and Orange Counties but is not
          among the highest emitted in any of the three counties. For PAFL, arsenic and lead
          are among the pollutants with the highest toxicity-weighted emissions. None of the
          metals sampled at PAFL are among the highest emitted pollutants in Orange County.


9.6    Summary of the 2011 Monitoring Data for the Florida Monitoring Sites

       Results from several of the data treatments described in this section include the

following:

       ปซป  Acetaldehyde and'formaldehyde failedscreens for AZFL andORFL, where only
          carbonyl compounds were sampled. Five pollutants (three carbonyls, naphthalene,
          and hexavalent chromium) failed screens for SKFL. Four pollutants (two carbonyls
          and two PAHs) failed screens for SYFL.  Arsenic was the only PMw metal to fail
          screens for PAFL.

       ปซป  Acetaldehyde had the highest annual average concentration for AZFL, formaldehyde
          had the highest annual average concentration for SKFL and SYFL, and the annual
          averages of these two pollutants were roughly the same for ORFL. Manganese and
          lead had the highest annual average concentration of the metals sampled at PAFL.

       ปซป  Formaldehyde concentrations were higher during the warmer months of the year at
          SYFL and ORFL.

       ปซป  Concentrations of formaldehyde have a decreasing trend at ORFL since the onset of
          sampling. In recent years, concentrations of hexavalent chromium have been
          increasing at SYFL. Concentrations of acetaldehyde decreased significantly between
          2010 and 2011 at AZFL and SKFL. Conversely, formaldehyde concentrations at
          SKFL  have increased between 2010 and 2011.
                                         9-62

-------
10.0   Site in Georgia
       This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS site in Georgia, and integrates these concentrations with
emissions, meteorological, and risk information. Data generated by sources other than ERG are
not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) for detailed discussions and definitions
regarding the various data analyses presented below.

10.1   Site Characterization
       This section characterizes the SDGA monitoring site by providing geographical and
physical information about the location of the site and the surrounding area. This information is
provided to give the reader insight regarding factors that may influence the air quality near the
site and assist in the interpretation of the ambient monitoring measurements.

       The SDGA monitoring site  is located in Decatur, Georgia, a suburb of Atlanta. Figure 10-
1 is a composite satellite image retrieved from ArcGIS Explorer showing the monitoring site in
its urban location. Figure 10-2 identifies nearby point source emissions locations by source
category, as reported in the 2008 NEI for point sources. Note that only sources within 10 miles
of the site are included in the facility counts provided in Figure 10-2. A 10-mile boundary was
chosen to give the reader an indication of which emissions sources and emissions source
categories could potentially have a  direct effect on the air quality at the monitoring site. Further,
this boundary provides both the proximity of emissions sources to the monitoring site as well as
the quantity of such sources within a given distance of the site.  Sources outside the 10-mile
radius are still visible on the map, but have been grayed out in order to show emissions  sources
just outside the boundary. Table  10-1 provides supplemental geographical information such as
land use, location setting, and locational coordinates.
                                           10-1

-------
Figure 10-1. Decatur, Georgia (SDGA) Monitoring Site

-------
Figure 10-2. NEI Point Sources Located Within 10 Miles of SDGA
           Legend
            •^  SDGA NATTS site
                                                                         64'10'0*W
                                        Note: Due to facility density and collocation, the total facilities
                                        displayed may not represent all facilities within the area of interest.
10 mile radius
                   County boundary
           Source Category Group (No. of Facilities)
            +   Aircraft Operations (22)
            6   Bakery (1)
            ฉ   Fabricated Metal Products (1)
            ---   Flexible Polyurethane Foam Production (2)
            F   Food Processing/Agriculture (1)
            V   Glass Manufacturing (1)
            ฎ   Institutional-school (1)
            •   Landfill (2)
            ?   Miscellaneous Commercial/lndustrial (3)
            •   Oil and/or Gas Production (1)
            S   Surface Coating (1)
                                        10-3

-------
                              Table 10-1. Geographical Information for the Georgia Monitoring Site

Site
Code

SDGA



AQS Code

13-089-0002



Location

Decatur



County

DeKalb

Micro- or
Metropolitan
Statistical Area

Atlanta-Sandy
Springs-Marietta,
CrA
Latitude
and
Longitude

33.688007,
-84.290325



Land Use

Residential


Location
Setting

Suburban



Additional Ambient Monitoring Information1
CO, SO2, NOy, NO, NO2, NOX, PAMS, Carbonyl
compounds, VOCs, O3, Meteorological parameters,
PMio, PM Coarse, PM10 Speciation, Black carbon,
PM2 5, and PM2 5 Speciation, Haze.
BOLD ITALICS = EPA-designated NATTS Site

-------
       SDGA is located on the DeKalb County Schools Environmental Education property off
Wildcat Road and is the South DeKalb NATTS site. Residential subdivisions, a greenhouse and
horse barn, an athletic field, and a high school surround the monitoring site. A golf course backs
up against the school property. Interstate-285 is located less than 1 mile north of the site, as
shown in Figure 10-1. As Figure 10-2 shows, only one point source (a bakery) is located in close
proximity to SDGA. Additional  sources are located primarily on the west side of the 10-mile
radius. The aircraft operations source category (which includes airports as well as small runways,
heliports, or landing pads) is the source category with the greatest number of emissions sources
within 10 miles of SDGA.

       Table 10-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Georgia monitoring site. Table 10-2 includes county-level
population and vehicle registration information. Table 10-2 also includes a county-level vehicle
registration-to-population ratio, which was calculated to represent the number of vehicles per
person within DeKalb County. In addition, the population within 10 miles of the site is
presented, based on postal code population data estimates. An estimate of 10-mile vehicle
ownership was determined by applying the county-level vehicle registration-to-population ratio
to the 10-mile population surrounding the monitoring site. Table 10-2 also contains traffic
volume information for SDGA. Finally, Table 10-2 presents the county-level daily VMT for
DeKalb County.
      Table 10-2. Population, Motor Vehicle, and Traffic Information for the Georgia
                                      Monitoring Site
Site
SDGA
Estimated
County
Population1
699,893
County-level
Vehicle
Registration2
472,535
Vehicles per
Person
(Registration:
Population)
0.68
Population
within 10
miles3
730,133
Estimated
10-mile
Vehicle
Ownership
492,952
Annual
Average
Daily
Traffic4
140,820
County-
level Daily
VMT5
20,187,000
  Bounty-level population estimate reflects 2011 data from the U.S. Census Bureau (Census Bureau, 2012b)
  2County-level vehicle registration reflects 2011  data from the Georgia Department of Revenue (GA DOR,
    2011)
  310-mile population estimate reflects 2011 data  from Xionetic (Xionetic, 2011)
  4AADT reflects 2011 data from the Georgia DOT (GA DOT, 201 la)
  5County-level VMT reflects 2011 data for all public roads from the Georgia DOT (GA DOT, 201 Ib)
  BOLD ITALICS = EPA-designated NATTS Site
                                            10-5

-------
       Observations from Table 10-2 include the following:
       •  SDGA's county-level population and vehicle registration are in the middle of the
          range compared to other counties with NMP sites. The same is also true for its
          10-mile population and estimated vehicle ownership.
       •  The vehicle-per-person ratio is among the lower ratios compared to other NMP sites.
       •  The traffic volume experienced near SDGA ranks eighth highest compared to other
          NMP monitoring sites. The traffic estimate provided is for 1-285, north of Clifton
          Spring Road.
       •  The daily VMT for DeKalb  County is in the middle of the range among counties with
          NMP sites (where VMT data were available).

10.2   Meteorological Characterization
       The following sections characterize the meteorological conditions near the monitoring
site in Georgia on sample days, as well  as over the course of the year.
10.2.1  Climate Summary
       Atlanta is the largest city in Georgia, and is located at the base of the Blue Ridge
Mountains. The Gulf of Mexico to the south is the major moisture source for weather systems
that move across the region. Both topographical features, in addition to the Atlantic Ocean to the
east, exert moderating influences on the area's climate, tempering cold air outbreaks from the
north as well as summer heat waves.  Summers are warm and humid while winters are relatively
mild, although snow is not uncommon. The semi-permanent Bermuda High Pressure offshore
over the Atlantic Ocean is a dominant weather feature affecting the Atlanta area, which pulls
warm, moist air into the region. Precipitation is ample, although autumn is  the driest season
(Bair, 1992 and GSCO, 1998).
                                          10-6

-------
10.2.2 Meteorological Conditions in 2011
       Hourly meteorological data from the NWS weather station nearest this site were retrieved
for 2011 (NCDC, 2011). The closest weather station to SDGA is located at W. B.
Hartsfield/Atlanta International Airport (WBAN 13874). Additional information about the
Hartsfield weather station, such as the distance between the site and the weather station, is
provided in Table 10-3. These data were used to determine how meteorological conditions on
sample days vary from conditions experienced throughout the year.

       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 the entire year for 2011. 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 appear slightly cooler and drier than weather
conditions throughout the year. This is likely a result of a number of make-up samples collected
at SDGA in the month of December.

10.2.3 Back Trajectory Analysis
       Figure 10-3 is the composite back trajectory map for days on which samples were
collected at the SDGA monitoring site in 2011. Included in Figure 10-3 are four back trajectories
per sample day. Figure  10-4 is the corresponding cluster analysis. An in-depth description of
these maps and how they were generated is presented in Section 3.5.2.1. For the composite map,
each line represents the 24-hour trajectory along which a parcel of air traveled toward the
monitoring site on a given sample day and time, based on an initial height of 50  meters AGL. For
the cluster analysis, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric circle around the site in Figures  10-3 and 10-4 represents 100 miles.
                                          10-7

-------
                          Table 10-3. Average Meteorological Conditions near the Georgia Monitoring Site
Closest NWS Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(ฐF)
Average
Temperature
(ฐF)
Average
Dew Point
Temperature
(ฐF)
Average
Wet Bulb
Temperature
(ฐF)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Decatur, Georgia - SDGA
W.B.
Hartsfield/ Atlanta
Intl. Airport
13874
(33.64, -84.43)
8.18
miles
239ฐ
(WSW)
Sample
Day
2011
71.3
ฑ4.0
73.6
ฑ 1.6
61.7
ฑ3.8
63.9
ฑ1.6
47.2
ฑ3.9
49.5
ฑ 1.5
54.0
ฑ3.4
55.9
ฑ1.4
62.4
ฑ2.9
62.9
ฑ 1.3
1017.5
ฑ1.3
1017.2
ฑ0.6
7.4
ฑ0.7
7.0
ฑ0.3
      1 Sample day averages are highlighted to help differentiate the sample day averages from the full-year averages.
o
oo

-------
Figure 10-3. 2011 Composite Back Trajectory Map for SDGA
    Figure 10-4. Back Trajectory Cluster Map for SDGA
                          10-9

-------
       Observations from Figures 10-3 and 10-4 include the following:
       •  The composite back trajectory map for SDGA looks like a pinwheel, indicating that
          back trajectories originated from a variety of directions around SDGA.
       •  Size-wise, the 24-hour air shed domain for SDGA is in the bottom-third compared to
          other NMP monitoring sites. While the farthest away a trajectory originated was
          central Missouri, or nearly 500 miles away, the average back trajectory length is
          200 miles. Eighty-four percent of back trajectories originated within 300 miles of the
          site. The longest trajectories tended to originate from the northwest and west, over
          Indiana, Illinois, Missouri, and Arkansas.
       •  The cluster analysis confirms that the longest trajectories originated from the
          northwest and west while the shortest trajectories originated from the north, east, and
          southwest. Over one-quarter of the back trajectories originated from the southwest
          over Alabama and tended to be less than 200 miles in length. Another 25 percent of
          trajectories originated from the north over Tennessee and North Carolina. Nearly
          20 percent of back trajectories originated from the northwest over Kentucky, Illinois,
          and Indiana. Fourteen percent originated to the east, nine percent originated to the
          south over the Florida Panhandle, and six percent originated westward over Arkansas.

10.2.4 Wind Rose Comparison
       Hourly surface wind data from the NWS weather station at Hartsfield International
Airport near SDGA were uploaded into a wind rose software program to produce customized
wind roses, as described in Section 3.5.2.2. A wind rose shows the frequency of wind directions
using "petals" positioned around a 16-point compass, and uses different colors to represent wind
speeds.
       Figure 10-5 presents a map showing the distance between the NWS station and SDGA,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 10-5 also presents three different wind roses for the
SDGA monitoring site. First, a historical wind rose representing 2001 to 2010 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 2011 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2011 is
presented. These can be used to identify the predominant wind speed and direction for 2011 and
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.
                                          10-10

-------
     Figure 10-5. Wind Roses for the Hartsfield International Airport Weather Station
                                  near SDGA
Distance between SDGA and NWS Station
2001-2010 Historical Wind Rose
          2011 Wind Rose
    Sample Day Wind Rose
                                     10-11

-------
       Observations from Figure 10-5 for SDGA include the following:
       •  The NWS weather station at Hartsfield International Airport is the closest weather
          station to SDGA. The weather station is located approximately 8.2 miles west-
          southwest of SDGA.
       •  The historical wind rose shows that winds from the west to north-northwest account
          for nearly 40 percent of wind observations. Easterly winds were also common. Winds
          from the northeast quadrant were rarely observed. Calm winds (< 2 knots) were
          observed for less than 10 percent of the hourly wind measurements.
       •  The wind patterns on both the full-year and sample day wind roses are similar to
          those of the historical wind rose, although northwesterly winds account for a greater
          percentage of wind observations on the sample day wind rose.

10.3   Pollutants  of Interest
       Site-specific "pollutants of interest"  were determined for SDGA in order to allow
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." 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. In addition, if any of the NATTS
MQO Core Analytes measured by the monitoring site did not meet the pollutant of interest
criteria based on the preliminary risk-based  screening, that pollutant was added to the list of site-
specific pollutants  of interest. A more in-depth description of the risk-based screening process is
presented in Section 3.2.

       Table 10-4  presents the results of the preliminary risk-based screening process for
SDGA. The pollutants that failed at least one screen and contributed to 95 percent of the total
failed screens for the monitoring site are shaded. NATTS MQO Core Analytes are bolded. Thus,
pollutants of interest are shaded and/or bolded. SDGA sampled for PAHs and hexavalent
chromium.
                                          10-12

-------
         Table 10-4. Risk-Based Screening Results for the Georgia Monitoring Site
Pollutant
Screening
Value
(Ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
Decatur, Georgia - SDGA
Naphthalene
Acenaphthene
Fluorene
0.029
0.011
0.011
Total
54
1
1
56
61
61
61
183
88.52
1.64
1.64
30.60
96.43
1.79
1.79
96.43
98.21
100.00

       Observations from Table 10-4 for SDGA include the following:
       •  Naphthalene, acenaphthene, and fluorene failed screens. Naphthalene failed the
          majority of the screens (roughly 96 percent), accounting for 54 of the 56 total failed
          screens; the other two pollutants failed only one screen each.
       •  Naphthalene was the only pollutant initially identified as a pollutant of interest based
          on the risk-based screening process. Benzo(a)pyrene and hexavalent chromium were
          added as pollutants of interest for SDGA because they are NATTS MQO Core
          Analytes, even though they did not fail any screens. These pollutants are not shown in
          Table 10-4 but are shown in subsequent  tables in the sections that follow.

10.4   Concentrations
      This section presents various concentration averages used to characterize pollution levels
at the Georgia monitoring site. Where applicable, the following calculations and data analyses
were performed: Time period-based concentration averages (quarterly and annual) are provided
for the pollutants of interest for SDGA, where the data meet the applicable criteria.
Concentration averages for select pollutants are also presented graphically to illustrate how the
site's concentrations compare to the program-level averages, as presented in Section 4.1. In
addition, concentration averages for select pollutants are presented from previous years of
sampling in order to characterize concentration trends at the site. Additional site-specific
statistical summaries for SDGA are provided in Appendices M and O.

10.4.1 2011 Concentration Averages
       Quarterly and annual concentration averages were calculated for the pollutants of interest
for SDGA, as described in Section 3.1.  The quarterly average of a particular pollutant is simply
the average concentration of the preprocessed daily measurements over a given calendar quarter.
Quarterly average concentrations include the substitution of zeros for all non-detects. A site must
have a minimum of 75 percent valid samples of the total number of samples possible within a
                                          10-13

-------
given quarter for a quarterly average to be calculated. An annual average includes all measured

detections and substituted zeros for non-detects for the entire year of sampling. Annual averages

were calculated for pollutants where three valid quarterly averages could be calculated and

where method completeness was greater than or equal to 85 percent, as presented in Section 2.4.

Quarterly and annual average concentrations for the Georgia monitoring site are presented in

Table 10-5, where applicable. Note that if a pollutant was not detected in a given calendar

quarter, the quarterly average simply reflects "0" because only zeros substituted for non-detects

were factored into the quarterly average concentration.
Table 10-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
                               the Georgia Monitoring Site
Pollutant
#of
Measured
Detections vs.
# of Samples
1st
Quarter
Average
(ng/m3)
2nd
Quarter
Average
(ng/m3)
3rd
Quarter
Average
(ng/m3)
4th
Quarter
Average
(ng/m3)
Annual
Average
(ng/m3)
Decatur, Georgia - SDGA
Benzo(a)pyrene
Hexavalent Chromium
Naphthalene
27/61
47/61
61/61
0.09
ฑ0.05
0.02
ฑ0.01
86.71
ฑ31.86
<0.01
ฑ0.01
0.02
ฑ0.01
104.86
ฑ36.88
0.02
ฑ0.02
0.01
ฑ0.01
92.53
ฑ35.91
0.08
ฑ0.06
0.01
ฑ0.01
82.09
ฑ 27.64
0.05
ฑ0.02
0.01
ฑ0.01
90.82
ฑ15.61
       Observations for SDGA from Table 10-5 include the following:

       •  The annual average concentration of naphthalene is significantly higher than the
          annual average concentrations of benzo(a)pyrene and hexavalent chromium.
          Concentrations of naphthalene measured at SDGA range from 10.3 ng/m3 to
          294 ng/m3. SDGA's annual average concentration of naphthalene ranks tenth highest
          among NMP sites sampling this pollutant (as shown in Table 4-11).

       •  The first and fourth quarter averages of benzo(a)pyrene are greater than the other
          quarterly averages and have relatively large confidence intervals associated with
          them. This pollutant was detected in less than half of the PAH samples collected (27
          out of 61). All but one of the 10 highest concentrations of this pollutant (those
          greater than 0.01 ng/m3) were measured in the first and fourth quarters of 2011.
          Conversely, this pollutant was detected only once during the second quarter and four
          times during the third quarter.

       •  Hexavalent chromium concentrations span an order of magnitude, ranging from
          0.0049 ng/m3 to 0.0382 ng/m3.
                                          10-14

-------
10.4.2 Concentration Comparison
       In order to better illustrate how a site's annual average concentrations compare to the
program-level averages,  a site-specific box plot was created for the selected NATTS MQO Core
Analytes listed in Section 3.5.3, where applicable. Thus, box plots for benzo(a)pyrene,
hexavalent chromium, and naphthalene were created for SDGA. Figures 10-6 through 10-8
overlay the site's minimum, annual average, and maximum concentrations onto the program-
level minimum, first quartile, median, average, third quartile, and maximum concentrations, as
described in Section 3.5.3.
       Figure 10-6. Program vs. Site-Specific Average Benzo(a)pyrene Concentration
    I
SDGA
               3.25
                         0.5
                                    0.75         1         1.25
                                          Concentration (ng/mi)
                                                                    1.5
                                                                               1.75
                Program:   IstQuartile     2ndQuartile     3rdQuartile     4thQuartile     Average
                Site:     Site Average     Site Minimum/Maximum

   Figure 10-7. Program vs. Site-Specific Average Hexavalent Chromium Concentration
                                              0.15
                                          Concentration (nฃ/m3)
               Program:  IstQuartile     2nd Quartile     3rd Quartile
                Site:     Site Average     Site Minimum/Maximum

                                                             4th Quartile
                                                                          Average
                                            10-15

-------
       Figure 10-8. Program vs. Site-Specific Average Naphthalene Concentration
SDGi
                                                        Program Max Concentration = 779 ng/m3
                   100
                           153
                                   200      250      300
                                       Concentration (ng/mi)
                                                           35:
                                                                           453
                                                                                   555
Program
Site:
: IstQuartile
Site Average
O
2ndQuartile 3rd Quartile 4thQuartile AVE
Site Minimum/Maximum
rage

      Observations from Figures 10-6 through 10-8 include the following:

         •  Figure 10-6 is the box plot for benzo(a)pyrene. Note that the first quartile for this
            pollutant is zero and is not visible on this box plot. This box plot shows that the
            annual average concentration for SDGA is less than the program-level average
            concentration but greater than the program-level median. Figure 10-6 also shows
            that the maximum concentration measured at SDGA is less than the maximum
            concentration measured across the program. Several non-detects of
            benzo(a)pyrene were measured at SDGA.

         •  Figure 10-7 is the box plot for hexavalent chromium. Figure 10-7 shows that the
            annual average concentration of hexavalent chromium for SDGA is less than the
            program-level average concentration. SDGA's annual average concentration is
            also less than the program-level median concentration. The maximum
            concentration measured at SDGA is significantly less than the program-level
            maximum concentration, but greater than the program-level third quartile
            concentration. Several non-detects of hexavalent chromium were measured at
            SDGA.

         •  Figure 10-8 is the box plot for naphthalene. Note that the program-level
            maximum concentration (779 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
            500 ng/m3. Figure 10-8 shows that the annual average concentration of
            naphthalene for SDGA is greater than the program-level average concentration.
            The maximum naphthalene concentration measured at SDGA is less than the
            program-level maximum concentration. There were no non-detects of naphthalene
            measured at SDGA or across the program.
                                        10-16

-------
10.4.3  Concentration Trends
       A site-specific trends evaluation was completed for sites that have sampled one or more
of the selected NATTS MQO Core Analytes for 5 consecutive years or longer, as described in
Section 3.5.4. SDGA has sampled hexavalent chromium under the NMP since 2005. Thus,
Figure 10-9 presents the annual statistical metrics for hexavalent chromium for SDGA. The
statistical metrics presented for assessing trends include the substitution of zeros for non-detects.
SDGA began sampling PAHs in 2007, but did not begin until April, which does not allow for the
completeness criteria specified in Section 3.5.4 to be met; thus, the trends analysis was not
conducted for the pollutants for these methods.

     Figure 10-9. Annual  Statistical Metrics for Hexavalent Chromium Concentrations
                                  Measured at SDGA



1
Concentration
i
a""
e
st

















r
* ^ 1

....^^H 	 ^^—

2006 2007 2008 2009 2010 2011
Year
• SthPercentile — Minimum — Median — Maximum • 95th Percentile •••*" Average
       Observations from Figure 10-9 include the following:
          •  Although hexavalent chromium sampling began in 2005 at SDGA, sampling did
             not begin until late February, which does not yield enough samples for the
             statistical metrics to be calculated for 2005, based on the criteria specified in
             Section 3.5.4. Thus, Figure 10-9 begins with 2006.
                                         10-17

-------
          •  Due to sampler issues, sampling for hexavalent chromium was discontinued from
             September 2007 through May 2008; therefore, no statistical metrics are presented
             for 2007 or 2008.
          •  The maximum concentration measured for each year shown has decreased by an
             order of magnitude (0.300 ng/m3 in 2006 to 0.0382 ng/m3 in 2011). Yet, the
             difference between the 5th and 95th percentiles exhibits little change over the last
             three years of sampling, indicating that the majority of the measurements fall
             within roughly the same range, at least since 2009.
          •  The difference between the median and average concentration for each year has
             decreased,  indicating decreasing variability within the measurements. For 2011,
             the difference between these two statistical parameters is 0.00011 ng/m3.
          •  The minimum, 5th percentile, and median concentrations for 2009 are all 0,
             indicating that at least 50 percent of the measurements are non-detects; as a result,
             zeros were substituted for more than half of the samples (for statistical purposes).
             The number of non-detects began to decrease after 2009, down to 19 percent for
             2010 and 14 percent for 2011.

10.5   Additional Risk-Based Screening Evaluations
       The following risk-based screening evaluations were conducted to characterize risk at the
SDGA monitoring site. Refer to Sections 3.3 and 3.5.5 for definitions and explanations regarding
the various toxicity factors, time frames, and calculations associated with these risk-based
screenings.

10.5.1  Risk-Based Screening Assessment Using MRLs
       A risk-based screening was conducted by comparing the concentration data from the
Gerogia monitoring site to the ATSDRMRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three  exposure periods: acute
(exposures of 1 to 14 days); intermediate (exposures of  15 to 364 days); and chronic (exposures
of 1 year or greater). The preprocessed daily measurements of the pollutants of interest were
compared  to the acute MRLs; the quarterly averages were compared to the intermediate MRLs;
and the annual averages were compared to the chronic MRLs.

       As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2011.
                                         10-18

-------
  10.5.2 Cancer Risk and Noncancer Hazard Approximations
        For the pollutants of interest for SDGA and where annual average concentrations could
  be calculated, risk was examined by calculating cancer risk and noncancer hazard
  approximations. These approximations can be used as risk estimates for cancer and noncancer
  effects attributable to the pollutants of interest. Although the use  of these approximations is
  limited, they may help identify where policy-makers may want to shift or confirm their air-
  monitoring priorities. Refer to Section 3.5.5.2 for an explanation  of how cancer risk and
  noncancer hazard approximations are calculated and what limitations are associated with them.
  Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
  approximations are presented in Table 10-6, where applicable. Cancer risk approximations are
  presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
  values.

              Table 10-6. Risk Approximations for the Georgia Monitoring Site



Pollutant


Cancer
URE
(Hg/m3)1


Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples


Annual
Average
(ng/m3)


Cancer Risk
Approximation
(in-a-million)

Noncancer
Hazard
Approximation
(HQ)
Decatur, Georgia - SDGA

Benzo(a)pyrene

Hexavalent Chromium

Naphthalene

0.00176

0.012

0.000034

—

0.0001

0.003

27/61

47/61

61/61
0.05
ฑ0.02
0.01
ฑ<0.01
90.82
ฑ15.61

0.09

0.17

3.09

—

<0.01

0.03
— = A Cancer URE or Noncancer RfC is not available
        Observations for SDGA from Table 10-6 include the following:
        •   Naphthalene was the only pollutant of interest for which the cancer risk
            approximation was greater than 1.0 in-a-million (3.09 in-a-million).
        •   Noncancer hazard approximations for naphthalene and hexavalent chromium were
            less than 1.0, indicating that no adverse health effects are expected from these
            individual pollutants. Benzo(a)pyrene does not have a noncancer RfC.
                                           10-19

-------
10.5.3 Risk-Based Emissions Assessment
       In addition to the risk-based screenings discussed above, Tables 10-7 and 10-8 present an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 10-7 presents the 10 pollutants with the highest emissions from the 2008 NEI, the 10
pollutants with the highest toxicity-weighted emissions, and the 10 pollutants with the highest
cancer risk approximations (in-a-million), as calculated from the annual averages provided in
Table 10-6. Table 10-8 presents similar information, but identifies the 10 pollutants with the
highest noncancer hazard approximations (HQ), also calculated from the annual averages
provided in Table 10-6.

       The pollutants listed in Tables 10-7 and 10-8 are limited to those that have cancer and
noncancer toxicity factors, respectively. As a result, although the actual value of the emissions is
the same, the highest emitted pollutants in the cancer table may be different from the noncancer
table. Further, the cancer risk and noncancer hazard approximations based on each site's annual
averages are limited to those pollutants for which each  respective site sampled. As discussed in
Section 10.3, SDGA sampled for PAHs and hexavalent chromium. In addition, the cancer risk
and noncancer hazard approximations are limited to those pollutants with enough data to meet
the criteria for annual  averages to be calculated. A more in-depth discussion of this analysis is
provided in Section 3.5.5.3. Similar to the cancer risk and noncancer hazard approximations, this
analysis may help policy-makers prioritize their air monitoring activities.
                                          10-20

-------
  Table 10-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
                                                  the Georgia Monitoring Site
o
to
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
Decatur, Georgia (DeKalb County) - SDGA
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Naphthalene
Dichloromethane
POM, Group 2b
POM, Group la
Methyl tert butyl ether
378.41
213.44
191.48
102.73
46.29
24.10
6.82
3.49
0.49
0.29
Benzene
Formaldehyde
1,3 -Butadiene
Naphthalene
Ethylbenzene
Hexavalent Chromium, PM
POM, Group 2b
Acetaldehyde
Arsenic, PM
POM, Group 5a
2.95E-03
2.49E-03
1.39E-03
8.19E-04
5.34E-04
3.82E-04
3.07E-04
2.26E-04
1.31E-04
1.01E-04
Naphthalene 3.09
Hexavalent Chromium 0.17
Benzo(a)pyrene 0.09


-------
  Table 10-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer

                                             RfCs for the Georgia Monitoring Site
to
to
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Noncancer
Hazard
Approximation
Pollutant (HQ)
Decatur, Georgia (DeKalb County) - SDGA
Toluene
Xylenes
Methanol
Benzene
Hexane
Ethylbenzene
Formaldehyde
Acetaldehyde
Ethylene glycol
1,3 -Butadiene
1,018.74
794.36
430.90
378.41
223.04
213.44
191.48
102.73
56.41
46.29
Acrolein
1,3 -Butadiene
Formaldehyde
Benzene
Acetaldehyde
Naphthalene
Xylenes
Lead, PM
2,4-Toluene diisocyanate
Arsenic, PM
576,341.15
23,142.89
19,539.04
12,613.63
11,414.67
8,033.73
7,943.60
4,023.67
2,140.67
2,029.50
Naphthalene 0.03
Hexavalent Chromium <0.01


-------
Observations from Table 10-7 include the following:

•  Benzene, ethylbenzene, and formaldehyde are the highest emitted pollutants with
   cancer UREs in DeKalb County.

•  The pollutants with the highest toxi city-weighted emissions (of the pollutants with
   cancer UREs) are benzene, formaldehyde, and 1,3-butadiene.

•  Seven of the highest emitted pollutants also have the highest toxi city-weighted
   emissions for DeKalb County.

•  Naphthalene, which has the highest cancer risk approximation for SDGA, has the
   fourth highest toxicity-weighted emissions and sixth highest emissions for DeKalb
   County.

•  Hexavalent chromium ranks sixth highest for toxicity-weighted emissions, but is not
   among the highest emitted pollutants in DeKalb County.

•  POM, Group 2b is the eighth highest emitted "pollutant"  in DeKalb County and ranks
   seventh for toxicity-weighted emissions. POM, Group 2b includes several PAHs
   sampled for at SDGA including acenaphthene, benzo(e)pyrene, fluorene, and
   perylene. Although acenaphthene and fluorene each failed a single screen, neither
   pollutant was identified as a pollutant of interest for SDGA.

•  Benzo(a)pyrene is part of POM,  Group 5a. POM, Group  5a ranks tenth for toxi city-
   based emissions, but is not among the highest emitted pollutants in DeKalb County.


Observations from Table 10-8 include the following:

•  Toluene, xylenes, and methanol 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.

•  Five of the highest emitted pollutants in DeKalb County also have the highest
   toxicity-weighted emissions.

•  While naphthalene is not one of the 10 highest emitted pollutants with a noncancer
   RfC in DeKalb  County, its toxicity-weighted emissions rank sixth. Hexavalent
   chromium does not appear on either emissions-based list.
                                  10-23

-------
10.6   Summary of the 2011 Monitoring Data for SDGA

       Results from several of the data treatments described in this section include the
following:

       ปซป  Naphthalene, acenaphthene, andfluorene failed screens for SDGA, although
          naphthalene accounted for the majority of failed screens.

       *ป*  Of the site-specific pollutants of the interest, naphthalene had the highest annual
          average concentration for SDGA.

       ปซป  Benzo(a)pyrene concentrations were highest during the colder months of the year.
                                         10-24

-------
11.0   Sites in Illinois
       This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS and UATMP sites in Illinois, and integrates these
concentrations with emissions, meteorological, and risk information. Data generated by sources
other than ERG are not included in the data analyses contained in this report. Readers are
encouraged to refer to Sections  1 through 4 and the glossary (Appendix P) for detailed
discussions and definitions regarding the various data analyses presented below.

11.1   Site Characterization
       This section characterizes the Illinois monitoring sites by providing geographical and
physical information about the location of the sites and the surrounding areas. This information
is provided to give the reader insight regarding factors that may influence the air quality near the
sites and assist in the interpretation of the ambient monitoring  measurements.

       Both Illinois sites are located in northwestern suburbs of Greater Chicago, with NBIL
located in Northbrook and SPIL located in Schiller Park. Figures  11-1 and 11-2 are composite
satellite images retrieved from ArcGIS Explorer showing the monitoring sites in their urban
locations. Figure 11-3 identifies nearby point source emissions locations by source category, as
reported in the 2008 NEI for point sources. Note that only sources within 10 miles of the sites are
included in the  facility counts provided in Figure 11-3. A 10-mile boundary was chosen to give
the reader an  indication of which emissions sources and emissions source categories could
potentially have a direct effect on the air quality at the monitoring sites. Further, this boundary
provides both the proximity of emissions sources to the monitoring sites as well as the quantity
of such sources within a given distance of the sites. Sources  outside the 10-mile radii are still
visible on the map, but have been grayed out in order to show  emissions sources just outside the
boundary. Table 11-1 provides supplemental geographical information such as land use, location
setting, and locational coordinates.
                                           11-1

-------
          Figure 11-1. Northbrook, Illinois (NBIL) Monitoring Site
i^Wromoll ;W?
$r-?&!|> ;ฃv, 4\v ซ ?M5 r^-'-';, qjfoif-
                         63

-------
Figure 11-2. Schiller Park, Illinois (SPIL) Monitoring Site

-------
 Figure  11-3.  NEI  Point  Sources Located Within  10  Miles of NBIL and  SPIL
              tO'Q-W      86'5'Q-W       88 Q'Q-W       $7J55'0~W      S7-SO'D~W       &J~4$'QrW      a7'4Q"Q"W      W'KQ'W      87'30'CrW
Legend
                                                                                                                        -,-  -  - *s
                                                  88'OTTW      87"55"CTW      87"50'D"W      87'45'CTW      97^0'0'W      B7"35XTVV
                                                                     Note: Due lo facility density and collocation, the total facilities
                                                                     displayed may not represent all: facilities within the area of interest,
         NBIL NATTS site
                                       SPIL UATMP Site
Source Category Group (No. of Facilities)
  :::   AtwastvE Product Manufacturing (2)
 •;?-"   Air-eondiliomngi'Rer'rfcjjeration (2)
 Hh   Aircraft Operations (32)
  I   Asphalt Pf Qcessingi'Roolmg Manufacturing (3)
  0   Aulo Body Shop/Painters (3)
 ffl   Au1amobile>Tfuck Manufacturing (9)
 0   Bakery <9)
  r   Brow&ry/DistilleryArVinery (1)
 .  :   Bnck Manufacturing i Structural Clay (3)
 f   Building Construction (6)
 B   Bulk Tertninals/BulJc Plants {10}
 C   Chemical Manufacturing {20)
 •   Concrete Batch Plant (20)
[XJ  Crematory •AnimaVHuman<12)
 0   Dry Cleaning Facility (51)
 6   Electrical Equipment (18)
  f   Electricity Generation via Combustion (6)
 E   Electroplating, Plating, Polishing Anodizing.  & Coloring (58)
  4   Engine Testing (2)
10  mile radius  |_   J County boundary
                          M  Mss-eeHaneous Manufacturing (86)
                          •  Oil andtorGas Production (1)
                         CD  Pharmaceutical Manufacturing (8)
                          i   Pipeline Compressor Station {1)
                          1   Primary Metal Production (24)
                          ^ซ  Printing. Coaling & Dyeing of Fabnc <1)
                          P   Printino/Puatehmg [90)
                          B  Pulp and Paper PlantWood Products (18)
                          R  Rubber and Miscelianeaus Plashes Products (25)
                          2   Secondary Metal Processing <5)
                          >   Solid Waste Disposal - Commercial/Institutional i' i
                         V  S1eel Mill (9)
                          S   Surface Coaling (41)
                          TT  TBiecominunDcalwns (t9)
                          J   U.-Ki so Mi  i-j!
                         ^^  Transportalion Equipment (9)
                          ^  Transportalion and Marketing of Petroleum Products (€)
                          I   Wa-5tewate r Treatment {Q)
                         W  Woodworh, Furniture. MillwonV & Wood Preserving (3)
                                                       <*>  Fabricated Meial Products (53)
                                                       Cv>  Flexible Polyurethane Foam Production (5|i
                                                       F  Food Processifia'AgrKuhure (3S'|
                                                       [ _j  Fumilure Planl (10)
                                                       jf  Gaaoline;Diesel S?rvHe Stalcort iSi
                                                       V  G'ass Manufacturing (3)
                                                       fV  Heating Ewpment Msnufsetunng (2)
                                                       [J]  hi>-.ji-.=il if.;.
                                                       ^  Hot Mix A&phalt Plant (&>
                                                       -%:  Industrial Macnmery and Equipment (33)
                                                      ^  tnsbtulional - school (30)
                                                       I   Iron and Steel Foundry (2)
                                                       ^  Laboratory (1)
                                                       A  Landfill*?)
                                                       |_   Latga Appliance Manufactufing 41)
                                                       X  Mine/Ouarry (10>
                                                       ^T  Mineral Produced)
                                                       5  Miscellaneous Coating ManuFacluring (1)
                                                       ?  Miscellaneous Cwnmercia'JInduslnal [571
                                                                    11-4

-------
                               Table 11-1. Geographical Information for the Illinois Monitoring Sites

Site
Code
NBIL


SPIL



AQS Code
17-031-4201


17-031-3103



Location
Northbrook


Schiller
Park



County
Cook
County


Cook
County

Micro- or
Metropolitan
Statistical Area
Chicago-Joliet-
Naperville, IL-IN-
WIMSA
(Chicago Div)
Chicago-Joliet-
Naperville, IL-IN-
WIMSA
(Chicago Div)
Latitude
and
Longitude
42.139996,
-87.799227


41.965193,
-87.876265



Land Use
Residential


Mobile


Location
Setting
Suburban


Suburban



Additional Ambient Monitoring Information1
TSP, TSP Metals, CO, Hg, SO2, NO, NO2, NOX, NH3,
PAMS, O3, Meteorological parameters, PM10, PM2.5,
PM2s Speciation.


TSP, TSP Metals, CO, NO, NO2, NOX, Meteorological
parameters, PM2 5.

BOLD ITALICS = EPA-designated NATTS Site

-------
       NBIL is located on the property of the Northbrook Water Filtration Station. Figure 11-1
shows that NBIL is located off State Highway 68, Dundee Road, near Exit 30 on 1-94 (the clover
leaf of which is located on the lower right hand side of Figure 11-1). A railway runs north-south
in front of the water filtration station and intersects Dundee Readjust south of the monitoring
site. The surrounding area is classified as suburban and residential. Commercial, residential, and
forested areas are nearby, as well as a country club  and golf course. The NBIL monitoring site is
the Chicago NATTS site.

       SPIL is located on the eastern edge of the Chicago-O'Hare International Airport between
Mannheim Road and 1-294, just north of the toll plaza. The nearest runway is less than  1/2 mile
from the site. The surrounding area is classified as suburban and mobile. Commercial and
residential areas are nearby and a railyard is located to the east of 1-294.

       NBIL and SPIL are located within approximately 12 miles of each other. Each site is
located within 10 miles of numerous point sources, although the quantity of emissions sources is
higher near SPIL than NBIL, as shown in Figure 11-3. The source categories with the largest
number of sources within 10 miles of the Illinois monitoring sites are printing and publishing;
fabricated metal products; electroplating, plating, polishing, anodizing, and coloring; and dry
cleaning. Few point sources are located within 2 miles of NBIL, with most of the sources located
farther west or south. The closest source to NBIL is plotted under the symbol for the site in
Figure 11-3; this source is a dry cleaning facility. Besides the airport, the closest point source to
SPIL is involved in electroplating, plating, polishing,  anodizing, and coloring.

       Table 11-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Illinois monitoring sites. Table  11-2 includes county-level
population and vehicle registration information. Table 11-2 also includes a county-level vehicle
registration-to-population ratio, which was calculated to represent the number of vehicles per
person within each monitoring site's residing  county.  In addition, the population within 10 miles
of each site is presented, based on postal code population data estimates. An estimate of 10-mile
vehicle ownership was then determined by applying the county-level vehicle registration-to-
population ratio to the 10-mile population surrounding each monitoring site.  Table 11-2 also
contains traffic volume information for each site. Finally, Table 11-2 presents the county-level
daily VMT for Cook County.
                                           11-6

-------
Table 11-2. Population, Motor Vehicle, and Traffic Information for the Illinois Monitoring
                                          Sites
Site
NBIL
SPIL
Estimated
County
Population1
5,217,080
County-level
Vehicle
Registration2
2,072,399
Vehicles per
Person
(Registration:
Population)
0.40
Population
within 10
miles3
890,037
2,028,028
Estimated
10-mile
Vehicle
Ownership
353,553
805,601
Annual
Average
Daily
Traffic4
34,600
190,000
County-
level Daily
VMT5
86,863,779
  Bounty-level population estimates reflect 2011 data from the U.S. Census Bureau (Census Bureau, 2012b)
  2County-level vehicle registration reflects 2011 data from the IL Secretary of State (IL SOS, 2011)
  310-mile population estimates reflect 2011 data from Xionetic (Xionetic, 2011)
  4AADT reflects 2010 data for SPIL and 2011 data for NBIL from the Illinois DOT (IL DOT, 2010 and 201 la)
  5County-level VMT reflects 2011 data from the Illinois DOT (IL DOT, 201 Ib)
  BOLD ITALICS = EPA-designated NATTS  Site
       Observations from Table 11-2 include the following:

       •  Cook County has the second highest county-level population (behind Los Angeles
          County) and fourth highest county-level vehicle registration (behind Los Angeles
          County, CA; Maricopa County, AZ; and Harris County, TX) compared to other
          counties with NMP sites.

       •  The vehicle-per-person ratio for these sites is the third lowest compared to other NMP
          sites.

       •  The 10-mile radius population and estimated vehicle ownership are much higher near
          SPIL than NBIL.

       •  SPIL experiences a higher annual average daily traffic volume than NBIL. SPIL's
          traffic volume is the fourth highest among all NMP sites, behind ELNJ,  CELA, and
          SEW A, while the traffic volume for NBIL is in the middle of the range among NMP
          sites. Traffic data for SPIL is provided for 1-294 at Lawrence Avenue; traffic data for
          NBIL is for Dundee Road near the railroad crossing.

       •  The Cook County daily VMT ranks third among counties with NMP sites, behind
          only Los Angeles County, CA and Maricopa County, AZ (where VMT data were
          available).


11.2   Meteorological Characterization

       The following  sections characterize the meteorological conditions near the monitoring

sites in Illinois on sample days,  as well as over the course of the year.
                                           11-7

-------
11.2.1  Climate Summary
       Daily weather fluctuations are common for the Chicago area. The proximity of Chicago
to Lake Michigan offers moderating effects from the continental climate of the region. In the
summertime, afternoon lake breezes can cool the city when winds from the south and southwest
push temperatures upward. In the winter, 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" (Bair, 1992).

11.2.2  Meteorological Conditions in 2011
       Hourly meteorological data from the NWS weather stations nearest these sites were
retrieved for 2011 (NCDC, 2011). The two closest weather stations are located at Palwaukee
Municipal Airport (near NBIL) and O'Hare International Airport (near SPIL), WBAN 04838 and
94846, respectively. Additional information about the Palwaukee and O'Hare weather stations,
such as the distance between the  sites and the weather stations, is provided in Table 11-3. These
data were used to determine how meteorological conditions on sample days vary from conditions
experienced throughout the year.

       Table  11-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature,  average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for the entire year for 2011. Also included in  Table 11-3  is the
95 percent confidence interval for each parameter. As shown in Table 11-3, temperatures on
sample days appear cooler than temperatures over the course of the year at both sites. For NBIL,
the difference may be attributable to make-up samples. More make-up samples were collected
during  the first and fourth (and colder) quarters of the year compared to those collected during
the second and third (and warmer) quarters of the year. Only two make-up samples were
collected at SPIL, both during December.
                                         11-8

-------
                   Table 11-3. Average Meteorological Conditions near the Illinois Monitoring Sites
Closest NWS
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site


Average
Type1
Average
Maximum
Temperature
(ฐF)

Average
Temperature
(ฐF)
Average
Dew Point
Temperature
(ฐF)
Average
Wet Bulb
Temperature
(ฐF)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Northbrook, Illinois - NBIL
Palwaukee
Municipal
Airport
04838
(42.12, -87.91)
5.27
miles

250ฐ
(WSW)
Sample
Day


2011
57.0
ฑ5.0


58.7
ฑ2.1
49.4
ฑ4.6


51.0
ฑ2.0
38.7
ฑ4.4


40.5
ฑ1.9
44.3
ฑ4.1


45.9
ฑ1.8
68.8
ฑ2.4


69.4
ฑ1.1
1016.8
ฑ1.5


1015.9
ฑ0.7
6.5
ฑ0.7


6.5
ฑ0.3
Schiller Park, Illinois - SPIL
O'Hare
International
Airport
94846
(41.99, -87.91)
2.32
miles

303ฐ
(WNW)
Sample
Day


2011
57.3
ฑ5.4


58.5
ฑ2.2
49.6
ฑ5.0


50.9
ฑ2.0
39.3
ฑ4.5


40.7
ฑ1.9
44.5
ฑ4.4


45.8
ฑ1.8
70.0
ฑ2.7


70.4
ฑ1.2
1016.0
ฑ1.6


1015.4
ฑ0.7
8.2
ฑ0.9


8.2
ฑ0.3
Sample day averages are highlighted to help differentiate the sample day averages from the full-year averages.

-------
11.2.3 Back Trajectory Analysis
       Figure  11-4 is the composite back trajectory map for days on which samples were
collected at the NBIL monitoring site in 2011. Included in Figure 11-4 are four back trajectories
per sample day. Figure 11-5 is the corresponding cluster analysis. Similarly, Figure 11-6 is the
composite back trajectory map for days on which samples were collected at SPIL and
Figure 11-7 is the corresponding cluster analysis. An in-depth description of these maps and how
they were generated is presented in  Section 3.5.2.1. For the composite maps, each line represents
the 24-hour trajectory along which a parcel of air traveled toward the monitoring site on a given
sample day and time, based on an initial height of 50 meters AGL. For the cluster analyses, each
line corresponds to a trajectory representative of a given cluster of back trajectories. Each
concentric circle around the sites in Figures 11-4 through 11-7 represents 100 miles.

               Figure 11-4. 2011 Composite Back Trajectory Map for NBIL
                                          11-10

-------
    Figure 11-5. Back Trajectory Cluster Map for NBIL
Figure 11-6. 2011 Composite Back Trajectory Map for SPIL
                         11-11

-------
            Figure 11-7. Back Trajectory Cluster Map for SPIL
Observations from Figures 11-4 through 11-7 include the following:

•  The composite back trajectory maps for NBIL and SPIL are similar to each other in
   back trajectory distribution, which is expected given their proximity to each other.

•  Back trajectories originated from a variety of directions at the sites, with the longest
   trajectories originating from the northwest and north. The predominant direction of
   trajectory origin appears to be from the northwest and north.

•  The 24-hour air shed domains for NBIL and SPIL were among the largest in size
   compared to other NMP sites. The longest back trajectory for each site was greater
   than 1,000 miles in length and originated over Montana. These two trajectories are
   the longest back trajectories computed among NMP sites. However, the average back
   trajectory length for these sites is approximately 270 miles and greater than
   80 percent of back trajectories originated within 400 miles of the sites.

•  The cluster map for NBIL is similar to the cluster map for SPIL in geographical
   distribution of the clusters but the percentage of trajectories representing each cluster
   varies. Sixty-two percent of back trajectories originated from a direction with a
   westerly component for each site. The trajectories originating over Iowa (or due west)
   were grouped with trajectories originating from the northwest for NBIL and with
   trajectories originating from the southwest for SPIL.
                                   11-12

-------
       •  Thirty-nine percent of back trajectories originated from a direction with an easterly
          component for each site. For NBIL, all back trajectories originating from the north
          and northeast are grouped together, regardless of the length of trajectory, and are
          represented by the cluster trajectory over Lake Superior (16 percent). Back
          trajectories originating from the east, southeast, and south were grouped together, as
          represented by the short cluster originating over Indiana (23 percent). For SPIL, the
          model grouped short trajectories with an easterly component together under the short
          cluster trajectory originating from the east (32 percent), while grouping only the
          longer trajectories from the north and northeast together under the longer cluster
          trajectory originating to the northeast (7 percent).

11.2.4 Wind Rose Comparison
       Hourly surface wind data from the NWS weather stations at Palwaukee Municipal
Airport (for NBIL) and O'Hare International Airport (for SPIL) were uploaded into a wind rose
software program to produce customized wind roses, as described in Section 3.5.2.2. A wind
rose shows the frequency of wind directions using "petals" positioned around a 16-point
compass, and uses different colors to represent wind speeds.

       Figure 11-8 presents a map showing the distance between the NWS station and NBIL,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 11-8 also presents three different wind roses for the
NBIL monitoring site. First, a historical wind rose representing 2001 to 2010 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 2011 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2011 is
presented.  These can be used to identify the predominant wind speed and direction for 2011 and
determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically. Figure 11-9 presents the distance map and three wind roses
for SPIL.
                                          11-13

-------
Figure 11-8. Wind Roses for the Palwaukee Municipal Airport Weather Station near NBIL
   Distance between NBIL and NWS Station
                                                  2001-2010 Historical Wind Rose
 X

 ^^X,
_ '
                                    ซ-


                           vtKliJJBes <*ปnr& *c5ปaCiN Im
                           _ :

            2011 Wind Rose
                                                      Sample Day Wind Rose
                                        11-14

-------
Figure 11-9. Wind Roses for the O'Hare International Airport Weather Station near SPIL
  Distance between SPIL and NWS Station
2001-2010 Historical Wind Rose
                                              WEST;
           2011 Wind Rose
    Sample Day Wind Rose
                                       11-15

-------
       Observations from Figure 11-8 for NBIL include the following:

       •  The Palwaukee Municipal Airport weather station is located approximately 5.3 miles
          west-southwest of NBIL.

       •  The historical wind rose shows that winds from a variety of directions were observed
          near NBIL, although winds from the south, south-southwest, and west accounted for
          one-quarter of wind observations. Winds from the east-southeast to south-southeast
          were observed the least often. Calm winds (<2 knots) were observed for
          approximately 15 percent of the hourly measurements.

       •  The 2011 wind rose exhibits similar patterns in wind directions as the historical wind
          rose, indicating that wind conditions in 2011 were similar to what is experienced
          historically.

       •  The sample day wind patterns generally resemble the full-year wind patterns,
          although there were fewer westerly wind observations.


       Observations from Figure 11-9 for SPIL include the following:

       •  The O'Hare International Airport weather station is located 2.3 miles west-northwest
          of SPIL. The bulk of the  airport property lies between the weather station and the
          monitoring site.

       •  The historical wind rose for SPIL shows that winds from a variety of directions were
          observed, although winds from the south to  southwest to west account for the highest
          percentage of observations (greater than 40 percent). Winds from these directions also
          tended to be the strongest. Winds from the southeast quadrant were observed the
          least. Calm winds (< 2 knots) were observed for less than 10 percent of the hourly
          measurements.

       •  The 2011 wind rose exhibits similar patterns in wind directions as the historical wind
          rose, although fewer wind observations from the southwest to west appear to be
          reflected in wind  observations from the northeast quadrant.

       •  The sample day wind patterns exhibit a more even distribution of wind observations,
          with most directions accounting for five to seven percent of the wind observations
          (with the exception of east-southeast and west-northwest).


11.3   Pollutants of Interest

       Site-specific "pollutants of interest" were determined for the Illinois monitoring sites in

order to allow 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

                                          11-16

-------
pollutant's total failed screens contribute to the top 95 percent of the site's total failed screens. In
addition, if any of the NATTS MQO Core Analytes measured by each monitoring site did not
meet the pollutant of interest criteria based on the preliminary risk-based screening, that pollutant
was added to the list of site-specific pollutants of interest. A more in-depth description of the
risk-based screening process is presented in Section 3.2.

       Table 11-4 presents the results of the preliminary risk-based screening process. The
pollutants that failed at least one screen and contributed to 95 percent of the total failed screens
for each monitoring site are shaded. NATTS MQO Core Analytes are bolded. Thus, pollutants of
interest are shaded and/or bolded. NBIL sampled for VOCs, carbonyl compounds, SNMOCs,
metals (PMi0), PAHs, and hexavalent chromium, and is one of only two NMP sites sampling for
all six pollutant groups. SPIL sampled for VOCs and carbonyl compounds  only.
         Table 11-4. Risk-Based Screening Results for the Illinois Monitoring Sites
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
Northbrook, Illinois - NBIL
Formaldehyde
Acet aldehyde
Benzene
Carbon Tetrachloride
Naphthalene
Arsenic (PM10)
1,3-Butadiene
Acrylonitrile
Manganese (PM10)
Fluorene
Acenaphthene
1 ,2-Dichloroethane
Fluoranthene
Chloroform
Ethylbenzene
Dichloromethane
/>-Dichlorobenzene
Hexachloro- 1 , 3 -butadiene
Nickel (PM10)
Benzo(a)pyrene
Lead (PM10)
Trichloroethylene
0.077
0.45
0.13
0.17
0.029
0.00023
0.03
0.015
0.005
0.011
0.011
0.038
0.011
9.8
0.4
7.7
0.091
0.045
0.0021
0.00057
0.015
0.2
62
60
55
55
53
46
37
35
31
21
20
18
15
13
9
6
4
4
4
2
2
2
62
62
55
55
61
53
41
35
53
61
61
18
61
55
55
55
18
4
53
59
53
24
100.00
96.77
100.00
100.00
86.89
86.79
90.24
100.00
58.49
34.43
32.79
100.00
24.59
23.64
16.36
10.91
22.22
100.00
7.55
3.39
3.77
8.33
11.15
10.79
9.89
9.89
9.53
8.27
6.65
6.29
5.58
3.78
3.60
3.24
2.70
2.34
1.62
1.08
0.72
0.72
0.72
0.36
0.36
0.36
11.15
21.94
31.83
41.73
51.26
59.53
66.19
72.48
78.06
81.83
85.43
88.67
91.37
93.71
95.32
96.40
97.12
97.84
98.56
98.92
99.28
99.64
                                          11-17

-------
Table 11-4. Risk-Based Screening Results for the Illinois Monitoring Sites (Continued)
Pollutant
1 , 1 ,2,2-Tetrachloroethane
Xylenes
Screening
Value
(Ug/m3)
0.017
10
Total
#of
Failed
Screens
1
1
556
#of
Measured
Detections
1
55
1110
%of
Screens
Failed
100.00
1.82
50.09
% of Total
Failures
0.18
0.18
Cumulative
%
Contribution
99.82
100.00

Schiller Park, Illinois - SPIL
Acet aldehyde
Formaldehyde
Benzene
Carbon Tetrachloride
1,3-Butadiene
Acrylonitrile
Trichloroethylene
p-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Hexachloro- 1 ,3 -butadiene
1 , 1 ,2,2-Tetrachloroethane
Carbon Bisulfide
Chloromethylbenzene
1 ,2-Dibromoethane
Propionaldehyde
0.45
0.077
0.13
0.17
0.03
0.015
0.2
0.091
0.038
0.4
0.045
0.017
70
0.02
0.0017
0.8
Total
62
62
56
56
55
47
21
15
15
7
3
2
1
1
1
1
405
62
62
56
56
55
47
42
33
15
55
3
2
56
1
1
62
608
100.00
100.00
100.00
100.00
100.00
100.00
50.00
45.45
100.00
12.73
100.00
100.00
1.79
100.00
100.00
1.61
66.61
15.31
15.31
13.83
13.83
13.58
11.60
5.19
3.70
3.70
1.73
0.74
0.49
0.25
0.25
0.25
0.25
15.31
30.62
44.44
58.27
71.85
83.46
88.64
92.35
96.05
97.78
98.52
99.01
99.26
99.51
99.75
100.00

    Observations from Table 11-4 include the following:

    •   Twenty-four pollutants, including 13 NATTS MQO Core Analytes, failed screens for
        NBIL. Approximately 50 percent of the measured detections of these pollutants failed
        screens.

    •   Based on the risk-based screening process, 15 pollutants, of which nine are NATTS
        MQO Core Analytes, were identified as pollutants of interest for NBIL. Four
        additional NATTS MQO Core Analytes (trichloroethylene, nickel, lead, and
        benzo(a)pyrene) were added to the pollutants of interest for NBIL, even though they
        did not contribute to 95 percent of the failed screens for NBIL. Five additional
        pollutants were added to the pollutants of interest for NBIL because they are NATTS
        MQO Core Analytes, even though they did not fail any screens (beryllium, cadmium,
        hexavalent chromium, tetrachloroethylene, and vinyl chloride). These five pollutants
        are not shown in Table 11-4 but are shown in subsequent tables in the sections that
        follow.

    •   Benzene, carbon tetrachloride, and formaldehyde were detected in every VOC or
        carbonyl compound sample collected at NBIL and failed 100 percent of screens.
        Additional pollutants also failed  100 percent of screens, but the detection rate was
        lower (such as acrylonitrile and 1,2-dichloroethane).
                                       11-18

-------
       •  Sixteen pollutants, including six NATTS MQO Core Analytes, failed screens for
          SPIL.
       •  Based on the risk-based screening process, nine pollutants, of which six are NATTS
          MQO Core Analytes, were identified as pollutants of interest for SPIL. Three
          additional NATTS MQO Core Analytes were added to the pollutants of interest for
          SPIL, even though they did not fail any screens (chloroform, tetrachloroethylene, and
          vinyl chloride). These pollutants are not shown in Table 11-4 but are shown in
          subsequent tables in the sections that follow.
       •  Acetaldehyde, benzene, carbon tetrachloride, and formaldehyde were detected in
          every VOC and carbonyl compound sample collected at SPIL and failed 100 percent
          of their screens. Other pollutants also failed 100 percent of screens, but the detection
          rate was lower.
       •  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.

11.4   Concentrations
       This section presents various concentration averages used to characterize pollution levels
at the Illinois monitoring sites. Where applicable, the following calculations and data analyses
were performed: Time period-based concentration averages (quarterly and annual) are provided
for the pollutants of interest for the Illinois monitoring sites, where the data meet the applicable
criteria. Concentration averages for select pollutants are also presented graphically for the sites to
illustrate how the sites' concentrations compare to the program-level  averages, as presented in
Section 4.1.  In addition, concentration averages for select pollutants are presented from previous
years of sampling in order to characterize concentration trends at the  sites. Additional site-
specific statistical summaries for NBIL and SPIL are provided in Appendices J through O.

11.4.1 2011 Concentration Averages
       Quarterly and annual concentration averages were calculated for the pollutants of interest
for each Illinois site, as described in Section 3.1. The quarterly average of a particular pollutant
is simply the average concentration of the preprocessed daily measurements over a given
calendar quarter. Quarterly average concentrations include the substitution of zeros for all non-
detects. A site must have a minimum  of 75 percent valid samples of 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
                                          11-19

-------
sampling. Annual averages were calculated for pollutants where three valid quarterly averages
could be calculated and where method completeness was greater than or equal to 85 percent, as
presented in Section 2.4. Quarterly and annual average concentrations for the Illinois monitoring
sites are presented in Table 11-5, where applicable. Note that concentrations of the PAHs,
metals,  and hexavalent chromium 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 11-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest
                              for the Illinois Monitoring Sites
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Ug/m3)
2nd
Quarter
Average
(Ug/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Ug/m3)
Northbrook, Illinois - NBIL
Acetaldehyde
Acrylonitrile
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
Acenaphthene a
Arsenic (PM10)a
62/62
35/55
55/55
41/55
55/55
55/55
18/55
55/55
62/62
53/55
24/55
8/55
61/61
53/53
1.00
ฑ0.25
0.14
ฑ0.05
0.99
ฑ0.56
0.29
ฑ0.39
0.56
ฑ0.05
2.51
ฑ 1.17
0.01
ฑ0.02
0.20
ฑ0.09
2.49
ฑ1.82
0.31
ฑ0.31
0.04
ฑ0.03
<0.01
ฑ<0.01
2.63
ฑ0.76
NA
1.08
ฑ0.32
0.21
ฑ0.04
0.55
ฑ0.10
0.03
ฑ0.02
0.64
ฑ0.06
12.41
ฑ4.97
0.01
ฑ0.02
0.45
ฑ0.41
2.04
ฑ0.95
0.57
ฑ0.28
0.04
ฑ0.04
<0.01
ฑ<0.01
10.73
ฑ5.57
0.68
ฑ0.29
1.60
ฑ0.35
NA
NA
NA
NA
NA
NA
NA
3.18
ฑ0.84
NA
NA
NA
29.14
ฑ9.26
0.80
ฑ0.29
1.87
ฑ0.52
0.01
ฑ0.03
0.69
ฑ0.16
0.08
ฑ0.04
0.69
ฑ0.06
3.79
ฑ4.46
0.07
ฑ0.02
0.26
ฑ0.08
2.39
ฑ1.10
0.18
ฑ0.06
0.04
ฑ0.03
<0.01
ฑ0.01
7.79
ฑ7.07
0.89
ฑ0.36
1.41
ฑ0.20
0.13
ฑ0.03
0.71
ฑ0.16
0.12
ฑ0.10
0.64
ฑ0.03
6.06
ฑ2.17
0.03
ฑ0.01
0.32
ฑ0.12
2.53
ฑ0.57
0.35
ฑ0.13
0.04
ฑ0.02
<0.01
ฑ<0.01
13.14
ฑ4.14
0.73
ฑ0.15
      NA = Not available due to the criteria for calculating a quarterly and/or annual average
      a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of
      viewing.
                                           11-20

-------
Table 11-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest
                      for the Illinois Monitoring Sites (Continued)
Pollutant
Benzo(a)pyrene a
Bery Ilium (PM10)a
Cadmium (PM10)a
Fluoranthene a
Fluorene a
Hexavalent Chromium3
Lead(PM10)a
Manganese (PM10) a
Naphthalene a
Nickel (PM10)a
#of
Measured
Detections
vs. # of
Samples
59/61
51/53
53/53
61/61
61/61
50/61
53/53
53/53
61/61
53/53
1st
Quarter
Average
(Ug/m3)
0.21
ฑ0.10
NA
NA
2.02
ฑ0.56
2.94
ฑ0.64
0.02
ฑ0.01
NA
NA
48.33
ฑ 13.63
NA
2nd
Quarter
Average
(Ug/m3)
0.14
ฑ0.05
0.01
ฑO.01
0.13
ฑ0.06
6.94
ฑ3.46
10.94
ฑ5.95
0.02
ฑ0.01
2.95
ฑ1.31
8.73
ฑ5.52
54.35
ฑ 19.54
1.20
ฑ0.19
3rd
Quarter
Average
(Ug/m3)
0.08
ฑ0.02
O.01
ฑO.01
0.12
ฑ0.05
17.28
ฑ5.24
30.24
ฑ9.82
0.01
ฑO.01
3.44
ฑ1.34
7.66
ฑ2.70
157.13
ฑ45.66
1.35
ฑ0.47
4th
Quarter
Average
(Ug/m3)
0.16
ฑ0.07
0.01
ฑO.01
0.19
ฑ0.07
2.78
ฑ1.74
7.33
ฑ6.21
0.02
ฑ0.01
6.46
ฑ3.34
11.45
ฑ6.47
127.06
ฑ103.17
1.46
ฑ0.59
Annual
Average
(Ug/m3)
0.15
ฑ0.03
0.01
ฑO.01
0.14
ฑ0.03
7.59
ฑ2.29
13.46
ฑ4.25
0.02
ฑO.01
4.16
ฑ1.07
8.30
ฑ2.29
99.39
ฑ29.75
1.27
ฑ0.21
Schiller Park, Illinois - SPIL
Acetaldehyde
Acrylonitrile
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
ฃ>-Dichlorobenzene
1 ,2-Dichloroethane
Formaldehyde
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
62/62
47/56
56/56
55/56
56/56
38/56
33/56
15/56
62/62
53/56
42/56
6/56
5.29
ฑ2.04
0.95
ฑ0.19
0.77
ฑ0.11
0.13
ฑ0.04
0.50
ฑ0.04
0.05
ฑ0.02
0.04
ฑ0.04
0.01
ฑ0.01
3.89
ฑ1.13
0.29
ฑ0.22
0.15
ฑ0.15
0.01
ฑ0.01
2.26
ฑ0.51
0.83
ฑ0.14
0.82
ฑ0.25
0.15
ฑ0.07
0.60
ฑ0.05
0.11
ฑ0.04
0.08
ฑ0.06
0.04
ฑ0.03
3.16
ฑ0.97
0.32
ฑ0.21
0.67
ฑ0.52
0.01
ฑ0.01
2.38
ฑ0.40
0.85
ฑ0.11
0.78
ฑ0.18
0.15
ฑ0.06
0.60
ฑ0.02
0.10
ฑ0.05
0.07
ฑ0.04
0.01
ฑ0.02
3.80
ฑ0.80
0.25
ฑ0.07
0.39
ฑ0.31
0.01
ฑ0.01
1.95
ฑ0.66
0.46
ฑ0.29
1.02
ฑ0.35
0.20
ฑ0.09
0.64
ฑ0.05
0.10
ฑ0.07
0.10
ฑ0.08
0.03
ฑ0.02
2.35
ฑ0.74
0.36
ฑ0.12
1.28
ฑ1.44
0.01
ฑ0.01
2.94
ฑ0.61
0.77
ฑ0.11
0.85
ฑ0.12
0.16
ฑ0.03
0.58
ฑ0.02
0.09
ฑ0.02
0.07
ฑ0.03
0.02
ฑ0.01
3.29
ฑ0.46
0.31
ฑ0.08
0.64
ฑ0.41
0.01
ฑ0.01
    NA = Not available due to the criteria for calculating a quarterly and/or annual average
    a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of
    viewing.
                                          11-21

-------
Observations for NBIL from Table 11-5 include the following:

•  The pollutants with the highest annual average concentrations by mass are chloroform
   (6.06 ฑ 2.17 |ig/m3), formaldehyde (2.53 ฑ 0.57 |ig/m3), and acetaldehyde
   (1.41 ฑ 0.20 |ig/m3). The annual average concentrations for the remaining pollutants
   of interest were less than 1.00 |ig/m3.

•  Third quarter average concentrations could not be calculated for the VOCs because
   fewer than 75 percent of samples were valid during this quarter.  For this same reason,
   first quarter average concentrations could not be calculated for the PMio metals.

•  The annual average concentration of chloroform for NBIL is unusually high
   compared to other sites sampling this pollutant. The second quarter average is
   considerably higher than the other available quarterly average concentrations and the
   confidence interval associated with the fourth quarter average concentration is greater
   than the quarterly average itself. These indicate that the measurements of chloroform
   are highly variable and that outliers are likely present. Chloroform concentrations
   range from 0.308 to 32.9 |ig/m3, with a median concentration of 2.05 |ig/m3. The
   maximum concentration of chloroform was measured at NBIL on December 2, 2011
   (32.9 |ig/m3), although a similar concentration was also measured June 26, 2011
   (32.8 |ig/m3). These are the highest concentrations of chloroform measured across the
   program and of the 12 concentrations greater than 10  |ig/m3 measured across the
   program, 11 were measured at NBIL. Of the 63 measurements of chloroform greater
   than 1 |ig/m3,44 were measured at NBIL (with 16 measured at DEMI, one at PXSS,
   and two at S4MO). These findings are consistent with findings discussed in previous
   NMP reports.

•  The first quarter average concentrations for both benzene and 1,3-butadiene are the
   highest of the quarterly averages and have relatively large confidence intervals
   compared to the other quarterly averages, particularly 1,3-butadiene. A review of the
   data shows that the two highest concentrations of benzene and 1,3-butadiene were
   measured on the first two sample dates in January. The maximum concentration of
   benzene was measured on January 9, 2011 (4.51 |ig/m3) and the  second highest
   concentration was measured on January 3, 2011 (2.15 |ig/m3). The maximum
   concentration of 1,3-butadiene was measured on January 9, 2011 (2.68 |ig/m3) and
   the second highest concentration was measured on January 3, 2011 (1.21 |ig/m3).
   NBIL's maximum 1,3-butadiene concentrations rank second and fourth among NMP
   sites sampling this pollutant and are the only two 1,3-butadiene concentrations
   measured at NBIL greater than 1 |ig/m3. For benzene, these two  concentrations are
   the only concentrations measured at NBIL greater than 2 |ig/m3 (and only four
   benzene concentrations measured at NBIL are greater than 1 jig/m3).

•  The second quarter average concentration of ethylbenzene is roughly twice as high as
   the other quarterly averages, where they could be calculated, and has a relatively
   large confidence interval associated with it. The maximum ethylbenzene
   concentration was measured on June 8, 2011 (3.20 |ig/m3) and is more than twice the
   next highest concentration  (1.18 |ig/m3, measured on August 25, 2011). All other
   measurements of this pollutant are less than 1 |ig/m3. The third highest concentration
                                   11-22

-------
   of ethylbenzene was measured on January 9, 2011, the same day the highest
   concentrations of benzene and 1,3-butadiene were measured.

•  The first quarter average concentration of tetrachloroethylene is equivalent to its
   confidence interval, indicating that outliers may be affecting this quarterly average.
   The confidence interval for the second quarter average is also relatively large
   compared to the average itself. A review of the data shows that concentrations of
   tetrachloroethylene range from 0.068 |ig/m3to 2.42 |ig/m3, with a median
   concentration of 0.177 |ig/m3. The maximum concentration was measured on
   March 16, 2011 (2.42 |ig/m3), although a similar concentration was also measured on
   May 21, 2011 (2.18 |ig/m3). These two tetrachloroethylene measurements rank
   second and third highest among all NMP sites sampling VOCs.

•  The third quarter average concentrations of acenaphthene, fluoranthene, and fluorene
   were significantly higher than the other quarterly averages. The maximum
   concentration of each of these pollutants was measured on July 5, 2011. The highest
   concentrations of these pollutants were measured in June, July, August, and October.

•  The third and fourth quarter average concentrations of naphthalene were more than
   twice the first and second quarterly  averages. The highest concentrations of
   naphthalene were measured in July, August,  and October. The maximum naphthalene
   concentration (799 ng/m3) was measured at NBIL on October 6, 2011, with the
   second highest concentration (322 ng/m3) measured on the following sample day
   (October 12, 2011). The concentration measured on October 6, 2011 is the maximum
   naphthalene concentration measured across all NMP sites sampling PAHs.

•  Among the PMi0 metals, manganese has the  highest annual average concentration
   (8.30 ฑ  2.29 ng/m3). The quarterly average concentrations for most of the PMio
   metals are highest for the fourth quarter, although the difference is not statistically
   significant among the quarters for most of them.  This is most notable for lead and
   manganese. A review of the highest concentrations for each of the metals reveals that
   some of the highest concentrations of the metals were measured in the sample
   collected October 6, 2011.

Observations for SPIL from Table 11-5 include the following:

•  The pollutants with the highest annual average concentrations by mass  are
   formaldehyde (3.29 ฑ 0.46 |ig/m3) and acetaldehyde (2.94 ฑ 0.61  |ig/m3). These are
   the only pollutants with annual average concentrations greater than  1 |ig/m3.

•  Concentrations of most of the pollutants of interest for SPIL did not vary  significantly
   across calendar quarters. However,  a few quarterly averages do stand out, as
   described below.

•  The first quarter average concentration of acetaldehyde is twice as high as the other
   quarterly averages  and has a relatively large confidence interval associated with it. A
   review of the data shows that seven of the eight highest concentrations  of
   acetaldehyde (those greater than 5 |ig/m3) were measured during the first quarter of
   2011, and ranged from 5.74 |ig/m3 (measured on January 21, 2011)  to 14.5 |ig/m3

                                  11-23

-------
          (measured on January 9, 2011). The two highest concentrations of acetaldehyde
          measured among NMP  sites sampling carbonyl compounds were measured at SPIL
          (14.5 |ig/m3andll.2|ig/m3).

       •  The fourth quarter average concentration of trichloroethylene is higher than the other
          quarters and all four quarterly averages of this pollutant have relatively large
          confidence intervals associated with them. This indicates that the concentrations of
          trichloroethylene are highly variable. A review of the data shows that the highest
          trichloroethylene concentration was measured on October 12, 2011 (8.40 |ig/m3) and
          is the highest trichloroethylene concentration measured among NMP sites sampling
          VOCs. Of the 11 concentrations  of trichloroethylene greater than 1 |ig/m3 across the
          program, nine of these were measured at SPIL. The two highest concentrations
          measured at SPIL (8.40 |ig/m3 and 7.22 |ig/m3) were both measured in October,
          leading to the relatively high fourth quarter average concentration. Trichloroethylene
          concentrations ranged from 0.0377 |ig/m3to  8.40 |ig/m3, with a median concentration
          of 0.145 |ig/m3 and 14 non-detects.


       Tables 4-9 through 4-12 present the  sites with the 10 highest annual average

concentrations for each of the program-level pollutants of interest. Observations for NBIL and

SPIL from those tables include the following:

       •  NBIL and SPIL appear  in Tables 4-9 through 4-12 a total of 27 times.

       •  As shown in Table 4-9,  NBIL's annual average concentration of chloroform is the
          highest annual average among all NMP sites sampling this pollutant. Further, the
          annual average concentration of chloroform for NBIL is nearly eight times higher
          than the next highest annual average (calculated for DEMI). This finding is similar to
          the findings in the 2010 NMP report, although the difference is significantly higher
          for 2011. NBIL also has the highest annual average concentration of vinyl chloride,
          even though it was only detected eight times.

       •  SPIL has the highest annual average concentration of trichloroethylene, which is
          seven times greater than the next highest annual average concentration of this
          pollutant (calculated for ELNJ).  This is also  similar to the 2010 NMP report.  SPIL
          also has the highest annual average concentration of acrylonitrile.

       •  SPIL has the second highest annual average concentration of acetaldehyde among
          NMP sites sampling carbonyl compounds, as shown in Table 4-10, behind only
          ELNJ.

       •  NBIL has the highest annual average concentration of fluorene and second highest
          annual average concentration of acenaphthene among NMP sites sampling PAHs, as
          shown in Table 4-11. Even though the maximum naphthalene concentration across
          the program was measured at NBIL, the annual average concentration of naphthalene
          for NBIL ranks sixth.
                                         11-24

-------
       •  As shown in Table 4-12, the annual average concentrations for NBIL were among the
          top five for all of the program-level PMio metal pollutants of interest. However, it is
          important to note that only nine sites sampled PMio metals and have enough data for
          annual averages to be calculated.

11.4.2 Concentration Comparison
       In order to better illustrate how a site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for the selected NATTS MQO Core
Analytes listed in Section 3.5.3, where applicable. Thus, box plots for acetaldehyde, benzene,
1,3-butadiene, and formaldehyde were created for both NBIL and SPIL. Box plots were also
created for arsenic, benzo(a)pyrene, hexavalent chromium, lead, manganese, and naphthalene for
NBIL. Figures 11-10 through 11-19 overlay the sites' minimum,  annual average, and maximum
concentrations onto the program-level minimum, first quartile, median, average, third quartile,
and maximum concentrations, as described in Section 3.5.3.
       Figure 11-10. Program vs. Site-Specific Average Acetaldehyde Concentrations
    E
                                        Concentration (^ig/m3)
                Program:   1st Quartile    2nd Quartile    3rd Quartile     4th Quartile     Average
                Site:     Site Average    Site Minimum/Maximum

                                          11-25

-------
     Figure 11-11. Program vs. Site-Specific Average Arsenic (PMio) Concentration
   •-H
                                        2        2.5         3
                                        Concentration (ng/m3)
               Program:   IstQuartile     2ndQuartile     3rdQuartile     4thQuartile     Average
               Site:      Site Average     Site Minimum/Maximum

        Figure 11-12. Program vs. Site-Specific Average Benzene Concentrations









Program Max Concentration = 23.8 ng/m3


SPIL
    •4
                                                           Program Max Concentration = 23.8 ug/m3
                                     4        5        6
                                        Concentration (ug/mi)
                                                                                       10
               Program:   IstQuartile     2ndQuartile     SrdQuartile     4thQuartile     Average
               Site:      Site Average     Site Minimum/Maximum

     Figure 11-13. Program vs. Site-Specific Average Benzo(a)pyrene Concentration
   E
NEIL
             0.25
                                  0.75         1         1.25
                                        Concentration (ng/mi)
                                                                  1.5
                                                                            1.75
               Program:   IstQuartile     2ndQuartile     SrdQuartile     4thQuartile     Average
               Site:      Site Average     Site Minimum/Maximum

                                          11-26

-------
      Figure 11-14. Program vs. Site-Specific Average 1,3-Butadiene Concentrations
NEIL
II
IF

_,
Program Max Concentration = 9. 51 (jg/ms i




                                                               Program Max Concentration = 9.51
                                                1.5
                                           Concentration (^
                Program:   IstQuartile     2ndQuartile     SrdQuartile      4thQuartile     Average
                Site:     Site Average     Site Minimum/Maximum

     Figure 11-15. Program vs. Site-Specific Average Formaldehyde Concentrations
   E
   id:
                                                IS
                                           Concentration
                Program:   IstQuartile     2ndQuartile     SrdQuartile     4thQuartile      Average
               Site:      Site Average     Site Minimum/Maximum

  Figure 11-16. Program vs. Site-Specific Average Hexavalent Chromium Concentration
                                               0.15
                                           Concentration (ng/m3)
                Program:   IstQuartile     2ndQuartile     SrdQuartile     4thQuartile      Average
               Site:      Site Average     Site Minimum/Maximum

                                            11-27

-------
       Figure 11-17. Program vs. Site-Specific Average Lead (PMi0) Concentration
NEIL
                           10
                                       15           20
                                         Concentration (ng/m3)
                                                                            3D
                                                                                        35
Program
Site:
: IstQuartile
Site Average
o
2ndQuartile SrdQuartile 4thQuartile Ave
• n
Site Minimum/Maximum
rage

    Figure 11-18. Program vs. Site-Specific Average Manganese (PMi0) Concentration
NEIL
                                                        ; Program Max Concentration = 395 ng/m3
                        50
                                   75
                                             100
                                         Concentration (ng/m3)
                                                        125
                                                                  150
                                                                             175
               Program:   IstQuartile    2ndQuartile     SrdQuartile     4thQuartile     Average
               Site:
                       Site Average    Site Minimum/Maximum

       Figure 11-19. Program vs. Site-Specific Average Naphthalene Concentration
NEIL
                                                           i Program Max Concentration = 779 ng/m3
           50
                    100
                            is;
                                     200       Z50       300
                                         Concentration (ng/m3)
                                                              350
                                                                               453
                                                                                       555
Program
Site:
: IstQuartile
Site Average
O
2ndQuartile SrdQuartile 4thQuartile Ave
Site Minimum/Maximum
rage

      Observations from Figures 11-10 through 11-19 include the following:

          •  Figure 11-10 shows that the annual average acetaldehyde concentration for SPIL
             is twice the annual average acetaldehyde concentration for NBIL. NBIL's annual
             average is less than both the program-level average and median concentrations
             while the annual average for SPIL is greater than the program-level average and
             third quartile. As discussed in the previous section, the two maximum
             acetaldehyde concentrations measured among NMP sites sampling this pollutant
             were measured at  SPIL. There were no non-detects of acetaldehyde measured at
             either site or across the program.
                                          11-28

-------
•  Figure 11-11 is the box plot for arsenic, which was measured at NBIL but not at
   SPIL. The box plot shows that the annual average concentration for NBIL is
   greater than the program-level average concentration but just less than the
   program-level third quartile. The maximum concentration measured at NBIL is
   less than the maximum concentration measured across the program. While a few
   non-detects of arsenic were measured among sites sampling PMio metals, none
   were measured at NBIL.

•  Figure 11-12 shows the box plots for benzene. Note that the program-level
   maximum concentration (23.8 |ig/m3) is not shown directly on the box plots
   because the scale of the box plots would be too large to readily observe data
   points at the lower end of the concentration range. Thus, the scale has been
   reduced to 10 |ig/m3.  Figure 11-12 shows that while SPIL's annual average
   benzene concentration is greater than NBIL's  annual average benzene
   concentration, the range of measurements is larger for NBIL than SPIL. However,
   both annual  averages are less than the program-level average concentration. There
   were no non-detects of benzene measured at either site or across the program.

•  Figure 11-13 is the box plot for benzo(a)pyrene. Note that the program-level first
   quartile for this pollutant is zero and is not visible on this box plot. This box plot
   shows that the annual average benzo(a)pyrene concentration for NBIL is greater
   than the program-level average concentration  and third quartile. Figure 11-13 also
   shows that the maximum concentration measured at NBIL is less than the
   maximum concentration measured across the program. Two non-detects of
   benzo(a)pyrene were measured at NBIL.

•  Similar to the benzene graph, the program-level maximum 1,3-butadiene
   concentration (9.51 |ig/m3) is not  shown directly on the box plots as the scale has
   been reduced to 3 |ig/m3 in Figure 11-14 to allow for the observation of data
   points at the lower end of the concentration range. Figure 11-14 shows that both
   sites'  annual average  1,3-butadiene concentrations are greater than the program-
   level average and, although difficult to discern in Figure 11-14, the program-level
   third quartile. As discussed in the previous section, the maximum 1,3-butadiene
   concentration for NBIL is the second highest among NMP sites sampling this
   pollutant. A single non-detect of 1,3-butadiene was measured at SPIL while
   14 non-detects were measured at NBIL.

•  Figure 11-15 presents the box plots for formaldehyde. The box plots show that
   NBIL's annual average formaldehyde concentration is less than the program-level
   average while SPIL's annual average formaldehyde concentration is greater than
   the program-level average  concentration. Figure 11-15 also shows that the range
   of concentrations measured is larger for NBIL than SPIL. There were no non-
   detects of formaldehyde measured at either site or across the program.
                               11-29

-------
•  Figure 11-16 shows that the annual average concentration of hexavalent
   chromium for NBIL is less than the program-level average and median
   concentrations. The maximum concentration measured at NBIL is less than the
   program-level maximum concentration. There were several non-detects of
   hexavalent chromium measured at NBIL and  across the program.

•  Figure 11-17 is the box plot for lead, which was measured at NBIL only. The box
   plot shows that the annual average concentration of lead for NBIL is greater than
   the program-level average but less than the program-level third quartile. While the
   maximum lead concentration measured at NBIL is less than the maximum
   concentration measured across the program, it was the sixth highest measurement
   of lead among NMP sites sampling PMio metals. There  were no non-detects of
   lead measured at NBIL or across the program.

•  Figure 11-18 is the box plot for manganese (PMio). Note that the program-level
   maximum concentration (395 ng/m3) is not shown directly on the box plot as the
   scale has been reduced to 200 ng/m3 to allow  for the observation of data points at
   the lower end of the concentration range. The box plot shows that the annual
   average concentration for NBIL is just less than the program-level average
   concentration. The maximum concentration measured at NBIL is considerably
   less than the maximum measured across the program. Although difficult to
   discern in Figure 11-18, there were no non-detects of manganese measured at
   NBIL or across the program.

•  Figure 11-19 is the box plot for naphthalene. The program-level maximum
   concentration (779 ng/m3) is not shown directly on the box plot as the scale has
   been reduced to 500 ng/m3 in order to allow for the observations of data points at
   the lower end of the concentration range. Figure  11-19 shows that the maximum
   concentration of naphthalene measured at NBIL is off the scale of the box plot;
   this measurement is the maximum naphthalene concentration measured among all
   NMP sites sampling this pollutant. The annual average concentration for NBIL is
   greater than the program-level average concentration but less than the program-
   level third quartile. There were no non-detects of naphthalene measured at NBIL
   or across the program.
                               11-30

-------
11.4.3  Concentration Trends
       A site-specific trends evaluation was completed for sites that have sampled one or more
of the selected NATTS MQO Core Analytes for 5 consecutive years or longer, as described in
Section 3.5.4. NBIL and SPIL have sampled VOCs under the NMP since 2003. Both sites have
also sampled carbonyl compounds since 2005. Additionally, NBIL has sampled PMio metals and
hexavalent chromium since 2005. Figures 11-20 through 11-27 present the annual statistical
metrics for acetaldehyde, arsenic, benzene,  1,3-butadiene, formaldehyde, hexavalent chromium,
lead, and manganese for NBIL, respectively. Figures 11-28 through 11-31 present the annual
statistical metrics for acetaldehyde, benzene, 1,3-butadiene, and formaldehyde for SPIL,
respectively. The statistical metrics presented for assessing trends include the substitution of
zeros for non-detects. NBIL began sampling PAHs under the NMP in 2008, which is less than
5 consecutive years; therefore, the trends analysis was  not conducted for the PAHs.

  Figure 11-20. Annual Statistical Metrics for Acetaldehyde Concentrations Measured at
                                         NBIL
                                       2008         2009
                                             Year
               • 5th Percentile    — Minimum    —  Median    — Maximum     • 95th Percentile   "^"Average
                                          11-31

-------
Figure 11-21. Annual Statistical Metrics for Arsenic (PMi0) Concentrations Measured at

                                            NBIL
      a

      1  2

      I
      3


      I.
                                                2008         2009


                                                Year
               •  5th Percentile    —  Minimum     — Median     — Maximum     • 95th Percentile   "^"Average
Figure 11-22. Annual Statistical Metrics for Benzene Concentrations Measured at NBIL
                  Bth Percentile    —  Minimum     — Median     — Maximum     • 95th Percentile   ..^.. Average
                                             11-32

-------
Figure 11-23. Annual Statistical Metrics for 1,3-Butadiene Concentrations Measured at
                                            NBIL
               —  Minimum     —  Median     - Maximurr
                                                    95th Percentile
                                                                                . Average
Figure 11-24. Annual Statistical Metrics for Formaldehyde Concentrations Measured at
                                            NBIL
      I
      e
      .9
      B
                    The maximum
                    concentration for
                    2006 is 91.7 Ug/m3
               *  5th Percentile    -  Minimum     —  Median     - Maximum     #  95th Percentile    ..+.. Average
                                             11-33

-------
 .Figure 11-25. Annual Statistical Metrics for Hexavalent Chromium Concentrations
                                  Measured at NBIL
              —  Minimum     — Median     -  Maximun
                                                 95th Percentile
                                                                5th Percentile    "^"Average
Figure 11-26. Annual Statistical Metrics for Lead (PMio) Concentrations Measured at
                                         NBIL
     a
     S 15
     s
     i
                                             zoos
                                             Year
              *  5th Percentile    - Minimum     —  Median     - Maximum
                                                               95th Percentile
                                                                            • Average
                                          11-34

-------
Figure 11-27. Annual Statistical Metrics for Manganese (PMi0) Concentrations Measured
                                           at NBIL
                                                2008

                                                Year
                • 5th Percentile     — Minimum     —  Median    —  Maximum
                                                                   95th Percentile
                                                                                . Average
 Figure 11-28. Annual Statistical Metrics for Acetaldehyde Concentrations Measured at
                                             SPIL
       r
       e
       .9
       B
       ซ 8
       s
       i
       I.
                                          2003           2009

                                                 Year
                * 5th Percentile     - Minimum     —  Median    -  Maximum     •  95th Percentile   ••*•• Average
                                             11-35

-------
Figure 11-29. Annual Statistical Metrics for Benzene Concentrations Measured at SPIL



tration (ug/m3)
Average Co noen
1 -










4
••


•
r










r f T 1
1 r
^ 1
••*-.. .— ,
•


*^ 	 -^^ 	 ^^.. 	 •ฃ• "• ^^
•••
c i ^ r t
2004 2005 2006 2007 2008 2009 2010 2011
Year
• 5th Percentile — Minimum — Median — Max mum • 95th Percentile ..+** Average
Figure 11-30. Annual Statistical Metrics for 1,3-Butadiene Concentrations Measured at
                                      SPIL



L
Concentration
3 i
Tt i
1

















-*•
2004






iii





r
i — l-n T r * *
i — ฐ^ ' T
^^ I
— •—
ซ•• 1 • 1 _._
F btd r ^ I r
2005 2006 2007 2008 2009 2010 2011
Year
• 5th Percentile — Minimum — Median — Maximum • 95th Percentile ..^.. Average
                                      11-36

-------
  Figure 11-31. Annual Statistical Metrics for Formaldehyde Concentrations Measured at
                                         SPIL
90 -
so -
70 -
m"
|5o-
J
• 50 -
1
a
| 40 -
<
30 -
20 -
10 -
0















The maximum |
rnnrpntratinnfnr !
2006 is 162 ug/mS. |




*...






""•[^fa i i^^ 1 ^ 	 |
^^m 	 •" l^^" , l^*" , l^^* . ]^^m
2006 2007 2008 2009 2010 2011
Year
• 5th Percentile — Minimum — Median — Maximum • 95th Percentile "^"Average
       Observations from Figure 11-20 for acetaldehyde measurements at NBIL include the

following:

       •   Although carbonyl compound sampling at NBIL began in 2005,  sampling did not
          begin until March, which does not yield enough samples for the  statistical metrics to
          be calculated for 2005, based on the criteria specified in Section  3.5.4. Thus,
          Figure  11-20 begins with 2006.

       •   The maximum acetaldehyde concentration (4.12 |ig/m3) was measured in 2011, as
          was the second highest concentration (3.47 |ig/m3). In fact, of the 19 acetaldehyde
          concentrations greater than 2 |ig/m3 measured at NBIL, 11 were  measured in 2011.

       •   All of the statistical metrics exhibit an increase from 2009 to 2010 and again for
          2011. The 95th percentile for 2011 is greater than the maximum concentration
          measured for most years of sampling.

       •   The difference between the average and median concentrations is at a maximum for
          2011, reflecting greater variability in the acetaldehyde measurements.

       •   The minimum concentration for each year is greater than zero, indicating that there
          were no non-detects of acetaldehyde reported for the years shown.
                                         11-37

-------
       Observations from Figure 11-21 for arsenic (PMio) measurements at NBIL include the
following:

       •   Metals sampling at NBIL began in January 2005.

       •   The maximum arsenic concentration was measured on July 12, 2009, although a
          similar concentration was also measured in 2010.

       •   The average concentrations exhibit little significant change over the course of
          sampling. The average concentrations increased from 2005 to 2006, reached a
          maximum for 2007 (0.86 ng/m3), and then returned to previous levels in 2008, after
          which the average concentration has remained steady. Since 2008, the average
          concentrations have ranged from 0.73 ng/m3 (2011) to 0.75 ng/m3 (2010).

       •   Note that 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.


       Observations from Figure 11-22 for benzene measurements at NBIL include the

following:

       •   Although sampling for VOCs at NBIL began in 2003, sampling did not begin until
          April, which does not yield enough samples for the statistical metrics to be calculated
          for 2003, based on the criteria specified in Section 3.5.4. In addition, sampling for
          VOCs was discontinued in October 2004 through the end of the year. Thus, Figure
          11-22 begins with 2005.

       •   The maximum benzene concentration was measured on January 9, 2011 and is the
          only measurement greater than 4 |ig/m3 measured at NBIL.

       •   The average (and median) concentration decreased significantly from 2005 to 2006,
          and continued to decreased through 2007, then remained at the same level for 2008
          and 2009. The average benzene concentration increased from 2009 to 2010. Even
          with the maximum concentration measured in 2011, the average concentration
          remained fairly static (and the 95th percentile  actually decreased).

       •   The minimum concentration for each year is greater than zero, indicating that there
          were no non-detects of benzene reported for the years shown.


       Observations from Figure 11-23 for 1,3-butadiene measurements at NBIL include the
following:

       •   The maximum 1,3-butadiene concentration was measured on the same day as the
          maximum benzene concentration, January 9, 2011 (2.68 |ig/m3). Only three
          concentrations greater than 1 |ig/m3 have been measured at NBIL, two in 2011 and
          one in 2010. All other measurements of 1,3-butadiene are less  than 0.35 |ig/m3.
                                         11-38

-------
       •  The average concentration changed little through 2009, after which an increase is
          shown.

       •  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 still within the same range. For example, for 2011, only three measurements
          were greater than the 95th percentile; this is also true for 2010.

       •  For each year shown, the minimum and 5th percentile are zero, indicating the presence
          of non-detects (at least 5 percent of the measurements). The number of non-detects
          reported has fluctuated over the years of sampling, from as high as 43 percent (2005)
          to as low as seven percent (2007).


       Observations from Figure 11-24 for formaldehyde measurements  at NBIL include the
following:

       •  The maximum formaldehyde concentration was measured on January 5, 2006
          (91.7 |ig/m3). However, the next five highest concentrations, ranging from 14.4 |ig/m3
          to 53.5 |ig/m3, were all measured in 2010. The only other formaldehyde measurement
          greater than  10 |ig/m3 was measured in 2011 (13.7 |ig/m3).

       •  The maximum concentration measured in 2006 is 20 times higher than the next
          highest concentration measured that year (4.46 |ig/m3).  The magnitude of this outlier
          explains why the average concentration for 2006 is greater than the 95th percentile.

       •  The statistical metrics for 2010 are also affected by the  higher concentrations;
          however, concentrations measured this year are higher in general, as indicated by
          seven-fold increase in the 95th  percentile. Although difficult to discern in
          Figure 11-24, the average concentration 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 difficult to discern in Figure 11-24, the minimum concentration for each
          year is greater than zero, indicating that there were no non-detects  of formaldehyde
          reported for the years shown.


       Observations from Figure 11-25 for hexavalent chromium measurements at NBIL include

the following:

       •  Hexavalent chromium sampling at NBIL began in January 2005.

       •  The maximum hexavalent chromium concentration was measured  on July 5, 2007
          (0.307 ng/m3). Only five additional measurements from NBIL are greater than
          0.1 ng/m3, and range from 0.108 ng/m3 to 0.235 ng/m3.  Four of the top  six
          concentrations were measured in 2006, with the fifth measured in 2005.
                                         11-39

-------
       •  Both the minimum concentration and 5th percentile for all years are zero, indicating
          the presence of non-detects. For 2009, the median concentration is also zero,
          indicating that at least 50 percent of the measurements are non-detects.

       •  The average (and median) concentration of hexavalent chromium decreased
          significantly between 2006 and 2009. The maximum,  95th percentile, median, and
          average concentrations all exhibit an increase for 2010. For 2011, the range of
          concentrations (based on the minimum and maximum concentrations as well as the
          5th and 95th percentiles) is more similar to the statistical metrics for 2009, although
          both the average and median concentrations are higher for 2011, most likely because
          the number of non-detects is significantly lower (from 56 percent non-detects in 2009
          down to 18 percent in 2011).


       Observations from Figure 11-26 for lead (PMio) measurements at NBIL include the
following:

       •  The maximum lead concentration was measured on July 5, 2007 (26.8 ng/m3), which
          is the  same day the maximum hexavalent chromium concentration was measured. The
          only other concentration greater than 20 ng/m3 was measured in October 2011.

       •  The average concentrations exhibit a decreasing trend from 2008 through 2010, but
          increased in 2011 back to 2008 levels.

       •  The average and median concentrations differ by nearly 1 ng/m3 for each year.

       •  The minimum concentration for each year is greater than zero,  indicating that there
          were no non-detects of lead reported since the onset of metals sampling.


       Observations from Figure 11-27 for manganese (PMio) measurements at NBIL include
the following:

       •  The maximum manganese concentration was measured on August 26, 2005
          (54.6  ng/m3). Concentrations in the 40-45 ng/m3 range have been measured in 2005,
          2008, and 2010.

       •  The average concentration decreased significantly from 2005 to 2006. Note that the
          average concentration for 2005 is greater than or similar to the  95th percentiles for
          some  of the years that follow. After 2005, changes in the averages are statistically
          insignificant.

       •  Nearly all of the  statistical metrics are at a minimum for 2009.

       •  The minimum concentration for each year is greater than zero,  indicating that there
          were no non-detects of manganese reported since the onset of metals sampling.
                                         11-40

-------
       Observations from Figure 11-28 for acetaldehyde measurements at SPIL include the
following:

       •  Although carbonyl compound sampling at SPIL began in 2005, sampling did not
          begin in earnest until March, which does not yield enough samples for the statistical
          metrics to be calculated for 2005, based on the criteria specified in Section 3.5.4.
          Thus, Figure 11-28 begins with 2006.

       •  The maximum acetaldehyde concentration was measured on January 9, 2011. The
          four highest concentrations of acetaldehyde were measured in 2011 and of the 10
          acetaldehyde concentrations greater than 5.0 |ig/m3 measured at SPIL, eight measured
          in 2011.

       •  The average concentration decreased from 2006 to 2007, then held steady through
          2009. The average concentration increased slightly in 2010 then increased
          significantly in 2011. All of the statistical metrics increased for 2011, including the
          95th percentile, indicating that the increases shown are not attributable to a handful of
          outliers.

       •  Note that the minimum concentration for each year is greater than zero, indicating
          that there were no non-detects of acetaldehyde reported for SPIL over the period
          shown.


       Observations from Figure 11-29 for benzene measurements at SPIL include the

following:

       •  Although sampling for VOCs  at SPIL began in 2003, sampling did not begin in until
          April 2003, which does not yield enough samples for the statistical metrics to be
          calculated for 2003, based on the criteria specified in Section 3.5.4. Thus,
          Figure 11-29 begins with 2004.

       •  The maximum benzene concentration was measured on October  13, 2005, although a
          similar concentration was also measured in February 2005.

       •  The average benzene concentration has decreased over the years, reaching a
          minimum of 0.68 |ig/m3 for 2009. The average concentration then increased for 2010,
          with a slight decrease for 2011.

       •  The minimum concentration for each year is greater than zero, indicating that no non-
          detects of benzene have been reported for SPIL over the period shown.
                                         11-41

-------
       Observations from Figure 11-30 for 1,3-butadiene measurements at SPIL include the
following:

       •   The maximum concentration of 1,3-butadiene was measured on February 3, 2005
          (1.29 |ig/m3) and is the only measurement greater than 1 |ig/m3. Only five
          concentrations greater than 0.5 |ig/m3 have been measured at SPIL, one in 2004, two
          in 2005, and two in 2011.

       •   The average concentration of 1,3-butadiene began to decrease in 2007 and reached a
          minimum in 2009 (0.085 |ig/m3). The increase from 2009 to 2010 is significant, with
          a 40 percent increase from 2009 levels. The median concentrations follow a similar
          pattern.

       •   The range of concentrations measured, as indicated by both the minimum and
          maximum concentrations and the 5th and 95th percentiles, has been decreasing over
          the years, but increased significantly for 2010, which continued into 2011.

       •   The detection rate for 1,3-butadiene has increased over time, ranging from
          approximately 45 percent non-detects in 2004 down to zero in 2008 and 2009 and one
          non-detect each for 2010 and 2011.


       Observations from Figure 11-31 for formaldehyde measurements at SPIL include the

following:

       •   The maximum formaldehyde concentration was measured on May 29, 2006 and is
          more than 10 times the maximum concentrations for any of the other years shown in
          Figure 11-31. Of the 26 formaldehyde concentrations greater than 10 |ig/m3, all but
          two were measured in 2006.

       •   The average concentration for 2006 is 13.76 |ig/m3. After 2006, the average
          concentration decreased each year, reaching a minimum of 1.85 |ig/m3 for 2009.
          Although difficult to discern in Figure 11-31, the average concentration has increased
          each year since 2009.
                                         11-42

-------
11.5   Additional Risk-Based Screening Evaluations
       The following risk-based screening evaluations were conducted to characterize risk at
each Illinois monitoring site. Refer to Sections 3.3 and 3.5.5 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.

11.5.1  Risk-Based Screening Assessment Using MRLs
       A risk-based screening was conducted by comparing the concentration data from the
Illinois monitoring sites to the ATSDR MRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 to 14 days); intermediate (exposures of 15 to 364 days); and chronic (exposures
of 1 year or greater). The preprocessed daily measurements of the pollutants of interest were
compared to the acute MRLs; the quarterly averages were compared to the intermediate MRLs;
and the annual averages were compared to the chronic MRLs.

       As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2011.

11.5.2  Cancer Risk and Noncancer Hazard Approximations
       For the pollutants of interest for the Illinois sites and where annual average
concentrations could be calculated, risk was examined by calculating cancer risk and noncancer
hazard approximations. These approximations can be used as risk estimates for cancer and
noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers may want to shift or
confirm their air-monitoring priorities.  Refer to Section 3.5.5.2 for an explanation of how cancer
risk and noncancer hazard approximations are calculated and what limitations are associated with
them. Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer
hazard approximations are presented in Table 11-6, where applicable. Cancer risk
approximations are presented as probabilities while the noncancer hazard approximations are
ratios and thus, unitless values.
                                          11-43

-------
                 Table 11-6. Risk Approximations for the Illinois Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
# of Measured
Detections vs.
# of Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Northbrook, Illinois - NBIL
Acenaphthene a
Acetaldehyde
Acrylonitrile
Arsenic (PM10)a
Benzene
Benzo(a)pyrene a
Bery Ilium (PM10)a
1,3 -Butadiene
Cadmium (PM10)a
Carbon Tetrachloride
Chloroform
1,2-Dichloroethane
Ethylbenzene
Fluoranthene a
Fluorene a
Formaldehyde
Hexavalent Chromium3
Lead(PM10)a
Manganese (PM10) a
Naphthalene a
Nickel (PM10)a
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
0.000088
0.0000022
0.000068
0.0043
0.0000078
0.00176
0.0024
0.00003
0.0018
0.000006
_
0.000026
0.0000025
0.000088
0.000088
0.000013
0.012
..
__
0.000034
0.00048
0.00000026
0.0000048
0.0000088
_
0.009
0.002
0.000015
0.03
_
0.00002
0.002
0.00001
0.1
0.098
2.4
1
_
_
0.0098
0.0001
0.00015
0.00005
0.003
0.00009
0.04
0.002
0.1
61/61
62/62
35/55
53/53
55/55
59/61
51/53
41/55
53/53
55/55
55/55
18/55
55/55
61/61
61/61
62/62
50/61
53/53
53/53
61/61
53/53
53/55
24/55
8/55
0.01
ฑ<0.01
1.41
ฑ0.20
0.13
ฑ0.03
<0.01
ฑ<0.01
0.71
ฑ0.16
<0.01
ฑ<0.01
<0.01
ฑ<0.01
0.12
ฑ0.10
<0.01
ฑ<0.01
0.64
ฑ0.03
6.06
ฑ2.17
0.03
ฑ0.01
0.32
ฑ0.12
0.01
ฑ<0.01
0.01
ฑ<0.01
2.53
ฑ0.57
<0.01
ฑ<0.01
<0.01
ฑ<0.01
0.01
ฑ<0.01
0.10
ฑ0.03
<0.01
ฑ<0.01
0.35
ฑ0.13
0.04
ฑ0.02
<0.01
ฑ<0.01
1.16
3.11
9.07
3.15
5.53
0.26
0.02
3.45
0.26
3.84
_
0.67
0.80
0.67
1.18
32.89
0.20
_
	
3.38
0.61
0.09
0.21
0.03
_
0.16
0.07
0.05
0.02
..
0.01
0.06
0.01
0.01
0.06
0.01
0.01
_
_
0.26
0.01
0.03
0.17
0.03
0.01
0.01
0.02
O.01
— = a Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 11-5.
                                                 11-44

-------
          Table 11-6. Risk Approximations for the Illinois Monitoring Sites (Continued)
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
# of Measured
Detections vs.
# of Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Schiller Park, Illinois - SPIL
Acetaldehyde
Acrylonitrile
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
ฃ>-Dichlorobenzene
1 ,2-Dichloroethane
Formaldehyde
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
0.0000022
0.000068
0.0000078
0.00003
0.000006
_
0.000011
0.000026
0.000013
0.00000026
0.0000048
0.0000088
0.009
0.002
0.03
0.002
0.1
0.098
0.8
2.4
0.0098
0.04
0.002
0.1
62/62
47/56
56/56
55/56
56/56
38/56
33/56
15/56
62/62
53/56
42/56
6/56
2.94
ฑ0.61
0.77
ฑ0.11
0.85
ฑ0.12
0.16
ฑ0.03
0.58
ฑ0.02
0.09
ฑ0.02
0.07
ฑ0.03
0.02
ฑ0.01
3.29
ฑ0.46
0.31
ฑ0.08
0.64
ฑ0.41
<0.01
ฑ<0.01
6.48
52.22
6.64
4.69
3.50
_
0.80
0.59
42.81
0.08
3.05
0.03
0.33
0.38
0.03
0.08
0.01
0.01
0.01
0.01
0.34
0.01
0.32
0.01
— = a Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 11-5.
          Observations for NBIL from Table 11-6 include the following:

          •   Chloroform, formaldehyde, and acetaldehyde are the pollutants with the highest
              annual average concentrations for NBIL.

          •   Formaldehyde, acrylonitrile, and benzene have the highest cancer risk
              approximations.

          •   None of NBIL's pollutants of interest have noncancer hazard approximations greater
              than 1.0, indicating that no adverse health effects are expected from these individual
              pollutants. The pollutant with the highest noncancer hazard approximation is
              formaldehyde (0.26).

          •   Note that chloroform, which has the highest annual average for NBIL, has no cancer
              URE.  The noncancer hazard approximation for this pollutant is low (0.06).
                                             11-45

-------
       Observations for SPIL from Table 11-6 include the following:

       •  Formaldehyde, acetaldehyde, and benzene are the pollutants with the highest annual
          average concentrations for SPIL.

       •  Acrylonitrile has the highest cancer risk approximation for SPIL (52.22 in-a-million),
          followed by formaldehyde (42.81 in-a-million). The cancer risk approximation for
          acrylonitrile is the fifth highest cancer risk approximation among all site-specific
          pollutants of interest and the only pollutant besides formaldehyde program-wide with
          a cancer risk approximation greater than 30 in-a-million.

       •  None of SPIL's pollutants of interest have noncancer hazard approximations greater
          than 1.0, indicating that no adverse health effects are expected from these individual
          pollutants. The pollutants with the highest noncancer hazard approximations are
          acrylonitrile and formaldehyde (0.38 and 0.34, respectively) although acetaldehyde
          and trichloroethylene have noncancer hazard approximations of similar magnitudes.

       •  In most cases, the risk approximations for  SPIL are greater than the risk
          approximations for NBIL. This is most apparent for acrylonitrile, formaldehyde, and
          trichloroethylene, with chloroform  being the exception.


11.5.3 Risk-Based Emissions Assessment

       In addition to the risk-based screenings discussed above, Tables 11-7 and 11-8 present an

evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.

Table 11-7 presents the 10 pollutants with the  highest emissions from the 2008 NEI, the 10

pollutants with the highest toxicity-weighted emissions, and the 10 pollutants with the highest

cancer risk approximations (in-a-million), as calculated from annual averages provided in

Table 11-6. Table 11-8 presents similar information, but identifies the 10 pollutants with the

highest noncancer hazard approximations (HQ), also calculated from annual averages provided

in Table  11-6.
                                          11-46

-------
Table 11-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
                                                the Illinois Monitoring Sites
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Northbrook, Illinois (Cook County) - NBIL
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Tetrachloroethylene
Trichloroethylene
Naphthalene
Dichloromethane
POM, Group 2b
1,445.06
1,055.89
791.28
689.38
209.30
190.73
134.75
126.41
93.18
20.55
Formaldehyde
Benzene
Hexavalent Chromium, PM
1,3 -Butadiene
Naphthalene
Cadmium, PM
Nickel, PM
Arsenic, PM
Ethylbenzene
POM, Group 2b
1.37E-02
1.13E-02
6.92E-03
6.28E-03
4.30E-03
2.67E-03
2.44E-03
2.17E-03
1.98E-03
1.81E-03
Formaldehyde
Acrylonitrile
Benzene
Carbon Tetrachloride
1,3 -Butadiene
Naphthalene
Arsenic
Acetaldehyde
Fluorene
Acenaphthene
32.89
9.07
5.53
3.84
3.45
3.38
3.15
3.11
1.18
1.16
Schiller Park, Illinois (Cook County) - SPIL
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Tetrachloroethylene
Trichloroethylene
Naphthalene
Dichloromethane
POM, Group 2b
1,445.06
1,055.89
791.28
689.38
209.30
190.73
134.75
126.41
93.18
20.55
Formaldehyde
Benzene
Hexavalent Chromium, PM
1,3 -Butadiene
Naphthalene
Cadmium, PM
Nickel, PM
Arsenic, PM
Ethylbenzene
POM, Group 2b
1.37E-02
1.13E-02
6.92E-03
6.28E-03
4.30E-03
2.67E-03
2.44E-03
2.17E-03
1.98E-03
1.81E-03
Acrylonitrile
Formaldehyde
Benzene
Acetaldehyde
1,3 -Butadiene
Carbon Tetrachloride
Trichloroethylene
ฃ>-Dichlorobenzene
1 ,2-Dichloroethane
Tetrachloroethylene
52.22
42.81
6.64
6.48
4.69
3.50
3.05
0.80
0.59
0.08

-------
  Table 11-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
                                             RfCs for the Illinois Monitoring Sites
oo
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer Hazard
Approximation
(HQ)
Northbrook, Illinois (Cook County) - NBIL
Toluene
Xylenes
Methanol
1,1,1 -Trichloroethane
Methyl isobutyl ketone
Benzene
Formaldehyde
Hexane
Ethylbenzene
Acetaldehyde
5,432.03
3,804.27
3,800.86
3,014.08
2,290.29
1,445.06
1,055.89
1,006.50
791.28
689.38
Acrolein
Manganese, PM
Cadmium, PM
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Trichloroethylene
Nickel, PM
Benzene
Naphthalene
3,705,826.42
161,905.82
148,156.47
107,743.48
104,651.38
76,598.28
67,372.78
56,590.58
48,168.61
42,137.52
Formaldehyde
Manganese
Acetaldehyde
Acrylonitrile
Chloroform
1,3 -Butadiene
Arsenic
Naphthalene
Lead
Benzene
0.26
0.17
0.16
0.07
0.06
0.06
0.05
0.03
0.03
0.02
Schiller Park, Illinois (Cook County) - SPIL
Toluene
Xylenes
Methanol
1,1,1 -Trichloroethane
Methyl isobutyl ketone
Benzene
Formaldehyde
Hexane
Ethylbenzene
Acetaldehyde
5,432.03
3,804.27
3,800.86
3,014.08
2,290.29
1,445.06
1,055.89
1,006.50
791.28
689.38
Acrolein
Manganese, PM
Cadmium, PM
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Trichloroethylene
Nickel, PM
Benzene
Naphthalene
3,705,826.42
161,905.82
148,156.47
107,743.48
104,651.38
76,598.28
67,372.78
56,590.58
48,168.61
42,137.52
Acrylonitrile
Formaldehyde
Acetaldehyde
Trichloroethylene
1,3 -Butadiene
Benzene
Tetrachloroethylene
Carbon Tetrachloride
Chloroform
/>-Dichlorobenzene
0.38
0.34
0.33
0.32
0.08
0.03
0.01
0.01
<0.01
<0.01

-------
       The pollutants listed in Tables 11-7 and 11-8 are limited to those that have cancer and

noncancer toxicity factors, respectively. As a result, although the actual value of the emissions is

the same, the highest emitted pollutants in the cancer table may be different from the noncancer

table. The cancer risk and noncancer hazard approximations based on each site's annual averages

are limited to those pollutants for which each respective site sampled. As discussed in

Section 11.3, SPIL sampled for VOCs and carbonyl compounds. NBIL sampled for these

pollutants as well, but also sampled for SNMOCs, PMi0 metals, PAHs, and hexavalent

chromium. In addition, the cancer risk and noncancer hazard approximations are limited to those

pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-

depth discussion of this analysis is provided in Section 3.5.5.3. Similar to the cancer risk and

noncancer hazard approximations, this analysis may help policy-makers prioritize their air

monitoring activities.


       Observations from Table 11-7 include the following:

       •  Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
          cancer UREs in Cook County.

       •  The pollutants with the highest toxicity-weighted emissions (of the pollutants with
          cancer UREs) for Cook County are formaldehyde, benzene, and hexavalent
          chromium.

       •  Six of the highest emitted pollutants in Cook County also have the highest toxicity -
          weighted emissions.

       •  For both monitoring sites, formaldehyde, acrylonitrile, and benzene have the highest
          cancer risk approximations, although not necessarily in that order. Formaldehyde and
          benzene are the top two pollutants on both emissions-based lists, while acrylonitrile
          appears on neither emissions-based list.

       •  Carbon tetrachloride, which also appears among the pollutants with the highest cancer
          risk approximations for both sites, does not appear on either emissions-based list.

       •  Several metals appear among the pollutants with the highest toxicity-weighted
          emissions, including arsenic, which has the seventh highest cancer risk approximation
          for NBIL. SPIL did not sample metals. None of these metals appear among the
          highest emitted pollutants.

       •  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.
                                          11-49

-------
       •  POM, Group 2b ranks tenth for both the quantity emitted and the toxicity-weighted
          emissions in Cook County. POM, Group 2b includes acenaphthene, fluorene, and
          fluoranthene, all three of which are pollutants of interest for NBIL.


       Observations from Table 11-8 include the following:

       •  Toluene, xylenes, and methanol are the highest emitted pollutants with noncancer
          RfCs in Cook County.

       •  The pollutants with the highest toxicity-weighted emissions (of the pollutants with
          noncancer RfCs) for Cook County are acrolein, manganese, and cadmium. Although
          acrolein was sampled for at both NBIL and SPIL, this pollutant was excluded from
          the pollutants of interest designation, and thus subsequent risk-based screening
          evaluations, due to questions about the consistency and reliability of the
          measurements, as discussed in Section 3.2.

       •  Three of the highest emitted pollutants also have the highest toxicity-weighted
          emissions (benzene, formaldehyde, and acetaldehyde).

       •  Formaldehyde, manganese, and acetaldehyde have the highest noncancer hazard
          approximations for NBIL (albeit less than an HQ of 1.0). Formaldehyde and
          acetaldehyde appear on both emissions-based lists while manganese has the second
          highest toxicity-weighted emissions, but is not among the highest emitted pollutants
          in Cook County.

       •  Acrylonitrile has the highest noncancer hazard approximation for SPIL (albeit less
          than an HQ of 1.0). This pollutant appears on neither emissions-based list.


11.6   Summary of the 2011 Monitoring Data for NBIL and SPIL

       Results from several of the data treatments described in this section include the

following:

       ปซป  Twenty-four pollutants, including 13 NA TTS MQO Core Analytes, failed screens for
          NBIL. Sixteen pollutants, including six NATTS MQO Core Analytes, failed screens for
          SPIL.

       ปซป  The pollutant with the highest annual average concentration among the pollutants of
          interest for NBIL was chloroform. NBIL's annual average concentration of
          chloroform was the highest annual average among allNMP sites sampling this
          pollutant.

       ปซป  The pollutant with the highest annual average concentration among the pollutants of
          interest for SPIL was formaldehyde.SPIL has the highest annual average
          concentration of acrylonitrile and trichloroethylene among NMP sites sampling these
          pollutants.
                                         11-50

-------
Concentrations of acetaldehyde and 1,5'-butadiene have been increasing in recent
years at NBIL and SPIL.
                               11-51

-------
12.0   Sites in Indiana
       This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the UATMP sites in Indiana, and integrates these concentrations with
emissions, meteorological, and risk information. Data generated by sources other than ERG are
not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) for detailed discussions and definitions
regarding the various data analyses presented below.

12.1   Site Characterization
       This section characterizes the Indiana monitoring sites by providing geographical and
physical information about the location of the sites and the surrounding areas. This information
is provided to give the reader insight regarding factors that may influence the air quality near the
sites and assist in the interpretation of the ambient monitoring measurements.

       One Indiana site (INDEM) is located in the Chicago-Naperville-Joliet, IL-IN-WI MSA,
while a second site (WPIN) is located in the Indianapolis-Carmel, IN MSA. Figures 12-1 and
12-3 are composite satellite images retrieved from ArcGIS Explorer showing the monitoring
sites in their urban locations. Figures 12-2 and 12-4 identify nearby point source emissions
locations by source category near INDEM and WPIN, respectively, as reported in the 2008 NEI
for point sources. Note that only sources within 10 miles of the sites are included in the facility
counts provided in Figures 12-2 and 12-4. A 10-mile boundary was chosen to give the reader an
indication of which emissions sources and emissions source categories could potentially have a
direct effect on the air quality at the monitoring sites. Further, this boundary provides both the
proximity of emissions sources to the monitoring sites as well as the quantity of such sources
within a given distance of the sites. Sources outside the 10-mile radius are still visible on each
map, but have been grayed out in order to show emissions sources just outside the boundary.
Table 12-1 provides supplemental geographical information such as land use,  location setting,
and locational coordinates.
                                           12-1

-------
                                        Figure 12-1. Gary, Indiana (INDEM) Monitoring Site
to
to

-------
    Figure 12-2. NEI Point Sources Located Within 10 Miles of INDEM
                                                    87 20'CTW         87'15'CrW         STIWW
                                               Note: Due to facility density and collocation, the total facilities
                                               displayed may not represent alt facilities within the area of interest.
INDEM UATMP site       10 mile radius |	| County boundary

                                          x   Mine/Quarry (3)
                                          V   Mineral Products (7)
                                          ?   Miscellaneous Commeraal/lndustrial (24)
                                          M   Miscellaneous Manufacturing (1)
                                          •   Oil and/or Gas Production (3)
                                          ^   Petroleum Refinery (1)
                                          1   Primary Metal Production (1)
                                          B   Pulp and Paper  Plant/Wood Products (1)
                                          R   Rubber and Miscellaneous Plastics Products (1)
                                          >   Solid Waste Disposal - Commercial/Institutional (1)
                                          V   Steel Mill (14)
                                          S   Surface Coating (1)
                                          *   Transportation and Marketing of Petroleum Products (6)
Source Category Group (No. of Facilities)
 -t1  Aircraft Operations (14)
  i   Asphalt Processing/Roofing Manufacturing (1)
  B   Bulk Terminals/Bulk Plants (7)
 c   Chemical Manufacturing (5)
 S  Coke Battery (1)
  *   Electricity Generation via Combustion (6)
 ฉ  Fabricated Metal Products (3)
 F   Food Processing/Agriculture (1)
 +  Gypsum Manufacturing (1)
 'i   Hot Mix Asphalt Plant (1)
 •   Landfill (1)
 >  Lime Manufacturing (1)
                                              12-3

-------
                                     Figure 12-3. Indianapolis, Indiana (WPIN) Monitoring Site
to

-------
Figure 12-4. NEI Point Sources Located Within 10 Miles of WPIN
        Legend
                                         86'10'D-W        aa'S'O'W       88'ffO'W        85"55"OTV
                                         Note: Due to facility density and collocation, the total facilities
                                         displayed may nol represent all facilities within the area of interest.
               WPIN UATMP site
                                         10 mile radius
         County boundary
 41
 B
 c
Source Category Group (No. of Facilities)
     Flexible Polyurethane Foam Production (2)
  A   Grain Handling (2)
  ^   Institutional -school (1)
                                                •
Iron and Steel Foundry (1)
Landfill (1)
Miscellaneous Commercial/Industrial (1)
Municipal Waste Combustor (1)
Oil and/or Gas Production (1)
Pharmaceutical Manufacturing (2)
Steel Mill(1)
Surface Coating (4}
Transportation Equipment (1)
Wastewater Treatment (1)
                                        12-5

-------
                               Table 12-1. Geographical Information for the Indiana Monitoring Sites
Site
Code
INDEM
WPIN
AQS Code
18-089-0022
18-097-0078
Location
Gary
Indianapolis
County
Lake
Marion
Micro- or
Metropolitan
Statistical Area
Chicago-Joliet-
Naperville, IL-IN-
WIMSA
(Gary Div)
Indianapolis-
Carmel, IN MSA
Latitude
and
Longitude
41.606667,
-87.304722
39.811097,
-86.114469
Land Use
Industrial
Residential
Location
Setting
Urban/City
Center
Suburban
Additional Ambient Monitoring Information1
VOCs, SO2, NO, NO2, NOX, PAMS, O3,
Meteorological parameters, PM10, Black carbon,
UV Carbon, PM2 5, and PM2 5 Speciation,
Tetrahydrofuran, 1,4-Dioxane.
TSP Metals, CO, VOCs, SNMOCs, SO2, NOy, NO,
O3, Meteorological parameters, PM10, Black carbon,
UV Carbon, PM2 5, and PM2 5 Speciation,
Tetrahydrofuran, 1,4-Dioxane, PM Coarse.
Data for additional pollutants are reported to AQS for these sites (EPA, 2012c); however, these data are not generated by ERG and are therefore not included in this report
to

-------
       INDEM is located in Gary, Indiana, a few miles east of the Indiana-Illinois border and
southeast of Chicago. Gary is located on the southernmost bank of Lake Michigan. The site is
located just north of 1-90, the edge of which can be seen in the bottom left portion of
Figure 12-1, and 1-65. Although INDEM resides on the Indiana Dunes National Lakeshore, the
surrounding area is highly industrialized, as shown in Figure 12-1, and several railroads
transverse the area. Figure 12-2 shows that the majority of point sources within 10 miles of
INDEM are located to the west of the site.  The sources closest to INDEM are a mine/quarry, a
steel mill, and a facility that falls into the miscellaneous commercial/industrial category. The
emissions source categories with the highest number of sources within 10 miles of INDEM
include steel mills; aircraft operations; mineral products; and bulk terminals and plants.

       WPIN is located in the parking lot of George Washington Park, near East 30th Street in
northeast Indianapolis. Figure 12-3 shows that the area surrounding WPIN is suburban and
residential, with little industry in close proximity. A church and a charitable organization are
located across the street from Washington Park, as is Oscar Charleston Park. Figure 12-4 shows
that the majority of point sources are located to the south and southwest of WPIN, towards the
center of Marion County. The source category with the highest number of sources near WPIN is
the aircraft operations source category, which include airports as well as small runways,
heliports, or landing pads. The sources closest to WPIN are an aircraft operations facility and a
fabricated metal products facility.

       Table 12-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Indiana monitoring sites. Table 12-2 includes county-level
population and vehicle registration information. Table 12-2 also includes a county-level vehicle
registration-to-population ratio, which was calculated to represent the number of vehicles per
person within each monitoring site's residing county. In addition, the population within 10 miles
of each site is presented, based on postal code population data estimates. An estimate of 10-mile
vehicle ownership was then determined by applying the county-level vehicle registration-to-
population ratio to the 10-mile population surrounding each monitoring site. Table 12-2 also
contains traffic volume information for each site. Finally,  Table 12-2 presents the county-level
daily VMT for Marion and Lake  Counties.
                                           12-7

-------
Table 12-2. Population, Motor Vehicle, and Traffic Information for the Indiana Monitoring
                                          Sites
Site
INDEM
WPIN
Estimated
County
Population1
495,558
911,296
County-level
Vehicle
Registration2
419,431
820,767
Vehicles per
Person
(Registration:
Population)
0.85
0.90
Population
within 10
miles3
411,932
797,291
Estimated
10-mile
Vehicle
Ownership
348,652
718,087
Annual
Average
Daily
Traffic4
34,240
143,970
County-
level Daily
VMT5
16,226,000
32,005,000
Bounty-level population estimates reflect 2011 data from the U.S. Census Bureau (Census Bureau, 2012b)
2County-level vehicle registration reflects 2011 data from the Indiana Bureau of Motor Vehicles (IN BMV, 2012)
310-mile population estimates reflect 2011 data from Xionetic (Xionetic, 2011)
4 AADT reflects 2010 data from the Indiana DOT (IN DOT, 2010)
5County-level VMT reflects 2011 data from the Indiana DOT (IN DOT, 2012)
       Observations from Table 12-2 include the following:

       •  Marion County has almost twice the county-level population and vehicle registration
          as Lake County. The 10-mile population and estimated vehicle registration follow a
          similar pattern as the county-level values.

       •  The county-level and 10-mile populations are in the middle third of populations
          among NMP sites as are the county-level and 10-mile vehicle registrations.

       •  The vehicle-per-person ratios for both Indiana sites are in the middle of the range
          among the ratios for NMP sites.

       •  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 while the
          traffic volume for INDEM is based on data from 1-90 at 12/20. The traffic volume
          near WPIN is the seventh highest among NMP sites.

       •  The VMT for Marion County is almost twice the VMT for Lake County. The Marion
          County VMT ranked tenth among counties with NMP sites, while the VMT for Lake
          County is in the middle of the range (19th).


12.2   Meteorological Characterization

       The following sections characterize the meteorological conditions near the monitoring

sites in Indiana on sample days, as well as over the course of the year.


12.2.1 Climate Summary

       The city of Gary is located to the southeast of Chicago, and at the southern-most tip of

Lake Michigan. Gary's proximity to Lake Michigan is an important factor controlling the

weather of the area. In the summer, warm temperatures can be suppressed, while cold winter

temperatures are often moderated. Winds that blow across Lake Michigan and over Gary in the
                                           12-8

-------
winter can provide abundant amounts of lake-effect snow while lake breezes can bring relief
from summer heat (Bair, 1992; Gary, 2013; and ISCO, 2002).

       The city of Indianapolis is located in the center of Indiana, and experiences a temperate
continental climate and frequently changing weather patterns. Summers are warm and often
humid, as moist air flows northward out of the Gulf of Mexico. Winters are chilly with
occasional Arctic outbreaks. Precipitation is spread rather evenly throughout the year, with much
of the spring and summer precipitation resulting from showers and thunderstorms. The
prevailing wind direction is southwesterly  (Bair, 1992 and ISCO, 2002).

12.2.2 Meteorological Conditions in 2011
       Hourly meteorological data from NWS weather stations nearest these sites were retrieved
for 2011 (NCDC, 2011). The two closest weather stations are located at Lansing Municipal
Airport (near INDEM) and Eagle Creek Airpark (near WPIN), WBAN 04879 and 53842,
respectively. Additional information about these weather stations, such as the distance between
the sites and the weather stations, is provided in Table  12-3. These data were used to determine
how meteorological conditions on sample days vary from conditions experienced throughout the
year.

       Table 12-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for the entire year for 2011. 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 were representative of average weather conditions
throughout the year for WPIN. For INDEM, sample days appear slightly cooler and less moist
than all days throughout 2011, although the difference  is not statistically significant.
                                          12-9

-------
                          Table 12-3. Average Meteorological Conditions near the Indiana Monitoring Sites
Closest NWS
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(ฐF)
Average
Temperature
(ฐF)
Average
Dew Point
Temperature
(ฐF)
Average
Wet Bulb
Temperature
(ฐF)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Gary, Indiana - INDEM |
Lansing Municipal
Airport
04879
(41.54, -87.52)
11.36
miles
241ฐ
(WSW)
Sample
Day
2011
58.0
ฑ5.8
59.3
ฑ2.2
49.6
ฑ5.2
51.1
ฑ2.0
40.3
ฑ4.8
42.2
ฑ1.9
45.0
ฑ4.7
46.6
ฑ 1.8
72.9
ฑ2.7
74.1
ฑ1.2
NA
NA
7.2
ฑ0.9
7.0
ฑ0.4
Indianapolis, Indiana - WPIN
Eagle Creek
Airpark
53842
(39.83, -86.30)
9.13
miles
270ฐ
(W)
Sample
Day
2011
61.3
ฑ5.2
62.3
ฑ2.1
53.1
ฑ5.0
54.1
ฑ1.9
42.9
ฑ4.7
43.6
ฑ1.8
47.9
ฑ4.5
48.7
ฑ1.7
71.1
ฑ3.1
70.3
ฑ1.2
1016.3
ฑ1.7
1015.6
ฑ0.7
5.6
ฑ0.7
5.7
ฑ0.3
to
o
Sample day averages are highlighted to help differentiate the sample day averages from the full-year averages.
NA= Sea level pressure was not recorded at the Lansing Municipal Airport

-------
12.2.3 Back Trajectory Analysis

       Figure 12-5 is the composite back trajectory map for days on which samples were
collected at the INDEM monitoring site in 2011. Included in Figure 12-5 are four back

trajectories per sample day. Figure 12-6 is the corresponding cluster analysis. Similarly,

Figure 12-7 is the composite back trajectory map for days on which samples were collected at
WPIN and Figure 12-8 is the corresponding cluster analysis. An in-depth description of these

maps and how they were generated is presented in Section 3.5.2.1. For the composite maps, each

line represents the 24-hour trajectory along which a parcel of air traveled toward the monitoring

site on a given sample day and time, based on an initial height of 50 meters AGL. For the cluster

analyses, each line corresponds to a trajectory representative of a given cluster of back

trajectories. Each concentric circle around the sites in Figures 12-5 through 12-8 represents

100 miles.


       Observations from Figures 12-5 and 12-6 for INDEM include the following:

       •  Back trajectories originated from a variety of directions at INDEM.

       •  The 24-hour  air  shed domain for INDEM was among the larger in size compared to
          other NMP sites, with an average trajectory length of 263 miles. The farthest away a
          back trajectory originated was over northwest North Dakota, or nearly 875 miles
          away. However, most trajectories (approximately 90 percent) originated within 450
          miles of INDEM, with the longest trajectories originating from the west, northwest,
          and north.

       •  The short cluster trajectory originating over Lake Michigan represents nearly  one-
          third of back trajectories and includes trajectories originating from a  direction with a
          northerly  component and within 200 miles of the sites. Twenty-two percent of
          trajectories originated from the southwest of INDEM over Illinois and Missouri.
          Another 21 percent of back trajectories originated from the west to northwest to north
          of the site. Fifteen percent of trajectories originated from the southeast quadrant while
          11 percent originated from the northeast quadrant.
                                          12-11

-------
Figure 12-5. 2011 Composite Back Trajectory Map for INDEM
    Figure 12-6. Back Trajectory Cluster Map for INDEM
                          12-12

-------
Figure 12-7. 2011 Composite Back Trajectory Map for WPIN
    Figure 12-8. Back Trajectory Cluster Map for WPIN
                          12-13

-------
       Observations from Figures 12-7 and 12-8 for WPIN include the following:
       •  The composite back trajectory map for WPIN shows that back trajectories originated
          from a variety of directions, similar to INDEM.
       •  The 24-hour air shed domain for WPIN is similar in size to many other NMP
          monitoring sites, with an average trajectory length of 225 miles. The farthest away a
          back trajectory originated was over eastern North Dakota, or greater than 750 miles
          away, although most trajectories (90 percent) originated within 400 miles of WPIN.
          Back trajectories originating to the northwest and north tended to be the longest.
       •  The cluster analysis for WPIN shows that back trajectories originating from the
          northwest account for nearly one-third of back trajectories but are split into two
          cluster trajectories to account for varying lengths of trajectories. The shorter cluster
          trajectory originating over Illinois (21 percent) represents back trajectories originating
          from the west, northwest, and north and generally less than 250 miles in length. The
          short cluster trajectory originating over northern Kentucky represents back
          trajectories originating from the east, southeast, and south but of varying distances.
          Back trajectories also originated southwestward over Illinois and Kentucky; over
          Ohio, Michigan, and the  eastern Great Lakes; and northward over Michigan and
          Lakes Michigan, Superior, and Huron, and parts of Ontario, Canada.

12.2.4 Wind Rose Comparison
       Hourly surface wind data from the NWS weather stations near the Indiana sites, as
presented in Section 12.2.2, were uploaded into a wind rose software program to produce
customized wind roses, as described in Section 3.5.2.2. A wind rose shows the frequency of wind
directions using "petals" positioned  around a 16-point compass, and uses different colors to
represent wind speeds.
       Figure 12-9 presents a map showing the distance between the NWS station and INDEM,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 12-9 also presents three different wind roses for the
INDEM monitoring site. First, a historical wind rose representing 2003 to 2010 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 2011 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2011 is
presented. These can be used to identify the predominant wind speed and direction for 2011 and
determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically. Figure  12-10 presents the distance map and three wind roses
for WPIN.
                                          12-14

-------
Figure 12-9. Wind Roses for the Lansing Municipal Airport Weather Station near INDEM
 Distance between INDEM and NWS Station
               2003-2010 Historical Wind Rose
  :
                        lTt.im.lI   _


                      •uปn inn-Minim ;
                ,H if.rtปi i
   '
;,,„'•••
            2011 Wind Rose
                   Sample Day Wind Rose
                                        12-15

-------
Figure 12-10. Wind Roses for the Indianapolis International Airport Weather Station near

                                       WPIN
  Distance between WPIN and NWS Station
2001-2010 Historical Wind Rose
                      \ GRAHDUIEW
              =  W MUSIC B
        ._/          . ',  rซK,,,0r O


        V,    \     ! *-    >,•;      ;
            2011 Wind Rose
    Sample Day Wind Rose
                                        12-16

-------
Observations from Figure 12-9 for INDEM include the following:

•  The NWS weather station at Lancing Municipal Airport is the closest weather station
   to INDEM, although it is located approximately 11.4 miles west-southwest of
   INDEM. The location of the weather station is just east of the Illinois-Indiana state
   line and farther inland than INDEM and thus, farther away from the influences of
   Lake Michigan than INDEM.

•  The historical wind rose for INDEM shows that winds from the south to south-
   southwest and west are the predominant wind directions over the 2003-2010 time
   frame. Northerly to northeasterly winds off Lake Michigan accounted for just less
   than 20 percent of the wind measurements, as did calm winds. The strongest winds
   blew from the south to southwest to west of INDEM.

•  The wind patterns shown on the 2011 wind rose resemble the wind patterns shown on
   the historical wind rose,  although there are slightly fewer calm winds.

•  The sample day wind rose  exhibits similarities with the historical and full-year wind
   roses, but differences are also  evident. The calm rate is lower and there is a higher
   percentage of winds from the southeast quadrant. Winds from the east-southeast to
   south-southeast account  for approximately three times as many wind observations on
   sample days as they did throughout the year.


Observations from Figure 12-10 for WPIN include the following:

•  The NWS weather station at Eagle Creek Airpark is the closest weather station to
   WPIN and is located approximately 9.1 miles west of WPIN. Eagle Creek Airpark is
   located on the southeast  edge of the Eagle Creek Reservoir.

•  Winds from the south, from the western quadrants, and from the north account for the
   majority  (nearly 60 percent) of wind observations from 2001-2010, while winds from
   the eastern quadrants were observed less than one-quarter of the time. Calm winds
   (< 2 knots) were observed for  nearly 18 percent of observations. The strongest winds
   tended to flow from the northwest.

•  The wind patterns on the 2011 wind rose resemble the historical wind patterns,
   although there were fewer northwesterly winds and more northerly winds.

•  The sample day wind rose has a significantly higher number of winds from the east  to
   south-southeast than the  full-year wind rose. A similar observation is noted for the
   INDEM sample day wind rose. There were also fewer northerly and westerly wind
   observations on sample days.
                                   12-17

-------
12.3   Pollutants of Interest
       Site-specific "pollutants of interest" were determined for the Indiana monitoring sites in
order to allow 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. In
addition, if any of the NATTS MQO Core Analytes measured by each monitoring site did not
meet the pollutant of interest criteria based on the preliminary risk-based screening, that pollutant
was added to the list of site-specific pollutants of interest. A more in-depth description of the
risk-based screening process is presented in Section 3.2.

       Table 12-4 presents the results of the  preliminary risk-based screening process for the
Indiana monitoring sites. The pollutants that failed at least one screen and contributed to 95
percent of the total failed screens for each monitoring site are shaded.  NATTS MQO Core
Analytes are bolded. Thus, pollutants of interest are shaded and/or bolded. INDEM and WPIN
sampled for carbonyl compounds only.

        Table 12-4. Risk-Based Screening Results for the Indiana Monitoring Sites
Pollutant
Screening
Value
(Ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
Gary, Indiana - INDEM
Acetaldehyde
Formaldehyde
Propionaldehyde
0.45
0.077
0.8
Total
57
57
1
115
57
57
57
171
100.00
100.00
1.75
67.25
49.57
49.57
0.87
49.57
99.13
100.00

Indianapolis, Indiana - WPIN
Acetaldehyde
Formaldehyde
0.45
0.077
Total
51
51
102
51
51
102
100.00
100.00
100.00
50.00
50.00
50.00
100.00

       Observations from Table 12-4 include the following:
       •  Formaldehyde, acetaldehyde, and propionaldehyde are the only carbonyl compounds
          with risk screening values.
                                          12-18

-------
       •  All three carbonyl compounds with risk screening values failed screens for INDEM.
          Acetaldehyde and formaldehyde failed 100 percent of screens while propionaldehyde
          failed only one screen.
       •  Acetaldehyde and formaldehyde failed screens for WPIN. They contributed equally
          to the total number of failed screens. Both pollutants failed 100 percent of total failed
          screens.

12.4   Concentrations
       This section presents various concentration averages used to characterize pollution levels
at the Indiana monitoring sites. Where applicable, the following calculations and data analyses
were performed: Time period-based concentration averages (quarterly and annual) are provided
for the pollutants of interest for the Indiana monitoring sites, where the data meet the applicable
criteria. Concentration averages for select pollutants are also presented graphically for the sites to
illustrate how the sites' concentrations compare to the program-level averages, as presented in
Section 4.1. In addition, concentration averages for select pollutants are presented from previous
years of sampling in order to characterize concentration trends at the sites. Additional site-
specific statistical summaries for INDEM and WPIN are provided in Appendix L.

12.4.1 2011 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 of the total number of samples
possible within a given quarter for a quarterly average to be calculated. An annual average
includes all  measured detections and substituted zeros for non-detects for the entire year of
sampling. Annual averages were calculated for pollutants where three valid quarterly averages
could be calculated and where method completeness was greater than or equal to 85 percent, as
presented in Section 2.4. Quarterly and annual average concentrations for the Indiana monitoring
sites are presented in Table 12-5, where applicable. Note that if a pollutant was not detected in a
given calendar quarter, the quarterly average simply reflects "0" because only  zeros substituted
for non-detects were factored into the quarterly  average concentration.
                                           12-19

-------
    Table 12-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest
                              for the Indiana Monitoring Sites
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Ug/m3)
2nd
Quarter
Average
(Ug/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Ug/m3)
Gary, Indiana - INDEM
Acetaldehyde
Formaldehyde
57/57
57/57
1.11
ฑ0.15
2.02
ฑ0.19
NA
NA
1.43
ฑ0.24
3.03
ฑ0.69
1.43
ฑ0.35
1.86
ฑ0.48
1.27
ฑ0.13
2.30
ฑ0.27
Indianapolis, Indiana - WPIN
Acetaldehyde
Formaldehyde
51/51
51/51
NA
NA
3.11
ฑ0.49
6.24
ฑ1.32
2.51
ฑ0.46
4.69
ฑ0.86
1.74
ฑ0.37
2.50
ฑ0.46
NA
NA
       Observations for the Indiana sites from Table 12-5 include the following:

       •  The annual average concentration of formaldehyde is greater than the annual average
          concentration of acetaldehyde for INDEM.

       •  Second quarter average concentrations could not be calculated for INDEM due to too
          many invalidated samples within that quarter.

       •  Intermittent sampler issues resulted in a lack of first quarter and annual averages for
          WPIN. However, Appendix L provides the pollutant-specific average concentrations
          for all valid samples collected over the entire sample period for each site.

       •  The second quarter averages of acetaldehyde and formaldehyde are higher than the
          other quarterly averages for WPIN. The confidence interval for the second quarter
          formaldehyde average is rather large, indicating the potential influence of outliers. A
          review of the data shows that the highest formaldehyde concentration for WPIN was
          measured on June 8, 2011 (11.1 |ig/m3). Four of the six highest concentrations of
          formaldehyde (those greater than 6 |ig/m3) were measured at WPIN during the second
          quarter. Concentrations of formaldehyde ranged from 1.41 |ig/m3to 11.1 |ig/m3 with a
          median concentration of 4.22 |ig/m3.


       Tables 4-9 through 4-12 present the sites with the 10 highest annual average

concentrations for each of the program-level pollutants of interest. Observations for the Indiana

sites from those tables include the following:
          INDEM does not appear in Table 4-10. Its annual average concentration of
                              -.th
          formaldehyde ranks 17 and its annual average concentration of acetaldehyde ranks
            •>nd
          22" among NMP sites sampling carbonyl compounds.
                                         12-20

-------
12.4.2 Concentration Comparison
       In order to better illustrate how a site's annual average concentrations compare to the
program-level averages,  a site-specific box plot was created for the selected NATTS MQO Core
Analytes listed in Section 3.5.3, where applicable. Thus, box plots for acetaldehyde and
formaldehyde were created for INDEM. Annual averages could not be calculated for WPIN,
therefore, this site has no figures in this section. Figures 12-11 and 12-12 overlay INDEM's
minimum, annual average, and maximum concentrations onto the program-level minimum, first
quartile, median, average, third quartile, and maximum concentrations, as described in
Section 3.5.3.
       Figure 12-11. Program vs. Site-Specific Average Acetaldehyde Concentration
 INDEM
                                          Concent ration (|
                Program:  1st Quartile    2nd Quartile     3rd Quartile     4th Quartile     Average
                Site:     Site Average    Site Minimum/Maximum

       Figure 12-12. Program vs. Site-Specific Average Formaldehyde Concentration
 INDEM
                                 13
                                               15
                                          Concentration {|
                Program:   1st Quartile     2nd Quartile     3rd Quartile     4th Quartile     Average
                Site:      Site Average    Site Minimum/Maximum

                                           12-21

-------
       Observations from Figures 12-11 and 12-12 include the following:

          •  Figure 12-11 shows that the annual average acetaldehyde concentration for
             INDEM is less than both program-level average concentration and median
             concentration. In fact, INDEM's annual average is just greater than the program-
             level first quartile (or 25th percentile). The maximum concentration of
             acetaldehyde measured at INDEM is considerably less than the maximum
             concentration measured across the program. There were no non-detects of
             acetaldehyde measured at INDEM or across the program.

          •  Figure 12-12 shows that INDEM's annual average formaldehyde concentration is
             less than the program-level average formaldehyde concentration but greater than
             the program-level median concentration. The maximum concentration of
             formaldehyde measured at INDEM is considerably less than the maximum
             concentration measured across the program. There were no non-detects of
             formaldehyde measured at INDEM or across the program.


12.4.3  Concentration Trends

       A site-specific trends evaluation was completed for sites that have sampled one or more

of the selected NATTS MQO Core Analytes for 5 consecutive years or longer, as described in

Section 3.5.4. INDEM has sampled carbonyl  compounds since 2004; thus, Figures 12-13 and

12-4 present the annual  statistical metrics for acetaldehyde and formaldehyde for INDEM,

respectively. The statistical metrics presented for assessing trends include the substitution of

zeros for non-detects. Although WPIN began sampling carbonyl compounds in 2007 and thus,

meets the criteria for a trends analysis to be performed, the lack of annual averages for 2011

prevents the analysis from being performed for this site.
                                         12-22

-------
  Figure 12-13. Annual Statistical Metrics for Acetaldehyde Concentrations
                             Measured at INDEM
 ฃ 10
                                             '•ป.
                                    2007       2008
                                        Year
         • 5th Percentile    —  Minimum     —  Median     — Maximum    •  95th Percentile   "^"Average
 Figure 12-14. Annual Statistical Metrics for Formaldehyde Concentrations
                             Measured at INDEM
J 300
         * 5th Percentile    -  Minimum     —  Median     - Maximum
                                                          95th Percentile
                                                                       • Average
                                     12-23

-------
       Observations from Figure 12-13 for acetaldehyde measurements at INDEM include the
following:

       •   Although carbonyl compound sampling began in 2003 at INDEM, sampling did not
          begin until June, which does not yield enough samples for the statistical metrics to be
          calculated for 2003, based on the criteria specified in Section 3.5.4. Thus,
          Figure  12-13 begins with 2004.

       •   In addition, there was a 3-month gap in sampling between September and November
          2005 at the INDEM site; therefore, no statistical metrics are presented for 2005.

       •   The maximum acetaldehyde concentration shown (13.8 |ig/m3) was measured at
          INDEM on June 14, 2004. Four additional concentrations measured at INDEM were
          greater than 10 |ig/m3 (one in 2006 and three in 2008).

       •   The average concentration of acetaldehyde decreased significantly from 2007 to
          2008, although the maximum and 95th percentile increased for 2008. With the
          exception of the minimum and 5th percentile, the statistical parameters decreased
          significantly from 2008 to 2009. The  average and  median concentrations decreased
          by more than half and the 95th percentile decreased by more than 80 percent during
          this time. The carbonyl compound samplers  were  switched out in 2009, which seems
          to have had a significant impact on the concentrations measured, particularly with
          respect to formaldehyde, which is discussed in more detail  below.

       •   The statistical parameters for 2010 are similar in magnitude to the statistical
          parameters for 2011.

       •   There have been no non-detects of acetaldehyde measured  at INDEM over the years
          shown.

       Observations from Figure 12-14 for formaldehyde measurements at INDEM include the

following:

       •   Five formaldehyde concentrations greater than 400 |ig/m3 were measured in the
          summer of 2008 (ranging from 414 |ig/m3to 499 |ig/m3). While these are extremely
          high values of formaldehyde, concentrations of formaldehyde have been historically
          high at this site, as shown by the statistics in Figure 12-14.  There have been 25
          measurements of formaldehyde greater than 100 |ig/m3  measured at INDEM, not
          including the excluded years (2003 and 2005).

       •   Prior to 2009, the maximum concentration for each year is  greater than 100 |ig/m3.
          Further, the median concentrations for years prior to 2008 are greater than 30 |ig/m3,
          indicating that at least half of the concentrations were greater than 30 |ig/m3.
                                         12-24

-------
       •  Although the average concentration doubled from 2007 to 2008, the median
          concentration decreased by more than half. This means that although the magnitude
          of the outliers is driving the average concentration upward, there were a larger
          number of concentrations at the lower end of the range as well. For 2008, 40 percent
          of measurements were less than 5 |ig/m3; for the years prior to 2008, the number of
          measurements less than 5 |ig/m3 ranged from none (2006) to two (2004).

       •  All the statistical metrics decreased significantly for 2009 and the years that follow.
          The average concentration for 2009 is 2.58 |ig/m3 and the average continued to
          decrease thereafter (although the differences are not statistically significant). In
          contrast to the previous bullet, the number of measurements greater than 5 |ig/m3
          ranged from one to two for each year between 2009 and 2011.

       •  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 2009-2011
          compared to previous years. Thus, the elevated concentrations from previous years
          were likely related to the multi-channel collection equipment and may not reflect the
          actual levels in ambient air. However, concentrations in the earlier years of sampling
          must have still been higher based on the number of concentrations greater than 5
          |ig/m3 before and after 2009, as discussed in the previous bullets.


12.5   Additional Risk-Based Screening Evaluations

       The following risk-based screening evaluations were conducted to characterize risk at

each Indiana monitoring site. Refer to Sections 3.3 and 3.5.5 for definitions and explanations

regarding the various toxicity factors, time frames, and calculations associated with these risk-

based screenings.
12.5.1  Risk-Based Screening Assessment Using MRLs
       A risk-based screening was conducted by comparing the concentration data from the

Indiana monitoring sites to the ATSDR MRLs, where available. As described in Section 3.3,

MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute

(exposures of 1 to 14 days); intermediate (exposures of 15 to 364 days); and chronic (exposures

of 1 year or greater). The preprocessed daily measurements of the pollutants of interest were
                                         12-25

-------
 compared to the acute MRLs; the quarterly averages were compared to the intermediate MRLs;
 and the annual averages were compared to the chronic MRLs.

       As discussed in Section 4.2.2, none of the measured detections or time-period average
 concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
 noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2011.

 12.5.2 Cancer Risk and Noncancer Hazard Approximations
       For the pollutants of interest for the Indiana sites and where annual average
 concentrations could be calculated, risk was examined by calculating cancer risk and noncancer
 hazard approximations.  These approximations can be used as risk estimates for cancer and
 noncancer effects attributable to the pollutants of interest. Although the use of these
 approximations is limited, they may help identify where policy-makers may want to shift or
 confirm their air-monitoring priorities. Refer to Section 3.5.5.2 for an explanation of how cancer
 risk and noncancer hazard approximations are calculated and what limitations are associated with
 them. Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer
 hazard approximations are presented in Table 12-6, where applicable. Cancer risk
 approximations are presented as probabilities while the noncancer hazard approximations are
 ratios and thus, unitless values.

             Table 12-6. Risk Approximations for the Indiana Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
# of Measured
Detections vs.
# of Samples
Annual
Average
(jig/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Gary, Indiana - INDEM
Acetaldehyde
Formaldehyde
0.0000022
0.000013
0.009
0.0098
57/57
57/57
1.27
ฑ0.13
2.30
ฑ0.27
2.79
29.88
0.14
0.23
Indianapolis, Indiana - WPIN
Acetaldehyde
Formaldehyde
0.0000022
0.000013
0.009
0.0098
51/51
51/51
NA
NA
NA
NA
NA
NA
NA = Not available due to the criteria for calculating an annual average.
                                          12-26

-------
       Observations for the Indiana sites from Table 12-6 include the following:
       •  The annual average concentration of formaldehyde for INDEM is greater than the
          annual average concentration of acetaldehyde.
       •  The cancer risk approximation for formaldehyde is an order of magnitude higher than
          the cancer risk approximation for acetaldehyde for INDEM.
       •  None of the pollutants of interest for INDEM have noncancer hazard approximations
          greater than 1.0, indicating that no adverse health effects are expected from these
          individual pollutants.
       •  Cancer risk and noncancer hazard approximations could not be calculated for the
          pollutants of interest for WPIN because annual averages are not available.

12.5.3 Risk-Based Emissions Assessment
       In addition to the risk-based screenings discussed above, Tables 12-7 and 12-8 present an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 12-7 presents the 10 pollutants with the highest emissions from the 2008 NEI, the 10
pollutants with the highest toxicity-weighted emissions, and the 10 pollutants with the highest
cancer risk approximations (in-a-million), as calculated from the annual averages provided in
Table 12-6. Table 12-8 presents similar information, but identifies the 10 pollutants with the
highest noncancer hazard approximations, also calculated from annual averages provided in
Table 12-6.
       The pollutants listed in Tables 12-7 and 12-8 are limited to those that have cancer and
noncancer toxicity factors, respectively. As a result, although the actual value of the emissions is
the same, the highest emitted pollutants in the cancer table may be different from the noncancer
table. Further, the cancer risk and noncancer hazard approximations based on each site's annual
averages are limited to those pollutants for which each respective site sampled. As discussed in
Section 12.3, INDEM and WPIN sampled for carbonyl compounds only. 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, cancer risk and noncancer hazard
approximations were not calculated for WPIN. A more in-depth discussion of this analysis is
provided in Section 3.5.5.3. Similar to the cancer risk and noncancer hazard approximations, this
analysis may help policy-makers prioritize their air monitoring activities.
                                          12-27

-------
  Table 12-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for

                                                 the Indiana Monitoring Sites
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
Gary, Indiana (Lake County) - INDEM
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
Coke Oven Emissions, PM
1,3 -Butadiene
Naphthalene
Dichloromethane
POM, Group 2b
Trichloroethylene
231.53
157.04
107.50
100.19
33.43
29.96
29.83
4.72
2.39
2.23
Coke Oven Emissions, PM
Formaldehyde
Hexavalent Chromium, PM
Benzene
Naphthalene
Arsenic, PM
1,3 -Butadiene
Nickel, PM
Ethylbenzene
Cadmium, PM
3.31E-02
2.04E-03
1.81E-03
1.81E-03
1.01E-03
9.93E-04
8.99E-04
6.66E-04
2.69E-04
2.63E-04
Formaldehyde 29.88
Acetaldehyde 2.79

Indianapolis, Indiana (Marion County) - WPIN
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
Tetrachloroethylene
Dichloromethane
Trichloroethylene
POM, Group 2b
489.79
356.47
217.73
205.36
67.18
34.96
14.44
7.84
6.85
6.52
Formaldehyde
Benzene
1,3 -Butadiene
Hexavalent Chromium, PM
Naphthalene
Arsenic, PM
POM, Group 3
POM, Group 2b
Ethylbenzene
Nickel, PM
4.63E-03
3.82E-03
2.02E-03
1.23E-03
1.19E-03
1.10E-03
6.97E-04
5.74E-04
5.44E-04
4.59E-04

to

to
oo

-------
  Table 12-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer

                                             RfCs for the Indiana Monitoring Sites
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted
Emissions (County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Noncancer Hazard
Approximation
Pollutant (HQ)
Gary, Indiana (Lake County) - INDEM
Toluene
Xylenes
Methanol
Ethylene glycol
Hexane
Benzene
Formaldehyde
Hydrochloric acid
Ethylbenzene
Acetaldehyde
676.97
528.24
315.13
271.89
231.90
231.53
157.04
153.19
107.50
100.19
Acrolein
Manganese, PM
Lead, PM
Cyanide Compounds, gas
Formaldehyde
Nickel, PM
Arsenic, PM
1,3 -Butadiene
Cadmium, PM
Acetaldehyde
503,495.00
468,665.08
109,768.53
36,721.25
16,024.72
15,405.79
15,399.37
14,981.21
14,626.32
11,131.89
Formaldehyde 0.23
Acetaldehyde 0.14

Indianapolis, Indiana (Marion County) - WPIN
Toluene
Xylenes
Methanol
Ethylene glycol
Benzene
Hydrochloric acid
Formaldehyde
Hexane
Ethylbenzene
Acetaldehyde
1,373.22
935.63
684.51
504.44
489.79
460.45
356.47
269.35
217.73
205.36
Acrolein
Formaldehyde
1,3 -Butadiene
Hydrochloric acid
Acetaldehyde
Arsenic, PM
Benzene
Naphthalene
Lead, PM
Nickel, PM
1,220,725.02
36,374.58
33,589.47
23,022.60
22,818.28
17,017.69
16,326.44
11,654.54
11,113.23
10,619.54

to

to
VO

-------
Observations from Table 12-7 include the following:

•  Benzene, formaldehyde, and ethylbenzene are the three highest emitted pollutants
   with cancer UREs in both Marion and Lake County, although the quantity emitted is
   roughly twice as high in Marion County.

•  Coke oven emissions, formaldehyde, and hexavalent chromium 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 both Lake and Marion County also have the
   highest toxicity-weighted emissions (although the pollutants are not the same between
   the counties).

•  While several metals (arsenic, cadmium, hexavalent chromium, and nickel) 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.

•  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.


Observations from Table 12-8 include the following:

•  While toluene is the highest emitted pollutant with a noncancer RfC in both counties,
   the toluene emissions in Marion County are nearly twice that of Lake County.
   Xylenes and methanol are the second and third highest emitted pollutants in both
   counties, with a similar pattern in the quantity emitted.

•  Acrolein is the pollutant with the highest toxicity-weighted emissions (of the
   pollutants with noncancer RfCs) for both counties. Manganese and lead rank second
   and third for Lake County, while formaldehyde and  1,3-butadiene rank second and
   third for Marion County.

•  Only two of the highest emitted pollutants in Lake County also have  the highest
   toxicity-weighted emissions (formaldehyde and acetaldehyde).  Several metals
   (manganese, lead, nickel, arsenic, and cadmium) are among the pollutants with the
   highest toxicity-weighted emissions for Lake  County, although none  of these appear
   among the highest emitted pollutants.

•  Four of the highest emitted pollutants in Marion County also have the highest
   toxicity-weighted emissions (formaldehyde, acetaldehyde, hydrochloric acid, and
   benzene).
                                   12-30

-------
12.6   Summary of the 2011 Monitoring Data for INDEM and WPIN

       Results from several of the data treatments described in this section include the
following:

       ปซป  Two carbonyl compounds failed screens for WPIN and three failed screens for
          INDEM.

       ปซป  The annual average concentration of formaldehyde is greater than the annual
          average concentration of acetaldehyde for INDEM. Annual averages concentrations
          could not be  calculated for WPIN.

       ปซป  Concentrations of formaldehyde and acetaldehyde exhibit a decreasing trend at
          INDEM.
                                         12-31

-------
13.0   Site in Kentucky
       This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS site in Kentucky, and integrates these concentrations
with emissions, meteorological, and risk information. Data generated by sources other than ERG
are not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) for detailed discussions and definitions
regarding the various data analyses presented below.

13.1   Site Characterization
       This section characterizes the Kentucky monitoring 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 Kentucky monitoring site is located near Grayson Lake in northeast Kentucky.
Figure 13-1 is a composite satellite image retrieved from ArcGIS Explorer showing the
monitoring site in its rural location. Figure 13-2 identifies nearby point source emissions
locations by source category, as reported in the 2008 NEI for point sources. Note that only
sources within 10 miles of the site are included in the facility counts provided in Figure 13-2. A
10-mile boundary was chosen to give the reader an indication of which emissions sources and
emissions source categories could potentially have a direct effect on the air quality at the
monitoring site. Further, this boundary provides both the proximity of emissions sources to the
monitoring site as well as the quantity of such sources within a given distance of the site. Sources
outside the 10-mile radius are still visible on the map, but have been grayed out in order to show
emissions sources just outside the boundary. Table 13-1 provides supplemental geographical
information such as land use, location setting, and locational coordinates.
                                           13-1

-------
                                     Figure 13-1. Grayson, Kentucky (GLKY) Monitoring Site
to

-------
Figure 13-2. NEI Point Sources Located Within 10 Miles of GLKY
                                      Mote; Due to facility density and education, the total facilities
                                      displayed may not represent all facilities within the area of interest,
        Legend
         ฉ  GLKY NATTS site
10 mile radius
                   County boundary
        Source Category Group (No. of Facilities)
          •f1   Aircraft Operations (1)
              Brick Manufacturing & Structural Clay (2)
          B   Bulk Terminals/Bulk Plants (1)
          F   Food Processing/Agriculture (1)
          *   Hot Mix Asphalt Plant (1)
          x   Mine/Quarry (1)
          ?   Miscellaneous Commercial/Industrial (1)
                                     13-3

-------
                            Table 13-1. Geographical Information for the Kentucky Monitoring Site

Site
Code
GLKY


AQS Code
21-043-005


Location
Gray son


County
Carter
Micro- or
Metropolitan
Statistical Area
Not in an MSA
Latitude
and
Longitude
38.238333,
-82.988333


Land Use
Residential

Location
Setting
Rural


Additional Ambient Monitoring Information1
Carbonyl compounds, O3, Meteorological parameters,
PMio, PM10 Speciation, PM25, andPM25 Speciation
BOLD ITALICS = EPA-designaled NATTS Site

-------
       Grayson Lake is located in northeast Kentucky, south of the town of Grayson, and west
of the Huntington-Ashland, WV-KY MSA. The Little Sandy River feeds into Grayson Lake,
which is a U.S. Army Corps of Engineers-managed project, and part of the Kentucky State Parks
system. The lake is narrow and winding, as shown in Figure 13-1, with sandstone cliffs rising to
up to 200 feet above the lake surface (KY, 2013 and ACE, 2013). The closest road to the
monitoring site is a service road feeding into Camp Grayson. This site serves as the Grayson
Lake NATTS site. Figure 13-2 shows that few point sources surround GLKY and that most of
them are on the outer periphery of the 10-mile radius around GLKY.  This is not surprising given
the rural nature of the area and that Grayson Lake is located roughly in the center of the 10-mile
radius in Figure 13-2, oriented from northeast to southwest. Sources within 10 miles of GLKY
are involved in aircraft operations, brick and structural clay manufacturing, food processing, and
mining, among others.

       Table 13-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Kentucky monitoring site. Table 13-2 includes county-level
population and vehicle registration information. Table 13-2 also includes a county-level vehicle
registration-to-population ratio, which was calculated to represent the number of vehicles per
person within the monitoring site's residing county. In addition, the population within 10 miles
of the site is presented, based on postal code population data estimates. An estimate of 10-mile
vehicle ownership was determined by applying the county-level vehicle registration-to-
population ratio to the 10-mile population surrounding the monitoring site. Table 13-2 also
contains traffic volume information. Finally, Table 13-2 presents the  county-level daily VMT for
Carter County.
     Table 13-2. Population, Motor Vehicle, and Traffic Information for the Kentucky
                                     Monitoring Site



Site
GLKY

Estimated
County
Population1
27,586

County-level
Vehicle
Registration2
32,398
Vehicles per
Person
(Registration:
Population)
1.17

Population
within 10
miles3
14,610
Estimated
10-mile
Vehicle
Ownership
17,159
Annual
Average
Daily
Traffic4
428
County-
level
Daily
VMT5
1,084,000
 Bounty-level population estimate reflects 2011 data from the U.S. Census Bureau (Census Bureau, 2012b)
 2County-level vehicle registration reflects 2011  data from the Kentucky Transportation Cabinet (KYTC, 2012a)
 310-mile population estimate reflects 2011 data  from Xionetic (Xionetic, 2011)
 4AADT reflects 2009 data from the Kentucky Transportation Cabinet (KYTC, 2009)
 5County-level VMT reflects 2011 data from the Kentucky Transportation Cabinet (KYTC, 2012b)
 BOLD ITALICS = EPA-designated NATTS Site
                                           13-5

-------
       Observations from Table 13-2 include the following:
       •  The Carter County population is the second lowest compared to counties with NMP
          sites (behind only UCSD in Union County, SD). The 10-mile population for GLKY is
          also on the low side compared to other sites. The corresponding vehicle ownership
          data mimicked these rankings. The rather low population and vehicle ownership
          compared to other NMP sites is not surprising given the rural nature of the
          surrounding area.
       •  The vehicle-per-person ratio is among the higher ratios compared to other NMP sites.
       •  The traffic data for GLKY is provided for the intersection of State Road 1496 with
          Camp Webb Road, one of several secondary roads leading to Grayson Lake. This site
          has the third lowest traffic volume among NMP sites.
       •  The daily VMT for Carter County is the second lowest compared to other counties
          with NMP sites (where VMT data were available), behind only Union County, SD.

13.2   Meteorological Characterization
       The following sections characterize the meteorological conditions  near the monitoring
site in Kentucky on sample days, as well as over the course of the year.
13.2.1  Climate Summary
       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, 2013).

13.2.2  Meteorological Conditions in 2011
       Hourly meteorological data from the NWS weather station nearest this site were retrieved
for 2011 (NCDC,  2011). The closest weather station to GLKY is located at Tri-State/Ml.
Ferguson Field Airport (WBAN 03860). Additional information about this weather station, such
as the distance between the site and the weather station, is provided in Table 13-3. These data
were used to determine how meteorological conditions on sample days vary from conditions
experienced throughout the year.
                                          13-6

-------
                    Table 13-3. Average Meteorological Conditions near the Kentucky Monitoring Site
Closest NWS
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(ฐF)
Average
Temperature
(ฐF)
Average
Dew Point
Temperature
(ฐF)
Average
Wet Bulb
Temperature
(ฐF)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Grayson, Kentucky - GLKY
Tri-St/MJ.
Ferguson Field
Airport
03860
(38.38, -82.56)
24.27
miles
58ฐ
(ENE)
Sample
Day
2011
65.3
ฑ4.4
65.7
ฑ1.8
55.6
ฑ4.1
56.1
ฑ1.7
45.7
ฑ4.6
46.5
ฑ1.8
50.7
ฑ4.0
51.2
ฑ1.6
72.6
ฑ3.3
73.4
ฑ1.3
1016.4
ฑ 1.5
1016.3
ฑ0.7
4.2
ฑ0.6
4.2
ฑ0.2
Sample day averages are highlighted to help differentiate the sample day averages from the full-year averages.

-------
       Table 13-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for the entire year for 2011. 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 near GLKY were representative of average weather
conditions throughout the year.

13.2.3 Back Trajectory Analysis
       Figure 13-3 is the composite back trajectory map for days on which samples were
collected at the GLKY monitoring site in 2011. Included in Figure  13-3 are four back trajectories
per sample day. Figure 13-4 is the corresponding cluster analysis. An in-depth description of
these maps and how they were generated is presented in Section 3.5.2.1. For the composite map,
each line represents the 24-hour trajectory along which a parcel of air traveled toward the
monitoring site on a given sample day and time, based on an initial height of 50 meters AGL. For
the cluster analysis, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric circle around the site in Figures 13-3 and 13-4 represents 100 miles.

       Observations from Figures 13-3 and 13-4 for GLKY include the following:
       •  Back trajectories originated from a variety of directions at GLKY. An imaginary line
          drawn north-south through the center of Figure 13-3 would show that a higher
          number of trajectories originated from a direction with a westerly component than an
          easterly component.
       •  The farthest away a back trajectory originated was over southeast Minnesota, or just
          greater than 600 miles  away; however, the average  trajectory length was 212 miles
          and 86 percent of trajectories originated within 350 miles of the monitoring site.
       •  Back trajectories originating to the west and northwest of GLKY account for nearly
          40 percent of back trajectories, but are represented by two clusters in Figure 13-4.
          One cluster (26 percent) includes back trajectories originating over Ohio and Indiana
          while the other cluster  (13 percent) includes longer back trajectories originating over
          Illinois or farther to the west or northwest. Another nearly 40 percent of back
          trajectories originated to the southwest of GLKY. This cluster trajectory includes
          back  trajectories of varying lengths. Fifteen percent of back trajectories originated to
          the north-northwest to north-northeast of GLKY, as represented by the cluster
          trajectory originating over Lake Erie. The cluster trajectory originating over Virginia
          (10 percent) includes shorter trajectories originating to the east over West Virginia
          and Virginia or longer trajectories originating to the southeast over North Carolina.
                                           13-8

-------
Figure 13-3. 2011 Composite Back Trajectory Map for GLKY
    Figure 13-4. Back Trajectory Cluster Map for GLKY
                           13-9

-------
13.2.4 Wind Rose Comparison
       Hourly surface wind data from the NWS weather station at the Tri-State/M. J. Ferguson
Field Airport near GLKY were uploaded into a wind rose software program to produce
customized wind roses, as described in Section 3.5.2.2. A wind rose shows the frequency of wind
directions using "petals" positioned around a 16-point compass, and uses different colors to
represent wind speeds.

       Figure 13-5 presents a map showing the distance between the NWS station and GLKY,
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
GLKY monitoring site. First, a historical wind rose representing 2001 to 2010 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 2011 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2011 is
presented. These can be used to identify the predominant wind speed and direction for 2011 and
determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.

       Observations from Figure 13-5 for GLKY include the following:
       •  The Tri-State/MJ. Ferguson Field weather station is located over 24  miles to the
          east-northeast of GLKY and just across the state border in West Virginia.
       •  The historical wind rose shows that calm winds were observed for more than
          23 percent of the hourly measurements near GLKY. Winds from the  south to
          southwest to west make up the majority of observations near GLKY, particularly
          those from south-southwest.
       •  The wind patterns on the 2011 wind rose are similar to those on the historical wind
          rose, but calm winds accounted for a slightly higher percentage of the wind
          observations in 2011 (27 percent).
       •  The sample day wind rose resembles both the historical and full-year wind roses,
          although with fewer southerly winds and more northeasterly winds. This indicates the
          wind conditions on sample days were generally representative of those experienced
          throughout 2011 and historically.
                                          13-10

-------
Figure 13-5. Wind Roses for the Tri-State/M. J. Ferguson Field Airport Weather Station
                                  near GLKY
 Distance between GLKY and NWS Station
2001-2010 Historical Wind Rose

          2011 Wind Rose
    Sample Day Wind Rose
                                     13-11

-------
13.3   Pollutants of Interest

       Site-specific "pollutants of interest" were determined for GLKY in order to allow

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." 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. In addition, if any of the NATTS

MQO Core Analytes measured by the monitoring site did not meet the pollutant of interest

criteria based on the preliminary risk-based screening, that pollutant was added to the list of site-

specific pollutants of interest. A more in-depth description of the risk-based screening process is
presented in Section 3.2.


       Table 13-4 presents the results of the preliminary risk-based screening process for

GLKY. The pollutants that failed at least one screen and contributed to 95 percent of the total

failed screens for the monitoring site are shaded. NATTS MQO Core Analytes are bolded.  Thus,

pollutants of interest are shaded and/or bolded. GLKY sampled for hexavalent chromium, PMio

metals, carbonyl compounds, PAHs, and VOCs.


       Observations from Table 13-4 include the following:

          •   GLKY sampled hexavalent chromium, VOCs, and PAHs throughout 2011, but
              did not begin  sampling metals through the NMP until May and carbonyl
              compounds until August.

          •   Fifteen pollutants failed screens for GLKY, including nine NATTS MQO Core
              Analytes.

          •   Ten pollutants were initially identified as pollutants of interest via the risk-based
              screening process, of which six are NATTS MQO Core Analytes. Naphthalene
              and nickel were added as pollutants of interest for GLKY because they are
              NATTS  Core Analytes, even though they did not contribute to 95 percent of the
              total failed screens. Nine additional pollutants were added to GLKY's pollutants
              of interest because they are NATTS MQO Core Analytes, even though they did
              not fail any screens. These nine pollutants  are not shown in Table 13-4 but are
              shown in subsequent tables in the sections that follow.

          •   Benzene and formaldehyde were detected in each VOC and carbonyl compound
              sample, respectively, and failed 100 percent of screens. Other pollutants also
              failed 100 percent of screens but were detected less frequently.
                                          13-12

-------
        Table 13-4. Risk-Based Screening Results for the Kentucky Monitoring Site
Pollutant
Screening
Value
(Ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
Grayson, Kentucky - GLKY
Benzene
Carbon Tetrachloride
1,3-Butadiene
Arsenic (PM10)
Formaldehyde
Acet aldehyde
1 ,2-Dichloroethane
Acrylonitrile
Hexachloro- 1 , 3 -butadiene
Manganese (PM10)
1 , 1 ,2,2-Tetrachloroethane
Naphthalene
1 ,2-Dibromoethane
ฃ>-Dichlorobenzene
Nickel (PM10)
0.13
0.17
0.03
0.00023
0.077
0.45
0.038
0.015
0.045
0.005
0.017
0.029
0.0017
0.091
0.0021
Total
61
60
36
31
26
22
21
20
8
6
4
3
2
2
1
303
61
61
43
41
26
26
21
20
10
41
4
61
2
21
41
479
100.00
98.36
83.72
75.61
100.00
84.62
100.00
100.00
80.00
14.63
100.00
4.92
100.00
9.52
2.44
63.26
20.13
19.80
11.88
10.23
8.58
7.26
6.93
6.60
2.64
1.98
1.32
0.99
0.66
0.66
0.33
20.13
39.93
51.82
62.05
70.63
77.89
84.82
91.42
94.06
96.04
97.36
98.35
99.01
99.67
100.00

13.4   Concentrations
       This section presents various concentration averages used to characterize pollution levels
at the Kentucky monitoring site. Where applicable, the following calculations and data analyses
were performed: Time period-based concentration averages (quarterly and annual) are provided
for the pollutants of interest for GLKY, where the data meet the applicable criteria.
Concentration averages for select pollutants are also presented graphically to illustrate how the
site's concentrations compare to the program-level averages, as presented in Section 4.1. In
addition, concentration averages for select pollutants are presented from previous years of
sampling in order to characterize concentration trends at the site. Additional site-specific
statistical summaries for GLKY are provided in Appendices J, L, M, N, and O.

13.4.1 2011 Concentration Averages
       Quarterly and annual concentration averages were calculated for the pollutants of interest
for the Kentucky 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 of the total number of samples
                                          13-13

-------
possible within a given quarter for a quarterly average to be calculated. An annual average
includes all measured detections and substituted zeros for non-detects for the entire year of
sampling. Annual averages were calculated for pollutants where three valid quarterly averages
could be calculated and where method completeness was greater than or equal to 85 percent, as
presented in Section 2.4. Quarterly and annual average concentrations for the Kentucky
monitoring site are presented in Table 13-5, where applicable. Note that concentrations of the
PAHs, metals, and hexavalent chromium for GLKY 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 13-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
                               the Kentucky Monitoring Site
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Hg/m3)
2nd
Quarter
Average
(Hg/m3)
3rd
Quarter
Average
(jig/m3)
4th
Quarter
Average
(jig/m3)
Annual
Average
(jig/m3)
Grayson, Kentucky - GLKY
Acetaldehyde
Acrylonitrile
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
1 ,2-Dichloroethane
Formaldehyde
Hexachloro- 1 , 3 -butadiene
Tetrachloroethylene
Trichloroethylene
26/26
20/61
61/61
43/61
61/61
15/61
21/61
26/26
10/61
30/61
4/61
NA
0
0.62
ฑ0.06
0.04
ฑ0.01
0.53
ฑ0.08
0.02
ฑ0.02
0.02
ฑ0.02
NA
0.01
ฑ0.02
0.06
ฑ0.02
0
NA
0.01
ฑ0.02
0.46
ฑ0.08
0.02
ฑ0.01
0.59
ฑ0.04
0.02
ฑ0.02
0.04
ฑ0.03
NA
0.02
ฑ0.02
0.04
ฑ0.02
0.01
ฑ0.01
NA
0.05
ฑ0.07
0.58
ฑ0.19
0.03
ฑ0.02
0.69
ฑ0.04
0.06
ฑ0.04
0.01
ฑ0.01
NA
0.02
ฑ0.02
0.02
ฑ0.02
0.01
ฑ0.01
0.78
ฑ0.21
0.58
ฑ0.07
0.69
ฑ0.30
0.05
ฑ0.01
0.69
ฑ0.05
0.01
ฑ0.01
0.05
ฑ0.02
0.77
ฑ0.25
0.01
ฑ0.01
0.03
ฑ0.02
0.01
ฑ0.01
NA
0.16
ฑ0.07
0.59
ฑ0.09
0.04
ฑ0.01
0.63
ฑ0.03
0.03
ฑ0.01
0.03
ฑ0.01
NA
0.01
ฑ0.01
0.04
ฑ0.01
0.01
ฑ0.01
    NA = Not available due to the criteria for calculating a quarterly and/or annual average.
    a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of
    viewing.
                                           13-14

-------
Table 13-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
                       the Kentucky Monitoring Site (Continued)
Pollutant
Vinyl Chloride
Arsenic (PM10)a
Benzo(a)pyrene a
Beryllium (PM10) a
Cadmium (PM10)a
Hexavalent Chromium a
Lead(PM10)a
Manganese (PM10)a
Naphthalene a
Nickel (PM10)a
#of
Measured
Detections
vs. # of
Samples
3/61
41/41
29/61
41/41
41/41
37/61
41/41
41/41
61/61
41/41
1st
Quarter
Average
(Ug/m3)
0.01
ฑ0.01
NA
0.04
ฑ0.03
NA
NA
0.01
ฑO.01
NA
NA
20.50
ฑ3.12
NA
2nd
Quarter
Average
(Ug/m3)
0.01
ฑ0.01
NA
0.01
ฑ0.01
NA
NA
0.01
ฑ0.01
NA
NA
12.78
ฑ3.26
NA
3rd
Quarter
Average
(Ug/m3)
0.01
ฑ0.01
0.50
ฑ0.15
0.01
ฑ0.01
0.01
ฑO.01
0.09
ฑ0.03
0.01
ฑO.01
2.01
ฑ0.82
2.73
ฑ0.84
12.89
ฑ3.19
0.28
ฑ0.07
4th
Quarter
Average
(Ug/m3)
0
0.50
ฑ0.16
0.06
ฑ0.04
0.01
ฑ0.01
0.10
ฑ0.04
0.01
ฑ0.01
2.58
ฑ1.07
4.17
ฑ2.68
20.44
ฑ3.68
0.55
ฑ0.42
Annual
Average
(Ug/m3)
0.01
ฑ0.01
NA
0.03
ฑ0.01
NA
NA
0.01
ฑO.01
NA
NA
16.59
ฑ1.84
NA
   NA = Not available due to the criteria for calculating a quarterly and/or annual average.
   a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of
   viewing.

      Observations for GLKY from Table 13-5 include the following:

      •   Annual average concentrations could not be calculated for the PMi0 metals because
          sampling did not begin until May 2011. This is also true for the carbonyl compounds
          because they did not begin sampling until August. However, Appendix L and
          Appendix N provide the pollutant-specific average concentrations for all valid metals
          and carbonyl compound samples collected over the entire sample period.

      •   None of the annual average concentrations for the pollutants of interest for GLKY,
          where they could be calculated, were greater than 1 |ig/m3. Carbon tetrachloride
          (0.63 ฑ 0.03  |ig/m3) and benzene (0.59 ฑ 0.09 |ig/m3) have the highest annual average
          concentrations for GLKY.

      •   The fourth quarter average acrylonitrile concentration is significantly higher than the
          other quarterly averages. This pollutant was not detected at all in the first quarter of
          2011, was detected twice in the second quarter,  and three times in the third. Thus, the
          other 15 measured detections were measured during the  fourth quarter. Sorting the
          measurements in a descending order almost puts the data in order by descending date,
          with the lowest concentrations measured earlier in the year and the highest
          concentrations measured late in the year.
                                          13-15

-------
       •  The fourth quarter average concentration of benzene is similar to the other quarterly
          averages in magnitude, but its confidence interval is much larger. The maximum
          concentration of benzene was measured at GLKY on November 29, 2011
          (2.63 |ig/m3). The next highest concentration measured during the fourth quarter was
          less than half that concentration (0.906 |ig/m3).

       •  Concentrations of benzo(a)pyrene were highest during the first and fourth quarters of
          2011, as indicated by the quarterly averages. Five concentrations greater than
          0.15 ng/m3 were measured at GLKY, two in January, one in October, and two in
          December. Of the 29 measured detections of benzo(a)pyrene, eight were measured
          during the first quarter, four were measured during the second quarter,  four were
          measured during the third quarter, and 13 were measured during the fourth quarter.

       •  Similar to benzo(a)pyrene, the quarterly averages for naphthalene indicate that
          concentrations tended to be lower during the warmer months and higher during the
          colder months.


       Tables 4-9 through 4-12 present the sites with the 10 highest annual average

concentrations for each of the program-level pollutants of interest. Observations for GLKY from

those tables include the following:

       •  GLKY appears in Table 4-9 for VOCs four times but does not appear in any of the
          other tables.

       •  GLKY has the third highest annual average concentration of acrylonitrile among
          NMP sites sampling VOCs, behind only SPIL and UCSD.

       •  GLKY ranked seventh for three VOCs: 1,2-dichloroethane, carbon tetrachloride, and
          hexachl oro-1,3 -butadi ene.


13.4.2  Concentration Comparison

       In order to better illustrate how a site's annual average concentrations compare to the

program-level averages, a site-specific box plot was created for the selected NATTS MQO Core

Analytes listed in Section 3.5.3, where applicable. Thus, box plots for benzene, benzo(a)pyrene,

1,3-butadiene, hexavalent chromium, and naphthalene were created for GLKY. Box plots were

not created for metals or carbonyl compounds because annual averages could not be calculated

for these pollutants. Figures 13-6 through 13-10 overlay the site's minimum, annual average, and

maximum concentrations onto the program-level minimum, first quartile, median, average, third

quartile, and maximum concentrations, as described in Section 3.5.3.
                                         13-16

-------
          Figure 13-6. Program vs. Site-Specific Average Benzene Concentration
GLKY
                                                                   Program Max Concentration = 23.8 ug/ms
                                         4         5
                                             Concentration
Program
Site:
: IstQuartile
Site Average
o
2ndQuartile SrdQuartile 4thQuartile Av(
• n
Site Minimum/Maximum
;rage

                                                                                                  10
      Figure 13-7. Program vs. Site-Specific Average Benzo(a)pyrene Concentration

                                      0.75          1          1.25
                                              Concentration (ng/mS)
                 Program:   IstQuartile     2ndQuartile     3rdQuartile     4thQuartile      Average
                 Site:      Site Average     Site Minimum/Maximum

       Figure 13-8. Program vs. Site-Specific Average 1,3-Butadiene Concentration
                                                                  i Program Max Concentration = 9.51
                                                                  L
                   2.5
                                                  1.5
                                             Concentration
                                                                                  2.5
                 Program:   IstQuartile     2ndQuartile     SrdQuartile     4thQuartile      Average
                 Site:      Site Average     Site Minimum/Maximum

                                               13-17

-------
      Figure 13-9. Program vs. Site-Specific Average Hexavalent Chromium
                                 Concentration
"-
                                        D.I 5
                                    Concentration (ng/m3)
           Program:   IstQuartile     2ndQuartile     SrdQuartile     4thQuartile     Average
           Site:      Site Average     Site Minimum/Maximum

   Figure 13-10. Program vs. Site-Specific Average Naphthalene Concentration
                                                      Program Max Concentration =779 ng/m3
                                200      250      300
                                    Concentration (ng/m3)
           Program:   IstQuartile     2ndQuartile     SrdQuartile     4thQuartile     Average
           Site:      Site Average     Site Minimum/Maximum

   Observations from Figures 13-6 through 13-10 include the following:

      •  Figure 13-6 is the box plot for benzene. Note that the program-level maximum
         concentration (23.8 |ig/m3) is not shown directly on the box plot because the scale
         of the box plot would be too large to readily observe data points at the lower end
         of the concentration range. Thus, the scale has been reduced to 10 |ig/m3. This
         box plot shows that the annual average concentration for GLKY is less than the
         program-level average concentration as well as the program-level median
         concentration. The annual average benzene concentration for GLKY is among the
         lowest annual averages for this pollutant. There were no non-detects of benzene
         measured at this site (or among sites sampling VOCs).

      •  Figure 13-7 is the box plot for benzo(a)pyrene. Note that the first quartile for this
         pollutant is zero and is not visible on this box plot. This box plot shows that the
         annual average concentration for GLKY is less than the program-level average
         concentration as well as the program-level median concentration.  Several non-
         detects of benzo(a)pyrene were measured at GLKY.

      •  Similar to benzene, the scale for 1,3-butadiene has been adjusted in Figure 13-8 as
         a result of a relatively large maximum concentration. The program-level
         maximum concentration (9.51 |ig/m3)  is not shown directly on the box plot in
         order to allow for observation of data points at the lower end  of the concentration
                                      13-18

-------
             range; thus, the scale has been reduced to 3.00 |ig/m3. Figure 13-8 shows that the
             annual average concentration of 1,3-butadiene for GLKY is less than both the
             program-level average and median concentrations. The maximum 1,3-butadiene
             concentration measured at GLKY is just greater than the program-level third
             quartile. Several non-detects of 1,3-butadiene were measured at GLKY.

          •  Figure 13-9 is the box plot for hexavalent chromium. This box plot shows that the
             annual average hexavalent chromium concentration for GLKY is just greater than
             the program-level first quartile. The annual average hexavalent chromium
             concentration for GLKY is among the lowest annual averages among NMP sites
             sampling this pollutant. The maximum hexavalent chromium concentration
             measured at GLKY is just greater than the program-level third quartile. Several
             non-detects of hexavalent chromium were measured at GLKY.

          •  Figure 13-10 is the box plot for naphthalene. Note that the program-level
             maximum concentration (779 ng/m3) is not shown directly on the box plot as the
             scale has been reduced to 500 ng/m3  to allow for observation of data points at the
             lower end of the concentration range. Figure 13-10 shows that the annual average
             concentration of naphthalene for GLKY is less than the program-level  first
             quartile. This site has the smallest range of naphthalene concentrations measured
             among all sites sampling this pollutant. Nearly the entire range of measurements
             is less than the program-level first quartile (only one measurement is greater than
             the program-level first quartile). This site has the third-lowest annual average
             concentration among NMP sites sampling PAHs.


13.4.3  Concentration Trends

       A site-specific trends evaluation was completed for sites that have sampled one or more

of the selected NATTS MQO Core Analytes for 5 consecutive years or longer, as described in
Section 3.5.4. Sampling at GLKY under the NMP began in 2008; therefore, a trends analysis was

not conducted for this site.
13.5   Additional Risk-Based Screening Evaluations

       The following risk-based screening evaluations were conducted to characterize risk at the

Kentucky monitoring site. Refer to Sections 3.3 and 3.5.5 for definitions and explanations

regarding the various toxicity factors, time frames, and calculations associated with these risk-

based screenings.
                                         13-19

-------
13.5.1  Risk-Based Screening Assessment Using MRLs
       A risk-based screening was conducted by comparing the concentration data from the
Kentucky monitoring site to the ATSDR MRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 to 14 days); intermediate (exposures of 15 to 364 days); and chronic (exposures
of 1 year or greater). The preprocessed daily measurements of the pollutants of interest were
compared to the acute MRLs; the quarterly averages were compared to the intermediate MRLs;
and the annual averages were compared to the chronic MRLs.

       As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2011.

13.5.2  Cancer Risk and Noncancer Hazard Approximations
       For the pollutants of interest for GLKY and where annual average concentrations could
be calculated, risk was examined by calculating cancer risk and noncancer hazard
approximations. These approximations can be used as risk estimates for cancer and noncancer
effects attributable to the pollutants of interest. Although the use of these approximations is
limited, they may help identify where policy-makers may want to shift or confirm their air-
monitoring priorities. Refer to Section 3.5.5.2 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.
                                         13-20

-------
                 Table 13-6. Risk Approximations for the Kentucky Monitoring Site
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Grayson, Kentucky - GLKY
Acetaldehyde
Acrylonitrile
Arsenic (PM10)a
Benzene
Benzo(a)pyrene a
Bery Ilium (PM10)a
1,3 -Butadiene
Cadmium (PM10) a
Carbon Tetrachloride
Chloroform
1 ,2-Dichloroethane
Formaldehyde
Hexachloro- 1 , 3 -butadiene
Hexavalent Chromium a
Lead(PM10)a
Manganese (PM10) a
Naphthalene a
Nickel (PM10)a
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
0.0000022
0.000068
0.0043
0.0000078
0.00176
0.0024
0.00003
0.0018
0.000006

0.000026
0.000013
0.000022
0.012


0.000034
0.00048
0.00000026
0.0000048
0.0000088
0.009
0.002
0.000015
0.03
_
0.00002
0.002
0.00001
0.1
0.098
2.4
0.0098
0.09
0.0001
0.00015
0.00005
0.003
0.00009
0.04
0.002
0.1
26/26
20/61
41/41
61/61
29/61
41/41
43/61
41/41
61/61
15/61
21/61
26/26
10/61
37/61
41/41
41/41
61/61
41/41
30/61
4/61
3/61
NA
0.16
ฑ0.07
NA
0.59
ฑ0.09
0.01
ฑ0.01
NA
0.04
ฑ0.01
NA
0.63
ฑ0.03
0.03
ฑ0.01
0.03
ฑ0.01
NA
0.01
ฑ0.01
O.01
ฑ0.01
NA
NA
0.02
ฑ0.01
NA
0.04
ฑ0.01
O.01
ฑO.01
0.01
ฑ0.01
NA
10.93
NA
4.57
0.05
NA
1.07
NA
3.75

0.72
NA
0.30
0.11
NA

0.56
NA
0.01
0.02
0.01
NA
0.08
NA
0.02
_
NA
0.02
NA
0.01
O.01
0.01
NA
0.01
O.01
NA
NA
0.01
NA
0.01
O.01
0.01
NA = Not available due to the criteria for calculating an annual average.
— = a Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 13-5.
                                                  13-21

-------
       Observations for GLKY from Table 13-6 include the following:
       •  The pollutants of interest with the highest annual average concentrations for GLKY
          are carbon tetrachloride, benzene, and acrylonitrile.
       •  The pollutants with the highest cancer risk approximations for GLKY are
          acrylonitrile, benzene, and carbon tetrachloride. Only one other pollutant of interest
          for GLKY has a cancer risk approximation greater than 1.0 in-a-million
          (1,3-butadiene).
       •  All of the noncancer hazard approximations for the pollutants of interest for GLKY
          are considerably less than an HQ of 1.0 (0.08 or less), indicating that no adverse
          health effects are expected from these individual pollutants. The highest noncancer
          hazard approximation was calculated for acrylonitrile.
       •  Annual averages, and therefore cancer risk and noncancer hazard approximations,
          could not be calculated for the metal and carbonyl compound pollutants of interest.

13.5.3 Risk-Based Emissions Assessment
       In addition to the risk-based screenings discussed above, Tables  13-7 and 13-8 present 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 2008 NEI, the 10
pollutants with the highest toxicity-weighted emissions, and the 10 pollutants with the highest
cancer risk approximations (in-a-million), as calculated from the annual  averages provided in
Table 13-6. Table 13-8 presents similar information, but identifies the 10 pollutants with the
highest noncancer hazard approximations (HQ), also calculated from annual averages provided
in Table  13-6.
                                          13-22

-------
  Table 13-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
                                                 the Kentucky Monitoring Site
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Grayson, Kentucky (Carter County) - GLKY
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
POM, Group 2b
Dichloromethane
POM, Group 6
POM, Group la
26.56
15.19
11.80
9.53
3.03
2.06
0.40
0.25
0.04
0.02
Benzene
Formaldehyde
1,3 -Butadiene
Naphthalene
POM, Group 2b
Hexavalent Chromium, PM
Ethylbenzene
POM, Group 3
Acetaldehyde
POM, Group 5a
2.07E-04
1.97E-04
9.09E-05
7.00E-05
3.56E-05
3.33E-05
2.95E-05
2.47E-05
2.10E-05
1.78E-05
Acrylonitrile
Benzene
Carbon Tetrachloride
1,3 -Butadiene
1 ,2-Dichloroethane
Naphthalene
Hexachloro- 1 , 3 -butadiene
Hexavalent Chromium
Benzo(a)pyrene
Trichloroethylene
10.93
4.57
3.75
1.07
0.72
0.56
0.30
0.11
0.05
0.02
OJ
to

-------
  Table 13-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
                                            RfCs for the Kentucky Monitoring Site
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer Hazard
Approximation
(HQ)
Grayson, Kentucky (Carter County) - GLKY
Toluene
Xylenes
Benzene
Methanol
Hexane
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Styrene
59.99
44.41
26.56
15.98
15.43
15.19
11.80
9.53
3.03
2.50
Acrolein
Formaldehyde
1,3 -Butadiene
Cyanide Compounds, gas
Acetaldehyde
Benzene
Naphthalene
Xylenes
Arsenic, PM
Propionaldehyde
52,572.66
1,549.95
1,514.80
1,335.64
1,058.92
885.42
686.00
444.15
138.10
111.83
Acrylonitrile
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Naphthalene
Trichloroethylene
Tetrachloroethylene
Chloroform
Hexachloro- 1 ,3 -butadiene
Hexavalent Chromium
0.08
0.02
0.02
0.01
0.01
<0.01
<0.01
<0.01
0.01
0.01
OJ
to

-------
       The pollutants listed in Tables 13-7 and 13-8 are limited to those that have cancer and

noncancer toxicity factors, respectively. As a result, although the actual value of the emissions is

the same, the highest emitted pollutants in the cancer table may be different from the noncancer

table. Further, the cancer risk and noncancer hazard approximations based on each site's annual

averages are limited to those pollutants for which each respective site sampled. As discussed in

Section 13.3, GLKY sampled for hexavalent chromium, carbonyl compounds, PMio metals,

PAHs, and VOCs. 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. As

mentioned in Section 13.5.2, because annual averages could not be calculated for the metals or

carbonyl compounds, cancer risk and noncancer hazard approximations were not calculated. A
more in-depth discussion of this analysis is provided in Section 3.5.5.3. Similar to the cancer risk

and noncancer hazard approximations, 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 highest emitted pollutants with
          cancer UREs in Carter County. The emissions for this county are low compared to
          other counties with NMP sites.

       •  Benzene, formaldehyde, and 1,3-butadiene  are the pollutants with the highest
          toxicity-weighted emissions (of the pollutants with cancer UREs) for Carter County.

       •  Seven of the highest emitted pollutants also have the highest toxicity-weighted
          emissions for Carter County. Note that benzene and formaldehyde top both lists.

       •  Acrylonitrile, which has the highest cancer  risk approximation among the pollutants
          of interest for GLKY, appears on neither emissions-based list. This is also true for
          carbon tetrachloride, which has the highest  annual average concentration and the third
          highest cancer risk approximation for GLKY.

       •  Benzene,  1,3-butadiene, and naphthalene appear on both emissions-based lists for
          Carter County and have one of the 10 highest cancer risk approximations for this site.

       •  Hexavalent chromium, which has the eighth highest cancer risk approximation, ranks
          sixth for toxicity-weighted emissions but is not one of the highest emitted pollutants
          in Carter County.

       •  Three POM Groups appear among the highest emitted pollutants (POM, Groups la,
          2b,  and 6). Three POM Groups also appear among the pollutants with the highest
          toxicity-weighted emissions (POM, Groups 2b, 3, and 5a). Benzo(a)pyrene, a

                                          13-25

-------
          pollutant of interest for GLKY, is part of POM Group 5a. Several pollutants, such
          acenaphthene, benzo(e)pyrene, and fluorene, are measured using Method TO-13 and
          are part of POM, Group 2b, which appears on both emissions-based lists. Several
          other pollutants, such benzo(a)anthracene and indeno(l,2,3-cd)pyrene, are also
          measured using Method TO-13 and are part of POM, Group 6, which is one of the
          highest emitted "pollutants" in Carter County.


       Observations from Table 13-8 include the following:

       •  Toluene, xylenes, and benzene are the highest emitted pollutants with noncancer
          RfCs in Carter County.

       •  The pollutant with the highest toxicity-weighted emissions (of the pollutants with
          noncancer RfCs) is acrolein. Although acrolein was sampled for at GLKY, this
          pollutant was excluded from the pollutants of interest designation, and thus
          subsequent risk-based screening evaluations, due to questions about the consistency
          and reliability of the measurements, as discussed in Section 3.2. Acrolein does not
          appear among Carter County's highest emitted pollutants.

       •  Five of the highest emitted pollutants also have the highest toxicity-weighted
          emissions for Carter County.

       •  Acrylonitrile, which has the highest noncancer hazard approximation among the
          pollutants of interest for GLKY, appears on neither emissions-based list. Conversely,
          benzene and 1,3-butadiene appear on all three lists.


13.6   Summary of the 2011 Monitoring Data for GLKY

       Results from several of the data treatments described in this section include the
following:

       *ป*  Fifteen pollutants, including nine NA TTS MQO  Core Analytes, failed screens for
          GLKY.

       ปซป  None of the annual average concentrations of the pollutants of interest for GLKY,
          where they could be calculated, were greater than 1 jug/m .

       ปซป  GLKY had the third highest annual average concentration of acrylonitrile among
          NMP sites sampling VOCs.

       ปซป  GLKY had the smallest range of naphthalene concentrations measured among all
          NMP sites sampling this pollutant.

       ปซป  Because sampling for PMw metals and car bony I compounds did not begin until
          May 2011 and August 2011, respectively, annual average concentrations could not be
          calculated for these pollutants.
                                         13-26

-------
 14.0  Site in Massachusetts
       This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS site in Massachusetts, and integrates these concentrations
with emissions, meteorological, and risk information. Data generated by sources other than ERG
are not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) for detailed discussions and definitions
regarding the various data analyses presented below.

14.1   Site Characterization
       This section characterizes the BOMA monitoring site by providing geographical and
physical information about the location of the site and the surrounding area. This information is
provided to give the reader insight regarding factors that may influence the air quality near the
site and assist in the interpretation of the ambient monitoring measurements.

       The BOMA monitoring site is located in Boston. Figure 14-1 is a composite satellite
image retrieved from ArcGIS Explorer showing the monitoring site in its urban location.
Figure 14-2 identifies nearby point source emissions locations by source category, as reported in
the 2008 NEI for point sources. Note that only sources within 10 miles of the site are included in
the facility counts provided in Figure 14-2. A 10-mile boundary was chosen to give the reader an
indication of which emissions sources and emissions source categories could potentially have a
direct effect on the air quality at the monitoring site. Further, this boundary provides both the
proximity of emissions sources to the monitoring site as well as the quantity of such sources
within a given distance of the site. Sources outside the 10-mile radius are still visible on the map,
but have been grayed out in order to show emissions sources just outside the boundary.
Table 14-1 provides  supplemental geographical information such as land use, location setting,
and locational coordinates.
                                           14-1

-------
Figure 14-1. Boston, Massachusetts (BOMA) Monitoring Site

-------
   Figure 14-2. NEI Point Sources Located Within  10 Miles of BOMA
?nd
 SOMANATTSsite
                         10 mile radius
                                                /V          7riQ'0"W
                                                       Note: Due to facility density and collocation, the total facilities
                                                       displayed may not represent a II facilities within the area of interest.

                                                 County boundary
e
B
C
•
e
e
I
E
                                                         I
                                                        *
Source Category Group (No. of Facilities)
•$1 Aerospace/Aircraft Manufacturing (1)
 v  Air-conditioning/Refrigeration (3)
 -t< Aircraft Operations (18)
 0  Auto Body Shop/Painters (1)
 A Automobile/Truck Manufacturing (1)
    Bakery (4)
    Bulk Terminals/Bulk Plants (7)
    Chemical Manufacturing (1)
    Concrete Batch Plant (1)
    Dry Cleaning (2)
    Electrical Equipment (3)
    Electricity Generation via Combustion (8)                  ^
    Electroplating. Plating, Polishing, Anodizing, and Coloring (4) p
 ฉ Fabricated Metal Products (5)                            A^
 F  Food Processing/Agriculture (1)                          >
 n Furniture Plant (1)
 f  Gasoline/Diesel Service Station (2)
 ffl  Hospital (4)
                                                       Hot Mix Asphalt Plant (2)
                                                       Institutional - school (39)
                                                       Iron and Steel Foundry (1)
                                                       Laboratory (2)
                                                    •  Landfill (1)
                                                    L  Large Appliance Manufacturing (1)
                                                    V Mineral Products (1)
                                                    ? Miscellaneous Commercial/Industrial (38)
                                                    M  Miscellaneous Manufacturing (6)
                                                    H Municipal Waste Combustor(l)
                                                    — Pharmaceutical Manufacturing (1)
                                                       Printing/Publishing (1)
                                                       Pulp and Paper PlanlWood Products (1)
                                                       Rubber and Miscellaneous Plastics Products (2)
                                                    ii. Ship Building and Repairing (2)
                                                    >  Solid Waste Disposal - Commercial/Institutional (2)
                                                    S  Surface  Coating (3)
                                                    Tt Telecommunications (4)
                                                    1  Wastewater Treatment {2)
                                                   14-3

-------
                           Table 14-1. Geographical Information for the Massachusetts Monitoring Site

Site
Code

BOMA



AQS Code

25-025-0042



Location

Boston



County

Suffolk

Micro- or
Metropolitan
Statistical Area
Boston-
Cambridge-
Quincy, MA-NH
MSA (Boston Div)
Latitude
and
Longitude

42.32944,
-71.0825



Land Use

Commercial


Location
Setting

Urban/City
Center



Additional Ambient Monitoring Information1
Lead (TSP), CO, VOCs, SO2, NO, NO2, NOX,
PAMS/NMOCs, Carbonyl compounds, O3,
Meteorological parameters, PM10, Black carbon,
PM2 5, PM2 5 Speciation.
BOLD ITALICS = EPA-designated NATTS Site

-------
       The BOMA monitoring site is located at Dudley Square in Roxbury, a southwest
neighborhood of Boston and is the Roxbury NATTS site. The surrounding area is commercial as
well as residential, as shown in Figure 14-1. The monitoring site is approximately 1.25 miles
south of 1-90 and 1 mile west of 1-93. The original purpose for the location of this site was to
measure population exposure to a city bus terminal located across the street from the monitoring
site. In recent years, the buses servicing the area were converted to compressed natural gas
(CNG). As Figure 14-2 shows, BOMA is located near a large number of point sources, with a
high density of sources located a few miles to the west, northwest, and north of the site. The
source categories with the highest number of emissions sources surrounding BOMA include
institutional facilities (schools); aircraft operations, which includes airports as well as small
runways, heliports, or landing pads; and electricity generating units (via combustion).

       Table 14-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Massachusetts monitoring site. Table 14-2 includes county-level
population and vehicle registration information. Table 14-2 also includes a county-level vehicle
registration-to-population ratio, which was calculated to represent the number of vehicles per
person within the monitoring site's residing county. In addition, the population within 10 miles
of the site is presented, based on postal code population data estimates. An estimate of 10-mile
vehicle ownership was determined by applying the county-level vehicle registration-to-
population ratio to the 10-mile population surrounding the monitoring site. Table 14-2 also
contains traffic volume information for BOMA. Finally, Table 14-2 presents the county-level
daily VMT for Suffolk County.
   Table 14-2. Population, Motor Vehicle, and Traffic Information for the Massachusetts
                                      Monitoring Site



Site
BOMA

Estimated
County
Population1
730,932

County-level
Vehicle
Registration2
481,199
Vehicles per
Person
(Registration:
Population)
0.66

Population
within 10
miles3
1,671,730
Estimated
10-mile
Vehicle
Ownership
1,100,560
Annual
Average
Daily
Traffic4
31,400

County-
level Daily
VMT5
10,695,874
 Bounty-level population estimate reflects 2011 data from the U.S. Census Bureau (Census Bureau, 2012b)
 2County-level vehicle registration reflects 2011 data from the Massachusetts Registry of Motor Vehicles (MA
  RMV, 2012)
 310-mile population estimate reflects 2011 data from Xionetic (Xionetic, 2011)
 4AADT reflects 2007 data from the Massachusetts DOT (MA DOT, 2007)
 5County-level VMT reflects 2011 data for from the Massachusetts DOT (MA DOT, 2013)
 BOLD ITALICS = EPA-designated NATTS Site
                                            14-5

-------
       Observations from Table 14-2 include the following:
       •  The Suffolk County population is in the middle of the range compared to other
          counties with NMP sites, while BOMA's 10-mile population is among the higher
          10-mile populations (ranking 6th).
       •  Similar to the populations, the Suffolk County vehicle registration is in the middle of
          the range compared to other counties with NMP sites, while its 10-mile estimated
          vehicle ownership is among the higher estimates (ranking 5th).
       •  The vehicle-per-person ratio is the fourth lowest ratio when compared to other NMP
          sites.
       •  The traffic volume experienced near BOMA is in the middle of the range compared to
          other NMP sites. The traffic estimate provided is for Melnea Cass Boulevard between
          Washington Street and Harrison Avenue.
       •  The daily VMT for Suffolk County is in the middle of the range compared to other
          counties with NMP sites (where VMT data were available).

14.2   Meteorological Characterization
       The following sections characterize the meteorological conditions near the monitoring
site in Massachusetts on sample days, as well as over the course of the year.
14.2.1  Climate Summary
       Boston's New England location ensures that the city experiences a fairly active weather
pattern. Storm systems frequently track across the region, bringing ample precipitation to the
area. The proximity to the Atlantic Ocean helps moderate temperatures, both in the summer and
the winter, while at the same time allowing winds to gust higher than they would farther inland.
Winds generally flow from the northwest in the winter and southwest in the summer. Coastal
storm systems called "Nor'easters," strong low pressure systems that produce heavy rain or snow
and winds up to hurricane strength along the Mid-Atlantic and northeast coastal states, often
produce the heaviest  snowfalls for the area (Bair, 1992 and NOAA, 2013a).
                                          14-6

-------
14.2.2 Meteorological Conditions in 2011
       Hourly meteorological data from the NWS weather station nearest this site were retrieved
for 2011 (NCDC, 2011). The closest weather station to BOMA is located at Logan International
Airport (WBAN 14739). Additional information about the Logan Airport weather station, such
as the distance between the site and the weather station, is provided in Table 14-3. These data
were used to determine how meteorological conditions on sample days vary from conditions
experienced throughout the year.

       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 the entire year for 2011. Also included in Table 14-3 is the
95 percent confidence interval for each parameter. As  shown in Table 14-3, average
meteorological conditions on sample days at BOMA were representative of average weather
conditions throughout the year.

14.2.3 Back Trajectory Analysis
       Figure 14-3 is the composite back trajectory map for days on which samples were
collected at the BOMA monitoring site in 2011. Included in Figure  14-3 are four back
trajectories per sample day. Figure 14-4 is the corresponding cluster analysis. An in-depth
description of these maps and how they were generated is presented in  Section 3.5.2.1. For the
composite map, each line represents the 24-hour trajectory along which a parcel of air traveled
toward the monitoring site on a given sample day and time, based on an initial height of
50 meters AGL. For the cluster analysis, each line corresponds to a  trajectory representative of a
given cluster of back trajectories. Each concentric circle around the site in Figures 14-3 and 14-4
represents 100 miles.
                                          14-7

-------
                       Table 14-3. Average Meteorological Conditions near the Massachusetts Monitoring Site
Closest NWS
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site


Average
Type1
Average
Maximum
Temperature
(ฐF)

Average
Temperature
(ฐF)
Average
Dew Point
Temperature
(ฐF)
Average
Wet Bulb
Temperature
(ฐF)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Boston, Massachusetts - BOMA
Logan
International
Airport
14739
(42.36, -71.01)
4.05
miles

42ฐ
(NE)
Sample
Day


2011
60.5
ฑ4.3


60.2
ฑ1.8
53.3
ฑ4.1


53.2
ฑ1.8
41.4
ฑ4.7


41.2
ฑ1.9
47.9
ฑ3.9


47.7
ฑ1.6
67.0
ฑ4.0


66.4
ฑ1.6
1014.5
ฑ2.0


1015.0
ฑ0.8
9.4
ฑ0.8


9.1
ฑ0.3
     Sample day averages are highlighted to help differentiate the sample day averages from the full-year averages.
oo

-------
Figure 14-3. 2011 Composite Back Trajectory Map for BOMA
    Figure 14-4. Back Trajectory Cluster Map for BOMA
                           14-9

-------
        Observations from Figures 14-3 and 14-4 include the following:

       •  The composite back trajectory map shows that back trajectories originated from a
          variety of directions at BOMA. The predominant direction of back trajectory origin is
          to the northwest and south of the site, with the longest trajectories originating to the
          west and northwest over Canada.

       •  The 24-hour air shed domain for BOMA is similar in size to other NMP monitoring
          sites. The farthest away a back trajectory originated was nearly 600 miles, over
          Quebec, Canada, although back trajectories of similar length also originated over
          southern Ontario, Canada. The average trajectory length was 251 miles and most back
          trajectories (84 percent) originated within 400 miles of the monitoring site.

       •  Nearly 40 percent of back trajectories originated within 200 miles of BOMA, as
          indicated  by the short cluster trajectory originating to the south of the site. This
          cluster trajectory represents back trajectories originating from all directions of
          relatively short length. It is important to recall that the HYSPLIT model includes both
          distance and direction when determining clusters. Another 21 percent of back
          trajectories originated to the north of BOMA, over the eastern Canadian provinces.
          Seventeen percent of back trajectories originated to the south of BOMA, although this
          cluster trajectory also includes back trajectories originating to the  southwest over the
          Mid-Atlantic states and Pennsylvania. The direction from which the least number of
          back trajectories originated was east, at eight percent.


14.2.4 Wind Rose Comparison

       Hourly surface wind data from the NWS weather station at Logan International Airport

near BOMA were uploaded into a wind rose software program to produce customized wind

roses, as described in Section 3.5.2.2. A wind rose shows the frequency of wind directions using

"petals" positioned around a 16-point compass, and uses different colors to represent wind

speeds.
       Figure 14-5 presents a map showing the distance between the NWS station and BOMA,

which may be useful for identifying topographical influences that may affect the meteorological

patterns experienced at this location. Figure 14-5 also presents three different wind roses for the

BOMA monitoring site. First, a historical wind rose representing 2001 to 2010 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 2011 is presented.

Next, a wind rose representing wind data for days on which samples were collected in 2011 is

presented. These can be used to identify the predominant wind speed and direction for 2011 and
                                          14-10

-------
determine if wind observations on sample days were representative of conditions experienced

over the entire year and historically.


       Observations from Figure 14-5 for BOMA include the following:

       •  The Logan International Airport weather station is located approximately 4 miles
          northeast of BOMA. Note that the airport is located on a peninsula on Boston Harbor
          with downtown Boston to the west, Chelsea to the north, and Winthrop to the east,
          while the BOMA monitoring site is located west of South Boston, farther inland.

       •  The historical wind rose shows that calm winds (< 2 knots) account for only three
          percent of wind observations. Winds with a westerly component (south-southwest to
          north-northwest) make up the bulk (nearly 60 percent) of winds greater than 2 knots.

       •  The wind patterns shown on the 2011 wind rose resemble the historical wind patterns,
          indicating that wind conditions during 2011 were typical of conditions experienced
          historically near BOMA.

       •  The sample day wind patterns resemble the full-year and historical wind patterns,
          indicating that wind conditions on sample days were representative of those
          experienced over the entire year and historically.
                                          14-11

-------
Figure 14-5. Wind Roses for the Logan International Airport Weather Station near BOMA
  Distance between BOMA and NWS Station
2001-2010 Historical Wind Rose
  L vv^x   /•
                                              ,'VEsr
            2011 Wind Rose
    Sample Day Wind Rose
                                      14-12

-------
14.3   Pollutants of Interest
       Site-specific "pollutants of interest" were determined for the Massachusetts monitoring
site in order to allow 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." 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. In
addition, if any of the NATTS MQO Core Analytes measured by the monitoring site did not
meet the pollutant of interest criteria based on the preliminary risk-based screening, that pollutant
was added to the list of site-specific pollutants of interest. A more in-depth description of the
risk-based screening process is presented in Section 3.2.

       Table 14-4 presents the results of the preliminary risk-based screening process for
BOMA. The pollutants that failed at least one screen and contributed to 95 percent of the total
failed screens for the monitoring site are shaded. NATTS MQO Core Analytes are bolded. Thus,
pollutants of interest are shaded and/or bolded. BOMA sampled for PMio metals, PAHs, and
hexavalent chromium.

      Table 14-4. Risk-Based Screening Results for the Massachusetts Monitoring Site
Pollutant
Screening
Value
(Hg/m3)
# of Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
Boston, Massachusetts - BOMA
Naphthalene
Arsenic (PM10)
Manganese (PM10)
Nickel (PM10)
Acenaphthene
Fluorene
Hexavalent Chromium
0.029
0.00023
0.005
0.0021
0.011
0.011
0.000083
Total
54
48
12
5
1
1
1
122
61
60
60
60
61
61
54
417
88.52
80.00
20.00
8.33
1.64
1.64
1.85
29.26
44.26
39.34
9.84
4.10
0.82
0.82
0.82
44.26
83.61
93.44
97.54
98.36
99.18
100.00

       Observations from Table 14-4 include the following:
       •   Seven pollutants failed at least one screen for BOMA; of these, five are NATTS
           MQO Core Analytes.
                                          14-13

-------
       •  Naphthalene accounted for nearly half of the total failed screens for BOMA.
          Naphthalene and arsenic together account for nearly 85 percent of the total failed
          screens.
       •  Four pollutants, naphthalene and three PMio metals, were initially identified as
          pollutants of interest for BOMA. Hexavalent chromium was added  to the pollutants
          of interest for BOMA because it is a NATTS MQO Core Analyte, even though it did
          not contribute to 95 percent of the failed screens. Benzo(a)pyrene, beryllium,
          cadmium, and lead were also added to BOMA's pollutants of interest because they
          are NATTS MQO Core Analytes, even though they did not fail any screens. These
          four pollutants are not shown in Table 14-4 but are shown in subsequent tables in the
          sections that follow.

14.4   Concentrations
       This section presents various concentration averages used to characterize pollution levels
at the Massachusetts monitoring site. Where applicable, the following calculations and data
analyses were performed:  Time period-based concentration averages (quarterly and annual) are
provided for the pollutants of interest for BOMA, where the data meet the applicable criteria.
Concentration averages for select pollutants are also presented graphically to illustrate how the
site's concentrations compare to the program-level averages, as presented in Section 4.1. In
addition, concentration averages for select pollutants are presented from previous years of
sampling in order to characterize concentration trends at the site. Additional site-specific
statistical summaries for BOMA are provided in Appendices M through O.

14.4.1 2011 Concentration Averages
       Quarterly and annual concentration averages were calculated for the pollutants of interest
for BOMA,  as described in Section 3.1. The quarterly average of a particular pollutant is simply
the average  concentration  of the preprocessed daily measurements over a given calendar quarter.
Quarterly average concentrations include the substitution of zeros for all non-detects. A site must
have a minimum of 75  percent valid samples of the total  number of samples possible within a
given quarter for a quarterly average to be calculated. An annual average includes all measured
detections and substituted zeros for non-detects for the entire year of sampling. Annual averages
were calculated for pollutants where three  valid quarterly averages could be calculated and
where method completeness was greater than or equal to 85 percent, as presented in Section 2.4.
Quarterly and annual average concentrations for BOMA are presented in Table 14-5, where
applicable. Note that if a pollutant was not detected in a given calendar quarter, the quarterly
                                          14-14

-------
average simply reflects "0" because only zeros substituted for non-detects were factored into the
quarterly average concentration.
          Table 14-5. Quarterly and Annual Average Concentrations of the Pollutants of
                     Interest for the Massachusetts Monitoring Site
Pollutant
#of
Measured
Detections vs.
# of Samples
1st
Quarter
Average
(ng/m3)
2nd
Quarter
Average
(ng/m3)
3rd
Quarter
Average
(ng/m3)
4th
Quarter
Average
(ng/m3)
Annual
Average
(ng/m3)
Boston, Massachusetts - BOMA
Arsenic (PM10)
Benzo(a)pyrene
Beryllium (PM10)
Cadmium (PM10)
Hexavalent Chromium
Lead (PM10)
Manganese (PM10)
Naphthalene
Nickel (PM10)
60/60
61/61
60/60
60/60
54/61
60/60
60/60
61/61
60/60
0.26
ฑ0.06
0.15
ฑ0.04
0.03
ฑ0.03
0.13
ฑ0.02
0.02
ฑ0.01
3.18
ฑ1.25
2.77
ฑ0.85
55.51
ฑ11.52
1.29
ฑ0.21
0.45
ฑ0.14
0.11
ฑ0.03
0.01
ฑ0.01
0.15
ฑ0.03
0.03
ฑ0.01
3.50
ฑ1.10
4.58
ฑ1.15
55.96
ฑ 12.84
1.68
ฑ0.45
0.61
ฑ0.16
0.07
ฑ0.03
0.01
ฑ0.01
0.14
ฑ0.02
0.04
ฑ0.02
2.89
ฑ0.54
3.42
ฑ0.64
75.12
ฑ 16.68
1.39
ฑ0.43
0.38
ฑ0.11
0.12
ฑ0.02
0.01
ฑ0.01
0.10
ฑ0.02
0.02
ฑ0.01
2.74
ฑ0.71
3.11
ฑ0.75
53.83
ฑ 12.85
1.16
ฑ0.20
0.43
ฑ0.07
0.11
ฑ0.02
0.01
ฑ0.01
0.13
ฑ0.01
0.03
ฑ0.01
3.07
ฑ0.44
3.48
ฑ0.44
60.35
ฑ6.84
1.38
ฑ0.17
       Observations for BOMA from Table 14-5 include the following:

       •  Naphthalene is the pollutant with the highest annual average concentration by mass
          (60.35 ฑ 6.84 ng/m3). The annual average concentrations for the remaining pollutants
          of interest are at least an order of magnitude lower.

       •  Of the PMio metals, manganese is the pollutant with the highest annual average
          concentration (3.48 ฑ 0.44 ng/m3).

       •  The first quarter concentration of beryllium (0.03 ฑ 0.03 ng/m3) is higher than the
          other quarterly averages (each of them is less than 0.01 ng/m3). A review of the data
          shows that the maximum concentration of beryllium was measured at BOMA on
          January 21, 2011 (0.202 ng/m3). This measurement is an order of magnitude higher
          than the next highest concentration measured at BOMA and the fifth highest
          beryllium  concentration measured across the program. In addition, all seven
          beryllium  concentrations greater than 0.01 ng/m3 were measured during the first
          quarter of 2011  (four in January, one in February, and two in March).
                                         14-15

-------
       •  The first and second quarter average lead concentrations are higher than the other
          quarterly averages and have larger confidence interval associated with them. A
          review of the data shows that the maximum lead concentration was measured on
          February 8, 2011 (10.4 ng/m3) although a similar concentration was also measured on
          June 8, 2011 (10.3 ng/m3). These two measurements are twice the next highest
          concentration (5.61 ng/m3). The median lead concentration for BOMA is 2.78 ng/m3.

       •  The second quarter average manganese concentration is higher than the other
          quarterly averages and has a larger confidence interval than the others. A review of
          the data shows that the two maximum manganese concentrations were measured on
          June 8, 2011 (9.48 ng/m3) and June 2, 2011 (7.08 ng/m3).


       Tables 4-9 through 4-12 present the sites with the 10 highest annual average

concentrations for each of the program-level pollutants of interest. Observations for BOMA from

those tables include the following:

       •  BOMA's annual average concentration of benzo(a)pyrene ranks seventh highest
          among sites sampling PAHs.

       •  BOMA's annual average concentration of beryllium ranks third highest among other
          sites sampling PMio metals. BOMA also ranks fourth for nickel and fifth for
          cadmium and lead.

       •  BOMA's annual average concentration of hexavalent chromium ranks seventh
          highest among sites sampling this pollutant.


14.4.2  Concentration Comparison

       In order to better illustrate how a site's annual  average concentrations compare to the

program-level averages, a site-specific box plot was created for the selected NATTS MQO Core

Analytes listed in Section 3.5.3, where applicable. Thus, box plots for arsenic, benzo(a)pyrene,

hexavalent chromium, lead, manganese, and naphthalene were created for BOMA. Figures 14-6

through 14-11 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 3.5.3.
                                         14-16

-------
      Figure 14-6. Program vs. Site-Specific Average Arsenic (PMi0) Concentration
EC IV a
              DL5
                                  15
                                             2         2.5
                                             Concentration (ng/m3)
                                                                           3.5
                                                                                               4.5
                Program:   IstQuartile     2ndQuartile      3rdQuartile      4thQuartile      Average
                Site:      Site Average     Site Minimum/Maximum

      Figure 14-7. Program vs. Site-Specific Average Benzo(a)pyrene Concentration
EC IV a
E
               035          0.5          0.75          1          1.25         1.5          1.75
                                             Concentration (ng/m3)
                Program:   IstQuartile     2ndQuartile      3rdQuartile      4thQuartile      Average
                Site:      Site Average     Site Minimum/Maximum

  Figure 14-8. Program vs. Site-Specific Average Hexavalent Chromium Concentration
    n-
                                  3.1
                                             Concentration (ng/m3)

                                                                               0.25
                                                                                               :.=
                Program:   IstQuartile     2ndQuartile      SrdQuartile      4thQuartile     Average
                Site:      Site Average     Site Minimum/Maximum
                                              14-17

-------
        Figure 14-9. Program vs. Site-Specific Average Lead (PMi0) Concentration
BCtv'i
                             i:
                                         15            20
                                          Concentration (ng/m3)
                                                                                           35
               Program:   IstQuartile     2ndQuartile     SrdQuartile     4thQuartile     Average
               Site:      Site Average     Site Minimum/Maximum

    Figure 14-10. Program vs. Site-Specific Average Manganese (PMi0) Concentration
EC IV A
                                                          ; Program Max Concentration = 395 ng/m3
                          50
                                     75
                                               100
                                           Concentration (ng/m3)
                                                          125
                                                                     153
                                                                                175
               Program:   IstQuartile     2ndQuartile     SrdQuartile     4thQuartile     Average
               Site:
Site Average
                                     Site Minimum/Maximum

                                                                                           2DD
       Figure 14-11. Program vs. Site-Specific Average Naphthalene Concentration
BO IV A

                                                              Program Max Concentration =779 ng/m3
                     100
                                      200       250       300
                                           Concentration (ng/m3)
                                                                 353
                                                                         4"'
                                                                                  45:
Program
Site:
: IstQuartile
Site Average
o
2ndQuartile SrdQuartile 4thQuartile Ave
Site Minimum/Maximum
rage

      Observations from Figures 14-6 through 14-11 include the following:

          •   Figure 14-6 shows that BOMA's annual average arsenic (PMio) concentration is
              less than the program-level average concentration for arsenic (PMio) but is similar
              to the program-level median concentration. The maximum concentration
              measured at BOMA is considerably less than the maximum concentration
              measured at the program level. There were no non-detects of arsenic measured at
              BOMA.
                                            14-18

-------
•  Figure 14-7 is the box plot for benzo(a)pyrene. Note that the first quartile for this
   pollutant is zero and is not visible on this box plot. This box plot shows that the
   annual average  concentration for BOM A is greater than the program-level
   average concentration and roughly equivalent to the program-level third quartile.
   Figure 14-7 also shows that the maximum concentration measured at BOMA is
   less than the maximum concentration measured across the program. There were
   no non-detects of benzo(a)pyrene measured at BOMA.

•  Figure 14-8 is the box plot for hexavalent chromium. This figure shows that the
   annual average  concentration of hexavalent chromium for BOMA is just greater
   than the program-level average concentration. The maximum concentration
   measured at BOMA is less than the maximum concentration measured at the
   program level. Several non-detects of hexavalent chromium were measured at
   BOMA.

•  Figure 14-9 shows that the annual average lead (PMio) concentration for BOMA
   is less than the program-level average concentration but greater than the program-
   level median concentration. The maximum lead concentration measured at
   BOMA is less than the maximum concentration measured at the program level.
   There were no non-detects of lead at BOMA or across the program.

•  Figure 14-10 is  the box plot for manganese (PMio). Note that the program-level
   maximum concentration (395 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
   200 ng/m3. Figure 14-10 shows that the range of manganese concentrations
   measured at BOMA is relatively small compared to the range of manganese
   concentrations measured across the program. Although difficult to discern in
   Figure 14-10, the annual average manganese (PMio) concentration for BOMA  is
   less than half the program-level average concentration and is less than the
   program-level median concentration as well. The maximum manganese
   concentration measured at BOMA is less than the program-level third quartile  and
   just greater than the program-level average concentration.

•  Similar to manganese, the program-level maximum concentration (799 ng/m3) is
   not shown directly on the box plot for naphthalene in Figure 14-11 as the scale
   has been reduced to 500 ng/m3 to  allow for observation of data points at the lower
   end of the concentration range. Figure 14-11 shows that the annual average
   naphthalene concentration for BOMA is less than both the program-level average
   and median concentrations. The maximum concentration measured at BOMA is
   considerably less than the maximum concentration measured at the program level.
   There were no non-detects of naphthalene measured at BOMA.
                               14-19

-------
14.4.3  Concentration Trends
       A site-specific trends evaluation was completed for sites that have sampled one or more
of the selected NATTS MQO Core Analytes for 5 consecutive years or longer, as described in
Section 3.5.4. BOMA has been sampling metals since 2003 and hexavalent chromium since
2005. Thus, Figures 14-12 through 14-15 present the annual statistical metrics for arsenic,
hexavalent chromium, lead, and manganese for BOMA, respectively. The statistical metrics
presented for calculating trends include the substitution of zeros for non-detects. BOMA began
sampling PAHs under the NMP in 2008; thus, the trends analysis was not conducted for the
PAHs because the 5 consecutive year criterion is not met.
        Figure 14-12. Annual Statistical Metrics for Arsenic (PMio) Concentrations
                                  Measured at BOMA



m"
E
Average Go nee n


















T



2005







b . ,
dta


	

2006
• SthPercentile
T
I


2007
- Minirrurr


"


— Median

...<




T
r^H 1 =L
1 r~^
^m '^ 	 ••• 	 ซ•-

2008 2009 2010 2011
Year
— Maximum • 95th Percentile *ซ•+.. Average
                                         14-20

-------
Figure 14-13. Annual Statistical Metrics for Hexavalent Chromium Concentrations

                               Measured at BOMA
    .a

    5
    S 0.3

    8
    8
    &

    S
    < 0.2
                                          2008

                                          Year
              5th Percentile    - Minimum     — Median     —  Maximum     • 95th Percentile   "^--Average
      Figure 14-14 Annual Statistical Metrics for Lead (PMio) Concentrations

                               Measured at BOMA



^
tration (ng/r
5 ;
rage Concen
g „
<
5 -





















2005




I

**m
*
2006
• 5th Percentile


	

MinilTLlIT













r
,— 5-,
	 ' 1 	 1
• 	 ^
^MH H^H
t 1 	 a—I i 	 f— ' I
2007 2008 2009 2010 2011
Year
— Median - Maximum • 95th Percentile ซซ•*" Average
                                       14-21

-------
      Figure 14-15. Annual Statistical Metrics for Manganese (PMi0) Concentrations
                                  Measured at BOMA
                                             :DOS
                                             Year
                            —  Minimum    —  Median
                                                — Maximum
                                                             95thPercentile
                                                                        ••*•• Average
       Observations from Figure 14-12 for arsenic measurements at BOMA include the

following:

       •  While PMio metals sampling began in 2003, data from that year were excluded from
          this analysis because sampling did not begin until October. In addition, samples were
          not collected in parts of April, May, September, and October 2004, which results in
          the dataset not meeting the criteria specified in Section 3.5.4; thus, 2004 is also
          excluded from Figure 14-12.

       •  The maximum arsenic concentration shown was measured on July 5, 2008. The next
          highest concentration measured is approximately half as high and was measured on
          July 4, 2006.

       •  Figure 14-12 shows that while there have been fluctuations in the average
          concentration of arsenic at BOMA, the differences are not significant. The average
          concentrations of arsenic have ranged from 0.36 ng/m3 (2010) to 0.61 ng/m3 (2008).
          The median concentrations follow a similar trend but with less variation, ranging
          from 0.34 ng/m3 (2010) to 0.46 ng/m3 (2008).

       •  The minimum concentration measured for each year is greater than zero, indicating
          that there have been no non-detects of arsenic measured at BOMA since 2005.
                                          14-22

-------
       Observations from Figure 14-13 for hexavalent chromium measurements at BOMA

include the following:

       •   The maximum hexavalent chromium concentration was measured in 2008
          (0.525 ng/m3). Less than 10 percent of hexavalent chromium concentrations measured
          at BOMA are greater than 0.1 ng/m3. At least one concentration greater than
          0.1 ng/m3 has been measured in each year since the onset of sampling, with 2005
          having the most (eight) and 2011 having the least (one).

       •   The range of measurements has varied each year, as indicated by both the maximum
          concentration and the 95th percentile. The 95th percentile for 2008 is greater than the
          maximum concentrations for 2010 and 2011.

       •   The average concentration decreased significantly from 2006 to 2007, then increased
          for 2008. A decreasing trend is also shown after 2008, although a slight increase is
          shown for 2011.

       •   The minimum and 5th percentile are both zero for each year of sampling, indicating
          the presence of non-detects. The percentage  of non-detects has varied between seven
          percent (2006) to 26 percent (2009).


       Observations from Figure 14-14 for lead measurements at BOMA include the following:

       •   The maximum lead concentration shown (37.9 ng/m3) was measured in 2007. Only
          three concentrations measured at BOMA are greater than 20 ng/m3 and these were
          measured in 2005, 2007, and 2008.

       •   The difference between the 5th and 95th percentiles for 2007 is nearly the same as
          2010 and 2011, where the range of measurements is significantly smaller. This
          indicates that the majority of measurements in 2007 fell within a relatively small
          range, despite the maximum concentration shown.

       •   The average concentration of lead exhibits an overall decreasing trend over the years
          of sampling, reaching a minimum for 2010 and increasing slightly for 2011.
          However, the variability in the measurements, particularly for 2007 and 2008, make it
          difficult to draw definitive conclusions about this dataset.

       •   The minimum concentration measured for each year is greater than zero, indicating
          that there were no non-detects of lead measured at BOMA since 2005.


       Observations from Figure 14-15 for manganese  measurements at BOMA include the

following:

       •   The maximum manganese concentration shown was measured on July 7, 2010
          (12.3 ng/m3). Only three manganese concentrations measured at BOMA are greater
          than 10 ng/m3, and these were measured in 2005, 2008, and 2010.
                                         14-23

-------
       •  Figure 14-15 shows that the average concentration of manganese decreased from
          2005 to 2006. Between 2006 and 2011, the average and median concentrations of
          manganese have changed relatively little. The average concentration for 2005 is
          4.44 ng/m3, after which it ranges from 3.17 ng/m3 (2009) to 3.67 ng/m3 (2006).
          Similarly, the median concentration for 2005 is 3.96 ng/m3, after which it ranges from
          2.83 ng/m3 (2010) to 3.37 ng/m3 (2011).
       •  The minimum concentration measured for each year is greater than zero, indicating
          that there were no non-detects of manganese measured at BOMA since 2005.

14.5   Additional Risk-Based Screening Evaluations
       The following risk-based screening evaluations were conducted to characterize risk at the
BOMA monitoring site. Refer to Sections 3.3  and 3.5.5 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.
14.5.1  Risk-Based Screening Assessment Using MRLs
       A risk-based screening was conducted by comparing the concentration data from the
Massachusetts monitoring site to the ATSDR MRLs, where available. As described in
Section 3.3, MRLs are noncancer health risk benchmarks and are defined for three exposure
periods: acute (exposures of 1 to 14 days); intermediate (exposures of 15 to 364 days); and
chronic (exposures of 1 year or greater). The preprocessed daily measurements of the pollutants
of interest were compared to the acute MRLs; the quarterly averages were compared to the
intermediate MRLs; and the annual averages were compared to the chronic MRLs.

       As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2011.

14.5.2  Cancer Risk and Noncancer Hazard Approximations
       For the pollutants of interest for BOMA and where annual average concentrations could
be calculated, risk was examined by calculating cancer risk and noncancer hazard
approximations. These approximations can be used as risk estimates for cancer and noncancer
effects attributable to the pollutants of interest. Although the use of these approximations is
limited, they may help identify where policy-makers may want to shift or confirm their air-
                                         14-24

-------
  monitoring priorities. Refer to Section 3.5.5.2 for an explanation of how cancer risk and

  noncancer hazard approximations are calculated and what limitations are associated with them.

  Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard

  approximations are presented in Table 14-6, where applicable. Cancer risk approximations are

  presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless

  values.


            Table 14-6. Risk Approximations for the Massachusetts Monitoring Site
Pollutant
Cancer
URE
(Hg/m3)-1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections vs.
# of Samples
Annual
Average
(ng/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Boston, Massachusetts - BOMA
Arsenic (PM10)
Benzo(a)pyrene
Beiy Ilium (PM10)
Cadmium (PM10)
Hexavalent Chromium
Lead (PM10)
Manganese (PM10)
Naphthalene
Nickel (PM10)
0.0043
0.00176
0.0024
0.0018
0.012


0.000034
0.00048
0.000015

0.00002
0.00001
0.0001
0.00015
0.00005
0.003
0.00009
60/60
61/61
60/60
60/60
54/61
60/60
60/60
61/61
60/60
0.43
ฑ0.07
0.11
ฑ0.02
0.01
ฑ0.01
0.13
ฑ0.01
0.03
ฑ0.01
3.07
ฑ0.44
3.48
ฑ0.44
60.35
ฑ6.84
1.38
ฑ0.17
1.85
0.20
0.02
0.23
0.31


2.05
0.66
0.03

0.01
0.01
0.01
0.02
0.07
0.02
0.02
— = a Cancer URE or Noncancer RfC is not available
         Observations for BOMA from Table 14-6 include the following:

         •   Naphthalene has the highest annual average concentration for BOMA. Manganese,
             lead, and nickel also have annual average concentrations greater than 1.0 ng/m3.

         •   Naphthalene and arsenic are the only pollutants of interest with cancer risk
             approximations greater than 1.0 in-a-million (2.05 in-a-million and 1.85 in-a-million,
             respectively).

         •   None of BOMA's pollutants of interest have noncancer hazard approximations
             greater than 1.0, indicating that no adverse health effects are expected due to these
             individual pollutants.
                                             14-25

-------
14.5.3 Risk-Based Emissions Assessment
       In addition to the risk-based screenings discussed above, Tables 14-7 and 14-8 present a
risk-based evaluation of county-level emissions based on cancer and noncancer toxicity,
respectively. Table 14-7 presents the  10 pollutants with the highest emissions from the 2008
NEI, the 10 pollutants with the highest toxicity-weighted emissions, and the 10 pollutants with
the highest cancer risk approximations (in-a-million), as calculated from the annual averages
provided in Table 14-6. Table 14-8 presents similar information, but identifies the 10 pollutants
with the highest noncancer hazard approximations (HQ), also calculated from annual averages
provided in Table 14-6.

       The pollutants listed in Tables 14-7 and 14-8 are limited to those that have cancer and
noncancer toxicity factors, respectively. As a result, although the actual value of the emissions is
the same, the highest emitted pollutants in the cancer table may be different from the noncancer
table. The cancer risk and noncancer hazard approximations based on the site's annual averages
are limited to those pollutants for which the site sampled. As discussed in Section 14.3, BOMA
sampled for PAHs, PMio metals, and hexavalent chromium. In addition, the cancer risk and
noncancer hazard approximations are limited to those pollutants with enough data to meet the
criteria for an annual average to be calculated. A more  in-depth discussion of this analysis is
provided in Section 3.5.5.3. Similar to the cancer risk and noncancer hazard approximations, this
analysis may help policy-makers prioritize their air monitoring activities.
                                          14-26

-------
  Table 14-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
                                               the Massachusetts Monitoring Site
-^
to
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Boston, Massachusetts (Suffolk County) - BOMA
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
Dichloromethane
POM, Group 2b
Methyl tert butyl ether
Nickel, PM
168.18
164.08
83.84
79.23
26.64
14.54
5.42
3.92
3.63
1.87
Formaldehyde
Benzene
Nickel, PM
1,3 -Butadiene
POM, Group 3
Naphthalene
Hexavalent Chromium, PM
POM, Group 2b
Arsenic, PM
Ethylbenzene
2.13E-03
1.31E-03
8.99E-04
7.99E-04
6.36E-04
4.94E-04
4.31E-04
3.45E-04
2.86E-04
2.10E-04
Naphthalene
Arsenic
Nickel
Hexavalent Chromium
Cadmium
Benzo(a)pyrene
Beryllium
2.05
1.85
0.66
0.31
0.23
0.20
0.02


-------
  Table 14-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer

                                          RfCs for the Massachusetts Monitoring Site
to
oo
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer Hazard
Approximation
(HQ)
Boston, Massachusetts (Suffolk County) - BOMA
Toluene
Xylenes
Benzene
Formaldehyde
Hexane
Ethylbenzene
Acetaldehyde
Ethylene glycol
1,3 -Butadiene
Glycol ethers, gas
456.02
372.68
168.18
164.08
97.07
83.84
79.23
44.58
26.64
18.42
Acrolein
Nickel, PM
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Benzene
Naphthalene
Arsenic, PM
Xylenes
Cadmium, PM
313,612.93
20,816.71
16,743.22
13,322.42
8,803.49
5,606.08
4,847.85
4,440.73
3,726.77
2,204.94
Manganese
Arsenic
Lead
Naphthalene
Nickel
Cadmium
Beryllium
Hexavalent Chromium
0.07
0.03
0.02
0.02
0.02
0.01
<0.01
<0.01


-------
Observations from Table 14-7 include the following:

•  Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
   cancer UREs in Suffolk County.

•  The pollutants with the highest toxi city-weighted emissions (of the pollutants with
   cancer UREs) are formaldehyde, benzene, and nickel.

•  Seven of the highest emitted pollutants also have the highest toxi city-weighted
   emissions.

•  Naphthalene and arsenic are the pollutants with the highest cancer risk
   approximations for BOMA. Naphthalene ranks sixth among the highest emitted
   pollutants and sixth for toxicity-weighted emissions. Arsenic ranks ninth on the list of
   highest toxicity-weighted emissions but is not among the highest emitted.

•  Nickel, which has the third highest cancer risk approximation (albeit low) for BOMA,
   appears on both emissions-based lists. Hexavalent chromium, which has the fourth
   highest cancer risk approximation, ranks seventh on the list of highest toxicity-
   weighted emissions but is not among the highest emitted.

•  POM, Group 2b ranks eighth for both quantity of emissions and toxicity-weighted
   emissions.  POM, Group 2b includes several PAHs sampled for at BOMA including
   acenaphthene, fluoranthene,  fluorene, and perylene.  None of the PAHs included in
   POM, Group 2b were identified as pollutants of interest for BOMA, although both
   acenaphthene and fluorene failed screens (one each). Benzo(a)pyrene is part of POM,
   Group 5a, which is not listed on either emissions-based list.


Observations from Table 14-8 include the following:

•  Toluene, xylenes, and benzene are the highest emitted pollutants with noncancer
   RfCs in Suffolk County.

•  The pollutants with the highest toxicity-weighted emissions (of the pollutants with
   noncancer RfCs) are acrolein, nickel,  and formaldehyde.

•  Five of the highest emitted pollutants  also have the highest toxicity-weighted
   emissions.

•  While nickel, arsenic, and cadmium are among  the pollutants with the highest
   toxicity-weighted emissions, no metals appear among the highest emitted pollutants.
   Nickel, which has the second highest toxicity-weighted emissions (of those pollutants
   with a noncancer RfC), has a negligible noncancer hazard approximation, as do the
   remaining pollutants of interest for BOMA.

•  Manganese, which has  the highest noncancer hazard approximation for BOMA
   (albeit low), appears on neither emissions-based list.
                                   14-29

-------
14.6   Summary of the 2011 Monitoring Data for BOMA

       Results from several of the data treatments described in this section include the
following:

       ปซป  Seven pollutants failed screens for BOMA, of which five are NATTSMQO Core
          Analytes. Naphthalene and arsenic account for a majority of the failed screens.

       ปซป  Naphthalene had the highest annual average concentration among the pollutants of
          interest for BOMA.

       ปซป  Concentrations of the metal pollutants of interest have not changed significantly over
          recent years.
                                          14-30

-------
15.0   Sites in Michigan
       This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS and UATMP sites in Michigan, and integrates these
concentrations with emissions, meteorological, and risk information. Data generated by sources
other than ERG are not included in the data analyses contained in this report. Readers are
encouraged to refer to Sections 1 through 4 and the glossary (Appendix P) for detailed
discussions and definitions regarding the various data analyses presented below.

15.1   Site Characterization
       This section characterizes the monitoring sites by providing geographical and physical
information about the locations of the sites and the surrounding areas. This information is
provided to give the reader insight regarding factors that may influence the air quality near the
sites and assist in the interpretation of the ambient monitoring measurements.

       The DEMI, RRMI, and SWMI monitoring sites are located in the Detroit-Warren-
Livonia, MI MSA. Figures 15-1 through 15-3 are the composite satellite images retrieved from
ArcGIS Explorer showing the monitoring sites in their urban locations. Figure 15-4 identifies
nearby point source emissions locations by source category, as reported in the 2008 NEI for
point sources. Note that only sources within 10 miles of the sites are included in the facility
counts provided in Figure  15-4. A 10-mile  boundary was chosen to give the reader an indication
of which emissions sources and emissions source categories could potentially have a direct effect
on the air quality at the monitoring sites. Further,  this boundary provides both the proximity of
emissions sources to the monitoring  sites as well as the quantity of such sources within a given
distance of the sites. Sources outside the 10-mile radii are still visible on the map, but have been
grayed out in order to show emissions sources just outside the boundary. Table 15-1 provides
supplemental geographical information such as land use, location setting, and locational
coordinates.
                                           15-1

-------
Figure 15-1. Dearborn, Michigan (DEMI) Monitoring Site

-------
Figure 15-2. River Rouge, Michigan (RRMI) Monitoring Site

-------
Figure 15-3. Detroit, Michigan (SWMI) Monitoring Site

-------
Figure  15-4.  NEI  Point Sources Located  Within  10 Miles  of DEMI, RRMI, and  SWMI
                    Legend

                      @    DEMI  NATTS site
                      ฎ    RRMI  UATMP site

Source Category Group (No. of Facilities)             |
  ':•  Air-condiliomrtg/Refrigeratfon (5)
 -ff  Aircraft Operations (9)                              [
  I   Asp
 Gypsum Manufacturing (1}
 Ho&pilal (2j
 Hot MIK Asphalt Plant (2)
 industrial Machinery and Equipment < 1}
 InsDtulKHial - school (7)
 iron and Steel Foundry (i)
 Laboratory (4)
 Landfill {1}
 Lime Manufacturing 0)
 Mine^Quarry (4)
 Mineral Products 44)
 Miscellaneous Commercial/Industrial (12)
 Municipal V&ste Comtw&tor (1)
                                                                                                  | County boundary
                                                                                               W
                                                                                                   Oil and/or Gas Prwluclion (2)
                                                                                                   Pelf oleum Refinery (1)
                                                                                                   Pnnling^PublisrunBCl)
                                                                                                   Rubber and Miscellaneous Plastics Products (3)
                                                                                                   Secondary Metal Processing (1)
                                                                                                   Solid ttfesle Disposal - Commercial'lnstituttonal (4)
                                                                                                   Steel Mill (3)
                                                                                                   Surface Coating (7)
                                                                                                   Telecommunications (t)
                                                                                                   Transportation Equipment (5)
                                                                                                   Tiansportation and Marketing ol Petroleum Products (2)
                                                                                                   V^stewater Treatmenl (1 )
                                                                                                   Woodwork. Furniture. Milhvofk ฃ Wood Preserving <1>
                                                                      15-5

-------
                             Table 15-1. Geographical Information for the Michigan Monitoring Sites
Site
Code
DEMI
RRMI
SWMI
AQS Code
26-163-0033
26-163-0005
26-163-0015
Location
Dearborn
River
Rouge
Detroit
County
Wayne
Wayne
Wayne
Micro- or
Metropolitan
Statistical Area
Detroit-Warren-
Livonia, MI MSA
(Detroit Div)
Detroit-Warren-
Livonia, MI MSA
(Detroit Div)
Detroit-Warren-
Livonia, MI MSA
(Detroit Div)
Latitude
and
Longitude
42.30754,
-83.14961
42.267222,
-83.132222
42.302778,
-83.106667
Land Use
Industrial
Industrial
Commercial
Location
Setting
Suburban
Suburban
Urban/City
Center
Additional Ambient Monitoring Information1
TSP Metals, Meteorological parameters, PM10, PM10
Speciation, PM2 5, and PM2 5 Speciation.
TSP Metals, Meteorological parameters, PM10, PM10
Manganese.
SO2, TSP Metals, Soil Index, VOCs, Meteorological
parameters, PM10, PM10 Manganese, PM25, and
PM2 5 Speciation
BOLD ITALICS = EPA-designated NATTS Site

-------
       DEMI is located at Paul Costea Park in Dearborn, just southwest of Detroit, and is the
Detroit NATTS site. The surrounding area is both suburban and industrial in nature. Figure 15-1
shows that a freight yard is located just west of the site and a residential neighborhood is located
to the east. Industrial sources such as automobile and steel manufacturing facilities are also
located in the vicinity. The monitoring site lies between two heavily traveled roadways, 1-75  and
1-94.

       RRMI is located at John Bilak Park in River Rouge, a southwestern suburb of Detroit,
less than 1 mile from the Detroit River and the U.S./Canadian border. The surrounding area is of
mixed usage, with residential properties surrounded by highly industrial ones (a freight yard is
located to the west of the site while the Port of Detroit is located just to the east and southeast,
just beyond the bottom right-hand side of Figure 15-2). This site is also downwind of a steel
manufacturing facility.

       SWMI is located on the property of Southwestern High  School in the city of Detroit.  The
high school's track can be seen just west of the site marker in Figure 15-3. Interstate-75 runs
northeast-southwest less than 0.3 miles north of SWMI. The surrounding area is considered
commercial and the site lies  approximately 1 mile north of Zug Island, a small, highly
industrialized area where the Rouge River empties into the Detroit River. This site is also less
than 1 mile northwest of the Detroit River and U.S./Canadian border.

       Figure 15-4  shows that DEMI, RRMI, and SWMI are located within a few miles of each
other. Numerous point sources surround these sites. A cluster of sources is located just southwest
of DEMI. Another cluster of sources is located just north of RRMI. The source categories with
the most point sources within 10 miles of the sites include the aircraft operations source
category, which includes airports as well as small runways, heliports, or landing pads; bulk
terminals and bulk plants; electricity generation via combustion; fabricated metals facilities;
institutional facilities (schools); and surface coating facilities. Although difficult to discern in
Figure 15-4, the closest  source to DEMI is involved in food processing; the closest source to
SWMI is involved in electricity generation via combustion; and the closest source to RRMI is
involved in asphalt processing/roofing manufacturing.
                                           15-7

-------
       Table 15-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Michigan monitoring sites. Table 15-2 includes county-level
population and vehicle registration information.  Table 15-2 also includes a county-level vehicle
registration-to-population ratio, which was calculated to represent the number of vehicles per
person within each monitoring site's residing county.  In addition, the population within 10 miles
of each site is presented, based on postal code population data estimates. An estimate of 10-mile
vehicle ownership was determined by applying the county-level vehicle registration-to-
population ratio to the 10-mile population surrounding each monitoring site. Table 15-2 also
contains traffic volume information for each site. Finally, Table 15-2 presents the county-level
daily VMT for Wayne County.
     Table 15-2. Population, Motor Vehicle, and Traffic Information for the Michigan
                                     Monitoring Sites
Site
DEMI
RRMI
SWMI
Estimated
County
Population1
1,802,096
County-level
Vehicle
Registration2
1,334,752
Vehicles per
Person
(Registration:
Population)
0.74
Population
within 10
miles3
1,046,574
773,610
974,585
Estimated
10-mile
Vehicle
Ownership
775,162
572,987
721,842
Annual
Average
Daily
Traffic4
92,800
98,500
93,000
County-
level Daily
VMT5
42,804,737
 Bounty-level population estimates reflect 2011 data from the U.S. Census Bureau (Census Bureau, 2012b)
 2County-level vehicle registration reflects 2011 data from the Michigan Department of State (MDS, 2011)
 310-mile population estimates reflect 2011 data from Xionetic (Xionetic, 2011)
 4AADT reflects 2011 data from the Michigan DOT (MI DOT, 2011)
 5County-level VMT reflects 2011 data for all public roads from the Michigan DOT (MI DOT, 2012)
 BOLD ITALICS = EPA-designated NATTS Site

       Observations from Table 15-2 include the following:
       •  Wayne County's population and vehicle registration both rank eighth among counties
          with NMP sites.

       •  The vehicle-per-person ratio for these sites is the same and is in the bottom third
          among NMP sites.
       •  Among the Michigan monitoring sites, the 10-mile population is highest near DEMI
          and lowest near RRMI.  The 10-mile populations rank between 10th for DEMI and
          21st for RRMI among NMP sites. The 10-mile estimated vehicle ownership rankings
          are similar to the 10-mile population rankings.
                                            15-8

-------
       •  The traffic volumes near the Michigan sites are similar to each other and rank 14th,
          15th, and 16th among NMP sites. Traffic for DEMI is provided for 1-94, between Ford
          Plant Road and Rotunda Drive; traffic data for RRMI is for 1-75 between Outer Drive
          and South Fort Street/M-85; and traffic data for SWMI is for 1-75 between Springwell
          Street and Livernois Avenue.
       •  The Wayne County daily VMT is the fifth highest VMT compared to other counties
          with NMP sites (where VMT data were available).

15.2   Meteorological Characterization
       The following sections characterize the meteorological conditions near the monitoring
sites in Michigan on sample days, as well as over the course of the year.
15.2.1  Climate Summary
       Detroit is located in a region of active weather. Winters tend to be cold and wet, while
summers are generally mild, although temperatures exceeding 90ฐF are not uncommon. Two
major influences on the city's weather include the urbanization of the area and Lake St. Clair to
the east. The lake tends to keep the Detroit area warmer in the winter and cooler in the summer
than more inland areas. The urban heat island keeps the city warmer than outlying areas. Winds
are often breezy and flow from the southwest on average. Precipitation is fairly well distributed
throughout the year, with summer precipitation coming primarily in the form of showers and
thunderstorms. Approximately 30 inches of snow falls on average during winter (Bair, 1992 and
MSU, 2013).

15.2.2  Meteorological Conditions in 2011
       Hourly meteorological data from the NWS weather station nearest these sites were
retrieved for 2011 (NCDC, 2011). The closest weather station to all three Detroit  sites is located
at Detroit City Airport (WBAN 14822). Additional information about this weather station, such
as the distance between the sites and the weather station, is provided in Table 15-3. These data
were used to determine how meteorological conditions on sample days vary from conditions
experienced throughout the year.
                                          15-9

-------
                 Table 15-3. Average Meteorological Conditions near the Michigan Monitoring Sites
Closest NWS Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(ฐF)
Average
Temperature
(ฐF)
Average
Dew Point
Temperature
(ฐF)
Average
Wet Bulb
Temperature
(ฐF)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average 1
Scalar
Wind
Speed
(kt)
Dearborn, Michigan - DEMI
Detroit City Airport
14822
(42.41, -83.01)
9.70
miles
54ฐ
(NE)
Sample
Day
2011
56.8
ฑ5.2
58.0
ฑ2.2
49.7
ฑ4.9
50.8
ฑ2.0
40.2
ฑ4.4
40.9
ฑ 1.9
45.0
ฑ4.3
45.9
ฑ1.8
72.6
ฑ3.1
71.6
ฑ 1.2
1016.0
ฑ1.6
1015.6
ฑ0.7
6.7
ฑ0.7
6.7
ฑ0.3 |
River Rouge, Michigan - RRMI |
Detroit City Airport
14822
(42.41, -83.01)
15.73
miles
32ฐ
(NNE)
Sample
Day
2011
57.6
ฑ5.4
58.0
ฑ2.2
50.4
ฑ5.1
50.8
ฑ2.0
40.8
ฑ4.5
40.9
ฑ1.9
45.6
ฑ4.4
45.9
ฑ1.8
72.4
ฑ3.3
71.6
ฑ1.2
Detroit, Michigan - SWMI
Detroit City Airport
14822
(42.41, -83.01)
11.98
miles
34ฐ
(NE)
Sample
Day
2011
58.3
ฑ7.9
58.0
ฑ2.2
50.7
ฑ7.4
50.8
ฑ2.0
41.6
ฑ6.5
40.9
ฑ1.9
46.0
ฑ6.4
45.9
ฑ1.8
73.7
ฑ4.6
71.6
ฑ1.2
1016.1
ฑ 1.7
1015.6
ฑ0.7
1016.4
ฑ2.4
1015.6
ฑ0.7
6.6
ฑ0.7
6.7
ฑ0.3
5.9
ฑ0.9
6.7
ฑ0.3
Sample day averages are highlighted to help differentiate the sample day averages from the full-year averages.

-------
       Table 15-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for the entire year for 2011. Also included in Table 15-3 is the
95 percent confidence interval for each parameter. Average meteorological conditions on sample
days near the Detroit sites were generally representative of average weather conditions
experienced throughout the year, although Table 15-3 shows that the sample day temperature
averages for SWMI were most like the full-year averages while sample day temperature averages
for DEMI were the least like the full-year averages. This is interesting because SWMI sampled
on a l-in-12 day  schedule and is the reason the sample day confidence intervals are larger than
the sample day confidence intervals for DEMI and RRMI, which sampled on a l-in-6 day
schedule. The bulk of the samples days for RRMI and DEMI are the same, although sampling at
RRMI did not begin until late January 2011.

15.2.3 Back Trajectory Analysis
       Figure 15-5 is the composite back trajectory map for days on which samples were
collected at the DEMI monitoring site in 2011. Included in Figure 15-5 are four back trajectories
per sample day. Figure 15-6 is the corresponding cluster analysis. An in-depth description of
these maps and how they were generated is presented in Section 3.5.2.1. For the composite map,
each line represents the 24-hour trajectory along which a parcel of air traveled toward the
monitoring site on a given sample day and time, based on an initial height of 50 meters AGL. For
the cluster analysis, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric circle around the site in Figures 15-5 and 15-6 represents 100 miles.
Figures 15-7 and 15-8 are the composite back trajectory map and corresponding cluster analysis
for RRMI. Figure 15-9 is the composite back trajectory map for SWMI but the cluster analysis
was not performed  for this site because there were fewer than 30 sample days.
                                         15-11

-------
Figure 15-5. 2011 Composite Back Trajectory Map for DEMI
    Figure 15-6. Back Trajectory Cluster Map for DEMI
                         15-12

-------
Figure 15-7. 2011 Composite Back Trajectory Map for RRMI
    Figure 15-8. Back Trajectory Cluster Map for RRMI
                         15-13

-------
              Figure 15-9. 2011 Composite Back Trajectory Map for SWMI
       Observations from Figures 15-5 through 15-9 for the Michigan sites include the
following:

       •  The composite back trajectory maps for DEMI and RRMI are similar to each other in
          trajectory distribution. This is expected given the close proximity to each other and
          the similarities in sample days. The composite map for SWMI resembles the maps for
          the other two sites but has roughly half the back trajectories because this site sampled
          on a l-in-12 day sampling schedule rather than a l-in-6 day.

       •  Back trajectories originated from a variety of directions at the Detroit monitoring
          sites. Back trajectories originating to the east of the sites tended to be shorter in length
          than trajectories from other directions.

       •  The 24-hour air shed domain for DEMI was similar in size to RRMI. The farthest
          away a back trajectory originated from DEMI was over central Ontario, Canada, or
          greater than 675 miles away. While the farthest away a back trajectory originated
          from RRMI was over South Dakota, or nearly 690 miles away, back trajectories of
          similar length also originated over Ontario. The average trajectory lengths for these
          two sites were just less than 270 miles. Approximately 88 percent of trajectories
          originated within 450 miles of the sites.
                                          15-14

-------
       •  For SWMI, the air shed domain was slightly smaller than those for DEMI and RRMI,
          with an average trajectory length of 243 miles with greater than 90 percent of back
          trajectories originating within 450 miles of the site. The farthest away a back
          trajectory originated was  553 miles over western Quebec, Canada.
       •  The cluster analyses for DEMI and RRMI are similar to each other. The main
          difference is how the HYSPLIT model grouped some of the shorter back trajectories
          with a westerly component. Both cluster maps show that the bulk of the back
          trajectories originated from a direction with a westerly component and that they
          varied in length and therefore geographical origin.  The cluster trajectories originating
          over western Wisconsin, southern Indiana, Lake Erie, and Quebec, Canada are similar
          in direction and origin for both sites although the percentages vary.

15.2.4 Wind Rose Comparison
       Hourly surface wind data from the NWS weather station at the Detroit City Airport were
uploaded into a wind rose software program to produce customized wind roses, as described in
Section 3.5.2.2. A wind rose shows the frequency of wind directions using "petals" positioned
around a 16-point compass, and uses different colors to represent wind speeds.

       Figure 15-10 presents a map showing the distance between the NWS station  and DEMI,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure  15-10 also presents three different wind roses for the
DEMI monitoring site. First, a historical wind rose representing 2001 to 2010 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 2011 is presented.
Next, a wind rose representing wind  data for days on which samples  were collected in 2011 is
presented. These can be used to identify the predominant wind speed and direction for 2011 and
determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.  Figures 15-11 and 15-12 present the distance maps and
wind roses for RRMI and SWMI.
                                          15-15

-------
 Figure 15-10. Wind Roses for the Detroit City Airport Weather Station near DEMI
Distance between DEMI and NWS Station
2001-2010 Historical Wind Rose
                                            IWEST
         2011 Wind Rose
    Sample Day Wind Rose
                                     15-16

-------
 Figure 15-11. Wind Roses for the Detroit City Airport Weather Station near RRMI
Distance between RRMI and NWS Station
2001-2010 Historical Wind Rose
                            MWS
                           Stilton
          2011 Wind Rose
                                                   Sample Day Wind Rose
                                     15-17

-------
 Figure 15-12. Wind Roses for the Detroit City Airport Weather Station near SWMI
Distance between SWMI and NWS Station
2001-2010 Historical Wind Rose
                              NWS
                               lion
                                 V'

                                             IWEST
          2011 Wind Rose
    Sample Day Wind Rose
                                     15-18

-------
       Observations from Figures 15-10 through 15-12 include the following:

       •  The NWS weather station at Detroit City Airport is the closest weather station to all
          three monitoring sites. This weather station is located to the northeast of the sites and
          ranges from 9.70 miles (DEMI) to 15.7 miles (RRMI) away from the sites. Most of
          the city of Detroit lies between the weather station and the monitoring sites.

       •  Because the Detroit City Airport weather station is the closest weather station to all
          three sites,  the historical and 2011 wind roses for DEMI are the same as those for
          RRMI and  SWMI.

       •  The historical wind roses for the Detroit sites show that winds from a variety of
          directions were observed, although winds from the northeast and southeast quadrants
          were observed less frequently than winds from other directions. Calm winds
          (< 2 knots) were observed for approximately 10 percent of the hourly measurements.

       •  The wind patterns on the 2011 wind roses resemble the historical wind patterns,
          although there were slightly fewer westerly to northwesterly winds and more
          northerly and north-northeasterly winds. This resemblance indicates that conditions
          during 2011 were consistent with those experienced historically.

       •  The sample day wind roses for DEMI and RRMI generally resemble the full-year
          wind rose,  although there were fewer winds from the southwest quadrant and more
          northerly to north-northeasterly winds on sample days.

       •  The wind patterns on the sample day wind rose for SWMI differ from those on the
          sample day wind roses for DEMI and RRMI. The calm rate is higher for SWMI and
          winds from the north, west-southwest, and west account for nearly one-third of the
          wind observations by themselves. Recall that the sample day wind rose for SWMI has
          half the wind observations than the sample day wind roses for DEMI and RRMI due
          to the sampling frequency.


15.3   Pollutants of Interest
       Site-specific "pollutants of interest" were determined for the Michigan monitoring  sites

in order to allow 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. In addition, if any of the NATTS MQO Core Analytes measured by each monitoring site

did not meet the pollutant of interest criteria based on the preliminary risk-based screening, that

pollutant was added to the list of site-specific pollutants of interest. A more in-depth description

of the risk-based screening process is presented in Section 3.2.

                                          15-19

-------
       Table 15-4 presents the results of the preliminary risk-based screening process for the
Michigan monitoring sites. The pollutants that failed at least one screen and contributed to
95 percent of the total failed screens for each monitoring site are shaded. NATTS MQO Core
Analytes are bolded. Thus, pollutants of interest are shaded and/or bolded. All three Detroit
monitoring sites sampled for carbonyl compounds; in addition, DEMI sampled for VOCs, PAHs,
and hexavalent chromium.
       Table 15-4. Risk-Based Screening Results for the Michigan Monitoring Sites
Pollutant
Screening
Value
(Ug/m3)
# of Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
Dearborn, Michigan - DEMI
Acet aldehyde
Formaldehyde
Benzene
Carbon Tetrachloride
Naphthalene
1,3-Butadiene
Ethylbenzene
Acenaphthene
Fluorene
1 ,2-Dichloroethane
Fluoranthene
Hexavalent Chromium
Dichloromethane
Benzo(a)pyrene
ฃ>-Dichlorobenzene
Hexachloro- 1 ,3 -butadiene
Acrylonitrile
Chloroprene
Xylenes
0.45
0.077
0.13
0.17
0.029
0.03
0.4
0.011
0.011
0.038
0.011
0.000083
7.7
0.00057
0.091
0.045
0.015
0.0021
10
Total
62
62
61
61
60
57
36
17
16
11
7
6
o
J
2
2
2
1
1
1
468
62
62
61
61
60
57
61
60
60
11
60
60
61
58
16
2
1
1
61
875
100.00
100.00
100.00
100.00
100.00
100.00
59.02
28.33
26.67
100.00
11.67
10.00
4.92
3.45
12.50
100.00
100.00
100.00
1.64
53.49
13.25
13.25
13.03
13.03
12.82
12.18
7.69
3.63
3.42
2.35
1.50
1.28
0.64
0.43
0.43
0.43
0.21
0.21
0.21
13.25
26.50
39.53
52.56
65.38
77.56
85.26
88.89
92.31
94.66
96.15
97.44
98.08
98.50
98.93
99.36
99.57
99.79
100.00

River Rouge, Michigan - RRMI
Acet aldehyde
Formaldehyde
0.45
0.077
Total
55
55
110
57
57
114
96.49
96.49
96.49
50.00
50.00
50.00
100.00

Detroit, Michigan - SWMI
Acet aldehyde
Formaldehyde
Propionaldehyde
0.45
0.077
0.8
Total
28
28
1
57
28
28
28
84
100.00
100.00
3.57
67.86
49.12
49.12
1.75
49.12
98.25
100.00

                                          15-20

-------
       Observations from Table 15-4 for DEMI include the following:

       •  Nineteen pollutants, of which eight are NATTS MQO Core Analytes, failed at least
          one screen for DEMI.

       •  Eleven pollutants contributed to 95 percent of the total failed screens for DEMI; of
          these, six are NATTS MQO Core Analytes. Hexavalent chromium and
          benzo(a)pyrene were added to the pollutants of interest for DEMI because they are
          NATTS MQO  Core Analytes, even though they did not contribute to 95 percent of
          the total failed  screens. Four additional pollutants (chloroform, tetrachloroethylene,
          trichloroethylene, and vinyl chloride) were also added to the list, even though they
          did not fail any screens, because they too are NATTS MQO Core Analytes. These
          four pollutants  are not shown in Table 15-4 but are shown in subsequent tables in the
          sections that follow.

       •  Of the pollutants failing screens, approximately 53  percent of the measured detections
          failed screens.  Ten pollutants failed 100  percent of their screens, although their
          detection rates  varied.

       •  The six pollutants failing the most screens contributed to over 75 percent of the total
          failed screens,  are  all NATTS MQO Core Analytes, and failed 100 percent of their
          screens.

       Observations from Table 15-4 for RRMI and SWMI include the following:

       •  Acetaldehyde and formaldehyde failed screens for RRMI. These pollutants
          contributed equally to the total failed screens.

       •  Acetaldehyde and formaldehyde failed the majority of screens for SWMI, each
          contributing to 49 percent of the total failed screens. Propionaldehyde failed a  single
          screen for SWMI.  These three pollutants are the only  carbonyl compounds with risk
          screening values.


15.4   Concentrations

       This section presents various concentration averages used to characterize pollution levels

at the Michigan monitoring sites. Where applicable, the following calculations and  data analyses

were performed: Time period-based concentration averages (quarterly and annual) are provided

for the pollutants of interest for the Michigan monitoring sites, where the data meet the

applicable criteria. Concentration averages for select pollutants are also presented graphically for

the sites to illustrate how the sites' concentrations compare to the program-level averages, as

presented in Section 4.1. In addition, concentration  averages for select pollutants are presented

from previous years of sampling in order to characterize concentration trends at the sites.

Additional site-specific statistical summaries for DEMI, RRMI, and SWMI are provided in

Appendices J, L, M, and O.

                                           15-21

-------
15.4.1 2011 Concentration Averages
       Quarterly and annual concentration averages were calculated for the pollutants of interest
for the Michigan sites, as described in Section 3.1. The quarterly average of a particular
pollutant is simply the average concentration of the preprocessed daily measurements over a
given calendar quarter. Quarterly average concentrations include the substitution of zeros for all
non-detects. A site must have a minimum of 75 percent valid samples of the total number of
samples possible within a given quarter for a quarterly average to be calculated. An annual
average includes all measured detections and substituted zeros for non-detects for the entire year
of sampling. Annual averages were calculated for pollutants where three valid quarterly averages
could be calculated and where method completeness was greater than or equal to 85 percent, as
presented in Section 2.4. Quarterly and annual average concentrations for the Michigan
monitoring sites are presented in Table 15-5, where applicable. Note that concentrations of the
PAHs and hexavalent chromium are presented in ng/m3 for ease of viewing. Also note that if a
pollutant was not detected in a given calendar quarter, the quarterly average simply reflects "0"
because only zeros substituted for non-detects were factored into the quarterly average
concentration.
         Table 15-5. Quarterly and Annual Average Concentrations of the Pollutants of
                           Interest for the Michigan Monitoring Sites
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(jig/m3)
Dearborn, Michi
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
62/62
61/61
57/61
61/61
61/61
11/61
61/61
62/62
1.67
ฑ0.16
0.84
ฑ0.10
0.10
ฑ0.01
0.59
ฑ0.05
0.52
ฑ0.10
0
0.42
ฑ0.13
2.01
ฑ0.23
2nd
Quarter
Average
(jig/m3)
3rd
Quarter
Average
(jig/m3)
4th
Quarter
Average
(jig/m3)
Annual
Average
(jig/m3)
gan - DEMI
1.80
ฑ0.29
0.88
ฑ0.26
0.08
ฑ0.03
0.63
ฑ0.05
1.58
ฑ0.24
0.03
ฑ0.03
0.45
ฑ0.13
2.04
ฑ0.76
1.77
ฑ0.30
0.90
ฑ0.18
0.09
ฑ0.03
0.70
ฑ0.04
0.68
ฑ0.19
0
0.91
ฑ0.40
4.14
ฑ0.90
1.63
ฑ0.37
0.84
ฑ0.17
0.10
ฑ0.03
0.69
ฑ0.04
0.52
ฑ0.16
0.03
ฑ0.02
0.64
ฑ0.24
2.17
ฑ0.49
1.72
ฑ0.14
0.86
ฑ0.09
0.09
ฑ0.01
0.65
ฑ0.02
0.82
ฑ0.14
0.01
ฑ0.01
0.61
ฑ0.13
2.60
ฑ0.39
      a Average concentrations provided for the pollutants below the blue line are presented in ng/m for
      ease of viewing.
                                          15-22

-------
   Table 15-5. Quarterly and Annual Average Concentrations of the Pollutants of
              Interest for the Michigan Monitoring Sites (Continued)
Pollutant
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
Acenaphthene3
Benzo(a)pyrenea
Fluoranthene3
Fluorene3
Hexavalent Chromium3
Naphthalene3
#of
Measured
Detections
vs. # of
Samples
56/61
5/61
8/61
60/60
58/60
60/60
60/60
60/61
60/60
1st
Quarter
Average
(Ug/m3)
0.18
ฑ0.05
0.01
ฑ0.01
0.01
ฑ0.01
3.67
ฑ1.69
0.15
ฑ0.07
2.05
ฑ0.50
3.66
ฑ0.91
0.04
ฑ0.02
98.14
ฑ 13.86
2nd
Quarter
Average
(Ug/m3)
0.15
ฑ0.06
0
0
15.04
ฑ11.35
0.26
ฑ0.26
6.52
ฑ5.02
12.46
ฑ9.32
0.05
ฑ0.02
136.83
ฑ43.19
3rd
Quarter
Average
(Ug/m3)
0.20
ฑ0.06
0.02
ฑ0.02
0.01
ฑ0.01
27.99
ฑ 14.22
0.13
ฑ0.06
10.26
ฑ3.87
22.15
ฑ9.80
0.05
ฑ0.02
217.74
ฑ 59.44
4th
Quarter
Average
(Ug/m3)
0.18
ฑ0.06
0
0.01
ฑ0.01
5.49
ฑ2.48
0.21
ฑ0.15
2.44
ฑ0.99
5.36
ฑ2.17
0.05
ฑ0.01
113.75
ฑ 32.02
Annual
Average
(Ug/m3)
0.18
ฑ0.03
0.01
ฑ0.01
0.01
ฑ0.01
13.42
ฑ5.17
0.19
ฑ0.07
5.45
ฑ1.78
11.19
ฑ3.84
0.05
ฑ0.01
143.35
ฑ23.07
River Rouge, Michigan - RRMI
Acetaldehyde
Formaldehyde
57/57
57/57
1.52
ฑ0.45
1.90
ฑ0.60
1.87
ฑ0.32
3.09
ฑ1.04
1.87
ฑ0.29
4.86
ฑ0.88
1.43
ฑ0.30
2.79
ฑ0.50
1.68
ฑ0.17
3.23
ฑ0.47
Detroit, Michigan - SWMI
Acetaldehyde
Formaldehyde
28/28
28/28
1.32
ฑ0.23
2.50
ฑ0.31
1.43
ฑ0.60
2.93
ฑ2.05
2.61
ฑ1.55
4.11
ฑ1.11
1.56
ฑ0.42
2.11
ฑ0.43
1.74
ฑ0.43
2.90
ฑ0.60
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for
ease of viewing.
 Observations for DEMI from Table 15-5 include the following:

 •  The pollutants with the highest annual average concentrations by mass are
    formaldehyde and acetaldehyde; all other annual average concentrations are less than
    1.0|ig/m3.

 •  The third quarter average concentration of formaldehyde is twice the other quarterly
    averages. A review of the data shows that the maximum concentration of
    formaldehyde was measured on July 20, 2011 (6.99 |ig/m3) although similar
    concentrations were also measured in June and August. Of the nine concentrations
    greater than 4 |ig/m3, eight were measured during the third quarter of 2011.
                                    15-23

-------
•  The second quarter average concentration of chloroform is three times higher than the
   other quarterly averages. A review of the data shows that the maximum concentration
   of chloroform was measured on June 20, 2011  (2.25 |ig/m3) although similar
   concentrations were also measured in May and June. Of the 16 chloroform
   concentrations greater than 1 |ig/m3, 13 were measured during the second quarter of
   2011.

•  The third quarter average concentration of ethylbenzene is higher than the other
   quarter averages and has a relatively high confidence interval associated with it. A
   review of the data shows that the highest concentration of this pollutant was measured
   on August 7, 2011 (3.34 |ig/m3) and is nearly twice the next highest concentration
   (1.91  |ig/m3 measured on August 13, 2011). This concentration is also the sixth
   highest ethylbenzene concentration measured among NMP sites sampling this
   pollutant. Of the nine ethylbenzene concentrations greater than 1 |ig/m3 measured at
   DEMI, more than half were measured during the third  quarter of 2011.

•  Similar observations for chloroform and ethylbenzene  were made in the 2010 NMP
   report.

•  The second and third quarter averages of acenaphthene, fluoranthene, and fluorene
   are significantly higher than the other quarterly averages and have relatively high
   confidence intervals associated with them. The highest concentrations of these
   pollutants were all measured on the same days, June 8, 2011 and July 2, 2011. The
   highest concentrations of these compounds were measured in June, July, and August,
   generally on the same days, although the order varied.

•  A similar trend is  shown for naphthalene. The maximum concentration of
   naphthalene was also measured on July 2, 2011. This was the fourth highest
   concentration of naphthalene measured across the program. Of the  12 highest
   concentrations of naphthalene measured at DEMI (those greater than 200 ng/m3), nine
   were measured during the third quarter of 2011 (plus two in June and one in October).

•  The second quarter average concentration of benzo(a)pyrene is the same as its
   associated confidence interval, an indicator of considerable variability in the
   concentrations measured. The maximum benzo(a)pyrene concentration (1.99 ng/m3)
   was measured on May 3, 2011 and is also the maximum concentration measured
   among NMP sites sampling PAHs. The next highest concentration measured during
   the second quarter of 2011 at DEMI was considerably  less (0.236 ng/m3) and the
   median concentration for the second quarter is  0.149 ng/m3. The fourth quarter
   average concentration and its associated confidence interval also reflect a relatively
   high level of variability in the measurements. The only other measurement greater
   than 1 ng/m3 at DEMI was measured on October 6, 2011.  The next highest
   concentration measured during the fourth quarter was 0.269 ng/m3 and the median
   concentration for the fourth quarter is 0.142 ng/m3.
                                   15-24

-------
       Observations for RRMI and SWMI from Table 15-5 include the following:

       •   The annual average concentration of acetaldehyde for RRMI is similar to the annual
          average concentration for SWMI. Both are similar to the annual average acetaldehyde
          concentration for DEMI. The annual formaldehyde averages for the sites vary a little
          more.

       •   The third quarter average concentration of acetaldehyde for SWMI is greater than the
          other quarterly averages and has a relatively large confidence interval associated with
          it. A review of the data shows that the maximum concentration (6.70 |ig/m3) was
          measured on August 7, 2011 and is more than twice the next highest concentration
          (2.85 |ig/m3, measured on August 19, 2011).

       •   The second and third quarter average concentrations of formaldehyde for RRMI and
          SWMI are higher than the other quarterly averages for 2011 and have relatively large
          confidence intervals associated with them. The maximum formaldehyde
          concentration was measured at RRMI on July 20, 2011 (8.25 |ig/m3), which is the
          same day the maximum concentration of formaldehyde was measured at DEMI.
          Because SWMI sampled on a l-in-12 day sampling schedule, no sample was
          collected at this site on July 20, 2011. However, the maximum formaldehyde
          concentration for SWMI (8.36 |ig/m3) was measured on the same day that the second
          highest concentration was measured at RRMI, June 8, 2011.  The highest
          concentrations of formaldehyde at both sites were measured between June and
          September.


       Tables 4-9 through 4-12  present the sites with the 10 highest annual average

concentrations for each of the program-level pollutants of interest. Observations for the

Michigan sites from those tables include the following:

       •   DEMI appears in Tables 4-9 through 4-12 a total of 10 times.

       •   DEMI has the second highest annual average concentration of chloroform, behind
          only NBIL, as shown in Table 4-9. However, the difference between the annual
          chloroform average for DEMI and NBIL is  significant.

       •   The annual average concentration of acenaphthene for DEMI is the highest among
          NMP sites sampling PAHs. DEMI's annual average concentrations of fluorene and
          naphthalene each rank second and the annual average concentration of
          benzo(a)pyrene ranks third among sites sampling PAHs.

       •   The annual average concentration of hexavalent chromium for DEMI ranks fourth
          highest among sites sampling this pollutant.

       •   The rankings for DEMI are similar to those for the 2010 NMP report.
                                         15-25

-------
       •  None of the Michigan sites appear among the sites with the highest annual average
          concentrations of acetaldehyde. RRMI ranks tenth for its annual average
          concentration of formaldehyde.

15.4.2 Concentration Comparison
       In order to better illustrate how a site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for the selected NATTS MQO Core
Analytes listed in Section 3.5.3, where applicable. Thus, box plots for acetaldehyde and
formaldehyde were created for all three Michigan sites. In addition, box plots for benzene,
benzo(a)pyrene, 1,3-butadiene, hexavalent chromium, and naphthalene were created for DEMI.
Figures 15-13 through 15-19 overlay the sites' minimum, annual average, and maximum
concentrations onto the program-level minimum, first quartile, median, average, third quartile,
and maximum concentrations, as described in Section 3.5.3.
       Figure 15-13. Program vs. Site-Specific Average Acetaldehyde Concentrations
 DEI'.'I
     •H-
 S'.'.'f.'l
                                                        ID
                                         Concentration
                Program:   IstQuartile     2ndQuartile     SrdQuartile     4thQuartile     Average
               Site:     Site Average    Site Minimum/Maximum

                                          15-26

-------
         Figure 15-14. Program vs. Site-Specific Average Benzene Concentration
DEf.'l


) 1 2 3
Program: IstQuartile
Site: Site Average
o

Program Max Concentration = 23. 8 ug/m3

4567891
Concentration (pg/mi)
2ndQuartile SrdQuartile 4thQuartile Average
• n n
Site Minimum/Maximum

     Figure 15-15. Program vs. Site-Specific Average Benzo(a)pyrene Concentration
           -o-
                                            Concentration (ng/m3)
                Program:   IstQuartile     2ndQuartile      SrdQuartile      4thQuartile     Average
                Site:      Site Average     Site Minimum/Maximum

      Figure 15-16. Program vs. Site-Specific Average 1,3-Butadiene Concentration
                                                                I  Program Max Concentration = 9.51 ng/
                  3.5
                                                 15
                                            Concentration

                                                                                2.5
                Program:   IstQuartile     2ndQuartile      SrdQuartile      4thQuartile     Average
                Site:      Site Average     Site Minimum/Maximum
                                              15-27

-------
      Figure 15-17. Program vs. Site-Specific Average Formaldehyde Concentrations
S'.'.'f.'l
                                   13
                                                   IS
                                              Concentration
                Program:    IstQuartile     2ndQuartile      SrdQuartile     4thQuartile      Average
                Site:
                         Site Average     Site Minimum/Maximum

  Figure 15-18. Program vs. Site-Specific Average Hexavalent Chromium Concentration
DEf.'l
1
                   3.35
                                   0.1
                                                  3.15
                                             Concentration (
                                                                  3.2
                                                                                 3.25
                                                                                                 3.3
                Program:    IstQuartile     2ndQuartile      SrdQuartile     4thQuartile      Average
                Site:      Site Average     Site Minimum/Maximum

       Figure 15-19. Program vs. Site-Specific Average Naphthalene Concentration
DEf.'l
                                                                  i Program Max Concentration = 779 ng/ms
                      100
                               153
                                         200        250       300
                                             Concentration (ng/m3)
                                                                     35C
                                                                              433
                                                                                       453
                                                                                                533
Program
Site:
: IstQuartile
Site Average
o
2ndQuartile SrdQuartile 4thQuartile Ave
• n
Site Minimum/Maximum
'rage

                                               15-28

-------
Observations from Figures 15-13 through 15-19 include the following:

   •   Figure 15-13 includes the box plots for acetaldehyde for all three sites. The box
       plots show that not only are all three sites' annual averages less than the program-
       level average concentration of acetaldehyde, they are all relatively similar to each
       other (less than 0.1 |ig/m3 separates them). However, the range of concentrations
       measured at SWMI is twice as wide as the range measured at DEMI and RRMI.
       Although no non-detects of acetaldehyde were measured at the Michigan sites or
       across the program, the two minimum concentrations of acetaldehyde were
       measured at RRMI.

   •   Figure 15-14 is the box plot for benzene. Note that the program-level maximum
       concentration (23.8 |ig/m3) is not shown directly on the box plot because the scale
       of the box plot would be too large to readily observe data points at the lower end
       of the concentration range. Thus, the scale has been reduced to 10 |ig/m3.
       Figure 15-14 shows that DEMI's annual average benzene concentration is just
       less than the program-level average concentration. The maximum concentration
       of benzene measured at DEMI is considerably less than the maximum
       concentration measured at the program level.  There were no non-detects of
       benzene measured at DEMI or across the program.

   •   Figure 15-15 is the box plot for benzo(a)pyrene for DEMI. Note that the program-
       level first quartile for this pollutant is zero and is not visible on this box plot. The
       box plot shows that the maximum concentration of benzo(a)pyrene was measured
       at DEMI. The annual average concentration for DEMI is greater than both the
       program-level average and third quartile. Recall that the annual average
       benzo(a)pyrene concentration for DEMI is the third highest annual  average
       concentration of this pollutant among sites sampling PAHs. Two non-detects of
       benzo(a)pyrene were measured at DEMI.

   •   Figure 15-16 is the box plot for 1,3-butadiene. Similar to the benzene box plot,
       the program-level maximum concentration (9.51 |ig/m3) is not shown directly on
       the box plot as the scale has been reduced to 3 |ig/m3 to allow for the observation
       of data points at the lower end of the concentration range. Figure 15-16 shows
       that the annual average concentration for DEMI is similar to the program-level
       average concentration. The maximum 1,3-butadiene concentration measured at
       DEMI is considerably less than the maximum concentration measured across the
       program. Four non-detects of 1,3-butadiene were measured at DEMI.

   •   Figure 15-17 includes the box plots for formaldehyde for all three sites. The box
       plots show that the annual averages for these three sites vary by less than
       0.70 |ig/m3. The annual average for DEMI is just less than the program-level
       average, the annual average for SWMI is similar to the program-level average,
       and the annual average for RRMI is just greater than the program-level average.
       All three annual averages are between the program-level median and third quartile
       (or between 50th and 75th percentile).The range of concentrations measured is
       similar for RRMI and SWMI and slightly smaller for DEMI. The maximum
       concentration measured at each site is less than the maximum concentration
       measured at the program level. Although no non-detects of formaldehyde were

                                  15-29

-------
             measured at the Michigan sites or across the program, the two minimum
             concentrations of formaldehyde were measured at RRMI.

          •  Figure 15-18 is the box plot for hexavalent chromium. The box plot shows that
             annual average concentration for DEMI is nearly twice the program-level average
             concentration.  The maximum concentration measured at DEMI is not the
             maximum concentration measured across the program, but it was the second
             highest. There was a single non-detect of this pollutant measured at DEMI.

          •  Figure 15-19 is the box plot for naphthalene.  Similar to the benzene and
             1,3-butadiene box plots, the program-level maximum concentration (779 ng/m3)
             is not shown directly on the box plot as the scale has been reduced to 500 ng/m3 to
             allow for the observation of data points at the lower end of the concentration
             range. Figure 15-19 shows that the annual average concentration of naphthalene
             for DEMI is nearly twice the program-level average concentration. Recall from
             the previous section that DEMI's annual average concentration is the second
             highest annual average among NMP sites sampling this pollutant. The maximum
             naphthalene concentration measured at DEMI is less than the maximum
             concentration measured across program, although it was the fourth highest
             naphthalene concentration measured across the program. The minimum
             naphthalene concentration measured at DEMI is greater than the program-level
             first quartile (25th percentile).


15.4.3  Concentration Trends

       A site-specific trends evaluation was completed for sites that have sampled one or more

of the selected NATTS MQO Core Analytes for 5 consecutive years or longer, as described in
Section 3.5.4. DEMI has sampled VOCs  and  carbonyl compounds under the NMP since 2003

and hexavalent chromium since 2005. Thus, Figures 15-20 through  15-24 present the annual

statistical metrics for acetaldehyde, benzene,  1,3-butadiene, formaldehyde, and hexavalent

chromium for DEMI. The statistical metrics presented for assessing trends include the

substitution of zeros for non-detects.
       Sampling for PAHs at DEMI did not begin until 2008, therefore a trends analysis was not

conducted for the pollutants for these methods.  Although RRMI and SWMI have sampled under

the NMP previously, they have not sampled continuously for 5 consecutive years; thus, a trends

analysis was not performed for these sites.
                                         15-30

-------
Figure 15-20. Annual Statistical Metrics for Acetaldehyde Concentrations
                           Measured at DEMI




ST
.a ^
c
1
3 4
|
< 3 _













-L
*•







••<
^
— i


r

>•-
•
'




l-l-,
1 — '*~~l 1 — *~~l

"*~ """- -fc 	 i* 	 "*
Y Y u, -r T
2003 2004 2005 2006 2007 2008 2009 2010 2011
Year
• 5th Pe re entile — Minimum — Median — Maximum * 95th Percentile ..^.. Average
   Figure 15-21. Annual Statistical Metrics for Benzene Concentrations
                           Measured at DEMI
               2005       2006
                                 2007       2008
                                     Year
       * 5th Percentile    - Minimum     — Median    - Maximum    • 95th Percentile   ••*•• Average
                                  15-31

-------
Figure 15-22. Annual Statistical Metrics for 1,3-Butadiene Concentrations
                         Measured at DEMI



tration (\ifjn\3)
S i
A i/e rage Concert
j e

0 -














+ -
^^m
2004




[

— • —
2005
* 5th Percentile








••
T

^.-, T E
r^ r^n
"* 	 *. 	 .
•

r f ? ' t
2006 2007 2003 2009 2010 2011
Year
— M nimum — Median — Maximum • 95th Percentile . .^.. Average
Figure 15-23. Annual Statistical Metrics for Formaldehyde Concentrations
                         Measured at DEMI



1
Concentration
I
5 -
















F

••

2003
*




I
1

-L.
T
2004
5th Percentile




•



••ป..


JL T f r^i
~^~
•--...I rn LJ

V ^^ ^^ U.J
2005 2006 2007 2008 2009 2010 2011
Year
— Minimum — Median — Maximum • 95th Percentile "^"Average
                               15-32

-------
    Figure 15-24. Annual Statistical Metrics for Hexavalent Chromium Concentrations
                                  Measured at DEMI
       c
       a
       E
       = '
       I
       3

       E
                                            2008
                                            Year
                 5th Percentile    — Minimum     —  Median    — Maximum
                                                            95th Percentile
                                                                        . Average
       Observations from Figure 15-20 for acetaldehyde measurements at DEMI include the
following:

       •   Carbonyl compounds have been sampled continuously at DEMI under the NMP since
          2003. The site began sampling on a l-in-12 day schedule in 2003 then changed to a
          l-in-6 day schedule in the spring of 2004.

       •   Carbonyl compound samples from the primary sampler were invalidated from
          March 13, 2007 through March 25, 2008 by the state of Michigan due to a leak in the
          sample line. With only 12 valid samples in 2007 and less than 85 percent in 2008, no
          statistical metrics are presented for 2007 or 2008.

       •   The maximum acetaldehyde concentration shown was measured in 2004
          (7.84 |ig/m3). Of the six concentrations greater than 5 |ig/m3 measured at DEMI, three
          were measured in 2004, two were measured in 2005, and one was measured in 2006.

       •   The average  concentration exhibits a decreasing trend after 2004 and an increasing
          trend after 2009.

       •   Even with the maximum concentration at a minimum for 2011, the median, average,
          and 95th percentile all exhibit slight increases from 2010 levels.

       •   There have been no non-detects of acetaldehyde measured at DEMI.

                                         15-33

-------
       Observations from Figure 15-21 for benzene measurements at DEMI include the
following:

       •   VOCs have been sampled continuously at DEMI under the NMP since 2003.
          However, the l-in-12 day schedule in 2003 combined with a number of invalids
          resulted in a low completeness; as a result, no statistical metrics are presented for
          2003 and Figure  15-21 begins with 2004.
          The three highest benzene concentrations were all measured in 2004 and ranged from
          5.44 |ig/m3to 7.62 |ig/m3. Only one other conc<
          measured at DEMI and was measured in 2007.
5.44 |ig/m3 to 7.62 |ig/m3. Only one other concentration greater than 5 |ig/m3 has been
       •  Both the median and average concentrations exhibit a steady decreasing trend over
          the period shown, reaching a minimum in 2009. Although a slight increase is shown
          for 2010 followed by a slight decrease in 2011, the changes in the later years are not
          statistically significant.

       •  The difference between the median and average concentrations has also decreased,
          indicating less variability in the measurements. For the last three years of sampling,
          less than 0.1 |ig/m3 separates these two statistical parameters.

       •  The minimum concentration is greater than zero for all years shown, indicating that
          this pollutant has been detected in every VOC sample collected at DEMI.


       Observations from Figure 15-22 for 1,3-butadiene measurements at DEMI include the

following:

       •  Recall that even though VOC sampling at DEMI began in 2003, Figure 15-22 begins
          with data from 2004 due to low completeness in 2003.

       •  The maximum 1,3-butadiene concentration was measured on October 18, 2004. This
          is the  only 1,3-butadiene measurement greater than 1 |ig/m3, although concentrations
          greater than 0.90 |ig/m3 were measured in 2004 and 2006.

       •  For 2004, the minimum, 5th percentile, and median concentrations are all zero,
          indicating that at least 50 percent of the measurements were non-detects. Yet at the
          same time, 2004 also has the two highest concentrations measured at the site and the
          maximum 95th percentile. This indicates there is a high level of variability in the
          measurements. Although there were fewer non-detects in 2006, as indicated by the
          increase in the median concentration, this year also reflects a high level of variability
          compared to other years.

       •  After  2006, the average 1,3-butadiene concentration decreased, reaching a minimum
          in 2009. Nearly all of the statistical parameters are at a minimum for 2009.
                                         15-34

-------
       •   The number of non-detects measured for 1,3-butadiene decreased significantly after
          the first few years of sampling. The number of non-detects decreased from 29 in
          2004, to 19 in 2005, to five in 2006.  Only one to two non-detects were measured in
          the following years until 2011, when four non-detects were measured.


       Observations from Figure 15-23 for formaldehyde measurements at DEMI include the
following:

       •   Recall that carbonyl compounds have been sampled continuously at DEMI under the
          NMP since 2003 but due to a leak in the sample line, samples collected between
          March  13, 2007 through March 25, 2008 were invalidated and thus, no statistical
          metrics are presented for 2007 or 2008.

       •   The maximum formaldehyde concentration shown was measured in 2005
          (33.1 |ig/m3). The next four highest concentrations measured at DEMI were also
          measured in 2005 and range from 13.3 |ig/m3 to 20.9 |ig/m3. The only other
          formaldehyde concentrations greater than 10 |ig/m3 were measured in 2004.
       •   The decrease in the average concentration shown between 2005 and 2006 is
          significant (from 5.35 |ig/m3to 2.92 |ig/m3). The average concentration for tt
          following did not vary significantly from the average concentration for 2006.
       •   Even though the difference between the 5th and 95th percentiles (the range into which
          the majority of the measurements fall) increased from 2010 to 2011, slight decreases
          are shown for the median and average concentrations.

       •   There have been no non-detects of formaldehyde measured at DEMI.


       Observations from Figure 15-24 for hexavalent chromium measurements at DEMI

include the following:

       •   The maximum hexavalent chromium  concentration was measured in 2006. The two
          highest hexavalent chromium concentrations for this site were measured on
          July 4, 2006 (0.496 ng/m3) and on July 5, 2008 (0.392 ng/m3). A similar
          concentration was also measured on January 1, 2009 (0.372 ng/m3).

       •   Although a decrease in the average concentration is shown from 2006 (0.068 ng/m3)
          to 2007 (0.042 ng/m3), the confidence intervals calculated are relatively large as a
          result of the highest concentrations and indicate that these changes are not statistically
          significant. However, the average concentration changed little after 2006, ranging
          from 0.036 ng/m3 in 2009 to 0.047 ng/m3 in 2011.

       •   The minimum concentrations and 5th percentiles for several years are zero, indicating
          the presence of non-detects.
                                         15-35

-------
15.5   Additional Risk-Based Screening Evaluations
       The following risk-based screening evaluations were conducted to characterize risk at the
Michigan monitoring sites. Refer to Sections 3.3 and 3.5.5 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.

15.5.1  Risk-Based Screening Assessment Using MRLs
       A risk-based screening was conducted by comparing the concentration data from the
Michigan monitoring sites to the ATSDR MRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 to 14 days); intermediate (exposures of 15 to 364 days); and chronic (exposures
of 1 year or greater). The preprocessed daily measurements of the pollutants of interest were
compared to the acute MRLs; the quarterly averages were compared to  the intermediate MRLs;
and the annual averages were compared to the chronic MRLs.

       As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2011.

15.5.2  Cancer Risk and  Noncancer Hazard Approximations
       For the pollutants of interest for the Michigan sites and where annual average
concentrations could be calculated, risk was examined by calculating cancer risk and noncancer
hazard approximations. These approximations can be used as risk estimates for cancer and
noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers  may want to shift or
confirm their air-monitoring priorities.  Refer to Section 3.5.5.2 for an explanation of how cancer
risk and noncancer hazard approximations are calculated and what limitations are associated with
them. Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer
hazard approximations are presented in Table 15-6, where applicable. Cancer risk
approximations are presented as probabilities while the noncancer hazard approximations are
ratios and thus, unitless values.
                                         15-36

-------
              Table 15-6. Risk Approximations for the Michigan Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Dearborn, Michigan - DEMI
Acenaphthene3
Acetaldehyde
Benzene
Benzo(a)pyrenea
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
1 ,2-Dichloroethane
Ethylbenzene
Fluoranthene3
Fluorene3
Formaldehyde
Hexavalent Chromium3
Naphthalene3
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
0.000088
0.0000022
0.0000078
0.00176
0.00003
0.000006

0.000026
0.0000025
0.000088
0.000088
0.000013
0.012
0.000034
0.00000026
0.0000048
0.0000088

0.009
0.03

0.002
0.1
0.098
2.4
1


0.0098
0.0001
0.003
0.04
0.002
0.1
60/60
62/62
61/61
58/60
57/61
61/61
61/61
11/61
61/61
60/60
60/60
62/62
60/61
60/60
56/61
5/61
8/61
0.01
ฑ0.01
1.72
ฑ0.14
0.86
ฑ0.09
<0.01
ฑ<0.01
0.09
ฑ0.01
0.65
ฑ0.02
0.82
ฑ0.14
0.01
ฑ0.01
0.61
ฑ0.13
0.01
ฑ<0.01
0.01
ฑ0.01
2.60
ฑ0.39
0.01
ฑ0.01
0.14
ฑ0.02
0.18
ฑ0.03
0.01
ฑ0.01
0.01
ฑ0.01
1.18
3.78
6.73
0.33
2.79
3.92

0.39
1.52
0.48
0.98
33.85
0.56
4.87
0.05
0.03
0.02

0.19
0.03

0.05
0.01
0.01
O.01
0.01


0.27
0.01
0.05
0.01
O.01
0.01
River Rouge, Michigan - RRMI
Acetaldehyde
Formaldehyde
0.0000022
0.000013
0.009
0.0098
57/57
57/57
1.68
ฑ0.17
3.23
ฑ0.47
3.69
41.97
0.19
0.33
Detroit, Michigan - SWMI
Acetaldehyde
Formaldehyde
0.0000022
0.000013
0.009
0.0098
28/28
28/28
1.74
ฑ0.43
2.90
ฑ0.60
3.82
37.70
0.19
0.30
 — = a Cancer URE or Noncancer RfC is not available.
3 For the annual average concentration of this pollutant in ng/m3, refer to Table 15-5.
                                              15-37

-------
        Observations from Table 15-6 include the following:
        •  Formaldehyde has the highest annual average concentration for each of the Michigan
           monitoring sites.  This pollutant also has the highest cancer risk approximation for
           each site, ranging from 33.85 in-a-million for DEMI to 41.97 in-a-million for RRMI.
        •  The range of cancer risk approximations for acetaldehyde was even tighter, ranging
           from 3.69 in-a-million for RRMI to 3.82 in-a-million for SWMI.
        •  Aside from formaldehyde, the pollutants with the highest cancer risk approximations
           for DEMI were benzene, naphthalene, and carbon tetrachloride.
        •  None of the pollutants of interest for the Michigan monitoring sites have noncancer
           hazard approximations greater than 1.0, indicating that no adverse health effects are
           expected from these individual pollutants. The pollutant with the highest noncancer
           hazard approximation for each site is formaldehyde.

 15.5.3  Risk-Based Emissions Assessment
        In addition to the risk-based screenings discussed above, Tables 15-7 and 15-8 present an
 evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
 Table 15-7 presents the 10 pollutants with the highest emissions from the 2008 NEI, the
 10 pollutants with the highest toxicity-weighted emissions, and the 10 pollutants with the highest
 cancer risk approximations (in-a-million), as calculated from the annual averages provided in
 Table 15-6. Table 15-8 presents similar information, but identifies the 10 pollutants with the
 highest noncancer hazard approximations (HQ),  also calculated from annual averages provided
 in Table 15-6.
        The pollutants listed in Table 15-7 and 15-8 are limited to those that have cancer and
noncancer toxicity factors, respectively. As a result, although the actual value of the emissions is
the same, the highest emitted pollutants in the cancer table may be different from  the noncancer
table. Further, the cancer risk and noncancer hazard approximations based on each site's annual
averages are limited to those pollutants for which each respective site sampled. As discussed in
Section 15.3, all three Michigan sites sampled carbonyl compounds; DEMI also sampled for
VOCs, PAHs, and hexavalent chromium. In addition, the cancer risk and noncancer hazard
approximations are limited to those pollutants with enough data to meet the  criteria for annual
averages to be calculated. A more in-depth discussion of this analysis is provided in
Section 3.5.5.3. Similar to the cancer risk and noncancer hazard approximations, this analysis
may help policy-makers prioritize their air monitoring activities.
                                           15-38

-------
  Table 15-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
                                                the Michigan Monitoring Sites
VO
Top 10 Total Emissions for Pollutants
with Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Dearborn, Michigan (Wayne County) - DEMI
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
Dichloromethane
Tetrachloroethylene
Coke Oven Emissions, PM
Trichloroethylene
1,134.11
612.43
500.03
334.62
134.38
74.43
52.69
47.69
35.72
13.67
Coke Oven Emissions, PM
Benzene
Formaldehyde
POM, Group 5a
Hexavalent Chromium, PM
1,3 -Butadiene
Arsenic, PM
Naphthalene
Nickel, PM
Ethylbenzene
3.54E-02
8.85E-03
7.96E-03
7.57E-03
6.69E-03
4.03E-03
3.81E-03
2.53E-03
1.42E-03
1.25E-03
Formaldehyde
Benzene
Naphthalene
Carbon Tetrachloride
Acetaldehyde
1,3 -Butadiene
Ethylbenzene
Acenaphthene
Fluorene
Hexavalent Chromium
33.85
6.73
4.87
3.92
3.78
2.79
1.52
1.18
0.98
0.56
River Rouge, Michigan (Wayne County) - RRMI
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
Dichloromethane
Tetrachloroethylene
Coke Oven Emissions, PM
Trichloroethylene
1,134.11
612.43
500.03
334.62
134.38
74.43
52.69
47.69
35.72
13.67
Coke Oven Emissions, PM
Benzene
Formaldehyde
POM, Group 5a
Hexavalent Chromium, PM
1,3 -Butadiene
Arsenic, PM
Naphthalene
Nickel, PM
Ethylbenzene
3.54E-02
8.85E-03
7.96E-03
7.57E-03
6.69E-03
4.03E-03
3.81E-03
2.53E-03
1.42E-03
1.25E-03
Formaldehyde
Acetaldehyde
41.97
3.69


-------
Table 15-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
                                        the Michigan Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants
with Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
Detroit, Michigan (Wayne County) - SWMI
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
Dichloromethane
Tetrachloroethylene
Coke Oven Emissions, PM
Trichloroethylene
1,134.11
612.43
500.03
334.62
134.38
74.43
52.69
47.69
35.72
13.67
Coke Oven Emissions, PM
Benzene
Formaldehyde
POM, Group 5a
Hexavalent Chromium, PM
1,3 -Butadiene
Arsenic, PM
Naphthalene
Nickel, PM
Ethylbenzene
3.54E-02
8.85E-03
7.96E-03
7.57E-03
6.69E-03
4.03E-03
3.81E-03
2.53E-03
1.42E-03
1.25E-03
Formaldehyde 37.70
Acetaldehyde 3.82


-------
Table 15-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
                                         RfCs for the Michigan Monitoring Sites
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer Hazard
Approximation
(HQ)
Dearborn, Michigan (Wayne County) - DEMI
Hydrochloric acid
Toluene
Xylenes
Methanol
Benzene
Formaldehyde
Hexane
Ethylbenzene
Acetaldehyde
Methyl isobutyl ketone
3,765.69
2,480.93
1,912.74
1,227.86
1,134.11
612.43
551.94
500.03
334.62
277.33
Acrolein
Hydrochloric acid
Manganese, PM
1,3 -Butadiene
Cyanide Compounds, gas
Formaldehyde
Arsenic, PM
Benzene
Acetaldehyde
Nickel, PM
1,866,893.10
188,284.52
98,186.30
67,191.86
64,535.63
62,493.26
58,998.69
37,803.81
37,180.09
32,767.48
Formaldehyde
Acetaldehyde
Naphthalene
1,3 -Butadiene
Benzene
Chloroform
Carbon Tetrachloride
Tetrachloroethylene
Trichloroethylene
Ethylbenzene
0.27
0.19
0.05
0.05
0.03
0.01
0.01
<0.01
<0.01
<0.01
River Rouge, Michigan (Wayne County) - RRMI
Hydrochloric acid
Toluene
Xylenes
Methanol
Benzene
Formaldehyde
Hexane
Ethylbenzene
Acetaldehyde
Methyl isobutyl ketone
3,765.69
2,480.93
1,912.74
1,227.86
1,134.11
612.43
551.94
500.03
334.62
277.33
Acrolein
Hydrochloric acid
Manganese, PM
1,3 -Butadiene
Cyanide Compounds, gas
Formaldehyde
Arsenic, PM
Benzene
Acetaldehyde
Nickel, PM
1,866,893.10
188,284.52
98,186.30
67,191.86
64,535.63
62,493.26
58,998.69
37,803.81
37,180.09
32,767.48
Formaldehyde
Acetaldehyde
0.33
0.19


-------
  Table 15-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
                                      RfCs for the Michigan Monitoring Sites (Continued)
to
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Noncancer Hazard
Approximation
Pollutant (HQ)
Detroit, Michigan (Wayne County) - SWMI
Hydrochloric acid
Toluene
Xylenes
Methanol
Benzene
Formaldehyde
Hexane
Ethylbenzene
Acetaldehyde
Methyl isobutyl ketone
3,765.69
2,480.93
1,912.74
1,227.86
1,134.11
612.43
551.94
500.03
334.62
277.33
Acrolein
Hydrochloric acid
Manganese, PM
1,3 -Butadiene
Cyanide Compounds, gas
Formaldehyde
Arsenic, PM
Benzene
Acetaldehyde
Nickel, PM
1,866,893.10
188,284.52
98,186.30
67,191.86
64,535.63
62,493.26
58,998.69
37,803.81
37,180.09
32,767.48
Formaldehyde 0.30
Acetaldehyde 0.19


-------
Observations from Table 15-7 include the following:

•  Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
   cancer UREs in Wayne County.

•  The pollutants with the highest toxi city-weighted emissions (of the pollutants with
   cancer UREs) for Wayne County are coke oven emissions, benzene, and
   formaldehyde.

•  Six of the highest emitted pollutants in Wayne County also have the highest toxicity-
   weighted emissions.

•  Formaldehyde has the highest cancer risk approximations for all three Michigan sites.
   This pollutant appears on both emissions-based lists. Acetaldehyde, the other
   pollutant of interest in common for all three Michigan sites, is one of the highest
   emitted pollutants but does not appear among those with the highest toxicity-
   weighted emissions.

•  In addition to formaldehyde, benzene, naphthalene, ethylbenzene, and 1,3-butadiene
   are among the pollutants with the highest cancer risk approximations for DEMI and
   appear on both emissions-based lists. Hexavalent chromium has the fifth highest
   toxicity-weighted emissions but does not appear among the highest emitted. Carbon
   tetrachloride  does not appear on either emissions-based list.

•  POM, Group 5a ranks fourth for toxicity-weighted emissions in Wayne County.
   POM, Group 5a includes benzo(a)pyrene, which has one of the lowest cancer risk
   approximations of the pollutants of interest for DEMI.


Observations from Table 15-8 include the following:

•  Hydrochloric acid, toluene, and xylenes are the highest emitted pollutants with
   noncancer RfCs in Wayne County. Wayne  County has the highest hydrochloric acid
   emissions of any county with an NMP site (by an order of magnitude).

•  The pollutants with the highest toxicity-weighted emissions (of the pollutants with
   noncancer RfCs) for Wayne County are acrolein, hydrochloric acid, and manganese.
   Although acrolein was sampled for at DEMI, this pollutant was excluded from the
   pollutants of interest designation and thus subsequent risk-based screening
   evaluations due to questions about the consistency and reliability of the
   measurements, as discussed in Section 3.2.

•  Four of the highest emitted pollutants in Wayne County also have the highest
   toxicity-weighted emissions.

•  The pollutant with the highest noncancer hazard approximation for all three Michigan
   sites is formaldehyde, although none of the pollutants of interest have associated
   noncancer hazard approximations greater than 1.0. Formaldehyde emissions rank
   sixth for Wayne County and  sixth for toxicity-weighted emissions.

                                   15-43

-------
       •  Acetaldehyde is the other pollutant these sites have in common; this pollutant also
          appears on both emissions-based lists. Acetaldehyde ranks ninth for both quantity
          emitted and toxicity-weighted emissions.

       •  Benzene is the only other pollutant that appears on all three lists for DEMI.


15.6   Summary of the 2011 Monitoring Data for DEMI, RRMI, and SWMI

       Results from several of the data treatments described in this section include the
following:

       ปซป  Nineteen pollutants, of which eight are NA TTS MQO Core Analytes, failed screens
          for DEMI. Acetaldehyde and formaldehyde both failed screens for RRMI and SWMI
          and propionaldehyde also failed a single screen for SWMI.

       *ป*  Of the site-specific pollutants of interest, formaldehyde had the highest annual
          average concentration for all three sites.

       ปซป  DEMI has the second highest annual average concentration of chloroform among
          NMP sites sampling VOCs. DEMI also has the highest annual average concentration
          of acenaphthene and the second the highest annual average concentrations of
          fluorene and naphthalene among NMP sites sampling PAHs.

       ปซป  Concentrations of benzene have been steadily decreasing at DEMI since sampling
          began at this site.
                                          15-44

-------
16.0   Site in Missouri
       This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS site in Missouri, and integrates these concentrations with
emissions, meteorological, and risk information. Data generated by sources other than ERG are
not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) for detailed discussions and definitions
regarding the various data analyses presented below.

16.1   Site Characterization
       This section characterizes the S4MO monitoring site by providing geographical and
physical information about the location of the site and the surrounding area. This information is
provided to give the reader insight regarding factors that may influence the air quality near the
site and assist in the interpretation of the ambient monitoring measurements.

       The S4MO monitoring site is located in the St. Louis, MO-IL MSA. Figure 16-1 is a
composite satellite image retrieved from ArcGIS Explorer showing the monitoring site in its
urban location. Figure 16-2 identifies nearby point source emissions locations by source
category, as reported in the 2008 NEI for point sources. Note that only sources within 10 miles
of the site are included in the facility counts provided in Figure 16-2. A 10-mile boundary was
chosen to give the reader an indication of which emissions sources and emissions source
categories could potentially have a direct effect on the air quality at the monitoring site. Further,
this boundary provides both the proximity of emissions sources to the monitoring site as well as
the quantity of such sources within a given distance of the site.  Sources outside the 10-mile
radius are still visible on the map, but have been grayed out in order to show emissions sources
just outside the boundary. Table 16-1 provides supplemental geographical information such as
land use, location setting, and locational coordinates.
                                           16-1

-------
                                       Figure 16-1. St. Louis, Missouri (S4MO) Monitoring Site
to

-------
          Figure  16-2.  NEI Point Sources Located Within 10 Miles of S4MO
                                                                                      arm'trw              so 5'trw
                                                                    Note: Due la facility density and collocation, the total facilities
                                                                    displayed may not represent all facilities within the area of interest.
                                     S4MO NATTS site
Source Category Group (No. of Facilities)
 i=   Air-con diiioning/Ref figuration (1)
 -fi   Aircraft Operations (21)
  I   Aspnatt Processing/Roofing Manufacturing (1)
 H   Automobile/Truck Manufacturing (1)
  $}   Bakery (1)
  y   BreweryDistilleryMAnery (1)
 ft   Buiktiny Construction i'2)
  B   Bulk Termmaf&'Bulk Plants (6)
 C   Chem.cal Manufacturing (9)
 •   Concrete Batcti Plant (11)
 E> ]  Crematory -Animal/Human (1)
 fฃ)   Dry Cteamng Facility (4)
      Electrical Equipment 42)
      Electricity Generation via Combustion (4)
      Electroplating, Plating. Polishing, Anodizing, & Coloring (2)
 <*)   Fabricated Metal Products (2)
 ;;^   Flexible Polyurethane Foam Production (1)
e
*
E
                                                       F
                                                      IT
                                                      A
                                                      it
                                                      ffl
                                                      *
                                                      *
                                                      o
         10 mile radius I
Toed Proce&&inaYArjricultLir.e (1C)
Gasoline/Diesel Service Station 42}
Grain Handling (?)
Heating Equipms-nt Manufacturing (T)
HospilaUl)
Hat Mix Asphuii Plan! i-i-
IrtrJustnal Macriinery and Equipmenl (1)
Inslitulional - prison (1)
Inslilulional - school <7J
Iron and Sleel Foundry (1)
Landfill (3)
Leather and Leather Products 41)
Marine Port 4 5)
Military Base/National Security Facilrty (1)
Mine/Quarry (6)
Mineral Products (1)
Miscellaneous CommerclaVlndustfral (25)
County boundary
  M   Miscellaneous Manufacturing (12)
  •   Oil and/or Gas Production (3)
  _  Pharmaceutical Manufacturing (2)
  1   Pnmary Metal Production (4)
  f   Prinlltg. Coaling SDyeng of Fabric ID
  P   Pnnting/Purjlisning (5)
  H   Pulp and Paper PlanlMttod ProOucts (2)
  R   Rubber and Miscellaneous Plastics Products (2)
  2   Secondary Metal Processing (2)
  <   Srte Remediation Actwlty (1)
  >   Solid Waste Disposal • Commerctal/lnsritulional (2)
 Y  Steel Mill (3)
  S   Surface Coabng (8)
  TT   Telecommunications (1)
  T   Textile Mill (2)
 $   Transportation and Marketing of Petroleum Products (3)
  >   V&slewater Treatment (5)
                                                                   16-3

-------
                             Table 16-1. Geographical Information for the Missouri Monitoring Site

Site
Code
S4MO

AQS Code
29-510-0085

Location
St. Louis

County
St. Louis
City
Micro- or
Metropolitan
Statistical Area
St. Louis, MO-IL
MSA
Latitude
and
Longitude
38.656436,
-90.198661

Land Use
Residential

Location
Setting
Urban/City
Center

Additional Ambient Monitoring Information1
CO, O3, Meteorological parameters, PM10, Black
carbon, PM2 5, PM25 Speciation, SO2, NOy, NO.
BOL D ITALICS = EPA-designaled NATTS Site

-------
       S4MO is located in central St. Louis. Figure 16-1 shows that the S4MO monitoring site is
located less than 1/4 mile west of 1-70. The Mississippi River, which separates Missouri and
Illinois, is less than 1 mile east of the site. Although the area directly around the monitoring site
is primarily residential, industrial facilities are located just on the other side of 1-70. Figure 16-2
shows that a large number of point sources are located within 10 miles of S4MO. The source
categories with the greatest number of point sources surrounding S4MO include aircraft
operations, which include airports as well as small runways, heliports, or landing pads;  food
processing facilities; and concrete batch plants. In the immediate vicinity of S4MO are  a
pharmaceutical manufacturing facility and a printing and publishing facility.

       Table 16-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Missouri monitoring site. Table  16-2 includes county-level
population and vehicle registration information. Table 16-2 also includes a county-level vehicle
registration-to-population ratio, which was calculated to represent the number of vehicles per
person within the monitoring site's residing county. In addition, the population within 10 miles
of the site is presented, based on postal  code population data estimates. An estimate of  10-mile
vehicle ownership was then determined by applying the county-level vehicle registration-to-
population ratio to the 10-mile population surrounding the monitoring site. Table 16-2 also
contains traffic volume information for S4MO. Finally, Table 16-2 presents the county-level
daily VMT for S4MO. Note that because the state of Missouri provides data within the  city of
St. Louis separately from St. Louis County, Table 16-2 includes the combination of the city and
county data for county-level statistics in order to compare these statistics with other sites'
county-level data.

     Table 16-2. Population, Motor Vehicle, and Traffic Information for the Missouri
                                      Monitoring Site



Site
S4MO

Estimated
County
Population1
1,316,761

County-level
Vehicle
Registration2
1,114,812
Vehicles per
Person
(Registration:
Population)
0.85

Population
within 10
miles3
796,065
Estimated
10-mile
Vehicle
Ownership
673,974
Annual
Average
Daily
Traffic4
79,558

County-
level Daily
VMT5
23,333,850
 Bounty-level population estimate reflects county and city 2011 data from the U.S. Census Bureau (Census
 Bureau, 2012b)
 2Vehicle registration reflects county and city 2011 data from the Missouri Dept of Revenue (MO DOR, 2012)
 310-mile population estimate reflects 2011 data from Xionetic (Xionetic, 2011)
 4AADT reflects 2011 data from the Missouri DOT (MO DOT, 2011)
 5VMT reflects county and city 2011 data for all public roads from the Missouri DOT (MO DOT, 2013)
 BOLD ITALICS = EPA-designated NATTS Site
                                            16-5

-------
       Observations from Table 16-2 include the following:
       •  S4MO's county-level population and vehicle registration rank 10th compared to other
          counties with NMP sites. The 10-mile population and vehicle ownership estimates for
          S4MO are in the middle third compared to other sites.
       •  The vehicle-per-person ratio is also in the middle of the range compared to other
          NMP sites.
       •  The traffic volume experienced near S4MO ranks 19th among other NMP sites. The
          traffic estimate provided is for 1-70 near Exit 249.
       •  The St. Louis City and County daily VMT ranks 12th among counties with NMP sites
          (where VMT data were available).

16.2   Meteorological Characterization
       The following sections characterize the meteorological conditions near the monitoring
site in Missouri on sample days, as well as over the course of the year.

16.2.1  Climate Summary
       The city of St. Louis is located along the Mississippi River, which acts as Missouri's
eastern border. St. Louis has a climate that is continental in nature, with cold, dry winters; warm,
somewhat wetter summers; and significant seasonal variability. Warm, moist air flowing
northward from the Gulf of Mexico alternating with cold, dry air marching southward  from
Canada and the northern U.S.  result in weather patterns that do not persist for very long. The
City of St. Louis experiences the urban heat island effect, retaining more heat within the city than
outlying areas (Bair, 1992 and MCC, 2013).

16.2.2  Meteorological Conditions in 2011
       Hourly meteorological data from the NWS weather station nearest this site were retrieved
for 2011 (NCDC, 2011). The closest weather station is located at St. Louis Downtown Airport
(WBAN 03960). Additional information about this weather station, such as the distance between
the site and the weather station, is provided in Table 16-3. These data were used to determine
how meteorological conditions on sample days vary from conditions experienced throughout the
year.
                                          16-6

-------
                     Table 16-3. Average Meteorological Conditions near the Missouri Monitoring Site
Closest NWS
Station
(WBAN and
Coordinates)
Distance
and
Direction
From Site
Average
Type1
Average
Maximum
Temperature
(ฐF)
Average
Temperature
(ฐF)
Average
Dew Point
Temperature
(ฐF)
Average
Wet Bulb
Temperature
(ฐF)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
St. Louis, Missouri - S4MO
St. Louis
Downtown
Airport
03960
(38.57, -90.16)
6.27
miles
156ฐ
(SSE)
Sample
Day
2011
64.8
ฑ5.3
66.5
ฑ2.1
55.6
ฑ4.9
56.7
ฑ1.9
44.7
ฑ4.7
46.0
ฑ1.8
50.0
ฑ4.4
51.1
ฑ1.7
69.7
ฑ3.0
70.5
ฑ1.2
1016.4
ฑ 1.8
1015.9
ฑ0.7
6.3
ฑ0.7
5.8
ฑ0.3
Sample day averages are highlighted to help differentiate the sample day averages from the full-year averages.

-------
       Table 16-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for the entire year for 2011. Also included in Table 16-3 is the
95 percent confidence interval for each parameter. Although average meteorological conditions
on sample days are not statistically different than the average meteorological conditions
throughout 2011, temperatures do appear slightly cooler on sample days, as shown in Table 16-3.
This is likely the result of five additional days in December 2011 when make-up samples were
collected.

16.2.3 Back Trajectory Analysis
       Figure 16-3 is the composite back trajectory map for days on which samples were
collected at the S4MO monitoring site in 2011. Included in Figure 16-3 are four back trajectories
per sample day. Figure 16-4 is the corresponding cluster analysis. An in-depth description of
these maps and how they were generated is presented in Section 3.5.2.1. For the composite map,
each line represents the 24-hour trajectory along which a parcel of air traveled toward the
monitoring site on a given sample day and time, based on an initial height of 50 meters AGL. For
the cluster analysis, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric circle around the site in Figures 16-3 and 16-4 represents 100 miles.
                                          16-8

-------
Figure 16-3. 2011 Composite Back Trajectory Map for S4MO
    Figure 16-4. Back Trajectory Cluster Map for S4MO
                          16-9

-------
       Observations from Figures 16-3 and 16-4 for S4MO include the following:

       •  Back trajectories originated from a variety of directions at S4MO, although back
          trajectories from the northwest and south-southeast to south-south west were most
          common.

       •  The 24-hour air  shed domain for S4MO is similar in size to other NMP sites. The
          farthest away a back trajectory originated was greater than 800 miles, over northern
          North Dakota. However, the average trajectory length was 239 miles and most back
          trajectories (87 percent) originated within 400 miles of the monitoring site.

       •  The cluster analysis shows that many (32 percent) back trajectories originated to the
          south to southwest of S4MO, although of varying lengths. Another 20 percent of back
          trajectories originated to the southeast to south of S4MO. Thus, greater than
          50 percent of back trajectories originated from a direction with a southerly
          component.

       •  The longest back trajectories originated to the northwest and north of the site. Twenty
          percent of back trajectories originated to the northwest of S4MO. Another five
          percent originated to the north over Wisconsin and the Great Lakes.

       •  Back trajectories originating from the east or the west tended to be shorter in length.
          The cluster trajectory originating to the west of S4MO (13 percent) represents back
          trajectories originating over the state of Missouri. The cluster trajectory originating to
          the northeast of  S4MO (12 percent) represents back trajectories that originate over the
          states of Illinois  and Indiana.


16.2.4 Wind Rose Comparison

       Hourly surface wind data from the NWS weather station at  St. Louis Downtown Airport

near S4MO were uploaded  into a wind rose software program to produce  customized wind  roses,

as described in Section 3.5.2.2. A wind rose shows the frequency of wind directions using

"petals" positioned around a 16-point compass, and uses different colors to represent wind

speeds.
       Figure 16-5 presents a map showing the distance between the NWS station and S4MO,

which may be useful for identifying topographical influences that may affect the meteorological

patterns experienced at this location. Figure 16-5 also presents three different wind roses for the

S4MO monitoring site. First, a historical wind rose representing 2001 to 2010 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 2011 is presented.

Next, a wind rose representing wind data for days on which samples were collected in 2011 is
                                          16-10

-------
presented. These can be used to identify the predominant wind speed and direction for 2011 and

determine if wind observations on sample days were representative of conditions experienced

over the entire year and historically.


       Observations from Figure 16-5 for S4MO include the following:

       •  The St. Louis Downtown Airport weather station is located approximately 6.3 miles
          south-southeast of S4MO. The weather station location is across the Mississippi River
          and state border in Illinois.

       •  The historical wind rose shows that winds from the southeast, south-southeast, and
          south were frequently observed near S4MO. Winds from these directions account for
          approximately 28 percent of observations. Calm winds (<2 knots) were observed for
          approximately 19 percent of the hourly wind measurements. Winds from the west to
          northwest to north account for the bulk of the remaining wind observations. The
          strongest winds were from the west to northwest.

       •  The wind patterns shown on the 2011 wind rose generally resemble those shown on
          the historical wind rose, although there were fewer southeasterly winds and more
          southerly winds.

       •  The sample day wind patterns also resemble the historical and full-year wind patterns,
          although there was a  higher percentage of south-southeasterly winds  and slightly
          fewer calm winds.

       •  The similarities in the wind patterns between the wind roses indicate that wind
          conditions on sample days were similar to wind conditions experienced throughout
          2011 and historically.
                                         16-11

-------
Figure 16-5. Wind Roses for the St. Louis Downtown Airport Weather Station near S4MO
  Distance between S4MO and NWS Station
2001-2010 Historical Wind Rose
           2011 Wind Rose
    Sample Day Wind Rose
                                      16-12

-------
16.3   Pollutants of Interest

       Site-specific "pollutants of interest" were determined for S4MO in order to allow 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." 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. In addition, if any of the NATTS

MQO Core Analytes measured by the monitoring site did not meet the pollutant of interest

criteria based on the preliminary risk-based screening, that pollutant was added to the list of site-

specific pollutants of interest. A more in-depth description of the risk-based screening process is
presented in Section 3.2.


       Table 16-4 presents the results of the preliminary risk-based screening process for S4MO.

The pollutants that failed at least one screen and contributed to 95 percent of the total failed

screens for the monitoring site  are shaded. NATTS MQO Core Analytes are bolded. Thus,

pollutants of interest are shaded and/or bolded. S4MO sampled for VOCs, PAHs, carbonyl

compounds, metals (PMio), and hexavalent chromium.


       Observations from Table 16-4 include the following:

       •  Twenty-five pollutants, of which 13 are NATTS MQO Core Analytes, failed at least
          one screen for S4MO. S4MO failed the greatest number of screens among all NMP
          sites.

       •  Formaldehyde, acetaldehyde, and benzene were detected in every carbonyl compound
          or VOC sample collected and failed 100 percent of screens for S4MO. Other
          pollutants also failed 100 percent of screens but were detected less frequently:
          1,3-butadiene, 1,2-dichloroethane, hexachloro-1,3-butadiene, 1,2-dibromoethane,
          1,1,2,2-tetrachloroethane, and chloromethylbenzene. With the exception of
          1,3-butadiene, these pollutants were detected in only a few samples.

       •  Sixteen pollutants were identified as pollutants of interest for S4MO based on the
          risk-based screening process; of these, 10 are NATTS MQO Core Analytes. Three
          additional pollutants (nickel, hexavalent chromium, and chloroform) were added to
          S4MO's  pollutants of interest because they are NATTS MQO Core Analytes, even
          though they did not contribute to 95 percent of S4MO's failed screens. Five more
          pollutants (beryllium, benzo(a)pyrene, tetrachloroethylene, trichloroethylene, and
          vinyl chloride) were also added to S4MO's pollutants of interest because they are
          NATTS MQO Core Analytes, even though they did not fail any screens. These five
                                          16-13

-------
          pollutants are not shown in Table 16-4 but are shown in subsequent tables in the
          sections that follow.
        Table 16-4. Risk-Based Screening Results for the Missouri Monitoring Site
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
St. Louis, Missouri - S4MO
Acetaldehyde
Formaldehyde
Naphthalene
Arsenic (PM10)
Benzene
Carbon Tetrachloride
1,3-Butadiene
Manganese (PM10)
ฃ>-Dichlorobenzene
Cadmium (PM10)
1 ,2-Dichloroethane
Ethylbenzene
Acenaphthene
Fluorene
Lead (PM10)
Hexachloro- 1 ,3 -butadiene
1 ,2-Dibromoethane
Nickel (PM10)
1 , 1 ,2,2-Tetrachloroethane
Propionaldehyde
Chloromethylbenzene
Hexavalent Chromium
Carbon Bisulfide
Chloroform
1 , 1 ,2-Trichloroethane
0.45
0.077
0.029
0.00023
0.13
0.17
0.03
0.005
0.091
0.00056
0.038
0.4
0.011
0.011
0.015
0.045
0.0017
0.0021
0.017
0.8
0.02
0.000083
70
9.8
0.0625
Total
59
59
59
57
57
56
52
44
29
20
18
15
12
12
12
11
7
6
6
4
2
2
1
1
1
602
59
59
61
59
57
57
52
59
47
59
18
57
61
61
59
11
7
59
6
59
2
60
57
44
2
1,132
100.00
100.00
96.72
96.61
100.00
98.25
100.00
74.58
61.70
33.90
100.00
26.32
19.67
19.67
20.34
100.00
100.00
10.17
100.00
6.78
100.00
3.33
1.75
2.27
50.00
53.18
9.80
9.80
9.80
9.47
9.47
9.30
8.64
7.31
4.82
3.32
2.99
2.49
1.99
1.99
1.99
1.83
1.16
1.00
1.00
0.66
0.33
0.33
0.17
0.17
0.17
9.80
19.60
29.40
38.87
48.34
57.64
66.28
73.59
78.41
81.73
84.72
87.21
89.20
91.20
93.19
95.02
96.18
97.18
98.17
98.84
99.17
99.50
99.67
99.83
100.00

16.4   Concentrations
       This section presents various concentration averages used to characterize pollution levels
at the Missouri monitoring site. Where applicable, the following calculations and data analyses
were performed: Time period-based concentration averages (quarterly and annual) are provided
for the pollutants of interest for S4MO, where the data meet the applicable criteria.
Concentration averages for select pollutants are also presented graphically to illustrate how the
site's concentrations compare to the program-level averages, as presented in Section 4.1. In
                                          16-14

-------
addition, concentration averages for select pollutants are presented from previous years of
sampling in order to characterize concentration trends at the site. Additional site-specific
statistical summaries for S4MO are provided in Appendices J, L, M, N, and O.

16.4.1 2011 Concentration Averages
       Quarterly and annual concentration averages were calculated for the pollutants of interest
for the Missouri site, as described in Section 3.1. The quarterly average of a particular pollutant
is simply the average concentration of the preprocessed daily measurements over a given
calendar quarter. Quarterly average concentrations include the  substitution of zeros for all non-
detects. A site must have a minimum of 75 percent valid samples of the total number of samples
possible within a given quarter for a quarterly average to be calculated. An annual average
includes all measured detections and substituted zeros for non-detects for the entire year of
sampling. Annual averages were calculated for pollutants where at least three valid quarterly
averages could be calculated and where method completeness was greater than or equal to
85 percent, as  presented in Section 2.4. Quarterly and annual average concentrations for S4MO
are presented in Table 16-5, where applicable. Note that concentrations of the PAHs, metals, and
hexavalent chromium are presented in ng/m3 for ease of viewing. Also note that if a pollutant
was not detected in a given calendar quarter, the quarterly average simply reflects "0" because
only zeros  substituted for non-detects  were factored into the quarterly average concentration.
Table 16-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
                               the Missouri Monitoring Site



Pollutant
#of
Measured
Detections vs.
# of Samples
1st
Quarter
Average
(Ug/m3)
2nd
Quarter
Average
(Ug/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)

Annual
Average
(Ug/m3)
St. Louis, Missouri - S4MO

Acetaldehyde

Benzene

1,3 -Butadiene

Carbon Tetrachloride

Chloroform

59/59

57/57

52/57

57/57

44/57
2.47
ฑ0.42
0.89
ฑ0.14
0.10
ฑ0.03
0.51
ฑ0.06
0.12
ฑ0.06
3.14
ฑ0.80
0.69
ฑ0.07
0.06
ฑ0.02
0.60
ฑ0.04
0.19
ฑ0.15
3.97
ฑ1.03
0.76
ฑ0.09
0.07
ฑ0.02
0.71
ฑ0.06
1.00
ฑ1.67
1.31
ฑ0.25
0.86
ฑ0.15
0.11
ฑ0.03
0.63
ฑ0.10
0.09
ฑ0.04
2.75
ฑ0.43
0.80
ฑ0.06
0.08
ฑ0.01
0.61
ฑ0.04
0.35
ฑ0.40
    NA = Not available due to the criteria for calculating a quarterly and/or annual average.
    a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of
    viewing.
                                           16-15

-------
Table 16-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
                        the Missouri Monitoring Site (Continued)
Pollutant
ฃ>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
Hexachloro- 1 , 3 -butadiene
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
Acenaphthene3
Arsenic (PM10)a
Benzo(a)pyrenea
Bery Ilium (PM10)a
Cadmium (PM10)a
Fluorene a
Hexavalent Chromium3
Lead(PM10)a
Manganese (PM10) a
Naphthalene a
Nickel (PM10)a
#of
Measured
Detections vs.
# of Samples
47/57
18/57
57/57
59/59
11/57
55/57
25/57
8/57
61/61
59/59
59/61
59/59
59/59
61/61
60/61
59/59
59/59
61/61
59/59
1st
Quarter
Average
(Hg/m3)
0.21
ฑ0.21
0.02
ฑ0.02
0.41
ฑ0.10
5.30
ฑ2.97
0.04
ฑ0.04
0.19
ฑ0.07
0.04
ฑ0.02
0.01
ฑ0.01
1.69
ฑ0.61
0.78
ฑ0.28
0.16
ฑ0.05
0.01
ฑ0.01
0.60
ฑ0.32
2.54
ฑ0.64
0.03
ฑ0.02
9.53
ฑ3.57
7.27
ฑ2.07
74.38
ฑ 27.68
1.00
ฑ0.19
2nd
Quarter
Average
(Ug/m3)
0.12
ฑ0.09
0.05
ฑ0.03
0.27
ฑ0.04
4.57
ฑ1.56
0.01
ฑ0.02
0.16
ฑ0.03
0.02
ฑ0.02
0.01
ฑ0.01
5.87
ฑ2.40
0.81
ฑ0.18
0.10
ฑ0.05
0.01
ฑ0.01
0.63
ฑ0.19
6.66
ฑ2.58
0.04
ฑ0.01
10.81
ฑ3.27
41.50
ฑ52.71
61.09
ฑ 16.32
1.53
ฑ0.46
3rd
Quarter
Average
(Ug/m3)
0.39
ฑ0.41
0.01
ฑ0.02
0.37
ฑ0.05
5.43
ฑ1.65
0.01
ฑ0.02
0.15
ฑ0.04
0.04
ฑ0.03
0.01
ฑ0.01
10.71
ฑ3.79
0.90
ฑ0.21
0.07
ฑ0.02
0.01
ฑ0.01
0.51
ฑ0.23
11.28
ฑ3.78
0.03
ฑ0.01
10.55
ฑ3.63
12.67
ฑ4.57
106.70
ฑ25.01
1.20
ฑ0.79
4th
Quarter
Average
(Ug/m3)
0.14
ฑ0.09
0.04
ฑ0.03
0.46
ฑ0.09
1.76
ฑ0.36
0.03
ฑ0.04
0.20
ฑ0.14
0.04
ฑ0.02
0.01
ฑ0.01
3.66
ฑ2.05
1.00
ฑ0.45
0.18
ฑ0.07
0.01
ฑ0.01
0.49
ฑ0.24
4.18
ฑ1.57
0.03
ฑ0.01
10.67
ฑ4.07
11.13
ฑ4.63
88.28
ฑ21.76
1.05
ฑ0.45
Annual
Average
(Ug/m3)
0.21
ฑ0.11
0.03
ฑ0.01
0.38
ฑ0.04
4.25
ฑ0.92
0.02
ฑ0.01
0.18
ฑ0.04
0.03
ฑ0.01
0.01
ฑ0.01
5.41
ฑ1.45
0.87
ฑ0.14
0.13
ฑ0.03
0.01
ฑ0.01
0.56
ฑ0.12
6.08
ฑ1.40
0.03
ฑ0.01
10.42
ฑ1.72
18.42
ฑ 13.18
83.82
ฑ11.75
1.20
ฑ0.26
    NA = Not available due to the criteria for calculating a quarterly and/or annual average.
    a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of
    viewing.
       Observations for S4MO from Table 16-5 include the following:

       •   The pollutants with the highest annual average concentrations by mass are
          formaldehyde (4.25 ฑ 0.92 |ig/m3) and acetaldehyde (2.75 ฑ 0.43 |ig/m3). These are
          the only pollutants of interest with annual averages greater than 1 |ig/m3.
                                           16-16

-------
•  The first, second, and third quarter average formaldehyde concentrations are
   significantly higher than the fourth quarter average and their confidence intervals
   reflect a relatively high-level of variability in the measurements. Concentrations of
   formaldehyde ranged from 0.714 |ig/m3to 17.8 |ig/m3. The two highest
   concentrations were measured on back-to-back sample days in February. Four
   concentrations greater than 10 |ig/m3 were measured at S4MO and at least one was
   measured in each quarter except the fourth quarter. Conversely, the six lowest
   measurements of formaldehyde were all measured during the fourth quarter of 2011.

•  The confidence interval for chloroform is greater than the annual average
   concentration it is associated with, indicating that there are likely outliers affecting
   this average. A review of the quarterly averages shows that any outliers were likely
   measured during the  third quarter of 2011. The maximum chloroform concentration
   was measured at S4MO on July 8, 2011 (11.5 |ig/m3). The next highest concentration
   measured at S4MO is an order of magnitude less (1.17 |ig/m3). S4MO is the only site
   besides NBIL with a chloroform concentration greater than 10 |ig/m3.

•  The confidence intervals associated with the first and third quarter averages of
   />-dichlorobenzene are greater than or equal to the averages themselves, indicating the
   likely influence of outliers. A review of the data shows that the maximum
   />-dichlorobenzene concentrations were measured in September (2.83 |ig/m3) and
   January (1.50 |ig/m3). These are the two highest/>-dichlorobenzene concentrations
   measured across all NMP sites sampling VOCs and represent two of the four
   concentrations of />-dichlorobenzene greater than  1 |ig/m3 measured across the
   program. For perspective, the median/>-dichlorobenzene concentration for S4MO is
   0.096 |ig/m3.

•  The quarterly averages of several of the VOCs (benzene, 1,3-butadiene,  ethylbenzene,
   and tetrachloroethylene) appear to be higher during the colder months of the year.
   However, the confidence intervals indicate that the differences between the quarterly
   averages are not statistically  significant.

•  The metals sampler was changed in July 2011, and Teflonฎ filters were used to
   collect metals samples rather than quartz filters. It does not appear that the
   instrumentation and filter changes resulted in a significant difference in the quarterly
   averages of the metals.

•  Manganese has the highest annual average concentration among the PMio  metals
   measured at S4MO. The confidence interval associated with this annual  average
   indicates a  high level of variability within the measurements. The second quarter
   average is more than three times higher than the next highest quarterly average
   concentration. Further, the confidence interval  for the second quarter average  is
   greater than the average concentration itself. A review of the data shows that the
   maximum concentration of manganese was measured on June 2, 2011 (395 ng/m3)
   and is the maximum  manganese concentration  measured among all NMP sites
   sampling metals. This measurement is three times greater than the  next highest
   concentration of manganese measured across the program (130 ng/m3, measured at
   PXSS) and six times greater than the next highest concentration of manganese

                                  16-17

-------
          measured at S4MO (65.3 ng/m3). For perspective, the median manganese
          concentration for S4MO is 10.0 ng/m3.

       •  Naphthalene has the highest annual average concentration among the PAHs measured
          at S4MO. The confidence intervals calculated for the naphthalene averages indicate
          that there is a high level of variability in the measurements. Concentrations of
          naphthalene measured at S4MO range from 18.0 ng/m3 to 238 ng/m3 with a median
          concentration of 72.2 ng/m3.

       •  Concentrations of acenaphthene and fluorene appear to be highest during the warmer
          months of the year, particularly the third quarter of 2011. The averages for these
          quarters have relatively large confidence intervals associated with them.  A review of
          the data shows that the maximum concentration of each pollutant was measured on
          July 2, 2011 (31.8 ng/m3 and 31.4 ng/m3, respectively). The July 2nd concentrations
          are almost twice the next highest concentrations measured at S4MO, which were
          measured on the same date in June. Of the concentrations of each pollutant greater
          than 10 ng/m3, the majority were measured during the third quarter, followed by the
          second quarter. Of the concentrations less than 2 ng/m3, the measurements were
          evenly split between the first and fourth quarters.


       •  At the beginning of 2013, the Missouri Department of Natural Resources discovered
          that a sampler contamination issue resulted in artificially elevated concentrations  of
          acrylonitrile from September 2010 through October 2012. Thus, the acrylonitrile
          results from this time period were invalidated, which includes all of the results for
          2011.

       Tables 4-9 through 4-12 present the sites with the 10 highest annual average

concentrations for each of the program-level pollutants of interest. Observations for S4MO from

those tables include the following:

       •  S4MO appears in Tables 4-9 through 4-12 a total of 21 times, the most of any NMP
          site.

       •  S4MO is listed among the sites with the highest annual average concentrations of all
          VOCs except benzene and 1,3-butadiene. S4MO has the highest annual average
          concentration of hexachloro-1,3-butadiene and the second highest annual average
          concentrations of 1,2-dichloroethane and/>-dichlorobenzene. This site also has the
          fourth highest annual average concentration of chloroform.

       •  S4MO has the second highest annual average concentration  of formaldehyde and the
          third highest annual average concentration of acetaldehyde.

       •  S4MO's annual average concentration of benzo(a)pyrene ranks sixth highest among
          NMP sites sampling PAHs. This site's annual average  acenaphthene and fluorene
          concentrations each rank seventh among NMP sites sampling PAHs.
                                         16-18

-------
       •  S4MO has the highest annual average concentrations of arsenic, cadmium, and lead
          among NMP sites sampling PMio metals (and those sampling TSP metals). This site's
          annual average concentration of manganese ranked second among NMP sites
          sampling PMio metals.

16.4.2 Concentration Comparison
       In order to better illustrate how a site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for the selected NATTS MQO Core
Analytes listed in Section 3.5.3, where applicable. Thus, box plots for acetaldehyde, arsenic,
benzene, benzo(a)pyrene, 1,3-butadiene, formaldehyde, hexavalent chromium, lead, manganese,
and naphthalene were created for S4MO. Figures 16-6 through 16-15 overlay the site's
minimum, annual average, and maximum concentrations onto the program-level minimum, first
quartile, median, average, third quartile, and maximum concentrations, as described in
Section 3.5.3.
        Figure 16-6. Program vs. Site-Specific Average Acetaldehyde Concentration
 S4MO
                                         Concentration (|
                Program:   1st Quartile    2nd Quartile    3rd Quartile     4th Quartile     Average
                Site:     Site Average    Site Minimum/Maximum

       Figure 16-7. Program vs. Site-Specific Average Arsenic (PMio) Concentration
                                                                                      16
 S4P/0
             0.5
                                         2        2.5
                                         Concentration (ng/m3)
                                                                    3.5
                                                                                      4.5
Program
Site:
: 1st Quartile
Site Average
o
2nd Quartile 3rd Quartile 4th Quartile Ave
• n
Site Minimum/Maximum
'rage

                                          16-19

-------
           Figure 16-8. Program vs. Site-Specific Average Benzene Concentration
S4f,'C
                                                                   j  Program Max Concentration = 23.8
                                           45
                                               Concentration
                 Program:    IstQuartile     2ndQuartile     SrdQuartile     4thQuartile     Average
                 Site:      Site Average     Site Minimum/Maximum

                                                                                                    10
      Figure 16-9. Program vs. Site-Specific Average Benzo(a)pyrene Concentration
                                        0.75           1           1.15
                                               Concentration (ng/m3)
                 Program:   IstQuartile     2ndQuartile     3rdQuartile     4thQuartile      Average
                 Site:      Site Average     Site Minimum/Maximum

       Figure 16-10. Program vs. Site-Specific Average 1,3-Butadiene Concentration
                                                                     Program Max Concentration = 9.51
                    5.5
                                                    1.5
                                               Concentration (jig/mi)
                 Program:    IstQuartile     2ndQuartile     SrdQuartile     4thQuartile     Average
                 Site:      Site Average     Site Minimum/Maximum

                                                16-20

-------
      Figure 16-11. Program vs. Site-Specific Average Formaldehyde Concentration
S4f,'C
                                    10              15
                                              Concentration
                 Program:   IstQuartile      2ndQuartile      SrdQuartile     4thQuartile     Average
                 Site:      Site Average      Site Minimum/Maximum

  Figure 16-12. Program vs. Site-Specific Average Hexavalent Chromium Concentration
S4f/C
                   0.05
                                   5.1
                                                   0.15
                                              Concentration (ng/m3)
                                                                   0.2
                                                                                  DL25
                 Program:   IstQuartile      2ndQuartile      SrdQuartile     4thQuartile     Average
                 Site:      Site Average      Site Minimum/Maximum

        Figure 16-13. Program vs. Site-Specific Average Lead (PMi0) Concentration
S4f,'C
                               -u-
                               10
                                             15            20
                                              Concentration (ng/m3)
                                                                        25
                                                                                     30
                 Program:   IstQuartile      2ndQuartile      SrdQuartile     4thQuartile     Average
                 Site:      Site Average      Site Minimum/Maximum

                                                                                                   35
                                                16-21

-------
   Figure 16-14. Program vs. Site-Specific Average Manganese (PMi0) Concentration
S4P/0



n
; Program Max Concentration = 395 ng/m3


;

                                            100        125
                                        Concentration (ng/mi)
                                                                150
                                                                           17E
                                                                                     200
              Program:   IstQuartile    2ndQuartile    SrdQuartile     4thQuartile     Average
              Site:     Site Average    Site Minimum/Maximum

      Figure 16-15. Program vs. Site-Specific Average Naphthalene Concentration
S4f/C
                                                          Program Max Concentration = 779 ng/m3
            50
                   100
                            is:
                                    200      250      300
                                        Concentration (ng/mi)
                                                            35:
                                                                     400
                                                                             45:
Program:
Site:
IstQuartile 2ndC
Site Average Site l\
o
      Observations from Figures 16-6 through 16-15 include the following:

         •   Figure 16-6 shows that S4MO's annual average acetaldehyde concentration is
             greater than the program-level average and third quartile. Although the maximum
             concentration measured at S4MO is not the maximum concentration measured
             across the program, it is the fifth highest acetaldehyde concentration measured
             among NMP sites sampling this pollutant. There were no non-detects of
             acetaldehyde measured at S4MO or across the program.

         •   Figure 16-7 shows that S4MO's annual average arsenic (PMio) concentration is
             greater than the program-level average and third quartile. Recall from the
             previous section that this site has the highest annual average arsenic concentration
             among NMP sites sampling metals. The maximum concentration measured at
             S4MO is the second highest concentration measured across the program. There
             were no non-detects of arsenic measured at S4MO.

         •   Figure 16-8 is the box plot for benzene. Note that the program-level maximum
             concentration (23.8 |ig/m3) is not shown directly on the box plot because the scale
             of the box plot would be too large to readily observe data points at the lower end
             of the concentration range. Thus, the scale has been reduced to 10 |ig/m3.
             Figure 16-8 for benzene shows that the annual average benzene concentration for
             S4MO is less than the program-level average concentration but greater than the
             program-level median concentration. The  maximum benzene concentration
                                         16-22

-------
   measured at S4MO is considerably less than the maximum concentration
   measured at the program level. There were no non-detects of benzene measured at
   S4MO or across the program.

•  Figure 16-9 is the box plot for benzo(a)pyrene. Note that the first quartile for this
   pollutant is zero and is not visible on this box plot. This box plot shows that the
   annual average concentration for S4MO is greater than the program-level average
   and third quartile. Recall from the previous section that the annual  average
   concentration for S4MO ranked third among  sites sampling benzo(a)pyrene.
   Figure 16-9 also shows that the maximum concentration measured at S4MO is
   less than the maximum concentration measured across the program. Two non-
   detects of benzo(a)pyrene were measured at S4MO.

•  Similar to the benzene graph, the program-level maximum 1,3-butadiene
   concentration (9.51 |ig/m3) is not shown directly on the box plot as the scale has
   been reduced to 3 |ig/m3 to allow for the observation of data points at the lower
   end of the concentration range. Figure 16-10  for 1,3-butadiene shows that the
   annual average concentration for S4MO is less than the program-level average
   concentration but greater than the program-level median concentration. The
   maximum concentration measured at S4MO is considerably less than the
   maximum concentration measured across the program. There were a few non-
   detects of 1,3-butadiene measured at S4MO.

•  Figure 16-11 for formaldehyde shows that the annual average concentration for
   S4MO is greater than the program-level average concentration as well as the
   program-level third quartile. The maximum formaldehyde concentration
   measured at S4MO is less than the maximum concentration measured across the
   program but ranks sixth highest among sites sampling formaldehyde. There were
   no non-detects of formaldehyde measured at  S4MO or across the program.

•  Figure 16-12 shows that the annual average concentration of hexavalent
   chromium for S4MO is greater than both the  program-level average concentration
   and program-level third quartile. The maximum concentration measured at S4MO
   is less than  the program-level maximum concentration. A single non-detect of
   hexavalent  chromium was measured at S4MO.

•  Figure 16-13 shows that S4MO's annual average lead (PMio) concentration is
   more than double the program-level average concentration. Recall  from the
   previous section that this site has the highest  annual average lead concentration
   among sites sampling metals. In addition, the maximum lead concentration
   measured at S4MO is the maximum concentration measured across the program.
   There were no non-detects of lead measured at S4MO or across the program.

•  Figure 16-14 is the box plot for manganese (PMio). Note that the program-level
   maximum concentration (395 ng/m3) is not shown directly on the box plot as the
   scale has been reduced to 200 ng/m3 to allow for the observation of data points at
   the lower end of the concentration range. Figure 16-14 shows that  S4MO's annual
   average manganese (PMio) concentration is greater than both the program-level

                               16-23

-------
             average concentration and third quartile. Recall from the previous section that this
             site has the second highest annual average manganese concentration among sites
             sampling PMio metals. The maximum concentration of manganese measured at
             S4MO is the maximum concentration measured across the program. There were
             no non-detects of manganese measured at S4MO or across the program.
          •  Figure 16-15 is the box plot for naphthalene. Note that the program-level
             maximum concentration (779 ng/m3) is not shown directly on the box plot as the
             scale has been reduced to 500 ng/m3 to allow for the observation of data points at
             the lower end of the concentration range. Figure 16-15 shows that the annual
             average naphthalene concentration for S4MO is just greater than the program-
             level average concentration. The maximum naphthalene concentration measured
             at S4MO is less than the program-level maximum concentration. There were no
             non-detects of naphthalene measured at S4MO or across the program.

16.4.3  Concentration Trends
       A site-specific trends evaluation was completed for sites that have sampled one or more
of the selected NATTS MQO Core Analytes for 5 consecutive years or longer, as described in
Section 3.5.4. S4MO has sampled VOCs and carbonyl compounds under the NMP since 2002,
PMio metals since 2003, and hexavalent chromium since 2005. Thus, Figures 16-16 through
16-23 present the annual statistical  metrics for acetaldehyde, arsenic, benzene, 1,3-butadiene,
formaldehyde, hexavalent chromium, lead, and manganese (respectively) for S4MO. The
statistical metrics presented for assessing trends include the substitution of zeros for non-detects.
S4MO began sampling PAHs under the NMP in 2008; therefore, the trends analysis was not
conducted for the PAHs.
                                         16-24

-------
 Figure 16-16. Annual Statistical Metrics for Acetaldehyde Concentrations
                           Measured at S4MO



1
Concentration
I








3




r
^^
r
* T
n f , ^U
* ... "^ [ ^.
^] p-| '••••ง• 	 .i. ™ •••-Li ....Ly" r"1 W
• t L-rJ !53 ^^ L-^-l
2003 2004 2005 2006 2007 2008 2009 2010 2011
Year
• 5th Percentile — Minimum — Median — Max mum • 95th Percentile "^"Average
Figure 16-17. Annual Statistical Metrics for Arsenic (PMio) Concentrations
                           Measured at S4MO
        -  Minimum     —  Median     -  Maximum     •  95th Percentile    # 5th Percentile   .. 4.. Average
                                   16-25

-------
   Figure 16-18. Annual Statistical Metrics for Benzene Concentrations
                            Measured at S4MO




ST
jncent ratio
O H
|
< 3 _














•


+ ..
— •—

.
r T
i
r ^^ r
r-l-, i
nn
"""^-•-•— r^n
'— •— ' 1 — •— ' ^™ ^^ — ซ-
-r UjJ ^ Lj-l I
2003 2004 2005 2006 2007 2008 2009 2010 2011
Year
• 5th Pe re entile — Minimum — Median — Max mum * 95th Percentile ..^.. Average
Figure 16-19. Annual Statistical Metrics for 1,3-Butadiene Concentrations
                            Measured at S4MO
      2003      2004      2005      2006
                                       2007      2003
                                       Year
                                                        2009      2010
       •  SthPercentile     - Minimum    —  Median     —  Maximum    •  95thPercentile   • •*** Average
                                    16-26

-------
    Figure 16-20. Annual Statistical Metrics for Formaldehyde Concentrations
                            Measured at S4MO





J 30 -
E
.a
c
i
a
S
<

5 -

0 -1







T
_pU_






I
I
r
r r I
^ - r
-L^ 	 ••*— • rH i c i
• f-1 a-1 ^^ 	 >— •-" 1— •-
2003 2004 2005 2006 2007 2008 2009 2010 2011
Year
4 5th Pe re entile — Minimum — Median — Max mum * 95th Percentile ..^.. Average
Figure 16-21. Annual Statistical Metrics for Hexavalent Chromium Concentrations
                            Measured at S4MO




E
.1
1
i
& 02
2 0.2
ฃ
<





















T



2005





• 5th Percentile














2006
r



2007
- Minimum















T
T

^

. . r
2008 2009 2010 2011
Year
— Median - Maximum • 95th Percentile ..+.. Average
                                   16-27

-------
   Figure 16-22 Annual Statistical Metrics for Lead (PMi0) Concentrations
                           Measured at S4MO



nation iii|T/nil.
: ง
Average Concen
i i




























+ "
	 • —


2004


0
_L

2305
5th Percentile















2006
- M
nimum








1337



Year
Median




1 — |

—
•
2008




Maximum




••

[


_*_••_ ijjfc.
— • — • E
2009 2010 2011
* 95th Percentile ..+.. Average
Figure 16-23. Annual Statistical Metrics for Manganese (PMio) Concentrations
                           Measured at S4MO



___
(ration (ng/i
i i
s
3
^














J







The maximum
concentrationfor
2008 is 734 ng/nn3





I
I

k--- — ...

2004 2005
I


2006


	

Minimum — Median

T
— • —
•IPI
2007


Year
Maximum •

3_











"• 	 | — 1 — |
^ 	 ^^n. 	
2008 2009







2010



95th Percentile • 5th Percentile ..*..

•M
3








>
2011
Average
                                  16-28

-------
Observations from Figure 16-16 for acetaldehyde include the following:

•  Because carbonyl compound sampling did not begin until December 2002, 2002 data
   were excluded from this analysis.

•  The maximum acetaldehyde concentration was measured in 2004 and is more than
   twice the next highest concentration (measured in 2007).

•  Even with the maximum concentration measured in 2004, nearly all of the statistical
   metrics decreased from 2003 to 2004. The maximum concentration measured in 2004
   (32.5 |ig/m3) is nearly six times higher than the next highest concentration
   (5.72 |ig/m3) measured in 2004 and the median concentration for that year is
   2.91 |ig/m3.

•  The average concentrations have an undulating pattern in Figure 16-16 and have
   fluctuated between 1.83 |ig/m3 (2008) and 4.10 |ig/m3 (2010).

•  Even though the maximum, 95th percentile, and average concentrations increased
   from 2005 to 2006, the median concentration actually decreased. Although there were
   four concentrations greater than 6 |ig/m3 measured in 2006, the number of
   concentrations less than 3 |ig/m3 increased from 36 in 2005 to 47 in 2006.

•  After a significant decrease from 2007 to 2008, the average acetaldehyde
   concentration began increasing. The average concentration for 2010 is the maximum
   average concentration shown across the years  of sampling. A significant decrease is
   shown from 2010 to 2011.

•  The minimum concentration measured is greater than zero for all years shown,
   indicating that there were no non-detects reported for acetaldehyde.


Observations from Figure 16-17 for arsenic include the following:

•  S4MO began sampling metals in July 2003. Because fewer than 85 percent of
   possible samples were collected in 2003, Figure 16-17 excludes data from 2003.

•  The maximum arsenic concentration was measured on December 26, 2007. Only five
   concentrations greater than  10 ng/m3 have been measured at  S4MO since 2004.

•  There is a slight downward trend in the average concentration of arsenic at S4MO,
   but the magnitude of the outliers measured in 2005, 2007, and 2009 makes this
   difficult to see. In addition,  confidence intervals calculated for the average
   concentrations are relatively large as  a result of the highest concentrations and
   indicate that these changes are not  statistically significant.

•  Many of the statistical parameters are at  a minimum for 2011.
                                  16-29

-------
•  Although difficult to discern in Figure 16-17, the minimum concentration measured is
   greater than zero for all years shown, indicating that there were no non-detects
   reported for arsenic.

Observations from Figure 16-18 for benzene measurements include the following:

•  Because VOC sampling did not begin until December 2002, 2002 data was excluded
   from this analysis.

•  All four benzene concentrations greater than 5 ug/m3 were measured in 2003.

•  The average concentrations exhibit an overall decreasing trend through 2007,
   representing a greater than 1 ug/m3 decrease, although the most significant changes
   occurred in the early years of sampling. The average concentration has varied
   between 0.80 ug/m3 (2011) and 1.03 ug/m3 (2010) since 2007.

•  The range of benzene measurements is smallest for 2011,  with a difference of
   approximately 1  ug/m3 between the minimum and maximum concentration measured.

•  The minimum concentration measured is greater than zero for each year, indicating
   that there were no non-detects reported for benzene since  2003.

Observations from Figure 16-19 for 1,3-butadiene include the following:

•  The maximum 1,3-butadiene concentration was measured at S4MO in 2003, although
   a similar concentration was also measured in 2008. These are the only two
   1,3-butadiene concentrations greater than 1.0 ug/m3 that have been measured at
   S4MO.

•  The minimum, 5th percentile, and median concentrations are all zero for 2003 and
   2004, indicating that at least 50 percent of the measurements were non-detects. The
   number of non-detects decreased after 2004, from a maximum of 66 percent in 2004
   to a minimum of zero percent in 2010.

•  Between 2004 and 2008, the average concentration changed very little,  ranging from
   0.079 ug/m3 (2005) to 0.095 ug/m3 (2006). Greater fluctuations are shown in the
   years that follow.

Observations from Figure 16-20 for formaldehyde include the following:

•  The maximum formaldehyde concentration (43.8 ug/m3) was measured in 2004 on
   the same day that the maximum acetaldehyde concentration was measured
   (August 31, 2004). This concentration is more than twice  the next highest
   concentration (17.8 ug/m3), which was measured in 2011. The six highest
   concentrations of formaldehyde were all measured in 2004 (2) or 2011 (4). Overall,
   21 measurements of formaldehyde greater than 10 ug/m3 have been measured at
   S4MO.
                                  16-30

-------
•  The average concentration has fluctuated over the years of sampling, ranging between
   2.46 |ig/m3 (2009) and 5.10 |ig/m3 (2004).

•  The average formaldehyde concentration increased significantly from 2010 to 2011.
   There were 11 concentrations of formaldehyde in 2011 that were greater than the
   maximum concentration for 2010.

•  The minimum concentration measured for each year is greater than zero, indicating
   that there were no non-detects of formaldehyde reported over the years shown.


Observations from Figure 16-21 for hexavalent chromium include the following:

•  The maximum hexavalent chromium concentration was measured on July 5, 2008
   (0.460 ng/m3), although a similar concentration was also measured on July 4, 2006
   (0.422 ng/m3). The third highest hexavalent chromium concentration was measured
   on July 4, 2010 (0.188 ng/m3), but is significantly less than the maximum
   concentrations measured in 2006 and 2008. While only 15 concentrations greater than
   0.1 ng/m3 have been measured at S4MO since the onset of sampling, at least one has
   been measured each year, with the exception of 2009.

•  After an initial increase from 2005 to 2006, the average and median concentrations
   exhibit a decreasing trend, with nearly all of the statistical parameters reaching a
   minimum in 2009. The average concentration increased from 2009 to 2010 and held
   steady for 2011.

•  With the exception of 2011, both the minimum concentration and 5th percentile for
   each year are zero, indicating the presence of non-detects (at least 5 percent). The
   percentage of non-detects has ranged from less than 2 percent (2011) to 43 percent
   (2009).


Observations from Figure 16-22 for lead include the following:

•  The maximum lead concentration was measured at S4MO in 2008 and was nearly
   100 ng/m3. The second highest concentration was also measured in 2008 (84.8 ng/m3)
   although a similar concentration was also measured in 2009.

•  Concentrations of lead measured at S4MO exhibit a relatively high level of
   variability, as illustrated by the minimum and maximum concentrations. The
   difference between these two statistical metrics has ranged from 30 ng/m3 (2011) to
   96 ng/m3 (2008).

•  The average concentration of lead at S4MO has fluctuated over the years and exhibits
   no real trend. The averages have ranged from 9.95 ng/m3 (2009) to 14.5 ng/m3
   (2006). The confidence intervals calculated for these averages are relatively large and
   therefore support the previous bullet. This site has had the highest annual average
   concentration of lead for the last several years compared to other NMP sites sampling
        metals under the NMP.
                                  16-31

-------
       •  The minimum concentration measured for each year is greater than zero, indicating
          that there were no non-detects of lead reported for the years shown.
       Observations from Figure 16-23 for manganese include the following:
       •  The maximum manganese concentration was measured on November 26, 2008
          (734 ng/m3) and is nearly twice the next highest concentration (395 ng/m3, measured
          in 2011). A similar concentration was also measured in 2004 (387 ng/m3).
       •  Four manganese concentrations greater than 100 ng/m3 have been measured at S4MO
          since 2004, each in a different year. For each of these years, the second highest
          concentration of manganese was an order of magnitude lower. For example, for 2011,
          the two maximum concentrations are 395 ng/m3 and 65.3 ng/m3.
       •  The average concentration of manganese has ranged from 12.5 ng/m3 (2007) to
          21.9 ng/m3 (2008). The median concentration, which is influenced less by outliers,
          has varied less, ranging from 6.82 ng/m3 (2009) to  11.6 ng/m3 (2006). The median
          concentration actually has a decreasing trend from  2006 to 2009, despite the outlier
          measured in 2008.
       •  The minimum concentration measured for each year is greater than zero, indicating
          that there were no non-detects of manganese reported for the years shown.

16.5   Additional Risk-Based Screening Evaluations
       The following risk-based screening evaluations were conducted to characterize risk at the
S4MO monitoring site. Refer to Sections 3.3 and 3.5.5 for  definitions and explanations regarding
the various toxicity factors, time frames, and calculations associated with these risk-based
screenings.
16.5.1  Risk-Based Screening Assessment Using MRLs
       A risk-based screening was conducted by comparing the concentration data from the
Missouri monitoring site to the ATSDRMRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks  and are defined for three exposure periods: acute
(exposures of 1 to 14 days); intermediate (exposures of 15 to 364 days); and chronic (exposures
of 1 year or greater). The preprocessed daily measurements of the pollutants of interest were
compared to the acute MRLs; the quarterly averages were compared to the intermediate MRLs;
and the annual averages were compared to the chronic MRLs.
                                         16-32

-------
          As discussed in Section 4.2.2, none of the measured detections or time-period average
   concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
   noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2011.

   16.5.2 Cancer Risk and Noncancer Hazard Approximations
          For the pollutants of interest for S4MO and where annual average concentrations could
   be calculated, risk was examined by calculating cancer risk and noncancer hazard
   approximations. These approximations can be used as risk estimates for cancer and noncancer
   effects attributable to the pollutants of interest. Although the use of these approximations is
   limited, they may help identify where policy-makers may want to shift or confirm their air-
   monitoring priorities. Refer to Section 3.5.5.2 for an explanation of how cancer risk and
   noncancer hazard approximations are calculated and what limitations are associated with them.
   Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
   approximations are presented in Table 16-6, where applicable. Cancer risk approximations are
   presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
   values.

                Table 16-6. Risk Approximations for the Missouri Monitoring Site
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
St. Louis, Missouri - S4MO
Acenaphthene3
Acetaldehyde
Arsenic (PM10)a
Benzene
Benzo(a)pyrenea
Beryllium (PM10)a
1,3 -Butadiene
Cadmium (PM10)a
0.000088
0.0000022
0.0043
0.0000078
0.00176
0.0024
0.00003
0.0018

0.009
0.000015
0.03
_
0.00002
0.002
0.00001
61/61
59/59
59/59
57/57
59/61
59/59
52/57
59/59
0.01
ฑ<0.01
2.75
ฑ0.43
<0.01
ฑ<0.01
0.80
ฑ0.06
<0.01
ฑ0.01
<0.01
ฑ<0.01
0.08
ฑ0.01
<0.01
ฑ<0.01
0.48
6.05
3.76
6.23
0.23
0.02
2.54
1.00

0.31
0.06
0.03
_
<0.01
0.04
0.06
— = a Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 16-5.
                                             16-33

-------
         Table 16-6. Risk Approximations for the Missouri Monitoring Site (Continued)
Pollutant
Carbon Tetrachloride
Chloroform
ฃ>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Fluorene3
Formaldehyde
Hexachloro- 1 ,3 -butadiene
Hexavalent Chromium3
Lead (PM10)a
Manganese (PM10)a
Naphthalene3
Nickel (PM10)a
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
Cancer
URE
(Hg/m3)1
0.000006

0.000011
0.000026
0.0000025
0.000088
0.000013
0.000022
0.012


0.000034
0.00048
0.00000026
0.0000048
0.0000088
Noncancer
RfC
(mg/m3)
0.1
0.098
0.8
2.4
1

0.0098
0.09
0.0001
0.00015
0.00005
0.003
0.00009
0.04
0.002
0.1
#of
Measured
Detections
vs. # of
Samples
57/57
44/57
47/57
18/57
57/57
61/61
59/59
11/57
60/61
59/59
59/59
61/61
59/59
55/57
25/57
8/57
Annual
Average
(Hg/m3)
0.61
ฑ0.04
0.35
ฑ0.40
0.21
ฑ0.11
0.03
ฑ0.01
0.38
ฑ0.04
0.01
ฑ<0.01
4.25
ฑ0.92
0.02
ฑ0.01
0.01
ฑ0.01
0.01
ฑO.01
0.02
ฑ0.01
0.08
ฑ0.01
0.01
ฑ0.01
0.18
ฑ0.04
0.03
ฑ0.01
O.01
ฑO.01
Cancer Risk
Approximation
(in-a-million)
3.68

2.35
0.84
0.94
0.53
55.21
0.50
0.40


2.85
0.58
0.05
0.17
0.03
Noncancer
Hazard
Approximation
(HQ)
0.01
0.01
O.01
0.01
O.01

0.43
O.01
0.01
0.07
0.37
0.03
0.01
O.01
0.02
O.01
— = a Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 16-5.

          Observations for S4MO from Table 16-6 include the following:

          •   The pollutants with the highest annual average concentrations for S4MO are
              formaldehyde, acetaldehyde, and benzene.

          •   Formaldehyde, benzene, and acetaldehyde have the highest cancer risk
              approximations for S4MO. The cancer risk approximation for formaldehyde
              (55.21 in-a-million) is nearly nine times higher than the cancer risk approximations
              for benzene and acetaldehyde, which were both approximately 6 in-a-million.
              S4MO's cancer risk approximation for formaldehyde is the second highest cancer risk
              approximation calculated among the site-specific pollutants of interest across the
              program.
                                             16-34

-------
       •  Naphthalene has the highest cancer risk approximation among the PAHs
          (2.58 in-a-million) and arsenic has the highest cancer risk approximation among the
          metals (3.76 in-a-million).
       •  None of the pollutants of interest for S4MO have noncancer hazard approximations
          greater than 1.0, indicating that no adverse health effects are expected from these
          individual pollutants. The pollutant with the highest noncancer hazard approximation
          is formaldehyde (0.43), which is the fourth highest noncancer hazard approximation
          calculated for a site-specific pollutant interest among NMP sites.

16.5.3 Risk-Based Emissions Assessment
       In addition to the risk-based screenings discussed above, Tables 16-7 and 16-8 present an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 16-7 presents the 10 pollutants with the highest emissions from the 2008 NEI, the
10 pollutants with the highest toxicity-weighted emissions, and the 10 pollutants with the highest
cancer risk approximations (in-a-million), as  calculated from the annual averages provided in
Table 16-6. Table 16-8 presents similar information, but identifies the 10 pollutants with the
highest noncancer hazard approximations, also calculated from annual averages provided in
Table 16-6.

       The pollutants listed in Tables 16-7 and 16-8 are limited to those that have cancer and
noncancer toxicity factors, respectively. As a result, although the actual value of the emissions is
the same, the highest emitted pollutants in the cancer table may be different from the noncancer
table. The cancer risk and noncancer hazard approximations based on the site's annual averages
are limited to those pollutants for which each respective site sampled. As discussed in
Section 16.3, S4MO sampled for VOCs, PAHs, carbonyl compounds, metals (PMi0), and
hexavalent chromium. In addition, the cancer risk and noncancer hazard approximations are
limited to those pollutants with enough data to meet the criteria for annual averages to be
calculated. A more in-depth discussion of this analysis is provided in Section 3.5.5.3. Similar to
the cancer risk and noncancer hazard approximations, this analysis may help policy-makers
prioritize their air monitoring activities.
                                          16-35

-------
  Table 16-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
                                                 the Missouri Monitoring Site
Oi
OJ
Oi
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
St. Louis, Missouri (St. Louis City) - S4MO
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Trichloroethylene
Naphthalene
Dichloromethane
POM, Group 2b
Methyl tert butyl ether
148.92
118.03
78.65
74.93
20.98
15.79
13.16
7.23
2.53
0.86
Hexavalent Chromium, PM
Formaldehyde
Arsenic, PM
Benzene
1,3 -Butadiene
Naphthalene
Nickel, PM
POM, Group 3
POM, Group 2b
Ethylbenzene
2.11E-03
1.53E-03
1.49E-03
1.16E-03
6.29E-04
4.47E-04
3.92E-04
3.24E-04
2.22E-04
1.97E-04
Formaldehyde
Benzene
Acetaldehyde
Arsenic
Carbon Tetrachloride
Naphthalene
1,3 -Butadiene
ฃ>-Dichlorobenzene
Cadmium
Ethylbenzene
55.21
6.23
6.05
3.76
3.68
2.85
2.54
2.35
1.00
0.94

-------
Table 16-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
                                          RfCs for the Missouri Monitoring Site
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer Hazard
Approximation
(HQ)
St. Louis, Missouri (St. Louis City) - S4MO
Toluene
Xylenes
Methanol
Benzene
Formaldehyde
Hexane
Hydrochloric acid
Ethylbenzene
Acetaldehyde
Methyl isobutyl ketone
522.43
358.32
289.83
148.92
118.03
116.82
109.87
78.65
74.93
73.11
Acrolein
Manganese, PM
Arsenic, PM
Formaldehyde
1,3 -Butadiene
Chlorine
Nickel, PM
Acetaldehyde
Trichloroethylene
Lead, PM
316,721.93
34,593.25
23,029.98
12,043.75
10,489.20
9,452.80
9,073.56
8,325.31
7,895.27
6,587.56
Formaldehyde
Manganese
Acetaldehyde
Lead
Arsenic
Cadmium
1,3 -Butadiene
Naphthalene
Benzene
Trichloroethylene
0.43
0.37
0.31
0.07
0.06
0.06
0.04
0.03
0.03
0.02

-------
Observations from Table 16-7 include the following:

•  Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
   cancer UREs in the city of St. Louis.

•  The pollutants with the highest toxi city-weighted emissions (of the pollutants with
   cancer UREs) are hexavalent chromium, formaldehyde, and arsenic.

•  Six of the highest emitted pollutants also have the highest toxi city-weighted
   emissions.

•  Five of the pollutants with the  highest cancer risk approximations for S4MO also
   appear on both emissions-based lists (formaldehyde, benzene, 1,3-butadiene,
   ethylbenzene, and naphthalene). Other pollutants with the highest cancer risk
   approximations for S4MO appear on one emissions-based list but not the other. While
   arsenic is not one of the highest emitted pollutants, it ranks third for its toxi city-
   weighted emissions. While acetaldehyde does not appear on the list of highest
   toxicity-weighted emissions, it is the fourth highest emitted pollutant in the city of
   St. Louis.

•  POM, Group 2b is the ninth highest emitted "pollutant" in St. Louis and ranks ninth
   for toxicity-weighted emissions. POM, Group 2b includes several PAHs sampled for
   at S4MO including acenaphthene  and fluorene, which are pollutants of interest for
   S4MO. These pollutants are not among those with the highest cancer risk
   approximations for S4MO.


Observations from Table 16-8 include the following:

•  Toluene, xylenes, and methanol are the highest emitted pollutants with noncancer
   RfCs in the city of St. Louis.

•  The pollutants with the highest toxicity-weighted emissions (of the pollutants with
   noncancer RfCs) are acrolein, manganese,  and arsenic. Although acrolein was
   sampled for at S4MO, this pollutant was excluded from the pollutants  of interest
   designation, and thus subsequent risk-based screening evaluations, due to questions
   about the consistency and reliability of the measurements, as discussed in Section 3.2.

•  Only two of the highest emitted pollutants  in the city of St.  Louis also  have the
   highest toxicity-weighted emissions (formaldehyde and acetaldehyde).

•  Formaldehyde and acetaldehyde are the pollutants with the highest and third highest
   noncancer hazard approximations for S4MO, respectively,  and are the only two
   pollutants of interest to appear on both emissions-based lists. Manganese, the
   pollutant with the second highest noncancer hazard approximation, is the pollutant
   with the  second highest toxicity-weighted emissions but is not one of the highest
   emitted.
                                   16-38

-------
16.6   Summary of the 2011 Monitoring Data for S4MO

       Results from several of the data treatments described in this section include the
following:

       ปซป  Twenty-five pollutants, of which 13 are NA TTS MQO Core Analytes, failed screens
          for S4MO.

       ปซป  Formaldehyde and acetaldehyde have the highest annual average concentrations for
          S4MO. These are the only pollutants of interest with annual averages greater than
          1 jug/m3.

       ปซป  S4MO had the highest annual average concentrations ofhexachloro-1,3-butadiene,
          arsenic, cadmium, and lead among all NMP sites sampling these pollutants. S4MO
          had the second highest annual average concentrations of 1,2-dichloroethane,
          p-dichlorobenzene, formaldehyde, and manganese among all NMP sites sampling
          these pollutants.

       ปซป  The trends analysis shows that concentrations of benzene have been decreasing at
          S4MO, particularly the earlier years of sampling.
                                         16-39

-------
17.0   Sites in New Jersey
       This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at UATMP and CSATAM sites in New Jersey, and integrates these
concentrations with emissions, meteorological, and risk information. Data generated by sources
other than ERG are not included in the data analyses contained in this report. Readers are
encouraged to refer to Sections 1 through 4 and the glossary (Appendix P) for detailed
discussions and definitions regarding the various data analyses presented below.

17.1   Site Characterization
       This section characterizes the New Jersey monitoring sites by providing geographical and
physical information about the location of the sites and the surrounding areas.  This information
is provided to give the reader insight regarding factors that may influence the air quality near the
sites and assist in the interpretation of the ambient monitoring data.

       The New Jersey sites are all located within the New York-Northern New Jersey-Long
Island, NY-NJ-PA MSA, although within different divisions. Figure 17-1 is a  composite satellite
image retrieved from ArcGIS Explorer showing the CFINJ monitoring site in its urban location.
Figure 17-2 identifies nearby point source emissions locations by source category, as reported in
the 2008 NEI for point sources. Figures 17-3 through 17-7 are the composite satellite maps and
emissions source maps for the remaining New Jersey monitoring sites. Note that only sources
within 10 miles of the sites are included in the facility counts provided in Figures 17-2, 17-5, and
17-7. A 10-mile boundary was chosen to give the reader an indication of which emissions
sources and emissions source categories could potentially have a direct effect on the air quality at
the monitoring sites. Further, this boundary provides both the proximity of emissions sources to
the monitoring sites as well as the quantity of such sources within a given distance of the sites.
Sources outside the 10-mile radii are still visible on the maps, but have been grayed out in order
to  show emissions sources just outside the boundary. Table 17-1 provides supplemental
geographical  information such as land use, location setting, and locational coordinates.
                                           17-1

-------
                                      Figure 17-1. Chester, New Jersey (CHNJ) Monitoring Site
to

-------
 Figure 17-2. NEI Point Sources Located Within 10 Miles of CHNJ
                           74'50'Q"W         74'45trW         7JU40B"W         74'35'CTW
                                         Mole: Due to facility density and collocation, the total facilities
                                         displayed may not represent all facilities within the area of interest.
Legend
 @   CHNJ UATMP site         10 mile radius |      | County boundary

Source Category Group (No. of Facilities)    F   Food Processing/Agriculture (1)
  -f   Aircraft Operations (12)                      (B   Hospital (1)
  i   Asphalt Processing/Roofing Manufacturing (1)   .-;   Mine/Quarry (1)
  o   Clay Ceramics Manufacturing (3)              M   Miscellaneous Manufacturing (2)
  •   Concrete Batch Plant (1)                    <=>   Pharmaceutical Manufacturing (1)
  ฉ   Fabricated Metal Products (1)                 H   Pulp and Paper Plant/Wood Products (2)
                                        17-3

-------
Figure 17-3. Elizabeth, New Jersey (ELNJ) Monitoring Site

-------
Figure 17-4. New Brunswick, New Jersey (NBNJ) Monitoring Site

-------
Figure  17-5. NEI  Point Sources  Located Within 10 Miles of ELNJ and  NBNJ
           7445-0-W     7440TO-W     7< 35TrW     74 3ffO"W      74'25'0-W      7a'20'0-W     74-151TW     74'M'ITW      74 SKTW
                                                                 Note: Due to facility density and collocation, the total facilities
                                                                 displayed may not represent all facilities within the area of interest.
       ELNJ UATMP site
                                       NBNJ  UATMP site
                                                                         10 mile radius         I County boundary
Source Category Group (No. of Facilities)
 ^   Aircraft Operations (44)
  I   Asphalt Pfocessing/Roofirm Manufacturing (3)
  Q   Auto Body ShoprPanters (1)
  ft   Bakery (2)
  Y   Brewery/Distillery Winery (1)
 f   BuiklirtQ Construction (\)
  B   Bulk Terminate/Bulk Plants < 13)
 C   Chemical Msnufaclunng (24}
 O   Clay Ceramics Manufacturing <3t
  ฃ   Electricity Generation via Combustion (21)
  E   Elec&oplalmg, Plating. Polishing. Anodizing, and Coloring <2)
 <•>   Fabricated Metal Products (12)
 Ci>   R&xibte Potyurcthane Foam Production (2)
  F   Food ProcessmoyAgncullurelS)
 [~j   Furniture Plant (2)
                                                         Gasoline/Diesel Service Station {1>
                                                         Gypsum Manufacturing |1 )
                                                         Heating Equipment Manufacturing ^ 1 >
                                                         Hot Mix Asphalt Plant (2)
                                                         indusErtat Machinery and Equipment (S)
                                                         insbliiiionai - school (14)
                                                         Laboratory (1)
                                                         Landfill (?)
                                                         Lumbertsawmilio}
                                                         Marine Port (1)
                                                         iWarine Vessel Loading Rack (1)
                                                         Mine/Quarry (1)
                                                         Miscellaneous Commercial/Industrial (24)
                                                         Miscellaneous Man ufactu n nri { 1 3)
                                                         Municipal Wisle Combustor (2)
•  Oil and/of Gas Production {1)
A  Petrateum Refinery (3)
._  Pharmaceutical Manufacturing (7)
 1  Pnmary Metal Production (1)
^  Printing. Coattng fi Dyeing of Fabrics (t)
P  Pnnting/Publi&riing (10)
(H  Pulp and Paper PlantMtood Products (13)
R  Rubber and Miscellaneous Plastics Products (12J
2  Secondary Metal Processing (1)
>  Solid Waste Disposal - Commercial/institutional (4)
V  Steel Mm (2)
S  Surface Coating (16)
ฉ  Tire Manufacture (1)
^  Transportation and Marketing of Petroleum Products (2)
I  Wastewaler Treatment {Bt
W  Woodwork. Furniture. Millwork & Wood Preserving (2)
                                                                17-6

-------
Figure 17-6. Paterson, New Jersey (PANJ) Monitoring Site

-------
Figure 17-7. NEI  Point  Sources Located Within  10 Miles  of PANJ
          Legend
74--1 STO'W           74' 10'0"W           74" 5'0"W           74'0'0'W
 Note: Due to facility density and collocation, the total facilities
 displayed may not represent all facilities within the area of interest.
                   PANJ CSATAM site
10 mile radius
County boundary
          Source Category Group (No. of Facilities)              ?
           •4<  Aircraft Operations (21)                              M
           B  Bakery (1)                                        •
           B  Bulk Terminals/Bulk Plants (1)
           C  Chemical Manufacturing (8)                           1
           O  Clay Ceramics Manufacturing (1)                       -t
           6  Electrical Equipment (4)                              P
           f  Electricity Generation via Combustion (4)                 ffi
           E  Electroplating. Plating. Polishing, Anodizing, & Coloring (i)    R
           ฉ  Fabricated Metal Products <3)                         <
           ^  Flexible Polyurethane Foam Production (3)              V
           F  Food Processing/Agriculture (3)                        S
           GJ  Hospital (3)                                       T
           $  Hot Mix Asphalt Plant (3)                            —
           •$•  industrial Machinery and Equipment (2)                  '
           ^  Institutional - school (2)                             W
              Miscellaneous CornmerciaVlrtdustrial (3)
              Miscellaneous Manufacturing (20)
              Oil and/or Gas Production (1)
              Pharmaceutical Manufacturing {4)
              Primary Metal Production (3)
              Printing, Coaling & Dyeing of Fabric f 1)
              Prinling/Publishirtg (t3)
              Pulp and Pa pe r PlanLWood Products (10)
              Rubber and Miscellaneous Plastics Products (4)
              Site Remediation Activity (1)
              Steel Mill {1)
              Surface Coating (S}
              Textile Mill (4)
              Transportation Equipment (1)
              Wastewater Treatment (1)
              Woodwork, Furniture, MiHwork & Wood Preserving (1J
                                                    17-8

-------
                             Table 17-1. Geographical Information for the New Jersey Monitoring Sites

Site
Code

CHNJ

ELNJ

NBNJ

PANJ


AQS Code

34-027-3001

34-039-0004

34-023-0006

34-031-0005


Location

Chester

Elizabeth

New
Brunswick

Paterson


County

Morris

Union

Middlesex

Passaic


Micro- or Metropolitan
Statistical Area
New York-Northern
New Jersey -Long Island,
NY-NJ-PAMSA
(Newark Div)
New York-Northern
New Jersey -Long Island,
NY-NJ-PAMSA
(Newark Div)
New York-Northern
New Jersey -Long Island,
NY-NJ-PA MSA
(Edison Div)
New York-Northern
New Jersey -Long Island,
NY-NJ-PAMSA
(New York Div)
Latitude
and
Longitude

40.78763,
-74.6763

40.64144,
-74.20836

40.472786,
-74.42251

40.918381,
-74.168092


Land Use

Agricultural

Industrial

Agricultural

Commercial


Location
Setting

Rural

Suburban

Rural

Urban/City
Center


Additional Ambient Monitoring
Information1

SO2, NO, NO2, O3, Meteorological
parameters, PM2 5, PM2.5 Speciation

CO, SO2, NO2, NOX, Meteorological
parameters, PM2 5, PM2 5 Speciation

Meteorological parameters, PM2 5, PM2 5
Speciation

Meteorological Parameters, PM2 5

:Data for additional pollutants are reported to AQS for these sites (EPA, 2012c); however, these data are not generated by ERG and are therefore not included in this
 report.

-------
       CHNJ is located in northern New Jersey, in the town of Chester, west of the New York
City metropolitan area. Figure 17-1 shows that CHNJ is located in an open area near Building 1
on the property of Bell Labs, which is owned by Alcatel-Lucent. The surrounding area is rural
and agricultural with a rolling topography, but surrounded by small neighborhoods. Although the
location is considered part of the New York City MSA, the site's location is outside most of the
urbanized areas. Figure 17-2 shows that few sources are close to CFINJ and that the source
category with the greatest number of emissions  sources surrounding CFINJ is the aircraft
operations category, which includes airports as well as small runways, heliports, or landing pads.
The source closest to CFINJ is in the pulp and paper/wood products source category.

       ELNJ is located in the city of Elizabeth,  which lies just south of Newark and west of
Newark Bay and Staten Island, New York. As Figure 17-3 shows, the monitoring site is located
just off Exit 13  of the New Jersey Turnpike (1-95), near the toll plaza. Interstate-278 intersects
the Turnpike here as well. The surrounding area is highly industrialized, with an oil refinery
located just southwest of the site. Additional industry is located to the southwest, west, and east,
while residential neighborhoods are located to the northwest and north of the site.

       NBNJ is located in New  Brunswick, less than 20 miles southwest of Elizabeth. The
monitoring site is located on the property of Rutgers University's Cook-Douglass campus, on a
horticultural farm. The surrounding area is agricultural and rural, although residential
neighborhoods are located to the east, across a branch of the Raritan River, as shown in
Figure 17-4. County Road 617 (Ryders Lane) and US-1 intersect just west of the  site and 1-95
runs northeast-southwest about 1 mile east of the site, part of which can be seen in the lower
right hand corner of Figure 17-4.

       Figure 17-5 shows that the outer portions of the 10-mile radii for ELNJ and NBNJ
intersect and that many emissions sources surround these two sites. The bulk of the emissions
sources are located in northern Middlesex County and northeastward toward New York City and
northern New Jersey. The source categories with the greatest number of emissions sources in the
vicinity of these sites include aircraft operations, chemical manufacturing, electricity generation
via combustion, and surface coating. The emissions sources in closest proximity to the ELNJ
monitoring site are in the miscellaneous manufacturing, wastewater treatment, chemical
manufacturing,  electricity generation via combustion, and petroleum refining categories. The
                                         17-10

-------
emissions sources in closest proximity to the NBNJ monitoring site are involved in aircraft
operations and pharmaceutical manufacturing.

       PANJ is located in northern New Jersey, in the town of Paterson, north of Newark and
between Clifton and Hackensack. The monitoring site is located at the local health department
with residential areas to the east and commercial areas to the west, as shown in Figure 17-6. The
Passaic River runs northeast-southwest just north of PANJ and is shown in the upper left corner
of Figure  17-6. Interstate-80 runs east-west less than 1 mile south of PANJ. Figure  17-7 shows
that the majority of point sources within 10 miles of PANJ are located to the southeast of the site.
Many of the point sources near PANJ are involved in aircraft operations, printing and publishing,
or pulp and paper products, although the source closest to PANJ falls in the miscellaneous
industries category.

       Table 17-2 presents additional  site-characterizing information, including indicators of
mobile source activity, for the New Jersey  monitoring sites.  Table  17-2 includes a county-level
population for each site. County-level vehicle registration data for Union, Morris, Passaic, and
Middlesex Counties were not available from the State of New Jersey. Thus, state-level vehicle
registration, which was obtained from  the Federal Highway Administration, was allocated to the
county level using the county-level proportion of the state population from the U.S. Census
Bureau. Table 17-2 also includes a county-level vehicle registration-to-population ratio, which
was calculated to represent the number of vehicles per person within each monitoring site's
residing county. In addition, the population within 10 miles of each site is presented, based on
postal code population data estimates.  An estimate of 10-mile vehicle ownership was then
determined by applying the county-level vehicle registration-to-population ratio to the 10-mile
population surrounding each monitoring site. Table 17-2 also contains traffic volume information
for each site. Finally, Table 17-2 presents the county-level daily VMT for Middlesex, Morris,
Passaic, and Union Counties.
                                          17-11

-------
     Table 17-2. Population, Motor Vehicle, and Traffic Information for the New Jersey
                                     Monitoring Sites
Site
CHNJ
ELNJ
NBNJ
PANJ
Estimated
County
Population1
494,976
539,494
814,217
502,007
County-level
Vehicle
Registration2
389,359
424,894
640,893
396,602
Vehicles per
Person
(Registration:
Population)
0.79
0.79
0.79
0.79
Population
within 10
miles3
237,740
2,189,758
810,434
1,356,675
Estimated
10-mile
Vehicle
Ownership
187,012
1,724,607
637,915
1,071,818
Annual
Average
Daily
Traffic4
12,917
250,000
114,322
22,272
County-
level Daily
VMT5
14,256,044
12,485,902
20,415,685
8,178,167
Bounty-level population estimates reflect 2011 data from the U.S. Census Bureau (Census Bureau, 2012b)
2County-level vehicle registration reflects ratios based on 2010 state-level vehicle registration data from the FHWA
 and the 2010 county-level proportion of the state population data (FHWA, 2011 and Census Bureau, 2011)
310-mile population estimates reflect 2011 data from Xionetic (Xionetic, 2011)
4AADT for ELNJ reflects 2006 data from NJ Department of Treasury and 2010 data from the New Jersey DOT for
 the other sites (Steer, 2008 and NJ DOT, 2010a)
5County-level VMT reflects 2010 data from the New Jersey DOT (NJ DOT, 2010b)
        Observations from Table 17-2 include the following:

        •  Middlesex County, where NBNJ is located, has the highest county-level population of
           the New Jersey sites while Morris County, where CHNJ is located, has the least.
           ELNJ has the highest 10-mile population among the four New Jersey sites while
           CHNJ has the least. The 10-mile populations for both ELNJ and PANJ are greater
           than 1 million people.

        •  Compared to NMP monitoring sites in other locations, the county-level populations
           are in the middle of the range. However, ELNJ has one of the highest 10-mile
           populations, ranking third among NMP sites. The 10-mile populations for the other
           New Jersey sites range from eight highest (PANJ) to 36th highest (CHNJ).

        •  The estimated county-level vehicle registration is highest for NBNJ while the vehicle
           ownerships across the remaining New Jersey sites are fairly similar to each other. The
           county-level registration estimates for all the sites are in the middle of the range
           compared to other NMP sites. ELNJ and PANJ have the highest 10-mile vehicle
           ownership estimates compared to the other New Jersey sites and rank second and
           sixth compared to other NMP sites.

        •  ELNJ and NBNJ experience a significantly higher average traffic volume than CHNJ
           and PANJ. Traffic data for ELNJ are provided for 1-95, between Exit 13 and 13 A;
           this is the highest traffic volume among all NMP sites. Traffic data for CHNJ are
           provided for Main Street (County Road 513) near Highway 206 in downtown
           Chester; traffic data for NBNJ are provided for US-1 near State Road 617 (Ryders
           Lane); and traffic data for PANJ are provided for Memorial Drive between Ellison
           Street and College Boulevard.
                                           17-12

-------
       •  Among the New Jersey counties with monitoring sites, VMT for Middlesex County is
          highest while VMT for Passaic County is the lowest. However, county-level VMT for
          the New Jersey counties are in the middle of the range compared to other counties
          with NMP sites (where VMT data were available).

17.2   Meteorological Characterization
       The following sections characterize the meteorological conditions near the monitoring
sites in New Jersey on sample days, as well as over the course of the year.

17.2.1  Climate Summary
       Frontal systems push across the state of New Jersey regularly, producing variable
weather conditions. The state's proximity to the Atlantic Ocean has a moderating effect on
temperature. Summers along the coast tend to be cooler than areas farther inland, while winters
tend to be warmer. Large urban areas within the state experience the urban heat island effect, in
which urban areas retain more heat than outlying areas. New Jersey's mid-Atlantic location also
allows for ample annual precipitation  and relatively high humidity. A southwesterly wind is most
common in the summer and a northwesterly wind is typical in the winter. Winds from the west
and northwest result in air masses that dry out,  stabilize, and warm as they move eastward from
higher elevations to sea level (Bair, 1992 and Rutgers, 2013).

17.2.2  Meteorological Conditions in 2011
       Hourly meteorological data from NWS  weather stations nearest these sites were retrieved
for 2011 (NCDC, 2011). The three closest weather stations are located at Somerville-Somerset
Airport (near CHNJ and NBNJ), Newark International Airport (near ELNJ), and Essex County
Airport (near PANJ), WBAN 54785, 14734, and 54743, respectively. Additional information
about these weather stations, such as the distance between the sites and the weather stations, is
provided in Table 17-3. These data were used to determine how meteorological conditions on
sample days vary from conditions experienced  throughout the year.
                                         17-13

-------
                  Table 17-3. Average Meteorological Conditions near the New Jersey Monitoring Sites
Closest NWS Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(ฐF)
Average
Temperature
(ฐF)
Average
Dew Point
Temperature
(ฐF)
Average
Wet Bulb
Temperature
(ฐF)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Chester, New Jersey - CHNJ
Somerville, New
Jersey/Somerset
Airport
54785
(40.62, -74.67)
11.30
miles
165ฐ
(SSE)
Sample
Day
2011
63.3
ฑ4.6
63.8
ฑ1.9
53.2
ฑ4.3
53.3
ฑ1.8
42.2
ฑ4.7
43.0
ฑ2.0
48.0
ฑ4.1
48.4
ฑ1.7
69.8
ฑ3.5
71.7
ฑ1.3
1015.8
ฑ1.8
1015.4
ฑ0.8
3.4
ฑ0.6
2.9
ฑ0.2
Elizabeth, New Jersey - ELNJ
Newark International
Airport
14734
(40.68, -74.17)
3.45
miles
20ฐ
(NNE)
Sample
Day
2011
64.3
ฑ4.6
65.2
ฑ 1.9
56.6
ฑ4.3
57.4
ฑ1.8
41.9
ฑ4.7
42.9
ฑ 1.9
49.6
ฑ3.9
50.3
ฑ1.6
61.0
ฑ3.8
61.6
ฑ 1.5
1015.8
ฑ1.9
1015.6
ฑ0.8
8.0
ฑ0.9
7.6
ฑ0.3
New Brunswick, New Jersey - NBNJ
Somerville, New
Jersey/Somerset
Airport
54785
(40.62, -74.67)
16.06
miles
297ฐ
(WNW)
Sample
Day
2011
63.0
ฑ4.3
63.8
ฑ1.9
53.1
ฑ4.1
53.3
ฑ1.8
42.4
ฑ4.5
43.0
ฑ2.0
48.1
ฑ3.9
48.4
ฑ1.7
70.7
ฑ3.4
71.7
ฑ1.3
1015.6
ฑ 1.8
1015.4
ฑ0.8
3.4
ฑ0.6
2.9
ฑ0.2
Paterson, New Jersey - PANJ
Essex County
Airport
54743
(40.88, -74.28)
6.39
miles
229ฐ
(SW)
Sample
Day
2011
48.1
ฑ8.2
62.9
ฑ1.9
39.8
ฑ7.8
53.9
ฑ 1.8
26.3
ฑ8.2
42.8
ฑ2.0
34.7
ฑ7.2
48.6
ฑ 1.7
61.7
ฑ8.2
69.5
ฑ1.5
1017.5
ฑ5.7
1016.3
ฑ0.8
5.1
ฑ 1.3
3.5
ฑ0.2
1 Sample day averages are highlighted to help differentiate the sample day averages from the full-year averages.

-------
       Table 17-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for the entire year for 2011. Also included in Table 17-3 is the
95 percent confidence interval for each parameter. As shown in Table 17-3, average
meteorological conditions on sample days were representative of average weather conditions
throughout the year for CHNJ, ELNJ, and NBNJ. It appears that sample days at PANJ were
cooler, drier, and windier than for the entire year as a whole. However, a 1-year monitoring
effort at PANJ was completed in May 2011, thereby missing the warmer months of the  year.

17.2.3 Back Trajectory Analysis
       Figure 17-8 is the composite back trajectory map for days on which samples were
collected at the CHNJ monitoring site in 2011. Included in Figure 17-8 are four back trajectories
per sample day. Figure 17-9 is the corresponding cluster analysis.  Similarly, Figures 17-10
through 17-13 are the composite back trajectory maps and corresponding cluster analyses for
ELNJ and NBNJ. Figure 17-14 is the composite back trajectory map for PANJ but a cluster
analysis was not performed for this site because there were fewer than 30 sample days.  An in-
depth description of these maps and how they were  generated is presented in  Section 3.5.2.1. For
the composite maps, each line represents the 24-hour trajectory along which a parcel of air
traveled toward the monitoring site on a given sample day and time, based on an initial  height of
50 meters AGL. For the cluster analyses, each line corresponds to a trajectory representative of a
given cluster of back trajectories. Each concentric circle around the sites in Figures 17-8 through
17-14 represents 100 miles.
                                         17-15

-------
Figure 17-8. 2011 Composite Back Trajectory Map for CHNJ
    Figure 17-9. Back Trajectory Cluster Map for CHNJ
                         17-16

-------
Figure 17-10. 2011 Composite Back Trajectory Map for ELNJ
    Figure 17-11. Back Trajectory Cluster Map for ELNJ
                                          "    •        \
                         17-17

-------
Figure 17-12. 2011 Composite Back Trajectory Map for NBNJ
    Figure 17-13. Back Trajectory Cluster Map for NBNJ
                          17-18

-------
       Figure 17-14. 2011 Composite Back Trajectory Map for PANJ
Observations from Figures 17-8 through 17-14 include the following:

•  Due to their relatively close proximity to each other and the standardization of sample
   days, the back trajectories shown on each composite back trajectory map for the New
   Jersey sites are similar to each other. The composite back trajectory map for PANJ
   includes one-third of the back trajectories compared to the other sites as PANJ
   stopped sampling in May 2011  and sampled on a l-in-12 day sachedule.

•  Back trajectories originated from a variety of directions at the sites. In general, the
   longest back trajectories originated from northwest of the monitoring sites.

•  The 24-hour air shed domains for the New Jersey sites were similar in size to each
   other. Back trajectories greater than 600 miles in length originated near Lake Superior
   and Ontario, Canada. The average trajectory length for these sites ranged from
   251 miles (CJTNJ) to 259 miles (NBNJ).
                                   17-19

-------
       •  The cluster trajectories for the New Jersey sites are similar to each other in
          geographical distribution, although the percentages vary. Each of the cluster maps has
          a relatively short cluster originating to the west of the sites over central Pennsylvania,
          representing approximately 30 percent of back trajectories. This cluster trajectory
          represents relatively short back trajectories (<250 miles) originating from a direction
          with a westerly component. Another roughly 20 percent of back trajectories
          originated from the west and northwest of the site but of longer length. For  CHNJ and
          ELNJ, these are split into two clusters, one from the west and one from the  northwest.
          Roughly 20 percent of back trajectories originated to the south, 20 percent to the east,
          and 10 percent to the north.
17.2.4 Wind Rose Comparison
       Hourly surface wind data from the NWS weather stations nearest the New Jersey sites, as
presented in Section 17.2.2, were uploaded into a wind rose software program to produce
customized wind roses, as described in Section 3.5.2.2. A wind rose shows the frequency of wind
directions using "petals" positioned around a 16-point compass, and uses different colors to
represent wind speeds.

       Figure 17-15 presents  a map showing the distance between the NWS station and CFtNJ,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 17-15 also presents three different wind roses for the
CFINJ monitoring site. First, a historical wind rose representing 2001 to 2010 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 2011 is presented.
Next, a wind rose representing wind data for days on which samples were collected in  2011 is
presented. These can be used to identify the predominant wind speed and direction for  2011 and
determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically. Figures 17-16 through 17-18 present the distance maps and
wind roses for ELNJ, NBNJ, and PANJ, respectively.
                                          17-20

-------
Figure 17-15. Wind Roses for the Summerville-Somerset Airport Weather Station near
                                    CHNJ
Distance between CHNJ and NWS Station
                               2001-2010 Historical Wind Rose
 NWS
Station
                           \     '+.
          2011 Wind Rose
                                   Sample Day Wind Rose
                                     17-21

-------
Figure 17-16. Wind Roses for the Newark International Airport Weather Station near
                                  ELNJ
Distance between ELNJ and NWS Station
         2001-2010 Historical Wind Rose
                     -*' i
    iff?         '^"
1
                             ~","J-
                                          /VEST
         2011 Wind Rose
             Sample Day Wind Rose
                                   17-22

-------
Figure 17-17. Wind Roses for the Summerville-Somerset Airport Weather Station near


                                    NBNJ
Distance between NBNJ and NWS Station
2001-2010 Historical Wind Rose
JEAOINOTON  ซ.ซ!ซป. B




           ' MANVJtftt ~>^_ ^


        ^Lg^^jpH*™;'?^*""***


 f*ซซ-ฃfc*^j T^v"'   '^      Edi?o(
               ?  EAST

         g MILL* TOME o**'1 L* TOMI
          2011 Wind Rose
    Sample Day Wind Rose
                                     17-23

-------
 Figure 17-18. Wind Roses for the Essex County Airport Weather Station near PANJ
Distance between PANJ and NWS Station
2001-2010 Historical Wind Rose
        N.
  ' ,-..-•ป  ••
                         •  T. -    7'
        -a~  /                   4
         2011 Wind Rose
    Sample Day Wind Rose
                                    17-24

-------
Observations from Figures 17-15 and 17-17 for CHNJ and NBNJ include the following:

•  The NWS weather station at Somerville/Somerset Airport is the closest weather
   station to both CHNJ and NBNJ. The Somerville/Somerset Airport weather station is
   located approximately 11.3  miles south-southeast of CHNJ and 16.1 miles
   west-northwest of NBNJ.

•  The wind data for the historical and full-year wind roses for CHNJ and NBNJ are
   identical because they are from the same weather station.

•  The historical wind roses for these sites show that calm winds accounted for greater
   than 40 percent of observations. For wind speeds greater than 2 knots, northerly
   winds were observed most frequently, accounting for nearly 10 percent of the
   observations, while winds from the southwest quadrant were rarely observed.

•  Calm winds account nearly  50 percent of the wind observations throughout 2011.
   Winds from the northwest quadrant, including northerly winds, account  for another
   one-quarter of wind observations throughout 2011.

•  Wind patterns on sample day wind roses resemble the full-year wind patterns, with
   even more wind observations from the northwest quadrant. The similarities in the
   wind patterns indicate that conditions on sample days were similar to conditions
   experienced near these sites over the course of 2011.

•  While the 2011 wind roses do exhibit the same prevalence for calm winds as the
   historical wind rose, they do not exhibit the same northerly predominance for wind
   speeds greater than 2 knots. Instead, there was an increase in winds from the
   northwest quadrant. A similar observation was made for 2009 in the 2008-2009 NMP
   report and in the 2010 NMP report.


Observations from Figure 17-16 for ELNJ include the following:

•  The Newark International Airport weather station is located approximately 3.5 miles
   north-northeast of ELNJ.

•  The historical wind rose shows that winds from a variety of directions were observed
   near ELNJ, although easterly winds and winds from the southeast quadrant were
   observed infrequently. Calm winds were observed for less than six percent of
   observations. The strongest winds were associated with westerly and northwesterly
   winds.

•  The wind patterns shown on the 2011 wind rose are similar to the historical wind
   patterns, although there are  some differences. For instance, the calm rate for 2011 is
   higher than the historical calm rate. Westerly and west-southwesterly winds were
   slightly less prominent in 2011 compared to historical observations.
                                   17-25

-------
       •  The sample day wind rose has a similar calm rate as the 2011 full-year wind rose, but
          there are a number of differences as well. Winds from the southwest quadrant were
          observed less while winds from the west-northwest and northwest were observed
          more frequently on sample days.


       Observations from Figure 17-18 for PANJ include the following:

       •  The Essex County Airport weather station is located approximately 6.4 miles
          southwest of PANJ.

       •  The historical wind rose shows that calm winds account for approximately one-third
          of the wind observations near PANJ. Winds from the western quadrants account for
          the majority of winds greater than 2 knots, particularly winds from the west-
          northwest and northwest. The strongest winds were associated with westerly to
          northwesterly winds.

       •  The 2011  wind rose shows that calm winds accounted for greater than 40 percent of
          wind observations in 2011 and that west-northwesterly to north-northwesterly winds
          account for the majority of wind observations greater than 2  knots. This represents a
          northward shift in the predominant wind direction compared to the historical wind
          rose.

       •  The sample day wind rose for PANJ exhibits several differences from the historical
          and full-year wind roses. The sample day wind rose has a lower percentage of calm
          winds. The percentage of northwesterly wind observations is double that shown in the
          full-year wind rose. There is also a higher percentage of winds from the east-northeast
          and fewer from the southwest quadrant. This wind rose likely reflects a seasonal
          pattern as it only includes sample days through May 2011 to correspond with the
          sample period.


17.3   Pollutants of Interest

       Site-specific "pollutants of interest" were determined for the New Jersey monitoring sites

in order to allow 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. In addition, if any of the NATTS MQO Core Analytes measured by each monitoring site

did not meet the pollutant of interest criteria based on the preliminary risk-based screening, that

pollutant was added to the list of site-specific pollutants of interest. A more in-depth description

of the risk-based screening process is presented in Section 3.2.
                                          17-26

-------
       Table 17-4 presents the results of the preliminary risk-based screening process for the
New Jersey sites. The pollutants that failed at least one screen and contributed to 95 percent of
the total failed screens for each monitoring site are shaded. NATTS MQO Core Analytes are
bolded. Thus, pollutants of interest are shaded and/or bolded. All three UATMP sites sampled
for VOCs and carbonyl compounds while PANJ sampled for VOCs only.
       Table 17-4. Risk-Based Screening Results for the New Jersey Monitoring Sites
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
Chester, New Jersey - CHNJ
Acetaldehyde
Benzene
Carbon Tetrachloride
Formaldehyde
Acrylonitrile
1,3-Butadiene
1,2-Dichloroethane
Hexachloro- 1 ,3 -butadiene
ฃ>-Dichlorobenzene
1 , 1 ,2,2-Tetrachloroethane
Chloromethylbenzene
1 ,2-Dibromoethane
Dichloromethane
Ethylbenzene
1 , 1 ,2-Trichloroethane
Trichloroethylene
Xylenes
0.45
0.13
0.17
0.077
0.015
0.03
0.038
0.045
0.091
0.017
0.02
0.0017
7.7
0.4
0.0625
0.2
10
Total
61
61
61
61
35
24
19
6
5
5
1
1
1
1
1
1
1
345
61
61
61
61
35
31
19
6
25
5
1
1
61
61
1
5
61
556
100.00
100.00
100.00
100.00
100.00
77.42
100.00
100.00
20.00
100.00
100.00
100.00
1.64
1.64
100.00
20.00
1.64
62.05
17.68
17.68
17.68
17.68
10.14
6.96
5.51
1.74
1.45
1.45
0.29
0.29
0.29
0.29
0.29
0.29
0.29
17.68
35.36
53.04
70.72
80.87
87.83
93.33
95.07
96.52
97.97
98.26
98.55
98.84
99.13
99.42
99.71
100.00

Elizabeth, New Jersey - ELNJ
Acetaldehyde
Benzene
Formaldehyde
Carbon Tetrachloride
1,3-Butadiene
Ethylbenzene
ฃ>-Dichlorobenzene
1,2-Dichloroethane
Acrylonitrile
Propionaldehyde
Hexachloro- 1 ,3 -butadiene
1 ,2-Dibromoethane
Trichloroethylene
0.45
0.13
0.077
0.17
0.03
0.4
0.091
0.038
0.015
0.8
0.045
0.0017
0.2
61
61
61
60
59
29
27
16
11
9
4
3
2
61
61
61
61
59
61
49
16
11
61
6
3
24
100.00
100.00
100.00
98.36
100.00
47.54
55.10
100.00
100.00
14.75
66.67
100.00
8.33
15.02
15.02
15.02
14.78
14.53
7.14
6.65
3.94
2.71
2.22
0.99
0.74
0.49
15.02
30.05
45.07
59.85
74.38
81.53
88.18
92.12
94.83
97.04
98.03
98.77
99.26
                                          17-27

-------
Table 17-4. Risk-Based Screening Results for the New Jersey Monitoring Sites (Continued)
Pollutant
Dichloromethane
1, 1,2,2-Tetrachloroethane
Xylenes
Screening
Value
(Ug/m3)
7.7
0.017
10
Total
#of
Failed
Screens
1
1
1
406
#of
Measured
Detections
61
1
61
657
%of
Screens
Failed
1.64
100.00
1.64
61.80
% of Total
Failures
0.25
0.25
0.25
Cumulative
%
Contribution
99.51
99.75
100.00

New Brunswick, New Jersey - NBNJ
Acetaldehyde
Formaldehyde
Benzene
Carbon Tetrachloride
1,3-Butadiene
Acrylonitrile
1 ,2-Dichloroethane
ฃ>-Dichlorobenzene
Hexachloro- 1 ,3 -butadiene
Ethylbenzene
1 , 1 ,2,2-Tetrachloroethane
1 ,2-Dibromoethane
Bromomethane
Chloromethylbenzene
Dichloromethane
Propionaldehyde
1 , 1 ,2-Trichloroethane
Xylenes
0.45
0.077
0.13
0.17
0.03
0.015
0.038
0.091
0.045
0.4
0.017
0.0017
0.5
0.02
7.7
0.8
0.0625
10
Total
62
62
58
57
39
25
19
15
8
7
4
2
1
1
1
1
1
1
364
62
62
58
58
42
25
19
39
9
58
4
2
37
1
58
62
1
58
655
100.00
100.00
100.00
98.28
92.86
100.00
100.00
38.46
88.89
12.07
100.00
100.00
2.70
100.00
1.72
1.61
100.00
1.72
55.57
17.03
17.03
15.93
15.66
10.71
6.87
5.22
4.12
2.20
1.92
1.10
0.55
0.27
0.27
0.27
0.27
0.27
0.27
17.03
34.07
50.00
65.66
76.37
83.24
88.46
92.58
94.78
96.70
97.80
98.35
98.63
98.90
99.18
99.45
99.73
100.00

Paterson, New Jersey - PANJ
Benzene
1,3-Butadiene
Carbon Tetrachloride
ฃ>-Dichlorobenzene
Ethylbenzene
1 ,2-Dibromoethane
Acrylonitrile
1 ,2-Dichloroethane
Hexachloro- 1 ,3 -butadiene
1 , 1 ,2,2-Tetrachloroethane
0.13
0.03
0.17
0.091
0.4
0.0017
0.015
0.038
0.045
0.017
Total
12
12
10
10
9
3
2
2
2
1
63
12
12
12
12
12
3
2
2
2
1
70
100.00
100.00
83.33
83.33
75.00
100.00
100.00
100.00
100.00
100.00
90.00
19.05
19.05
15.87
15.87
14.29
4.76
3.17
3.17
3.17
1.59
19.05
38.10
53.97
69.84
84.13
88.89
92.06
95.24
98.41
100.00

      Observations from Table 17-4 include the following:

      •   Seventeen pollutants failed at least one screen for CHNJ (including six NATTS MQO
          Core Analytes); 16 failed screens for ELNJ (including six NATTS MQO Core
          Analytes); 18 failed screens for NBNJ (including five NATTS MQO Core Analytes);
          and 10 failed screens for PANJ (including three NATTS MQO Core Analytes).
                                        17-28

-------
•  The risk-based screening process identified eight pollutants of interest for CHNJ (of
   which five are NATTS MQO Core Analytes). Trichloroethylene was added as a
   pollutant of interest because it is a NATTS MQO Core Analyte, even though it did
   not contribute to 95 percent of failed screens. Chloroform, tetrachloroethylene, and
   vinyl chloride were added as pollutants of interest because they are also NATTS
   MQO Core Analytes, even though they did not fail any screens. These three
   pollutants are not shown in Table 17-4 but are shown in subsequent tables in the
   sections that follow.

•  The risk-based screening process identified 10 pollutants of interest for ELNJ (of
   which five are NATTS MQO Core Analytes). Trichloroethylene was added as a
   pollutant of interest because it is a NATTS MQO Core Analyte, even though it did
   not contribute to 95 percent of failed screens. Chloroform, tetrachloroethylene, and
   vinyl chloride were also added because they are NATTS MQO Core Analytes, even
   though they did not fail any screens. These three pollutants are not shown in
   Table 17-4 but are shown in subsequent tables in the sections that follow.

•  The risk-based screening process identified 10 pollutants of interest for NBNJ (of
   which five are NATTS MQO Core Analytes). Chloroform, tetrachloroethylene,
   trichloroethylene, and vinyl chloride were added as pollutants of interest because they
   are also NATTS MQO Core Analytes, even though they did not fail any  screens.
   These four pollutants are not shown in Table 17-4 but are shown in subsequent tables
   in the sections that follow.

•  The risk-based screening process identified nine pollutants of interest for PANJ (of
   which three are NATTS MQO Core Analytes). Chloroform, tetrachloroethylene,
   trichloroethylene, and vinyl chloride were added as pollutants of interest because they
   are also NATTS MQO Core Analytes, even though they did not fail any  screens.
   These four pollutants are not shown in Table 17-4 but are shown in subsequent tables
   in the sections that follow.

•  For CHNJ, many of the pollutants failed a single screen. This may indicate that the
   concentrations on a single day may be the cause. A review of the data shows that
   many of these failed screens were for the same sample day. The concentrations of
   dichloromethane, ethylbenzene, trichloroethylene, and xylenes all failed  the screen
   for July 8, 2011. 1,2-Dibromoethane and 1,1,2-trichloroethane failed screen a single
   for May 27, 2011.

•  The concentrations of dichloromethane, trichloroethylene, and xylenes all failed the
   screen for July 8, 2011 for ELNJ, which is the same date as several of the pollutants
   of interest for CHNJ. 1,2-Dibromoethane also failed the screen for May 27, 2011 for
   ELNJ.

•  While there are no similarities in failed screens for NBNJ with CJrENJ or ELNJ, there
   are some common sample days among the pollutants failing relatively few  screens.
   Chioromethylbenzene, 1,2-dibromoethane, and 1,1,2-trichloroethane each failed the
   screen for November 2, 2011, as did 1,1,2,2-tetrachloroethane. Dichloromethane and
   xylenes each failed the screen for April 9, 2011.

                                   17-29

-------
       •  Many of the pollutants listed for PANJ failed screens on February 20, 2011;
          May 9, 2011; and May 15,2011.
17.4   Concentrations
       This section presents various concentration averages used to characterize pollution levels
at the New Jersey monitoring sites. Where applicable, the following calculations and data
analyses were performed: Time period-based concentration averages (quarterly and annual) are
provided for the pollutants of interest for the New Jersey monitoring sites, where the data meet
the applicable criteria. Concentration averages for select pollutants are also presented graphically
for the sites  to illustrate how the sites' concentrations compare to the program-level averages, as
presented in Section 4.1. In addition, concentration averages for select pollutants are presented
from previous years of sampling in order to characterize concentration trends at the sites.
Additional site-specific statistical summaries for the New Jersey sites are provided in
Appendices J and L.

17.4.1 2011 Concentration Averages
       Quarterly and annual concentration averages were calculated for the pollutants of interest
for each New Jersey site, as described in Section 3.1. The quarterly average of a particular
pollutant is simply the average concentration of the preprocessed daily measurements over a
given calendar quarter. Quarterly average concentrations include the substitution of zeros for all
non-detects. A site must have a minimum of 75  percent valid samples of the total number of
samples possible within a given quarter for a quarterly average to be calculated. An annual
average includes all measured detections and substituted zeros for non-detects for the entire year
of sampling. Annual averages were calculated for pollutants where three valid quarterly averages
could be calculated and where method completeness was greater than or equal to 85 percent, as
presented in Section 2.4. Quarterly and annual average concentrations for the New Jersey
monitoring sites are presented in Table 17-5, where applicable. Note that if a pollutant was not
detected in a given calendar quarter, the quarterly average simply reflects "0" because only zeros
substituted for non-detects were factored into the quarterly average concentration.
                                           17-30

-------
Table 17-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
                             the New Jersey Monitoring Sites
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Ug/m3)
2nd
Quarter
Average
(Ug/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Ug/m3)
Chester, New Jersey - CHNJ
Acetaldehyde
Acrylonitrile
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
1 ,2-Dichloroethane
Formaldehyde
Hexachloro- 1 ,3 -butadiene
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
61/61
35/61
61/61
31/61
61/61
50/61
19/61
61/61
6/61
40/61
5/61
5/61
1.15
ฑ0.21
0.08
ฑ0.03
0.60
ฑ0.08
0.02
ฑ0.01
0.55
ฑ0.04
0.08
ฑ0.01
0.01
ฑ0.01
0.75
ฑ0.14
0.02
ฑ0.02
0.08
ฑ0.01
0
0
1.82
ฑ0.49
0.12
ฑ0.05
0.44
ฑ0.09
0.01
ฑ0.01
0.62
ฑ0.07
0.11
ฑ0.03
0.03
ฑ0.03
2.87
ฑ1.48
0.02
ฑ0.03
0.10
ฑ0.06
0.01
ฑ0.01
<0.01
ฑ<0.01
1.72
ฑ0.43
0.19
ฑ0.17
0.50
ฑ0.22
0.03
ฑ0.03
0.68
ฑ0.08
0.13
ฑ0.11
0.05
ฑ0.07
4.16
ฑ1.73
0.01
ฑ0.01
0.12
ฑ0.13
0.17
ฑ0.34
<0.01
ฑ0.01
1.76
ฑ0.41
0.03
ฑ0.03
0.57
ฑ0.09
0.04
ฑ0.01
0.65
ฑ0.07
0.10
ฑ0.02
0.05
ฑ0.02
1.66
ฑ0.21
0
0.07
ฑ0.03
0.01
ฑ0.01
<0.01
ฑ<0.01
1.61
ฑ0.20
0.10
ฑ0.05
0.53
ฑ0.07
0.02
ฑ0.01
0.62
ฑ0.03
0.10
ฑ0.03
0.03
ฑ0.02
2.37
ฑ0.63
0.01
ฑ0.01
0.09
ฑ0.04
0.05
ฑ0.09
<0.01
ฑ<0.01
Elizabeth, New Jersey - ELNJ
Acetaldehyde
Acrylonitrile
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
ฃ>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
61/61
11/61
61/61
59/61
61/61
48/61
49/61
16/61
61/61
61/61
2.15
ฑ0.40
0.01
ฑ0.02
0.92
ฑ0.14
0.11
ฑ0.03
0.50
ฑ0.04
0.08
ฑ0.02
0.06
ฑ0.02
0.01
ฑ0.02
0.30
ฑ0.06
2.06
ฑ0.46
3.44
ฑ0.85
0.10
ฑ0.06
0.91
ฑ0.20
0.12
ฑ0.03
0.62
ฑ0.06
0.18
ฑ0.06
0.15
ฑ0.08
0.03
ฑ0.03
0.46
ฑ0.13
3.63
ฑ 1.06
4.11
ฑ0.78
0
1.18
ฑ0.41
0.14
ฑ0.03
0.67
ฑ0.04
0.18
ฑ0.09
0.11
ฑ0.05
0.03
ฑ0.04
0.82
ฑ0.62
4.70
ฑ0.91
3.28
ฑ0.85
0.04
ฑ0.08
1.09
ฑ0.28
0.17
ฑ0.04
0.66
ฑ0.10
0.12
ฑ0.04
0.08
ฑ0.03
0.04
ฑ0.03
0.45
ฑ0.12
3.41
ฑ0.62
3.24
ฑ0.39
0.04
ฑ0.03
1.03
ฑ0.13
0.14
ฑ0.02
0.61
ฑ0.04
0.14
ฑ0.03
0.10
ฑ0.02
0.03
ฑ0.01
0.51
ฑ0.16
3.45
ฑ0.44
     NA = Not available due to the criteria for calculating a quarterly and/or annual average.
                                          17-31

-------
Table 17-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
                      the New Jersey Monitoring Sites (Continued)
Pollutant
Propionaldehyde
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
#of
Measured
Detections
vs. # of
Samples
61/61
56/61
24/61
4/61
1st
Quarter
Average
(Ug/m3)
0.28
ฑ0.06
0.14
ฑ0.04
0.03
ฑ0.02
<0.01
ฑ<0.01
2nd
Quarter
Average
(Ug/m3)
0.51
ฑ0.14
0.25
ฑ0.09
0.05
ฑ0.04
<0.01
ฑ<0.01
3rd
Quarter
Average
(Ug/m3)
0.67
ฑ0.13
0.24
ฑ0.16
0.27
ฑ0.39
O.01
ฑO.01
4th
Quarter
Average
(Ug/m3)
0.37
ฑ0.09
0.18
ฑ0.05
0.03
ฑ0.02
O.01
ฑO.01
Annual
Average
(Ug/m3)
0.45
ฑ0.06
0.20
ฑ0.05
0.09
ฑ0.09
O.01
ฑO.01
New Brunswick, New Jersey - NBNJ
Acetaldehyde
Acrylonitrile
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
ฃ>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
Hexachloro- 1 ,3 -butadiene
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
62/62
25/58
58/58
42/58
58/58
48/58
39/58
19/58
58/58
62/62
9/58
50/58
15/58
7/58
2.38
ฑ0.55
0.08
ฑ0.04
0.84
ฑ0.23
0.05
ฑ0.03
0.45
ฑ0.07
0.07
ฑ0.03
0.04
ฑ0.03
<0.01
ฑ0.01
0.24
ฑ0.04
1.41
ฑ0.31
0.03
ฑ0.03
0.10
ฑ0.03
0.02
ฑ0.02
0.01
ฑ0.01
2.90
ฑ0.70
0.14
ฑ0.06
0.65
ฑ0.27
0.03
ฑ0.03
0.56
ฑ0.09
0.17
ฑ0.05
0.10
ฑ0.06
0.03
ฑ0.04
1.22
ฑ1.95
2.70
ฑ2.05
0.02
ฑ0.03
0.16
ฑ0.08
0.03
ฑ0.04
0.01
ฑ0.01
2.74
ฑ0.43
0.17
ฑ0.21
0.51
ฑ0.08
0.04
ฑ0.02
0.69
ฑ0.05
0.16
ฑ0.06
0.07
ฑ0.03
O.01
ฑ0.01
0.39
ฑ0.14
4.94
ฑ4.02
0.01
ฑ0.01
0.11
ฑ0.04
0.02
ฑ0.03
0.01
ฑ0.01
1.92
ฑ0.34
0.02
ฑ0.03
0.76
ฑ0.17
0.08
ฑ0.02
0.63
ฑ0.08
0.11
ฑ0.03
0.05
ฑ0.02
0.06
ฑ0.02
0.28
ฑ0.06
1.98
ฑ1.29
0.01
ฑ0.02
0.13
ฑ0.04
0.02
ฑ0.02
0.01
ฑ0.01
2.49
ฑ0.27
0.09
ฑ0.05
0.70
ฑ0.10
0.06
ฑ0.01
0.59
ฑ0.04
0.13
ฑ0.02
0.06
ฑ0.02
0.03
ฑ0.01
0.49
ฑ0.39
2.71
ฑ1.10
0.02
ฑ0.01
0.13
ฑ0.02
0.02
ฑ0.01
0.01
ฑO.01
Paterson, New Jersey - PANJ
Acrylonitrile
Benzene
1,3 -Butadiene
Carbon Tetrachloride
2/12
12/12
12/12
12/12
0
1.20
ฑ0.15
0.19
ฑ0.02
0.42
ฑ0.17
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
     NA = Not available due to the criteria for calculating a quarterly and/or annual average.
                                          17-32

-------
Table 17-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
                     the New Jersey Monitoring Sites (Continued)


Pollutant

Chloroform

1 ,2-Dibromoethane

ฃ>-Dichlorobenzene
1 ,2-Dichloroethane

Ethylbenzene

Hexachloro- 1 , 3 -butadiene

Tetrachloroethylene

Trichloroethylene

Vinyl Chloride
#of
Measured
Detections
vs. # of
Samples

5/12

3/12

12/12
2/12

12/12

2/12

12/12

7/12

3/12

1st
Quarter
Average
(Ug/m3)
0.05
ฑ0.05
0.01
ฑ0.03
0.13
ฑ0.03
0
0.42
ฑ0.08
0.02
ฑ0.04
0.19
ฑ0.06
0.05
ฑ0.03
0.01
ฑ0.01

2nd
Quarter
Average
(Ug/m3)

NA

NA

NA
NA

NA

NA

NA

NA

NA

3rd
Quarter
Average
(Ug/m3)

NA

NA

NA
NA

NA

NA

NA

NA

NA

4th
Quarter
Average
(Ug/m3)

NA

NA

NA
NA

NA

NA

NA

NA

NA


Annual
Average
(Ug/m3)

NA

NA

NA
NA

NA

NA

NA

NA

NA
     NA = Not available due to the criteria for calculating a quarterly and/or annual average.
      Observations for CHNJ from Table 17-5 include the following:

      •   The pollutants of interest with the highest annual average concentrations by mass are
          formaldehyde, acetaldehyde, and carbon tetrachloride. Only the two carbonyl
          compounds have annual average concentrations greater than 1 |ig/m3.

      •   Concentrations of formaldehyde appear to be higher during the warmer months, based
          on the quarterly averages. The second and third quarter averages are higher than the
          other quarterly averages and have significantly large confidence intervals. A review
          of the data shows that the maximum formaldehyde concentration was measured on
          July 26, 2011 (15.45 |ig/m3). The  18 highest formaldehyde concentrations (those
          greater than 2.25 |ig/m3) were all measured between May and September.
          Conversely, of the 19 concentrations less than 1 |ig/m3, 17 were measured in the  first
          quarter, one was measured in the second quarter, and one was measured in the fourth
          quarter.

      •   Concentrations of acrylonitrile appear highest during the third quarter of 2011, but the
          confidence interval for this average is nearly as high as the average itself. A review of
          the data shows that the maximum  acrylonitrile concentration was measured on
          August 19, 2011 (1.37 |ig/m3) and is an order of magnitude higher than the next
          highest concentration (0.294 |ig/m3, measured on August 1, 2011). The three highest
          measurements of acrylonitrile were all measured in August.
                                         17-33

-------
•  Several of the VOCs have third quarter average concentrations that are greater than
   the other quarterly average concentrations and/or have relatively large confidence
   intervals. For many of these, the maximum concentration was measured on
   July 8, 2011. And for most of these, the difference between the July 8, 2011
   measurement and the next highest concentration is significant. For example, the
   benzene concentration on July 8, 2011 (2.07 |ig/m3) is twice the next highest
   concentration (1.01 |ig/m3, measured on December 5, 2011).  As a second example,
   the tetrachloroethylene concentration measured on July 8, 2011 (1.09 |ig/m3) is nearly
   four times higher than the next highest concentration (0.299 |ig/m3, measured on
   May 3, 2011). This trend can also be seen in the data for 1,3-butadiene, chloroform,
   1,2-dichloroethane, and trichloroethylene. This date was also discussed in
   Section 17.3.
Observations for ELNJ from Table 17-5 include the following:

•  The pollutants of interest with the highest annual average concentrations by mass are
   formaldehyde, acetaldehyde, and benzene. These are the only pollutants with annual
   average concentrations greater than 1 |ig/m3.

•  The concentrations of many of the pollutants of interest for ELNJ appear to be higher
   during the warmer months of the year, as illustrated by the quarterly average
   concentrations, particularly the carbonyl compounds. However, most of the
   differences are not statistically significant.

•  Acrylonitrile was detected in only 11 samples collected at ELNJ and spanned an
   order of magnitude (0.0969 |ig/m3to 0.605  |ig/m3). There were two measured
   detections during the first quarter, eight in the second quarter, none in the third
   quarter, and one in the fourth quarter, which happened to be the maximum
   concentration measured at ELNJ. Thus, the fourth quarter average was determined
   from one single measurement and 15 non-detects.

•  The third quarter average ethylbenzene concentration is twice the other quarterly
   averages and has a relatively large confidence interval. A review of the data shows
   that the maximum concentration of ethylbenzene was measured on July 8, 2011
   (5.00 |ig/m3) and is more than five times the next highest concentration
   (0.979 |ig/m3). Several of the VOCs were highest on July 8, 2011, including
   chloroform, 1,2-dichloroethane, tetrachloroethylene, and trichloroethylene.


Observations for NBNJ from Table 17-5 include the following:

•  The pollutants of interest with the highest annual average concentrations by mass are
   formaldehyde, acetaldehyde, and benzene. Acetaldehyde and formaldehyde are the
   only pollutants with annual average concentrations greater than 1 |ig/m3.

•  The second, third, and fourth quarter average concentrations of formaldehyde have
   relatively large confidence intervals associated with them relative to the averages
   themselves. Concentrations of formaldehyde measured at NBNJ range from
   0.81 |ig/m3to 27.7 |ig/m3. The maximum concentration of formaldehyde was

                                   17-34

-------
          measured at NBNJ on September 6, 2011 and is the maximum concentration
          measured across the program. Three additional formaldehyde concentrations greater
          than 10 |ig/m3 were measured at NBNJ, one each in the second, third, and fourth
          quarters of 2011.

       •  The confidence interval associated with third quarter average concentration of
          acrylonitrile is larger than the average itself, indicating that the concentration average
          is likely influenced by outliers. The maximum concentration of acrylonitrile is
          1.34 |ig/m3 and was measured on September 24, 2011. This concentration is more
          than four times the next highest concentration measured at NBNJ (0.307 |ig/m3
          measured on October 6, 2011). This pollutant was detected in less than half of the
          VOC samples collected and concentrations of acrylonitrile range from 0.083 |ig/m3 to
          1.34|ig/m3.

       •  The second quarter ethylbenzene concentration is more than three times the other
          quarterly averages and its confidence interval is greater than the average itself. A
          review of the data shows that the maximum concentration of ethylbenzene was
          measured at NBNJ on April 9, 2011 (11.5 |ig/m3). This measurement is more than
          10 times the next highest concentration measured at NBNJ (1.07 |ig/m3) and is the
          highest concentration measured across the program. Several of the VOCs were
          highest on April 9, 2011, including l,2-dichloroethane,/>-dichlorobenzene,
          tetrachloroethylene, and trichloroethylene.


       Observations for PANJ from Table 17-5 include the following:

       •  VOC sampling at PANJ ended in the middle of May 2011. Thus,  only first quarter
          average concentrations could be calculated (second quarter concentrations were not
          calculated because there were not enough samples collected to meet the completeness
          criteria). However, Appendix J provides the pollutant-specific average  concentration
          for all valid samples collected at PANJ over the entire sample period.

       •  Benzene is the pollutant with the highest first quarter average concentration for PANJ
          and is the only pollutant for which a first quarter average is greater than 1 |ig/m3.


       Tables 4-9 through 4-12 present the sites with the 10 highest annual average

concentrations for each of the program-level pollutants of interest. Observations for the New

Jersey sites from those tables include the following:

       •  The New Jersey sites appear in Table 4-9 for VOCs a total of 23 times  (CJrtNJ - 7;
          ELNJ - 8; and NBNJ - 8). At least one New Jersey  site appears among the rankings
          for each of the program-level pollutants of interest  and all three New Jersey sites
          appear for acrylonitrile, chloroform, 1,2-dichloroethane, and trichloroethylene.

       •  CJrDSTJ has the highest annual average concentration of 1,2-dichloroethane among
          NMP sites sampling VOCs. The highest ranking for ELNJ is second for
          trichloroethylene; the highest ranking for NBNJ is second for vinyl chloride.
                                          17-35

-------
       •  The New Jersey sites appear in Table 4-10 for carbonyl compounds a total of three
          times. ELNJ and NBNJ have the highest and fifth highest annual average
          concentrations of acetaldehyde, respectively, while ELNJ has the eighth highest
          annual average concentration of formaldehyde.

17.4.2 Concentration Comparison
       In order to better illustrate how a site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for the selected NATTS MQO Core
Analytes listed in Section 3.5.3, where applicable. Thus, box plots for acetaldehyde, benzene,
1,3-butadiene, and formaldehyde were created for CHNJ, ELNJ, and NBNJ. Box plots were not
created for PANJ because annual averages could not be calculated for this site. Figures 17-19
through 17-22 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 3.5.3.
       Figure 17-19. Program vs. Site-Specific Average Acetaldehyde Concentrations
 CHNJ
 ELNJ
 NBNJ
                                   6          8          10
                                         Concentration !•: r":v
                Program:   1st Quartile     2nd Quartile     3rd Quartile     4th Quartile     Average
               Site:
                       Site Average    Site Minimum/Maximum

                                          17-36

-------
          Figure 17-20. Program vs. Site-Specific Average Benzene Concentrations
                                                                      i  Program Max Concentration = 23.8 ug/m3
                                                                      i  Program Max Concentration = 23.8 ug/rm3
NBNJ   ^—|	1                                             I Program Max Concentration = 23.8 ug/nv

    0123456789
                                                 Concentration (|ig/m3)
                  Program:   IstQuartile      2ndQuartile      SrdQuartile      4thQuartile      Average
                               •             •           D           D           I
                  Site:       Site Average     Site Minimum/Maximum
                               o	—                                 	

       Figure 17-21. Program  vs. Site-Specific Average 1,3-Butadiene Concentrations

    — i	                                                          i Program Max Concentration = 9.51 ug/m3


    ^m i                                                                [
    M_b~.	i                                                        I prฐgramMa)( Concentration = 9.51 ug/m3



    tl                                                              ,
NBNJ ^J	1                                                          [ Projjram Max Concentration = 9.51
                     3.5
                                                       1.5
                                                 Concentration (
                                                                                        2.5
                  Program:   IstQuartile      2ndQuartile      SrdQuartile      4thQuartile      Average
                  Site:       Site Average     Site Minimum/Maximum

                                                   17-37

-------
     Figure 17-22. Program vs. Site-Specific Average Formaldehyde Concentrations
NBNJ
                                            15
                                        Concentration (|ig/m3)
                                                                       25
              Program:  IstQuartile     2ndQuartile     3rd Quartile    4thQuartile     Average
              Site:     Site Average     Site Minimum/Maximum

      Observations from Figures 17-19 through 17-22 include the following:

         •   Figure 17-19 shows that while CUNJ's annual average acetaldehyde
             concentration is less than the program-level average concentration, the annual
             averages for ELNJ and NBNJ are greater than the program-level average
             concentration. In addition, the annual average for NBNJ is equivalent to the
             program-level third quartile while the annual average for ELNJ is greater than the
             program-level third quartile. The range of acetaldehyde measurements is greatest
             for ELNJ and least for CFINJ. There were no non-detects of acetaldehyde
             measured at the New Jersey sites or across the program.

         •   Figure 17-20 presents the box plots for benzene. Note that the program-level
             maximum concentration (23.8  |ig/m3) is not shown directly on the box plots
             because the scale of the box plots would be too large to readily observe data
             points at the lower end of the concentration range.  Thus, the scale has been
             reduced to 10 |ig/m3. This figure shows that the annual average benzene
             concentration for CFINJ is just greater than the  program-level first quartile. This
             site has one of the lowest annual average benzene concentrations among sites
             sampling benzene. NBNJ's annual average benzene concentration is less the
             program-level average but equivalent to the program-level median concentration.
             The annual average concentration for ELNJ is just greater than the program-level
             average concentration. The range of concentrations measured for ELNJ is roughly
             twice the range measured at CFINJ and NBNJ. There were no non-detects of
             benzene measured at the New Jersey sites or across the program.
                                         17-38

-------
          •  Figure 17-21 presents the box plots for 1,3-butadiene. Similar to the benzene box
             plots, the program-level maximum concentration (9.51 |ig/m3) is not shown
             directly on the box plots as the scale has been reduced to 3 |ig/m3 to allow for the
             observation of data points at the lower end of the concentration range. The
             statistical metrics shown for each site follow a similar trend as the metrics on the
             benzene box plots. The annual average 1,3-butadiene concentration for CHNJ is
             greater than the program-level first quartile but less than the program-level
             median concentration.  The annual average concentration for NBNJ is just less
             than the program-level median concentration. ELNJ's annual average
             concentration is greater than the program-level third quartile.  The range of
             measurements was smallest for NBNJ and greatest for ELNJ. Non-detects were
             measured at all three sites, but the number of non-detects varied significantly.
             Approximately half of the 1,3-butadiene measurements at CHNJ were non-
             detects, approximately one-quarter of the measurements at NBNJ were non-
             detects, and two non-detects were measured at ELNJ.

          •  Figure 17-22 for formaldehyde shows that while the annual average
             concentrations of formaldehyde for CHNJ and NBNJ are less than the program-
             level average, the annual average for ELNJ is greater than the program-level
             average concentration. Even though the range of measurements for ELNJ is the
             smallest among the New Jersey sites, ELNJ has the highest annual average
             concentration of formaldehyde among the three sites. Although the maximum
             formaldehyde concentration measured across the program was measured at
             NBNJ, this site's annual average ranked 13th among NMP sites sampling carbonyl
             compounds. By comparison, ELNJ ranked 8th and CHNJ 16th. There were no
             non-detects of formaldehyde measured at the New Jersey sites or across the
             program.


17.4.3  Concentration Trends

       A site-specific trends evaluation was completed for sites that have sampled one or more

of the selected NATTS MQO Core Analytes for 5 consecutive years or longer, as described in

Section 3.5.4. CHNJ, ELNJ, and NBNJ have sampled VOCs and carbonyl compounds under the

NMP for many years. ELNJ has sampled under the NMP since 2000 and CHNJ and NBNJ since

2001. Thus, Figures 17-23 through 17-34 present the annual statistical metrics for acetaldehyde,

benzene, 1,3-butadiene, and formaldehyde for CHNJ, ELNJ, and NBNJ, respectively. The

statistical metrics presented for assessing trends include the substitution of zeros for non-detects.

Because sampling at PANJ is being conducted for a one-year period ending in May 2011, a

trends analysis could not be performed.
                                         17-39

-------
Figure 17-23. Annual Statistical Metrics for Acetaldehyde Concentrations
                            Measured at CHNJ



1
E
.9
E
Average Concen
) |


0





rh
^* ™




•
I r-L F^

'-j-' 1— •- .—a-, ,=5=, i_o_, — ฃ- i-^-. i_^j — r-
2003 2004 2005 2006 2007 2008 2009 2010 2011
Year
• 5th Percentile — Minimum — Median — Maximum • 95th Percentile "^"Average
   Figure 17-24. Annual Statistical Metrics for Benzene Concentrations
                            Measured at CHNJ
                                       •*••
      2003      2004      2005     2006      2007      2008      2DD9      2010      2011
                                       Year
       *  5th Percentile     - Minimum     — Median     -  Maximum     • 95th Percentile   ••*•• Average
                                   17-40

-------
Figure 17-25. Annual Statistical Metrics for 1,3-Butadiene Concentrations
                         Measured at CHNJ



rn
i
[Vincent ration
3 c
I
<

0







•


_ r
T r
n u i a A r*i i
V" 	 	 .^H 	 A 	 ^^
— •• 	 1 	 H^ 	 1 	 	 1 	 ' t ' 	 1 	 ! — •— ! 	 1 	 !— •— ! 	 , 	 ' • ' 	 1— — l..fcl 	 , 	 ! — •— 1 	 ,
2003 2004 2005 2006 2007 2008 2009 2010 2011
Year
- Minimum — Median - Maximum • 95th Percentile • 5th Percentile "^"Average
Figure 17-26. Annual Statistical Metrics for Formaldehyde Concentrations
                         Measured at CHNJ



1
1
e
.9
E
|
3
3
s
<


0












•

^^ 	 <
L- l"rป =5= =•= =^*= —i- •-•-1 ^^ i=s=











r
~

rh n rh i r^i ""
' ...t. 	 I.AU 	 ...fa 	 ..fc.1 	 ซป^ 	 ^^_ 	 	
2003 2004 2005 2006 2007 2008 2009 2010 2011
Year
* 5th Percentile - Minimum — Median - Max mum • 95th Percentile . .+.. Average
                                17-41

-------
 Figure 17-27. Annual Statistical Metrics for Acetaldehyde Concentrations

                                Measured at ELNJ
!

I'
e
s
8
                                                                          *••
       2001     2002    2003     2004
                                    2005     2006     2007     2008



                                            Year
                                                                  2009    2010     2011
         •  5th Percentile     — Minimum     —  Median     — Maximum
                                                               95th Percentile
                                                                             • Average
    Figure 17-28. Annual Statistical Metrics for Benzene Concentrations

                                Measured at ELNJ
                                            The maximum benzene


                                            concentration for 2008


                                               is 34.3 ^g/m3
r
e
.9

B
ซ 8
8

i
      ฑ
       2001     2002    2003     2004    2005     2006     2007     2008     2009    2010     2011



                                            Year
         *  5th Percentile     - Minimum     —  Median     - Maximum     •  95th Percentile   • •*•• Average
                                        17-42

-------
 Figure 17-29. Annual Statistical Metrics for 1,3-Butadiene Concentrations
                               Measured at ELNJ
                           =E
      2001    2002    2003     2004     2005     2006     2007     2008    2009    2010    2011
                                          Year
         •  5th Percentile     — Minimum     — Median     — Maximum
                                                            95th Percentile   "^"Average
Figure 17-30. Annual Statistical Metrics for Formaldehyde Concentrations
                               Measured at ELNJ
r
B
a
•S S
8
i
      I
                            r   T
"r
T
      2001    2002     2003     2004
                                   2005     2006     2007    2003    2009     2010     2011
                                          Year
         *  5th Percentile     - Minimum     — Median     - Maximum
                                                            95th Percentile
                                                                          • Average
                                      17-43

-------
 Figure 17-31. Annual Statistical Metrics for Acetaldehyde Concentrations
                              Measured at NBNJ
                              The maximum acetaldehyde
                               concentration for 2004 is
                                   111 ng/m3
1
r




— p—




1
























	



r
—4—








r


m*m
'-?-'







T

1

— •-






      2002     2003      2004     2005     2006     2007     2008      2009     2010     2011
         #  5th Percentile    —  Minimum     — Median    —  Maximum    •  95th Percentile   ..^.. Average
    Figure 17-32. Annual Statistical Metrics for Benzene Concentrations
                              Measured at NBNJ



m"
rat ton \wJ\
i L
rage Co nee n
ซE











J


•
••


f T
~ฑ
fa 	 •+. ... . rn






" 1
ป— •• r*n

i i V^ — ^
1 i L,


.
J T
2002 2003 2004 20O5 2006 2007 2008 20O9 2010 2011
Year
* 5th Percentile - Minimum — Median - Maximum * 95th Percentile . . ^. . Average
                                     17-44

-------
  Figure 17-33. Annual Statistical Metrics for 1,3-Butadiene Concentrations

                                 Measured at NBNJ
"ง 0.3




|



•S 0.25


I

3
        2002      2003     2004     2005      2006      2007      200S      2009     2010     2011



                                             Year
          •  5th Percentile     — Minimum     —  Median     — Maximum     •  95th Percentile   "^"Average
  Figure 17-34. Annual Statistical Metrics for Formaldehyde Concentrations

                                 Measured at NBNJ
                              The maximum


                              formaldehyde


                             concentration for


                            2004 is 96.1 ug/m3.
 i
 e
 .9

 s
 ฃ :
        2002      2003     2004
                                2005      2006      2007



                                             Year
                                                         2008      2009     2010     2011
          *  5th Percentile     - Minimum     —  Median     - Maximum
                                                                95th Percentile
                                                                              • Average
                                         17-45

-------
       Observations from Figure 17-23 for acetaldehyde measurements at CHNJ include the
following:

       •  Although carbonyl compound sampling at CHNJ began in 2001, sampling did not
          begin until May, which does not yield enough samples for the statistical metrics to be
          calculated for 2001, based on the criteria specified in Section 3.5.4. In addition, data
          from 2002 are not provided due to a completeness less than 85 percent in 2002. Thus,
          Figure 17-23 begins with 2003.

       •  The maximum acetaldehyde concentration was measured in 2004. The second and
          third highest concentrations were measured in 2004 and 2005; excluding these three
          concentrations, all other acetaldehyde concentrations measured at CFDSTJ were less
          than 5 |ig/m3.

       •  Although difficult to discern in Figure 17-23, a decreasing trend in the average and
          median acetaldehyde  concentrations is shown though 2006, after which the median
          and average concentrations leveled out until 2011, when an increase is noted.
          However, the high concentrations measured in 2004 and 2005 result in confidence
          intervals that are relatively large and indicate that these changes are not statistically
          significant.

       •  All the statistical metrics calculated exhibit an increase from 2010 to 2011.

       •  There have been no non-detects of acetaldehyde measured at CHNJ over the period
          shown.


       Observations from Figure 17-24 for benzene measurements at CHNJ include the

following:

       •  Similar to carbonyl compounds, VOC sampling at CHNJ began in 2001. However,
          sampling did not begin until May, which does not yield enough samples for the
          statistical metrics to be calculated for 2001, based on the criteria specified in
          Section 3.5.4. In addition, data from 2002 and 2005 are not provided due to a
          completeness less than 85 percent for those years.

       •  Only six benzene concentrations greater than 2 |ig/m3 have been measured at CHNJ
          during the years shown. Two were measured in 2008, three in 2009, and one in 2011.

       •  The average and median concentrations exhibit a decreasing trend through 2007,
          although no data is shown for 2005. Even though an increase in the average
          concentration is shown from 2007 to 2008, confidence intervals calculated indicate
          that the changes are not statistically significant. Since 2006, the average concentration
          has ranged from 0.47 |ig/m3 (2007) to 0.59 |ig/m3 (2008).
                                         17-46

-------
       Observations from Figure 17-25 for 1,3-butadiene measurements at CHNJ include the
following:

       •   The maximum 1,3-butadiene concentration shown was measured in 2003 and is
          nearly twice the next highest concentration, which was measured in 2008. The third
          highest concentration was measured in 2011. Only five concentrations measured at
          CHNJ are greater than 0.2 |ig/m3.

       •   For 2003 and 2004, the minimum, first quarter, and median concentrations are all
          zero. This is because 88 percent of the measurements were non-detects for 2003 and
          96 percent were non-detects for 2004. The percentage of non-detects decreased to less
          than 40 percent for 2006; this decrease continued, reaching a minimum in 2008, after
          which an increasing number of non-detects was reported. For 2010, the median was
          again zero, indicating that at least 50 percent of the measurements were non-detects.

       •   The average and median concentrations have a decreasing trend from 2008 through
          2010. All of the statistical metrics exhibit an increase for 2011.


       Observations from Figure 17-26 for formaldehyde measurements at CFDSTJ include the

following:

       •   The statistical metrics presented for formaldehyde mimic those for acetaldehyde in
          Figure 17-23 in regards to trending.

       •   The maximum formaldehyde concentration was measured in 2004. This  concentration
          of formaldehyde is nearly four times the maximum concentrations shown for other
          periods excluding  2004. The second highest concentration was also measured in
          2004, but was nearly half the magnitude. These two maximum concentrations were
          measured on the same days as the maximum acetaldehyde concentrations. The third
          highest concentration of formaldehyde was measured in 2011, although similar
          concentrations were also measured in 2003 and 2007.

       •   Although difficult to discern in Figure 17-26, a decreasing trend  in the average and
          median formaldehyde concentrations is shown though 2006, after which the median
          and average concentrations leveled out until 2010, when the average concentration
          reached a minimum. The average concentration then increased for 2011. However,
          the high concentrations measured in 2004 result in confidence intervals that are
          relatively large, making the identification of trends difficult.

       •   Similar to acetaldehyde, all of the statistical metrics calculated for formaldehyde
          exhibit an increase from 2010 to 2011. The 95th percentile for 2011  is greater than  the
          maximum concentration for 2010.

       •   There have been no non-detects of formaldehyde measured at CHNJ over the period
          shown.
                                         17-47

-------
       Observations from Figure 17-27 for acetaldehyde measurements at ELNJ include the
following:

       •   ELNJ is the longest running UATMP site. Carbonyl compound sampling at ELNJ
          began in January 2000. However, sporadic sampling at the beginning of 2000
          combined with a l-in-12 day sampling schedule led to a completeness less than
          85 percent. Thus, Figure 17-27 begins with 2001. Completeness was also low in 2003
          due to a 1-month period when samples were not collected in January 2003; thus, no
          2003 data are presented in Figure 11-21.

       •   The maximum acetaldehyde concentration shown was measured in 2007, although a
          concentration of similar magnitude was also measured in 2005. In total, 22
          concentrations greater than 10 |ig/m3 have been measured at ELNJ, all of which were
          measured prior to 2008.

       •   The average concentration of acetaldehyde has a steadily increasing trend through
          2007, after which a significant decrease is exhibited. Although an increasing trend is
          also shown between 2008 and 2011, these averages are roughly half the magnitude of
          those shown before 2008.

       •   There have been no non-detects of acetaldehyde measured at ELNJ over the period
          shown.


       Observations from Figure 17-28 for benzene measurements at ELNJ include the

following:

       •   VOC sampling at ELNJ began in January 2000. However, sporadic sampling at the
          beginning of 2000 combined with a l-in-12 day sampling schedule led to a
          completeness less than 85 percent. Thus,  Figure 17-28 also begins with 2001.

       •   The maximum benzene concentration (34.3 |ig/m3) was measured in 2008 and is
          more than four times higher than the next highest concentration (measured in 2009).
          The third highest concentration was also measured in 2009. Only five benzene
          concentrations greater than 5 |ig/m3 have been measured at ELNJ.

       •   A decreasing trend in the average and median concentrations is shown through 2007.
          With the exception of the median concentration, all of the statistical parameters
          exhibit an increase for 2008. If the maximum concentration for 2008 was removed
          from the data set, the average concentration would exhibit a negligible increase for
          2008. Thus, it is this single concentration that is driving the average concentration.
          The median concentration is influenced less by outliers, as this statistical parameter
          represents the midpoint, or 50th percentile, of a data set. That the median did not
          change between 2007 and 2008 is further proof that this single outlier is driving the
          average concentration upward.
                                         17-48

-------
       •   Even though some of the highest concentrations were measured in 2009, as indicated
          by the maximum and 95th percentile, the average concentration decreased from 2008
          to 2009, likely a result of the sheer magnitude of the outlier affecting the 2008
          calculations.

       •   Figure 17-28 shows that benzene concentrations for 2010 and 2011 returned to levels
          similar to 2007.


       Observations from Figure 17-29 for 1,3-butadiene measurements at ELNJ include the
following:

       •   The maximum concentration of 1,3-butadiene was measured in 2009 and is nearly
          two and a half times the next highest concentration (measured in 2001). These two
          concentrations are the only concentrations of 1,3-butadiene measured at ELNJ greater
          than 1 |ig/m3.

       •   Figure 17-29 shows a decreasing trend in the average concentration through 2004,
          then a leveling off of average concentrations that continues through the 2011. Even
          with the higher concentration measured in 2009, the average concentration for 2009 is
          similar to the average concentration for 2008. Between 2004 and 2011, the average
          concentration has ranged from 0.11 |ig/m3 (2004) to 0.16 |ig/m3 (2009).

       •   Even with the maximum concentration measured in 2009, the difference between the
          5th and 95th percentiles has been decreasing since the onset of sampling, reaching a
          minimum in 2011. This indicates an overall decrease in the majority of concentrations
          measured at ELNJ.


       Observations from Figure 17-30 for formaldehyde measurements at ELNJ include the
following:

       •   The maximum formaldehyde concentration shown was measured in 2010, as was the
          second highest concentration. A total of 11 concentrations greater than 10 |ig/m3 have
          been measured at ELNJ.

       •   Figure 17-30 shows that there was an increase in formaldehyde concentrations from
          2002 to 2004, although there is no data shown for 2003. Between 2004 and 2007,
          there was relatively little change in the average formaldehyde concentration. Similar
          to acetaldehyde, the average concentration of formaldehyde decreased significantly
          between 2007 and 2008, after which an increasing trend is shown. While the trends
          graph for acetaldehyde shows a continued increase for 2011, the average
          formaldehyde concentration decreased for 2011.


       Observations from Figure 17-31 for acetaldehyde measurements atNBNJ include the
following:

       •   Although carbonyl compound sampling at NBNJ began in 2001, sampling did not
          begin until May, which does not yield enough samples for the statistical metrics to be

                                         17-49

-------
          calculated for 2001, based on the criteria specified in Section 3.5.4. Thus,
          Figure 17-31 begins with 2002.

       •  The maximum acetaldehyde concentration was measured in 2004. This concentration
          (111 |ig/m3) is nearly seven times higher, and an order of magnitude higher, than the
          next highest concentration (16.2 |ig/m3 measured in 2005).

       •  Of the 29 concentrations greater than 8 |ig/m3, 28 were measured in 2004 or 2005
          (and the other was measured in 2008). This, along with the outlier concentration
          measured in 2004, explains the significant increase in the statistical metrics from
          2003 to 2004. Even without an outlier for 2005, most of the statistical metrics for
          2005 exhibit slight increases from 2004 levels.  The average, however, does not. If the
          outlier was removed from the data set for 2004, the average concentration for 2004
          would be slightly less than the average concentration for 2005.

       •  The average concentration decreases significantly from 2005 to 2006 and reaches a
          minimum for 2007, as do all of the other statistical parameters. After 2007,  the
          average concentration fluctuates between 2 |ig/m3 and 3 |ig/m3.

       •  There have been no non-detects of acetaldehyde measured at NBNJ.


       Observations from Figure 17-32 for benzene measurements at NBNJ include the

following:

       •  VOC sampling at NBNJ also began in May 2001. Because seven months of sampling
          does not yield enough samples for the statistical metrics to be calculated for 2001,
          based on the criteria specified in Section 3.5.4,  Figure 17-32 also begins with 2002.

       •  The maximum benzene concentration was measured in 2002, but similar
          concentrations were also measured in 2005 and 2009. These are the only three
          concentrations of benzene greater than 3 |ig/m3 measured at NBNJ.

       •  Although a slight decreasing trend is shown between 2002 and 2004, a significant
          decrease is shown between 2005 and 2007, where several of the statistical parameters
          reached a minimum. The average  concentration increased slightly for 2008, after
          which little change is observed. The average concentration ranges from 0.65 |ig/m3
          (2010) to 0.70  |ig/m3 (2011) between 2008 and 2011.


       Observations from Figure 17-33 for 1,3-butadiene measurements  at NBNJ include the

following:

       •  The maximum 1,3-butadiene concentration was measured in 2005 and is the only
          measurement greater than 0.35 |ig/m3 measured at NBNJ.

       •  The minimum, 5th percentile, and median concentrations are all zero for 2002 through
          2004. This indicates that at least 50 percent of the measurements were non-detects for
          these years. The number of non-detects began to decrease in 2005 (47 percent) and

                                         17-50

-------
          reached a minimum in 2008 (2 percent). The number of non-detects increased to
          29 percent for 2011. The increase in non-detects for 2011 is evident, at least in part,
          from the decrease in the 5th percentile shown from 2010 to 2011.

       •  The average concentration of 1,3-butadiene at NBNJ decreased significantly from
          2003 to 2004. This is likely a result of the change in the number of non-detects as
          well as a reduction in the range of measurements. The number of non-detects
          increased from 69 percent to greater than 90 percent from 2003 to 2004. Thus, many
          zeros were substituted into this average. Conversely, the increase in the average
          concentration shown from  2004  to 2005 results from a combination of fewer non-
          detects and a larger range of measurements. The average concentration of
          1,3-butadiene after 2004 exhibits little change and ranges from 0.046 |ig/m3 (2009) to
          0.057 |ig/m3 (2008).


       Observations from Figure 17-34 for formaldehyde measurements at NBNJ include the
following:

       •  The maximum formaldehyde concentration (96.1 |ig/m3) was measured on the same
          day in 2004 that the highest acetaldehyde concentration was measured
          (August 31, 2004). This concentration of formaldehyde is more than three times the
          next highest concentration  (27.7 |ig/m3, measured in 2011). Concentrations greater
          than 20 |ig/m3 have been measured in 2004, 2006, 2009,  and 2011.

       •  Similar to acetaldehyde, several  of the statistical metrics exhibit increases from 2003
          to 2004. Also similar to acetaldehyde, while the average  concentration decreased
          from 2004 to 2005, many of the  other statistical metrics did not, indicating that
          concentrations on a whole were  higher in 2005 and not just influenced by outlier(s).

       •  After 2005, concentrations of formaldehyde decreased steadily, reaching a minimum
          in 2008. This year also had the smallest range of formaldehyde measurements. After
          2008, the average concentration fluctuated. Although the average concentration for
          2011 is not statistically different than the average concentrations for the  last few years
          of sampling, the 95th percentile for 2011 increased significantly, doubling or tripling
          compared to previous years. This indicates that more of the measurements are falling
          into a higher range than in  previous years.


17.5   Additional Risk-Based Screening Evaluations

       The following risk-based screening  evaluations were conducted to characterize risk at

each New Jersey monitoring site. Refer to Sections 3.3 and 3.5.5 for definitions and explanations

regarding the various toxicity factors,  time frames, and calculations  associated with these risk-

based screenings.
                                         17-51

-------
17.5.1  Risk-Based Screening Assessment Using MRLs
       A risk-based screening was conducted by comparing the concentration data from the New
Jersey monitoring sites to the ATSDR MRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 to 14 days); intermediate (exposures of 15 to 364 days); and chronic (exposures
of 1 year or greater). The preprocessed daily measurements of the pollutants of interest were
compared  to the acute MRLs; the quarterly averages were  compared to the intermediate MRLs;
and the annual averages were compared to the chronic MRLs.

       As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2011.

17.5.2  Cancer Risk and Noncancer Hazard Approximations
       For the pollutants of interest for the New Jersey sites and where annual average
concentrations could be calculated, risk was examined by calculating cancer risk and noncancer
hazard approximations.  These approximations can be used as risk estimates for cancer and
noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers may want to shift or
confirm their air-monitoring priorities. Refer to Section 3.5.5.2 for an explanation of how  cancer
risk and noncancer hazard approximations are calculated and what limitations are associated with
them. Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer
hazard approximations are presented in Table 17-6, where applicable. Cancer risk
approximations are presented as probabilities while the noncancer hazard approximations  are
ratios and  thus, unitless values.
                                         17-52

-------
                Table 17-6. Risk Approximations for the New Jersey Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Chester, New Jersey - CHNJ
Acetaldehyde
Acrylonitrile
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
1 ,2-Dichloroethane
Formaldehyde
Hexachloro- 1 ,3 -butadiene
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
0.0000022
0.000068
0.0000078
0.00003
0.000006

0.000026
0.000013
0.000022
0.00000026
0.0000048
0.0000088
0.009
0.002
0.03
0.002
0.1
0.098
2.4
0.0098
0.09
0.04
0.002
0.1
61/61
35/61
61/61
31/61
61/61
50/61
19/61
61/61
6/61
40/61
5/61
5/61
1.61
ฑ0.20
0.10
ฑ0.05
0.53
ฑ0.07
0.02
ฑ0.01
0.62
ฑ0.03
0.10
ฑ0.03
0.03
ฑ0.02
2.37
ฑ0.63
0.01
ฑ0.01
0.09
ฑ0.04
0.05
ฑ0.09
<0.01
ฑ<0.01
3.55
7.13
4.12
0.71
3.74

0.88
30.80
0.24
0.02
0.23
0.02
0.18
0.05
0.02
0.01
0.01
0.01
0.01
0.24
0.01
O.01
0.02
O.01
Elizabeth, New Jersey - ELNJ
Acetaldehyde
Acrylonitrile
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
ฃ>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
0.0000022
0.000068
0.0000078
0.00003
0.000006

0.000011
0.000026
0.0000025
0.000013
0.009
0.002
0.03
0.002
0.1
0.098
0.8
2.4
1
0.0098
61/61
11/61
61/61
59/61
61/61
48/61
49/61
16/61
61/61
61/61
3.24
ฑ0.39
0.04
ฑ0.03
1.03
ฑ0.13
0.14
ฑ0.02
0.61
ฑ0.04
0.14
ฑ0.03
0.10
ฑ0.02
0.03
ฑ0.01
0.51
ฑ0.16
3.45
ฑ0.44
7.14
2.50
8.00
4.05
3.68

1.10
0.73
1.27
44.84
0.36
0.02
0.03
0.07
0.01
O.01
0.01
O.01
0.01
0.35
NA = Not available due to the criteria for calculating an annual average
— = a Cancer URE or Noncancer RfC is not available
                                                 17-53

-------
          Table 17-6. Risk Approximations for the New Jersey Monitoring Sites (Continued)
Pollutant
Propionaldehyde
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
Cancer
URE
(Hg/m3)1

0.00000026
0.0000048
0.0000088
Noncancer
RfC
(mg/m3)
0.008
0.04
0.002
0.1
#of
Measured
Detections
vs. # of
Samples
61/61
56/61
24/61
4/61
Annual
Average
(Hg/m3)
0.45
ฑ0.06
0.20
ฑ0.05
0.09
ฑ0.09
0.01
ฑ0.01
Cancer Risk
Approximation
(in-a-million)

0.05
0.45
0.01
Noncancer
Hazard
Approximation
(HQ)
0.06
0.01
0.05
0.01
New Brunswick, New Jersey - NBNJ
Acetaldehyde
Acrylonitrile
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
ฃ>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
Hexachloro- 1 ,3 -butadiene
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
0.0000022
0.000068
0.0000078
0.00003
0.000006

0.000011
0.000026
0.0000025
0.000013
0.000022
0.00000026
0.0000048
0.0000088
0.009
0.002
0.03
0.002
0.1
0.098
0.8
2.4
1
0.0098
0.09
0.04
0.002
0.1
62/62
25/58
58/58
42/58
58/58
48/58
39/58
19/58
58/58
62/62
9/58
50/58
15/58
7/58
2.49
ฑ0.27
0.09
ฑ0.05
0.70
ฑ0.10
0.06
ฑ0.01
0.59
ฑ0.04
0.13
ฑ0.02
0.06
ฑ0.02
0.03
ฑ0.01
0.49
ฑ0.39
2.71
ฑ 1.10
0.02
ฑ0.01
0.13
ฑ0.02
0.02
ฑ0.01
O.01
ฑO.01
5.47
6.20
5.45
1.66
3.52

0.68
0.79
1.23
35.23
0.33
0.03
0.11
0.03
0.28
0.05
0.02
0.03
0.01
O.01
0.01
O.01
0.01
0.28
0.01
O.01
0.01
O.01
Paterson, New Jersey - PANJ
Acrylonitrile
Benzene
1,3 -Butadiene
Carbon Tetrachloride
0.000068
0.0000078
0.00003
0.000006
0.002
0.03
0.002
0.1
2/12
12/12
12/12
12/12
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA = Not available due to the criteria for calculating an annual average
— = a Cancer URE or Noncancer RfC is not available
                                                 17-54

-------
          Table 17-6. Risk Approximations for the New Jersey Monitoring Sites (Continued)
Pollutant
Chloroform
1 ,2-Dibromoethane
ฃ>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Hexachloro- 1 , 3 -butadiene
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
Cancer
URE
(Hg/m3)1

0.0006
0.000011
0.000026
0.0000025
0.000022
0.00000026
0.0000048
0.0000088
Noncancer
RfC
(mg/m3)
0.098
0.009
0.8
2.4
1
0.09
0.04
0.002
0.1
#of
Measured
Detections
vs. # of
Samples
5/12
3/12
12/12
2/12
12/12
2/12
12/12
7/12
3/12
Annual
Average
(Hg/m3)
NA
NA
NA
NA
NA
NA
NA
NA
NA
Cancer Risk
Approximation
(in-a-million)
NA
NA
NA
NA
NA
NA
NA
NA
NA
Noncancer
Hazard
Approximation
(HQ)
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA = Not available due to the criteria for calculating an annual average
- = a Cancer URE or Noncancer RfC is not available
     Observations from Table 17-6 include the following:
            •  For CHNJ, the pollutants with the highest annual averages are formaldehyde,
               acetaldehyde, and carbon tetrachloride. Formaldehyde has the highest cancer risk
               approximation for this site, followed by acrylonitrile and benzene. The cancer risk
               approximation for formaldehyde is at least an order of magnitude higher than the
               approximations for the other pollutants of interest. None of the pollutants of interest
               for CHNJ have noncancer hazard approximations greater than 1.0, indicating that no
               adverse health effects are expected from these individual pollutants.

            •  For ELNJ, the pollutants with the highest annual averages are formaldehyde,
               acetaldehyde, and benzene. These three pollutants also have the highest cancer risk
               approximations for this site, although the cancer risk approximation for benzene is
               greater than the cancer risk approximation for acetaldehyde. None of the pollutants of
               interest for ELNJ have noncancer hazard approximations greater than 1.0, indicating
               that no adverse health effects are expected from these individual pollutants.

            •  For NBNJ, the pollutants with the highest annual averages are formaldehyde,
               acetaldehyde, and benzene. Formaldehyde has the highest cancer risk approximation
               for NBNJ, followed by acrylonitrile and acetaldehyde (with benzene nearly
               equivalent to acetaldehyde). None of the pollutants of interest for NBNJ have
               noncancer hazard approximations greater than 1.0, indicating that no adverse health
               effects are expected from these individual pollutants.
                                               17-55

-------
       •  Because annual averages could not be calculated for PANJ, cancer risk and noncancer
          hazard approximations could not be calculated either.
17.5.3 Risk-Based Emissions Assessment
       In addition to the risk-based screenings discussed above, Tables 17-7 and 17-8 present an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 17-7 presents the 10 pollutants with the highest emissions from the 2008 NEI, the
10 pollutants with the highest toxicity-weighted emissions, and the 10 pollutants with the highest
cancer risk approximations (in-a-million), as calculated from the annual averages provided in
Table 17-6. Table 17-8 presents similar information, but identifies the 10 pollutants with the
highest noncancer hazard approximations, also calculated from annual averages provided in
Table 17-6.

       The pollutants listed in  Tables 17-7 and 17-8 are limited to those that have cancer and
noncancer toxicity factors, respectively. As a result, although the actual value of the emissions is
the same, the highest emitted pollutants in the cancer table may be different from the noncancer
table. The cancer risk and noncancer hazard approximations based on each site's annual averages
are limited to those pollutants for which each respective site sampled. As discussed in
Section 17.3, CHNJ, ELNJ, and NBNJ sampled for VOCs and carbonyl compounds. 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. The completeness criteria were not
met by PANJ because of the short sampling duration; as a result, annual  averages, and thus
cancer risk and noncancer hazard approximations, were not calculated for this site. A more in-
depth discussion of this analysis is provided in Section 3.5.5.3. Similar to the cancer risk and
noncancer hazard approximations, this analysis may help policy-makers  prioritize their air
monitoring activities.
                                          17-56

-------
Table 17-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
                                             the New Jersey Monitoring Sites
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Chester, New Jersey (Morris County) - CHNJ
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Naphthalene
Dichloromethane
POM, Group 2b
POM, Group la
POM, Group 6
207.44
118.39
111.67
75.87
32.41
12.97
9.08
1.98
0.32
0.16
Benzene
Formaldehyde
1,3 -Butadiene
Naphthalene
Hexavalent Chromium, PM
Ethylbenzene
POM, Group 2b
Acetaldehyde
Arsenic, PM
POM, Group 3
1.62E-03
1.45E-03
9.72E-04
4.41E-04
3.11E-04
2.96E-04
1.74E-04
1.67E-04
1.31E-04
1.24E-04
Formaldehyde
Acrylonitrile
Benzene
Carbon Tetrachloride
Acetaldehyde
1 ,2-Dichloroethane
1,3 -Butadiene
Hexachloro- 1 ,3 -butadiene
Trichloroethylene
Tetrachloroethylene
30.80
7.13
4.12
3.74
3.55
0.88
0.71
0.24
0.23
0.02
Elizabeth, New Jersey (Union County) - ELNJ
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
Dichloromethane
1,3 -Butadiene
Naphthalene
POM, Group 2b
Nickel, PM
Propylene oxide
160.93
111.28
89.90
67.31
41.31
23.48
11.41
1.57
1.36
0.70
Formaldehyde
Benzene
1,3 -Butadiene
Nickel, PM
Hexavalent Chromium, PM
Naphthalene
Arsenic, PM
Ethylbenzene
Acetaldehyde
POM, Group 2b
1.45E-03
1.26E-03
7.04E-04
6.52E-04
6.03E-04
3.88E-04
2.37E-04
2.25E-04
1.48E-04
1.38E-04
Formaldehyde
Benzene
Acetaldehyde
1,3 -Butadiene
Carbon Tetrachloride
Acrylonitrile
Ethylbenzene
ฃ>-Dichlorobenzene
1 ,2-Dichloroethane
Trichloroethylene
44.84
8.00
7.14
4.05
3.68
2.50
1.27
1.10
0.73
0.45

-------
  Table 17-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
                                          the New Jersey Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
New Brunswick, New Jersey (Middlesex County) - NBNJ
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
Dichloromethane
POM, Group 2b
Ethylene oxide
Tetrachloroethylene
263.33
175.06
146.78
107.72
40.00
20.62
7.26
2.82
1.05
0.96
Formaldehyde
Benzene
1,3 -Butadiene
Naphthalene
Hexavalent Chromium, PM
Ethylbenzene
Nickel, PM
POM, Group 2b
Acetaldehyde
Arsenic, PM
2.28E-03
2.05E-03
1.20E-03
7.01E-04
4.61E-04
3.67E-04
2.67E-04
2.48E-04
2.37E-04
1.84E-04
Formaldehyde
Acrylonitrile
Acetaldehyde
Benzene
Carbon Tetrachloride
1,3 -Butadiene
Ethylbenzene
1 ,2-Dichloroethane
ฃ>-Dichlorobenzene
Hexachloro- 1 ,3 -butadiene
35.23
6.20
5.47
5.45
3.52
1.66
1.23
0.79
0.68
0.33
Paterson, New Jersey (Passaic County) - PANJ
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Naphthalene
Dichloromethane
POM, Group 2b
POM, Group la
POM, Group 6
115.84
63.19
62.14
41.52
17.51
7.26
3.63
1.09
0.32
0.09
Benzene
Formaldehyde
1,3 -Butadiene
Naphthalene
Hexavalent Chromium, PM
Ethylbenzene
POM, Group 2b
Acetaldehyde
POM, Group 3
Arsenic, PM
9.04E-04
8.08E-04
5.25E-04
2.47E-04
1.63E-04
1.58E-04
9.60E-05
9.14E-05
8.90E-05
7.09E-05

oo

-------
  Table 17-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
                                           RfCs for the New Jersey Monitoring Sites
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer Hazard
Approximation
(HQ)
Chester, New Jersey (Morris County) - CHNJ
Toluene
Xylenes
Benzene
Hexane
Ethylbenzene
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Ethylene glycol
Naphthalene
624.27
493.82
207.44
131.84
118.39
111.67
75.87
32.41
29.66
12.97
Acrolein
1,3 -Butadiene
Formaldehyde
Acetaldehyde
Benzene
Xylenes
Lead, PM
Naphthalene
Arsenic, PM
Propionaldehyde
302,596.87
16,206.30
11,395.19
8,429.94
6,914.65
4,938.24
4,349.55
4,323.69
2,026.03
1,006.84
Formaldehyde
Acetaldehyde
Acrylonitrile
Trichloroethylene
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Tetrachloroethylene
Chloroform
Hexachloro- 1 ,3 -butadiene
0.24
0.18
0.05
0.02
0.02
0.01
0.01
0.01
0.01
0.01
Elizabeth, New Jersey (Union County) - ELNJ
Toluene
Xylenes
Benzene
Formaldehyde
Hexane
Ethylbenzene
Acetaldehyde
Dichloromethane
Ethylene glycol
Hydrochloric acid
471.64
352.06
160.93
111.28
109.81
89.90
67.31
41.31
36.52
24.34
Acrolein
Nickel, PM
1,3 -Butadiene
Formaldehyde
Acetaldehyde
Benzene
Naphthalene
Manganese, PM
Arsenic, PM
Xylenes
267,278.15
15,081.94
11,739.58
11,354.92
7,478.86
5,364.23
3,803.12
3,755.20
3,671.47
3,520.63
Acetaldehyde
Formaldehyde
1,3 -Butadiene
Propionaldehyde
Trichloroethylene
Benzene
Acrylonitrile
Carbon Tetrachloride
Tetrachloroethylene
Chloroform
0.36
0.35
0.07
0.06
0.05
0.03
0.02
0.01
0.01
0.01
VO

-------
Table 17-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
                                   RfCs for the New Jersey Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer Hazard
Approximation
(HQ)
New Brunswick, New Jersey (Middlesex County) - NBNJ
Toluene
Xylenes
Benzene
Hexane
Formaldehyde
Ethylbenzene
Acetaldehyde
Ethylene glycol
1,3 -Butadiene
Glycol ethers, gas
776.52
593.79
263.33
227.61
175.06
146.78
107.72
49.05
40.00
37.20
Acrolein
1,3 -Butadiene
Formaldehyde
Manganese, PM
Acetaldehyde
Benzene
Naphthalene
Titanium tetrachloride
Nickel, PM
Xylenes
445,148.01
20,000.36
17,862.97
12,571.41
11,968.98
8,777.75
6,872.93
6,385.00
6,175.26
5,937.95
Formaldehyde
Acetaldehyde
Acrylonitrile
1,3 -Butadiene
Benzene
Trichloroethylene
Carbon Tetrachloride
Tetrachloroethylene
Chloroform
Ethylbenzene
0.28
0.28
0.05
0.03
0.02
0.01
0.01
0.00
0.00
0.00
Paterson, New Jersey (Passaic County) - PANJ
Toluene
Xylenes
Benzene
Hexane
Ethylbenzene
Formaldehyde
Acetaldehyde
Ethylene glycol
1,3 -Butadiene
Glycol ethers, gas
358.91
286.81
115.84
71.05
63.19
62.14
41.52
29.87
17.51
13.40
Acrolein
1,3 -Butadiene
Formaldehyde
Acetaldehyde
Benzene
Xylenes
Naphthalene
Arsenic, PM
Lead, PM
Glycol ethers, gas
154,681.77
8,754.09
6,341.18
4,613.84
3,861.24
2,868.11
2,419.97
1,099.65
797.37
670.22


-------
Observations from Table 17-7 include the following:

•  Benzene is the highest emitted pollutant with a cancer URE in all four New Jersey
   counties, followed by ethylbenzene, formaldehyde, and acetaldehyde (although not
   necessarily in that order).

•  Benzene, formaldehyde, and 1,3-butadiene are the pollutants with the highest
   toxi city-weighted emissions (of the pollutants with cancer UREs) for all four New
   Jersey counties, although not necessarily in that order.

•  Seven of the 10 highest emitted pollutants in Morris, Middlesex, and Passaic Counties
   also have the highest toxi city-weighted  emissions while eight of the highest emitted
   pollutants in Union County also have the highest toxicity-weighted emissions.

•  Formaldehyde is the pollutant with the highest cancer risk approximations for CFINJ,
   ELNJ, and NBNJ. This pollutant also appears at or near the top of both emissions-
   based lists. Acetaldehyde, benzene, and 1,3-butadiene also appear on all three lists for
   these sites. Conversely, carbon tetrachloride and acrylonitrile appear on neither
   emissions-based list for the New  Jersey sites but appear among the  pollutants with the
   highest cancer risk approximations for each site.


Observations from Table 17-8 include the following:

•  Toluene, xylenes, and benzene are the highest emitted pollutants with noncancer
   RfCs for all four New Jersey counties.

•  Acrolein is the pollutant with the highest toxicity-weighted emissions (of the
   pollutants with noncancer RfCs) for all four counties but is not among the highest
   emitted pollutants for any of the New Jersey counties. Although acrolein was sampled
   for at all four sites, this pollutant  was excluded from the pollutant of interest
   designation, and thus subsequent risk-based screening evaluations,  due to questions
   about the consistency and reliability of the measurements, as discussed in Section 3.2.

•  Behind acrolein, 1,3-butadiene and formaldehyde have the highest noncancer
   toxicity-weighted emissions for three of the four New Jersey counties. For Union
   County (ELNJ), nickel has the second highest noncancer toxicity-weighted emissions,
   followed by 1,3-butadiene and formaldehyde.

•  The number of pollutants in common between the highest emitted pollutants and
   those with the highest toxicity-weighted emissions ranged from four (Union County)
   to six (Passaic and Morris Counties).

•  Formaldehyde and acetaldehyde are among the pollutants with the highest noncancer
   hazard approximations for CFINJ, ELNJ, and NBNJ (although all were less than an
   HQ of 1.0). These pollutants also appear among the pollutants with the highest
   emissions and toxicity-weighted emissions for all four counties. Benzene and
   1,3-butadiene also appear on all three lists for these sites.
                                   17-61

-------
17.6   Summary of the 2011 Monitoring Data for the New Jersey Monitoring Sites

       Results from several of the data treatments described in this section include the
following:

       ปซป  Seventeen pollutants failed at least one screen for CHNJ; 16 failed screens for ELNJ;
          18 failed screens for NBNJ; and 10 failed screens for PANJ.

       ปซป  Formaldehyde and acetaldehyde had the highest annual average concentrations for
          CHNJ, ELNJ, and NBNJ. Annual average concentrations could not be calculated for
          PANJ.

       ปซป  The annual average acetaldehyde concentration for ELNJ is the highest annual
          average among NMP sites sampling this pollutant. The maximum concentration of
          formaldehyde measured across the program was measured at NBNJ.

       ปซป  Several of the pollutants for which a trends analysis was performed exhibit slight
          increasing trends from 2010 to 2011 for the New Jersey sites.
                                        17-62

-------
18.0   Sites in New York
       This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS sites in New York, and integrates these concentrations
with emissions, meteorological, and risk information. Data generated by sources other than ERG
are not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) for detailed discussions and definitions
regarding the various data analyses presented below.

18.1   Site Characterization
       This section characterizes the New York monitoring sites by providing geographical and
physical information about the locations of the sites and the surrounding areas. This information
is provided to give the reader insight regarding factors that may influence the air quality near the
sites and assist in the interpretation of the ambient monitoring measurements.

       The New York monitoring sites are located in New York City (MONY) and Rochester
(ROCH). Figures 18-1 and 18-3 are composite satellite images retrieved from ArcGIS Explorer
showing the monitoring sites in their urban locations. Figures 18-2 and  18-4 identify nearby
point source emissions locations by source category, as reported in the 2008 NEI for point
sources. Note that only sources within 10 miles of the sites are included in the facility counts
provided in Figures 18-2 and 18-4. A 10-mile boundary was chosen to give the reader an
indication of which emissions sources and emissions source categories could potentially have a
direct effect on the air quality at the monitoring sites. Further, this boundary provides both the
proximity of emissions sources to the monitoring sites as well as the quantity of such sources
within a given distance of the sites. Sources outside the 10-mile radius are still visible on each
map, but have been grayed out in order to show emissions sources just outside the boundary.
Table 18-1 provides supplemental geographical information  such as land use, location setting,
and locational coordinates.
                                           18-1

-------
                                Figure 18-1. New York City, New York (MONY) Monitoring Site
oo
to

-------
Figure 18-2. NEI Point Sources Located Within 10 Miles of MONY
                                              73-55VW           73'50'Q"W            73'45'0*W           73'40'0-W
  Legend
           Note: Due to facility density and collocation, the total facilities
           displayed may not represent all facilities within the area of interest.
          MONY NATTS site
  Source Category Group (No. of Facilities)
    •   Abtastve Product Manufacturing (1)
   •?'?  Air-condrtioning/Refrigeralion (3)
   •&  Aircraft Operations (21)
   &  Automobite'Truck Manufacturing (1)
   $  Bake^(l)
   B  Bulk Terminals/Bulk Plants (2)
   C  Chemical Manufacturing (5)
   6  Etectrical Equipment (2)
   ฃ   Electncity Generation via Combustion (16)
   E  Electroplating. Plating, Polishing. Anodizing, & Coloring (1)
   ฉ  Fabricated Metal Products (2)
   ?ฃ"  Flexibte Polyurethane Foam Production (2)
   F  Food Processing/Agriculture (3)
   fl~  Gasoline/Diesel Service Station (3)
   i-V  Heating Equipment Manufacturing (2)
   O  Hospital (4)
10 mile radius |	| County boundary
              ^  Industrial Ma.chinery and Equipment (1)
              O  Institutional • prison (1)
              ^  Institutional - school (21)
              ?  Miscellaneous Commercial/Industrial (26)
              M  Miscellaneous Manufacturing (11)
              •   Oil and/or Gas Production (1)
              en  Pharmaceutical Manufacturing (2)
              ^  Printing, Coating & Dyeing of Fabric (1)
              P   Printing/Publishing (10)
              B  Pulp and Paper Pinny Wood Products <5)
              R   Rubber and Miscellaneous Plastics Products (2)
              ^  Ship Building and Repairing (1)
              >   Solid Wasle D4Sposal - CommorciaVlnstilutional (1)
              S   Surface Coating (5)
              T   Tex tile Mil I (2)
               •   •/•„• I-,'.-•„•.• j!>- Treatment (G)
                                                     18-3

-------
                                    Figure 18-3. Rochester, New York (ROCH) Monitoring Site
oo

-------
Figure 18-4. NEI Point Sources Located Within 10 Miles of ROCH
        Legend
       77"35'0'W        77'3010"W         77'25XTW
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
              ROCH NATTS site  O   10 mile radius     ] County boundary
                Source Category Group (No. of Facilities)
                      Air-conditioning/Refrigeration (1)
                  •f1   Aircraft Operations (6)
                  c   Chemical Manufacturing (2)
                  f   Electricity Generation via Combustion (1}
                  t<   Heating Equipment Manufacturing (2)
                  •   Landfill (2)
                  P   Printing/Publishing (2)
                  R   Rubber and Miscellaneous Plastics Products (1)
                  *   Wastewater Treatment (1)
                                      18-5

-------
                              Table 18-1. Geographical Information for the New York Monitoring Sites

Site
Code

MONY

ROCH

AQS Code

36-005-0080

36-055-1007

Location

New York

Rochester

County

Bronx

Monroe
Micro- or
Metropolitan
Statistical Area
New York-
Northern New
Jersey-Long Island,
NY-NJ-PAMSA
(New York Div)
Rochester, NY
MSA
Latitude
and
Longitude

40.83606,
-73.92009

43.146198,
-77.54813

Land Use

Residential

Residential

Location
Setting

Urban/City
Center

Urban/City
Center

Additional Ambient Monitoring Information1

Carbonyl Compounds, VOCs, Meteorological
Parameters, Black carbon, PM10 Speciation, PM25.

CO, SO2, VOCs, Carbonyl compounds, O3,
Meteorological parameters, Black Carbon, PM10,
PM10 Speciation, PM25, and PM25 Speciation.
  BOLD ITALICS = EPA-designated NATTS Site
oo

-------
       The MONY site is located at the Morrisania Neighborhood Family Care center. This site
is considered the Bronx (#2) NATTS site and is a relocation of the previous location. MONY is
located less than three-quarters of a mile south of 1-95, one-half mile east of 1-87 and east of the
Harlem River, which separates the island of Manhattan from the Bronx. Part of the Harlem River
can be seen in the upper left-hand corner of Figure 18-1. The Hudson River is just a few blocks
farther west. The area surrounding MONY is primarily residential, although commercial  areas
are located along Jerome  Avenue and East 167th Street. Figure 18-2 shows the numerous point
sources that are located within 10 miles of the site. The bulk of the emissions sources are located
to the south and west of the site. The source categories with the highest number of emissions
sources surrounding MONY include aircraft operations, which include airports as  well as small
runways, heliports, or landing pads;  electricity generation via combustion; schools; and printing
and publishing. The point source closest to MONY is a medical school.

       ROCH is located on the east side of Rochester, in western New York, at a power
substation. Rochester is approximately halfway between Syracuse and Buffalo, and Lake
Ontario lies to the north. Although the area north and west of the site is primarily residential, as
shown in Figure 18-3, a railroad transverses the area just south of the site, and 1-590  and 1-490
intersect farther south with commercial  areas adjacent to this corridor. The site is used by
researchers from several universities for short-term air monitoring studies and is the Rochester
NATTS site. As Figure 18-4 shows, the relatively few point sources within a 10-mile radius of
ROCH are located primarily on the west side of the 10-mile radius. The aircraft operations
source category is the category with the highest number of emissions sources surrounding
ROCH, although there are also landfills, chemical manufacturers, printing and publishing
facilities, and heating equipment manufacturers nearby, to name a few.

       Table 18-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the New York monitoring sites. Table  18-2 includes county-level
population and vehicle registration information. Table 18-2 also includes a county-level vehicle
registration-to-population ratio, which was calculated to represent the number of vehicles per
person within each monitoring site's residing county. In addition, the population within 10 miles
of each site is presented, based on postal code population data estimates. An estimate of 10-mile
vehicle ownership was determined by applying the county-level vehicle registration-to-
population ratio to the 10-mile population surrounding each monitoring site. Table 18-2 also
                                          18-7

-------
contains traffic volume information for each site. Finally, Table 18-2 presents the county-level

daily VMT for Bronx and Monroe Counties.
     Table 18-2. Population, Motor Vehicle, and Traffic Information for the New York
                                    Monitoring Sites
Site
MONY
ROCH
Estimated
County
Population1
1,392,002
745,625
County-level
Vehicle
Registration2
246,748
550,992
Vehicles per
Person
(Registration:
Population)
0.18
0.74
Population
within 10
miles3
5,684,739
650,600
Estimated
10-mile
Vehicle
Ownership
1,007,684
480,772
Annual
Average
Daily
Traffic4
91,465
86,198
County-
level Daily
VMT5
9,698,000
17,772,000
Bounty-level population estimates reflect 2011 data from the U.S. Census Bureau (Census Bureau, 2012b)
2County-level vehicle registration reflects 2011 data from the New York State DMV (NYS DMV, 2011)
310-mile population estimates reflect 2011 data from Xionetic (Xionetic, 2011)
4AADT reflects 2010 data from the New York State DOT (NYS DOT, 2010)
5 County-level VMT reflects 2009 data from the New York State DOT (NYS DOT, 2012)
BOLD ITALICS = EPA-designaled NATTS Site
       Observations from Table 18-2 include the following:

       •  Bronx County has the ninth highest county-level population among counties with
          NMP sites, but the 10-mile radius population for MONY is the highest among all
          NMP sites.
          County-level vehicle ownership for Bronx County is in the middle of the range
          among NMP sites. Although the 10-mile ownership estimate is one of the highest
          estimates among NMP sites, given the large population living within 10 miles of
          MONY, the vehicle-per-person ratio is very low (0.18), which is the lowest vehicle-
          per-person ratio calculated. This may seem surprising given the high population, but
          may be explained by the use of mass transportation systems.

          The population surrounding ROCH is lower than the population surrounding MONY.
          However, the county-level vehicle ownership for ROCH is double the vehicle
          ownership near MONY. The same is not true of the 10-mile ownership estimates. The
          county-level and 10-mile population and vehicle ownership data for ROCH are in the
          middle of the range compared to other NMP  sites.
                                                                                      rth
The traffic volume near MONY and ROCH are not that different from each other.
Compared to other NMP sites, the traffic volumes near MONY and ROCH rank 171
and 18th, respectively. The traffic data for MONY are provided for 1-87 between the
Bronx Expressway and Macombs Bridge and the traffic data for ROCH are provided
for 1-490 at 1-590.
       •  County-level daily VMT for Monroe County is nearly twice the VMT for Bronx
          County. These VMT are in the middle of the range compared to other sites.
                                           18-8

-------
18.2   Meteorological Characterization
       The following sections characterize the meteorological conditions near the monitoring
sites in New York on sample days, as well as over the course of the year.

18.2.1  Climate Summary
       Weather conditions are somewhat variable in New York City as frontal systems
frequently affect the area. Precipitation is spread fairly evenly throughout the year, with
thunderstorms in the summer and fall and more significant rain or snow events in the winter and
spring. The proximity to the Atlantic Ocean offers a moderating influence from cold outbreaks as
well as the summertime heat. The urban heat island effect tends to keep the city warmer than
outlying areas. Both influences result in a relatively  small diurnal range of temperatures. In
addition, air sinking down from the mountains to the west can help drive temperatures higher
during warm spells (Bair, 1992).

       Rochester is located in western New York and borders Lake Ontario's south side.
Elevation increases significantly from the shore to the southern-most parts of the city, rising over
800 feet. While the lake acts as a moderating influence on the city's temperatures, both in the
summer and the winter, it also plays a major factor in the city's precipitation patterns. Lake
effect snow enhances the area's snowfall totals, although snowfall rates tend to be higher near
Lake Ontario than farther inland. Spring and summer tend to be sunny while cloudy conditions
are prevalent in the fall and winter (Bair, 1992 and NOAA, 2012).

18.2.2  Meteorological Conditions in 2011
       Hourly meteorological data from NWS weather stations nearest these sites were retrieved
for 2011  (NCDC, 2011). The closest weather stations are located at Central Park (near MONY)
and Greater Rochester International Airport (near ROCH), WBAN 94728 and 14768,
respectively. Additional information about these weather stations, such as the distance between
the sites and the weather stations, is provided in Table 18-3. These data were used to determine
how meteorological conditions on sample days vary from conditions experienced throughout the
year.
                                          18-9

-------
                           Table 18-3. Average Meteorological Conditions near the New York Monitoring Sites
Closest NWS
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(ฐF)
Average
Temperature
(ฐF)
Average
Dew Point
Temperature
(ฐF)
Average
Wet Bulb
Temperature
(ฐF)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
New York City, New York - MONY
Central Park
94728
(40.78,73.97)
4.35
miles
199ฐ
(SSW)
Sample
Day
2011
62.6
ฑ4.5
63.1
+ 1.8
55.4
ฑ4.2
56.1
+ 1.7
42.4
ฑ4.9
43.6
+ 2.0
49.4
ฑ4.0
50.2
+ 1.6
64.6
ฑ4.1
65.5
+ 1.7
1015.6
ฑ1.9
1015.5
+ 0.8
5.2
ฑ0.6
5.0
+ 0.2
Rochester, New York - ROCH |
Greater
Rochester Intl.
Airport
14768
(43.12, -77.68)
6.46
miles
240ฐ
(WSW)
Sample
Day
2011
56.4
ฑ5.0
57.8
+ 2.0
48.6
ฑ4.6
49.7
+ 1.9
39.0
ฑ4.4
40.1
+ 1.8
44.0
ฑ4.2
45.1
+ 1.7
72.0
ฑ2.7
71.8
+ 1.1
1015.8
ฑ1.8
1015.3
+ 0.8
7.5
ฑ0.8
7.2
+ 0.3
oo
I


o
        Sample day averages are highlighted to help differentiate the sample day averages from the full-year averages.

-------
       Table 18-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for the entire year for 2011. Also included in Table 18-3 is the
95 percent confidence interval for each parameter. Table 18-3 shows that meteorological
conditions near MONY on sample days were representative of average weather conditions
experienced throughout the year. Conditions on sample days near ROCH appear slightly cooler
than conditions experienced throughout the year, although the difference is not statistically
significant.

18.2.3 Back Trajectory Analysis
       Figure 18-5 is the composite back trajectory map for days on which samples were
collected at the MONY monitoring site in 2011. Included in Figure 18-5 are four back
trajectories per sample day. Figure 18-6 is the corresponding cluster analysis. Similarly,
Figure 18-7 is the composite back trajectory map for days on which samples were collected at
ROCH and Figure 18-8 is the corresponding cluster analysis. An in-depth description of these
maps and  how they were generated is presented in Section 3.5.2.1. For the composite maps, each
line represents the 24-hour trajectory along which a parcel of air traveled toward the monitoring
site on a given sample day and time, based on an initial height of 50 meters AGL. For the cluster
analyses, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric  circle around the sites in Figures 18-5 through 18-8 represents
100 miles.
                                          18-11

-------
Figure 18-5. 2011 Composite Back Trajectory Map for MONY
    Figure 18-6. Back Trajectory Cluster Map for MONY
                                         a    HO
                          18-12

-------
Figure 18-7. 2011 Composite Back Trajectory Map for ROCH
    Figure 18-8. Back Trajectory Cluster Map for ROCH
                          18-13

-------
Observations from Figures 18-5 and 18-6 for MONY include the following:

•  Back trajectories originated from a variety of directions at MONY. Back trajectories
   most frequently originated to the northwest of MONY.

•  The 24-hour air shed domain for MONY is similar in size to other NMP sites.
   Although the farthest away a back trajectory originated was over Ontario, Canada, or
   less than 650 miles away, the average trajectory length was 254 miles and 93 percent
   of trajectories originated within 400 miles of the site.

•  The cluster analysis shows that approximately 20 percent of back trajectories
   originated to the northwest of MONY over the Great Lakes and Canada. Twelve
   percent originated over the offshore waters of the Mid-Atlantic states. Shorter back
   trajectories, generally 200 miles in length or less, originating between these two
   cluster trajectories are  represented by the short cluster trajectory originating over
   southeast Pennsylvania. This cluster trajectory also includes a few longer trajectories
   originating over Kentucky, West Virginia, and Virginia. Another 15 percent of back
   trajectories originated to the east of MONY over the Atlantic Ocean and 13 percent of
   back trajectories originated to the north of the site, primarily over northern New
   York, Vermont, and Quebec, Canada.


Observations from Figures 18-7 and 18-8 for ROCH include the following:

•  Back trajectories originated from a variety of directions at ROCH, although relatively
   few originated from the northeast of ROCH.

•  The 24-hour air shed domain for ROCH was comparable in size to MONY and other
   NMP sites. The farthest away a trajectory originated was over western Kentucky, or
   approximately 650 miles away. However, the average trajectory length was 252 miles
   and 86 percent of back trajectories originated within 400 miles of the site.

•  The cluster analysis shows that nearly 50 percent of back trajectories originated from
   a direction with a westerly component. These back trajectories are represented by
   three cluster trajectories in Figure 18-8: 1) longer trajectories originating towards
   Lake Michigan (16 percent), 2) longer trajectories originating towards Ohio
   (10 percent), and 3) relatively short trajectories (< 200 miles in length) originating
   from nearly any direction with a westerly component (25 percent). This third cluster
   also includes a few shorter trajectories originating to the south of ROCH.

•  Nearly one-quarter of back trajectories originated over Ontario and Quebec, Canada.
   Nearly 20 percent of back trajectories originated from a direction with an easterly
   component. This cluster trajectory also includes the longer trajectories originating
   offshore,  east  of Long  Island, New York. Seven percent of trajectories originated to
   the south  of ROCH over West Virginia, Virginia, and  Maryland.
                                    18-14

-------
18.2.4 Wind Rose Comparison
       Hourly surface wind data from the weather stations at Central Park (for MONY) and
Greater Rochester International Airport (for ROCH) were uploaded into a wind rose software
program to produce customized wind roses, as described in Section 3.5.2.2. A wind rose shows
the frequency of wind directions using "petals" positioned around a 16-point compass, and uses
different colors to represent wind speeds.

       Figure 18-9 presents a map showing the distance between the NWS station and MONY,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 18-9 also presents three different wind roses for the
MONY monitoring site.  First, a historical wind rose representing 2001 to 2010 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 2011 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2011 is
presented. These can be used to identify the predominant wind speed and direction for 2011 and
determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically. Figure 18-10 presents the distance map and three wind roses
for ROCH.

       Observations from Figure 18-9 for MONY include the following:
       •   The weather station at Central Park is located 4.35 miles south-south west of MONY.
       •   The historical wind rose shows that winds from the west and west-northwest account
          for greater than 25 percent of wind observations. Winds from the northeast and east-
          northeast account for another 15 percent of observations. Calm winds (<2 knots) were
          observed for nearly 13 percent of the hourly measurements near MONY.
       •   The 2011 full-year wind rose shares many similarities with the historical wind rose,
          such as the prominence of winds from the west. However, the percentage of calm
          winds increased for 2011 and the percentages of each of the primary wind directions
          decreased slightly.
       •   The sample day wind patterns resemble the wind patterns on the 2011 full-year and
          historical wind roses, indicating that wind conditions on sample days were similar to
          those experienced throughout 2011 and historically. However, some differences
          include an even higher percentage of calm winds (nearly 19 percent) and fewer
          observations from the west-southwest.
                                          18-15

-------
     Figure 18-9. Wind Roses for the Central Park Weather Station near MONY
Distance between MONY and NWS Station
2001-2010 Historical Wind Rose
                                             WEST!
          2011 Wind Rose
    Sample Day Wind Rose
WEST
                                     18-16

-------
Figure 18-10. Wind Roses for the Greater Rochester International Airport Weather Station
                                   near ROCH
  Distance between ROCH and NWS Station
2001-2010 Historical Wind Rose
            2011 Wind Rose
    Sample Day Wind Rose
                                       18-17

-------
       Observations from Figure 18-10 for ROCH include the following:
       •  The Rochester International Airport weather station is located approximately
          6.5 miles west-southwest of ROCH, with much of the southern half of the city of
          Rochester between them.
       •  The historical wind rose shows that winds from the south-southwest to west were
          frequently observed, accounting for nearly 50 percent of the wind observations. Calm
          winds were observed for less than nine percent of the hourly measurements near
          ROCH, while the strongest winds were most frequently observed with west-
          southwesterly and westerly winds.
       •  The wind patterns shown on the 2011 wind rose are similar to the historical wind
          patterns for ROCH, although south-southwesterly and west-southwesterly winds were
          observed less frequently and a higher percentage of calm winds were observed
          (nearly 12 percent).
       •  The sample day wind patterns  are similar to those shown on the full-year wind rose,
          although the  percentages differ somewhat.

18.3   Pollutants of Interest
       Site-specific "pollutants of interest" were determined for the New York monitoring sites
in order to allow 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. In addition, if any of the NATTS MQO Core Analytes measured by each monitoring site
did not meet the pollutant of interest criteria based on the preliminary risk-based screening, that
pollutant was added to the list of site-specific pollutants of interest. A more in-depth description
of the risk-based screening process is presented in Section 3.2.
       Table 18-4 presents the results of the preliminary risk-based screening process for the
New York monitoring sites. The pollutants that failed at least one screen and contributed to
95 percent of the total failed screens for each monitoring site are shaded. NATTS MQO Core
Analytes are bolded. Thus, pollutants of interest are shaded and/or bolded. MONY and ROCH
both sampled for hexavalent chromium and PAHs.
                                          18-18

-------
Table 18-4. Risk-Based Screening Results for the New York Monitoring Sites
Pollutant
Screening
Value
(Ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
New York City, New York - MONY
Naphthalene
Acenaphthene
Fluorene
Fluoranthene
Hexavalent Chromium
Acenaphthylene
Benzo(a)pyrene
0.029
0.011
0.011
0.011
0.000083
0.011
0.00057
Total
60
22
20
8
3
1
1
115
60
60
60
60
60
50
59
409
100.00
36.67
33.33
13.33
5.00
2.00
1.69
28.12
52.17
19.13
17.39
6.96
2.61
0.87
0.87
52.17
71.30
88.70
95.65
98.26
99.13
100.00

Rochester, New York - ROCH
Naphthalene
Acenaphthene
Fluorene
Fluoranthene
0.029
0.011
0.011
0.011
Total
46
22
20
9
97
58
58
58
58
232
79.31
37.93
34.48
15.52
41.81
47.42
22.68
20.62
9.28
47.42
70.10
90.72
100.00

Observations from Table 18-4 include the following:

•  For MONY, seven pollutants, of which three are NATTS MQO Core Analytes, failed
   screens. The risk-based screening process identified four of the pollutants that failed
   at least one screen as pollutants of interest. Hexavalent chromium and benzo(a)pyrene
   were added as pollutants of interest for MONY because they are NATTS MQO Core
   Analytes, even though they did not contribute to 95 percent of the total failed screens.

•  For ROCH, only four pollutants failed screens.  The risk-based screening process
   identified all of the pollutants that failed a screen as pollutants of interest. This is
   because it took the combined percentages of failed screens to reach the 95 percent
   criteria specified in Section 3.2. Hexavalent chromium and benzo(a)pyrene were
   added  as pollutants of interest for ROCH because they are NATTS MQO Core
   Analytes, even though they did not fail any screens. These two pollutants are not
   shown in Table 18-4 but are shown in subsequent tables in the sections that follow.

•  For both sites, naphthalene, acenaphthene, fluorene, and fluoranthene are the four
   pollutants that failed the most screens. Aside from naphthalene, these pollutants failed
   nearly the same number of screens for these sites.

•  Naphthalene failed the majority of screens for each site. Naphthalene accounts for
   47 percent of failed screens for ROCH and 52 percent of failed screens for MONY.
                                   18-19

-------
18.4   Concentrations
       This section presents various concentration averages used to characterize pollution levels
at the New York monitoring sites. Where applicable, the following calculations and data
analyses were performed: Time period-based concentration averages (quarterly and annual) are
provided for the pollutants of interest for the New York sites, where the data meet the applicable
criteria. Concentration averages for select pollutants are also presented graphically for each site
to illustrate how the sites' concentrations compare to the program-level averages, as presented in
Section 4.1. In addition, concentration averages for select pollutants are presented from previous
years of sampling in order to characterize concentration trends at each site. Additional site-
specific statistical summaries for MONY and ROCH are provided in Appendices M and O.

18.4.1 2011 Concentration Averages
       Quarterly and annual concentration averages were calculated for the pollutants of interest
for each New York site, as described in Section 3.1. The quarterly average of a particular
pollutant is simply the average concentration of the preprocessed daily measurements over a
given calendar quarter. Quarterly average concentrations include the substitution of zeros for all
non-detects. A site must have a minimum of 75 percent valid samples of the total number of
samples possible within a given quarter for a quarterly average to be calculated. An annual
average includes all measured detections and substituted zeros for non-detects for the entire year
of sampling. Annual averages were calculated for pollutants where three valid quarterly averages
could be calculated and where method completeness was greater than or equal to  85 percent, as
presented in Section 2.4. Quarterly and annual  average concentrations for the New York
monitoring sites are presented in Table 18-5, where applicable. Note that if a pollutant was not
detected in a given calendar quarter, the  quarterly average simply reflects "0" because only zeros
substituted for non-detects were factored into the quarterly average concentration.
                                           18-20

-------
  Table 18-5. Quarterly and Annual Average Concentrations of the Pollutants of
                   Interest for the New York Monitoring Sites
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(ng/m3)
2nd
Quarter
Average
(ng/m3)
3rd
Quarter
Average
(ng/m3)
4th
Quarter
Average
(ng/m3)
Annual
Average
(ng/m3)
New York City, New York - MONY
Acenaphthene
Benzo(a)pyrene
Fluoranthene
Fluorene
Hexavalent Chromium
Naphthalene
60/60
59/60
60/60
60/60
60/61
60/60
2.28
ฑ0.39
0.29
ฑ0.06
4.75
ฑ0.75
4.39
ฑ0.65
0.06
ฑ0.01
108.37
ฑ23.45
11.51
ฑ3.31
0.11
ฑ0.03
7.03
ฑ1.89
11.62
ฑ3.36
0.04
ฑ0.01
130.27
ฑ29.50
16.36
ฑ3.22
0.12
ฑ0.06
9.72
ฑ1.95
15.27
ฑ2.87
0.03
ฑ0.01
170.41
ฑ34.78
5.81
ฑ1.49
0.28
ฑ0.08
4.97
ฑ1.27
6.60
ฑ1.42
0.03
ฑ0.01
133.59
ฑ33.66
8.99
ฑ1.80
0.20
ฑ0.04
6.62
ฑ0.89
9.47
ฑ1.55
0.04
ฑ0.01
135.66
ฑ 15.50
Rochester, New York - ROCH
Acenaphthene
Benzo(a)pyrene
Fluoranthene
Fluorene
Hexavalent Chromium
Naphthalene
58/58
42/58
58/58
58/58
41/56
58/58
1.83
ฑ0.69
0.08
ฑ0.06
1.35
ฑ0.42
2.06
ฑ0.55
0.01
ฑ0.01
31.28
ฑ5.74
13.74
ฑ6.91
0.05
ฑ0.03
6.76
ฑ4.56
11.19
ฑ6.11
0.01
ฑ0.01
53.56
ฑ 14.95
27.43
ฑ7.09
0.08
ฑ0.07
12.01
ฑ3.64
21.41
ฑ5.63
0.01
ฑ0.01
103.28
ฑ24.56
5.55
ฑ2.69
0.15
ฑ0.04
2.37
ฑ0.57
4.59
ฑ 1.77
0.01
ฑ0.01
56.69
ฑ 10.08
12.29
ฑ3.55
0.09
ฑ0.03
5.68
ฑ1.76
9.92
ฑ2.81
0.01
ฑ0.01
61.85
ฑ10.12
Observations from Table 18-5 include the following:

•  The annual average concentration of naphthalene is the highest annual average among
   the pollutants of interest for both New York sites. The annual average naphthalene
   concentration for MONY is more than twice the annual average for ROCH.

•  For both sites, the third quarter average concentration of naphthalene is greater than
   the other quarterly averages. For ROCH, the eight highest concentrations of
   naphthalene (those greater than 100 ng/m3) were all measured in July or August. For
   MONY, the maximum naphthalene concentration (330 ng/m3) was measured on
   December 5, 2011, although a similar concentration was also measured on
   July 20, 2011 (318 ng/m3). The seven concentrations greater than 200 ng/m3
   measured at MONY were spread across all four quarters of 2011. This indicates a
   relatively high-level of variability within the measurements for MONY. This is also
   supported by the relatively large confidence intervals calculated for the quarterly
   averages.
                                   18-21

-------
       •   For both New York sites, concentrations of acenaphthene, fluoranthene, and fluorene
          were highest during the warmer months of the year. However, the confidence
          intervals calculated for these sites indicate that there is a high level of variability in
          the measurements. For example, acenaphthene concentrations for ROCH range from
          0.34 ng/m3 to 50.3 ng/m3 with a median concentration of 5.49 ng/m3. The 15 highest
          concentrations were measured at ROCH between June and September while 13 of the
          15 lowest concentrations were measured between January and March or November
          and December.

       •   Conversely, benzo(a)pyrene measurements tended to be measured during the colder
          months of the year, particularly for MONY. The single non-detect of benzo(a)pyrene
          and the measurements less than  0.1 ng/m3 were measured at MONY between April
          and October while all but one of the concentrations greater than 0.25 ng/m3 were
          measured between January and March or October and December.
       Tables 4-9 through 4-12 present the sites with the 10 highest annual average

concentrations for each of the program-level pollutants of interest. Observations for MONY and

ROCH from those tables include the following:

       •   MONY and ROCH appear in Tables 4-11  seven times, appearing among the top 10
          for each PAH with the exception of ROCH's annual average concentration of
          naphthalene.

       •   MONY has the second highest annual average concentration of benzo(a)pyrene, the
          third highest annual average concentration naphthalene, the fourth highest
          concentration of fluorene, and the fifth highest concentration of acenaphthene among
          NMP sites sampling PAHs.

       •   ROCH has the second highest annual average concentrations of acenaphthene and
          fluorene and the ninth highest annual average concentration of benzo(a)pyrene among
          NMP sites sampling PAHs.

       •   MONY has the fourth highest annual average concentration of hexavalent chromium
          among sites sampling this pollutant, while the annual average for ROCH ranks lower
          (19th).


18.4.2  Concentration Comparison

       In order to better illustrate how a site's annual average concentrations compare to the

program-level averages, a site-specific box plot was created for the selected NATTS MQO Core

Analytes listed in Section 3.5.3, where applicable. Thus, box plots for benzo(a)pyrene,

hexavalent chromium, and naphthalene were created for each site. Figures 18-11 through 18-13

overlay the sites' minimum, annual average, and maximum concentrations onto the program-
                                         18-22

-------
level minimum, first quartile, median, average, third quartile, and maximum concentrations, as
described in Section 3.5.3.
      Figure 18-11. Program vs. Site-Specific Average Benzo(a)pyrene Concentrations
                                                  1         1.25
                                               Concentration (ng/m3)
                 Program:   IstQuartile     2ndQuartile     3rdQuartile     4thQuartile     Average
                 Site:      Site Average     Site Minimum/Maximum

  Figure 18-12. Program vs. Site-Specific Average Hexavalent Chromium Concentrations
 RCO
                   0.35
                                   0.1
                                                 3.15
                                             Concentration (
                                                                 3.2
                                                                               3.25
                                                                                               3.3
                 Program:   IstQuartile     2nd Quartile     3rd Quartile     4th Quartile     Average
                 Site:      Site Average     Site Minimum/Maximum
                            o
                                              18-23

-------
Figure 18-13. Program vs. Site-Specific Average Naphthalene Concentrations
                                                    : ProgramMaxConcentration = 779 ng/m-
                                                  i Program Max Concentration = 779 ng/m3
              100       150      200      250      300      350      433      450

                                  Concentration (nf/m3)
         Program:   IstQuartile    2ndQuartile     3rdQuartile     4thQuartile    Average
         Site:     Site Average    Site Minimum/Maximum

   Observations from Figures 18-11 through 18-13 include the following:

   •   Figure 18-11 presents the box plots for benzo(a)pyrene. Note that the program-
       level first quartile for this pollutant is zero and is not visible on the box plots. The
       box plots show that the annual average concentration for MONY is greater the
       annual average concentration for ROCH, although both are greater than the
       program-level average concentration (but just barely for ROCH). Although the
       range of measurements is slightly greater for MONY than ROCH, the maximum
       concentrations of benzo(a)pyrene for both sites are considerably less than the
       maximum concentration measured across the program. Several non-detects of
       benzo(a)pyrene were measured at ROCH while a single non-detect was measured
       at MONY.

   •   Figure 18-12 presents the box plots for hexavalent chromium. The range of
       measurements for MONY is roughly twice the range of measurements for ROCH.
       The annual average concentration for MONY is greater than the program-level
       average concentration and third quartile while the annual average for ROCH is
       less than the program-level median concentration.  Several non-detects of
       hexavalent chromium were measured at ROCH while a single non-detect was
       measured at MONY.

   •   Figure 18-13 presents the box plots for naphthalene. Note that the program-level
       maximum concentration (779 ng/m3) is not shown directly on the box plots
       because the scale of the box plots would be too large to readily observe data
       points at the lower end of the concentration range. Thus, the scale has been
       reduced to 500 ng/m3. Figure 18-13 shows that the annual average naphthalene
       concentration for MONY is twice the annual average concentration for ROCH.
       While the annual average for ROCH is similar to the program-level median
       concentration, the annual average for MONY is greater than the program-level
       third quartile. The maximum naphthalene concentration measured at MONY is
       twice the maximum concentration measured at ROCH, although both are less than
                                   18-24

-------
             the program-level maximum concentration. There were no non-detects of
             naphthalene measured at either site. The minimum concentration of naphthalene
             measured at MONY (53.9 ng/m3) is the highest minimum concentration among
             sites sampling this pollutant.

18.4.3  Concentration Trends
       A site-specific trends evaluation was completed for sites that have sampled one or more
of the selected NATTS MQO Core Analytes for 5 consecutive years or longer, as described in
Section 3.5.4. Although ROCH has sampled hexavalent chromium under the NMP since 2007,
sampling did not begin until late in the year, which does not allow for the statistical metrics to be
calculated. As a result, a trends analysis was not performed because there would be fewer than
5 years of statistical metrics provided. In addition, sampling for PAHs at ROCH did not begin
until 2008. The MONY site began sampling under the NMP in 2010. Thus, a trends analysis was
not conducted for either site.

18.5   Additional Risk-Based Screening Evaluations
       The following risk-based screening evaluations were conducted to characterize risk at
each New York monitoring site. Refer to Sections 3.3 and 3.5.5 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.

18.5.1  Risk-Based Screening Assessment Using MRLs
       A risk-based screening was conducted by comparing the concentration data from the New
York monitoring sites to the ATSDR MRLs, where available. As described in Section 3.3, MRLs
are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 to 14 days); intermediate (exposures  of 15 to 364 days); and chronic (exposures
of 1 year or greater). The preprocessed daily measurements of the pollutants of interest were
compared to the acute MRLs; the quarterly averages  were compared to the intermediate MRLs;
and the annual averages were compared to the chronic MRLs.

       As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2011.
                                         18-25

-------
  18.5.2 Cancer Risk and Noncancer Hazard Approximations
        For the pollutants of interest for the New York sites and where annual average
  concentrations could be calculated, risk was examined by calculating cancer risk and noncancer
  hazard approximations. These approximations can be used as risk estimates for cancer and
  noncancer effects attributable to the pollutants of interest. Although the use of these
  approximations is limited, they may help identify where policy-makers may want to shift or
  confirm their air-monitoring priorities. Refer to Section 3.5.5.2 for an explanation of how cancer
  risk and noncancer hazard approximations are calculated and what limitations are associated with
  them. Annual  averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer
  hazard approximations are presented in Table  18-6, where applicable. Cancer risk
  approximations are presented as probabilities while the noncancer hazard approximations are
  ratios and thus, unitless values.

             Table 18-6. Risk Approximations for the New York Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
New York City, New York - MONY
Acenaphthene
Benzo(a)pyrene
Fluoranthene
Fluorene
Hexavalent Chromium
Naphthalene
0.000088
0.00176
0.000088
0.000088
0.012
0.000034




0.0001
0.003
60/60
59/60
60/60
60/60
60/61
60/60
8.99
ฑ 1.80
0.20
ฑ0.04
6.62
ฑ0.89
9.47
ฑ1.55
0.04
ฑ0.01
135.66
ฑ 15.50
0.79
0.35
0.58
0.83
0.49
4.61




<0.01
0.05
Rochester, New York - ROCH
Acenaphthene
Benzo(a)pyrene
Fluoranthene
Fluorene
Hexavalent Chromium
Naphthalene
0.000088
0.00176
0.000088
0.000088
0.012
0.000034
_



0.0001
0.003
58/58
42/58
58/58
58/58
41/56
58/58
12.29
ฑ3.55
0.09
ฑ0.03
5.68
ฑ1.76
9.92
ฑ2.81
0.01
ฑ0.01
61.85
ฑ10.12
1.08
0.16
0.50
0.87
0.14
2.10
_



0.01
0.02
- = a Cancer URE or Noncancer RfC is not available
                                           18-26

-------
Observations for the New York sites from Table 18-6 include the following:
       •  Naphthalene has the highest annual average concentration for both sites.
       •  Naphthalene also has the highest cancer risk approximations for both sites
          (4.61 in-a-million) for MONY and 2.10 in-a-million for ROCH). For MONY, this is
          the only pollutant of interest with a cancer risk approximation greater than
          1-in-a-million. For ROCH, acenaphthene also has a cancer risk approximation greater
          than 1 in-a-million (1.08 in-a-million).
       •  None of the pollutants of interest for either New York monitoring site have noncancer
          hazard approximations greater than 1.0, indicating that no adverse health effects are
          expected from these individual pollutants.

18.5.3 Risk-Based Emissions Assessment
       In addition to the  risk-based screenings discussed above, Tables 18-7 and 18-8 present an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 18-7 presents the 10 pollutants with the highest emissions from the 2008 NEI, the
10 pollutants with the highest toxicity-weighted emissions, and the 10 pollutants with the highest
cancer risk approximations (in-a-million), as calculated from the annual averages provided in
Table 18-6. Table 18-8 presents similar information, but identifies the 10 pollutants with the
highest noncancer hazard approximations (HQ), also calculated from annual averages provided
in Table  18-6.
       The pollutants listed in Tables 18-7 and 18-8 are limited to those that have cancer and
noncancer toxicity factors, respectively. As a result, although the actual value of the emissions is
the same, the highest emitted pollutants in the cancer table may be different from the noncancer
table. The cancer risk and noncancer hazard approximations based on each site's annual averages
are limited to those pollutants for which each respective site sampled. As discussed in
Section 18.3, both New York sites sampled PAHs and hexavalent chromium. In addition, the
cancer risk and noncancer hazard approximations are limited to those pollutants with enough
data to meet the criteria for annual averages to be calculated. A more in-depth discussion of this
analysis is provided in Section 3.5.5.3. Similar to the cancer risk and noncancer hazard
approximations, this analysis may help policy-makers prioritize their air monitoring activities.
                                          18-27

-------
  Table 18-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for

                                                the New York Monitoring Sites
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
New York City, New York (Bronx County) - MONY
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
Naphthalene
1,3 -Butadiene
Dichloromethane
POM, Group 2b
Tetrachloroethylene
POM, Group la
225.51
171.25
104.61
62.07
23.44
22.86
12.53
2.17
1.28
0.93
Benzene
Formaldehyde
Naphthalene
1,3 -Butadiene
Ethylbenzene
POM, Group 3
Hexavalent Chromium, PM
Arsenic, PM
POM, Group 2b
Acetaldehyde
1.76E-03
1.36E-03
7.97E-04
6.86E-04
4.28E-04
3.35E-04
3.18E-04
2.23E-04
1.91E-04
1.37E-04
Naphthalene
Fluorene
Acenaphthene
Fluoranthene
Hexavalent Chromium
Benzo(a)pyrene
4.61
0.83
0.79
0.58
0.49
0.35

Rochester, New York (Monroe County) - ROCH
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Dichloromethane
Naphthalene
POM, Group 2b
Tetrachloroethylene
POM, Group 6
397.57
198.79
197.58
106.73
53.16
47.06
29.93
6.47
1.62
0.67
Benzene
Formaldehyde
1,3 -Butadiene
POM, Group 3
Naphthalene
Hexavalent Chromium, PM
Arsenic, PM
POM, Group 2b
Ethylbenzene
POM, Group 5a
3.10E-03
2.57E-03
1.59E-03
1.36E-03
1.02E-03
7.52E-04
5.79E-04
5.70E-04
4.97E-04
3.51E-04
Naphthalene
Acenaphthene
Fluorene
Fluoranthene
Benzo(a)pyrene
Hexavalent Chromium
2.10
1.08
0.87
0.50
0.16
0.14

oo

to
oo

-------
  Table 18-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer

                                           RfCs for the New York Monitoring Sites
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations Based
on Annual Average Concentrations
(Site-Specific)
Noncancer
Hazard
Approximation
Pollutant (HQ)
New York City, New York (Bronx County) - MONY
Methanol
Toluene
Xylenes
Benzene
Hexane
Ethylene glycol
Ethylbenzene
Formaldehyde
Acetaldehyde
Hydrochloric acid
808.80
801.90
456.53
225.51
197.82
194.12
171.25
104.61
62.07
55.88
Acrolein
1,3 -Butadiene
Formaldehyde
Naphthalene
Benzene
Acetaldehyde
Xylenes
Cadmium, PM
Arsenic, PM
Hydrochloric acid
999,876.74
11,427.60
10,674.18
7,813.40
7,517.03
6,896.87
4,565.34
3,703.33
3,455.69
2,793.99
Naphthalene 0.05
Hexavalent Chromium <0.01

Rochester, New York (Monroe County) - ROCH
Toluene
Xylenes
Methanol
Benzene
Hexane
Hydrochloric acid
Ethylbenzene
Formaldehyde
Acetaldehyde
Ethylene glycol
944.15
669.05
479.68
397.57
220.52
200.68
198.79
197.58
106.73
104.78
Acrolein
1,3 -Butadiene
Formaldehyde
Benzene
Acetaldehyde
Hydrochloric acid
Naphthalene
Arsenic, PM
Chlorine
Manganese, PM
641,820.42
26,578.28
20,160.87
13,252.21
11,858.72
10,033.82
9,976.12
8,971.06
7,563.24
7,454.48
Naphthalene 0.02
Hexavalent Chromium <0.01

oo

to
VO

-------
Observations from Table 18-7 include the following:

•  Benzene, ethylbenzene, and formaldehyde are the highest emitted pollutants with
   cancer UREs in Bronx and Monroe Counties.

•  Benzene and formaldehyde are the pollutants with the highest toxicity-weighted
   emissions (of the pollutants with cancer UREs) for both New York counties.

•  Seven of the highest emitted pollutants also have the highest toxicity-weighted
   emissions for Bronx County; six of the highest emitted pollutants also have the
   highest toxicity-weighted emissions for Monroe County.

•  Naphthalene, which has the highest cancer risk approximation for both sites, appears
   on both emissions-based lists. Hexavalent chromium appears among the pollutants
   with the highest toxicity-weighted emissions for both counties, but is not among the
   highest emitted.

•  Emissions of several POM Groups rank among the highest emitted pollutants as well
   as the highest toxicity-weighted emissions for both New York counties. POM, Group
   2b appears on both emissions-based lists for both counties and includes several PAHs
   sampled for at these sites, including acenaphthene, fluoranthene, and fluorene. POM,
   Group 5a has one of the highest toxicity-weighted emissions for Monroe County and
   includes benzo(a)pyrene. POM, Group 6, which ranks 10th for quntity emitted for
   Monroe County, includes benzo(a)anthracene, benzo(b)fluoranthene,
   benzo(k)fluoranthene, and indeno(l,2,3-cd)pyrene.


Observations from Table 18-8 include the following:

•  Methanol, toluene, and xylenes are the highest emitted pollutants with noncancer
   RfCs in both Bronx and Monroe Counties, although not necessarily in that order.

•  The pollutants with the highest toxicity-weighted emissions (of the pollutants with
   noncancer RfCs) are acrolein, 1,3-butadiene, and formaldehyde for both counties.

•  Five of the highest emitted pollutants in Bronx County are also among the pollutants
   with the highest toxicity-weighted emissions; four of the highest emitted pollutants in
   Monroe County are also among the pollutants with the highest toxicity-weighted
   emissions.

•  Naphthalene is among the pollutants with the highest toxicity-weighted emissions for
   each county, but is not among the highest emitted pollutants with a noncancer toxicity
   factor for either county. Hexavalent chromium does not appear on either emissions-
   based list for either New York county. None of the other pollutants of interest for
   either site have noncancer RfCs.
                                   18-30

-------
18.6   Summary of the 2011 Monitoring Data for MONY and ROCH

       Results from several of the data treatments described in this section include the
following:

       ปซป  Seven pollutants failed screens for MONY while four pollutants failed screens for
          ROCH.  The same four pollutants were identified as pollutants of interest via the risk-
          based screening process for MONY and ROCH.

       ปซป  Naphthalene had the highest annual average concentration among the pollutants of
          interest for both sites.

       ปซป  For both sites, concentrations ofacenaphthene, fluoranthene, andfluorene were
          highest during the warmer months of the year while concentrations of benzo(a)pyrene
          were highest during the colder months of the year.
                                          18-31

-------
19.0   Sites in Oklahoma
       This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the UATMP sites in Oklahoma, and integrates these concentrations
with emissions, meteorological, and risk information. Data generated by sources other than ERG
are not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) for detailed discussions and definitions
regarding the various data analyses presented below.

19.1   Site Characterization
       This section characterizes the Oklahoma monitoring sites by providing geographical and
physical information about the locations of the sites and the surrounding areas. This information
is provided to give the reader insight regarding factors that may influence the air quality near the
sites and assist in the interpretation of the ambient monitoring measurements.

       Two Oklahoma sites (TOOK and TMOK) are located in the Tulsa, Oklahoma MSA.
Another site,  PROK, is located east of the Tulsa area in Pry or Creek, Oklahoma. There are also
two sites in the Oklahoma City, Oklahoma MSA; one site is located in Oklahoma City (OCOK)
and another is located just outside Oklahoma City in Midwest City (MWOK).

       Figures 19-1 and 19-2 are composite satellite images retrieved from ArcGIS Explorer
showing the Tulsa monitoring sites in their urban locations. Figure 19-3 identifies nearby point
source emissions locations by source category, as reported in the 2008 NEI for point sources.
Note that only sources within 10 miles of the sites are included in the facility counts provided in
Figure 19-3. A 10-mile boundary was chosen to give the reader an indication of which emissions
sources and emissions source categories could potentially have a direct effect on the air quality at
the monitoring sites. Further, this boundary provides both the proximity of emissions sources to
the monitoring sites as well as the quantity of such sources within a given distance of the sites.
Sources outside the 10-mile radii are still visible on the map, but have been grayed  out in order
to show emissions sources just outside the boundary. Figures 19-4 through 19-8 are the
composite satellite maps and emissions source maps for the Pry or Creek and Oklahoma City
sites. Table 19-1 provides supplemental geographical information such as land use, location
setting, and locational coordinates.
                                          19-1

-------
                                      Figure 19-1. Tulsa, Oklahoma (TOOK) Monitoring Site
to

-------
Figure 19-2. Tulsa, Oklahoma (TMOK) Monitoring Site

-------
Figure 19-3. NEI Point Sources Located Within  10 Miles of TMOK and TOOK
Legend
 @  TMOK UATMP site  ฎ   TOOK UATMP site

Source Category Group (No. of Facilities)
 •J"   Aerospace/Aircraft Manufacturing (1)
 -fr1   Aircraft Operations (14)
 W   Automobile/Truck Manufacturing (1)
  *   Electricity Generation via Combustion (2)
 W   Glass Manufacturing (2)
                                                     ee-O'trw        es-sso-w
                                                 Note: Due to facility density and collocation, the total facilities
                                                 displayed may not represent all facilities within the area of interest
                                                       10 mile radius
County boundary
                                                ?   Miscellaneous Commercial/Industrial (1)
                                                H   Municipal Waste Combustor (1)
                                                *   Petroleum Refinery (2)
                                                7   Portland Cement Manufacturing (1)
                                                V   Steel Mill (1)
                                                *   Transportation and Marketing  of Petroleum Products (1)
                                               19-4

-------
Figure 19-4. Pryor Creek, Oklahoma (PROK) Monitoring Site

-------
Figure 19-5. NEI Point Sources Located Within 10 Miles of PROK
Legend
 ฉ  PROK UATMP site
                          fl6"25'0"W       fl5-2ff(rw       95115'0"W       95 •1 ffO'W       95' 5'CTW
                                      Note: Due to facility density and collocation, the total facilities
                                      displayed may not represent all facilities within the area of interest.
                                 0   10 mile radius
County boundary
                 Source Category Group (No. of Facilities)
                  41   Aircraft Operations (5)
                  c   Chemical Manufacturing (1)
                  *   Electricity Generation via Combustion (2)
                  F   Food Processing/Agriculture (1)
                  7   Portland Cement Manufacturing (1)
                                     19-6

-------
                            Figure 19-6. Midwest City, Oklahoma (MWOK) Monitoring Site
V ป >^                 -y  I 'LfjA—m
u '//

-------
                                 Figure 19-7. Oklahoma City, Oklahoma (OCOK) Monitoring Site
oo
                                  .A

-------
Figure 19-8. NEI Point Sources Located Within 10 Miles of MWOK and OCOK
                97-45'0-W     97"40'0*W     97'35'Q*W     S7i3C<10"W     S/'JS'O'W     97'20'0"W
                                                Note: Due to facility density and collocation. th9 total facilities
   Legend
     Sfe   MWOK UATMP site
             displayed may not represent all facilities within the area of interest.

OCOK UATMP site        10 mile radius |     | County boundary
                         Source Category Group (No. of Facilities)
                           41   Aircraft Operations (19)
                           IB   Bakery (1)
                           F   Food Processing/Agriculture (1)
                           A   Military Base/National Security Facility (1)
                           •   Oil and/or Gas Production (2)
                           P   Printing/Publishing (1)
                                              19-9

-------
                             Table 19-1. Geographical Information for the Oklahoma Monitoring Sites
Site
Code
TOOK
TMOK
PROK
MWOK
OCOK
AQS Code
40-143-0235
40-143-1127
40-097-0187
40-109-0041
40-109-1037
Location
Tulsa
Tulsa
Pryor
Creek
Midwest
City
Oklahoma
City
County
Tulsa
Tulsa
Mayes
Oklahoma
Oklahoma
Micro- or
Metropolitan
Statistical Area
Tulsa, OK MSA
Tulsa, OK MSA
Not in an MSA
Oklahoma City,
OK MSA
Oklahoma City,
OK MSA
Latitude
and
Longitude
36.126945,
-95.998941
36.204902,
-95.976537
36.292941,
-95.303409
35.437641,
-97.387254
35.614131,
-97.475083
Land Use
Industrial
Residential
Industrial
Commercial
Residential
Location
Setting
Urban/City
Center
Urban/City
Center
Suburban
Urban/City
Center
Suburban
Additional Ambient Monitoring Information1
SO2 and H2S.
CO, SO2, NOy, NO, NO2, NOX, O3, Meteorological
parameters, PM10, PM Coarse, PM25, and PM25
Speciation.
None.
None.
SO2, NO, NO2, NOX, O3, Meteorological parameters,
PMio, PM2 5, and PM2 5 Speciation.
Data for additional pollutants are reported to AQS for these sites (EPA, 2012c); however, these data are not generated by ERG and are therefore not included in this report
VO



o

-------
       TOOK is located in West Tulsa, on the southwest side of the Arkansas River. The site is
located in the parking lot of the Public Works building. The monitoring site is positioned
between the Arkansas River and 1-244, which runs parallel to Southwest Boulevard. The
surrounding area is primarily industrial. As shown in Figure  19-1, an oil refinery is located just
south of the site. Another refinery is located to the northwest of the site, on the other side of
1-244. A rail yard is located on the opposite side of 1-244.

       TMOK is located in north Tulsa on the property of Fire Station Number 24. As shown in
Figure 19-2, the intersection of North Peoria Avenue (Highway 11) and East 36th Street North
lies just to the northeast of the site. The surrounding area is primarily residential, with wooded
areas just to the east, an early childhood education facility and an elementary  school to the south,
and a park to the west.

       Figure 19-3 shows that the Tulsa sites are located approximately 5 miles apart, with
TMOK to the north and TOOK to the south. Most of the emissions sources are clustered around
TOOK, while there are no point sources within a couple miles of TMOK.  The source category
with the greatest number of sources surrounding the Tulsa sites is the aircraft operations source
category, which includes airports as well as small runways, heliports, or landing pads. Point
sources closest to TOOK include petroleum  refineries, a municipal waste combustor, and a
facility generating electricity via combustion.

       PROK is located on the eastern edge of the town of Pry or Creek, on the property of Pry or
Creek High School. Residential areas are located to the northwest, west, and south of the site,
while agricultural areas are located to the east, as shown in Figure 19-4. The monitoring site is
located due north (and downwind) of an industrial  park located a few miles to the south.
Figure 19-5 shows  that there are relatively few emissions sources surrounding PROK and that
the aircraft operations source category has the greatest number of emissions sources near the site.
An aircraft operations facility is located a quarter mile north  of PROK but is located under the
site symbol in Figure 19-5. The aforementioned industrial park is represented in Figure 19-5 by
the chemical manufacturing and food processing/agriculture  facilities located to the south of
PROK.
                                          19-11

-------
       The MWOK monitoring site is located in Midwest City, southeast of Oklahoma City. The
site is located in a commercial area on South Midwest Boulevard just north of 1-40. This site is
located at a school enrollment center just north of Tinker Air Force Base, the northern portion of
which can be seen just south of 1-40 in Figure  19-6. Residential areas are located to the northwest
and north, while an extension of Tinker AFB is located to the east.

       OCOK is located in northern Oklahoma City, on the property of Oklahoma Christian
University of Science and Arts.  The site is located in the northwest corner of the University, near
the athletic fields. The areas surrounding the university are primarily residential. Heavily
traveled roadways such as 1-35 and 1-44 to the east and John Kilpatrick Turnpike to the south are
within a few miles of the site, although outside the boundaries of Figure  19-7.

       Figure 19-8 shows that MWOK and  OCOK are approximately 13 miles apart and that
most of the point sources located within 10 miles of them are located between the sites in the
center of Oklahoma City (west and northwest of MWOK and south of OCOK). The source
category with the greatest number of sources surrounding the two sites is the aircraft operations
source category. The point source closest to MWOK is the military base; the source closest to
OCOK is a heliport.

       Table 19-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Oklahoma monitoring sites. Table 19-2 includes county-level
population and vehicle registration information. Table 19-2 also includes a county-level vehicle
registration-to-population ratio,  which was calculated to represent the number of vehicles per
person within each monitoring site's residing county. In addition, the population within 10 miles
of each site is presented, based on postal code population data estimates. An estimate of 10-mile
vehicle ownership was determined by applying the county-level vehicle registration-to-
population ratio to the 10-mile population surrounding each monitoring site. Table 19-2 also
contains traffic volume information for each site. Finally, Table 19-2 presents the county-level
daily VMT for Tulsa, Mayes, and Oklahoma Counties.
                                          19-12

-------
     Table 19-2. Population, Motor Vehicle, and Traffic Information for the Oklahoma
                                    Monitoring Sites
Site
TOOK
TMOK
PROK
MWOK
OCOK
Estimated
County
Population1
610,599
41,389
732,371
County-level
Vehicle
Registration2
603,926
39,968
832,160
Vehicles per
Person
(Registration:
Population)
0.99
0.97
1.14
Population
within 10
miles3
453,918
327,223
30,326
376,168
378,154
Estimated
10-mile
Vehicle
Ownership
448,957
323,647
29,285
427,423
429,679
Annual
Average
Daily
Traffic4
63,000
12,600
15,100
40,900
40,900
County-
level Daily
VMT5
20,348,926
1,656,458
27,190,328
Bounty-level population estimates reflect 2011 data from the U.S. Census Bureau (Census Bureau, 2012b)
2County-level vehicle registration reflects 2011 data from the Oklahoma Tax Commission (OKTC, 2011)
310-mile population estimates reflect 2011 data from Xionetic (Xionetic, 2011)
4AADT reflects 2011 data from the Oklahoma DOT (OK DOT, 2011)
5County-level VMT reflects 2011 data from the Oklahoma DOT (OK DOT, 2012)
       Observations from Table 19-2 include the following:

       •  The Mayes County population is significantly lower than the populations for Tulsa
          and Oklahoma County. This is also true of the 10-mile populations. Compared to
          other NMP monitoring sites, the Tulsa and Oklahoma City populations are in the
          middle of the range, while Pry or Creek's population is on the low end.

       •  The Mayes County vehicle registration is also significantly lower than vehicle
          registration for Tulsa and Oklahoma Counties. Similar observations can be made for
          the 10-mile vehicle registration  estimates. These observations are expected given the
          rural nature of the area surrounding PROK compared to the urban location of the
          Tulsa and Oklahoma City sites.  Compared to other NMP  monitoring sites, the vehicle
          ownership estimates follow a  similar pattern as the populations.

       •  The county-level registration-to-population ratios range from 0.97 vehicles per person
          for PROK to 1.14 vehicles per person for MWOK and OCOK.

       •  The traffic volume passing the TMOK site is the lowest among the Oklahoma
          monitoring sites and is similar to the traffic passing the PROK site, while the traffic
          passing by TOOK is the highest of the five sites. The traffic data for TMOK and
          PROK are in the bottom third among NMP sites while the traffic for TOOK is in the
          middle of the range. Although the traffic counts are the same for MWOK and OCOK,
          the counts represent different  locations. The following list provides the roadways or
          intersections from which the traffic data were obtained: TOOK -1-244, north of the
          split with US-75; TMOK - East 36th Street North near North Peoria Avenue; PROK -
          Highway 69, south of Graham Avenue (Highway 20); MWOK -1-40, west of Tinker
          Air Force Base; OCOK - Highway 77 north of Highway  44 (before the bend at West
          33rd Street).
                                          19-13

-------
       •  County-level VMT is greatest for Oklahoma County and ranks 11th compared to other
          NMP sites. VMT is the smallest for Mayes County and is among the smallest VMT
          compared to other sites.

19.2   Meteorological Characterization
       The following sections characterize the meteorological conditions near the monitoring
sites in Oklahoma on sample days, as well as over the course of the year.
19.2.1  Climate Summary
       Tulsa is located in northeast Oklahoma, just southeast of the Osage Indian Reservation,
and along the Arkansas River. Pry or Creek is also in northeast Oklahoma, approximately
30 miles east of Tulsa. Oklahoma City is located in the center of the state. These areas are
characterized by a continental climate, with very warm summers and cool winters. Precipitation
is generally concentrated in the spring and summer months, with spring as the wettest season,
although precipitation amounts generally decrease across the state from east to west. Spring and
summer precipitation usually results from showers and thunderstorms, while fall and winter
precipitation accompanies frontal systems. A southerly wind prevails for much of the year,
bringing warm, moist air northward from the Gulf of Mexico. Oklahoma is part of "Tornado
Alley," where severe thunderstorms capable of producing strong winds, hail, and tornadoes
occur more frequently than other areas around the county; tornadoes are more prevalent here
than any other region in the U.S. (Bair, 1992; NCDC, 2013;  and NOAA, 2013b).

19.2.2  Meteorological Conditions in 2011
       Hourly meteorological data from NWS weather stations nearest these sites were retrieved
for 2011 (NCDC, 2011). The closest weather stations to the Tulsa sites are located at Richard
Lloyd Jones Jr. Airport (near TOOK) and Tulsa International Airport (near TMOK), WBAN
53908 and 13968, respectively. The closest weather station to the Pryor Creek site is located at
Claremore Regional Airport, WBAN 53940. The two closest weather stations to the Oklahoma
City sites are located at Tinker Air Force Base Airport (near MWOK) and Wiley Post Airport
(near OCOK), WBAN 13919 and 03954, respectively. Additional information about these
weather stations, such as the distance between the sites and the weather stations, is provided in
Table 19-3. These data were used to determine how meteorological conditions on sample days
vary from conditions experienced throughout the year.
                                         19-14

-------
       Table 19-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for the entire year for 2011. Also included in Table 19-3 is the
95 percent confidence interval for each parameter. As shown in Table 19-3, average
meteorological conditions on sample days were representative of average weather conditions
throughout the year for most of the Oklahoma monitoring sites. The differences are greatest for
MWOK, where sample days appear slightly warmer than conditions experienced throughout the
year, but the difference is not statistically significant. Sampling was discontinued at MWOK at
the end of November 2011 in order to move the instruments to a new sampling location, which
likely explains these differences.

19.2.3 Back Trajectory Analysis
       Figure 19-9 is the composite back trajectory map for days on which samples were
collected at the TOOK monitoring site in 2011. Included in Figure 19-9 are four back trajectories
per sample day. Figure 19-10 is the corresponding cluster analysis. Similarly, Figures 19-11
through 19-18 are the composite back trajectory maps for days on which samples were collected
at the remaining Oklahoma sites and the corresponding cluster analyses. An in-depth description
of these maps and how they were generated is presented in Section 3.5.2.1. For the composite
maps, each line represents the 24-hour trajectory along which a parcel of air traveled toward the
monitoring site on a given sample day and time, based on an initial height of 50 meters AGL. For
the cluster analyses, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric circle around the sites in Figures 19-9 through 19-18 represents
100 miles.
                                          19-15

-------
                         Table 19-3. Average Meteorological Conditions near the Oklahoma Monitoring Sites
Closest NWS
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(ฐF)
Average
Temperature
(ฐF)
Average
Dew Point
Temperature
(ฐF)
Average
Wet Bulb
Temperature
(ฐF)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Tulsa, Oklahoma - TOOK
Richard Lloyd
Jones Jr.
Airport
53908
(36.04, -95.98)
6.12
miles
172ฐ
(S)
Sample
Day
2011
74.5
ฑ5.3
74.1
+ 2.2
62.7
ฑ5.2
62.1
+ 2.1
47.0
ฑ4.5
46.8
+ 1.8
53.9
ฑ4.3
53.6
+ 1.7
60.8
ฑ2.9
61.6
+ 1.3
1015.8
ฑ2.1
1015.8
ฑ0.7
6.2
ฑ0.9
5.6
+ 0.3
Tulsa, Oklahoma - TMOK
Tulsa
International
Airport
13968
(36.20, -95.89)
4.81
miles
96ฐ
(E)
Sample
Day
2011
73.4
ฑ5.4
73.3
+ 2.2
62.7
ฑ5.2
62.5
+ 2.1
46.1
ฑ4.3
46.0
+ 1.8
53.4
ฑ4.2
53.3
+ 1.7
58.8
ฑ3.1
59.2
+ 1.4
1014.7
ฑ2.1
1014.6
+ 0.7
9.1
ฑ1.1
8.4
+ 0.4
Pryor Creek, Oklahoma - PROK
Claremore
Regional
Airport
53940
(36.29, -95.47)
8.66
miles
270ฐ
(W)
Sample
Day
2011
70.4
ฑ5.2
71.0
+ 2.1
58.8
ฑ5.1
59.7
+ 2.0
45.9
ฑ4.3
46.9
+ 1.8
51.6
ฑ4.3
52.6
+ 1.7
66.4
ฑ2.9
66.7
+ 1.3
NA
NA
7.5
ฑ1.0
6.8
+ 0.4
Midwest City, Oklahoma - MWOK
Tinker
AFB/Airport
13919
(35.42, -97.39)
1.57
miles
178ฐ
(S)
Sample
Day
2011
76.3
ฑ5.5
74.6
+ 2.2
64.4
ฑ5.3
63.0
+ 2.1
47.0
ฑ4.5
45.8
+ 1.8
54.6
ฑ4.3
53.5
+ 1.7
57.8
ฑ3.6
58.5
+ 1.6
1013.7
ฑ2.0
1014.6
ฑ0.8
11.3
ฑ1.2
10.3
+ 0.4
Oklahoma City, Oklahoma - OCOK
Wiley Post
Airport
03954
(35.53, -97.65)
10.68
miles
240ฐ
(WSW)
Sample
Day
2011
74.5
ฑ5.6
74.5
+ 2.2
63.1
ฑ5.4
63.1
+ 2.2
44.5
ฑ4.3
44.4
+ 1.7
52.8
ฑ4.2
52.8
+ 1.7
55.7
ฑ3.4
55.8
+ 1.6
1014.4
ฑ2.2
1014.3
ฑ0.8
11.9
ฑ1.3
10.9
+ 0.4
VO
      Sample day averages are highlighted to help differentiate the sample day averages from the full-year averages.
      NA = Sea level pressure was not recorded at the Claremore Regional Airport

-------
Figure 19-9. 2011 Composite Back Trajectory Map for TOOK
   Figure 19-10. Back Trajectory Cluster Map for TOOK
                          19-17

-------
Figure 19-11. 2011 Composite Back Trajectory Map for TMOK
    Figure 19-12. Back Trajectory Cluster Map for TMOK
                           19-18

-------
Figure 19-13. 2011 Composite Back Trajectory Map for PROK
    Figure 19-14. Back Trajectory Cluster Map for PROK
                          19-19

-------
Figure 19-15. 2011 Composite Back Trajectory Map for MWOK
    Figure 19-16. Back Trajectory Cluster Map for MWOK
                           19-20

-------
Figure 19-17. 2011 Composite Back Trajectory Map for OCOK
    Figure 19-18. Back Trajectory Cluster Map for OCOK
                           19-21

-------
Observations from Figures 19-9 through 19-18 include the following:

•  The back trajectory maps for the Tulsa sites, the Pry or Creek site, and the Oklahoma
   City sites are similar to each other in back trajectory distribution. This is somewhat
   expected, given their relatively close proximity to each other and the similarity in
   sample days, although not all sites sampled on the exact same days over the period.

•  The air shed domains for the Oklahoma City and Midwest City sites were among the
   largest compared to other NMP sites, ranking second and third, respectively, based on
   average trajectory length. The farthest away a back trajectory originated was over
   northern Minnesota, or approximately 800 miles away, with an average trajectory
   length of nearly 300 miles. The air shed domains for TOOK, TMOK, and PROK
   round out the top 10 sites based on air shed domain size compared to other NMP
   sites.  For these sites, the farthest away a back trajectory originated was over north-
   central South Dakota, or greater than 650 miles away. The average trajectory length
   for these sites ranged from 268 miles (PROK) to 277 miles (TOOK).

•  Each  of the sites exhibits a strong tendency for back trajectories to originate from the
   south-southeast to south-southwest of the sites and from the northwest to northeast of
   the sites.

•  For the Tulsa and Pry or Creek sites, greater than 60 percent of back trajectories
   originated from the southeast to southwest, generally over eastern Texas. Another
   one-third of trajectories originated generally from the west to northwest to north of
   the sites, but of varying lengths. The remaining back trajectories originated from the
   northeast to east, generally over the  state of Missouri.

•  The cluster analysis maps for the Oklahoma City and Midwest City sites are similar
   to each other and to the cluster maps for the Tulsa and Pry or Creeks sites in cluster
   distribution patterns. One difference between the cluster maps for OCOK and
   MWOK is the number of cluster trajectories. Back trajectories originating over the
   panhandles of Texas and Oklahoma are represented by a single cluster for OCOK
   (12 percent). For MWOK, the back trajectories originating over the panhandle of
   Texas are represented by their own cluster trajectory (13 percent), while those
   originating over the panhandle of Oklahoma are included with shorter, northward-
   originating trajectories as well as others originating within the state of Oklahoma.
   These are represented by the cluster trajectory originating just north of the Kansas-
   Oklahoma border (12 percent).
                                   19-22

-------
19.2.4 Wind Rose Comparison
       Hourly surface wind data from the NWS weather stations at Richard Lloyd Jones Junior
Airport (for TOOK), Tulsa International Airport (for TMOK), Claremore Regional Airport (for
PROK), Wiley Post Airport (for OCOK), and Tinker Air Force Base (for MWOK) were
uploaded into a wind rose software program to produce customized wind roses, as described in
Section 3.5.2.2. A wind rose shows the frequency of wind directions using "petals" positioned
around a 16-point compass, and uses different colors to represent wind speeds.

       Figure 19-19 presents a map showing the distance between the NWS station and TOOK,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 19-19 also presents three different wind roses for the
TOOK monitoring site. First, a historical wind rose representing 2001 to 2010 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 2011 is presented.
Next, a wind rose representing wind data for days on which samples  were collected in 2011 is
presented. These can be used to identify the predominant wind speed and direction for 2011 and
determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically. Figures 19-20 through 19-23 present the distance map and
wind roses for the remaining Oklahoma sites.
                                          19-23

-------
Figure 19-19. Wind Roses for the Richard Lloyd Jones Jr. Airport Weather Station near
                                       TOOK
 Distance between TOOK and NWS Station
2001-2010 Historical Wind Rose
           2011 Wind Rose
                          35%

                    *"ปi   28%

                 " "'-, 21%
                    1 4%

                 7%  ';   :    :
                                WIND SPEED
                                (Knot!)
                                n ป=
                                ^| 17 • 21
                                ^| 11 - 17
                                ^| 7- 11
                                n 4.7
                                Hi 2- 4
                                Calms: 25.07%
    Sample Day Wind Rose
                                        19-24

-------
  Figure 19-20. Wind Roses for the Tulsa International Airport Weather Station near
                                      TMOK
Distance between TMOK and NWS Station
2001-2010 Historical Wind Rose
   i *   *
  \>\l   I   •
                                               IWEST
                                                                         30%

                                                                  *"*,,  24%

                                                                "~N   18%
                                                                  1 2%

                                                                s%   ';  ':   :
                                                                               WIND SPEED
                                                                               (Knots)

                                                                               o *=
                                                                               ^| 17 • 21
                                                                               ^| 11 - 17
                                                                               ^| 7- 11
                                                                               n 4-7
           2011 Wind Rose
    Sample Day Wind Rose
                                       19-25

-------
 Figure 19-21. Wind Roses for the Claremore Regional Airport Weather Station near
                                        PROK
Distance between PROK and NWS Station
2003-2010 Historical Wind Rose
                                                                                 WIND SPEED
                                                                                 (Knots)
                                                                                 O *=
                                                                                 ^| 17 • 21
                                                                                 ^| 11 - 17
                                                                                 ^| 7- 11
                                                                                 CH 4-7
                                                                                 !• 2- 4
                                                                                 Calms: 15.89%
           2011 Wind Rose
    Sample Day Wind Rose
'WEST ;
                          25%
                    "**,   20%
                     1 5%
                    1 0%
                                VjlND SPEED
                                (Knots)
                                ^| 17 - 21
                                •I 11 • 17
                                JH 7- 11
                                |  | 4- 7

                                Calms: 19.89ฐ4
                                         19-26

-------
Figure 19-22. Wind Roses for the Tinker Air Force Base Airport Weather Station near
                                    MWOK
Distance between MWOK and NWS Station
2006-2010 Historical Wind Rose
          2011 Wind Rose
    Sample Day Wind Rose
                                     19-27

-------
  Figure 19-23. Wind Roses for the Wiley Post Airport Weather Station near OCOK
Distance between OCOK and NWS Station
2001-2010 Historical Wind Rose
                J*^5 1 "HE VILLAGE

         ^""t,^^  W   w Wiซltiปป •!ป*
        .,j€''''.,   a 4*c™ซa.
           2011 Wind Rose
    Sample Day Wind Rose
                                                 /VEST
                                                                                 WIND SPEED
                                                                                 (Knots)
                                                                                 n -22
                                                                                 H 17 • 21
                                                                                 ^| 11 • 17
                                                                                 ^| 7- 11
                                                                                 n 4-7
                                                                                 ^| 2- 4
                                                                                 Calms: 3.21 %
                                         19-28

-------
Observations from Figures 19-19 through 19-23 include the following:

•  The distance maps show that the distances between the sites and the weather stations
   varies from 1.6 miles between Tinker Air Force Base and MWOK to 10.7 miles
   between OCOK and the Wiley Post Airport.

•  Even though the historical data are from five different weather stations, the wind
   patterns shown on wind roses for the Oklahoma sites are similar to each other. Each
   of the historical wind roses shows that southerly winds prevailed near each Oklahoma
   monitoring site, accounting for roughly one-fifth to one-quarter of the observations
   among the historical time periods. The historical wind roses varied in the percentage
   of calm winds (<2 knots) observed, ranging from as little as three percent at the
   Tinker Air Force Base (MWOK) to as high as 25 percent at the Richard Lloyd Jones
   Jr. Airport (TOOK). Further, calms winds, winds from the south-southeast to south-
   southwest, and winds from the north-northwest to north-northeast account for the
   majority of wind observations while winds from the west or east were rarely observed
   near each site.

•  For TOOK, the 2011 wind patterns are similar to the historical wind patterns, as are
   the sample day wind patterns, although a higher percentage of southerly winds were
   observed on sample days.  These similarities indicate that conditions on sample days
   were  representative of those experienced over the entire year and historically.

•  For TMOK, the 2011 wind patterns resemble the historical wind patterns, although
   there was a slightly higher percentage of southeasterly and south-southeasterly winds
   observed in 2011. The sample day wind patterns are similar to the full-year wind
   patterns, although a higher percentage of southerly winds were observed on sample
   days. These similarities indicate that conditions on sample days were representative
   of conditions experienced throughout the year and historically.

•  For PROK, the historical wind rose includes eight years worth of data. The 2011 wind
   patterns are nearly identical to the historical wind patterns. The one difference is the
   slightly reduced number of south-southwesterly wind observations for 2011, which
   are reflected in the higher calm rate. The sample day wind rose for PROK is similar
   to the historical and full-year wind roses, although a higher percentage of southerly
   winds were observed on sample days. These similarities indicate that conditions on
   sample  days were representative of conditions experienced throughout the year and
   historically.

•  For MWOK, the historical wind rose includes only  five years worth of data. The 2011
   wind patterns resemble the historical wind patterns, although there were slightly
   fewer south-southwesterly wind observations and slightly more northwesterly winds
   observations in 2011. The sample day wind patterns resemble the historical and the
   full-year wind patterns roses, although there are a few subtle differences. The sample
   day wind rose has a higher percentage of southerly winds, as well as a higher
   percentage of south-southeasterly winds. The calm rate for sample days
   (1.36 percent) is roughly half the calm rate on the full-year wind rose (2.96 percent),
   although both are low percentages.
                                   19-29

-------
       •  For OCOK, the wind patterns shown on the 2011 wind rose are nearly identical to the
          historical wind patterns. The sample day wind rose for OCOK is similar to both the
          historical and full-year wind roses, but there was a slightly higher percentage of
          southerly winds observed in 2011 and a slightly reduced calm rate. However, the
          similarities indicate that conditions on sample days were representative of those
          experienced over the entire year and historically.

19.3   Pollutants of Interest
       Site-specific "pollutants of interest" were determined for the Oklahoma monitoring sites
in order to allow 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. In addition, if any of the NATTS MQO Core Analytes measured by each monitoring site
did not meet the pollutant of interest  criteria based on the preliminary risk-based screening, that
pollutant was added to the list of site-specific pollutants of interest. A more in-depth description
of the risk-based screening process is presented in Section 3.2.

       Table 19-4 presents the results of the preliminary risk-based screening process for each
Oklahoma monitoring site. The pollutants that failed at least one screen and contributed to 95
percent of the total failed screens for each monitoring site are shaded. NATTS MQO Core
Analytes are bolded. Thus, pollutants of interest are shaded and/or bolded. The five Oklahoma
sites sampled for VOCs, carbonyl compounds,  and metals (TSP).
                                           19-30

-------
Table 19-4. Risk-Based Screening Results for the Oklahoma Monitoring Sites
Pollutant
Screening
Value
(Ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
Tulsa, Oklahoma - TOOK
Benzene
Carbon Tetrachloride
Acetaldehyde
Formaldehyde
Manganese (TSP)
Arsenic (TSP)
1,3-Butadiene
p-Dichlorobenzene
Ethylbenzene
Nickel (TSP)
Propionaldehyde
Cadmium (TSP)
1 ,2-Dichloroethane
Hexachloro- 1 ,3 -butadiene
Acrylonitrile
1 ,2-Dibromoethane
Cobalt (TSP)
Lead (TSP)
1 , 1 ,2,2-Tetrachloroethane
Xylenes
0.13
0.17
0.45
0.077
0.005
0.00023
0.03
0.091
0.4
0.0021
0.8
0.00056
0.038
0.045
0.015
0.0017
0.01
0.015
0.017
10
Total
57
57
56
56
56
55
51
43
39
18
7
6
6
5
3
2
1
1
1
1
521
57
57
56
56
56
56
52
51
57
56
56
56
6
5
3
2
56
56
1
57
852
100.00
100.00
100.00
100.00
100.00
98.21
98.08
84.31
68.42
32.14
12.50
10.71
100.00
100.00
100.00
100.00
1.79
1.79
100.00
1.75
61.15
10.94
10.94
10.75
10.75
10.75
10.56
9.79
8.25
7.49
3.45
1.34
1.15
1.15
0.96
0.58
0.38
0.19
0.19
0.19
0.19
10.94
21.88
32.63
43.38
54.13
64.68
74.47
82.73
90.21
93.67
95.01
96.16
97.31
98.27
98.85
99.23
99.42
99.62
99.81
100.00

Tulsa, Oklahoma - TMOK
Acetaldehyde
Formaldehyde
Benzene
Carbon Tetrachloride
Manganese (TSP)
Arsenic (TSP)
1,3-Butadiene
Ethylbenzene
ฃ>-Dichlorobenzene
Acrylonitrile
1 ,2-Dichloroethane
Nickel (TSP)
Propionaldehyde
Hexachloro- 1 ,3 -butadiene
Cadmium (TSP)
0.45
0.077
0.13
0.17
0.005
0.00023
0.03
0.4
0.091
0.015
0.038
0.0021
0.8
0.045
0.00056
61
61
60
60
58
56
49
36
24
10
10
10
8
o
J
1
61
61
60
60
58
58
51
60
47
10
10
58
61
3
58
100.00
100.00
100.00
100.00
100.00
96.55
96.08
60.00
51.06
100.00
100.00
17.24
13.11
100.00
1.72
11.94
11.94
11.74
11.74
11.35
10.96
9.59
7.05
4.70
1.96
1.96
1.96
1.57
0.59
0.20
11.94
23.87
35.62
47.36
58.71
69.67
79.26
86.30
91.00
92.95
94.91
96.87
98.43
99.02
99.22
                                 19-31

-------
Table 19-4. Risk-Based Screening Results for the Oklahoma Monitoring Sites (Continued)
Pollutant
Chloromethylbenzene
Cobalt (TSP)
1 ,2-Dibromoethane
Lead (TSP)
Screening
Value
(Ug/m3)
0.02
0.01
0.0017
0.015
Total
#of
Failed
Screens
1
1
1
1
511
#of
Measured
Detections
1
58
1
58
834
%of
Screens
Failed
100.00
1.72
100.00
1.72
61.27
% of Total
Failures
0.20
0.20
0.20
0.20
Cumulative
%
Contribution
99.41
99.61
99.80
100.00

Pryor Creek, Oklahoma - PROK
Acetaldehyde
Formaldehyde
Benzene
Carbon Tetrachloride
Arsenic (TSP)
Manganese (TSP)
p-Dichlorobenzene
1,3-Butadiene
1 ,2-Dichloroethane
Hexachloro- 1 ,3 -butadiene
Acrylonitrile
1, 1,2,2-Tetrachloroethane
Bromomethane
Ethylbenzene
Xylenes
0.45
0.077
0.13
0.17
0.00023
0.005
0.091
0.03
0.038
0.045
0.015
0.017
0.5
0.4
10
Total
58
58
56
56
51
48
36
30
17
4
3
2
1
1
1
422
58
58
56
56
56
56
47
33
17
4
3
2
34
56
56
592
100.00
100.00
100.00
100.00
91.07
85.71
76.60
90.91
100.00
100.00
100.00
100.00
2.94
1.79
1.79
71.28
13.74
13.74
13.27
13.27
12.09
11.37
8.53
7.11
4.03
0.95
0.71
0.47
0.24
0.24
0.24
13.74
27.49
40.76
54.03
66.11
77.49
86.02
93.13
97.16
98.10
98.82
99.29
99.53
99.76
100.00

Midwest City, Oklahoma - MWOK
Acetaldehyde
Benzene
Carbon Tetrachloride
Formaldehyde
Manganese (TSP)
Arsenic (TSP)
1,3-Butadiene
ฃ>-Dichlorobenzene
1 ,2-Dichloroethane
Acrylonitrile
Ethylbenzene
Hexachloro- 1 ,3 -butadiene
Nickel (TSP)
1 ,2-Dibromoethane
cis- 1 , 3 -D ichloropropene
0.45
0.13
0.17
0.077
0.005
0.00023
0.03
0.091
0.038
0.015
0.4
0.045
0.0021
0.0017
0.25
56
56
56
56
52
48
36
26
9
5
5
4
4
1
1
56
56
56
56
56
56
40
50
9
5
56
4
56
1
1
100.00
100.00
100.00
100.00
92.86
85.71
90.00
52.00
100.00
100.00
8.93
100.00
7.14
100.00
100.00
13.43
13.43
13.43
13.43
12.47
11.51
8.63
6.24
2.16
1.20
1.20
0.96
0.96
0.24
0.24
13.43
26.86
40.29
53.72
66.19
77.70
86.33
92.57
94.72
95.92
97.12
98.08
99.04
99.28
99.52
                                      19-32

-------
Table 19-4. Risk-Based Screening Results for the Oklahoma Monitoring Sites (Continued)
Pollutant
trans- 1 , 3 -Dichloropropene
1 , 1 ,2,2-Tetrachloroethane
Screening
Value
(Ug/m3)
0.25
0.017
Total
#of
Failed
Screens
1
1
417
#of
Measured
Detections
1
1
560
%of
Screens
Failed
100.00
100.00
74.46
% of Total
Failures
0.24
0.24
Cumulative
%
Contribution
99.76
100.00

Oklahoma City, Oklahoma - OCOK
Benzene
Acetaldehyde
Formaldehyde
Carbon Tetrachloride
Arsenic (TSP)
Manganese (TSP)
1,3-Butadiene
p-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Acrylonitrile
Propionaldehyde
Hexachloro- 1 ,3 -butadiene
Cadmium (TSP)
1, 1,2,2-Tetrachloroethane
1 ,2-Dibromoethane
0.13
0.45
0.077
0.17
0.00023
0.005
0.03
0.091
0.038
0.4
0.015
0.8
0.045
0.00056
0.017
0.0017
Total
61
60
60
59
56
56
37
16
14
13
9
7
4
2
2
1
457
61
60
60
61
61
61
41
39
14
61
9
60
4
61
2
1
656
100.00
100.00
100.00
96.72
91.80
91.80
90.24
41.03
100.00
21.31
100.00
11.67
100.00
3.28
100.00
100.00
69.66
13.35
13.13
13.13
12.91
12.25
12.25
8.10
3.50
3.06
2.84
1.97
1.53
0.88
0.44
0.44
0.22
13.35
26.48
39.61
52.52
64.77
77.02
85.12
88.62
91.68
94.53
96.50
98.03
98.91
99.34
99.78
100.00

      Observations from Table 19-4 include the following:

      •  Twenty pollutants failed at least one screen for TOOK; 19 pollutants failed screens
         for TMOK; 15 pollutants failed screens for PROK; 17 pollutants failed screens for
         MWOK; and 16 pollutants failed screens for OCOK.

      •  The risk-based screening process identified 11 pollutants of interest for TOOK, of
         which eight are NATTS MQO Core Analytes. Cadmium and lead were added to the
         pollutants of interest for TOOK because they  are NATTS MQO Core Analytes, even
         though they did not contribute to 95 percent of the total failed screens. Five additional
         pollutants (beryllium, chloroform, tetrachloroethylene, trichloroethylene, and vinyl
         chloride) were added to the pollutants of interest for TOOK because they are NATTS
         MQO Core Analytes, even though they did not fail any screens. These five pollutants
         do not appear  in Table  19-4 but are shown in  subsequent tables in the sections that
         follow.

      •  The risk-based screening process identified 12 pollutants of interest for TMOK, of
         which eight are NATTS MQO Core Analytes. Cadmium and lead were added to the
         pollutants of interest for TMOK because they are NATTS MQO Core Analytes, even
         though they did not contribute to 95 percent of the total failed screens. Five additional
         pollutants (beryllium, chloroform, tetrachloroethylene, trichloroethylene, and vinyl
                                         19-33

-------
   chloride) were added to the pollutants of interest for TMOK because they are NATTS
   MQO Core Analytes, even though they did not fail any screens. These five pollutants
   do not appear in Table 19-4 but are shown in subsequent tables in the sections that
   follow.

•  The risk-based screening process identified nine pollutants of interest for PROK, of
   which  seven are NATTS MQO Core Analytes. An additional eight pollutants
   (beryllium, cadmium, chloroform, lead, nickel, tetrachloroethylene, trichloroethylene,
   and vinyl chloride) were added to the pollutants of interest for PROK because they
   are NATTS MQO Core Analytes, even though they did not fail any screens. These six
   pollutants do not appear in Table 19-4 but are shown in subsequent tables in the
   sections that follow.

•  The risk-based screening process identified 11 pollutants of interest for MWOK, of
   which  seven are NATTS MQO Core Analytes. Nickel was added to the pollutants of
   interest for MWOK because it is a NATTS MQO Core Analyte, even though it did
   not contribute to 95 percent of the total failed screens. An additional seven pollutants
   (beryllium, cadmium, chloroform, lead, tetrachloroethylene, trichloroethylene, and
   vinyl chloride) were added to the pollutants of interest for MWOK because they are
   NATTS MQO Core Analytes, even though they did not fail any screens. These seven
   pollutants do not appear in Table 19-4 but are shown in subsequent tables in the
   sections that follow.

•  The risk-based screening process identified 11 pollutants of interest for OCOK, of
   which  seven are NATTS MQO Core Analytes. Cadmium was  added to the pollutants
   of interest for OCOK because it is a NATTS MQO Core Analyte, even though it did
   not contribute to 95 percent of the total failed screens. Seven additional pollutants
   (beryllium, chloroform, lead, nickel,  tetrachloroethylene, trichloroethylene, and vinyl
   chloride) were added to the pollutants of interest for OCOK because they are NATTS
   MQO Core Analytes, even though they did not fail any screens. These pollutants do
   not appear in Table 19-4 but are shown in subsequent tables in the sections that
   follow.

•  Acetaldehyde, benzene, and formaldehyde each failed 100 percent of screens for each
   site.

•  TOOK and TMOK failed the fourth and fifth highest number of screens among all
   NMP sites, although the other Oklahoma sites ranked 10th (OCOK), 11th (PROK), and
   12th (MWOK), as shown in Table 4-8.
                                  19-34

-------
19.4   Concentrations
       This section presents various concentration averages used to characterize pollution levels
at the Oklahoma monitoring sites. Where applicable, the following calculations and data analyses
were performed: Time period-based concentration averages (quarterly and annual) are provided
for the pollutants of interest for the Oklahoma monitoring sites, where the data meet the
applicable criteria. Concentration averages for select pollutants are also presented graphically for
the sites to illustrate how the sites' concentrations compare to the program-level averages, as
presented in Section 4.1.  In addition, concentration averages for select pollutants are presented
from previous years of sampling in order to characterize concentration trends at the sites.
Additional site-specific statistical summaries for each of the Oklahoma sites are provided in
Appendices J, L, and N.

19.4.1 2011 Concentration Averages
       Quarterly and annual concentration averages were calculated for the pollutants of interest
for each Oklahoma site, as described in Section  3.1. The quarterly average of a particular
pollutant is simply the average concentration of the preprocessed  daily  measurements over a
given calendar quarter. Quarterly average concentrations include the substitution of zeros for all
non-detects. A site must have a minimum of 75  percent valid samples of the total number of
samples possible within a given quarter for a quarterly average to be calculated. An annual
average includes all measured detections and substituted zeros for non-detects for the entire year
of sampling. Annual averages were calculated for pollutants where three valid quarterly averages
could be calculated and where method completeness was greater than or equal to 85 percent, as
presented in Section 2.4.  Quarterly and annual average concentrations for the Oklahoma
monitoring sites are presented in Table 19-5, where applicable. Note that concentrations of the
TSP metals  are presented in ng/m3 for ease of viewing. Also note that if a pollutant was not
detected in a given calendar quarter, the  quarterly average simply reflects "0" because only zeros
substituted for non-detects were factored into the quarterly average concentration.
                                          19-35

-------
   Table 19-5. Quarterly and Annual Average Concentrations of the Pollutants of
                     Interest for the Oklahoma Monitoring Sites
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Ug/m3)
2nd
Quarter
Average
(Ug/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Ug/m3)
Tulsa, Oklahoma - TOOK
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
ฃ>-Dichlorobenzene
Ethylbenzene
Formaldehyde
Propionaldehyde
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
Arsenic (TSP)a
Bery Ilium (TSP)a
Cadmium (TSP) a
Lead (TSP) a
Manganese (TSP) a
Nickel (TSP) a
56/56
57/57
52/57
57/57
17/57
51/57
57/57
56/56
56/56
47/57
8/57
3/57
56/56
56/56
56/56
56/56
56/56
56/56
NA
4.03
ฑ2.22
0.11
ฑ0.04
0.47
ฑ0.06
0.03
ฑ0.03
0.11
ฑ0.02
0.52
ฑ0.16
NA
NA
0.12
ฑ0.05
0.01
ฑ0.01
0
0.77
ฑ0.15
0.02
ฑ0.01
0.34
ฑ0.16
5.91
ฑ1.94
21.84
ฑ5.42
1.49
ฑ0.31
2.27
ฑ0.46
4.39
ฑ3.01
0.05
ฑ0.02
0.68
ฑ0.15
0.04
ฑ0.05
0.06
ฑ0.03
0.42
ฑ0.08
4.22
ฑ1.16
0.49
ฑ0.12
0.13
ฑ0.09
0.01
ฑ0.01
O.01
ฑO.01
0.70
ฑ0.14
0.03
ฑ0.01
0.39
ฑ0.21
6.41
ฑ2.21
29.22
ฑ7.69
1.69
ฑ0.41
4.49
ฑ0.95
3.81
ฑ1.53
0.09
ฑ0.02
0.64
ฑ0.03
0.04
ฑ0.04
0.29
ฑ0.06
0.69
ฑ0.24
5.61
ฑ1.21
0.77
ฑ0.18
0.14
ฑ0.09
0.01
ฑ0.01
O.01
ฑO.01
0.77
ฑ0.10
0.03
ฑ0.01
0.18
ฑ0.04
5.43
ฑ1.20
37.56
ฑ8.63
2.06
ฑ0.32
2.11
ฑ0.49
2.33
ฑ0.80
0.12
ฑ0.05
0.68
ฑ0.06
0.06
ฑ0.04
0.16
ฑ0.02
1.03
ฑ0.37
2.16
ฑ0.40
0.37
ฑ0.08
0.11
ฑ0.04
0.02
ฑ0.02
0
0.82
ฑ0.25
0.02
ฑ0.01
0.35
ฑ0.22
5.75
ฑ1.98
29.53
ฑ 13.51
1.70
ฑ0.52
2.75
ฑ0.41
3.59
ฑ0.98
0.09
ฑ0.02
0.63
ฑ0.05
0.04
ฑ0.02
0.15
ฑ0.03
0.68
ฑ0.13
3.74
ฑ0.57
0.51
ฑ0.07
0.12
ฑ0.03
0.01
ฑ0.01
O.01
ฑO.01
0.76
ฑ0.08
0.03
ฑ0.01
0.31
ฑ0.08
5.87
ฑ0.87
30.09
ฑ4.58
1.75
ฑ0.19
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for
ease of viewing.
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
                                       19-36

-------
   Table 19-5. Quarterly and Annual Average Concentrations of the Pollutants of
              Interest for the Oklahoma Monitoring Sites (Continued)
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Ug/m3)
2nd
Quarter
Average
(Ug/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Ug/m3)
Tulsa, Oklahoma - TMOK
Acetaldehyde
Acrylonitrile
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
ฃ>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
Arsenic (TSP)a
Bery Ilium (TSP)a
Cadmium (TSP) a
Lead (TSP) a
Manganese (TSP) a
Nickel (TSP) a
61/61
10/60
60/60
51/60
60/60
28/60
47/60
10/60
60/60
61/61
40/60
10/60
1/60
58/58
58/58
58/58
58/58
58/58
58/58
1.51
ฑ0.27
0.01
ฑ0.02
1.34
ฑ0.31
0.10
ฑ0.03
0.50
ฑ0.07
0.04
ฑ0.02
0.08
ฑ0.03
0
0.48
ฑ0.14
2.15
ฑ0.38
0.08
ฑ0.03
0.02
ฑ0.02
0
0.61
ฑ0.16
0.01
ฑ<0.01
0.27
ฑ0.05
5.08
ฑ1.16
14.93
ฑ5.09
1.21
ฑ0.30
2.00
ฑ0.43
0.16
ฑ0.18
1.14
ฑ0.30
0.04
ฑ0.03
0.61
ฑ0.06
0.05
ฑ0.03
0.05
ฑ0.04
0.03
ฑ0.02
0.40
ฑ0.11
4.44
ฑ1.35
0.04
ฑ0.03
0.02
ฑ0.02
0.01
ฑ0.01
0.59
ฑ0.12
0.02
ฑ0.01
0.20
ฑ0.03
4.76
ฑ1.12
24.93
ฑ6.60
1.58
ฑ0.38
4.05
ฑ0.67
0.02
ฑ0.03
1.54
ฑ0.37
0.13
ฑ0.04
0.65
ฑ0.04
0.15
ฑ0.11
0.11
ฑ0.03
0.01
ฑ0.02
0.63
ฑ0.13
6.04
ฑ1.31
0.07
ฑ0.03
0.01
ฑ0.02
0
0.58
ฑ0.12
0.02
ฑO.01
0.17
ฑ0.04
4.72
ฑ0.89
23.30
ฑ5.19
1.56
ฑ0.33
1.89
ฑ0.45
0.11
ฑ0.22
1.41
ฑ0.48
0.14
ฑ0.07
0.64
ฑ0.06
0.06
ฑ0.04
0.10
ฑ0.03
0.04
ฑ0.03
0.68
ฑ0.26
2.81
ฑ0.58
0.09
ฑ0.03
0.01
ฑ0.02
0
0.75
ฑ0.24
0.01
ฑO.01
0.24
ฑ0.08
5.13
ฑ2.00
18.29
ฑ7.30
1.33
ฑ0.52
2.40
ฑ0.34
0.08
ฑ0.07
1.35
ฑ0.18
0.10
ฑ0.02
0.60
ฑ0.03
0.07
ฑ0.03
0.08
ฑ0.02
0.02
ฑ0.01
0.55
ฑ0.09
3.93
ฑ0.62
0.07
ฑ0.02
0.01
ฑ0.01
0.01
ฑ0.01
0.63
ฑ0.08
0.02
ฑO.01
0.22
ฑ0.03
4.91
ฑ0.64
20.52
ฑ3.06
1.42
ฑ0.19
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for
ease of viewing.
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
                                      19-37

-------
   Table 19-5. Quarterly and Annual Average Concentrations of the Pollutants of
              Interest for the Oklahoma Monitoring Sites (Continued)
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Ug/m3)
2nd
Quarter
Average
(Ug/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Ug/m3)
Pryor Creek, Oklahoma - PROK
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
ฃ>-Dichlorobenzene
1 ,2-Dichloroethane
Formaldehyde
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
Arsenic (TSP)a
Bery Ilium (TSP)a
Cadmium (TSP) a
Lead (TSP) a
Manganese (TSP) a
Nickel (TSP) a
58/58
56/56
33/56
56/56
15/56
47/56
17/56
58/58
28/56
2/56
3/56
56/56
56/56
56/56
56/56
56/56
56/56
1.35
ฑ0.21
0.75
ฑ0.10
0.04
ฑ0.03
0.55
ฑ0.05
0.01
ฑ0.02
0.07
ฑ0.03
0.01
ฑ0.02
2.38
ฑ0.35
0.06
ฑ0.02
0
0
0.57
ฑ0.14
0.02
ฑ0.01
0.23
ฑ0.07
3.16
ฑ0.72
8.28
ฑ2.76
0.75
ฑ0.17
1.52
ฑ0.32
0.60
ฑ0.11
0.01
ฑ0.01
0.63
ฑ0.06
0.06
ฑ0.06
0.07
ฑ0.05
0.02
ฑ0.03
4.52
ฑ1.65
0.02
ฑ0.02
0
0.01
ฑ0.01
0.45
ฑ0.10
0.04
ฑ0.02
0.13
ฑ0.04
2.46
ฑ0.60
11.95
ฑ3.87
0.89
ฑ0.24
2.43
ฑ0.38
0.70
ฑ0.27
0.03
ฑ0.02
0.66
ฑ0.05
0.01
ฑ0.02
0.54
ฑ0.21
0.01
ฑ0.02
5.96
ฑ1.43
0.01
ฑ0.01
0.01
ฑ0.02
0.01
ฑ0.01
0.53
ฑ0.16
0.03
ฑ0.01
0.13
ฑ0.03
2.57
ฑ0.62
15.28
ฑ5.10
0.88
ฑ0.14
1.38
ฑ0.26
0.67
ฑ0.14
0.05
ฑ0.03
0.65
ฑ0.06
0.05
ฑ0.03
0.17
ฑ0.02
0.06
ฑ0.02
2.43
ฑ0.73
0.05
ฑ0.02
O.01
ฑ0.01
0.01
ฑ0.01
0.59
ฑ0.14
0.02
ฑ0.01
0.14
ฑ0.03
2.58
ฑ0.52
9.41
ฑ2.23
0.64
ฑ0.14
1.68
ฑ0.18
0.68
ฑ0.08
0.04
ฑ0.01
0.62
ฑ0.03
0.03
ฑ0.02
0.20
ฑ0.07
0.03
ฑ0.01
3.84
ฑ0.67
0.03
ฑ0.01
O.01
ฑO.01
0.01
ฑ0.01
0.54
ฑ0.07
0.03
ฑ0.01
0.16
ฑ0.02
2.69
ฑ0.30
11.31
ฑ1.88
0.79
ฑ0.09
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for
ease of viewing.
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
                                      19-38

-------
   Table 19-5. Quarterly and Annual Average Concentrations of the Pollutants of
              Interest for the Oklahoma Monitoring Sites (Continued)
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Ug/m3)
2nd
Quarter
Average
(Ug/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Ug/m3)
Midwest City, Oklahoma - MWOK
Acetaldehyde
Acrylonitrile
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
ฃ>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
Arsenic (TSP)a
Bery Ilium (TSP)a
Cadmium (TSP) a
Lead (TSP) a
Manganese (TSP) a
Nickel (TSP) a
56/56
5/56
56/56
40/56
56/56
23/56
50/56
9/56
56/56
56/56
40/56
1/56
1/56
56/56
56/56
56/56
56/56
56/56
56/56
1.52
ฑ0.44
0.01
ฑ0.02
0.84
ฑ0.16
0.09
ฑ0.05
0.49
ฑ0.07
0.04
ฑ0.02
0.09
ฑ0.02
0
0.23
ฑ0.09
2.61
ฑ0.47
0.25
ฑ0.19
0.01
ฑ0.02
0
0.39
ฑ0.08
0.01
ฑ0.01
0.11
ฑ0.02
2.33
ฑ0.34
9.17
ฑ2.63
0.93
ฑ0.24
1.66
ฑ0.27
0.02
ฑ0.02
0.64
ฑ0.09
0.04
ฑ0.02
0.62
ฑ0.06
0.05
ฑ0.03
0.09
ฑ0.03
0.03
ฑ0.02
0.18
ฑ0.03
3.85
ฑ0.96
0.13
ฑ0.08
0
0
0.42
ฑ0.06
0.02
ฑ0.01
0.18
ฑ0.06
2.79
ฑ0.50
15.24
ฑ3.98
1.54
ฑ0.63
2.98
ฑ0.39
0
0.69
ฑ0.11
0.04
ฑ0.02
0.66
ฑ0.06
0.01
ฑ0.02
0.10
ฑ0.03
0
0.25
ฑ0.04
5.99
ฑ0.97
0.08
ฑ0.09
0
0.01
ฑ0.01
0.32
ฑ0.06
0.03
ฑ0.01
0.08
ฑ0.02
2.35
ฑ0.50
16.17
ฑ4.43
1.26
ฑ0.30
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
2.03
ฑ0.24
0.01
ฑ0.01
0.73
ฑ0.07
0.06
ฑ0.02
0.60
ฑ0.03
0.04
ฑ0.01
0.09
ฑ0.01
0.01
ฑ0.01
0.25
ฑ0.04
4.05
ฑ0.53
0.15
ฑ0.06
O.01
ฑO.01
0.01
ฑ0.01
0.40
ฑ0.04
0.02
ฑO.01
0.12
ฑ0.02
2.48
ฑ0.22
13.26
ฑ1.94
1.20
ฑ0.21
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for
ease of viewing.
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
                                      19-39

-------
   Table 19-5. Quarterly and Annual Average Concentrations of the Pollutants of
              Interest for the Oklahoma Monitoring Sites (Continued)
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Ug/m3)
2nd
Quarter
Average
(Ug/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Ug/m3)
Oklahoma City, Oklahoma - OCOK
Acetaldehyde
Acrylonitrile
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
ฃ>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
Arsenic (TSP)a
Bery Ilium (TSP)a
Cadmium (TSP) a
Lead (TSP) a
Manganese (TSP) a
Nickel (TSP) a
60/60
9/61
61/61
41/61
61/61
13/61
39/61
14/61
61/61
60/60
40/61
1/61
3/61
61/61
61/61
61/61
61/61
61/61
61/61
1.44
ฑ0.27
0.02
ฑ0.04
0.75
ฑ0.11
0.04
ฑ0.03
0.46
ฑ0.11
0.03
ฑ0.02
0.04
ฑ0.02
0
0.23
ฑ0.05
2.01
ฑ0.36
0.11
ฑ0.06
0
0
0.47
ฑ0.10
0.02
ฑ0.01
0.15
ฑ0.07
2.87
ฑ0.54
12.42
ฑ3.49
0.85
ฑ0.15
2.50
ฑ0.79
0.23
ฑ0.20
0.70
ฑ0.15
0.01
ฑ0.01
0.61
ฑ0.08
0.02
ฑ0.02
0.01
ฑ0.02
0.03
ฑ0.03
0.25
ฑ0.07
5.70
ฑ2.56
0.04
ฑ0.03
0
0
0.48
ฑ0.11
0.02
ฑ0.01
0.15
ฑ0.09
2.78
ฑ0.48
16.94
ฑ4.71
1.03
ฑ0.20
3.69
ฑ0.50
0.01
ฑ0.02
1.12
ฑ0.79
0.64
ฑ1.21
0.65
ฑ0.05
0.01
ฑ0.01
0.10
ฑ0.03
0.01
ฑ0.01
0.38
ฑ0.06
5.82
ฑ0.91
0.05
ฑ0.03
0
0.01
ฑ0.01
0.36
ฑ0.07
0.02
ฑO.01
0.08
ฑ0.05
2.15
ฑ0.34
16.28
ฑ4.79
0.79
ฑ0.12
1.86
ฑ0.37
0.02
ฑ0.04
0.93
ฑ0.21
0.08
ฑ0.04
0.61
ฑ0.06
0.02
ฑ0.02
0.07
ฑ0.02
0.09
ฑ0.09
0.38
ฑ0.11
2.46
ฑ0.61
0.13
ฑ0.05
0.01
ฑ0.01
0.01
ฑ0.01
0.57
ฑ0.22
0.01
ฑO.01
0.12
ฑ0.03
2.68
ฑ0.73
10.35
ฑ2.51
0.70
ฑ0.10
2.41
ฑ0.33
0.07
ฑ0.05
0.88
ฑ0.21
0.20
ฑ0.31
0.58
ฑ0.04
0.02
ฑ0.01
0.06
ฑ0.01
0.03
ฑ0.02
0.31
ฑ0.04
4.06
ฑ0.80
0.08
ฑ0.02
O.01
ฑO.01
0.01
ฑ0.01
0.47
ฑ0.07
0.02
ฑO.01
0.13
ฑ0.03
2.61
ฑ0.26
14.04
ฑ2.01
0.84
ฑ0.08
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for
ease of viewing.
NA = Not available due to the criteria for calculating a quarterly and/or annual average.
                                      19-40

-------
Observations for all five Oklahoma sites from Table 19-5 include the following:

•  Formaldehyde has the highest annual average concentration by mass for each site,
   followed by acetaldehyde and benzene, with one exception. The annual average
   concentration of benzene is greater than the annual average concentration of
   acetaldehyde for TOOK.

•  The annual average concentrations of formaldehyde range from 3.74 ฑ 0.57 |ig/m3 for
   TOOK to 4.06 ฑ 0.80 |ig/m3 for OCOK, although the annual averages for MWOK
   and OCOK are nearly identical. The annual average concentration of acetaldehyde
   ranged from 1.68 ฑ 0.18 |ig/m3 for PROK to 2.75 ฑ 0.41 |ig/m3  for TOOK.

•  Concentrations of the carbonyl compounds, formaldehyde in particular, tended to be
   highest in the summer months and lowest in the winter months.  However, the
   relatively large confidence intervals associated with these averages indicate that, in
   most cases, the differences are not statistically significant.

•  The annual average concentration of benzene is highest for TOOK and TMOK. These
   are the only two Oklahoma sites for which the annual average is greater than 1 |ig/m3.
   However, the annual average for TOOK (3.59 ฑ 0.98 |ig/m3) is nearly three times the
   annual average for TMOK (1.35 ฑ 0.18 Hg/m3).

•  The annual average concentration of manganese is  the highest of the TSP metals for
   each site, followed by lead and nickel. The annual average  manganese concentrations
   range from 11.31 ฑ 1.88 |ig/m3 for PROK to 30.09 ฑ 4.58 |ig/m3 for TOOK.

•  Concentrations of the TSP metals tended to be higher at TOOK and TMOK than
   PROK, OCOK, and MWOK.


Observations for TOOK from Table 19-5 include the following:

•  The first, second, and third quarter average concentrations of benzene for TOOK have
   relatively large confidence intervals associated with them, particularly the second
   quarter. The maximum benzene concentration was  measured on April 3, 2011
   (23.8  |ig/m3) and is nearly twice the next highest concentration of benzene
   (12.6  |ig/m3), measured on January 15, 2011.  The April 3rd measurement is also the
   maximum concentration of benzene measured across the program.  Of the four
   concentrations of benzene greater than 10 |ig/m3 measured  at TOOK, two were
   measured in the first quarter and one each in the second and third quarters. Of the
   27 concentrations of benzene greater than 4 |ig/m3 measured across the program, 19
   were measured at TOOK. No other NMP site had more than two benzene
   measurements greater than 4 |ig/m3 (OCOK had one). Similar observations for the
   benzene measurements collected at TOOK were made in the 2010 NMP report.
                                  19-41

-------
•  Concentrations ofp-dichlorobenzene were highest in the third quarter of 2011 at
   TOOK. A review of the data shows that the four highest concentrations of this
   pollutant were measured in July 2011  and of the 15 concentrations greater than
   0.2 |ig/m3,  12 were measured between July and September.

•  The fourth quarter average concentration of ethylbenzene is higher than the other
   quarterly averages. A review of the data shows that six of the seven highest
   concentrations measured at TOOK were measured during the fourth quarter of 2011,
   including the only two concentrations greater than 2 |ig/m3.

•  Although the third quarter average concentration of manganese is greater than the
   fourth quarter average, the fourth quarter has a larger confidence interval  associated
   with it. A review of the data shows that the maximum concentration of manganese
   was measured at TOOK on October 6, 2011 (104 ng/m3). The second highest
   concentration was measured on the  previous sample day, September 30, 2011
   (64.5 ng/m3). Of the 14 manganese  concentrations greater than 40 ng/m3,  nine were
   measured during the third quarter of 2011, while only one was measured during the
   fourth quarter. Thus, manganese concentrations were generally higher during the third
   quarter, while a single outlier is driving the fourth quarter average concentration.

•  TOOK does not  have first quarter averages for the carbonyl  compounds due to
   complications at the site during this time frame.


Observations for TMOK from Table 19-5 include the following:

•  Each of the quarterly average concentrations of acrylonitrile has a confidence interval
   greater than the average itself, particularly the second and fourth quarters. A review
   of the data shows that concentrations of acrylonitrile spanned an order of magnitude,
   ranging from 0.109  |ig/m3 to 1.69 |ig/m3. Two concentrations greater than 1 |ig/m3
   were measured at TMOK, one in June and one in December. The December 29, 2011
   measurement (1.69  |ig/m3) is the maximum acrylonitrile concentration measured
   among all NMP  sites sampling VOCs (while the June 14, 2011 measurement ranked
   seventh across the program). However, acrylonitrile was detected only 10 times at
   TMOK, leading  to the substitution of many zeros for non-detects in the calculations.

•  The third quarter average chloroform concentration is two to three times higher than
   the other quarterly average concentrations. A review of the data shows that five of the
   six concentrations greater than 2 |ig/m3 were measured during the third quarter of
   2011. In addition, the third quarter has the fewest non-detects of chloroform (6),
   while the number of non-detects in the other quarters range from seven (first quarter)
   to 10 (fourth quarter).


•  Each of the quarterly average concentrations of manganese has a relatively large
   confidence interval  associated with it, particularly the fourth quarter average. A
   review of the data shows that the maximum concentration of manganese was
   measured at TMOK on October 6, 2011 (65.6 ng/m3), which is the same day the
   maximum manganese concentration was measured at TOOK. The next highest
   concentration measured during the fourth quarter was considerably less (25.1 ng/m3).

                                  19-42

-------
   A manganese measurement greater than 40 ng/m3 was measured in three of the four
   calendar quarters. The second quarter of 2011 has the greatest number of
   concentrations greater than 30 ng/m3 (5), followed by four in the third quarter and one
   each in the first and fourth quarters. This explains the relatively high-level of
   variability shown in the quarterly averages.


Observations for PROK from Table 19-5 include the following:

•  The third quarter average concentration ofp-dichlorobenzene is more than seven
   times higher than the other quarterly averages. A review of the data shows that the
   only two measurements greater than 1 |ig/m3 were both measured in July. These two
   measurements account for half of the/>-dichlorobenzene concentrations greater than
   1 |ig/m3 measured across all NMP sites sampling VOCs. The eight highest
   concentrations ofp-dichlorobenzene measured at PROK (those greater than
   0.35 |ig/m3) were measured during the third quarter.

•  The third quarter average manganese concentration is greater than the other quarterly
   averages and has a relatively large confidence interval associated with it. The
   maximum manganese concentration (43.8 ng/m3) was measured on
   September 30, 2011, which is the day the second and third highest manganese
   concentrations were measured at TOOK and TMOK, respectively.  The second
   highest manganese concentration measured at PROK was half as high (23.0 ng/m3),
   but also measured during the third quarter.


Observations for MWOK from Table 19-5  include the following:

•  The annual average concentration of formaldehyde for MWOK is  nearly identical to
   the annual average concentration for OCOK, which has the highest annual average
   concentration of formaldehyde among the Oklahoma monitoring sites. The maximum
   concentration of formaldehyde was measured at MWOK on August 25, 2011
   (7.10 |ig/m3), although the five highest concentrations were all measured in July or
   August. Twelve of the 17 concentrations greater than 5 |ig/m3 were measured during
   the third quarter, with four in the second quarter and one in the fourth quarter.

•  The first quarter average concentration of tetrachloroethylene is two and three times
   greater than the second and third quarter averages, respectively, and has a relatively
   large confidence interval associated with it. A review of the data shows that the
   maximum concentration of this pollutant was measured on March 4, 2011
   (1.45 |ig/m3). This concentration is the  only measurement greater than 1  |ig/m3
   measured at MWOK  and is nearly three times higher than the next highest
   tetrachloroethylene concentration (0.571 |ig/m3). This measurement is also the ninth
   highest tetrachloroethylene concentration measured across all NMP sites sampling
   VOCs.

•  There are no fourth quarter average concentrations in Table 19-5 for MWOK because
   sampling was discontinued at this site at the end of November.
                                  19-43

-------
       Observations for OCOK from Table 19-5 include the following:

       •  The second quarter acrylonitrile average for OCOK is higher than the other quarterly
          averages and has a relatively large confidence interval associated with it. The
          maximum concentration of acrylonitrile was measured on May 9, 2011 (1.27 |ig/m3)
          and is nearly twice the next highest concentration (0.644 |ig/m3) measured on
          June 20, 2011. The May 9th measurement is the ninth highest concentration measured
          among sites sampling VOCs and OCOK is one of only five sites to measure an
          acrylonitrile concentration greater than 1 |ig/m3 (TMOK is one of the other four).
          However, acrylonitrile was detected only nine times at OCOK, leading to the
          substitution of many zeros for non-detects in the calculations, which explains the
          relatively large confidence intervals for each quarterly average.

       •  The third quarter average concentrations of benzene and 1,3-butadiene  are higher than
          the other quarterly averages and have relatively large confidence intervals,
          particularly for 1,3-butadiene. The maximum concentration of each pollutant was
          measured on September 18, 2011. The benzene concentration for this date
          (6.81 |ig/m3) is nearly four times higher than the next highest concentration measured
          at OCOK and the eighth highest benzene concentration  measured across the program.
          The 1,3-butadiene concentration for this date (9.51  |ig/m3) is nearly forty times
          higher than the next highest concentration measured at OCOK. This 1,3-butadiene
          concentration is not only the maximum concentration measured across the program,
          but is more than three times higher than the next highest concentration measured
          across the program (2.68 |ig/m3, measured at NBIL).

       •  OCOK has the highest annual average concentration of formaldehyde among the
          Oklahoma sites. The maximum formaldehyde concentration measured at OCOK was
          measured on May 9, 2011 (19.6 |ig/m3) and is the third highest concentration of
          formaldehyde measured among NMP sites sampling carbonyl compounds. The
          second and third quarter average concentrations are significantly greater than the
          other quarterly averages. Although the second and third quarter averages are similar
          to each other, the second quarter average has a relatively large confidence interval
          associated with it, indicating that outlier(s) may be  affecting the  average
          concentration.  The three highest formaldehyde concentrations measured at OCOK
          were measured between April 27, 2011 and May 9, 2011 and range from 9.10 |ig/m3
          to 19.6 |ig/m3. Of the 20 formaldehyde concentrations greater than 5 |ig/m3, only one
          was measured outside of the  second or third quarter. The range of formaldehyde
          concentrations for the second quarter is 1.06 |ig/m3 to 19.6 |ig/m3 with a median
          concentration of 4.55 |ig/m3. The range of formaldehyde concentrations for the third
          quarter is 2.90 |ig/m3 to 8.59 |ig/m3 with a  median concentration of 6.07 |ig/m3.


       Tables 4-9 through 4-12 present the sites with the 10 highest annual average

concentrations for each of the program-level pollutants of interest. Observations for the

Oklahoma sites include the following:

       •  The Oklahoma sites appear in Tables 4-9 through 4-12 a total of 57 times. However,
          because they are the only sites sampling TSP metals, all five sites appear for each
          metal, accounting for 30 of the appearances.

                                         19-44

-------
       •  TOOK has the highest annual average of concentration of benzene among all NMP
          sites sampling this pollutant. Similar findings were observed in the 2010 NMP report.
          The annual average for TMOK ranks fourth.

       •  The annual average concentrations for four of the five Oklahoma sites ranked among
          the highest annual average concentrations ofp-dichlorobenzene, with PROK ranking
          the highest at third. OCOK, the only site that does not appear in Table 4-9 for
          />-dichlorobenzene, ranks 12th.

       •  OCOK ranks third highest for 1,2-dichloroethane and 1,3-butadiene. TOOK ranks
          third highest for ethylbenzene and MWOK ranks third highest for
          hexachl oro-1,3 -butadi ene.

       •  All five Oklahoma sites appear in Table 4-10 for their annual average concentrations
          of formaldehyde, ranking third through seventh among sites sampling carbonyl
          compounds.  TOOK ranks fourth for acetaldehyde while OCOK and TMOK rank
          seventh and eighth, respectively.

       •  Of the six TSP metals shown in Table 4-12, TOOK has the highest annual average
          concentration of five of them among the Oklahoma sites.  The only exception is for
          beryllium, where PROK ranks highest and TOOK ranks second.


19.4.2  Concentration Comparison

       In order to better illustrate how a site's annual average concentrations compare to the

program-level averages, a site-specific box plot was created for the selected NATTS MQO Core

Analytes listed in Section 3.5.3, where applicable. Thus, box plots for acetaldehyde,  arsenic,

benzene, 1,3-butadiene, formaldehyde, lead, and manganese were created for the Oklahoma

sites. Figures 19-24 through 19-30 overlay the sites' minimum, annual average, and maximum

concentrations onto the  program-level minimum, first quartile, median, average, third quartile,

and maximum concentrations, as described in  Section 3.5.3.
                                         19-45

-------
      Figure 19-24. Program vs. Site-Specific Average Acetaldehyde Concentrations
FF.CK
[•,•",'•; CK
OCOK
                                        6           B
                                              Concentration (|
                                                                10
                                                                                                   16
                 Program:   IstQuartile      2ndQuartile      SrdQuartile     4thQuartile     Average
                 Site:      Site Average     Site Minimum/Maximum

                                                19-46

-------
      Figure 19-25. Program vs. Site-Specific Average Arsenic (TSP) Concentrations
FF.CK
[•,",'•; CK
OCOK
                j.25
                            3.5
                                        0.75           1           1.25
                                                Concentration (ng/m3)
                                                                             1.5
                                                                                         1.75
                  Program:   IstQuartile      2ndQuartile      3rdQuartile      4thQuartile      Average
                 Site:       Site Average     Site Minimum/Maximum

                                                 19-47

-------
Figure 19-26. Program vs. Site-Specific Average Benzene Concentrations
TOOK




TMOK



PROK



MWOK


OCOK


;














•







1
i Program Max Concentration = 23. 8 ng/m3 ]



— Program Max Concentration = 23. 8 |ag/nna
1 1


~! ! Program Max Concentration = 23. 8 uE/m3



L i Program Max Concentration = 23. 8 ng/m3


i
J _,
\ Program Max Concentration = 23. 8 ng/m3 j

12345 S7B91
Concentration (^g/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
• • D D
Site: Site Average Site Minimum/Maximum
o —
                               19-48

-------
Figure 19-27. Program vs. Site-Specific Average 1,3-Butadiene Concentrations

-rrk UA
r

i
TMOK B-O 	



PRCK H I
r

i
MWOK ปJ-



.
"r"K H-

D


Program Max Concentration 9.51 ng/ms



Program Max Concentratfon = 9.51 ng/ms



i Program Max Concentration = 9.51 (Jg/m3



i Program Max Concentration = 9.51 (Jg/m-



i Program Max Concentratin = 9.51 ng/m3


0.5 1 1.5 2 2.5 :
Concentration (^g/m3)
Program: IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
• • D D
Site: Site Average Site Minimum/Maximum
o —
                                  19-49

-------
     Figure 19-28. Program vs. Site-Specific Average Formaldehyde Concentrations
    mฑ
FF.CK
[•,•",'•; CK
OCOK
                              10
                                           15
                                       Concentration (|
                                                        20
                                                                     25
                                                                                   50
              Program:   IstQuartile    2ndQuartile     SrdQuartile     4thQuartile     Average
              Site:      Site Average    Site Minimum/Maximum
                        o
                                        19-50

-------
        Figure 19-29. Program vs. Site-Specific Average Lead (TSP) Concentrations
FF.CK
[•,",'•; CK
OCOK
                                                    10         12
                                               Concentration (ng/m3)
                                                                                 16
                                                                                           IE
                 Program:    IstQuartile     2ndQuartile     3rdQuartile      4thQuartile      Average
                 Site:      Site Average     Site Minimum/Maximum

                                                19-51

-------
   Figure 19-30. Program vs. Site-Specific Average Manganese (TSP) Concentrations
FRCK
[•,•",'•; CK
OCOK
                              40
                                            ฃ3
                                       Concentration (ng/m3)
                                                                       153
                                                                                    123
              Program:   IstQuartile     2ndQuartile    SrdQuartile     4thQuartile     Average
              Site:     Site Average     Site Minimum/Maximum

      Observations from Figures 19-24 through 19-30 include the following:

         •  Figure 19-24 shows that the annual average acetaldehyde concentrations for
            TOOK, TMOK, and OCOK are greater than the program-level average for
            acetaldehyde. The annual average for TOOK is also greater than the program-
            level third quartile. The annual average acetaldehyde concentration for MWOK is
            similar to the program-level average while the annual average for PROK is less
            than the program-level average and equivalent to the program-level median
            concentration. The range of acetaldehyde concentrations is largest for TOOK and
            smallest for PROK. There were no non-detects of acetaldehyde reported for the
            Oklahoma sites or across the program.

         •  Because the Oklahoma sites are the only sites sampling TSP metals, Figure 19-25
            compares the individual Oklahoma site data against the combined  Oklahoma data.
            Figure 19-25 shows that the annual average arsenic (TSP) concentration is
            greatest for TOOK and least for MWOK. This figure also shows that the range of
            measurements of arsenic is largest for OCOK, where the maximum arsenic (TSP)
            concentration was measured, although similar concentrations were also measured
                                        19-52

-------
   at TOOK and TMOK. The minimum arsenic concentration measured among the
   five sites sampling TSP metals was measured at PROK.

•  Figure 19-26 presents the box plots for benzene. Note that the program-level
   maximum concentration (23.8 |ig/m3) is not shown directly on each 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
   10 |ig/m3. Figure 19-26 shows that the annual average concentration of benzene
   for TOOK and TMOK are greater than the program-level average while the
   annual average concentration of benzene for PROK, MWOK, and OCOK are less
   than the program-level average. The annual average benzene concentration for
   TOOK is at least twice the other sites' annual averages. The maximum benzene
   concentration measured at TOOK is the maximum benzene concentration
   measured across the program. There were no non-detects of benzene measured at
   the Oklahoma sites.

•  Figure 19-27 presents the box plots for 1,3-butadiene. Similar to the box plots for
   benzene, the program-level maximum concentration (9.51 |ig/m3) is not shown
   directly on the box plots as the scale has been reduced to 3 |ig/m3 to allow for the
   observation of data points at the lower end of the concentration range. The
   maximum concentration of 1,3-buadiene measured across the program was
   measured at OCOK. The annual average 1,3-butadiene concentration for OCOK
   is twice the program-level average concentration, while the annual averages for
   TOOK and TMOK are similar to  the program-level average and the annual
   averages for PROK and MWOK  are less than the program-level  average. Several
   non-detects of 1,3-butadiene were measured at the Oklahoma sites, as indicated
   by the minimum concentration for each site.

•  Figure 19-28 shows that the annual average formaldehyde concentration for each
   Oklahoma site is greater than the  program-level  average concentration and third
   quartile. Although the annual average concentrations of formaldehyde did not
   vary significantly among the Oklahoma sites, the maximum concentration
   measured at OCOK is roughly twice the maximum concentration measured at the
   other four sites, although all are less than the maximum concentration measured
   across the program. There were no non-detects of formaldehyde  measured at the
   Oklahoma sites or across the program.

•  Because the Oklahoma sites are the only monitoring sites sampling TSP metals,
   Figure 19-29 compares the individual  Oklahoma site lead data against the
   combined Oklahoma data. Figure 19-29 shows that the annual average lead (TSP)
   concentration is greatest for TOOK and TMOK  and lowest for PROK, MWOK,
   and OCOK (note that the annual averages for these three sites are not that
   different from each other). This figure also shows that the range  of lead
   measurements was greatest for TOOK and TMOK and smallest for MWOK. The
   maximum manganese (TSP) concentration was measured at TOOK.
                              19-53

-------
             Figure 19-30 compares the individual Oklahoma site manganese data against the
             combined Oklahoma data. Figure 19-30 shows that the annual average manganese
             (TSP) concentration is highest for TOOK and lowest for PROK. Figure 19-30
             also shows that the range of manganese measurements was greatest for TOOK
             and smallest for MWOK.
19.4.3  Concentration Trends
       A site-specific trends evaluation was completed for sites that have sampled one or more
of the selected NATTS MQO Core Analytes for 5 consecutive years or longer, as described in
Section 3.5.4. TOOK has sampled TSP metals, carbonyl compounds, and VOCs since 2006;
thus, Figures 19-31 through 19-37 present the annual statistical metrics for acetaldehyde, arsenic,
benzene,  1,3-butadiene, formaldehyde, lead, and manganese, respectively. The statistical metrics
presented for assessing trends include the substitution of zeros for non-detects.
        Figure 19-31. Annual Statistical Metrics for Acetaldehyde Concentrations
                                 Measured at TOOK




1 c
I 6
e
a
3
ft 4 _
E q
2 -
1 -


















ฃ f
2007 2008





^

20O9
Year
# 5th Percentile — Minimum — Median



	 * 	


	






T





I 1
2010 2011
Maximum • 95th Percentile .,+.. Average
                                         19-54

-------
Figure 19-32. Annual Statistical Metrics for Arsenic (TSP) Concentrations
                             Measured at TOOK
                                         2009
                                         Year
           5th Percentile    —  Minimum     — Median    —  Maximum     * 95th Percentile   +.^.* Average
   Figure 19-33. Annual Statistical Metrics for Benzene Concentrations
                             Measured at TOOK
                                         2009
                                         Year
         #  5th Percentile    —  Minimum     — Median    —  Maximum
                                                          95th Percentile   ..^.. Average
                                     19-55

-------
Figure 19-34. Annual Statistical Metrics for 1,3-Butadiene Concentrations
                         Measured at TOOK



.'
t rat ion (ME/I
e
rage ton con
5
<

0.05 -





"






*"*••-

r




•
^7





••




•
'


	














.•••*






2007 2008 20O9 2010 2011
Year
• 5th Percentile — Minimum — Median
— Maximum •
95t




age
Figure 19-35. Annual Statistical Metrics for Formaldehyde Concentrations
                         Measured at TOOK



B
a
i
3
1













<
••



^....
••






....^


••.ป„,
••••





	






^
I







••



••










...•*



!




2007 2008 2009 2010 2011
Year
# 5th Percentile — Min mum — Median
— Maximum 4
95th Percentile ••*•
Average
                                19-56

-------
   Figure 19-36. Annual Statistical Metrics for Lead (TSP) Concentrations
                            Measured at TOOK



ralk.ii (ng/ni3)
', j
i
I
15 jn













•d





b.

2007





•




I
I *
>.. 1 I


ft • — ป — • ' — • — ' *
200B 20O9 2010 2011
Year
• 5th Percentile - Min
,. . , .
	 1
Figure 19-37. Annual Statistical Metrics for Manganese (TSP) Concentrations
                            Measured at TOOK
          #  5th Percentile    — Minimum    —  Median    —  Maximum
                                                       95th Percentile   ..^.. Average
                                    19-57

-------
       Observations from Figure 19-31 for acetaldehyde measurements at TOOK include the
following:

       •   Although TOOK began sampling carbonyl compounds in January 2006, equipment
          complications at the onset of sampling resulted in fewer than 85 percent valid
          samples in 2006; thus, Figure 19-31 excludes data from 2006 per the criteria specified
          in Section 3.5.4.

       •   The maximum concentration of acetaldehyde was measured in 2011 (8.95 |ig/m3).
          The five highest concentrations were all measured in 2011. Of the 16 acetaldehyde
          concentrations greater than 4 |ig/m3 measured at TOOK, half were measured in 2011,
          five were measured in 2010, and one was measured in each of the other years shown.

       •   The average concentration exhibits an increasing trend since 2009. Nearly all of the
          statistical metrics shown have been increasing since 2009.

       •   The minimum concentration measured for each year shown is greater than zero,
          indicating that there were no non-detects of acetaldehyde reported over the years at
          TOOK.


       Observations from Figure 19-32 for arsenic (TSP) measurements at TOOK include the
following:

       •   Although TOOK began sampling TSP  metals in 2006, sampling did not begin until
          October, which does not yield enough samples for the statistical metrics to be
          calculated; thus, Figure 19-32 excludes data from 2006 per the criteria specified in
          Section 3.5.4.

       •   The two highest concentrations of arsenic were measured at TOOK in September
          2007. These are the only two concentrations greater than 4 ng/m3 measured at TOOK.

       •   The average and median concentrations exhibit a decreasing trend from 2007 to 2009.
          The average concentration of arsenic did not change significantly from 2009 to 2010
          while the median decreased slightly. Although a slight increase in the average and
          median concentrations is shown for 2011, the maximum and 95th percentiles actually
          decreased and the median for 2011 is actually slightly greater than the average
          concentration.

       •   The minimum concentration measured for each year shown is greater than zero,
          indicating that there were no non-detects of arsenic reported over the years at TOOK.


       Observations from Figure 19-33 for benzene measurements at TOOK include the

following:

       •   Although TOOK began sampling VOCs in January 2006, equipment complications at
          the onset of sampling resulted in fewer than 85 percent valid samples in 2006; thus,
          Figure 19-33 excludes data from 2006.

                                         19-58

-------
       •  The maximum concentration of benzene was measured in 2011 (23.8 |ig/m3). The
          four highest concentrations were all measured in 2011 and are greater than 10 |ig/m3.
          The 95th percentile for 2011 is greater than the maximum concentration for each of
          the previous years.

       •  The average benzene concentration has fluctuated over the years. After a substantial
          decrease from 2008 to 2009, most of the statistical parameters increased for 2010, an
          increase that continued into 2011.

       •  The difference between the average and median concentrations nearly tripled from
          2010 to 2011. This is a further indication of the increasing variability of the 2011
          benzene measurements. The median represents the mid-point of the dataset, which
          increased by more than 0.5 |ig/m3 from 2010 to 2011.  The average, which is
          influenced more by outliers, such as the maximum concentration measured in 2011,
          increased by more than 1.25 |ig/m3 from 2010 to 2011.

       •  The minimum concentration measured for each year shown is greater than zero,
          indicating that there were no non-detects of benzene reported over the years at
          TOOK.


       Observations from Figure 19-34 for 1,3-butadiene measurements at TOOK include the
following:

       •  Similar to other pollutants, the maximum concentration of 1,3-butadiene was
          measured in 2011  (0.34 |ig/m3), although a similar concentration was also measured
          in 2007 (0.33 |ig/m3).

       •  After an initial decrease from 2007 to 2008 and little change in 2009, the average
          concentration began to increase, with the greatest increase occurring from 2010 to
          2011. With the exception of the minimum and 5th percentile, all of the statistical
          metrics increased for 2011.

       •  The minimum concentration for 2007 is greater than zero, indicating that non-detects
          of 1,3-butadiene were not reported. For 2008 and 2009, the minimum concentration
          shown  is zero, indicating that at least one non-detect was measured during those
          years. For 2010 and 2011, both the minimum concentration and 5th percentile are
          zero, indicating that additional non-detects were reported. The percentage of non-
          detects for 2010 and 2011 is less than nine percent.


       Observations from Figure 19-35 for formaldehyde measurements at TOOK include the
following:

       •  The maximum concentration of formaldehyde (10.1 |ig/m3) was measured at TOOK
          on August 19, 2011, the same day the maximum acetaldehyde concentration was
          measured.
                                         19-59

-------
       •  Similar to acetaldehyde, an increasing trend is shown for formaldehyde from 2009 to
          2010 and 2011. However, due to the higher level of variability in the formaldehyde
          measurements as a whole, the difference is not statistically significant.

       •  The minimum concentration measured for each year shown is greater than zero,
          indicating that there were no non-detects of formaldehyde reported over the years at
          TOOK.


       Observations from Figure 19-36 for lead (TSP) measurements at TOOK include the
following:

       •  The maximum concentration  of lead was measured in 2008 (50.5 ng/m3). Four of the
          five highest concentrations of lead were measured in 2008.

       •  Although most of the statistical parameters increased from 2007 to 2008, the median
          concentration actually decreased. This indicates that the higher concentrations
          measured in 2008 were likely driving the average concentration while a higher
          percentage of measurements were actually lower than in 2007.

       •  A significant decrease is  shown for most of the statistical parameters from 2008 to
          2009, with additional slight decreases for 2010.

       •  Nearly all of the statistical parameters increased for 2011.

       •  The minimum concentration measured for each year shown is greater than zero,
          indicating that there were no non-detects of lead reported over the years at TOOK.


       Observations from Figure 19-37 for manganese (TSP) measurements at TOOK  include
the following:

       •  The maximum concentration  of manganese was measured in 2007 (131 ng/m3),
          although another measurement greater than 100 ng/m3 was also measured in 2011
          (104 ng/m3).

       •  A steady decreasing trend in the average and median concentrations through 2009
          was followed by an increasing trend for 2010 and 2011. The average concentration
          for 2007 is similar to the  average concentration for 2011.

       •  The minimum concentration measured for each year shown is greater than zero,
          indicating that there were no non-detects of manganese reported over the years at
          TOOK.
                                         19-60

-------
19.5   Additional Risk-Based Screening Evaluations
       The following risk-based screening evaluations were conducted to characterize risk at
each Oklahoma monitoring site. Refer to Sections 3.3 and 3.5.5 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.

19.5.1  Risk-Based Screening Assessment Using MRLs
       A risk-based screening was conducted by comparing the concentration data from the
Oklahoma monitoring  sites to the ATSDR MRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 to 14 days); intermediate (exposures of 15 to 364 days); and chronic (exposures
of 1 year or greater). The preprocessed daily measurements of the pollutants of interest were
compared to the acute MRLs; the quarterly averages were compared to the intermediate MRLs;
and the annual averages were compared  to the chronic MRLs.

       As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2011.

19.5.2  Cancer Risk and Noncancer Hazard Approximations
       For the pollutants of interest for the Oklahoma monitoring sites and where annual
average concentrations could be calculated, risk was examined by calculating cancer risk and
noncancer hazard approximations. These approximations can be used as risk estimates for cancer
and noncancer effects attributable to the  pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers may want to shift or
confirm their air-monitoring priorities. Refer to Section 3.5.5.2 for an explanation of how cancer
risk and noncancer hazard approximations are calculated and what limitations are associated  with
them. Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer
hazard approximations are presented in Table 19-6, where applicable. Cancer risk
approximations are presented as probabilities while the noncancer hazard approximations are
ratios and thus, unitless values.
                                         19-61

-------
             Table 19-6. Risk Approximations for the Oklahoma Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
# of Measured
Detections vs.
# of Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Tulsa, Oklahoma - TOOK
Acetaldehyde
Arsenic (TSP)a
Benzene
Beryllium (TSP)a
1,3 -Butadiene
Cadmium (TSP)a
Carbon Tetrachloride
Chloroform
ฃ>-Dichlorobenzene
Ethylbenzene
Formaldehyde
Lead (TSP)a
Manganese (TSP)a
Nickel (TSP)a
Propionaldehyde
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
0.0000022
0.0043
0.0000078
0.0024
0.00003
0.0018
0.000006

0.000011
0.0000025
0.000013


0.00048

0.00000026
0.0000048
0.0000088
0.009
0.000015
0.03
0.00002
0.002
0.00001
0.1
0.098
0.8
1
0.0098
0.00015
0.00005
0.00009
0.008
0.04
0.002
0.1
56/56
56/56
57/57
56/56
52/57
56/56
57/57
17/57
51/57
57/57
56/56
56/56
56/56
56/56
56/56
47/57
8/57
3/57
2.75
ฑ0.41
0.01
ฑ0.01
3.59
ฑ0.98
0.01
ฑO.01
0.09
ฑ0.02
O.01
ฑO.01
0.63
ฑ0.05
0.04
ฑ0.02
0.15
ฑ0.03
0.68
ฑ0.13
3.74
ฑ0.57
0.01
ฑO.01
0.03
ฑ0.01
O.01
ฑO.01
0.51
ฑ0.07
0.12
ฑ0.03
0.01
ฑ0.01
O.01
ฑO.01
6.04
3.27
28.02
0.06
2.70
0.55
3.76

1.70
1.70
48.62


0.84

0.03
0.05
0.01
0.31
0.05
0.12
O.01
0.05
0.03
0.01
O.01
0.01
O.01
0.38
0.04
0.60
0.02
0.06
O.01
0.01
O.01
- = a Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 19-5.
                                              19-62

-------
       Table 19-6. Risk Approximations for the Oklahoma Monitoring Sites (Continued)
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
# of Measured
Detections vs.
# of Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Tulsa, Oklahoma - TMOK
Acetaldehyde
Acrylonitrile
Arsenic (TSP)a
Benzene
Beryllium (TSP)a
1,3 -Butadiene
Cadmium (TSP) a
Carbon Tetrachloride
Chloroform
ฃ>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
Lead (TSP)a
Manganese (TSP)a
Nickel (TSP)a
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
0.0000022
0.000068
0.0043
0.0000078
0.0024
0.00003
0.0018
0.000006

0.000011
0.000026
0.0000025
0.000013


0.00048
0.00000026
0.0000048
0.0000088
0.009
0.002
0.000015
0.03
0.00002
0.002
0.00001
0.1
0.098
0.8
2.4
1
0.0098
0.00015
0.00005
0.00009
0.04
0.002
0.1
61/61
10/60
58/58
60/60
58/58
51/60
58/58
60/60
28/60
47/60
10/60
60/60
61/61
58/58
58/58
58/58
40/60
10/60
1/60
2.40
ฑ0.34
0.08
ฑ0.07
<0.01
ฑ0.01
1.35
ฑ0.18
0.01
ฑ0.01
0.10
ฑ0.02
0.01
ฑ0.01
0.60
ฑ0.03
0.07
ฑ0.03
0.08
ฑ0.02
0.02
ฑ0.01
0.55
ฑ0.09
3.93
ฑ0.62
O.01
ฑO.01
0.02
ฑ0.01
O.01
ฑO.01
0.07
ฑ0.02
0.01
ฑ0.01
0.01
ฑ0.01
5.28
5.36
2.73
10.54
0.04
3.07
0.39
3.61

0.91
0.47
1.37
51.15


0.68
0.02
0.06
0.01
0.27
0.04
0.04
0.05
0.01
0.05
0.02
0.01
0.01
O.01
0.01
O.01
0.40
0.03
0.41
0.02
0.01
0.01
0.01
— = a Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 19-5.
                                              19-63

-------
       Table 19-6. Risk Approximations for the Oklahoma Monitoring Sites (Continued)
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
# of Measured
Detections vs.
# of Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Pryor Creek, Oklahoma - PROK
Acetaldehyde
Arsenic (TSP) a
Benzene
Beryllium (TSP)a
1,3 -Butadiene
Cadmium (TSP)a
Carbon Tetrachloride
Chloroform
ฃ>-Dichlorobenzene
1 ,2-Dichloroethane
Formaldehyde
Lead (TSP)a
Manganese (TSP)a
Nickel (TSP)a
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
0.0000022
0.0043
0.0000078
0.0024
0.00003
0.0018
0.000006

0.000011
0.000026
0.000013


0.00048
0.00000026
0.0000048
0.0000088
0.009
0.000015
0.03
0.00002
0.002
0.00001
0.1
0.098
0.8
2.4
0.0098
0.00015
0.00005
0.00009
0.04
0.002
0.1
58/58
56/56
56/56
56/56
33/56
56/56
56/56
15/56
47/56
17/56
58/58
56/56
56/56
56/56
28/56
2/56
3/56
1.68
ฑ0.18
0.01
ฑ0.01
0.68
ฑ0.08
0.01
ฑO.01
0.04
ฑ0.01
O.01
ฑO.01
0.62
ฑ0.03
0.03
ฑ0.02
0.20
ฑ0.07
0.03
ฑ0.01
3.84
ฑ0.67
O.01
ฑO.01
0.01
ฑ0.01
O.01
ฑO.01
0.03
ฑ0.01
O.01
ฑO.01
0.01
ฑ0.01
3.70
2.32
5.28
0.07
1.07
0.28
3.73

2.20
0.73
49.86


0.38
0.01
0.01
0.01
0.19
0.04
0.02
O.01
0.02
0.02
0.01
O.01
0.01
O.01
0.39
0.02
0.23
0.01
0.01
O.01
0.01
- = a Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 19-5.
                                              19-64

-------
       Table 19-6. Risk Approximations for the Oklahoma Monitoring Sites (Continued)
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
# of Measured
Detections vs.
# of Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Midwest City, Oklahoma - MWOK
Acetaldehyde
Acrylonitrile
Arsenic (TSP)a
Benzene
Beryllium (TSP)a
1,3 -Butadiene
Cadmium (TSP)a
Carbon Tetrachloride
Chloroform
ฃ>-Dichlorobenzene
1,2-Dichloroethane
Ethylbenzene
Formaldehyde
Lead (TSP)a
Manganese (TSP)a
Nickel (TSP)a
Tetrachloroethylene
Trichloroethylene
0.0000022
0.000068
0.0043
0.0000078
0.0024
0.00003
0.0018
0.000006

0.000011
0.000026
0.0000025
0.000013


0.00048
0.00000026
0.0000048
0.009
0.002
0.000015
0.03
0.00002
0.002
0.00001
0.1
0.098
0.8
2.4
1
0.0098
0.00015
0.00005
0.00009
0.04
0.002
56/56
5/56
56/56
56/56
56/56
40/56
56/56
56/56
23/56
50/56
9/56
56/56
56/56
56/56
56/56
56/56
40/56
1/56
2.03
ฑ0.24
0.01
ฑ0.01
<0.01
ฑ0.01
0.73
ฑ0.07
0.01
ฑ0.01
0.06
ฑ0.02
0.01
ฑ0.01
0.60
ฑ0.03
0.04
ฑ0.01
0.09
ฑ0.01
0.01
ฑ0.01
0.25
ฑ0.04
4.05
ฑ0.53
O.01
ฑ0.01
0.01
ฑ0.01
O.01
ฑ0.01
0.15
ฑ0.06
O.01
ฑO.01
4.47
0.76
1.71
5.73
0.05
1.81
0.22
3.60

0.99
0.39
0.61
52.69


0.58
0.04
0.01
0.23
0.01
0.03
0.02
0.01
0.03
0.01
0.01
0.01
O.01
0.01
O.01
0.41
0.02
0.27
0.01
0.01
O.01
- = a Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 19-5.
                                              19-65

-------
       Table 19-6. Risk Approximations for the Oklahoma Monitoring Sites (Continued)
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
# of Measured
Detections vs.
# of Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Oklahoma City, Oklahoma - OCOK
Acetaldehyde
Acrylonitrile
Arsenic (TSP)a
Benzene
Beryllium (TSP)a
1,3 -Butadiene
Cadmium (TSP)a
Carbon Tetrachloride
Chloroform
ฃ>-Dichlorobenzene
1,2-Dichloroethane
Ethylbenzene
Formaldehyde
Lead (TSP)a
Manganese (TSP)a
Nickel (TSP)a
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
0.0000022
0.000068
0.0043
0.0000078
0.0024
0.00003
0.0018
0.000006

0.000011
0.000026
0.0000025
0.000013


0.00048
0.00000026
0.0000048
0.0000088
0.009
0.002
0.000015
0.03
0.00002
0.002
0.00001
0.1
0.098
0.8
2.4
1
0.0098
0.00015
0.00005
0.00009
0.04
0.002
0.1
60/60
9/61
61/61
61/61
61/61
41/61
61/61
61/61
13/61
39/61
14/61
61/61
60/60
61/61
61/61
61/61
40/61
1/61
3/61
2.41
ฑ0.33
0.07
ฑ0.05
<0.01
ฑ0.01
0.88
ฑ0.21
0.01
ฑ0.01
0.20
ฑ0.31
0.01
ฑ0.01
0.58
ฑ0.04
0.02
ฑ0.01
0.06
ฑ0.01
0.03
ฑ0.02
0.31
ฑ0.04
4.06
ฑ0.80
O.01
ฑO.01
0.01
ฑ0.01
O.01
ฑO.01
0.08
ฑ0.02
O.01
ฑO.01
0.01
ฑ0.01
5.30
4.74
2.02
6.85
0.04
6.03
0.23
3.50

0.62
0.81
0.78
52.80


0.40
0.02
0.01
0.01
0.27
0.03
0.03
0.03
0.01
0.10
0.01
0.01
0.01
O.01
0.01
O.01
0.41
0.02
0.28
0.01
0.01
O.01
0.01
- = a Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 19-5.
                                              19-66

-------
       Observations from Table 19-6 include the following:

       •  Formaldehyde has the highest annual average concentration by mass for each site.
          Among the TSP metals, the annual average concentration of manganese is the highest
          for each site.

       •  Formaldehyde and benzene have the highest cancer risk approximations among the
          Oklahoma monitoring sites. Formaldehyde cancer risk approximations range from
          48.62 in-a-million for TOOK to 52.80 in-a-million for OCOK. The cancer risk
          approximations for OCOK and MWOK rank third and fourth highest among all
          cancer risk approximations program-wide. Benzene cancer risk approximations range
          from 5.28 in-a-million for PROK to 28.02 in-a-million for TOOK. The benzene
          cancer risk approximation for TOOK is the highest benzene cancer risk
          approximation program-wide.

       •  Among the metals, arsenic has the highest cancer risk approximations for all of the
          Oklahoma monitoring sites, ranging from 1.71 in-a-million for MWOK to
          3.27 in-a-million for TOOK.

       •  None of the pollutants of interest have noncancer hazard approximations greater than
          1.0, indicating that no adverse health effects are expected from these individual
          pollutants. Among the noncancer hazard approximations for the Oklahoma sites,
          formaldehyde, manganese, and acetaldehyde have the highest noncancer hazard
          approximations for each site (albeit less than 1.0).


19.5.3  Risk-Based Emissions Assessment

       In addition to the risk-based screenings discussed above, Tables 19-7 and 19-8 present an

evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.

Table 19-7 presents the 10 pollutants with the highest emissions from the 2008 NEI, the

10 pollutants with the highest toxicity-weighted emissions,  and the 10 pollutants with the highest

cancer risk approximations (in-a-million), as calculated from the annual averages provided in

Table 19-6. Table 19-8 presents similar information, but identifies the 10 pollutants with the

highest noncancer hazard approximations (HQ), also calculated from annual averages provided

in Table 19-6.
                                         19-67

-------
  Table 19-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
                                               the Oklahoma Monitoring Sites
OO
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Tulsa, Oklahoma (Tulsa County) - TOOK
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Tetrachloroethylene
Naphthalene
Dichloromethane
POM, Group 2b
Nickel, PM
398.15
236.92
198.13
105.13
52.47
41.33
23.57
5.51
3.59
0.40
Benzene
Formaldehyde
1,3 -Butadiene
Hexavalent Chromium, PM
Naphthalene
Ethylbenzene
POM, Group 2b
Acetaldehyde
POM, Group 3
Nickel, PM
3.11E-03
2.58E-03
1.57E-03
1.11E-03
8.02E-04
5.92E-04
3.16E-04
2.31E-04
2.30E-04
1.90E-04
Formaldehyde
Benzene
Acetaldehyde
Carbon Tetrachloride
Arsenic
1,3 -Butadiene
Ethylbenzene
ฃ>-Dichlorobenzene
Nickel
Cadmium
48.62
28.02
6.04
3.76
3.27
2.70
1.70
1.70
0.84
0.55
Tulsa, Oklahoma (Tulsa County) - TMOK
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Tetrachloroethylene
Naphthalene
Dichloromethane
POM, Group 2b
Nickel, PM
398.15
236.92
198.13
105.13
52.47
41.33
23.57
5.51
3.59
0.40
Benzene
Formaldehyde
1,3 -Butadiene
Hexavalent Chromium, PM
Naphthalene
Ethylbenzene
POM, Group 2b
Acetaldehyde
POM, Group 3
Nickel, PM
3.11E-03
2.58E-03
1.57E-03
1.11E-03
8.02E-04
5.92E-04
3.16E-04
2.31E-04
2.30E-04
1.90E-04
Formaldehyde
Benzene
Acrylonitrile
Acetaldehyde
Carbon Tetrachloride
1,3 -Butadiene
Arsenic
Ethylbenzene
ฃ>-Dichlorobenzene
Nickel
51.15
10.54
5.36
5.28
3.61
3.07
2.73
1.37
0.91
0.68

-------
  Table 19-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
                                          the Oklahoma Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Pryor Creek, Oklahoma (Mayes County) - PROK
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
Chloromethylbenzene
Nickel, PM
Dichloromethane
Arsenic, PM
37.28
24.25
17.46
12.40
3.72
2.04
1.60
1.17
1.03
0.53
Arsenic, PM
Hexavalent Chromium, PM
Nickel, PM
Formaldehyde
Benzene
1,3 -Butadiene
Beryllium, PM
Cadmium, PM
Chloromethylbenzene
Naphthalene
2.29E-03
9.45E-04
5.62E-04
3.15E-04
2.91E-04
1.11E-04
8.46E-05
8.35E-05
7.85E-05
6.93E-05
Formaldehyde
Benzene
Carbon Tetrachloride
Acetaldehyde
Arsenic
ฃ>-Dichlorobenzene
1,3 -Butadiene
1 ,2-Dichloroethane
Nickel
Cadmium
49.86
5.28
3.73
3.70
2.32
2.20
1.07
0.73
0.38
0.28
Midwest City, Oklahoma (Oklahoma County) - MWOK
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Naphthalene
Dichloromethane
Tetrachloroethylene
POM, Group 2b
Bis(2-ethylhexyl)phthalate (DEHP), gas
447.79
275.02
250.63
137.04
59.70
29.28
15.08
10.79
4.51
0.49
Benzene
Formaldehyde
1,3 -Butadiene
Naphthalene
Ethylbenzene
Hexavalent Chromium, PM
POM, Group 2b
Acetaldehyde
POM, Group 3
Arsenic, PM
3.49E-03
3.26E-03
1.79E-03
9.96E-04
6.88E-04
6.16E-04
3.97E-04
3.01E-04
2.64E-04
1.85E-04
Formaldehyde
Benzene
Acetaldehyde
Carbon Tetrachloride
1,3 -Butadiene
Arsenic
ฃ>-Dichlorobenzene
Acrylonitrile
Ethylbenzene
Nickel
52.69
5.73
4.47
3.60
1.81
1.71
0.99
0.76
0.61
0.58
VO

-------
Table 19-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
                                       the Oklahoma Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Oklahoma City, Oklahoma (Oklahoma County) - OCOK
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Naphthalene
Dichloromethane
Tetrachloroethylene
POM, Group 2b
Bis(2-ethylhexyl)phthalate (DEHP), gas
447.79
275.02
250.63
137.04
59.70
29.28
15.08
10.79
4.51
0.49
Benzene
Formaldehyde
1,3 -Butadiene
Naphthalene
Ethylbenzene
Hexavalent Chromium, PM
POM, Group 2b
Acetaldehyde
POM, Group 3
Arsenic, PM
3.49E-03
3.26E-03
1.79E-03
9.96E-04
6.88E-04
6.16E-04
3.97E-04
3.01E-04
2.64E-04
1.85E-04
Formaldehyde
Benzene
1,3 -Butadiene
Acetaldehyde
Acrylonitrile
Carbon Tetrachloride
Arsenic
1 ,2-Dichloroethane
Ethylbenzene
/>-Dichlorobenzene
52.80
6.85
6.03
5.30
4.74
3.50
2.02
0.81
0.78
0.62

-------
Table 19-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
                                         RfCs for the Oklahoma Monitoring Sites
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Tulsa, Oklahoma (Tulsa County) - TOOK
Toluene
Xylenes
Methanol
Benzene
Hexane
Ethylbenzene
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Ethylene glycol
1,180.06
902.52
402.45
398.15
293.67
236.92
198.13
105.13
52.47
50.77
Acrolein
1,3 -Butadiene
Formaldehyde
Manganese, PM
Benzene
Lead, PM
Acetaldehyde
Xylenes
Naphthalene
Cobalt , PM
577,473.50
26,235.74
20,217.32
18,832.62
13,271.68
13,200.35
11,681.56
9,025.21
7,857.87
5,351.77
Manganese
Formaldehyde
Acetaldehyde
Benzene
Propionaldehyde
Arsenic
1,3 -Butadiene
Lead
Cadmium
Nickel
0.60
0.38
0.31
0.12
0.06
0.05
0.05
0.04
0.03
0.02
Tulsa, Oklahoma (Tulsa County) - TMOK
Toluene
Xylenes
Methanol
Benzene
Hexane
Ethylbenzene
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Ethylene glycol
1,180.06
902.52
402.45
398.15
293.67
236.92
198.13
105.13
52.47
50.77
Acrolein
1,3 -Butadiene
Formaldehyde
Manganese, PM
Benzene
Lead, PM
Acetaldehyde
Xylenes
Naphthalene
Cobalt , PM
577,473.50
26,235.74
20,217.32
18,832.62
13,271.68
13,200.35
11,681.56
9,025.21
7,857.87
5,351.77
Manganese
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Benzene
Arsenic
Acrylonitrile
Lead
Cadmium
Nickel
0.41
0.40
0.27
0.05
0.05
0.04
0.04
0.03
0.02
0.02

-------
  Table 19-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
                                     RfCs for the Oklahoma Monitoring Sites (Continued)
to
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Pryor Creek, Oklahoma (Mayes County) - PROK
Hydrochloric acid
Toluene
Xylenes
Benzene
Cyanide Compounds, gas
Hydrofluoric acid
Formaldehyde
Methanol
Hexane
Ethylbenzene
145.18
89.08
72.70
37.28
34.90
26.35
24.25
23.55
22.30
17.46
Acrolein
Chlorine
Cyanide Compounds, gas
Arsenic, PM
Nickel, PM
Manganese, PM
Hydrochloric acid
Cyanide Compounds, PM
Cadmium, PM
Lead, PM
91,374.65
61,006.67
43,619.17
35,508.85
13,015.34
8,604.28
7,259.15
7,154.96
4,636.70
4,369.13
Formaldehyde
Manganese
Acetaldehyde
Arsenic
Benzene
Lead
1,3 -Butadiene
Cadmium
Nickel
Carbon Tetrachloride
0.39
0.23
0.19
0.04
0.02
0.02
0.02
0.02
0.01
0.01
Midwest City, Oklahoma (Oklahoma County) - MWOK
Toluene
Xylenes
Methanol
Benzene
Hexane
Ethylbenzene
Formaldehyde
Acetaldehyde
Ethylene glycol
1,3 -Butadiene
1,373.20
1,058.17
468.16
447.79
337.57
275.02
250.63
137.04
61.51
59.70
Acrolein
1,3 -Butadiene
Formaldehyde
Acetaldehyde
Benzene
Xylenes
Naphthalene
Arsenic, PM
Lead, PM
Propionaldehyde
875,997.95
29,851.09
25,574.80
15,226.70
14,926.42
10,581.67
9,759.95
2,874.57
2,417.07
2,079.24
Formaldehyde
Manganese
Acetaldehyde
1,3 -Butadiene
Arsenic
Benzene
Lead
Nickel
Cadmium
Carbon Tetrachloride
0.41
0.27
0.23
0.03
0.03
0.02
0.02
0.01
0.01
0.01

-------
Table 19-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
                                   RfCs for the Oklahoma Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Oklahoma City, Oklahoma (Oklahoma County) - OCOK
Toluene
Xylenes
Methanol
Benzene
Hexane
Ethylbenzene
Formaldehyde
Acetaldehyde
Ethylene glycol
1,3 -Butadiene
1,373.20
1,058.17
468.16
447.79
337.57
275.02
250.63
137.04
61.51
59.70
Acrolein
1,3 -Butadiene
Formaldehyde
Acetaldehyde
Benzene
Xylenes
Naphthalene
Arsenic, PM
Lead, PM
Propionaldehyde
875,997.95
29,851.09
25,574.80
15,226.70
14,926.42
10,581.67
9,759.95
2,874.57
2,417.07
2,079.24
Formaldehyde
Manganese
Acetaldehyde
1,3 -Butadiene
Acrylonitrile
Arsenic
Benzene
Lead
Cadmium
Nickel
0.41
0.28
0.27
0.10
0.03
0.03
0.03
0.02
0.01
0.01

-------
       The pollutants listed in Tables 19-7 and 19-8 are limited to those that have cancer and

noncancer toxicity factors, respectively. As a result, although the actual value of the emissions is

the same, the highest emitted pollutants in the cancer table may be different from the noncancer

table. The cancer risk and noncancer hazard approximations based on each site's annual averages

are limited to those pollutants for which each respective site sampled. As discussed in

Section 19.3, the Oklahoma sites sampled VOCs, carbonyl compounds, and TSP metals. 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, as discussed in

previous sections. A more in-depth discussion of this analysis is provided in Section 3.5.5.3.

Similar to the cancer risk and noncancer hazard approximations, this analysis may help policy-

makers prioritize their air monitoring activities.


       Observations from Table 19-7 include the following:

       •  Benzene is the highest emitted pollutant with a cancer URE in Mayes, Oklahoma, and
          Tulsa Counties, followed by ethylbenzene and formaldehyde in Oklahoma and Tulsa
          Counties and formaldehyde and ethylbenzene in Mayes County. The emissions of
          these pollutants in Mayes County are an order of magnitude lower than the emissions
          for Oklahoma and Tulsa Counties.

       •  The pollutants with the highest toxicity-weighted emissions (of the pollutants with
          cancer UREs) for Oklahoma and Tulsa Counties are benzene, formaldehyde, and
          1,3-butadiene. The pollutants with the highest toxicity-weighted emissions for Mayes
          County are arsenic, hexavalent chromium, and nickel.

       •  Eight of the highest emitted pollutants in Tulsa County also have the highest toxicity-
          weighted emissions. Six of the highest emitted pollutants in Mayes County also have
          the highest toxicity-weighted emissions. Seven of the highest emitted pollutants in
          Oklahoma County also have the highest toxicity-weighted emissions.

       •  Formaldehyde and benzene have the highest cancer risk approximations among the
          Oklahoma sites' pollutants of interest. These pollutants appear on both emissions-
          based lists for all five sites. Conversely, carbon tetrachloride, another pollutant with
          relatively  high cancer risk approximations,  does not appear on either emissions-based
          list.

       •  While hexavalent chromium is among the pollutants with the highest toxicity-
          weighted emissions for each county, it is not among the highest emitted pollutants.
          This indicates that lower emissions can translate to higher risk levels.

       •  The toxicity-weighted pollutants listed for Mayes County are considerably different
          than for the other two counties. There are five metals listed for Mayes County while
          the other counties only have two  each. In addition, there  are no POM Groups listed

                                         19-74

-------
          for Mayes County, while POM, Groups 2b and 3 appear for Oklahoma and Tulsa
          Counties.


       Observations from Table 19-8 include the following:

       •  Toluene, xylenes, and methanol are the highest emitted pollutants with noncancer
          RfCs in Oklahoma and Tulsa Counties. Hydrochloric acid, toluene, and xylenes are
          the highest emitted pollutants with noncancer RfCs in Mayes County. Note that the
          quantity emitted is much higher in Tulsa and Oklahoma Counties than in Mayes
          County.

       •  Acrolein is the pollutant with the highest toxi city-weighted emissions (of the
          pollutants with noncancer RfCs) for all three counties. Yet, this pollutant is not
          among the highest emitted pollutants for any of the three counties. This indicates that
          lower emissions can translate to higher risk levels. Acrolein was sampled for at  all of
          the Oklahoma sites, but this pollutant was excluded from the pollutants of interest
          designation, and thus subsequent risk-based screening evaluations, due to questions
          about the consistency and reliability of the measurements,  as discussed in Section 3.2.

       •  Two of the highest emitted pollutants in Mayes County also have the highest toxicity-
          weighted emissions; five of the highest emitted pollutants in Tulsa and Oklahoma
          Counties also have the highest toxicity-weighted emissions.

       •  Five of the 10 pollutants with the highest noncancer toxicity-weighted emissions in
          Mayes County were metals. Cyanide compounds, gaseous and particulate, account for
          two more.

       •  Formaldehyde and manganese have the highest noncancer hazard approximations
          among the Oklahoma sites. Formaldehyde appears on both emissions-based lists for
          Tulsa and Oklahoma Counties but ranks  111  for toxicity-weighted emissions for
          Mayes County and therefore does not appear in Table 19-8 in  that column.
          Manganese appears among the pollutants with the highest toxicity-weighted
          emissions for Tulsa and Mayes Counties but ranks 14th for toxicity-weighted
          emissions for Oklahoma County. There are no metals listed among the highest
          emitted pollutants for any of the three counties.

       •  It is important to note that for the metals, the emissions-based lists are PMi0 while the
          Oklahoma sites sampled TSP metals.


19.6   Summary of the 2011 Monitoring Data for the Oklahoma Monitoring Sites

       Results from several of the data treatments described in this section include the
following:

       ปซป  Twenty pollutants failed at least one screen for TOOK; 19 pollutants failed screens
          for TMOK; 15 pollutants failed screens for PROK; 17 pollutants failed screens for
          MWOK; and 16 pollutants failed screens for OCOK.
                                         19-75

-------
Formaldehyde had the highest annual average concentration by mass for each site.
Among the TSP metals, the annual average concentration of manganese was the
highest for each site.

TOOK had the highest annual average of concentration of benzene among all NMP
sites sampling this pollutant. Annual averages of formaldehyde for all five Oklahoma
sites rank among the highest annual average concentrations of formaldehyde
program-wide.

Concentrations of several of the NATTSMQO Core Analytes exhibit increasing
trends at TOOK.
                               19-76

-------
20.0   Site in Rhode Island
       This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS site in Rhode Island, and integrates these concentrations
with emissions, meteorological, and risk information. Data generated by sources other than ERG
are not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) for detailed discussions and definitions
regarding the various data analyses presented below.

20.1   Site Characterization
       This section characterizes the Rhode Island monitoring site by providing geographical
and physical information about the location of the site and the surrounding area. This
information is provided to give the reader insight regarding factors that may influence the air
quality near the site and assist in the interpretation of the ambient monitoring measurements.

       The PRRI monitoring  site is located in south Providence. Figure 20-1 is a composite
satellite image retrieved from ArcGIS Explorer showing the monitoring site in its urban location.
Figure 20-2 identifies nearby point source emissions locations by source category, as reported in
the 2008 NEI for point sources. Note that only sources within 10 miles  of the site are included in
the facility counts provided in Figure 20-2. A 10-mile boundary was chosen to give the reader an
indication of which emissions sources and emissions source categories  could potentially have a
direct effect on the air quality at the monitoring site. Further, this boundary provides both the
proximity of emissions sources to the monitoring site as well as the quantity of such sources
within a given distance of the site. Sources outside the 10-mile radius are still visible on the map,
but have been grayed out in order to show emissions sources just outside the boundary.
Table 20-1 provides supplemental geographical information such as land use, location setting,
and locational coordinates.
                                          20-1

-------
                                   Figure 20-1. Providence, Rhode Island (PRRI) Monitoring Site
to
o

-------
           Figure 20-2. NEI Point  Sources Located Within 10 Miles of PRRI
                       Legend
                                                                  7V SO'fTW            71 "KWVf            71
                                                                 Note: Due to faculty density and collocation, the total facilities
                                                                 displayed may nol ropjcsont all facilities within the area of interest
                               PRRI  NATTS site           10 mile radius
                                      County boundary
Source Category Group (No. of Facilities)
 H^  Aircran Operations (13)
 $3  Automobile/Truck Manufacturing (2)
 ฃ  Battery (2)
 ฃ  Boat Manufacturing (1)
 C  Chemical Manufacturing {7}
 •  Concrete Batch Plant (2)
 X Crematory -Animal/Human (1)
 (D  Dry Cleaning Facility (37)
 6  Eleclncal Equipment (4)
 ฃ  Eleclncity Gerseralion via Combustion (4)
 E  Electroplating. Plating, Polishing, Anodizing & Coloring (24)
 <•>  Fabricated Metal Products (25)
id>  Flexible Polyurethane Foam Production (I >
 F   Food Processing/Agncullure {4)
[T]  Furniture Plani (2)
jf   Gasolinej'Diesel Service Station (4)
fV  Glass Manufacturing (1)
(3  Hospital (3)
J|   Hot Mix Asphalt Plant (4)
-^  Industnal Machinery and Equipment (5)
^  In stilutional - school (11)
 A   LandMI (1i
?   Miscellaneous Commercial/Industrial {21}
M   Miscellaneous Manufacturing (45)
4  Oil and/or Gas Production 11)
•-—.•  Pharmaceutical Manufacturing (1)
 1  Primary Metal Production (6)
P  Printing/Publishing (11)
El  PulP and Paper Plant/Wood Products (4)
R  Rubber and Miscellaneous Plastics Products (10)
 <  Site Remediation Activity (1 >
S  Surface Coaling (3)
TT  Telecommuntcatjons (4)
T  Textile Mril (12)
 I  V&stewater Treatment (1)
W  Waochvork Furmlure Milhvork & Wood Preserving (1)
                                                                20-3

-------
                                     Table 20-1. Geographical Information for the Rhode Island Monitoring Site

Site
Code
PRRI

AQS Code
44-007-0022

Location
Providence

County
Providence
Micro- or
Metropolitan
Statistical Area
Providence-New
Bedford-Fall
River, RI-MA
MSA
Latitude
and
Longitude
41.807949,
-71.415

Land Use
Residential

Location
Setting
Urban/City
Center

Additional Ambient Monitoring Information1
PAMS, VOCs, Carbonyl Compounds, Meteorological
parameters, PM10, PM10 Speciation, Black Carbon,
PM2 5, and PM2 5 Speciation, Germanium.
    1 Data for additional pollutants are reported to AQS for PRRI (EPA, 2012c); however, these data are not generated by ERG and are therefore not included in this report.

    BOLD ITALICS = EPA-designaled NATTS Site
to
o

-------
       Figure 20-1 shows that the areas to the west and south of PRRI are residential, but areas
to the north and east are commercial. A hospital lies to the northeast of the site, just north of
Dudley Street. About 1/2 mile to the east 1-95 runs north-south, then turns northwestward,
entering downtown Providence. Narragansett Bay and the Port of Providence are a few tenths of
a mile farther to the east, just on the other side of 1-95. Figure 20-2 shows that a large number of
point sources are located within 10 miles of PRRI, especially to the north of the site. Many of
these sources seem to parallel 1-95. The source categories with the greatest number of point
sources within 10 miles of PRRI include dry cleaners; fabricated metals products facilities;
electroplating, plating,  polishing, anodizing, and coloring facilities;  and aircraft operations,
which includes airports as well as small runways, heliports, or landing pads.

       Table 20-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Rhode Island monitoring site. Table 20-2 includes a county-level
population for the site.  County-level vehicle registration data for Providence County were not
available from the State of Rhode Island. Thus, state-level vehicle registration, which was
obtained from the Federal Highway Administration, was allocated to the county level using the
county-level proportion of the state population from the U.S. Census Bureau. Table 20-2 also
includes a county-level vehicle registration-to-population ratio, which was calculated to
represent the number of vehicles per person within the monitoring site's residing county. In
addition, the population within 10 miles of the site is presented, based on postal code population
data estimates. An estimate of 10-mile vehicle ownership was then determined by applying the
county-level vehicle registration-to-population ratio to the 10-mile population surrounding the
monitoring site. Table 20-2 also contains traffic volume information for PRRI. County-level
VMT data were not readily available for Providence County.
                                           20-5

-------
   Table 20-2. Population, Motor Vehicle, and Traffic Information for the Rhode Island
                                      Monitoring Site



Site
PRRI

Estimated
County
Population1
626,709

County-level
Vehicle
Registration2
485,837
Vehicles per
Person
(Registration:
Population)
0.78

Population
within 10
miles3
657,586
Estimated
10-mile
Vehicle
Ownership
509,773
Annual
Average
Daily
Traffic4
136,800

County-
level Daily
VMT5
~
1 County -level population estimate reflects 20 1 1 data from the U. S. Census Bureau (Census Bureau, 20 12b)
2County-level vehicle registration reflects a ratio based on 2010 state-level vehicle registration data from the
 FHWAandthe 2010 county -level proportion of the state population data (FHW A, 2011 and Census Bureau, 2011)
3 10-mile population estimate reflects 2011 data from Xionetic (Xionetic, 2011)
4AADT reflects 2009 data from the Rhode Island DOT (RI DOT, 2009)
5County -level VMT was not available for this site.
BOLD ITALICS = EPA-de signaled NATTS Site
       Observations from Table 20-2 include the following:

       •   Providence County's population is in the middle of the range compared to other
           counties with NMP sites, as is the 10-mile population.

       •   The estimated county-level vehicle registration is in the middle of the range compared
           to other counties with NMP sites,  as is the 10-mile ownership estimate.

       •   The vehicle-per-person ratio is in the bottom third compared to other NMP sites.

       •   The traffic volume experienced near PRRI is the ninth highest compared to other
           NMP monitoring sites. The traffic estimate provided is for 1-95 near the 1-195
           interchange.


20.2   Meteorological Characterization

       The following sections characterize the meteorological conditions near the monitoring

site in Rhode Island on sample days, as well as over the course of the year.
20.2.1 Climate Summary

       Providence is a coastal city on the Narragansett Bay, which opens to the Rhode Island

Sound and the Atlantic Ocean. The city's proximity to the Sound and the Atlantic Ocean temper

cold air outbreaks, and breezes off the ocean moderate summertime heat. On average, southerly

and southwesterly winds in the summer become northwesterly in the winter. Precipitation in

Providence is well distributed throughout the year. Weather is fairly variable as frequent storm

systems affect the New England region (Bair,  1992).
                                           20-6

-------
20.2.2 Meteorological Conditions in 2011
       Hourly meteorological data from the NWS weather station nearest PRRI were retrieved
for 2011 (NCDC, 2011). The closest weather station is located at Theodore F. Green State
Airport (WBAN 14765). Additional information about the T.F. Green weather station, such as
the distance between the site and the weather station, is provided in Table 20-3. These data were
used to determine how meteorological conditions on sample days vary from conditions
experienced throughout the year.

       Table 20-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for the entire year. Also included in  Table 20-3 is the 95 percent
confidence interval for each parameter. As shown in Table 20-3, average meteorological
conditions on sample days appear slightly cooler than average weather conditions throughout the
year, although the differences are not statistically significant. This is likely the result of several
make-up samples collected during the first quarter of the year.

20.2.3 Back Trajectory Analysis
       Figure 20-3 is the composite back trajectory map for days on which samples were
collected at the PRRI monitoring site in 2011. Included in Figure 20-3 are four back trajectories
per sample day. Figure 20-4 is the corresponding cluster analysis. An in-depth description of
these maps and how they were generated is presented in Section 3.5.2.1. For the composite map,
each line represents the 24-hour trajectory along which a parcel  of air traveled toward the
monitoring site on a given sample day and time, based on an initial height of 50 meters AGL. For
the cluster analysis, each line corresponds to a trajectory representative of a given cluster  of back
trajectories. Each concentric circle around the site in Figures 20-3 and 20-4 represents 100 miles.
                                          20-7

-------
to
o

oo
                           Table 20-3. Average Meteorological Conditions near the Rhode Island Monitoring Site
Closest NWS
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(ฐF)
Average
Temperature
(ฐF)
Average
Dew Point
Temperature
(ฐF)
Average
Wet Bulb
Temperature
(ฐF)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Providence, Rhode Island - PRRI
Theodore F.
Green State
Airport
14765
(41.72, -71.43)
6.01
miles
173ฐ
(S)
Sample
Day
2011
59.5
ฑ4.3
60.9
+ 1.8
51.3
ฑ4.1
52.6
+ 1.8
40.2
ฑ4.7
41.9
+ 1.9
46.4
ฑ4.0
47.7
+ 1.7
69.0
ฑ3.8
69.9
+ 1.5
1014.9
ฑ2.0
1015.4
+ 0.8
7.6
ฑ0.7
7.0
+ 0.3
         Sample day averages are highlighted to help differentiate the sample day averages from the full-year averages.

-------
Figure 20-3. 2011 Composite Back Trajectory Map for PRRI
    Figure 20-4. Back Trajectory Cluster Map for PRRI
                         20-9

-------
       Observations from Figures 20-3 and 20-4 for PRRI include the following:

       •  Back trajectories  originated from a variety  of directions at PRRI, although the
          majority  of back trajectories originated  from  the  northwest to  north  and south-
          southeast to south-southwest.

       •  The airshed domain for PRRI was among the larger in size compared to other NMP
          sites, based on the average trajectory length. The average trajectory  length was
          265 miles long, although the farthest away a back trajectory originated was off the
          North Carolina coast and over the Atlantic Ocean, or greater than 600 miles away.
          However, 87 percent of back trajectories originating within 450 miles of the site.

       •  The cluster analysis shows that nearly 50 percent of back trajectories originated from
          the west,  northwest, and north, although of differing lengths, as represented  by  three
          cluster  trajectories. One  originates over south Ontario,  Canada  (11 percent) and
          represents longer trajectories originating over  the  Great  Lakes region. The  short
          cluster trajectory originating over New York (16 percent) represents back trajectories
          originating over northern New York, Vermont, and New Hampshire as well as shorter
          northward originating trajectories. The  cluster trajectory originating to the  north
          (20 percent) represents longer trajectories originating over Quebec, Canada, as well
          as shorter back trajectories originating over Maine and the Gulf of Maine.  Twenty-
          nine percent of back trajectories originated to the southwest of PRRI, although this
          cluster trajectory also represents short back trajectories (generally 100-200  miles in
          length)  originating from the south to southwest to west of the site. Twelve percent of
          back trajectories originated over the offshore waters of the Mid-Atlantic states, while
          another 12 percent originated from a direction with an easterly component.


20.2.4 Wind Rose Comparison
       Hourly surface wind data from the NWS weather station at T.F. Green Airport near PRRI

were uploaded into a wind rose software program to produce customized wind roses, as

described in Section 3.5.2.2. A wind rose shows the frequency of wind directions using "petals"

positioned around a 16-point compass, and uses different colors to represent wind speeds.
       Figure 20-5 presents a map showing the distance between the NWS station and PRRI,

which may be useful for identifying topographical influences that may affect the meteorological

patterns experienced at this location. Figure 20-5 also presents three different wind roses for the

PRRI monitoring site. First, a historical wind rose representing 2001 to 2010 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 2011 is presented.

Next, a wind rose representing wind data for days on which samples were collected in 2011 is

presented. These can be used to identify the predominant wind speed and direction for 2011 and
                                          20-10

-------
determine if wind observations on sample days were representative of conditions experienced

over the entire year and historically.


       Observations from Figure 20-5 for PRRI include the following:

       •  The NWS weather station at T.F. Green Airport is located approximately 6 miles
          south of PRRI.

       •  The historical wind rose shows that while westerly winds were observed the most
          (approximately 12 percent of observations), winds from the western quadrants, due
          north, and due south are common near PRRI. Calm winds (< 2 knots) account for less
          than nine percent of the hourly measurements.

       •  The wind patterns shown on the 2011 wind rose are similar to the historical wind
          patterns,  with winds from the western quadrants, due north, and due south prevalent
          near PRRI, although there are some slight differences in the percentages. Also, the
          calm rate for 2011 is nearly 12 percent, which is  slightly higher than the calm rate for
          the historical wind rose.

       •  The wind patterns shown on the sample day wind rose continue the prevalence of
          winds from the western quadrants and due north  and due south, although the
          percentages vary for some directions. The sample day calm rate is similar to the full-
          year calm rate. These similarities indicate that conditions on sample days were
          generally representative of conditions experienced throughout the year and
          historically.
                                         20-11

-------
Figure 20-5. Wind Roses for the T.F. Green State Airport Weather Station near PRRI
 Distance between PRRI and NWS Station
2001-2010 Historical Wind Rose
                                             .'VEST
          2011 Wind Rose
    Sample Day Wind Rose
WEST
                                     20-12

-------
20.3   Pollutants of Interest
       Site-specific "pollutants of interest" were determined for the Rhode Island monitoring
site in order to allow 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." 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. In
addition, if any of the NATTS MQO Core Analytes measured by the monitoring site did not
meet the pollutant of interest criteria based on the preliminary risk-based screening, that pollutant
was added to the list of site-specific pollutants of interest. A more in-depth description of the
risk-based screening process is presented in  Section 3.2.

       Table 20-4 presents the results of the preliminary risk-based screening process for PRRI.
The pollutants that failed at least one screen and contributed to 95 percent of the total failed
screens for PRRI are shaded. NATTS MQO Core Analytes are bolded. Thus, pollutants  of
interest are shaded and/or bolded. PRRI sampled for PAHs and hexavalent chromium.

      Table 20-4. Risk-Based Screening Results for the Rhode Island Monitoring Site
Pollutant
Screening
Value
(Ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Providence, Rhode Island - PRRI
Naphthalene
Fluorene
Acenaphthene
Fluoranthene
Hexavalent Chromium
0.029
0.011
0.011
0.011
0.000083
Total
57
4
3
1
1
66
57
57
57
57
52
280
100.00
7.02
5.26
1.75
1.92
23.57
86.36
6.06
4.55
1.52
1.52
86.36
92.42
96.97
98.48
100.00

       Observations from Table 20-4 include the following:
       •  Five pollutants failed screens for PRRI. Naphthalene failed 100 percent of its screens
          and accounted for 86 percent of PRRI's total failed screens.
       •  Naphthalene, fluorene, and acenaphthene were identified as the pollutants of interest
          for PRRI based on the risk-based screening process. Hexavalent chromium was added
          to the pollutants of interest for PRRI because is it a NATTS MQO Core Analyte, even
          though it did not contribute to 95 percent of failed screens. Benzo(a)pyrene was
          added to the pollutants of interest for PRRI because it is also a NATTS MQO Core
                                          20-13

-------
          Analyte, even though it did not fail any screens. Benzo(a)pyrene is not shown in
          Table 20-4 but is shown in subsequent tables in the sections that follow.

20.4   Concentrations
       This section presents various concentration averages used to characterize pollution levels
at the Rhode Island monitoring site. Where applicable, the following calculations and data
analyses were performed: Time period-based concentration averages (quarterly and annual) are
provided for the pollutants of interest for PRRI, where the data meet the applicable criteria.
Concentration averages for select pollutants are also presented graphically for the site to
illustrate how the site's concentrations compare to the program-level averages, as presented in
Section 4.1. In addition, concentration averages for select pollutants are presented from previous
years of sampling in order to characterize concentration trends at the site. Additional site-specific
statistical summaries for PRRI are provided in Appendices M and O.

20.4.1 2011 Concentration Averages
       Quarterly and annual concentration averages were calculated for the pollutants of interest
for the Rhode Island site, as described in Section 3.1. The quarterly average of a particular
pollutant is simply the average concentration of the preprocessed daily measurements over a
given calendar quarter. Quarterly average concentrations include the substitution of zeros for all
non-detects. A site must have a minimum of 75 percent valid samples of the total number of
samples possible within a given quarter for a quarterly average to be calculated. An annual
average includes all measured detections and substituted zeros for non-detects for the entire year
of sampling. Annual averages were calculated for pollutants where three valid quarterly averages
could be calculated and where method completeness was greater than or equal to 85 percent, as
presented in Section 2.4.  Quarterly and annual average concentrations for PRRI are presented in
Table 20-5, where applicable. Note that if a pollutant was not detected in a given calendar
quarter, the quarterly average simply reflects "0" because only zeros substituted for non-detects
were factored into the quarterly average  concentration.
                                          20-14

-------
  Table 20-5. Quarterly and Annual Average Concentrations of the Pollutants of
                  Interest for the Rhode Island Monitoring Site



Pollutant
#of
Measured
Detections
vs. # of
Samples

1st
Quarter
Average
(ng/m3)

2nd
Quarter
Average
(ng/m3)

3rd
Quarter
Average
(ng/m3)

4th
Quarter
Average
(ng/m3)


Annual
Average
(ng/m3)
Providence, Rhode Island - PRRI

Acenaphthene

Benzo(a)pyrene

Fluorene

Hexavalent Chromium

Naphthalene

57/57

56/57

57/57

52/57

57/57
1.39
ฑ0.30
0.20
ฑ0.07
2.78
ฑ0.50
0.02
ฑ0.01
84.66
ฑ 22.05
4.35
ฑ1.60
0.17
ฑ0.06
5.97
ฑ2.25
0.02
ฑ0.01
70.92
ฑ 20.09
6.61
ฑ2.06
0.10
ฑ0.03
8.13
ฑ2.24
0.03
ฑ0.02
108.89
ฑ38.22
2.33
ฑ0.79
0.16
ฑ0.06
3.13
ฑ0.74
0.02
ฑ0.01
100.52
ฑ 32.64
3.65
ฑ0.83
0.16
ฑ0.03
4.97
ฑ0.96
0.02
ฑ0.01
91.41
ฑ 14.27
Observations for PRRI from Table 20-5 include the following:

•  The annual average concentration of naphthalene is significantly higher than the
   annual averages of the other pollutants of interest.

•  Although the third and fourth quarter average concentrations of naphthalene are
   higher than the other quarterly averages, the associated confidence intervals indicate
   that there is a high level of variability  associated with the naphthalene concentrations.
   Naphthalene measurements range from 29.7 ng/m3 to 277 ng/m3, with a median
   concentration of 78.5 ng/m3. Of the six concentrations of naphthalene greater than
   150 ng/m3, one was measured in February, one in July, two in August, and two in
   November. The minimum concentration of naphthalene was also measured in
   November.

•  Concentrations of acenaphthene and fluorene are highest during the warmer months
   of the year, as indicated by the second and third quarter averages, although they have
   relatively high levels of variability associated with them.  The 11 concentrations of
   fluorene greater than 7 ng/m3 were all measured between May and August, while the
   20 concentrations less than 3 ng/m3 were measured between January and April or
   October and December. A similar trend is exhibited by the acenaphthene
   measurements.

•  Although the annual average benzo(a)pyrene concentration is relatively low
   compared to the other pollutants of interest for PRRI, this annual average is the fourth
   highest annual average benzo(a)pyrene concentration among sites sampling PAHs, as
   shown in Table 4-11.
                                  20-15

-------
20.4.2 Concentration Comparison
       In order to better illustrate how a site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for the selected NATTS MQO Core
Analytes listed in Section 3.5.3, where applicable. Thus, box plots were created for
benzo(a)pyrene, hexavalent chromium, and naphthalene for PRRI. Figures 20-6 through 20-8
overlay the site's minimum, annual average, and maximum concentrations onto the program-
level minimum, first quartile, median, average, third quartile, and maximum concentrations, as
described in Section 3.5.3.
       Figure 20-6. Program vs. Site-Specific Average Benzo(a)pyrene Concentration
 PRRI
              3.25
                         3.5
                                    0.75         1
                                          Concentration (
                                                         1.25
                                                                    1.5
                                                                               1.75
                Program:   IstQuartile     2ndQuartile     3rdQuartile     4thQuartile     Average
                Site:      Site Average     Site Minimum/Maximum

Figure 20-7. Program vs. Site-Specific Average Hexavalent Chromium Concentration
 PRRI
I
                  0.05
                                0.1
                                              0.15
                                          Concentration (ng/m3)
                                                             0.2
                                                                           DL25
                                                                                          3.3
                Program:   IstQuartile     2ndQuartile     SrdQuartile     4thQuartile     Average
                Site:     Site Average     Site Minimum/Maximum

                                            20-16

-------
       Figure 20-8. Program vs. Site-Specific Average Naphthalene Concentration
PRRI
                                                        i Program Max Concentration = 779 ng/m3
           50
                   100
                           is:
                                   200       Z50      300
                                       Concentration (ng/m3)
                                                           = 50
                                                                   400
                                                                           453
                                                                                   555
              Program:   IstQuartile     2ndQuartile    3rd Quartile     4thQuartile     Average
              Site:
                      Site Average     Site Minimum/Maximum

      Observations from Figures 20-6 through 20-8 include the following:

         •  Figure 20-6 is the box plot for benzo(a)pyrene. Note that the program-level first
            quartile for this pollutant is zero and is not visible on the box plot. This box plot
            shows that the annual average concentration for PRRI is greater than the program-
            level average concentration. The maximum concentration measured at PRRI is
            considerably less than the maximum concentration measured across the program.
            A single non-detect of benzo(a)pyrene was measured at PRRI.

         •  Figure 20-7 is the box plot for hexavalent chromium. The annual  average
            concentration of hexavalent chromium for PRRI is just less than the program-
            level average but greater than program-level median concentration. The maximum
            concentration measured at PRRI is less than the program-level maximum
            concentration, although the maximum hexavalent chromium concentration for
            PRRI is the  ninth highest concentration measured among NMP sites sampling this
            pollutant. There were five non-detects of hexavalent chromium measured at
            PRRI.

         •  Figure 20-8 is the box plot for naphthalene. Note that the program-level
            maximum concentration (779 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
            500 ng/m3. Figure  20-8 shows that the annual average naphthalene concentration
            for PRRI is  greater than the program-level average concentration. The maximum
            naphthalene concentration measured at PRRI is less than the maximum
            concentration measured at the program-level while the minimum  concentration
            measured at PRRI  is just less  than the program-level first quartile. There were no
            non-detects  of naphthalene measured at PRRI.
                                        20-17

-------
20.4.3  Concentration Trends
       A site-specific trends evaluation was completed for sites that have sampled one or more
of the selected NATTS MQO Core Analytes for 5 consecutive years or longer, as described in
Section 3.5.4. PRRI has sampled hexavalent chromium under the NMP since 2005. Thus,
Figure 20-9 presents the annual statistical metrics for hexavalent chromium for PRRI. The
statistical metrics presented for assessing trends include the substitution of zeros for non-detects.
A trends analysis was not performed for the PAHs because PAH sampling did not begin at PRRI
until 2008.
    Figure 20-9. Annual Statistical Metrics for Hexavalent Chromium Concentrations
                                  Measured at PRRI
                                            2008
                                            Year
                -  Minimum
                                   - Maximum
                                             * 95thPercentile
                                                                       • Average
       Observations from Figure 20-9 for hexavalent chromium measurements at PRRI include
the following:
       •  The maximum hexavalent chromium concentration was measured on
          August 28, 2007 (0.193 ng/m3), although a similar concentration was also measured
          on July 4, 2006 (0.192 ng/m3). The third highest concentration was measured in 2011
          (July 20, 2011,0.147 ng/m3).
                                         20-18

-------
       •  The average concentration of hexavalent chromium has fluctuated over the years of
          sampling, with the average at a maximum in 2006 (0.027 ng/m3) and a minimum in
          2009 (0.007 ng/m3). However, an increasing trend is shown over the last two year
          years of sampling.
       •  For each year shown, the minimum and 5th percentile are zero, indicating the presence
          of non-detects. The number of non-detects reported has varied by year, from as low as
          nine percent in 2011 to as high as 65 percent in 2009. This explains why the median
          concentration is also zero for 2009.

20.5   Additional Risk-Based Screening Evaluations
       The following risk-based screening evaluations were conducted to characterize risk at the
PRRI monitoring site. Refer to Sections 3.3 and 3.5.5 for definitions and explanations regarding
the various toxicity factors, time frames, and calculations associated with these risk-based
screenings.

20.5.1 Risk-Based Screening Assessment Using MRLs
       A risk-based screening was conducted by comparing the concentration data from the
Rhode Island monitoring site to the ATSDR MRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 to 14 days); intermediate (exposures of 15  to 364 days); and chronic (exposures
of 1 year or greater). The preprocessed daily measurements of the pollutants of interest were
compared  to the acute MRLs; the quarterly averages were compared to the intermediate MRLs;
and the annual averages were compared to the chronic MRLs.

       As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2011.

20.5.2 Cancer Risk and Noncancer Hazard Approximations
       For the pollutants of interest for the Rhode Island monitoring site and where annual
average concentrations could be calculated, risk was examined by calculating cancer risk and
noncancer hazard approximations. These approximations can be used as risk estimates for cancer
and noncancer effects attributable to the pollutants of interest.  Although the use of these
approximations is limited, they may help identify where  policy-makers may want to shift or
                                         20-19

-------
  confirm their air-monitoring priorities. Refer to Section 3.5.5.2 for an explanation of how cancer

  risk and noncancer hazard approximations are calculated and what limitations are associated with
  them. Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer

  hazard approximations are presented in Table 20-6, where applicable. Cancer risk

  approximations are presented as probabilities while the noncancer hazard approximations are

  ratios and thus, unitless values.


            Table 20-6. Risk Approximations for the Rhode Island Monitoring Site
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Providence, Rhode Island - PRRI
Acenaphthene
Benzo(a)pyrene
Fluorene
Hexavalent Chromium
Naphthalene
0.000088
0.00176
0.000088
0.012
0.000034



0.0001
0.003
57/57
56/57
57/57
52/57
57/57
3.65
ฑ0.83
0.16
ฑ0.03
4.97
ฑ0.96
0.02
ฑ0.01
91.41
ฑ 14.27
0.32
0.28
0.44
0.27
3.11



O.01
0.03
— = a Cancer URE or Noncancer RfC is not available
        Observations for PRRI from Table 20-6 include the following:

        •   As discussed in the previous section, naphthalene has the highest annual average
            concentration among the pollutants of interest for PRRI, followed by fluorene and
            acenaphthene.

        •   The cancer risk approximation for naphthalene (3.11 in-a-million) is the highest
            cancer risk approximation among the pollutants of interest for PRRI and the only
            cancer risk approximation greater than 1-in-a-million.

        •   Only two of the five pollutants of interest for PRRI have noncancer RfCs (hexavalent
            chromium and naphthalene).The noncancer hazard approximations for naphthalene
            and hexavalent chromium are negligible (0.03 in-a-million and <0.01 in-a-million,
            respectively), indicating that no adverse health effects are expected from these
            individual pollutants.
                                           20-20

-------
20.5.3 Risk-Based Emissions Assessment
       In addition to the risk-based screenings discussed above, Tables 20-7 and 20-8 present an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 20-7 presents the 10 pollutants with the highest emissions from the 2008 NEI, the
10 pollutants with the highest toxicity-weighted emissions, and the 10 pollutants with the highest
cancer risk approximations (in-a-million), as calculated from the annual averages provided in
Table 20-6. Table 20-8 presents similar information, but identifies the 10 pollutants with the
highest noncancer hazard approximations, also calculated from annual averages provided in
Table 20-6.

       The pollutants listed in Tables 20-7 and 20-8 are limited to those that have cancer and
noncancer toxicity factors, respectively. As a result, although the actual value of the emissions is
the same, the highest emitted pollutants in the cancer table may be different from the noncancer
table. The cancer risk and noncancer hazard approximations based on the site's annual averages
are limited to those pollutants for which each respective site sampled. As discussed in
Section 20.3, PRRI sampled for PAHs and hexavalent chromium only. In addition, the cancer
risk and noncancer hazard approximations are limited to those pollutants with enough data to
meet the criteria for annual averages to be calculated. A more in-depth discussion of this analysis
is provided in Section 3.5.5.3. Similar to the cancer risk and noncancer hazard approximations,
this analysis may help policy-makers prioritize their air monitoring activities.
                                          20-21

-------
       Table 20-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer

                                          UREs for the Rhode Island Monitoring Site
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Providence, Rhode Island (Providence County) - PRRI
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Tetrachloroethylene
Naphthalene
Dichloromethane
Trichloroethylene
POM, Group 2b
204.49
155.91
96.49
84.05
30.18
27.44
17.46
8.03
6.19
4.97
Formaldehyde
Benzene
POM, Group 3
1,3 -Butadiene
Naphthalene
POM, Group 2b
Hexavalent Chromium, PM
POM, Group 5a
Ethylbenzene
Arsenic, PM
2.03E-03
1.60E-03
1.18E-03
9.05E-04
5.93E-04
4.37E-04
3.47E-04
2.51E-04
2.41E-04
1.98E-04
Naphthalene
Fluorene
Acenaphthene
Benzo(a)pyrene
Hexavalent Chromium
3.11
0.44
0.32
0.28
0.27

to
o

to
to

-------
       Table 20-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with

                                     Noncancer RfCs for the Rhode Island Monitoring Site
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Noncancer Hazard
Approximation
Pollutant (HQ)
Providence, Rhode Island (Providence County) - PRRI
Toluene
Xylenes
Methanol
Benzene
Formaldehyde
Hexane
Ethylbenzene
Acetaldehyde
Ethylene glycol
1,3 -Butadiene
576.48
398.70
368.54
204.49
155.91
128.27
96.49
84.05
38.49
30.18
Acrolein
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Benzene
Naphthalene
Xylenes
Nickel, PM
Trichloroethylene
Arsenic, PM
335,397.31
15,908.87
15,090.50
9,339.31
6,816.26
5,818.50
3,986.98
3,977.28
3,096.76
3,072.81
Naphthalene 0.03
Hexavalent Chromium O.01

to
o

to

-------
Observations from Table 20-7 include the following:

•  Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
   cancer UREs in Providence County.

•  Formaldehyde is also the pollutant with the highest toxicity-weighted emissions (of
   the pollutants with cancer UREs), followed by benzene and POM, Group 3.

•  Six of the highest emitted pollutants also have the highest toxicity-weighted
   emissions for Providence County.

•  Naphthalene, which has the highest cancer risk approximation among the pollutants
   of interest for PRRI, has the seventh highest emissions and the fifth highest toxicity-
   weighted emissions.

•  POM, Group 2b is the tenth highest emitted "pollutant" in Providence County and
   ranks sixth for toxicity-weighted emissions. POM, Group 2b includes several PAHs
   sampled for at PRRI including acenaphthene, fluoranthene, and fluorene, all of which
   failed at least one screen for PRRI.

•  POM, Group 5a ranks eighth for toxicity-weighted emissions. POM, Group 5a
   includes benzo(a)pyrene, another pollutant of interest for PRRI. POM, Group 5a is
   not among the highest emitted "pollutants"  in Providence County.

•  Hexavalent chromium, which is also one of the pollutants of interest for PRRI, has
   the seventh highest toxicity-weighted emissions for Providence  County, but does not
   appear among the highest emitted pollutants.


Observations from Table 20-8 include the following:

•  Toluene, xylenes, and methanol are the highest emitted pollutants with noncancer
   RfCs in Providence County.

•  The pollutants with the highest toxicity-weighted emissions (of the pollutants with
   noncancer RfCs) are acrolein, formaldehyde, and 1,3-butadiene.

•  Five of the highest emitted pollutants in Providence County also have the highest
   toxicity-weighted emissions.

•  While naphthalene ranks sixth among the pollutants with the highest toxicity-
   weighted emissions, it is not one of the highest emitted pollutants (with a noncancer
   RfC) in Providence County. Hexavalent chromium does not appear on either
   emissions-based list. These are the only two pollutants of interest with noncancer
   RfCs for PRRI.
                                  20-24

-------
20.6   Summary of the 2011 Monitoring Data for PRRI

       Results from several of the data treatments described in this section include the
following:

       ปซป  Five pollutants failed at least one screen for PRRI, with naphthalene accounting for
          the majority of the failed screens.

       *ป*  Of the site-specific pollutants of the interest, naphthalene had the highest annual
          average concentration for PRRI.

       ปซป  Concentrations ofhexavalent chromium have an increasing trend for the most recent
          years of sampling at PRRI.
                                         20-25

-------
21.0   Site in South Carolina
       This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS site in South Carolina, and integrates these
concentrations with emissions, meteorological, and risk information. Data generated by sources
other than ERG are not included in the data analyses contained in this report. Readers are
encouraged to refer to Sections 1 through 4 and the glossary (Appendix P) for detailed
discussions and definitions regarding the various data analyses presented below.

21.1   Site Characterization
       This section characterizes the South Carolina monitoring site by providing geographical
and physical information about the location of the site and the surrounding area. This
information is provided to give the reader insight regarding factors that may influence the air
quality near the site and assist in the interpretation of the ambient monitoring measurements.

       CHSC is located in central Chesterfield County, South Carolina. Figure 21-1 is a
composite satellite image retrieved from ArcGIS Explorer showing the monitoring site in its
rural location. Figure 21-2 identifies nearby point source emissions locations by source category,
as reported in the 2008 NEI for point sources. Note that only sources within 10 miles of the site
are included in the facility counts provided in Figure 21-2. A 10-mile boundary was chosen
to give the reader an indication of which emissions sources and emissions source categories
could potentially have a direct effect on the air quality at the monitoring site. Further, this
boundary provides both the proximity of emissions sources to the monitoring site as well as the
quantity of such sources within a given distance of the site. Sources outside the 10-mile radius
are still visible on the map, but have been grayed out in order to show emissions sources just
outside the boundary. Table 21-1 provides  supplemental geographical information such as land
use, location setting, and locational coordinates.
                                          21-1

-------
                                 Figure 21-1. Chesterfield, South Carolina (CHSC) Monitoring Site
to

-------
Figure 21-2. NEI Point Sources Located Within 10 Miles of CHSC
       Legend
                        eo"2ffo-w         eg115'crw         s>' itrtrw         srs'trw
                                     Note: Due to facility density and collocation, the total facilities
                                     displayed may not represent all facilities within the area of interest.
             CHSC NATTS site       10 mile radius [	| County boundary

                 Source Category Group (No. of Facilities)
                   •f   Aircraft Operations (1)
                   -$•   Industrial Machinery and Equipment (1)
                                    21-3

-------
                             Table 21-1. Geographical Information for the South Carolina Monitoring Site

Site
Code
CHSC

AQS Code
45-021-0001

Location
Not in a
city

County
Chesterfield
Micro- or
Metropolitan
Statistical Area
Not in an MSA
Latitude
and
Longitude
34.615367,
-80.198789

Land Use
Forest

Location
Setting
Rural

Additional Ambient Monitoring Information1
TSP, TSP Metals, VOCs, O3, Meteorological
parameters, PM10, PM10 Speciation, PM25, and
PM2 5 Speciation, Carbonyl Compounds,
Hexachlorobutadiene.
  BOLD ITALICS = EPA-designated NATTS Site
to

-------
        CHSC is located about 15 miles south of the North Carolina/South Carolina border,
 between the towns of McBee and Chesterfield. The monitoring site is located near the Ruby fire
 tower and, as Figure 21-1 shows, is located just off State Highway 145. The surrounding area is
 rural in nature and is part of the Carolina Sandhills Wildlife Refuge. Figure 21-2 shows that few
 point sources are located within  10 miles of CHSC, the closest of which is the Wild Irish Rose
 Airport.

        Table 21-2 presents additional site-characterizing information, including indicators of
 mobile source activity, for the South Carolina monitoring site. Table 21-2 includes county-level
 population and vehicle registration information. Table 21-2 also includes a county-level vehicle
 registration-to-population ratio, which was calculated to represent the number of vehicles per
 person within each monitoring site's residing county. In addition, the population within 10 miles
 of the site is presented, based on postal code population data estimates. An estimate of 10-mile
 vehicle ownership was then determined by applying the county-level vehicle registration-to-
 population ratio to the 10-mile population surrounding the monitoring site. Table 21-2 also
 contains traffic volume information for CHSC. Finally, Table 21-2 presents the daily VMT for
 Chesterfield County.
   Table 21-2. Population, Motor Vehicle, and Traffic Information for the South Carolina
                                      Monitoring Site



Site
CHSC

Estimated
County
Population1
46,557

County-level
Vehicle
Registration2
40,792
Vehicles per
Person
(Registration:
Population)
0.88

Population
within 10
miles3
5,538
Estimated
10-mile
Vehicle
Ownership
4,852
Annual
Average
Daily
Traffic4
550

County-
level Daily
VMT5
1,276,517
Bounty-level population estimate reflects 2011 data from the U.S. Census Bureau (Census Bureau, 2012b)
2County-level vehicle registration reflects 2011 data from the South Carolina DMV (SC DMV, 2011)
310-mile population estimate reflects 2011 data from Xionetic (Xionetic, 2011)
4AADT reflects 2011 data from the South Carolina DOT (SC DOT, 2012a)
5County-level VMT reflects 2011 data from the South Carolina DOT (SC DOT, 2012b)
BOLD ITALICS = EPA-designated NATTS Site

        Observations from Table 21-2 include the following:
        •  Chesterfield County's population is among the lowest compared to other counties
           with NMP sites. This site's 10-mile population is the second lowest among NMP
           sites, behind only CAMS  85 (in Texas). Similar rankings were found for both the
           county-level and 10-mile vehicle ownerships.

        •  The vehicle-per-person ratio is in the middle of the range among NMP sites.
                                            21-5

-------
       •  The traffic volume experienced near CHSC ranks among the lowest compared to
          other NMP monitoring sites. The traffic estimate provided is for State Highway 145
          between State Highway 109 and US-1.
       •  The daily VMT for Chesterfield County is the third lowest VMT compared to other
          counties with NMP sites (where VMT data were available).

21.2   Meteorological Characterization
       The following sections characterize the meteorological conditions near the monitoring
site in South Carolina on sample days, as well as over the course of the year.

21.2.1  Climate Summary
       The town of Chesterfield is located just south of the North Carolina/South Carolina
border, about 35 miles northwest of the city of Florence. Although the area experiences all four
seasons, South Carolina's southeastern location ensures mild winters and long, hot summers.
Summers are dominated by the Bermuda high pressure system over the Atlantic Ocean, which
allows southwesterly winds to prevail, bringing in warm, moist air out of the Gulf of Mexico.
During winter, winds out of the southwest shift northeasterly after frontal systems move across
the area. Chesterfield County leads the state in the average number of sleet and freezing rain
events per year (Bair, 1992 and SC SCO, 2013).

21.2.2  Meteorological Conditions in 2011
       Hourly meteorological data from the NWS weather  station nearest this site were retrieved
for 2011 (NCDC, 2011). The closest weather station with adequate data is located at the Monroe
Airport in Monroe, North  Carolina (WBAN 53872). Additional information about the Monroe
Airport weather station, such as the distance between the site and the weather station, is provided
in Table 21-3. These data  were used to determine how meteorological conditions on sample days
vary from conditions experienced throughout the year.
                                          21-6

-------
                          Table 21-3. Average Meteorological Conditions near the South Carolina Monitoring Site
Closest NWS
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(ฐF)
Average
Temperature
(ฐF)
Average
Dew Point
Temperature
(ฐF)
Average
Wet Bulb
Temperature
(ฐF)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Chesterfield, South Carolina - CHSC
Monroe Airport
53872
(35.02, -80.62)
35.81
miles
311ฐ
(NW)
Sample
Day
2011
71.4
ฑ4.1
72.3
+ 1.6
61.4
ฑ3.9
61.7
+ 1.6
49.6
ฑ4.5
50.4
+ 1.7
55.3
ฑ3.7
55.6
+ 1.4
69.1
ฑ3.6
70.2
+ 1.3
1017.6
ฑ 1.6
1017.7
+ 0.7
4.8
ฑ0.6
4.8
+ 0.3
         Sample day averages are highlighted to help differentiate the sample day averages from the full-year averages.
to

-------
       Table 21-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for the entire year. Also included in Table 21-3 is the 95 percent
confidence interval for each parameter. As shown in Table 21-3, average meteorological
conditions on sample days were representative of average weather conditions throughout the
year.

21.2.3 Back Trajectory Analysis
       Figure 21-3 is the composite back trajectory map for days on which samples were
collected at the CHSC monitoring site in 2011.  Included in Figure 21-3 are four back trajectories
per sample day. Figure 21-4 is the corresponding cluster analysis. An in-depth description of
these maps and how they were generated is presented in Section 3.5.2.1. For the composite map,
each line represents the 24-hour trajectory along which a parcel of air traveled toward the
monitoring site on a given sample day and time, based on an initial height of 50 meters AGL. For
the cluster analysis, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric circle around the site in Figures 21-3 and 21-4 represents 100 miles.

       Observations from Figures 21-3 and 21-4 for CHSC include the following:
       •  Back trajectories originated from a variety of directions at CHSC.
       •  The 24-hour air shed domain for CHSC was similar in size to other NMP monitoring
          sites.  The farthest away a back trajectory originated was over central Michigan, or
          greater than 600 miles away. However, the average trajectory length was 206 miles
          and 81 percent of back trajectories originated within 300 miles of the site.
       •  The cluster analysis shows that 38 percent of back trajectories originated from the
          southwest to west to northwest of CHSC, primarily over western North and  South
          Carolina and northern Georgia. Although most of these back trajectories originated
          less than 300 miles away from the site, a few longer back trajectories originating to
          the west of the site are also included. Eleven percent of back trajectories also
          originated from the northwest of CHSC but of longer length (greater than 300 miles
          away). Another 26 percent of back trajectories originated to the north to northeast of
          CHSC, varying between 300 and 600 miles in length. Nearly 20 percent of back
          trajectories are represented by the short cluster originating near the coast of  South
          Carolina. These back trajectories originated to the northeast,  east, and southeast of
          CHSC, over eastern North and South Carolina and their offshore waters. Fewer than
           10 percent of back trajectories originated farther south, over the offshore waters of
          Georgia and Florida.

                                          21-8

-------
Figure 21-3. 2011 Composite Back Trajectory Map for CHSC
    Figure 21-4. Back Trajectory Cluster Map for CHSC
                         21-9

-------
21.2.4 Wind Rose Comparison
       Hourly surface wind data from the NWS weather station at Monroe Airport near CHSC
were uploaded into a wind rose software program to produce customized wind roses, as
described in Section 3.5.2.2. A wind rose shows the frequency of wind directions using "petals"
positioned around a 16-point compass, and uses different colors to represent wind speeds.

       Figure 21-5 presents a map showing the distance between the NWS station and CHSC,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 21-5 also presents three different wind roses for the
CHSC monitoring site. First, a historical wind rose representing 2001 to 2010 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 2011 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2011 is
presented. These can be used to identify the predominant wind speed and direction for 2011 and
determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.

       Observations from Figure 21-5 for CHSC include the following:
       •  The Monroe Airport weather station is located across the North Carolina/South
          Carolina border, approximately 36 miles northwest of CHSC.
       •  The historical wind rose for CHSC shows that calm winds (< 2 knots) account for
          23 percent of the hourly measurements.  Winds from the south-southwest to west
          account for approximately one-third of observations, just slightly more than winds
          from the north to east-northeast. Winds from the southeast quadrant are generally not
          observed.
       •  The wind patterns shown on the 2011 wind rose for CHSC are similar to the historical
          wind patterns, although there were slightly more calm observations and fewer winds
          observations from the northeast quadrant. This indicates that wind conditions in 2011
          were similar to what is expected climatologically near this site.
       •  The sample day wind patterns for 2011 also resemble the historical and full-year wind
          patterns.  However, there were fewer observations from the southwest quadrant and
          more observations from the northeast quadrant compared to the full-year wind rose.
                                         21-10

-------
    Figure 21-5. Wind Roses for the Monroe Airport Weather Station near CHSC
Distance between CHSC and NWS Station
2001-2010 Historical Wind Rose
         2011 Wind Rose
    Sample Day Wind Rose
                                    21-11

-------
21.3   Pollutants of Interest
       Site-specific "pollutants of interest" were determined for the South Carolina monitoring
site in order to allow 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." 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. In
addition, if any of the NATTS MQO Core Analytes measured by the monitoring site did not
meet the pollutant of interest criteria based on the preliminary risk-based screening, that pollutant
was added to the list of site-specific pollutants of interest. A more in-depth description of the
risk-based screening process is presented in  Section 3.2.

       Table 21-4 presents the results of the preliminary risk-based screening process for CHSC.
The pollutants that failed at least one screen and contributed to 95 percent of the total failed
screens for the monitoring site are shaded. NATTS MQO Core Analytes are bolded. Thus,
pollutants of interest are shaded and/or bolded. CHSC sampled hexavalent chromium and PAHs.

     Table 21-4. Risk-Based Screening Results for the South Carolina Monitoring Site
Pollutant
Screening
Value
(Ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
Chesterfield, South Carolina - CHSC
Naphthalene
0.029
Total
3
3
60
60
5.00
5.00
100.00
100.00

       Observations from Table 21-4 include the following:
       •  Naphthalene was the only pollutant to fail screens for CHSC. This pollutant was
          detected in all 60 valid samples collected at CHSC and failed three screens, or
          approximately 5 percent of screens.
       •  This site has the third lowest number of failed screens (3) among all NMP sites.
       •  Benzo(a)pyrene and hexavalent chromium were added to CHSC's pollutants of
          interest because they are NATTS MQO Core Analytes, even though they did not fail
          any screens. These pollutants are not shown in Table 21-4 but are shown in
          subsequent tables in the sections that follow.
                                          21-12

-------
21.4   Concentrations
       This section presents various concentration averages used to characterize pollution levels
at the South Carolina monitoring site. Where applicable, the following calculations and data
analyses were performed: Time period-based concentration averages (quarterly and annual) are
provided for the pollutants of interest for CHSC, where the data meet the applicable criteria.
Concentration averages for select pollutants are also presented graphically for the site to
illustrate how the site's concentrations compare to the program-level averages, as presented in
Section 4.1. In addition, concentration averages for select pollutants are presented from previous
years of sampling in order to characterize concentration trends at the site. Additional site-specific
statistical summaries for CHSC are provided in Appendices M and O.

21.4.1 2011 Concentration Averages
       Quarterly and annual concentration averages were calculated for the pollutants of interest
for the South Carolina  site, as described in Section 3.1.  The quarterly average of a particular
pollutant is simply the  average concentration of the preprocessed daily measurements over a
given calendar quarter. Quarterly average concentrations include the substitution of zeros for all
non-detects. A site must have a minimum of 75 percent valid samples of the total number of
samples possible within a given quarter for a quarterly average to be calculated. An annual
average includes all measured detections and substituted zeros for non-detects for the entire year
of sampling. Annual averages were calculated for pollutants where three valid quarterly averages
could be calculated and where method completeness was greater than or equal to 85 percent, as
presented in Section 2.4. Quarterly and annual  average concentrations for CHSC are  presented in
Table 21-5, where applicable. Note that if a pollutant was not detected in a given calendar
quarter, the quarterly average simply reflects "0" because only zeros substituted for non-detects
were factored into the quarterly average concentration.
                                          21-13

-------
  Table 21-5. Quarterly and Annual Average Concentrations of the Pollutants of
                 Interest for the South Carolina Monitoring Site



Pollutant
#of
Measured
Detections
vs. # of
Samples

1st
Quarter
Average
(ng/m3)

2nd
Quarter
Average
(ng/m3)

3rd
Quarter
Average
(ng/m3)

4th
Quarter
Average
(ng/m3)


Annual
Average
(ng/m3)
Chesterfield, South Carolina - CHSC

Benzo(a)pyrene

Hexavalent Chromium

Naphthalene

7/60

29/60

60/60
0.02
ฑ0.03
<0.01
ฑ<0.01
25.11
ฑ 15.57

0
0.01
ฑ<0.01
12.38
ฑ2.45
0.03
ฑ0.05
0.01
ฑ<0.01
14.37
ฑ3.96
0.01
ฑ0.01
<0.01
ฑ<0.01
13.81
ฑ3.82
0.01
ฑ0.02
0.01
ฑ<0.01
16.42
ฑ4.13
Observations for CHSC from Table 21-5 include the following:

•  The annual average concentration of naphthalene is three orders of magnitude higher
   than the annual average concentrations of hexavalent chromium and benzo(a)pyrene.

•  Compared to other NMP sites, CHSC has the second lowest annual average
   concentration for each of these pollutants, second only to UNVT for each.

•  Benzo(a)pyrene was infrequently detected at CHSC. This pollutant was detected
   twice in the first quarter, was not detected at all during the second quarter, was
   detected once in the third quarter, and was detected three times in the fourth quarter
   of 2011.

•  Although hexavalent chromium was detected in all four quarters of 2011, it was
   detected in fewer than half of the samples collected (29 out of 60). The measurements
   ranged from 0.0023 ng/m3 to 0.0244 ng/m3, with two-thirds of the concentrations
   measured during the second (10) and third (10) quarters of 2011.

•  Naphthalene was detected in every sample collected at CHSC. The first quarter
   average concentration is roughly twice the other quarterly averages and has a
   relatively large confidence interval associated with it. The maximum naphthalene
   concentration was measured on March 10, 2011 (122 ng/m3) and is more than twice
   the next highest measurement (49.3 ng/m3), measured on the following sample day.
   The concentrations measured at CHSC ranged from 4.54 ng/m3 to 122 ng/m3,  with a
   median concentration of 11.75 ng/m3.
                                  21-14

-------
21.4.2 Concentration Comparison
       In order to better illustrate how a site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for the selected NATTS MQO Core
Analytes listed in Section 3.5.3, where applicable. Thus, box plots for benzo(a)pyrene,
hexavalent chromium, and naphthalene were created for CHSC. Figures 21-6 through 21-8
overlay the site's minimum, annual average, and maximum concentrations onto the program-
level minimum, first quartile, median, average, third quartile, and maximum concentrations, as
described in Section 3.5.3.
       Figure 21-6. Program vs. Site-Specific Average Benzo(a)pyrene Concentration
 Cr-SC
              0.25
                         0.5
                                    0.75         1         1.25
                                          Concentration (ng/m3J
                                                                    1.5
                                                                               1.75
                Program:   IstQuartile     2ndQuartile     3rdQuartile     4thQuartile     Average
                Site:     Site Average     Site Minimum/Maximum

   Figure 21-7. Program vs. Site-Specific Average Hexavalent Chromium Concentration
 CHSC
                  0.05
                                3.1
                                              5.15
                                          Concentration (ng/mi)
                                                                           DL25
                Program:   IstQuartile     2ndQuartile     SrdQuartile     4thQuartile     Average
                Site:     Site Average     Site Minimum/Maximum

                                            21-15

-------
       Figure 21-8. Program vs. Site-Specific Average Naphthalene Concentration
CHSC
                                                         Program Max Concentration =779 ng/m3
                   100
                           150
                                   200      250      300
                                       Concentration (ng/m3)
                                                            350
                                                                            453
                                                                                    555
              Program:  IstQuartile     2ndQuartile     3rd Quartile     4thQuartile    Average
              Site:     Site Average    Site Minimum/Maximum

      Observations from Figures 21-6 through 21-8 include the following:

         •   Figure 21-6 is the box plot for benzo(a)pyrene. Note that the first quartile for this
             pollutant is zero and is not visible on this box plot. This box plot shows that the
             annual average concentration for CHSC is less than both the program-level
             average and median concentrations. Figure 21-6 also shows that the maximum
             concentration measured at CHSC is considerably less than the maximum
             concentration measured across the program. Several non-detects of
             benzo(a)pyrene were measured at CHSC.

         •   Figure 21-7 is the box plot for hexavalent chromium and shows that the annual
             average concentration of hexavalent chromium for CHSC is less than the
             program-level first quartile (25th percentile). Further, the maximum concentration
             measured at CHSC is just greater than the program-level average concentration.
             More than half of the measurements of hexavalent chromium for CHSC were
             non-detects.

         •   Figure 21-8 is the box plot for naphthalene. Note that the program-level
             maximum concentration (779 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
             500 ng/m3. Figure 21-8 shows that the annual naphthalene average for CHSC is
             less than the program-level first quartile. The maximum naphthalene
             concentration measured at CHSC is considerably less than the program-level
             maximum concentration. There were no non-detects of naphthalene measured at
             CHSC or across the program.
                                         21-16

-------
21.4.3  Concentration Trends
       A site-specific trends evaluation was completed for sites that have sampled one or more
of the selected NATTS MQO Core Analytes for 5 consecutive years or longer, as described in
Section 3.5.4. CHSC has sampled hexavalent chromium under the NMP since 2005. Thus,
Figure 21-9 presents the annual statistical metrics for hexavalent chromium for CHSC. The
statistical metrics presented for assessing trends include the substitution of zeros for non-detects.
PAHs are excluded from this analysis because sampling for PAHs did not begin until 2008.
    Figure 21-9. Annual Statistical Metrics for Hexavalent Chromium Concentrations
                                 Measured at CHSC


(ration (ng/m3)
s c
Average Concen











••

T
1 ^
I* i
*

*• 	 - r

••*•-. i
	 	 * 	 •*
	 ^ 	 * 	
2005 2006 2007 200S 2009 2010 2011
Year
# 5th Pe re entile — Minimum — Median — Maximum # 95th Percentile ..^.. Average
       Observations from Figure 21-9 for hexavalent chromium measurements at CHSC include
the following:
       •   Sampling for hexavalent chromium at CHSC began in January 2005.
       •   The maximum concentration of hexavalent chromium was measured on
          March 23, 2005 (0.147 ng/m3). The maximum concentration of hexavalent chromium
          measured in subsequent time periods was considerably lower (by at least half). The
          eight highest concentrations of hexavalent chromium were all measured in 2005 and
          2006.
                                        21-17

-------
       •  The minimum, 5th percentile, and median concentrations are zero for each year
          shown, indicating that at least 50 of the measurements collected at CHSC were non-
          detects. The percentage of non-detects has varied from 52 percent (2011) to
          91 percent (2009).
       •  The maximum concentration, 95th percentile, and average concentration of hexavalent
          chromium decreased significantly through 2007. Little change is shown in the
          average concentration between 2007 and 2009. The range of concentrations measured
          nearly tripled from 2009 to 2010; thus, the average concentration exhibits an increase
          from 2009 to 2010, with little change in the average concentration from 2010 to 2011.

21.5   Additional Risk-Based Screening Evaluations
       The following risk-based screening evaluations were conducted to characterize risk at the
CHSC monitoring site. Refer to Sections 3.3 and 3.5.5 for definitions and explanations regarding
the various toxicity factors, time frames, and calculations associated with these risk-based
screenings.

21.5.1 Risk-Based Screening Assessment Using MRLs
       A risk-based screening was conducted by comparing the concentration data from the
South Carolina monitoring site to the ATSDR MRLs, where available. As described in
Section 3.3, MRLs are noncancer health risk benchmarks and are defined for three exposure
periods: acute (exposures of 1 to 14 days); intermediate (exposures of 15 to 364 days); and
chronic (exposures of 1 year or greater). The preprocessed daily measurements of the pollutants
of interest were compared to the acute MRLs; the quarterly averages were compared to the
intermediate MRLs; and the annual averages were compared to the chronic MRLs.

       As discussed in Section 4.2.2, none of the measured detections  or time-period average
concentrations for any of the monitoring sites  were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2011.

 21.5.2 Cancer Risk and Noncancer Hazard Approximations
       For the pollutants of interest for the South Carolina monitoring site and where annual
average concentrations could be calculated, risk was examined by calculating cancer risk and
noncancer hazard approximations.  These approximations can be used as risk estimates for cancer
and noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers may want to shift or

                                         21-18

-------
 confirm their air-monitoring priorities. Refer to Section 3.5.5.2 for an explanation of how cancer
 risk and noncancer hazard approximations are calculated and what limitations are associated with
 them. Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer
 hazard approximations are presented in Table 21-6, where applicable. Cancer risk
 approximations are presented as probabilities while the noncancer hazard approximations are
 ratios and thus, unitless values.

          Table 21-6. Risk  Approximations for the South Carolina Monitoring Site
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Chesterfield, South Carolina - CHSC
Benzo(a)pyrene
Hexavalent Chromium
Naphthalene
0.00176
0.012
0.000034

0.0001
0.003
7/60
29/60
60/60
0.01
ฑ0.02
0.01
ฑ<0.01
16.42
ฑ4.13
0.03
0.08
0.56

<0.01
0.01
— = a Cancer URE or Noncancer RfC is not available
        Observations for CHSC from Table 21-6 include the following:
        •  The cancer risk approximations for the pollutants of interest for CHSC are all less
           than 1 in-a-million, with the highest cancer risk approximation calculated for
           naphthalene (0.56 in-a-million).
        •  The noncancer hazard approximations for the pollutants of interest are very low (0.01
           or less), indicating that no adverse health effects are expected from these individual
           pollutants. Because benzo(a)pyrene has no RfC, a noncancer hazard approximation
           could not be calculated.

 21.5.3  Risk-Based Emissions Assessment
        In addition to the risk-based screenings discussed above, Tables 21-7 and 21-8 present an
 evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
 Table 21-7 presents the 10 pollutants with the highest emissions from the 2008 NEI, the
 10 pollutants with the highest toxicity-weighted emissions, and the 10 pollutants with the highest
 cancer risk approximations (in-a-million), as calculated from annual averages provided in
 Table 21-6. Table 21-8 presents similar information, but identifies the 10 pollutants with the
 highest noncancer hazard approximations (HQ), also calculated from the annual averages
 provided in Table 21-6.
                                           21-19

-------
   Table 21-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for

                                               the South Carolina Monitoring Site
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Chesterfield, South Carolina (Chesterfield County) - CHSC
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Naphthalene
Trichloroethylene
Dichloromethane
POM, Group 2b
Ethylene oxide
28.92
14.89
13.15
7.77
3.17
1.50
0.47
0.40
0.32
0.07
Benzene
Formaldehyde
1,3 -Butadiene
Hexavalent Chromium, PM
Naphthalene
Ethylbenzene
POM, Group 2b
POM, Group 3
POM, Group 5a
Acetaldehyde
2.26E-04
1.71E-04
9.52E-05
7.65E-05
5.11E-05
3.72E-05
2.81E-05
2.20E-05
1.92E-05
1.71E-05
Naphthalene
Hexavalent Chromium
Benzo(a)pyrene
0.56
0.08
0.03

to
o

-------
   Table 21-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
                                          RfCs for the South Carolina Monitoring Site
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Noncancer Hazard
Approximation
Pollutant (HQ)
Chesterfield, South Carolina (Chesterfield County) - CHSC
Toluene
Xylenes
Benzene
Methanol
Hexane
Ethylbenzene
Formaldehyde
Acetaldehyde
Methyl isobutyl ketone
Ethylene glycol
78.42
60.74
28.92
28.81
19.60
14.89
13.15
7.77
6.32
3.86
Acrolein
Cyanide Compounds, gas
1,3 -Butadiene
Formaldehyde
Benzene
Acetaldehyde
Xylenes
Naphthalene
Lead, PM
Manganese, PM
42,691.89
1,852.73
1,586.49
1,341.80
963.86
863.64
607.44
500.57
432.31
287.58
Naphthalene 0.01
Hexavalent Chromium O.01

to
to

-------
       The pollutants listed in Tables 21-7 and 21-8 are limited to those that have cancer and

noncancer toxicity factors, respectively. As a result, although the actual value of the emissions is

the same, the highest emitted pollutants in the cancer table may be different from the noncancer

table. The cancer risk and noncancer hazard approximations based on the site's annual averages

are limited to those pollutants for which the site sampled. As discussed in Section 21.3, CHSC
sampled for PAHs and hexavalent chromium only. In addition, the cancer risk and noncancer

hazard approximations are limited to those pollutants with enough data to meet the criteria for

annual averages to be calculated. A more in-depth discussion of this analysis is provided in

Section 3.5.5.3. Similar to the cancer risk and noncancer hazard approximations, this analysis

may help policy-makers prioritize their air monitoring activities.


       Observations from Table 21-7 include the following:

       •   Benzene, ethylbenzene, and formaldehyde are the highest emitted pollutants with
          cancer UREs in Chesterfield County.

       •   Benzene, formaldehyde, and 1,3-butadiene are the pollutants with the highest
          toxicity-weighted emissions (of the pollutants with cancer UREs) for Chesterfield
          County.

       •   Seven of the highest emitted pollutants also have the highest toxicity-weighted
          emissions for Chesterfield County.

       •   Naphthalene appears on all three lists, with the sixth highest emissions, the fifth
          highest toxicity-weighted emissions, and highest cancer risk approximation among
          the pollutants of interest for CHSC.

       •   Hexavalent chromium ranks fourth for its toxicity-weighted emissions, but is not
          among the highest emitted pollutants.

       •   Several POM Groups appear among the pollutants with the highest emissions and
          toxicity-weighted emissions. POM, Group 2b includes several PAHs sampled for at
          CHSC including acenaphthylene, fluoranthene, and perylene. POM, Group 5a
          includes benzo(a)pyrene, which is one of CHSC's pollutants of interest. POM,
          Group 5a ranks ninth for toxicity weighted emissions but is not among the highest
          emitted. POM, Group 3 does not include any pollutants sampled for at CHSC.


       Observations from Table 21-8 include the following:

       •   Toluene, xylenes, and benzene are the highest emitted  pollutants with noncancer
          RfCs in Chesterfield County.
                                         21-22

-------
       •  The pollutants with the highest toxi city-weighted emissions (of the pollutants with
          noncancer RfCs) are acrolein, cyanide compounds (gaseous), and 1,3-butadiene.

       •  Four of the highest emitted pollutants in Chesterfield County also have the highest
          toxicity-weighted emissions.

       •  Naphthalene does not appear among the highest emitted pollutants with a noncancer
          toxicity factor, but ranks eighth among the pollutants with the highest toxicity-
          weighted emissions. Hexavalent chromium does not appear on either emissions-based
          list. These are the only two pollutants of interest with noncancer RfCs for CHSC.


21.6   Summary of the 2011 Monitoring Data for CHSC

       Results from several of the data treatments described in this section include the
following:

       ปซป  Naphthalene was the only pollutant to fail screens for CHSC. This site has the third
          lowest number of failed screens (3) among allNMP sites.

       *ป*  Of the site-specific pollutants of the interest, naphthalene had the highest annual
          average concentration for CHSC; however, it was second lowest compared to other
          NMP sites sampling naphthalene.

       ปซป  Concentrations ofhexavalent chromium increased from 2009 to 2010 and then held
          steady for 2011.
                                         21-23

-------
22.0   Sites in South Dakota
       This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the UATMP sites in South Dakota, and integrates these
concentrations with emissions, meteorological, and risk information. Data generated by sources
other than ERG are not included in the data analyses contained in this report. Readers are
encouraged to refer to Sections 1 through 4 and the glossary (Appendix P) for detailed
discussions and definitions regarding the various data analyses presented below.

22.1   Site Characterization
       This section characterizes the monitoring sites by providing geographical and physical
information about the location of the sites and the surrounding areas.  This information is
provided to give the reader insight regarding factors that may influence the air quality near the
sites and assist in the interpretation of the ambient monitoring measurements.

       There are two South Dakota monitoring sites. One monitoring site is located in Sioux
Falls, South Dakota (SSSD) while the other is located in Union County (UCSD). Figures 22-1
and 22-3 are composite satellite images retrieved from ArcGIS Explorer showing the monitoring
sites in their urban and rural locations. Figures 22-2 and 22-4 identify nearby point source
emissions locations by source category, as  reported in the 2008 NEI for point sources. Note that
only sources within 10 miles of the sites are included in the facility counts provided in
Figures 22-2 and 22-4. A 10-mile boundary was chosen to give the reader an indication of which
emissions sources and emissions source categories could potentially have a direct effect on the
air quality at the monitoring sites. Further,  this boundary provides both the proximity of
emissions sources to the monitoring  sites as well as the quantity of such sources within a given
distance of the sites.  Sources outside each 10-mile radius are still visible on the maps, but have
been grayed out in order to show emissions sources just outside the boundaries. Table 22-1
provides supplemental geographical  information such as land use, location setting, and locational
coordinates.
                                          22-1

-------
                                  Figure 22-1. Sioux Falls, South Dakota (SSSD) Monitoring Site
to
to

-------
Figure 22-2. NEI Point Sources Located Within 10 Miles of SSSD
       Legend

        &b  SSSD UATMP site
                                   5 45'CTVV       96'40'0-W       96'35'0"W       96'30'CTW
                                     Note: Due to facility density and collocation, the total facilities
                                     displayed may not represent all facilities within the area of interest.
10 mile radius
                   County boundary
                Source Category Group (No. of Facilities)
                  -f1   Aircraft Operations (7)
                  •*ป•   Transportation Equipment (1)
                                    22-3

-------
                                Figure 22-3. Union County, South Dakota (UCSD) Monitoring Site
to
to

-------
Figure 22-4. NEI Point Sources Located Within 10 Miles of UCSD
       Legend
                                   96-J5X1-W       96'4WW       96'35'0-W
                                    Note: Due to facility density and collocation the total facilities
                                    displayed may nol represent ali facilities within the area of interest.
             UCSD UATM P site
                                    10 mile radius
                                                       County boundary
                  Source Category Group (No. of Facilities)
                    -f1   Aircraft Operations (1)
                                    22-5

-------
                                Table 22-1. Geographical Information for the South Dakota Monitoring Sites

Site
Code
SSSD
UCSD

AQS Code
46-099-0008
46-127-0001

Location
Sioux Falls
Not in a
City

County
Minnehaha
Union
Micro- or
Metropolitan
Statistical
Area
Sioux Falls, SD
MSA
Sioux City, IA-
NE-SD MSA

Latitude and
Longitude
43.54792,
-96.700769
42.751518,
-96.707208

Land Use
Commercial
Agricultural

Location
Setting
Urban/City
Center
Rural

Additional Ambient Monitoring Information1
SO2, NO, NO2, NOX, O3, Meteorological
parameters, PM10, PM2 5, and PM25 Speciation.
CO, SO2, NO, NO2, NOX, Meteorological
parameters, PM10, and PM25.
    1 Data for additional pollutants are reported to AQS for SSSD and UCSD (EPA, 2012c); however, these data are not generated by ERG and are therefore not included

    in this report.
to
to

-------
       SSSD is located on the east side of Sioux Falls, in eastern South Dakota. The monitoring
site is located at the South Dakota School for the Deaf. The surrounding area is mixed usage,
with both commercial and residential areas surrounding the site. SSSD is less than 1/2 mile from
the intersection of Highway 42 (East 10th Street) and 1-229, as shown in Figure 22-1. As
Figure 22-2 shows, few emissions sources are located within 10 miles of SSSD. There are only
two source categories shown in Figure 22-2, the aircraft operations category, which includes
airports as well as small runways, heliports, or landing pads, and the transportation equipment
category. The emissions source closest to SSSD is a hospital heliport.

       UCSD is located in Union County, the southeastern-most county of the state, where the
South Dakota state border follows the Missouri River and comes to a point near Sioux City, Iowa
at the Nebraska and Iowa borders. The UCSD monitoring site is located in a rural and
agricultural area in the town of Brule, north of Elk Point and west of Vermillion. As shown in
Figure 22-3, the monitoring site is located on a residential property surrounded by agricultural
fields. Interstate-29 runs northwest-southeast through the center of Union County and lies less
than 1.5 miles west of UCSD. Figure 22-4 shows that there is a single point source located
within 10 miles of the site. However, UCSD is south of a proposed power plant and oil refinery.
The purpose of monitoring at UCSD is to collect air quality data before, during, and after the
construction of the proposed power plant and oil refinery (SD DENR, 2011).

       Table 22-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the South Dakota monitoring sites. Table 22-2 includes county-level
population and vehicle registration information. Table 22-2 also includes a county-level vehicle
registration-to-population ratio, which was calculated to represent the number of vehicles per
person within each monitoring site's residing county. In addition, the population within 10 miles
of each site is presented, based on postal code population data estimates. An estimate of 10-mile
vehicle ownership was then determined by applying the county-level vehicle registration-to-
population ratio to the 10-mile population surrounding each monitoring site. Table 22-2 also
contains traffic volume information for each site. Finally, Table 22-2 presents the county-level
daily VMT for Minnehaha and Union Counties.
                                          22-7

-------
    Table 22-2. Population, Motor Vehicle, and Traffic Information for the South Dakota
                                     Monitoring Sites
Site
SSSD
UCSD
Estimated
County
Population1
171,752
14,651
County-level
Vehicle
Registration2
210,914
25,419
Vehicles per
Person
(Registration:
Population)
1.23
1.73
Population
within 10
miles3
186,954
5,577
Estimated
10-mile
Vehicle
Ownership
229,582
9,676
Annual
Average
Daily
Traffic4
18,700
156
County-
level Daily
VMT5
3,751,886
808,049
Bounty-level population estimates reflect 2011 data from the U.S. Census Bureau (Census Bureau, 2012b)
2County-level vehicle registration reflects 2011 data from the South Dakota Department of Revenue (SDDOR, 2012)
310-mile population estimates reflect 2011 data from Xionetic (Xionetic, 2011)
4AADT reflects 2011 data for SSSD and 2007 data for UCSD from the South Dakota DOT (SD DOT, 2007 and
 2011)
5County-level VMT reflects 2011 data from the South Dakota DOT (SD DOT, 2012)

        Observations from Table 22-2 include the following:

        •   Although SSSD's county-level population is an order of magnitude higher than the
           county-level population for UCSD, both county-level populations are in the bottom
           third compared to other counties with NMP sites, with UCSD ranking last. The
           10-mile population for each site is also on the low side compared to other NMP sites,
           particularly for UCSD.

        •   SSSD's county-level vehicle registration is an order of magnitude higher than the
           county-level vehicle registration for UCSD, but both of the county-level vehicle
           registrations are on the low side compared to other counties with NMP sites. Union
           County's vehicle registration is the lowest of all NMP counties, while Minnehaha
           County is in the bottom third. The 10-mile vehicle ownership estimates for SSSD and
           UCSD rank slightly higher among NMP sites than county-level vehicle ownerships.

        •   The vehicle-per-person ratios for UCSD and SSSD are among the highest compared
           to other NMP sites, ranking first and second, respectively. This indicates that
           residents likely own multiple vehicles.

        •   The traffic volume for SSSD is two orders of magnitude higher than the traffic
           volume for UCSD. The traffic near UCSD is the second lowest among all NMP sites,
           behind only BRCO. Traffic data for SSSD are provided for East 10th Avenue between
           South Mable Avenue and South Highland Avenue; traffic data for UCSD are
           provided for 475th Avenue near 317th Street.

        •   The daily VMT for Minnehaha County is more than four times the VMT for Union
           County. The daily VMT for Union County is the lowest among NMP sites (where
           VMT was available). The VMT for Minnehaha County ranks tenth lowest among
           counties with NMP sites (where VMT was available).
                                           22-8

-------
22.2   Meteorological Characterization
       The following sections characterize the meteorological conditions near the monitoring
sites in South Dakota on sample days, as well as over the course of the year.

22.2.1  Climate Summary
       The Sioux Falls area has a continental climate, with cold winters, warm summers, and
often drastic day-to-day variations. Precipitation varies throughout the year, with the spring and
summer seasons receiving more than half of the annual rainfall. On average, a south wind blows
in the summer and fall and a northwest wind blows in the winter and spring. Flooding is often a
concern in the area during springtime when snow begins to melt, although a flood control
system, including levees and a diversion channel, was constructed to reduce the flood threat
within the city limits and to divert water from the Big Sioux River and Skunk Creek around the
city (Bair, 1992).

       Sioux City is located just north of the Missouri River where the Iowa border meets the
Nebraska and South Dakota borders. The climate near Sioux City is generally continental in
nature, with warm summers and cold, relatively dry winters. Precipitation is concentrated in the
spring and summer months. Wind direction varies with season, with southeasterly to southerly
winds in spring and summer, and northwesterly winds in fall and winter (Bair, 1992).

22.2.2  Meteorological Conditions in 2011
       Hourly meteorological data from the NWS weather stations nearest these sites were
retrieved for 2011 (NCDC, 2011). The closest weather stations are located at Joe Foss Field
Airport (near SSSD) and Sioux Gateway Airport (near UCSD), WBAN 14944 and 14943,
respectively. Additional information about these weather stations, such as the distance between
the sites and the weather stations, is provided in Table 22-3. These data were used to determine
how meteorological conditions on sample days vary from conditions experienced throughout the
year.
                                          22-9

-------
                          Table 22-3. Average Meteorological Conditions near the South Dakota Monitoring Sites
Closest NWS
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(ฐF)
Average
Temperature
(ฐF)
Average
Dew Point
Temperature
(ฐF)
Average
Wet Bulb
Temperature
(ฐF)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Sioux Falls, South Dakota - SSSD
Joe Foss Field
Airport
14944
(43.58, -96.75)
3.20
miles
309ฐ
(NW)
Sample
Day
2011
56.3
ฑ5.9
55.9
+ 2.4
46.8
ฑ5.7
46.4
+ 2.3
37.0
ฑ5.5
36.1
+ 2.2
42.1
ฑ5.3
41.5
+ 2.1
71.8
ฑ3.0
70.4
+ 1.2
1014.3
ฑ2.1
1015.3
+ 0.8
8.6
ฑ1.0
8.2
+ 0.4
Union County, South Dakota - UCSD
Sioux
Gateway/Col.
Bud Day Field
Airport
14943
(42.39, -96.38)
29.45
miles
148ฐ
(SSE)
Sample
Day
2011
60.1
ฑ6.0
59.5
+ 2.4
49.0
ฑ5.7
48.7
+ 2.3
38.8
ฑ5.4
37.9
+ 2.2
43.9
ฑ5.2
43.4
+ 2.1
71.1
ฑ3.1
69.5
+ 1.2
1014.9
ฑ2.3
1015.4
+ 0.8
8.6
ฑ0.9
8.3
+ 0.4
to
to
Sample day averages are highlighted to help differentiate the sample day averages from the full-year averages.

-------
       Table 22-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for the entire year. Also included in Table 22-3 is the 95 percent
confidence interval for each parameter. As shown in Table 22-3, average meteorological
conditions on sample days were representative of average weather conditions throughout the year
for both sites.

22.2.3 Back Trajectory Analysis
       Figure 22-5 is the composite back trajectory map for days on which samples were
collected at the SSSD monitoring site in 2011. Included in Figure 22-5 are four back trajectories
per sample day. Figure 22-6 is the corresponding cluster analysis. Similarly, Figure 22-7 is the
composite back trajectory map for days on which samples were collected at UCSD and
Figure 22-8 is the corresponding cluster analysis. An in-depth description of these maps and how
they were generated is presented in  Section 3.5.2.1. For the composite maps, each line represents
the 24-hour trajectory along which a parcel of air traveled toward the monitoring site on a given
sample day and time, based on an initial height of 50 meters AGL. For the cluster analyses,  each
line corresponds to a trajectory representative of a given cluster of back trajectories. Each
concentric circle around the sites in Figures 22-5 through 22-8 represents 100 miles.

       Observations from Figures 22-5 and 22-6 for SSSD include the following:
       •  Back trajectories originated from a variety of directions at the SSSD site, but rarely
          from  due north.
       •  The 24-hour air shed domain for SSSD  is among the larger air sheds compared to the
          other NMP monitoring sites. The farthest away a back trajectory originated was
          southwest Montana, or greater than 700 miles away, although the average back
          trajectory length was nearly 282 miles and 88 percent of back trajectories originated
          within 450 miles of the site.
       •  The cluster analysis shows that back trajectories originating from the southwest to
          northwest of SSSD account for one-quarter of the back trajectories. Another one-
          quarter of back trajectories originated to the northwest to north of SSSD, primarily
          over North Dakota.  The  shorter cluster (34 percent) originating to the southeast of
           SSSD represents shorter trajectories (< 300 miles) originating from a variety of
          directions, although primarily over the state of Iowa. Eight percent of trajectories
          originated from the  south of SSSD over Kansas and Missouri and another eight
          percent originated to the northeast of SSSD.
                                         22-11

-------
Figure 22-5. 2011 Composite Back Trajectory Map for SSSD
    Figure 22-6. Back Trajectory Cluster Map for SSSD
                         22-12

-------
Figure 22-7. 2011 Composite Back Trajectory Map for UCSD
    Figure 22-8. Back Trajectory Cluster Map for UCSD
                         22-13

-------
       Observations from Figures 22-7 and 22-8 for UCSD include the following:

       •  Back trajectories originated from a variety of directions at the UCSD monitoring site,
          with the longest back trajectories originating to the northwest of UCSD.

       •  The 24-hour air shed domain for UCSD was larger in size compared to SSSD and
          most other NMP sites. Two back trajectories originated farther than 900 miles away
          (one over northwest Montana and one over northeast Oregon. The average back
          trajectory length was greater than 300 miles; this is the only NMP site  for which this
          is true. However, 85 percent of the trajectories originated within 450 miles of the site.

       •  The cluster analysis for UCSD shows that nearly 50 percent of back trajectories
          originated to the west to northwest to north of the site, although these back
          trajectories are represented by three cluster trajectories. One cluster trajectory
          (12 percent) includes back trajectories originating to the west of UCSD over central
          and western North and South Dakota and of relatively short length (<350 miles).
          Another cluster trajectory (12 percent) includes back trajectories originating from the
          northwest, primarily over Montana, and are relatively long (> 400 miles) in length.
          The third cluster trajectory (25 percent) includes mostly shorter trajectories
          originating to the northwest to north of UCSD and over North Dakota  and northern
          South Dakota. Roughly one-third of back trajectories (34 percent) originated to the
          east and southeast over Iowa and Missouri. These were generally less than 300 miles
          in length. Another 13 percent of back trajectories originated to the south of UCSD
          and finally, three percent originated northeastward over Minnesota, Lake Superior,
          and the Upper Peninsula of Michigan.


22.2.4 Wind Rose Comparison

       Hourly surface wind data from the NWS weather stations at Joe Foss Field (for SSSD)

and Sioux Gateway (for UCSD) Airports were uploaded into a wind rose software program to

produce customized wind roses, as described in Section 3.5.2.2. A wind rose shows the

frequency of wind directions using "petals" positioned  around a 16-point compass, and uses

different colors to represent wind speeds.
       Figure 22-9 presents a map showing the distance between the NWS station and SSSD,

which may be useful for identifying topographical influences that may affect the meteorological

patterns experienced at this location. Figure 22-9 also presents three different wind roses for the

SSSD monitoring site. First, a historical wind rose representing 2001 to 2010 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 2011 is presented.

Next, a wind rose representing wind data for days on which samples were collected in 2011 is

presented. These can be used to identify the predominant wind speed and direction for 2011 and

                                         22-14

-------
determine if wind observations on sample days were representative of conditions experienced

over the entire year and historically. Figure 22-10 presents the distance map and three wind roses
forUCSD.


       Observations from Figure 22-9 for SSSD include the following:

       •  The Joe Foss Field Airport weather station is located approximately 3.2 miles
          northwest of SSSD.

       •  The historical wind rose shows that winds from a variety of directions were observed
          near SSSD, although winds from the south were observed the most (13 percent), and
          southwesterly and west-southwesterly winds observed the least (less than 3 percent).
          Calm winds were observed for approximately 11 percent of the observations. The
          strongest winds tend to be from the south or the northwest quadrant.

       •  The 2011 wind patterns are very similar to the historical wind patterns, indicating that
          wind conditions in 2011 near SSSD are similar to historical wind conditions.

       •  The sample day wind rose also resembles the historical and full-year wind roses, but
          does exhibit some differences. The sample day wind rose has a higher percentage of
          winds from the southeast quadrant and fewer from the northwest quadrant. In
          addition, winds greater than 22 knots were observed with easterly winds as well as
          winds from the southeast to south, and northwest to north-northwest.


       Observations from Figure 22-10 for UCSD include the following:

       •  The Sioux Gateway Airport weather station is located  approximately 29 miles
          south-southeast of UCSD, across the state border in Sioux City, Iowa. The weather
          station is located less than a mile from the Missouri River.

       •  The historical wind rose shows that winds from the southeast and northwest quadrants
          were observed the most near UCSD. Calm winds were observed for less than eight
          percent of the observations. The strongest winds tend to be from the south or the
          northwest quadrant.

       •  The 2011 wind patterns are similar to the historical wind patterns, although the calm
          rate is slightly higher for 2011 (approximately 10 percent).

       •  The sample day wind patterns resemble the historical and full-year wind patterns, but
          have a higher percentage of east-southeasterly and southeasterly wind observations.
                                         22-15

-------
 Figure 22-9. Wind Roses for the Joe Foss Field Airport Weather Station near SSSD
Distance between SSSD and NWS Station
2001-2010 Historical Wind Rose
                      <-

                        ' ..... ili< i/fe  I
                       -
         2011 Wind Rose
    Sample Day Wind Rose
                                    22-16

-------
Figure 22-10. Wind Roses for the Sioux Gateway Airport Weather Station near UCSD
Distance between UCSD and NWS Station
2001-2010 Historical Wind Rose
         2011 Wind Rose
    Sample Day Wind Rose
                                    22-17

-------
22.3   Pollutants of Interest
       Site-specific "pollutants of interest" were determined for the South Dakota monitoring
sites in order to allow 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. In addition, if any of the NATTS MQO Core Analytes measured by each monitoring site
did not meet the pollutant of interest criteria based on the preliminary risk-based screening, that
pollutant was added to the list of site-specific pollutants of interest. A more in-depth description
of the risk-based screening process is presented in Section 3.2.

       Table 22-4 presents the results of the preliminary risk-based screening process for the
South Dakota monitoring sites. The pollutants that failed at least one screen and contributed to
95 percent of the total failed screens for each monitoring site are shaded. NATTS MQO Core
Analytes are bolded. Thus, pollutants of interest are shaded and/or bolded. SSSD and UCSD
sampled for VOCs, SNMOCs, and carbonyl compounds.
     Table 22-4. Risk-Based Screening Results for the South Dakota Monitoring Sites
Pollutant
Screening
Value
(Ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Sioux Falls, South Dakota - SSSD
Benzene
Acetaldehyde
Formaldehyde
Carbon Tetrachloride
1,3-Butadiene
1 ,2-Dichloroethane
Acrylonitrile
Ethylbenzene
Hexachloro- 1 ,3 -butadiene
1 , 1 ,2,2-Tetrachloroethane
Chloromethylbenzene
ฃ>-Dichlorobenzene
Dichloromethane
Propionaldehyde
0.13
0.45
0.077
0.17
0.03
0.038
0.015
0.4
0.045
0.017
0.02
0.091
7.7
0.8
Total
60
59
59
55
41
17
13
6
3
2
1
1
1
1
319
60
59
59
59
44
17
13
60
3
2
1
19
43
59
498
100.00
100.00
100.00
93.22
93.18
100.00
100.00
10.00
100.00
100.00
100.00
5.26
2.33
1.69
64.06
18.81
18.50
18.50
17.24
12.85
5.33
4.08
1.88
0.94
0.63
0.31
0.31
0.31
0.31
18.81
37.30
55.80
73.04
85.89
91.22
95.30
97.18
98.12
98.75
99.06
99.37
99.69
100.00

                                          22-18

-------
Table 22-4. Risk-Based Screening Results for the South Dakota Monitoring Sites
                                 (Continued)
Pollutant
Screening
Value
(Ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Union County, South Dakota - UCSD
Formaldehyde
Acetaldehyde
Benzene
Carbon Tetrachloride
Acrylonitrile
1 ,2-Dichloroethane
Hexachloro- 1 ,3 -butadiene
1,3-Butadiene
Ethylbenzene
1 ,2-Dibromoethane
1 , 1 ,2,2-Tetrachloroethane
Trichloroethylene
ฃ>-Dichlorobenzene
0.077
0.45
0.13
0.17
0.015
0.038
0.045
0.03
0.4
0.0017
0.017
0.2
0.091
Total
61
60
56
54
32
18
5
4
3
2
2
2
1
300
61
61
56
56
32
18
5
5
56
2
2
4
11
369
100.00
98.36
100.00
96.43
100.00
100.00
100.00
80.00
5.36
100.00
100.00
50.00
9.09
81.30
20.33
20.00
18.67
18.00
10.67
6.00
1.67
1.33
1.00
0.67
0.67
0.67
0.33
20.33
40.33
59.00
77.00
87.67
93.67
95.33
96.67
97.67
98.33
99.00
99.67
100.00

 Observations from Table 22-4 include the following:

 •   Fourteen pollutants failed at least one screen for SSSD; of these, five are NATTS
     MQO Core Analytes. Thirteen pollutants failed screens for UCSD, of which six are
     also NATTS MQO Core Analytes. Of the pollutants failing screens, the sites share
     11 pollutants in common in Table 22-4.

 •   For SSSD, seven pollutants (of which five are NATTS MQO Core Analytes) were
     identified as pollutants of interest by the risk-based screening process. Chloroform,
     tetrachloroethylene, trichloroethylene, and vinyl chloride were added to SSSD's
     pollutants of interest because they are NATTS MQO Core Analytes, even though
     they did not fail any screens. These pollutants are not shown in Table 22-4 but are
     shown in subsequent tables in the sections that follow.

 •   For UCSD, seven pollutants (of which four are NATTS MQO Core Analytes) were
     identified as pollutants of interest by the risk-based screening process.
     Trichloroethylene and 1,3-butadiene were added to the pollutants of interest for
     UCSD because they are NATTS MQO Core Analytes, even though they did not
     contribute to 95 percent of the total failed screens. Chloroform, tetrachloroethylene,
     and vinyl chloride were added to UCSD's pollutants  of interest because they are
     NATTS MQO Core Analytes, even though they did not fail any screens. These three
     pollutants are not shown in Table 22-4 but are shown in subsequent tables in the
     sections that follow.
                                    22-19

-------
       •  Formaldehyde and benzene were detected in every valid sample collected at UCSD
          and SSSD and failed 100 percent of screens. Other pollutants, such as acrylonitrile,
          1,2-dichloroethane, and hexachloro-1,3-butadiene also failed 100 percent of screens
          for each site but were detected less frequently.
       •  Recall from Section 3.2 that if a pollutant was measured by both the TO-15 and
          SNMOC methods at the same site, the TO-15 results were used for the risk-based
          screening process. As the South Dakota sites sampled both VOCs (TO-15) and
          SNMOCs, the TO-15 results were used for the 12 pollutants these methods have in
          common.

22.4   Concentrations
       This section presents various concentration averages used to characterize pollution levels
at the South Dakota monitoring sites. Where applicable, the following calculations and data
analyses were performed: Time period-based concentration averages (quarterly and annual) are
provided for the pollutants of interest for the South Dakota  monitoring sites, where the data meet
the applicable criteria. Concentration averages for select pollutants are also presented graphically
for the sites  to illustrate how the sites' concentrations compare to the program-level averages, as
presented in Section 4.1. In addition, concentration averages for select pollutants are presented
from previous years of sampling in order to characterize concentration trends at the sites.
Additional site-specific statistical summaries for SSSD and UCSD are provided in Appendices J
through L.

22.4.1 2011 Concentration Averages
       Quarterly and annual concentration averages were calculated for the pollutants of interest
for each South Dakota site, as described in Section 3.1. The quarterly average of a particular
pollutant is simply the average concentration of the preprocessed daily measurements over a
given calendar quarter. Quarterly average concentrations include the substitution of zeros for all
non-detects. A site must have a minimum of 75 percent valid samples of the total number of
samples possible within a given quarter for a quarterly average to be calculated. An annual
average includes all measured detections and substituted zeros for non-detects for the entire year
of sampling. Annual averages were calculated for pollutants where three valid quarterly averages
could be calculated and where method completeness was greater than or equal to 85 percent, as
presented in Section 2.4. Quarterly and annual average concentrations for the South Dakota
monitoring sites are presented in Table 22-5,  where applicable. Note that if a pollutant was not
                                          22-20

-------
detected in a given calendar quarter, the quarterly average simply reflects "0" because only zeros

substituted for non-detects were factored into the quarterly average concentration.
          Table 22-5. Quarterly and Annual Average Concentrations of the Pollutants of
                     Interest for the South Dakota Monitoring Sites
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Hg/m3)
2nd
Quarter
Average
(Hg/m3)
3rd
Quarter
Average
(jig/m3)
4th
Quarter
Average
(jig/m3)
Annual
Average
(jig/m3)
Sioux Falls, South Dakota - SSSD
Acetaldehyde
Acrylonitrile
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
1 ,2-Dichloroethane
Formaldehyde
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
59/59
13/60
60/60
44/60
59/60
28/60
17/60
59/59
46/60
3/60
1/60
1.90
ฑ0.68
0.02
ฑ0.05
0.69
ฑ0.10
0.03
ฑ0.01
0.32
ฑ0.11
0.02
ฑ0.02
0.01
ฑ0.02
0.98
ฑ0.30
0.08
ฑ0.03
<0.01
ฑ0.01
0.01
ฑ0.01
1.31
ฑ0.23
0.06
ฑ0.04
0.60
ฑ0.12
0.03
ฑ0.02
0.65
ฑ0.05
0.04
ฑ0.03
0.02
ฑ0.02
2.53
ฑ0.71
0.09
ฑ0.03
0
0
1.56
ฑ0.36
0.05
ฑ0.04
0.56
ฑ0.09
0.04
ฑ0.02
0.64
ฑ0.04
0.04
ฑ0.03
0.01
ฑ0.01
2.50
ฑ0.38
0.06
ฑ0.03
0
0
1.51
ฑ0.23
0
0.67
ฑ0.13
0.07
ฑ0.02
0.57
ฑ0.07
0.07
ฑ0.02
0.06
ฑ0.02
2.05
ฑ0.45
0.09
ฑ0.03
O.01
ฑ0.01
0
1.57
ฑ0.21
0.03
ฑ0.02
0.63
ฑ0.05
0.04
ฑ0.01
0.54
ฑ0.05
0.04
ฑ0.01
0.02
ฑ0.01
2.01
ฑ0.28
0.08
ฑ0.01
O.01
ฑO.01
0.01
ฑ0.01
Union County, South Dakota - UCSD
Acetaldehyde
Acrylonitrile
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
1 ,2-Dichloroethane
Formaldehyde
Hexachloro- 1 ,3 -butadiene
61/61
32/56
56/56
5/56
56/56
14/56
18/56
61/61
5/56
1.26
ฑ0.24
0.25
ฑ0.03
0.50
ฑ0.06
0.01
ฑ0.01
0.47
ฑ0.08
0.03
ฑ0.02
0.01
ฑ0.02
0.97
ฑ0.19
0
1.30
ฑ0.25
0.34
ฑ0.03
0.39
ฑ0.06
0.01
ฑ0.01
0.59
ฑ0.09
0.01
ฑ0.01
0.02
ฑ0.02
1.66
ฑ0.42
0.02
ฑ0.03
2.71
ฑ0.87
NA
NA
NA
NA
NA
NA
1.72
ฑ0.34
NA
3.85
ฑ0.74
0
0.39
ฑ0.06
0.01
ฑ0.01
0.63
ฑ0.05
0.04
ฑ0.02
0.06
ฑ0.02
1.60
ฑ0.57
0.02
ฑ0.03
2.33
ฑ0.40
0.17
ฑ0.04
0.40
ฑ0.03
0.01
ฑ0.01
0.58
ฑ0.04
0.02
ฑ0.01
0.03
ฑ0.01
1.50
ฑ0.21
0.01
ฑ0.01
                                         22-21

-------
   Table 22-5. Quarterly and Annual Average Concentrations of the Pollutants of
         Interest for the South Dakota Monitoring Sites (Continued)



Pollutant

Tetrachloroethylene

Trichloroethylene

Vinyl Chloride
#of
Measured
Detections
vs. # of
Samples

23/56

4/56

3/56

1st
Quarter
Average
(Ug/m3)
0.05
ฑ0.02
0.06
ฑ0.08
0.01
ฑ0.01

2nd
Quarter
Average
(Ug/m3)
0.03
ฑ0.02
0.01
ฑ0.02
0.01
ฑ0.01

3rd
Quarter
Average
(Ug/m3)

NA

NA

NA

4th
Quarter
Average
(Ug/m3)
0.03
ฑ0.02

0

0


Annual
Average
(Ug/m3)
0.03
ฑ0.01
0.02
ฑ0.02
0.01
ฑ0.01
Observations for SSSD from Table 22-5 include the following:

•  The pollutants with the highest annual average concentrations by mass are
   formaldehyde (2.01 ฑ 0.28 |ig/m3) and acetaldehyde (1.57 ฑ 0.21 |ig/m3). These are
   the only two pollutants of interest with an annual average greater than 1.0 |ig/m3.

•  The first quarter acetaldehyde average is higher than the other quarterly averages and
   has a relatively large confidence interval associated with it. The maximum
   acetaldehyde concentration was measured on March 10, 2011 (5.83 |ig/m3). This
   concentration is one of only two acetaldehyde concentrations greater than 3 |ig/m3
   measured at SSSD (the second was measured in September). Of the 10 acetaldehyde
   concentrations greater than or equal to 2 |ig/m3 measured at SSSD, six were measured
   during the first quarter of 2011.

•  Several of the VOC pollutants of interest have large confidence intervals relative to
   their quarterly and annual averages. Most of these pollutants were detected in fewer
   than half of the valid samples collected. For example, the confidence interval for the
   first quarter average of acrylonitrile is greater than the average itself. This quarterly
   average includes one measured detection of acrylonitrile (0.373 |ig/m3, measured on
   March 22, 2011) and 14 non-detects. The other  12  measured detections of
   acrylonitrile were from samples collected across the second and third quarters (there
   were no measured detections of acrylonitrile in the fourth quarter of the year).
   1,2-Dichloroethane is another example. This pollutant was detected 17 times in
   samples collected in 2011. Two were measured during the first quarter, three in the
   second, two in the third, and 10 in the fourth. These quarterly averages are a result of
   many zeros substituted for non-detects combined with relatively low measurements.
   For 1,2-dichloroethane, the concentrations ranged from 0.0446 |ig/m3 to 0.11  |ig/m3.


Observations for UCSD from Table 22-5 include the following:

•  The pollutants with the highest annual average concentrations by mass are
   acetaldehyde (2.33 ฑ 0.40 |ig/m3) and formaldehyde (1.50 ฑ 0.21 |ig/m3). These are
   the only two pollutants of interest with an annual average greater than 1.0 |ig/m3.
                                   22-22

-------
       •  For acetaldehyde, the third and fourth quarter averages are two and three times higher
          than the first and second quarter averages. A review of the data shows that the
          maximum concentration of acetaldehyde was measured on August 31, 2011
          (6.93 |ig/m3). Two additional concentrations greater than 6 |ig/m3 were also measured
          in October and November. All 16 concentrations greater than 3 |ig/m3 were measured
          between August 31 and December 31, 2011.  Conversely, all  eight measurements less
          than 1 |ig/m3 were measured between January and April. Formaldehyde does not
          follow a similar trend.

       •  Third quarter averages for the VOCs could not be calculated because sampler issues
          resulted in several canister samples outside pressure limits.

       •  With the exception of benzene, acrylonitrile, and carbon tetrachloride, many of the
          VOCs were detected in fewer than half of the valid samples collected.

       Tables 4-9 through 4-12 present the sites with the 10 highest annual average

concentrations for each of the program-level pollutants of interest. Observations for SSSD and

UCSD from those tables include the following:

       •  None  of the annual average concentrations of the pollutants of interest for SSSD
          appear in Tables 4-9 through 4-12.

       •  UCSD has the second highest annual average concentration of acrylonitrile among
          NMP  sites sampling VOCs, as shown in Table 4-9. This site has the fourth highest
          number  of failed screens for this pollutant among all sites sampling VOCs.

       •  UCSD also appears in Table 4-9 for hexachloro-l,3-butadiene, vinyl chloride,
          trichloroethylene, and 1,2-dichloroethane. However, these pollutants were detected in
          one-third or less of the valid samples collected.

       •  The annual average concentration of acetaldehyde for UCSD ranks ninth among
          NMP  sites sampling carbonyl compounds. By comparison, the annual average
          formaldehyde concentration for UCSD ranks 23rd compared to other NMP sites.


22.4.2 Concentration Comparison

       In order to better illustrate how a site's annual average concentrations compare to the

program-level averages, a site-specific box plot was created for the selected NATTS MQO Core

Analytes listed in Section 3.5.3, where applicable. Thus, box plots for acetaldehyde, benzene,

1,3-butadiene,  and formaldehyde were created for both SSSD and UCSD. Figures 22-11 through

22-14 overlay the sites' minimum, annual average, and maximum concentrations onto the

program-level minimum, first quartile, median, average, third quartile, and maximum

concentrations, as described in Section 3.5.3.
                                         22-23

-------
      Figure 22-11. Program vs. Site-Specific Average Acetaldehyde Concentrations
UCSD
                                                                10
                                                                            12
                                                                                        14
                                              Concentration (^
                 Program:    IstQuartile     2ndQuartile      SrdQuartile      4thQuartile     Average
                 Site:      Site Average     Site Minimum/Maximum

                                                                                                    16
         Figure 22-12. Program vs. Site-Specific Average Benzene Concentrations
                                                                    Program Max Concentration = 23.8 ug/m3
UCSD

1
••










Program Max Concentration = 23.8 ug/rn3


i
j

                                          4         5
                                               Concentration (|
                                                                                                    10
                 Program:    IstQuartile     2ndQuartile      SrdQuartile      4thQuartile     Average
                 Site:      Site Average     Site Minimum/Maximum

                                                22-24

-------
      Figure 22-13. Program vs. Site-Specific Average 1,3-Butadiene Concentrations
                                                            Program Max Concentration = 9.51 ug/m3
UCSD
                                                            Program Max Concentration = 9.51 ug/m3
                 3.5
                                              15
                                         Concentration IVg,''m3)
                                                                          2.5
               Program:  IstQuartile    2ndQuartile     SrdQuartile     4thQuartile     Average
               Site:
Site Average
                                   Site Minimum/Maximum

     Figure 22-14. Program vs. Site-Specific Average Formaldehyde Concentrations
UCSD
                                10
                                              15
                                         Concentration (
                                                                          25
                                                                                        30
               Program:   IstQuartile     2ndQuartile     SrdQuartile     4thQuartile     Average
               Site:      Site Average    Site Minimum/Maximum

      Observations from Figures 22-11 through 22-14 include the following:

          •   Figure 22-11 shows that the annual average acetaldehyde concentration for UCSD
             is greater than the program-level average while the annual average for SSSD is
             less than the program-level average and median concentrations. The range of
             concentrations measured is larger at UCSD than at SSSD, although the maximum
             concentration measured at both sites is less than the maximum concentration
             measured across the program. There were no non-detects of acetaldehyde
             measured at either site or across the program.

          •   The program-level maximum benzene concentration (23.8 |ig/m3) is not shown
             directly on the box plots in Figure 22-12 because the scale of the box plots would
             be too large to readily observe data points at the lower end of the concentration
             range. Thus, the scale has been reduced to  10 |ig/m3. This figure shows that the
                                          22-25

-------
             annual average benzene concentrations for both sites are less than the program-
             level average and median concentrations of benzene. Further, the annual average
             for UCSD is less than the program-level first quartile (25th percentile) and the
             maximum benzene concentration measured at this site is just less than the
             program-level median concentration. UCSD has the lowest annual average
             benzene concentration among NMP sites sampling this pollutant. There were no
             non-detects of benzene measured at either site or across the program.

          •  Similar to the benzene graph, the program-level maximum 1,3-butadiene
             concentration (9.51 |ig/m3) is not shown directly on the box plots in Figure 22-13
             as the scale has been reduced to 3 |ig/m3 to allow for the observation of data
             points at the lower end of the concentration range. This figure shows that the
             annual average 1,3-butadiene concentrations for both sites are less than the
             program-level average and median concentrations. The annual average for UCSD
             is actually an order of magnitude less than the program-level median
             concentration. Further, the maximum 1,3-butadiene measured at UCSD is also
             less than the program-level median concentration. This site has lowest annual
             average 1,3-butadiene concentration among NMP sites sampling this pollutant.

          •  Figure 22-14 shows that although the annual average formaldehyde concentration
             for SSSD is greater than the annual average for UCSD, the annual averages for
             both sites are less than the program-level average and median concentrations. The
             annual average for UCSD is equivalent to the program-level first quartile. The
             maximum formaldehyde concentration measured at each site is considerably less
             than the maximum concentration measured across the program.  There were no
             non-detects of formaldehyde measured at either site.


22.4.3  Concentration Trends

       A site-specific trends evaluation was completed for sites that have sampled one or more

of the selected NATTS MQO Core Analytes for 5 consecutive  years or longer, as described in
Section 3.5.4. Sampling at SSSD began in 2008  after a re-location from a previous sampling site.

Sampling at UCSD under the NMP began in 2009 and was completed at the  end of 2011. Thus, a

trends analysis was not performed for either site.
22.5   Additional Risk-Based Screening Evaluations

       The following risk-based screening evaluations were conducted to characterize risk at the

South Dakota monitoring sites. Refer to Sections 3.3 and 3.5.5 for definitions and explanations

regarding the various toxicity factors, time frames, and calculations associated with these risk-

based screenings.
                                         22-26

-------
22.5.1  Risk-Based Screening Assessment Using MRLs
       A risk-based screening was conducted by comparing the concentration data from the
South Dakota monitoring sites to the ATSDR MRLs, where available. As described in
Section 3.3, MRLs are noncancer health risk benchmarks and are defined for three exposure
periods: acute (exposures of 1 to 14 days); intermediate (exposures of 15 to 364 days); and
chronic (exposures of 1 year or greater). The preprocessed daily measurements of the pollutants
of interest were compared to the acute MRLs; the quarterly averages were compared to the
intermediate MRLs; and the annual averages were compared to the chronic MRLs.

       As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2011.

22.5.2  Cancer Risk and Noncancer Hazard Approximations
       For the pollutants of interest for the South Dakota 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 may want to shift or
confirm their air-monitoring priorities. Refer to Section 3.5.5.2 for an explanation of how cancer
risk and noncancer hazard approximations are calculated and what limitations are  associated with
them. Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer
hazard approximations are presented in Table 22-6, where applicable. Cancer risk
approximations are presented as probabilities while the noncancer hazard approximations are
ratios and thus, unitless values.
                                         22-27

-------
            Table 22-6. Risk Approximations for the South Dakota Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Sioux Falls, South Dakota - SSSD
Acetaldehyde
Acrylonitrile
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
1 ,2-Dichloroethane
Formaldehyde
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
0.0000022
0.000068
0.0000078
0.00003
0.000006

0.000026
0.000013
0.00000026
0.0000048
0.0000088
0.009
0.002
0.03
0.002
0.1
0.098
2.4
0.0098
0.04
0.002
0.1
59/59
13/60
60/60
44/60
59/60
28/60
17/60
59/59
46/60
3/60
1/60
1.57
ฑ0.21
0.03
ฑ0.02
0.63
ฑ0.05
0.04
ฑ0.01
0.54
ฑ0.05
0.04
ฑ0.01
0.02
ฑ0.01
2.01
ฑ0.28
0.08
ฑ0.01
<0.01
ฑ<0.01
0.01
ฑ0.01
3.46
2.33
4.92
1.33
3.26

0.62
26.08
0.02
0.01
0.01
0.17
0.02
0.02
0.02
0.01
0.01
0.01
0.20
0.01
O.01
0.01
Union County, South Dakota - UCSD
Acetaldehyde
Acrylonitrile
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
1 ,2-Dichloroethane
Formaldehyde
Hexachloro- 1 ,3 -butadiene
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
0.0000022
0.000068
0.0000078
0.00003
0.000006

0.000026
0.000013
0.000022
0.00000026
0.0000048
0.0000088
0.009
0.002
0.03
0.002
0.1
0.098
2.4
0.0098
0.09
0.04
0.002
0.1
61/61
32/56
56/56
5/56
56/56
14/56
18/56
61/61
5/56
23/56
4/56
3/56
2.33
ฑ0.40
0.17
ฑ0.04
0.40
ฑ0.03
0.01
ฑ0.01
0.58
ฑ0.04
0.02
ฑ0.01
0.03
ฑ0.01
1.50
ฑ0.21
0.01
ฑ0.01
0.03
ฑ0.01
0.02
ฑ0.02
0.01
ฑ0.01
5.12
11.77
3.11
0.09
3.47

0.69
19.53
0.28
0.01
0.08
0.01
0.26
0.09
0.01
0.01
0.01
0.01
O.01
0.15
O.01
0.01
0.01
0.01
— = a Cancer URE or Noncancer RfC is not available
                                           22-28

-------
       Observations from Table 22-6 for SSSD include the following:

       •  The pollutants with the highest annual average concentrations for SSSD are
          formaldehyde, acetaldehyde, and benzene.

       •  Formaldehyde, benzene, and acetaldehyde have the highest cancer risk
          approximations among this site's pollutants of interest, although formaldehyde's
          cancer risk approximation is an order of magnitude higher than the cancer risk
          approximations for the other pollutants.

       •  None of the noncancer hazard approximations were greater than 1.0, indicating that
          no adverse health effects are expected from these individual pollutants.


       Observations from Table 22-6 for UCSD include the following:

       •  The pollutants with the highest annual average concentrations for UCSD are
          acetaldehyde, formaldehyde, and carbon tetrachloride.

       •  Formaldehyde has the highest cancer risk approximation for UCSD, followed by
          acrylonitrile. The cancer risk approximation for acrylonitrile (11.77 in-a-million) is
          the second highest cancer risk approximation for this pollutant. Although acrylonitrile
          has a much lower annual average concentration than formaldehyde, the cancer risk
          approximations are both greater than  10 in-a-million, indicating the relative toxicity
          of this pollutant.

       •  None of the noncancer hazard approximations for the pollutants of interest for UCSD
          were greater than 1.0, indicating that no adverse health effects are expected from
          these individual pollutants.


22.5.3 Risk-Based Emissions Assessment

       In addition to the risk-based screenings discussed above, Tables 22-7 and 22-8 present an

evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.

Table 22-7 presents the 10 pollutants with the highest emissions from the 2008 NEI, the 10

pollutants with the highest toxicity-weighted emissions, and the 10 pollutants with the highest

cancer risk approximations (in-a-million), as calculated from the annual averages provided in

Table 22-6. Table 22-8 presents similar information, but identifies the 10 pollutants with the

highest noncancer hazards approximations (HQ), also calculated from annual averages provided

in Table 22-6.
                                          22-29

-------
       Table 22-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer

                                          UREs for the South Dakota Monitoring Sites
Top 10 Total Emissions for Pollutants
with Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Sioux Falls, South Dakota (Minnehaha County) - SSSD
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
Dichloromethane
POM, Group 2b
POM, Group 6
POM, Group la
84.66
56.95
54.47
36.84
12.38
6.21
1.65
1.34
0.12
0.12
Formaldehyde
Benzene
1,3 -Butadiene
Naphthalene
POM, Group 3
Hexavalent Chromium, PM
Ethylbenzene
POM, Group 2b
Acetaldehyde
POM, Group 5a
7.40E-04
6.60E-04
3.72E-04
2.11E-04
1.94E-04
1.48E-04
1.36E-04
1.18E-04
8.10E-05
5.09E-05
Formaldehyde
Benzene
Acetaldehyde
Carbon Tetrachloride
Acrylonitrile
1,3 -Butadiene
1 ,2-Dichloroethane
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
26.08
4.92
3.46
3.26
2.33
1.33
0.62
0.02
0.01
<0.01
Union County, South Dakota (Union County) - UCSD
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
POM, Group 2b
Dichloromethane
POM, Group 6
POM, Group la
15.98
12.81
10.98
8.37
2.12
1.15
0.23
0.13
0.02
0.01
Formaldehyde
Benzene
1,3 -Butadiene
Naphthalene
Ethylbenzene
POM, Group 2b
Acetaldehyde
POM, Group 3
Hexavalent Chromium, PM
Arsenic, PM
1.67E-04
1.25E-04
6.36E-05
3.90E-05
2.74E-05
2.04E-05
1.84E-05
1.71E-05
1.56E-05
1.05E-05
Formaldehyde
Acrylonitrile
Acetaldehyde
Carbon Tetrachloride
Benzene
1 ,2-Dichloroethane
Hexachloro- 1 ,3 -butadiene
1,3 -Butadiene
Trichloroethylene
Vinyl Chloride
19.53
11.77
5.12
3.47
3.11
0.69
0.28
0.09
0.08
0.01
to
to

-------
       Table 22-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with

                                     Noncancer RfCs for the South Dakota Monitoring Sites
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer Hazard
Approximation
(HQ)
Sioux Falls, South Dakota (Minnehaha County) - SSSD
Toluene
Xylenes
Methanol
Benzene
Formaldehyde
Ethylbenzene
Hexane
Acetaldehyde
Ethylene glycol
1,3 -Butadiene
247.74
232.25
104.42
84.66
56.95
54.47
49.47
36.84
13.66
12.38
Acrolein
1,3 -Butadiene
Formaldehyde
Acetaldehyde
Benzene
Xylenes
Naphthalene
Lead, PM
Arsenic, PM
Propionaldehyde
148,988.96
6,191.87
5,811.72
4,093.12
2,821.95
2,322.48
2,068.40
871.54
550.67
461.12
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Benzene
Acrylonitrile
Carbon Tetrachloride
Tetrachloroethylene
Trichloroethylene
Chloroform
1 ,2-Dichloroethane
0.20
0.17
0.02
0.02
0.02
0.01
0.01
0.01
0.01
0.01
Union County, South Dakota (Union County) - UCSD
Toluene
Xylenes
Benzene
Formaldehyde
Ethylbenzene
Hexane
Acetaldehyde
Methanol
1,3 -Butadiene
1,1,1 -Trichloroethane
49.76
48.01
15.98
12.81
10.98
8.85
8.37
8.23
2.12
1.31
Acrolein
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Cyanide Compounds, gas
Benzene
Xylenes
Naphthalene
Arsenic, PM
Propionaldehyde
33,649.12
1,307.04
1,060.53
930.04
603.39
532.80
480.09
382.14
162.43
119.46
Acetaldehyde
Formaldehyde
Acrylonitrile
Benzene
Trichloroethylene
Carbon Tetrachloride
1,3 -Butadiene
Tetrachloroethylene
Chloroform
Hexachloro- 1 , 3 -butadiene
0.26
0.15
0.09
0.01
0.01
0.01
0.01
0.01
0.01
0.01
to
to

-------
       The pollutants listed in Tables 22-7 and 22-8 are limited to those that have cancer and

noncancer toxicity factors, respectively. As a result, although the actual value of the emissions is

the same, the highest emitted pollutants in the cancer table may be different from the noncancer

table. The cancer risk and noncancer hazard approximations based on each site's annual averages

are limited to those pollutants for which each respective site sampled. As discussed in

Section 22.3, SSSD and UCSD sampled for VOCs, SNMOCs, and carbonyl compounds. 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, as discussed in

previous sections. A more in-depth discussion of this analysis is provided in Section 3.5.5.3.

Similar to the cancer risk and noncancer hazard approximations, this analysis may help policy-

makers prioritize their air monitoring activities.


       Observations from Table 22-7 include the following:

       •  Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
          cancer UREs in Minnehaha and Union Counties. The emissions are higher in
          Minnehaha County than in Union County, although both counties have relatively low
          emissions of these pollutants. Union County has the lowest emissions of these three
          pollutants among all counties with NMP sites.

       •  Formaldehyde, benzene, and 1,3-butaidene are the pollutants with the highest
          toxicity-weighted emissions  (of the pollutants with cancer UREs) for both counties.

       •  Seven of the highest emitted pollutants also have the highest toxicity-weighted
          emissions for Minnehaha County. The  same seven pollutants appear on both
          emissions-based lists for Union County.

       •  Formaldehyde is the pollutant with the highest cancer risk approximation for SSSD;
          this pollutant also appears on both emissions-based lists. This is also true for
          acetaldehyde, benzene, and 1,3-butadiene. Conversely, several pollutants, including
          carbon tetrachloride and acrylonitrile, appear on neither emissions-based list but are
          among the pollutants with the highest cancer risk approximations for SSSD.

       •  Formaldehyde has the highest cancer risk approximation for UCSD and appears on
          both emissions-based lists. Conversely, acrylonitrile, which has the second highest
          cancer risk approximation for UCSD, appears on neither emissions-based list.


       Observations from Table 22-8 include the following:

       •  Toluene and xylenes are the highest emitted pollutants with noncancer RfCs in both
          Minnehaha and Union Counties. The emissions of these pollutants were an order of
          magnitude higher in Minnehaha County than in Union County.
                                         22-32

-------
       •  Acrolein is the pollutant with the highest toxi city-weighted emissions (of the
          pollutants with noncancer RfCs) for both counties. Although acrolein was sampled
          for at SSSD and UCSD, this pollutant was excluded from the pollutants of interest
          designation, and thus subsequent risk-based screening evaluations, due to questions
          about the consistency and reliability of the measurements, as discussed in Section 3.2.
          Acrolein is not one of the highest emitted pollutants in Minnehaha or Union Counties.

       •  Five of the highest emitted pollutants also have the highest toxi city-weighted
          emissions for Minnehaha County. The same five pollutants appear on both emissions-
          based lists for Union County.

       •  Formaldehyde and acetaldehyde, which have the highest noncancer hazard
          approximations for SSSD and UCSD, appear on both emissions-based lists. Benzene
          and 1,3-butadiene also appear on all three lists for each South Dakota monitoring site
          while acrylonitrile appears on neither emissions-based list.


22.6   Summary of the 2011 Monitoring Data for SSSD and UCSD

       Results from several of the data treatments described in this section include the
following:

       ปซป  Fourteen pollutants failed at least one screen for SSSD and 13 pollutants failed at
          least one screen for UCSD.

       ปซป  For both sites, formaldehyde and acetaldehyde are the only pollutants for which the
          annual average concentrations were greater than 1 jug/m3.

       *ป*  UCSD has  the second highest annual average concentration of acrylonitrile among
          NMP sites sampling VOCs. Conversely,  UCSD has the lowest annual average
          concentrations of benzene and 1,3-butadiene among sites sampling these pollutants.
                                         22-33

-------
23.0   Sites in Texas
       This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS sites in Texas, and integrates these concentrations with
emissions, meteorological, and risk information. Data generated by sources other than ERG are
not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) for detailed discussions and definitions
regarding the various data analyses presented below.

23.1   Site Characterization
       This section characterizes the CAMS 35 and CAMS 85 monitoring sites by providing
geographical  and physical information about the location of the sites and the surrounding areas.
This information is provided to give the reader insight regarding factors that may influence the
air quality near the sites  and assist in the interpretation of the ambient monitoring measurements.

       The CAMS 35 monitoring site is located in the Houston-Sugarland-Baytown, Texas
MSA and CAMS 85 is part of the Marshall, Texas MSA. Figures 23-1 and 23-3 are composite
satellite images retrieved from ArcGIS Explorer showing the monitoring sites in their urban and
rural locations. Figures 23-2  and 23-4 identify nearby point source emissions locations by source
category for each site, as reported in the 2008 NEI for point sources. Note that only sources
within 10 miles of the sites are included in the facility counts provided in Figures 23-2  and 23-4.
A 10-mile boundary was chosen to give the reader an indication of which emissions sources and
emissions source categories could potentially have a direct effect on the air quality at the
monitoring sites. Further, this boundary provides both the proximity of emissions sources to the
monitoring sites as well  as the quantity of such sources within a given distance of the sites.
Sources outside the 10-mile radius are still visible on each map, but have been grayed out in
order to show emissions sources just outside the boundary. Table 23-1 provides supplemental
geographical  information such as land use, location setting, and locational coordinates.
                                          23-1

-------
                                     Figure 23-1. Deer Park, Texas (CAMS 35) Monitoring Site
to

-------
     Figure 23-2. NEI Point Sources Located Within 10 Miles of CAMS 35
              Legend
               ฉ   CAMS 35 NATTS site
   O'w            95"5'(rw            95- Q'trw
Note: Due to facility donsitv and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
10 mile radius
                       County boundaries
Source Category Group (No. of Facilities)               ป
  -f1   Aircraft Operations (24)                                 ?
  I   Asphalt Processing/Roofing Manufacturing (1)              M
  B   Bulk Terminals/Bulk Plants (6)                            •
  C   Chemical Manufacturing (59)                             a
  *   Electricity Generation via Combustion (8)                   B
  E   Electroplating, Plating. Polishing, Anodizing, & Coloring (2)    R
  <•>   Fabricated Metal Products (8)                            ฑ
  IV   Glass Manufacturing (1)                                 •
  •v   Heating Equipment Manufacturing (1)                     •ซ&
  •   Landfill (2)                                            i
           Marine Port (4)
           Miscellaneous Commercial/Industrial (22)
           Miscellaneous Manufacturing (2)
           Oil and/or Gas Production (11)
           Petroleum Refinery (5)
           Pulp and Paper Plant/Wood Products (3)
           Rubber and Miscellaneous Plastics Products (19)
           Ship Building and Repairing (4)
           Stationary Combustion Turbines (1)
           Transportation and Marketing of Petroleum Products (6)
           Wastewater Treatment (2)
                                                   23-3

-------
Figure 23-3. Karnack, Texas (CAMS 85) Monitoring Site

-------
Figure 23-4. NEI Point Sources Located Within 10 Miles of CAMS 85
       Legend
                               94"15t>"W        W10WV        W5'B"W        94rO'0"W
                                       Note: Due to facilny density and collocation, the total facilities
                                       displayed may not represent all facilities within the area of interest.
             CAMS 85 NATTS site  Q   10 mile radius
County boundaries
                   Source Category Group (No. of Facilities)
                     -f   Aircraft Operations (3)
                     •   Oil and/or Gas Production (2)
                                      23-5

-------
                                    Table 23-1. Geographical Information for the Texas Monitoring Sites
Site Code
CAMS 35
CAMS 85
AQS Code
48-201-1039
48-203-0002
Location
Deer Park
Karnack
County
Harris
Harrison
Micro- or
Metropolitan
Statistical Area
Houston-Sugar
Land-Baytown,
TXMSA
Marshall, TX
MSA
Latitude
and
Longitude
29.670046,
-95.128485
32.669004,
-94.167449
Land Use
Residential
Agricultural
Location
Setting
Suburban
Rural
Additional Ambient Monitoring Information1
Haze, CO, NOy, NO, NO2, NOX, PAMS, NMOCs,
VOCs, Carbonyl compounds, O3, Meteorological
parameters, PM10, PM Coarse, PM10 Speciation,
PM2 5, and PM2 5 Speciation, SO2, SVOCs.
SVOCs, NO2, NO, NOX, PAMS, NMOCs, Carbonyl
Compounds, VOCs, O3, Meteorological parameters,
PMio, PM10 Speciation, PM25, PM2 5 Speciation.
  :Data for additional pollutants are reported to AQS for these sites (EPA, 2012c); however, these data are not generated by ERG and are therefore not included in this report.
  BOLD ITALICS = EPA-designated NATTS Site
to

-------
       The CAMS 35 monitoring site is located in Deer Park, southeast of Houston, in east
Texas. This site serves as the Houston NATTS Site. The site is located at Brown Memorial Park,
in a primarily residential area, as shown in Figure 23-1. Major thoroughfares are near the site,
including Beltway 8 (1.5 miles to the west) and Highway 225 (nearly 3 miles to the north).
Galveston Bay is located to the east and southeast of the site and the Houston Ship Channel,
which runs from the Bay westward towards downtown Houston, is located to the north on the
other side of Highway 225. The east side of Houston has significant industry, including several
oil refineries. As Figure 23-2 shows, the point source located closest to the CAMS 35 monitoring
site is a heliport at San Jacinto College in Pasadena. However, a large number of emissions
sources are located roughly along a line that runs east to west just north of the site (or along the
Houston Ship Channel). A second cluster of emissions sources is located to the southeast of the
monitoring site. The source category with the greatest number of sources (59) surrounding
CAMS 35 is chemical manufacturing. Other source categories with a number of sources around
CAMS 35 include aircraft operations, which include airports as well as small runways, heliports,
or landing pads; rubber and miscellaneous plastics products; and oil and gas production.

       The CAMS 85 NATTS site is located in Karnack, in northeast Texas. The monitoring site
is about 10 miles northeast of Marshall, Texas and about 7 miles from the Texas-Louisiana
border. This site is located on the property of the Longhorn Army Ammunition Plant near the
intersection of FM Road 134 and Spur Road 449 (Taylor Avenue), as shown in Figure 23-3. The
surrounding area is rural and agricultural. As Figure 23-4 shows, there are few point sources
within 10 miles of CAMS 85 and these sources  fall into two source categories: aircraft
operations and oil and gas production. The  closest source to CAMS 85 is the Fly-N-Fish Lodge
Airport.

       Table 23-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Texas monitoring sites. Table 23-2 includes county-level
population and vehicle registration information. Table 23-2 also includes a county-level vehicle
registration-to-population ratio, which was  calculated to represent the number of vehicles per
person within each monitoring site's residing county. In addition, the population within 10 miles
of each site is presented, based on postal code population data estimates. An estimate of 10-mile
vehicle ownership was then determined by  applying the county-level vehicle registration-to-
population ratio to the 10-mile population surrounding each monitoring site. Table 23-2 also
                                          23-7

-------
 contains traffic volume information for each site. Finally, Table 23-2 presents the county-level
 daily VMT for Harris and Harrison Counties.
           Table 23-2. Population, Motor Vehicle, and Traffic Information for the Texas
                                      Monitoring Sites
Site
CAMS 35
CAMS 85
Estimated
County
Population1
4,180,894
66,296
County-level
Vehicle
Registration2
3,164,173
70,585
Vehicles per
Person
(Registration:
Population)
0.76
1.06
Population
within 10
miles3
698,184
2,264
Estimated
10-mile
Vehicle
Ownership
528,398
2,410
Annual
Average
Daily
Traffic4
31,043
1,250
County-level
Daily VMT5
56,650,489
2,578,700
Bounty-level population estimates reflect 2011 data from the U.S. Census Bureau (Census Bureau, 2012b)
2County-level vehicle registration reflects 2011 data from the Texas Department of Motor Vehicles (TX DMV, 2012)
310-mile population estimates reflect 2011 data from Xionetic (Xionetic, 2011)
4AADT reflects 2010 data for CAMS 85 from the Texas DOT and 2004 data for CAMS 35 from Harris County Public
 Infrastructure Department (TX DOT, 2011 and HCPID, 2012)
5County-level VMT reflects 2011 data from the Texas DOT (TX DOT, 2012)
BOLD ITALICS = EPA-designated NATTS Site

        Observations from Table 23-2 include the following:

        •  The population and vehicle ownership counts are significantly higher for CAMS 35
           than CAMS 85. Compared to other counties with NMP monitoring sites, Harris
           County is third highest for both county-level population and county-level vehicle
           ownership. Conversely, Harrison County is among the lowest for both county-level
           population and vehicle ownership.

        •  The 10-mile populations for both CAMS 35 and CAMS 85 do not reflect the
           magnitude of the county-level populations, indicating that these sites are not located
           near the centers of highest population density.  The 10-mile population for CAMS 35
           is in the middle of the range compared to other NMP sites, while the 10-mile
           population for CAMS 85 is the lowest among all NMP sites. The estimated  10-mile
           vehicle ownerships for each site exhibit similar rankings.

        •  The vehicle-per-person ratio for CAMS 85 is higher than for CAMS 35. Compared to
           other sites, the ratio for CAMS 85 is in the top third while the ratio for CAMS 35 is in
           the bottom third.

        •  The traffic volume passing CAMS 35 is significantly higher than the traffic volume
           passing CAMS 85. The traffic volume for CAMS  35  is in the middle of the range
           compared to  other NMP sites while the traffic volume near CAMS 85 is among the
           lower traffic volumes for NMP sites. Traffic data for CAMS 35 are provided for
           Spencer Highway between Red Bluff Road and Underwood Road; the traffic data for
           CAMS 85 are provided for FM Road 134.
                                            23-8

-------
23.2   Meteorological Characterization
       The following sections characterize the meteorological conditions near the monitoring
sites in Texas on sample days, as well as over the course of the year.

23.2.1  Climate Summary
       The eastern third of Texas is characterized by a subtropical humid climate, with the
climate becoming more continental in nature farther north and west. The proximity to the Gulf of
Mexico acts as a moderating influence as temperatures soar in the summer or dip in the winter.
Areas closer to the coast, such as Houston, remain slightly cooler in the summer than
neighboring areas to the north. The reverse is also true, as coastal areas are warmer in the winter
than areas farther inland, although East Texas winters are relatively mild. The onshore flow from
the Gulf of Mexico also allows humidity levels to remain high in East Texas, particularly near
the coast. The winds flow out of the Gulf of Mexico a majority of the year, with the winter
months being the exception, as frontal systems allow colder air to filter in from the north.
Abundant rainfall is also typical  of the region, again due in part to the nearness to the Gulf of
Mexico. Severe weather is most  common in spring, particularly in May, and tropical systems can
be a threat to the state during the summer and fall. Snowfall is rare in East Texas but ice storms
are more common in northeast Texas than in other parts of the state (Bair, 1992 and TAMU,
2013).

23.2.2  Meteorological Conditions in 2011
       Hourly meteorological data from the NWS weather stations nearest these sites were
retrieved for 2011 (NCDC, 2011). The closest weather station to CAMS  35 is located at William
P. Hobby Airport, WBAN 12918; the closest weather station to CAMS 85 is located at
Shreveport Regional Airport, WBAN 13957. Additional information about the  Hobby Airport
and Shreveport Regional Airport weather stations, such as the distance between the  sites and the
weather stations, is provided in Table 23-3. These data were used to determine  how
meteorological conditions on sample days vary from conditions experienced throughout the year.
                                          23-9

-------
                        Table 23-3. Average Meteorological Conditions near the Texas Monitoring Sites
Closest NWS
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
Average
Temperature
Average
Dew Point
Temperature
Average
Wet Bulb
Temperature
Average
Relative
Humidity
Average
Sea Level
Pressure
(mb)
Average
Scalar
Wind
Speed
(kt)
Deer Park, Texas - CAMS 35
William P.
Hobby Airport
12918
(29.65, -95.28)
8.85
miles
258ฐ
(WSW)
Sample
Day
2011
80.1
ฑ3.8
81.2
+ 1.5
70.6
ฑ3.6
71.7
+ 1.4
56.8
ฑ4.0
58.6
+ 1.5
62.7
ฑ3.4
64.0
+ 1.3
65.8
ฑ3.1
67.1
+ 1.3
1016.7
ฑ1.5
1016.3
+ 0.6
7.2
ฑ0.8
7.1
+ 0.3
Karnack, Texas - CAMS 85
Shreveport
Regional
Airport
13957
(32.45, -93.82)
24.46
miles
127ฐ
(SE)
Sample
Day
2011
77.9
ฑ4.8
79.4
+ 1.9
66.4
ฑ4.6
68.0
+ 1.8
51.0
ฑ4.0
52.7
+ 1.5
57.6
ฑ3.8
59.1
+ 1.4
62.5
ฑ3.4
62.6
+ 1.4
1016.2
ฑ 1.6
1015.7
+ 0.6
6.9
ฑ0.8
6.7
+ 0.3
Sample day averages are highlighted to help differentiate the sample day averages from the full-year averages.
to




o

-------
       Table 23-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for the entire year. Also included in Table 23-3 is the 95 percent
confidence interval for each parameter. Although the differences are not statistically significant,
the average temperature and moisture parameters for both sites are slightly higher for all of 2011
than they are for sample days alone.

23.2.3 Back Trajectory Analysis
       Figure 23-5 is the composite back trajectory map for days on which samples were
collected at the CAMS 35 monitoring site in 2011. Included in Figure 23-5 are four back
trajectories per sample day. Figure 23-6 is the corresponding cluster analysis. Figure 23-7 is the
composite back trajectory map for days on which samples were collected at CAMS 85 and
Figure 23-8 is the corresponding cluster analysis. An in-depth description of these maps and how
they were generated is presented in Section  3.5.2.1. For the composite maps, each line represents
the 24-hour trajectory along which a parcel  of air traveled toward the monitoring site on a given
sample day and time, based on an  initial height of 50 meters AGL. For the cluster analyses, each
line corresponds to a trajectory representative of a given cluster of back trajectories. Each
concentric circle around the sites in Figures 23-5 through 23-8 represents 100 miles.

       Observations from Figures 23-5 and 23-6 for CAMS 35 include the following:
       •  Back trajectories originated from a variety of directions at the CAMS 35 monitoring
           site, although the majority of trajectories originated over the Gulf of Mexico or to the
          north of the site and rarely to the west of the site.
       •  The 24-hour air shed domain for CAMS 35 is among the largest in size compared to
          other NMP monitoring sites. The average trajectory length was 278 miles, which is
          the fifth highest average trajectory length among NMP sites. Two back trajectories
          originated greater than 800 miles away, over central Nebraska. Approximately
           82 percent of back trajectories originated within 400 miles of the  site.
       •  The cluster analysis shows that greater than 50 percent of back trajectories originated
          over the Gulf of Mexico, although the position over the Gulf and the trajectory length
          varies. Another common trajectory origin is from the northwest to northeast
          (25 percent).  The short cluster trajectory originating near the Louisiana border
          (21 percent) represents back trajectories originating primarily over east Texas and
          Louisiana as well as relatively short trajectories originating from other directions and
          generally within 300 miles of CAMS 35.

                                          23-11

-------
Figure 23-5. 2011 Composite Back Trajectory Map for CAMS 35
    Figure 23-6. Back Trajectory Cluster Map for CAMS 35
                           23-12

-------
Figure 23-7. 2011 Composite Back Trajectory Map for CAMS 85
    Figure 23-8. Back Trajectory Cluster Map for CAMS 85
                           23-13

-------
       Observations from Figures 23-7 and 23-8 for CAMS 85 include the following:
       •  Back trajectories originated from a variety of directions at the CAMS 85 monitoring
          site, although back trajectories originating to the east and west are rare. The
          composite map for CAMS 85 resembles the composite map for CAMS 35 in the
          direction of trajectory origin.
       •  The 24-hour air shed domain for CAMS 85 is slightly smaller in size compared to
          CAMS 35 but is similar in size to many other NMP monitoring sites. The farthest
          away a trajectory originated was 800 miles away, near the Wyoming/Nebraska
          border. However, the average trajectory length is 260 miles and most trajectories
          (86 percent) originated less than 400 miles from CAMS 85.
       •  The cluster analysis for CAMS  85 shows that that 44 percent of back trajectories
          originated to the south of the site, as indicated by the short cluster (30 percent)
          representing back trajectories originating over East Texas and the longer cluster
          (14 percent) originating over the Gulf of Mexico. Another common trajectory origin
          is from the southeast over Louisiana (24 percent). Additionally, 33 percent of back
          trajectories originated to the northwest to northeast of the site, as indicated by the
          short cluster (16 percent) representing relatively short back trajectories originating
          over Arkansas and the longer cluster (17 percent) originating over the central Plains.

23.2.4 Wind Rose Comparison
       Hourly surface wind data from the NWS weather stations at Hobby Airport near
CAMS 35 and Shreveport Regional Airport near CAMS 85 were uploaded into a wind rose
software program to produce customized wind roses, as described in Section 3.5.2.2. A wind
rose shows the frequency of wind directions using "petals" positioned around a 16-point
compass, and uses different colors to represent wind speeds.
       Figure 23-9 presents a map showing the distance between the NWS station and
CAMS 35, which may be useful for identifying topographical influences that may affect the
meteorological patterns experienced at this location. Figure 23-9 also presents three different
wind roses for the CAMS 35 monitoring site. First, a historical wind rose representing 2001 to
2010 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
2011 is presented. Next, a wind rose representing wind data for days on which samples were
collected in 2011 is presented. These can be used to identify the predominant wind speed and
direction for 2011 and determine if wind observations on sample days were representative of
conditions experienced over the entire year and historically. Figure 23-10 presents the distance
map and three wind roses for CAMS 85.
                                          23-14

-------
Figure 23-9. Wind Roses for the William P. Hobby Airport Weather Station near CAMS 35
 Distance between CAMS 35 and NWS Station
2001-2010 Historical Wind Rose
    i    \       C
            2011 Wind Rose
    Sample Day Wind Rose
                                      23-15

-------
  Figure 23-10. Wind Roses for the Shreveport Regional Airport Weather Station near
                                      CAMS 85
Distance between CAMS 85 and NWS Station
2001-2010 Historical Wind Rose
                                                IWEST
                                                                               WIND SPEED
                                                                               (Knots)
                                                                               O *=
                                                                               ^| 17 • 21
                                                                               ^| 11 - 17
                                                                               ^| 7- 11
                                                                               CH 4-7
                                                                               !• 2- 4
                                                                               Calms: 18.58%
            2011 Wind Rose
    Sample Day Wind Rose
                                        23-16

-------
       Observations from Figure 23-9 for CAMS 35 include the following:

       •  The Hobby Airport weather station is located approximately 8.9 miles west-southwest
          of CAMS 35.

       •  The historical wind rose shows that winds from the southeast quadrant, including
          both easterly and southerly winds, prevailed near the CAMS 35 site. Northerly winds
          were also observed often. Calm winds (<2 knots) were observed for approximately
          14 percent of the wind measurements.

       •  Winds from the southeast quadrant were frequently observed in 2011 as well, but the
          number of south-southeasterly and southerly wind observations was higher in 2011
          than historically. This is also true for the sample day wind rose.


       Observations from Figure 23-10 for CAMS 85 include the following:

       •  The Shreveport Regional Airport weather station is located across the Texas-
          Louisiana border, approximately 24.5 miles southeast of CAMS 85.

       •  The wind patterns on the historical wind rose for CAMS 85 resemble those on the
          historical wind rose for CAMS 35. The historical wind rose shows that winds from
          the southeast to south account for approximately 30 percent of the wind observations
          near the CAMS 85. Northerly winds were also observed fairly often. Calm winds
          were observed for approximately 17 percent of the wind measurements.

       •  The wind patterns shown on the 2011 wind rose are similar to the historical wind
          patterns, although the number of southerly winds increased from roughly 13 percent
          to 20 percent. A similar trend is shown for the CAMS 35 wind rose.

       •  The sample day wind patterns resemble the full-year wind patterns, indicating that
          wind conditions on sample days were representative of those experienced throughout
          2011.


23.3   Pollutants of Interest

       Site-specific "pollutants of interest" were determined for the Texas monitoring sites in

order to allow 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. In

addition, if any of the NATTS MQO Core Analytes measured by each monitoring site did not

meet the pollutant of interest criteria based on the preliminary risk-based screening, that pollutant
                                          23-17

-------
was added to the list of site-specific pollutants of interest. A more in-depth description of the

risk-based screening process is presented in Section 3.2.


       Table 23-4 presents the results of the preliminary risk-based screening process for

CAMS 35 and CAMS 85. The pollutants that failed at least one screen and contributed to

95 percent of the total failed screens for each monitoring site are shaded. NATTS MQO Core

Analytes are bolded. Thus, pollutants of interest are shaded and/or bolded. CAMS 35 sampled

for hexavalent chromium and PAHs while CAMS 85 sampled for hexavalent chromium only.
         Table 23-4. Risk-Based Screening Results for the Texas Monitoring Sites
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Deer Park, Texas - CAMS 35
Naphthalene
Fluorene
Acenaphthene
Hexavalent Chromium
Fluoranthene
0.029
0.011
0.011
0.000083
0.011
Total
53
7
6
6
o
J
75
58
58
58
60
58
292
91.38
12.07
10.34
10.00
5.17
25.68
70.67
9.33
8.00
8.00
4.00
70.67
80.00
88.00
96.00
100.00

Karnack, Texas - CAMS 85
Hexavalent Chromium
0.000083
Total
0
0
50
50
0.00
0.00
0.00
0.00

       Observations from Table 23-4 include the following:

       •  Five pollutants, including two NATTS MQO Core Analytes, failed at least one screen
          for CAMS 35. Naphthalene contributed to 71 percent of the total number of failed
          screens for CAMS 35.

       •  Naphthalene, fluorene, acenaphthene, and hexavalent chromium were initially
          identified as pollutants of interest for CAMS 35. Benzo(a)pyrene was added to the
          pollutants of interest for CAMS 35 because it is a NATTS MQO Core Analyte, even
          though it did not fail any screens. Benzo(a)pyrene is not shown in Table 23-4 but is
          shown in subsequent tables in the sections that follow.

       •  Hexavalent chromium is the only pollutant sampled for at CAMS 85. This pollutant
          did not fail any  screens  during the 2011 monitoring effort. However, because it is a
          NATTS MQO Core Analyte, and because it is the only pollutant sampled for at this
          site, hexavalent chromium is the pollutant of interest for CAMS 85.
                                         23-18

-------
23.4   Concentrations
       This section presents various concentration averages used to characterize pollution levels
at the Texas monitoring sites. Where applicable, the following calculations and data analyses
were performed: Time period-based concentration averages (quarterly and annual) are provided
for the pollutants of interest for the Texas monitoring sites, where the data meet the applicable
criteria. Concentration averages for select pollutants are also presented graphically for the sites to
illustrate how the sites' concentrations compare to the program-level averages, as presented in
Section 4.1. In addition, concentration averages for select pollutants are presented from previous
years of sampling in order to characterize concentration trends at the sites. Additional site-
specific statistical summaries for CAMS 35 and CAMS 85 are provided in Appendices M and O.

23.4.1 2011 Concentration Averages
       Quarterly and  annual concentration averages were calculated for the pollutants of interest
for each Texas 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 of 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 Texas monitoring sites are
presented in Table 23-5, where applicable. Note that if a pollutant was not detected in a given
calendar quarter, the quarterly average simply reflects "0" because only zeros substituted for
non-detects were factored into the quarterly average concentration.
                                          23-19

-------
  Table 23-5. Quarterly and Annual Average Concentrations of the Pollutants of
                     Interest for the Texas Monitoring Sites
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(ng/m3)
2nd
Quarter
Average
(ng/m3)
3rd
Quarter
Average
(ng/m3)
4th
Quarter
Average
(ng/m3)
Annual
Average
(ng/m3)
Deer Park, Texas - CAMS 35
Acenaphthene
Benzo(a)pyrene
Fluorene
Hexavalent Chromium
Naphthalene
58/58
23/58
58/58
60/60
58/58
2.45
ฑ0.72
0.02
ฑ0.02
3.05
ฑ0.76
0.04
ฑ0.01
81.21
ฑ27.31
4.94
ฑ1.58
0.01
ฑ0.01
5.57
ฑ2.08
0.05
ฑ0.01
52.65
ฑ 18.88
14.08
ฑ7.24
0.02
ฑ0.01
17.00
ฑ9.06
0.06
ฑ0.01
113.19
ฑ35.95
3.76
ฑ0.77
0.04
ฑ0.02
4.42
ฑ0.83
0.04
ฑ0.01
126.77
ฑ 49.62
6.58
ฑ2.30
0.02
ฑ0.01
7.84
ฑ2.85
0.05
ฑ0.01
94.14
ฑ 18.02
Karnack, Texas - CAMS 85
Hexavalent Chromium
50/61
0.02
ฑ0.01
0.03
ฑ<0.01
0.02
ฑ0.01
0.01
ฑ0.01
0.02
ฑ<0.01
Observations from Table 23-5 include the following:

•  Naphthalene's annual average concentration is significantly higher than the annual
   averages for the other pollutants of interest for CAMS 35.

•  Although naphthalene concentrations appear to be higher during the second half of
   the year, the confidence intervals indicate that concentrations of this pollutant have a
   considerable amount of variability associated with them. The maximum concentration
   of naphthalene was measured at CAMS 35 on November 11, 2011 (299 ng/m3), and
   additional concentrations greater than 200 ng/m3 were also measured in August (2),
   November (1), and December (1). While concentrations greater than 100 ng/m3 were
   measured in each quarter of 2011, most were measured during the second half of the
   year (four in the first quarter, one in the second, seven in the third, and seven in the
   fourth).

•  The third quarter average concentrations for acenaphthene and fluorene were
   significantly higher than the other quarterly averages and have relatively large
   confidence intervals associated with them. Concentrations of acenaphthene ranged
   from 1.02 ng/m3 to 52.9 ng/m3, with the five highest concentrations all measured in
   August. All 13 concentrations greater than 6 ng/m3 were measured between June and
   September. Conversely, the nine concentrations of acenaphthene less than 2 ng/m3
   were all measured in January, February, or December. A similar trend is exhibited by
   fluorene.

•  Benzo(a)pyrene was detected in fewer than half of the valid PAH samples collected at
   CAMS 35. Measured detections ranged from 0.0196 ng/m3 to 0.151 ng/m3, with only
   one measurement greater than 0.1 ng/m3.
                                  23-20

-------
       •   Concentrations of hexavalent chromium were higher at CAMS 35 than at CAMS 85.
          The hexavalent chromium concentrations ranged from 0.0019 ng/m3 to 0.159 ng/m3
          at CAMS 35, with a total of three measurements greater than 0.1 ng/m3. Hexavalent
          chromium concentrations ranged from a few non-detects to 0.0522 ng/m3 at
          CAMS 85. Hexavalent chromium was detected in 100 percent of the valid samples
          collected at CAMS 35, while the detection rate was lower at CAMS 85 (82 percent).

       •   The annual average concentration of hexavalent chromium for CAMS 35 for 2011 is
          similar to the annual average for 2010. This same is not true for CAMS 85. In 2010,
          CAMS 85 had the highest annual average concentration of hexavalent chromium
          (0.31 ฑ 0.16 ng/m3) among all NMP sites sampling this pollutant and was six times
          higher than the annual average concentration for the Deer Park site. For 2011, the
          annual average concentration of hexavalent chromium for CAMS 85 is an order of
          magnitude lower (0.02 ฑ <0.01 ng/m3). This may be attributable to the use of stainless
          steel filter holders used in the sampler which may have contaminated the samples.
          Changing to a Teflonฎ filter holder has resulted in a decrease in hexavalent chromium
          concentrations at CAMS 85. The filter holder was exchanged at the end of 2010.


       Tables 4-9 through 4-12 present the sites with the  10 highest annual average

concentrations for each of the program-level pollutants of interest. Observations for CAMS 35

and CAMS 85 from those tables include the following:

       •   CAMS 35 appears in Table 4-11 three times. The annual average fluorene
          concentration for CAMS 35 ranks fifth highest among sites sampling PAHs.
          CAMS 35 also ranks sixth for acenaphthene and seventh for naphthalene.

       •   As shown in Table 4-12, the annual average hexavalent chromium concentration for
          CAMS 35 is the second highest annual average hexavalent chromium concentration
          among NMP sites sampling this pollutant, behind only PXSS. The annual average
          hexavalent chromium concentration for CAMS 85 ranks tenth highest among NMP
          sites.

23.4.2  Concentration Comparison

       In order to better illustrate how a site's annual average concentrations compare to the

program-level averages, a site-specific box plot was created for the selected NATTS MQO Core

Analytes listed in Section 3.5.3, where applicable. Thus, box plots for benzo(a)pyrene,

hexavalent chromium, and naphthalene were created for CAMS 35. A box plot for hexavalent

chromium was also created for CAMS 85. Figures 23-11 through 23-13 overlay the sites'

minimum, annual average, and maximum  concentrations onto the program-level minimum, first

quartile, median,  average, third quartile, and maximum concentrations, as described in

Section 3.5.3.
                                         23-21

-------
      Figure 23-11. Program vs. Site-Specific Average Benzo(a)pyrene Concentration
CAMS 35
      F
                 DL25
                             0.5
                                         0.75           1           1.25
                                                Concentration (ne/m3)
                                                                             15
                                                                                         1.75
                 Program:   IstQuartile      2ndQuartile     3rdQuartile     4thQuartile     Average
                 Site:      Site Average     Site Minimum/Maximum

 Figure 23-12. Program vs. Site-Specific Average Hexavalent Chromium Concentrations
      1
CAMS 85
                     0.05
                                     DL1
                                                     0.15
                                                Concentration (ng/m3J
                                                                     0.2
                                                                                     0.25
                                                                                                     DL3
                 Program:    IstQuartile     2ndQuartile     SrdQuartile      4thQuartile      Average
                 Site:      Site Average     Site Minimum/Maximum

       Figure 23-13. Program vs. Site-Specific Average Naphthalene Concentration
CAMS 35
                                                               H     Program Max Concentration =779 ng/m3
                        100
                                  is:
                                           200        250       300
                                                Concentration (ng/m3)
                                                                        350
                                                                                 403
                                                                                          45;
                                                                                                    5DD
                 Program:    IstQuartile     2ndQuartile     SrdQuartile      4thQuartile      Average
                 Site:      Site Average     Site Minimum/Maximum

                                                23-22

-------
       Observations from Figures 23-11 through 23-13 include the following:

          •  Figure 23-11 is the box plot for benzo(a)pyrene for CAMS 35. Note that the first
             quartile for this pollutant is zero and is not visible on this box plot. This box plot
             shows that the annual average concentration for CAMS 35 is less than both the
             program-level average and median concentrations. Figure 23-11 also shows that
             the maximum concentration measured at CAMS 35 is considerably less than the
             maximum concentration measured across the program. More than half of the
             measurements of benzo(a)pyrene were non-detects at CAMS 35, as discussed in
             the previous section.

          •  Figure 23-12 for hexavalent chromium shows both sites, as both CAMS 35 and
             CAMS 85 sampled this pollutant. The annual average concentration for CAMS 35
             is greater than the program-level third quartile (75th percentile). The annual
             average concentration for CAMS 85 is less than the program-level average but
             greater than the program-level median concentration. The concentration range
             measured at CAMS 35 is more than three times the concentration range for
             CAMS 85, although the maximum concentration measured at both sites is  less
             than the maximum concentration measured across the program. There were no
             non-detects of hexavalent chromium measured at CAMS 35 while 11 non-detects
             were reported for CAMS 85.

          •  Figure 23-13 is the box plot for naphthalene for CAMS 35. Note that the
             program-level maximum concentration (779 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 500 ng/m3. Figure 23-13 shows that the annual average naphthalene
             concentration for CAMS 35 is greater than the program-level average
             concentration. The maximum naphthalene concentration measured at CAMS 35 is
             significantly less than the program-level maximum concentration. There were no
             non-detects of naphthalene measured at CAMS 35 or across the program.


23.4.3  Concentration Trends

       A site-specific trends evaluation was completed for sites that have sampled one or more

of the selected NATTS MQO Core Analytes for 5 consecutive years or longer, as described in

Section 3.5.4. CAMS 35 has not sampled PAHs continuously for 5 years as part of the NMP  and
both sites began sampling hexavalent chromium under the NMP in 2010. Therefore, a trends

analysis was not conducted.
                                         23-23

-------
23.5   Additional Risk-Based Screening Evaluations
       The following risk-based screening evaluations were conducted to characterize risk at the
Texas monitoring sites. Refer to Sections 3.3 and 3.5.5 for definitions and explanations regarding
the various toxicity factors, time frames, and calculations associated with these risk-based
screenings.

23.5.1  Risk-Based Screening Assessment Using MRLs
       A risk-based screening was conducted by comparing the concentration data from the
Texas monitoring sites to the ATSDRMRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 to 14 days); intermediate (exposures of 15 to 364 days); and chronic (exposures
of 1 year or greater). The preprocessed daily measurements of the pollutants of interest were
compared  to the acute MRLs; the quarterly averages were compared to  the intermediate MRLs;
and the annual averages were compared to the chronic MRLs.

       As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the  monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2011.

23.5.2  Cancer Risk and Noncancer Hazard Approximations
       For the pollutants of interest for the Texas 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  may want to shift or
confirm their air-monitoring priorities.  Refer to Section 3.5.5.2 for an explanation of how cancer
risk and noncancer hazard approximations are calculated and what limitations are associated with
them. Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer
hazard approximations are presented in Table 23-6, where applicable. Cancer risk
approximations are presented as probabilities while the noncancer hazard approximations are
ratios and  thus, unitless values.
                                         23-24

-------
              Table 23-6. Risk Approximations for the Texas Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Deer Park, Texas - CAMS 35
Acenaphthene
Benzo(a)pyrene
Fluorene
Hexavalent Chromium
Naphthalene
0.000088
0.00176
0.000088
0.012
0.000034



0.0001
0.003
58/58
23/58
58/58
60/60
58/58
6.58
ฑ2.30
0.02
ฑ0.01
7.84
ฑ2.85
0.05
ฑ0.01
94.14
ฑ 18.02
0.58
0.04
0.69
0.59
3.20



0.01
0.03
Karnack, Texas - CAMS 85
Hexavalent Chromium
0.012
0.0001
50/61
0.02
ฑ0.01
0.26
0.01
— = a Cancer URE or Noncancer RfC is not available
        Observations from Table 23-6 include the following:
        •   The cancer risk approximation for naphthalene for CAMS 35 is 3.20 in-a-million,
           based on the annual average. This is the only cancer risk approximation greater than
           1.0 in-a-million for CAMS 35.
        •   The cancer risk approximation for hexavalent chromium for CAMS 85
           (0.26 in-a-million) is half the cancer risk approximation for hexavalent chromium for
           CAMS 35 (0.59 in-a-million), although both are less than 1 in-a-million.
        •   The noncancer hazard approximations for CAMS 35 and CAMS 85, where they could
           be calculated, are less than 1.0, indicating that no adverse health effects are expected
           from these individual pollutants.

 23.5.3  Risk-Based Emissions Assessment
        In addition to the risk-based screenings discussed above, Tables 23-7 and 23-8 present an
 evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
 Table 23-7 presents the 10 pollutants with the highest emissions from the 2008 NEI, the
 10 pollutants with the highest toxicity-weighted emissions, and the 10 pollutants with the highest
 cancer risk approximations (in-a-million), as calculated from the annual averages provided in
 Table 23-6. Table 23-8 presents similar information, but identifies the 10 pollutants with the
 highest noncancer hazard approximations (HQ), also calculated from annual averages provided
 in Table 23-6.
                                           23-25

-------
Table 23-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
                                                the Texas Monitoring Sites
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
Deer Park, Texas (Harris County) - CAMS 35
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Methyl tert butyl ether
Naphthalene
Propylene oxide
Dichloromethane
Tetrachloroethylene
1,418.35
1,260.11
827.79
699.55
446.24
168.21
142.12
85.80
77.80
23.32
Formaldehyde
1,3 -Butadiene
Benzene
Hexavalent Chromium, PM
Naphthalene
Nickel, PM
Arsenic, PM
Ethylbenzene
Acetaldehyde
POM, Group 2b
1.64E-02
1.34E-02
1.11E-02
1.04E-02
4.83E-03
4.62E-03
2.26E-03
2.07E-03
1.54E-03
1.44E-03
Naphthalene 3.20
Fluorene 0.69
Hexavalent Chromium 0.59
Acenaphthene 0.58
Benzo(a)pyrene 0.04

Karnack, Texas (Harrison County) - CAMS 85
Formaldehyde
Benzene
Acetaldehyde
Ethylbenzene
1,3 -Butadiene
Naphthalene
Ethylene oxide
Dichloromethane
Chloromethylbenzene
Carbon tetrachloride
108.38
64.37
55.52
37.76
15.22
13.37
9.90
3.67
1.31
1.08
Hexavalent Chromium, PM
Formaldehyde
Ethylene oxide
Benzene
1,3 -Butadiene
Naphthalene
Nickel, PM
Arsenic, PM
Acetaldehyde
Ethylbenzene
6.35E-03
1.41E-03
8.72E-04
5.02E-04
4.57E-04
4.54E-04
3.29E-04
1.34E-04
1.22E-04
9.44E-05
Hexavalent Chromium 0.26

to
to

-------
Table 23-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
                                           RfCs for the Texas Monitoring Sites
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Noncancer Hazard
Approximation
Pollutant (HQ)
Deer Park, Texas (Harris County) - CAMS 35
Toluene
Xylenes
Methanol
Hexane
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Styrene
4,428.76
3,249.06
2,707.58
1,559.23
1,418.35
1,260.11
827.79
699.55
446.24
359.89
Acrolein
1,3 -Butadiene
Formaldehyde
Nickel, PM
Acetaldehyde
Titanium tetrachloride
Hexamethylene-l,6-diisocyanate, gas
Chlorine
Naphthalene
Benzene
4,253,301.87
223,117.95
128,583.14
106,919.25
77,728.16
77,037.49
62,970.00
57,589.67
47,373.42
47,278.41
Naphthalene 0.03
Hexavalent Chromium O.01

Karnack, Texas (Harrison County) - CAMS 85
Toluene
Xylenes
Formaldehyde
Benzene
Ethylene glycol
Acetaldehyde
Hexane
Methanol
Chloromethane
Ethylbenzene
155.11
142.12
108.38
64.37
63.56
55.52
53.09
48.73
40.86
37.76
Acrolein
Hexamethylene-l,6-diisocyanate, gas
Manganese, PM
Chlorine
Formaldehyde
Cyanide Compounds, PM
Nickel, PM
1,3 -Butadiene
Acetaldehyde
Hexavalent Chromium, PM
647,128.44
31,490.00
23,210.90
22,445.25
11,059.67
9,776.20
7,623.51
7,611.02
6,168.54
5,294.84
Hexavalent Chromium O.01

to
to

-------
       The pollutants listed in Tables 23-7 and 23-8 are limited to those that have cancer and

noncancer toxicity factors, respectively. As a result, although the actual value of the emissions is

the same, the highest emitted pollutants in the cancer table may be different from the noncancer

table. The cancer risk and noncancer hazard approximations based on each site's annual averages

are limited to those pollutants for which each respective site sampled. As discussed in
Section 23.3, both Texas monitoring sites sampled hexavalent chromium; CAMS 35 also

sampled for PAHs. In addition, the cancer risk and noncancer hazard approximations are limited

to those pollutants with enough data to meet the criteria for annual averages to be calculated. A
more in-depth discussion of this analysis is provided in Section 3.5.5.3. Similar to the cancer risk

and noncancer hazard approximations, this analysis may help policy-makers prioritize their air

monitoring activities.


       Observations from Table 23-7 include the following:

       •   Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
          cancer UREs in Harris County. Formaldehyde, benzene, and acetaldehyde are the
          highest emitted pollutants with cancer UREs in Harrison County. The magnitude of
          the emissions is significantly higher in Harris County than Harrison County.

       •   The pollutants with the highest toxicity-weighted emissions (of the pollutants with
          cancer UREs) for Harris County are formaldehyde, 1,3-butadiene, and benzene. The
          pollutants with the highest toxicity-weighted emissions for Harrison County are
          hexavalent chromium, formaldehyde, and ethylene oxide.

       •   Six of the highest emitted pollutants in Harris County also have the highest toxicity-
          weighted emissions while seven of the highest emitted pollutants in Harrison  County
          also have the highest toxicity-weighted emissions.

       •   Naphthalene is the only pollutant of interest for CAMS 35 that appears on both
          emissions-based lists for Harris County. Although hexavalent chromium, which has
          the third highest cancer risk approximation for CAMS 35,  ranks fourth for toxicity-
          weighted emissions, this pollutant is not one of the highest emitted in Harris County.

       •   POM, Group 2b ranks tenth for toxicity-weighted emissions in Harris County. POM,
          Group 2b includes several PAHs sampled for at CAMS 35 including acenaphthene
          and fluorene, both  pollutants of interest for CAMS 35. Benzo(a)pyrene, another
          pollutant of interest for CAMS 35, is part of POM, Group  5a,  which does not appear
          on either emissions-based list for Harris County.

       •   Hexavalent chromium, the only pollutant of interest for CAMS 85, is the pollutant
          with the  highest toxicity-weighted emissions for Harrison County, but is not among
          the 10 highest emitted (its emissions rank 13th).
                                         23-28

-------
       Observations from Table 23-8 include the following:

       •  Toluene, xylenes, and methanol are the highest emitted pollutants with noncancer
          RfCs in Harris County. Toluene, xylenes, and formaldehyde are the highest emitted
          pollutants in Harrison County. The magnitude of the emissions is significantly higher
          in Harris County than Harrison County.

       •  The pollutant with the highest toxicity-weighted emissions (of the pollutants with
          noncancer RfCs) for both counties is acrolein.

       •  Four of the highest emitted pollutants also have the highest toxicity-weighted
          emissions for Harris County while only two of the highest emitted pollutants also
          have the highest toxicity-weighted emissions for Harrison County.

       •  Hexamethylene-l,6-diisocyante gas appears among the pollutants with the highest
          toxicity-weighted emissions for both sites. This pollutant only appears among the
          pollutants with the highest toxicity-weighted emissions for one additional NMP site
          (CELA).

       •  Naphthalene ranks ninth for toxicity-weighted emissions in Harris County but is not
          one of the highest emitted pollutants with a noncancer RfC. Hexavalent chromium
          does not appear on either emissions-based list for Harris County. These are the only
          two pollutants of interest for CAMS 35 with noncancer toxicity factors.

       •  Hexavalent chromium ranks tenth for toxicity-weighted emissions in Harrison County
          but is not one of the highest emitted (with a noncancer RfC).


23.6   Summary of the 2011 Monitoring Data for CAMS 35 and CAMS 85

       Results from several of the data treatments described in this section include the
following:

       ปซป  Five pollutants failed at least one screen for CAMS 35, with naphthalene accounting
          for 71 percent of the total failed screens. Hexavalent chromium, the only pollutant
          sampled for at CAMS 85, did not fail any screens.

       *ป*  Of the site-specific pollutants of the interest, naphthalene had the highest annual
          average concentration for CAMS 35.

       ปซป  Concentrations of acenaphthene andfluorene were highest during the summer
          months  at CAMS 35.

       ปซป  Concentrations ofhexavalent chromium measured at CAMS 85 decreased
          significantly from 2010 to 2011. This is a result of replacing the stainless steel filter
          holder in the sampler with a Teflonฎ filter holder.
                                         23-29

-------
24.0   Site in Utah
       This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS site in Utah, and integrates these concentrations with
emissions, meteorological, and risk information. Data generated by sources other than ERG are
not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) for detailed discussions and definitions
regarding the various data analyses presented below.

24.1   Site Characterization
       This section characterizes the BTUT 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.

       BTUT is located in Bountiful, in northern Utah. Figure 24-1 is a composite satellite
image retrieved from ArcGIS Explorer showing the monitoring site in its urban location.
Figure 24-2 identifies nearby point source emissions locations by source category, as reported in
the 2008 NEI for point sources. Note that only sources within 10 miles  of the site are included in
the facility counts provided in Figure 24-2. A 10-mile boundary was chosen to give the reader an
indication of which emissions sources  and emissions source categories  could potentially have a
direct effect on the air quality at the monitoring site. Further, this boundary  provides both the
proximity of emissions sources to the monitoring site as well as the quantity of such sources
within a given distance of the site. Sources outside the 10-mile radius are still visible on the map,
but have been grayed out in order to show emissions sources just outside the boundary.
Table 24-1 provides supplemental geographical information such as land use, location setting,
and locational coordinates.
                                          24-1

-------
                                       Figure 24-1. Bountiful, Utah (BTUT) Monitoring Site
                                                          Tl
to

-------
Figure 24-2. NEI Point Sources Located Within 10 Miles of BTUT
         112'IQ'CrW       112'5'trW
                                                trssxrw       nt-so'trw       nv^xrw
                                  111 55'fJ-W       lirSCNTW       111'45'CrW       111'40'Q"W
                                        Note: Due to facility density and collocation, the total facilities
                                        displayed may not represent all facilities within the area of interest.
               BTUT NATTS site   O   10 mile radius |     |  County boundary
                  Source Category Group (No.  of Facilities)
                         Air-conditioning/Refrigeration  (1 )
                    -f    Aircraft Operations (8)
                    *    Electricity Generation via Combustion (2)
                    ฎ   Institutional - school (1 )
                    ?    Miscellaneous Commercial/Industrial (2)
                    M    Miscellaneous Manufacturing  (1)
                    ^    Petroleum Refinery (5)
                                       24-3

-------
                                  Table 24-1. Geographical Information for the Utah Monitoring Site

Site
Code
BTUT

AQS Code
49-011-0004

Location
Bountiful

County
Davis
Micro- or
Metropolitan
Statistical Area
Ogden-Clearfield,
UTMSA
Latitude
and
Longitude
40.902967,
-111.884467

Land Use
Residential

Location
Setting
Suburban

Additional Ambient Monitoring Information1
SO2, NO, NO2, NOX, PAMS, O3, Meteorological
parameters, PM10, PM2 5, and PM2 5 Speciation.
  BOLD ITALICS = EPA-designated NATTS Site
to
-k

-------
       Bountiful is north of Salt Lake City, and is situated in a valley between the Great Salt
Lake to the west and the Wasatch Mountains to the east. Figure 24-1 shows that BTUT is located
on the property of Viewmont High School, in a primarily residential area. The site is located
about one-third of a mile from 1-15, which runs north-south through most of the surrounding
urban area including Salt Lake City, Clearfield, and Ogden. Figure 24-2 shows that all of the
point sources near BTUT are located to the south of the site. The facilities surrounding BTUT
are involved in a variety of industries, although the source categories with the greatest number of
point sources surrounding BTUT are aircraft operations, which include airports as well as small
runways, heliports,  or landing pads, and petroleum refineries. The source closest to BTUT
generates electricity via combustion.

       Table 24-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Utah monitoring site. Table 24-2 includes county-level population
and vehicle registration information. Table 24-2 also includes a county-level vehicle registration-
to-population ratio,  which was calculated to represent the number of vehicles per person within
the monitoring site's residing county. In addition, the population within 10 miles of the site is
presented, based on postal code population data estimates. An estimate of 10-mile vehicle
ownership was determined by applying the county-level vehicle registration-to-population ratio
to the 10-mile population surrounding the monitoring site. Table 24-2 also contains traffic
volume information for BTUT. Finally, Table 24-2 presents the county-level daily VMT for
Davis County.
 Table 24-2. Population, Motor Vehicle, and Traffic Information for the Utah Monitoring
                                            Site



Site
BTUT

Estimated
County
Population1
311,811

County-level
Vehicle
Registration2
239,582
Vehicles per
Person
(Registration:
Population)
0.77

Population
within 10
miles3
268,749
Estimated
10-mile
Vehicle
Ownership
206,495
Annual
Average
Daily
Traffic4
113,955
County-
level
Daily
VMT5
6,866,779
  Bounty-level population estimate reflects 2011 data from the U.S. Census Bureau (Census Bureau, 2012b)
  2County-level vehicle registration reflects 2011 data from the Utah Tax Commission (UT TC, 2011)
  310-mile population estimate reflects 2011 data from Xionetic (Xionetic, 2011)
  4AADT reflects 2010 data from the Utah DOT (UT DOT, 2010)
  5County-level VMT reflects 2011 data from the Utah DOT (UT DOT, 2012)
  BOLD ITALICS = EPA-designated NATTS Site
                                           24-5

-------
       Observations from Table 24-2 include the following:
       •  Davis County's population is in the mid-to-low end of the range, as is its 10-mile
          population, compared to counties with NMP sites. The county-level vehicle
          registration and 10-mile ownership estimate rankings are similar to the population
          rankings.
       •  The vehicle-per-person ratio (0.78) is in the bottom third of the range compared to
          other NMP sites.
       •  The traffic volume experienced near BTUT is in the top third compared to other NMP
          monitoring sites. The traffic estimate provided is for the intersection of 1-15 with
          US-89, just west of the site.
       •  The daily VMT for Davis County is  on the mid-to-low end compared to counties with
          NMP sites (where VMT was available).

24.2   Meteorological Characterization
       The following sections characterize the meteorological  conditions near the monitoring
site in Utah on sample days, as well as over the  course of the year.
24.2.1  Climate Summary
       The Salt Lake City area's climate can be described as semi-arid and continental with
considerable seasonal variations. Summers are hot and dry while winters are cold and snow is
common. The area is generally dry, with spring as the wettest season, and sunshine prevails
across the area during much of the year. Precipitation that does fall can be enhanced over the
eastern parts of the valley as storm systems move up the side of the Wasatch Mountains, located
to the east. Surrounding mountains protect the valley from winter storm systems moving in from
the southwest or north,  preventing cold air outbreaks. The Great Salt Lake tends to have a
moderating influence on the area's temperature. Moderate winds flow out of the southeast on
average, although there is a valley breeze/lake breeze system that affects the area. High pressure
systems that occasionally settle over the area can result in stagnation episodes (Bair, 1992 and
WRCC, 2013).
                                          24-6

-------
24.2.2 Meteorological Conditions in 2011
       Hourly meteorological data from the NWS weather station nearest BTUT were retrieved
for 2011 (NCDC, 2011). The closest weather station is located at Salt Lake City International
Airport (WBAN 24127). Additional information about the Salt Lake City International Airport
weather station, such as the distance between the site and the weather station, is provided in
Table 24-3. These data were used to determine how meteorological  conditions on sample days
vary from conditions experienced throughout the year.

       Table 24-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and  average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for the entire year. Also included in Table 24-3 is the 95 percent
confidence interval for each parameter. As shown in Table 24-3, average meteorological
conditions on sample days appear cooler than average weather conditions experienced
throughout the year, although the differences are not statistically significant.  This is likely due to
a number of make-up samples collected in November and December of 2011.

24.2.3 Back Trajectory Analysis
       Figure 24-3 is the composite back trajectory map for days on which samples were
collected at the BTUT monitoring site in 2011. Included in Figure 24-3 are four back trajectories
per sample day. Figure 24-4 is the corresponding cluster analysis. An in-depth description of
these maps and how they were generated is presented in Section 3.5.2.1. For the composite map,
each line represents the 24-hour trajectory along which  a parcel of air traveled toward the
monitoring site on a given sample day and time, based on an initial height of 50 meters AGL. For
the cluster analysis, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric circle around the site in Figures 24-3 and 24-4 represents 100 miles.
                                          24-7

-------
to
J^.

oo
                         Table 24-3. Average Meteorological Conditions near the Utah Monitoring Site
Closest NWS Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(ฐF)
Average
Temperature
(ฐF)
Average
Dew Point
Temperature
(ฐF)
Average
Wet Bulb
Temperature
(ฐF)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Bountiful, Utah - BTUT
Salt Lake City
International
24127
(40.79, -111.97)
8.98
miles
217ฐ
(SW)
Sample
Day
2011
58.8
ฑ5.1
61.4
+ 2.1
49.2
ฑ4.5
51.6
+ 1.9
31.7
ฑ2.9
33.2
+ 1.2
40.6
ฑ3.3
42.3
+ 1.4
57.2
ฑ4.3
55.5
+ 1.8
1016.1
ฑ2.1
1015.6
+ 0.9
6.6
ฑ0.8
6.7
+ 0.3
Sample day averages are highlighted to help differentiate the sample day averages from the full-year averages.

-------
Figure 24-3. 2011 Composite Back Trajectory Map for BTUT
    Figure 24-4. Back Trajectory Cluster Map for BTUT
                          24-9

-------
       Observations from Figures 24-3 and 24-4 include the following:

       •  Back trajectories originated from a variety of directions at BTUT. Back trajectories
          originating from a direction with a westerly component tended to be longer than those
          originating from a direction with an easterly component.

       •  Similar to other sites located in the inter-mountain west, the 24-hour air shed domain
          for BTUT is smaller in size than many other NMP monitoring sites. The farthest
          away a back trajectory originated was over the Mojave Desert, or just less than
          500 miles away. However,  the average trajectory length was 176 miles and nearly
          87 percent of back trajectories originated within 300 miles of the site.

       •  The cluster analysis shows  that nearly one-third of back trajectories are represented
          by the  short cluster trajectory originating just southwest of the site. This cluster
          represents back trajectories originating within roughly 150 miles of BTUT and over
          the northern half of Utah. Thirty percent of back trajectories originated to the south of
          BTUT, although of varying distances, as indicated by the shorter cluster trajectory
          (23 percent), which represents back trajectories originating over the southern half of
          Utah, and the longer cluster trajectory (7 percent), which represents longer back
          trajectories originating primarily over northern Arizona. Back trajectories also
          originated to the west, northwest, and north of BTUT, and are also represented by two
          cluster trajectories; one representing shorter back trajectories  originating primarily
          over southern Idaho, and one representing  longer back trajectories originating to the
          west and northwest of BTUT over northeast Nevada and southwest Idaho.


24.2.4 Wind Rose Comparison

       Hourly surface wind data from the NWS weather station at Salt Lake City International

Airport near BTUT were uploaded into a wind rose software program to  produce customized

wind roses, as described in Section 3.5.2.2. A wind rose shows the frequency of wind directions

using "petals" positioned around a 16-point compass,  and uses different colors to represent wind

speeds.
       Figure 24-5 presents a map showing the distance between the NWS station and BTUT,

which may be useful for identifying topographical influences that may affect the meteorological

patterns experienced at this location. Figure 24-5 also presents three different wind roses for the

BTUT monitoring site. First, a historical wind rose representing 2001 to 2010 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 2011 is presented.

Next, a wind rose representing wind data for days on which samples were collected in 2011 is

presented. These can be used to identify the predominant wind speed and  direction for 2011 and
                                          24-10

-------
to determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.

       Observations from Figure 24-5 for BTUT include the following:
       •  The Salt Lake City International Airport weather station is located approximately
          9 miles southwest of BTUT.
       •  The historical wind rose shows that southeasterly, south-southeasterly, and southerly
          winds were prevalent near BTUT, accounting for approximately 40 percent of the
          wind observations. Winds from the north-northwest and north were also common.
          Calm winds (<2 knots) were observed for approximately 11  percent of the hourly
          measurements from 2001-2010. The strongest wind speeds were observed with south-
          southeasterly and southerly winds.
       •  The wind patterns shown on the 2011 wind rose are similar to the historical wind
          patterns, indicating that wind conditions in 2011 were similar to wind conditions
          experienced historically near BTUT.
       •  The wind patterns shown on the sample day wind rose resemble the 2011 wind
          patterns, indicating that conditions on sample days were representative of those
          experienced over the entire year (and historically).
jjcuutina, inui^aung uiiau ^uiiuiuuna un aaiiipic uay
experienced over the entire year (and historically)
24.3   Pollutants of Interest
       Site-specific "pollutants of interest" were determined for the BTUT monitoring site in
order to allow 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." 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. In addition, if any
of the NATTS MQO Core Analytes measured by the BTUT monitoring site did not meet the
pollutant of interest criteria based  on the preliminary risk-based screening, that pollutant was
added to the list of site-specific pollutants of interest. A more in-depth description of the risk-
based screening process is presented in Section 3.2.
                                          24-11

-------
Figure 24-5. Wind Roses for the Salt Lake City International Airport Weather Station near

                                      BTUT
  Distance between BTUT and NWS Station
2001-2010 Historical Wind Rose
       rj
        Stall
                                i :
            2011 Wind Rose
    Sample Day Wind Rose
                                       24-12

-------
       Table 24-4 presents the results of the preliminary risk-based screening process for BTUT.
The pollutants that failed at least one screen and contributed to 95 percent of the total failed
screens for the monitoring site are shaded. NATTS MQO Core Analytes are bolded. Thus,
pollutants of interest are shaded and/or bolded. BTUT sampled for VOCs, carbonyl compounds,
SNMOCs, PAHs, metals (PMio), and hexavalent chromium and is one of only two sites sampling
the entire suite of pollutants under the NMP (NBIL is the other).

          Table 24-4. Risk-Based Screening Results for the Utah Monitoring Site
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Bountiful, Utah - BTUT
Acet aldehyde
Benzene
Carbon Tetrachloride
Formaldehyde
1,3-Butadiene
Arsenic (PM10)
Manganese (PM10)
Naphthalene
Ethylbenzene
1 ,2-Dichloroethane
Nickel (PM10)
Dichloromethane
Acrylonitrile
/>-Dichlorobenzene
Hexachloro- 1 , 3 -butadiene
Propionaldehyde
Cadmium (PM10)
1 , 1 ,2,2-Tetrachloroethane
Hexavalent Chromium
Lead (PM10)
Xylenes
0.45
0.13
0.17
0.077
0.03
0.00023
0.005
0.029
0.4
0.038
0.0021
7.7
0.015
0.091
0.045
0.8
0.00056
0.017
0.000083
0.015
10
Total
60
60
60
60
48
43
38
38
25
15
14
13
7
6
6
6
4
4
1
1
1
510
60
60
60
60
49
58
60
62
60
15
60
60
7
29
6
60
60
4
53
60
60
1,003
100.00
100.00
100.00
100.00
97.96
74.14
63.33
61.29
41.67
100.00
23.33
21.67
100.00
20.69
100.00
10.00
6.67
100.00
1.89
1.67
1.67
50.85
11.76
11.76
11.76
11.76
9.41
8.43
7.45
7.45
4.90
2.94
2.75
2.55
1.37
1.18
1.18
1.18
0.78
0.78
0.20
0.20
0.20
11.76
23.53
35.29
47.06
56.47
64.90
72.35
79.80
84.71
87.65
90.39
92.94
94.31
95.49
96.67
97.84
98.63
99.41
99.61
99.80
100.00

       Observations from Table 24-4 include the following:
       •   Twenty-one pollutants, of which 12 are NATTS MQO Core Analytes, failed at least
          one screen for BTUT.
       •   The risk-based screening process identified 16 pollutants of interest for BTUT, of
          which nine are NATTS MQO Core Analytes. Three pollutants (cadmium, hexavalent
          chromium, and lead) were added to BTUT's pollutants of interest because they are
                                        24-13

-------
          NATTS MQO Core Analytes, even though they did not contribute to 95 percent of
          the total failed screens. Six additional pollutants were added to BTUT's pollutants of
          interest because they are also NATTS MQO Core Analytes, even though they did not
          fail any screens: beryllium, benzo(a)pyrene, chloroform, tetrachloroethylene,
          trichloroethylene, and vinyl chloride. These six pollutants are not shown in
          Table 24-4 but are shown in subsequent tables in the sections that follow.
       •  Nearly 50 percent of measured detections failed screens (of the pollutants that failed
          at least one screen) for BTUT.
       •  Acetaldehyde, benzene, carbon tetrachloride, and formaldehyde were detected in
          every valid carbonyl compound and VOC sample collected at BTUT and failed
          100 percent of screens. Other pollutants also failed 100 percent of screens but were
          detected much less frequently.
       •  Recall from Section 3.2 that if a pollutant was measured by both the TO-15 and
          SNMOC methods at the same site, the TO-15 results were used for the risk-based
          screening process. As BTUT sampled both VOCs (TO-15) and SNMOCs, the TO-15
          results were used for the 12 pollutants these methods have in common.

24.4   Concentrations
       This  section presents various concentration averages used to characterize pollution levels
at the Utah monitoring site. Where applicable, the following calculations and data analyses were
performed: Time period-based concentration averages (quarterly and annual) are provided for the
pollutants of interest for BTUT, where the data meet the applicable criteria. Concentration
averages for select pollutants are also presented graphically for this site to illustrate how the
site's concentrations compare to the program-level averages, as presented in Section 4.1. In
addition, concentration averages for select pollutants are presented from previous years of
sampling in order to characterize concentration trends  at the site. Additional  site-specific
statistical summaries for BTUT are provided in Appendix J through Appendix O.

24.4.1  2011 Concentration Averages
       Quarterly and annual concentration averages were calculated for the pollutants of interest
for BTUT, as described in Section 3.1. The quarterly average of a particular pollutant is simply
the average concentration of the preprocessed daily measurements over a given calendar quarter.
Quarterly average concentrations include the substitution of zeros for all non-detects. A site must
have a minimum of 75 percent valid  samples of 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
                                         24-14

-------
were calculated for pollutants where three valid quarterly averages could be calculated and
where method completeness was greater than or equal to 85 percent, as presented in Section 2.4.
Quarterly and annual average concentrations for the Utah monitoring site are presented in
Table 24-5, where applicable. Note that concentrations of the PAHs, metals, and hexavalent
chromium are presented in ng/m3 for ease of viewing. Also note that if a pollutant was not
detected in a given calendar quarter, the quarterly average simply reflects "0" because only zeros
substituted for non-detects were factored into the quarterly average concentration.
 Table 24-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
                                 the Utah Monitoring Site
Pollutant
# of Measured
Detections vs.
# of Samples
1st
Quarter
Average
(Hg/m3)
2nd
Quarter
Average
(jig/m3)
3rd
Quarter
Average
(jig/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(jig/m3)
Bountiful, Utah - BTUT
Acetaldehyde
Acrylonitrile
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
ฃ>-Dichlorobenzene
1 ,2-Dichloroethane
Dichloromethane
Ethylbenzene
Formaldehyde
Hexachloro- 1 ,3 -butadiene
Propionaldehyde
Tetrachloroethylene
Trichloroethylene
60/60
7/60
60/60
49/60
60/60
29/60
29/60
15/60
60/60
60/60
60/60
6/60
60/60
41/60
11/60
3.30
ฑ 1.06
NA
NA
NA
NA
NA
NA
NA
NA
NA
8.37
ฑ3.63
NA
0.67
ฑ0.20
NA
NA
1.33
ฑ0.28
0.03
ฑ0.04
0.79
ฑ0.14
0.04
ฑ0.03
0.60
ฑ0.06
0.03
ฑ0.03
0.01
ฑ0.02
0.02
ฑ0.02
6.18
ฑ3.94
0.27
ฑ0.05
3.01
ฑ1.10
0.02
ฑ0.03
0.32
ฑ0.06
0.05
ฑ0.03
0
2.12
ฑ0.50
0.02
ฑ0.03
1.03
ฑ0.22
0.05
ฑ0.02
0.64
ฑ0.07
0.03
ฑ0.03
0.02
ฑ0.03
0.01
ฑ0.02
96.53
ฑ 117.86
0.37
ฑ0.05
4.42
ฑ1.56
0
0.47
ฑ0.09
0.06
ฑ0.03
0
1.94
ฑ0.43
0
1.44
ฑ0.30
0.17
ฑ0.04
0.64
ฑ0.07
0.10
ฑ0.02
0.05
ฑ0.01
0.05
ฑ0.02
1.92
ฑ1.84
0.55
ฑ0.11
2.13
ฑ0.23
0.01
ฑ0.01
0.36
ฑ0.08
0.14
ฑ0.04
0.04
ฑ0.02
2.19
ฑ0.35
0.02
ฑ0.01
1.14
ฑ0.15
0.10
ฑ0.02
0.61
ฑ0.03
0.06
ฑ0.02
0.05
ฑ0.02
0.03
ฑ0.01
53.90
ฑ 50.64
0.47
ฑ0.13
4.49
ฑ1.15
0.01
ฑ0.01
0.46
ฑ0.07
0.10
ฑ0.02
0.02
ฑ0.01
 a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
 NA = Not available due to the criteria for calculating a quarterly and/or annual average
                                           24-15

-------
Table 24-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
                         the Utah Monitoring Site (Continued)
Pollutant
Vinyl Chloride
Arsenic (PM10)a
Benzo(a)pyrene a
Bery Ilium (PM10)a
Cadmium (PM10)a
Hexavalent Chromium3
Lead(PM10)a
Manganese (PM10) a
Naphthalene a
Nickel (PM10)a
# of Measured
Detections vs.
# of Samples
1/60
58/60
20/62
53/60
60/60
53/61
60/60
60/60
62/62
60/60
1st
Quarter
Average
(Hg/m3)
NA
0.71
ฑ0.62
0.06
ฑ0.06
0.01
ฑ0.01
0.15
ฑ0.13
0.02
ฑ0.01
4.14
ฑ3.27
4.76
ฑ1.88
54.33
ฑ21.23
1.98
ฑ0.81
2nd
Quarter
Average
(jig/m3)
0
0.36
ฑ0.20
0
0.01
ฑ0.01
0.07
ฑ0.03
0.02
ฑ0.01
2.26
ฑ 1.25
6.39
ฑ3.04
25.28
ฑ5.96
0.98
ฑ0.23
3rd
Quarter
Average
(jig/m3)
0
0.43
ฑ0.09
0.02
ฑ0.01
0.02
ฑ0.01
0.09
ฑ0.02
0.02
ฑ0.02
2.62
ฑ0.58
10.17
ฑ1.90
46.19
ฑ9.36
1.80
ฑ0.74
4th
Quarter
Average
(Ug/m3)
0
0.87
ฑ0.50
0.10
ฑ0.05
0.01
ฑ0.01
0.27
ฑ0.15
0.02
ฑ0.01
5.15
ฑ2.16
8.64
ฑ2.80
62.83
ฑ18.12
2.19
ฑ0.83
Annual
Average
(jig/m3)
O.01
ฑ<0.01
0.59
ฑ0.19
0.04
ฑ0.02
0.01
ฑ0.01
0.14
ฑ0.05
0.02
ฑ0.01
3.51
ฑ0.98
7.58
ฑ1.27
47.39
ฑ7.88
1.73
ฑ0.35
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
NA = Not available due to the criteria for calculating a quarterly and/or annual average
      Observations for BTUT from Table 24-5 include the following:

      •   The pollutants with the highest annual average concentrations by mass are
          dichloromethane, formaldehyde, acetaldehyde, and benzene, consistent with the last
          few years of sampling. The annual average for dichloromethane is significantly
          higher than the annual averages of the other pollutants.

      •   Dichloromethane has the highest annual average for BTUT, but also has a very large
          confidence interval associated it, as do the quarterly averages. This indicates the
          likely presence of outliers. The concentrations of dichloromethane at BTUT range
          from 0.313 |ig/m3to 1,214 |ig/m3. One measurement of this pollutant is greater than
          1,000 |ig/m3, three are greater than 500 |ig/m3, six are greater than 75.0 |ig/m3, and
          12 are greater than 10 |ig/m3. The top six measurements are the six highest
          concentrations of dichloromethane measured across the program. However, the
          median concentration of dichloromethane for BTUT is 0.919 |ig/m3, as over half of
          the measurements are less than 1 |ig/m3.

      •   There are no first quarter averages for the VOCs because sampler issues during this
          time resulted in canisters with pressures outside of the tolerance limits.

      •   The first quarter average concentrations of the three carbonyl compound pollutants of
          interest are higher than the other quarterly averages and have relatively large
                                         24-16

-------
   confidence intervals. A review of the data shows that the highest concentrations of all
   three pollutants were measured in February and March. The maximum concentration
   of formaldehyde (19.9 |ig/m3) was measured at BTUT on March 4, 2011, which is the
   day the second highest concentrations of acetaldehyde and propionaldehyde were
   measured. This formaldehyde concentration is the second highest formaldehyde
   concentration measured across the program. The maximum concentrations of
   acetaldehyde (7.06 |ig/m3) and propionaldehyde (1.44 |ig/m3) were measured at
   BTUT on February 26,  2011, which is the day the second highest concentration of
   formaldehyde was measured. This propionaldehyde concentration is the highest
   propionaldehyde concentration measured across the program (although a
   concentration of similar magnitude was also measured at TOOK).

•  In addition to the first quarter, the second and third quarter average concentrations of
   formaldehyde also have relatively large confidence intervals associated with them.
   BTUT is the site with the highest number of formaldehyde measurements greater than
   5 |ig/m3 (15), which were measured in February (5), March (3), June (3), and July (4).

•  Many of the VOC pollutants of interest are highest during the fourth quarter of 2011,
   as illustrated by the quarterly averages. The highest concentrations of these pollutants
   were measured  in January, November, or December, or were detected most frequently
   during these months. For example, 1,2-dichloroethane was detected 11 times (out of a
   total of 15) during the fourth quarter and trichloroethylene was not detected outside
   the first or fourth quarters. Of the eight benzene concentrations greater than 2 |ig/m3,
   all but one was  measured in January, November, or December. Similarly, of the
   15 ethylbenzene concentrations greater than 0.5 |ig/m3, all but one was measured in
   the first or fourth quarter of 2011. The 19 highest concentrations of 1,3-butadiene
   were also measured in the first or fourth quarters of 2011.

•  Concentrations  of naphthalene and benzo(a)pyrene appear highest in the colder
   months of the year. The ten  highest concentrations of naphthalene (those greater  than
   90 ng/m3) were all measured in the first and fourth quarters of 2011. Although
   benzo(a)pyrene was detected in only 20  samples, 14 of these were measured during
   the first and fourth quarters  of 2011, with the highest concentrations measured during
   these quarters.

•  This trend continues for some of the metals. The maximum concentrations of arsenic,
   cadmium, and lead were all  measured on January 15, 2011, with additional "higher"
   concentrations measured in November and December. The arsenic concentration
   measured on this date for BTUT is the maximum arsenic concentration measured
   among sites sampling metals. A  similar trend was noted in the 2010 NMP report. This
   is not true for manganese, beryllium, or nickel.

•  Concentrations  of hexavalent chromium ranged from 0.0055 ng/m3 to 0.13 ng/m3.
   The maximum concentration of hexavalent chromium was measured at BTUT on
   July 2, 2011, and is twice the next highest concentration (0.06 ng/m3, measured on
   January 21, 2011).
                                  24-17

-------
       Tables 4-9 through 4-12 present the sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for BTUT from
those tables include the following:
       •   BTUT appears in Table 4-9 through 4-12 a total of 13 times for the program-lev el
          pollutants of interest.
       •   BTUT does not rank higher than eighth among the program-level VOC pollutants of
          interest shown in Table 4-9.
       •   BTUT has the highest annual average concentration of formaldehyde and tenth
          highest annual average concentration of acetaldehyde among NMP sites sampling
          carbonyl compounds, as shown in Table 4-10.
       •   BTUT does not appear in Table 4-11 for PAHs. This site's annual average
          concentrations of the PAHs are among the lower averages for sites sampling PAHs.
       •   BTUT ranks in the top five for each of the PMio metals shown in Table 4-12. It is
          important, however, to note that only nine sites sampled PMio metals. BTUT does not
          appear in Table 4-12 for hexavalent chromium.

24.4.2  Concentration Comparison
       In order to better illustrate how a site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for the selected NATTS MQO Core
Analytes listed in Section 3.5.3, where applicable. Thus, box plots for acetaldehyde, arsenic,
benzene, benzo(a)pyrene, 1,3-butadiene, formaldehyde, hexavalent chromium, lead, manganese,
and naphthalene were created for BTUT. Figures 24-6 through 24-15 overlay the site's
minimum, annual average, and maximum concentrations onto the program-level minimum, first
quartile, median, average, third quartile, and maximum concentrations, as described in
Section 3.5.3.
       Figure 24-6. Program vs. Site-Specific Average Acetaldehyde Concentration
Program:
Site:

1st Quartile
Site Average
0
2nd Quartile
3rd Quartile
n
4th Quartile Average
n
^m i 	 i
Site Minimum/Maximum




                                        24-18

-------
Figure 24-7. Program vs. Site-Specific Average Arsenic (PMi0) Concentration
       0.5
                            1.5
                                       2         2.5
                                       Concentration (ng/mi)
                                                                      3.5
Program
Site:
: IstQuartile
Site Average
o
2ndQuartile 3rdQuartile 4thQuartile Ave
• n
Site Minimum/Maximum
'rage

                                                                                           4.5
    Figure 24-8. Program vs. Site-Specific Average Benzene Concentration
                                                            Program Max Concentration = 23.8 ug/n
                                   456
                                       Concentration (pg/mi)
                                                                                           10
          Program:   IstQuartile     2ndQuartile     SrdQuartile      4thQuartile     Average
          Site:      Site Average     Site Minimum/Maximum

Figure 24-9. Program vs. Site-Specific Average Benzo(a)pyrene Concentration

                                0.75           1           1.25
                                       Concentration (ng/mj)
           Program:   IstQuartile     2ndQuartile     3rdQuartile      4thQuartile     Average
          Site:      Site Average     Site Minimum/Maximum

                                        24-19

-------
       Figure 24-10. Program vs. Site-Specific Average 1,3-Butadiene Concentration
B~U~
E
                                                                    Program Max Concentration = 9.51 ng/m3
                   0.5
                                                   1.5
                                              Concentration (jig/mi)
                                                                                   2.5
                 Program:    IstQuartile      2ndQuartile      SrdQuartile     4thQuartile     Average
                 Site:
                          Site Average     Site Minimum/Maximum
                             o
      Figure 24-11. Program vs. Site-Specific Average Formaldehyde Concentration
                                   10
                                                   15
                                              Concentration [[Jg/m3)
                                                                                   IS
                 Program:   IstQuartile      2ndQuartile      SrdQuartile      4thQuartile     Average
                 Site:
                          Site Average     Site Minimum/Maximum
                             o
                                                                                                   30
  Figure 24-12. Program vs. Site-Specific Average Hexavalent Chromium Concentration
    K
                   0.05
                                   0.1
                                                   0.15
                                              Concentration (ng/m3)
                                                                                  O.Z5
                 Program:   IstQuartile      2ndQuartile      SrdQuartile      4thQuartile     Average
                 Site:      Site Average     Site Minimum/Maximum
                             o
                                                24-20

-------
       Figure 24-13. Program vs. Site-Specific Average Lead (PMi0) Concentration
trir
                            10
                                         15           20
                                          Concentration (ng/mi)
                                                                 25
                                                                                          35
               Program:   IstQuartile     2ndQuartile     SrdQuartile      4thQuartile     Average
               Site:      Site Average    Site Minimum/Maximum

    Figure 24-14. Program vs. Site-Specific Average Manganese (PMi0) Concentration
E
                                                          ; Program Max Concentration = 395 ng/m3 |
                                    75
                                              100
                                          Concentration (ng/m3)
                                                         125
                                                                               175
               Program:   IstQuartile     2ndQuartile     SrdQuartile      4thQuartile     Average
               Site:      Site Average    Site Minimum/Maximum

                                                                                          sec-
              24-15. Program vs. Site-Specific Average Naphthalene Concentration
trir

                                                             Program Max Concentration =779 ng/m3
            5:
                    100
                             150
                                      200       250       300
                                          Concentration (ng/m3)
                                                                350
                                                                        400
                                                                                 453
                                                                                          sec-
Program
Site:
: IstQuartile
Site Average
o
2ndQuartile SrdQuartile 4thQuartile Ave
Site Minimum/Maximum
rage

      Observations from Figures 24-6 through 24-14 include the following:

          •  Figure 24-6 shows that the annual average acetaldehyde concentration for BTUT
             is just greater than the program-level average concentration. The maximum
             concentration measured at BTUT is less than the maximum acetaldehyde
             concentration measured at the program-level. There were no non-detects of
             acetaldehyde measured at BTUT or across the program.
                                           24-21

-------
•  Figure 24-7 shows that the maximum concentration of arsenic (PMio) across the
   program was measured at BTUT. Yet, the annual average arsenic concentration is
   equivalent to the program-level average arsenic concentration. Two non-detects of
   arsenic were measured at BTUT.

•  Figure 24-8 is the box plot for benzene. Note that the program-level maximum
   benzene concentration (23.8 |ig/m3) is not shown directly on the box plot because
   the scale of the box plot would be too large to readily observe data points at the
   lower end of the concentration range. Thus, the scale has been reduced to
   10 |ig/m3. Figure 24-8 shows that the annual average concentration for BTUT is
   greater than the program-level average concentration and third quartile
   (75th percentile). The maximum concentration of benzene measured at BTUT is
   considerably less than the maximum concentration measured across the program.
   There were no non-detects of benzene measured at BTUT or across the program.

•  Figure 24-9 is the box plot for benzo(a)pyrene. Note that the program-level first
   quartile for this pollutant is zero and is not visible on this box plot. The box plot
   shows that the annual average concentration for BTUT is less than the program-
   level average concentration but greater than the program-level median
   concentration. Figure 24-9 also shows that the maximum concentration measured
   at BTUT is considerably less than the maximum concentration measured across
   the program. A number of non-detects of benzo(a)pyrene were measured at
   BTUT.

•  Similar to the box plot for benzene, the program-level maximum 1,3-butadiene
   concentration (9.51 |ig/m3)  is not shown directly on the box plot in Figure 24-10;
   thus, the scale has been reduced to 3 |ig/m3 in order to allow for the observation of
   data points at the lower end of the concentration range. Figure 24-10 for
   1,3-butadiene shows that the annual average concentration for BTUT is similar to
   the program-level average concentration and that both are just less than the
   program-level third quartile. The maximum concentration of 1,3-butadiene
   measured at BTUT is considerably  less than the maximum concentration
   measured across the program. Several non-detects of 1,3-butadiene were
   measured at BTUT.

•  Figure 24-11 shows that BTUT's annual average formaldehyde concentration is
   greater than the program-level average and greater than the program-level third
   quartile. Although the maximum concentration of formaldehyde measured at
   BTUT is not the maximum  measured across the program, it is the second highest
   formaldehyde concentration measured among NMP sites sampling carbonyl
   compounds. There were no non-detects of formaldehyde measured at BTUT or
   across the program.

•  Figure 24-12 is the box plot for hexavalent chromium. The annual average
   concentration for BTUT is less than the program-level average but greater than
   the program-level median concentration. The maximum concentration measured
   at BTUT is less than the program-level maximum concentration. There were a
   few non-detects of hexavalent chromium measured at BTUT.

                              24-22

-------
          •  Figure 24-13 shows that the annual average concentration of lead (PMio) for
             BTUT is less than the program-level average concentration. Although the
             maximum concentration measured at BTUT is less than the program-level
             maximum concentration, it is the third highest lead concentration measured
             among NMP sites sampling metals. There were no non-detects of lead measured
             at BTUT or across the program.

          •  Figure 24-14 is the box plot for manganese (PMio). The program-level maximum
             manganese concentration (395  ng/m3) is not shown directly on the box plot as the
             scale has been reduced to 200 ng/m3  in order to allow for the observation of data
             points at the lower end of the concentration range. Figure 24-14  shows that the
             annual average concentration of manganese (PMio) for BTUT is less than the
             program-level average concentration. The maximum concentration measured at
             BTUT is considerably less than the program-level maximum concentration.
             Although difficult to discern in Figure 24-14, there were no non-detects  of
             manganese measured at BTUT.

          •  Figure 24-15 is the box plot for naphthalene. The program-level maximum
             naphthalene concentration (779 ng/m3) is not shown directly on the box  plot in
             Figure 24-15 as the scale has been reduced to 500 ng/m3 in order to allow for the
             observation of data points at the lower end of the concentration range.
             Figure 24-15 shows that the annual average naphthalene concentration for BTUT
             is less than both the program-level average and median concentrations. Of the
             23 sites  sampling PAHs, the annual average concentration of naphthalene for
             BTUT ranks 19th. The maximum naphthalene concentration measured at BTUT is
             considerably less than the program-level maximum concentration. There were no
             non-detects of naphthalene measured at BTUT or across the program.


24.4.3  Concentration Trends

       A site-specific trends evaluation was completed for sites that have sampled one  or more

of the selected NATTS MQO Core Analytes for 5 consecutive years or longer, as described in

Section 3.5.4. BTUT has  sampled carbonyl  compounds, VOCs, metals, and SNMOCs as part of

the NMP since 2003. BTUT has also sampled hexavalent chromium since 2005. Thus,

Figures 24-16 through 24-23 present the annual statistical metrics for acetaldehyde,  arsenic,

benzene, 1,3-butadiene, formaldehyde, hexavalent chromium, lead, and manganese for

BTUT, respectively. The statistical metrics presented for assessing trends include the substitution

of zeros for non-detects. Sampling for PAHs at BTUT did not begin until 2008; thus, a  trends

analysis was not conducted for the PAHs because this method does not meet the 5 consecutive

year criteria.
                                        24-23

-------
 Figure 24-16. Annual Statistical Metrics for Acetaldehyde Concentrations
                            Measured at BTUT
        2004      2005       2006       2007       2008      2009      2010       2011
                                      Year
         #  5th Percentile    — Minimum    — Median     —  Maximum     •  95th Percentile   .. + .. Average
Figure 24-17. Annual Statistical Metrics for Arsenic (PMio) Concentrations
                            Measured at BTUT




1
|l2
B
a
i
3
1



o -













The maximum arsenic j
concentration for




• •


*..
•
••


" i i A

2004 2005 2006 2007 2008 2009 2010 2011
Year
• 5th Percentile - Minimum — Median — Maximum • 95th Percentile ...+*. Average
                                   24-24

-------
Figure 24-18. Annual Statistical Metrics for Benzene Concentrations Measured at BTUT




'm
.a b
1
i
I 4 -
3 4
1
la


i -



-

i
— ^ ,— 1-,






r
I
,_







E r
^^ [

*• 	 ^
""^** 	 * 	 — *" ^^ 	 •*•• 	 A

-•- f — r-
™— — — — ^ta—
| 1 L_^ , — t- z E
2004 2005 2006 2007 2008 2009 2010 2011
Year
• 5th Percentile — Minimum — Median - Maximum • 95th Percentile +*^.. Average
       Figure 24-19. Annual Statistical Metrics for 1,3-Butadiene Concentrations
                               Measured at BTUT




tration (ME/i
3 i
1
3
EO 3 -
•S

o -




'








*"
^^
2004


.4 •*'
— (


..
T
3 —


T
" r~^

	 *•• 	 	 t , 	 ป-• 	
^^^ -••+ 	 ^....^ - ••ป 	 •ป
^^^^^ ^^^^i
f • •
2005 2006 2007 2008 2009 2010 2011
Year
• 5th Percentile — Minimum — Median — Max mum • 95th Percentile ..+.. Average
                                     24-25

-------
Figure 24-20. Annual Statistical Metrics for Formaldehyde Concentrations Measured at
                                    BTUT




t
B
a
!25
1
3
8,n -
c 20





5 -
















I
— 1
20








[ r-I-,
1

T


„ 	 at* 	 <*•-...
•-1 L-t-1 • PT5 e^3 P55 ^^ HP
04 2005 2006 2007 2008 2009 2010 2011
Year
• 5th Percentile — Minimum — Median — Maximum • 95th Percentile •••+•• Average
  Figure 24-21. Annual Statistical Metrics for Hexavalent Chromium Concentrations
                              Measured at BTUT
0.5
0.45
0.4
0.35
E
"5 0.3
s
.:
•S 0.25
1
3
|o,
ฃ
<
0.15
0.1
0.05
0










T
[




••
T T
_



^... .-^ .....j.^... ^^.. ..^^ ....^^ 	 ^
2005 2006 2007 2003 2009 2010 2011
Year
• 5th Percentile - Minimum — Median — Maximum • 95th Percentile ..+** Average
                                     24-26

-------
   Figure 24-22. Annual Statistical Metrics for Lead (PMi0) Concentrations
                                Measured at BTUT
   I15
   B
   a
   S

   I
   8
          2004       2005       2006       2007       200B       2009       2010       2011
                                            Year
              5th Percentile     - Minimum     — Median    —  Maximum    •  95th Percentile   "^"Average
Figure 24-23. Annual Statistical Metrics for Manganese (PMio) Concentrations
                                Measured at BTUT
          2004       2005       2006       2007       2008       2009       2010       2011
                                            Year
              5th Percentile    -  Minimum     — Median     -  Maximum    •  95th Percentile   "-4*+Average
                                        24-27

-------
Observations from Figure 24-16 for acetaldehyde measurements include the following:

•  Although sampling for carbonyl compounds at BTUT began in 2003, sampling did
   not begin until July, which does not yield enough samples for the statistical metrics to
   be calculated, based on the criteria specified in Section 3.5.4. Thus, Figure 24-16
   begins with 2004.

•  The maximum acetaldehyde concentration was measured in 2004 (32.7 |ig/m3). The
   second highest concentration of acetaldehyde measured at BTUT is the maximum
   shown for 2008 (20.0 |ig/m3). All acetaldehyde concentrations greater than 8 |ig/m3
   were measured prior to 2009.

•  The average concentration exhibits a steady decreasing trend beginning with 2006
   and continuing through 2009, after which the average concentration held steady
   around 2 |ig/m3.

•  Even with the second highest concentration measured (20.0 |ig/m3), nearly all of the
   statistical metrics exhibit a decrease from 2007 to 2008. The second highest
   concentration measured in 2008 (4.17 |ig/m3) was considerably less than the
   maximum concentration and only five acetaldehyde concentrations greater than
   3 |ig/m3 were measured that year, which is the least of any year of sampling.

•  There have been no non-detects of acetaldehyde measured at BTUT.


Observations from Figure 24-17 for arsenic measurements include the following:

•  Although sampling for PMio metals at BTUT began in 2003, sampling did not begin
   until July, which does not yield enough samples for the statistical metrics to be
   calculated, based on the criteria specified in Section 3.5.4. Thus, Figure 24-17 begins
   with 2004.

•  The maximum arsenic concentration was measured in 2004.  The maximum
   concentration measured (33.0 ng/m3) is nearly twice the next highest concentration
   (16.8 ng/m3), also measured in 2004. The three highest measurements of arsenic were
   all measured at BTUT in 2004; further, eight of the 12 highest concentrations of
   arsenic (those greater than 5 ng/m3) were measured in 2004.  Of these 12, eight were
   measured in the first quarter of the calendar year and four were measured during the
   fourth  quarter of the calendar year, supporting the tendency discussed in
   Section 24.4.1.

•  The average concentration of arsenic decreased significantly from 2004 to 2005, a
   trend that continued into 2006. After 2004, the average arsenic concentration has
   fluctuated between 0.59 ng/m3 (2011) to 1.13 ng/m3 (2009). The statistical parameters
   for 2007 and 2009 are being driven primarily by a single "high" measurement. If
   2007 and 2009 are excluded,  the averages then range from 0.59 ng/m3 to 0.71 ng/m3.
   If the maximum concentrations measured in 2007 and 2009 were removed, the
   average concentrations from 2005 to 2011 would all be less than 1 ng/m3. The
   maximum concentrations  for  2007 and 2009 were both measured in January.
                                  24-28

-------
•  Non-detects of arsenic were not measured for half of the years of sampling. For those
   years with non-detects, the greatest percentage were measured in 2011 (5 percent).


Observations from Figures 24-18 for benzene include the following:

•  Although sampling for VOCs at BTUT began in 2003, sampling did not begin until
   July, which does not yield enough samples for the statistical metrics to be calculated,
   based on the criteria specified in Section 3.5.4. Thus, Figure 24-18 begins with 2004.

•  The maximum concentration of benzene shown was measured in 2009 (8.16 |ig/m3).
   The next highest concentration (6.56 |ig/m3) was also measured in 2009, although
   concentrations greater than 6 |ig/m3 were also measured in 2005 and 2007.

•  The average and median concentrations have a decreasing trend through 2007. An
   increasing trend in the average is then shown through 2009, after which another
   decreasing trend follows. The average benzene concentration ranges from 1.14 |ig/m3
   (2011)tol.87|ig/m3(2004).

•  Although the average concentration increased for 2009, the median concentration
   decreased. This average is being driven by the higher concentrations measured in
   2009, as discussed above.

•  There have been no non-detects of benzene measured at BTUT.
Observations from Figures 24-19 for 1,3-butadiene include the following:

•  The maximum concentration of 1,3-butadiene shown was measured in 2005
   (0.75 |ig/m3). The second highest concentration was also measured in 2005
   (0.53 |ig/m3), although a similar measurement was also collected in 2006. These are
   the only concentrations of 1,3-butadiene greater than 0.5  |ig/m3.

•  The minimum, 5th percentile, and median concentrations  are all zero for the 2004,
   indicating that at least 50 percent of the measurements were non-detects. The
   detection rate of 1,3-butadiene increased after 2004, as indicated by the increase in
   the median concentration and then the 5th percentile. The percentage of non-detects
   decreased from 75 percent for 2004 to 0 percent for 2008 and 2009. The percentage
   of non-detects increased to 7 percent for 2010 and 18 percent for 2011.

•  The average concentration has changed little over the years of sampling and ranges
   from 0.10 |ig/m3 to 0.12  |ig/m3  between 2005 and 2011.
                                  24-29

-------
       Observations from Figure 24-20 for formaldehyde measurements include the following:

       •   The maximum formaldehyde concentration was measured in 2004 (45.4 |ig/m3), on
          the same day as the highest acetaldehyde concentration, August 31, 2004. This
          measurement is more than twice the next highest concentration (19.9 |ig/m3),
          measured in 2011. Concentrations greater than 15 |ig/m3 were measured in 2004,
          2005, 2006, 2007, and 2011.

       •   Although the maximum concentration decreased significantly from 2004 to 2005, the
          other statistical metrics exhibit an increase from 2004 to 2005. The median, or
          midpoint concentration, increased by  2 |ig/m3 from 2004 to 2005, indicating that
          concentrations ran higher in 2005 than 2004 rather than being driven by an outlier, as
          in 2004. As an illustration, there were 11 concentrations greater than 5 |ig/m3 in 2004
          compared to 31 in 2005.

       •   After 2005, the average concentration began to decrease, reaching a minimum for
          2008. A steady increasing trend in the average formaldehyde concentration, as well as
          most other statistical parameters, is shown after 2008.

       •   Although the maximum, 95th percentile, and average concentrations increased for
          2011, the median actually decreased.  Although the concentrations on the upper end of
          the range are higher than they had been in recent years of sampling, the number of
          concentrations on the lower end of the range is also higher for 2011. The number of
          measurements less than 2.5 |ig/m3 for 2011 (30) is the second highest, behind only
          2008 (37).

       •   There have been no non-detects of formaldehyde measured at BTUT.

       Observations from Figure 24-21 for hexavalent chromium measurements include the
following:

       •   The maximum hexavalent chromium  concentration was measured on July 4, 2006
          (0.45 ng/m3). The next highest concentration, measured in 2010, is roughly half as
          high.

       •   The minimum and 5th percentile are both zero for each year of sampling, indicating
          the presence of non-detects. The number of non-detects has varied over the years,
          ranging from eight percent (2006) to 38 percent (2009). The decrease exhibited by
          most of the statistical metrics from 2008 to 2009 may result, at least in part, from the
          increase in non-detects (up from 10 percent 2008), and thus, zeros substituted into the
          calculations.

       •   The average hexavalent chromium concentration has fluctuated over the years of
          sampling, ranging from 0.019 ng/m3 for 2009 to 0.037 ng/m3 for 2008.
                                         24-30

-------
       Observations from Figure 24-22 for lead measurements include the following:

       •  Although the maximum concentration of lead was measured in 2011 (23.2 ng/m3),
          similar concentrations were also measured in 2004 and 2009.

       •  The statistical metrics shown in Figure 24-22 reflect a relatively high level of
          variability in the lead concentrations measured at BTUT. One indicator of this is the
          difference between the average and median concentrations. This difference is at a
          minimum for 2008 (0.22 ng/m3) and at a maximum for 2009 (1.33 ng/m3), with a
          difference for half of the years of sampling greater than 1 ng/m3.

       •  The average and median concentrations have an overall decreasing trend over the
          period of sampling, although this is difficult to discern in Figure 24-22 due to the
          fluctuations in the maximum concentrations and 95th percentiles.


       Observations from Figure 24-23 for manganese measurements include the following:

       •  The maximum manganese concentration (40.4 ng/m3) was measured in 2004,
          although the next highest concentration, measured in 2007, is similar in magnitude
          (36.0 ng/m3). Concentrations greater than 20 ng/m3 have been measured in 2004,
          2007, 2008, 2009, and 2011.

       •  The average concentration decreased from 2004 to 2005, after which an increase is
          shown through 2007, although these changes are not statistically significant.
          However, a significant decrease in manganese concentrations is shown from 2008
          through 2010. The median concentration follows a similar trend.

       •  With the exception of the 5th percentile, all of the statistical metrics exhibit an
          increase for 2011. The number of concentrations greater than 10 ng/m3 increased
          from six in 2010 to 18 in 2011.


24.5   Additional Risk-Based Screening Evaluations

       The following risk-based screening evaluations were conducted to characterize risk at the

BTUT monitoring site. Refer to Sections 3.3 and 3.5.5 for definitions and explanations regarding

the various toxicity factors, time frames, and calculations associated with these risk-based

screenings.
24.5.1  Risk-Based Screening Assessment Using MRLs
       A risk-based screening was conducted by comparing the concentration data from the

Utah monitoring site to the ATSDRMRLs, where available. As described in Section 3.3, MRLs

are noncancer health risk benchmarks and are defined for three exposure periods: acute

(exposures of 1 to 14 days); intermediate (exposures of 15 to 364 days); and chronic (exposures

of 1 year or greater). The preprocessed daily measurements of the pollutants of interest were

                                         24-31

-------
 compared to the acute MRLs; the quarterly averages were compared to the intermediate MRLs;
 and the annual averages were compared to the chronic MRLs.

       As discussed in Section 4.2.2, none of the measured detections or time-period average
 concentrations for any of the monitoring sites are greater than their respective ATSDR MRL
 noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2011.

 24.5.2 Cancer Risk and Noncancer Hazard Approximations
       For the pollutants of interest for BTUT and where annual average concentrations could
 be calculated, risk was examined by calculating cancer risk and noncancer hazard
 approximations. These approximations can be used as risk estimates for cancer and noncancer
 effects attributable to the pollutants of interest. Although the use of these approximations is
 limited, they may help identify where policy-makers may want to shift or confirm their air-
 monitoring priorities. Refer to Section 3.5.5.2 for an explanation of how cancer risk and
 noncancer hazard approximations are calculated and what limitations are associated with them.
 Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
 approximations are presented in Table 24-6, where applicable. Cancer risk approximations are
 presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
 values.

               Table 24-6. Risk Approximations for the Utah Monitoring Site
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(jig/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Bountiful, Utah - BTUT
Acetaldehyde
Acrylonitrile
Arsenic (PM10)a
Benzene
Benzo(a)pyrenea
Beryllium (PM10)a
0.0000022
0.000068
0.0043
0.0000078
0.00176
0.0024
0.009
0.002
0.000015
0.03

0.00002
60/60
7/60
58/60
60/60
20/62
53/60
2.19
ฑ0.35
0.02
ฑ0.01
<0.01
ฑ<0.01
1.14
ฑ0.15
<0.01
ฑ<0.01
0.01
ฑ0.01
4.83
1.10
2.52
8.90
0.08
0.03
0.24
0.01
0.04
0.04

0.01
— = a Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 24-5.
                                          24-32

-------
        Table 24-6. Risk Approximations for the Utah Monitoring Site (Continued)
Pollutant
1,3 -Butadiene
Cadmium (PM10)a
Carbon Tetrachloride
Chloroform
ฃ>-Dichlorobenzene
1 ,2-Dichloroethane
Dichloromethane
Ethylbenzene
Formaldehyde
Hexachloro- 1 ,3 -butadiene
Hexavalent Chromium3
Lead (PM10)a
Manganese (PM10)a
Naphthalene3
Nickel (PM10)a
Propionaldehyde
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
Cancer
URE
(Hg/m3)1
0.00003
0.0018
0.000006

0.000011
0.000026
0.00000013
0.0000025
0.000013
0.000022
0.012


0.000034
0.00048

0.00000026
0.0000048
0.0000088
Noncancer
RfC
(mg/m3)
0.002
0.00001
0.1
0.098
0.8
2.4
0.6
1
0.0098
0.09
0.0001
0.00015
0.00005
0.003
0.00009
0.008
0.04
0.002
0.1
#of
Measured
Detections
vs. # of
Samples
49/60
60/60
60/60
29/60
29/60
15/60
60/60
60/60
60/60
6/60
53/61
60/60
60/60
62/62
60/60
60/60
41/60
11/60
1/60
Annual
Average
(Hg/m3)
0.10
ฑ0.02
0.01
ฑ0.01
0.61
ฑ0.03
0.06
ฑ0.02
0.05
ฑ0.02
0.03
ฑ0.01
53.90
ฑ 50.64
0.47
ฑ0.13
4.49
ฑ1.15
0.01
ฑ0.01
0.01
ฑ0.01
O.01
ฑO.01
0.01
ฑ0.01
0.05
ฑ0.01
0.01
ฑ0.01
0.46
ฑ0.07
0.10
ฑ0.02
0.02
ฑ0.01
0.01
ฑ0.01
Cancer Risk
Approximation
(in-a-million)
2.98
0.25
3.68

0.50
0.72
7.01
1.17
58.42
0.22
0.26


1.61
0.83

0.03
0.07
0.01
Noncancer
Hazard
Approximation
(HQ)
0.05
0.01
0.01
O.01
0.01
O.01
0.09
O.01
0.46
O.01
0.01
0.02
0.15
0.02
0.02
0.06
0.01
0.01
0.01
— = a Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 24-5.
        Observations for BTUT from Table 24-6 include the following:

        •  The pollutants with the highest annual average concentrations are dichloromethane,
           formaldehyde, acetaldehyde, and benzene, as discussed in Section 24.4.1.

        •  The pollutants with the highest cancer risk approximations are formaldehyde,
           benzene, dichloromethane, and acetaldehyde. The cancer risk approximation for
                                           24-33

-------
          formaldehyde is eight times greater than the cancer risk approximation for
          dichloromethane, even with an annual average concentration of dichloromethane
          nearly 12 times greater than the annual average concentration of formaldehyde. This
          is an indication of the toxicity potential of formaldehyde vs. dichloromethane.
       •  There were no pollutants of interest with noncancer hazard approximations greater
          than 1.0, indicating that no adverse health effects are expected from these individual
          pollutants. The highest noncancer hazard approximation was calculated for
          formaldehyde (0.46), which is the second highest noncancer hazard approximation
          calculated among the site-specific pollutants of interest with noncancer toxicity
          factors.

24.5.3 Risk-Based Emissions Assessment
       In addition to the risk-based screenings discussed above, Tables 24-7 and 24-8 present an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 24-7 presents the 10 pollutants with the highest emissions from the 2008 NEI, the
10 pollutants with the highest toxicity-weighted emissions, and the 10 pollutants with the highest
cancer risk approximations (in-a-million), as calculated from the annual averages provided in
Table 24-6. Table 24-8 presents similar information, but identifies the 10 pollutants with the
highest noncancer hazard approximations (HQ), also calculated from annual averages provided
in Table 24-6.

       The pollutants listed in Tables 24-7 and 24-8 are limited to those that have cancer and
noncancer toxicity factors, respectively. As a result, although the actual value of the emissions is
the same, the highest emitted pollutants in the cancer table may be different from the noncancer
table. The cancer risk and noncancer hazard approximations based on each site's annual averages
are limited to those pollutants for which each respective site sampled. As discussed  in
Section 24.3, BTUT sampled for VOCs, carbonyl compounds, SNMOCs, metals (PMio), PAHs,
and hexavalent chromium. In addition, the cancer risk and noncancer hazard approximations are
limited to those pollutants with enough data to meet the criteria for annual averages to be
calculated. A more in-depth discussion of this analysis is provided in Section 3.5.5.3. Similar to
the cancer risk and noncancer hazard approximations, this analysis may help policy-makers
prioritize their air monitoring activities.
                                          24-34

-------
     Table 24-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
                                                     the Utah Monitoring Site
to
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Bountiful, Utah (Davis County) - BTUT
Benzene
Ethylbenzene
Formaldehyde
Acetaldehyde
1,3 -Butadiene
Naphthalene
Dichloromethane
POM, Group 2b
POM, Group la
POM, Group 6
174.72
80.37
76.71
44.32
20.55
10.19
2.66
1.59
0.20
0.14
Benzene
Formaldehyde
1,3 -Butadiene
Naphthalene
Ethylbenzene
POM, Group 3
POM, Group 2b
Hexavalent Chromium, PM
Acetaldehyde
POM, Group 5a
1.36E-03
9.97E-04
6.17E-04
3.46E-04
2.01E-04
1.87E-04
1.40E-04
1.31E-04
9.75E-05
7.17E-05
Formaldehyde
Benzene
Dichloromethane
Acetaldehyde
Carbon Tetrachloride
1,3 -Butadiene
Arsenic
Naphthalene
Ethylbenzene
Acrylonitrile
58.42
8.90
7.01
4.83
3.68
2.98
2.52
1.61
1.17
1.10

-------
     Table 24-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
                                                 RfCs for the Utah Monitoring Site
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Bountiful, Utah (Davis County) - BTUT
Toluene
Xylenes
Benzene
Methanol
Hexane
Ethylbenzene
Formaldehyde
Acetaldehyde
Ethylene glycol
1,3 -Butadiene
403.75
329.01
174.72
171.92
100.56
80.37
76.71
44.32
22.52
20.55
Acrolein
1,3 -Butadiene
Formaldehyde
Benzene
Acetaldehyde
Naphthalene
Xylenes
Lead, PM
Arsenic, PM
Propionaldehyde
215,538.97
10,275.13
7,827.57
5,823.93
4,924.23
3,395.03
3,290.06
1,295.06
736.79
619.63
Formaldehyde
Acetaldehyde
Manganese
Dichloromethane
Propionaldehyde
1,3 -Butadiene
Arsenic
Benzene
Lead
Nickel
0.46
0.24
0.15
0.09
0.06
0.05
0.04
0.04
0.02
0.02
to

-------
Observations from Table 24-7 include the following:

•  Benzene, ethylbenzene, and formaldehyde are the highest emitted pollutants with
   cancer UREs in Davis County.

•  The pollutants with the highest toxi city-weighted emissions (of the pollutants with
   cancer UREs) are benzene, formaldehyde, 1,3-butadiene.

•  Seven of the highest emitted pollutants also have the highest toxi city-weighted
   emissions.

•  Formaldehyde and benzene, which have the highest and second highest cancer risk
   approximations for BTUT, appear near the top of both emissions-based lists.
   Acetaldehyde, 1,3-butadiene, naphthalene, and ethylbenzene also appear on all three
   lists in Table 24-7. Dichloromethane, which has the third highest cancer risk
   approximation for BTUT, ranks seventh for emissions in Davis County but does not
   have one of the  highest toxicity-weighted emissions (it ranks 19th). Carbon
   tetrachloride, which has the fifth highest cancer risk approximation for BTUT,
   appears on neither  emissions-based list.

•  POM, Group 2b is  the eighth highest emitted "pollutant" in Davis County and ranks
   seventh for toxicity-weighted emissions. POM, Group 2b includes several PAHs
   sampled for at BTUT including acenaphthylene, fluoranthene,  and perylene. None of
   the PAHs included in POM, Group 2b failed screens for BTUT.

•  POM, Group 5a ranks tenth for toxicity-weighted emissions in Davis County. POM,
   Group 5a includes benzo(a)pyrene, which is one of BTUT's pollutants of interest.
   POM, Group 5a is not one of the highest emitted pollutants in Davis County and is
   not among the pollutants with the highest cancer risk approximations for BTUT.


Observations from Table 24-8 include the following:

•  Toluene, xylenes, and benzene are the highest emitted pollutants with noncancer
   RfCs in Davis County.

•  The pollutants with the highest toxicity-weighted emissions (of the pollutants with
   noncancer RfCs) are acrolein, 1,3-butadiene, and formaldehyde. Although acrolein
   was sampled for at BTUT, this pollutant was excluded from the pollutants of interest
   designation, and thus subsequent risk-based screening evaluations, due to questions
   about the consistency and reliability of the measurements, as discussed in Section 3.2.

•  Five of the highest emitted pollutants also have the highest toxicity-weighted
   emissions.

•  Although less than 1.0, formaldehyde, acetaldehyde, and manganese have the highest
   noncancer hazard approximations for BTUT. Formaldehyde and acetaldehyde rank
   third and fifth (respectively) for toxicity-weighted emissions and seventh and eighth
                                  24-37

-------
          (respectively) for total emissions. Dichloromethane, which has the fourth highest
          noncancer hazard approximation, appears on neither emissions-based list.


24.6   Summary of the 2011 Monitoring Data for BTUT

       Results from several of the data treatments described in this section include the

following:

       ปซป  Twenty-one pollutants failed at least one screen for BTUT; of these, 12 were NATTS
          MQO Core Analytes.

       ปซป  Dichloromethane had the highest annual average concentration among the pollutants
          of interest for BTUT, followed by formaldehyde, acetaldehyde, and benzene.

       ปซป   The annual average formaldehyde concentration for BTUT is the highest annual
          average among NMP sites sampling carbonyl compounds.

       ปซป  Formaldehyde concentrations at BTUT have been increasing in the last few years of
          sampling while benzene concentrations have decreased in the last few years. There is
          an overall decreasing trend in concentrations of lead at BTUT.
                                         24-38

-------
25.0   Sites in Vermont
       This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the UATMP and NATTS sites in Vermont, and integrates these
concentrations with emissions, meteorological, and risk information. Data generated by sources
other than ERG are not included in the data analyses contained in this report. Readers are
encouraged to refer to Sections 1 through 4 and the glossary (Appendix P) for detailed
discussions and definitions regarding the various data analyses presented below.

25.1   Site Characterization
       This section characterizes the Vermont monitoring sites by providing geographical and
physical information about the location of the sites and the surrounding areas. This information
is provided to give the reader insight regarding factors that may influence the air quality near the
sites and assist in the interpretation of the ambient monitoring measurements.

       The Vermont NATTS site (UNVT) and one of the UATMP sites (BURVT) are located in
northwest Vermont in the Burlington-South Burlington, VT MSA. The third site is located
farther south in Rutland, Vermont. Figures 25-1 and 25-2 are the composite satellite images
retrieved from ArcGIS Explorer showing the Burlington monitoring sites in their urban and rural
locations.  Figure 25-3 identifies nearby point source emissions locations by source category, as
reported in the 2008 NEI for point sources. Note that only sources within 10 miles of the sites are
included in the facility counts provided in Figure 25-3. A 10-mile boundary was chosen to give
the reader an  indication of which emissions sources and emissions source categories could
potentially have a direct effect on the air quality at the monitoring sites. Further, this boundary
provides both the proximity of emissions sources to the monitoring sites as well as the quantity
of such sources within a given distance of the sites. Sources outside the 10-mile radii are still
visible on the map, but have been grayed out in order to show emissions sources just outside the
boundary. Figures 25-4 and 25-5 are the composite satellite image and emissions sources map
for the Rutland site. Table 25-1 provides supplemental geographical information such as land
use, location setting,  and locational coordinates.
                                          25-1

-------
                                    Figure 25-1. Burlington, Vermont (BURVT) Monitoring Site
to

-------
                                     Figure 25-2. Underbill, Vermont (UNVT) Monitoring Site
to

I
OJ

-------
Figure 25-3. NEI Point Sources Located Within 10 Miles of BURVT and UNVT
                                73-Sf-W    73'OX-W    TZ'WOTW   72-5trtn/V    TJ'45'CTW    72'WCrW
Legend
 ฉ  BURVT UATMP site
                                               Note: Due to facility density and collocation, the total facilities
                                               displayed may not represent all facilities within the area of interest.
                                  UNVT NATTS site   O   10 mile radius f   1 County boundary
   Source Category Group (No. of Facilities)
     -f1   Aircraft Operations (13)
     e   Electrical Equipment (2)
                                                 &   Institutional - school (1)
                                                 ฎ   Laboratory (1)
                                                  A   Military Base/National Security Facility (1)
          Electricity Generation via Combustion (2)       M   Miscellaneous Manufacturing (1)
     •   Gasoline/Diesel Service Station (1 )
     4   Hot Mix Asphalt Plant (1)
     %   Industrial Machinery and Equipment (2)
                                                  1    Primary Metal Production (1)
                                                  P    Printing/Publishing (2)
                                                  TI    Telecommunications (1)
                                              25-4

-------
                                      Figure 25-4. Rutland, Vermont (RUVT) Monitoring Site
to

I
(^

-------
Figure 25-5. NEI Point Sources Located Within 10 Miles of RUVT
   TS'lO'tTW        73 S'O'W        73"0'(TW
                                                      72'50'CTVV        72-45'Q'W
          73-151TW       73MO-0-W
        Legend
                                    73'5'O-W        TJ'OtrW        TTteWt       72-SOXnW
                                       Note: Due to facility density and collocation, the total facilities
                                       displayed may not represent all facilities within the area of interest.
              RUVT UATMP site        10 mile radius |      | County boundary

              Source Category Group (No. of Facilities)
                *   Aerospace/Aircraft Manufacturing (2)
                -fr1   Aircraft Operations (4)
                o   Clay Ceramics Manufacturing (1)
                ป   Hot Mix Asphalt Plant (1)
                5   Miscellaneous Coating Manufacturing (1}
                     Plywood, Particleboard, OSB (1)
                H   Pulp and Paper Plant/Wood Products (1)
                w   Woodwork, Furniture, Millwork & Wood  Preserving (1)
                                      25-6

-------
                              Table 25-1. Geographical Information for the Vermont Monitoring Sites
Site
Code
BURVT
RUVT
UNVT
AQS Code
50-007-0014
50-021-0002
50-007-0007
Location
Burlington
Rutland
Underbill
County
Chittenden
Rutland
Chittenden
Micro- or
Metropolitan
Statistical Area
Burlington-South
Burlington, VT
MSA
Rutland, VT MSA
Burlington-South
Burlington, VT
MSA
Latitude
and
Longitude
44.476202,
-73.210383
43.608056,
-72.982778
44.52839,
-72.86884
Land Use
Commercial
Commercial
Forest
Location
Setting
Urban/City
Center
Urban/City
Center
Rural
Additional Ambient Monitoring Information1
CO, NO, NO2, NOX, Carbonyl compounds,
Meteorological parameters, PM10, PM25.
CO, SO2, NO, NO2, NOX, Carbonyl compounds,
Meteorological parameters, PM10, PM25.
Haze, Sulfate, SO2, O3, Meteorological parameters,
PM10, PM Coarse, PM25, and PM25 Speciation.
Data for additional pollutants are reported to AQS for these sites (EPA, 2012c); however, these data are not generated by ERG and are therefore not included in this report
BOLD ITALICS = EPA-designated NATTS Site
to

-------
       BURVT is located in a municipal parking lot in downtown Burlington near the
intersection of Main Street and South Winooski Avenue. This location is less than 1 mile east of
Burlington Bay on Lake Champlain. The areas to the west are commercial while the areas to the
east are residential, as shown in Figure 25-1. Route 2 (Main Street) and Route 7 (South Willard
Street) intersect two blocks east of the monitoring site and 1-89 runs north-south just over 1 mile
east of the site. Between the two roadways and the interstate lies the University of Vermont.

       The UNVT monitoring  site is located on the Proctor Maple Research Farm in Underhill,
Vermont, which is east of the Burlington area. Mount Mansfield, the highest peak in Vermont,
lies to the east in Underhill State Park, less than 3  miles away. The Underhill Artillery Range is a
few miles to the south. Figure 25-2 shows that the area surrounding the site is rural in nature and
heavily forested.  This site is intended to serve as a background site for the region for trends
assessment, standards compliance, and long-range transport assessment.

       As Figure 25-3 shows, UNVT and BURVT are located approximately 16  miles apart.
Most of the emissions sources are located between these two sites, although closer to BURVT.
The source category with the greatest number of emissions sources surrounding these sites is the
aircraft operations source category, which includes airports as well as small runways,  heliports,
or landing pads. The sources closest to BURVT are a medical school/hospital, an airport, and
two facilities generating electricity via combustion. The sources  closest to UNVT are  private
airports.

       The RUVT monitoring  site is located in Rutland, in central Vermont. The city of Rutland
is in  a valley between the Green Mountains to the  east and Taconic Mountains to the west. The
monitoring site is located in the courthouse parking lot in downtown Rutland, just north of West
Street.  Commercial areas are located to the  east and south, while residential areas are located to
the north and west, as shown in Figure 25-4. A railway parallels  Route 4 coming  into  Rutland
from the west, crosses under Route 4, then meanders around a shopping plaza just south of
Route 4. The north junction of Route 4 and Route  7 is approximately 1/2 mile east of the site.
Figure 25-5 shows that most of the emissions sources near RUVT are located along Route 4 and
Route 7, just south of the monitoring site. The source categories  with the greatest number of
sources include aircraft operations and aerospace/aircraft manufacturing. The source closest to
RUVT is an aerospace/aircraft manufacturer.
                                          25-8

-------
       Table 25-2 presents additional site-characterizing information, including indicators of

mobile source activity, for the Vermont monitoring sites. Table 25-2 includes county-level

population and vehicle registration information. Table 25-2 also includes a county-level vehicle

registration-to-population ratio, which was calculated to represent the number of vehicles per

person within each monitoring site's residing county. In addition, the population within 10 miles

of each site is presented, based on postal code population data estimates. An estimate of 10-mile

vehicle ownership was then determined by applying the county-level vehicle registration-to-

population ratio to the 10-mile population surrounding each monitoring site. Table 25-2 also

contains traffic volume information for each site. Finally, Table 25-2 presents the  county-level

daily VMT for Chittenden and Rutland Counties.
     Table 25-2. Population, Motor Vehicle, and Traffic Information for the Vermont
                                     Monitoring Sites
Site
BURVT
UNVT
RUVT
Estimated
County
Population1
157,491
61,289
County-level
Vehicle
Registration2
169,767
70,900
Vehicles per
Person
(Registration:
Population)
1.08
1.16
Population
within 10
miles3
120,787
33,622
34,662
Estimated
10-mile
Vehicle
Ownership
130,202
36,243
40,098
Annual
Average
Daily
Traffic4
14,000
1,110
7,200
County-
level Daily
VMT5
4,027,945
1,766,027
 Bounty-level population estimates reflect 2011 data from the U.S. Census Bureau (Census Bureau, 2012b)
 2County-level vehicle registration reflects 2012 data from the Vermont DMV (VT DMV, 2012)
 310-mile population estimates reflect 2011 data from Xionetic (Xionetic, 2011)
 4AADT reflects 2007 and 2011 data for BURVT and UNVT, respectively, from the Chittenden County Regional
  Planning Commission (CCRPC, 2007 and 2011) and 2010 data for RUVT from Vermont Agency of
  Transportation (VTrans, 2011)
 5County-level VMT reflects 2010 data from the Vermont Agency of Transportation (Vtrans, 2010)
 BOLD ITALICS = EPA-designaled NATTS Site
       Observations from Table 25-2 include the following:

       •   The population for Chittenden County is more than twice the population of Rutland
           County. The populations for both counties are in the bottom third compared to other
           counties with NMP sites. UNVT's 10-mile population is the lowest among the three
           Vermont sites, although it is similar to the 10-mile population surrounding RUVT.
           The 10-mile populations for BURVT and UNVT show that BURVT is near the center
           of population for Chittenden County while UNVT is not.

       •   Although similar patterns are  shown in the vehicle ownership data, the number of
           vehicles registered in each county is higher than the population counts, leading to
           relatively large vehicle-per-person ratios. This indicates that some people own more
           than one vehicle.
                                           25-9

-------
       •  The traffic volume is highest near BURVT and lowest near UNVT among the
          Vermont sites. The traffic estimates near these sites are in the bottom third compared
          to the traffic near other NMP sites. The traffic estimate for BURVT is provided for
          Main Street between South Union Street and South Willard Street; Pleasant Valley
          Road, north of Harvey Road for UNVT; and US-4 Business between Merchants Row
          and Grove Street for RUVT.
       •  Even though the county-level daily VMT for Chittenden County is more than twice
          the VMT for Rutland County, both VMTs are in the bottom third compared to other
          counties with NMP monitoring sites (where VMT data were available).

25.2   Meteorological Characterization
       The following sections characterize the meteorological conditions near the monitoring
sites in Vermont on sample days, as well as over the course of the year.
25.2.1  Climate Summary
       The city of Burlington resides just to the east of Lake Champlain in northwest Vermont.
Lake Champlain has a moderating affect on the city, keeping the city slightly warmer in winter
than it could be given its New England location. The town of Underhill is located to the east of
Burlington but still within the Burlington MSA. The city of Rutland is located 60 miles south of
the Burlington area. Rutland resides within the same climatic division of Vermont as Burlington,
but misses the moderating influences of Lake Champlain. The state of Vermont is affected by
most storm systems that track across the  country, producing variable weather and often cloudy
skies. Summers in Vermont are pleasant, with warm days and cool nights, escaping much of the
heat and humidity much of the East Coast experiences. Winters are warmer in the Champlain
Valley  region than in other portions of the state but snow is common state-wide. Precipitation is
evenly  distributed throughout the year. Average annual winds parallel the valleys, generally from
the south ahead of advancing weather systems, or from the north behind these systems. These
storm systems tend to be moderated somewhat due to the Adirondacks to the west and Green
Mountains to the east (Bair, 1992; NCDC, 2013; NOAA, 2013c).
                                         25-10

-------
25.2.2 Meteorological Conditions in 2011
       Hourly meteorological data from the NWS weather stations nearest the Vermont
monitoring sites were retrieved for 2011 (NCDC, 2011). The closest weather station to BURVT
is located at Burlington International Airport; nearest RUVT is Rutland State Airport; and
nearest UNVT is Morrisville-Stowe State Airport (WBANs 14742, 94737, and 54771,
respectively). Additional information about these weather stations, such as the distance between
the sites and the weather stations, is provided in Table 25-3. These data were used to determine
how meteorological conditions on sample days vary from conditions experienced throughout the
year.

       Table 25-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for the entire year for 2011. Also included in Table 25-3 is the
95 percent confidence interval for each parameter. As shown in Table 25-3, average
meteorological conditions on sample days at UNVT were representative of average weather
conditions throughout the year. Temperatures on sample days at BURVT and RUVT appear
slightly warmer than those for all of 2011, although the differences are not statistically
significant.

25.2.3 Back Trajectory Analysis
       Figure 25-6 is the composite back trajectory map for days on which samples were
collected at the BURVT monitoring site in 2011. Included in Figure 25-6 are four back
trajectories per sample day. Figure 25-7 is the corresponding cluster analysis.  Similarly,
Figures 25-8 through 25-11 are the composite back trajectory maps and corresponding cluster
analyses for RUVT and UNVT. An in-depth description of these maps and how they were
generated is presented in Section 3.5.2.1. For the composite maps, each line represents the
24-hour trajectory along which a parcel of air traveled toward the monitoring site on a given
sample day and time, based on an initial height of 50 meters AGL. For the cluster analyses, each
line corresponds to a trajectory representative of a given cluster of back trajectories. Each
concentric circle around the sites in Figures 25-6 through 25-11 represents 100 miles.
                                         25-11

-------
                           Table 25-3. Average Meteorological Conditions near the Vermont Monitoring Sites
Closest NWS Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(ฐF)
Average
Temperature
(ฐF)
Average
Dew Point
Temperature
(ฐF)
Average
Wet Bulb
Temperature
(ฐF)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar
Wind
Speed
(kt)
Burlington, Vermont - BURVT
Burlington Intl.
Airport
14742
(44.48, -73.16)
2.38
miles
87ฐ
(E)
Sample
Day
2011
57.4
ฑ7.3
55.9
ฑ2.2
48.9
ฑ6.7
47.9
ฑ2.0
39.1
ฑ6.7
38.2
ฑ2.0
44.4
ฑ6.2
43.5
ฑ1.9
71.9
ฑ4.7
71.7
ฑ1.2
1015.5
ฑ3.1
1014.9
ฑ0.8
6.4
ฑ1.1
6.3
ฑ0.3
Rutland, Vermont - RUVT
Rutland State Airport
94737
(43.53, -72.95)
5.60
miles
149ฐ
(SSE)
Sample
Day
2011
56.6
ฑ6.1
54.8
ฑ2.0
48.9
ฑ5.8
46.4
ฑ1.9
39.1
ฑ6.2
36.4
ฑ2.0
44.4
ฑ5.5
41.9
ฑ1.8
71.4
ฑ4.4
70.3
ฑ1.2
NA
NA
6.6
ฑ1.0
6.1
ฑ0.3
Underbill, Vermont - UNVT
Morrisville - Sto we
State Airport
54771
(44.53, -72.61)
11.84
miles
78ฐ
(E)
Sample
Day
2011
54.1
ฑ4.7
54.7
+ 2.2
44.9
ฑ4.2
45.1
+ 2.0
36.5
ฑ4.4
36.7
+ 2.1
41.2
ฑ4.0
41.3
+ 1.9
74.9
ฑ2.7
75.1
+ 1.1
1015.9
ฑ1.8
1015.6
+ 0.8
3.2
ฑ0.5
3.0
+ 0.2
to
        NA = Sea level pressure was not recorded at the Rutland State Airport
                                                                                    -year averages.

-------
Figure 25-6. 2011 Composite Back Trajectory Map for BURVT
    Figure 25-7. Back Trajectory Cluster Map for BURVT
                          25-13

-------
Figure 25-8. 2011 Composite Back Trajectory Map for RUVT
    Figure 25-9. Back Trajectory Cluster Map for RUVT
                          25-14

-------
Figure 25-10. 2011 Composite Back Trajectory Map for UNVT
    Figure 25-11. Back Trajectory Cluster Map for UNVT
                          25-15

-------
       Observations from Figures 25-6 through 25-11 for the Vermont monitoring sites include
the following:

       •  The composite back trajectories maps for the Vermont sites are similar to each other,
          which is not unexpected given their relatively close proximity to each other.
          However, BURVT and RUVT sampled on a l-in-12 schedule, yielding roughly half
          as many sample days for these sites as UNVT.

       •  The composite back trajectory maps show that the majority of back trajectories
          originated from a direction with a westerly component.

       •  For BURVT and RUVT, the farthest away a back trajectory originated was over
          south-central Ontario, Canada, or greater than 500 miles away. For UNVT, the
          farthest away a back trajectory originated was central West Virginia, or nearly
          600 miles away. The average trajectory length varied from 220 miles for RUVT to
          231 miles for BURVT and UNVT. Most back trajectories (roughly 90  percent)
          originated within  400 miles of each site.

       •  The cluster analyses for BURVT and RUVT are fairly similar to each other
          direct! onally, although the percentages vary. For BURVT, 34 percent of back
          trajectories originated to the north of the site, generally over Quebec, Canada.  For
          RUVT, this percentage is 26 percent. Between 7 and 9 percent of back trajectories
          originated to the west of the sites over the Greater Lakes region and  southern Ontario,
          Canada. For BURVT, 30 percent of back trajectories originated from the southwest
          quadrant, including south and west, primarily over New York and Pennsylvania. For
          RUVT, this percentage is higher, at 41 percent. The cluster trajectory originating to
          the east of the sites and representing about one-quarter of the back trajectories,
          includes longer back trajectories originating over the Atlantic Ocean, shorter
          trajectories originating over New Hampshire, as well as trajectories originating from
          other directions but generally less than 100 miles in length.

       •  The back trajectories for UNVT are represented by five cluster trajectories rather than
          four in Figure 25-11. Twenty-eight percent of back trajectories originated to the
          northwest of UNVT and another seven percent originated to the north to northeast of
          UNVT. These back trajectories are combined in the cluster analyses for BURVT and
          RUVT. Twenty percent of back trajectories originated to  the southwest to west of
          UNVT over the Great Lakes region and southern Ontario, Canada. Just less than one-
          quarter of back trajectories originated to the south to southwest of UNVT over the
          Northeast and Mid-Atlantic states. The same percentage of back trajectories is
          represented by the short cluster trajectory originating over New Hampshire. Similar to
          the other two Vermont sites, this cluster trajectory represents longer back trajectories
          originating over the Atlantic Ocean, shorter trajectories originating over the New
          England states, as well as back trajectories originating from other directions but
          generally less than 100 miles in  length.
                                          25-16

-------
25.2.4 Wind Rose Comparison
       Hourly surface wind data from the NWS weather stations at Burlington International
Airport (for BURVT), Rutland State Airport (for RUVT), and Morrisville-Stowe State Airport
(for UNVT) were uploaded into a wind rose software program to produce customized wind
roses, as described in Section 3.5.2.2. A wind rose shows the frequency of wind directions using
"petals" positioned around a 16-point compass, and uses different colors to represent wind
speeds.

       Figure 25-12 presents a map showing the distance between the NWS station and
BURVT, which may be useful for identifying topographical influences that may affect the
meteorological patterns experienced at this location. Figure 25-12 also presents three different
wind roses for the BURVT monitoring site. First, a historical wind rose representing 2001 to
2010 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
2011 is presented. Next, a wind rose representing wind data for days on which samples were
collected in 2011 is presented. These can be used to identify the predominant wind speed and
direction in 2011 and determine if wind  observations on sample days were representative of
conditions experienced over the entire year and historically. Figures 25-13 and 25-14 present the
three wind roses and distance maps for the RUVT and UNVT monitoring sites.

       Observations from Figure 25-12  for BURVT include the following:
       •    The Burlington International Airport weather station is located approximately
           2.8 miles east of BURVT.
       •    The historical wind rose shows that southerly winds are prevalent near BURVT,
           accounting for nearly 22 percent of the hourly measurements. Calm winds (< 2 knots)
           account for another 19 percent of measurements. Winds from the northwest quadrant,
           including north, account for another 30 percent of the wind observations. Winds from
           the eastern quadrants are rarely observed.
       •    The wind patterns shown on the 2011 wind rose are similar to the historical wind
           patterns, indicating that conditions observed during 2011 were similar to those
           observed over the last 10 years.
       •    The sample day wind rose shows that wind conditions on sample days were similar to
           those experienced throughout 2011, although northwesterly to northerly winds
           accounted for a higher percentage of the hourly wind measurements.
                                         25-17

-------
    Figure 25-12. Wind Roses for the Burlington International Airport Weather Station

                                    near BURVT
Distance between BURVT and NWS Station
2001-2010 Historical Wind Rose
                    *     .    *  „
                   ... bid,  V  S
                   , Ellington   ^*v X 5_
                           '• .?  j     -
          2011 Wind Rose
                                                      Sample Day Wind Rose
                                      25-18

-------
  Figure 25-13. Wind Roses for the Rutland State Airport Weather Station near RUVT
Distance between RUVT and NWS Station
2003-2010 Historical Wind Rose
         2011 Wind Rose
                                                    Sample Day Wind Rose
                                    25-19

-------
 Figure 25-14. Wind Roses for the Morrisville-Stowe State Airport Weather Station near
                                      UNVT
Distance between UNVT and NWS Station
2001-2010 Historical Wind Rose
                                    N
                                  +
          2011 Wind Rose
    Sample Day Wind Rose
                                     25-20

-------
       Observations from Figure 25-13 for RUVT include the following:

       •  The Rutland State Airport weather station is located approximately 5.6 miles south-
          southeast of RUVT.

       •  The historical wind rose shows that east-southeasterly and southeasterly winds were
          prevalent near RUVT, as these directions account for over a quarter of the hourly
          measurements. Winds from the southwest and northwest quadrants were also
          observed while winds from the northeast quadrant were almost never observed. Calm
          winds were observed for 17 percent of the hourly measurements.

       •  The wind patterns shown on the 2011 wind rose are similar to the historical wind
          patterns, although a higher percentage of winds from the southeast and fewer east-
          southeasterly winds were observed in 2011.

       •  The sample day wind rose exhibits similar wind patterns as the historical and full-
          year wind roses, but with even fewer east-southeasterly winds and more south-
          southeasterly and southerly wind observations.


       Observations from Figure 25-14 for UNVT include the following:

       •  The Morrisville-Stowe Airport weather station is located approximately 12 miles east
          of UNVT. Between the site and the weather station lie the Green Mountains.

       •  The historical wind rose shows that calm winds were prevalent near UNVT, as calm
          winds were observed for nearly 45 percent of the hourly measurements. Winds from
          the northwest to north account for 20 percent of the wind  observations greater than
          2 knots. Winds from the south to south-southwest account for another 15 percent of
          observations.

       •  The wind patterns shown on the 2011 wind rose are similar to the historical wind
          patterns, although calm winds account for almost 50 percent of the observations.

       •  The sample day wind rose shows that wind conditions on sample days were similar to
          those experienced throughout 2011, although winds from the northwest to north
          account for a greater percentage of wind measurements on sample days.


25.3   Pollutants of Interest

       Site-specific "pollutants of interest" were determined for the Vermont monitoring sites in

order to allow  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. In

addition, if any of the NATTS MQO Core Analytes measured by  each monitoring site did not

                                         25-21

-------
meet the pollutant of interest criteria based on the preliminary risk-based screening, that pollutant
was added to the list of site-specific pollutants of interest. A more in-depth description of the
risk-based screening process is presented in Section 3.2.

       Table 25-4 presents the results of the preliminary risk-based screening process for the
Vermont monitoring sites. The pollutants that failed at least one screen and contributed to 95
percent of the total failed screens for each monitoring site are shaded. NATTS MQO Core
Analytes are bolded. Thus, pollutants of interest are shaded and/or bolded. BURVT and RUVT
sampled for VOCs only while UNVT sampled for VOCs, hexavalent chromium, PAHs, and
metals (PMio).
        Table 25-4. Risk-Based Screening Results for the Vermont Monitoring Sites
Pollutant
Screening
Value
(Ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Burlington, Vermont - BURVT
Benzene
1,3-Butadiene
Carbon Tetrachloride
ฃ>-Dichlorobenzene
1 ,2-Dichloroethane
Hexachloro- 1 ,3 -butadiene
Acrylonitrile
Ethylbenzene
1, 1,2,2-Tetrachloroethane
Bromomethane
1 ,2-Dibromoethane
0.13
0.03
0.17
0.091
0.038
0.045
0.015
0.4
0.017
0.5
0.0017
Total
30
30
30
7
7
5
4
4
4
1
1
123
30
30
30
27
7
6
4
30
4
22
1
191
100.00
100.00
100.00
25.93
100.00
83.33
100.00
13.33
100.00
4.55
100.00
64.40
24.39
24.39
24.39
5.69
5.69
4.07
3.25
3.25
3.25
0.81
0.81
24.39
48.78
73.17
78.86
84.55
88.62
91.87
95.12
98.37
99.19
100.00

Rutland, Vermont - RUVT
Benzene
Carbon Tetrachloride
1,3-Butadiene
1 ,2-Dichloroethane
Ethylbenzene
Acrylonitrile
Hexachloro- 1 ,3 -butadiene
0.13
0.17
0.03
0.038
0.4
0.015
0.045
Total
30
30
25
4
4
2
1
96
30
30
25
4
30
2
1
122
100.00
100.00
100.00
100.00
13.33
100.00
100.00
78.69
31.25
31.25
26.04
4.17
4.17
2.08
1.04
31.25
62.50
88.54
92.71
96.88
98.96
100.00

                                         25-22

-------
Table 25-4. Risk-Based Screening Results for the Vermont Monitoring Sites (Continued)
Pollutant
Screening
Value
(Ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Underbill, Vermont - UNVT
Benzene
Carbon Tetrachloride
Arsenic (PM10)
1 ,2-Dichloroethane
Acrylonitrile
1,3-Butadiene
Naphthalene
Hexachloro- 1 ,3 -butadiene
Manganese (PM10)
1 , 1 ,2,2-Tetrachloroethane
Bromomethane
ฃ>-Dichlorobenzene
Nickel (PM10)
0.13
0.17
0.00023
0.038
0.015
0.03
0.029
0.045
0.005
0.017
0.5
0.091
0.0021
Total
60
58
23
12
6
5
4
3
3
2
1
1
1
179
60
60
54
12
6
13
60
4
57
2
28
10
57
423
100.00
96.67
42.59
100.00
100.00
38.46
6.67
75.00
5.26
100.00
3.57
10.00
1.75
42.32
33.52
32.40
12.85
6.70
3.35
2.79
2.23
1.68
1.68
1.12
0.56
0.56
0.56
33.52
65.92
78.77
85.47
88.83
91.62
93.85
95.53
97.21
98.32
98.88
99.44
100.00

     Observations from Table 25-4 include the following:

     •   A total of 11 pollutants, including three NATTS MQO Core Analytes, failed screens
         for BURVT. Seven pollutants, including the same three NATTS MQO Core
         Analytes, failed screens for RUVT. The seven pollutants failing screens for RUVT
         also failed at least one screen for BURVT. Thirteen pollutants, of which seven are
         NATTS MQO Core Analytes, failed screens for UNVT.

     •   The preliminary risk-based screening process identified nine pollutants of interest for
         BURVT. Even though the 95 percent criteria is met by ethylbenzene,
         1,1,2,2-tetrachloroethane contributed equally to the number of failed screens (4), thus,
         this pollutant was also designated as a pollutant of interest, as  discussed in
         Section 3.2. Four additional pollutants (chloroform, tetrachloroethylene, and
         trichloroethylene, and vinyl chloride) were added as pollutants of interest for BURVT
         because they are NATTS MQO Core Analytes, even though these pollutants did not
         fail any screens. These pollutants are not shown in Table 25-4 but are shown in
         subsequent tables in the sections that follow.

     •   The preliminary risk-based screening process identified five pollutants of interest for
         RUVT (benzene, carbon tetrachloride,  1,3-butadiene, 1,2-dichloroethane, and
         ethylbenzene). Three additional pollutants (chloroform, tetrachloroethylene, and
         trichloroethylene) were added as pollutants of interest because they are NATTS MQO
         Core Analytes, even though these pollutants did not fail any screens. These pollutants
         are not shown in Table 25-4 but are shown in subsequent tables in the sections that
         follow. Vinyl chloride is also a NATTS MQO Core Analyte, but because this
         pollutant was not detected at RUVT, it was not added to the pollutants of interest.
                                        25-23

-------
       •  The preliminary risk-based screening process identified nine pollutants of interest for
          UNVT (six VOCs, two metals, and one PAH). Nickel was added to UNVT's
          pollutants of interest because it is a NATTS MQO Core Analyte, even though it did
          not contribute to 95 percent of UNVT's total failed screens. Nine additional
          pollutants (three PMi0 metals, four VOCs, one PAH, and hexavalent chromium) were
          added as pollutants of interest for UNVT because they are NATTS MQO Core
          Analytes, even though these pollutants did not fail any screens. These pollutants are
          not shown in Table 25-4 but are shown in subsequent tables in the sections that
          follow.
       •  Benzene, carbon tetrachloride, 1,3-butadiene, and  1,2-dichloroethane were identified
          as pollutants of interest for each of the Vermont monitoring sites.

25.4   Concentrations
       This section presents various concentration averages used to characterize pollution levels
at the Vermont monitoring sites. Where applicable, the following calculations and data analyses
were performed: Time period-based concentration averages (quarterly and annual) are provided
for the pollutants of interest for the Vermont sites, where the data meet the applicable criteria.
Concentration averages for select pollutants are also presented graphically for each site to
illustrate how the sites' concentrations compare to the program-level averages, as presented in
Section 4.1. In addition, concentration averages for select pollutants are presented from previous
years of sampling in order to characterize concentration trends at the sites. Additional site-
specific statistical summaries for BURVT, RUVT, and UNVT are provided in Appendices J, M,
N,and O.

25.4.1 2011 Concentration Averages
       Quarterly and annual concentration averages were calculated for the pollutants of interest
for each Vermont site, as described in Section 3.1. The quarterly average of a particular
pollutant is simply the average concentration of the preprocessed daily measurements over a
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  of the total number of
samples possible within a given quarter for a quarterly average to be calculated. An annual
average includes all measured detections and substituted zeros for non-detects for the entire year
of sampling. Annual averages were calculated for pollutants where three valid quarterly averages
could be calculated and where method completeness was greater than or equal to 85 percent, as
presented in Section 2.4. Quarterly and annual average concentrations for the Vermont
                                          25-24

-------
monitoring sites are presented in Table 25-5, where applicable. Note that concentrations of the
PAHs, metals, and hexavalent chromium for UNVT are presented in ng/m3 for ease of viewing.
Also note that if a pollutant was not detected in a given calendar quarter, the quarterly average
simply reflects "0" because only zeros substituted for non-detects were factored into the
quarterly average concentration.
 Table 25-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
                               the Vermont Monitoring Sites
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(jig/m3)
2nd
Quarter
Average
(jig/m3)
3rd
Quarter
Average
(jig/m3)
4th
Quarter
Average
(jig/m3)
Annual
Average
(jig/m3)
Burlington, Vermont - BURVT
Acrylonitrile
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
ฃ>-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Hexachloro-1 ,3 -butadiene
1 , 1 ,2,2-Tetrachloroethane
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
4/30
30/30
30/30
30/30
17/30
27/30
7/30
30/30
6/30
4/30
23/30
2/30
1/30
0.01
ฑ0.03
0.99
ฑ0.18
0.11
ฑ0.03
0.49
ฑ0.08
0.05
ฑ0.05
0.07
ฑ0.03
0
0.28
ฑ0.06
0.03
ฑ0.05
0.01
ฑ0.02
0.11
ฑ0.06
0
0
0.03
ฑ0.06
0.61
ฑ0.13
0.09
ฑ0.03
0.58
ฑ0.10
0.07
ฑ0.05
0.08
ฑ0.03
0.02
ฑ0.03
0.25
ฑ0.08
0.01
ฑ0.03
0
0.06
ฑ0.06
0.01
ฑ0.01
0
0.05
ฑ0.11
0.75
ฑ0.19
0.10
ฑ0.03
0.65
ฑ0.04
0.08
ฑ0.07
0.07
ฑ0.03
0.01
ฑ0.02
0.35
ฑ0.09
0.01
ฑ0.02
0.01
ฑ0.02
0.18
ฑ0.21
0
<0.01
ฑ0.01
0
0.79
ฑ0.15
0.10
ฑ0.03
0.62
ฑ0.07
0.08
ฑ0.05
0.06
ฑ0.01
0.04
ฑ0.03
0.32
ฑ0.08
0.02
ฑ0.03
0.01
ฑ0.02
0.13
ฑ0.06
0.01
ฑ0.01
0
0.02
ฑ0.03
0.78
ฑ0.09
0.10
ฑ0.01
0.59
ฑ0.04
0.07
ฑ0.02
0.07
ฑ0.01
0.02
ฑ0.01
0.30
ฑ0.04
0.02
ฑ0.01
0.01
ฑ0.01
0.12
ฑ0.05
0.01
ฑ0.01
O.01
ฑ0.01
    NA = Not available due to the criteria for calculating a quarterly and/or annual average.
    a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of
    viewing.
                                           25-25

-------
Table 25-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
                        the Vermont Monitoring Sites (Continued)
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Ug/m3)
2nd
Quarter
Average
(Ug/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Ug/m3)
Rutland, Vermont - RUVT
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
1 ,2-Dichloroethane
Ethylbenzene
Tetrachloroethylene
Trichloroethylene
30/30
25/30
30/30
9/30
4/30
30/30
23/30
2/30
0.98
ฑ0.21
0.11
ฑ0.05
0.54
ฑ0.02
0
0
0.25
ฑ0.06
0.09
ฑ0.03
0
0.70
ฑ0.39
0.05
ฑ0.03
0.60
ฑ0.07
0.05
ฑ0.05
0.02
ฑ0.03
0.25
ฑ0.06
0.14
ฑ0.12
O.01
ฑ0.01
0.71
ฑ0.06
0.07
ฑ0.02
0.60
ฑ0.12
0.07
ฑ0.07
0
0.33
ฑ0.02
0.09
ฑ0.06
0
1.26
ฑ0.38
0.17
ฑ0.08
0.68
ฑ0.08
0.05
ฑ0.06
0.02
ฑ0.03
0.38
ฑ0.13
0.14
ฑ0.03
O.01
ฑ0.01
0.89
ฑ0.16
0.09
ฑ0.03
0.61
ฑ0.04
0.04
ฑ0.03
0.01
ฑ0.01
0.30
ฑ0.04
0.12
ฑ0.04
O.01
ฑO.01
Underbill, Vermont - UNVT
Acrylonitrile
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
1 ,2-Dichloroethane
Hexachloro- 1 , 3 -butadiene
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
Arsenic (PM10)a
Benzo(a)pyrene a
Bery Ilium (PM10)a
6/60
60/60
13/60
60/60
17/60
12/60
4/60
19/60
1/60
1/60
55/58
13/60
46/58
0.01
ฑ0.01
0.59
ฑ0.11
0.01
ฑ0.01
0.46
ฑ0.09
0.03
ฑ0.02
O.01
ฑ0.01
0.01
ฑ0.02
0.03
ฑ0.02
0
0
0.23
ฑ0.10
0.04
ฑ0.03
0.01
ฑO.01
0.01
ฑ0.01
0.33
ฑ0.03
0.01
ฑ0.01
0.54
ฑ0.09
0.03
ฑ0.03
0.02
ฑ0.02
0.01
ฑ0.02
0.01
ฑ0.01
0
0.01
ฑ0.01
0.19
ฑ0.08
0.01
ฑ0.01
0.01
ฑO.01
0.01
ฑ0.01
0.31
ฑ0.06
0.01
ฑ0.01
0.63
ฑ0.04
0.01
ฑ0.02
0
0.01
ฑ0.01
0.02
ฑ0.02
0
0
NA
0
NA
0.01
ฑ0.02
0.43
ฑ0.06
0.01
ฑ0.01
0.64
ฑ0.05
0.03
ฑ0.02
0.04
ฑ0.02
O.01
ฑ0.01
0.02
ฑ0.02
O.01
ฑO.01
0
0.31
ฑ0.12
0.01
ฑ0.01
O.01
ฑO.01
0.01
ฑ0.01
0.42
ฑ0.04
0.01
ฑ0.01
0.57
ฑ0.04
0.03
ฑ0.01
0.02
ฑ0.01
0.01
ฑ0.01
0.02
ฑ0.01
O.01
ฑO.01
0.01
ฑ0.01
0.25
ฑ0.06
0.01
ฑ0.01
0.01
ฑO.01
   NA = Not available due to the criteria for calculating a quarterly and/or annual average.
   a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of
   viewing.
                                           25-26

-------
Table 25-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
                       the Vermont Monitoring Sites (Continued)



Pollutant

Cadmium (PM10)a

Hexavalent Chromium3

Lead(PM10)a

Manganese (PM10) a

Naphthalene a

Nickel (PM10)a
#of
Measured
Detections
vs. # of
Samples

58/58

17/60

58/58

58/58

60/60

58/58

1st
Quarter
Average
(Ug/m3)
0.07
ฑ0.03
<0.01
ฑ<0.01
1.50
ฑ0.64
1.37
ฑ0.42
27.87
ฑ 12.64
0.45
ฑ0.29

2nd
Quarter
Average
(Ug/m3)
0.05
ฑ0.01
0.01
ฑ0.01
0.94
ฑ0.27
2.06
ฑ0.70
6.43
ฑ1.53
0.33
ฑ0.09

3rd
Quarter
Average
(Ug/m3)

NA
<0.01
ฑ<0.01

NA

NA
5.44
ฑ1.68

NA

4th
Quarter
Average
(Ug/m3)
0.08
ฑ0.03
<0.01
ฑ<0.01
1.88
ฑ0.54
2.00
ฑ0.63
13.23
ฑ4.58
0.74
ฑ0.15


Annual
Average
(Ug/m3)
0.07
ฑ0.01
<0.01
ฑ<0.01
1.46
ฑ0.29
1.84
ฑ0.32
13.37
ฑ3.96
0.53
ฑ0.10
   NA = Not available due to the criteria for calculating a quarterly and/or annual average.
   a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of
   viewing.


      Observations for BURVT and RUVT from Table 25-5 include the following:

      •  BURVT and RUVT sampled VOCs on a l-in-12 day schedule.

      •  For both sites, the pollutants with the highest annual average concentrations are
         benzene, carbon tetrachloride, and ethylbenzene, although all of the annual average
         concentrations for the pollutants of interest for both sites are less than 1 |ig/m3.

      •  Several of the VOCs listed for BURVT were detected relatively few times, resulting
         in relatively large confidence intervals for the quarterly and annual averages, some of
         which are greater than the averages themselves. Examples include acrylonitrile,
         1,2-dichloroethane, hexachloro-l,3-butadiene, and 1,1,2,2-tetrachloroethane. This is
         also true from some of RUVT's pollutants of interest.

      •  The first quarter benzene average for BURVT is higher than the other quarterly
         averages. A review of the data shows that the maximum concentration of benzene
         was measured on February 2, 2011  (1.38 |ig/m3). Of the five benzene concentrations
         greater than 1  |ig/m3 measured at BURVT, three were measured in the first quarter of
         2011 (with one each measured in the third and fourth quarter).

      •  The third quarter average concentration of tetrachloroethylene for BURVT is greater
         than the other quarterly averages and has a confidence interval that is greater than the
         averages itself. A review of the data shows that the maximum concentration of
         tetrachloroethylene was measured on September 24, 2011 (0.75  |ig/m3), which is
         more than twice the next highest concentration (0.31 |ig/m3, measured on
         October 6, 2011). All other measurements of tetrachloroethylene are 0.20 |ig/m3 or
         less.
                                         25-27

-------
•  Several of the quarterly averages of benzene for RUVT have relatively large
   confidence intervals associated with them. Concentrations of benzene measured at
   RUVT range from 0.311 |ig/m3 to 2.07 |ig/m3. The maximum benzene concentration
   was measured on June 17, 2011  and is the only concentration of benzene greater than
   1 |ig/m3 measured in the second quarter. A total of eight benzene concentrations
   greater than 1 |ig/m3 were measured at RUVT, with two measured in the first quarter,
   one (the maximum) was measured in the second, and five were measured in the
   fourth quarter.

•  Although the second and fourth  quarter average concentrations of tetrachloroethylene
   are the same for RUVT,  the second quarter average has a larger confidence interval
   associated with it. A review of the data shows that the three highest concentrations of
   tetrachloroethylene were measured in April, May, and June. These are three of only
   four measurements greater than 0.20 |ig/m3 measured at this site (with the fourth
   being measured in October).


Observations for UNVT from Table 25-5 include the following:

•  UNVT sampled VOCs, PAHs, PMio metals, and hexavalent chromium on a l-in-6
   day schedule.

•  Carbon tetrachloride and benzene have highest annual average concentrations.
   Similar to BURVT and RUVT, all of the annual average concentrations for the
   pollutants of interest for UNVT  are less than 1 |ig/m3.

•  Of the metals, manganese and lead have the highest annual average concentrations.
   For the PAHs, naphthalene has the highest annual average concentration.

•  The first quarter benzene average for UNVT is higher than the other quarterly
   averages, although the differences are not  statistically significant.  The maximum
   concentration of benzene was measured on February 27, 2011 (1.31 |ig/m3) and is the
   only benzene measurement greater than 1  |ig/m3. The second highest benzene
   concentration measured at UNVT was collected on the following sample day and was
   half as high (0.755 |ig/m3). Of the 16 measurements greater than 0.5 |ig/m3, 11 were
   measured during the first quarter, one was measured in third quarter, and four were
   measured during the fourth quarter.

•  Similar to the other Vermont sites, several of the VOCs listed for UNVT were
   detected relatively few times, resulting in relatively large confidence intervals for the
   quarterly and annual averages. Examples include acrylonitrile, 1,2-dichloroethane,
   and hexachloro-1,3-butadiene.

•  Third quarter averages for the PMio metals are not provided in Table 25-5. This is a
   result of flooding and damage to the metals sampler as well as the Vermont
   Department of Conservation laboratory in the wake of Hurricane Irene at the end of
   August 2011. While a few missed samples are noted for all methods sampled at
   UNVT, metals were the most affected because individual samples were lost in the
   flood at the laboratory.

                                  25-28

-------
       •  Concentrations of naphthalene at UNVT tended to be higher during the colder months
          of the year. The maximum concentration of naphthalene was measured at UNVT on
          January 27, 2011 (85.8 ng/m3). The three highest concentrations of naphthalene were
          measured at UNVT in January and February and of the 15 highest concentrations
          (those greater than 15 ng/m3), 10 were measured during the first quarter of the year
          and the other five were measured during the fourth quarter of 2011.

       •  The first quarter benzo(a)pyrene average for UNVT is higher than the other quarterly
          averages and has a relatively large confidence interval associated with it. The
          maximum concentration of this pollutant (0.162 ng/m3) was measured on
          January 21, 2011 and is one of only three concentrations greater than 0.1 ng/m3
          measured at UNVT. These three measurements were made on the same days as the
          highest naphthalene concentrations.


       Tables 4-9 through 4-12 present the sites with the 10 highest annual average

concentrations for each of the program-level pollutants of interest. Observations for the Vermont

monitoring sites from those tables include the following:

       •  BURVT appears twice in Table 4-9 for VOCs. BURVT has the second highest annual
          average concentration of hexachloro-l,3-butadiene and the tenth highest annual
          average concentration ofp-dichlorobenzene among NMP sites sampling VOCs.

       •  RUVT appears once in Table 4-9. RUVT has the tenth highest annual average
          concentration of benzene among NMP sites sampling VOCs.

       •  UNVT appears only in Table 4-12 for PMio metals. However, because only nine
          NMP sites sampled PMio metals, all nine sites appear in Table 4-12. UNVT ranks
          eighth or ninth for each of the six program-wide metal pollutants of interest.

       •  Compared to other NMP sites, UNVT has some of the lowest annual average
          concentrations for each of the program-wide pollutants of interest. For the VOCs,
          UNVT ranks no higher than 16th. For the PAHs and hexavalent chromium, UNVT
          ranks last for each pollutant.


25.4.2  Concentration Comparison

       In order to better illustrate how a site's annual average concentrations compare to the

program-level averages, a site-specific box plot was created for the selected NATTS MQO Core

Analytes listed in Section 3.5.3, where applicable. Thus, box plots for benzene and 1,3-butadiene

were created for BURVT, RUVT, and UNVT. Box plots were also created for arsenic,

benzo(a)pyrene, hexavalent chromium, manganese, lead, and naphthalene for UNVT.

Figures 25-15 through 25-22 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 3.5.3.

                                         25-29

-------
      Figure 25-15. Program vs. Site-Specific Average Arsenic (PMi0) Concentration
UNVT
              3.5
                                   15
                                              2          Z.5
                                              Concentration (ng/m3)
                                                                             3.5
                 Program:   IstQuartile      2ndQuartile     SrdQuartile      4thQuartile     Average
                 Site:      Site Average     Site Minimum/Maximum

                                                                                                  4.5
         Figure 25-16. Program vs. Site-Specific Average Benzene Concentrations
                                                                  i Program Max Concentration = 23.8 ug/m3
•
	
! Program Max (
1 1



                                                                  	-,
                                                                  i Program Max Concentration = 23.8 |-ig/m-
                                          45
                                              Concentration (t
Progra m :
Site:

IstQuartile
Site Average
0
2ndQuartile
SrdQuartile
n
4thQuartile Average
n
^m i 	 i
Site Minimum/Maximum



                                                                                                   10
      Figure 25-17. Program vs. Site-Specific Average Benzo(a)pyrene Concentration
                                       0.75          1
                                              Concentration (ng/m3)
                 Program:   IstQuartile      2ndQuartile     SrdQuartile      4thQuartile     Average
                 Site:
                          Site Average     Site Minimum/Maximum

                                               25-30

-------
      Figure 25-18. Program vs. Site-Specific Average 1,3-Butadiene Concentrations
    E
Program Max Concentration = 3.51 ug/m3
RUVT
                                                               | Program Max Concentration = 9.51 Lig/m3


I
0
^^^ ^^^

0.5 1 1.5
Concentration (|jg/m3)
Program: IstQuartile 2ndQuartile SrdQuartile
• • •
Site: Site Average Site Minimum/Maximum
o —


Program Max Concentration = 9.51 ng/m3

2
4thQuartile Av(
15
;rage
3
  Figure 25-19. Program vs. Site-Specific Average Hexavalent Chromium Concentration
    •
                                 D.I
                                                3.15
                                           Concentration (ng/m3)
                                                                                            :•.=
                Program:   IstQuartile     2ndQuartile     SrdQuartile     4thQuartile     Average
                Site:      Site Average     Site Minimum/Maximum

       Figure 25-20. Program vs. Site-Specific Average Lead (PMio) Concentration
                                          15           20
                                           Concentration (ng/m3)
Program
Site:
: IstQuartile
Site Average
O
2ndQuartile SrdQuartile 4thQuartile Ave
• n
Site Minimum/Maximum
rage

                                            25-31

-------
    Figure 25-21. Program vs. Site-Specific Average Manganese (PMi0) Concentration
UNVT

1
•

! Program Max ConcEntration = 395 ng/m3


i

                                  75         100        125
                                        Concentration (ng/mi)
                                                                 150
                                                                           175
                                                                                     200
               Program:   IstQuartile     2ndQuartile     SrdQuartile     4thQuartile     Average
              Site:     Site Average    Site Minimum/Maximum

      Figure 25-22. Program vs. Site-Specific Average Naphthalene Concentration
                                                          i Program Max Concentration =779 ng/m-
                                    200      250      3OO
                                        Concentration (ng/m3)
               Program:   IstQuartile     2ndQuartile     SrdQuartile     4thQuartile     Average
              Site:     Site Average    Site Minimum/Maximum

      Observations from Figures 25-15 through 25-22 include the following:

         •   Figure 25-15 shows that UNVT's annual average arsenic (PMio) concentration is
             roughly equivalent to the program-level first quartile (25th percentile). The annual
             average concentration of arsenic for UNVT is the lowest annual average
             concentration among NMP sites sampling this pollutant. The maximum arsenic
             concentration measured at UNVT is less than the program-level average
             concentration. A few non-detects of arsenic were measured at UNVT.

         •   Figure 25-16 for benzene shows all three Vermont sites. Note that the program-
             level maximum concentration (23.8 |ig/m3) is not shown directly on the box plots
             because the scale of the box plots would be too large to readily observe data
             points at the lower end of the concentration range. Thus, the scale has been
             reduced to  10 |ig/m3. This figure shows that the annual average concentration of
             benzene is highest for RUVT and lowest for UNVT and that all three annual
             averages are less than  the program-level average benzene concentration.
             Figure 25-16 also shows that UNVT's annual average benzene concentration is
             less than the program-level average, median, and first quartile  concentrations (and
             is the second lowest among all NMP sites sampling benzene).  The range of
             benzene measurements is smallest for BURVT and largest for  RUVT, although
             there were no non-detects of benzene measured at the Vermont sites. Note that the
             minimum benzene concentration across the program was measured at UNVT.
                                         25-32

-------
•  Figure 25-17 is the box plot for benzo(a)pyrene for UNVT. Note that the first
   quartile for this pollutant is zero and is not visible on the box plot. This box plot
   shows that the annual average concentration for UNVT is less than the program-
   level average and median concentrations. The maximum concentration measured
   at UNVT is considerably less than the maximum concentration measured across
   the program. Nearly 80 percent of the measurements at UNVT were non-detects.

•  Figure 25-18 for 1,3-butadiene also shows all three sites. Similar to the benzene
   box plots, the program-level maximum concentration (9.51 |ig/m3) is not shown
   directly on the box plots as the scale has been reduced to 3 |ig/m3 in order to
   allow for the observation of data points at the lower end of the concentration
   range. The annual average concentration for BURVT is similar to the annual
   average for RUVT, even though the range of measurements is higher for RUVT,
   and both are roughly equivalent to the program-level average concentration. The
   annual average for UNVT is an order of magnitude lower than the other two
   Vermont sites. The maximum concentration measured at each site is considerably
   less than the maximum 1,3-butadiene concentration measured across the program.
   Five non-detects of 1,3-butadiene were measured at RUVT and 80  percent of the
   measurements were non-detects for UNVT. Conversely, the minimum
   concentration of 1,3-butadiene measured at BURVT is equivalent to the program-
   level median concentration.

•  Figure 25-19 is the box plot for hexavalent chromium for UNVT. This figure
   shows that the annual average concentration of hexavalent chromium for UNVT
   is less than the program-level first quartile and is the lowest annual average
   hexavalent chromium concentration among NMP sites sampling this pollutant.
   The maximum concentration measured at UNVT is just greater than the program-
   level average concentration. Nearly 70 percent of the measurements of hexavalent
   chromium were non-detects.

•  Figure 25-20 is the box plot for lead (PMio) for UNVT. This figure shows that the
   annual average concentration of lead for UNVT is just less than the program-level
   first quartile and is the lowest annual average lead concentration among NMP
   sites sampling this pollutant. The maximum concentration measured at UNVT is
   the lowest maximum concentration among NMP sites sampling lead. The
   minimum concentration measured at UNVT is the lowest minimum concentration
   among NMP sites sampling lead.

•  Figure 25-21 is the box plot for manganese (PMio) for UNVT. Note that the
   program-level maximum concentration (395 ng/m3) is not shown directly on the
   box plot as the scale has been reduced to 200 ng/m3 in order to allow for the
   observation of data points at the lower end of the concentration range. This figure
   shows that, similar to other metals, the annual average concentration of
   manganese (PMio) for UNVT is less than the program-level first quartile. The
   annual average concentration of manganese for UNVT is the lowest annual
   average concentration among NMP sites sampling this pollutant. There were no
   non-detects of manganese measured at UNVT.
                              25-33

-------
          •  Figure 25-22 is the box plot for naphthalene for UNVT. Note that the program-
             level maximum concentration (799 ng/m3) is not shown directly on the box plot as
             the scale has been reduced to 500 ng/m3 in order to allow for the observation of
             data points at the lower end of the concentration range. The annual average for
             UNVT is less than the program-level first quartile and ranks lowest among all
             sites sampling this pollutant. The maximum naphthalene concentration measured
             at UNVT is less than the program-level third quartile and just greater than the
             program-level average concentration. It is also the lowest maximum concentration
             among NMP sites sampling naphthalene.


25.4.3  Concentration Trends

       A site-specific trends evaluation was completed for sites that have sampled one or more

of the selected NATTS MQO Core Analytes for 5 consecutive years or longer, as described in
Section 3.5.4. UNVT has sampled hexavalent chromium under the NMP since 2005. Thus,

Figure 25-23 presents the annual statistical metrics for hexavalent chromium for UNVT. The

statistical metrics presented for calculating trends include the substitution of zeros for non-

detects. Sampling of the other methods did not begin until 2008 and thus does not meet the
5-consecutive year criterion.
    Figure 25-23. Annual Statistical Metrics for Hexavalent Chromium Concentrations
                                  Measured at UNVT
               • 5th Percentile    —  Minimum    — Median    — Maximum
                                                           •  95th Percentile
                                                                         * Average
                                         25-34

-------
       Observations from Figure 25-23 for hexavalent chromium measurements at UNVT

include the following:

       •   The maximum hexavalent chromium concentration was measured at UNVT on
          June 16, 2006 (0.399 ng/m3). The next highest hexavalent chromium concentration
          was measured on April 22, 2005  (0.101 ng/m3). All other measurements of this
          pollutant are less than 0.1 ng/m3.

       •   For all time frames shown, the minimum, 5th percentile, and median concentrations
          are zero, indicating that at least 50 percent of the measurements are non-detects. The
          percentage of non-detects has varied over the years of sampling, from as low as
          63 percent in 2006 to as high as 95 percent in 2009.

       •   The second highest concentration measured in  2006 is an order of magnitude less
          than the maximum concentration measured that year.  The 95th percentile decreased by
          almost half from 2005 to 2006. Both are an indication that the maximum
          concentration is driving the average concentration measured at UNVT for 2006. If
          this measurement were removed from the calculations, Figure 25-23 would show a
          decreasing trend beginning with 2006.

       •   The average concentration decreased from 2006 to 2007 and again in 2008. Fewer
          than 10 percent of the samples collected in 2008 and 2009 had measurable levels of
          hexavalent chromium, which explains why even the average and 95th percentile are
          nearly zero for these years. The number of non-detects decreased for 2010 and 2011
          to between 70 and 80 percent, allowing the average concentration to increase. The
          average concentrations for 2010 and 2011 are around 0.004 ng/m3.


25.5   Additional Risk-Based Screening Evaluations

       The following risk-based screening evaluations were conducted to characterize risk at the

Vermont monitoring sites. Refer to Sections 3.3 and 3.5.5  for definitions and explanations

regarding the various toxicity factors, time frames, and calculations associated with these risk-

based screenings.


25.5.1  Risk-Based Screening Assessment Using MRLs
       A risk-based screening was conducted by comparing the concentration data from the

Vermont monitoring sites to the ATSDR MRLs, where available. As described in Section 3.3,

MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute

(exposures of 1 to 14 days); intermediate (exposures of 15 to 364 days); and chronic (exposures

of 1 year or greater). The preprocessed daily measurements of the pollutants of interest were

compared to  the acute MRLs; the quarterly averages were compared to the intermediate MRLs;

and the annual averages were compared to the chronic MRLs.
                                         25-35

-------
       As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2011.

25.5.2 Cancer Risk and Noncancer Hazard Approximations
       For the pollutants of interest for the Vermont 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 may want to shift or
confirm their air-monitoring priorities.  Refer to Section 3.5.5.2 for an explanation of how  cancer
risk and noncancer hazard approximations are calculated and what limitations are associated with
them. Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer
hazard approximations are presented in Table 25-6, where applicable. Cancer risk
approximations are presented as probabilities while the noncancer hazard approximations  are
ratios and thus, unitless values.

       Observations from Table 25-6 include the following:
       •  For BURVT, benzene and carbon tetrachloride have the highest annual average
          concentrations. These two pollutants also have the highest cancer risk approximations
          for BURVT (6.08 in-a-million and 3.52 in-a-million, respectively).
       •  Similar to BURVT, benzene and carbon tetrachloride have the highest annual average
          concentrations for RUVT. These two pollutants also have the highest cancer risk
          approximations for RUVT (6.94 in-a-million and 3.65 in-a-million, respectively).
       •  Carbon tetrachloride and benzene have the highest annual average concentrations for
          UNVT.  These two pollutants also have the highest cancer risk approximations  for
          UNVT (3.41 in-a-million and 3.25 in-a-million, respectively).
       •  The noncancer hazard approximations for the pollutants of interest for all three
          Vermont sites are all considerably less than 1.0, indicating that no adverse health
          effects are expected from these individual pollutants.
                                         25-36

-------
                 Table 25-6. Risk Approximations for the Vermont Monitoring Sites
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Burlington, Vermont - BURVT
Acrylonitrile
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
p-Dichlorobenzene
1 ,2-Dichloroethane
Ethylbenzene
Hexachloro- 1 , 3 -butadiene
1 , 1 ,2,2-Tetrachloroethane
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
0.000068
0.0000078
0.00003
0.000006

0.000011
0.000026
0.0000025
0.000022
0.000058
0.00000026
0.0000048
0.0000088
0.002
0.03
0.002
0.1
0.098
0.8
2.4
1
0.09

0.04
0.002
0.1
4/30
30/30
30/30
30/30
17/30
27/30
7/30
30/30
6/30
4/30
23/30
2/30
1/30
0.02
ฑ0.03
0.78
ฑ0.09
0.10
ฑ0.01
0.59
ฑ0.04
0.07
ฑ0.02
0.07
ฑ0.01
0.02
ฑ0.01
0.30
ฑ0.04
0.02
ฑ0.01
0.01
ฑ0.01
0.12
ฑ0.05
<0.01
ฑ<0.01
0.01
ฑ0.01
1.60
6.08
2.98
3.52

0.77
0.47
0.75
0.41
0.41
0.03
0.02
0.01
0.01
0.03
0.05
0.01
O.01
0.01
0.01
O.01
0.01

0.01
O.01
0.01
Rutland, Vermont - RUVT
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
1 ,2-Dichloroethane
Ethylbenzene
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
0.0000078
0.00003
0.000006

0.000026
0.0000025
0.00000026
0.0000048
0.0000088
0.03
0.002
0.1
0.098
2.4
1
0.04
0.002
0.1
30/30
25/30
30/30
9/30
4/30
30/30
23/30
2/30
0/30
0.89
ฑ0.16
0.09
ฑ0.03
0.61
ฑ0.04
0.04
ฑ0.03
0.01
ฑ0.01
0.30
ฑ0.04
0.12
ฑ0.04
0.01
ฑ0.01
O.01
ฑO.01
6.94
2.80
3.65

0.29
0.76
0.03
0.01
O.01
0.03
0.05
0.01
0.01
O.01
0.01
O.01
0.01
O.01
— = a Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 25-5.
                                                 25-37

-------
          Table 25-6. Risk Approximations for the Vermont Monitoring Sites (Continued)
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Underbill, Vermont - UNVT
Acrylonitrile
Arsenic (PM10)a
Benzene
Benzo(a)pyrene a
Beryllium (PM10) a
1,3 -Butadiene
Cadmium (PM10)a
Carbon Tetrachloride
Chloroform
1 ,2-Dichloroethane
Hexachloro- 1 ,3 -butadiene
Hexavalent Chromium a
Lead(PM10)a
Manganese (PM10)a
Naphthalene a
Nickel (PM10)a
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
0.000068
0.0043
0.0000078
0.00176
0.0024
0.00003
0.0018
0.000006

0.000026
0.000022
0.012


0.000034
0.00048
0.00000026
0.0000048
0.0000088
0.002
0.000015
0.03

0.00002
0.002
0.00001
0.1
0.098
2.4
0.09
0.0001
0.00015
0.00005
0.003
0.00009
0.04
0.002
0.1
6/60
54/57
60/60
13/60
45/57
13/60
57/57
60/60
17/60
12/60
4/60
17/60
57/57
57/57
60/60
57/57
19/60
1/60
1/60
0.01
ฑ0.01
0.01
ฑ<0.01
0.42
ฑ0.04
<0.01
ฑ<0.01
0.01
ฑ0.01
0.01
ฑ0.01
0.01
ฑ0.01
0.57
ฑ0.04
0.03
ฑ0.01
0.02
ฑ0.01
0.01
ฑ0.01
O.01
ฑO.01
0.01
ฑ0.01
O.01
ฑ0.01
0.01
ฑ0.01
O.01
ฑ0.01
0.02
ฑ0.01
O.01
ฑ0.01
0.01
ฑ0.01
0.41
1.09
3.25
0.03
0.01
0.21
0.12
3.41

0.40
0.14
0.05


0.45
0.25
0.01
O.01
0.01
O.01
0.02
0.01

0.01
O.01
0.01
0.01
0.01
O.01
0.01
O.01
0.01
0.04
0.01
0.01
0.01
O.01
0.01
— = a Cancer URE or Noncancer RfC is not available.
a For the annual average concentration of this pollutant in ng/m3, refer to Table 25-5.
                                                25-38

-------
25.5.3 Risk-Based Emissions Assessment
       In addition to the risk-based screenings discussed above, Tables 25-7 and 25-8 present an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 25-7 presents the 10 pollutants with the highest emissions from the 2008 NEI, the 10
pollutants with the highest toxicity-weighted emissions, and the 10 pollutants with the highest
cancer risk approximations (in-a-million) for the Vermont monitoring sites, as calculated from
the annual averages provided in Table 25-6. Table 25-8 presents similar information, but
identifies the 10 pollutants with the highest noncancer hazard approximations (HQ), also
calculated from annual averages provided in Table 25-6.

       The pollutants listed in Tables 25-7 and 25-8 are limited to those that have cancer and
noncancer toxicity factors, respectively. As a result, although the actual value of the emissions is
the same, the highest emitted pollutants in the cancer table may be different from the noncancer
table. The cancer risk and noncancer hazard approximations based on each site's annual averages
are limited to those pollutants for which each respective site sampled. As discussed in
Section 25.3, UNVT  sampled for VOCs, PAHs, metals (PMio), and hexavalent chromium;
BURVT  and RUVT sampled for VOCs only. In addition, the cancer risk and noncancer hazard
approximations are limited to those pollutants with enough data to meet the criteria for annual
averages to be calculated. A more in-depth discussion of this analysis is provided in
Section 3.5.5.3. Similar to the cancer risk and noncancer hazard approximations, this analysis
may help policy-makers prioritize their air monitoring activities.
                                          25-39

-------
Table 25-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for

                                              the Vermont Monitoring Sites
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Burlington, Vermont (Chittenden County) - BURVT
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Dichloromethane
Naphthalene
POM, Group 2b
POM, Group 6
POM, Group la
117.39
57.01
42.71
30.61
12.99
7.10
6.54
1.69
0.20
0.12
Benzene
Formaldehyde
1,3 -Butadiene
Hexavalent Chromium, PM
POM, Group 3
Arsenic, PM
Naphthalene
POM, Group 2b
POM, Group 5a
Ethylbenzene
9.16E-04
7.41E-04
3.90E-04
2.57E-04
2.46E-04
2.44E-04
2.22E-04
1.49E-04
1.24E-04
1.07E-04
Benzene
Carbon Tetrachloride
1,3 -Butadiene
Acrylonitrile
ฃ>-Dichlorobenzene
Ethylbenzene
1 ,2-Dichloroethane
1 , 1 ,2,2-Tetrachloroethane
Hexachloro- 1 , 3 -butadiene
Tetrachloroethylene
6.08
3.52
2.98
1.60
0.77
0.75
0.47
0.41
0.41
0.03
Underbill, Vermont (Chittenden County) - UNVT
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Dichloromethane
Naphthalene
POM, Group 2b
POM, Group 6
POM, Group la
117.39
57.01
42.71
30.61
12.99
7.10
6.54
1.69
0.20
0.12
Benzene
Formaldehyde
1,3 -Butadiene
Hexavalent Chromium, PM
POM, Group 3
Arsenic, PM
Naphthalene
POM, Group 2b
POM, Group 5a
Ethylbenzene
9.16E-04
7.41E-04
3.90E-04
2.57E-04
2.46E-04
2.44E-04
2.22E-04
1.49E-04
1.24E-04
1.07E-04
Carbon Tetrachloride
Benzene
Arsenic
Naphthalene
Acrylonitrile
1 ,2-Dichloroethane
Nickel
1,3 -Butadiene
Hexachloro- 1 , 3 -butadiene
Cadmium
3.41
3.25
1.09
0.45
0.41
0.40
0.25
0.21
0.14
0.12
to
v\

Jฑ
o

-------
Table 25-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
                                        the Vermont Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Rutland, Vermont (Rutland County) - RUVT
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
POM, Group 2b
Dichloromethane
POM, Group 6
POM, Group la
54.68
24.42
20.01
14.97
6.47
3.14
0.93
0.47
0.12
0.06
Benzene
Formaldehyde
1,3 -Butadiene
POM, Group 3
Hexavalent Chromium, PM
Naphthalene
POM, Group 2b
POM, Group 5a
Ethylbenzene
Acetaldehyde
4.26E-04
3.17E-04
1.94E-04
1.38E-04
1.24E-04
1.07E-04
8.17E-05
6.57E-05
5.00E-05
3.29E-05
Benzene
Carbon Tetrachloride
1,3 -Butadiene
Ethylbenzene
1 ,2-Dichloroethane
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
6.94
3.65
2.80
0.76
0.29
0.03
0.01
0.00

to
-k

-------
Table 25-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer

                                          RfCs for the Vermont Monitoring Sites
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations (Site-
Specific)
Pollutant
Noncancer Hazard
Approximation
(HQ)
Burlington, Vermont (Chittenden County) - BURVT
Toluene
Xylenes
Benzene
Methanol
Formaldehyde
Ethylbenzene
Hexane
Hydrochloric acid
Acetaldehyde
1,3 -Butadiene
221.14
170.08
117.39
88.67
57.01
42.71
41.71
41.61
30.61
12.99
Acrolein
Manganese, PM
Chlorine
1,3 -Butadiene
Formaldehyde
Benzene
Arsenic, PM
Acetaldehyde
Cyanide Compounds, gas
Naphthalene
552,694.88
67,632.70
13,450.30
6,493.30
5,817.27
3,912.93
3,782.88
3,400.90
2,550.52
2,180.85
1,3 -Butadiene
Benzene
Acrylonitrile
Carbon Tetrachloride
Tetrachloroethylene
Trichloroethylene
Chloroform
Ethylbenzene
Hexachloro- 1 , 3 -butadiene
ฃ>-Dichlorobenzene
0.05
0.03
0.01
0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
Underbill, Vermont (Chittenden County) - UNVT
Toluene
Xylenes
Benzene
Methanol
Formaldehyde
Ethylbenzene
Hexane
Hydrochloric acid
Acetaldehyde
1,3 -Butadiene
221.14
170.08
117.39
88.67
57.01
42.71
41.71
41.61
30.61
12.99
Acrolein
Manganese, PM
Chlorine
1,3 -Butadiene
Formaldehyde
Benzene
Arsenic, PM
Acetaldehyde
Cyanide Compounds, gas
Naphthalene
552,694.88
67,632.70
13,450.30
6,493.30
5,817.27
3,912.93
3,782.88
3,400.90
2,550.52
2,180.85
Manganese
Arsenic
Benzene
Lead
Cadmium
Nickel
Carbon Tetrachloride
Naphthalene
1,3 -Butadiene
Acrylonitrile
0.04
0.02
0.01
0.01
0.01
0.01
0.01
<0.01
<0.01
<0.01
to
v\

Jฑ
to

-------
Table 25-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
                                    RfCs for the Vermont Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer Hazard
Approximation
(HQ)
Rutland, Vermont (Rutland County) - RUVT
Toluene
Xylenes
Benzene
Methanol
Formaldehyde
Ethylbenzene
Hexane
Acetaldehyde
1,3 -Butadiene
Styrene
121.10
77.74
54.68
37.29
24.42
20.01
17.83
14.97
6.47
5.25
Acrolein
1,3 -Butadiene
Formaldehyde
Cyanide Compounds, gas
Benzene
Acetaldehyde
Naphthalene
Xylenes
Arsenic, PM
Lead, PM
74,401.74
3,234.62
2,492.15
2,311.36
1,822.62
1,663.70
1,045.42
777.44
462.98
399.93
1,3 -Butadiene
Benzene
Carbon Tetrachloride
Tetrachloroethylene
Trichloroethylene
Chloroform
Ethylbenzene
1 ,2-Dichloroethane
Vinyl Chloride
0.05
0.03
0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01

to

-------
Observations from Table 25-7 include the following:

•  Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
   cancer UREs in Chittenden and Rutland Counties, although the emissions in
   Chittenden County were nearly twice those in Rutland County.

•  Benzene, formaldehyde, and 1.3-butadiene are the pollutants with the highest
   toxi city-weighted emissions (of the pollutants with cancer UREs) for both counties.

•  Six of the highest emitted pollutants also have the highest toxi city-weighted
   emissions for Chittenden County while seven of the highest emitted pollutants also
   have the highest toxicity-weighted emissions for Rutland County.

•  Benzene and carbon tetrachloride have the highest cancer risk approximations for all
   three sites. Benzene topped both emissions-based lists for both counties, while carbon
   tetrachloride appeared on neither emissions-based list for either county.
   1,3-Butadiene also appears on all three lists for each site. Ethylbenzene appears on all
   three lists for BURVT and RUVT but is not a pollutant of interest for UNVT.
   Naphthalene also appears on all three lists for UNVT.

•  Arsenic has the third highest cancer risk approximation and ranks sixth for toxicity-
   weighted emissions, but is not one of the highest emitted.

•  Benzo(a)pyrene is part of POM, Group 5a and is one of UNVT's pollutants of
   interest. POM, Group 5a has the ninth highest toxicity-weighted emissions but is not
   among the highest emitted in Chittenden County.

•  POM, Group 2b ranks eighth for both quantity emitted and its toxicity-weighted
   emissions in Chittenden County. POM, Group 2b includes several PAHs sampled for
   at UNVT including acenaphthylene, fluoranthene, and perylene. None of the PAHs
   included in POM, Group 2b failed screens for UNVT. POM, Groups la, 3, and 6 also
   appear in Table 25-7, but only POM, Group 6 includes PAHs sampled for at UNVT
   (benzo(a)anthracene, for example), but none of these pollutants failed screens.


Observations from Table 25-8 include the following:

•  Toluene, xylenes, and benzene are the highest emitted pollutants with noncancer
   RfCs in Chittenden and Rutland Counties, although the emissions in Chittenden
   County were nearly twice those in Rutland County.

•  Acrolein is the pollutant with the highest toxicity-weighted emissions (of the
   pollutants with noncancer RfCs) for both Chittenden and Rutland Counties. Although
   acrolein was sampled for at all three sites, this pollutant was excluded from the
   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.
                                   25-44

-------
       •  Four of the highest emitted pollutants for Chittenden County also have the highest
          toxi city-weighted emissions while five of the highest emitted pollutants for Rutland
          County also have the highest toxicity-weighted emissions.

       •  Although very low, 1,3-butadiene and benzene have the highest noncancer hazard
          approximations for BURVT and RUVT. Benzene and 1,3-butadiene appear on both
          emissions-based lists.

       •  Although very low, manganese and arsenic have the highest noncancer hazard
          approximations for UNVT. While these pollutants rank second and seventh among
          the toxicity-weighted emissions for Chittenden County, respectively, neither pollutant
          appears among the highest emitted.


25.6   Summary of the 2011 Monitoring Data for the Vermont Monitoring Sites

       Results from several of the data treatments described in this section include the
following:

       ปซป  A total of 11 pollutants failed screens for BURVT; seven pollutants failed screens for
          RUVT; and 13 pollutants failed screens for UNVT.

       ปซป  None of the annual average concentrations of the pollutants of interest for the
          Vermont monitoring sites were greater than 1 jug/m3.

       ปซป  The annual average concentrations for several of UNVT's pollutants of interest were
          the lowest annual averages among allNMP sites sampling those pollutants.
                                         25-45

-------
26.0   Site in Virginia
       This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS site in Virginia, and integrates these concentrations with
emissions, meteorological, and risk information. Data generated by sources other than ERG are
not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) for detailed discussions and definitions
regarding the various data analyses presented below.

26.1   Site Characterization
       This section characterizes the Virginia monitoring site by providing geographical and
physical information about the location of the site and the surrounding area. This information is
provided to give the reader insight regarding factors that may influence the air quality near the
site and assist in the interpretation of the ambient monitoring measurements.

       The RIVA monitoring site is located just outside the Richmond, Virginia city limits.
Figure 26-1 is a composite satellite image retrieved from ArcGIS Explorer showing the
monitoring site in its urban  location. Figure 26-2 identifies nearby point source emissions
locations by source category, as reported in the 2008 NEI for point sources. Note that only
sources within 10 miles of the site are included in the facility counts provided in Figure 26-2. A
10-mile boundary was chosen to give the reader an indication of which emissions sources and
emissions source categories could potentially have a direct effect on the air quality  at the
monitoring site. Further, this boundary provides both the proximity of emissions sources to the
monitoring site as well as the quantity of such sources within a  given distance of the site. Sources
outside the 10-mile radius are still visible on the map, but have  been grayed out in order to show
emissions sources just outside the boundary. Table 26-1 provides supplemental geographical
information such as land use, location setting, and locational coordinates.
                                           26-1

-------
                                      Figure 26-1. Richmond, Virginia (RIVA) Monitoring Site
to
ON

to

-------
      Figure 26-2. NEI Point Sources Located Within 10 Miles of RIVA
              Legend
RIVA NATTS site
                                          7r3Q'fl"W         77c25'
-------
                                Table 26-1. Geographical Information for the Virginia Monitoring Site

Site
Code

RIVA



AQS Code

51-087-0014



Location

Richmond



County

Henrico

Micro- or
Metropolitan
Statistical Area

Richmond, VA
MSA

Latitude
and
Longitude

37.55655,
-77.400411



Land Use

Residential


Location
Setting

Suburban



Additional Ambient Monitoring Information1
TSP Metals, SO2, NOy, NO, NO2, NOX, PAMS,
NMOCs, VOCs, Carbonyl compounds, O3,
Meteorological parameters, PM10, PM10 Metals,
PM Coarse, PM2 5, and PM2 5 Speciation, CO,
Tetrahydrofuran.
  BOLD ITALICS = EPA-designated NATTS Site
to
-k

-------
       The RIVA monitoring site is located just northeast of the capital city of Richmond, in
east-central Virginia. The site is located at the MathScience Innovation Center in a residential
area less than 1/4 mile from 1-64. The 1-64 interchange with Mechanicsville Turnpike (360) is
less than 1/2 mile southwest of the site, as shown in Figure 26-1. Beyond the residential areas
surrounding the school property are a golf course to the southeast, a high school to the south (on
the southside of 1-64), and commercial areas to the west. As Figure 26-2 shows, RIVA is located
near several point sources, most of which are located to the southeast and south of the site and
within the city of Richmond. The sources closest to RIVA are a fabricated metal products facility
and a heliport at the Medical College of Virginia. The source categories with the greatest number
of emissions sources within 10 miles of RIVA are aircraft operations, which include airports as
well as small runways, heliports, or landing pads; printing and publishing facilities; bulk
terminals and bulk plants; and facilities generating electricity via combustion.

       Table 26-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Virginia monitoring site. Table 26-2 includes county-level
population and vehicle registration information. Table 26-2 also includes a county-level vehicle
registration-to-population ratio, which was calculated to represent the number of vehicles per
person within the monitoring site's residing county. In addition, the population within  10 miles
of the site is presented, based on postal code population data estimates. An estimate of 10-mile
vehicle ownership was determined by applying the county-level vehicle registration-to-
population ratio to the 10-mile population surrounding the monitoring site. Table 26-2 also
contains traffic volume information for RIVA. Finally, Table 26-2 presents the county-level
daily VMT for Henrico County.

      Table  26-2. Population, Motor Vehicle, and Traffic Information for the Virginia
                                      Monitoring  Site



Site
RIVA

Estimated
County
Population1
310,445

County-level
Vehicle
Registration2
354,721
Vehicles per
Person
(Registration:
Population)
1.14

Population
within 10
miles3
476,219
Estimated
10-mile
Vehicle
Ownership
544,138
Annual
Average
Daily
Traffic4
73,000

County-
level Daily
VMT5
8,246,774
Bounty-level population estimate reflects 2011 data from the U.S. Census Bureau (Census Bureau, 2012b)
2County-level vehicle registration reflects 2011 data from the Revenue Division of the County of Henrico (Henrico
 County, 2012)
310-mile population estimate reflects 2011 data from Xionetic (Xionetic, 2011)
4AADT reflects 2011 data from the Virginia DOT (VA DOT, 2011)
5County-level VMT reflects 2011 data from the Virginia DOT (VA DOT, 2012)
BOLD ITALICS = EPA-designated NATTS Site
                                            26-5

-------
       Observations from Table 26-2 include the following:
       •  RIVA's county-level population is in the lower third of the range compared to other
          counties with NMP sites while its 10-mile population is in the middle of the range
          among NMP sites.
       •  The county-level and 10-mile vehicle ownerships are in the middle of the range
          compared to other NMP sites.
       •  The vehicle-per-person ratio is among the higher ratios compared to other NMP sites.
       •  The traffic volume experienced near RIVA is in the middle of the range compared to
          other NMP monitoring sites. The traffic volume provided is for the interchange of
          US-360 (Mechanicsville Turnpike) and 1-64.
       •  The daily VMT for Henrico County is in the middle of the range compared to other
          counties with NMP sites (where VMT data are available).

26.2   Meteorological Characterization
       The following sections characterize the meteorological conditions near the monitoring
site in Virginia on sample days, as well as over the course of the year.
26.2.1  Climate Summary
       The city of Richmond is located in east-central Virginia, east of the Blue Ridge
Mountains and west of the Chesapeake Bay. The James River flows through the west, center, and
south parts of town. Richmond has a modified continental climate. Winters tend to be mild, as
the mountains act as a barrier to cold air and the proximity to the Atlantic Ocean prevents
temperatures from plummeting too low.  Summers are warm and humid, also due to these
influences. Precipitation is well distributed throughout the year (Bair, 1992).

26.2.2  Meteorological Conditions in 2011
       Hourly meteorological data from the NWS weather station nearest this site were retrieved
for 2011 (NCDC, 2011). The closest weather station is located at Richmond International Airport
(WBAN 13740). Additional information about the Richmond International Airport weather
station, such as the distance between the site and the weather station, is provided in Table 26-3.
These data were used to determine how meteorological conditions on sample days vary from
conditions experienced throughout the year.
                                          26-6

-------
                           Table 26-3. Average Meteorological Conditions near the Virginia Monitoring Site
Closest NWS Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(ฐF)
Average
Temperature
(ฐF)
Average
Dew Point
Temperature
(ฐF)
Average
Wet Bulb
Temperature
(ฐF)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Richmond, Virginia - RTVA
Richmond
International Airport
13740
(37.51, -77.32)
5.16
miles
118ฐ
(ESE)
Sample
Day
2011
69.8
ฑ4.2
70.6
+ 1.7
60.0
ฑ4.0
60.6
+ 1.7
47.2
ฑ4.8
47.9
+ 1.8
53.6
ฑ3.9
54.1
+ 1.6
66.8
ฑ3.9
66.7
+ 1.4
1016.9
ฑ1.7
1017.0
+ 0.7
6.4
ฑ0.8
6.2
+ 0.3
       1 Sample day averages are highlighted to help differentiate the sample day averages from the full-year averages.
to

-------
       Table 26-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for the entire year. Also included in Table 26-3 is the 95 percent
confidence interval for each parameter. As shown in Table 26-3, average meteorological
conditions on sample days were representative of average weather conditions throughout the
year.

26.2.3 Back Trajectory Analysis
       Figure 26-3 is the composite back trajectory map for days on which samples were
collected at the RIVA monitoring site in 2011. Included in Figure 26-3 are four back trajectories
per sample day. Figure 26-4 is the corresponding cluster analysis. An in-depth description of
these maps and how they were generated is presented in Section 3.5.2.1. For the composite map,
each line represents the 24-hour trajectory along which a parcel of air traveled toward the
monitoring site on a given sample day and time, based on an initial height of 50 meters AGL.  For
the cluster analysis, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric circle around the site in Figures 26-3 and 26-4 represents 100 miles.

       Observations from Figures 26-3 and Figure 26-4 for RIVA include the following:
       •  Back trajectories originated from a variety of directions near RIVA, although a large
          number of them originated to the northwest.
       •  The 24-hour air shed domain for RIVA was similar in size to many other NMP
          monitoring sites. The farthest away a trajectory originated was over the Upper
          Peninsula of Michigan, or over 700 miles away. However, the average trajectory
          length is 242 miles and most back trajectories (87 percent) originated within
          400 miles of the site.
       •  The cluster analysis shows that 28 percent of back trajectories originated from the
          northwest of RIVA over Michigan, Ohio, and Indiana. Seventeen percent originated
          from the southwest to west and includes back trajectories of varying lengths. The
          cluster trajectory originating over eastern North Carolina (36 percent) represents back
          trajectories originating from the east, southeast, and south, primarily over eastern
          North Carolina but also over the offshore waters of Virginia, North Carolina, and
           South Carolina. Another  19 percent of back trajectories originated from the north-
          northwest, north, and northeast of the site.
                                          26-8

-------
Figure 26-3. 2011 Composite Back Trajectory Map for RIVA
    Figure 26-4. Back Trajectory Cluster Map for RIVA
                          26-9

-------
26.2.4 Wind Rose Comparison
       Hourly surface wind data from the NWS weather station at Richmond International
Airport near RIVA were uploaded into a wind rose software program to produce customized
wind roses, as described in Section 3.5.2.2. A wind rose shows the frequency of wind directions
using "petals" positioned around a 16-point compass, and uses different colors to represent wind
speeds.

       Figure 26-5 presents a map showing the distance between the NWS station and RIVA,
which may be useful for identifying topographical influences that may affect the meteorological
patterns experienced at this location. Figure 26-5 also presents three different wind roses for the
RIVA monitoring site. First, a historical wind rose representing 2001 to 2010 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 2011 is presented.
Next, a wind rose representing wind data for days on which samples were collected in 2011 is
presented. These can be used to identify the predominant wind speed and direction for 2011 and
determine if wind observations on sample days were representative of conditions experienced
over the entire year and historically.

       Observations from Figure 26-5 for RIVA include the following:
       •  The Richmond International weather station is located approximately 5.1 miles
          east-southeast of RIVA.
       •  The historical wind rose shows that the most commonly observed wind direction is
          north, although winds from the north-northeast, south, south-southwest, and
          southwest were also frequently observed. Winds from the southeast quadrant were
          observed the least. Calm winds (< 2 knots) were observed for approximately
          15 percent of the hourly wind measurements.
       •  The 2011 wind rose resembles the historical wind rose in some ways but exhibits
          differences as well. Southerly and south-southwesterly winds were more prominent in
          2011, although northerly winds were still frequently observed.
       •  Northerly winds prevailed on sample days near RIVA, although southerly winds still
          accounted for greater than 10 percent of the wind observations. A higher percentage
          of winds from the northwest quadrant were observed on sample days compared to the
          entire year and historically. There were also fewer wind observations from the
          southwest quadrant on sample days compared to the entire year and historically.
                                         26-10

-------
Figure 26-5. Wind Roses for the Richmond International Airport Weather Station near
                                    RIVA
 Distance between RIVA and NWS Station
2001-2010 Historical Wind Rose
              v
                                :\
          2011 Wind Rose
    Sample Day Wind Rose
                                    26-11

-------
26.3   Pollutants of Interest
       Site-specific "pollutants of interest" were determined for the Virginia monitoring site in
order to allow 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." 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. In addition, if any
of the NATTS MQO Core Analytes measured by the monitoring site did not meet the pollutant
of interest criteria based on the preliminary risk-based screening, that pollutant was added to the
list of site-specific pollutants of interest. A more in-depth description of the risk-based screening
process is presented in Section 3.2.

       Table 26-4 presents the results of the preliminary risk-based screening process for RIVA.
The pollutants that failed at least one screen and contributed to 95 percent of the total failed
screens for the monitoring site are shaded. NATTS MQO Core Analytes are bolded. Thus,
pollutants of interest are shaded and/or bolded. RIVA sampled for PAHs and hexavalent
chromium.

         Table 26-4. Risk-Based Screening Results for the Virginia Monitoring Site
Pollutant
Screening
Value
(Hg/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
Richmond, Virginia - RIVA
Naphthalene
Acenaphthene
Benzo(a)pyrene
Fluorene
0.029
0.011
0.00057
0.011
Total
54
1
1
1
57
61
61
30
61
213
88.52
1.64
3.33
1.64
26.76
94.74
1.75
1.75
1.75
94.74
96.49
98.25
100.00

       Observations from Table 26-4 include the following:
       •  Although four PAHs failed screens for RIVA, naphthalene contributed to roughly
          95 percent of the total failed screens, while the other pollutants accounted for a single
          failed screen each.
       •  The risk-based screening process identified all four PAHs failing screens as the
          pollutants of interest for RIVA. This is because two of the pollutants contribute to the
          minimum 95 percent criteria discussed in Section 3.2, but because benzo(a)pyrene
          and fluorene contributed equally to the number of failed screens as acenaphthene,
                                          26-12

-------
          these pollutants were also designated as pollutants of interest. Hexavalent chromium
          was also added to the pollutants of interest for RIVA because it is a NATTS MQO
          Core Analyte, even though it did not fail any screens. This pollutant is not shown in
          Table 26-4 but is shown in subsequent tables in the sections that follow.

26.4   Concentrations
       This section presents various concentration  averages used to characterize pollution levels
at the Virginia monitoring site. Where applicable, the following calculations and data analyses
were performed: Time period-based concentration averages (quarterly  and annual) are provided
for the pollutants of interest for RIVA, where the data meet the applicable criteria. Concentration
averages for select pollutants are also presented graphically to illustrate how the site's
concentrations compare to the program-level averages, as presented in Section 4.1. In addition,
concentration averages for select pollutants are presented from previous years of sampling in
order to characterize concentration trends at the site. Additional site-specific statistical
summaries for RIVA are provided in Appendices M and O.

26.4.1 2011 Concentration Averages
       Quarterly and annual concentration averages were calculated for the pollutants of interest
for RIVA, as described in  Section 3.1. The quarterly average of a particular pollutant is simply
the average concentration of the preprocessed daily measurements over a given calendar quarter.
Quarterly average concentrations include the substitution of zeros for all  non-detects. A site must
have a minimum of 75 percent valid samples of the total number of samples possible within a
given quarter for a quarterly average to be calculated. An annual average includes all measured
detections and substituted  zeros for non-detects for the entire year of sampling. Annual averages
were calculated for pollutants where three  valid quarterly averages could be calculated and
where method completeness was greater than or equal to 85 percent, as presented in Section 2.4.
Quarterly and annual average concentrations for the Virginia monitoring site are presented in
Table 26-5, 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.
                                          26-13

-------
Table 26-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
                             the Virginia Monitoring Site



Pollutant
#of
Measured
Detections
vs. # of
Samples

1st
Quarter
Average
(ng/m3)

2nd
Quarter
Average
(ng/m3)

3rd
Quarter
Average
(ng/m3)

4th
Quarter
Average
(ng/m3)


Annual
Average
(ng/m3)
Richmond, Virginia - RJVA

Acenaphthene

Benzo(a)pyrene

Fluorene

Hexavalent Chromium

Naphthalene

61/61

30/61

61/61

48/61

61/61
1.17
ฑ0.21
0.12
ฑ0.07
2.18
ฑ0.34
0.01
ฑ<0.01
72.80
ฑ 20.97
4.53
ฑ1.59
0.01
ฑ0.01
4.83
ฑ1.57
0.01
ฑ<0.01
76.51
ฑ 22.49
4.66
ฑ1.14
0.08
ฑ0.15
4.92
ฑ1.11
0.02
ฑ0.01
70.33
ฑ 15.69
2.05
ฑ0.79
0.10
ฑ0.06
2.69
ฑ0.66
0.01
ฑ<0.01
93.27
ฑ 26.43
3.13
ฑ0.64
0.08
ฑ0.04
3.68
ฑ0.58
0.01
ฑ<0.01
78.10
ฑ 10.38
      Observations for RIVA from Table 26-5 include the following:

      •   The annual average concentration of naphthalene is significantly higher than the
          annual average concentrations of the remaining pollutants of interest.

      •   The quarterly averages of naphthalene have a relatively high-level of variability
          associated with them, as indicated by the confidence intervals. The maximum
          concentration measured at RIVA was measured on May 9, 2011 (175 ng/m3),
          although 17 concentrations greater than or equal to 100 ng/m3 were measured at this
          site. Six of these were measured during the fourth quarter of 2011, while the other
          quarters have three or four each.

      •   The quarterly averages of acenaphthene and fluorene are higher in the warmer months
          of the year and lower in the colder months of the year.  The maximum concentrations
          of both of these pollutants were measured on June 8, 2011. Of the 26 acenaphthene
          concentrations greater than 3 ng/m3, 11 were measured in the second quarter of 2011
          and 12 were measured during the third quarter of 2011. Conversely, all but two of the
          11 concentrations less than 1 ng/m3 were measured in the first or fourth quarters of
          2011. A similar pattern is shown in the concentrations  of fluorene.

      •   The maximum benzo(a)pyrene concentration was measured  at RIVA on
          August 16, 2011 (1.14 ng/m3) and is one of only seven concentrations greater than
          1  ng/m3 for this PAH measured across the program. The second highest concentration
          of this pollutant measured at RIVA was half as high (0.515 ng/m3) and was measured
          in March. Aside from the August 16th measurement, concentrations of
          benzo(a)pyrene tended to be higher during the colder months of the year. Of the
          30 measured detections of this pollutant,  12 were measured during the first quarter,
          three were measured in the second, five were measured in the third, and 10 were
          measured in the fourth. Conversely, of the 31  non-detects, three were measured
          during the first quarter, 12 were measured in the second, 11 were measured in the
          third, and five were measured in the fourth.
                                        26-14

-------
       •  Hexavalent chromium concentrations ranged from 0.0011 ng/m3 to 0.0466 ng/m3,
          with the five highest concentrations all measured during the third quarter of 2011.

26.4.2 Concentration Comparison
       In order to better illustrate how a site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for the selected NATTS MQO Core
Analytes listed in Section 3.5.3, where applicable. Thus, box plots for benzo(a)pyrene,
hexavalent chromium, and naphthalene were created for RIVA. Figures 26-6 through 26-8
overlay the site's minimum, annual average, and maximum concentrations onto the program-
level minimum, first quartile, median, average, third quartile, and maximum concentrations, as
described in Section 3.5.3.
      Figure 26-6. Program vs. Site-Specific Average Benzo(a)pyrene Concentration
                                  0.75         1         1.25
                                         Concentration (ng/m3)
Program
Site:
: 1st Quartile
Site Average
o
2nd Quartile 3rd Quartile 4th Quartile Ave
Site Minimum/Maximum

'rage

Figure 26-7. Program vs. Site-Specific Average Hexavalent Chromium Concentration
                                             0.15
                                         Concentration (ng/m3)
               Program:   1st Quartile    2nd Quartile    3rd Quartile     4th Quartile    Average
               Site:     Site Average    Site Minimum/Maximum

                                          26-15

-------
        Figure 26-8. Program vs. Site-Specific Average Naphthalene Concentration
 F.lvA
                                                         Program Max Concentration = 779 ng/m3
            50
                    100
                            150
                                    200       150       300
                                        Concentration {ng/m3)
                                                            = 50
                                                                            450
                                                                                    555
Program
Site:
: IstQuartile
Site Average
O
2ndQuartile SrdQuartile 4thQuartile AVE
Site Minimum/Maximum
rage

       Observations from Figures 26-6 through 26-8 include the following:

          •  Figure 26-6 is the box plot for benzo(a)pyrene. Note that the first quartile for this
             pollutant is zero and is not visible on this box plot. This box plot shows that the
             annual average concentration for RIVA is just less than the program-level average
             concentration. Figure 26-6 also shows that the maximum concentration measured
             at RIVA is less than the maximum concentration measured across the program.
             Half of the measurements of benzo(a)pyrene were non-detects.

          •  Figure 26-7 is the box plot for hexavalent chromium. Figure 26-7  shows that the
             annual average concentration of hexavalent chromium for RIVA is less than both
             the program-level average and median concentrations. This site has one of the
             lowest annual average concentrations of hexavalent chromium among NMP sites
             sampling this pollutant. The maximum concentration measured at RIVA is
             considerably less than the maximum concentration measured across the program.
             Several non-detects of hexavalent chromium were measured at RIVA.

          •  Figure 26-8 is the box plot for naphthalene. Note that the program-level
             maximum concentration (799 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
             500 ng/m3. Figure 26-8 shows that the annual average concentration of
             naphthalene for RIVA is just less than the program-level average concentration.
             The maximum naphthalene concentration measured at RIVA is considerably less
             than the program-level maximum concentration. There were no non-detects of
             naphthalene measured at RIVA or across the program.


26.4.3  Concentration Trends

       A site-specific trends evaluation was completed for sites that have sampled one or more

of the selected NATTS MQO Core Analytes for 5 consecutive years or longer, as described in

Section 3.5.4. RIVA did not begin sampling PAHs or hexavalent chromium under the NMP until

October 2008; therefore, the trends analysis was not conducted.
                                         26-16

-------
26.5   Additional Risk-Based Screening Evaluations
       The following risk-based screening evaluations were conducted to characterize risk at the
RIVA monitoring site. Refer to Sections 3.3 and 3.5.5 for definitions and explanations regarding
the various toxicity factors, time frames, and calculations associated with these risk-based
screenings.

26.5.1  Risk-Based Screening Assessment Using MRLs
       A risk-based screening was conducted by comparing the concentration data from the
Virginia monitoring site to the ATSDR MRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 to 14 days); intermediate (exposures of 15 to 364 days); and chronic (exposures
of 1 year or greater). The preprocessed daily measurements of the pollutants of interest were
compared  to the acute MRLs; the quarterly averages were compared to the intermediate MRLs;
and the annual averages were compared to the chronic MRLs.

       As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2011.

26.5.2  Cancer Risk and Noncancer Hazard Approximations
       For the pollutants of interest for RIVA and where annual average concentrations could
be calculated, risk was examined by calculating cancer risk and noncancer hazard
approximations. These approximations can be used as risk estimates for cancer and noncancer
effects attributable to the pollutants of interest. Although the use of these approximations is
limited, they may help identify where policy-makers may want to shift or confirm their air-
monitoring priorities. Refer to Section 3.5.5.2 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 26-6, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
                                         26-17

-------
             Table 26-6. Risk Approximations for the Virginia Monitoring Site
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Richmond, Virginia - RJVA
Acenaphthene
Benzo(a)pyrene
Fluorene
Hexavalent Chromium
Naphthalene
0.000088
0.00176
0.000088
0.012
0.000034



0.0001
0.003
61/61
30/61
61/61
48/61
61/61
3.13
ฑ0.64
0.08
ฑ0.04
3.68
ฑ0.58
0.01
ฑ0.01
78.10
ฑ 10.38
0.28
0.14
0.32
0.15
2.66



0.01
0.03
— = a Cancer URE or Noncancer RfC is not available

        Observations for RIVA from Table 26-6 include the following:
        •  The pollutant with the highest annual average concentration for RIVA is naphthalene,
           followed by fluorene and acenaphthene, although the annual average for naphthalene
           is significantly higher than the other annual average concentrations.
        •  The cancer risk approximation for naphthalene is 2.66 in-a-million. The cancer risk
           approximations for the remaining pollutants of interest are less than 1.0 in-a-million.
        •  Only two of the pollutants of interest for RIVA have noncancer toxicity factors. The
           noncancer hazard approximations for hexavalent chromium and naphthalene are
           considerably less than 1.0, indicating that no adverse health effects are expected from
           these individual pollutants.

 26.5.3  Risk-Based Emissions Assessment
        In addition to the risk-based screenings discussed above, Tables 26-7 and 26-8 present an
 evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
 Table 26-7 presents the 10 pollutants with the highest emissions from the 2008 NEI, the
 10 pollutants with the highest toxicity-weighted emissions, and the 10  pollutants with the highest
 cancer  risk approximations (in-a-million), as calculated from the annual averages provided in
 Table 26-6. Table 26-8 presents similar information,  but identifies the  10 pollutants with the
 highest noncancer hazard approximations (HQ), also calculated from annual averages provided
 in Table 26-6.
                                           26-18

-------
  Table 26-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
                                                  the Virginia Monitoring Site
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Richmond, Virginia (Henrico County) - RTVA
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
Dichloromethane
POM, Group 2b
Tetrachloroethylene
POM, Group la
115.26
90.51
64.92
60.31
19.67
9.77
2.81
1.91
0.38
0.19
Formaldehyde
Benzene
1,3 -Butadiene
Naphthalene
POM, Group 3
Hexavalent Chromium, PM
POM, Group 2b
Ethylbenzene
Acetaldehyde
Arsenic, PM
1.18E-03
8.99E-04
5.90E-04
3.32E-04
2.60E-04
2.50E-04
1.68E-04
1.62E-04
1.33E-04
6.94E-05
Naphthalene
Fluorene
Acenaphthene
Hexavalent Chromium
Benzo(a)pyrene
2.66
0.32
0.28
0.15
0.14

to

-------
  Table 26-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer

                                             RfCs for the Virginia Monitoring Site
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations (Site-
Specific)
Noncancer Hazard
Approximation
Pollutant (HQ)
Richmond, Virginia (Henrico County) - RTVA
Toluene
Xylenes
Methanol
Benzene
Formaldehyde
Hexane
Ethylbenzene
Acetaldehyde
Ethylene glycol
1,3 -Butadiene
752.59
248.81
175.23
115.26
90.51
75.32
64.92
60.31
21.10
19.67
Acrolein
1,3 -Butadiene
Formaldehyde
Acetaldehyde
Benzene
Naphthalene
Xylenes
Arsenic, PM
Lead, PM
Propionaldehyde
216,262.42
9,835.00
9,235.99
6,700.91
3,842.01
3,257.58
2,488.08
1,076.66
792.69
689.88
Naphthalene 0.03
Hexavalent Chromium O.01

to
to
o

-------
       The pollutants listed in Tables 26-7 and 26-8 are limited to those that have cancer and

noncancer toxicity factors, respectively. As a result, although the actual value of the emissions is

the same, the highest emitted pollutants in the cancer table may be different from the noncancer

table. The cancer risk and noncancer hazard approximations based on each site's annual averages

are limited to those pollutants for which each respective site sampled. As discussed in
Section 26.3, RIVA sampled for PAHs and hexavalent chromium.  In addition, the cancer risk

and noncancer hazard approximations are limited to those pollutants with enough data to meet

the criteria for annual averages to be calculated. A more in-depth discussion of this analysis is
provided in Section 3.5.5.3. Similar to the cancer risk and noncancer hazard approximations, this

analysis may help policy-makers prioritize their air monitoring activities.


       Observations from Table 26-7 include the following:

       •  Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
          cancer UREs in Henrico County.

       •  The pollutants with the highest toxicity-weighted emissions (of the pollutants with
          cancer UREs) are formaldehyde, benzene, and 1,3-butadiene.

       •  Seven of the highest emitted pollutants also have the highest toxicity-weighted
          emissions for Henrico County.

       •  Naphthalene, which is the pollutant with the highest cancer risk approximation for
          RIVA, has the sixth highest emissions and the fourth highest toxicity-weighted
          emissions for Henrico County.

       •  POM, Group 2b is the eighth highest emitted "pollutant" in Henrico County and ranks
          seventh for toxicity-weighted emissions. POM, Group 2b includes several PAHs
          sampled for at RIVA including acenaphthene and fluorene.

       •  Hexavalent chromium does not appear among the highest emitted pollutants, but
          ranks sixth for the toxicity-weighted emissions.


       Observations from Table 26-8 include the following:

       •  Toluene, xylenes, and methanol are the highest emitted pollutants with noncancer
          RfCs in Henrico County.

       •  The pollutants with the highest toxicity-weighted emissions (of the pollutants with
          noncancer RfCs) are acrolein, 1,3-butadiene, and formaldehyde.

       •  Five of the highest emitted pollutants  in Henrico County also have the highest
          toxicity-weighted emissions.

                                         26-21

-------
       •  Naphthalene has the sixth highest toxicity-weighted emissions for Henrico County
          but is not among the highest emitted pollutants with a noncancer toxicity factor in
          Henrico County.

       •  Hexavalent chromium appears on neither emissions-based list.


26.6   Summary of the 2011 Monitoring Data for RIVA

       Results from several of the data treatments described in this section include the
following:

       ปซป  Although four pollutants failed screens for RIVA, naphthalene failed the majority of
          screens.

       ปซป  The annual average concentration of naphthalene was significantly higher than the
          annual average concentrations of the other pollutants of interest.

       ปซป  Benzo(a)pyrene concentrations appear higher during the colder months of the year
          while concentrations of acenaphthene andfluorene appear higher during the warmer
          months of the year.
                                         26-22

-------
27.0   Site in Washington
       This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS site in Washington, and integrates these concentrations
with emissions, meteorological, and risk information. Data generated by sources other than ERG
are not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) for detailed discussions and definitions
regarding the various data analyses presented below.

27.1   Site Characterization
       This section characterizes the monitoring site by providing geographical and physical
information about the location of the site and the surrounding area. This information is provided
to give the reader insight regarding factors that may influence the air quality near the site and
assist in the interpretation of the ambient monitoring measurements.

       The NATTS site in Washington is located in Seattle. Figure 27-1 is a composite satellite
image retrieved from ArcGIS Explorer showing the monitoring site in its urban location.
Figure 27-2 identifies nearby point source emissions locations by source category, as reported in
the 2008 NEI for point sources. Note that only sources within 10 miles of the site are included in
the facility counts provided in Figure 27-2. A 10-mile boundary was chosen to give the reader an
indication of which emissions sources  and emissions source categories could potentially have a
direct effect on the air quality at the monitoring site. Further, this boundary provides both the
proximity of emissions sources to the monitoring site as well as the quantity of such sources
within a given distance of the site. Sources outside the 10-mile radius are still visible on the map,
but have been grayed out in order to show emissions sources just outside the boundary.
Table 27-1  provides supplemental geographical information such as land use, location setting,
and locational coordinates.
                                           27-1

-------
                                       Figure 27-1. Seattle, Washington (SEWA) Monitoring Site
to

-------
     Figure 27-2. NEI Point Sources Located Within 10 Miles of SEWA
            122-4OT-W      122-35-0-W      122'30irYV     122'25'0-W
                                                                     122'15'CTW     122'10'0"W
           122-3CTO-W     I22-2S-0-W      122'20'0-W
               Legend
                                                         12T1DWV      I22'5'0"W      122"0'Crw
                                            Note: Due- to facility density and collocation the total facilities
                                            displayed may not represent all facilities within the area of interest.
                @   SEWA NATTS site        10 mile radius |      | County boundary
Source Category Group (No. of Facilities)   v   Glass Manufacturing (1)
  "i<   Aerospace/Aircraft Manufacturing (2)
  ซ>   Air-conditioning/Refrigeration (1)
  •+1   Aircraft Operations (25)
  H   Automobile/Truck Manufacturing (1)
  IB    Bakery (2)
       Brick Manufacturing & Structural Clay (1)
^   Institutional - school (1)
?   Miscellaneous Commercial/Industrial (2)
7   Portland Cement Manufacturing (2)
A   Ship Building and Repairing (1)
V   Steel Mill (1)
'   Wastewater Treatment (1)
                                            27-3

-------
                              Table 27-1. Geographical Information for the Washington Monitoring Site
Site
Code
SEWA
AQS Code
53-033-0080
Location
Seattle
County
King
Micro- or
Metropolitan
Statistical Area
Seattle-Tacoma-
Bellevue, WA
MSA (Seattle Div)
Latitude
and
Longitude
47.568333,
-122.308056
Land Use
Industrial
Location
Setting
Suburban
Additional Ambient Monitoring Information1
Haze, CO, SO2, NOy, NO, O3, Meteorological
parameters, PM Coarse, PM10, Black Carbon, PM2 5,
PM2 5 Speciation.
  BOLD ITALICS = EPA-designated NATTS Site
to
-k

-------
        The SEWA monitoring site is located in Seattle, at the southeast corner of the Beacon
 Hill Reservoir. The reservoir and the Jefferson Park Golf Course to the east are separated by
 Beacon Avenue. A middle school and a hospital can be seen to the south of the site in the
 bottom-most portion of Figure 27-1. The site is surrounded by residential neighborhoods to the
 west, north, and east. Interstate-5, which runs north-south through Seattle, is less than 1 mile to
 the west and intersects with 1-90 a couple of miles to the north of the site. The area to the west of
 1-5 is industrial while the area to the east is primarily residential. Although the emissions sources
 within  10 miles of the site are involved in a variety of industries, the aircraft operations source
 category, which includes airports, as well as small runways, heliports, or landing pads, has the
 greatest number of sources. The point source located closest to SEWA is a bakery.

        Table 27-2 presents additional site-characterizing information, including indicators of
 mobile source activity, for the  Washington monitoring site. Table 27-2 includes county-level
 population and vehicle registration information. Table 27-2 also includes a county-level vehicle
 registration-to-population ratio, which was calculated to represent the number of vehicles per
 person  within the monitoring site's residing county. In addition, the population within  10 miles
 of the site is presented, based on postal code population data estimates. An estimate of 10-mile
 vehicle ownership was determined by applying the county-level vehicle registration-to-
 population ratio to the  10-mile population surrounding the monitoring site. Table 27-2 also
 contains traffic volume information for SEWA. Finally, Table 27-2 presents the county-level
 daily VMT for King County.
     Table 27-2. Population, Motor Vehicle, and Traffic Information for the Washington
                                       Monitoring Site
Site
SEWA
Estimated
County
Population1
1,969,722
County-level
Vehicle
Registration2
1,783,335
Vehicles per
Person
(Registration:
Population)
0.91
Population
within 10
miles3
983,171
Estimated
10-mile
Vehicle
Ownership
890,137
Annual
Average
Daily
Traffic4
226,000
County-
level Daily
VMT5
23,282,703
County-level population estimate reflects 2011 data from the U.S. Census Bureau (Census Bureau, 2012b)
2County-level vehicle registration reflects 2011 data from the Washington Dept of Licensing (WA DOL, 2011)
' 10-mile population estimate reflects 2011 data from Xionetic (Xionetic, 2011)
'AADT reflects 2011 data from the Washington DOT (WA DOT, 2011)
'County-level VMT reflects 2011 data from the Washington DOT (WA DOT, 2011)
BOLD ITALICS = EPA-designated NATTS Site
                                            27-5

-------
       Observations from Table 27-2 include the following:
       •  King County has the sixth highest county-level population among counties with NMP
          sites. The 10-mile population estimate for SEWA ranks lower but still in the top third
          among NMP sites.
       •  The county-level and 10-mile vehicle registration counts for SEWA mimick the
          rankings of the county-level and 10-mile populations.
       •  The vehicle-per-person ratio for SEWA is in the middle of the range compared to
          other NMP sites.
       •  The traffic volume experienced near SEWA is the third highest compared to other
          NMP monitoring sites. The traffic estimate  provided is for 1-5 near Spokane Street.
       •  The daily VMT for King County is in the top third compared to other counties with
          NMP sites (where VMT data were available).

27.2   Meteorological Characterization
       The following sections characterize the meteorological conditions near the monitoring
site in Washington on sample days, as well as over the course of the year.
27.2.1  Climate Summary
       The city of Seattle is located between Puget Sound and Lake Washington. The entire
urban area is situated between the Olympic Mountains to the west and the Cascades to the east.
The area experiences a mild climate as the mountains moderate storm systems that move into the
Pacific Northwest and both the mountains and the Sound shield the city from temperature
extremes. Although the city is known for its cloudy, rainy conditions, actual precipitation totals
tend to be lower compared to many locations east of the Rocky Mountains. The winter months
are the wettest and the summer months the driest. Prevailing winds are out of the southwest
(Bair, 1992).
                                          27-6

-------
27.2.2 Meteorological Conditions in 2011
       Hourly meteorological data from the NWS weather station nearest SEWA were retrieved
for 2011 (NCDC, 2011). The closest weather station to SEWA is located at Boeing Field/King
County International Airport (WBAN 24234). Additional information about this weather station,
such as the distance between the site and the weather station, is provided in Table 27-3. These
data were used to determine how meteorological conditions on sample days vary from conditions
experienced throughout the year.

       Table 27-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind  (average scalar wind speed) information for days
samples were collected and for the entire year. Also included in Table 27-3 is the 95 percent
confidence interval for each parameter. As shown in Table 27-3, average meteorological
conditions on sample days appear slightly cooler and drier than average weather conditions
experienced throughout the year, although the differences are not statistically significant.

27.2.3 Back Trajectory Analysis
       Figure 27-3 is the composite back trajectory map for days on which samples were
collected at the SEWA monitoring site in 2011.  Included in Figure 27-3 are four back trajectories
per sample day. Figure 27-4 is the corresponding cluster analysis. An in-depth description of
these maps and how they were generated is presented in Section 3.5.2.1. For the composite map,
each line represents the 24-hour trajectory along which a parcel of air traveled toward the
monitoring site on a given sample day and time, based on an initial height of 50 meters AGL. For
the cluster analysis, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric circle around the site in Figures 27-3 and 27-4 represents 100 miles.
                                          27-7

-------
to

-------
Figure 27-3. 2011 Composite Back Trajectory Map for SEWA
    Figure 27-4. Back Trajectory Cluster Map for SEWA
                          27-9

-------
       Observations from Figures 27-3 and 27-4 for SEWA include the following:

       •  Back trajectories originated from a variety of directions from SEWA, although less
          frequently from the northeast quadrant.

       •  The 24-hour air shed domain for SEWA is somewhat smaller in size than many other
          NMP sites. Although the longest trajectory originated 800 miles away over the
          Pacific Ocean, the average trajectory length was less than 200 miles long and
          86 percent of trajectories originated within 300 miles of the site.

       •  The cluster analysis shows that 35 percent of back trajectories originated over the
          Pacific Ocean, but are represented by three cluster trajectories. One represents back
          trajectories originating well offshore (4 percent), one represents back trajectories
          originating over the offshore waters of southwest Washington and Oregon
          (17 percent), and one represents back trajectories originating south and west of
          Vancouver Island (14 percent). Twenty percent of back trajectories originated over
          northwest Washington  and British Columbia, Canada, and generally  less than
          200 miles from the site. Another 16 percent of back trajectories originated over
          central Washington, 18 percent originated over south-central Washington and north-
          central Oregon, and 11 percent originated to the south of SEWA over western Oregon
          and northern California.


27.2.4 Wind Rose Comparison

       Hourly surface wind data from the NWS weather station at Boeing Field/King County

International Airport were uploaded into a wind rose software program to produce customized

wind roses, as described in Section 3.5.2.2. A wind rose shows the frequency of wind directions

using "petals" positioned around a 16-point compass, and uses different colors to represent wind

speeds.
       Figure 27-5 presents a map showing the distance between the NWS station and SEWA,

which may be useful for identifying topographical influences that can affect the meteorological

patterns experienced at this location. Figure 27-5 also presents three different wind roses for the

SEWA monitoring site. First, a historical wind rose representing 2001 to 2010 wind data is

presented, which shows the predominant surface wind speed and direction over an extended

period of time. Second, a wind rose representing wind data for all of 2011 is presented. Next, a

wind rose representing wind data for days on which samples were collected in 2011 is presented.

These can be used to identify the predominant wind speed and direction for 2011 and determine

if wind observations on sample days were representative of conditions experienced over the

entire year and historically.
                                          27-10

-------
Figure 27-5. Wind Roses for the Boeing Field/King County International Airport Weather
                                Station near SEWA
  Distance between SEWA and NWS Station
2001-2010 Historical Wind Rose
           2011 Wind Rose
    Sample Day Wind Rose
                                      27-11

-------
       Observations from Figure 27-5 for SEWA include the following:
       •  The Boeing Field/King County Airport weather station is located approximately
          2.7 miles south of SEWA.
       •  The historical wind rose shows that southeasterly, south-southeasterly, and southerly
          winds were frequently observed, accounting for nearly 40 percent of observations.
          Calm winds (< 2 knots) accounted for 24 percent of wind observations near SEWA.
       •  The wind patterns shown on the 2011 wind rose are similar to the historical wind
          patterns, although the percentage of calm winds is slightly higher (nearly 28 percent)
          in 2011.
       •  The wind patterns shown on the sample day wind rose resemble the historical and
          2011  wind patterns, indicating that conditions on sample days were representative of
          those experienced over the entire year and historically.

27.3   Pollutants  of Interest
       Site-specific "pollutants of interest" were determined for the Washington monitoring site
in order to allow 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." 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. In
addition, if any of the NATTS MQO Core Analytes measured by the monitoring site did not
meet the pollutant of interest criteria based on the preliminary risk-based screening, that pollutant
was added to the list of site-specific pollutants of interest.  A more in-depth description of the
risk-based screening process is presented in Section 3.2.
       Table 27-4 presents the results of the preliminary risk-based screening process for
SEWA. The pollutants that failed at least one screen and contributed to 95 percent of the total
failed screens are shaded. NATTS MQO Core Analytes are bolded. Thus, pollutants of interest
are shaded and/or bolded. SEWA sampled for PMi0 metals, VOCs, PAHs, carbonyl compounds,
and hexavalent chromium.
                                          27-12

-------
Table 27-4. Risk-Based Screening Results for the Washington Monitoring Site
Pollutant
Screening
Value
(Ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
Seattle, Washington - SEWA
Benzene
Carbon Tetrachloride
Formaldehyde
Acet aldehyde
Arsenic (PM10)
1,3-Butadiene
Naphthalene
Manganese (PM10)
1 ,2-Dichloroethane
Nickel (PM10)
Ethylbenzene
Dichloromethane
Hexavalent Chromium
Benzo(a)pyrene
Cadmium (PM10)
Acenaphthene
Acrylonitrile
1 , 1 ,2,2-Tetrachloroethane
0.13
0.17
0.077
0.45
0.00023
0.03
0.029
0.005
0.038
0.0021
0.4
7.7
0.000083
0.00057
0.00056
0.011
0.015
0.017
Total
61
61
60
58
56
52
51
24
17
14
12
5
4
2
2
1
1
1
482
61
61
60
60
61
52
60
61
17
61
61
61
59
25
61
60
1
1
883
100.00
100.00
100.00
96.67
91.80
100.00
85.00
39.34
100.00
22.95
19.67
8.20
6.78
8.00
3.28
1.67
100.00
100.00
54.59
12.66
12.66
12.45
12.03
11.62
10.79
10.58
4.98
3.53
2.90
2.49
1.04
0.83
0.41
0.41
0.21
0.21
0.21
12.66
25.31
37.76
49.79
61.41
72.20
82.78
87.76
91.29
94.19
96.68
97.72
98.55
98.96
99.38
99.59
99.79
100.00

Observations from Table 27-4 for SEWA include the following:

•   Eighteen pollutants failed at least one screen for SEWA, of which 12 are NATTS
    MQO Core Analytes.

•   The risk-based screening process identified 11 pollutants of interest, of which all but
    two are NATTS MQO Core Analytes. Hexavalent chromium, benzo(a)pyrene, and
    cadmium were added to SEWA's pollutants of interest because they are NATTS
    MQO Core Analytes, even though they did not contribute to 95 percent of the total
    failed screens. Six additional pollutants were added to SEWA's pollutants of interest
    because they are NATTS MQO Core Analytes, even though they did not fail any
    screens (beryllium, chloroform, lead, tetrachloroethylene, trichloroethylene, and vinyl
    chloride). These six pollutants are not shown in Table 27-4 but are shown in
    subsequent tables in the sections that follow.

•   Benzene, 1,3-butadiene, carbon tetrachloride, and formaldehyde  failed 100 percent of
    screens for SEWA. 1,2-Dichloroethane, acrylonitrile, and 1,1,2,2-tetrachloroethane
    also failed 100 percent of screens for SEWA, but were detected less frequently.
                                   27-13

-------
27.4   Concentrations
       This section presents various concentration averages used to characterize pollution levels
at the Washington monitoring site. Where applicable, the following calculations and data
analyses were performed: Time period-based concentration averages (quarterly and annual) are
provided for the pollutants of interest for the Washington monitoring site, where the data meet
the applicable criteria. Concentration averages for select pollutants are also presented graphically
to illustrate how the site's concentrations compare to the program-level averages, as presented in
Section 4.1. In addition, concentration averages for select pollutants are presented from previous
years of sampling in order to characterize concentration trends at the site. Additional site-specific
statistical summaries for SEWA are provided in Appendices J, L, M, N, and O.

27.4.1 2011 Concentration Averages
       Quarterly and annual concentration averages were calculated for the pollutants of interest
for SEWA, as described in Section 3.1. The quarterly average of a particular pollutant is simply
the average  concentration of the preprocessed daily measurements over a given calendar quarter.
Quarterly average concentrations include the substitution of zeros for all  non-detects. A site must
have a minimum of 75 percent valid samples of the total number of samples possible within a
given quarter for a quarterly average to be calculated. An annual average includes all measured
detections and substituted zeros for non-detects for the entire year of sampling. Annual averages
were calculated for pollutants where three valid quarterly averages could be calculated and
where method completeness was greater than or equal to 85 percent, as presented in Section 2.4.
Quarterly and annual average concentrations for the Washington monitoring site are presented in
Table 27-5,  where applicable. Note that concentrations of the PAHs, metals, and hexavalent
chromium are presented in ng/m3 for ease of viewing. Also note that if a  pollutant was not
detected in a given calendar quarter, the quarterly average simply reflects "0" because only zeros
substituted for non-detects were factored into the quarterly average concentration.
                                          27-14

-------
Table 27-5. Quarterly and Annual Average Concentrations of the Pollutants of Interest for
                                the SEWA Monitoring Site
Pollutant
#of
Measured
Detections
vs. # of
Samples
1st
Quarter
Average
(Ug/m3)
2nd
Quarter
Average
(Ug/m3)
3rd
Quarter
Average
(Ug/m3)
4th
Quarter
Average
(Ug/m3)
Annual
Average
(Ug/m3)
Seattle, Washington - SEWA
Acetaldehyde
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
Arsenic (PM10)a
Benzo(a)pyrene a
Bery Ilium (PM10)a
Cadmium (PM10)a
Hexavalent Chromium3
Lead(PM10)a
Manganese (PM10) a
Naphthalene a
Nickel (PM10)a
60/60
61/61
52/61
61/61
57/61
17/61
61/61
60/60
46/61
5/61
1/61
61/61
25/60
61/61
61/61
59/61
61/61
61/61
60/60
61/61
0.80
ฑ0.23
0.88
ฑ0.27
0.11
ฑ0.06
0.55
ฑ0.04
0.10
ฑ0.01
0
0.30
ฑ0.14
0.77
ฑ0.22
0.14
ฑ0.07
0.01
ฑ0.02
0
0.57
ฑ0.25
0.08
ฑ0.08
0.02
ฑ0.02
0.17
ฑ0.13
0.04
ฑ0.02
2.83
ฑ1.11
7.36
ฑ4.65
82.50
ฑ 42.28
1.19
ฑ0.51
0.80
ฑ0.26
0.56
ฑ0.17
0.04
ฑ0.02
0.65
ฑ0.06
0.14
ฑ0.02
0.04
ฑ0.03
0.20
ฑ0.04
0.77
ฑ0.15
0.03
ฑ0.03
0
0
0.46
ฑ0.17
0.01
ฑ<0.01
<0.01
ฑ0.01
0.07
ฑ0.02
0.02
ฑ0.01
2.20
ฑ0.44
5.57
ฑ3.31
36.18
ฑ8.84
1.97
ฑ1.11
1.17
ฑ0.19
0.42
ฑ0.07
0.06
ฑ0.02
0.66
ฑ0.03
0.14
ฑ0.04
O.01
ฑ0.01
0.28
ฑ0.05
0.99
ฑ0.16
0.08
ฑ0.03
0
0.01
ฑ0.01
0.50
ฑ0.11
0.01
ฑ0.01
O.01
ฑ0.01
0.06
ฑ0.01
0.02
ฑ0.01
2.49
ฑ0.29
9.36
ฑ5.08
65.76
ฑ20.36
3.39
ฑ1.22
0.96
ฑ0.12
0.96
ฑ0.21
0.14
ฑ0.04
0.72
ฑ0.05
0.11
ฑ0.02
0.05
ฑ0.02
0.38
ฑ0.08
0.82
ฑ0.20
0.12
ฑ0.04
0.01
ฑ0.01
0
1.05
ฑ0.26
0.22
ฑ0.17
O.01
ฑ0.01
0.11
ฑ0.03
0.05
ฑ0.02
3.84
ฑ1.14
9.76
ฑ6.85
99.48
ฑ 29.66
1.07
ฑ0.22
0.94
ฑ0.10
0.71
ฑ0.11
0.09
ฑ0.02
0.65
ฑ0.03
0.12
ฑ0.01
0.02
ฑ0.01
0.30
ฑ0.04
0.84
ฑ0.09
0.09
ฑ0.02
0.01
ฑ0.01
0.01
ฑ0.01
0.66
ฑ0.12
0.08
ฑ0.05
0.01
ฑ0.01
0.10
ฑ0.03
0.03
ฑ0.01
2.89
ฑ0.44
8.17
ฑ2.57
72.81
ฑ 14.58
1.90
ฑ0.46
      a Average concentrations provided for the pollutants below the blue line are presented in ng/m for
       ease of viewing.
                                           27-15

-------
Observations from Table 27-5 include the following:

•  The annual average concentrations for all of SEWA's pollutants of interest are less
   than 1.0 |ig/m3. The pollutants with the highest annual average concentrations by
   mass are acetaldehyde (0.94 ฑ 0.10 |ig/m3), formaldehyde (0.84 ฑ 0.09 |ig/m3),
   benzene (0.71 ฑ0.11 |ig/m3), and carbon tetrachloride (0.65 ฑ 0.03 |ig/m3).

•  Even though acetaldehyde and formaldehyde have the highest annual average
   concentrations among SEWA's pollutants of interest, these annual averages are the
   lowest among other NMP sites sampling carbonyl compounds, similar to previous
   years.

•  Several of the VOCs and PAHs appear to be higher during the colder months of year.
   However, the confidence intervals indicate that most of these differences are not
   statistically significant and that there is considerable variability in the  measurements.

•  The fourth quarter average concentration of arsenic is greater than the other quarterly
   averages. A review of the data shows that of the 14 concentrations greater than
   1 ng/m3 measured at SEW A, 10 were measured during the fourth quarter. The
   maximum arsenic concentration was measured on December 14, 2011 (2.04 ng/m3).
   The maximum concentrations of lead (9.45 ng/m3) and manganese (48.7 ng/m3) were
   also measured on this date.

•  The fourth quarter average concentration of lead is also greater than the other
   quarterly averages of lead and both the first and fourth quarter averages have large
   confidence intervals associated with them. A review of the data shows that
   concentrations of lead range from 1.12 ng/m3 to 9.45 ng/m3, with a median
   concentration of 2.47 ng/m3. Of the 10 concentrations greater than 3.50 ng/m3
   measured at SEW A, seven were measured during the fourth quarter and three in the
   first quarter of 2011.

•  The quarterly averages of manganese have a high-level of variability in the
   measurements, as indicated by the associated confidence intervals. Concentrations of
   manganese range from 0.647 ng/m3 to 48.7 ng/m3, with a median concentration of
   4.01 ng/m3. Of the 10 concentrations greater than 15 ng/m3 measured at SEW A, two
   were measured during the first quarter of 2011, one in the second quarter, three in the
   third quarter, and four in the fourth quarter.

•  The second and third quarter average concentrations of nickel have large confidence
   intervals associated with them. A review of the data shows that concentrations of
   nickel range from 0.431  ng/m3 to 7.97 ng/m3. The maximum concentration of nickel
   was measured on August 7, 2011 and is the maximum nickel concentration measured
   among NMP sites sampling this pollutant. Of the 13 nickel concentrations greater
   than 5 ng/m3 measured across the program, seven were measured at SEWA. All but
   one of the 13 concentrations greater than 3 ng/m3 measured at SEWA were measured
   in the second and third quarters of 2011, explaining the relatively large confidence
   intervals associated with those quarterly averages.
                                  27-16

-------
       •  The first and fourth quarter average concentrations of naphthalene are greater than the
          other quarterly averages and the first quarter average has a large confidence interval
          associated with it. A review of the data shows that the maximum concentration of
          naphthalene was measured on January 27, 2011 (308 ng/m3) and that two additional
          concentrations greater than 200 ng/m3 were measured in November and December.
          All but two of the 12 concentrations greater than 100 ng/m3 were measured in either
          the first or fourth quarters of 2011. Conversely, all but one of the nine concentrations
          less than 30 ng/m3 were measured in the second or third quarter of 2011.

       •  The first and fourth quarter average concentrations of benzo(a)pyrene are greater than
          the other quarterly averages, particularly the fourth quarter, and both averages have
          relatively large confidence intervals associated with them. A review of the data shows
          that the maximum concentration was measured on November 29, 2011 (1.30 ng/m3).
          This is one of seven concentrations of benzo(a)pyrene greater than 1 ng/m3 measured
          across the program. The next highest concentration measured at SEWA is roughly
          half as high (0.764 ng/m3, measured on December 14, 2011). Of the 14
          benzo(a)pyrene concentrations greater than 0.1  ng/m3, four were measured in January
          and February and 10 were measured in November and December.

       •  No samples were collected at SEWA between June 20, 2011 and July 13, 2011.
          However, make-up samples were collected in the latter half of July.


       Tables 4-9 through 4-12 present the sites with the 10 highest annual average

concentrations for each of the program-level pollutants of interest. Observations for SEWA from

those tables include the following:

       •  As shown in Table 4-12, SEWA has the highest annual average concentration of
          nickel among all sites sampling metals (PMio and TSP) for the  second year in a row.

       •  Recall that only nine sites sampled PMio metals; as a result every site sampling PMio
          metals appears in Table 4-12  for each metal.

       •  SEWA has the fourth highest concentrations of arsenic and manganese and ranks fifth
          highest for hexavalent chromium.
                                         27-17

-------
27.4.2 Concentration Comparison
       In order to better illustrate how a site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for the selected NATTS MQO Core
Analytes listed in Section 3.5.3, where applicable. Thus, box plots for acetaldehyde, arsenic,
benzene, benzo(a)pyrene, 1,3-butadiene, formaldehyde, hexavalent chromium, lead, manganese,
and naphthalene were created for SEWA. Figures 27-6 through 27-15 overlay the site's
minimum, annual average, and maximum concentrations onto the program-level minimum, first
quartile, median, average, third quartile, and maximum concentrations for each pollutant, as
described in Section 3.5.3.
        Figure 27-6. Program vs. Site-Specific Average Acetaldehyde Concentration
 SEWA
•-
0


2 4 6 B 10 12
Concentration (u=ym3)
Program: 1st Quartile 2nd Quartile 3rd Quartile
• D D
Site: Site Average Site Minimum/Maximum
o —
4th Quartile
Av(
14
;rage
1
       Figure 27-7. Program vs. Site-Specific Average Arsenic (PMio) Concentration
    m
SEWA
             0.5
                       1        1.5        2        2.5         3        3.5        4        4.5
                                         Concentration (ng/m3)
                Program:   IstQuartile     2ndQuartile     3rdQuartile    4thQuartile    Average
                Site:      Site Average     Site Minimum/Maximum

                                          27-18

-------
          Figure 27-8. Program vs. Site-Specific Average Benzene Concentration
SE'.'rt
•
D
L 1
P 1
1

I PrnirramMavrnnrpntratinn = 5^.
I
B ug/m3

1234567891
Concentration (p€/m3)
Program:
Site:
IstQuartile 2ndQuartile SrdQuartile 4thQuartile Average
• • D D
Site Average Site Minimum/Maximum
o —

      Figure 27-9. Program vs. Site-Specific Average Benzo(a)pyrene Concentration
                                      0.75           1          1.15
                                             Concentration (ng/m3)
                 Program:   IstQuartile      2ndQuartile     3rdQuartile     4thQuartile      Average
                Site:      Site Average     Site Minimum/Maximum

       Figure 27-10. Program vs. Site-Specific Average 1,3-Butadiene Concentration
    I
Program Max Concentration = 9.51 |ig/m3
                   3.5
                                                  1.5
                                             Concentration (jig/mi)
                Program:    IstQuartile      2ndQuartile     SrdQuartile     4thQuartile      Average
                Site:      Site Average     Site Minimum/Maximum

                                              27-19

-------
      Figure 27-11. Program vs. Site-Specific Average Formaldehyde Concentration
SE'.'rt
D
                                    10
                                                   15
                                              Concentration
                 Program:   IstQuartile      2ndQuartile     SrdQuartile     4thQuartile      Average
                 Site:      Site Average      Site Minimum/Maximum

  Figure 27-12. Program vs. Site-Specific Average Hexavalent Chromium Concentration
SE'.'•;'i.
                   0.05
                                   0.1
                                                   0.15
                                              Concentration (ng/mi)
                                                                   3.2
                                                                                  DL25
                 Program:   IstQuartile      2ndQuartile     SrdQuartile     4thQuartile      Average
                 Site:      Site Average      Site Minimum/Maximum

        Figure 27-13. Program vs. Site-Specific Average Lead (PMi0) Concentration
SE'.'rt
                                             15            20
                                              Concentration (ng/m3)
                                                                        25
                                                                                     33
                 Program:   IstQuartile      2ndQuartile      SrdQuartile     4thQuartile      Average
                 Site:      Site Average      Site Minimum/Maximum

                                                                                                   35
                                               27-20

-------
    Figure 27-14. Program vs. Site-Specific Average Manganese (PMi0) Concentration
SE'.'rt
                                                       I Program Max CoricEirtratior = 395 ng/m3
                                            100        125
                                        Concentration (ng/mi)
                                                                 150
                                                                           17E
                                                                                     200
              Program:   IstQuartile     2ndQuartile     SrdQuartile    4thQuartile     Average
              Site:     Site Average    Site Minimum/Maximum

      Figure 27-15. Program vs. Site-Specific Average Naphthalene Concentration
SEWA

                                                          Program Max Concentration = 779 ng/m3
            50
                   100
                            150
                                    200      250      300
                                        Concentration (ng/mi)
                                                             350
                                                                     400
                                                                             450
                                                                                     5M
Program:
Site:
IstQuartile 2ndC
Site Average Site l\
O
      Observations from Figures 27-6 through 27-15 include the following:

         •   Figure 27-6 shows that SEWA's annual average acetaldehyde concentration is
             considerably less than the program-level average for acetaldehyde and is actually
             less than the program-level first quartile (25th percentile). Even the maximum
             acetaldehyde concentration measured at SEWA less than the program-level
             average concentration. This site has the lowest annual average concentration of
             acetaldehyde among NMP sites sampling carbonyl compounds.

         •   Figure 27-7 shows that SEWA's annual average arsenic (PMio) concentration is
             greater than the program-level average concentration of arsenic (PMio). The
             maximum arsenic concentration measured at SEWA is roughly half the maximum
             concentration measured across the program. There were no non-detects of arsenic
             measured at SEWA.

         •   Figure 27-8 is the box plot for benzene. Note that the program-level maximum
             concentration (23.8 |ig/m3) is not shown directly on the box plot because the scale
             of the box plot would be too large to readily observe data points at the lower end
             of the concentration range. Thus, the scale has been reduced to 10 |ig/m3.
             Figure 27-8 shows that the annual average concentration for SEWA is less than
             the program-level average concentration but greater than the program-level
             median concentration. The maximum benzene concentration measured at SEWA
             is considerably less than the maximum benzene concentration measured across
                                         27-21

-------
   the program. There were no non-detects of benzene measured at SEWA or across
   the program.

•  Figure 27-9 is the box plot for benzo(a)pyrene. Note that the first quartile for this
   pollutant is zero and is not visible on this box plot. This box plot shows that the
   annual average concentration for SEWA is roughly equivalent to the program-
   level average concentration. Although the maximum concentration measured at
   SEWA is less than the maximum concentration measured across the program, it is
   one of the higher measurements. Nearly 60 percent of the benzo(a)pyrene
   measurements at SEWA were non-detects.

•  Figure 27-10 is the box plot for 1,3-butadiene. Similar to the benzene box plot,
   the program-level maximum concentration (9.51 |ig/m3) is not shown directly on
   the box plot as the scale has been reduced to 3 |ig/m3 in order to allow for the
   observation of data points at the lower end of the concentration range. This figure
   shows that the annual average concentration for SEWA is less than the program-
   level average concentration but greater than the program-level median
   concentration. Figure 27-10 also shows that the maximum 1,3-butadiene
   concentration measured at SEWA is considerably less than the maximum
   concentration measured across the program. A few non-detects of 1,3-butadiene
   were measured at SEWA.

•  Figure 27-11 shows that SEWA's annual average formaldehyde concentration is
   less than the program-level first quartile, similar to acetaldehyde. The maximum
   formaldehyde concentration measured at SEWA is less than the program-level
   median concentration. This site has the lowest annual average concentration of
   formaldehyde among NMP sites sampling carbonyl compounds and the second
   smallest range of concentrations.

•  Figure 27-12 shows that the annual average concentration of hexavalent
   chromium for SEWA is greater than the program-level average concentration and
   the program-level third quartile. The maximum hexavalent chromium
   concentration measured at SEWA is less than the maximum hexavalent chromium
   concentration measured across the program. Two non-detects of hexavalent
   chromium were measured at SEWA.

•  Figure 27-13 shows that the annual average concentration of lead (PMio) for
   SEWA is less than the program-level average but just greater than the program-
   level median concentration. The maximum lead concentration measured at SEWA
   is considerably less than the maximum lead concentration measured across the
   program. There were no non-detects of lead measured at SEWA.

•  Figure 27-14 is the box plot for manganese. The program-level maximum
   concentration (395 ng/m3) is not shown directly on the box plot as the scale has
   been reduced to 200 |ig/m3 in order to allow for the observation of data points at
   the lower end of the concentration range. This figure shows that the annual
   average concentration of manganese (PMio) for SEWA is less than the program-
   level average concentration. The maximum manganese concentration measured at

                              27-22

-------
              SEWA is considerably less than the maximum concentration measured across the
              program. There were no non-detects of manganese measured at SEWA.

          •   Figure 27-15 is the box plot for naphthalene. The program-level maximum
              concentration (779 ng/m3) is not shown directly on the box plot as the scale has
              been reduced to 500 |ig/m3 in order to allow for the observation of data points at
              the lower end of the concentration range. Figure 27-15 shows that the annual
              average concentration of naphthalene for SEWA is less than the program-level
              average concentration. The maximum naphthalene concentration measured at
              SEWA is less than the program-level maximum concentration. There were no
              non-detects of naphthalene measured at SEWA or across the program.


27.4.3  Concentration Trends

       A site-specific trends evaluation was completed for sites that have sampled one or more

of the selected NATTS MQO Core Analytes for 5 consecutive years or longer, as described in
Section 3.5.4. Sampling for hexavalent chromium at SEWA began in 2005 while sampling  for

PMio metals, VOCs, and carbonyl compounds began in 2007; thus, Figures 27-16 through 27-23

present the annual statistical metrics for acetaldehyde, arsenic, benzene, 1,3-butadiene,

formaldehyde, hexavalent chromium, lead, and manganese, respectively.  The statistical metrics

presented for assessing trends include the substitution of zeros for non-detects. A trends analysis

was not conducted for the PAHs because sampling for PAHs did not begin at SEWA until 2008.


       Observations from Figure 27-16 for acetaldehyde measurements at SEWA include the
following:

       •   The maximum acetaldehyde concentration was measured at SEWA on July 17, 2007
          (9.73 |ig/m3). The next highest concentration was considerably less (3.36 |ig/m3,
          measured in September 2009). These are the only concentrations greater than 3 |ig/m3
          measured at SEWA.

       •   The average acetaldehyde concentration ranges from 0.80 |ig/m3 (2010) to
          0.98 |ig/m3 (2009). Confidence intervals calculated indicate that the average
          concentrations are not statistically different, although the range of measurements is
          lower in more recent years.

       •   There have been no non-detects of acetaldehyde measured since the onset of
          sampling.
                                         27-23

-------
 Figure 27-16. Annual Statistical Metrics for Acetaldehyde Concentrations
                           Measured at SEWA




? c
J 6
E
.9
C
1
a
S

1 -

















-•
T
r t
•

^M •*• •"•• ""'
1 — •— ' 1
2007 2008 2009 2010 2011
Year
• 5th Pe re entile — Minimum — Median — Maximum * 95th Percentile ..^.. Average
Figure 27-17. Annual Statistical Metrics for Arsenic (PMio) Concentrations
                           Measured at SEWA
        *  5th Percentile     -  Minimum     —  Median     -  Maximum
                                                       95thPercentile   ..+.. Average
                                   27-24

-------
   Figure 27-18. Annual Statistical Metrics for Benzene Concentrations
                           Measured at SEWA
!
1
                                      2009
                                      Year
        •  5th Percentile    —  Minimum    — Median    — Maximum
                                                       95th Percentile
                                                                   . Average
Figure 27-19. Annual Statistical Metrics for 1,3-Butadiene Concentrations
                           Measured at SEWA




I
E
.a
E
I
8
a04
c ฐ-4
1











I
T
1 -




••


^ T

ป.... ... ^^^^^
— i
I — f — * *
2007 2008 2009 2010 2011
Year
* SthPercentile - Minimum — Median - Maximum • 95th Percentile ..+.. Average
                                  27-25

-------
    Figure 27-20. Annual Statistical Metrics for Formaldehyde Concentrations
                                 Measured at SEWA

                                            2009
                                            Year
            •  5th Percentile     — Minimum     — Median     —  Maximum
                                                             95th Percentile   "^"Average
Figure 27-21. Annual Statistical Metrics for Hexavalent Chromium Concentrations
                                 Measured at SEWA
   a
   •ฃ 0.15
                                            zoos
                                            Year
            *  SthPercentile     - Minimum     — Median     -  Maximum
                                                             95th Percentile
                                                                           • Average
                                        27-26

-------
   Figure 27-22. Annual Statistical Metrics for Lead (PMi0) Concentrations
                           Measured at SEWA



1
Concentration
I















<
M




>....
•
t
2007




* SthPercentile


••








^

^g ••••^^^ 	 ^^_
F"1^"
f I *
2008 2009 2010 2011
Year
Minimum — Median — Maximum • 95th Percentile "^"Average
Figure 27-23. Annual Statistical Metrics for Manganese (PMio) Concentrations
                           Measured at SEWA




I
.9
E
|
i
•
2 40

10 -



















••

1 	 " 	 1
[
*

* 	 ^
	 • 	 ' ' 	 1 	 ' 1 • ! 1^^^^ 1 	 • 	
2007 2008 2009 2010 2011
Year
* SthPercentile - Minimum — Median - Maximum • 95th Percentile ..+.. Average
                                 27-27

-------
       Observations from Figure 27-17 for arsenic (PMio) measurements at SEWA include the
following:

       •   The maximum arsenic concentration was measured at SEWA on January 19, 2009
          (2.69 ng/m3), although a similar concentration was also measured in 2007
          (2.56 g/m3).

       •   The average concentration has ranged from 0.58 ng/m3 (2010) to 0.76 |ig/m3 (2007).
          The very slight decreasing trend exhibited by the average and median concentrations
          is not statistically significant.

       •   There have been no non-detects of arsenic measured since the onset of sampling,
          including in 2008, where it appears the minimum concentration is zero. For 2008, the
          minimum is 0.011 ng/m3.


       Observations from Figure 27-18 for benzene measurements at SEWA include the
following:

       •   The maximum benzene concentration was measured at SEWA on January 19, 2009
                    3
                                                                               3
(5.38 |ig/m ), which is the same day the maximum arsenic concentration was
measured. The next highest concentration was roughly half as high (2.48 |ig/nr*,
measured in January 2011). Only five benzene concentrations greater than 2 |ig/m
have been measured at SEWA.
                                                                                  3
       •   The average concentration of benzene ranges from 0.69 |ig/m3 (2010) to 0.81 |ig/m3
          (2009). If the maximum concentration measured in 2009 was removed from the
          calculation, the averages would vary by less than 0.1 |ig/m3.

       •   Most of the statistical metrics increased at least slightly from 2010 to 2011, with the
          exception of the median. The median decreased because the number of concentrations
          at the lower end of the range increased from 2010 to 2011  while the average
          concentration increased because it is being driven by the higher concentrations
          measured in 2011 (the maximum concentration increased by 1  |ig/m3 from 2010 to
          2011).

       •   There have been no non-detects of benzene measured since the onset of sampling.


       Observations from Figure 27-19 for 1,3-butadiene measurements at SEWA include the
following:

       •   The maximum 1,3-butadiene concentration was measured at SEWA on the same day
          as the maximum benzene concentration was measured, January 19, 2009
          (0.89 |ig/m3). The next highest concentration was roughly half as high (0.46 |ig/m3)
          and was measured on the same day in January 2011 as the second highest benzene
          concentration.
                                        27-28

-------
       •   The number non-detects measured at SEWA has been increasing since the onset of
          sampling, from 0 percent in 2007 to 15 percent in 2011.

       •   Nearly all of the statistical metrics exhibit an increase from 2010 to 2011. The
          average concentration of 1,3-butadiene is at a maximum for 2011 (0.089 |ig/m3),
          although confidence intervals calculated for the average concentrations indicate that
          the averages are not statistically different.


       Observations from Figure 27-20 for formaldehyde measurements at SEWA include the
following:

       •   The maximum formaldehyde concentration was measured at SEWA on
          January 13, 2009 (16.6 |ig/m3). The next highest concentration (9.43 |ig/m3) was
          measured on the same day in 2007 as the maximum acetaldehyde concentration. Only
          one other formaldehyde concentration greater than  3 |ig/m3 has been measured at
          SEWA and was also measured in 2009.

       •   The level of variability in the measurements decreased significantly from 2009 to
          2010. The difference between the average and median concentrations is less than
          0.1 |ig/m3 for both 2010 and 2011.

       •   Although difficult to discern in Figure 27-20, the average concentration increased
          slightly from 2010 to 2011, although the average concentrations for both years are the
          lowest annual average concentrations of formaldehyde among all NMP sites sampling
          this pollutant.

       •   There have been no non-detects of formaldehyde measured since the onset of
          sampling.


       Observations from Figure 27-21 for hexavalent chromium measurements at SEWA

include the following:

       •   Although SEWA began sampling hexavalent chromium in January 2005, sampling
          was discontinued for an eight-month period in 2006 from March through October.
          There is no data provided for 2006 in Figure 27-21  because four months is not
          considered enough to be representative of an entire year.

       •   The maximum hexavalent chromium concentration shown was measured on
          January 19, 2009 (0.232 ng/m3), the same  day that the maximum concentration of
          several pollutants were measured at SEWA. A similar concentration was measured in
          2005 (0.224 ng/m3), also  in January.

       •   The minimum and 5th percentiles are zero  for each  year of sampling shown except
          2011, indicating the presence of non-detects. The percentage of non-detects has
          ranged from three percent (2011) to 21 percent (2009).
                                         27-29

-------
       •  The average concentration has a decreasing trend beginning in 2008 that continues
          through 2010. The median, however, reaches a minimum in 2009 rather than 2010,
          even with the maximum hexavalent chromium concentration measured. This is partly
          a result of the number of non-detects and thus, zeroes substituted into the calculation.
          The number of non-detects is 21 percent for 2009, which decreased to 15 percent for
          2010. There were also a higher number of "lower" measurements in 2009; 18
          concentrations less than 0.02 ng/m3 were measured in 2009 as opposed to 12 in 2010.

       •  Nearly all of the statistical metrics exhibit increases from 2010 to 2011.


       Observations from Figure 27-22 for lead (PMio) measurements at SEWA include the
following:

       •  The maximum lead concentration was measured at SEWA on February 24, 2008
          (31.7 ng/m3). Only one additional concentration measured  at SEWA is greater than
          20 ng/m3 (20.8 ng/m3 measured on July 5, 2007).

       •  A decreasing trend in the average and median lead concentrations is shown in
          Figure 27-22. Nearly all of the statistical metrics decreased from 2008 to  2009 and
          again for 2010. Slight increases in the statistical metrics are shown for 2011, although
          the difference in the average concentrations between 2010  and 2011 is not statistically
          significant.

       •  There have  been no non-detects of lead measured since the onset of sampling,
          including in 2008, where it appears the minimum concentration is zero. For 2008, the
          minimum is 0.11 ng/m3.


       Observations from Figure 27-23 for manganese (PMio) measurements at SEWA include
the following:

       •  The three highest manganese concentrations measured at SEWA were all measured in
          2007 and are the only three measurements greater than 50 ng/m3 measured at this site,
          although  the maximum concentration measured in 2011 is just less than 50 ng/m3.

       •  A steady  decreasing trend in the average manganese concentration is shown through
          2010.

       •  Most of the statistical metrics increased from 2010 to 2011. Although the 95th
          percentile more than doubled and the average increased by 40 percent, the median
          concentration changed only slightly.

       •  There have  been no non-detects of manganese measured since the onset of sampling.
                                         27-30

-------
27.5   Additional Risk-Based Screening Evaluations
       The following risk-based screening evaluations were conducted to characterize risk at the
Washington monitoring site. Refer to Sections 3.3 and 3.5.5 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.

27.5.1  Risk-Based Screening Assessment Using MRLs
       A risk-based screening was conducted by comparing the concentration data from the
Washington monitoring site to the ATSDR MRLs, where available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 to 14 days); intermediate (exposures of 15 to 364 days); and chronic (exposures
of 1 year or greater). The preprocessed daily measurements of the pollutants of interest were
compared to the acute MRLs; the quarterly averages were compared to the intermediate MRLs;
and the annual averages were compared to the chronic MRLs.

       As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2011.

27.5.2  Cancer Risk and Noncancer Hazard Approximations
       For the pollutants of interest for the Washington site and where annual average
concentrations could be calculated, risk was examined by calculating cancer risk and noncancer
hazard approximations. These approximations can be used as risk estimates for cancer and
noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers may want to shift or
confirm their air-monitoring priorities.  Refer to Section 3.5.5.2 for an explanation of how cancer
risk and noncancer hazard approximations are calculated and what limitations are associated with
them. Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer
hazard approximations are presented in Table 27-6, where applicable. Cancer risk
approximations are presented as probabilities while the noncancer hazard approximations are
ratios and thus, unitless values.
                                         27-31

-------
             Table 27-6. Risk Approximations for the Washington Monitoring Site
Pollutant
Cancer
URE
(Hg/m3)1
Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples
Annual
Average
(Hg/m3)
Cancer Risk
Approximation
(in-a-million)
Noncancer
Hazard
Approximation
(HQ)
Seattle, Washington - SEWA
Acetaldehyde
Arsenic (PM10)a
Benzene
Benzo(a)pyrene a
Bery Ilium (PM10)a
1,3 -Butadiene
Cadmium (PM10) a
Carbon Tetrachloride
Chloroform
1 ,2-Dichloroethane
Ethylbenzene
Formaldehyde
Hexavalent Chromium3
Lead(PM10)a
Manganese (PM10) a
Naphthalene a
Nickel (PM10)a
Tetrachloroethylene
Trichloroethylene
Vinyl Chloride
0.0000022
0.0043
0.0000078
0.00176
0.0024
0.00003
0.0018
0.000006

0.000026
0.0000025
0.000013
0.012


0.000034
0.00048
0.00000026
0.0000048
0.0000088
0.009
0.000015
0.03

0.00002
0.002
0.00001
0.1
0.098
2.4
1
0.0098
0.0001
0.00015
0.00005
0.003
0.00009
0.04
0.002
0.1
60/60
61/61
61/61
25/60
61/61
52/61
61/61
61/61
57/61
17/61
61/61
60/60
59/61
61/61
61/61
60/60
61/61
46/61
5/61
1/61
0.94
ฑ0.10
O.01
ฑ<0.01
0.71
ฑ0.11
O.01
ฑ<0.01
0.01
ฑ0.01
0.09
ฑ0.02
0.01
ฑ0.01
0.65
ฑ0.03
0.12
ฑ0.01
0.02
ฑ0.01
0.30
ฑ0.04
0.84
ฑ0.09
0.01
ฑ0.01
O.01
ฑO.01
0.01
ฑ0.01
0.07
ฑ0.01
0.01
ฑ0.01
0.09
ฑ0.02
0.01
ฑ0.01
O.01
ฑ0.01
2.08
2.86
5.56
0.15
0.02
2.66
0.18
3.90

0.63
0.74
10.95
0.40


2.48
0.91
0.02
0.03
0.01
0.10
0.04
0.02

0.01
0.04
0.01
0.01
0.01
O.01
0.01
0.09
0.01
0.02
0.16
0.02
0.02
O.01
O.01
0.01
- = a Cancer URE or Noncancer RfC is not available
a For the annual average concentration of this pollutant in ng/m3, refer to Table 27-5.
                                               27-32

-------
       Observations from Table 27-6 for SEWA include the following:
       •  The pollutants with the highest annual averages for SEWA are acetaldehyde,
          formaldehyde, benzene, and carbon tetrachloride.
       •  The pollutants with the highest cancer risk approximations are formaldehyde,
          benzene, carbon tetrachloride, and arsenic. Although the pollutant with the highest
          cancer risk approximation is formaldehyde, its cancer risk approximation is the
          lowest among NMP sites sampling carbonyl compounds.
       •  The noncancer hazard approximations for SEWA are all less than 1.0, with the
          highest calculated for manganese (0.16), indicating that no adverse health effects are
          expected from these individual pollutants.

27.5.3 Risk-Based Emissions Assessment
       In addition to the risk-based screenings discussed above, Tables 27-7 and 27-8 present an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 27-7 presents the 10 pollutants with the highest emissions from the 2008 NEI, the
10 pollutants with the highest toxicity-weighted emissions, and the 10 pollutants with the highest
cancer risk approximations  (in-a-million), as calculated from the annual averages provided in
Table 27-6. Table 27-8 presents similar information, but identifies the 10 pollutants with the
highest noncancer hazard approximations (HQ),  also calculated from annual averages provided
in Table 27-6.
       The pollutants listed in Table 27-7 and 27-8 are limited to those that have cancer and
noncancer toxicity factors, respectively. As a result, although the actual value of the emissions is
the same, the highest emitted pollutants in the cancer table may be different from the noncancer
table. Further, the cancer risk and noncancer hazard approximations based on each site's annual
averages are limited to those pollutants for which each respective site sampled. As discussed in
Section 27.3, SEWA sampled for VOCs, carbonyl compounds, PAHs, metals (PMio), and
hexavalent chromium. In addition, the cancer risk and noncancer hazard approximations are
limited to those pollutants with enough data to meet the criteria for annual averages to be
calculated. A more in-depth discussion of this analysis is provided in Section 3.5.5.3.  Similar to
the cancer risk and noncancer hazard approximations, this analysis may help policy-makers
prioritize their air monitoring activities.
                                          27-33

-------
    Table 27-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
                                                 the Washington Monitoring Site
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Pollutant
Cancer Risk
Approximation
(in-a-million)
Seattle, Washington (King County) - SEWA
Benzene
Formaldehyde
Ethylbenzene
Acetaldehyde
1,3 -Butadiene
Naphthalene
POM, Group 2b
Dichloromethane
POM, Group 6
Nickel, PM
1,542.30
849.03
654.82
430.54
190.09
99.14
28.31
17.29
2.99
2.83
Benzene
Formaldehyde
POM, Group 3
1,3 -Butadiene
Naphthalene
POM, Group 2b
Ethylbenzene
Nickel, PM
Hexavalent Chromium, PM
POM, Group 5a
1.20E-02
1.10E-02
8.19E-03
5.70E-03
3.37E-03
2.49E-03
1.64E-03
1.36E-03
1.34E-03
1.11E-03
Formaldehyde
Benzene
Carbon Tetrachloride
Arsenic
1,3 -Butadiene
Naphthalene
Acetaldehyde
Nickel
Ethylbenzene
1 ,2-Dichloroethane
10.95
5.56
3.90
2.86
2.66
2.48
2.08
0.91
0.74
0.63
to

-------
    Table 27-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
                                            RfCs for the Washington Monitoring Site
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Noncancer Hazard
Approximation
(HQ)
Seattle, Washington (King County) - SEWA
Toluene
Xylenes
Benzene
Methanol
Formaldehyde
Ethylbenzene
Hexane
Acetaldehyde
1,3 -Butadiene
Ethylene glycol
3,282.31
2,629.14
1,542.30
1,092.81
849.03
654.82
642.62
430.54
190.09
142.99
Acrolein
1,3 -Butadiene
Formaldehyde
Benzene
Acetaldehyde
Naphthalene
Nickel, PM
Xylenes
Lead, PM
Arsenic, PM
2,237,728.52
95,046.24
86,635.26
51,409.86
47,837.23
33,047.76
31,499.47
26,291.40
15,384.07
10,669.99
Manganese
Acetaldehyde
Formaldehyde
Arsenic
1,3 -Butadiene
Naphthalene
Benzene
Nickel
Lead
Cadmium
0.16
0.10
0.09
0.04
0.04
0.02
0.02
0.02
0.02
0.01
to

-------
Observations from Table 27-7 for SEWA include the following:

•  Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
   cancer UREs in King County.

•  The pollutants with the highest toxi city-weighted emissions (of the pollutants with
   cancer UREs) for King County are benzene, formaldehyde, and POM, Group 3.
   POM, Group 3 does not include any pollutants sampled for at SEWA.

•  Seven of the highest emitted pollutants also have the highest toxi city-weighted
   emissions for King County.

•  Formaldehyde and benzene have the highest cancer risk approximations for SEWA.
   These two pollutants top both emissions-based lists as well. Naphthalene,
   1,3-butadiene, ethylbenzene, and nickel also appear on all three lists.

•  Carbon tetrachloride and arsenic, which rank third and fourth for cancer risk
   approximations for SEWA, do not appear on either emissions-based list.
   Acetaldehyde has the seventh highest cancer risk approximation and ranks fourth for
   total emissions, but ranks 11th for toxicity-weighted emissions and thus, does not
   appear in this column in Table 27-7.

•  POM, Group 2b is the seventh highest emitted "pollutant" in King County and ranks
   sixth for toxicity-weighted emissions. POM, Group 2b includes several PAHs
   sampled for at SEWA including acenaphthene, fluorene, and perylene. Although none
   of the PAHs included in POM, Group 2b were identified as pollutants of interest for
   SEWA, acenaphthene failed one screen for SEWA.

•  POM, Group 5a ranks tenth for toxicity-weighted emissions for King County. POM,
   Group 5a includes benzo(a)pyrene and is not one of the highest "pollutants" emitted
   in King County.


Observations from Table 27-8 for SEWA include the following:

•  Toluene, xylenes, and benzene are the highest emitted pollutants with noncancer
   RfCs in King County.

•  Acrolein is the pollutant with the highest toxicity-weighted emissions (of the
   pollutants with noncancer RfCs) for King County, followed by 1,3-butadiene and
   formaldehyde. Although acrolein was sampled for at SEWA, this pollutant was
   excluded from the pollutants of interest designation, and thus subsequent risk-based
   screening evaluations, due to questions about the consistency and reliability of the
   measurements, as discussed in Section 3.2.

•  Five of the highest emitted pollutants  also have the highest toxicity-weighted
   emissions for King County.
                                  27-36

-------
       •  Manganese, which has the highest noncancer hazard approximation for SEW A, does
          not appear on either emissions-based list for King County. Acetaldehyde,
          formaldehyde, 1,3-butadiene, and benzene appear on all three lists in Table 27-8.

       •  Three of the four metals that appear among the highest noncancer hazard
          approximations for SEWA are also among the pollutants with the highest toxicity-
          weighted emissions (arsenic, nickel, and lead). However, none of these are among the
          highest emitted pollutants (with a noncancer RfC) in King County.

       •  Naphthalene also has one of the highest noncancer hazard  approximations for SEWA
          and appears among the pollutants with the highest toxicity-weighted emissions but is
          not one of the highest emitted (with a noncancer RfC).


27.6   Summary of the 2011 Monitoring Data for SEWA

       Results from several of the data treatments described in this section include the
following:

       *ป*  Eighteen pollutants failed at least one screen for SEWA, of which 12 are NATTS
          MQO Core Analytes.

       ปซป  Acetaldehyde had the highest annual average concentration for SEWA, although all
          of the pollutants of interest for SEWA had annual average  concentrations less than
          1 ng/m3.

       ปซป  The annual average concentration of nickel for SEWA is the highest among NMP
          sites sampling metals.

       ปซป  The average concentration for each of the pollutants for which a trends analysis was
          performed exhibits a slight increase from 2010 to 2011.
                                         27-37

-------
28.0   Site in Wisconsin
       This section examines the spatial and temporal characteristics of the ambient monitoring
concentrations measured at the NATTS site in Wisconsin, and integrates these concentrations
with emissions, meteorological, and risk information. Data generated by sources other than ERG
are not included in the data analyses contained in this report. Readers are encouraged to refer to
Sections 1 through 4 and the glossary (Appendix P) for detailed discussions and definitions
regarding the various data analyses presented below.

28.1   Site Characterization
       This section characterizes the monitoring site by providing geographical and physical
information about the location of the site and the surrounding area. This information is provided
to give the reader insight regarding factors that may influence the air quality near the site and
assist in the interpretation of the ambient monitoring measurements.

       The HOWI site is located in Horicon, Wisconsin and is the relocated Mayville NATTS
site. Figure 28-1 is a composite satellite image retrieved from ArcGIS Explorer showing the
monitoring site in its rural location. Figure 28-2  identifies nearby point source emissions
locations by source category, as reported in the 2008 NEI for point sources. Note that only
sources within 10 miles of the site are included in the facility counts provided in Figure 28-2.
A 10-mile boundary was chosen to give the  reader an indication of which emissions sources and
emissions source categories could potentially have a direct effect on the air quality at the
monitoring site. Further, this boundary provides  both the proximity of emissions sources to the
monitoring site as well as the quantity of such sources within a given distance of the site. Sources
outside the 10-mile radius are still visible on the map, but have been grayed out in order to show
emissions sources just outside the boundary. Table 28-1 provides supplemental geographical
information such as land use, location setting, and locational coordinates.
                                           28-1

-------
                                       Figure 28-1. Horicon, Wisconsin (HOWI) Monitoring Site
to
oo

to

-------
       Figure 28-2. NEI Point Sources Located Within 10 Miles of HOWI
                Legend
                      HOWI NATTS site
Source Category Group (No. of Facilities)
  -f   Aircraft Operations (3)
  B   Automobile/Truck Manufacturing (1)
  E   Electroplating, Plating, Polishing, Anodizing, & Coloring (1}
  0   Fabricated Metal Products (5)
  F   Food Processing/Agriculture (2)
  •   Lime Manufacturing (1)
              5   Miscellaneous Coating Manufacturing (1)
              M   Miscellaneous Manufacturing (1)
              P   Printing/Publishing (1)
                                               28-3

-------
                               Table 28-1. Geographical Information for the Wisconsin Monitoring Site

Site
Code
HOW

AQS Code
55-027-0007

Location
Horicon

County
Dodge
Micro- or
Metropolitan
Statistical Area
Beaver Dam, WI
MSA
Latitude
and
Longitude
43.466111,
-88.621111

Land Use
Agricultural

Location
Setting
Rural

Additional Ambient Monitoring Information1
CO, SO2, NOy, NO, VOCs, Carbonyl compounds, O3,
Meteorological parameters, PM10, PM10 Metals,
PM2 5, and PM2 5 Speciation, SVOCs, PM Coarse.
  BOLD ITALICS = EPA-designated NATTS Site
to
oo

-------
       The HOWI site is located just north of the town of Horicon, in southeast Wisconsin,
within the boundaries of the Horicon Marsh Wildlife Area. HOWI is located about 46 miles
northwest of Milwaukee and roughly 48 miles northeast of Madison. The surrounding area is
rural and agricultural in nature, although a residential subdivision is located just south of the site.
The HOWI monitoring site serves as a rural background site. However, the area is affected by
nearby urban areas, and thus, could show the effects on the wildlife sanctuary. State Highway 28,
which can be seen on the lower right-hand side of Figure 28-1, is the closest major roadway.
Figure 28-2 shows that a cluster of point sources is located just south and west of HOWI, in the
town of Horicon. The closest point source to HOWI is an automobile and truck manufacturing
facility. The source categories with the most emissions sources within 10 miles of HOWI are
fabricated metal products facilities; aircraft operations, which include airports as well as small
runways, heliports, and landing pads; and food processing/agricultural facilities.

       Table 28-2 presents additional site-characterizing information, including indicators of
mobile source activity, for the Wisconsin monitoring site. Table 28-2 includes county-level
population and vehicle registration information. Table 28-2 also includes a county-level vehicle
registration-to-population ratio, which was  calculated to represent the number of vehicles per
person within the monitoring site's residing county. In addition, the population within 10 miles
of the site is presented, based on postal code population data estimates. An estimate of 10-mile
vehicle ownership was determined by applying the county-level vehicle registration-to-
population ratio to the 10-mile population surrounding the monitoring site. Table 28-2 also
contains traffic volume information for HOWI. Finally, Table 28-2 presents the county-level
daily VMT for Dodge County.

     Table 28-2. Population, Motor Vehicle, and Traffic Information for the Wisconsin
                                     Monitoring  Site



Site
HOWI

Estimated
County
Population1
88,661

County-level
Vehicle
Registration2
100,176
Vehicles per
Person
(Registration:
Population)
1.13

Population
within 10
miles3
21,990
Estimated
10-mile
Vehicle
Ownership
24,846
Annual
Average
Daily
Traffic4
5,000

County-
level Daily
VMT5
2,626,054
 Bounty-level population estimate reflects 2011 data from the U.S. Census Bureau (Census Bureau, 2012b)
 2County-level vehicle registration reflects 2011 data from the Wisconsin DOT (WI DOT, 2011)
 310-mile population estimate reflects 2011 data from Xionetic (Xionetic, 2011)
 4AADT reflects 2008 data from the Wisconsin DOT (WI DOT, 2008)
 5County-level VMT reflects 2011 data from the Wisconsin DOT (WI DOT, 2012)
 BOLD ITALICS = EPA-designated NATTS Site
                                           28-5

-------
       Observations from Table 28-2 include the following:
       •  Dodge County's population is in the bottom-third compared to other counties with
          NMP sites. This is also true of its 10-mile population.
       •  The county-level vehicle registration is also on the low end compared to other
          counties with NMP sites. This is also true of its estimated 10-mile vehicle ownership.
       •  The vehicle-per-person ratio is slightly greater than one vehicle per person. This ratio
          ranks among the higher ratios for NMP sites.
       •  The traffic volume near HOWI is also on the low end compared to other NMP sites.
          The traffic estimate provided is for State Road 28 near State Road 33 on the east side
          of Horicon.
       •  The daily VMT for Dodge County is on the low side compared to other counties with
          NMP sites (where VMT data were available).

28.2   Meteorological Characterization
       The following sections characterize the meteorological conditions near the monitoring
site in Wisconsin on sample days, as well as  over the course of the year.
28.2.1  Climate Summary
       The town of Horicon is located in southeast Wisconsin, between the towns of West Bend
and Beaver Dam, and about 40 miles west of Lake Michigan. This area is far enough inland to
limit some of the moderating influences of Lake Michigan on the area's climate. This area
experiences a highly variable, continental climate as weather systems frequently track across the
region. Precipitation falls predominantly in the spring and summer months. Winters are cold and
predominantly dry, although snowfall is common. Lake effect snows can occur with winds with
a northeasterly and easterly component, although lake effect snows are often reduced this far
inland. Summers tend to be mild, although southerly winds out of the Gulf of Mexico can
occasionally advect warm, humid air into the area (Bair, 1992).
                                          28-6

-------
28.2.2 Meteorological Conditions in 2011
       Hourly meteorological data from the NWS weather station nearest this site were retrieved
for 2011 (NCDC, 2011). The closest weather station is located at Dodge County Airport
(WBAN 04898). Additional information about the Dodge County Airport weather station, such
as the distance between the site and the weather station, is provided in Table 28-3. These data
were used to determine how meteorological conditions on sample days vary from conditions
experienced throughout the year.

       Table 28-3 presents average temperature (average maximum and average daily), moisture
(average dew point temperature, average wet bulb temperature, and average relative humidity),
pressure (average sea level pressure), and wind (average scalar wind speed) information for days
samples were collected and for the entire year. Also included in Table 28-3 is the 95 percent
confidence interval for each parameter. As shown in Table 28-3, average meteorological
conditions on sample days appear cooler and drier than average weather conditions throughout
2011, although the differences are not statistically significant.

28.2.3 Back Trajectory Analysis
       Figure 28-3 is the composite back trajectory map for days on which samples were
collected at the HOWI monitoring site in 2011. Included in Figure 28-3 are four back trajectories
per sample day. Figure 28-4 is the corresponding cluster analysis. An in-depth description of
these maps and how they were generated is presented in  Section 3.5.2.1. For the composite map,
each line represents the 24-hour trajectory along which a parcel of air traveled toward the
monitoring site on a given sample day and time, based on an initial height of 50 meters AGL. For
the cluster analysis, each line corresponds to a trajectory representative of a given cluster of back
trajectories. Each concentric circle around the site in Figures 28-3 and 28-4 represents 100 miles.
                                          28-7

-------
to
oo

oo
                          Table 28-3. Average Meteorological Conditions near the Wisconsin Monitoring Site
Closest NWS
Station
(WBAN and
Coordinates)
Distance
and
Direction
from Site
Average
Type1
Average
Maximum
Temperature
(ฐF)
Average
Temperature
(ฐF)
Average
Dew Point
Temperature
(ฐF)
Average
Wet Bulb
Temperature
(ฐF)
Average
Relative
Humidity
(%)
Average
Sea Level
Pressure
(mb)
Average
Scalar Wind
Speed
(kt)
Horicon, Wisconsin - HOWI
Dodge County
Airport
04898
(43.43, -88.70)
4.64
miles
229ฐ
(SW)
Sample
Day
2011
52.3
ฑ5.5
54.5
+ 2.2
44.9
ฑ5.1
46.6
+ 2.0
34.3
ฑ4.6
36.2
+ 1.8
40.0
ฑ4.5
41.8
+ 1.8
69.7
ฑ3.2
70.7
+ 1.3
NA
NA
6.9
ฑ0.9
6.9
+ 0.3
      Sample day averages are highlighted to help differentiate the sample day averages from the full-year averages.

      NA = Sea level pressure was not recorded at the Dodge County Airport

-------
Figure 28-3. 2011 Composite Back Trajectory Map for HOWI
                             -7
    Figure 28-4. Back Trajectory Cluster Map for HOWI
                          28-9

-------
       Observations from Figures 28-3 and 28-4 for HOWI include the following:

       •  Back trajectories originated from a variety of directions at HOWI, although a
          majority of the back trajectories originate from a direction with a westerly
          component.

       •  The 24-hour air  shed domain for HOWI is among the larger in size compared to other
          NMP monitoring sites. The farthest away a back trajectory originated was north-
          central Montana, or approximately 950 miles away. However, the average trajectory
          length was 274 miles and most trajectories (88 percent) originated within 450 miles
          of the site.

       •  The cluster analysis shows that about 16 percent of the back trajectories originated
          from the north and northeast of HOWI, although of varying distances (10 percent
          originated over Lakes Michigan and Superior while another 6 percent originated over
          Ontario, Canada). Another 13 percent of back trajectories originated to the north-
          northwest to north of HOWI. Ten percent of back trajectories originated to the
          northwest and greater than 400 miles away of HOWI. Twenty-seven percent of back
          trajectories are represented by the short trajectory originating to the west of the site.
          The individual back trajectories represented by this cluster trajectory originated from
          the northwest, west, and southwest and within 300 miles of the site. Thirteen percent
          of back trajectories originated from the southeast to southwest of the site over Illinois,
          Missouri, and Iowa while another 21  percent originated to the east to southeast over
          Michigan, Ohio, and Indiana.


28.2.4 Wind Rose Comparison

       Hourly surface wind data from the NWS weather station at Dodge County Airport near

HOWI were uploaded into a wind rose software program to produce customized wind roses, as

described in Section 3.5.2.2. A wind rose shows the frequency of wind directions using "petals"

positioned around a 16-point compass, and uses different colors to represent wind speeds.
       Figure 28-5 presents a map showing the distance between the NWS station and HOWI,

which may be useful for identifying topographical influences that can affect the meteorological

patterns experienced at this location. Figure 28-5 also presents three different wind roses for the

HOWI monitoring site. First, a historical wind rose representing 2003 to 2010 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 2011 is presented.

Next, a wind rose representing wind data for days on which samples were collected in 2011 is

presented. These can be used to identify the predominant wind speed and direction for 2011 and

determine if wind observations on sample days were representative of conditions experienced

over the entire year and historically.

                                          28-10

-------
 Figure 28-5. Wind Roses for the Dodge County Airport Weather Station near HOWI
Distance between HOWI and NWS Station
2003-2010 Historical Wind Rose
          2011 Wind Rose
    Sample Day Wind Rose
                                    28-11

-------
       Observations from Figure 28-5 for HOWI include the following:
       •  The Dodge County Airport weather station is located approximately 4.6 miles
          southwest of HOWI.
       •  The historical wind rose shows that winds from a variety of directions were observed
          near HOWI. Winds from the south, southwest quadrant, and west account for
          one-third of wind observations. The strongest wind speeds were associated with
          southerly to west-southwesterly winds. Calm winds (<2 knots) were observed for
          nearly 15 percent of the hourly measurements.
       •  The wind patterns shown on the 2011 wind rose resemble the historical wind patterns,
          although winds from the north were observed more frequently.
       •  The sample day wind rose shows that winds from the north were observed even more
          frequently on sample days and that a higher percentage of strong (> 22 knots) winds
          were observed for these northerly winds. In addition, winds from the east to southeast
          were observed more frequently than winds from the southwest quadrant.

28.3   Pollutants of Interest
       Site-specific "pollutants of interest" were determined for the Wisconsin monitoring site
in order to allow 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." 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.  In
addition, if any of the NATTS MQO Core Analytes measured by the monitoring site did not
meet the pollutant of interest criteria based on the preliminary risk-based screening, that pollutant
was added to the list of site-specific pollutants of interest. A more in-depth description of the
risk-based screening process is presented in Section 3.2.
        Table 28-4. Risk-Based Screening Results for the Wisconsin Monitoring Site
Pollutant
Screening
Value
(Ug/m3)
#of
Failed
Screens
#of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
Horicon, Wisconsin - HOWI
Hexavalent Chromium
0.000083
Total
0
0
41
41
0.00
0.00
0.00
0.00

                                          28-12

-------
       Table 28-4 presents the results of the preliminary risk-based screening process for
HOWL Observations from Table 28-4 include the following:
       •  Hexavalent chromium was detected in 41 of the 61 valid samples collected at HOWL
       •  Hexavalent chromium did not fail any screens. However, hexavalent chromium is the
          pollutant of interest for HOWI because it is a NATTS MQO Core Analyte and
          because it is the only pollutant sampled for at HOWI.

28.4   Concentrations
       This section presents various concentration averages used to characterize pollution levels
at the Wisconsin monitoring site. Where applicable, the following calculations and data analyses
were performed: Time period-based concentration averages (quarterly and annual) are provided
for the pollutant of interest for the Wisconsin monitoring site, where the data meet the applicable
criteria. Concentration averages for the pollutants of interest are also presented graphically for
the site to illustrate how the site's concentrations compare to the program-level averages, as
presented in Section 4.1. In addition, concentration averages for select pollutants are presented
from previous years of sampling in order to characterize concentration trends at the site.
Additional site-specific statistical summaries for HOWI are provided in Appendix O.

28.4.1 2011 Concentration Averages
       Quarterly and annual concentration averages were calculated for the pollutants of interest
for the Wisconsin 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 of 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 HOWI are presented in
Table 28-5, where applicable. Note that if a pollutant was not detected in a given calendar
quarter, the quarterly average simply reflects "0" because only zeros substituted for non-detects
were factored into the quarterly average  concentration.
                                          28-13

-------
         Table 28-5. Quarterly and Annual Average Concentrations of the Pollutants of
                          Interest for the Wisconsin 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)
Horicon, Wisconsin - HOWI

Hexavalent Chromium

41/61
0.004
ฑ 0.004
0.023
ฑ 0.009
0.022
ฑ 0.006
0.010
ฑ 0.007
0.015
ฑ 0.004
      NA = Not available due to the criteria for calculating a quarterly and/or annual average

       Observations for HOWI from Table 28-5 include the following:
       •   Concentrations of hexavalent chromium spanned an order of magnitude, ranging from
          0.0078 ng/m3 to 0.072 ng/m3. The maximum concentration was measured on
          May 9, 2011.
       •   Hexavalent chromium was most often detected in the warmer months of the year.
          Nine non-detects were measured during the first quarter of 2011, two each in the
          second and third quarters, and seven were measured in the fourth quarter of the year.
       •   The annual average concentration of hexavalent chromium for HOWI was on the low
          side compared to other NMP sites sampling this pollutant, ranking 15th out 22.

28.4.2 Concentration Comparison
       In order to better illustrate how a site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for the selected NATTS MQO Core
Analytes listed in Section 3.5.3, where applicable. Thus, a box plot for hexavalent chromium was
created for HOWI. Figure 28-6 overlays the site's minimum, annual average, and maximum
concentrations onto the program-level minimum, first quartile, median, average, third quartile,
and maximum concentrations, as described in Section 3.5.3.
                                         28-14

-------
   Figure 28-6. Program vs. Site-Specific Average Hexavalent Chromium Concentration
 HOWl
                 0.05
                               0.1
                                            0.15
                                        Concentration (ng/mi)
                                                         0.2
                                                                       DL25
                                                                                    0.3
Program
Site:
: IstQuartile
	
Site Average
o
2ndQuartile SrdQuartile 4thQuartile AVE
Site Minimum/Maximum
rage

       Observations from Figure 28-6 include the following:
          •  Figure 28-6 shows that the annual average concentration of hexavalent chromium
             for HOWI is less than the program-level average as well as the program-level
             median concentration. The maximum concentration measured at HOWI is
             considerably less than the program-level maximum concentration. There were 20
             non-detects of hexavalent chromium measured at HOWI.

28.4.3  Concentration Trends
       A site-specific trends evaluation was completed for sites that have sampled one or more
of the selected NATTS MQO Core Analytes for 5 consecutive years or longer, as described in
Section 3.5.4. Because sampling under the NMP did not begin until December 2009 at HOWI, a
trends analysis was not conducted for this site.

28.5   Additional Risk-Based Screening Evaluations
       The following risk-based screening evaluations were conducted to characterize risk at the
Wisconsin monitoring site. Refer to Sections 3.3 and 3.5.5 for definitions and explanations
regarding the various toxicity factors, time frames, and calculations associated with these risk-
based screenings.

28.5.1  Risk-Based Screening Assessment Using MRLs
       A risk-based screening was conducted by comparing the concentration data from the
Wisconsin monitoring site to the ATSDR MRLs, where  available. As described in Section 3.3,
MRLs are noncancer health risk benchmarks and are defined for three exposure periods: acute
(exposures of 1 to 14 days); intermediate (exposures of 15 to 364 days); and chronic (exposures
of 1 year or greater). The preprocessed daily measurements of the pollutants of interest were
                                         28-15

-------
compared to the acute MRLs; the quarterly averages were compared to the intermediate MRLs;
and the annual average was compared to the chronic MRLs.

       As discussed in Section 4.2.2, none of the measured detections or time-period average
concentrations for any of the monitoring sites were greater than their respective ATSDR MRL
noncancer health risk benchmarks for any of the pollutants measured under the NMP for 2011.

28.5.2  Cancer Risk and Noncancer Hazard Approximations
       For the pollutants of interest for the Wisconsin 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 may want to shift or
confirm their air-monitoring priorities. Refer to Section 3.5.5.2 for an explanation of how cancer
risk and noncancer hazard approximations are calculated and what limitations  are associated with
them. Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer
hazard approximations are  presented in Table 28-6, where applicable. Cancer risk
approximations are presented as probabilities while the noncancer hazard  approximations are
ratios and thus, unitless values.

           Table  28-6. Risk Approximations for the Wisconsin Monitoring Site



Pollutant


Cancer
URE
(Hg/m3)1


Noncancer
RfC
(mg/m3)
#of
Measured
Detections
vs. # of
Samples


Annual
Average
(ng/m3)


Cancer Risk
Approximation
(in-a-million)

Noncancer
Hazard
Approximation
(HQ)
Horicon, Wisconsin - HOWI

Hexavalent Chromium

0.012

0.0001

41/61
0.015
ฑ 0.004

0.18

0.01
       Observations for HOWI from Table 28-6 include the following:
       •  The cancer risk approximation for hexavalent chromium is less than 1.0 in-a-million
          (0.18 in-a-million).
       •  The noncancer hazard approximation for hexavalent chromium is less than 1.0,
          indicating that no adverse health effects are expected from this pollutant.
                                         28-16

-------
28.5.3 Risk-Based Emissions Assessment
       In addition to the risk-based screenings discussed above, Tables 28-7 and 28-8 present an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 28-7 presents the 10 pollutants with the highest emissions from the 2008 NEI, the
10 pollutants with the highest toxicity-weighted emissions, and the cancer risk approximation for
hexavalent chromium, as calculated from the annual average provided in Table 28-6. Table 28-8
presents similar information, but identifies the noncancer hazard approximation for hexavalent
chromium, also calculated from annual average provided in Table 28-6.

       The pollutants listed in Tables 28-7 and 28-8 are limited to those that have cancer and
noncancer toxicity factors, respectively. As a result, although the actual value  of the emissions is
the same, the highest emitted pollutants in the cancer table may be different from the noncancer
table. The cancer risk and noncancer hazard approximations based on each site's annual averages
are limited to those pollutants for which each respective site sampled. As discussed in
Section 28.3, HOWI sampled for hexavalent chromium only. In addition, the cancer risk and
noncancer hazard approximations are limited to those pollutants with enough data to meet the
criteria for annual averages to be calculated. A more in-depth discussion of this analysis is
provided in Section 3.5.5.3.  Similar to the cancer risk and noncancer hazard approximations, this
analysis may help policy-makers prioritize their air monitoring activities.

       Observations from Table 28-7 include the following:
       •  Benzene, formaldehyde, and acetaldehyde are the highest emitted pollutants with
          cancer UREs in Dodge County.
       •  Benzene is the pollutant with the highest toxicity-weighted emissions (of the
          pollutants with cancer UREs), followed by formaldehyde and hexavalent chromium.
       •  Seven of the highest emitted pollutants in Dodge County also have the highest
          toxicity-weighted emissions.
       •  Hexavalent chromium, which is the only pollutant sampled for at HOWI, has the third
          highest toxicity-weighted emissions for Dodge County, but is not among the highest
          emitted. Hexavalent chromium emissions in Dodge County rank 17th.
       •  Several POM Groups rank among Dodge County's  highest total emissions and
          toxicity-weighted emissions.
                                         28-17

-------
  Table 28-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for

                                                 the Wisconsin Monitoring Site
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Cancer
Toxicity
Weight
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)
Cancer Risk
Approximation
Pollutant (in-a-million)
Horicon, Wisconsin (Dodge County) - HOWI
Benzene
Formaldehyde
Acetaldehyde
Ethylbenzene
1,3 -Butadiene
Naphthalene
POM, Group 2b
Tetrachloroethylene
Trichloroethylene
Dichloromethane
90.74
47.98
31.44
26.35
9.67
6.11
1.31
1.00
0.85
0.82
Benzene
Formaldehyde
Hexavalent Chromium, PM
1,3 -Butadiene
Naphthalene
POM, Group 3
POM, Group 2b
POM, Group 5a
Acetaldehyde
Ethylbenzene
7.08E-04
6.24E-04
3.70E-04
2.90E-04
2.08E-04
2.07E-04
1.15E-04
1.12E-04
6.92E-05
6.59E-05
Hexavalent Chromium 0.18

to
oo
oo

-------
  Table 28-8. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer

                                            RfCs for the Wisconsin Monitoring Site
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Pollutant
Emissions
(tpy)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Pollutant
Noncancer
Toxicity
Weight
Top 10 Noncancer Hazard Approximations Based
on Annual Average Concentrations
(Site-Specific)
Noncancer Hazard
Approximation
Pollutant (HQ)
Horicon, Wisconsin (Dodge County) - HOWI
Toluene
Xylenes
Methanol
Benzene
Formaldehyde
Methyl isobutyl ketone
Acetaldehyde
Hexane
Hydrochloric acid
Ethylbenzene
329.03
146.74
98.40
90.74
47.98
33.47
31.44
31.15
26.97
26.35
Acrolein
Manganese, PM
Cyanide Compounds, gas
Formaldehyde
1,3 -Butadiene
Acetaldehyde
Benzene
Chlorine
Naphthalene
Xylenes
163,416.51
6,223.90
5,737.27
4,895.94
4,837.37
3,493.32
3,024.81
2,097.57
2,036.14
1,467.40
Hexavalent Chromium O.01

to
oo

-------
       Observations from Table 28-8 include the following:

       •   Toluene, xylenes, and methanol are the highest emitted pollutants with noncancer
          RfCs in Dodge County.

       •   The pollutants with the highest toxi city-weighted emissions (of the pollutants with
          noncancer RfCs) are acrolein, manganese, and cyanide compounds (gaseous).

       •   Five of the highest emitted pollutants in Dodge County also have the highest toxi city-
          weighted emissions.

       •   Hexavalent chromium  does not appear among the pollutants with the highest
          emissions or toxi city-weighted emissions. This pollutant's emissions rank 32nd and its
          toxi city-weighted emissions rank 19th (among the pollutants with noncancer RfCs).


28.6   Summary of the 2011  Monitoring Data for HOWI

       Results from several of the data treatments described in this section include the
following:

       ปซป  Hexavalent chromium was the only pollutant sampled for at HO WI and did not fail
          any screens.

       ปซป  Concentrations ofhexavalent chromium measured at HO WI ranged from
          0.0078 ng/m3 to 0.072 ng/m3.  The annual average concentration ofhexavalent
          chromium ranked 15th compared to other NMP sites sampling this pollutant.

       ปซป  Hexavalent chromium has the third highest toxicity-weighted emissions for Dodge
          County, but is not among the highest emitted.
                                         28-20

-------
29.0   Data Quality
       This section discusses the data quality of the ambient air measurements that constitute the
2011 NMP dataset. Each monitoring program under the NMP has its own specific DQO(s) which
have been established and approved by EPA, consistent with the specific data use needs of the
individual monitoring program. Because the DQOs are program-specific and the ERG laboratory
is contracted to perform services for a subset of the overall program participants, attainment of
the individual program DQO(s) is not assessed in this report.  This section establishes data
quality through the assessment of Data Quality Indicators (DQI) in the form of Measurement
Quality Objectives (MQOs) specific to the program elements conducted by the ERG laboratory.
MQOs are designed to control and evaluate the various phases of the measurement process
(sampling, preparation, analysis, etc.) to ensure that the total measurement quality meets the
overall program data quality needs. In accordance with ERG's EPA-approved QAPP (ERG,
2011), the following MQOs were assessed: completeness, precision, and accuracy (also called
bias).

       The quality assessments presented in this section show that the 2011 monitoring data are
of a known and high quality, consistent with the intended data use. The overall method-specific
completeness was greater than 90 percent for each method. The method precision for collocated
and duplicate analyses met the precision MQO of 15 percent CV for all methods. The analytical
precision for replicate analyses also met the precision MQO of 15 percent CV. Audit samples
show that ERG is meeting the accuracy requirements of the NATTS TAD (EPA, 2009b). These
data quality indicators are discussed in further detail in the following sections.
29.1   Completeness
       Completeness refers to the number of valid samples actually collected and analyzed
compared to the number of total samples scheduled to be collected and analyzed. The MQO for
completeness based on the EPA-approved QAPP specifies that at least 85 percent of samples
collected at a given monitoring site must be analyzed successfully to be considered sufficient for
data trends analysis (ERG, 2011). The MQO of 85 percent completeness was met by all but
seven out of 123  site-method combinations. Completeness statistics are presented and discussed
more thoroughly in Section 2.4.
                                          29-1

-------
29.2   Method Precision
       Precision defines the level of agreement between independent measurements performed
according to identical protocols and procedures. Method precision, which includes sampling and
analytical precision, quantifies random errors associated with collecting ambient air samples and
analyzing the samples in the laboratory. Method precision is evaluated by comparing
concentrations measured in duplicate or collocated samples. A duplicate sample is a sample
collected simultaneously with a primary sample using the same sampling system (i.e., two
separate samples through the same sampling system at the same time). This simultaneous
collection is typically achieved by teeing the line from the sampler to two canisters (or other
sampling media) and doubling the flow rate applied to achieve integration over the 24-hour
collection period. Collocated samples are samples collected simultaneously using two
independent collection systems at the same location at the same time.

       Both approaches provide valuable, but different, assessments of method precision:
       •   Analysis of duplicate samples  provides information on the potential for variability (or
          precision) expected from a single collection system (intra-system assessment).
       •   Analysis of collocated samples provides information on the potential for variability
          (or precision) expected between different collection systems (inter-system
          assessment).

       During the 2011 sampling year, duplicate and collocated samples were collected on at
least 10 percent of the scheduled sample days, as outlined in the EPA-approved QAPP. This
provides a minimum of six pairs of either duplicate or collocated samples per site and method.
For the VOC, SNMOC, and carbonyl compound methods, samples may be duplicate or
collocated.  For PAHs, metals, and hexavalent chromium, only collocated samples may be
collected due to limitations of the sampling media/instrumentation. These duplicate or collocated
samples were then analyzed in replicate. Replicate measurements are repeated analyses
performed on a duplicate or collocated pair of samples and are discussed in greater detail in
Section 29.3.
                                          29-2

-------
       Method precision was calculated by comparing the concentrations of the
duplicates/collocates for each pollutant. The CV for duplicate or collocated samples was
calculated for each pollutant and each site. The following approach was employed to estimate
how closely the collected and analyzed samples agree with one another:

    Coefficient of Variation (CV) provides a relative measure of data dispersion compared to the
    mean. A coefficient of variation of one percent would indicate that the analytical results
    could vary slightly due to sampling error, while a variation of 50 percent means that the
    results are more imprecise.
                                                  (p-r)    l2
                                                  2n
                                       i
          Where:
                 p = the primary result from a duplicate or collocated pair;
                 r = the secondary result from a duplicate or collocated pair;
                 n = the number of valid data pairs (the 2 adjusts for the fact that there are two
                     values with error).
       Coefficients of variation were based on every pair of duplicate or collocated samples
collected during the program year. However,  only results at or above the MDL were used in
these calculations. Thus, the number of pairs included in the calculations varies significantly
from pollutant to pollutant. This is a change in procedure compared to NMP reports prior to
2010, where 1/2 MDL was substituted for non-detects. To  make an overall estimate of method
precision, program-level average CVs were calculated as follows:
          •   A pollutant-specific CV was calculated for each monitoring site.
          •   A site-specific CV was calculated for each method.
          •   A method-specific CV was calculated and compared to the precision MQO.

       Table 29-1 presents the 2011 NMP method precision for VOCs, SNMOCs, carbonyl
compounds, PAHs, metals, and hexavalent chromium, presented as the average CV (expressed as
a percentage). Each analytical method met the program MQO of 15 percent CV for method
precision. This table also includes the number of pairs that were included in the calculation of the
method precision. The number of pairs including those with concentrations less than the MDL is
also included in Table 29-1 to provide an indication of the effect that excluding those with
concentrations less than the MDL has on the population of pairs in the dataset.

                                          29-3

-------
                   Table 29-1. Method Precision by Analytical Method
Method/Pollutant
Group
voc
(TO- 15)
SNMOC
Carbonyl Compounds
(TO- 11 A)
PAH
(TO-13)
Metals Analysis
(Method IO-3. 5)
Hexavalent Chromium
(EPA-approved method)
MQO
Average
Coefficient of
Variation
(%)
11.88
11.59
6.83
11.83
13.58
14.54
Number of
Pairs Included
in the
Calculation
2,971
1,176
1,540
375
1,506
110
Total Number
of Pairs Without
the > MDL
exclusion
3,773
1,614
1,541
488
1,854
118
15.00 percent CV
       Tables 29-2 through 29-7 present method precision for VOCs, SNMOCs, carbonyl
compounds, PAHs, metals, and hexavalent chromium, respectively, as the CV per pollutant per
site, and the average CV per site, per pollutant, and per method. Also included in these tables is
the number of duplicate and/or collocated pairs included in the CV calculations. The shaded
rows in each table identify the NATTS MQO Core Analytes for each method. CVs exceeding the
15 percent MQO are bolded in each table. The CVs that exceed the program MQO for precision
are often driven by relatively low concentrations, even though they are greater than the MDL, as
these may result in relatively large CVs (i.e., the concentration difference between 0.01 ng/m3
and 0.02 ng/m3 is 100 percent).

29.2.1  VOC Method Precision
       Table 29-2 presents the method precision for all duplicate and collocated VOC samples
as the CV per pollutant per site, the average CV per site, the average CV per pollutant, and the
overall average CV for NMP sites sampling VOCs. The pollutant-specific CV ranged from
0 percent for a few pollutants for several sites to 134.72 percent (carbon disulfide for GLKY).
The pollutant-specific average CV ranged from 0 percent (a few pollutants) to 42.77 percent
(dibromochloromethane). Note that in these cases,  the number of pairs included in the
calculations are low. The site-specific average CV  ranged from 7.26 percent for S4MO to
20.00 percent for GLKY. The overall average method precision for VOCs was 11.88 percent.
                                         29-4

-------
Table 29-2. VOC Method Precision: Coefficient of Variation Based on Duplicate and
                              Collocated Samples by Site
Pollutant
Acetylene
Acrylonitrile
tert-Amyl Methyl Ether
Benzene
Bromochloromethane
Bromodichloromethane
Bromoform
Bromomethane
1,3 -Butadiene
Carbon Bisulfide
Carbon Tetrachloride
Chlorobenzene
Chloroethane
Chloroform
Chloromethane
Chloromethylbenzene
Chloroprene
Dibromochloromethane
1,2-Dibromoethane
/w-Dichlorobenzene
o-Dichlorobenzene
ฃ>-Dichlorobenzene
Dichlorodifluoromethane
1 , 1 -Dichloroethane
1 ,2-Dichloroethane
1 , 1 -Dichloroethene
cis- 1 ,2-Dichloroethylene
trans- 1 ,2-Dichloroethylene
Dichloromethane
1 ,2-Dichloropropane
cis- 1 ,3 -Dichloropropene
trans- 1 , 3 -Dichloropropene
Dichlorotetrafluoroethane
Ethyl Acrylate
Ethyl tert-Butyl Ether
Ethylbenzene
Hexachloro- 1 , 3 -butadiene
Methyl Ethyl Ketone
Methyl Isobutyl Ketone
Methyl Methacrylate
Methyl ter/-Butyl Ether
BTUT
5.51
NA
NA
17.92
NA
NA
NA
31.09
10.60
35.12
6.48
NA
39.12
3.57
6.39
NA
NA
NA
NA
NA
NA
NA
4.70
NA
10.28
0.00
NA
0.00
13.31
NA
NA
NA
7.47
NA
NA
8.27
NA
NA
15.67
NA
NA
BURVT
5.96
25.98
NA
8.61
NA
NA
NA
11.96
6.58
NA
22.80
NA
5.66
14.59
14.08
NA
NA
NA
NA
NA
NA
NA
7.36
NA
12.46
NA
NA
NA
38.38
NA
NA
NA
7.65
NA
NA
9.63
NA
NA
25.98
NA
NA
CHNJ
5.81
21.63
NA
6.36
NA
NA
NA
21.94
19.37
18.31
11.85
NA
18.86
13.41
10.91
NA
NA
NA
NA
NA
NA
NA
7.01
NA
0.00
NA
NA
NA
22.91
NA
NA
NA
14.02
NA
NA
8.53
NA
NA
10.88
NA
NA
DEMI
6.09
NA
NA
4.24
NA
NA
NA
3.37
6.71
18.74
7.94
NA
11.22
34.72
6.97
NA
NA
NA
NA
NA
NA
NA
6.54
NA
3.63
NA
NA
NA
16.61
NA
NA
NA
6.29
NA
NA
7.74
NA
NA
18.05
NA
NA
ELNJ
8.41
37.52
NA
9.93
NA
NA
NA
14.70
9.09
9.11
11.93
NA
8.32
9.56
8.43
NA
NA
NA
NA
NA
NA
3.82
8.32
NA
0.00
NA
NA
NA
9.59
NA
NA
NA
8.78
NA
NA
6.29
NA
NA
30.11
3.93
7.14
GLKY
5.27
81.91
NA
6.14
NA
NA
NA
8.69
20.13
134.72
4.34
NA
NA
4.76
5.98
NA
NA
NA
NA
NA
NA
NA
4.69
NA
4.04
NA
NA
NA
43.48
NA
NA
NA
3.21
NA
NA
15.26
NA
NA
32.47
NA
NA
 NA = No pairs with concentrations greater than or equal to the MDL
 BOLD ITALICS = EPA-designated NATTS Site
 Gray shading indicates NATTS MQO Core Analyte
 Orange shading indicates the site-specific average CV for this method; the overall average CV for this method
 is calculated from the site-specific averages and provided in the final column of the table.
                                          29-5

-------
Table 29-2. VOC Method Precision: Coefficient of Variation Based on Duplicate and
                       Collocated Samples by Site (Continued)
Pollutant
w-Octane
Propylene
Styrene
1 , 1 ,2,2-Tetrachloroethane
Tetrachloroethylene
Toluene
1 ,2,4-Trichlorobenzene
1,1,1 -Trichloroethane
1 , 1 ,2-Trichloroethane
Trichloroethylene
Trichlorofluoromethane
Trichlorotrifluoroethane
1 ,2,4-Trimethylbenzene
1 ,3 ,5-Trimethylbenzene
Vinyl chloride
m,p-Xylene
o-Xylene
Average by Site
BTUT
11.63
11.23
10.85
NA
6.67
8.14
NA
NA
NA
NA
5.18
4.78
8.97
16.52
NA
8.27
9.58
11.33
BURVT
11.26
13.72
11.18
NA
5.52
12.82
NA
NA
NA
NA
4.99
4.98
11.08
6.86
NA
11.54
9.80
12.36
CHNJ
12.50
7.21
9.07
NA
13.82
14.42
NA
NA
NA
NA
7.12
8.40
6.34
7.71
NA
9.08
8.22
11.69
DEMI
6.68
6.37
5.55
NA
5.99
11.00
NA
NA
NA
NA
4.15
3.51
12.03
7.10
NA
8.62
8.96
9.19
ELNJ
7.76
10.05
11.25
NA
9.69
5.10
NA
NA
NA
NA
8.44
8.08
8.92
10.93
NA
6.45
6.98
9.95
GLKY
15.70
11.64
11.80
NA
NA
15.50
NA
NA
NA
NA
3.11
2.91
9.09
NA
NA
20.64
14.41
20.00
 NA = No pairs with concentrations greater than or equal to the MDL
 BOLD ITALICS = EPA-designated NATTS Site
 Gray shading indicates NATTS MQO Core Analyte
 Orange shading indicates the site-specific average CV for this method; the overall average CV for this method
 is calculated from the site-specific averages and provided in the final column of the table.
                                          29-6

-------
Table 29-2. VOC Method Precision: Coefficient of Variation Based on Duplicate and
                       Collocated Samples by Site (Continued)
Pollutant
Acetylene
Acrylonitrile
tert-Amyl Methyl Ether
Benzene
Bromochloromethane
Bromodichloromethane
Bromoform
Bromomethane
1,3 -Butadiene
Carbon Bisulfide
Carbon Tetrachloride
Chlorobenzene
Chloroethane
Chloroform
Chloromethane
Chloromethylbenzene
Chloroprene
Dibromochloromethane
1,2-Dibromoethane
/w-Dichlorobenzene
o-Dichlorobenzene
ฃ>-Dichlorobenzene
Dichlorodifluoromethane
1 , 1 -Dichloroethane
1 ,2-Dichloroethane
1 , 1 -Dichloroethene
cis- 1 ,2-Dichloroethylene
trans- 1 ,2-Dichloroethylene
Dichloromethane
1 ,2-Dichloropropane
cis- 1 ,3 -Dichloropropene
trans- 1 , 3 -Dichloropropene
Dichlorotetrafluoroethane
Ethyl Acrylate
Ethyl tert-Butyl Ether
Ethylbenzene
Hexachloro- 1 ,3 -butadiene
Methyl Ethyl Ketone
Methyl Isobutyl Ketone
Methyl Methacrylate
Methyl tert-Butyl Ether
GPCO
3.10
1.91
NA
5.76
NA
NA
NA
11.58
5.48
10.17
27.27
NA
2.24
7.07
1.77
NA
NA
NA
NA
NA
NA
NA
1.92
NA
0.00
NA
NA
NA
21.33
NA
NA
NA
4.66
NA
NA
29.96
NA
NA
15.80
NA
NA
MWOK
3.44
NA
NA
3.94
NA
NA
NA
57.64
3.81
NA
8.39
NA
NA
4.04
7.71
NA
NA
NA
NA
NA
NA
NA
3.13
NA
0.00
NA
NA
NA
8.82
NA
NA
NA
4.14
NA
NA
6.17
NA
NA
17.78
NA
NA
NBIL
2.16
13.58
NA
15.01
NA
41.30
34.49
10.68
6.17
7.78
20.65
NA
10.27
34.50
4.35
NA
NA
42.77
NA
NA
NA
NA
2.77
NA
5.00
NA
NA
NA
30.01
NA
NA
NA
4.74
NA
NA
6.72
NA
NA
24.84
NA
NA
NBNJ
8.12
NA
NA
23.48
NA
NA
NA
18.21
7.19
15.72
11.88
NA
40.41
11.61
6.08
NA
NA
NA
NA
NA
NA
NA
6.38
NA
3.29
NA
NA
NA
14.40
NA
NA
NA
13.33
NA
NA
13.01
NA
NA
10.03
NA
6.15
OCOK
36.98
NA
NA
53.15
NA
NA
NA
13.55
70.59
NA
20.39
NA
NA
0.00
2.26
NA
NA
NA
NA
NA
NA
NA
2.32
NA
NA
NA
NA
NA
7.42
NA
NA
NA
5.43
NA
NA
14.88
NA
NA
25.67
NA
NA
PANJ
6.97
NA
NA
13.66
NA
NA
NA
8.84
2.28
NA
10.88
NA
19.00
1.75
4.81
NA
NA
NA
NA
NA
NA
72.04
8.69
NA
0.00
NA
NA
NA
0.20
NA
NA
NA
9.87
NA
NA
9.85
NA
NA
39.31
NA
NA
 NA = No pairs with concentrations greater than or equal to the MDL
 BOLD ITALICS = EPA-designated NATTS Site
 Gray shading indicates NATTS MQO Core Analyte
 Orange shading indicates the site-specific average CV for this method; the overall average CV for this method
 is calculated from the site-specific averages and provided in the final column of the table.
                                          29-7

-------
Table 29-2. VOC Method Precision: Coefficient of Variation Based on Duplicate and
                       Collocated Samples by Site (Continued)
Pollutant
w-Octane
Propylene
Styrene
1, 1,2,2-Tetrachloroethane
Tetrachloroethylene
Toluene
1 ,2,4-Trichlorobenzene
1,1,1 -Trichloroethane
1 , 1 ,2-Trichloroethane
Trichloroethylene
Trichlorofluoromethane
Trichlorotrifluoroethane
1 ,2,4-Trimethylbenzene
1 ,3 ,5-Trimethylbenzene
Vinyl chloride
m,p-Xylene
o-Xylene
Average by Site
GPCO
19.10
2.94
19.64
NA
6.85
15.86
NA
NA
NA
1.17
2.83
3.14
54.82
53.68
NA
39.37
38.85
14.58
MWOK
8.97
5.89
8.40
NA
7.96
5.71
NA
NA
NA
NA
3.33
4.31
13.85
9.85
NA
8.02
12.84
9.09
NBIL
13.91
4.53
8.52
NA
5.16
6.38
NA
NA
NA
NA
4.37
1.86
11.81
10.44
NA
6.77
6.76
13.28
NBNJ
8.91
14.87
10.10
NA
16.00
9.35
NA
NA
NA
NA
6.13
12.31
6.93
8.93
NA
14.41
13.57
12.25
OCOK
10.05
56.41
26.16
NA
0.75
32.49
NA
NA
NA
NA
3.11
4.43
17.64
12.03
NA
13.16
25.33
19.75
PANJ
19.96
0.37
10.88
NA
5.14
19.92
NA
NA
NA
NA
10.07
7.07
32.12
25.63
NA
17.95
20.20
14.52
 NA = No pairs with concentrations greater than or equal to the MDL
 BOLD ITALICS = EPA-designated NATTS Site
 Gray shading indicates NATTS MQO Core Analyte
 Orange shading indicates the site-specific average CV for this method; the overall average CV for this method
 is calculated from the site-specific averages and provided in the final column of the table.
                                          29-8

-------
Table 29-2. VOC Method Precision: Coefficient of Variation Based on Duplicate and
                       Collocated Samples by Site (Continued)
Pollutant
Acetylene
Acrylonitrile
tert-Amyl Methyl Ether
Benzene
Bromochloromethane
Bromodichloromethane
Bromoform
Bromomethane
1,3 -Butadiene
Carbon Bisulfide
Carbon Tetrachloride
Chlorobenzene
Chloroethane
Chloroform
Chloromethane
Chloromethylbenzene
Chloroprene
Dibromochloromethane
1,2-Dibromoethane
/w-Dichlorobenzene
o-Dichlorobenzene
ฃ>-Dichlorobenzene
Dichlorodifluoromethane
1 , 1 -Dichloroethane
1 ,2-Dichloroethane
1 , 1 -Dichloroethene
cis- 1 ,2-Dichloroethylene
/raซ5-l,2-Dichloroethylene
Dichloromethane
1 ,2-Dichloropropane
cis- 1 ,3 -Dichloropropene
trans- 1 , 3 -Dichloropropene
Dichlorotetrafluoroethane
Ethyl Aery late
Ethyl tert-Butyl Ether
Ethylbenzene
Hexachloro- 1 ,3 -butadiene
Methyl Ethyl Ketone
Methyl Isobutyl Ketone
Methyl Methacrylate
Methyl ter/-Butyl Ether
PROK
9.63
33.67
NA
11.22
NA
NA
NA
8.99
10.24
NA
25.33
NA
5.66
NA
8.53
NA
NA
NA
NA
NA
NA
44.84
7.73
NA
14.43
NA
NA
NA
8.60
NA
NA
NA
8.70
NA
NA
12.88
NA
NA
21.81
NA
NA
PXSS
4.70
NA
NA
6.96
NA
NA
NA
3.41
4.11
51.10
8.10
NA
0.00
5.14
3.47
NA
NA
NA
NA
NA
NA
2.44
3.50
NA
6.73
NA
NA
NA
58.38
NA
NA
NA
5.61
NA
NA
4.06
NA
NA
17.20
NA
NA
S4MO
4.91
NA
NA
5.83
NA
NA
NA
8.50
5.26
5.16
6.54
NA
3.45
3.74
7.17
NA
NA
NA
NA
NA
NA
5.74
3.66
NA
8.87
NA
NA
NA
10.94
NA
NA
NA
4.55
NA
NA
4.82
NA
NA
7.58
NA
NA
SEWA
4.00
NA
NA
4.62
NA
NA
NA
11.65
5.02
NA
13.14
NA
NA
16.59
4.35
NA
NA
NA
NA
NA
NA
NA
4.01
NA
11.47
NA
NA
NA
68.48
NA
NA
NA
30.09
NA
NA
5.87
NA
NA
20.19
NA
NA
SPIL
3.02
4.81
NA
8.71
NA
NA
NA
4.35
28.49
3.80
12.38
NA
5.24
14.29
3.91
NA
NA
NA
NA
NA
NA
NA
5.65
NA
9.67
NA
NA
NA
9.73
NA
NA
NA
3.42
NA
NA
32.32
NA
NA
7.33
NA
NA
SSSD
5.55
NA
NA
9.01
NA
NA
NA
9.35
7.39
NA
43.22
NA
4.88
0.00
6.19
NA
NA
NA
NA
NA
NA
NA
4.80
NA
3.86
NA
NA
NA
11.40
NA
NA
NA
6.00
NA
NA
3.77
NA
NA
23.38
NA
26.93
 NA = No pairs with concentrations greater than or equal to the MDL
 BOLD ITALICS = EPA-designated NATTS Site
 Gray shading indicates NATTS MQO Core Analyte
 Orange shading indicates the site-specific average CV for this method; the overall average CV for this method
 is calculated from the site-specific averages and provided in the final column of the table.
                                          29-9

-------
Table 29-2. VOC Method Precision: Coefficient of Variation Based on Duplicate and
                       Collocated Samples by Site (Continued)
Pollutant
w-Octane
Propylene
Styrene
1 , 1 ,2,2-Tetrachloroethane
Tetrachloroethylene
Toluene
1 ,2,4-Trichlorobenzene
1,1,1 -Trichloroethane
1 , 1 ,2-Trichloroethane
Trichloroethylene
Trichlorofluoromethane
Trichlorotrifluoroethane
1 ,2,4-Trimethylbenzene
1 ,3 ,5-Trimethylbenzene
Vinyl chloride
m,p-Xylene
o-Xylene
Average by Site
PROK
27.83
7.35
11.33
NA
NA
10.40
NA
NA
NA
NA
8.00
8.60
12.69
NA
NA
12.37
16.50
14.47
PXSS
6.15
3.78
5.21
NA
5.47
4.83
NA
NA
NA
NA
3.32
4.03
4.27
3.08
NA
4.88
4.14
8.67
S4MO
7.13
6.76
19.41
NA
11.67
3.79
NA
NA
NA
NA
4.05
5.02
13.22
24.70
0.00
4.31
6.60
7.26
SEWA
11.41
4.11
10.68
NA
0.88
7.13
NA
NA
NA
5.44
4.20
5.46
6.61
1.71
NA
5.01
5.74
10.71
SPIL
7.17
16.63
4.05
NA
3.76
6.68
NA
NA
NA
7.67
5.14
5.57
12.85
3.54
NA
32.71
8.60
9.70
SSSD
10.20
8.77
5.83
NA
NA
7.52
NA
NA
NA
NA
4.25
4.86
6.57
5.26
NA
5.59
4.98
9.18
 NA = No pairs with concentrations greater than or equal to the MDL
 BOLD ITALICS = EPA-designated NATTS Site
 Gray shading indicates NATTS MQO Core Analyte
 Orange shading indicates the site-specific average CV for this method; the overall average CV for this method
 is calculated from the site-specific averages and provided in the final column of the table.
                                         29-10

-------
Table 29-2. VOC Method Precision: Coefficient of Variation Based on Duplicate and
                       Collocated Samples by Site (Continued)
Pollutant
Acetylene
Acrylonitrile
tert-Amyl Methyl Ether
Benzene
Bromochloromethane
Bromodichloromethane
Bromoform
Bromomethane
1,3 -Butadiene
Carbon Bisulfide
Carbon Tetrachloride
Chlorobenzene
Chloroethane
Chloroform
Chloromethane
Chloromethylbenzene
Chloroprene
Dibromochloromethane
1 ,2-Dibromoethane
/w-Dichlorobenzene
o-Dichlorobenzene
ฃ>-Dichlorobenzene
Dichlorodifluoromethane
1 , 1 -Dichloroethane
1 ,2-Dichloroethane
1 , 1 -Dichloroethene
cis- 1 ,2-Dichloroethylene
trans- 1 ,2-Dichloroethylene
Dichloromethane
1 ,2-Dichloropropane
cis- 1 ,3 -Dichloropropene
trans- 1 ,3 -Dichloropropene
Dichlorotetrafluoroethane
Ethyl Acrylate
Ethyl tert-Butyl Ether
Ethylbenzene
Hexachloro- 1 ,3 -butadiene
Methyl Ethyl Ketone
Methyl Isobutyl Ketone
Methyl Methacrylate
Methyl tert-Butyl Ether
TMOK
3.85
17.16
NA
6.50
NA
NA
NA
9.01
6.15
31.95
27.46
NA
5.24
6.43
3.92
NA
NA
NA
NA
NA
NA
NA
4.15
NA
2.57
NA
NA
NA
4.95
NA
NA
NA
3.60
NA
NA
15.25
NA
NA
51.20
NA
NA
TOOK
5.84
NA
NA
12.52
NA
NA
NA
8.74
13.05
NA
8.21
NA
22.52
3.14
6.98
NA
NA
NA
NA
NA
NA
26.05
4.30
NA
0.00
NA
NA
NA
13.52
NA
NA
NA
7.90
NA
NA
9.11
NA
NA
10.53
NA
NA
UCSD
6.32
22.48
NA
9.13
NA
NA
NA
7.41
0.00
8.17
10.85
NA
18.27
NA
6.67
NA
NA
NA
NA
NA
NA
NA
5.51
NA
5.39
NA
NA
NA
17.19
NA
NA
NA
5.04
NA
NA
8.83
NA
NA
6.95
NA
NA
#of
Pairs
139
17
NA
139
NA
6
1
89
123
56
138
NA
22
62
139
NA
NA
5
NA
NA
NA
11
139
NA
35
1
NA
1
136
NA
NA
NA
138
NA
NA
138
NA
NA
101
1
5
Average by
Pollutant
6.94
26.07
NA
11.56
NA
41.30
34.49
13.51
11.80
26.91
15.24
NA
12.96
9.94
6.24
NA
NA
42.77
NA
NA
NA
25.82
5.10
NA
5.08
0.00
NA
0.00
20.46
NA
NA
NA
7.83
NA
NA
11.10
NA
NA
20.61
3.93
13.41
     NA = No pairs with concentrations greater than or equal to the MDL
     BOLD ITALICS = EPA-designated NATTS Site
     Gray shading indicates NATTS MQO Core Analyte
     Orange shading indicates the site-specific average CV for this method; the overall average CV for this
     method is calculated from the site-specific averages and provided in the final column of the table.
                                         29-11

-------
   Table 29-2. VOC Method Precision: Coefficient of Variation Based on Duplicate and
                         Collocated Samples by Site (Continued)
Pollutant
w-Octane
Propylene
Styrene
1, 1,2,2-Tetrachloroethane
Tetrachloroethylene
Toluene
1 ,2,4-Trichlorobenzene
1,1,1 -Trichloroethane
1,1,2-Trichloroethane
Trichloroethylene
Trichlorofluoromethane
Trichlorotrifluoroethane
1 ,2,4-Trimethylbenzene
1 , 3 ,5 -Trimethy Ibenzene
Vinyl chloride
m,p-Xylene
o-Xylene
Average by Site
TMOK
8.51
5.59
21.97
NA
8.00
15.13
NA
NA
NA
NA
3.54
6.34
30.10
8.50
NA
20.61
26.26
13.11
TOOK
10.31
38.67
3.79
NA
8.32
16.00
NA
NA
NA
NA
6.21
4.89
5.09
5.56
NA
6.83
7.68
10.22
UCSD
8.53
15.05
19.03
NA
3.82
6.54
NA
NA
NA
NA
5.64
5.14
0.00
NA
0.00
5.08
4.52
8.14
#of
Pairs
134
139
112
NA
61
139
NA
NA
NA
5
139
139
126
65
2
135
133
-
Average by
Pollutant
11.60
12.00
11.65
NA
6.97
11.18
NA
NA
NA
4.76
5.10
5.51
13.57
12.34
0.00
12.46
12.41
11.88
        NA = No pairs with concentrations greater than or equal to the MDL
        BOLD ITALICS = EPA-designated NATTS Site
        Gray shading indicates NATTS MQO Core Analyte
        Orange shading indicates the site-specific average CV for this method; the overall average CV for this
        method is calculated from the site-specific averages and provided in the final column of the table.

29.2.2 SNMOC Method Precision
       The SNMOC method precision for duplicate and collocated samples is presented in
Table 29-3 as the CV per pollutant per site, the average CV per site, the average CV per
pollutant, and the overall average CV for NMP sites sampling SNMOCs. The results from
duplicate and collocated samples show low- to high-level variability among pollutants and sites,
ranging from a CV of 0 percent (1,3-butadiene for RICO) to 111.81 percent (cyclopentene for
RICO). The pollutant-specific average CV ranged from 0.76 percent (w-tridecane) to
36.81 percent (cyclopentene). Note that in these cases, the number of pairs included in the
calculations is low. The site-specific average CV ranged from 6.69 percent for RICO to
16.27 percent for PACO, with an overall method average of 11.59 percent.
                                          29-12

-------
    Table 29-3. SNMOC Method Precision: Coefficient of Variation
          Based on Duplicate and Collocated Samples by Site
Pollutant
Acetylene
Benzene
1,3 -Butadiene
w-Butane
c/s-2-Butene
/raws-2-Butene
Cyclohexane
Cyclopentane
Cyclopentene
w-Decane
1-Decene
/w-Diethylbenzene
ฃ>-Diethylbenzene
2,2-Dimethylbutane
2,3 -Dimethylbutane
2,3 -Dimethylpentane
2,4-Dimethylpentane
w-Dodecane
1-Dodecene
Ethane
2-Ethyl-l-butene
Ethylbenzene
Ethylene
/w-Ethyltoluene
o-Ethyltoluene
ฃ>-Ethyltoluene
w-Heptane
1-Heptene
w-Hexane
1-Hexene
c/s-2-Hexene
/raซs-2-Hexene
Isobutane
Isobutene/ 1 -Butene
Isopentane
Isoprene
Isopropylbenzene
2-Methyl-l-butene
3-Methyl-l-butene
2-Methyl- 1 -pentene
BMCO
4.20
22.62
NA
NA
4.36
14.36
8.57
6.45
9.75
7.24
NA
3.45
38.12
7.55
NA
NA
9.00
1.89
0.96
1.93
NA
18.13
1.15
12.65
9.30
NA
20.04
NA
20.87
NA
NA
NA
2.14
NA
7.14
NA
NA
4.03
NA
NA
BTUT
8.69
5.52
10.57
1.68
4.23
50.21
6.25
9.97
39.65
9.69
NA
22.32
16.52
6.08
11.78
8.88
11.83
8.36
18.21
1.04
NA
6.34
4.40
8.11
6.02
11.13
4.56
NA
3.46
NA
NA
NA
1.47
12.05
9.62
9.43
NA
7.31
NA
NA
NBIL
5.66
14.77
NA
8.94
NA
18.46
9.98
13.04
15.48
24.07
29.73
14.44
20.92
9.21
7.34
11.15
7.69
33.55
NA
22.69
NA
10.36
5.26
18.53
11.82
13.54
11.93
NA
11.54
NA
NA
NA
10.74
11.60
11.44
12.08
NA
NA
NA
NA
PACO
18.44
6.46
NA
3.24
5.37
34.08
36.80
18.69
7.37
6.79
NA
9.29
12.67
22.02
2.83
4.48
5.91
9.73
NA
35.08
NA
9.79
2.57
8.25
4.99
4.00
32.07
3.88
34.26
NA
NA
NA
39.12
6.21
28.73
6.70
3.43
NA
NA
NA
RICO
1.66
10.17
0.00
5.83
7.22
6.57
0.39
1.32
111.81
4.14
NA
5.10
17.46
4.82
5.36
4.27
9.64
2.81
8.34
0.36
NA
3.99
2.81
1.47
20.72
8.00
1.99
0.45
4.05
NA
NA
NA
2.17
NA
8.26
2.89
NA
1.76
NA
NA
NA = No pairs with concentrations greater than or equal to the MDL
BOLD ITALICS = EPA-designaled NATTS Site
Orange shading indicates the site-specific average CV for this method; the overall average CV for
this method is calculated from the site-specific averages and provided in the final column of the table.
                                  29-13

-------
    Table 29-3. SNMOC Method Precision: Coefficient of Variation
    Based on Duplicate and Collocated Samples by Site (Continued)
Pollutant
4-Methyl- 1 -pentene
2-Methyl-2-butene
Methylcyclohexane
Methylcyclopentane
2-Methylheptane
3-Methylheptane
2-Methylhexane
3-Methylhexane
2-Methylpentane
3-Methylpentane
w-Nonane
1-Nonene
w-Octane
1-Octene
w-Pentane
1 -Pentene
c/s-2-Pentene
/raws-2-Pentene
a-Pinene
6-Pinene
Propane
w-Propylbenzene
Propylene
Propyne
Styrene
Toluene
w-Tridecane
1-Tridecene
1 ,2,3 -Trimethylbenzene
1 ,2,4-Trimethylbenzene
1 ,3 ,5-Trimethylbenzene
2,2,3-Trimethylpentane
2,2,4-Trimethylpentane
2,3 ,4-Trimethylpentane
w-Undecane
1-Undecene
7w-Xylene//?-Xylene
o-Xylene
Average by Site
BMCO
NA
7.74
15.38
14.63
13.95
31.78
13.08
10.64
15.53
12.16
20.71
NA
9.62
NA
12.74
2.90
NA
NA
NA
NA
7.05
5.71
10.71
NA
NA
37.22
NA
NA
22.81
16.26
26.94
NA
NA
19.21
4.53
NA
NA
25.65
12.61
BTUT
NA
9.85
4.52
3.98
7.13
5.22
5.92
5.25
5.00
6.22
7.72
NA
11.73
NA
11.42
14.86
NA
6.08
32.81
2.42
1.56
3.81
6.44
NA
4.06
6.11
0.76
NA
8.99
6.66
1.77
2.37
8.49
6.78
10.48
3.39
9.26
5.68
8.92
NBIL
NA
7.59
16.85
10.24
40.13
29.85
23.77
14.86
11.62
9.11
19.61
33.64
15.90
NA
5.64
17.33
NA
2.02
5.58
NA
19.53
15.61
5.11
29.19
NA
8.03
NA
NA
17.02
13.79
26.54
0.14
9.18
5.17
24.29
41.10
8.98
11.73
15.17
PACO
NA
11.80
30.65
31.27
26.98
26.69
2.21
25.51
20.66
28.04
15.72
10.90
30.03
8.64
57.12
24.34
NA
5.81
NA
NA
39.92
7.14
7.95
NA
NA
27.29
NA
NA
2.13
6.45
2.64
1.65
1.75
11.39
3.54
67.91
0.67
15.98
16.27
RICO
NA
1.78
1.52
0.70
1.28
6.31
2.51
5.01
2.57
0.84
2.35
NA
2.22
NA
3.63
5.26
9.30
3.59
7.64
NA
2.07
4.88
13.60
NA
NA
1.83
NA
NA
4.25
2.76
2.57
NA
17.52
1.68
13.74
NA
4.48
3.25
6.69
NA = No pairs with concentrations greater than or equal to the MDL
BOLD ITALICS = EPA-designaled NATTS Site
Orange shading indicates the site-specific average CV for this method; the overall average CV for
this method is calculated from the site-specific averages and provided in the final column of the table.
                                  29-14

-------
Table 29-3. SNMOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site (Continued)
Pollutant
Acetylene
Benzene
1,3 -Butadiene
w-Butane
c/s-2-Butene
/raซs-2-Butene
Cyclohexane
Cyclopentane
Cyclopentene
w-Decane
1-Decene
/w-Diethylbenzene
/>-Diethylbenzene
2,2-Dimethylbutane
2,3 -Dimethylbutane
2,3 -Dimethylpentane
2,4-Dimethylpentane
w-Dodecane
1-Dodecene
Ethane
2-Ethyl-l-butene
Ethylbenzene
Ethylene
/w-Ethyltoluene
o-Ethyltoluene
/>-Ethyltoluene
w-Heptane
1-Heptene
w-Hexane
1-Hexene
c/s-2-Hexene
/raซs-2-Hexene
Isobutane
Isobutene/ 1 -Butene
Isopentane
Isoprene
Isopropylbenzene
2-Methyl-l-butene
3 -Methyl- 1 -butene
2-Methyl- 1 -pentene
SSSD
2.28
8.23
NA
2.98
4.19
5.83
11.88
21.14
NA
2.54
NA
13.14
13.70
9.18
5.36
1.36
3.57
8.46
52.37
1.40
NA
12.75
3.21
5.85
2.50
3.36
3.04
NA
7.44
NA
NA
NA
19.56
23.99
25.45
2.58
NA
14.61
NA
NA
UCSD
3.38
8.39
NA
2.73
NA
15.99
3.04
17.78
NA
NA
NA
15.39
4.43
3.53
4.30
NA
NA
10.69
90.37
1.52
NA
13.86
4.70
NA
NA
NA
3.26
NA
5.59
NA
NA
NA
5.13
5.38
15.82
3.19
NA
NA
NA
NA
#of
Pairs
30
29
3
20
10
16
22
24
6
19
1
25
21
26
16
15
16
19
8
30
0
26
30
19
15
8
22
3
29
0
0
0
30
11
28
16
1
10
0
0
Average
by
Pollutant
6.33
10.88
5.28
4.23
5.07
20.79
10.99
12.62
36.81
9.08
29.73
11.87
17.69
8.91
6.16
6.03
7.94
10.78
34.05
9.15
NA
10.74
3.44
9.14
9.23
8.00
10.98
2.17
12.46
NA
NA
NA
11.48
11.85
15.21
6.15
3.43
6.93
NA
NA
NA = No pairs with concentrations greater than or equal to the MDL
BOLD ITALICS = EPA-designated NATTS Site
Orange shading indicates the site-specific average CV for this method; the overall
average CV for this method is calculated from the site-specific averages and provided
in the final column of the table.
                              29-15

-------
Table 29-3. SNMOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site (Continued)
Pollutant
4-Methyl- 1 -pentene
2-Methyl-2-butene
Methylcyclohexane
Methylcyclopentane
2-Methylheptane
3-Methylheptane
2-Methylhexane
3 -Methy Ihexane
2-Methylpentane
3-Methylpentane
w-Nonane
1-Nonene
w-Octane
1-Octene
w-Pentane
1 -Pentene
cis-2 -Pentene
/raซs-2-Pentene
a-Pinene
6-Pinene
Propane
w-Propylbenzene
Propylene
Propyne
Styrene
Toluene
w-Tridecane
1-Tridecene
1 ,2,3 -Trimethylbenzene
1 ,2,4-Trimethylbenzene
1 ,3 ,5-Trimethylbenzene
2,2,3 -Trimethylpentane
2,2,4-Trimethylpentane
2,3 ,4-Trimethylpentane
w-Undecane
1-Undecene
/w-Xylene/p-Xylene
o-Xylene
Average by Site
SSSD
NA
NA
3.24
7.61
NA
NA
18.37
11.79
23.06
10.01
1.25
NA
3.00
NA
47.27
18.88
NA
7.37
3.89
NA
1.40
NA
5.38
NA
1.72
13.23
NA
NA
6.83
12.98
6.76
NA
10.00
17.81
20.84
NA
1.04
7.20
10.43
UCSD
NA
NA
0.68
13.90
NA
NA
10.78
18.73
5.68
5.24
NA
NA
NA
NA
19.65
30.02
NA
NA
NA
NA
2.25
NA
11.97
NA
17.40
7.22
NA
NA
NA
9.76
NA
NA
3.58
NA
NA
14.34
3.91
4.30
11.00
#of
Pairs
0
13
22
27
12
12
26
26
30
28
18
3
18
1
30
22
1
17
5
1
30
7
30
1
3
29
2
0
14
27
12
4
24
18
16
4
15
24
~
Average
by
Pollutant
NA
7.75
10.41
11.76
17.89
19.97
10.95
13.11
12.02
10.23
11.23
22.27
12.09
8.64
22.49
16.23
9.30
4.97
12.48
2.42
10.54
7.43
8.74
29.19
7.73
14.42
0.76
NA
10.34
9.81
11.20
1.39
8.42
10.34
12.90
31.68
4.72
10.54
11.59
NA = No pairs with concentrations greater than or equal to the MDL
BOLD ITALICS = EPA-designaled NATTS Site
Orange shading indicates the site-specific average CV for this method; the overall
average CV for this method is calculated from the site-specific averages and provided
in the final column of the table.
                              29-16

-------
29.2.3 Carbonyl Compound Method Precision
       Table 29-4 presents the method precision for duplicate and collocated carbonyl
compound samples as the CV per pollutant per site, the average CV per site, the average CV per
pollutant,  and the overall average CV for NMP sites sampling carbonyl compounds. The
duplicate and collocated sample results show low- to mid-level variability among the sites,
ranging from a CV of 0.86 percent (formaldehyde for MWOK) to 49.32 percent (tolualdehydes
for UCSD). The pollutant-specific average CV ranged from 3.32 percent (acetaldehyde) to
17.74 percent (tolualdehydes). The site-specific average CV ranged from 3.92 percent for OCOK
to 11.59 percent for UCSD. The average CV for every site is less than the program MQO of
15 percent. The overall average method precision was 6.83 percent for carbonyl compounds.

   Table 29-4. Carbonyl Compound Method Precision: Coefficient of Variation Based on
                        Duplicate and Collocated Samples by Site
Pollutant
Acetaldehyde
Acetone
Benzaldehyde
Butyraldehyde
Crotonaldehyde
2,5-Dimethylbenzaldehyde
Formaldehyde
Hexaldehyde
Isovaleraldehyde
Propionaldehyde
Tolualdehydes
Valeraldehyde
Average by Site
AZFL
3.62
7.03
7.14
5.31
2.61
NA
6.28
7.19
NA
2.80
18.13
8.91
6.90
BTUT
2.30
3.53
4.88
4.24
1.66
NA
5.64
8.15
NA
2.60
26.30
4.69
6.40
CHNJ
2.41
6.37
7.46
3.91
3.53
NA
3.58
4.18
NA
4.88
9.49
1.35
4.71
DEMI
10.13
4.57
9.12
12.76
16.58
NA
11.67
11.00
NA
8.35
9.93
9.91
10.40
ELNJ
1.80
10.82
3.76
5.64
5.88
NA
2.37
10.39
NA
2.12
22.21
4.24
6.92
GLKY
3.13
5.96
7.66
4.23
5.94
NA
2.51
2.86
NA
3.31
8.90
8.93
5.34
GPCO
2.08
6.34
5.34
4.50
3.90
NA
2.21
7.69
NA
3.17
36.57
6.23
7.81
 NA = No pairs with concentrations greater than or equal to the MDL
 BOLD ITALICS = EPA-designaled NATTS Site
 Gray shading indicates NATTS MQO Core Analyte
 Orange shading indicates the site-specific average CV for this method; the overall average CV for this method is
 calculated from the site-specific averages and provided in the final column of the table.
                                          29-17

-------
  Table 29-4. Carbonyl Compound Method Precision: Coefficient of Variation Based on
                  Duplicate and Collocated Samples by Site (Continued)
Pollutant
Acetaldehyde
Acetone
Benzaldehyde
Butyraldehyde
Crotonaldehyde
2,5-Dimethylbenzaldehyde
Formaldehyde
Hexaldehyde
Isovaleraldehyde
Propionaldehyde
Tolualdehydes
Valeraldehyde
Average by Site
INDEM
7.65
9.23
9.96
5.81
7.15
NA
9.55
8.68
NA
6.45
8.34
7.96
8.08
MWOK
1.82
10.78
6.04
6.34
2.69
NA
0.86
9.20
NA
4.35
7.02
8.58
5.77
NBIL
2.25
7.49
5.91
6.91
4.20
NA
2.89
7.62
NA
2.01
10.32
5.03
5.46
NBNJ
8.57
4.76
6.51
7.64
5.94
NA
6.50
10.72
NA
7.01
37.94
8.47
10.41
OCOK
1.57
4.83
4.81
1.62
3.64
NA
1.90
5.70
NA
1.45
9.41
4.24
3.92
ORFL
7.19
16.62
7.14
7.98
7.71
NA
3.79
10.05
NA
4.57
36.96
6.31
10.83
PROK
1.17
2.80
9.12
4.79
2.66
NA
2.16
4.90
NA
3.16
6.87
6.60
4.42
NA = No pairs with concentrations greater than or equal to the MDL
BOLD ITALICS = EPA-designated NATTS Site
Gray shading indicates NATTS MQO Core Analyte
Orange shading indicates the site-specific average CV for this method; the overall average CV for this method is
calculated from the site-specific averages and provided in the final column of the table.


  Table 29-4. Carbonyl Compound Method Precision: Coefficient of Variation Based on
                  Duplicate and Collocated Samples by Site (Continued)
Pollutant
Acetaldehyde
Acetone
Benzaldehyde
Butyraldehyde
Crotonaldehyde
2,5-Dimethylbenzaldehyde
Formaldehyde
Hexaldehyde
Isovaleraldehyde
Propionaldehyde
Tolualdehydes
Valeraldehyde
Average by Site
PXSS
2.64
4.43
3.61
3.21
3.08
NA
5.82
5.77
NA
3.36
9.46
8.30
4.97
S4MO
1.39
3.46
4.98
3.92
3.70
NA
3.19
7.36
NA
4.18
24.50
6.71
6.34
SEWA
1.84
1.89
12.41
3.58
7.54
NA
3.96
6.38
NA
4.66
8.85
3.75
5.49
SKFL
2.58
8.69
8.15
8.18
6.83
NA
2.64
8.05
NA
2.83
37.07
5.44
9.05
SPIL
1.74
8.10
4.10
4.16
8.10
NA
3.21
5.65
NA
1.89
9.41
4.43
5.08
SSSD
4.41
18.22
7.03
6.01
3.97
NA
8.29
9.06
NA
5.37
11.33
7.00
8.07
SYFL
2.07
7.56
9.07
4.13
4.81
NA
2.81
7.37
NA
3.07
13.73
6.94
6.16
NA = No pairs with concentrations greater than or equal to the MDL
BOLD ITALICS = EPA-designated NATTS Site
Gray shading indicates NATTS MQO Core Analyte
Orange shading indicates the site-specific average CV for this method; the overall average CV for this method is
calculated from the site-specific averages and provided in the final column of the table.
                                           29-18

-------
   Table 29-4. Carbonyl Compound Method Precision: Coefficient of Variation Based on
                  Duplicate and Collocated Samples by Site (Continued)
Pollutant
Acetaldehyde
Acetone
Benzaldehyde
Butyraldehyde
Crotonaldehyde
2,5-Dimethylbenzaldehyde
Formaldehyde
Hexaldehyde
Isovaleraldehyde
Propionaldehyde
Tolualdehydes
Valeraldehyde
Average by Site
TMOK
2.02
4.21
3.10
3.64
2.51
NA
1.79
6.17
NA
4.23
9.70
3.17
4.05
TOOK
1.89
1.60
7.85
3.42
3.08
NA
1.02
7.03
NA
2.33
7.05
6.77
4.20
UCSD
2.25
12.68
6.53
7.15
7.50
NA
6.62
12.55
NA
5.03
49.32
6.23
11.59
WPIN
4.61
8.63
5.24
6.79
10.05
NA
4.93
11.31
NA
5.89
14.72
11.20
8.34
# of Pairs
158
158
157
157
153
0
158
157
0
158
131
153
~
Average by
Pollutant
3.32
7.22
6.68
5.43
5.41
NA
4.25
7.80
NA
3.96
17.74
6.46
6.83
   NA = No pairs with concentrations greater than or equal to the MDL
   BOLD ITALICS = EPA-designaled NATTS Site
   Gray shading indicates NATTS MQO Core Analyte
   Orange shading indicates the site-specific average CV for this method; the overall average CV for this method
   is calculated from the site-specific averages and provided in the final column of the table.

29.2.4  PAH Method Precision
       The method precision results for collocated PAH samples are shown in Table 29-5 as the
CV per pollutant per site, the average CV per site, the average CV per pollutant, and the overall
average CV for NMP sites sampling PAHs. All samples evaluated in this section are collocated
samples. Collocated systems were the responsibility of the participating agency for sites
sampling PAHs. Thus, collocated samples were not collected at most PAH sites because few
sites had collocated samplers. Therefore, the  method precision data for PAHs is based on only
five sites for 2011.

       The results from collocated samples show low- to mid-level variability among sites,
ranging from a CV of 0.38 percent (benzo(a)anthracene for SDGA) to 58.36 percent (anthracene
for DEMI). The pollutant-specific average  CV ranged from 1.65 percent (cyclopenta[cd]pyrene)
to 21.17 percent (anthracene). The site-specific average CV ranged from 5.98 percent for SEWA
to 19.42 percent for RUCA. The overall average method precision was 11.83 percent.
                                          29-19

-------
  Table 29-5. PAH Method Precision: Coefficient of Variation Based on Collocated Samples
                                           by Site
Pollutant
Acenaphthene
Acenaphthylene
Anthracene
Benzo(a)anthracene
Benzo(a)pyrene
Benzo(b)fluoranthene
Benzo(e)pyrene
Benzo(g,h,i)perylene
Benzo(k)fluoranthene
Chrysene
Coronene
Cyclopenta|cd|pyrene
Dibenz(a,h)anthracene
Fluoranthene
Fluorene
9-Fluorenone
Indeno(l,2,3-cd)pyrene
Naphthalene
Perylene
Phenanthrene
Pyrene
Retene
Average by Site
DEMI
7.47
10.29
58.36
49.45
4.36
6.24
6.42
7.69
7.00
7.29
2.51
NA
NA
4.61
5.13
4.84
7.02
8.59
NA
5.77
5.60
20.10
12.04
RUCA
26.38
33.31
6.32
NA
NA
13.93
6.74
10.06
NA
8.38
NA
NA
NA
24.59
24.32
26.96
NA
15.49
NA
28.02
27.07
20.28
19.42
SDGA
10.45
11.41
7.12
0.38
5.22
5.09
9.00
11.21
1.75
15.48
NA
1.09
NA
5.91
7.88
6.12
17.93
6.41
NA
8.33
19.90
16.60
8.80
SEWA
7.16
8.82
12.89
3.70
12.66
6.38
1.12
0.41
5.32
7.25
1.92
2.20
NA
5.33
6.69
6.81
2.23
5.98
NA
5.47
7.44
9.75
5.98
SYFL
11.42
13.02
NA
9.82
13.78
15.29
19.75
15.67
9.63
10.43
1.00
NA
NA
12.67
11.82
13.26
13.14
15.27
NA
11.79
13.39
21.07
12.90
#of
Pairs
29
11
13
8
7
23
16
14
8
27
o
J
2
0
29
29
29
12
29
0
29
29
28
~
Average by
Pollutant
12.58
15.37
21.17
15.84
9.00
9.38
8.61
9.01
5.92
9.77
1.81
1.65
NA
10.62
11.17
11.60
10.08
10.34
NA
11.88
14.68
17.56
11.83
NA=No pairs with concentrations greater than or equal to the MDL
BOLD ITALICS = EPA-designaled NATTS Site
Gray shading indicates NATTS MQO Core Analyte
Orange shading indicates the site-specific average CV for this method; the overall average CV for this method is
calculated from the site-specific averages and provided in the final column of the table.

 29.2.5 Metals Method Precision
        The method precision for all collocated metals samples are presented in Table 29-6 as the
 CV per pollutant per site, the average CV per site, the average CV per pollutant, and the overall
 average CV for NMP sites sampling metals. All samples evaluated in this section are collocated
 samples. The results from collocated samples show low- to mid-level variability among sites,
 ranging from a CV of 0 percent (nickel for UNVT and beryllium for GLKY) to 45.18 percent
 (mercury for UNVT). The pollutant-specific average CV ranged from 5.16 percent (chromium)
                                            29-20

-------
 to 26.44 percent (mercury). The site-specific average CV ranged from 8.70 percent for BOMA to
 20.84 percent for GLKY. The overall average method precision for metals was 13.58 percent.
     Table 29-6. Metals Method Precision: Coefficient of Variation Based on Collocated
                                      Samples by Site
Pollutant
Antimony
Arsenic
Beryllium
Cadmium
Chromium
Cobalt
Lead
Manganese
Mercury
Nickel
Selenium
Average by Site
BOMA
3.26
4.06
18.02
18.07
5.43
12.25
6.43
3.69
15.22
4.89
4.37
8.70
BTUT
6.79
11.01
11.11
15.67
NA
15.02
13.09
10.38
20.20
3.13
18.57
12.50
GLKY
19.08
21.90
0.00
20.56
NA
31.64
14.99
16.15
27.65
43.46
12.99
20.84
S4MO
5.48
11.80
22.00
8.64
2.13
18.74
5.13
5.46
36.97
43.24
10.10
15.43
TOOK
8.86
5.95
7.47
17.90
7.93
14.56
7.85
8.28
13.45
9.07
5.23
9.69
UNVT
9.05
23.20
NA
16.15
NA
NA
4.55
4.16
45.18
0.00
12.34
14.33
#of
Pairs
169
158
74
168
66
156
169
169
115
99
163
~
Average by
Pollutant
8.75
12.99
11.72
16.17
5.16
18.44
8.67
8.02
26.44
17.30
10.60
13.58
NA = No pairs with concentrations greater than or equal to the MDL
BOLD ITALICS = EPA-designated NATTS Site
Gray shading indicates NATTS MQO Core Analyte
Orange shading indicates the site-specific average CV for this method; the overall average CV for this method is
calculated from the site-specific averages and provided in the final column of the table.

 29.2.6  Hexavalent Chromium Method Precision
        Table 29-7 presents the method  precision results from collocated hexavalent chromium
 samples as the CV per site and the overall average CV for NMP sites sampling hexavalent
 chromium. All samples evaluated in this section are collocated samples. Hexavalent chromium is
 a NATTS MQO Core Analyte and the sites shown in Table 29-6 are collocated NATTS sites.
 The site-specific CV ranged from 6.04 percent for SYFL to 31.21 percent for PXSS, with an
 overall average method precision of 14.54 percent.
                                           29-21

-------
Table 29-7. Hexavalent Chromium Method Precision: Coefficient of Variation Based on
                             Collocated Samples by Site
Site
BOMA
BTUT
CAMS 35
CHSC
DEMI
GLKY
GPCO
HOW
MONY
NBIL
PRRI
PXSS
RIVA
ROCH
S4MO
SDGA
SEWA
SKFL
SYFL
UNVT
WADC
# of Pairs
Average by Site
Average CV
27.62
22.35
10.97
8.86
10.93
20.41
10.91
9.10
8.25
18.21
17.14
31.21
11.46
19.56
7.86
16.80
11.03
10.80
6.04
15.45
10.31
110
14.54
                            BOLD ITALICS = EPA-de signaled
                            NATTS Site
                            Orange shading indicates the average CV
                            for this method
                                       29-22

-------
29.3   Analytical Precision
       Analytical precision is a measurement of random errors associated with the process of
analyzing environmental samples. These errors may result from various factors, including
random "noise" inherent to analytical instruments. Laboratories can evaluate the analytical
precision of ambient air samples by comparing concentrations measured during multiple
analyses of a single  sample (i.e., replicate samples). CVs were calculated for every replicate
analysis run on duplicate or collocated samples collected during the program year. However,
only results at or above the MDL were used in these calculations, similar to the calculation of
method precision discussed in Section 29.2.

       Table 29-8 presents the 2011 NMP analytical precision for VOCs, SNMOCs, carbonyl
compounds, PAHs, metals, and hexavalent chromium, presented as average CV (expressed as a
percentage). The analytical precision averaged across all sites collecting duplicate or collocated
samples met the program MQO, which is 15 percent CV. The analytical precision for all six
methods was less than 8 percent.  This table also includes the number of pairs that were included
in the calculation of the analytical precision. The number of pairs including those with
concentrations less than the MDL is also included in Table 29-8 to provide an indication of the
effect that excluding those with concentrations less than the MDL has on the population of pairs
in the dataset.
                  Table 29-8. Analytical Precision by Analytical Method
Method/Pollutant
Group
voc
(TO-15)
SNMOC
Carbonyl Compounds
(TO-11A)
PAH
(TO-13)
Metals Analysis
(Method IO-3. 5)
Hexavalent Chromium
(EPA-approved method)
MQO
Average
Coefficient of
Variation
(%)
6.60
7.39
2.43
4.61
7.91
6.69
Number of
Pairs Included
in the
Calculation
6,174
2,437
3,178
761
3,074
223
Total Number of
Pairs Without
the > MDL
exclusion
7,397
3,059
3,178
941
3,721
225
15. 00 percent CV
                                          29-23

-------
       Tables 29-9 through 29-14 present analytical precision for VOCs, SNMOCs, carbonyl
compounds, PAHs, metals, hexavalent chromium, respectively, as the CV per pollutant per site,
and the average CV per pollutant, per site, and per method. Pollutants exceeding the 15 percent
MQO for CV are bolded in each table. In Tables 29-9 through 29-14, the number of pairs in
comparison to the respective tables listed for duplicate or collocated analyses in Tables 29-2
through 29-7, is approximately twice as high because each sample produces a replicate for each
duplicate (or collocated) sample.  The replicate analyses of both duplicate and collocated samples
indicate that the analytical precision level is within the program MQOs. The shaded rows in each
table identify the NATTS MQO Core Analytes for each method.

29.3.1  VOC Analytical Precision
       Table 29-9 presents analytical precision results from replicate analyses of all duplicate
and collocated VOC samples as the CV per pollutant per site, the average CV per site, the
average CV per pollutant, and the overall average CV for NMP sites sampling VOCs. The
analytical precision results from replicate analyses of all duplicate and collocated samples show
that, for most of the pollutants, the VOC analytical precision was within the program MQO of
15 percent. The CV ranged from  0 percent for several pollutants and several sites to
26.61 percent (methyl isobutyl ketone for TMOK). The pollutant-specific average CV ranged
from 3.01 percent (1,1,1-trichloroethane) to 20.47 percent (ฃrami-l,2-dichloroethylene). Note that
in these cases, the number of pairs included in the calculations are low. The site-specific average
CV ranged from 4.31 percent for GPCO to 9.18 percent for PROK. The overall average
analytical precision was 6.60 percent.
                                          29-24

-------
Table 29-9. VOC Analytical Precision: Coefficient of Variation Based on Replicate
                                   Analyses by Site
Pollutant
Acetylene
Acrylonitrile
tert-Amyl Methyl Ether
Benzene
Bromochloromethane
Bromodichloromethane
Bromoform
Bromomethane
1,3 -Butadiene
Carbon Bisulfide
Carbon Tetrachloride
Chlorobenzene
Chloroethane
Chloroform
Chloromethane
Chloromethylbenzene
Chloroprene
Dibromochloromethane
1 ,2-Dibromoethane
m -Dichlorobenzene
o-Dichlorobenzene
^-Dichlorobenzene
Dichlorodifluoromethane
1 , 1 -Dichloroethane
1 ,2-Dichloroethane
1 , 1 -Dichloroethene
c/s-l,2-Dichloroethylene
trans- 1 ,2-Dichloroethylene
Dichloromethane
1 ,2-Dichloropropane
cis- 1 ,3 -Dichloropropene
trans- 1 ,3 -Dichloropropene
Dichlorotetrafluoroethane
Ethyl Aery late
Ethyl tert-Butyl Ether
Ethylbenzene
Hexachloro- 1 ,3 -butadiene
Methyl Ethyl Ketone
Methyl Isobutyl Ketone
Methyl Methacrylate
BTUT
7.75
NA
NA
5.24
NA
NA
NA
7.47
9.42
5.85
5.84
NA
6.07
12.57
5.49
NA
NA
NA
NA
NA
NA
NA
5.35
NA
10.09
0.00
NA
20.47
6.59
NA
NA
NA
11.21
NA
NA
6.00
NA
NA
6.75
NA
BURVT
3.60
5.41
NA
5.04
NA
NA
NA
10.34
5.44
NA
4.78
NA
14.36
11.04
3.48
NA
NA
NA
NA
NA
NA
NA
3.48
NA
9.97
NA
NA
NA
4.56
NA
NA
NA
8.12
NA
NA
7.07
NA
NA
7.36
NA
CHNJ
6.21
4.62
NA
6.01
NA
NA
NA
15.71
18.58
6.62
6.43
NA
10.45
12.04
6.39
NA
NA
NA
NA
NA
NA
NA
6.19
NA
8.45
NA
NA
NA
7.18
NA
NA
NA
12.42
NA
NA
8.17
NA
NA
11.54
8.95
DEMI
4.03
NA
NA
6.45
NA
NA
NA
10.83
7.56
4.03
6.31
NA
3.02
6.04
5.81
NA
NA
NA
NA
NA
NA
NA
5.79
NA
5.56
NA
NA
NA
4.78
NA
NA
NA
12.67
NA
NA
7.53
NA
NA
10.41
NA
ELNJ
5.26
8.88
NA
5.27
NA
NA
NA
11.50
6.20
4.58
6.70
NA
0.00
5.43
4.78
NA
NA
NA
NA
NA
NA
2.70
4.73
NA
1.89
NA
NA
NA
5.13
NA
NA
NA
6.96
NA
NA
5.72
NA
NA
16.20
4.47
GLKY
5.35
4.42
NA
6.69
NA
NA
NA
12.27
6.44
6.04
6.56
9.43
NA
6.81
4.47
NA
NA
NA
NA
NA
NA
NA
4.85
13.69
6.18
NA
NA
NA
4.70
NA
NA
NA
8.94
NA
NA
6.74
NA
NA
9.94
NA
 NA = No pairs with concentrations greater than or equal to the MDL
 BOLD ITALICS = EPA-designated NATTS Site
 Gray shading indicates NATTS MQO Core Analyte
 Orange shading indicates the site-specific average CV for this method; the overall average CV for this
 method is calculated from the site-specific averages and provided in the final column of the table.
                                         29-25

-------
Table 29-9. VOC Analytical Precision: Coefficient of Variation Based on Replicate
                            Analyses by Site (Continued)
Pollutant
Methyl fer/-Butyl Ether
w-Octane
Propylene
Styrene
1 , 1 ,2,2-Tetrachloroethane
Tetrachloroethylene
Toluene
1 ,2,4-Trichlorobenzene
1,1,1 -Trichloroethane
1 , 1 ,2-Trichloroethane
Trichloroethylene
Trichlorofluoromethane
Trichlorotrifluoroethane
1 ,2,4-Trimethylbenzene
1,3,5 -Trimethy Ibenzene
Vinyl chloride
m,p-Xylene
o-Xylene
Average by Site
BTUT
NA
5.42
5.60
8.43
NA
4.70
4.71
NA
NA
NA
NA
5.94
8.30
6.47
8.20
NA
5.06
6.16
7.18
BURVT
NA
11.93
4.30
13.45
NA
7.19
4.27
NA
NA
NA
NA
3.66
6.33
8.55
6.14
NA
6.13
7.53
7.06
CHNJ
NA
7.34
5.25
7.49
NA
16.63
6.30
NA
NA
NA
NA
6.29
7.00
8.45
8.93
NA
8.68
16.21
9.09
DEMI
NA
16.65
5.55
9.26
NA
5.16
4.98
NA
NA
NA
NA
5.39
6.49
6.54
5.27
NA
6.60
9.83
7.02
ELNJ
5.61
7.36
4.96
12.02
NA
7.91
4.38
NA
NA
NA
NA
4.19
5.37
8.01
11.14
NA
5.46
8.99
6.39
GLKY
9.12
6.98
5.82
6.66
NA
NA
5.16
NA
NA
NA
NA
4.60
5.45
7.78
13.26
20.20
7.21
7.46
7.70
 NA = No pairs with concentrations greater than or equal to the MDL
 BOLD ITALICS = EPA-designated NATTS Site
 Gray shading indicates NATTS MQO Core Analyte
 Orange shading indicates the site-specific average CV for this method; the overall average CV for this
 method is calculated from the site-specific averages and provided in the final column of the table.
                                         29-26

-------
Table 29-9. VOC Analytical Precision: Coefficient of Variation Based on Replicate
                            Analyses by Site (Continued)
Pollutant
Acetylene
Acrylonitrile
tert-Amyl Methyl Ether
Benzene
Bromochloromethane
Bromodichloromethane
Bromoform
Bromomethane
1,3 -Butadiene
Carbon Bisulfide
Carbon Tetrachloride
Chlorobenzene
Chloroethane
Chloroform
Chloromethane
Chloromethylbenzene
Chloroprene
Dibromochloromethane
1 ,2-Dibromoethane
/w-Dichlorobenzene
o-Dichlorobenzene
ฃ>-Dichlorobenzene
Dichlorodifluoromethane
1 , 1 -Dichloroethane
1 ,2-Dichloroethane
1 , 1 -Dichloroethene
cis- 1 ,2-Dichloroethylene
/raซ5-l,2-Dichloroethylene
Dichloromethane
1 ,2-Dichloropropane
cis- 1 ,3 -Dichloropropene
trans- 1 , 3 -Dichloropropene
Dichlorotetrafluoroethane
Ethyl Aery late
Ethyl tert-Butyl Ether
Ethylbenzene
Hexachloro- 1 ,3 -butadiene
Methyl Ethyl Ketone
Methyl Isobutyl Ketone
Methyl Methacrylate
GPCO
3.50
4.22
NA
4.55
NA
NA
NA
6.93
3.22
3.54
4.86
NA
3.62
4.88
3.58
NA
NA
NA
NA
NA
NA
NA
3.31
NA
1.96
NA
NA
NA
6.72
NA
NA
NA
3.24
NA
NA
3.27
NA
NA
5.31
9.03
MWOK
5.05
9.75
NA
4.71
NA
NA
NA
6.30
6.56
1.95
5.29
NA
5.24
7.36
2.50
NA
NA
NA
NA
NA
NA
3.82
2.50
NA
6.46
NA
NA
NA
4.13
NA
NA
NA
5.41
NA
NA
4.39
NA
NA
3.82
NA
NBIL
5.67
4.78
NA
5.78
NA
6.52
7.71
10.23
4.60
3.92
9.48
NA
8.77
4.02
3.89
NA
NA
3.98
NA
NA
NA
NA
3.94
NA
9.66
NA
NA
NA
6.39
NA
NA
NA
4.65
NA
NA
3.75
NA
NA
10.20
NA
NBNJ
5.23
3.71
NA
4.62
NA
NA
NA
7.47
3.89
4.58
5.01
NA
1.37
7.06
4.30
NA
NA
NA
NA
NA
NA
NA
4.28
NA
5.74
6.15
NA
NA
3.71
NA
NA
NA
6.51
NA
7.44
4.70
NA
NA
6.23
NA
OCOK
3.30
NA
NA
5.58
NA
NA
NA
11.91
8.45
NA
7.01
NA
NA
3.23
2.58
NA
NA
NA
NA
NA
NA
NA
2.56
NA
NA
NA
NA
NA
6.31
NA
NA
NA
7.82
NA
NA
7.13
NA
NA
8.32
NA
PANJ
4.45
NA
NA
5.93
NA
NA
NA
3.23
5.28
NA
6.55
NA
5.30
3.54
4.94
NA
NA
NA
NA
NA
NA
5.77
4.91
NA
5.92
NA
NA
NA
4.72
NA
NA
NA
6.98
NA
NA
4.67
NA
NA
3.69
NA
 NA = No pairs with concentrations greater than or equal to the MDL
 BOLD ITALICS = EPA-designated NATTS Site
 Gray shading indicates NATTS MQO Core Analyte
 Orange shading indicates the site-specific average CV for this method; the overall average CV for this
 method is calculated from the site-specific averages and provided in the final column of the table.
                                         29-27

-------
Table 29-9. VOC Analytical Precision: Coefficient of Variation Based on Replicate
                            Analyses by Site (Continued)
Pollutant
Methyl fer/-Butyl Ether
w-Octane
Propylene
Styrene
1, 1,2,2-Tetrachloroethane
Tetrachloroethylene
Toluene
1 ,2,4-Trichlorobenzene
1,1,1 -Trichloroethane
1 , 1 ,2-Trichloroethane
Trichloroethylene
Trichlorofluoromethane
Trichlorotrifluoroethane
1 ,2,4-Trimethylbenzene
1 ,3 ,5-Trimethylbenzene
Vinyl chloride
m,p-Xylene
o-Xylene
Average by Site
GPCO
NA
5.08
4.41
3.42
NA
4.66
3.30
NA
NA
NA
4.98
3.70
5.33
3.55
4.19
NA
3.62
3.11
4.31
MWOK
NA
4.29
2.48
7.82
NA
5.64
3.32
NA
NA
NA
NA
2.88
4.40
7.72
6.25
NA
4.29
7.58
5.07
NBIL
NA
6.63
3.98
6.94
NA
6.11
3.92
NA
NA
NA
NA
3.51
3.55
4.43
6.43
NA
4.13
4.34
5.73
NBNJ
3.92
6.98
4.16
7.33
NA
6.71
4.47
NA
NA
NA
NA
4.37
4.18
6.45
8.77
11.79
5.02
6.16
5.56
OCOK
NA
8.75
3.25
13.70
NA
8.78
4.51
NA
NA
NA
NA
3.13
7.29
10.05
9.84
NA
7.51
13.46
7.15
PANJ
NA
5.07
4.82
3.61
NA
6.47
5.92
NA
3.01
NA
NA
5.41
4.75
5.61
4.58
8.32
5.16
5.30
5.14
 NA = No pairs with concentrations greater than or equal to the MDL
 BOLD ITALICS = EPA-designated NATTS Site
 Gray shading indicates NATTS MQO Core Analyte
 Orange shading indicates the site-specific average CV for this method; the overall average CV for this
 method is calculated from the site-specific averages and provided in the final column of the table.
                                         29-28

-------
Table 29-9. VOC Analytical Precision: Coefficient of Variation Based on Replicate
                            Analyses by Site (Continued)
Pollutant
Acetylene
Acrylonitrile
tert-Amyl Methyl Ether
Benzene
Bromochloromethane
Bromodichloromethane
Bromoform
Bromomethane
1,3 -Butadiene
Carbon Bisulfide
Carbon Tetrachloride
Chlorobenzene
Chloroethane
Chloroform
Chloromethane
Chloromethylbenzene
Chloroprene
Dibromochloromethane
1,2-Dibromoethane
/w-Dichlorobenzene
o-Dichlorobenzene
ฃ>-Dichlorobenzene
Dichlorodifluoromethane
1 , 1 -Dichloroethane
1 ,2-Dichloroethane
1 , 1 -Dichloroethene
cis- 1 ,2-Dichloroethylene
trans- 1 ,2-Dichloroethylene
Dichloromethane
1 ,2-Dichloropropane
cis- 1, 3-D ichloropropene
trans- 1 ,3 -Dichloropropene
Dichlorotetrafluoroethane
Ethyl Aery late
Ethyl tert-Butyl Ether
Ethylbenzene
Hexachloro- 1 , 3 -butadiene
Methyl Ethyl Ketone
Methyl Isobutyl Ketone
Methyl Methacrylate
PROK
8.28
4.91
NA
7.12
NA
NA
NA
8.25
7.03
NA
7.32
NA
9.41
NA
6.35
NA
NA
NA
NA
NA
NA
9.09
6.28
NA
10.23
NA
NA
NA
6.97
NA
NA
NA
9.05
NA
NA
11.11
NA
NA
7.70
NA
PXSS
4.96
2.76
NA
6.60
NA
NA
NA
8.87
5.24
6.52
6.66
NA
7.44
4.86
4.63
NA
NA
NA
NA
NA
NA
6.24
4.58
NA
5.25
NA
NA
NA
6.98
NA
NA
NA
7.44
NA
NA
4.67
NA
NA
9.51
NA
S4MO
4.71
NA
NA
5.10
NA
NA
NA
16.11
7.20
4.14
4.87
NA
3.70
8.05
4.72
NA
NA
NA
NA
NA
NA
4.24
4.59
NA
9.95
NA
NA
NA
4.37
NA
NA
NA
9.52
NA
NA
5.44
NA
NA
6.43
NA
SEWA
3.16
7.97
NA
3.10
NA
NA
NA
12.78
7.36
NA
4.56
NA
0.00
8.37
3.36
NA
NA
NA
NA
NA
NA
NA
3.16
NA
2.86
NA
NA
NA
3.15
NA
NA
NA
5.29
NA
NA
4.09
NA
NA
9.22
NA
SPIL
4.92
5.79
NA
7.08
NA
NA
NA
7.88
5.53
4.68
6.55
NA
14.17
6.73
5.16
NA
NA
NA
NA
NA
NA
NA
5.28
NA
12.20
NA
NA
NA
5.04
NA
NA
NA
8.21
NA
NA
8.88
NA
NA
6.39
NA
SSSD
4.88
13.24
NA
4.24
NA
NA
NA
7.20
8.44
1.79
3.86
NA
9.75
8.03
4.05
NA
NA
NA
NA
NA
NA
NA
3.83
NA
8.52
NA
NA
NA
5.06
NA
NA
NA
7.43
NA
NA
5.00
NA
NA
7.74
NA
 NA = No pairs with concentrations greater than or equal to the MDL
 BOLD ITALICS = EPA-designated NATTS Site
 Gray shading indicates NATTS MQO Core Analyte
 Orange shading indicates the site-specific average CV for this method; the overall average CV for this
 method is calculated from the site-specific averages and provided in the final column of the table.
                                         29-29

-------
Table 29-9. VOC Analytical Precision: Coefficient of Variation Based on Replicate
                            Analyses by Site (Continued)
Pollutant
Methyl ferMSutyl Ether
w-Octane
Propylene
Styrene
1 , 1 ,2,2-Tetrachloroethane
Tetrachloroethylene
Toluene
1 ,2,4-Trichlorobenzene
1,1,1 -Trichloroethane
1, 1,2-Trichloroethane
Trichloroethylene
Trichlorofluoromethane
Trichlorotrifluoroethane
1 ,2,4-Trimethylbenzene
1,3,5 -Trimethy Ibenzene
Vinyl chloride
m,p-Xylene
o-Xylene
Average by Site
PROK
NA
17.76
7.19
15.25
NA
NA
6.72
NA
NA
NA
NA
6.26
7.05
15.67
NA
NA
10.68
14.60
9.18
PXSS
NA
6.26
4.51
9.23
NA
4.84
4.29
NA
NA
NA
NA
4.08
6.55
5.45
6.17
NA
4.70
5.35
5.88
S4MO
NA
6.00
5.14
15.26
NA
8.96
4.11
NA
NA
NA
NA
3.89
7.03
11.27
19.15
6.79
5.00
7.03
7.24
SEWA
NA
11.33
5.69
8.36
NA
8.75
1.98
NA
NA
NA
0.00
2.51
4.90
6.59
7.44
NA
3.45
9.82
5.53
SPIL
NA
10.49
5.86
10.66
NA
6.86
5.94
NA
NA
NA
6.52
4.58
5.23
11.48
10.28
NA
8.08
10.72
7.54
SSSD
7.58
6.46
4.37
5.93
NA
8.52
3.77
NA
NA
NA
NA
3.14
5.78
5.76
5.33
NA
5.26
5.30
6.08
 NA = No pairs with concentrations greater than or equal to the MDL
 BOLD ITALICS = EPA-designated NATTS Site
 Gray shading indicates NATTS MQO Core Analyte
 Orange shading indicates the site-specific average CV for this method; the overall average CV for this
 method is calculated from the site-specific averages and provided in the final column of the table.
                                         29-30

-------
Table 29-9. VOC Analytical Precision: Coefficient of Variation Based on Replicate
                            Analyses by Site (Continued)
Pollutant
Acetylene
Acrylonitrile
tert-Amyl Methyl Ether
Benzene
Bromochloromethane
Bromodichloromethane
Bromoform
Bromomethane
1,3 -Butadiene
Carbon Bisulfide
Carbon Tetrachloride
Chlorobenzene
Chloroethane
Chloroform
Chloromethane
Chloromethylbenzene
Chloroprene
Dibromochloromethane
1 ,2-Dibromoethane
/w-Dichlorobenzene
o-Dichlorobenzene
ฃ>-Dichlorobenzene
Dichlorodifluoromethane
1 , 1 -Dichloroethane
1 ,2-Dichloroethane
1 , 1 -Dichloroethene
cis- 1 ,2-Dichloroethylene
trans- 1 ,2-Dichloroethylene
Dichloromethane
1 ,2-Dichloropropane
cis- 1 ,3 -Dichloropropene
trans- 1 ,3 -Dichloropropene
Dichlorotetrafluoroethane
Ethyl Acrylate
Ethyl tert-Butyl Ether
Ethylbenzene
Hexachloro- 1 ,3 -butadiene
Methyl Ethyl Ketone
Methyl Isobutyl Ketone
Methyl Methacrylate
TMOK
2.97
14.40
NA
5.05
NA
NA
NA
8.19
5.37
2.90
5.55
NA
5.06
2.33
3.32
NA
NA
NA
NA
NA
NA
6.12
3.17
NA
1.82
NA
NA
NA
5.18
NA
NA
NA
5.92
NA
NA
4.24
NA
NA
26.61
NA
TOOK
5.88
NA
NA
7.32
NA
NA
NA
8.92
6.59
0.44
6.82
NA
7.19
2.27
5.72
NA
NA
NA
NA
NA
NA
4.23
5.93
NA
3.62
NA
NA
NA
5.18
NA
NA
NA
6.54
NA
NA
6.68
NA
NA
7.80
NA
UCSD
6.80
4.34
NA
6.81
NA
NA
NA
5.79
9.60
3.14
6.49
NA
6.30
12.24
5.78
NA
NA
NA
NA
NA
NA
NA
5.90
NA
8.58
NA
NA
NA
5.17
NA
NA
NA
6.45
NA
NA
5.08
NA
NA
6.99
NA
#of
Pairs
283
45
0
283
0
12
2
191
249
129
282
1
50
144
283
0
0
11
0
0
0
29
283
1
69
o
6
0
2
277
0
0
0
280
0
1
281
0
0
216
4
Average
by
Pollutant
5.05
6.61
NA
5.63
NA
6.52
7.71
9.44
7.05
4.05
6.07
9.43
6.38
6.84
4.54
NA
NA
3.98
NA
NA
NA
5.28
4.51
13.69
6.75
3.07
NA
20.47
5.33
NA
NA
NA
7.66
NA
7.44
5.92
NA
NA
8.96
7.48
     NA = No pairs with concentrations greater than or equal to the MDL
     BOLD ITALICS = EPA-designated NATTS Site
     Gray shading indicates NATTS MQO Core Analyte
     Orange shading indicates the site-specific average CV for this method; the overall average CV for
     this method is calculated from the site-specific averages and provided in the final column of the table.
                                         29-31

-------
     Table 29-9. VOC Analytical Precision: Coefficient of Variation Based on Replicate
                               Analyses by Site (Continued)
Pollutant
Methyl tert-Butyl Ether
w-Octane
Propylene
Styrene
1 , 1 ,2,2-Tetrachloroethane
Tetrachloroethylene
Toluene
1 ,2,4-Trichlorobenzene
1,1,1 -Trichloroethane
1 , 1 ,2-Trichloroethane
Trichloroethylene
Trichlorofluoromethane
Trichlorotrifluoroethane
1 ,2,4-Trimethylbenzene
1 ,3 ,5-Trimethylbenzene
Vinyl chloride
m,p-Xylene
o-Xylene
Average by Site
TMOK
NA
6.16
3.70
12.46
NA
9.05
4.24
NA
NA
NA
NA
3.00
7.23
5.15
7.96
NA
4.07
14.70
6.64
TOOK
NA
6.91
5.41
9.43
NA
9.64
6.33
NA
NA
NA
NA
6.76
5.40
5.60
6.42
NA
6.69
6.83
6.17
UCSD
NA
10.60
5.37
5.25
NA
9.01
4.62
NA
NA
NA
5.44
5.20
4.83
8.74
NA
16.85
5.59
7.50
6.95
#of
Pairs
12
278
283
241
0
129
283
0
1
0
10
282
283
262
141
8
277
273
-
Average
by
Pollutant
6.56
8.31
4.85
9.14
NA
7.66
4.63
NA
3.01
NA
4.24
4.40
5.83
7.59
8.20
12.79
5.83
8.48
6.60
          NA = No pairs with concentrations greater than or equal to the MDL
          BOLD ITALICS = EPA-designaled NATTS Site
          Gray shading indicates NATTS MQO Core Analyte
          Orange shading indicates the site-specific average CV for this method; the overall average CV for
          this method is calculated from the site-specific averages and provided in the final column of the table.

29.3.2 SNMOC Analytical Precision
       Table 29-10 presents analytical precision results from replicate analyses of all duplicate
and collocated SNMOC samples as the CV per pollutant per site, the average CV per site, the
average CV per pollutant, and the overall average CV for NMP sites sampling SNMOCs. The
CV ranged from 0 percent (2-methyl-2-butene for BMCO) to 43.71 percent (1-dodecene for
NBIL). The pollutant-specific average CV ranged from 0.37 percent (1-hexene) to 18.93 percent
(1-dodecene). The site-specific average CV ranged from 4.47 percent for PACO to 16.14 percent
for BMCO. The overall program average CV was 7.39 percent.
                                           29-32

-------
Table 29-10. SNMOC Analytical Precision: Coefficient of Variation Based on Replicate
                                     Analyses by Site
Pollutant
Acetylene
Benzene
1,3 -Butadiene
w-Butane
c/s-2-Butene
trans-2-Butene
Cyclohexane
Cyclopentane
Cyclopentene
w-Decane
1-Decene
w-Diethylbenzene
p-Diethylbenzene
2,2-Dimethylbutane
2,3 -Dimethylbutane
2,3 -Dimethylpentane
2,4-Dimethylpentane
w-Dodecane
1-Dodecene
Ethane
2-Ethyl-l-butene
Ethylbenzene
Ethylene
OT-Ethyltoluene
o-Ethyltoluene
p-Ethyltoluene
w-Heptane
1-Heptene
w-Hexane
1-Hexene
c/s-2-Hexene
trans-2-Hexene
Isobutane
Isobutene/ 1 -Butene
Isopentane
Isoprene
Isopropylbenzene
2-Methyl- 1 -butene
3 -Methyl- 1 -butene
BMCO
15.60
19.92
NA
NA
1.14
6.91
19.11
11.11
2.99
18.02
NA
5.85
20.70
16.09
NA
NA
15.39
14.08
36.60
19.42
NA
17.66
10.91
4.25
NA
8.91
18.74
23.95
20.94
NA
NA
NA
20.27
NA
19.00
NA
NA
20.20
NA
BTUT
23.90
3.83
8.51
1.10
7.38
4.51
5.12
5.20
12.75
4.52
NA
9.43
15.21
7.80
8.96
5.99
8.63
3.23
5.82
25.25
NA
7.32
28.33
3.63
8.83
8.57
2.59
NA
2.99
0.56
NA
NA
1.37
14.23
1.90
2.61
NA
8.43
NA
NBIL
2.64
3.97
NA
1.53
NA
5.97
12.13
11.78
2.46
6.37
2.40
6.63
16.02
7.09
2.82
6.57
4.85
9.99
43.71
0.46
NA
5.22
1.98
6.33
15.29
10.04
6.14
NA
3.42
NA
NA
NA
1.89
4.74
1.78
1.88
2.22
NA
NA
PACO
3.66
2.16
NA
1.13
3.44
3.86
3.90
6.87
10.74
2.50
NA
4.76
7.87
7.83
1.56
2.84
3.84
6.74
0.62
0.86
NA
4.61
1.50
2.19
6.87
5.63
1.78
1.75
2.80
0.19
NA
NA
1.09
6.61
5.68
16.23
4.84
NA
NA
RICO
2.07
2.44
4.22
0.35
12.46
2.55
1.37
3.29
11.70
5.77
NA
9.48
4.52
8.00
6.24
7.78
9.49
8.71
7.18
0.45
NA
7.79
1.24
7.03
11.81
7.41
2.63
3.13
0.77
NA
NA
NA
1.06
NA
1.63
5.50
NA
3.43
NA
          NA = No pairs with concentrations greater than or equal to the MDL
          BOLD ITALICS = EPA-designated NATTS Site
          Orange shading indicates the site-specific average CV for this method; the overall average CV for
          this method is calculated from the site-specific averages and provided in the final column of the table.
                                           29-33

-------
Table 29-10. SNMOC Analytical Precision: Coefficient of Variation Based on Replicate
                              Analyses by Site (Continued)
Pollutant
2-Methyl- 1 -pentene
4-Methyl- 1 -pentene
2-Methyl-2-butene
Methylcyclohexane
Methylcyclopentane
2-Methylheptane
3-Methylheptane
2-Methylhexane
3 -Methy Ihexane
2-Methylpentane
3-Methylpentane
w-Nonane
1-Nonene
w-Octane
1-Octene
w-Pentane
1 -Pentene
c/s-2-Pentene
trans-2-Pentene
a-Pinene
&-Pinene
Propane
w-Propylbenzene
Propylene
Propyne
Styrene
Toluene
w-Tridecane
1-Tridecene
1 ,2,3 -Trimethylbenzene
1 ,2,4-Trimethylbenzene
1 ,3 ,5-Trimethylbenzene
2,2,3 -Trimethylpentane
2,2,4-Trimethylpentane
2,3 ,4-Trimethylpentane
w-Undecane
1-Undecene
w-Xylene/^-Xylene
o-Xylene
Average by Site
BMCO
NA
NA
0.00
21.56
21.32
19.29
25.42
22.60
23.73
18.00
19.62
21.86
NA
24.58
NA
20.42
23.91
NA
2.56
NA
NA
19.12
NA
15.48
NA
NA
19.07
NA
NA
8.06
14.19
0.43
NA
NA
18.58
16.17
NA
NA
11.14
16.14
BTUT
NA
NA
4.56
4.73
4.08
11.13
6.69
3.54
7.75
4.42
4.68
6.37
NA
5.45
2.75
3.06
5.25
NA
6.90
13.09
2.36
0.30
5.50
6.48
NA
10.46
1.91
7.54
NA
12.73
5.51
6.43
6.02
10.51
5.16
9.98
2.93
4.47
7.09
7.05
NBIL
NA
NA
9.79
13.26
3.35
17.65
7.49
14.42
9.87
3.88
6.74
5.84
13.51
6.69
NA
2.49
7.64
NA
11.96
3.39
NA
0.36
20.84
4.19
8.74
8.20
3.26
NA
NA
16.22
8.35
7.87
4.72
13.08
10.66
6.72
1.32
3.71
4.92
7.57
PACO
NA
NA
7.82
1.79
1.70
1.86
3.84
1.84
7.34
5.51
1.53
1.76
5.72
5.15
4.61
2.18
15.59
NA
21.16
NA
NA
0.78
5.38
2.15
NA
NA
1.64
NA
NA
7.09
1.97
7.66
4.94
1.20
6.24
3.44
2.60
2.57
3.07
4.47
RICO
NA
NA
4.62
0.99
0.56
4.64
4.92
5.47
2.39
0.38
1.60
7.08
NA
3.74
NA
0.70
6.40
5.92
4.63
4.86
NA
0.25
11.34
1.24
NA
NA
4.14
NA
NA
10.18
4.87
5.39
NA
15.81
5.16
11.98
NA
1.64
6.11
5.06
          NA = No pairs with concentrations greater than or equal to the MDL
          BOLD ITALICS = EPA-designated NATTS Site
          Orange shading indicates the site-specific average CV for this method; the overall average CV for
          this method is calculated from the site-specific averages and provided in the final column of the table.
                                           29-34

-------
Table 29-10. SNMOC Analytical Precision: Coefficient of Variation Based on Replicate
                              Analyses by Site (Continued)
Pollutant
Acetylene
Benzene
1,3 -Butadiene
w-Butane
c/s-2-Butene
/raws-2-Butene
Cyclohexane
Cyclopentane
Cyclopentene
w-Decane
1-Decene
/w-Diethylbenzene
ฃ>-Diethylbenzene
2,2-Dimethylbutane
2,3 -Dimethylbutane
2,3 -Dimethylpentane
2,4-Dimethylpentane
w-Dodecane
1-Dodecene
Ethane
2-Ethyl-l-butene
Ethylbenzene
Ethylene
/w-Ethyltoluene
o-Ethyltoluene
ฃ>-Ethyltoluene
w-Heptane
1-Heptene
w-Hexane
1-Hexene
c/s-2-Hexene
/raซs-2-Hexene
Isobutane
Isobutene/ 1 -Butene
Isopentane
Isoprene
Isopropylbenzene
2-Methyl-l-butene
3-Methyl-l-butene
SSSD
3.86
3.30
NA
2.55
3.90
8.64
3.71
7.13
NA
2.78
NA
6.21
10.71
11.47
7.92
1.79
6.67
4.32
21.25
1.24
NA
10.83
1.24
5.43
1.35
4.70
4.35
NA
5.63
NA
4.76
NA
1.98
5.06
5.76
3.72
NA
7.81
NA
UCSD
5.21
4.23
NA
2.52
NA
6.56
3.23
5.55
NA
NA
NA
5.22
16.08
8.22
4.50
NA
NA
1.77
17.31
0.85
NA
7.70
2.39
NA
NA
NA
3.35
NA
4.83
NA
NA
NA
2.67
1.56
2.46
6.70
NA
3.16
NA
#of
Pairs
61
60
7
40
19
31
45
49
9
38
2
52
45
52
32
32
34
38
18
61
0
53
61
38
29
22
44
8
60
2
1
0
61
26
58
31
2
22
0
Average by
Pollutant
8.13
5.69
6.36
1.53
5.66
5.57
6.94
7.28
8.13
6.66
2.40
6.80
13.01
9.50
5.33
4.99
8.14
6.98
18.93
6.93
NA
8.73
6.80
4.81
8.83
7.54
5.65
9.61
5.91
0.37
4.76
NA
4.33
6.44
5.46
6.11
3.53
8.61
NA
           NA = No pairs with concentrations greater than or equal to the MDL
           BOLD ITALICS = EPA-designaled NATTS Site
           Orange shading indicates the site-specific average CV for this method; the overall average CV for
           this method is calculated from the site-specific averages and provided in the final column of the table.
                                           29-35

-------
Table 29-10. SNMOC Analytical Precision: Coefficient of Variation Based on Replicate
                              Analyses by Site (Continued)
Pollutant
2-Methyl- 1 -pentene
4-Methyl- 1 -pentene
2-Methyl-2-butene
Methylcyclohexane
Methylcyclopentane
2-Methylheptane
3-Methylheptane
2-Methylhexane
3-Methylhexane
2-Methylpentane
3-Methylpentane
w-Nonane
1-Nonene
w-Octane
1-Octene
w-Pentane
1 -Pentene
c/s-2-Pentene
/raws-2-Pentene
a-Pinene
6-Pinene
Propane
w-Propylbenzene
Propylene
Propyne
Styrene
Toluene
w-Tridecane
1-Tridecene
1 ,2,3 -Trimethylbenzene
1 ,2,4-Trimethylbenzene
1 ,3 ,5-Trimethylbenzene
2,2,3 -Trimethylpentane
2,2,4-Trimethylpentane
2,3 ,4-Trimethylpentane
w-Undecane
1-Undecene
7w-Xylene//?-Xylene
o-Xylene
Average by Site
SSSD
NA
NA
3.73
5.95
4.26
7.00
5.76
3.40
8.78
5.64
4.20
2.75
NA
3.15
NA
2.67
13.10
NA
7.35
4.82
NA
0.87
1.01
2.89
NA
2.35
5.11
NA
NA
5.71
5.62
5.56
NA
5.34
7.42
6.89
NA
2.48
5.26
5.33
UCSD
NA
AN
2.49
1.55
4.71
NA
NA
4.25
16.23
4.59
5.77
NA
NA
2.38
NA
6.47
18.65
NA
NA
NA
NA
0.82
NA
3.53
NA
0.36
2.61
NA
NA
22.52
7.86
NA
NA
4.62
NA
6.70
6.73
10.57
11.62
6.07
#of
Pairs
0
0
27
45
57
26
27
54
54
61
58
37
7
40
o
6
61
46
3
34
12
o
J
61
15
61
2
6
59
4
0
32
55
25
9
49
36
38
8
30
49
~
Average by
Pollutant
NA
NA
4.72
7.12
5.71
10.26
9.02
7.93
10.87
6.06
6.31
7.61
9.62
7.31
3.68
5.43
12.94
5.92
9.09
6.54
2.36
3.22
8.81
5.14
8.74
5.34
5.39
7.54
NA
11.79
6.91
5.56
5.23
8.43
8.87
8.84
3.40
4.24
7.03
7.39
           NA = No pairs with concentrations greater than or equal to the MDL
           BOLD ITALICS = EPA-designated NATTS Site
           Orange shading indicates the site-specific average CV for this method; the overall average CV for
           this method is calculated from the site-specific averages and provided in the final column of the table.
                                           29-36

-------
29.3.3 Carbonyl Compound Analytical Precision
       Table 29-11 presents the analytical precision results from replicate analyses of duplicate
and collocated carbonyl compound samples as the CV per pollutant per site, the average CV per
site, the average CV per pollutant, and the overall average CV for NMP sites sampling carbonyl
compounds. The overall average variability was 2.43 percent, which is well within the program
MQO of 15 percent CV. The analytical precision results from replicate analyses of duplicate and
collocated samples range from 0.25 percent (acetaldehyde for CHNJ) to 7.48 percent
(benzaldehyde for GLKY). The pollutant-specific average CV ranged from 0.73 percent
(acetone) to 4.08 percent (tolualdehydes). The site-specific average CV ranged from 1.91 percent
for PXSS to 2.97 percent for INDEM.

 Table 29-11. Carbonyl Compound Analytical Precision: Coefficient of Variation Based on
                                Replicate Analyses by Site
Pollutant
Acetaldehyde
Acetone
Benzaldehyde
Butyraldehyde
Crotonaldehyde
2,5-Dimethylbenzaldehyde
Formaldehyde
Hexaldehyde
Isovaleraldehyde
Propionaldehyde
rolualdehydes
Valeraldehyde
Average by Site
AZFL
1.03
1.53
4.35
3.40
2.69
NA
1.61
4.08
NA
2.21
4.79
3.96
2.96
BTUT
0.44
0.38
3.77
1.53
2.11
NA
0.84
2.98
NA
1.39
4.17
2.83
2.04
CHNJ
0.25
0.35
2.72
1.95
1.97
NA
0.50
3.45
NA
2.33
4.55
3.39
2.15
DEMI
0.29
0.49
3.42
2.83
3.07
NA
1.30
4.21
NA
1.81
4.33
4.45
2.62
ELNJ
0.76
0.67
2.40
0.92
2.42
NA
1.22
3.98
NA
1.42
4.89
3.38
2.21
GLKY
0.93
0.60
7.48
2.69
2.87
NA
0.71
3.33
NA
1.90
3.60
4.92
2.90
GPCO
0.59
0.67
1.35
2.72
2.53
NA
0.92
3.36
NA
1.40
3.80
3.08
2.04
 NA = No pairs with concentrations greater than or equal to the MDL
 BOLD ITALICS = EPA-designated NATTS Site
 Gray shading indicates NATTS MQO Core Analyte
 Orange shading indicates the site-specific average CV for this method; the overall average CV for this method is
 calculated from the site-specific averages and provided in the final column of the table.
                                          29-37

-------
Table 29-11. Carbonyl Compound Analytical Precision: Coefficient of Variation Based on
                          Replicate Analyses by Site (Continued)
Pollutant
Acetaldehyde
Acetone
Benzaldehyde
Butyraldehyde
Crotonaldehyde
2,5-Dimethylbenzaldehyde
Formaldehyde
Hexaldehyde
Isovaleraldehyde
Propionaldehyde
Tolualdehydes
Valeraldehyde
Average by Site
INDEM
1.47
0.69
4.60
2.46
3.05
NA
1.29
4.31
NA
2.59
4.73
4.52
2.97
MWOK
0.59
0.38
3.30
2.24
0.99
NA
0.49
3.22
NA
2.10
4.09
4.13
2.15
NBIL
0.95
1.19
3.82
2.59
4.86
NA
0.35
3.48
NA
2.60
2.94
4.34
2.71
NBNJ
0.84
0.42
3.62
2.55
2.61
NA
0.48
4.09
NA
1.99
2.91
1.60
2.11
OCOK
0.58
0.38
2.79
1.82
2.48
NA
0.61
3.04
NA
1.08
4.99
2.76
2.05
ORFL
0.52
1.02
3.62
3.11
1.56
NA
0.79
3.96
NA
2.23
5.06
4.51
2.64
PROK
0.41
0.55
3.44
1.76
1.74
NA
1.29
2.98
NA
2.08
3.25
2.72
2.02
 NA = No pairs with concentrations greater than or equal to the MDL
BOLD ITALICS = EPA-designated NATTS Site
Gray shading indicates NATTS MQO Core Analyte
Orange shading indicates the site-specific average CV for this method; the overall average CV for this method is
calculated from the site-specific averages and provided in the final column of the table.


Table 29-11. Carbonyl Compound Analytical Precision: Coefficient of Variation Based on
                           Replicate Analyses by Site (Continued)
Pollutant
Acetaldehyde
Acetone
Benzaldehyde
Butyraldehyde
Crotonaldehyde
2,5-Dimethylbenzaldehyde
Formaldehyde
Hexaldehyde
Isovaleraldehyde
Propionaldehyde
rolualdehydes
Valeraldehyde
Average by Site
PXSS
1.23
0.89
1.82
1.22
1.93
NA
0.92
3.64
NA
1.64
3.81
2.03
1.91
S4MO
0.47
0.79
3.65
1.43
1.99
NA
0.43
3.87
NA
2.02
4.70
2.82
2.22
SEWA
1.26
1.59
3.86
1.90
3.83
NA
0.78
4.00
NA
2.69
2.42
3.01
2.53
SKFL
0.62
0.79
3.70
3.90
2.43
NA
0.64
3.85
NA
2.94
4.50
3.38
2.68
SPIL
0.27
0.56
3.13
1.90
2.90
NA
0.56
2.21
NA
1.82
4.20
3.17
2.07
SSSD
2.14
0.54
4.21
3.22
3.77
NA
0.81
3.49
NA
1.54
4.96
3.26
2.80
SYFL
0.56
0.56
3.32
2.91
3.51
NA
0.49
2.62
NA
2.02
4.53
4.28
2.48
 NA = No pairs with concentrations greater than or equal to the MDL
BOLD ITALICS = EPA-designated NATTS Site
Gray shading indicates NATTS MQO Core Analyte
Orange shading indicates the site-specific average CV for this method; the overall average CV for this method is
calculated from the site-specific averages and provided in the final column of the table.
                                            29-38

-------
 Table 29-11. Carbonyl Compound Analytical Precision: Coefficient of Variation Based on
                          Replicate Analyses by Site (Continued)
Pollutant
Acetaldehyde
Acetone
Benzaldehyde
Butyraldehyde
Crotonaldehyde
2,5-Dimethylbenzaldehyde
Formaldehyde
Hexaldehyde
Isovaleraldehyde
Propionaldehyde
Tolualdehydes
Valeraldehyde
Average by Site
TMOK
0.47
0.55
2.89
2.73
1.79
NA
0.77
3.89
NA
2.30
3.25
2.60
2.12
TOOK
1.44
0.77
3.49
2.80
2.63
NA
1.25
4.71
NA
2.79
3.89
4.75
2.85
UCSD
0.59
1.01
4.34
2.91
4.25
NA
0.98
3.12
NA
3.12
3.11
4.69
2.81
WPIN
1.06
0.89
4.01
2.27
3.34
NA
1.10
3.49
NA
2.61
4.56
3.86
2.72
# of Pairs
326
326
324
324
316
0
326
324
0
326
269
317
~
Average
by
Pollutant
0.79
0.73
3.56
2.39
2.69
NA
0.84
3.57
NA
2.10
4.08
3.54
2.43
       NA = No pairs with concentrations greater than or equal to the MDL
      BOLD ITALICS = EPA-designaled NATTS Site
      Gray shading indicates NATTS MQO Core Analyte
      Orange shading indicates the site-specific average CV for this method; the overall average CV for this
      method is calculated from the site-specific averages and provided in the final column of the table.

29.3.4 PAH Analytical Precision
       Table 29-12 presents analytical precision results from replicate analyses of all collocated
PAH samples  as the CV per pollutant per site, the average CV per site, the average CV per
pollutant, and  the overall average CV for NMP sites sampling PAHs. The analytical precision
results from replicate analysis of collocated samples show low- to mid-level variability among
sites, ranging from a CV of 0.82 percent (benzo(a)anthracene for SYFL) to 41.40 percent
(anthracene for DEMI). The pollutant-specific average  CV ranged from 1.21 percent
(dibenz(a,h)anthracene) to 15.30 percent (anthracene). The site-specific average CV ranged from
3.15 percent for SEWA to 7.41  percent for DEMI. The  overall  average CV for all sites was
4.61 percent.
                                          29-39

-------
    Table 29-12. PAH Analytical Precision: Coefficient of Variation Based on Replicate
                                      Analyses by Site
Pollutant
Acenaphthene
Acenaphthylene
Anthracene
Benzo(a)anthracene
Benzo(a)pyrene
Benzo(b)fluoranthene
Benzo(e)pyrene
Benzo(g,h,i)perylene
Benzo(k)fluoranthene
Chrysene
Coronene
Cyclopenta[cd]pyrene
Dibenz(a,h)anthracene
Fluoranthene
Fluorene
9-Fluorenone
Indeno( 1 ,2,3 -cd)pyrene
Naphthalene
Perylene
Phenanthrene
Pyrene
Retene
Average by Site
DEMI
8.73
3.26
41.40
36.75
4.41
2.75
2.51
2.79
2.61
1.71
3.96
NA
NA
3.43
3.38
3.34
4.55
2.85
NA
1.61
3.80
6.86
7.41
RUCA
2.73
2.09
4.19
2.43
NA
2.63
2.81
1.33
NA
3.65
NA
NA
NA
3.16
3.90
2.58
9.17
2.56
NA
1.11
3.79
3.50
3.23
SDGA
5.13
1.89
20.80
1.39
3.83
4.16
4.61
1.30
4.42
2.90
NA
2.21
NA
3.94
3.69
4.05
9.13
4.11
NA
1.74
4.06
4.46
4.62
SEWA
4.81
7.90
2.89
2.35
6.27
4.86
1.74
2.12
4.20
2.28
2.97
1.56
NA
2.61
3.38
2.96
0.91
2.51
NA
1.35
2.60
2.79
3.15
SYFL
2.57
13.34
7.21
0.82
7.42
3.68
5.91
3.35
7.24
3.24
10.94
NA
1.21
2.89
2.80
2.92
4.94
3.30
NA
2.22
2.78
4.31
4.66
# of Pairs
58
22
28
17
15
47
35
27
16
55
6
4
1
58
58
58
25
58
0
58
58
57
~
Average
by
Pollutant
4.80
5.70
15.30
8.75
5.48
3.62
3.52
2.18
4.62
2.76
5.96
1.89
1.21
3.21
3.43
3.17
5.74
3.06
NA
1.61
3.40
4.39
4.61
  NA = No pairs with concentrations greater than or equal to the MDL
  BOLD ITALICS = EPA-designated NATTS Site
  Gray shading indicates NATTS MQO Core Analyte
  Orange shading indicates the site-specific average CV for this method; the overall average CV for this method is
  calculated from the site-specific averages and provided in the final column of the table.


29.3.5 Metals Analytical Precision

       Table 29-13 presents analytical precision results from replicate analyses of all collocated

metals samples as the CV per pollutant per site, the average CV per site, the average CV per

pollutant, and the overall average CV for NMP sites sampling metals. The results from replicate

analyses of collocated samples show low- to mid-level variability among sites, ranging from a

CV of 0.30 percent (chromium for S4MO) to 34.02 percent (mercury for UNVT). The pollutant-

specific average CV ranged from 1.29 percent (lead) to 16.52 percent (mercury). The site-
                                           29-40

-------
specific average CV ranged from 3.01 percent for TOOK to 12.56 percent for GLKY. The
overall average analytical precision was 7.91 percent.
   Table 29-13. Metals Analytical Precision: Coefficient of Variation Based on Replicate
                                     Analyses by Site
Pollutant
Antimony
Arsenic
Beryllium
Cadmium
Chromium
Cobalt
Lead
Manganese
Mercury
Nickel
Selenium
Average by Site
BOMA
1.40
2.09
14.31
6.81
2.83
6.43
1.50
1.13
9.27
1.80
2.68
4.57
BTUT
1.17
11.01
24.85
7.53
NA
10.19
0.72
1.08
20.20
3.24
17.70
9.77
GLKY
2.85
11.23
17.82
4.95
NA
29.23
1.03
0.81
17.53
33.36
6.77
12.56
S4MO
0.96
7.21
13.30
2.79
0.30
19.55
0.61
1.15
11.51
9.58
8.85
6.89
TOOK
2.16
1.98
4.58
2.27
3.03
2.22
2.38
2.22
6.59
3.74
1.97
3.01
UNVT
2.24
16.36
NA
5.77
NA
15.71
1.52
3.44
34.02
2.86
13.83
10.64
# of Pairs
342
323
160
340
133
317
342
342
245
201
329
-
Average
by
Pollutant
1.80
8.32
14.97
5.02
2.05
13.89
1.29
1.64
16.52
9.10
8.63
7.91
 NA = No pairs with concentrations greater than or equal to the MDL
 BOLD ITALICS = EPA-designated NATTS Site
 Gray shading indicates NATTS MQO Core Analyte
 Orange shading indicates the site-specific average CV for this method; the overall average CV for this method is
 calculated from the site-specific averages and provided in the final column of the table.

29.3.6 Hexavalent Chromium Analytical Precision
       Table 29-14 presents analytical precision results from replicate analyses of all collocated
hexavalent chromium samples as the CV per site and the overall average CV for NMP sites
sampling hexavalent chromium. Hexavalent chromium is a NATTS MQO Core Analyte and the
sites shown are NATTS sites. The range of variability for hexavalent chromium was 3.56 percent
for HOWI to 13.83 percent for BTUT, with an overall average analytical precision of
6.69 percent.
                                           29-41

-------
Table 29-14. Hexavalent Chromium Analytical Precision: Coefficient of Variation Based on
                                Replicate Analyses by Site
Site
BOMA
BTUT
CAMS 35
CHSC
DEMI
GLKY
GPCO
HOW
MONY
NBIL
PRRI
PXSS
RIVA
ROCH
S4MO
SDGA
SEWA
SKFL
SYFL
UNVT
WADC
# of Pairs
Average by Site
Average CV
8.29
13.83
5.22
7.32
4.56
5.30
8.03
3.56
5.54
6.44
7.19
4.92
6.26
9.40
6.57
5.44
5.64
5.71
5.27
7.74
8.34
223
6.69
                               BOLD ITALICS = EPA-designated
                               NATTS Site
                               Orange shading indicates the average CV
                               for this method
29.4   Accuracy
       Laboratories typically evaluate their accuracy (or bias) by analyzing audit samples that
are prepared by an external source. The pollutants and the respective concentrations of the audit
samples are unknown to the laboratory. The laboratory analyzes the samples and the external
source compares the measured concentrations to the reference concentrations of those audit
samples and calculates a percent difference. Accuracy, or bias, indicates the extent to which
experimental measurements represent their corresponding "true" or "actual" values.
                                          29-42

-------
       Laboratories participating in the NATTS program are provided with proficiency test (PT)
audit samples for VOCs, carbonyl compounds, PAHs, metals, and hexavalent chromium, which
are used to quantitatively measure analytical accuracy. Tables 29-15 through 29-19 present
ERG's results from the 2011 NATTS PT audit samples for VOCs, carbonyl compounds, PAHs,
metals, hexavalent chromium, respectively. The program MQO for the percent difference from
the true value is ฑ 25 percent, and the values exceeding this criterion are bolded in the tables.
Shaded rows present results for NATTS MQO Core Analytes. The percent difference calculation
is:
                               Percent Difference =
:100
    Where:
          X\ is the analytical result from the laboratory;
          Xi is the true concentration of the audit sample

    Table 29-15. VOC NATTS PT Audit Samples-Percent Difference from True Value
Pollutant
Acrolein
Benzene
1,3 -Butadiene
Carbon Tetrachloride
Chloroform
1 ,2-Dibromoethane
1 ,2-Dichloroethane
Dichloromethane
1 ,2-Dichloropropane
cis- 1 ,3 -Dichloropropene
trans- 1 ,3 -Dichloropropene
1 , 1 ,2,2-Tetrachloroethane
retrachloroethylene
rrichloroethylene
Vinyl Chloride
November 2011
-20.3
10.2
10.9
25.4
-9.7
5.7
-11.8
6.4
2.8
21.5
41.3
1.1
0.8
-8.3
0.3
                        Gray shading indicates NATTS MQO Core Analyte
                                         29-43

-------
Table 29-16. Carbonyl Compound NATTS PT Audit Samples-Percent Difference from
                                  True Value
Pollutant
Acetaldehyde
Benzaldehyde
Formaldehyde
Propionaldehyde
November 2011
3.0
8.0
-3.4
-14.6
                        Gray shading indicates NATTS MQO Core
                        Analyte
  Table 29-17. PAH NATTS PT Audit Samples-Percent Difference from True Value
Pollutant
Acenaphthene
Anthracene
Benzo(a)pyrene
Fluoranthene
Fluorene
Naphthalene
Phenanthrene
Pyrene
July 2011
-11.9
-13.6
-2.1
-9.4
-6.9
-13.9
-7.9
-6.9
                  Gray shading indicates NATTS MQO Core Analyte
 Table 29-18. Metals NATTS PT Audit Samples-Percent Difference from True Value
Pollutant
Arsenic
Beryllium
Cadmium
Cobalt
Lead
Manganese
Nickel
Selenium
July 2011
1.1
-8.2
-6.0
-4.3
-6.3
-4.0
-6.8
-15.6
                        Gray shading indicates NATTS MQO Core
                        Analyte
 Table 29-19. Hexavalent Chromium PT Audit Samples-Percent Difference from True
                         Value Across Multiple Samples
Pollutant
Hexavalent Chromium
July 2011
Concentration
#1
-2.6
Concentration
#2
-6.0
Concentration
#3
-1.7
Concentration
#4
-0.8
      Gray shading indicates NATTS MQO Core Analyte
                                     29-44

-------
       The accuracy of the 2011 monitoring data can also be assessed qualitatively by reviewing

the accuracy of the monitoring methods and how they were implemented:

       •      The sampling and analytical methods used in the 2011 monitoring effort have
              been approved by EPA for accurately measuring ambient levels of various
              pollutants - an approval that is based on many years of research into the
              development of ambient air monitoring methodologies.

       •      When collecting and analyzing ambient air samples, all field sampling staff and
              laboratory analysts are required to strictly adhere to quality control and quality
              assurance guidelines detailed in the respective monitoring methods. This strict
              adherence to the well-documented sampling and analytical methods suggests that
              the 2011 monitoring data accurately represent ambient air quality.
                                         29-45

-------
30.0   Results, Conclusions, and Recommendations

       The following discussion summarizes the results of the data analyses contained in this

report, renders conclusions based on those results, and presents recommendations applicable to

future air monitoring efforts. As demonstrated by the results of the data analyses discussed
throughout this report, NMP monitoring data offer a wealth of information for assessing air

quality by evaluating trends, patterns, correlations, and the potential for health risk and should

ultimately assist a wide range of audiences understand the complex nature of air pollution.


30.1   Summary of Results
       Analyses of the 2011 monitoring data identified the following notable results,

observations, trends, and patterns in the program-level and state- and site-specific air monitoring

data.


30.1.1  National-level Results Summary

       •  Number of participating NATTS sites. Twenty-five of the 51 sites are EPA-designated
          NATTS sites (BOMA, BTUT, CAMS 35, CAMS 85, CELA, CHSC, DEMI, GLKY,
          GPCO, HOWI, MONY, NBIL, PRRI, PXSS, RIVA, ROCH, RUCA, S4MO, SDGA,
          SEW A, SJJCA, SKFL, SYFL, UNVT, and WADC).

       •  Total number of samples collected and analyzed.  Over 8,800 samples were collected
          yielding over 218,900 valid measurements of air toxics.

       •  Detects. The detection of a given pollutant is subject to the sensitivity limitation
          associated with the analytical methods used and the limitations of the instruments.
          Simply stated, a method detection limit is the lowest concentration of a target
          pollutant that can be measured and reported with  99 percent confidence that the
          pollutant concentration is greater than zero. Approximately 54 percent of the reported
          measurements were above the associated MDLs.  Of the 177 pollutants monitored,
          only two pollutants were not detected over the  course of the 2011 monitoring effort:
          c/5-l,2-dichloroethylene and 2,5-dimethylbenzaldehyde.

       •  Program-level Pollutants of Interest. The pollutants of interest at the program-level
          are based on the number of exceedances, or "failures," of the risk screening values. In
          addition, the 18 NATTS MQO Core Analytes (excluding acrolein) are inherently
          classified as pollutants of interest. Only two NATTS MQO Core Analytes (beryllium
          and vinyl chloride) did not fail any screens. Hexachloro-1,3-butadiene is new to the
          program-level  pollutants of interest for 2011.
                                          30-1

-------
       •  Noncancer Risk-Based Screening using A TSDR MRLs. Where an MRL was available,
          all of the preprocessed daily measurements were less than the associated ATSDR
          acute MRLs. Additionally, all of the quarterly or annual average concentrations of the
          pollutants with MRLs were less than the associated ATSDR intermediate or chronic
          MRLs.

       •  Mobile Sources. Site-specific hydrocarbon concentrations had positive correlations
          with county-level and 10-mile motor vehicle ownership data, traffic data, and VMT
          data. While these correlations were not "strong", they do indicate that hydrocarbon
          concentrations tend to increase with increasing motor vehicle activity data.

       •  Carbon Tetrachloride. Although production of carbon tetrachloride has declined
          sharply over the last 30 years due to its role as an ozone depleting substance, it has a
          relatively long atmospheric lifetime and thus, is present in at similar levels at any
          given location. NMP sites are located in a variety of locations across the county with
          difference purposes behind the monitoring at each site. The relative uniformity in the
          concentrations of carbon tetrachloride across the program confirms the ubiquitous
          nature of this pollutant and is an indication of the representativeness of the data
          generated under the program.

       •  Seasonal Trends. Formaldehyde concentrations tended to be highest during the third
          quarter of 2011, or during the period from July to September. Acenaphthene and
          fluorene  concentrations exhibit a similar pattern. Conversely, benzene concentrations
          tended to be higher during the first or fourth quarters of the year, or between January
          through March and October through December. Benzo(a)pyrene and 1,3-butadiene
          concentrations exhibit a similar trend.  Arsenic concentrations tended to be highest
          between  October and December.


30.1.2  State-level Results Summary


       Arizona.

       •  The Arizona monitoring sites are located in Phoenix. PXSS is a NATTS site; SPAZ is
          a UATMP site.

       •  PXSS sampled for  VOCs, carbonyl compounds, PAHs,  metals (PMio), and
          hexavalent chromium. SPAZ sampled for VOCs only.

       •  Twenty-three pollutants, of which 13 are NATTS MQO Core Analytes, failed screens
          for PXSS. PXSS failed the second highest number of screens among all NMP sites.

       •  Ten pollutants failed screens for SPAZ, of which four are NATTS MQO Core
          Analytes.
                                          30-2

-------
•  Of the pollutants of interest for PXSS, benzene had the highest annual average
   concentration and was the only pollutant with an annual average concentration greater
   than 1 |ig/m3 for this site.

•  Xylenes had the highest annual average concentration for SPAZ. Benzene and
   ethylbenzene also had annual average concentrations greater than 1 |ig/m3.

•  PXSS had the highest annual average concentration of tetrachloroethylene,
   hexavalent chromium, beryllium, and manganese among NMP sites sampling these
   pollutants.

•  SPAZ had the highest annual average concentrations of 1,3-butadiene,
   />-dichlorobenzene, and ethylbenzene among NMP sites sampling these pollutants.

•  Annual average concentrations could not be calculated for the carbonyl compounds
   for PXSS due to a sampler problem that led to the invalidation of carbonyl
   compounds samples through the  end  of March 2011.

•  Sampling for metals (PMi0) and  hexavalent chromium has occurred at PXSS for at
   least 5 consecutive years; thus, a trends analysis was  conducted for select pollutants.
   The concentrations of arsenic, lead, and manganese increased for 2011 from 2010
   levels.

•  Benzene and 1,3-butadiene had the highest cancer risk approximations for PXSS  and
   SPAZ. None of the pollutants of interest for either site had a noncancer hazard
   approximation greater than 1.0.

•  Benzene was the highest emitted pollutant with a cancer toxicity factor in Maricopa
   County, while toluene was the highest emitted pollutant with a noncancer toxicity
   factor. Formaldehyde had the highest cancer toxicity-weighted emissions, while
   acrolein had the highest noncancer toxicity-weighted emissions for Maricopa County.


California.

•  The three California monitoring  sites are located in Los Angeles (CELA), Rubidoux
   (RUCA), and San Jose (SJJCA).  All three are NATTS sites.

•  CELA and RUCA sampled for PAHs only. SJJCA sampled for PAHs and metals
   (PMio).

•  Three pollutants failed screens for CELA, of which one (naphthalene) is a NATTS
   MQO Core Analyte. Naphthalene was the only pollutant to fail screens for RUCA.
   Six pollutants failed screens for SJJCA, of which five are NATTS MQO Core
   Analytes.
                                   30-2

-------
•  Naphthalene had the highest annual average concentration for each site. The annual
   average concentration of naphthalene for CELA was higher than the annual average
   concentrations for RUCA and SJJCA, and fourth highest compared to all NMP sites
   sampling naphthalene.

•  Of the pollutants of interest for each site, naphthalene exhibited the highest cancer
   risk approximation for all three California sites. The noncancer hazard approximation
   for each pollutant of interest was less than 1.0 for all three sites.

•  Formaldehyde was the highest emitted pollutant with a cancer toxicity factor in Los
   Angeles, Riverside, and Santa Clara Counties; formaldehyde also had the highest
   cancer toxicity-weighted emissions for Los Angeles and Santa Clara Counties while
   hexavalent chromium had the highest cancer toxicity-weighted emissions for
   Riverside County.

•  1,1,1 -Trichloroethane was the highest emitted pollutant with a noncancer toxicity
   factor in Los Angeles County, while toluene was the highest emitted pollutant with a
   noncancer toxicity factor in Riverside and Santa Clara Counties. Acrolein had the
   highest noncancer toxicity-weighted emissions for all three counties.


Colorado.

•  The NATTS site in Colorado is located in Grand Junction (GPCO). There are also
   four UATMP sites located northeast of Grand Junction in Garfield County. The sites
   are located in the towns of Battlement Mesa (BMCO), Silt (BRCO), Parachute
   (PACO), and Rifle (RICO).

•  GPCO sampled for VOCs, carbonyl compounds, PAHs, and hexavalent chromium.
   The Garfield County  sites sampled for SNMOCs and carbonyl compounds.

•  Twenty-three pollutants failed at least one screen for GPCO, of which eight are
   NATTS MQO Core Analytes. The number of pollutants that failed screens for the
   Garfield County sites ranged from four (BRCO) to five (BMCO, PACO, and RICO).

•  Of the pollutants of interest for GPCO, formaldehyde had the highest annual average
   concentration, followed by acetaldehyde and benzene.

•  Annual average concentrations for the carbonyl compounds for the Garfield County
   sites could not be calculated because these sites did not meet the necessary
   completeness criteria. The same is true for SNMOCs for BMCO. Benzene had the
   highest annual average concentration for the three remaining Garfield County sites;
   PACO had the highest annual average benzene concentration of all three sites.

•  GPCO had the highest annual average concentrations of naphthalene and
   benzo(a)pyrene among sites sampling PAHs. GPCO also had the fourth highest
   annual average ethylbenzene concentration among all NMP sites sampling this

                                   30-4

-------
   pollutant. PACO, GPCO, and RICO are listed among the NMP sites with the highest
   annual average concentrations for sites that sampled benzene.

•  VOC, carbonyl compound, and hexavalent chromium sampling has occurred at
   GPCO for at least 5 consecutive years; thus, a trends analysis was conducted for
   select pollutants. After several years without significant change, formaldehyde
   concentrations decreased significantly from 2009 to 2010, then held steady in 2011.
   Benzene concentrations exhibit an overall decreasing trend over recent years.
   Concentrations of hexavalent chromium have increased at GPCO since 2009.

•  Formaldehyde had the highest cancer risk approximation for GPCO. Benzene had the
   highest cancer risk approximation for each of the three Garfield County sites where
   an annual average could be calculated (BRCO, PACO, and RICO). All noncancer
   hazard approximations were less than  1.0 for all five Colorado sites.

•  Benzene was the highest emitted pollutant with a cancer toxicity factor in Mesa
   County, while formaldehyde was the highest emitted pollutant with a cancer toxicity
   factor in Garfield County. Formaldehyde  had the highest cancer toxicity-weighted
   emissions for both counties.

•  While toluene was the highest emitted pollutant with a noncancer toxicity factor for
   both Mesa and Garfield Counties, acrolein had the highest noncancer toxicity -
   emissions.
District of Columbia

•  The Washington, D.C. monitoring site (WADC) is a NATTS site.

•  WADC sampled for hexavalent chromium and PAHs. The only pollutant to fail
   screens for WADC was naphthalene.

•  The pollutant with the highest annual average concentration for WADC was
   naphthalene, which was significantly higher than the annual average concentrations
   for benzo(a)pyrene and hexavalent chromium.

•  Hexavalent chromium sampling has occurred at WADC for at least 5 consecutive
   years; thus, a trends analysis was conducted. The average concentration of hexavalent
   chromium increased significantly from 2009 to 2010 then held steady for 2011.

•  Naphthalene had the only cancer risk approximation for WADC greater than
   1.0 in-a-million. None of the pollutants of interest had a noncancer hazard
   approximation greater than 1.0.
                                   30-5

-------
•  Benzene was the highest emitted pollutant with a cancer toxicity factor in the District
   of Columbia, while toluene was the highest emitted pollutant with a noncancer
   toxicity factor. Formaldehyde had the highest cancer toxicity-weighted emissions,
   while acrolein had the highest noncancer toxicity-weighted emissions in the District.


Florida.

•  Three of the Florida monitoring sites are located in the Tampa-St. Petersburg-
   Clearwater MSA (SYFL, AZFL, and SKFL) and two are located in the Orlando-
   Kissimmee MSA (ORFL and PAFL). SKFL and SYFL are NATTS sites.

•  AZFL and ORFL sampled for carbonyl compounds only. SKFL and SYFL sampled
   for hexavalent chromium and PAHs in addition to carbonyl compounds. PAFL
   sampled only metals (PMio).

•  Acetaldehyde and formaldehyde were the only pollutants to fail screens for AZFL
   and ORFL, where only carbonyl compounds were sampled. Five pollutants failed
   screens for SKFL and four pollutants failed screens for SYFL. Arsenic was the only
   pollutant to fail screens for PAFL, where only metals were sampled.

•  Acetaldehyde had the highest annual average concentration for AZFL, while
   formaldehyde had the highest annual average concentration for SKFL, SYFL, and
   ORFL. SYFL had the highest annual average concentration of formaldehyde among
   the Florida sites. Manganese and lead had the highest annual  average concentrations
   for PAFL.

•  Carbonyl compound sampling has been conducted at AZFL, ORFL, SKFL, and
   SYFL for at least 5 consecutive years; thus a trends analysis was conducted for
   acetaldehyde and formaldehyde. Hexavalent chromium sampling has also been
   conducted at SYFL for at least 5 consecutive years. From 2010 to 2011, acetaldehyde
   concentrations have decreased at AZFL, SKFL, and SYFL while acetaldehyde
   concentrations have increased at ORFL. Concentrations of formaldehyde increased
   from 2010 to 2011 at SKFL, but have been steadily decreasing at ORFL since the
   onset of sampling. Hexavalent chromium concentrations have been increasing since
   2009 at SYFL.

•  For the Florida sites sampling carbonyl compounds, formaldehyde had the highest
   cancer risk approximations. Arsenic had the highest cancer risk approximation for
   PAFL. All noncancer hazard approximations for the pollutants  of interest for the
   Florida  sites were less than 1.0.

•  Benzene was the highest emitted pollutant with a cancer toxicity factor in all three
   Florida  counties. Benzene also had the highest cancer toxicity-weighted emissions for
   Pinellas County, while formaldehyde had the highest cancer toxicity-weighted
   emissions for Hillsborough and Orange Counties.
                                  30-6

-------
•  Toluene was the highest emitted pollutant with a noncancer toxicity factor in all three
   Florida counties. Acrolein had the highest noncancer toxicity-weighted emissions for
   all three counties.


Georgia.

•  The SDGA monitoring site located in Decatur, south of Atlanta, is a NATTS site.

•  SDGA sampled for PAHs and hexavalent chromium. Naphthalene, acenaphthene, and
   fluorene failed screens for SDGA, with naphthalene accounting for the majority of
   the total failed screens.

•  Of the pollutants of interest for SDGA, naphthalene had the highest annual average
   concentration, ranking tenth among NMP sites sampling PAHs.

•  Hexavalent chromium sampling has occurred at SDGA for at least 5 consecutive
   years; thus, a trends analysis was conducted. The range of hexavalent chromium
   measurements has decreased since the onset of sampling in 2006.

•  Naphthalene was the only pollutant of interest with a cancer risk approximation
   greater than 1.0 in-a-million. None of the pollutants of interest for SDGA had a
   noncancer hazard approximation greater than 1.0.

•  Benzene was the highest emitted pollutant with a cancer toxicity factor in DeKalb
   County, while toluene was the highest emitted pollutant with a noncancer toxicity
   factor. Benzene also had the highest cancer toxicity-weighted emissions, while
   acrolein had the highest noncancer toxicity-weighted emissions for DeKalb County.


Illinois.

•  The Illinois monitoring sites are located near Chicago. NBIL is a NATTS site located
   in Northbrook and SPIL is a UATMP site located in Schiller Park.

•  Both Illinois sites sampled for VOCs and carbonyl compounds. NBIL also sampled
   for SNMOCs, PAHs, hexavalent chromium, and metals (PMio).

•  Twenty-four pollutants failed screens for NBIL, of which 13 are NATTS MQO Core
   Analytes.  Sixteen pollutants failed screens for SPIL, of which six  are NATTS MQO
   Core Analytes.

•  Of the pollutants of interest for NBIL, chloroform had the highest annual average
   concentration. This is also the highest annual average concentration of chloroform
   among NMP sites sampling this pollutant. NBIL also had the highest annual average
   concentration of fluorene among all NMP sites sampling PAHs.
                                   30-7

-------
•  Formaldehyde had the highest annual average concentration for SPIL. SPIL had the
   highest annual average concentrations of trichloroethylene and acrylonitrile among
   NMP sites sampling these pollutants.

•  VOC and carbonyl compound sampling have been conducted at NBIL and SPIL for
   at least 5 consecutive years. In addition, metals (PMio) and hexavalent chromium
   sampling have been conducted at NBIL for at least 5 consecutive years. Thus, a
   trends analysis was conducted for these methods for both sites. Concentrations of
   acetaldehyde and 1,3-butadiene have an increasing trend at both sites in recent years.
   The concentrations for several of the pollutants for which a trends analysis was
   performed were at a minimum in 2009, particularly for SPIL.

•  Formaldehyde had the highest cancer risk approximation for NBIL, while
   acrylonitrile had the highest cancer risk approximation for SPIL. All none
   hazard approximations for the pollutants of interest for the Illinois sites w
acrylomtrile nad tne nignest cancer risk approximation tor brlL. All noncancer
hazard approximations for the pollutants of interest for the Illinois sites were less
than 1 0
•  Benzene was the highest emitted pollutant with a cancer toxicity factor in Cook
   County, while formaldehyde had the highest cancer toxicity-weighted emissions.
   Toluene was the highest emitted pollutant with a noncancer toxicity factor, while
   acrolein had the highest noncancer toxicity-weighted emissions for Cook County.


Indiana.

•  There are two Indiana monitoring sites, one located in Indianapolis (WPIN), and a
   second located in Gary, near Chicago (INDEM). Both are UATMP sites.

•  WPIN and INDEM sampled for carbonyl compounds only.

•  Formaldehyde and acetaldehyde failed screens for both INDEM and WPIN;
   propionaldehyde also failed a single screen for INDEM.

•  Of the pollutants of interest, formaldehyde had the highest annual average
   concentration  for  INDEM. Annual average concentrations for WPIN could not be
   calculated due to intermittent sampler issues.

•  Carbonyl compound sampling has been conducted at INDEM for at least
   5 consecutive  years; thus, a trends analysis was conducted for acetaldehyde and
   formaldehyde. The concentration of both acetaldehyde and formaldehyde decreased
   dramatically at INDEM between 2008 and 2009, with little change since.

•  Formaldehyde had the highest cancer risk approximation for INDEM. Neither of the
   pollutants of interest for INDEM had a noncancer hazard approximation greater than
   1.0. Annual average concentrations could not be calculated for WPIN, therefore,
   cancer risk and noncancer hazard approximations could not be calculated.
                                   30-8

-------
•  Benzene was the highest emitted pollutant with a cancer toxicity factor in Marion and
   Lake Counties. Coke oven emissions (PM) had the highest cancer toxicity-weighted
   emissions for Lake County while formaldehyde had the highest cancer toxicity-
   weighted emissions for Marion County.

•  Toluene was the highest emitted pollutant with a noncancer toxicity factor in both
   Lake and Marion Counties. Acrolein had the highest noncancer toxicity-weighted
   emissions for both counties.


Kentucky.

•  The Kentucky monitoring  site is located near Grayson Lake, south of Grayson,
   Kentucky (GLKY).  GLKY is a NATTS site.

•  GLKY sampled for hexavalent chromium, metals (PMio), carbonyl compounds,
   PAHs, and VOCs. Fifteen  pollutants failed screens for GLKY, of which nine are
   NATTS MQO Core Analytes.

•  Annual averages could not be calculated for metals (PMio) and carbonyl compounds
   because sampling did not begin until May and  August, respectively. Of the pollutants
   of interest for which annual averages could be  calculated, carbon tetrachloride and
   benzene had the highest annual average concentrations for GLKY, though none of the
   calculated annual average concentrations were greater than 1 |ig/m3.

•  Acrylonitrile had the highest cancer risk approximation for GLKY, followed by
   benzene and carbon tetrachloride.  None of the pollutants of interest for GLKY had
   noncancer hazard approximations greater than  1.0.

•  Benzene was the highest emitted pollutant with a cancer toxicity factor in Carter
   County and had the highest cancer toxicity-weighted emissions. Toluene was the
   highest emitted pollutant with a noncancer toxicity factor, while acrolein had the
   highest noncancer toxicity-weighted emissions in Carter County.


Massachusetts.

•  The Massachusetts monitoring site (BOMA) is a NATTS site located in Boston.

•  BOMA sampled for metals (PMio), PAHs, and hexavalent chromium.

•  Seven pollutants failed screens for BOMA, of which five are NATTS MQO Core
   Analytes. Naphthalene accounted for nearly half of the site's failed screens.

•  Of the pollutants of interest, naphthalene had the highest annual average
   concentration. The annual  average beryllium concentration for BOMA ranked third
   highest among sites sampling metals (PMio).
                                   30-9

-------
•  Metals and hexavalent chromium sampling have been conducted at BOMA for at
   least 5 consecutive years; thus, a trends analysis was conducted for arsenic,
   hexavalent chromium, and manganese. The concentrations of these pollutants have
   changed little in recent years.

•  The only pollutants of interest with cancer risk approximations greater than 1.0 in-a-
   million were arsenic and naphthalene. None of the pollutants of interest for BOMA
   had noncancer hazard approximations greater than 1.0.

•  Benzene was the highest emitted pollutant with a cancer toxicity  factor in Suffolk
   County, while formaldehyde had the highest cancer toxicity-weighted emissions.
   Toluene was the highest emitted pollutant with a noncancer toxicity factor in Suffolk
   County, while acrolein had the highest noncancer toxicity-weighted emissions.


Michigan.

•  The three Michigan monitoring sites are located in the Detroit area. DEMI is a
   NATTS site located in Dearborn. RRMI and SWMI are UATMP sites located  in
   River Rouge and Detroit,  respectively.

•  All three Michigan sites sampled carbonyl compounds; DEMI also sampled VOCs,
   PAHs, and hexavalent chromium.

•  Nineteen pollutants failed screens for DEMI, of which eight are NATTS MQO Core
   Analytes. Acetaldehyde and formaldehyde both failed screens for RRMI and SWMI;
   propionaldehyde also failed one screen for SWMI.

•  Formaldehyde had the highest annual average concentration for all three Michigan
   sites. Compared to other NMP  sites, DEMI had the second highest annual average
   concentration of chloroform among sites sampling VOCs. DEMI also had the highest
   annual average  concentration of acenaphthene and the second highest annual average
   concentrations of fluorene and  naphthalene among sites sampling PAHs. DEMI also
   has the fourth highest annual average concentration of hexavalent chromium.

•  Hexavalent chromium, carbonyl compound, and VOC sampling has been conducted
   at DEMI for at least 5 consecutive years; thus, a trends analysis was conducted for
   select pollutants. The most notable trend is for benzene. Benzene concentrations
   exhibit a steady decreasing trend although concentrations have leveled out in recent
   years.

•  Formaldehyde had the highest cancer risk approximation for all three sites. None of
   the pollutants of interest for the Michigan sites had a noncancer hazard approximation
   greater than 1.0.

•  Benzene was the highest emitted pollutant with a cancer toxicity  factor in Wayne
   County, while coke oven  emissions had the highest cancer toxicity-weighted

                                   30-10

-------
   emissions. Hydrochloric acid was the highest emitted pollutant with a noncancer
   toxicity factor in Wayne County, while acrolein had the highest noncancer toxicity -
   weighted emissions.


Missouri.

•  The NATTS site in Missouri (S4MO) is located in St. Louis.

•  S4MO sampled for VOCs, carbonyl compounds, PAHs, metals (PMio), and
   hexavalent chromium.

•  Twenty-five pollutants failed at least one screen for S4MO, of which 13 are NATTS
   MQO Core Analytes. S4MO failed the greatest number of screens among NMP sites.

•  Of the pollutants of interest, formaldehyde and acetaldehyde had the highest annual
   average concentrations for S4MO. S4MO had the highest annual average
   concentrations of hexachloro-1,3-butadiene, arsenic, cadmium, and lead among all
   NMP sites sampling these pollutants.

•  Carbonyl compound, VOC, metals (PMio), and hexavalent chromium sampling have
   been conducted at S4MO for at least 5 consecutive years; thus, a trends analysis was
   conducted for select pollutants. Concentrations of acetaldehyde and benzene
   decreased from 2010 to 2011. Concentrations of formaldehyde increased significantly
   from 2010 to 2011.

•  Formaldehyde had the highest cancer risk approximation for S4MO. None of the
   pollutants of interest for S4MO had a noncancer hazard approximation greater
   than 1.0.

•  Benzene was the highest emitted pollutant with a cancer toxicity factor in St. Louis
   (city), while toluene was the highest emitted pollutant with a noncancer toxicity
   factor. Hexavalent chromium had the highest cancer toxicity-weighted emissions,
   while acrolein had the highest noncancer toxicity-weighted emissions in St. Louis
   (city).


New Jersey.

•  The four UATMP sites in New Jersey are located in Chester (CHNJ), Elizabeth
   (ELNJ), New Brunswick (NBNJ), and Paterson (PANJ).

•  CHNJ, ELNJ, and NBNJ sampled for VOCs and carbonyl compounds, while PANJ
   sampled for VOCs only.

•  Seventeen pollutants failed at least one screen for CHNJ;  16 pollutants failed at least
   one screen for ELNJ; 18 pollutants failed screens for NBNJ; and 10 failed screens for
   PANJ.

                                  30-11

-------
•  Of the pollutants of interest, formaldehyde had the highest annual average
   concentration for CHNJ, ELNJ, and NBNJ. Annual average concentrations could not
   be calculated for PANJ due to a combination of a shortened sampling duration
   (sampling ended in the middle of May) and a l-in-12 day sampling schedule.

•  Compared to other NMP sites, CHNJ had the highest annual average concentration of
   1,2-dichloroethane among sites sampling VOCs and ELNJ had the highest annual
   average concentration of formaldehyde among sites sampling carbonyl compounds.

•  Carbonyl compound and VOC sampling has been conducted at CHNJ, ELNJ, and
   NBNJ for at least 5 consecutive years; thus, a trends analysis was conducted for select
   pollutants. Several of the pollutants for which a trends analysis was conducted exhibit
   slight increasing trends from 2010 to 2011, most notably 1,3-butadiene concentrations
   at CHNJ and acetaldehyde concentrations for ELNJ.

•  Formaldehyde had the highest cancer risk approximations for CHNJ, ELNJ, and
   NBNJ. None of the pollutants of interest for any of the New Jersey sites had
   noncancer hazard approximations greater than 1.0. Cancer risk and noncancer hazard
   approximations were not available for PANJ because annual average concentrations
   could not be calculated.

•  Benzene was the highest emitted pollutant with a cancer URE in Union, Middlesex,
   Morris, and Passaic Counties. Benzene also had the highest toxicity-weighted
   emissions for Morris and Passaic Counties, while formaldehyde had the highest
   toxicity-weighted emissions for Union and Middlesex Counties.

•  Toluene was the highest emitted pollutant with a noncancer toxicity factor in all four
   counties, while acrolein had the highest noncancer toxicity-weighted emissions for
   each county.


New York.

•  The two New York monitoring sites are located in New York City (MONY) and
   Rochester (ROCH). Both are NATTS  monitoring sites.

•  Both New York sites sampled PAHs and hexavalent chromium.

•  Seven pollutants failed screens for MONY and four pollutants failed screens for
   ROCH. Naphthalene failed the majority of screens for both sites.

•  Naphthalene had the highest annual average concentration for both MONY and
   ROCH.

•  Naphthalene had the highest cancer risk approximation for both sites. None of the
   pollutants of interest for either New York site had noncancer hazard approximations
   greater than  1.0.

                                  30-12

-------
•  Benzene was the highest emitted pollutant with a cancer toxicity factor for Bronx and
   Monroe Counties and had the highest cancer toxicity-weighted emissions for both
   counties.

•  Methanol was the highest emitted pollutant with a noncancer toxicity factor in Bronx
   County, while toluene was the highest emitted pollutant with a noncancer toxicity
   factor in Monroe County. Acrolein had the highest noncancer toxicity-weighted
   emissions for both counties.


Oklahoma.

•  There are five UATMP sites in Oklahoma: two located in Tulsa (TOOK and TMOK),
   one in Pryor Creek (PROK), one in Oklahoma City (OCOK), and one in the
   Oklahoma City suburb of Midwest City (MWOK).

•  Each of the Oklahoma sites sampled for VOCs, carbonyls compounds, and metals
   (TSP).

•  Twenty pollutants failed screens for TOOK; 19 failed screens for TMOK; 15 failed
   screens for PROK; 17 failed screens for MWOK; and 16 failed screens for OCOK.

•  Of the pollutants of interest, formaldehyde had the highest annual average
   concentration for each Oklahoma site.

•  TOOK had the highest annual  average concentration of benzene  among NMP sites
   sampling this pollutant. The five Oklahoma sites account for the  third through
   seventh highest annual average concentrations of formaldehyde among NMP sites
   sampling carbonyl compounds.

•  TOOK has sampled carbonyl compounds, VOCs, and TSP metals for at least
   5 consecutive years, therefore a trends analysis was conducted for select pollutants.
   Concentrations of acetaldehyde, benzene, 1,3-butadiene, and manganese (TSP)
   exhibit an increasing trend since 2009.

•  Formaldehyde  and benzene had the highest cancer risk approximations for all of the
   Oklahoma monitoring sites.  The benzene cancer risk approximation for TOOK is the
   highest benzene cancer risk approximation program-wide. Arsenic had the highest
   cancer risk approximations among the metals. None of the pollutants of interest for
   the Oklahoma sites had a noncancer hazard approximation greater than 1.0.

•  Benzene was the highest emitted pollutant with a cancer toxicity factor in Mayes,
   Oklahoma, and Tulsa Counties. Arsenic had the highest cancer toxicity-weighted
   emissions for Mayes County while benzene had the highest cancer toxicity-weighted
   emissions for Oklahoma and Tulsa Counties.
                                  30-13

-------
•  Toluene was the highest emitted pollutant with a noncancer toxicity factor in Tulsa
   and Oklahoma Counties, while hydrochloric acid was the highest emitted pollutant
   with a noncancer toxicity factor in Mayes County.  Acrolein had the highest
   noncancer toxicity-weighted emissions for all three counties.


Rhode Island.

•  The Rhode Island monitoring site (PRRI) is located in Providence and is a NATTS
   site.

•  PRRI sampled for PAHs and hexavalent chromium.

•  Five pollutants failed screens for PRRI; 86 percent of failed screens are attributable to
   naphthalene.

•  The annual average concentration of naphthalene was significantly higher than the
   annual averages for the other pollutants of interest.

•  Hexavalent chromium sampling  has been conducted at PRRI for at least
   5 consecutive years; thus, a trends analysis was conducted. Concentrations of
   hexavalent chromium exhibit an increasing trend for the last 2 years of sampling.

•  Naphthalene had the highest cancer risk approximation for PRRI and is the only one
   greater than 1.0 in-a-million; all  noncancer hazard  approximations for PRRI were less
   than 1.0.

•  Benzene was the highest emitted pollutant with a cancer toxicity factor in Providence
   County, while formaldehyde had the highest cancer toxicity-weighted emissions.
   Toluene was the highest emitted pollutant with a noncancer toxicity factor, while
   acrolein had the highest noncancer toxicity-weighted emissions for Providence
   County.


South Carolina.

•  The South Carolina monitoring site (CHSC) is located near Chesterfield and is a
   NATTS site.

•  CHSC sampled for hexavalent chromium and PAHs.

•  Naphthalene was the only pollutant to fail screens  for CHSC. Naphthalene failed
   three screens out of 60 measured detections.

•  The annual average concentration of naphthalene was significantly higher than the
   annual average concentrations of the other two NATTS MQO Core Analytes.
   Compared to other NMP sites sampling PAHs and hexavalent chromium, CHSC had
   some of the lowest annual average concentrations.

                                  30-14

-------
•  Hexavalent chromium sampling has been conducted at CHSC for at least
   5 consecutive years; thus, a trends analysis was conducted. Hexavalent chromium
   concentrations increased significantly from 2009 to 2010 then held steady for 2011.

•  The cancer risk approximations for the pollutants of interest for CHSC were less
   than 1 in-a-million; all noncancer hazard approximations for CHSC were less than
   1.0.

•  Benzene was the highest emitted pollutant with a cancer toxicity factor in
   Chesterfield County and had the highest cancer toxicity-weighted emissions. Toluene
   was the highest emitted pollutant with a noncancer toxicity factor, while acrolein had
   the highest noncancer toxicity-weighted emissions.


South Dakota.

•  The UATMP sites in South Dakota are located in Sioux Falls (SSSD) and Union
   County (UCSD).

•  Both  South Dakota sites sampled for VOCs, SNMOCs, and carbonyl compounds.

•  Fourteen pollutants failed screens for SSSD, of which five are NATTS MQO Core
   Analytes. Thirteen pollutants failed screens for UCSD, of which six are NATTS
   MQO Core Analytes.

•  Formaldehyde and acetaldehyde had the highest annual average concentrations for
   both SSSD and UCSD and are the only two pollutants with annual averages greater
   than 1.0 |ig/m3 for these sites. UCSD had the second highest concentration of
   acrylonitrile among NMP sites sampling VOCs.

•  Formaldehyde had the highest cancer risk approximations for both sites. None of the
   pollutants of interest for either South Dakota site had a noncancer hazard
   approximation greater than 1.0.

•  Benzene was the highest emitted pollutant with a cancer toxicity factor in Minnehaha
   and Union Counties, while formaldehyde had the highest toxicity-weighted emissions
   for both counties.

•  Toluene was the highest emitted pollutant with a noncancer toxicity factor in
   Minnehaha and Union Counties, while acrolein had the highest noncancer toxicity-
   weighted emissions for both counties.


Texas.

•  There are two NATTS sites in Texas: one in Deer Park (CAMS 35) and one in
   Karnack (CAMS 85).
                                  30-15

-------
•  The CAMS 35 site sampled for PAHs and hexavalent chromium, while CAMS 85
   sampled for hexavalent chromium only.

•  Five pollutants failed screens for CAMS 35, with naphthalene contributing to nearly
   71 percent of the total failed screens. Hexavalent chromium did not fail any screens
   for CAMS 85.

•  Of the pollutants of interest for CAMS 35, naphthalene had the highest annual
   average concentration and is significantly higher than the annual averages for the
   other pollutants of interest. The annual average concentration of hexavalent
   chromium for CAMS 85 is less than half the annual average concentration for
   CAMS 35. The annual average concentration of hexavalent chromium for CAMS 85
   is an order of magnitude lower than its annual average for 2010. This may be
   attributable to the use of stainless steel filter holders used in the sampler which may
   have contaminated the samples. Changing to a Teflonฎ filter holder has resulted in a
   decrease in hexavalent chromium concentrations at CAMS 85.

•  Naphthalene had the highest cancer risk approximation among the pollutants of
   interest for CAMS 35 and was the only pollutant with a cancer risk approximation
   greater than 1 in-a-million for both sites. None of the pollutants of interest for either
   Texas site had a noncancer hazard approximation greater than 1.0.

•  Benzene was the highest emitted pollutant with a cancer toxicity factor in Harris
   County, while formaldehyde had the highest cancer toxicity-weighted emissions.
   Formaldehyde was the highest emitted pollutant with a cancer toxicity factor in
   Harrison County, while hexavalent chromium had the highest cancer toxicity-
   weighted emissions.

•  Toluene was the highest emitted pollutant with a noncancer toxicity factor in both
   counties, while  acrolein had the highest noncancer toxicity-weighted emissions.


Utah.

•  The NATTS site in Utah is located in Bountiful, north of Salt Lake City (BTUT).

•  BTUT sampled for VOCs, carbonyl compounds, SNMOCs, PAHs, metals (PMio),
   and hexavalent  chromium.

•  Twenty-one pollutants failed screens for BTUT, of which 12 are NATTS MQO Core
   Analytes.

•  Of the pollutants of interest, dichloromethane had the highest annual average
   concentration for BTUT, followed by formaldehyde, acetaldehyde, and benzene.
   BTUT had the highest annual average concentration of formaldehyde among NMP
   sites sampling carbonyl compounds.
                                  30-16

-------
•  Sampling for carbonyl compounds, VOCs, SNMOCs, metals (PMio), and hexavalent
   chromium have been conducted at BTUT for at least 5 consecutive years; thus, a
   trends analysis was conducted for select pollutants. Concentrations of formaldehyde
   have been increasing in recent years, while concentrations of benzene have decreased.
   Concentrations of lead have an overall decreasing trend since the onset of sampling.

•  The pollutant with the highest cancer risk approximation for BTUT is formaldehyde.
   None of the pollutants of interest had noncancer hazard approximations greater than
   1.0.

•  Benzene was the highest emitted pollutant with a cancer toxicity factor in Davis
   County and had the highest cancer toxicity-weighted emissions. Toluene was the
   highest emitted pollutant with a noncancer toxicity factor, while acrolein had the
   highest noncancer toxicity-weighted emissions for Davis County.


Vermont.

•  Two Vermont monitoring sites are located in or near Burlington (BURVT and
   UNVT); a third monitoring site is located in Rutland (RUVT). UNVT is a NATTS
   monitoring site.

•  UNVT sampled for VOCs, hexavalent chromium, PAHs, and metals (PMio). BURVT
   and RUVT sampled for VOCs only.

•  Eleven pollutants failed screens for BURVT and seven failed screens for RUVT.
   Thirteen pollutants failed screens for UNVT.

•  Benzene had the highest annual average concentration for BURVT and RUVT, while
   carbon tetrachloride had the highest annual average concentration for UNVT. Annual
   average concentrations of the pollutants of interest for UNVT were among the lowest
   compared to NMP sites sampling the same pollutants.

•  UNVT has sampled  hexavalent chromium for at least 5 consecutive years; thus, a
   trends analysis was conducted. The number of non-detects has decreased in recent
   years, resulting in an increase in the average concentration of hexavalent chromium
   for UNVT.

•  Benzene and carbon tetrachloride have the highest cancer risk approximations for the
   Vermont monitoring sites. None of the noncancer hazard approximations were greater
   than an HQ of 1.0.

•  Benzene was the highest emitted pollutant with a cancer toxicity factor in Chittenden
   and Rutland Counties and also had the highest cancer toxicity-weighted emissions for
   both counties. Toluene was the highest emitted pollutant with a noncancer toxicity
   factor in both counties, while acrolein had the highest noncancer toxicity-weighted
   emissions.

                                  30-17

-------
Virginia.

•  The NATTS site in Virginia is located near Richmond (RIVA).

•  RIVA sampled for PAHs and hexavalent chromium.

•  Four PAHs failed screens for RIVA, although naphthalene contributed to nearly
   95 percent of the total failed screens. Hexavalent chromium did not fail any screens.

•  Of the pollutants of interest, naphthalene had the highest annual average
   concentration.

•  Naphthalene had the highest cancer risk approximation for RIVA and is the only one
   with a cancer risk approximation greater than 1 in-a-million. None of the pollutants of
   interest for RIVA had a noncancer hazard approximation greater than 1.0.

•  Benzene was the highest emitted pollutant with a cancer toxicity factor in Henrico
   County, while  formaldehyde had the highest cancer toxicity-weighted emissions.
   Toluene was the highest emitted pollutant with a noncancer toxicity factor in Henrico
   County, while  acrolein had the highest noncancer toxicity-weighted emissions.


Washington.

•  The NATTS site in Washington is located in Seattle (SEWA).

•  SEWA sampled for VOCs, carbonyl compounds, PAHs, metals (PMi0), and
   hexavalent chromium.

•  Eighteen pollutants failed screens for SEWA, of which 12 are NATTS MQO Core
   Analytes.

•  Of the pollutants of interest for SEWA, acetaldehyde and formaldehyde had the
   highest annual average concentrations, although they are the lowest annual averages
   among NMP sites sampling carbonyl compounds. SEWA had the highest annual
   average concentration of nickel among NMP sites sampling metals (PMio).

•  Carbonyl compound, VOC, metals (PMio), and hexavalent chromium sampling has
   been conducted at BTUT for at least 5 consecutive years; thus, a trends analysis was
   conducted for select pollutants. Although most of the selected pollutants exhibit
   increases from 2010 to 2011, the increase for formaldehyde is the only one that is
   statistically significant.

•  Formaldehyde had the highest cancer risk approximation for SEWA. All of the
   noncancer hazard approximations for the pollutants of interest for SEWA sites were
   less than an HQ of 1.0.
                                  30-18

-------
       •  Benzene was the highest emitted pollutant with a cancer toxicity factor in King
          County and had the highest cancer toxicity-weighted emissions. Toluene was the
          highest emitted pollutant with a noncancer toxicity factor in King County, while
          acrolein had the highest noncancer toxicity-weighted emissions.


       Wisconsin.

       •  The Wisconsin monitoring site is located in Horicon (HOWI) and is a NATTS site.

       •  HOWI sampled for hexavalent chromium only.

       •  Hexavalent chromium was detected in greater than 65 percent of samples collected
          but did not fail any screens.

       •  The annual average concentration of hexavalent chromium was on the low side
          compared to other NMP sites sampling hexavalent chromium, ranking 15th of out
          22 sites.

       •  The cancer risk approximation for hexavalent chromium is less than 1 in-a-million
          and the noncancer hazard approximation for hexavalent chromium is less than an HQ
          of 1.0.

       •  Benzene was the highest emitted pollutant with a cancer toxicity factor in Dodge
          County and had the highest cancer toxicity-weighted emissions. Toluene was the
          highest emitted pollutant with a noncancer toxicity factor in Dodge County, while
          acrolein had the highest noncancer toxicity-weighted emissions.


30.1.3  Composite  Site-level Results Summary

       •  Twenty-eight pollutants were identified as site-specific pollutants of interest,  based
          on the risk-based screening process. Acetaldehyde and formaldehyde were the two
          most common  pollutants of interest  among the monitoring sites. All 31 sites that
          sampled carbonyl compounds had acetaldehyde and formaldehyde as pollutants of
          interest.  Benzene, 1,3-butadiene, and carbon tetrachloride were the most common
          VOC  pollutants of interest. Every site that sampled benzene (28) had it as a pollutant
          of interest. All  but one site that sampled PAHs (22) had naphthalene as a pollutant of
          interest (based  on the risk-based screening process).

       •  Concentrations from two sites, CAMS 85 and HOWI, did not fail any screens.
          However, only hexavalent chromium was sampled at these two sites. Hexavalent
          chromium failed 65 percent of screens for CAMS 85 in 2010 but did not fail any for
          2011. This difference is a result of replacing the stainless steel filter holder in the
          sampler with a Teflonฎ filter holder. A similar exchange was made at the PXSS site,
          where the number of failed screens was halved from 2010 to 2011.
                                         30-19

-------
•  Formaldehyde frequently had the highest site-specific annual average concentration
   among the site-specific pollutants of interest; formaldehyde had the highest annual
   average concentration for 19 sites. Naphthalene had the next highest at 12.

•  The toxicity factor for formaldehyde used in the preliminary risk-based screening
   process, the cancer risk approximation calculations, and the toxicity-weighting of
   emissions decreased substantially since the 2007 report. This translated to a much
   higher toxicity potential for formaldehyde, leading to more failed screens, higher
   cancer risk approximations, and relatively higher toxicity-weighted emissions values
   for the 2008-2009 through 2011 reports than in previous reports.

•  Formaldehyde, naphthalene, and benzene tended to have the highest cancer risk
   approximations on a site-specific basis. The cancer risk approximation calculated for
   BTUT from the annual average concentration of formaldehyde (58.42 in-a-million) is
   the highest of all annual average-based cancer risk approximations. Four other sites
   exhibited cancer risk approximations greater than 50 in-a-million for formaldehyde
   (S4MO, OCOK, MWOK, and TMOK). One additional cancer risk approximation was
   greater than 50 in-a-million, which was calculated from SPIL's annual average
   concentration of acrylonitrile.

•  Carbon tetrachloride often had relatively high cancer risk approximations based on
   annual average concentrations among the monitoring sites, ranging between 3 and 4
   in-a-million across the sites sampling VOCs,  but tended to have relatively low
   emissions and toxicity-weighted emissions according to the NEI.  This pollutant
   appears only once in the emissions-based tables for counties with NMP sites
   (CAMS 85).

•  None of the noncancer hazard approximations were greater than 1.0. The noncancer
   hazard approximation calculated for TOOK's annual average concentration of
   manganese (an HQ of 0.60) was the highest of all annual average-based noncancer
   hazard approximations. Formaldehyde and naphthalene along with manganese tended
   to have the highest noncancer hazard approximations on a site-specific basis.

•  Benzene, formaldehyde, and ethylbenzene often had the highest county-level
   emissions for participating counties (of those with a cancer URE). Both benzene and
   formaldehyde typically had the highest toxicity-weighted emissions, along with
   1,3-butadiene (of those with a cancer URE).

•  Toluene, xylenes, and methanol were often the highest emitted pollutants (of those
   with a noncancer RfC), although they rarely had top 10 toxicity-weighted emissions.
   Acrolein tended to have the highest toxicity-weighted emissions of pollutants with
   noncancer RfCs, although acrolein emissions were relatively low when compared to
   other pollutants.  However, due to the high toxicity of this pollutant, even low
   emissions translated into high noncancer toxicity-weighted emissions; the toxicity-
   weighted value was often several orders of magnitude higher than other pollutants.
   Acrolein is a national noncancer risk driver according to NATA. Besides acrolein,

                                   30-20

-------
          formaldehyde and 1,3-butadiene tended to have the highest toxicity-weighted
          emissions among the pollutants with noncancer RfCs.

 30.1.4 Data Quality Results Summary
       Completeness, precision, and accuracy were assessed for the 2011 monitoring effort. The
quality assessments presented in this report show that the 2011 monitoring data are of a known
and high quality, consistent with the intended data use.

       To the largest extent, ambient air concentration data sets met MQO for completeness.
Only seven out of 123 site- and method-specific data sets failed to comply with the MQO of
85 percent completeness while 58 data sets achieved 100 percent completeness.

       Method precision and analytical precision were determined for the 2011 NMP monitoring
efforts using CV calculations based on duplicate, collocated, and replicate samples. The
precision for each analytical method utilized during the 2011 NMP was within the MQO of
85 percent. The method precision presented in this report is based on analytical results greater
than or equal to the sample- and pollutant-specific MDL.

       Analytical method accuracy is ensured by using proven methods, as demonstrated by
third-party analysis of proficiency test audit samples, and following strict quality control and
quality assurance guidelines.

30.2   Conclusions
       Conclusions resulting from the data analyses of the data generated from the 2011 NMP
monitoring efforts are presented below.

       •  There are a large number of concentrations that are greater than their respective risk
          screening values, particularly for many of the NATTS MQO Core Analytes.
          However, there were no instances where the preprocessed daily measurements or
          time-period average concentrations were greater than the ATSDR MRL noncancer
          health risk benchmarks.
                                         30-21

-------
•  Where annual averages could be calculated and for those pollutants with available
   cancer UREs, none of the cancer risk approximations were greater than 100 in-a-
   million; 33 were greater than 10 in-a-million (24 for formaldehyde, six for benzene,
   and three for acrylonitrile); and less than half were greater than 1.0 in-a-million.

•  Where annual averages could be calculated and for those pollutants with available
   noncancer RfCs, none of the noncancer hazard approximations were greater than 1.0.

•  When comparing the highest emitted pollutants for a specific county to the pollutants
   with the highest toxicity-weighted emissions, the listed pollutants were more similar
   for the pollutants with cancer UREs than for pollutants with noncancer RfCs. This
   indicates that pollutants with cancer UREs that are emitted in higher quantities are
   often more toxic than pollutants emitted in lower quantities; conversely, the highest
   emitted pollutants with noncancer RfCs are not necessarily the most toxic. For
   example, toluene was the noncancer pollutant that was emitted in the
   highest quantities for many NMP counties, but was not one of the pollutants
   with highest toxicity-weighted emissions for any listed county. Conversely, while
   acrolein had the highest noncancer toxicity-weighted emissions for every NMP
   county, it was among the highest emitted pollutants for only one county.

•  The number of states and sites participating in the  NMP changes from year to year.
   Yet, many of the data analyses utilized in this report require data from year-round (or
   nearly year-round) sampling. Of the 51 sites whose data are included in the 2011
   report, only two sites sampled for an abbreviated duration (due to site initialization
   and/or site closure/relocation). Of the 123 site-method combinations, only six site-
   method combinations did not cover the entire year. As a result, the number of time-
   period averages and subsequent risk-based analyses that could not be calculated
   decreased significantly for 2011 compared to 2010 (and 2010 was improved from the
   2008-2009 monitoring effort). Fewer data gaps allow for more complete results and
   inter-site comparisons.

•  Of the 51 monitoring sites participating in the 2011 NMP, only two sampled for all
   six available pollutant groups under the national program (BTUT and NBIL). Another
   four sites sampled all five pollutant groups required for NATTS sites. The wide range
   of pollutant groups sampled among the sites, which is often a result of different
   purposes behind the monitoring at the sites, makes it difficult to draw definitive
   conclusions regarding air toxics in ambient air in a global manner.

•  This report strives to represent the best laboratory  practices and utilize the best data
   analysis techniques available. Examples for 2011 include the improvement of MDLs
   and the incorporation of updated values for various toxicity factors. This can lead to
   adjusting the focus of the report to concentrate on  the air quality issues of highest
   concern. Thus, the NMP report is dynamic in nature  and  scope; yet this approach may
   prevent the direct comparison of the current report to past reports. There are two
   major differences between reports prior to 2010 and  the 2011 report. First, all
   statistical calculations include zero substitution for non-detect results (rather than just

                                   30-22

-------
          those calculations related to risk). Second, the detect criteria applied to quarterly
          averages was removed and replaced with a completeness criteria, allowing for the
          calculation of quarterly average concentrations for those pollutants detected less
          frequently than others. The only significant differences between the 2010 and 2011
          reports are in regards to the trend analysis. For the 2011 report, this analysis utilized
          yearly average concentrations rather than 3-year rolling average concentrations. In
          addition, the list of pollutants for which this analysis was performed was extended to
          include lead.


30.3   Recommendations
       Based on the conclusions from the 2011 NMP, a number of recommendations for future

ambient air monitoring efforts are presented below.

       •  Continue participation in the National Monitoring Programs.  Ongoing ambient air
          monitoring at fixed locations can provide insight into long-term trends in air quality
          and the potential for air pollution to cause adverse health effects among the general
          population. Therefore, state and local  agencies should be encouraged to either 1)
          develop and implement their own ambient air monitoring programs based on proven,
          consistent sampling and analysis methods and EPA technical and quality assurance
          guidance, or 2) consider long-term participation in the NMP.

       •  Participate in the National Monitoring Programs year-round. Many of the analyses
          presented in the 2011 report require a full year of data to be most useful and
          representative of conditions experienced at each specified location. Therefore, state
          and local agencies should be encouraged to implement year-long ambient air
          monitoring programs in addition to participating in future monitoring  efforts. As
          discussed above, there was marked improvement in this area for 2011.

       •  Monitor for additional pollutant groups based on the results of data analyses in the
          annual report. The risk-based analysis where county-level emissions are weighted
          based on toxicity identifies those pollutants whose emissions may result in adverse
          health effects in a specific area. If a site is not sampling for a pollutant or pollutant
          group identified as particularly hazardous for a given area, the agency responsible for
          that site should consider sampling for those compounds.

       •  Strive to  develop standard conventions for interpreting air monitoring data. The lack
          of consistent approaches to present and summarize ambient air monitoring data
          complicates direct comparisons between different studies. Thought should be given to
          the feasibility of establishing standard approaches for analyzing and reporting air
          monitoring data for programs with similar objectives.

       •  Continue to identify and implement improvements to the sampling and analytical
          methods. The improvements made to the analytical methods prior to the 1999-2000
          UATMP allowed for the measurement of ambient air concentrations of 11 pollutants
          that were not  measured during  previous programs. This improvement  provides

                                         30-23

-------
   sponsoring agencies and a variety of interested parties with important information
   about air quality within their area. Further research is encouraged to identify other
   method improvements that would allow for the characterization of an even wider
   range of components in air pollution and enhance the ability of the methods to
   quantify all cancer and noncancer pollutants to at least their levels of concern (risk
   screening concentrations).

•  Require consistency in sampling and analytical methods. The development of the
   NATTS program has shown that there are inconsistencies in collection and analytical
   methods that make data comparison difficult across agencies.  Requiring agencies to
   use specified and accepted measurement methods, consistent with the guidelines
   presented in the most recent version of the NATTS TAD, is integral to the
   identification of trends and measuring the effectiveness of regulation.

•  Perform case studies based on findings from the  annual report. Often, the annual
   report identifies an interesting tendency or trend, or highlights an event at a particular
   site(s). For example, chloroform concentrations have been highest at NBIL for
   multiple report years. Further examination of the data in conjunction with
   meteorological phenomena and potential emissions events or incidents, or further site
   characterization may help identify state and local agencies pinpoint issues affecting
   air quality in their area.

•  Consider more rigorous study of the effect of automobile emissions on ambient air
   quality using multiple years of data. Because many NMP sites have generated years
   of continuous data, a real opportunity exists to evaluate the importance and impact of
   automobile emissions on ambient air quality. Suggested areas of study include
   additional signature compound assessments and  parking lot characterizations.

•  Develop and/or verify HAP and VOC emissions inventories. State/local/tribal
   agencies should use the data collected from NMP sites to develop and validate
   emissions inventories, or at the very least, identify and/or verify emissions sources of
   concern. Ideally, state/local/tribal agencies would compare the ambient monitoring
   results with an emissions inventory for source  category completeness. The emissions
   inventory could then be used to develop modeled concentrations useful to compare
   against ambient monitoring data.

•  Promulgate ambient air standards for HAPs. Concentrations of several pollutants
   sampled during the 2011  program year were greater than risk  screening values
   developed by various government agencies. One way to reduce the  risk to human
   health would be to develop standards similar to the NAAQS for pollutants that
   frequently exceed published risk screening levels.

•  Incorporate/Update Risk in State Implementation Plans (SIPs). Use risk calculations
   to design State  Implementation Plans to implement policies that reduce the potential
   for human health risk. This would be easier to  enforce if ambient standards for certain
   HAPs were developed (refer to above recommendation).

                                  30-24

-------
31.0   References

ACE, 2013. U.S. Army Corps of Engineers. Huntington District, Grayson Lake website.
       http://www.lrh.usace.army.mil/Missions/Recreation/Kentuckv/GravsonLake.aspx
       Date Last Accessed: 4/30/2013.

ASTM, 2013. ASTM, International. ASTM D6209 Standard Test Method for Determination of
       Gaseous and Particulate Polycyclic Aromatic Hydrocarbons in Ambient Air (Collection
       on Sorbent-Backed Filters with Gas Chromatographic/Mass Spectrometric Analysis).
       http://www.astm.org/Standards/D6209.htmDate Last Accessed: 7/11/2013.

ATSDR, 2012a. Agency for Toxics Substances and Disease Registry. February 2012.  Minimal
       Risk Levels (MRLs). http://www.atsdr.cdc.gov/mrls/pdfs/atsdr_mrls_february_2012.pdf

ATSDR, 2012b. Agency for Toxics Substances and Disease Registry. Toxic Substance Portal
       website, Minimal Risk Levels (MRLs). http://www.atsdr.cdc.gov/mrls/index.asp Date
       Last Accessed: 11/14/2012.

AZ DOT, 2010. Arizona Department of Transportation. State Highway Traffic Log, AADT
       2010. http://mpd.azdot.gov/mpd/data/aadt.asp Date Last Accessed: 10/18/2012.

AZ DOT, 201 la. Arizona  Department of Transportation, Motor Vehicle Division. June 30, 2011.
       Currently Registered Vehicles by County Listing.
       http://www.azdot.gov/mvd/statistics/registeredVehicles.asp Date Last Accessed:
       8/21/2012.

AZ DOT, 201 Ib. Arizona Department of Transportation, Motor Vehicle Division. August 2011.
       Daily Vehicle-Miles of Travel (DVMT) by County with Population Estimates for the
       Year ended December 31, 2010. http://www.azdot.gov/mpd/data/hpms.asp Date Last
       Accessed: 3/26/2012.

Bair, 1992. Bair, Frank E.  1992. Weather of U.S. Cities, 4th ed.  Gale Research Inc. Detroit, MI.

Boubel, et al., 1994. Boubel, R.W., D.L. Fox, D.B. Turner, and A.C.  Stern. 1994. Fundamentals
       of Air Pollution, 3rd ed. San Diego, CA: Academic Press,  pp 263.

CA DMV, 2011. California Department of Motor Vehicles. 2011. Estimated Vehicles Registered
       by County for the Period of January 1 through December 31,  2011.
       http://www.dmv.ca.gov/about/profile/est fees_pd by county.pdf Date Last Accessed:
       8/21/2012.

CA DOT, 2011. State of California Department of Transportation. 2011 All Traffic Volumes on
       California State Highway System. Sacramento, CA.
       http://www.dot.ca.gov/hq/traffops/saferesr/trafdata/index.htm Date Last Accessed:
       10/18/2012.
                                         31-1

-------
CA DOT, 2012. State of California Department of Transportation. October 2012. 2011
       California Public Road Data, http://www.dot.ca.gov/hq/tsip/hpms/index.php Date Last
       Accessed: 11/16/2012.

CCRPC, 2007. Chittenden County Regional Planning Commission. Transportation Data-Annual
       Average Daily Traffic website.  2007 Data, http://www.ccrpcvt.org/data/traffic.php
       Date Last Accessed: 10/23/2012.

CCRPC, 2011. Chittenden County Regional Planning Commission. Transportation Data-Annual
       Average Daily Traffic website.  2011 Data, http://www.ccrpcvt.org/data/traffic.php
       Date Last Accessed: 10/23/2012.

Census Bureau, 2010. U.S. Census Bureau.  September 2010. Metropolitan and Micropolitan
       Statistical Areas and Components, December 2009, With Codes.
       http://www.census.gov/population/metro/files/lists/2009/Listl.txtDate Last Accessed:
       11/5/2012.

Census Bureau, 2011. U.S. Census Bureau. Intercensal Estimates of the Resident Population for
       Counties: April 1, 2000 to July 1, 2010. Released September 2011.
       http://www.census.gov/popest/data/intercensal/county/CO-ESTOOINT-01.html Date Last
       Accessed: 3/26/2012.

Census Bureau, 2012a. U.S. Census Bureau. Metropolitan and Micropolitan Statistical Areas
       website, http://www.census.gov/population/metro/ Date Last Accessed: 1/9/2012.

Census Bureau, 2012b. U.S. Census Bureau. Annual Estimates of the Resident Population for
       Counties: April 1, 2010 to July 1, 2011. Released April 2012.
       http://www.census.gov/popest/data/counties/totals/2011/CO-EST2011-01.html Date Last
       Accessed: 8/17/2012.

CO DOR, 2011. Colorado Department of Revenue. December 21,2011. Colorado Department
       of Revenue 2011 Annual Report. Denver, CO.
       http://www.colorado.gov/cs/Satellite?c=Page&cid=1213 86797503 5&pagename=Revenu
       e-Main%2FXRMLavout Date Last Accessed: 8/21/2012.

CO DOT, 201 la. Colorado Department of Transportation. Online Transportation Information
       System (OTIS), Traffic Data Explorer. 2011 data, http://dtdapps.coloradodot.info/otis
       Date Last Accessed: 10/18/2012.

CO DOT, 201 Ib. Colorado Department of Transportation.  Online Transportation Information
       System (OTIS). 2011 State Highway Statistics - Daily Vehicle Miles of Travel (DVMT)
       for All Vehicles by County. http://dtdapps.coloradodot.info/otisDate Last Accessed:
       10/30/2012.

DC DOT, 2011. District Department of Transportation. February 2011.  2009 Traffic Volumes.
       Washington, D.C.
       http://ddot.dc. gov/DC/DDOT/About+DDOT/Maps/Traffic+Volume+Maps Date Last
       Accessed: 10/18/2012.
                                         31-2

-------
DC DOT, 2012. Email from Nana Bailey-Thomas, District Department of Transportation, FOIA
       Officer. September 11, 2012. 2011 Data, dated 7/9/12.

Draxler, R.R. and Rolph, G.D., 1997. Draxler, R.R. and Rolph, G.D. 1997. Description of the
       HYSPLIT_4 modeling system. NOAA Air Resources Laboratory. NOAA Tech. Memo.
       ERL ARL-224.  Silver Spring, MD. 24 pp. http://www.arl.noaa.gov/HYSPLIT_pubs.php

Draxler, R.R. and Rolph, G.D., 1998. Draxler, R.R. and Rolph, G.D. 1998. An overview of the
       HYSPLIT_4 modeling system of trajectories, dispersion, and deposition. Aust. Meteor.
       Mag. 47, 295-308. http://www.arl.noaa.gov/HYSPLIT  pubs.php

Draxler, R.R., 1999. Draxler, R.R.  1999. HYSPLIT4 user's guide. NOAA Air Resources
       Laboratory. NOAA Tech. Memo. ERL ARL-230.  Silver Spring, MD.
       http://www.arl.noaa.gov/HYSPLIT_pubs.php

Draxler, R.R., et al, 2009. Draxler, R.R., et al. January 2009. HYSPLIT 4 User's Guide, Version
       4.9. NOAA Air Resources Laboratory. Silver Spring, MD.
       http://www.arl.noaa.gov/documents/reports/hysplit user guide.pdf

EPA, 1998. U.S. EPA.  September 1998. Technical Assistance Document for Sampling and
       Analysis of Ozone Precursors. EPA/600-R-98/161. Research Triangle Park, NC.
       http://www.epa.gov/ttnamtil/files/ambient/pams/newtad.pdfDate Last Accessed:
       11/5/2012.

EPA, 1999a. U.S. EPA. January, 1999. Compendium Method TO-15: Determination of Volatile
       Organic Compounds (VOC) in Air Collected in Specially-Prepared Canisters and
       Analyzed by Gas Chromatography/Mass Spectrometry (GC/MS). EPA/625/R-96/010b.
       Cincinnati, OH. http://www.epa.gov/ttn/amtic/airtox.html Date Last Accessed:
       11/5/2012.

EPA, 1999b. U.S. EPA. January, 1999. Compendium Method TO-11A: Determination of
       Formaldehyde in Ambient Air Using Adsorbent Cartridge Followed by High
       Performance Liquid Chromatography (HPLC). EPA/625/R-96/010b. Cincinnati, OH.
       http://www.epa.gov/ttn/amtic/airtox.html Date Last Accessed: 11/5/2012.

EPA, 1999c. U.S. EPA. January 1999. Compendium Method TO-13A: Determination of
       Polycyclic Aromatic Hydrocarbons (PAHs) in Ambient Air Using Gas
       Chromatography/Mass Spectrometry (GC/MS). EPA/625/R-96/010b. Cincinnati, OH.
       http://www.epa.gov/ttn/amtic/airtox.html Date Last Accessed: 11/5/2012.

EPA, 1999d. U.S. EPA. June 1999. Compendium Method IO-3.5: Determination of Metals in
       Ambient Particulate Matter Using Inductively Coupled Plasma/Mass Spectrometry
       (ICP/MS). EPA/625/R-96/010a. Cincinnati, OH.
       http://www.epa.gov/ttn/amtic/files/ambient/inorganic/mthd-3-5.pdf Date Last Accessed:
       11/5/2012.
                                        31-3

-------
EPA, 2006. U.S. EPA. December 2006. Standard Operating Procedure for the Determination of
       Hexavalent Chromium in Ambient Air Analyzed by Ion Chromatography (1C).
       EPA 68-D-00-264. Research Triangle Park, NC.
       http://www.epa.gov/ttn/amtic/airtox.html Date Last Accessed: 11/5/2012.

EPA, 2007. U.S. EPA. May 2007.  1990-2002 NEI HAP Trends: Success of CAA Air Toxic
       Programs in Reducing HAP Emissions and Risk. Paper presented by Anne Pope at the
       16th Annual International Emission Inventory Conference. Raleigh, NC.
       http://www.epa.gOv/ttn/chief/conference/eil6/session6/a.pope.pdfDate Last Accessed:
       12/10/2012.

EPA, 2009a. U.S. EPA.  April 2009. UATMP, NATTS, CSATAM, PAMS, andNMOC Support,
       Contract No. EP-D-09-48.  Pages  1-2 through 1-3.

EPA, 2009b. U.S. EPA.  April 1, 2009. Technical Assistance Document for the  National Air
       Toxics Trends Stations Program, Revision 2. Research Triangle Park, NC.
       http://www.epa.gov/ttn/amtic/airtox.html Date Last Accessed: 10/19/2012.

EPA, 2009c. U.S. EPA.  September 10, 2009.  Schools Air Toxics Monitoring Activity (2009)
       Uses of Health Effects Information in Evaluating Sample Results.
       http://www.epa.gov/schoolair/pdfs/UsesOfHealthEffectsInfoinEvalSampleResults.pdf

EPA, 2010a. U.S. EPA.  October 2010. A Preliminary Risk-based Screening Approach for Air
       Toxics Monitoring Data  Sets, version 2. EPA-904-B-06-001.  Atlanta, GA.
       http://www.epa.gov/region4/air/airtoxic/Screening 111610 KMEL.pdf Date Last
       Accessed: 11/14/2012.

EPA, 2010b. U.S. EPA.  April 2010.  Status of Acrolein Monitoring at Schools website.
       http://www.epa.gov/schoolair/acrolein.html Date Last Accessed: 11/14/2012.

EPA, 2011 a. U.S. EPA.  June 2011. Technical Report for School: Assessing Outdoor Air Near
       Schools: Riggins School, North Birmingham Elementary School, and Lewis Elementary
       School (Birmingham, Alabama),  http://www.epa.gov/schoolair/NorthBirmi.html
       Date Last Accessed: 11/6/2012.

EPA, 201 Ib. U.S EPA. March 2011. The 2005 National-Scale Air Toxics Assessment (NATA)
       glosssary website, http://www.epa.gov/ttn/atw/natamain/glossl .html Date Last Accessed:
       12/10/2012.

EPA, 201 Ic. U.S. EPA.  January 2011. An Overview of Methods for EPAs National-Scale Air
       Toxics Assessment, Appendix I. http://www.epa.gov/ttn/atw/nata2005/methods.html
       Date Last Accessed: 12/10/2012.

EPA, 2012a. U.S. EPA.  Air Toxics - National Air Toxics Trends Stations website.
       http://www.epa.gov/ttn/amtic/natts.html Date Last Accessed:  10/19/2012.
                                         31-4

-------
EPA, 2012b. U.S. EPA. April 2012. Air Toxics-National Air Toxics Trends Stations website,
       Listing of NATTS sites, http://www.epa.gov/ttn/amtic/natts.html Date Last Accessed:
       10/19/2012.

EPA, 2012c. U.S. EPA. Technology Transfer Network, Air Quality System (AQS) website.
       http://www.epa.gov/ttn/airs/airsaqs/ Date Last Accessed:  10/19/2012.

EPA, 2012d. U.S. EPA. April 10, 2012. 2008 National Emissions Inventory, Version 2.
       http://www.epa.gov/ttn/chief/net/2008inventory.html Date Last Accessed: 11/5/2012.

EPA, 2012e. U.S. EPA. 1986. 40 CFR Part 136, Appendix B. Definition and Procedure for the
       Determination of the Method Detection Limit, Revision 1.11. http://www.ecfr.gov/cgi-
       bin/text-idx?c=ecfr&tpl=/ecfrbrowse/Title40/40cfrl36 main O2.tpl Date Last Accessed:
       11/5/2012.

EPA, 2012f. U.S. EPA. September 2012. National Air Toxics Trends Stations (NATTS)
       Network Assessment, Revised Draft.  Prepared by Eastern Research Group, Inc.

EPA, 2012g. U.S. EPA. Technology Transfer Network, About Air Toxics website.
       http://www.epa.gov/ttn/atw/allabout.html Date Last Accessed: 11/14/2012.

EPA, 2012h. Email from Ted Palma, U.S.  EPA, OAQPS. July 2, 2012. Dose Response Library
       spreadsheet.

EPA, 20121. U.S. EPA. Sources of Pollutants in the Ambient Air - Mobile Sources website.
       http://www.epa.gov/apti/course422/ap3a.html Date Last Accessed: 12/4/2012.

EPA, 2012J. U.S. EPA. Greenhouse Gas Emissions website.
       http://epa.gov/climatechange/ghgemissions/gases.html Date Last Accessed: 12/10/2012.

ERG, 2011. Eastern Research Group, Inc.  2011. Support for the  EPA National Monitoring
       Programs (UATMP, NATTS, CSATAM, PAMS, and NMOC Support), Quality
       Assurance Project Plan, Category 1.  Contract No. EP-D-09-048. Morrisville, NC.

FAC, 2007. Report of the Federal Advisory Committee on Detection and Quantitation
       Approaches and Uses In Clean Water Act Programs, Appendix D: DQ FAC Single
       Laboratory Procedure v2.4. December 28, 2007.  Submitted to the U.S. EPA.
       http://water.epa.gov/scitech/methods/cwa/det/upload/fmal-report-200712.pdfDate Last
       Accessed: 11/5/2012.

FCC, 2013. Florida Climate Center, Office of the State Climatologist. Florida State University.
       Special Topics website, Climate of Florida, http://climatecenter.fsu.edu/topics/specials
       Date Last Accessed: 4/11/2013.

FHWA, 2011. Federal Highway Administration. December 2011. Highway Statistics 2010.
       State Motor Vehicle Registrations-2010 (MV-1).
       http://www.fhwa.dot.gov/policyinformation/statistics/2010/ Date Last Accessed:
       10/18/2012.
                                         31-5

-------
FL DHSMV, 2011. Florida Department of Highway Safety and Motor Vehicles. Registered
       Vehicles by County as of May 1, 2011.
       http://www.flhsmv.gov/html/reports and  statistics/CVR/10-1 l/CVR-05-2011 .pdf
       Date Last Accessed: 8/21/2012.

FL DOT, 2007. Florida Department of Transportation. October 2007. Florida Traffic Monitoring
       Handbook, Chapter 4. http://www.dot.state.fl.us/planning/statistics/tmh/Date Last
       Accessed: 12/4/2012.

FL DOT, 201 la. Florida Department of Transportation. Florida Traffic Online Mapping
       Application, 2011 AADTs. http://www.dot.state.fl.us/planning/statistics/trafficdata/ Date
       Last Accessed: 10/18/2012.

FL DOT, 201 Ib. Florida Department of Transportation. Public Road Mileage and Miles
       Traveled, 2011. http://www.dot.state.fl.us/planning/statistics/mileage-rpts/public.shtm
       Date Last Accessed: 10/30/2012.

GA DOR, 2011. Georgia Department of Revenue, Motor Vehicle Division. Statistics-Vehicle
       Registration website.  Registrations as of November 30, 2011.
       http://motor.etax.dor.ga.gov/stats/renewalsstats.aspx Date Last Accessed: 12/5/2011.

GA DOT, 201 la. Georgia Department of Transportation. Georgia's State Traffic and Report
       Statistics (STARS) mapping application. 2011 Published Data.
       http://www.dot.state.ga.us/statistics/stars/Pages/default.aspx Date Last Accessed:
       10/18/2012.

GA DOT, 201 Ib. Georgia Department of Transportation. Mileage by Route and Road System
       Report 445 for 2011.
       http://www.dot.state.ga.us/statistics/RoadData/Pages/400Series.aspx Date Last Accessed:
       10/30/2012.

Gary, 2013. Gary Chamber of Commerce.  About Gary - Geography and Climate website.
       http://www.garychamber.com/ Date Last Accessed: 4/29/2013.

GCPH, 2010. Garfield County Public Heath Department. June 2010. Garfield County 2010 Air
       Quality Monitoring Summary. Rifle, CO. http://www.garfieldcountyaq.net/Date Last
       Accessed: 4/2/2013.

GCRBD, 2002. Garfield County Road and Bridge Department. 2002 Transportation Study,
       County Road System, Average Daily Traffic, Garfield County, Colorado.
       http://www.garfield-county.com/road-bridge/countv-roads.aspxDate Last Accessed:
       10/18/2012.

GSCO, 1998. Georgia State Climate Office, University of Georgia.  1998. Climatology of the
       Georgia Piedmont. Athens, GA.
       http://www.rivercenter.uga.edu/education/summit/general/climate/piedmont.pdfDate
       Last Accessed: 4/16/2013.
                                         31-6

-------
HCPID, 2012. Harris County Public Infrastructure Department, Architecture and Engineering
       Department, Traffic and Transportation Group. February 2012.  Traffic Volumes Report,
       Update 2011. Data from December 2004. Houston, TX.
       http://www.eng.hctx.net/traffic/traffic.htm Date Last Accessed: 10/22/2012.

Henrico County, 2012. Email from Deborah Oxenreider, County of Henrico, Revenue Division.
       October 18, 2012. 2011 Data.

IL DOT, 2010. Illinois Department of Transportation. Getting Around Illinois website. Average
       Annual Daily Traffic map viewer. 2010 data.
       http://www.gettingaroundillinois.com/gai.htm?mt=aadt Date Last Accessed: 10/18/2012.

IL DOT, 201 la. Illinois Department of Transportation. Getting Around Illinois website. Average
       Annual Daily Traffic map viewer. 2011 data.
       http://www.gettingaroundillinois.com/gai.htm?mt=aadt Date Last Accessed: 10/18/2012.

IL DOT, 201 Ib. Illinois Department of Transportation. Illinois Travel Statistics 2011.
       http://www.dot.il.gov/adttravelstats.html Date Last Accessed:  10/30/2012.

IL SOS, 2011. Illinois Secretary of State, Vehicle Services Department. 2011 Registration
       Counts by  County.
       http://www.cyberdriveillinois.com/departments/vehicles/title_and_registration/home.html
       Date Last Accessed: 8/21/2012.

IN BMV, 2012. Email from Julie Fletcher, Indiana Bureau of Motor Vehicles Communications
       Department. October 19, 2012.

IN DOT, 2010. Indiana Department of Transportation. Average Daily Traffic and Commercial
       Vehicles Interactive map. 2010 Data,  http://www.in.gov/indot/2350.htm Date Last
       Accessed:  10/19/2012.

IN DOT, 2012. Indiana Department of Transportation. Mileage and DVMT by Year, County,
       and System. Dated October 9, 2012. http://www.in.gov/indot/2469.htmDate Last
       Accessed:  10/30/2012.

IPCC, 2012. Intergovernmental Panel on Climate Change. 2012. Errata for the Working Group I
       Fourth Assesment Report. May 7, 2012.
       http://www.ipcc.ch/publications_and_data/publications_ipcc_fourth_assessment_report_
       wgl report the_physical science basis.htm Date Last Accessed: 7/10/2013.

ISCO, 2002. Indiana State Climate Office, Purdue University. About  Indiana Climate website.
       December 2002. http://climate.agry.purdue.edu/climate/narrative.asp Date Last Accessed:
       4/29/2013.

KY, 2013. Kentucky State Parks.  Grayson Lake History website.
       http://parks.ky.gov/parks/recreationparks/grayson-lake/history.aspx Date Last Accessed:
       4/30/2012.
                                          31-7

-------
KYTC, 2009. Kentucky Transportation Cabinet. Traffic Counts mapping application. 2009 Data.
      http://transportation.ky.gov/Planning/Pages/count-maps.aspxDate Last Accessed:
       10/19/2012.

KYTC, 2012a. Email from Ainsley Snyder, Kentucky Transportation Cabinet, Division of Motor
      Vehicle Licensing. October 26, 2012.

KYTC, 2012b. Kentucky Transportation Cabinet. July 15, 2012. 2011 DVMT (in thousands)
      and Mileage by County and Functional Class.
      http://transportation.ky.gov/Planning/Pages/Roadwav-Information-and-Data.aspx
      Date Last Accessed: 10/30/2012.

LADCO, 2003. Lake Michigan Air Directors Consortium. December 2003. Phase II Air Toxics
      Monitoring Data: Analyses and Network Design Recommendations. Final technical
      report prepared by Battelle Memorial Institute and Sonoma Technology, Inc.
      http://www.ladco.org/reports/toxics/battelle_2/Date Last Accessed: 4/22/2013.

Lakes, 2011. Lakes Environmental. June 2011.  WRPLOT View 7.0.0.
      http://www.weblakes.com/products/wrplot/index.html

MA DOT, 2007. Massachusetts Department of Transportation, Highway Division. Traffic
      Counts for Boston website. 2007 Data.
      http://www.mhd.state.ma.us/traffic.asp?f=l&C=BOSTONDate Last Accessed:
       10/19/2012.

MA DOT, 2013. Email from Bob Frey, Massachusetts Department of Transportation, Office of
      Transportation Planning. March 7, 2013.

MA RMV, 2012. Fax from Michael Hooton, Massachusetts Department of Transportation,
      Registry of Motor Vehicles. October 23, 2012.

MCC, 2013. Missouri Climate Center, University of Missouri. Climate of Missouri website.
      http://climate.missouri.edu/climate.php Date Last  Accessed: 8/15/2013.

MDS, 2011. Email from Paul T. Kelly, Michigan Department of State. December 7, 2011.
      Michigan Department of State Non-Expired Total Registrations on File as of 11/12/2011.

MI DOT, 2011. Michigan Department of Transportation.  Traffic Monitoring Information System
      (TMIS) website. Annual Average Daily Traffic Report (2011).
      http://mdotnetpublic.state.mi.us/tmispublic/Search.aspx Date Last Accessed: 10/19/2012.

MI DOT, 2012. Email from Don Howe, Michigan Department of Transportation, Bureau of
      Transporation Planning. October 31, 2012. 2011 AVMT Data.

MO DOR, 2012. Missouri Department of Revenue. Total Motor Vehicles by Kind within County
      (DM45080-01), 2011. Run Date 1/6/12. http://dor.mo.gov/publicreports/index.php
                                         31-8

-------
MO DOT, 2011. Missouri Department of Transportation, Transportation Planning. 2011
       St. Louis District Traffic Volume and Commercial Vehicle Count Map.
       http://www.modot.org/safety/trafficvolumemaps.htm Date Last Accessed: 10/19/2012.

MO DOT, 2013. Email from Tina Schmidtz, Missouri Department of Transportation.
       May 17, 2013. 2011 VMT Data for St. Louis City and County.

MSU, 2013. Michigan State Climatologisf s Office, Michigan State University. Climate of
       Dearborn Narrative. http://climate.geo.msu.edu/stations/2015/Date Last Accessed:
       5/7/2013.

NCDC, 2010. National Climatic Data Center. Subscription to Quality Controlled Local
       Climatological Data. 2010 Data. https://nes.ncdc.noaa.gov/subscriptions.html Date Last
       Accessed: 3/19/2012.

NCDC, 2011. National Climatic Data Center. Subscription to Quality Controlled Local
       Climatological Data. 2011 Data. https://nes.ncdc.noaa.gov/subscriptions.html Date Last
       Accessed: 10/19/2012.

NCDC, 2013. National Climatic Data Center.  U.S. Climate Normals website, Climate of the
       States, Climatography of the United States No. 60. http://humcane.ncdc.noaa. gov/cgi-
       bin/climatenormals/climatenormals.pl?directive=prod_select2&prodtype=CLIM60&subr
       num Date Last Accessed: 5/28/2013.

NJ DOT, 2010a. New Jersey Department of Transportation. Roadway Information and Traffic
       Counts mapping application. 2010 Data.
       http://www.state.nj .us/transportation/refdata/roadway/traffic_counts/ Date Last Accessed:
       10/19/2012.

NJ DOT, 2010b. New Jersey Department of Transportation. New Jersey's Roadway Mileage and
       Daily VMT by Functional Classification Distributed by County. 2010 Data.
       http://www.state.nj.us/transportation/refdata/roadway/vmt.shtmDate Last Accessed:
       7/2/2012.

NOAA, 1999. NOAA, National Weather Service.  April 1999.  Miguel Miller. NOAA Technical
       Memorandum,  NWS-WR259, Climate of San Jose.
       http://www.wrh.noaa.gov/mtr/sfd_sjc_climate/sjc/SJC_CLIMATE3.php Date Last
       Accessed: 3/27/2013.

NOAA, 2012. NOAA, National Weather Service, Buffalo, NY. Rochester Climate Narrative.
       http://www.nws.noaa.gov/climate/local data.php?wfo=buf Date Last Accessed:
       8/27/2012.

NOAA, 2013a. NOAA, National Weather Service.  NWS glossary website.
       http://weather.gov/glossary/ Date Last Accessed: 5/3/2013.
                                         31-9

-------
NOAA, 2013b. NOAA, Storm Prediction Center. Roger Edwards. The Online Tornado FAQ,
       Frequently Asked Questions about Tornadoes website.
       http://www.spc.noaa.gov/faq/tornado/ Date Last Accessed: 5/28/2013.

NOAA, 2013c. NOAA, Climate Prediction Center. Climate Divisions with Counties website.
       http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/regional_monitoring/CLIM
        DIVS/states  counties climate-divisions.shtml Date Last Accessed: 6/13/2013.

NYS DMV, 2011. New York State Department of Motor Vehicles. Statistics-Vehicle
       Registrations in Force-2011. http://www.dmv.ny.gov/stats.htmDate Last Accessed:
       8/23/2012.

NYS DOT, 2010. New York State Department of Transportation. Traffic Data Viewer mapping
       application. 2010 Data. https://www.dot.ny.gov/tdv Date Last Accessed:  10/19/2012.

NYS DOT, 2012. Email from Patrick Lentlie, New York State Department of Transportation,
       Office of Environment. December 5, 2012. 2009 DVMT Data.

OK DOT, 2011. Oklahoma Department of Transportation, Planning and Research Division. 2011
       Annual Average Daily Traffic Oklahoma Highway  System maps.
       http://www.okladot.state.ok.us/maps/aadt/index.htm Date Last Accessed: 10/19/2012.

OK DOT, 2012. Email from Ron Maxwell, Oklahoma Department of Transportation.
       August 20, 2012. 2011 DVMT Data spreadsheet.

OKTC, 2011. Oklahoma Tax Commission, Motor Vehicle Division.  Annual Vehicle
       Registration Report, Fiscal Year July 1, 2010-June 30, 2011. Oklahoma City, OK.
       http://www.tax.ok.gov/annrpts.html Date Last Accessed: 8/23/2012.

OR DOT, 2012. Oregon Department of Transportation.  Transportation Data website, Definition
       of VMT. http://www.oregon.gov/ODOT/TD/TDATA/tsm/vmtpage.shtmlDate Last
       Accessed: 12/4/2012.

Pagano, P. and Gauvreau, K, 2000. Pagano, M. and Gauvreau, K. 2000. Principles of
       Biostatistics, 2nd edition. Pacific Grove, CA: Duxbury.

Peltier, 2012. Peltier Techinical Services, Inc. Box and Whisker Plot Utility. Downloaded
       April 5, 2012. http://peltiertech.com/Utilitv/

RI DOT, 2009. Rhode Island Department of Transportation. December 2009. State Highway
       Map of Rhode Island, Traffic Flow Map 2009.
       http://www.dot.ri.gov/engineering/gis/maps.asp Date Last Accessed: 10/19/2012.

Rutgers, 2013. Office of the New Jersey State Climatologist, Rutgers University.  New Jersey
       Climate Overview website,  http://climate.rutgers.edu/stateclim_v 1/njclimoverview.html
       Date Last Accessed: 5/14/2013.
                                         31-10

-------
SC DMV, 2011. South Carolina Department of Motor Vehicles. Historical Information and
       Statistics website. S.C. Registered Vehicles (12/2011).
       http://www.scdmvonline.com/DMVNew/default.aspx?n=historical_information
       Date Last Accessed: 8/23/2012.

SC DOT, 2012a.  South Carolina Department of Transportation. Average Annual Daily Traffic
       website. Average Daily Traffic for Map Sales, Chesterfield County. July 18, 2012.
       http://www.dot.state.sc.us/getting/aadt.asp Date Last Accessed: 10/22/2012.

SC DOT, 2012b.  Email from Angela Hance, South Carolina Department of Transportation.
       October 30, 2012. 2011 DVMT Data.

SC SCO, 2013. South Carolina Department of Natural Resources, State Climate Office. South
       Carolina Climate website.
       http://www.dnr.sc.gov/climate/sco/ClimateData/cli sc climate.php Date Last Accessed:
       6/3/2013.

SD DENR,  2011. South Dakota Department of Environment and Natural Resources. South
       Dakota Ambient Air Monitoring Annual Network Plan 2011.
       http://denr.sd.gov/des/aq/monitoring/state-mo.aspx Date Last Accessed: 6/3/2013.

SD DOR, 2012. South Dakota Department of Revenue, Motor Vehicle Division. Number of
       Titles Awarded by Year by County website. 2011 Data.
       http://www.state.sd.us/drr2/motorvehicle/title/history_of_titlesl2.htm Date Last
       Accessed: 8/23/2012.

SD DOT, 2007. South Dakota Department of Transportation. Traffic website.  Union County
       Map, 2007 Data. http://www.sddot.com/transportation/highways/traffic/Default.aspx
       Date Last Accessed: 10/22/2012.

SD DOT, 2011. South Dakota Department of Transportation. Traffic website.  Sioux Falls-
       Brandon Map, 2011 Data.
       http://www.sddot.com/transportation/highways/traffic/Default.aspxDate Last Accessed:
       10/22/2012.

SD DOT, 2012. South Dakota Department of Transportation. March 23, 2012.  South Dakota
       Vehicle Miles of Travel Report by County - All Vehicles, 2011 Data.
       http://sddot.com/transportation/highways/traffic/ Date Last Accessed: 10/30/2012.

Steer, 2008. Steer Davies Gleave.  January 2008.  New Jersey Traffic  and Revenue Study, New
       Jersey Turnpike & Route 440 Asset Appraisal. Final Report prepared for State of New
       Jersey Department of Treasury by Steer Davies Gleave in association with CRA
       International and EDR Group.
       Appraisal, http://slic.njstatelib.org/slic  files/digidocs/f491/f4912008d.pdf Date Last
       Accessed: 10/19/2012.
                                         31-11

-------
TAMU, 2013. Office of the State Climatologist, Texas A&M University.  John Nielsen-
       Gammon. Climate of Texas.
       http://climatexas.tamu.edu/files/osc_pubs/climate_of_texas.pdf Date Last Accessed:
       6/5/2013.

TX DMV, 2012. Email from Diane Emrick-Dodson, Texas Department of Motor Vehicles,
       Vehicle Titles and Registration Division. August 27, 2012. Vehicle Registration and
       License Transactions and Related Collections by County Calendar Year January 1, 2011
       to December 31, 2011.

TX DOT, 2011. Texas Department of Transportation. District Traffic Maps - 2011. Atlanta Base
       Sheet, http://www.txdot.gov/travel/traffic maps/2010.htm Date Last Accessed:
       10/22/2012.

TX DOT, 2012. Texas Department of Transportation. District and County Statistics (DISCOS)
       website. Daily Vehicle Miles (9/1/10 thru 8/31/11) for the Atlanta and Houston Districts.
       http://www.txdot.gov/inside-txdot/division/fmance/discos.html Date Last Accessed:
       11/1/2012.

UT DOT, 2010. Utah Department of Transportation. Traffic Maps for 2010 website, Ogden
       Urbanized Area map. http://www.udot.utah.gov/main/f?p=100:pg:0::::V,T:,3523
       Date Last Accessed: 10/22/2012.

UT DOT, 2012. Utah Department of Transportation. September 2012. 2011 Vehicle Miles of
       Travel (VMT) by Functional Class by County.
       http://www.udot.utah.gov/main/f?p=100:pg:0::::V,T:,530 Date Last Accessed:
       10/30/2012.

UT TC, 2011. Utah Tax Commission. February 4, 2011. Utah Current Registrations 2011,
       On-road Registrations by County and Vehicle Type, http://tax.utah.gov/esu/mv-
       registrations Date Last Accessed: 8/23/2012.

VA DOT, 2011. Virginina Department of Transportation, Traffic Engineering Division.
       Average Daily Traffic Volumes with Vehicle Classification Data on Interstate, Arterial,
       and Primary Routes 2011. http://www.virginiadot.org/info/2011 traffic data.asp
       Date Last Accessed: 10/22/2012.

VA DOT, 2012. Virginia Department of Transportation. August 2012.  Daily Vehicle Miles
       Traveled (DVMT) by Physical Jurisdiction, with Towns Combined into Counties
       (VMT 1206). 2011 data.
       http://www.virginiadot.org/info/201 l_traffic_data_daily_vehicle_miles_traveled.asp
       Date Last Accessed: 10/30/2012

VT DMV, 2012. Email from Judy Eastman, Vermont Department of Motor Vehicles.
       October 19, 2012.  2012 Data.
                                         31-12

-------
Vtrans, 2010. Vermont Agency of Transportation, Highway Research.  2010 Miles and Annual
       Vehicle Miles Traveled (AVMT) by County and Functional Class.
       http://www.aot.state.vt.us/planning/documents/highresearch/publications/pub.htm
       Date Last Accessed: 7/3/2012.

Vtrans, 2011. Vermont Agency of Transportation, Policy, Planning and Intermodal Development
       Division, Traffic Research Unit.  May 2011. 2010 (Route Log) AADTs State Highways.
       http://www.aot.state.vt.us/Planning/Documents/TrafResearch/Publications/pub.htm
       Date Last Accessed: 10/23/2012.

WA DOL, 2011. Washington State Department of Licensing. Vehicle/Vessel Fee Collection and
       Vehicle Counts Program website. Motor Vehicle Registration by Class and County.
       Calendar Year 2011. http://www.dol.wa.gov/about/vehvesselreports.html Date  Last
       Accessed:  10/18/2012.

WA DOT, 2011. Washington State Department of Transportation. 2011 Annual Traffic Report.
       Includes AVMT and traffic data.
       http://www.wsdot.wa.gov/mapsdata/travel/annualtrafficreport.htm Date Last Accessed:
       10/31/2012.

WI DOT, 2008. Wisconsin Department of Transportation. 2008 City of Horicon Dodge County
       Annual Average Daily Traffic map.
       http://www.dot.wisconsin.gov/travel/counts/index.htm Date Last Accessed: 10/22/2012.

WI DOT, 2011. Wisconsin Department of Transportation, Division of Motor Vehicles. Facts and
       Figures 2011, A Reference Guide. Madison, WI.
       http://www.dot.wisconsin.gov/drivers/facts.htm Date Last Accessed: 9/7/2012.

WI DOT, 2012. Wisconsin Department of Transportation. September 2012. 2011 Vehicle Miles
       of Travel (VMT) by County, http://www.dot.wisconsin.gov/travel/counts/vmt.htm
       Date Last Accessed: 10/31/2012.

WRCC, 2013. Western Region Climate Center. Climate Narrative of the States: Climate of
       Alaska, Arizona, California, Colorado, Utah.
       http://www.wrcc.dri.edu/CLIMATEDATA.html Date Last Accessed: 3/22/2013.

Xionetic, 2011. Xionetic Technologies. ZIPFindฎ Deluxe 5.0. Data obtained 12/28/2011.
       http://xionetic.com/ZipFindDeluxe.aspx
                                         31-13

-------
United States                             Office of Air Quality Planning and Standards             Publication No. EPA-454/R-13-007a
Environmental Protection                        Air Quality Assessment Division                                       August 2013
Agency                                          Research Triangle Park, NC

-------