NATIONAL AIR TOXICS TRENDS STATIONS
QUALITY ASSURANCE ANNUAL REPORT
CALENDER YEAR 2009
FINAL
Environmental Protection Agency
Office of Air Quality, Planning and Standards
Air Quality Analysis Division
109 TW Alexander Drive
Research Triangle Park, NC 27711

-------
FORWARD
In the Spring 2011, Research Triangle Institute (RTI) prepared a technical report under Contract
No. EP-D-08-047 Work Assignment 3-04. The report describes the Quality Assurance (QA)
data collected within the calendar year 2009. The report was prepared for Dennis K. Mikel,
Work Assignment Manager within the Office of Air Quality Planning and Standards (OAQPS)
in Research Triangle Park, North Carolina. The draft report was written by Larry Michael and
Jeff Nichols of RTI and EPA submitted it to the State and Local air toxics community for review
and comment. RTI addressed the comments that were submitted and sent this final report to the
Work Assignment Manager. This is the final report.
Additional work on this report was provided by AQAD staff.
Comments and questions should be submitted to:
Dennis K. Mikel
EPA-OAQPS-AQAD
919-541-5511 or;
Mikel.dennisk@epa.gov

-------
NATIONAL AIR TOXICS TRENDS STATIONS
QUALITY ASSURANCE ANNUAL REPORT
CALENDAR YEAR 2009
FINAL REPORT
Prepared by:
RTI International
For:
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Air Quality Analysis Division
109 TW Alexander Drive
Research Triangle Park, NC 27711
Under:
U.S. EPA Contract EP-D-08-047
Work Assignment 03-04, Task 10

-------
ACKNOWLEDGEMENTS
This report was prepared by RTI International, under Work Assignment 03-04, Task 10 for EPA
contract EP-D-08-047. Substantial contributions to the determination of the laboratories
associated with specific monitoring sites and to the acquisition of proficiency testing results were
provided by Dennis Mikel and Mike Jones of U.S. EPA.

-------
TABLE OF CONTENTS
Section	Page
1.0 Introduction	1
2.0 NATTS Quality Assurance Data for Calendar Year 2009 	2
2.1	Measurement Quality Objectives	5
2.2	Completeness of NATTS Data	6
2.3	Precision of NATTS Data	11
2.3.1	Analytical Precision Results	15
2.3.2	Overall Precision Results	16
2.4	Laboratory Bias Data Based on Proficiency Testing Samples	34
2.5	Flow Audit Results from Instrument Performance Audits	38
2.6	Method Detection Limit Data	41
3.0 Summary	51
4.0 Recommendations	52
5.0 References	53
LIST OF TABLES
Number	Page
Table 1. EPA Region Numbers, NATTS Sites, Site Type, and Air Quality Systems Site
Codes	3
Table 2. The 23 Hazardous Air Pollutants and their Air Quality Systems Parameter
Codes	4
Table 3. Measurement Quality Objectives for the NATTS Program	5
Table 4. Data Sources Used to Evaluate Measurement Quality Objectives	5
Table 5. Percentage Completeness of the 2009 AQS Dataset by Site for Seven Hazardous
Air Pollutants	7
Table 6. Parameter Occurrence Codes by NATTS Site and Analyte Type	13
Table 7. Laboratories Performing Analyses for the Different Analyte Types for Each
NATTS Site in 2009	14
Table 8. Laboratory Abbreviations and Descriptions for NATTS Laboratories	15
Table 9. Analytical Precision for Replicate Analyses of 2009 NATTS Data	17
Table 10. Overall Precision for Primary and Collocated Samples from 2009	24
Table 11. Performance Testing Bias Results for VOCs in 2009 NATTS Laboratories	35
Table 12. Proficiency Testing Bias Results for Carbonyls in 2009 NATTS Laboratories	36
Table 13. Proficiency Testing Bias Results for Metals in 2009 NATTS Laboratories	36
Table 14. Proficiency Testing Program Participation for 2009	38
Table 15. Flow Audit Results from 2009 Instrument Performance Audits	39

-------
Table 16. Method Detection Limits by Site and Overall for Calendar Year 2009 (VOCs
and Carbonyls: (J,g/m3; Metals: ng/m3)	43
Table 17. Comparison of Method Detection Limits for Metals between High- and Low-
Volume Samplers in Calendar Year 2009	51
Table 18. Summary Statistics for Method Detection Limits across All Reporting NATTS
Laboratories for 2009	51
LIST OF FIGURES
Figure 1. Completeness for Benzene at NATTS Sample Collection Sites in 2009 (MQO
reference indicated at 85%)	8
Figure 2. Completeness for 1,3-Butadiene at NATTS Sample Collection Sites in 2009
(MQO reference indicated at 85%)	8
Figure 3. Completeness for Acrolein at NATTS Sample Collection Sites in 2009 (MQO
reference indicated at 85%)	9
Figure 4. Completeness for Formaldehyde at NATTS Sample Collection Sites in 2009
(MQO reference indicated at 85%)	9
Figure 5. Completeness for Naphthalene at NATTS Sample Collection Sites in 2009
(MQO reference indicated at 85%)	10
Figure 6. Completeness for Chromium (VI) at NATTS Sample Collection Sites in 2009
(MQO reference indicated at 85%)	10
Figure 7. Completeness for Arsenic at NATTS Sample Collection Sites in 2009 (MQO
reference indicated at 85%)	11
Figure 8. Analytical Precision Summary for Benzene at NATTS Sample Collection Sites
in 2009 (MQO reference indicated at 15%)	20
Figure 9. Analytical Precision Summary for 1,3-Butadiene at NATTS Sample Collection
Sites in 2009	20
Figure 10. Analytical Precision Summary for Acrolein at NATTS Sample Collection
Sites in 2009 (MQO reference indicated at 15%)	21
Figure 11. Analytical Precision Summary for Formaldehyde at NATTS Sample
Collection Sites in 2009	21
Figure 12. Analytical Precision Summary for Naphthalene at NATTS Sample Collection
Sites in 2009	22
Figure 13. Analytical Precision Summary for Chromium (VI) at NATTS Sample
Collection Sites in 2009 (MQO reference indicated at 15%)	22
Figure 14. Analytical Precision Summary for Arsenic at NATTS Sample Collection Sites
in 2009	23
Figure 15. Overall Precision Summary for Benzene at NATTS Sample Collection Sites in
200 (MQO reference indicated at 15%)	31
Figure 16. Overall Precision Summary for 1,3-Butadiene at NATTS Sample Collection
Sites in 2009 (MQO reference indicated at 15%)	31
Figure 17. Overall Precision Summary for Acrolein at NATTS Sample Collection Sites in
2009 (MQO reference indicated at 15%)	32
Figure 18. Overall Precision Summary for Formaldehyde at NATTS Sample Collection
Sites in 2009 (MQO reference indicated at 15%)	32
Figure 19. Overall Precision Summary for Naphthalene at NATTS Sample Collection
Sites in 2009 (MQO reference indicated at 15%)	33

-------
Figure 20. Overall Precision Summary for Chromium (VI) atNATTS Sample Collection
Sites in 2009 (MQO reference indicated at 15%)	33
Figure 21. Overall Precision Summary for Arsenic at NATTS Sample Collection Sites in
2009 (MQO reference indicated at 15%)	34
Figure 22. Distribution of Laboratory Bias by Analyte for Proficiency Testing Data from
2009	38
Figure 23. Summary of Instrument Performance Flow Audit Results for 2009	42
Figure 24. Distribution of Method Detection Limits for Carbonyls for 2009 NATTS Data
(dashed line indicates MQO target MDL for formaldehyde; > 1.5 x IQR are
identified as blue stars in top display)	45
Figure 25. Distribution of Method Detection Limits for Metals for 2009 NATTS Data
(dashed line indicates MQO target MDL for arsenic; > 1.5 x IQR are identified
as blue stars in top display)	46
Figure 26. Distribution of Method Detection Limits for Arsenic for 2009 NATTS Data
(dashed line indicates MQO target MDL for arsenic; > 1.5 x IQR are identified
as blue stars in top display)	47
Figure 27. Distribution of Method Detection Limits for VOCs for 2009 NATTS Data
(dashed line indicates MQO target MDL for benzene; > 1.5 x IQR are identified
as blue stars in top display)	48
Figure 28. Distribution of Method Detection Limits for PAHs for 2009 NATTS Data
(> 1.5 x IQR are identified as blue stars in top display)	49

-------
1.0 INTRODUCTION
As mandated under the Government Performance Results Act, the U.S. Environmental
Protection Agency (EPA) is focused on reducing risk of cancer and other serious health effects
associated with hazardous air pollutants (HAPs) by achieving a 75% reduction in air toxics
emissions chemicals, based on 1993 levels. The current inventory of HAPs includes 188
chemicals regulated under the Clean Air Act that have been linked to numerous adverse human
health and ecological effects, including cancer, neurological effects, reproductive effects, and
developmental effects. Current agency attention is targeting risk reduction associated with
human exposure to air toxics.
The National Air Toxics Trends Station (NATTS) network was established to create a
database of air quality data to assess progress in reducing ambient concentrations of air toxics
and concomitant exposure-associated risk. During 2009, the NATTS network consisted of
27 stations in the contiguous 48 states. To ensure the quality of the data collected under the
NATTS network, EPA has implemented a Quality System comprising two primary components:
(1) Technical Systems Audits (TSAs) and (2) Instrument Performance Audits (IPAs) for both the
network stations and the associated sample analysis laboratories. As an integral part of the
Quality System, EPA has also instituted semiannual analysis of proficiency testing (PT) samples
for volatile organic compounds (VOCs) and carbonyls and annual analysis of PT samples for
metals and polycyclic aromatic hydrocarbons (PAHs) to provide quantitative assessment of
laboratory performance and to ensure that sampling and analysis techniques are consistent with
precision, bias, and method detection limits (MDLs) specified by the NATTS Measurement
Quality Objectives (MQOs).
This report describes and summarizes the quality assurance (QA) data generated by the
NATTS program during calendar year 2009. For data retrieved from EPA's Air Quality Systems
(AQS) database, only data collected in 2009 and posted prior to May 7, 2011, are included.
Although this report contains substantive information about air concentrations of 27 different
chemicals of interest, it focuses primarily on results for four classes of toxic ambient air
constituents (VOCs, carbonyls, PAHs, and PMi0 metals) as represented by seven pollutants:
benzene, 1,3-butadiene, formaldehyde, acrolein, naphthalene, chromium (VI), andPMio arsenic.
At the request of EPA, these seven pollutants were selected as having particular interest by virtue
of associated health risk and the frequency of their occurrence at measurable concentrations.
Although no group of compounds can provide unequivocal representation of their respective
compound groups, these seven analytes were selected by EPA as reasonable representatives of
the four main categories of HAPs routinely measured in the NATTS program and thus provide
the framework for this summary report. It is presumed that if the NATTS program can meet the
Data Quality Objectives (DQOs) for these seven compounds, the additional 20 compounds of
concern will be of comparable quality by virtue of the representativeness of the physicochemical
properties and the consistency of the collection and analysis methodologies of these seven
compounds. Because monitoring for PAHs and chromium (VI) is new, or relatively new, at
many sites, QA results may be unavailable for some MQOs at some sites.
1

-------
The comprehensive information in this Quality Assurance Annual Report (QAAR) was
compiled from data acquired from numerous sources. The following general categories of
information are presented:
•	Descriptive background information on the AQS site identities, compounds of
interest, and MQOs;
•	Assessment of the completeness of the data available in the AQS database;
•	Precision estimates, independently, for analytical and overall sampling error
computed for as many of the 27 applicable compounds and for as many of the
27 NATTS sites as available for calendar year 2009;
•	Evaluation of an analytical laboratory's accuracy (or bias), based on analysis of blind
audit PT samples for many of the 27 compounds;
•	Field bias data, which are expressed as the differences between actual and measured
sampler flow readings for each of the four different sampler types associated with
VOCs, carbonyls, PAHs, and PMi0 metals, for primary and collocated samplers
(where available) at the three sites visited during the IP As conducted during calendar
year 2009; and
•	MDL data for each site and/or analytical laboratory. In contrast to the manner in
which these data were compiled in 2009, the AQS database, specifically the
ALT MDL variable, was used as the primary source of MDLs for 2009. However,
because this MDL field in AQS is not a required field, it was necessary to augment
the information with direct contacts to several NATTS state and local agencies and
affiliated laboratories to compile MDL data for the 27 compounds of interest at all
sites. This modification improved both acquisition efficiency and the accuracy of the
MDL data.
Where possible, all data analyses were performed in SAS, version 9.2. Direct compilation
of MDLs from individual laboratories was performed using Microsoft Excel.
2.0 NATTS QUALITY ASSURANCE DATA FOR CALENDAR YEAR 2009
The NATTS network included 27 sites in 2009. Table 1 presents the EPA Regions in
which the sites are located, a descriptive location of the sites (site identifier), the urban or rural
character of each site, and the unique AQS identification code [1],
Although a city and state are typically used as the site identifier, the county name is used
for the two Florida sites on either side of Tampa Bay and for Harrison County, TX. Historical
consistency has been maintained for the Grand Junction, CO, site, where two separate codes are
used, one for VOCs and carbonyls (-0018) and the other for metals (-0017). This convention is
unique to this site and is used because the organics and metals samplers are present at two
separate physical locations at the sampling site.
2

-------
Table 1. EPA Region Numbers, NATTS Sites, Site Type, and Air Quality Systems Site
Codes.
EPA Region
Site Identifier
Type
AQS Site Code
1
Boston-Roxbury, MA
Urban
25-025-0042
1
Underbill, VT
Rural
50-007-0007
1
Providence, Rl
Urban
44-007-0022
II
Bronx, NY
Urban
36-005-0110
II
Rochester, NY
Urban
36-055-1007
III
Washington, DC
Urban
11-001-0043
III
Richmond, VA
Urban
51-087-0014
IV
Chesterfield, SC
Rural
45-025-0001
IV
Decatur, GA
Urban
13-089-0002
IV
Grayson Lake, KY
Rural
21-043-0500
IV
Hillsborough County, FL
Urban
12-057-3002
IV
Pinellas County, FL
Urban
12-103-0026
V
Dearborn, Ml
Urban
26-163-0033
V
Mayvilie, Wl
Rural
55-027-0007
V
Northbrook, IL
Urban
17-031-4201
VI
Deer Park, TX
Urban
48-201-1039
VI
Harrison County, TX
Rural
48-203-0002
VII
St. Louis, MO
Urban
29-510-0085
VIII
Bountiful, UT
Urban
49-011-0004
VIII
Grand Junction, CO
Rural
08-077-0017a, -0018b
IX
Phoenix, AZ
Urban
04-013-9997
IX
San Jose, CA
Urban
06-085-0005
IX
Rubidoux, CA
Urban
06-065-8001
IX
Los Angeles, CA
Urban
06-037-1103
X
La Grande, OR
Rural
41-061-0119
X
Portland, OR
Urban
41-051-0246
X
Seattle, WA
Urban
53-033-0080
' Metals only.
b VOCs, carbonyls, PAHs, and Cr(VI) only.
The 27 specific HAPs measured in the NATTS program, presented in Table 2 along with
their unique AQS identification codes, are compounds that EPA has identified as being of
significant health concern. These include 16 VOCs, 2 carbonyls, 2 PAHs, 6 PMio metals, and
chromium (VI). Succinct abbreviations of each chemical name are provided to facilitate table
and figure creation and interpretation throughout this report.
3

-------
Table 2. The 23 Hazardous Air Pollutants" and their Air Quality Systems Parameter
Codes.
Analyte


AQS
Compound
Abbreviation3
Compound Name
Exact AQS Label
Code
Group
BENZb
benzene
Benzene
45201
voc
BUTAb
1,3-butadiene
1,3-Butadiene
43218
voc
CTET
carbon tetrachloride
Carbon Tetrachloride
43804
voc
CLFRM
chloroform
Chloroform
43803
voc
EDB
1,2-dibromoethane
Ethylene Dibromide
43843
voc
DCP
1,2-dichloropropane
1,2-Dichloropropane
43829
voc
EDC
1,2-dichloroethane
Ethylene Dichloride
43815
voc
MECL
dichloromethane
Dichloromethane
43802
voc
TCE1122
1,1,2,2-tetrachloroethane
1,1,2,2-Tetrachloroethane
43818
voc
PERC
tetrachloroethylene
Tetrachloroethylene
43817
voc
TCE
trichloroethylene
Trichloroethylene
43824
voc
VCM
vinyl chloride
Vinyl Chloride
43860
voc
cDCPEN
cis-1,3-dichloropropene
Cis-1,3-Dichloropropylene
43831
voc
tDCPEN
trans-1,3-dichloropropene
Trans-1,3-Dichloropropylene
43830
voc
ACROc,e
acrolein
Acrolein
43505d
VOCc
ACROd,e
acrolein
Acrolein
435098
VOCc
ACRY
acrylonitrile
Acrylonitrile
43704
voc
NAPHb
naphthalene
Naphthalene (Tsp) Stp

PAH
BaP
benzo[a]pyrene
Benzo[A]Pyrene (Tsp) Stp

PAH
FORMb
formaldehyde
Formaldehyde
43502
Carbonyl
ACET
acetaldehyde
Acetaldehyde
43503
Carbonyl
Asb
arsenic
Arsenic Pm10 Stp
82103
Metal
Be
beryllium
Beryllium Pm10 Stp
82105
Metal
Cd
cadmium
Cadmium Pm10 Stp
82110
Metal
Pb
lead
Lead Pm10 Stp
82128
Metal
Mn
manganese
Manganese Pm10 Stp
82132
Metal
Ni
nickel
Nickel Pm10 Stp
82136
Metal
CrVlb
chromium (VI)
Chromium (VI) Tsp Stp
12115
Metal
Asf
arsenic
Arsenic Pm10 Lc
85103
Metal
Bef
beryllium
Beryllium Pm10 Lc
85105
Metal
Cdf
cadmium
Cadmium Pm10 Lc
85110
Metal
Pbf
lead
Lead Pm10 Lc
85128
Metal
Mnf
manganese
Manganese Pm10 Lc
85132
Metal
Nif
nickel
Nickel Pm10 Lc
85136
Metal
CrVlf
chromium (VI)
Chromium (VI) Tsp Lc
14115
Metal
a Mercury has been intentionally excluded, per U.S. EPA directive.
b Results presented are representative of completeness for other chemicals in this class.
c Unverified results.
11 Verified results.
e Completeness based on verified and unverified results.
f Some sites reported results for metal analytes at local conditions (lc), instead of STP (Stp), using these parameter codes. For this report, data
reported in Stp and lc units are combined, under the assumption that the difference between the two values is negligible.
4

-------
2.1 Measurement Quality Objectives
MQOs for completeness, precision, laboratory bias, and MDLs, established for the
NATTS network to ensure data quality within the network, were unchanged from 2008 [2], The
stated DQO for the NATTS program is "to be able to detect a 15 percent difference (trend)
between two consecutive 3-year annual mean concentrations within acceptable levels of decision
error" [3], MQOs for the four compounds of primary importance to the NATTS program
(benzene, 1,3-butadiene, formaldehyde, PMio arsenic) are summarized in Table 3. MQOs for the
three additional analytes of interest (acrolein, naphthalene, and chromium [VI]) have not been
assigned by EPA.
Table 3. Measurement Quality Objectives for the NATTS Program.
Compound
Completeness
Precision
(Coefficient of
Variation)
Laboratory Bias
Method Detection
Limit (MDL)
benzene
> 85%
< 15%
< 25%
0.016 |jg/m3
1,3-butadiene
> 85%
< 15%
< 25%
0.013 |jg/m3
formaldehyde
> 85%
< 15%
< 25%
0.0074 |jg/m3a
arsenic
> 85%
< 15%
< 25%
0.217 ng/m3b
a Assumes a sampling volume of 1,000 L.
b Assumes high-volume sampling with a sampling volume of 1,627 m3 (1.13 m3/min [40 ft3/min] for 24 hours) and that one-eighth of the sampled
area of the filter is extracted for analysis.
As intended by the NATTS network, the MQOs require that
(1)	sampling occurs every 6th day;
(2)	sampling is successful 85% of the time;
(3)	precision, as measured by the coefficient of variation (CV), is within 15% based on
duplicate and collocated samples; and
(4)	laboratory (measurement) bias is less than 25%, based on laboratory PT results.
Furthermore, actual MDLs, as reported by the laboratories supporting the NATTS sites or
their sponsoring federal, state, or municipal agencies, are compared with the target MDLs as
listed in the applicable edition of the NATTS Technical Assistance Document (TAD) [4],
Data acquired to assess compliance with the above stated MQOs were derived from a
variety of sources. These sources are given in Table 4.
Table 4. Data Sources Used to Evaluate Measurement Quality Objectives.
Measurement Quality Objective
Data Source
Completeness
AOS
Analytical and Overall Precision
AOS
Bias—Laboratory
Proficiency testing results reported by Alion
Bias—Field
Audits of sampler flow rates conducted by RTI International
MDL
AOS augmented with information from the analytical laboratories
5

-------
Data retrievals from AQS for relevant samples collected in 2009 and uploaded to the
AQS database prior to May 7, 2011, were analyzed to assess completeness and to estimate
precision from results of replicate analyses and collocated and duplicate sampling. PT samples
were distributed by EPA contractor Alion Science, Inc., to participating laboratories for
determination of analytical bias. Field bias was evaluated by independent measurement of
sampler flow rates with National Institute of Standards and Technology (NIST)-traceable
flowmeters during on-site IP As. Finally, MDL data were extracted from AQS, where present,
and augmented by values obtained by direct contact with the individual laboratories.
2.2 Completeness of NATTS Data
The AQS database was queried for data records corresponding to relevant samples
collected from the 27 NATTS sites during calendar year 2009 and entered into the AQS database
prior to May 7, 2011. This posting period to AQS was extended from the original July 1, 2010
deadline for the 2009 report to allow for posting of data unintentionally excluded by a single
NATTS site. Any data that might have been contributed to AQS by participating laboratories
after May 7, 2011, are not reflected in the completeness calculations presented in Table 5 below.
Specifically, completeness of the 2009 AQS dataset was assessed for seven compounds
representative of the entire suite of 27 compounds presented previously in Table 2: benzene, 1,3-
butadiene, acrolein, formaldehyde, naphthalene, chromium (VI), and arsenic. Based on the
NATTS requirement of a l-in-6 day sample collection frequency, 60 records for the primary
parameter occurrence code (POC) would represent 100% completeness. Depending on the first
date of collection in 2009, some sites might exhibit slightly greater than 100% completeness if
61 samples were collected during that year. For purposes of this completeness calculation, non-
detects were counted equivalently with measurable values. Conversely, missing values were not
counted toward the percentage complete. Completeness statistics were not adjusted for
abbreviated collection periods because all sites were operated for the entire 12 months during
2009.
Completeness statistics were computed for primary samples or, if the primary
measurement was missing, for the collocated samples collected at the same location during the
same sampling period. To ensure that only a single record was included for each site and date,
the maximum value of the measurements was retained across primary and collocated samples. In
this way, if one of the measurements was missing and the other was not detected/measured, the
maximum would capture the not detected/measured record. If both primary and collocated
records contained a missing value, only one record would be tallied for the completeness count.
Finally, if both records contained a not detected or measured value, the larger of the two would
be captured for the completeness count. Because sample collection at some locations was
performed more frequently to meet the requirements of other sampling networks or for other
specific purposes, only records that occurred at the required l-in-6 day sample collection
frequency (days 0, 6, 12, 18, 24, 30, etc.), starting with the first collection date for each site in
calendar year 2009, were counted. For this and other reasons, it is not possible to discern from
the AQS database when makeup samples are collected. The individual enumeration of valid
samples from each and every site would be an extremely tedious task and presumes that only
NATTS sample records are present in the database for a given parameter occurrence code.
Therefore, to account for makeup samples collected near the time of the scheduled collection
date, the interval of days since the last collection event was allowed to vary between 4 and 8. No
6

-------
correction was applied for compound-specific missing data (e.g., the value for benzene was
missing, but the value for dichloromethane was non-missing). It is assumed that this discrepancy
does not significantly distort the percentage completeness.
The results of the completeness assessment are presented for each collection location and
representative compound in Table 5 and in Figures 1 through 7. Mean and median completeness
values across all NATTS laboratories for a given analyte and across all analytes for a given site
are also presented. In cases where no data were reported, the particular analyte class was not
collected at that NATTS site, as indicated by table notes.
Table 5. Percentage Completeness of the 2009 AQS Dataset by Site for Seven Hazardous
Air Pollutants.

Parameter Code -»
45201
43218
43502
43505
17141
12115
82103
AQS Site ID
Site Name
BENZ
BUTA
FORM
ACRO
NAPH
CRVI
AS
25-025-0042
Boston-Roxbury, MA
97
97
102
97
100
98
95
49-011-0004
Bountiful, UT
95
95
95
95
100
97
92
36-005-0110
Bronx, NY
102
102
102
102
97
98
NR
45-025-0001
Chesterfield, SC
102
102
102
102
102
100
102
26-163-0033
Dearborn, Ml
93
93
97
93
97
97
102
13-089-0002
Decatur, GA
102
102
102
102
100
93
102
48-201-1039
Deer Park, TX
102
102
98
102
100
100
102
08-077-0017b, -0018°
Grand Junction, CO
85
85
95
85
98
98
102
21-043-0500
Grayson Lake, KY
93
93
87
93
97
98
95
48-203-0002
Harrison County, TX
100
100
97
97
95
98
102
12-057-3002
Hillsborough County, FL
100
100
100
100
102
102
102
41-061-0119
La Grande, OR
70
82
97
NR
70
92
88
06-037-1103
Los Angeles, CA
100
100
100
100
100
NR
NR
55-027-0007
Mayville, Wl
88
88
92
88
98
97
72
17-031-4201
Northbrook, IL
102
102
102
102
97
100
102
04-013-9997
Phoenix, AZ
95
95
98
95
95
97
95
12-103-0026
Pinellas County, FL
102
102
102
102
98
102
102
41-051-0246
Portland, OR
65
80
102
NR
68
98
98
44-007-0022
Providence, Rl
102
102
102
102
97
102
102
51-087-0014
Richmond, VA
100
100
102
100
100
100
100
36-055-1007
Rochester, NY
102
102
102
102
100
100
NR
06-065-8001
Rubidoux, CA
98
98
102
98
95
NR
NR
06-085-0005
San Jose, CA
93
93
102
93
102
NR
100
53-033-0080
Seattle, WA
93
93
93
93
97
100
93
29-510-0085
St. Louis, MO
102
100
100
102
95
97
102
50-007-0007
Underhill, VT
102
102
102
102
97
98
NRa
11-001-0043
Washington, DC
100
100
NR
100
97
102
102

Mean
96
97
99
98
96
99
98

Std. Dev.
9
6
4
5
8
2
7

Median
100
100
101
100
97
98
102
a Not reported for this site.
b Metals only.
c Carbonyls, VOCs, and PAHs only.
7

-------
CMPDALIAS = BENZ
Site Name
Boston, MA
Bountiful, UT
Bronx, NY
Chesterfield, SC
Dearborn, Ml
Decatur, GA
Deer Park, TX
Grand Junction, CO
Grayson Lake, KY
Harrison County, TX
Hillsborough County, FL
La Grande, OR
Los Angeles, CA
Mayville, Wl
Northbrook, IL
Phoenix, AZ
Pinellas County, FL
Portland, OR
Providence, Rl
Richmond, VA
Rochester, NY
Rubidoux, CA
San Jose, CA
Seattle, WA
St. Louis, MO
Underhill, VT
Washington, DC
10
20 30
40
50
60
70
80
90 100 110
Percent Complete
Figure 1. Completeness for Benzene at NATTS Sample Collection Sites in 2009
(MQO reference indicated at 85%).
CMPDALIAS = BUTA
Site Name
Boston, MA
Bountiful, UT
Bronx, NY
Chesterfield, SC
Dearborn, Ml
Decatur, GA
Deer Park, TX
Grand Junction, CO
Grayson Lake, KY
Harrison County, TX
Hillsborough County, FL
La Grande, OR
Los Angeles, CA
Mayville, Wl
Northbrook, IL
Phoenix, AZ
Pinellas County, FL
Portland, OR
Providence, Rl
Richmond, VA
Rochester, NY
Rubidoux, CA
San Jose, CA
Seattle, WA
St. Louis, MO
Underhill, VT
Washington, DC
10
20
30
40
50
60
70
80
90 100
110
Percent Complete
Figure 2. Completeness for 1,3-Butadiene at NATTS Sample Collection Sites in 2009
(MQO reference indicated at 85%).
8

-------
CMPDALIAS=ACRO
Site Name
Boston, MA
Bountiful, UT
Bronx, NY
Chesterfield, SC
Dearborn, Ml
Decatur, GA
Deer Park, TX
Grand Junction, CO
Grayson Lake, KY
Harrison County, TX
Hillsborough County, FL
Los Angeles, CA
Mayville, Wl
Northbrook, IL
Phoenix, AZ
Pinellas Counly, FL
Providence, Rl
Richmond, VA
Rochester, NY
Rubidoux, CA
San Jose, CA
Seattle, WA
St. Louis, MO
Underhill, VT
Washington, DC
10
20 30
40
50
60
70
80
90 100 110
Percent Complete
Figure 3. Completeness for Acrolein at NATTS Sample Collection Sites in 2009
(MQO reference indicated at 85%).
CMPDALIAS=FORM
Site Name
Boston, MA
Bountiful, UT
Bronx, NY
Chesterfield, SC
Dearborn, Ml
Decatur, GA
Deer Park, TX
Grand Junction, CO
Grayson Lake, KY		
Harrison County, TX
Hillsborough County, FL
La Grande, OR
Los Angeles, CA
Mayville, Wl		
Northbrook, IL
Phoenix, AZ
Pinellas County, FL
Portland, OR
Providence, Rl	1
Richmond, VA
Rochester, NY
Rubidoux, CA
San Jose, CA
Seattle, WA
St. Louis, MO
Underhill. VT
I 1 1 1 1 I 1 1 1 1 I 1 1 1 1 I 1 1 1 1 I 1 1 1 1 I 1 1 1 1 I 1 1 1 1 I 1 1 1 1 I	| i i i i | i i i i |
0 10 20 30 40 50 60 70 80 90 100 110
Percent Complete
Figure 4. Completeness for Formaldehyde at NATTS Sample Collection Sites in 2009
(MQO reference indicated at 85%).
9

-------
CM PDALI AS = NAPH
Site Name
Boston, MA
Bountiful, UT
Bronx, NY
Chesterfield, SC
Dearborn, Ml
Decatur, GA
Deer Park, TX
Grand Junction, CO
Grayson Lake, KY
Harrison County, TX
Hillsborough County, FL
La Grande, OR
Los Angeles, CA
Mayville, Wl
Northbrook, IL
Phoenix, AZ
Pinellas County, FL
Portland, OR
Providence, Rl
Richmond, VA
Rochester, NY
Rubidoux, CA
San Jose, CA
Seattle, WA
St. Louis, MO
Underhill, VT
Washington, DC
10
20 30
40
50
60
70
80
90 100 110
Percent Complete
Figure 5. Completeness for Naphthalene at NATTS Sample Collection Sites in 2009
(MQO reference indicated at 85%).
CMPDALIAS=CRVI
Site Name
Boston, MA
Bountiful, UT
Bronx, NY
Chesterfield, SC
Dearborn, Ml
Decatur, GA
Deer Park, TX
Grand Junction, CO
Grayson Lake, KY
Harrison County, TX
Hillsborough County, FL
La Grande, OR
Mayville, Wl
Northbrook, IL
Phoenix, AZ
Pinellas County FL
Portland, OR
Providence, Rl
Richmond, VA
Rochester, NY
Seattle, WA
St. Louis, MO
Underhill, VT
Washington, DC
10
20
30
40
50
60
70
80
90 100
110
Percent Complete
Figure 6. Completeness for Chromium (VI) at NATTS Sample Collection Sites in 2009
(MQO reference indicated at 85%).
10

-------
CMPDALIAS=AS
Site Name
Boston, MA
Bountiful, UT
Chesterfield, SC
Dearborn, Ml
Decatur, GA
Deer Park, IX
Grand Junction, CO
Grayson Lake, KY
Harrison County, IX
Hillsborough County, FL
La Grande, OR
Mayville, Wl
Northbrook, IL
Phoenix, AZ
Pinellas County, FL
Portland, OR
Providence, Rl
Richmond, VA
San Jose, CA
Seattle, WA
St. Louis, MO
Washington, DC
I 1 1 1 1 I 1 1 1 1 I 1 1 1 1 I 1 1 1 1 I 1 1 1 1 I 1 1 1 1 I 1 1 1 1 I 1 1 1 1 I 1 1 " 1 1 I 1 1 1 1 I 1 1 1 1 I
0 10 20 30 40 50 60 70 80 90 100 110
Percent Complete
Figure 7. Completeness for Arsenic at NATTS Sample Collection Sites in 2009
(MQO reference indicated at 85%).
With the exceptions of benzene and naphthalene at the two Oregon sites and arsenic at
the Mayville, WI, site, all sites exhibited consistently high completeness statistics for all analytes
and met the MQO of 85% completeness. The preponderance of completeness metrics at 102%
reflects the fact that most sites collected 61 samples during 2009 and completeness is based on
the collection of 60 samples.
2.3 Precision of NATTS Data
Three basic sample types are collected at NATTS sites:
•	Primary samples—a single sample that represents a particular sampling event.
•	Duplicate samples—a replicate sample, collected simultaneously with the primary
sample, that represents a second measurement from the same sample stream (e.g., the
inlet stream of an outdoor air monitor) but employs an independent sample collection
device (e.g., sampling pump) and collection substrate (e.g., filter) from the primary
sample. Duplicate samples provide the basis for assessing the aggregate variability
associated with the collection device, sampling substrate, and sample analysis.
•	Collocated samples—a replicate sample, collected simultaneously with the primary
sample, that represents a second measurement from a completely independent (but
spatially close, usually 1 to 2 meters away from the primary sampler) sample stream,
collection device, and collection substrate from the primary sample. Collocated
samples provide the basis for assessing the total variability associated with all
11

-------
components of the sample collection and analysis scheme; thus, the analyst can
assume that the air collected by the primary and collocated samplers is absolutely
identical in its composition. Samples collected at different sites violate this basic
premise of collocation and were excluded from these precision analyses at the
direction of EPA.
•	Replicate Sampling:
Replicate sampling refers, generally, to both duplicate and collocated sample collections
as described above and as differentiated within the AQS database. Precision assessments
associated with replicate sampling are distinctly different from those associated with replicate
analyses as the latter are derived from a second chemical analysis of a single sample and the
former are derived from single chemical analyses of two different samples. For this report,
precision analyses were performed exclusively on NATTS sites; surrogate, non-NATTS sites
with collocated samplers have not been included. The methodological precision for the NATTS
data was assessed from both analytical (i.e., instrumental) and overall (i.e., instrumental +
sampling) perspectives. Analytical precision measures the variability in reported results due
exclusively to differences in analytical instrument performance and was estimated by comparing
results from two analyses of a single sample, whether that sample be primary, duplicate, or
collocated. Overall sampling precision was assessed by comparing the results from primary and
collocated samples or from primary and duplicate samples and accounts for the combined
variability associated with sample collection and sample analysis. Despite the differences, albeit
subtle, between duplicate and collocated samples, this report provides separate overall precision
estimates for these two replicate sample types.
For the purposes of these precision assessments, the AQS database was queried for two
distinct record types: RP records and RD records. RP records contain data for various types of
replicate samples and analyses associated with a particular sampling date, site, and chemical
parameter. Different types of replicates are identified by the value of the precision ID variable
(PRECISID) according to the following scheme:
•	PRECISID = 1: Collocated sample data
•	PRECISID = 2: Replicate analysis of a primary sample
•	PRECISID = 3: Replicate analysis of a collocated sample
For this report, analytical precision was computed from the replicate pairs of data coded
with either Precision ID 2 or 3. Overall precision was computed using the data in the raw data
records as described below.
In addition to the replicate records, raw data (AQS RD) transactions provide a second
source of primary and collocated data in AQS. Using the POCs shown for each NATTS site
listed in Table 6, it is possible to distinguish among primary, duplicate, and collocated sampling
events. For example, primary samples collected at the Chesterfield, SC, NATTS site are assigned
a parameter occurrence code of 1, while collocated samples collected at the same site are
assigned a parameter occurrence code of 2. This results in the creation of two distinct records for
each sampling event at which a collocated sample is collected. Duplicate samples are similarly
identified. Because the assignment of a particular POC is made at the discretion of each NATTS
12

-------
site, extensive effort was required to ensure that the POCs for each site were correctly identified.
POCs for primary, duplicate, and collocated samples of each chemical class were determined by
hierarchical exploration of three principal pieces of information:
1)	POCs used by each NATTS collection site in 2007 and 2008 were used as the
reference for POCs assigned in 2009.
2)	POCs assigned in previous years were confirmed by results of frequency analysis
performed on RD records for samples collected in 2009.
3)	Discrepancies and/or uncertainties about POC assignments were resolved by direct
contact with NATTS administrators for specific collection sites.
Multiple POCs for a given site, analyte, and sample type reflect a number of factors
unique to a site during 200 , largely made for reasons known only to the NATTS site
administrators. Overall precision estimates were computed by comparing primary and collocated
records for a particular site, chemical parameter, and sample collection date.
Table 6. Parameter Occurrence Codes by NATTS Site and Analyte Type.






Parameter Occurrence Codes (POCs)3















Chromium


AQS Site

voc

Carbonyls
Metals

PAHs


(VI)
Region
Site Identifier
Code
Pb
Dc
Cd
P
D
C
P D C
P
D
C
P
D C
1
Boston, MA
25-025-0042
10
11

3
4

6 7
6


6
7
1
Underhill, VT
50-007-0007
1


1


1
6


6
7
1
Providence, Rl
44-007-0022
2


5

7
1 2
6


6
7
II
Bronx, NY
36-005-0110
2


2


1 2
6


6
7
II
Rochester, NY
36-055-1007
2


2


1
6


6
7
III
Washington, DC
11-001-0043
4


1


1
1


1
2
III
Richmond, VA
51-087-0014
4
1

2


1
6


6
7
IV
Chesterfield, SC
45-025-0001
1

2
1

2
1 2
6


6
7
IV
Decatur, GAe
13-089-0002
1,3

2,4
2

3
1 2
6

7
6
7
IV
Grayson Lake, KY
21-043-0500
1
2

1
2

1 2
6


6
7
IV
Hillsborough
County, FL
12-057-3002
1


6


5
6

7
6
7
IV
Pinellas County,
FL
12-103-0026
1


6


5
6

7
6
7
V
Dearborn, Ml
26-163-0033
1

2
1

2
1 9
1

2
1
2
V
Mayville, Wl
55-027-0007
1


1
2

1 2
6


6
7
V
Northbrook, IL
17-031-4201
6

7
6

7
6 7
6


6
7
VI
Deer Park, TX
48-201-1039
2

3
3


1
1
2
6
1
2
VI
Harrison County,
TX
48-203-0002
1


1


1
1


1

VII
St. Louis, MO
29-510-0085
6


6


6 7
6


6
7
VIII
Bountiful, UT
49-011-0004
6


6


1 2
6


6
7
VIII
Grand Junction,
CO
08-077-0017,
-0018
6


6


3 4
6


6
7
IX
Phoenix, AZ
04-013-9997
6

7
30

31
1
3


6
7
IX
Los Angeles, CA
06-037-1103
4

5
4

5

6




IX
Rubidoux, CA
06-065-8001
4

5
4

5

6
7



IX
San Jose, CA
06-085-0005
3

5
3

1
1
1




X
La Grande, OR
41-061-0119
7


7


7
7


7

13

-------
X Portland, OR 41-051-0246 7 97 97 97 97
X	Seattle, WA 53-033-0080 6	7 6	7 6	7 6 7	6 7
a As reported by the NATTS site administrator. Multiple POCs reflect different analytes or changes in assignments during the monitoring year.
b P = Primary
c D = Duplicate
11 C = Collocated
e Benzene on POCs 3 and 4; all other VOCs on POCs 1 and 2.
Table 7, complemented by Table 8, presents the laboratories that analyzed specific
sample types for each NATTS site. Of particular note is the fact that some laboratories provided
analytical chemistry services for multiple NATTS sites. Laboratory codes presented in Table 8
were assigned by Alion Science, Inc., to track PT samples and their results.
Table 7. Laboratories Performing Analyses for the Different Analyte Types for Each
NATTS Site in 2009.
Site Identifier
VOCsa
Carbonyls
Metals
PAHs
Chromium (VI)
Boston-Roxbury, MA
RIDOH
MADEP
ERG
ERG
ERG
Underbill, VT
ERG
VTDEC
ERG
ERG
ERG
Providence, Rl
RIDOH
RIDOH
RIDOH
ERG
ERG
Bronx, NY
NYSDEC
NYSDEC
RTI
ERG
ERG
Rochester, NY
NYSDEC
NYSDEC
RTI
ERG
ERG
Washington, DC
MDE
PAMSL
WVDEP
ERG
ERG
Richmond, VA
VA DCLS
VA DCLS
VA DCLS
ERG
ERG
Chesterfield, SC
SCDHEC
SCDHEC
SCDHEC
ERG
ERG
Decatur, GA
GADNR
GADNR
GADNR
ERG
ERG
Grayson Lake, KY
KYDES
KYDES
KYDES
ERG
ERG
Hillsborough County, FL
PCDEM
ERG
EPCHC
ERG
ERG
Pinellas County, FL
PCDEM
ERG
EPCHC
ERG
ERG
Dearborn, Ml
ERG
ERG
MIDEQ
ERG
ERG
Mayville, Wl
WSLH
WSLH
WSLH
ERG
ERG
Northbrook, IL
ERG
ERG
ERG
ERG
ERG
Deer Park, TX
TCEQ
TCEQ
TCEQ
TCEQ
TCEQ
Harrison County, TX
TCEQ
TCEQ
TCEQ
TCEQ
TCEQ
St. Louis, MO
ERG
ERG
ERG
ERG
ERG
Bountiful, UT
ERG
ERG
ERG
ERG
ERG
Grand Junction, CO
ERG
ERG
IMLASb
ERG
ERG
Phoenix, AZ
ERG
ERG
ERG
ERG
ERG
San Jose, CA
BAAQMD
BAAQMD
ERG
ERG
CARB
Rubidoux, CA
SCAQMD
SCAQMD
SCAQMD
ERG
CARB
Los Angeles, CA
SCAQMD
SCAQMD
SCAQMD
ERG
CARB
La Grande, OR
ODEQ
ODEQ
ODEQ
ODEQ
ODEQ
Portland, OR
ODEQ
ODEQ
ODEQ
ODEQ
ODEQ
Seattle, WA
ERG
ERG
ERG
ERG
ERG
a Includes acrolein.
b Switching from IMLAS to CO State Lab effective January 2010.
14

-------
Table 8. Laboratory Abbreviations and Descriptions for NATTS Laboratories.

Laboratory

Laboratory Code(s)
Abbreviation
Laboratory Description
01-01-C,V,M
RIDOH
Rhode Island Department of Health
01-02-C.V
VTDEC
Vermont Department of Environmental Conservation
01-03-C
MADEP
Massachusetts Department of Environmental Protection
01-04-M
USEPAR1
U.S. EPA Region 1 Laboratory
02-01-C.V
NYSDEC
New York State Department of Environmental Conservation
03-01-V
MDE
Maryland Department of the Environment
03-01-C
PAMSL
Philadelphia Air Management Services Laboratory
03-01-M
WVDEP
West Virginia Department of Environmental Protection
04-01-M
EPCHC
Environmental Protection Commission of Hillsborough County
04-01-V
PCDEM
Pinellas County Department of Environmental Management
04-02-C,M,V,P
SCDHEC
South Carolina Department of Health and Environmental
Control
04-03-C,M,V
KYDES
Kentucky Division of Environmental Services
04-04-C,M,V
GADNR
Georgia Department of Natural Resources
05-01-M
MIDEQ
Michigan Department of Environmental Quality
05-03-C,M,V
WSLH
Wisconsin State Laboratory of Hygiene
06-01-C,M,V,P
TCEQ
Texas Commission on Environmental Quality
08-02-M
IMLAS
IML Air Science Laboratory
09-03-C,V
BAAQMD
Bay Area Air Quality Management District
10-02-C,M,V
ODEQ
Oregon Department of Environmental Quality
11-01-C,M,V;
ERGa
Eastern Research Group
11-02-M
RTI
RTI International
a
VA DCLS
Virginia Division of Consolidated Laboratory Services
a
SCAQMD
South Coast Air Quality Management District
b
CARB
California Air Resources Board
1 Did not participate in the PT program in 2009.
' Cr(VI) not included in the PT program during 2009.
2.3.1 A nalytical Precision Results
Analytical precision was computed from the results of the primary and collocated
samples and their respective replicate analyses extracted from RP records in the AQS database.
This measure of agreement, expressed as the percentage coefficient of variation (% CV), is
defined algebraically in Eq. 1:
%CV = 100-
z
(Pi -'))
05 -iPi +rt)_
2 n
(Eq. 1)
15

-------
where
Pi = the result of the principal analysis on sample z,
Tj = the result of the replicate analysis on sample z, and
n = the number of principal-replicate analysis pairs.
The analytical precision for all measured HAPs analyzed in samples collected in calendar
year 200 is presented in Table 9 with selected analytes summarized graphically in Figures 8
through 14.
As in previous reporting years, the agreement between replicate analyses of the same
samples is highly variable across sites/laboratories but largely still within the MQO guidelines.
Although only reported for three sites, arsenic agreement is below 2%. Showing marked
improvement since 2008, nearly all laboratories show agreement within the MQO for chromium
(VI). Conversely, agreement between formaldehyde reanalyses is quite variable, albeit
consistently well within the MQO for all sites. Although an MQO is not assigned for PAHs,
naphthalene exhibits agreements below 5% for all reporting sites, well below CVs reported for
VOCs.
2.3.2 Overall Precision Results
Overall precision was computed from the results of the primary, duplicate, and collocated
samples extracted from RD records in the AQS database. This measure of agreement, expressed
as the % CV, is defined algebraically in Eq. 2:
%CV = 100
1
Z
(Pr~ri)
0-5 • (pi + rt)
2 n
(Eq. 2)
where
pt = the result of the principal analysis on primary sample z,
rt = the result of the principal analysis on collocated sample z, and
n = the number of primary-collocated sample pairs.
The overall precision results for samples collected in calendar year 2009 are presented in
Table 10 and summarized graphically in Figures 15 through 21. For cases where either the
primary or collocated sample yielded a result of zero or had a value below the reported MDL, the
data pairs were excluded from the overall precision estimate. All data pairs with measurable
values were included in the computation.
16

-------
Table 9. Analytical Precision" for Replicate Analyses of 2009 NATTS Data.
AQS Site Code
Site Description
BENZ
BUTA
CTET
CLFRM
EDB
DCP
EDC
MECL
TCE1122
PERC
TCE
VCM
04-013-9997
Phoenix, AZ
5.1
7.5
9.1
6.2
—
—
—
6.8

5.8
15.1
20.2


(11)
(11)
(11)
(9)



(11)
—
(11)
(2)
(1)
06-065-8001
Rubidoux, CA
	b
—
—
—
—
—
—
—
—
—
—
—
06-085-0005
San Jose, CA
7.2
0
6.2
8.5
—
—
—
9.9

2.8
0



(12)
(2)
(12)
(9)



(9)
—
(11)
(6)
—
08-077-0017
Grand Junction, CO
—
—
—
—
—
—
—
—
—
—
—
—
08-077-0018
Grand Junction, CO
10.3
4.8
5.8
11.4



5.5

18.5
4.7



(10)
(10)
(10)
(10)



(10)

(10
(6)

11-001-0043
Washington, DC
—
—
—
—
—
—
—
—
—
—
—
—
12-057-3002
Hillsborough County,
FL
—
—
—
—
—
—
—
—
—
—
—
—
12-103-0026
Pinellas County, FL
3.3
12.9
2.9
17.3
27.2

7.5
15.3
21.1
9.7
18.1
28.1


(39)
(39)
(39)
(39)
(13)
(0)
(37)
(39)
(33)
(39)
(28)
(12)
13-089-0002
Decatur, GA
—
—
—
—
—
—
—
—
—
—
—
—
17-031-4201
Northbrook, IL
5.4
13.1
2.7
4.7
—
—
12.8
2.8

9.6
11.3



(5)
(3)
(3)
(3)


(2)
(3)
—
(3)
(3)
—
21-043-0500
Grayson Lake, KY
—
—
—
—
—
—
—
—
—
—
—
—
25-025-0042
Boston, MA
—
—
—
—
—
—
—
—
—
—
—
—
26-163-0033
Dearborn, Ml
6.6
7.2
4.7
5.8
—
—
—
4.2
—
7.3
17
0


(16)
(16)
(16)
(16)



(16)

(16)
(4)
(3)
29-510-0085
St. Louis, MO
5.2
11.6
13.8
9.1
—
—
—
12.3
—
4.9
8.5
20.4


(12)
(12
(12
(12)



(12)

(12)
(6)
(3)
36-005-0110
Bronx, NY
—
—
—
—
—
—
—
—
—
—
—
—
36-055-1007
Rochester, NY
—
—
—
—
—
—
—
—
—
—
—
—
44-007-0022
Providence, Rl
—
—
—
—
—
—
—
—
—
—
—
—
45-025-0001
Chesterfield, SC
—
—
—
—
—
—
—
—
—
—
—
—
49-011-0004
Bountiful, UT
30.9
11.3
13
12.1
—
—
—
14.1

13
54.7



(10)
(9)
(6)
(6)



(6)
—
(6)
(4)
—
50-007-0007
Underhill, VT
—
—
—
—
—
—
—
—
—
—
—
—
51-087-0014
Richmond, VA
—
—
—
—
—
—
—
—
—
—
—
—
53-033-0080
Seattle, WA
6.5
8.5
4.4
10.7
—
—
—
4.1

4.6
6.7



(8)
(8)
(8)
(8)



(8)
—
(8)
(2)
—
55-027-0007
Mayville, Wl
—
—
—
—
—
—
—
—
—
—
—
—

Overall Mean
10.5
10.5
7.1
12.5
27.2

7.8
11.2
21.1
9.3
19.8
24.2


(123)
(110)
(117)
(112)
(13)
—
(39)
(114)
(33)
(116)
(61)
(19)
(continued)

-------
Table 9. Analytical Precision" for Replicate Analyses of 2009 NATTS Data (continued).
AQS Site Code
Site Description
cDCPEN
tDCPEN
ACRO
ACRY
NAPH
BaP
FORM
ACET
AS
BE
CD
PB
MN
Ni
CRVI
04-013-9997
Phoenix, AZ
—
—
7.4
(11)
8.9
(3)
b
—
0.5
(13)
0.8
(13)
—
—
—
—
—
—
6
(12)
06-065-8001
Rubidoux, CA
—
—
—
—
4.2
(8)
8.6
(3)
—
—
—
—
—
—
—
—
—
06-085-0005
San Jose, CA
—
—
—
—
—
—
1.3
(10)
0.6
(10)
—
—
—
—
—
—
—
08-077-0017
Grand Junction, CO
—
—
—
—
—
—
—
—
—
—
—
—
—
—
10.8
(6)
08-077-0018
Grand Junction, CO
—
—
13.6
(10)
12.6
(2)
—
—
1.5
(8)
0.9
(8)
—
—
—
—
—
—
—
11-001-0043
Washington, DC
—
—
—
—
—
—
—
—
—
—
—
—
—
—
9.4
(6)
12-057-3002
Hillsborough
County, FL
—
—
—
—
3.2
(66)
7.7
(18)
4.9
(12)
1.5
(12)
—
—
—
—
—
—
12.8
(6)
12-103-0026
Pinellas County, FL
32.9
(11)
34.3
(13)
11
(39)
30.8
(37)
—
—
3.4
(12)
4.4
(12)
—
—
—
—
—
—
7.3
(4)
13-089-0002
Decatur, GA
—
—
—
—
2.7
(12)
3.6
(1)
—
—
—
—
—
—
—
—
8.5
(9)
17-031-4201
Northbrook, IL
—
—
10.5
(3)
—
—
—
0.5
(10)
0.4
(10)
1.2
(48)
18.6
(43)
6
(48)
1
(48)
1.1
(48)
4
(48)
10.5
(6)
21-043-0500
Grayson Lake, KY
—
—
—
—
—
—
—
—
—
—
—
—
—
—
7.2
(4)
25-025-0042
Boston, MA
—
—
—
—
—
—
—
—
1.4
(66)
36.6
(52)
9.1
(66)
0.8
(66)
1.1
(66)
1.7
(66)
9.1
(7)
26-163-0033
Dearborn, Ml
—
—
6.7
(16)
—
1.3
(11)
3.7
(11)
0.3
(10)
0.6
(10)
—
—
—
—
—
—
4.2
(12)
29-510-0085
St. Louis, MO
—
—
25.3
(12)
13.9
(6)
—
—
1.5
(14)
1.8
(14)
2
(19)
22.6
(19)
0.7
(19)
1.4
(19)
0.7
(19)
1.8
(19)
6.7
(11)
36-005-0110
Bronx, NY



—
—
—
—
—
—
—
—
—
—
—
4.5
(10)
36-055-1007
Rochester, NY



—
—
—
—
—
—
—
—
—
—
—
10.2
(2)
(continued)

-------
Table 9. Analytical Precision" for Replicate Analyses of 2009 NATTS Data (continued).
AQS Site Code
Site Description
cDCPEN
tDCPEN
ACRO
ACRY
NAPH
BaP
FORM
ACET
AS
BE
CD
PB
MN
Ni
CRVI
44-007-0022
Providence, Rl
—
—
—
—
—
—
—
—
—
—
—
—
—
—
7.2
(4)
45-025-0001
Chesterfield, SC
—
—
—
—
—
—
—
—
—
—
—
—
—
—
15
(6)
49-011-0004
Bountiful, UT
—
—
16.9
(6)
13.9
(2)
—
—
7.6
(10)
3.4
(10)
—
—
—
—
—
—
4.4
(8)
50-007-0007
Underhill, VT
—
—
—
—
—
—
—
—
—
—
—
—
—
—
8.5
(2)
51-087-0014
Richmond, VA
—
—
—
—
—
—
—
—
—
—
—
—
—
—
3.1
(5)
53-033-0080
Seattle, WA
—
—
7.3
(8)

2
(11)
3.7
(4)
0.5
(12)
1.3
(12)
—
—
—
—
—
—
4.7
(9)
55-027-0007
Mayville, Wl
—
—
—
—
—
—
—
—
—
—
—
—
—
—
10.6
(4)

Overall Mean
32.9
(11)
34.3
(13)
13.1
(106)
27.2
(50)
3
(108)
6.4
(37)
3.1
(111)
2
(111)
1.4
(133)
28.8
(114)
7.4
(133)
1
(133)
1.1
(133)
2.8
(133)
8.1
(133)
a Expressed as percentage coefficient of variation (%CV) with number of contributing data pairs presented in parentheses. Metals results
are reported at STP at most sites and lcs at others.
b Sample not collected or analyte not reported.
c Across all sites.

-------
Site Name
Bountiful, UT
Dearborn, Ml
Grand Junction, CO
NorHibrook, IL
Phoenix, AZ
Pinellas County, FL
San Jose, CA
Seattle, WA
St. Louis, MO
0	10	20	30	40
Percent CV
Figure 8. Analytical Precision Summary for Benzene at NATTS Sample Collection Sites in
2009 (MQO reference indicated at 15%).
CMPDALIAS=BUTA
Site Name
Bountiful, UT
Dearborn, Ml
Grand Junction, CO
Northbrook, IL
Phoenix, AZ
Pinellas County, FL
Seattle, WA
St. Louis, MO
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Percent CV
Figure 9. Analytical Precision Summary for 1,3-Butadiene at NATTS Sample Collection
Sites in 2009.
20

-------
Site Name
Bountiful, UT
Dearborn, Ml
Grand Junction, CO
Norttibrook, IL
Phoenix, AZ
Pinellas County, FL
Seattle, WA
St. Louis, MO
0	10	20	30
Percent CV
Figure 10. Analytical Precision Summary for Acrolein at NATTS Sample Collection Sites
in 2009 (MQO reference indicated at 15%).
CMPDALIAS= FORM
Site Name
Bountiful, UT
Dearborn, Ml
Grand Junction, CO
Hillsborough County, FL
Northbrook, IL
Phoenix, AZ
Pinellas County, FL
San Jose, CA
Seattle, WA
St. Louis, MO
012345678
Percent CV
Figure 11. Analytical Precision Summary for Formaldehyde at NATTS Sample Collection
Sites in 2009.
21

-------
CM PDALIAS = N APH
Site Name
Dearborn, Ml
Decatur, GA
Hillsborough County, FL
Rubidoux, CA
Seattle, WA
2	3
Percent CV
Figure 12. Analytical Precision Summary for Naphthalene at NATTS Sample Collection
Sites in 2009.
CMPDALIAS=CRVI
Site Name
Boston, MA
Bountiful, UT
Bronx, NY
Chesterfield, SC
Dearborn, Ml
Decatur, GA
Grand Junction, CO
Grayson Lake, KY
Hillsborough County, FL
Mayville, Wl
Norlhbrook, IL
Phoenix, AZ
Pinellas County, FL
Providence, Rl
Richmond, VA
Rochester, NY
Seattle, WA
St. Louis, MO
Underhill, VT
Washington, DC
8 9 10 11 12 13 14 15 16
Percent CV
Figure 13. Analytical Precision Summary for Chromium (VI) at NATTS Sample Collection
Sites in 2009 (MQO reference indicated at 15%).
22

-------
CMPDALIAS = AS
Site Name
Boston, MA
Northbrook, IL
St. Louis, MO
0	12	3
Percent CV
Figure 14. Analytical Precision Summary for Arsenic at NATTS Sample Collection Sites in
200 .
Examination of Figures 15 through 21 reveals that aggregate precision associated with
sample collection and analysis varies widely by collection site and analyte. Not unexpectedly,
the aggregate variability observed is substantially greater than the analytical variability shown in
Figures 8 through 14. Variability seen for many sites may reflect the presence of extreme values.
No attempt has been made to elucidate this cause through careful review of the individual data
pairs for each analyte and site. With the exception of acrolein where only one site achieved the
MQO in 2009, some sites achieve the MQOs, and some sites do not for most analytes,
suggesting that the 15% threshold is a reasonable target for the MQO. The fact that many sites
exhibit percentage CVs above the MQO target level suggests that the collection methodology
contributes significantly to the overall variability in the data for a given site and analyte. Without
identifying specific sites, the percentages of sites with percentage CV above the MQO threshold
are 50%, 60%, 67%, 36%, 9%, and 64% for arsenic, chromium (VI), benzene, 1,3-butadiene,
acrolein, and formaldehyde, respectively. These percentages are consistent with variations in
collection and analysis challenges posed by different analytes, with more problematic analytes
(e.g., butadiene, and acrolein) showing poorer attainment of the MQO. That fact not
withstanding, the percentage CVs computed across sites by analyte are somewhat misleading
because they may be influenced by atypically large CVs at selected sites. The QA report of the
NATTS stations for 2006 [5] warned of the danger of extracting duplicate and collocated results
using only the RP records. For that reason—and despite the considerable difficulty in
determining the specific primary, duplicate, and collocated POCs for each site—the data
presented here are based primarily on the RD records. The sole exception was the duplicate data
for VOCs from the Washington, DC, site that were uploaded to AQS only as RP records and
were, therefore, extracted as such.
23

-------
Table 10. Overall Precision" for Primary and Collocated Samples from 2009.
AQS Site ID
Site Description
Duplicate
Type
BENZ
BUTA
CTET
CLFRM
EDB
DCP
EDC
MECL
TCE1122
PERC
TCE
VCM
04-013-9997
Phoenix, AZ
Collocate
14.7
U)
9.2
U)
14.4
(7)
6.8
(6)
	b
—
—
50.7
U)
—
9.9
U)
6.1
(1)
—
04-013-9997
Phoenix, AZ
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
06-037-1103
Los Angeles, CA
Collocate
10.8
(27)
24.5
(28)
"
23.4
(28)
"
"
"
27
(28)
"
28.7
(25
26
(26)
"
06-037-1103
Los Angeles, CA
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
06-065-8001
Rubidoux, CA
Collocate
19
(23)
25.5
(24)
—
23.7
(24)
	
—
—
62.6
(2)
—
41.7
(23)
0
(3)
—
06-065-8001
Rubidoux, CA
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
06-085-0005
San Jose, CA
Collocate
13
(25)
60
(4)
—
55.6
(20)
	
—
—
48.8
(22)
—
44.6
(25)
37.9
(11)
—
06-085-0005
San Jose, CA
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
08-077-0017
Grand Junction, CO
Collocate
—
—
—
—
—
—
—
—
—
—
—
—
08-077-0017
Grand Junction, CO
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
08-077-0018
Grand Junction, CO
Collocate
—
—
—
—
—
—
—
—
—
—
—
—
08-077-0018
Grand Junction, CO
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
11-001-0043
Washington, DC
Collocate
—
—
—
—
—
—
—
—
—
—
—
—
11-001-0043
Washington, DC
Duplicate
7.4
(59)
23.9
(58)
6.9
(59)
6
(59)
0
(6)
15.7
(27)
17.6
(59)
12.5
(47)
10.8
(19)
6.8
(59)
0
(41)
13.1
(13)
12-057-3002
Hillsborough
County, FL
Collocate
—
—
—
—
—
—
—
—
—
—
—
—
12-057-3002
Hillsborough
County, FL
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
12-103-0026
Pinellas County, FL
Collocate
80.6
(16)
56
(16)
102
(16)
99.6
(16)
105
(5)
—
85.7
(15)
77.4
(16)
106
(13)
107
(16)
98.8
(13)
51.6
(5)
12-103-0026
Pinellas County, FL
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
13-089-0002
Decatur, GA
Collocate
24.2
(60)
—
9.5
(21)
15
(9)
—
—
0
(1)
88.7
(1)
—
21.9
(4)
0
(1)
—
13-089-0002
Decatur, GA
Duplicate


—
—
—
—
—
—
—
—
—
—
(continued)

-------
Table 10. Overall Precision" for Primary and Collocated Samples from 2009 (continued).
AQS Site ID
Site Description
Duplicate
Type
BENZ
BUTA
CTET
CLFRM
EDB
DCP
EDC
MECL
TCE1122
PERC
TCE
VCM
17-031-4201
Northbrook, IL
Collocate
25.8
(1)
—
9.1
(1)
23.6
(1)
—
—
0
(1)
107
(1)
—
34.4
(1)
31.4
(1)
—
17-031-4201
Northbrook, IL
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
21-043-0500
Grayson Lake, KY
Collocate
—
—
—
—
—
—
—
—
—
—
—
—
21-043-0500
Grayson Lake, KY
Duplicate
4.7
(4)
—
—
—
—
—
—
44.7
(11)
—
—
—
—
25-025-0042
Boston, MA
Collocate
—
—
—
—
—
—
—
—
—
—
—
—
25-025-0042
Boston, MA
Duplicate
4.8
(31)
27.3
(31)
3.2
(31)
3.6
(31)
23.6
(4)
11.2
(23
4.3
(31)
19.3
(31)
26.2
(10)
5.1
(31)
9.7
(30)
4.5
(20)
26-163-0033
Dearborn, Ml
Collocate
9.3
(8)
10.5
(8)
3.8
(8)
39.4
(8)
—
—
—
10.1
(8)
—
6.9
(8)
15.5
(2)
0
(1)
26-163-0033
Dearborn, Ml
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
29-510-0085
St. Louis, MO
Collocate
—
—
—
—
—
—
—
—
—
—
—
—
29-510-0085
St. Louis, MO
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
36-005-0110
Bronx, NY
Collocate
—
—
—
—
—
—
—
—
—
—
—
—
36-005-0110
Bronx, NY
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
36-055-1007
Rochester, NY
Collocate
—
—
—
—
—
—
—
—
—
—
—
—
36-055-1007
Rochester, NY
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
41-051-0246
Portland, OR
Collocate
12.7
(15
5.9
(3)
9.4
(15)
32.6
(1)
—
—
—
38.3
(26)
—
15.7
(2)
—
—
41-051-0246
Portland, OR
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
41-061-0119
La Grande, OR
Collocate
—
—
—
—
—
—
—
—
—
—
—
—
41-061-0119
La Grande, OR
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
44-007-0022
Providence, Rl
Collocate
—
—
—
—
—
—
—
—
—
—
—
—
44-007-0022
Providence, Rl
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
45-025-0001
Chesterfield, SC
Collocate
36.4
(12)
—
0
(11)
—
—
—
—
31.3
(24)
—
—
—
—
45-025-0001
Chesterfield, SC
Duplicate

—
—
—
—
—
—
—
—
—
—
—
(continued)

-------
Table 10. Overall Precision" for Primary and Collocated Samples from 2009 (continued).
AQS Site ID
Site Description
Duplicate
Type
BENZ
BUTA
CTET
CLFRM
EDB
DCP
EDC
MECL
TCE1122
PERC
TCE
VCM
48-201-1039
Deer Park, TX
Collocate
11.6
(55)
26
(16)
7
(55)
12.5
(53)
23.6
(4)
34.6
(5)
17.3
(29)
10.7
(51)
21.1
(5)
15.7
(42)
28.3
(20)
15.8
(20)
48-201-1039
Deer Park, TX
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
48-203-0002
Harrison County,
TX
Collocate
—
—
—
—
—
—
—
—
—
—
—
—
48-203-0002
Harrison County,
TX
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
49-011-0004
Bountiful, UT
Collocate
—
—
—
—
—
—
—
—
—
—
—
—
49-011-0004
Bountiful, UT
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
50-007-0007
Underhill, VT
Collocate
—
—
—
—
—
—
—
—
—
—
—
—
50-007-0007
Underhill, VT
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
51-087-0014
Richmond, VA
Collocate
—
—
—
—
—
—
—
—
—
—
—
—
51-087-0014
Richmond, VA
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
53-033-0080
Seattle, WA
Collocate
9.8
(4)
4.5
(4)
5.3
(4)
32.6
(4)
—
—
—
40.5
(4)
—
5.2
(4)
9.4
(1)
—
53-033-0080
Seattle, WA
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
55-027-0007
Mayville, Wl
Collocate
—
—
—
—
—
—
—
—
—
—
—
—
55-027-0007
Mayville, Wl
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—

Overall Mean
All Dups.
25.3
(288)
30.7
(141)
32.2
(169)
37.2
(201)
67.5
(13)
17.8
(28)
39.4
(77)
40.1
(254)
74.2
(28)
40.5
(188)
40.7
(109)
20.2
(46)
(continued)

-------
Table 10. Overall Precision" for Primary and Collocated Samples from 2009 (continued).
AQS Site ID
Site Description
Duplicate
Type
cDCPEN
tDCPEN
ACRO
ACRY
NAPH
BaP
FORM
ACET
AS
BE
CD
PB
MN
Nl
CRVI
04-013-9997
Phoenix, AZ
Collocate
—
—
20.7
U)
118
(1)
—
—
4.2
(6)
5.8
(6)






15.1
(6)
04-013-9997
Phoenix, AZ
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
06-037-1103
Los Angeles, CA
Collocate
—
—
89.9
(28)
—
—
—
36
(26)
31.2
(26)






15.1
(6)
06-037-1103
Los Angeles, CA
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
06-065-8001
Rubidoux, CA
Collocate
—
—
85.9
(24)
—
—
—
24.4
(29)
28.2
(29)






15.1
(6)
06-065-8001
Rubidoux, CA
Duplicate
—
—
—
—
17.2
(3)
27
(1)









06-085-0005
San Jose, CA
Collocate
—
—
—
—
—
—
24.1
(30)
—
—
—
—
—
—
—
15.1
(6)
06-085-0005
San Jose, CA
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
08-077-0017
Grand Junction,
CO
Collocate
—
—
—
—
—
—
—
—
—
4.5
(3)
4.6
(3)
29.4
(17
7.1
(36)
37
(23)
34.3
(3)
08-077-0017
Grand Junction,
CO
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
08-077-0018
Grand Junction,
CO
Collocate
—
—
—
—
—
—
—
—
—
—
—
—
7.1
(36)
37
(23)
34.3
(3)
08-077-0018
Grand Junction,
CO
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
11-001-0043
Washington, DC
Collocate
—
—
—
—
—
—
—
—
—
—
—
—
—
—
8
(3)
11-001-0043
Washington, DC
Duplicate
0
(3)
0
(3)
26.1
(50)
37.2
(49)











12-057-3002
Hillsborough
County, FL
Collocate
—
—
—
—
20.8
(33)
39.6
(8)
—
—
15.2
(59)
0
(59)
22.3
(59)
16.1
(59)
12.8
(59)
12
(59)
20.7
(3)
12-057-3002
Hillsborough
County, FL
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
(continued)

-------
Table 10. Overall Precision" for Primary and Collocated Samples from 2009 (continued).
AQS Site ID
Site Description
Duplicate
Type
cDCPEN
tDCPEN
ACRO
ACRY
NAPH
BaP
FORM
ACET
AS
BE
CD
PB
MN
Nl
CRVI
12-103-0026
Pinellas County, FL
Collocate
90.2
(6)
90.5
(7)
58.8
(16)
104
(14)










7.5
(2)
12-103-0026
Pinellas County, FL
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
13-089-0002
Decatur, GA
Collocate
—
—
—
—
11.8
(5)
—
41.5
(24)
14.3
(22)
18.3
(16)
84.9
(1)
22
(24)
27.2
(25)
24.6
(25)
15.2
(25)
14.1
(4)
13-089-0002
Decatur, GA
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
17-031-4201
Northbrook, IL
Collocate
—
—
3.7
(1)
—
—
—
4.9
(5)
2.5
(5)
17.3
(23)
43.9
(20)
24
(23)
19
(23)
10.7
(23)
23.8
(23)
35
(3)
17-031-4201
Northbrook, IL
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
21-043-0500
Grayson Lake, KY
Collocate
—
—
—
—
—
—
—
—
—
—
—
—
—
—
9.4
(2)
21-043-0500
Grayson Lake, KY
Duplicate
—
—
—
—
—
—
15.4
(27)
15.2
(27)
6.6
(26)
—
—
6.8
(56)
45.7
(50)
44.9
(3)
—
25-025-0042
Boston, MA
Collocate
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
25-025-0042
Boston, MA
Duplicate
0
(2)
47.1
(1)
19.1
(31)
58.5
(27)
—
—
7.5
(31)
28.5
(31)
3
(33)
45.3
(25)
27.9
(33)
5.9
(33)
5.9
(33)
8.7
(33)
10.8
(3)
26-163-0033
Dearborn, Ml
Collocate
—
—
26.9
(8)
—
5.3
(5)
13.8
(5)
5.3
(5)
3.8
(5)
12.3
(56)
21
(48)
15
(54)
(0)
7.8
(56)
24.3
(55)
8.9
(6)
26-163-0033
Dearborn, Ml
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
29-510-0085
St. Louis, MO
Collocate
—
—
—
—
—
—
—
—
4.7
(11)
21.5
(11)
20.9
(11)
9.1
(11)
7.4
(11)
17.6
(11)
11.4
(6)
29-510-0085
St. Louis, MO
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
36-005-0110
Bronx, NY
Collocate
—
—
—
—
—
—
—
—
7.4
(51)
15.8
(51)
37.4
(49)
4.1
(51)
3.7
(51)
7.3
(51)
7.6
(5)
36-005-0110
Bronx, NY
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
36-055-1007
Rochester, NY
Collocate
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
36-055-1007
Rochester, NY
Duplicate
—
—


—
—
—
—
—
—
—
—
—
—
—
41-051-0246
Portland, OR
Collocate
—
—


18.7
(31)
5.5
(11)
7
(47)
6.7
(47)
8.6
(40)
17.4
(40)
17.7
(40)
8.4
(40)
8.7
(40)
8.1
(40)
—
41-051-0246
Portland, OR
Duplicate




—



—
—
—
—
—
—
—
(continued)

-------
Table 10. Overall Precision" for Primary and Collocated Samples from 2009 (continued).
AQS Site ID
Site Description
Duplicate
Type
cDCPEN
tDCPEN
ACRO
ACRY
NAPH
BaP
FORM
ACET
AS
BE
CD
PB
MN
Nl
CRVI
41-061-0119
La Grande, OR
Collocate
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
41-061-0119
La Grande, OR
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
44-007-0022
Providence, Rl
Collocate
—
—
—
—
—
—
10.6
(24)
15.5
(24)
25.6
(24)
34.1
(16)
55.4
(12)
15.2
(26)
11.8
(26)
16.4
(26)
10.1
(1)
44-007-0022
Providence, Rl
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
45-025-0001
Chesterfield, SC
Collocate
—
—
—
—
—
—
10.4
(61)
12.7
(60)
25.8
(48)
—
45.7
(30)
39.1
(35)
14.4
(48)
66.6
(36)
—
45-025-0001
Chesterfield, SC
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
—
—
26.5
(3)
48-201-1039
Deer Park, TX
Collocate
23.6
(4)
33.3
(2)
102
(53)
—
37.4
(56
36.1
(30)








33.1
(23)
48-201-1039
Deer Park, TX
Duplicate
—
—
—
—
11.1
(56
35.1
(49)









48-203-0002
Harrison County,
TX
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
48-203-0002
Harrison County,
TX
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
49-011-0004
Bountiful, UT
Collocate
—
—
—
—
—
—
—
—
27.4
(2)
—
0
(1)
7.2
(3)
14.5
(3)
69.9
(1)
41.9
(4)
49-011-0004
Bountiful, UT
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
50-007-0007
Underhill, VT
Collocate
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
50-007-0007
Underhill, VT
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
—
—
7.5
(1)
51-087-0014
Richmond, VA
Collocate
—
—
—
—
—
—
—
—
—
—
—
—
—
—
6.7
(2)
51-087-0014
Richmond, VA
Duplicate
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
53-033-0080
Seattle, WA
Collocate
—
—
28.7
(4)
—
—
—
5.7
(6)
2.7
(6)







53-033-0080
Seattle, WA
Duplicate
—
—
—
—
13.2
(6)
5
(2)








16.8
(4)
(continued)

-------
Table 10. Overall Precision" for Primary and Collocated Samples from 2009 (continued).
AQS Site ID
Site Description
Duplicate
Type
cDCPEN
tDCPEN
ACRO
ACRY
NAPH
BaP
FORM
ACET
AS
BE
CD
PB
MN
Nl
CRVI
55-027-0007
Mayville, Wl
Collocate
—
—
—
—
—
—
—
—
12.5
(2)
—
2.6
(1)
35.5
(1)
21.7
(1)
12.8
(2)
7.5
(2)
55-027-0007
Mayville, Wl
Duplicate
—
—
—
—
—
—
3.8
(5)
4.5
(5)








Overall Mean
All Dups.
65.2
(12)
78.6
(10)
74
(196)
78.4
(42)
24.1
(195)
32.8
(106)
20
(326)
19.6
(323)
18.7
(99)
15.4
(54)
40
(82)
25.9
(103)
9.6
(135)
42
(110)
24.2
(83)
a Expressed as percentage coefficient of variation (%CV) with number of contributing data pairs presented in parentheses. Metals results are reported at STP at most sites and local conditions at others.
b Sample either not collected or analyte not reported.
c Across all sites.
LtJ
O

-------
Site Name
Boston, MA
Chesterfield, SC
Dearborn, Ml
Decatur, GA
Deer Park, TX
Grayson Lake, KY
Los Angeles, CA
Northbrook, IL
Phoenix, AZ
Pinellas County, FL
Portland, OR
Rubidoux, CA
San Jose, CA
Seattle, WA
Washington, DC
0 10 20 30 40 50 60 70 80 90
Percent CV
Figure 15. Overall Precision Summary for Benzene at NATTS Sample Collection Sites in
200 (MQO reference indicated at 15%).
CMPDALIAS=BUTA
Site Name
Boston, MA
Dearborn, Ml
Deer Park, TX
Los Angeles, CA
Phoenix, AZ
Pinellas County, FL
Portland, OR
Rubidoux, CA
San Jose, CA
Seattle, WA
Washington, DC
0	10	20	30	40	50	60	70
Percent CV
Figure 16. Overall Precision Summary for 1,3-Butadiene at NATTS Sample Collection
Sites in 2009 (MQO reference indicated at 15%).
31

-------
Site Name
Boston, MA
Dearborn, Ml
Deer Park, TX
Los Angeles, CA
Northbrook, IL
Phoenix, AZ
Pinellas County, FL
Rubidoux, CA
San Jose, CA
Seattle, WA
Washington, DC
0 10 20 30 40 50 60 70 80 90 100 110
Percent CV
Figure 17. Overall Precision Summary for Acrolein at NATTS Sample Collection Sites in
2009 (MQO reference indicated at 15%).
CMPDALIAS= FORM
Site Name
Boston, MA
Chesterfield, SC
Dearborn, Ml
Decatur, GA
Grayson Lake, KY
Los Angeles, CA
Mayville, Wl
Northbrook, IL
Phoenix, AZ
Portland, OR
Providence, Rl
Rubidoux, CA
San Jose, CA
Seattle, WA
0	10	20	30	40	50
Percent CV
Figure 18. Overall Precision Summary for Formaldehyde at NATTS Sample Collection
Sites in 2009 (MQO reference indicated at 15%).
32

-------
Site Name
Dearborn, Ml
Decatur, GA
Deer Park, TX
Hillsborough County, FL
Portland, OR
Rubldoux, CA
Seattle, WA
0	10	20	30
Percent CV
Figure 19. Overall Precision Summary for Naphthalene at NATTS Sample Collection Sites
in 2009 (MQO reference indicated at 15%).
CMPDALIAS=CRVI
Site Name
Boston, MA
Bountiful, UT
Bronx, NY
Chesterfield, SC
Dearborn, Ml
Decatur, GA
Deer Park, TX
Grand Junction, CO
Grayson Lake, KY
Hillsborough County, FL
Mayville, Wl
Norlhbrook, IL
Phoenix, AZ
Pinellas County, FL
Providence, Rl
Richmond, VA
Seattle, WA
St. Louis, MO
Underhill, VT
Washington, DC
Percent CV
Figure 20. Overall Precision Summary for Chromium (VI) at NATTS Sample Collection
Sites in 200 (MQO reference indicated at 15%).
33

-------
CMPDALIAS=AS
Site Name
Boston, MA
Bountiful, UT
Chesterfield, SC
Dearborn, Ml
Decatur, GA
Grayson Lake, KY
Hillsborough County, FL
Mayville, Wl
Northbrook, IL
Portland, OR
Providence, Rl
St. Louis, MO
30
Percent CV
Figure 21. Overall Precision Summary for Arsenic at NATTS Sample Collection Sites in
2009 (MQO reference indicated at 15%).
2.4 Laboratory Bias Data Based on Proficiency Testing Samples
PT audits of participating NATTS sample analysis laboratories were conducted
semiannually for VOCs and carbonyls and annually for metals and PAHs in 2009. Alion Science,
Inc., under contract to EPA (Contract No. 68-D03-006), generated "spiked" samples containing
known amounts of the HAPs of interest and delivered these spiked samples to each laboratory in
2009 for each of the VOC, carbonyl, and metals analyte groups. Following chemical analyses,
the participating laboratories returned their results to Alion, which, in turn, prepared reports
comparing the laboratory-measured values to the stated (known) values for each proficiency
testing sample. The results of these PT sample analyses were provided to RTI International by
EPA for calendar year 2009.
Laboratory bias is defined as the percentage difference between the laboratory's
measured value and the known value for the audit sample:
^	Measured - Known
%I)ifference =	100
Known
(Eq. 3)
Tables 11 through 13 present the results of the PT samples for all compounds analyzed.
To reflect overall bias independent of direction, the mean of the absolute value of the bias, along
with the minimum and maximum bias values, is presented in the bottom and right-hand
summaries for the individual tabulated values. Figure 22 shows boxplots summarizing laboratory
bias results for all the participating laboratories across the five compounds for which PT data
34

-------
Table 11. Performance Testing Bias Results" for VOCs in 2009 NATTS Laboratories.
Laboratory
Code
Lab Description
BENZ
BUTA
CTET
CLFR
M
EDB
DCP
EDC
MECL
TCE1
122
PERC
TCE
VCM
c-
CPEN
t-
CPEN
ACRO
Mean
Abs. Bias
(across
anaiylesf
Min.
Max.
01-01-V
Rl Dept. of Health
Laboratories
0.41
-11.9
-9.21
-3.43
-9.36
-8.72
-11.7
-10.7
1.79
-9.66
-13.7
-10.3
-12.1
-12.4
87.9
15.3
-21.0
186.22
02-01-C
NYS DEC BAQS
-0.77
-12.2
0.26
-15.4
-11.4
-11.3
-11.6
-6.86
1.55
-9.62
-15.9
-9.83
-17.2
-13.4
-6.09
10.6
-24.8
7.00
03-01-V
Maryland
Department of the
Environment
5.98
2.26
4.61
6.00
-8.32
-2.28
-1.90
10.7
-2.56
6.82
-3.04
12.8
-12.3
-9.17
3.41
6.97
-15.9
13.4
03-02-V
Virginia Division of
Consolidated
Laboratory Services
-21.2
-16.6
-17.6
-29.0
-0.19
-7.84
-24.1
-23.0
-4.26
-6.50
-9.36
-20.6
-14.5
-3.80
14.2
18.8
-32.5
51.9
04-01-V
Pinellas County
DEM AQ
-9.12
-6.76
-9.57
-13.1
-5.95
-12.4
-13.4
-15.5
-6.91
-11.3
-10.3
-5.84
-7.86
-8.08
-0.71
10.5
-27.6
14.1
04-02-V
SC Dept of HEC,
Div. of AQ Analysis
-24.3
19.7
-34.5
-27.9
16.8
-8.37
-7.23
-10.3
97.5
-5.96
-8.62
-20.7
-25.6
9.88
67.1
26.8
-38.0
112
04-03-V
KY Div. of
Environmental
Services
-3.08
-5.83
-4.87
-4.06
-2.18
-11.7
-7.56
-4.88
-1.99
-1.10
-3.28
1.75
-4.53
-4.61
11.3
6.80
-16.9
15.2
04-04-V
GA DNR EPD
Laboratory
-15.3
-17.8
-12.3
-17.5
-11.8
-15.7
-13.5
-12.4
-19.8
-15.9
-12.1
-8.55
-15.8
-13.7
3.15
14.3
-21.7
15.6
05-01-V
Ml DEQ Lab
-9.51
8.89
-2.22
-9.37
-6.76
-11.2
-12.2
-13.0
-0.76
-3.05
-3.22
11.7
-12.7
-3.60

10.3
-24.4
20.9
05-03-V
Wisconsin DNR
-12.2
-12.8
10.9
-19.3
-24.0
-2.37
-14.8
10.3
-25.6
-18.2
-9.47
-1.28
-21.0
-12.2
27.3
15.3
-29.5
34.3
06-01-V
Texas CEQ
-3.44
-1.44
4.35
-17.9
-17.9
-28.0
-20.3
-15.1
-1.27
-21.9
-13.4
-3.91
2.78
3.80
6.67
12.2
-30.5
14.8
09-03-V
Bay Area Air Quality
Management
District
-10.2
-23.1
0.56
-9.69
-4.06
C
-17.2
-21.0
"
-9.58
4.09
-5.32
"
"
117
21.5
-32.2
152
09-06-C
San Diego County
Air Pollution Control
District
-10.9
-9.62
-6.14
-14.4
-3.29
-14.9
-9.29
-10.3
-7.40
-8.64
-9.74
-10.8
-3.39
6.25
1.03
8.95
-25.6
13.2
10-02-V
Oregon DEQ Lab
-31.8
-10.9
-45.9
-26.7
-41.6
-30.2
-24.2
14.0
-45.2
-43.
-36.8
1.03
-35.4
-42.5
-
33.4
-61.6
29.7
11-01-V
ERG
6.16
-0.52
1.85
5.44
-1.50
-2.22
2.69
5.66
-7.25
1.11
2.71
-0.06
1.63
-4.02
-14.7
4.66
-18.5
10.4

Mean Abs. Bias
(across laboratories)
11.8
15.3
12.7
14.9
11.8
12.9
12.8
12.4
16.4
14.0
12.6
9.67
13.3
12.5
30.3
14.1



Minimum
-35.3
-42.9
-52.8
-30.3
-45.8
-33.7
-32.5
-32.2
-61.6
-32.5
-40.3
-26.8
-40.2
-50.5
-23.6




Maximum
12.0
21.1
20.0
7.45
16.8
5.26
3.42
29.7
112
3.42
14.6
20.9
11.0
14.8
186



a Computed as the mean of the individual percent differences.
b Computed as the mean of the absolute values of the individual percent differences.
c Analyte not reported.

-------
Table 12. Proficiency Testing Bias Results" for Carbonyls in 2009 NATTS Laboratories.
Laboratory
Code
Laboratory Description
FORM
ACET
Mean Abs. Bias
(across
anaiylesf
Min.
Max.
01-01-C
Rl Dept. of Health Laboratories
-5.30
-1.15
9.22
-14.3
12.0
01-02-C
Vermont DEC Environmental Lab
-23.7
-25.6
24.7
-36.3
-11.1
01-03-C
MADEP
-9.56
-10.9
10.2
-15.4
-6.40
02-01-C
NYSDEC BAQS
-13.4
-15.4
14.4
-22.9
-8.00

Philadelphia Air Management Services
-6.83
-9.37
8.10
-13.1
-5.60
03-01-C
Laboratory





03-02-C
Virginia Division of Consolidated Laboratory
-6.54
-11.0
8.79
-14.9
-6.22

Services





04-02-C
SC Dept of HEC, Div. of AQ Analysis
1.62
-7.95
4.83
-14.3
2.67
04-03-C
KY Div. of Environmental Services
-13.7
-20.0
16.8
-40.0
0.00
04-04-C
GADNR,EPD Laboratory
-7.78
-22.1
14.9
-25.7
-3.56
05-01-C
Ml DEQ Lab
-2.73
-8.00
5.37
-12.0
-0.57
05-03-C
Wisconsin DNR
-5.40
-9.15
10.3
-14.3
-2.22
06-01-C
Texas CEQ
-12.4
-17.4
14.9
-21.1
-11.6
09-03-C
Bay Area Air Quality Management District
-7.84
0.06
4.75
-11.1
1.71
09-06-C
San Diego County Air Pollution Control District
-2.74
-6.94
4.84
-13.7
-0.16
09-08-C
South Coast Air Quality Management District
-15.7
-14.2
15.0
-21.1
-10.2
10-02-C
Oregon DEQ Lab
-10.8
-14.3
12.5
-20.6
-8.00
11-01-C
ERG
-7.81
-10.1
8.93
-17.7
-2.40

Mean Abs.Bias (across laboratories)
9.66
12.7
11.2



Minimum
-36.3
-40.0




Maximum
2.67
12.0



a Computed as the mean of the individual percent differences.
b Computed as the mean of the absolute values of the individual percent differences.
were compiled: 1,3-butadiene, formaldehyde, acrolein, benzene, and arsenic. In this figure, the
bottom and top of the "box" represent the 25th and 75th percentiles, respectively; the horizontal
line inside the box represents the median value; the diamond symbol represents the mean; the top
and bottom "whiskers" extend to a length of 1.5 times the interquartile range (IQR). The IQR is
defined as the distance between the 25* and 75th percentiles of the distribution of values. The
reference line in this figure represents the MQO bias goal of 25%. To maintain figure clarity,
only labs whose results fell outside of a window defined by 1.5 x IQR are identified on the
graphical display. Selected results that fell outside of the IQR are identified by their laboratory
ID number assigned by Alion; a cross-reference between the NATTS site and assigned
laboratory codes is provided above in Tables 7 and 8. A laboratory's results were included in the
summary analysis only if the laboratory provided analysis of a particular sample type. Although
some individual laboratories report PT sample concentrations that exhibit bias beyond the
NATTS MQO, the profound majority of laboratories demonstrate laboratory biases for benzene,
1,3-butadiene, formaldehyde, and arsenic that are well within the MQO limit of ±25%. The
biases for benzene, 1,3-butadiene, and formaldehyde are slightly negative, implying a smaller
measured result than expected; biases for acrolein and arsenic are nominally positive. Percentage
participation in the PT program (Table 14) was 100% for all compound classes.
36

-------
Table 13. Proficiency Testing Bias" Results for Metals in 2009 NATTS Laboratories.








Mean Abs.










Bias


Laboratory
Code
Lab
Description
AS
BE
CD
PB
MN
Nl
(across
analytesf
Min.
Max.
01-01-M
Rl Dept. of
Health
Laboratories
-24.4
-8.46
-21.5
-36.3
-41.2
-32.9
27.5
-41.2
-8.46
03-01-M
WVDEP
Division of Air
Quality
23.7
41.6
27.7
182
3.690
25.7
50.7
3.69
182
03-02-C
Virginia Division
of Consolidated
Laboratory
Services
25.6
37.8
27.2
156
-5.30
35.6
48.0
-5.30
156
04-01 -M
Environmental
Protection
Comm. of
Hillsborough
Co.
12.5
37.3
24.1
-3.79
4.84
12.9
15.9
-3.79
37.3
04-02- M
SC Dept of
HEC, Div. of AQ
Analysis
3.95
9.67
-0.55
-5.15
0.81
-9.29
4.90
-9.29
9.67
04-03-M
KY Div. of
Environmental
Services
12.0
32.0
17.2
-6.23
-4.84
4.86
12.8
-6.23
32.0
04-04-M
GA DNREPD
Laboratory
-0.34
2.94
-6.93
32.4
-27.8
-13.9
14.0
-27.8
32.4
05-01-M
Ml DEQ Lab
-22.7
-11.9
-17.9
-32.3
-11.3
-18.6
19.1
-32.3
-11.3
05-03-M
Wisconsin DNR
-3.40
10.5
9.12
-14.1
-28.3
-13.1
13.2
-28.3
10.5
06-01-M
Texas CEQ
4.81
10.6
6.57
-15.8
-26.2
-6.21
11.7
-26.2
10.6
08-02-M
IML Air Science
2.75
19.9
6.57
0.00
-0.23
-8.00
6.24
-8.00
19.9
09-08- M
South Coast Air
Quality
Management
District
-28.2
-47.7
-19.9
-29.7
-44.0
-40.4
35.0
-47.7
-19.9
10-02-M
Oregon DEQ
Lab
-12.5
-7.08
-8.94
-32.9
-41.9
-32.9
22.7
-41.9
-7.08
11-01-M
ERG
-14.8
-5.53
-16.2
-30.6
-37.7
-28.9
22. 3
-37.7
-5.53
11-02-M
RTI
International
-1.37
3.45
-4.74
-24.5
-34.8
-29.3
16.4
-34.8
3.45

Mean Abs.Bias
13.5
20.4
15.1
37.6
20.0
20.4
21.2



(across
laboratories)










Minimum
-28.2
-47.7
-21.5
-36.3
-44.0
-40.4




Maximum
25.6
41.6
27.7
181
7.83
35.6



a Computed as the mean of the individual percent differences.
b Computed as the mean of the absolute values of the individual percent differences.
37

-------
Distribution of PERCDIFF by CMPDALIAS
P-1
g
a
BUTA
1	r
FORM	ACRO
BENZ
T
AS
Analyte
Figure 22. Distribution of Laboratory Bias by Analyte for Proficiency Testing Data from
200 .
Participation in the laboratory PT program during 200 by all NATTS-affiliated
laboratories is shown in Table 14. All participating laboratories completed the PT sample
analyses.
Table 14. Proficiency Testing Program Participation for 2009.

Percentage
Compound Class
Participation
Carbonyls
100
Metals
100
VOCs
100
2.5 Flow Audit Results from Instrument Performance Audits
Six NATTS field sites (Dearborn, MI; Deer Park, TX; Kamack, TX; Mayville, WI;
Northbrook, IL; St. Louis MO) were audited during calendar year 200 for canister, carbonyl,
PMio, chromium (VI), and PAH samplers. The IPA involves independent measurements of flow
rates on all resident sampler types at the NATTS site using certified flow, temperature, and
pressure instruments.
38

-------
Sampler flows were measured using a calibrated volumetric flow measurement device
with flow rates subsequently corrected to the standard conditions of 25 EC and 1 atm.
Comparison of the site-recorded and similarly corrected flow rate to the audited flow rate
afforded calculation of field bias. For this purpose, field bias is defined as the percentage
difference between the corrected site flow (Fsc) and the corrected audit flow (Fac):
Fs — Fa
%Difference = —		— • 100	(Eq. 4)
Fac
The results from the flow audits conducted at six sites during calendar year 2009, along
with the relevant sampling techniques, are shown in Table 15. The specific sampler audited (i.e.,
primary or collocated) is identified in column 3, with no audits performed on canister samplers.
If present during the audit, collocated samplers were also audited. Because canister and carbonyl
samplers may have multiple flow channels to facilitate duplicate sampling, all active channels
were also subjected to a flow audit. PMi0 samplers have only primary channels.
Table 15. Flow Audit Results from 2009 Instrument Performance Audits.
Site Identifier
Method
Sampler
Channel
Measurements
Percentage
Difference

Canister3
Primary
NA
Not performed0


Carbonylb
Primary
1
Site: 0.7412 L/min (actual)
Auditd: 0.7488 L/min (actual)
-1.0

Carbonyl
Duplicate/Collocated
2
Site: 0.7500 L/min (actual)
Audit: 0.7635 L/min (actual)
-1.8

PMioe
Primary
NA
Site: 41.66 ft3/min (STP)
Audit: 40.63 ft3/min (STP)
2.5
Dearborn, Ml
PMioe
Duplicate/Collocated
NA
Site: 40.71 ft3/min (STP)
Audit: 39.23 ft3/min (STP)
3.8

Cr(VI)
Primary
NA
Site: 14.4 L/min (actual)
Audit: 13.58 L/min (actual)
6.0

Cr(VI)
Duplicate/Collocated
NA
Site: 14.7 L/min (actual)
Audit: 13.8 L/min (actual)
6.6

PAH
Primary
NA
Site: 6.91 ft3/min (STP)
Audit: 7.25 ft3/min (STP)
-4.7

PAH
Duplicate/Collocated
NA
Site: 7.30 ft3/min (STP)
Audit: 7.32 ft3/min (STP)
-0.3

Canister3
Primary
NA
Not performed


Carbonylb
Primary
1
Site: 1.125 L/min (actual)
Audit: 1.09 L/min (actual)
3.2
Deer Park, TX
Carbonyl
Duplicate/Collocated
2
Site: 1.134 L/min (actual)
Audit: 1.08 L/min (actual)
5.0

PMioe
Primary
NA
Site: 40.29 ft3/min (STP)
Audit: 40.29 ft3/min (STP)
0.0

PM10
Duplicate/Collocated
NA
Site: 40.29 ft3/min (STP)
Audit: 40.41 ft3/min (STP)
-0.3
(continued)
39

-------
Table 15. Flow Audit Results from 2009 Instrument Performance Audits (continued).
Site Identifier
Method
Sampler
Channel
Measurements
Percentage
Difference

Cr(VI)
Primary
NA
Site: 11.51 L/min (actual)
Audit: 11.58 L/min (actual)
-0.6
Deer Park, TX
Cr(VI)
Duplicate/Collocated
NA
Site: 11.51 L/min (actual)
Audit: 11.58 L/min (actual)
-0.6
(continued)
PAH
Primary
NA
Site: 9.83 ft3/min (STP)
Audit: 9.84 ft3/min (STP)
-0.1

PAH
Duplicate/Collocated
NA
Site: 8.3 ft3/min (STP)
Audit: 8.4 ft3/min (STP)
-1.2

Canister3
Primary
NA
Not performed


Carbonylb
Primary
1
Site: 1.099 L/min (actual)
Auditl: 1.109 L/min (actual)
-0.9
Harrison County,
TX
Carbonyl
PM10e
Duplicate/Collocated
Primary
2
NA
Site: 1.107 L/min (actual)
Audit: 1.112 L/min (actual)
Site: 39.98 ft3/min (STP)
Audit: 39.96 ft3/min (STP)
o o

Cr(VI)
Primary
NA
Site: 11.86 L/min (actual)
Audit: 12.13 L/min (actual)
-2.2

PAH
Primary
NA
Site: 8.57 ft3/min (STP)
Audit: 8.49 ft3/min (STP)
0.9

Canister3
Primary
NA
Not performed


Carbonylb
Primary
1
Site: 0.6997 L/min (actual)
Audit: 0.7356 L/min (actual)
-4.9

Carbonyl
Duplicate/Collocated
2
Site: 0.7 L/min (actual)
Audit: 0.7196 L/min (actual)
-2.7

PM10e
Primary
NA
Site: 32.96 ft3/min (STP)
Audit: 32.16 ft3/min (STP)
2.5
Mayville, Wl
PM10e
Duplicate/Collocated
NA
Site: 32.96 ft3/min (STP)
Audit: 32.59 ft3/min (STP)
1.1

Cr(VI)
Primary
NA
Site: 14.7 L/min (actual)
Audit: 15.15 L/min (actual)
-3.0

Cr(VI)
Duplicate/Collocated
NA
Site: 14.7 L/min (actual)
Audit: 15.01 L/min (actual)
-2.1

PAH
Primary
NA
Site: 6.36 ft3/min (STP)
Audit: 6.52 ft3/min (STP)
-2.5

PAH
Duplicate/Collocated
NA
Site: 6.61 ft3/min (STP)
Audit: 6.61 ft3/min (STP)
0.0

Canister3
Primary
NA
Not performed


Carbonylb
Primary
1
Site: 1.18 L/min (actual)
Audit: 1.21 L/min (actual)
-2.5

PM10e
Primary
NA
Site: 41.7 ft3/min (STP)
Audit: 42.8 ft3/min (STP)
-2.6
Northbrook, IL
Cr(VI)
Primary
NA
Site: 14.4 L/min (actual)
Audit: 13.86 L/min (actual)
3.9

Cr(VI)
Duplicate/Collocated
NA
Site: 15.1 L/min (actual)
Audit: 14.76 L/min (actual)
2.3

PAH
Primary
NA
Site: 8.08 ft3/min (STP)
Audit: 8.23 ft3/min (STP)
-1.8
(continued)
40

-------
Table 15. Flow Audit Results from 2009 Instrument Performance Audits (continued).
Site Identifier
Method
Sampler
Channel
Measurements
Percentage
Difference

Canister3
Primary
NA
Not performed


Carbonylb
Primary
1
Site: 0.7493 L/min (actual)
Audit:0.771 L/min (actual)
-2.8

Carbonylb
Collocated/Duplicate
1
Site: 0.8897 L/min (actual)
Audit:0.9222 L/min (actual)
-3.5
St. Louis, MO
PMioe
Primary
NA
Site: 41.32 ft3/min (STP)
Audit: 39.71 ft3/min (STP)
4.1

Cr(VI)
Primary
NA
Site: 14.26 L/min (actual)
Audit: 14.31 L/min (actual)
-0.3

Cr(VI)
Duplicate/Collocated
NA
Site: 14.72 L/min (actual)
Audit: 14.46 L/min (actual)
1.8

PAH
Primary
NA
Site: 7.77 ft3/min (STP)
Audit: 7.78 ft3/min (STP)
-0.1
a VOC sampler.
b Carbonyl cartridge.
d Performed by RTI International.
11 Audit not performed for this sampler type.
6 Filter sample for PMi0 metals.
A graphical summary of the flow audit results is presented in Figure 23. All flow rate
measurements were within ±10% of the audit flow rate; most were within 5%.
Accuracy of flow rates for carbonyl and PMio samplers is critical for determining sample
concentration. Conversely, because only an aliquot of the canister volume is analyzed, the
accuracy of canister sampler flow rates is less important. However, a constant flow rate across
the 24-hour sampling interval is critical to achieving a linearly representative integrated sample.
The field bias audit of a VOC sampler flow rate is a random check of this time-integrated value.
2.6 Method Detection Limit Data
During compilation of 2007 QA data, sunstantial effort was invested in acquiring the
MDL data through direct contacts with each contributing laboratory. For the 2008 and 2009
results, the AQS database, specifically the ALTMDL variable in the RD record types, served as
the primary source of laboratory-based MDL data. Because this is not a required field in AQS,
approximately 85% of the MDL data were acquired from this source. Because AQS allows the
posting of MDL data in a variety of units, even within chemical classes, all AQS-acquired MDLs
3	3
were standardized to ng/m for metals, PAHs, and chromium (VI) and (J,g/m for carbonyls and
VOCs. The balance of the MDLs (i.e., those values not posted to AQS) was requested from
direct contact with each laboratory known to be providing analytical services. Multiple e-mail
requests with some laboratory contacts were needed to obtain the full complement of MDL data.
After careful review of the received materials from each laboratory, the spreadsheet information
was compiled into a database from which subsequent data analyses could be performed.
41

-------
Sample Type	Site Id.
Carbonyl Dearborn, MI
Deer Park, TX
Karnack, TX
Mayvile, Wl
Norttibrook, IL
St Louis, MO
Cr (VI) Dearborn, MI
Deer Park, TX
Karnack, TX
Mayvile, Wl
Narttibrook, IL
St Louis, MO
PAH	Dearborn, MI
Deer Park, TX
Karnack, TX
Mayvile, Wl
Narttibrook, IL
St Louis, MO
PM10 Dearborn, MI
Deer Park, TX
Karnack, TX
Mayvile, Wl
Narttibrook, IL
St Louis, MO
Figure 23. Summary of Instrument Performance Flow Audit Results for 2009.
For this report and by generally accepted conventions, MDLs are defined as the detection
threshold for a given analyte based on the mathematical combination of all aspects of the sample
collection and analysis process. Thus, they reflect, among other factors, the collected sample
volume for each sample, the size of the subsample subjected to analysis, and any sample
dilutions that may be associated with the analysis methodology. Using the AQS database as the
primary source of the MDL information does not, in and of itself, ensure consistency of the data,
but consistency of the data derived largely from posted information is considered vastly
improved over the same data obtained through individual laboratory requests. There is, however,
no unequivocal way to discern from the existing data if the MDLs provided reflect the MDL
(i.e., taking into account sampling and analysis components) or if they reflect only instrumental
detection limits. These concerns notwithstanding, the MDL results presented in this report are
mean values computed from either individual AQS-posted values or directly from laboratory
contacts and are presented under the assumption that each laboratory reported actual method
detection limits that incorporated both instrumental and sampling considerations. In cases where
the data were acquired by direct laboratory contact and unit conversions were needed, the data
were converted to the same units specified above. The MDL data for individual sites, in addition
to the mean across all sites reporting data, are shown in Table 16.
Box and whisker plots and complementary scatter plots, shown in Figures 24 through 28,
illustrate the MDLs for carbonyls, metals, arsenic, VOCs, and PAHs, respectively. The MQOs
for benzene, 1,3-butadiene, formaldehyde, and arsenic are added to each plot for reference. Labs
whose results fell outside of a window defined by 1.5 x IQR are identified by blue asterisks on
42

-------
Table 16. Method Detection Limits by Site and Overall for Calendar Year 2009 (VOCs and Carbonyls: jig/m3; Metals: ng/m3).
Site Name
AQS Site Code
BENZ
BUTA
CTET
CLFRM
EDB
DCP
EDC
MECL
TCE 1122
PERC
TCE
VCM
c DCPEN
t DCPEN
Phoenix, AZ
04-013-9997
0.125
0.122
0.014
0.011
0.008
0.015
0.009
0.03
0.023
0.022
0.012
0.006
0.015
0.015
Los Angeles, CA
06-037-1103
	b
0.176
0.63
0.355
0.77
0.463
0.406
0.348
	b
0.469
0.39
0.256
0.455
0.455
Rubidoux, CA
06-065-8001
	b
0.172
0.63
0.344
0.77
0.463
0.406
0.348
	b
0.452
0.378
0.256
0.455
0.455
San Jose, CA
06-085-0005
0.112
0.113
0.063
0.068
0.077

0.406
0.348
	b
0.047
0.074
0.256
0.455
0.455
Grand Junction, CO
08-077-0017
b
b
b
b
b
b
b
b
b
b
b
b
b
b
Grand Junction, CO
08-077-0018
0.021
0.007
0.013
0.01
0.008
0.014
0.008
0.028
0.021
0.02
0.011
0.005
0.014
0.014
Washington, DC
11-001-0043
0.059
0.055
0.157
0.122
0.23
0.138
0.121
0.121
0.239
0.203
0.134
0.076
0.067
0.113
Hillsborough County, FL
12-057-3002
0.071
0.053
0.139
0.083
0.123
0.083
0.061
0.084
0.089
0.122
0.108
0.051
0.068
0.073
Pinellas County, FL
12-103-0026
0.071
0.053
0.139
0.083
0.123
0.083
0.061
0.084
0.089
0.122
0.108
0.051
0.068
0.073
Decatur, GA
13-089-0002
0.119
0.074
0.053
0.097
0.142
0.16
0.113
6.959
0.169
0.149
0.231
0.052
0.116
0.11
Northbrook, IL
17-031-4201
0.125
0.122
0.013
0.01
0.008
0.014
0.008
0.028
0.021
0.02
0.011
0.005
0.014
0.014
Grayson Lake, KY
21-043-0500
0.128
0.2
0.189
0.098
0.231
0.185
0.284
0.07
0.275
0.204
0.108
0.256
0.182
0.227
Boston, MA
25-025-0042
0.022
0.014
0.055
0.037
0.087
0.034
0.043
0.047
0.235
0.059
0.044
0.022
0.029
0.022
Dearborn, Ml
26-163-0033
0.021
0.007
0.013
0.01
0.008
0.014
0.008
0.028
0.021
0.02
0.011
0.005
0.014
0.014
St. Louis, MO
29-510-0085
0.021
0.007
0.013
0.01
0.008
0.014
0.008
0.028
0.021
0.02
0.011
0.005
0.014
0.014
Bronx, NY
36-005-0110
0.032
0.044
0.063
0.049
0.077
0.093
0.041
0.035
0.069
0.068
0.054
0.026
0.045
0.045
Rochester, NY
36-055-1007
0.032
0.044
0.063
0.049
0.077
0.093
0.041
0.035
0.069
0.068
0.054
0.026
0.045
0.045
Portland, OR
41-051-0246
0.143
0.222
0.315
0.245
b
0.232
b
0.261
b
0.34
0.269
0.154
b
b
La Grande, OR
41-061-0119
0.143
0.222
0.315
0.245
b
0.232
b
0.261
b
0.34
0.269
0.154
b
b
Providence, Rl
44-007-0022
0.024
0.014
0.055
0.037
0.087
0.034
0.043
0.047
0.235
0.059
0.044
0.022
0.029
0.022
Chesterfield, SC
45-025-0001
0.576
0.532
1.072
0.44
1.925
0.741
2.434
0.626
0.895
1.155
0.969
0.41
0.455
0.546
Deer Park, TX
48-201-1039
0.864
0.599
1.702
1.028
1.54
0.787
1.095
0.487
1.376
1.631
1.562
0.435
0.91
0.91
Harrison County, TX
48-203-0002
0.864
0.599
1.702
1.028
1.54
0.787
1.095
0.487
1.376
1.631
1.562
0.435
0.91
0.91
Bountiful, UT
49-011-0004
0.125
0.122
0.013
0.01
0.008
0.014
0.008
0.028
0.021
0.02
0.011
0.005
0.014
0.014
Underhill, VT
50-007-0007
0.019
0.02
0.027
0.021
0.049
0.018
0.017
0.034
0.085
0.028
0.02
0.015
0.02
0.026
Richmond, VA
51-087-0014
0.193
0.069
0.195
0.176
0.416
0.144
0.089
0.153
0.275
0.231
0.14
0.087
0.209
0.136
Seattle, WA
53-033-0080
0.021
0.007
0.013
0.01
0.008
0.014
0.008
0.028
0.021
0.02
0.011
0.005
0.014
0.014
Mayville, Wl
55-027-0007
0.357
0.222
0.63
0.489
0.77
0.463
0.406
0.348
0.688
0.68
0.539
0.256
0.455
0.455
Geometric Mean

0.089
0.071
0.104
0.076
0.103
0.091
0.075
0.114
0.121
0.122
0.091
0.049
0.08
0.081
Arithmetic Mean

0.179
0.147
0.318
0.202
0.379
0.215
0.293
0.419
0.304
0.317
0.274
0.128
0.212
0.217
Standard Deviation

0.24
0.169
0.473
0.279
0.546
0.252
0.531
1.294
0.415
0.456
0.425
0.143
0.271
0.274
Median

0.116
0.093
0.101
0.083
0.105
0.093
0.061
0.084
0.089
0.122
0.108
0.051
0.068
0.073
(continued)

-------
Table 16. Method Detection Limits by Site and Overall for Calendar Year 2009 (VOCs and Carbonyls: jig/m3; Metals: ng/m3)
	(additional analytes) (continued).	
Site Name
AQS Site Code
ACRY
ACRO
FORM
ACET
NAPH
BaP
AS
BE
CD
PB
MN
Nl
CRVI
Phoenix, AZ
04-013-9997
0.036
0.038
0.046
0.042
0.384
0.092
0.014
0.03
0.226
0.07
0.231
1.066
0.004
Los Angeles, CA
06-037-1103
0.653
0.388
b
b
0.378
0.093
b
b
b
b
b
b
b
Rubidoux, CA
06-065-8001
0.653
0.401
0.123
0.181
0.28
0.069
b
b
b
b
b
b
b
San Jose, CA
06-085-0005
0.324
0.426
0.123
0.181
0.337
0.082
0.024
0.002
0.046
0.438
0.378
0.658
	b
Grand Junction, CO
08-077-0017
	b
	b
	b
	b
	b
	b
1.583
0.205
0.179
1.112
0.392
0.45
0.004
Grand Junction, CO
08-077-0018
0.033
0.034
0.01
0.009
0.408
0.099
	b
	b
	b
	b
	b
	b
	b
Washington, DC
11-001-0043
0.054
b
0.026
0.029
0.383
0.093
1.18
0.18
0.31
0.84
0.19
0.78
0.005
Hillsborough County, FL
12-057-3002
0.061
0.143
0.011
0.01
0.336
0.081
0.46
0.2
0.15
1.04
0.14
0.92
0.005
Pinellas County, FL
12-103-0026
0.061
0.143
0.01
0.009
0.294
0.072
0.46
0.2
0.15
1.04
0.14
0.92
0.005
Decatur, GA
13-089-0002
	b
0.052
1.079
1.079
0.332
0.081
0.268
0.018
0.009
0.027
0.036
0.059
0.005
Northbrook, IL
17-031-4201
0.033
0.034
0.006
0.005
0.278
0.067
0.009
0.002
0.029
0.056
0.057
0.132
0.005
Grayson Lake, KY
21-043-0500
	b
0.391
0.008
0.022
0.277
0.067
0.413
0.253
0.243
0.247
0.414
0.578
0.4
Boston, MA
25-025-0042
b
0.116
0.08
0.108
0.258
0.062
0.009
0.002
0.029
0.056
0.057
0.138
0.005
Dearborn, Ml
26-163-0033
0.033
0.034
0.009
0.009
0.324
0.079
0.173
0.244
0.166
0.22
0.412
0.16
0.005
St. Louis, MO
29-510-0085
0.033
0.034
0.01
0.009
0.27
0.067
0.016
0.002
0.037
0.816
0.108
0.517
0.005
Bronx, NY
36-005-0110
	b
0.069
0.012
0.018
0.302
0.075
0.521
0.521
0.26
0.26
0.521
0.521
0.005
Rochester, NY
36-055-1007
	b
0.069
0.012
0.018
0.231
0.056
0.521
0.521
0.26
0.26
0.521
0.521
0.005
Portland, OR
41-051-0246
	b
	b
0.118
0.03
1.206
0.251
0.033
0.003
0.033
0.334
0.334
0.334
0.033
La Grande, OR
41-061-0119
b
b
0.122
0.03
1.801
0.255
0.037
0.004
0.037
0.367
0.367
0.367
0.036
Providence, Rl
44-007-0022
0.265
0.116
0.046
0.027
0.381
0.094
0.057
0.014
0.016
0.024
0.026
0.017
0.005
Chesterfield, SC
45-025-0001
b
1.953
0.249
0.197
0.347
0.085
0.031
0.001
0.001
0.003
0.002
0.003
0.005
Deer Park, TX
48-201-1039
	b
0.23
0.098
0.145
0.341
0.083
0.015
0.18
0.098
1.3
0.57
2.3
0.012
Harrison County, TX
48-203-0002
	b
0.23
0.098
0.145
	b
	b
0.015
0.18
0.098
1.3
0.57
2.3
0.012
Bountiful, UT
49-011-0004
0.033
0.034
0.009
0.008
0.431
0.105
0.031
0.045
0.19
0.098
0.205
0.835
0.004
Underhill, VT
50-007-0007
0.037
0.04
0.014
0.017
0.33
0.081
0.011
0.001
0.066
0.591
0.258
1.223
0.005
Richmond, VA
51-087-0014
0.089
0.179
0.055
0.043
0.379
0.092
0.037
0.011
0.007
0.032
0.042
0.312
0.005
Seattle, WA
53-033-0080
0.033
0.034
0.014
0.013
0.276
0.067
0.01
0.003
0.027
0.062
0.073
0.138
0.005
Mayville, Wl
55-027-0007
	b
0.23
0.331
1
0.283
0.069
0.028
0.01
0.011
0.004
0.015
0.093
0.005
Geometric Mean

0.075
0.116
0.04
0.042
0.358
0.085
0.062
0.023
0.053
0.159
0.138
0.304
0.007
Arithmetic Mean

0.152
0.226
0.115
0.162
0.412
0.092
0.23
0.109
0.104
0.409
0.237
0.594
0.023
Standard Deviation

0.214
0.383
0.217
0.317
0.329
0.048
0.388
0.152
0.097
0.441
0.189
0.609
0.079
Median

0.046
0.116
0.046
0.029
0.332
0.081
0.032
0.016
0.056
0.254
0.197
0.484
0.005
' Meets MQO.
b Not reported.

-------
Carbonyl Analyte
0.250
CO
"I
3
c
o
B
¦o
o
0.188
0.125
0.063
0.000
X
ACET
X
I
FORM
AAA Boston, MA
~ TT Bronx, NY
~	~~ Chesterfield, SC
Deer Park, TX
Grayson Lake, KY
+ + + Harris County, TX
¦ ¦ ¦ Hillsborough County, FL
~	~ ~ Pinelas County, FL
•	•• Rochester, NY
XXX Washington, DC
Carbonyl Analyte
Figure 24. Distribution of Method Detection Limits for Carbonyls for 2009 NATTS Data
(dashed line indicates MQO target MDL for formaldehyde; > 1.5 x IQR are identified as
blue stars in top display).
45

-------
#
<5>

Metal Analyte

3.00
CO
"I
3
2.25
c
o
B
¦o
o
1.50
0.75
*
0.00 *
X
$
~
BE
CD
¦
ir
i
~
*
_ _•
*
CRVI
MN
Nl
PB
AAA Boston, MA
~ TT Bronx, NY
~	~~ Chesterfield, SC
Deer Park, TX
Grayson Lake, KY
+ + + Harris County, TX
¦ ¦ ¦ Hillsborough County, FL
~	~ ~ Pinelas County, FL
•	•• Rochester, NY
XXX Washington, DC
Metal Analyte
Figure 25. Distribution of Method Detection Limits for Metals for 2009 NATTS Data
(dashed line indicates MQO target MDL for arsenic; > 1.5 x IQR are identified as blue
stars in top display).
46

-------
1.25-
co
E
"Si
c
1.00
c
o
B
a>
K
Q
"D
O
J=
ft
0.75
0.50
0.25
&
Metal Analyte
1.2
CO
E:
O)
c
c
o
¦O
o
£
0.9
0.6
0.3
A
0.0 "
AAA Boston, MA
~ TT Bronx, NY
~	~A ChBsterfield, SC
Deer Park, TX
Grayson Lake, KY
+ + + Harris County, TX
¦ ¦ ¦ Hillsborough County, FL
~	~ ~ Pinelas County, FL
•	•• Rochester, NY
XXX Washington, DC
AS
Metal Analyte
Figure 26. Distribution of Method Detection Limits for Arsenic for 2009 NATTS Data
(dashed line indicates MQO target MDL for arsenic.
47

-------
E 2.0
D)
3
1.5
1.0
0.5
CO
ll
3
c
o
'¦£3
a
¦a
o
3.00
2.25
1.50
0.75
0.00

VOC Analyte


v


~
*




*



*
*
~




*










#





1

f
C
1
1
1
%
L
I
1
1
1
1
A
1
A
1
B
1
B
1
C
1
C
1
D
1
E
i
E
1
M
1
P
1
T
1
V
C
C
E
U
L
T
C
D
D
E
E
C
c
R
R
N
T
F
E
P
B
C
C
R
E
M
O
Y
Z
A
R
M
T



L
C


~ ~~
TTT
»»»
+ + +
¦ ¦¦
~ ~ ~
XXX
###
Boston, MA
Bronx, NY
CfiBsteffield, SC
Deer Park, TX
Grayson Lake, KY
Harris County, TX
Harrison County, TX
Hillsborough County, FL
Pinelas County, FL
Rochester, NY
Washington, DC
VOC Analyte
Figure 27. Distribution of Method Detection Limits for VOCs for 2009 NATTS Data
(dashed line indicates MQO target MDL for benzene; > 1.5 x IQR are identified as blue
stars in top display).
48

-------
0.4
E
0.3
E
~
|
a>
Q
TJ
o
ft
0.1

/

0.390
0.305
CO
.E
O)
c
o 0.220
¦8
¦O
o
PAH Analyte
x
4
~
~
0.135
0.050
~
x
x
w
•
I
BaP
NAPH
AAA Boston, MA
~ TT Bronx, NY
~	~A ChBsterfield, SC
Deer Park, TX
Grayson Lake, KY
+ + + Harris County, TX
¦ ¦ ¦ Hillsborough County, FL
~	~ ~ Pinelas County, FL
•	•• Rochester, NY
XXX Washington, DC
PAH Analyte
Figure 28. Distribution of Method Detection Limits for PAHs for 2009 NATTS Data.
the graphical display. The IQR is defined as the distance between the 25th and 75th percentiles of
the distribution of values.
49

-------
Review of the graphically displayed MDL results reveals a number of interesting features
largely consistent with previous reporting years. MDLs for carbonyls (Figure 24) show
appreciable spread across laboratories but almost universally greater than the MQO. The
proximity of the mean MDL to the bottom of the distribution reflects the MDLs for many
laboratories close to, albeit above, the MQO; values for four laboratories are significantly higher
for both formaldehyde and acetaldehyde. Metals results in Figures 25 and 26 show MDL values
falling within the MQO for some analytes, notably Be, Cd, and chromium (VI), but substantially
above for others (Mn, Ni, Pb), with relatively few values outside the IQR. Arsenic performance
was well within the MQOs for some laboratories but substantially outside for others. The
consistency and magnitude of MDLs reported for chromium (VI) is particularly noteworthy and
may reflect the fact that only three laboratories are performing this analysis for all NATTS sites.
VOCs show much greater variability in MDLs across laboratories than other analyte groups, with
two laboratories (three sites) accounting for most of the spread in the distribution (Figure 27). As
was found for most analytes, a high proportion of MDLs for VOCs occurred above the MQO.
Lastly, MDLs for PAHs, while universally above the MQO, were highly clustered for both
benzo[a]pyrene and naphthalene, again reflecting that the analysis was performed by only three
labs (Figure 28).
As reported by the metals analysis laboratories for 2008, 19 NATTS sites (San Jose, CA;
Washington, DC; Boston-Roxbury, MA; Decatur, GA; Hillsborough County, FL; Pinellas
County, FL; Dearborn, MI; Mayville, WI; Northbrook, IL; Harrison County, TX; St. Louis, MO;
La Grande, OR; Portland, OR; Seattle, WA; Providence, RI; Chesterfield, SC; Deer Park, TX;
Underhill, VT; Richmond, VA) collected high-volume PMi0 metals on 8 in. x 10 in. quartz fiber
filters. Seven sites reported using low-volume PMio metals sampling on 47 mm Teflon filters
(Bronx, NY; Rochester, NY; Bountiful, UT; Grand Junction, CO; Phoenix, AZ; Hazard, KY;
Grayson Lake, KY). The remaining sites either did not collect PMio samples for metals analysis
or did not report the type of sampling implemented.
Comparison of MDLs for the two sampling approaches is meaningful only when the
analysis laboratory is the same for the two sites; otherwise the variability in MDLs is an
aggregate effect of sample collection and sample analysis. The metals results provided by the
ERG laboratory, which analyzes samples of both types, offer a unique opportunity to examine
MDLs between high- and low-volume sampling without the influence of cross-laboratory
instrumental detection limit variability. Table 17 shows the MDLs for each of the PMio metal
analytes. The enhanced MDLs for the higher volume samples are consistent with the 100-fold
increase in sample size over the lower volume samples. At the direction of EPA, the computation
of the MDLs by the laboratories for quartz sampling media was changed, allowing for
adjustment of the significantly higher background levels for this collection medium. Overall,
variability in MDLs among laboratories, shown in Table 18, is very large and suggests
significant differences in analytical performance as well as collection volumes.
50

-------
Table 17. Comparison of Method Detection Limits Reported by ERG Laboratory for
Metals between High- and Low-Volume Samplers in Calendar Year 2009.

Method Detection Limits (ng/m3)
Median (Std. Dev.)
MDL Ratio
(High/Low)
Analyte
2000 m3
Samples3
20 m3
Samples'1
Arsenic
0.0090 (0.0239)
0.0200 (0.0169)
0.45
Beryllium
0.0020 (0.0014)
0.0300 (0.0259)
0.067
Cadmium
0.0290 (0.0185)
0.2300 (0.0713)
0.13
Chromium (VI)
0.0046 (0.0008)
0.0043 (0.0006)
1.07

0.0170 (0.0170)
1.1600 (0.0381)
0.015
Manganese
0.0570 (0.2795)
0.2400 (0.0630)
0.24
Nickel
0.1320 (0.6573)
1.0300 (0.4958)
0.13
Lead
0.0560 (0.7785)
0.0700 (0.0677)
0.80
'Based on six sites conducting high-volume PMi0 sampling.
b Based on two sites conducting low-volume PMi0 sampling.
The geometric mean MDLs (Table 18) for the select analytes—benzene, 1,3-butadiene,
and formaldehyde—do not meet the target MQO for MDLs. Conversely, the MDL for arsenic
falls within the target MQO.
Table 18. Summary Statistics for Method Detection Limits across All Reporting NATTS
Laboratories for 2009.


Selected Analyte

MDL
Benzene, (pg/m3)
1,3-butadiene,
(Mg/m3)
Formaldehyde,
(Mg/m3)
Arsenic, (ng/m3)
Geometric Mean
0.063
0.0369
0.0402
0.0639
Arithmetic Mean
0.103
0.0773
0.1132
0.5119
Standard Deviation
0.102
0.0786
0.2377
1.1656
Minimum
0.010
0.0067
0.0049
0.0090
Median
0.071
0.0532
0.0400
0.0340
Maximum
0.357
0.2217
1.0800
5.3311
MQO
0.016
0.013
0.0074
0.217
Ratio of Geo. Mean
to MQO
3.9
2.8
5.4
0.3
3.0 SUMMARY
Based on four HAPs representative of the various chemical classes—benzene, 1,3-
butadiene, formaldehyde, and arsenic, the following summary comments are appropriate for the
2009 NATTS data.
51

-------
1.	Excluding NATTS sites intentionally not collecting data for a particular analyte class
(e.g., PMio metals), the mean completeness percentages of data reported into AQS
across all NATTS sites were 96%, 97%, 98%, 96%, 99%, 99%, and 98% for benzene,
1,3-butadiene, acrolein, naphthalene, formaldehyde, chromium (VI), and arsenic,
respectively. Completeness statistics reported in 2009 for naphthalene and chromium
(VI) were dramatically improved over those reported in 2008, exceeding the MQO
for nearly every site. Overall, the MQO was achieved for all seven analytes.
2.	With a few exceptions as noted in the text of this report, analytical precision among
sites for which replicate analyses were available was found to be below the 15%
MQO threshold for all analytes used to reflect their respective chemical classes.
Analytical precision for acrolein was somewhat more variable than other analytes but
still within the 15% threshold. As expected, the frequency of cases where the MQO
threshold was exceeded was distinctly greater for overall precision (i.e., including
sampling and analysis) among all analytes and particularly for acrolein, an analyte
presenting unique collection and analysis challenges. Estimates of overall precision
included both duplicate and collocated samples.
3.	Laboratory performance, as assessed by the percentage difference between the
laboratory measurement and the certified sample concentration of the proficiency
testing samples, was within the ±25% MQO for most analytes (i.e., benzene, 1,3-
butadiene, formaldehyde, and arsenic) and for laboratories with available data from
2009. The poorest performance across all laboratories and analytes was observed for
lead (37.6%>) and acrolein (30.3%). The proportion of laboratories participating in the
2009 performance testing program was 100% for all chemical classes—a significant
improvement over participation in 2008. Laboratories not performing analyses of a
particular analyte were excluded from these statistics.
4.	Without exception, sampler flows measured during IP As conducted at NATTS field
sites showed less than ±10% difference from their site-recorded values.
5.	Among all measures of data quality, MDLs were substantially greater than the
corresponding MQOs and showed substantial variability for any given analyte across
sites (i.e., laboratories). Only the mean value for arsenic fell within the MQO
threshold when all laboratories were considered together. The ratios of the cross-
network geometric means to the corresponding MQOs were 3.9, 2.8, 5.4, and 0.3 for
benzene, 1,3-butadiene, formaldehyde, and arsenic, respectively.
4.0 RECOMMENDATIONS
The information, both analytical results and site characteristics, for the NATTS network
samples present in the AQS database was acquired successfully, based on a thorough
understanding of the database structure. Based on knowledge of POC assignments in previous
years, the POCs for the primary, duplicate, and collocated samples were assigned with greater
facility than in previous years. However, as in previous years, acquiring MDL data for
laboratories not posting MDLs to AQS directly was still problematic.
The POCs are present in the AQS database, but the associated sample type information
(e.g., primary, duplicate, or collocated) is not. Because POCs are assigned by either the agency
monitoring a particular NATTS site or the laboratory uploading the data to AQS, and are largely
52

-------
nonstandardized across NATTS sites [6] (see Table 6), the inclusion of a field in the AQS
database to specify whether a particular POC is "primary," "duplicate," or "collocated" would be
of enormous benefit to the utility of the AQS data and would greatly streamline the analyses
reported here.
Summary statistics created for this report reflect the overall condition of the data but
may, in some cases, be unduly influenced by selected extreme values. Instances where the
summary statistics fall outside of the MQOs warrant further investigation of the individual data
points as deemed appropriate by EPA.
The acquisition and assembly of MDL information was again aided dramatically through
the extraction of the ALTMDL field for RD records in the AQS database. Only instances where
this optional field was not populated by the contributing laboratory (-15%) required direct
contacts with individual laboratory supervisors. Changing the character of this AQS field to
"required" would completely eliminate the need for this follow-up step. Lastly, AQS accepts
data in a variety of units at the discretion of the agency performing the upload. This requires very
careful scrutiny of the UNIT variable so that MDL measurements can be standardized
algebraically prior to data analysis. Standardization of MDLs posted in the ambiguous "ppbC"
unit is particularly problematic.
5.0 REFERENCES
1.	U.S. Environmental Protection Agency. (2007). National Air Toxics Trends Stations Quality
Assurance Annual Report - Calendar Year 2006. Prepared by Battelle Memorial Institute.
2.	U.S. Environmental Protection Agency. (2009). National Air Toxics Trends Stations Quality
Assurance Annual Report - Calendar Year 2007. Prepared by RTI International.
3.	U.S. Environmental Protection Agency. (2010). National Air Toxics Trends Stations Quality
Assurance Annual Report - Calendar Year 2008. Prepared by RTI International.
4.	U.S. Environmental Protection Agency. (October 17, 2008). AQS Data Coding Manual
(Version 2.33). Available at
http://www.epa.gov/ttn/airs/airsaqs/manuals/AQS%20Data%20Coding%20Manual.pdf
5.	Eastern Research Group. (January 1, 2007). Technical Assistance Document for the National
Ambient Air Toxics Trends and Assessment Program.
6.	U.S. Environmental Protection Agency. (July 2004). Final Draft, July 2004, National
Monitoring Strategy, Air Toxics Component. Available at
http://www.epa.gov/ttn/amtic/files/ambient/airtox/atstrat804.pdf
53

-------