United States Office of Air Quality EPA-454/R-02-004
Environmental Protection Planning and Standards April 2002
Agency Research Triangle Park, NC 27711
Office of Air and Radiation
Air Toxics Pilot Study Laboratory
Intercomparison
Quality Assurance Report
-------
A cknowledgments
This report is the product of the EPA Office of Air Quality Planning and Standards. The following
individuals are acknowledged for their contributions.
Author
Dennis Mikel, OAQPS-EMAD-MQAG, Research Triangle Park, North Carolina
Reviewers
Office of Air Quality Planning and Standards
Joann Rice, Sharon Nizich
Contributions
The author would like to acknowledge and thank Mr. Michael Miguel and his staff of the California Air
Resources Board for providing the analysis and coordination of the canisters for the Volatile Organic
Compounds and Mr. Rudy Eden and his staff of South Coast Air Quality Management District for
providing the filter strips for the metals analysis.
Comments and questions can be directed to:
Dennis Mikel, EPA-OAQPS-EMAD, Mail Drop C339-02, Research Triangle Park, NC 27711
email address: mikel .dennisk@,epa.gov Phone Number: 919-541-5511.
-------
Number
Figures
Description PaSe No-
I Data Recovery of Pilot Labs
2 Air Toxics Pilot 200 1 - VOCs Reported ^
3 Distribution of Compounds Analyzed
4 1,3 -butadiene
5 Carbon Tetrachloride
6 Tetrachloroethylene (PCE) '
7 Benzene
g Dichloromethane
9 Chloroform
10 1,2-dichloropropane
n Vinyl Chloride 9
12 Trichloroethylene
13 Metals- Round Robin |^
14 Metals Round Robin - Beryllium
15 Metals Round Robin - Cadmium
15 Metals Round Robin - Chromium
17 Metals Round Robin - Lead ;^
18 Metals Round Robin - Manganese
19 Metals Round Robin - Nickel
in
-------
Executive Summary
A laboratory "round robin" intercomparison was conducted by the US Environmental Protection
Agency (EPA) during the National Air Toxics Pilot Study (NATPS). Volatile Organic
Compounds (VOCs) and elemental metallic compounds were collected and distributed to the
laboratories that participated in the study. The data were sent to EPA's Office of Air Quality,
Planning and Standards (OAQPS). The results were tabulated and statistically analyzed. The
following statements can be made about the results of the intercomparison:
For the VOCs, an issue has been identified concerning the identification and reporting of
compounds. EPA will need to establish clear guidelines on what and how compounds
are identified and reported.
• 1,3-butadiene values reported on the west coast were significantly different from the
values reported on the east coast. Stability of this compound may be a factor. However,
several laboratories did not report any values for this compound, therefore more data are
needed from across the nation to discern whether this is a problem.
• Most of the compounds that were statistically analyzed were within 2 standard
deviations of the mean.
• The Hillsborough County laboratory reported a beryllium value approximately 100 times
the mean of the other values. It is believed that there may have been contamination or a
dilution error with this metals sample.
Background
To address the concerns about the prevalence of air toxics emissions and to meet EPA's
strategic goals, a national air toxics program has been designed to characterize, prioritize, and
equitably address the impacts of Hazardous Air Pollutants (HAPs) on the public health and
environment. The national air toxics program seeks to address air toxics problems through a
combination of activities and authorities, including regulatory approaches and voluntary
partnerships. One of the key activities is the National Air Toxics Assessment (NAT A). NATA
activities will help EPA identify areas of concern, characterize human health and ecosystem
risks and track progress of trends.
As outlined in the air toxics monitoring "Concept Paper1," the role of ambient monitoring is to
support NATA activities includes:
characterization of ambient concentrations and deposition in representative monitoring
areas;
provide data to support and evaluate dispersion and deposition models; and
• establish trends and evaluate the effectiveness of HAP reduction strategies.
The Concept Paper identifies 18 "core" compounds that have been identified through the
NATA as being of particular risk to the health the nations. These are: Benzene, 1,3-butadiene,
carbon tetrachloride, chloroform, 1,2-dichloropropane, dichloromethane, tetrachloroethylene,
trichloroethylene, vinyl chloride, beryllium, cadmium, chromium, lead, manganese, nickel,
acetaldehyde, formaldehyde and arsenic. Arsenic was not included in this intercomparison due
to the difficulty of analysis. Continued method development research is required for arsenic.
-------
To help further our understanding of monitoring the core HAPS, EPA embarked on the NATPS
in CY 2001. This initial pilot monitoring together with data analysis of existing measurements
was needed to provided information on spatial and temporal variability of ambient air toxics.
This information will aid in providing state and local air agencies important information about
their particular network needs. The pilot monitoring program will also provide very useful
information to help the EPA design a long-term national air toxics monitoring network.
In order to provide consistency in the data set generated by the NATPS, a laboratory work
group was formed to discuss the details regarding procedures to be used for measurements. This
group of laboratory, State and local, Regional and EPA representatives recommended in early
2001 to create a "round robin " interlaboratory program that would provide Performance
Evaluation samples (PEs) to the laboratories that are involved in the NATPS.
The purpose of this document is to discuss the procedures used and the results of the
interlaboratory comparison.
Procedures
For the Volatile Organic Compounds, (VOCs) the California Air Resources Board (CARB)
collected ambient air samples in SUMMA™ passivated 6 liter canisters from the following
laboratories: Bay Area Air Quality Management District (BAAQMD), State of Michigan
Department of Environmental Quality (MDEQ), Pinellas County Air Quality Division
(PCAQD), Maryland Department of Environmental Laboratory (MDEL), Rhode Island
Department of Health (RIDOH), Eastern Research Group (ERG), New Mexico Department of
Health (NMDOH), South Coast Air Quality Management District (SCAQMD) and Washington
State University Laboratory (WSU). The canisters arrived at CARB cleaned and evacuated.
CARB filled each canister with an urban ambient sample, from Azusa California, analyzed their
own canister and shipped the rest of the canisters to the participating laboratories. The analysis
of the canisters was performed by Toxic Organic Method TO-14A or 152 The results were then
sent to CARB, who culminated the data. The data were sent to each laboratory and to OAQPS.
The results are discussed in the next section. Please note that BAAQMD was not part of the
NATPS, however, this agency did participate in the CARB-VOC intercomparison. The results
were forwarded to OAQPS, therefore the BAAQMD results are included here for comparison.
For the elemental metallic compounds, the SCAQMD collected collocated Fine Particle - 10
micron or less (PM10) samples. The filter media (quartz) was cut into 1" by 4" strips that were
sent directly to OAQPS. The OAQPS Quality Assurance Manager (QAM) for the National Air
Toxics Program divided up the filter strips and sent them to all of the laboratories that were
performing metals analysis. These laboratories were: Hillsborough County Environmental
Protection Commission (HCEPC), Research Triangle Institute (RTI) through the State of Rhode
Island and ERG, MDEQ, Energy Northwest through the State of Washington, NMDOH and the
State of West Virginia. Analysis results were sent directly to OAQPS QAM. These laboratories
analyzed the samples using Inductively Coupled Plasma (ICP) method Inorganic Compendium
IO-3.53.
-------
There are two aldehyde compounds, acrolein and formaldehyde, that were not included in the
intercomparison study. Methods for delivery and capture of these compounds are still under
development and were not included in this study.
The statistical analysis was performed by calculating the mean value for each compound, then
calculating the standard error (i.e., standard deviation). In order to see distribution of the data,
one and two standard errors were calculated about the mean. The resulting data and the error
bars are illustrated in Figures 4-19.
Results and Findings
VOCs
For the VOCs, the QAM decided to examine the data recovery for the sample. Figure 1 and 2
illustrate that a discrepancy exists on what laboratories report. Figure 1 illustrates that some of
the laboratories reported as high as 83.3% (Rhode Island) of the total number of compounds
reported, while some laboratories only reported 26.6% (New Mexico). Figure 3 also bears out
this trend. This graph illustrates that only 2 laboratories reported 17 compounds. Figure 2
shows that for most of the compounds, only 2 laboratories reported those. A simple calculation
shows that 40% of all compounds were only reported by two laboratories. There may be several
reasons for this pattern:
A laboratory may get a peak in their chromatogram, but the GC operator may not be able
to identify the peak;
A laboratory may get a peak in a their chromatogram, but not report it because it is not
required to do so;
A laboratory may not have the correct mix of standards to identify every peak;
• A laboratory may be unwilling to report a value that is below or near their Method
Detection Limit (MDL).
This bring to the surface the issue of reporting of values. There must be standardization across
the network a target list of compounds, the calibration standards used by all of the laboratories,
identification techniques and reporting standards. The EPA must work with the S/L/T agencies
that are in the air toxic network on these issues.
The laboratories that participated in the intercomparison reported 9 of the 9 VOCs that were
requested. Vinyl chloride was reported by only one laboratory, and this values was below their
MDL. All other laboratories reported vinyl chloride as below their MDL or not detected. All
other requested VOCs were analyzed statistically. The QAM calculated the standard error for
each compound. The standard error was then plotted against the values and the mean. For most
of the compounds, the values were within 2 standard deviations of the mean. However, there
are several exceptions (please see figures 4 - 12):
• For 1,3-butadiene, the WSU result was outside of 2 standard deviations about the mean.
In addition, the CARB result was near the mean. The four laboratories in the eastern
part of the country reported levels much lower than CARB or WSU. This discrepancy
-------
may be due to stability of 1,3-butadiene in canisters. This should be investigated further.
Among the Pilot laboratories, the values reported for Benzene, Carbon Tetrachloride,
Tetrachloroethylene, Trichloroethylene and chloroform were all within 2 standard
deviations about the mean.
• For dichloromethane, the SCAQMD reported value was outside of 2 standard deviations.
For 1,2-dichloropropane, only two laboratories reported for this compound, both values
at 0.01 ppb-v. Since these values were the same, the standard error is zero.
Metals
Metallic compounds were collected on quartz filters using collocated PM10 samplers. Each
laboratory that performed metals analysis during the pilot study was sent 2 sample strips.
OAQPS did not receive results from the West Virginia or New Mexico laboratories. The same
statistical test was performed for the metals as for the VOCs. All other labs reported values.
Figures 13-19 illustrate the results:
• For beryllium, the HCEPC reported a value that was an order of magnitude 100 higher
(275 ng/strip) than the average of the other laboratories (2.59 ng/strip). The QAM
believes there was a dilution error or the filter strip was contaminated. All other samples
ranged from 5 ng/strip to 0.95 ng/strip.
• For cadmium, all of the reported values were within 1 standard deviation. HCEPC did
not report a value for cadmium.
• For chromium, manganese and nickel, all reported values were within 2 standard
deviations of the mean. For lead, HCEPC value was outside of the 2 standard deviations
limit.
Recovery and analysis for chromium has been reported in the literature as being
problematic4. Reference 4 reports that the recovery of Chromium from Standard
Reference Material 1648 as being 23%. It would be expected that this metal may be
problematic for the laboratories to extract, recover and analyze and the statistical
analysis would show large variability. However, as Figure 15 illustrates, this does not
appear to be a problem. All values were within 2 standard deviations. Three of the
laboratories reported values that were very close (i.e., 389.3, 391.0 and 368.0 ng/strip).
Two laboratories reported values of 300.0 and 300.5 ng/strip. It appears that all of the
laboratories are getting consistent recoveries through their extraction procedures and
comparable analysis for chromium.
-------
Graphs
Data Recovery of Pilot Labs
100
<£
\*'
^
-------
Distribution of VOCs Analyzed
20
15
10
10
6543
Labs that analyzed
Figure 3
.a
a.
a.
-0.2
1,3-butadiene
'S////S #
Figure 4
2STD
1 STD
1 STD
2 STD
-------
Carbon Tetrachloride
0.2
0.15
0.1
•g. 0.05
Q.
-0.05
-0.1
0°'
*• •
Finn rp Fi
0.75
0.5
.a
Q.
Q.
0.25
Tetrachloroethylene (PCE)
r\v fV* ^S*
o° -^ ^P^
A^ ^?
Figure 6
-------
Benzene
2.5
2
I 1-5
Q.
1
0.5
2STD
1 STD
1 STD
2 STD
*•
Figure 7
Dichloromethane
2.5
1.5
0.5
4
£
^\
*
3?
^
*
0°'
? 4
^
^
*
$
p^
°°S
/>
/
/
•^
/
2 STD
1 STD
1 STD
2 STD
Figure 8
-------
Chloroform
0.2
0.175
0.15
0.1
i u, t
& 0.075
0.05
0.
-0.025
*>" £
2 STD
1 STD
2 STD
.
^ o AS *^^ *^o >\^?
^^ '^^ w^ K? ^*^ ^^
^
<*
Figure 1 0
Note: Since both reported values are the
same, the Standard deviation is zero.
-------
Vinyl Chloride
U.U If.
0.01
0.008
4 0.006
Q.
0.004
0.002
0
<
•
Figure 11
Note: Only one lab reported a value, therefore, the
standard error is zero.
Trichloroethylene
2STD
1STD
1STD
2 SID
-0.04
Figure 12
10
-------
&,
VI
2500
2000
1500
1000
500
u
Metals - Round Robin
m
^ y
• Washington
I I Michigan
O HCEPC
D Rhode Is.
D ERG
Figure 13
Metal Round Robin - Berylium
c/>
300
200
100
0
-100
2STD
1 STD
1 STD
2 STD
Figure 14
11
-------
Metal Round Robin - Cadmium
Q. 30
*
2 STD
1 STD
2 STD
^
Figure 15
Metal Round Robin - Chromium
ouu
400
^ inn
^ 200
100
n
-
-
—
—
~
—
—
-
2 STD
o I U
1 STD
o I U
s*'
Figure 16
12
-------
Metal Round Robin - Lead
IUUU
900
ann
ynn
.9. 600
3« 500
O)
c 400
300
200
100
0
—
—
—
—
—
—
—
—
—
— I
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
— I
—
—
_
-
-
-
-
2 STD
1 STD
1 STD
2 STD
^
4s"
<^
Figure 1 7
Metal Round Robin - Manganese
2500
2000
.0. 1500
I
2 SI
1 SI
1 SI
2 S'
O)
1000
500
0 LI
*?
s**
^r
Figure 18
13
-------
Metal Round Robin - Nickel
uuu
A f\f\
400
a. 300
i_
-i-j
.00
O)
c 200
100
n
_
—
2 STD
1 STD
1 STD
2 STD
xP*
Figure 19
14
-------
References
1. Air Toxics Monitoring Concept Paper" (Draft), February 29, 2000. This file can be
found at the following EPA Internet site: http://www.epa.gov/ttn/amtic/airtxfil.html.
2. Compendium Method for Determination of Toxics Organic Compounds in Air, United
States Environmental Protection Agency Section TO-14A and TO-15A, January 1999.
This document can be found on the following Internet address:
http://www.epa.gov/ttn/amtic/airtox.html.
3. Compendium Method for Determination of Inorganic Compounds in Air, United States
Environmental Protection Agency Section TO-14A and TO-15A, June 1999. This
document can be found on the following Internet address:
http://www.epa.gov/ttn/amtic/inorg.html
4. Harper, Sharon, Walling, Joseph F., Holland, David and Pranger, Louis J. , "Simplex
Optimization of Multielement Ultrasonic Extraction of Atmospheric Particles"
Analytical Chemistry Volume 55, Number 9, August 1983.
15
-------
TECHNICAL REPORT DATA
(Please read Instructions on reverse before completing)
1. REPORT NO. 2.
EPA-454/R-02-004
4. TITLE AND SUBTITLE
Air Toxics Pilot Study Laboratory Intercomparison
Quality Assurance Report
7. AUTHOR(S)
Dennis Mikel
9. PERFORMING ORGANIZATION NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Research Triangle Park, NC 2771 1
12. SPONSORING AGENCY NAME AND ADDRESS
Director
Office of Air Quality Planning and Standards
Office of Air and Radiation
U.S. Environmental Protection Agency
Research Triangle Park, NC 2771 1
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
04/02
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
1 1 . CONTRACT/GRANT NO.
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/200/04
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The Air Toxics Pilot Study Laboratory Intercomparison Quality Assurance Report compares the results of two
Performance Evaluation (PE) studies that were conducted for the Air Toxics Pilot studies. Blind ambient air
toxics samples were sent to 10 laboratories that participated in the CY01 Air Toxics Pilot Study. The
laboratories analyzed the samples and forwarded their results to EPA. EPA Air Toxics QA Manager
statistically analyzed the results and provided comments and suggestions for the improvements for a national air
toxics program.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
Air Quality Monitoring
Quality Assurance
18. DISTRIBUTION STATEMENT
Release Unlimited
b. IDENTIFIERS/OPEN ENDED TERMS c. COSATI Field/Group
Air Pollution control
1 9. SECURITY CLASS (Report) 2 1 . NO. OF PAGES
Unclassified
20. SECURITY CLASS (Page) 22. PRICE
Unclassified
16
------- |