United States Office of Air Quality EPA454/R-01-003
Environmental Protection Planning and Standards February 2001
Agency Research Triangle Park, NC 27711
AIR
Pilot City Air Toxics
Measurements Summary
This document is intended only for use by the participants in the FY2000 and FY2001 Ambient
Air Toxics Pilot Monitoring Program and does not constitute guidance which is generally
applicable to State and local agencies. Its purpose is to ensure consistency among Pilot
monitoring project measurements so that analyses of the resulting data can be evaluated based on
minimal variables.
-------
Pilot City Air Toxics Monitoring Study
Table of Contents
Foreword 1
1.0 Background 2
2.0 Method Detection Limit (MDL) 3
2.1 Volatile Organic Compound (VOC) HAPs 4
2.2 Carbonyl Compounds 5
2.3 Metals and Compounds 5
3.0 Uncertainty (Precision) 6
4.0 National Performance Audit Program (NPAP) - Bias 8
5.0 Stability and Hold Times 8
5.1 Volatile Organic Compound (VOC) HAPs 8
5.2 Carbonyl Compounds 9
5.3 Metal Compounds 9
6.0 Measurement Procedures 9
6.1 Volatile Organic Compounds (VOCs ) 9
6.1.1 Humidification of VOC Canister Calibration Standards 10
6.1.2 Canister Certification 10
6.2 Carbonyl Analysis 10
6.3 Metals Analysis 12
6.3.1 Extractable versus Total Metals 12
7.0 Data Reporting 14
7.1 Lowest Calibration Level (LCL) 14
8.0 Clarification of Terminology 15
8.1 Detection Limits 15
8.1.1 Method Detection Limit (MDL) 15
8.1.2 Limit of Detection (LOD) 15
8.1.3 Detection Limit (DL) 16
8.1.4 Instrument Detection Limit (IDL) 16
8.2 Quantitation Limits 16
8.2.1 Minimum Level (ML) 16
-------
Pilot City Air Toxics Monitoring Study
8.2.2 Limit of Quantitation (LOQ) 17
8.2.3 Practical Quantitation Limit (PQL) 17
8.2.4 Minimum Reporting Level (MRL) 17
-------
Pilot City Air Toxics Monitoring Study
Appendix A
Compendium Method TO-15 A, "Determination of Volatile Organic Compounds in Air
Collected in Specially-prepared Canisters and Analyzed by Gas Chromatography/Mass
Spectrometry, GC/MS
Appendix B
Procedure for Humidification, Section 2.3.4.3.1 taken from the PAMS Technical Assistance
Document for the Sampling and Analysis of Ozone Precursors,
EPA/600-R-98/161
Appendix C
Compendium Method TO-11 A, "Determination of Formaldehyde in Ambient Air Using
Adsorbent Cartridge Followed by High Performance Liquid Chromatography, HPLC"
Appendix D
DRAFT Guidance for Carbonyl Measurements at PAMS
Appendix E
"Methodology for Determining Carbonyl Compounds in Ambient Air", Section 5 from the
PAMS Technical Assistance Document for the Sampling and Analysis of Ozone,
Precursors, EPA/600-R-98/161
Appendix F
Inorganic Compendium Method IO-3.5, "Determination of Metals in Ambient Particulate
Matter Using Inductively Coupled Plasma/Mass Spectrometry, ICP/MS"
Appendix G
Inorganic Compendium Method IO-3.1, "Selection, Preparation and Extraction of Filter
Material"
Appendix H
Appendix B to Part 136 - "Definition and Procedure for the Determination of the Method
Detection Limit", Revision 1.11
-------
Pilot City Air Toxics Monitoring Study
List of Tables
Table 1. VOC Analysis via TO-15A GC/MS for Air Toxics Pilot Monitoring Programs 9
Table 2. Carbonyl Analysis via TO-11A HPLC for Air Toxics Pilot Monitoring Program ..11
Table 3. TSP Analysis via IO-3.5 ICP/MS for Air Toxics Pilot Monitoring Program 13
-------
Pilot City Air Toxics Monitoring Study
Foreword
In order to provide consistency in the data set generated by the Pilot Toxics Monitoring
network, a laboratory measurements work group was formed to discuss the procedures to be used for
measurements. This group of laboratory, State, local, Regional and EPA representatives had a series of
discussions to define critical details of the measurement procedures needed to provide data to meet the
needs of the NAT A, the data users, and the national air toxics Pilot City monitoring program. The
purpose of this document is to outline the procedures that the laboratory work group have defined. The
document is to be used as a supplement to the EPA Compendium of Methods identified for use by the
Pilot City network. Specifically, Method TO-15 A, "Determination of Volatile Organic Compounds in
Air Collected in Specially Prepared Canisters and Analyzed by Gas Chromatography/Mass
Spectrometry, GC/MS."; TO-11A, "Determination of Formaldehyde in Ambient Air Using Adsorbent
Cartridge Followed by High Performance Liquid Chromatography, HPLC"; and IO-3.5,
"Determination of Metals in Ambient Particulate Matter Using Inductively Coupled Plasma/Mass
Spectrometry, ICP/MS".
Although the details of the procedures described below are not entirely consistent with the
Compendium of Methods, items called out in this document are specific to the data quality goals of the
Pilot monitoring program. This guideline is provided to assist states in implementation of the Pilot
monitoring network. This document is not policy and does not contain legally binding requirements, nor
is it regulation. It is intended only for use by the participants in the FY2000 and FY2001 Ambient Air
Toxics Pilot Monitoring Program and does not constitute guidance which is generally applicable to
State and local agencies. Its purpose is to ensure consistency among Pilot monitoring project
measurements so that analyses of the resulting data can be evaluated based on minimal variables. This
document is intended for use by those already familiar with the analysis of field samples for volatile
organic compounds (VOCs), carbonyl compounds and metals.
As this document is being issued, a Data Management work group is being convened to identify
and resolve issues related to reporting of data (e.g., concentration data reporting units, AIRS method
codes, etc.). Please refer to the reports of this group for clarification of data management issues.
-------
Pilot City Air Toxics Monitoring Study
1.0 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 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 (NATA). 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 Paper", posted at
http://www.epa.gov/ttn/amtic/files/ambient/airtox/cncp-sab.pdf the role of ambient monitoring 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.
In addition, initial pilot monitoring together with data analysis of existing measurements will be needed to
provide 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 Pilot City Program, a
laboratory work group was formed to discuss the details regarding procedures to be used for
measurements. This group of laboratory, State, local, Regional and EPA representatives met to define
critical details of the measurement procedures needed to provide data that will meet the needs of the
NATA, the data users, and the national air toxics monitoring program. The primary goal of the
laboratory work group was to develop consistent procedures for use by all cities participating in the
Pilot Study in order to maximize the data comparability. Consistency and comparability of data is very
February 2001 2
-------
Pilot City Air Toxics Monitoring Study
important due to differences in data reporting procedures, method detection limit determination, and
other issues that may create an artificial bias in the data base as a result of the preponderance of values
not detected.
The purpose of this document is to outline the procedures that the laboratory measurements
work group have defined. The document is to be used as a supplement to the EPA Compendium of
Methods identified for use by the Pilot City network. Specifically, Method TO-15 A, "Determination of
Volatile Organic Compounds in Air Collected in Specially Prepared Canisters and Analyzed by Gas
Chromatography/Mass Spectrometry, GC/MS."; TO-11A, "Determination of Formaldehyde in
Ambient Air Using Adsorbent Cartridge Followed by High Performance Liquid Chromatography,
HPLC"; and IO-3.5, "Determination of Metals in Ambient Particulate Matter Using Inductively
Coupled Plasma/Mass Spectrometry, ICP/MS". This document is intended to provide guidance to
those who are already familiar with the analysis of field samples for VOCs, carbonyls and metals.
Although the details described below are not entirely consistent with the Compendium of
Methods, items called out in this document are specific procedures needed to meet the data quality
goals of the Pilot monitoring program. This guideline is provided to assist states in implementation of the
Pilot monitoring network. This document is not policy and does not contain legally binding
requirements, nor is it regulation. While it presents recommendations and suggestions regarding
techniques for the measurement of toxic air pollutants for the Pilot Air Toxics Monitoring network, it
may not be appropriate for other situations.
2.0 Method Detection Limit (MDL)
It is recognized that laboratories may obtain varying detection limits based on the procedure
used and level of the standard chosen for the method detection limit (MDL) study. It is also recognized
that data measured below the detection limit has a high level of uncertainty and in theory cannot be
reliably measured or quantitatively distinguished from zero or instrument noise. One of the key goals of
this pilot program is to gather measurement data for use in evaluating the issue of calculating annual
averages with data sets containing several observations less than the MDL. Annual-average
concentrations and comparisons to modeled estimates can be highly uncertain when a large percentage
of the measurements are below the MDL. To estimate annual average concentrations from monitoring
data, the data user generally substitutes 1A MDL for those observations reported as less than MDL. In
February 2001 3
-------
Pilot City Air Toxics Monitoring Study
order to gather additional data to help improve the annual average determinations and shed light on the
quality of data at and below the MDL, "uncensored" data will be reported by the Pilot City
laboratories. An important facet of this "uncensored" data set will be the determination and reporting of
the uncertainty associated with data. The uncertainty estimates will be determined from data generated
by collocated monitors for precision.
A quote from the recent Science Advisory Board (SAB) review: "Just because an analytical
result is below the MDL does not mean that the laboratory has not been able to measure a value,
but rather that the measurement has less reliability than others that are above the MDL.
Subcommittee members stated that it is more useful to have laboratories report all data with
associated uncertainties than to have laboratories censor the data. " Although the values less than
the MDL cannot be reliably measured or quantitatively distinguished from zero, they have potential
value in computation of certain summary statistics (e.g. annual average concentration). The SAB review
can be obtained from the SAB web site under FY2000 full reports at
http://www.epa.gov/sab/fiscalOO.htm
The guidance given in 40 CFR Appendix B to Part 136, "Definition and procedure for
determination of the method detection limit"(See Appendix H of this document), will be used. Method
detection limit is defined as the minimum concentration of a substance that can be measured and
reported with 99% confidence that the analyte concentration is greater than zero and is determined
from analysis of a sample in a given matrix containing the analyte.
2.1 VolatUe Organic Compound (VOC) HAPs
Estimates of the method detection limits for the volatile organic compound (VOC) HAPs will be
determined in the following manner:
• A minimum of seven aliquots of the sample (individual canister samples) will be prepared and
each processed through the entire analytical method.
• The MDL should be determined on an annual basis, as a minimum, and when significant
instrument changes or maintenance occurs.
Canisters should be humidified prior to MDL determination; refer to section on humidification
contained in this document.
Individual canisters will be analyzed over a minimum period of 2 days (no maximum period is
specified).
February 2001 4
-------
Pilot City Air Toxics Monitoring Study
All computations are made according to the defined method with the final results in the method
reporting units (ppbv for VOC).
• The guidance in 40CFR will be used to determine the suggested concentration ranges for the
individual canister (1 to 5 times the estimated detection limit), which should correspond to
approximately 0.1 to 0.5 ppbv.
• Reasonableness of the calculated MDL will be determined using the iterative procedure as
described in 40 CFR Appendix B, section 7, which involves preparing additional standards at
the calculated MDL and analyzing. This may be difficult to implement with calculated MDLs as
low as 0.02 ppbv. Laboratory managers will be using the iterative procedure along with their
technical expertise and judgement to determine whether the calculated MDL is adequately
representative of the instrument capabilities.
2.2 Carbonyl Compounds
Estimates of the method detection limits for the carbonyl HAPs (formaldehyde and
acetaldehyde) will be determined in the following manner:
• A minimum of seven cartridges are spiked with derivatized compounds. Underivatized
compounds may be used at the discretion of the laboratory.
• The MDL should be determined on an annual basis, as a minimum, and when significant
instrument changes or maintenance occurs.
Individual spiked cartridges are extracted and analyzed no sooner than 24 hours after spiking.
• Each cartridge is processed through the entire analytical method. All computations are made
according to the defined method with the final results in the method reporting units (total ug
converted to ppbv based on typical sample volume in L for a 24-hr sample).
• The guidance in 40CFR will be used to determine the suggested concentration ranges for the
individual cartridges (1 to 5 times the estimated detection limit), which should correspond to
about 0.03 to 0.15 jig per cartridge for formaldehyde and 0.05 to 0.25 • g per cartridge for
acetaldehyde (based on 500-L sample volume).
Reasonableness of the calculated MDL will be determined using the iterative procedure as
described in 40 CFR Appendix B, section 7, which involves preparing additional standards at
the calculated MDL and analyzing. Laboratory managers will be using the iterative procedure
along with their technical expertise and judgement to determine whether the calculated MDL is
February 2001 5
-------
Pilot City Air Toxics Monitoring Study
adequately representative of instrument capabilities.
2.3 Metals and Compounds
The detection limits stated in Table 1 of IO-3 for ICP/MS are sufficient to meet the needs of
the Toxics Pilot Monitoring program. Estimates of the method detection limits for the HAP metals will
be determined using 40 CFR Part 136, Appendix B and in a similar manner as described for the VOCs
and carbonyls. A minimum of seven strips from seven individual filters will be spiked with solutions
containing the core metal compounds at a level of 3 to 5 times the expected detection limit. Using the
detection limits given in IO-3, this corresponds to about 0.03 to 0.05 ng/L.
3.0 Uncertainty (Precision)
The SAB recommends that uncensored data be reported with an associated level of
uncertainty. For the Pilot monitoring program, this uncertainty will be determined from data collected
for precision estimates; however, a measure of uncertainty will not be established or reported with each
individual measurement. Procedures available to provide data for uncertainty and estimates of precision
include the use of collocated samples and replicate analyses. Precision is a measurement of mutual
agreement among individual measurements made under prescribed similar conditions. No special
adjustments, calibrations, or maintenance of the instruments should be made. Precision checks should
be made prior to any routine or special adjustments, calibrations or maintenance.
The types of precision determinations that will be made for the FIAPs include:
replicate analyses;
• collocated samples; and
• inter- laboratory precision checks or "round-robin" analyses.
A minimum of 10% of the total number of samples will be collected in duplicate (collocated)
during the Pilot monitoring program for the urban area networks. For the small city/rural component,
collocated samples will be collected on a 1 in 12 day schedule for a minimum of 30 samples, as
resources allow. Replicate analyses will be performed on all collocated samples to provide "nested
duplicates" in order to provide an assessment of sampling and analytical precision for the study.
Measures of precision will also be fulfilled using collocated samples that are processed and
analyzed by different organizations to provide inter-laboratory precision information for the
February 2001 6
-------
Pilot City Air Toxics Monitoring Study
measurement process. A sample exchange program that involves inter-laboratory precision gives
important information concerning inconsistencies that may exist. Interpretation of these data must be
based on clear understanding and knowledge of how the data were obtained. Any differences in the
methodologies (i.e., detection limits, analytical column, calibration procedures, etc.) used to analyze the
exchange sample must be clarified in order to interpret and resolve any inconsistencies in the results.
Precision for inter-laboratory exchange samples is calculated in the same manner as precision for
replicate analyses or collocated samples. Round-robin sample analysis will occur twice over the course
of the Pilot program around the March and November time frame for VOCs and metals. Region 2,
Wisconsin, Michigan and South Coast Air Quality Management District (AQMD) have volunteered
and agreed to provide round-robin samples at this time.
A mechanism for providing round-robin samples for carbonyls has not been identified.
Technical limitations of sampling exist in relationship to collecting multiple, simultaneous ambient air
DNPH cartridges for this purpose. During the Pilot monitoring program, a round-robin comparison for
carbonyls will not be performed.
When evaluating the precision measurements, laboratories must consider each individual target
compound because precision will be compound-dependent with an influence of physical and chemical
properties (such as vapor pressure and reactivity). At very low concentrations, those at or below the
detection limit, agreement between measurements are expected to be poor.
Data pairs where the compound is detected in both samples can be evaluated for percent
difference. To make a comparison of two values (i.e., duplicates or replicates) for precision, the
Relative Percent Difference (RPD) is a more meaningful statistic than relative standard deviation (RSD),
since the number of available measurements is only two.
Where:
Yj = larger of the two observed values
X; = smaller of the two observed values
February 2001
-------
Pilot City Air Toxics Monitoring Study
Measures of precision are to be reported with the measurement data, in order to provide the data users
with information to evaluate the uncertainty. If precision is calculated from three or more values (e.g.,
annual precision), RSD should be used.
RSD= (s/x)xlOO
Where:
S = standard deviation of replicate values
X = mean of replicate values
The results from the various components of the quality assurance program are a vital part of the
database generated by the pilot projects. Unfortunately, AIRS does not currently accept this type of
information. It is the recommendation of the laboratory measurements work group that another group
be created to address data management issues and determine the components of the data package that
will be submitted by the Pilot City laboratories to the data analysis contractor.
4.0 National Performance Audit Program (NPAP) - Bias
The NPAP will be unable to provide any audit or "check" samples due to reduction in budget
for FY2001. The option does exist to allow agencies to "buy in" to the program. If an agency has
resources for audit samples for carbonyl and VOC audits, contact the NPAP coordinator Mark
Shanis, EPA, OAQPS at 919-541-1323
5.0 Stability and Hold Times
5.1 VolatUe Organic Compound (VOC) HAPs
The guidance given in TO-15 and data from the Office of Research and Development (ORD)
obtained as part of another toxics subcommittee are used to support the hold time for the core and max
pollutants at 30 days. The ORD data below gives the percent change over a 30 day period for each
pollutant. The concentration tested is given in parentheses. Methylene chloride seemed to be the least
stable in this data set.
February 2001 8
-------
Pilot City Air Toxics Monitoring Study
Core List
% Change Max. List % Change
(concentration) (concentration)
benzene
1,3 -butadiene
carbon tetrachloride
chloroform
1 ,2-dichloropropane
methylene chloride
tetrachloroethylene
trichloroethylene
vinyl chloride
5% (0.6
17% (2.
12% (0.
9% (1.2
15% (2.
27% (3.
12% (1
13% (1
18% (0.
ppbv) 1,2-dibromoethane 8%(1.2ppbv)
5 ppbv) 1,3-dichloropropene 8% (0.5 ppbv)
9 ppbv) 1,2-dichloroethane 9% (0.9 ppbv)
ppbv) 1,1,2,2-tetrachloroethane 12% (2.1 ppbv)
4 ppbv)
6 ppbv)
pbbv)
ppbv)
8 ppbv)
5.2 Carbonyl Compounds
The specified hold time in TO-11A will be used for the Pilot program. DNPH-coated
cartridges will be extracted within 2 weeks and the extracts should be analyzed within 30 days.
5.3 Metal Compounds
The hold time of 180 days as specified in IO-3 for filters will be used. Metals should be very
stable as long as the filters are handled and stored properly.
6.0 Measurement Procedures
6.1 Volatile Organic Compounds (VOCs )
Table 1 provides an outline of the specific procedures to be followed for the analysis of VOCs
by Compendium Method TO-15, "Determination of Volatile Organic Compounds in Air Collected in
Specially Prepared Canisters and Analyzed by Gas Chromatography/Mass Spectrometry, GC/MS."
A copy of this document is given in Appendix A. The procedures outlined in Table 1 also apply to the
use of Compendium Method TO-14A, "Determination of VOCs in Ambient Air Using Specially
Prepared Canisters with Subsequent Analysis by Gas Chromatography."
February 2001
-------
Pilot City Air Toxics Monitoring Study
Table 1. VOC Analysis via TO-15A GC/MS for Air Toxics Pilot Monitoring Program
Item
Canister Type
Canister Certification
Canister Transport
Canister Storage
Canister Hold Time
Method Detection Limit
Field Duplicates or Collocated Samples
Analytical Instrumentation
Blanks
• Instrument blank
Replicate Analyses
Description
SUMMA or equivalent
No target analyte > MDL; 1 canister selected per
batch (batch size determined by laboratory)
Ambient conditions
Ambient conditions
30 days
40 CFR Appendix B to Part 136
Minimum of 7 low level canister standards
analyzed over minimum of 2 day period; MDL to
be determined on annual basis at minimum
10% of total samples for urban network or 1/12
for small city; 30 minimum per network
GC/MS or GC
Performed after instrument calibration
Performed on collocated (duplicate) samples
6.1.1 Humidification of VOC Canister Calibration Standards
The guidance on humidification is given in the PAMS Technical Assistance Document (TAD),
EPA/600-R-98/161, Section 2.3.4.3.1. This guidance (Appendix B) will be adopted for the toxics
program. The TAD gives information on procedures for determining the appropriate amount of water to
attain an adequate level of humidity in the sample canister without condensation. As stated in the TAD,
low pressure (30 psig) calibration standards prepared in canisters ideally should have a minimum
amount of water vapor (• 20% relative humidity) to ensure sample integrity, but not enough water to
cause condensation of water vapor in the canister (• 33% relative humidity). To achieve these
conditions in a 6-liter canister at 70°F, between 66 and 110 jiL should be added. Calculations are
included in the guidance in order to determine the amount of water needed at varying pressures and
temperatures.
February 2001 10
-------
Pilot City Air Toxics Monitoring Study
6.1.2 Canister Certification
Canisters will be cleaned in accordance with the laboratory's normal procedures and TO-15.
Canisters will be acceptable for use if no target analyte is present at a level greater than the specified
MDL as determined in accordance with 40 CFR Part 136, Appendix B. One canister will be randomly
selected, or the canister known to be the "dirtiest" will be selected from each batch of canisters
cleaned. If the canister meets the acceptance criteria, the entire batch is considered acceptable and
therefore, ready for use. No additional flags or blank subtraction will be applied to the reported data.
6.2 Carbonyl Analysis
Table 2 outlines the specific procedures to be followed for the analysis of VOCs by
Compendium Method TO-11 A, "Determination of Formaldehyde in Ambient Air Using Adsorbent
Cartridge Followed by High Performance Liquid Chromatography, HPLC" (Appendix C).
Considerations were given to procedures also outlined in the compendium method, as well as those
provided in the Section 5, "Methodology for Determining Carbonyl Compounds in Ambient
Air"(Appendix D), of the PAMS Technical Assistance Document for the Sampling and Analysis of
Ozone Precursors, EPA/600-R-98/161; and draft guidance given in the "Guidance for Carbonyl
Measurements at PAMS" (Appendix E).
Table 2. Carbonyl Analysis via TO-11A HPLC for Air Toxics Pilot Monitoring Program
Item
Cartridge Type
Cartridge Lot Certification
Cartridge Sample Transport
Cartridge Storage
Hold Time
Description
DNPH-coated silica gel with ozone scrubber
Minimum of 3 selected per Lot. Formaldehyde not
> 0.15 ug per cartridge
Acetaldehyde not > 0.10 ug per cartridge
Ambient conditions
Refrigerated conditions
Cartridges extracted within 2 weeks; extracts
analyzed within 30 days
February 2001
11
-------
Pilot City Air Toxics Monitoring Study
Field Blanks
Trip Blanks
Method Detection Limit
Field Duplicates or Collocated Samples
Frequency = (N)/4; where N is the number of
field samples. < 0.15 ug formaldehyde and < 0.10
ug acetaldehyde
Optional - normally used to resolve issues
identified from field blanks.
40 CFR Appendix B to Part 136
Minimum of 7 derivatized spiked cartridges;
extracted no sooner than 24 hours after spiking;
MDL must be determined on annual basis at a
minimum. If labs successful spiking underivatized
components then - OK
10% of total samples for urban network or 1/12
for small city; 30 minimum per network
February 2001
12
-------
Pilot City Air Toxics Monitoring Study
Analytical Instrumentation
Blanks
• Instrument blank
• Reagent blank
Replicate Analyses
HPLC (High performance liquid
chromatography)
Performed after instrument calibration
Performed for each new Lot of reagent
Performed on duplicate (collocated) samples
6.3 Metals Analysis
Table 3 outlines the specific procedures to be followed for the analysis of TSP (total suspended
particulate) filters by Inorganic Compendium Method IO-3.5, "Determination of Metals in Ambient
Particulate Matter Using Inductively Coupled Plasma/Mass Spectrometry, ICP/MS" (Appendix F).
Considerations were also given to procedures outlined in 40 CFR, Part 50 Appendix G, "Reference
method for the determination of lead in suspended particulate matter collected from ambient air".
Consistency between Part 50, Appendix G and IO-3.5 were maintained where appropriate for filter
handling and other sampling-related procedures. For guidance related to the preparation of filter
material, Inorganic Compendium Method IO-3.1, "Selection, Preparation and Extraction of Filter
Material" is included in Appendix G.
6.3.1 Extractable versus Total Metals
Total metals (dissolution) indicates that the particulate and its matrix, as well as the filter, are
completely dissolved and results in a clear solution. This usually results in a fairly high level of solids in
solution and is often more difficult to analyze. Glass and quartz fiber filters would required the use of
hydrofluoric acid (HF) which means a more difficult and dangerous extraction process. "Extractables"
are just the compounds of the metal that dissolve into the solution you use for extraction. Different metal
compounds are extracted with nitric acid than with hydrochloric acid, or combinations, and the amounts
will vary depending on whether a hot plate, microwave or ultrasonic bath is used. Using different
extraction methods can complicate the interpretation of the data. Total metals determination is
considered more costly, difficult to perform and subject to greater background interference. Hot acid
extraction with HNO3 / HC1 to determine "extractable" metals will be the procedure used for the
Toxics Pilot Monitoring Program.
February 2001
13
-------
Pilot City Air Toxics Monitoring Study
Table 3. TSP Analysis via IO-3.5 ICP/MS for Air Toxics Pilot Monitoring Program
Item
Filter type
Filter QC
« Method (reagent) Blank
Filter Lot Blank
• Matrix Spike
• Lab control (LC) blank
• Lab control sample (LCS)
Filter cutting procedure
Filter Transport
Filter Storage
Filter Hold Time
Field Blanks
Extraction procedure
Extraction Efficiency
Method Detection Limit
Duplicate Filter Strips (Precision)
Description
Quartz Filters. Based on results from filter contamination study
- glass fiber are also acceptable for the "core" pollutant list of
metals only
per Method 10-3.1, Table 7 and Part 50, Appendix G
1 per 24 samples; reagents only
Lots >500 (20-30 selected at random); Lots <500 (lesser
number can be taken) Filters analyzed for target species
1 per 20 samples; spiked filter
1 per extraction day; manufactured filter blank certified below
NIST traceable detection limits
1 per extraction day
Pizza cutter preferred (as represented in 10-3.1, Figures 1 and
2). Strip width of 1 inch. Do not unfold filter as specified in
IO-3.
Ship under ambient conditions in protective envelope
15-30°C
180 days
1/10 filters or 10%
Hot acid extraction with HN03 / HC1 - extractable metals
Target 75-125% using NIST SRM 3087a, 2677a, or 1648 as
appropriate
40 CFR Appendix B to Part 136
Minimum of 7 filters; MDL must be determined on annual basis
as a minimum
10 % of total samples for urban network or 1/12 for small city;
30 minimum per network
February 2001
14
-------
Pilot City Air Toxics Monitoring Study
Analytical Instrumentation
ICP/MS
Blanks
Instrument Blank
Reagent Blank
Rinse Blank
As outlined in 10-3.5, page 3.5-11
Interferences - ICP/MS
• As interference by Argon on
ICP/MS
• Isobaric elemental
interference
• Abundance sensitivity
• Isobaric polyatomic Ion
interference
• Physical interferences
• Memory interferences
Several identified in 10-3.5, pages 3.5-4 for ICP/MS and
recommendations that labs should be aware of
7.0 Data Reporting
All data for the pilot study will be reported without "screening" or "censoring" the data below
detection or reporting limits. All measurements detected by the instrument will be reported with a
qualifying flag for those values below the lowest calibration level (LCL) - see below. The "7" data
qualifying flag has been established in the AIRS-AQS for this purpose. Only flag 7 will be used, which
also covers those values below the calculated MDL. Values analyzed for, but not detected, will be
reported as ND. Measures of precision as defined under the "Uncertainty" section of this document will
be reported along with the data set. Data reporting units to be defined by the Data Management work
group. A Data Management work group is being convened to discuss and clarify issues related to data
reporting. Please refer to the discussions and outcome of this work group for guidance on these issues.
7.1 Lowest Calibration Level (LCL)
Also often referred to as the minimum reporting level (MRL). Defined as the minimum
concentration that can be reported as a quantitated value for a target analyte in a sample following
analysis. For the purposes of the Pilot City study, data will be quantitated and reported below this level.
This will be the level at which the data below will be flagged indicating a level of uncertainty and still
February 2001
15
-------
Pilot City Air Toxics Monitoring Study
useful for certain statistical purposes. This will be equivalent to the concentration of the lowest
calibration standard which can only be used if acceptable quality control criteria for this standard are
met. This is established at a level 3 times the MDL. Reference: Perchlorate in drinking water method
http://www.epa.gov/OGWDW/methods/met314.pdf
8.0 Clarification of Terminology
There appear to be many different acronyms in use for the quantitation of instrument sensitivity
and reporting of data. A sampling: PQL, MDL, LOD, LOQ, IDL, ... Many of the terms are used to
refer to the same thing and typically are used for water quality analyses; however, there are generally 2
distinct classes: detection limits and quantitation limits. For use in our discussions I decided to compile
some information from a variety of sources which describes some of the terminology (acronyms) used.
However, I do not try to address the issue of the variety of methods that can be used to determine
these detection or quantitation limits and how that impacts the values obtained by a specific laboratory.
Clarification: For the purposes of the Pilot Toxics monitoring network, we are using the
terms method detection limit (MDL) and lowest calibration level (LCL), which is very similar
to the MRL given below.
8.1 Detection Limits
8.1.1 Method Detection Limit (MDL)
EPA definition: the minimum concentration of a substance that can be measured and reported
with 99% confidence that the analyte concentration is greater than zero and is determined from analysis
of a sample in a given matrix containing the analyte. (Part 136, App. B) Determined by taking a
minimum of seven aliquots of the sample (in case of air sample analysis we are using individual canister
samples) to be used to calculate the method detection limit and process each through the entire
analytical method. Make all computations according to the defined method with the final results in the
method reporting units (ppbv).
Compendium Method TO-15 and TO-11A; Method 314.0 and 1631 (Bob Avery) also refer
to 40 CFR, Part 136, Appendix B
8.1.2 Limit of Detection (LOD)
Lowest concentration of an analyte that the analytical process can reliably detect. (Anal.
Chem., Vol. 52, No. 14, December 1980)
February 2001 16
-------
Pilot City Air Toxics Monitoring Study
A number, expressed in units of concentration (or amount), that describes the lowest
concentration level (or amount) of the element that an analyst can determine to be statistically different
from the analytical blank. (Anal. Chem., Vol. 55, No. 7, June 1983 in reference to IUPAC definition in
Spectrochem. ActaB 1978, 33B, 242)
The lowest concentration of an analyte in a sample that can be detected but not necessarily
quantitated. Approximately 2 or 3 times the signal-to-noise (S/N) ratios. (LC/GC Vol. 16, No. 10,
October 1998 take from U.S. Pharmacopeia Conference, 1995)
The smallest observed signal (x) that with a reliability !-• can be considered as being a signal
caused by the component to be measured. When the observed signal is smaller than x, however, it
cannot be stated that the component is absent. It can only be said with a reliability !-• that the
concentration of the component will be less than a certain value. (Quality Control in Analytical
Chemistry, Vol. 60, Kateman and Pijpers, John Wiley & Sons, 1981)
8.1.3 Detection Limit (DL)
Minimum concentration of an analyte that can be measured above instrument background.
Estimates of concentrations at which one can be fairly certain that the compound is present.
Concentrations below this limit may not be detected. Concentrations above this limit are almost
certainly detected. http://www.wcaslab.com/TECIi/DETLIM.IJrM
8.1.4 Instrument Detection Limit (IDL)
Lowest concentration that can be detected by an instrument without correction for the effects of
the sample matrix or method-specific parameters such as sample preparation. (Web reference dated
1/27/999 - www.pwl .netcom.com/~qaa/DETLIM.html- appears to be no longer available)
8.2 Quantitation Limits
8.2.1 Minimum Level (ML)
The lowest level at which the entire analytical system must give a recognizable signal and
acceptable calibration point for the analyte. It is equivalent to the concentration of the lowest calibration
standard, assuming that all method-specific sample weights, volumes, and cleanup procedures have
been employed. Calculated by multiplying the MDL by 3.18 and rounding the result to the number
nearest to (1,2, or 5) x 10", where n is an integer. Method 1631 (from Bob Avery at
http ://www. epa.gov/ost/methods/1631 fina!2 .pdf)
February 2001 17
-------
Pilot City Air Toxics Monitoring Study
8.2.2 Limit of Quantitation (LOQ)
A minimum criterion or region for quantitation that should be clearly above the detection limit.
The lowest concentration of an analyte in a sample that can be determined (quantitated) with acceptable
precision and accuracy under the stated operational conditions of the method. Approximately 10 times
the signal-to-noise (S/N) ratios. (LC/GC Vol. 16, No. 10, October 1998 take from U.S.
Pharmacopeia Conference, 1995)
8.2.3 Practical Quantitation Limit (PQL)
The lowest concentration of an analyte that can be reliably measured within specified limits of
precision and accuracy during routine laboratory operating conditions. The Agency has used the PQL
to estimate or evaluate the minimum concentration at which most laboratories can be expected to
reliably measure a specific chemical contaminant during day-to-day analyses of drinking water samples.
(EPA Office of Water web site www.epa.gov/ogwdwOOO/standard/review/methods.htmn
Normally determined as 3 to 10 times the MDL and is considered the lowest concentration that
can be accurately measured , as opposed to just detected.
http ://www. wcaslab. com/TECH/DETLEVI.HTM
8.2.4 Minimum Reporting Level (MRL)
The minimum concentration that can be reported as a quantitated value for a target analyte in a
sample following analysis. This defined concentration can be no lower than the concentration of the
lowest calibration standard and can only be used if acceptable quality control criteria for this standard
are met. Established at a level either 3 times the MDL or at a concentration which would yield a
response greater than a signal-to-noise ratio of five. (Perchlorate in drinking water method
htto://www.eoa.gov/OGWDW/methods/met314.odf)
February 2001 18
-------
Pilot City Air Toxics Monitoring Study
Appendix A
Compendium Method TO-15A, "Determination of Volatile Organic
Compounds in Air Collected in Specially-prepared Canisters and Analyzed
by Gas Chromatography/Mass Spectrometry, GC/MS
See: http://www.epa.gov/ttn/amtic/files/ambient/airtox/to-15r.pdf
February 2001
-------
Pilot City Air Toxics Monitoring Study
Appendix B
Procedure for Humidification, Section 2.3.4.3.1 taken from the PAMS
Technical Assistance Document for the Sampling and Analysis of Ozone
Precursors, EPA/600-R-98/161
See: http://www.epa.gov/ttn/amtic/files/ambient/patns/newtad.pdf
February 2001
-------
Pilot City Air Toxics Monitoring Study
Appendix C
Compendium Method TO-11A, "Determination of Formaldehyde in
Ambient Air Using Adsorbent Cartridge Followed by High Performance
Liquid Chromatography, HPLC"
See: htto://www.epa.gov/ttn/amtic/files/ambient/airtox/to-11 ar.pdf
February 2001
-------
Pilot City Air Toxics Monitoring Study
Appendix D
DRAFT
Guidance for Carbonyl Measurements at PAMS
prepared for EPA under contract 68-D-98-030
February 2001
-------
DRAFT
Guidance for Carbonyl
Measurements at Photochemical
Assessment Monitoring
Stations (PAMS)
September 30,1999
Prepared for EPA under
Contract 68-D-98-030
Work Assignment 2-02
Emissions, Monitoring, and Analysis Division
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
-------
EPA Disclaimer
The information in this document has been funded wholly or in part by the United States Environmental
Protection Agency under Contract 68-D-98-030. It has not been subject to the Agency's peer and
administrative review. Mention of trade names or commercial products does not constitute
endorsement or recommendation for use.
-------
Contents
Acknowledgment iv
Guidance for Carbonyl Measurements at Photochemical Assessment
Monitoring Stations (PAMS) 1
I. Sampling System „,„..,; 3
II. Cartridges .., 9
III. Analytical System : ':•;. 12
IV. Blanks , '. . ^, .\. 17
V. Data Reporting ,,,,.,.. , 19
VI. Concluding Remarks ,,..., '. 20
Figure
1. Control chart for plotting cartridge field blank values versus time 22
Tables
1. Sampler quality control criteria checklist . : 7
2. Analysis system quality control criteria checklist , 14
3. Minimum number of Hanks per field samples 19
4. Proposed checklist table for tracking cartridges 21
111
-------
Acknowledgment
We thank the many participants for their time in reviewing our outline and the draft document.
The participants were staff involved in a previous work assignment involving a series of conference calls
to assess carbonyl measurements at PAMS. Their enthusiasm and willingness to share experiences and
knowledge were quite evident during the review process.
IV
-------
Guidance for Carbonyl Measurements
at Photochemical Assessment
Monitoring Stations (PAMS)
The determination of ambient concentrations of carbonyl compounds is a requirement of 40
CFR Part 58, Subpart E, enhanced ozone network monitoring programs.(1) The U.S. Environmental
Protection Agency (EPA) has established a Photochemical Assessment Monitoring Stations (PAMS)
program to provide routine measurements of selected volatile organic compounds (VOCs) and
carbonyl species. The PAMS program currently recommends the sampling and analysis of 55 VOCs
and three carbonyl compounds: formaldehyde, acetaldehydes and acetone. The measurement of
acetone is now optional (see PAMSGRAM, Volume 16). For the measurement of carbonyl species,
States are required to obtain 3-hour and 24-hour integrated samples, at collection frequencies specified
for each type of enhanced ozone monitoring site.
The measurement method for carbonyls in PAMS is based on U.S. EPA Compendium Method
TO-11 A, which incorporates the use of sorbent cartridges coated with 2,4-dinitrophenylhydrazine
(DNPH) for sample collection.(2) The analyses are performed with high performance liquid
chromatography (HPLC). The two sorbents described in the compendium method are silica gel and
octadecylsilane bonded silica substrate (CIS). For consistency, silica gel is recommended for use in
the PAMS program. For the PAMS program,
/•••;'/.
carbonyl methodology is further explained in the Technical Assistance Document (TAD), which more
thoroughly discusses specific topics including monitoring instrumentation, ozone scrubbers, and
.-•" . '• ' '
cartridge blanks.(3) The guidance provided here supercedes that given in the TAD where applicable.
Currently, numerous State, federal, and private laboratories are conducting carbonyl sampling
and analytical activities as part of the PAMS program. However, there are concerns about the existing
carbonyl database and data quality in general. As a result of these concerns, a series of conference
calls were conducted with several such groups, representing a wide range of procedures used during
the sampling and analysis efforts.(4)
The mechanical integrity of field sampling devices and the lack of field audit and sampling
protocols are key issues for PAMS carbonyl measurements. One concern is the failure of aging
-------
carbonyl sampling equipment as a result of leaks, which are often extremely difficult to detect and may
go unnoticed until data quality or other quality assurance/quality control (QA/QC) results indicate a
problem. Collocated sampling with duplicate equipment is one QA/QC approach to evaluate sampling
abnormalities, but State agencies often do not have the necessary extra equipment. Current
performance audits for the carbonyl sampling in the PAMS program employ a DNPH-silica gel
cartridge spiked with selected carbonyl derivatives. National Institute of Standards and Technology
(NIST) gas-phase carbonyl standards are not currently available for method Calibration and bias
determinations. Consequently, until these standards are available, the integrity of carbonyl sampling
equipment cannot be completely tested.
Suitable QA/QC procedures are particularly important in light of discrepancies observed
among nominally identical carbonyl sampling and analytical systems operating at some of the PAMS
sites. The present PAMS carbonyl sampling methodology could benefit from the development of
procedures to enhance sampling precision and accuracy. Greater standardization of sampling and
analysis techniques should result in better data comparability from different sites, more consistent
assessment of data quality, and better estimation of seasonal and long-term trends in air quality.
Detailed procedural guidance for existing PAMS equipment is critical to addressing the measurement
issues.
The purpose of this document is to provide guidance for use by Agencies in order to obtain
more consistency in conducting carbonyl monitoring in the PAMS program. This document identifies
critical requirements for the collection and analysis of formaldehyde, acetaldehyde, and acetone, and
addresses the necessary QA/QC procedures to assure good quality data for the PAMS program. It
focuses on five subject areas: sampling system, sampling cartridges, analytical system, blanks and data
reporting. This document is not intended to replace TO-11A and the PAMS TAD, but is intended to
outline, clarify and emphasize important and critical aspects of the cartridge carbonyl methodology
essential in obtaining good quality data. The following specific guidance is given to help improve the
qualityof the PAMS carbonyl data collected:
-------
Commercially available DNPH-cartridges and sampling equipment are to be used.
Cartridge field blank subtractions are not be required.
• Flow rates for the 3-hr and 24-hr sampling are specified.
• Cartridge shipping procedures are clearly defined.
• Analytical information provided to EPA will be posted on the PAMS website (see
Section IH.A.2).
A routine ozone scrubber change-out schedule is specified.
The following sections provide the rationale and further details for implementation of these
PAMS carbonyl monitoring recommendations.
I. Sampling System ,:,
This section focuses on the physical requirements, the calibration, and the operation of the
carbonyl sampling system. At a minimum, the following components should be included in the PAMS
carbonyl sampler. These items are also described in Section 5.2 of the PAMS Technical Assistance
Document.(3)
A. Carbonyl Sampling System
Carbonyl samplers should be constructed so that all material coming in contact with the
sampled air is glass, stainless steel, and/or Teflon™.
• A heated inlet line to the sampler is strongly recommended to prevent condensation of
water and/or organic compounds. The material of construction for the inlet line should
be stainless steel, or Teflon™. The elevated temperature of the inlet should be at • 50
±15'C.
» A denuder or cartridge type ozone scrubber is required to remove ambient ozone from
the sample stream. If a copper coil denuder is used, then it should be wrapped with a
cord heater and controlled to an elevated temperature (• 50 ± 15» C) to prevent
condensation of water and/or organic compounds in the sampling line.
-------
Inlet check valves, solenoid valves, or a multi-port rotary valve are recommended to
direct sample to, and to isolate, the individual sampling cartridges. Diffusive sampling
may occur if such valves are not present and operational.
A multiport cartridge assembly is recommended to support multi-event sampling and
allow for easy insertion and removal of DNPH sampling cartridges.
The outlet side of the sample cartridge assemblies also must be equipped with check
valves (or equivalent) to isolate individual sampling cartridges,
A mass flow controller or mass flow meter/control valve must be used to maintain
constant flow rate over the specified sampling period. :
An oil-less vacuum pump, capable of achieving a pressure drop of -25 inches Hg, is
necessary to draw sample through the sampling cartridge during collection.
An event control and data acquisition device is required to allow unattended operation
of the collection system and to record sampling event information such as start and stop
times, collection flow rates, etc.
Although the above list consists of generally available standard components of air sampling equipment,
proper assembly requires tedious and time-consuming testing and evaluation. It is strongly
recommended that future users consider commercially available instruments that have been tested and
evaluated to meet carbonyl sampling requirements, .
A separate commercial sampler also should be used for the 24-hour time integrated samples,
unless a single commercial unit is equipped to perform both types of sampling. At a minimum, the 24-hr
sampler should contain the same components as the 3-hr sampler except that the multi-port cartridge
assembly is not needed. Commercial samplers that can be automatically leak checked are highly
preferred.
Current commercial vendors of carbonyl sampling systems include:
ATEC Atmospheric Technology, P.O. Box 8062, Calabasas, CA 91372-8062, (310)
457-2671
Atmospheric Analysis and Consulting (AAC) Inc., 4572 Telephone Road, Suite 920,
Ventura, CA 93003, (805) 650-1642
-------
Mllipore/Waters Chromatography, P.O. Box 9162, Marlborough, MA 01752-9748,
(800) 252-4752
Scientific Instrumentation Specialists, P.O. Box 8941, Moscow, ID, (209) 882-3860
SKC Inc., 334 Valley View Road, Eighty Four, PA 15330-9614, (800) 752-8472
Supelco, Supelco Park, Bellefonte, PA 16823-0048, (800) 247-6628
XonTech, Inc., 6862 Hayvenhurst Avenue, Van Nuys, CA 91406, (818) 787-7380
The mention of vendor names does not constitute endorsement or recommendation by the U.S. EPA.
Each user should evaluate the system to make educated purchases and determine if it meets the
individual's needs. :
B. Requirements for an Ozone Scrubber
The EPA has previously determined through laboratory tests that ozone present in ambient air
interferes with the measurement of carbonyl compounds when using Method TO-11 A. As stated in the
Technical Assistance Document,(3) ozonfecan interfere with carbonyl analyses in three ways:
The ozone reacts with the DNPH on the cartridge, making the DNPH unavailable for
derivatizing carbonyl compounds
The ozone degrades the carbonyl derivatives formed on the cartridge during sampling
• If the analytical separation is insufficient, the DNPH degradation products can coelute
with the target carbonyl derivatives.
The extent of interference depends upon the ambient concentration of both ozone and the carbonyl
compounds, and on the' duration of sampling. Carbonyl compound losses can be as high as 50 percent
on days when the ambient ozone concentration reaches 120 ppbv.(3) As a result it is mandatory that an
ozone scrubber be used for carbonyl sampling in the PAMS program and that it be properly
maintained.
Two types of scrubbers have been developed • the ozone denuder and the ozone cartridge
scrubber. Both scrubbers use potassium iodide (KI) as the scrubbing agent, and their designs
-------
effectively allow for the removal of ozone at sampling flow rates up to 1 liter/minute. Details on the
equipment and preparation of these scrubbers are provided in the TAD, TO-11A, and PAMSGRam,
Volume 12(5) documentation. Below is a brief description and recommended change-out time of each
device.
The ozone denuder is a copper tube coated internally with a saturated solution of KI. The tube
is coiled and housed in a temperature-controlled chamber that is maintained at elevated temperature
(• 50 ± 15* C). The elevated temperature prevents condensation of water vapor and organic
compounds in the coil during sampling. The ozone denuder as described in the TAD has a usable
lifetime of up to 100,000 ppb-hours. This lifetime period was determined during laboratory tests using
controlled relative humidity (RH) conditions. Denuder performance may be affected by the variable
pollutant and RH conditions in the ambient atmosphere. On a conservative basis, however, the
scrubber should be effective for up to 30 days of continuous ambient air sampling. To assure consistent
performance, replacement of the ozone denuder is recommended after the equivalent of 30 days of use,
e.g., six months of sampling on every sixth day. The scrubber is reusable, and the re-coating procedure
is described in the TAD.
The second type of ozone removal device described in the TAD is the cartridge scrubber. This
device is commercially available (e.g., Supelco, Waters) and is filled with approximately 1 gram of
ACS reagent grade KI (the cartridge is identical in size and shape to the precoated DNPH silica
cartridges). The scrubber cartridge is positioned at the sample inlet, just ahead of the DNPH-coated
cartridge. During high humidity/temperature conditions, it is recommended that the cartridge scrubber
be maintained at elevated temperature (~50 ± 15* C) to prevent condensation of water vapor and
organic compounds. According to the TAD, the theoretical removal capacity for ozone is 200 mg,
based upon the assumption of 100 percent consumption of KI. As a result, change-out of the cartridge
scrubber every three weeks is recommended.
C. Sample Probe Line and Connection to Primary Manifold
The primary sampling manifold must meet the criteria for the PAMS network. These criteria
can be found in Section 5.2.3 of the Technical Assistance Document.(3) The carbonyl sampler should
be connected to the primary manifold using a 1/4 inch O.D. heated line that is made of stainless steel or
-------
Teflon™. The ozone scrubber and carbonyl sample cartridge should be placed as close as possible to
the primary manifold. The carbonyl sampling line should be connected to the primary manifold at a
location that is downstream of the connection line used for VOC sampling (in order to minimize the
possibility of acetonitrile solvent back-diffusing into the VOC sampling line).
D. Calibration and Operation of Carbonyl Sampling System
Procedures for the calibration and operation of the carbonyl sampling system include
implementation of field check procedures, operation at specified flow rates, flow checks, and
employment of calibration gases to challenge the sampler. Table 1 provides a checklist to assess the
performance of the sampling system.
Table 1. Sampler Quality Control Criteria Checklist
Parameter
Flow Check
Mass Flpw Controller
(or mass flow meter)
Leak Check ,
Sampler Blank
Collocated Samples
Backup Cartridges(a)
Trip Blanks
Field Blanks
Sampler Challenge
(With Gas Mixtures)
Frequency
Each Sampling Event,
Pre- and Post-Checks
"Start, Midpoint, End of
Season1 - '• •
Before Each Sampling
Event
Pre- and Post- Seasons
10% of Field Samples
1,0% of Field Samples
Square Root of the
Number of Field Samples
Square Root of the
Number of Field Samples
Pre- and Post- Seasons
- ' •••>>>>>
Limits
3 hr, 1.0 liter/minute
(±20%) '
24 hr, 0.1 3 liter/minute
,(±20%)
100 ±10% (Reference
Meter)
No Air Flow
<0. 1 5 (ig Formaldehyde/
Cartridge
± 20 %
• 10% of Total on Backup
Cartridge
<0. 1 5 (ig Formaldehyde/
Cartridge
<0. 1 5 ng Formaldehyde/
Cartridge
70 to 130% Recovery
Corrective Action
Repair/Exchange Unit
Repair/Recalibrate Unit
Recheck for Leaks,
Modify as Necessary
Clean or Replace Sampler
Mark Samples as Suspect
Use Backup Cartridges
for All Samples
Evaluate Sampling and
Analysis Procedures,
Purchase New Batch
Evaluate Sampling and
Analysis Procedures,
Purchase New Batch
Clean or Replace Sampler
-------
(a) Not needed if recommended flow rates are used; see Section ID.2.
-------
1. Implementation of field check procedures
The TAD mentions several key activities that should be performed to assure proper
operation of the carbonyl sampler; however, re-emphasis and further details are
provided here. First, a leak check should be performed before each sampling event.
The sampler should be on for at least 10 minutes prior to the leak test. The inlet line
should then be sealed and the mass flow controller (or mass flow meter) readout from
the sampler should drop to zero (within a few minutes). If leaks are detected, recheck,
tighten, and/or modify the system. Once the absence of leaks has been confirmed, the
inlet line is opened, and sampler flow should be checked with a NIST-traceable flow
meter to assure that the target flow rate is achieved (1.0 liter/minute for 3-hr sample,
0.13 liter/minute for 24-hr sample).
Second, the mass flow controller for each sampler should be checked at the
start, at the midpoint, and at the end of each ozone season. For acceptability, the
calibration reading should be within 100 ±10 percent of the reading from aMST-
certified flow meter. Deviations from this range should be noted and the mass flow
controller should be recalibrated or exchanged. Third, a sampling system blank check
should be performed as a pre- and post-season validation of the performance of the
sampler. This check is perforrned by obtaining a 3-hr cartridge sample while supplying
aldehyde-free air to the sampler inlet. It is recommended that aldehyde-free air be
generated by placing a DHPH cartridge at the inlet to the sampler. The sampler itself is
then operated,at 1.0 liter/minute flow rate over its normal 3-hr sampling period. The
amount of ialdehydes found in the resulting sample must originate from within the
sampling system and can be compared to ambient levels. The current requirement is
that the, system blank check loading should be less than the Method TO-11A
acceptance criteria (<0.15 jig formaldehyde/cartridge). If not, the data need to be
: qualified and the sampler should be cleaned/exchanged. The user should contact the
vendor for specific cleaning instructions.
As part of the normal QC activities for field sampling, it is recommended that
the following samples also be collected: collocated samples, backup cartridges, trip
-------
blanks, and field blanks. However, if the flow rates of 1.0 liter/minute for the 3-hr
sampling and 0.13 liter/minute for the 24-hr sampling are used, then backup cartridges
are not necessary. Each of these blank types should be collected on a frequency as
shown in Table 1. Table 1 summarizes the critical QC activities and is intended to
replace Table 5-3 in the TAD for these nine parameters.
2. Operation at specified flow rates
For carbonyl measurements in the PAMs program, the target collection volume through
a DNPH cartridge is • 180 liters of air. Thus, for the 3-hr sample, thd required flow
rate is 1.0 liter/minute (± 20 percent). For the 24-hr sample, the required flow rate is
0.13 liter/minute (± 20 percent). As indicated in Table 1, if these flow rates are used,
then backup cartridges are not required.
3. Employment of calibration gases
Commercially available calibration cylinders have been prepared that contain stable
ppb levels of aldehydes. It is recommended that these calibration gas cylinders be
purchased from a specialty gas vendor and used to challenge the field sampling units.
Percent recovered should be within 100 ± 30 percent of the delivered quantity of
carbonyl (based on the stated cylinder value). If the recovery values are outside this
range, appropriate troubleshooting procedures should be initiated. For additional
comparability, the cylinders should be exchanged across PAMS sites.
II. Cartridges
This section is intended to re-emphasize important TAD and TO-11A information addressing
the acquisition, handling, shipping, and storage of DNPH-coated cartridges.
A. Preparation/Acquisition of Cartridges
Cartridges should be acquired in bulk quantities from commercial vendors. Preparation of
cartridges by individual laboratories is tedious, labor intensive, requiring clean room conditions, and is
10
-------
not recommended. Noncommercial preparation is likely to result in more lot- to-lot variability than is
found in commercially prepared cartridges. This approach is counter-productive to improving
consistency in the PAMS program. Information on the cartridges such as vendor, quantity received,
date of receipt, lot number, and expiration date should be recorded in a laboratory note book.
Major commercial suppliers of DNPH-coated cartridges include:
Supelco, Supelco Park, Bennefonte, PA 16823-0048, (800) 247-6628
Millipore/Waters Chromatography, P.O. Box 9162, Marlborough, MA 01752-9748,
(800)252-4752 • ,
• Atmospheric Analysis and Consulting (AAG) Inc., 4572 Telephone Road, Suite 920,
Ventura, CA 93003, (805) 650-1642 \
SKC Inc., 334 Valley View Road, Eighty Four, PA 15330-9614, (800) 752-8472.
The mention of vendor names does not constitute endorsement or recommendation by the U.S. EPA.
Each user should evaluate the cartridges to determine if they meet the program's needs.
The receiving laboratory should certify acceptability of the cartridge lot by following the blank
analysis procedure specified in the Technical Assistance Document/3' For a minimum of three cartridge
lot blanks analyzed, the average blank value plus three times the standard deviation of the blank values
(i.e., • + 3s) must meet the criteria for acceptance set out in Method TO-11A, which are
formaldehyde <0.15 jig/cartridge
acetaldehyde <0.10 jig/cartridge
/acetone* <0.10 jig/cartridge
* Nt>te: analysis for acetone is now optional for PAMS.
The certification blank value and lot number must be recorded in the laboratory record book and the
cartridge lot rejected and returned to the vendor if any acceptance value is not met.
B. Handling Cartridges
Biological processes produce carbonyl species from the skin and breath. Therefore, gloves
should be worn when handling cartridges. Polyethylene gloves (or equivalent) are recommended during
11
-------
field usage of the cartridges. Nitrile gloves are recommended to protect the hands of the laboratory
chemist during the extraction of the cartridge with acetonitrile.
The field operator and laboratory chemist should minimize the time that unsealed cartridges are
exposed to the environment. Diffusive sampling does occur when the cartridge caps are removed and
can be significant depending upon background concentrations of the carbonyl species. DNPH is also
light sensitive, and the cartridges should be protected from direct light by retaining them in the sealed foil
pouch provided by the manufacturer or by covering with aluminum foil or similar material. Finally, to
further reduce the possibility of contamination, avoid writing directly on the cartridges or placing
adhesives onto the cartridges.
C. Shipping and Storing Cartridges
At a minimum, the shipment and storage of cartridges for the PAMS program should follow the
guidelines indicated below:
All cartridges should be stored in a dedicated refrigerator (4» C) until use • adhere to
vendor's expiration dates for use of cartridges.
All cartridges should contain sealing caps (or plugs). Make sure caps are in place •
discard any cartridge found with a missing cap.
All commercial cartridges should be transported inside their original shipping containers
(as shown in Figure"5 of the TO-11A document). Some commercial containers include
sealed foil pouches and glass culture tubes for individual cartridges; others include
polypropylene holders equipped with foam inserts for holding multiple tubes.
If the original shipping container is unavailable, friction-top metal cans should be used.
The cans should be partially filled with a layer of activated charcoal.
The shipping container should be padded with either polyethylene-air bubble padding
or clean laboratory tissue paper. Polyurethane foam or inked paper should never be
used as padding material.
Cold packs are not required for cartridge shipment. Bulk shipment at room
temperature with second-day delivery is acceptable.
Cartridges should be stored in a dedicated refrigerator (4» C) upon arrival at the
laboratory or field site.
12
-------
III. Analytical System
This section focuses on the equipment requirements and the calibration and operation of the
DNPH-cartridge extraction and HPLC analysis systems.
13
-------
The analytical system should include an adequate cartridge extraction apparatus as well as a
high performance liquid chromatographic system.
1. Cartridge extraction
The highest purity acetonitrile should be used to extract the sampled cartridges. Some
commercial manufacturers sell a carbonyl-free acetonitrile which is preferred. The
solvent lot should be analyzed upon receipt to determine the initial purity level, and then
periodically re-analyzed over the life of the bottle to track the aldehyde buildup over
time. '
All glassware should be cleaned by rinsing with acetonitrile, then dried by heating to
6O C in a vacuum oven. The use of a nitrogen-purged glove box (bag) further reduces
the risk of contamination.
The sampled cartridge should be fore-flushed with acetonitrile to extract the derivatized
carbonyls. The alternative back-flush elution approach is not recommended because it
sometimes adds particulate materials also collected on the cartridge to the acetonitrile
extract solution. During analysis, the particles can cause premature sample valve failure
and can increase the column back pressure. Because the acetonitrile holdup volume is
T- 0.3 nil, an extraction volume of 5 ml is recommended.
2. HPLC analysis system
, : Section 11.3.1 of the TO-11A document specifies the HPLC operating parameters and
an isocratic elution program is adequate for sample analysis when formaldehyde is the
: only carbonyl of interest. For more complex carbonyl samples, Section 14.3.1 of the
TO-11A document describes an HPLC gradient elution program that will resolve
acetaldehyde, acetone, propionaldehyde, and the higher molecular weight carbonyls.
More recently, several commercial vendors have demonstrated similar separation
14
-------
capabilities using their own specific brand of column and operating conditions. For
documentation and PAMS network consistency purposes, all PAMS measurement
groups are encouraged to provide the following information to Nash Gerald at
rice.ioann@,epa.gov:
• Name of Organization
• HPLC (type/manufacturer)
Detector (type/manufacturer and wavelength)
Data Handling System (type/manufacturer)
• Analytical Column (type/manufacturer)
• Guard Column (type/manufacturer)
Column Operating Temperature
Mobile Phase - Gradient Conditions (isocxatic conditions)
• Solvents (manufacturer and lot number)
• Column Flow Rate/Column Head Pressure
HPLC Run Time/Representative Calibration Run
• Sample Injection Volume
Calibration Results • MDL, Range, R2, etc,
This information will be tabulated and posted on the PAMS homepage at
www.epa.gov/oar/6aqps/pams.
Although acetone is no longer a required target compound, it is recommended that
calibration data continue to be examined for the separation of the three C-3 carbonyl
species that may be present in the chromatogram (acrolein*, acetone, and
propionaldehyde). The resolution of these three peaks should be tracked over time to
evaluate column performance (for the lowest calibration mixture, each valley between
the three successive peaks should be less than 50 percent of the highest peak). Further
decreases in resolution and/or excessive column pressure buildup indicate the need for
column replacement or refurbishing.
* Note: Method TO-11A no longer considered applicable to acrolein.(6)
A. Calibration of Analytical System and Implementation of QA/QC Procedures
15
-------
To ensure consistency across the PAMS network, the following calibration procedures are
recommended. The frequency, acceptance criteria, and corrective actions associated with these
procedures are shown in Table 2. Many of the table items are updated parameters from Table 5-3 of
the TAD.
Table 2. Analysis System Quality Control Criteria Checklist
Parameter
Multipoint Calibration
Check Standard
Method Detection Limit
Replicate Injections
NPAP Audit Samples
Resolution of C-3 Carbonyl
Species(a)
Matrix Spike
Laboratory/Extraction Blank
Frequency
Every 6 Months
Daily
Annually
Daily
One to Three
Times Per Year
X
Daily
Each Lot
Each Extraction
Batch
Limits
R2 > 0,99
± 10%
<0.1 ppborO,22
•g for a 180-liter
Sample
± 10%
-23% to +22%
Valley Between
Peaks « 50% of
Highest Peak
± 30%
• Lot Certification
Blank
Corrective Action
Recalibrate
Recalibrate
Check/Service Instrument
Check/Service Instrument
Recalibrate
Change Column Program/
Change Column
Check Against New
Matrix Spike
Check Laboratory
Processes
(a) Recommended for labs that continue to monitor the C-3 carbonyl species.
16
-------
1. Calibration standards
Calibration standards should be purchased from commercial vendors. Material can be
purchased as solid DNPH-carbonyl derivatives or as dilute liquid mixtures. The liquid
mixtures are generally supplied in a range from 1 to 50 |ig/ml as the carbonyl
compound. Further dilutions should be made with volumetric glassware.
2. Multipoint calibration
A working standard range from 10 ng/ml up to 2000 ng/ml should be targeted. A
multipoint calibration is recommended every six months, and a minimum of fite
calibration points (including zero) should be used. Analyses at each point should be in
triplicate. A linear least-squares fit of the data should be conducted and an R2 value of
0.99 or better should be attained. The slope of the calibration curve for each
component provides a response factor (RF).
3. Calibration check standard
A separate, independent calibration standard near the expected levels of the target
carbonyl concentrations should be used for daily calibration checks of the analytical
System. The day-to-day variation of the components should be within ±10 percent of
the initial calibration value. If greater variability is observed, a fresh check standard
should be prepared. If results with the fresh standard deviate from the original
calibration curve slope by more than 15 percent, then a new multipoint curve should be
constructed. A plot of daily values on a Quality Control Chart should be made and
used by the analyst to check on long-term performance of the analytical system.
4, Method Detection Limits (MDLs)
The MDL determination should be done on an annual basis using the procedures
specified in the Code of Federal Regulations (40 CFR Part 136B).(7) In brief, a low
level standard is prepared at a concentration that is approximately two to five times the
estimated MDL. The standard is injected seven times. The average concentration is
17
-------
calculated from the original calibration curve. The standard deviation and the
appropriate student t-value are used to calculate the MDL as described in the CFR.
C. Operation/Performance of Analytical System and Implementation of QA/QC
Procedures
1. Daily precision checks
Precision checks should be done on a daily basis and should include analyses of both
standard and sample. The precision (as relative standard deviation) should be within 10
percent, based on three replicate injections.
2. Chromatogram checks
Daily inspections should be made to check if retention times are drifting. A control
chart should be used to determine if trends are occurring.
3. NPAP - spiked cartridges/performance audit samples
The National Performance Assessment Program (NPAP) is an ongoing program to
check analytical accuracy of participating laboratories. Cartridges spiked with know
amounts of liquid carbonyls (unclerivatized) are prepared and distributed to analytical
laboratories. All PAMS participants should perform analyses on the NPAP audit
samples. It is recommended that laboratories participate at least once per year.
1. Matrix spike
A matrix spike test is recommended per cartridge lot. This procedure involves spiking
cartridges (at least three) with non-derivatized carbonyls, and provides an evaluation of
both the derivatization and the extraction processes. The underivatized carbonyls
••'. should be obtained from commercial vendors. A target acceptance criterion is 70 to
130 percent recovery. If this criterion is not met, then the analyst should recheck the
matrix standard mixture against a new mixture.
18
-------
5. Laboratory/extraction blank
Laboratory/extraction blanks should be analyzed for each batch of cartridges that are
extracted. The results from these analyzed samples will indicate the combined
cartridge, solvent, and glassware contamination level for each carbonyl compound.
This QC activity will guide the analyst in verifying that laboratory operations are being
conducted appropriately.
6. Internal Standard (IS)
An internal standard, such as the cyclohexanone-DNPH derivative, is recommended as
another means of tracking instrument performance. The IS is not used for calibration
purposes but rather to track detector response and certify the injection of each sample
vial. The cyclohexanone derivative can be added to the acetonitrile prior to cartridge
extraction.
7. Acetonitrile purity
Acetonitrile used for extractions should be evaluated upon receipt and periodically
during use as described in Section O.A.1 of this document. A carbonyl free grade of
acetonitrile should be used.
IV. Blanks
To ensure the quality of the data and to obtain more consistent results, the collection of sample
blanks is necessary. As indicate!! in the TAD, there are four types of blanks: lot certification blanks,
field blanks, trip blanks, and sampling system blanks. In this section, the purpose of each type of blank
is described, the number of blanks necessary is discussed, and finally, procedures to be used in
reporting the blank data are provided.
C. Types of Blanks
19
-------
Lot Certification Blanks • Certification blanks consist of a minimum of three
laboratory blank cartridges that are eluted with acetonitrile and analyzed to verify
acceptability of a specified cartridge lot number from a commercial vendor.
Certification blank analysis is required for each cartridge lot number.
Field Blanks • Field blanks are blank cartridges that are sent to the field, connected
to the sampling system, and treated identically to the samples except that no air is
drawn through the cartridge. Field blanks are used to assess the background carbonyl
level for cartridges used during the ambient sample collection process.
Trip Blanks • Trip blanks are cartridges of the same lot nuriiber that are sent to the
field, stored, and returned to the laboratory with the sampled cartridges; Trip blafik
cartridges are not connected to the sampling system. Trip blanks are optional and are
intended to be used to resolve contamination problems determined from the field
blanks. Trip blanks can be used to determine Whfithpr the contamination occurred
during the sampling process or during the shipping arid storage process.
Sampling System Blanks • Pre- and post-season validation of the performance of
the sampling system is necessary. These system blanks are used to assess the
contamination level of the sampler itself, as described in Section I.D.1 of this document.
B. Certification Blank
The blank value associated with the cartridge acceptance criteria is discussed here. The criteria
for certification are taken from TO-11A and are very conservative; most results will be well within
these values. For the certification Blanks to be acceptable, the following criteria must be met:
» Formaldehyde: <0.15 jig per cartridge*
• Acetaldehyde; <0.10 jig per cartridge
Acetone: <0.30 jig per cartridge.
* The equivalent formaldehyde concentration for a 180-liter sample volume is 0.679 ppbv.
If the analysis of the three unsampled DNPH cartridges provide blank values with a mean plus three
standard deviations (• + 3s) that is less than the above criteria, then the sample lot is acceptable. If the
value is above the criteria, then additional blanks must be processed. The sample lot cannot be used
20
-------
unless the above criteria have been met. Field blank values should be consistently less than 2 times the
mean value of the certification blanks (i.e., < 2»).
A sampling system blank also should be determined for each sampler prior to and after the
ozone season. To collect a sampling system blank, the system is challenged with carbonyl- free air.
Carbonyl-free air can be generated by passing the incoming air through acidic DNPH solution in a
bubbling device, or through DNPH-coated cartridges. Alternatively, air containing a predetermined
level of carbonyls can be used. The same sampling procedure used for actual samples should be used
for the system blank (e.g., flow rate, time - see Section I.D.2).
C. Frequency of Obtaining Blank Data
As discussed earlier, a minimum of three laboratory blanks from each lot of DNPH cartridges
are required for certification of that lot. Also, as stated in the TAD, it is recommended that a number of
cartridges equal to the square root of the total number of samples be analyzed as field blanks. Table 3
gives a few examples of the minimum number of blanks per field samples. Table 4 provides a guideline
for tracking the certified and field blanks over time. Critical information for the table includes the
vendor; date of receipt of cartridge; lot number; expiration, extraction, and analysis dates; and lot
certified and field blank values.
Table 3. Minimum Number of Blanks Per Field Samples
Number of
Field Samples
50 : * ; '
100
200
Lab Blanks for
Certification of Sample Lot
3
3
3
Field Blanks Required
(square root of sample size)
7
10
14
V. Data Reporting
The collection and analysis of field blanks should be distributed over the entire period that the
cartridge lot is used for ambient air sampling. The data from Table 4 should be used to evaluate
21
-------
background carbonyl buildup over time. A trend plot also can be used to track background values
versus time. The trend plot should show the certified blank mean value and the ± 3s level for that
particular lot. Field blank values are then plotted in the trend plot as they become available (see Figure
1).
The PAMS Technical Assistance Document (Section 5.3.2) requires subtraction of the lot average field
blank value from all samples. However, this approach can sometimes obscure the relative magnitude of
the blanks and sample results. As a result, for data reduction and reporting purposes, field blank
subtraction for PAMS should not normally be done. The following approach is recommended:
1. If the field blank values are all within the • ± 3s range, then blank subtraction is not
necessary. Field blank results must be reported with the appropriate data set.
2. If the field blank mass loadings exceed the » ± 3s range but are still less than the
certification criteria (e.g., formaldehyde <0.15 jag per cartridge), then blank subtraction
is again not necessary. However, it is recommended that a new sample lot of cartridges
be integrated into the program immediately. Field blank results must be reported with
the appropriate data set.
3. If the field blank mass loadings exceed the certification criteria (e.g. formaldehyde
>0.15 jj,g per cartridge), then blank subtraction should be done and the sample lot of
cartridges should be phased out as quickly as possible. Field blank data must be
reported with the appropriate data set.
YI. Concluding Remarks
The purpose of this document is to provide more consistency in conducting carbonyl monitoring
in the PAMS program. This document focuses on improving consistency in five subject areas: sampling
system, sampling cartridges, analytical system, blanks and data reporting. This is considered a working
22
-------
document and PAMS participants are encouraged to provide comments and suggest improvements.
Please send any comments to : rice.joann@epa.gov.
23
-------
Table 4. Proposed Checklist Table for Tracking Cartridges
Vendor
Date of
Receipt
Lot
Number
Expiration
Date
-
Extraction
Date
,, ,___- ..... .
Analysis
Date
Lot
Certified
Blank
Value
Field Blank
Value
24
-------
0
0
^5
.p
^
c
T3
re
2
O)
0
c
re
c
1r\ r\
UU
90 -
80 -
70 -
60 -
c r\
OU
40 -
30 -
20 -
10 -
C
x
X X
x x x
x
I I 1 I
) 2 4 6 8 1
Day
0
Figure 1. Control Chart for Plotting Cartridge Field Blank Values Versus Time
(Certified blank mean is shown as the solid line with dashed
lines as the+/-3 standard deviation values)
25
-------
References
1. Code of Federal Regulations. Title 40, Part 58. Ambient Air Quality Surveillance, Final Rule
Federal Register, Vol. 58, No. 28, February 12, 1993.
2. Compendium of Methods for the Determination of Toxic Organic Compounds in Ambient
Air-TO-11A, U.S. Environmental Protection Agency, EPA/625R-96/Q10b5 Research Triangle
Park,NC, January 1997.
3. Technical Assistance Document for Sampling and Analysis of Ozone Precursors, U,S.
Environmental Protection Agency, EPA/600-R-98/161, Research Triangle Park, NC,
September 1998.
4. Assessment of Carbonyl Measurements at Photochemical Assessment Monitoring Stations
(PAMS) - Conference Calls, Battelle Report to IIS. EPA, Contract 68-D-98-030, Work
Assignment 2-03, January, 1999.
5. The Use of Kl-coated Copper Ozone Denuders for Carbonyl Measurements at PAMS. EPA
PAMSGRam, Volume #125 October 13,1998.
6. Compendium of Methods for the Determination of Toxic Organic Compounds in Ambient Air -
Second Edition, ADDENDUM, April 15, 1999.
7. U.S. Environmental Protection Agency. Code of Federal Regulations. Title 40, Chapter 1, Part
136, Appendix B, Office of the Federal Register, July 1, 1987.
26
-------
Pilot City Air Toxics Monitoring Study Guidelines DRAFT
Appendix E
'Methodology for Determining Carbonyl Compounds in Ambient Air'
Section 5 from the PAMS Technical Assistance Document for the
Sampling and Analysis of Ozone Precursors,
EPA/600-R-98/161
See: http://www.epa.gov/ttn/atntic/files/atnbient/pams/newtad.pdf
February 2001
-------
Pilot City Air Toxics Monitoring Study Guidelines DRAFT
Appendix F
Inorganic Compendium Method IO-3.5, "Determination of Metals in
Ambient Particulate Matter Using Inductively Coupled Plasma/Mass
Spectrometry, ICP/MS"
See: htto ://www.epa. gov/ttn/amtic/files/ambient/inorganic/mthd-3-5 .pdf
February 2001
-------
Pilot City Air Toxics Monitoring Study Guidelines DRAFT
Appendix G
Inorganic Compendium Method IO-3.1, "Selection, Preparation and
Extraction of Filter Material"
See: http://www.epa.gov/ttn/amtic/files/atnbient/inorganic/tnthd-3-l.pdf
February 2001
-------
Pilot City Air Toxics Monitoring Study Guidelines DRAFT
Appendix H
40 CFR Appendix B to Part 136 - "Definition and Procedure for
the Determination of the Method Detection Limit", Revision 1.11
February 2001
-------
Pt. 136, App. B
CfK Chi. I CM-QO Edition}
i ABLE 6. — Acin E:rM ^,1-i-
pD'.inn P023W-
' aiid 1-12 P
fS-'fi J5 'CJ
~r>™' 71
1:>-i J"5 55-' Ei
rtri-- so
'rl-8'f S7-2C3
ID ^31
7 ~^Z 1C-2S3
"iM!>u'U'
GS 1.1
I-.!,- 10
1:5 'la
US 13
t'j-- 1'j
Ci • ',7
U^ 13
SCO. • Z.T
J7-.TOS
•ifj-415
a™Sl2j'ti±jri: rial's:: an •cs'i'si.r rscovarv ilHfeaaj'tniiiLf bx.
^"^•flriiiiy r,j3c,-sry far ki.r rsco'jrirv nn.ri.-,j ijiti&Hjb.
EQD=Er%enL GJlJtlir.ec. Diwricn.
ns-ra sunr.iFGfilnn: Imt Is oilisida U't laiiy- Ihtil Kin hn -nra-siiiij J rs Ifltly.
CL. 26, iSJ-t; iil)FR IJS2, 6ffi, Jan. •!. JB3ii, as
G2 i"R 43tK, Sept. !5, WH7; «S I 'K 3044, Jan. IS. 2COO]
..I sil M l'''R23IO!. Jude 3fl. I&-IK;
IX B "t'U PA'JT (J6—DEF1KITIOK lytical method be iz^in'lul in the
METHOD
I.I!
'i'iii MD1_ nr-roins-rl by liiia prci:e:Jir-« l-s
!N!~.il li; jurlpr flio sig/liPiCiui-e of ^ r;lnglB
''
:£.IHI i._y i ri ii IjroEid variety-
Ttii.1 r[:ctjiod detection Minn. {MTJT.) is dj- -iieji:ii:nL i i ic 11 nit In. To aicomp lisli. i.lii-.. ilio
TuKu sis uht nninlmuTTi r.onk'CjUI-OUOii ulfc tj JuliL-hfdtu-y WOLJ irindc- Hr^^c- Of ix^U'i-iTJienl:-
yubaLiuite t.hst. run be me&Sureii uiid ftpiiit.cd iJidtpaoiIfcii!..
V.'itll OC% CDrtfiflftnri- lh;it Iho ^nnlyte COil- ^...nJ.
K^jK.i-,-.;.Uiri it HF^atci tfi^ui siiti ant] 1.1 wtur- f> "LiJ L
rpinod Tr-nni ^Jialyriis D^ » .ijirrj nit In £ giv'sn ' • ^-^^.*^ -1 n c-'^i inniTtc of rJie d^st&nl.ic n' i rnlt
(cj 'i'Jit: 4.rn renepnt (blajxkf wai.Ki- ciiiNJilnJ/iy, struracjiCEl mfissut-eiiKTi'i o:' Oic Eiialyis in
SnfLyte Lu w^.!,l:u».';iT^r cwit ™'~" "
The MRF. for aji aiisilv Lit-nl
x!riurR TTify (i:)'I'Eiiitreglnn of chc Staildaiii cui VK
iiry iiy u function of sample type. Tlitiim!<:•- Ji<:r<: is ,1 JslglllficjijiL i:luri|io in ;;C:L
1^.1 ro rccjuiros a completft. ^v-^iiOc, iir.ti well J.CL a brEak t:i i:h^ ^;I%rK^ ni rt-ir
leFI/iefl mial^(.ri;i). irKthcd. I. .5 eEssnr.isl cun.'E.
hill iill sample processing S(AJI^ itf Lte ana- (d) IiiitL"iniejujul liiriiLEiLions.
-------
Environmental Protection Agency
i l iu- recognised that tlie experience i>f I !i«
analyst is important to 1:liis [n'ru^ijin. l'sy the. (iii-ma»:i:
of interrfiring B.-jenlei (in1i«-'<;r^iPK), 'JJiL"
interfeienf. ttnnrsnvvar.isn '*•; [jrusuppoied 10
he iMi-maily iii*inl;uLud i^ rcprosentatlve
Min|.jl(*,s Ml'iv given rruiLrlx
3- (n) If the ?»iDL i» Lu Ue cicLcrmin«d In TU-
SLg-^nt (tiK-jikJ M'slui', prx-'p^rr: n lJ'->i^rg^Oi'lv''
ftrfliiihiiii h 11 ia i >• tr: in rci.grnt v.'acsi") at a c:fu\-
uftnlr-MliUTi wtiictt :s at least -equal rn <»• in
:.hu si'ir.c conrontratlDit ran^a ts i.li!i imi;-
>ri'.iLr;<1 mtthcd deteciioo I Unit:. lkni.:r.iiiiTri«sii1
b^t'VCCJX 1 £U'J^ Ij ijliftt?i List! H"i(,hfl^L^Xl EHjOLjlOG
cicU;cLiiiri lrrmi:..J Proceed to Slop -1,
(h) TK l-hn MTlL IE to be Jjlurrnincid ir. ,in-
odiar sample matrix, aii;iJ.y£L' Lh2 &Eiinple. If
the iueaj>iiLV!il !««! ;if ihu EniJytc is In the
j'efinniiin.-'iiiikMi r'iuyo of one :o five tiinKS f.ln>
vs Li i r lined tlu.citlon Slr.Ht, pmraafT 1.t> % I. up 4 ,
IF the jlK-£.iLu.td ieuMl ul ini;.i|ylu Is ]ci;<:
Lhi>n Uiu ^:lim;Lcrrrt rtftcct.ail lUniU, H-Jd H
KnnK'n ,Tmmnf of Enalj-'C* ix> yi'i;jK It,'.: Icwn;
of F.tiHlyts fc-ERvaeji niw a-iirj ."ivt times the es-
If C.nrt ii'icutiiiwl Jcvc) of EEislyte is
r.np.ii '\-,'c.- l;mcs £ha estiniaterJ
lirrlc, r^ni:-: ore (•«••!> OIJUULJS.
(i) Ohf.-Hii filiocliei1 iSuijjlu ivil.h :i I
Icval of auai^te ~ai T.ILR r,\mu*, Tri^r-rl?,- if
(2) The saraple of.y I ft n?ii,;l -,i^ is fo
ininlng the raetinirr fl-uscifwi limit if tha
^tjulyj.H tev^i i.lu',y irat exceed U clnies Llii>
IViDL. nf f.lici nnal>tO iJi rcs^junL vvatBr. The
vsirinnro of the analytical uiuLtiud t;h;irrrjni! 3;
tbc Enialj-xa •ccji-cenl.niH.ii.m inurenscs from the
?.-IE'L. liEJice the Ml 31 , itaLr.'i-;r.iiu:d und^-r
PL 13o, App, B
:TS may not truly
ariance ntlowe/ ana'v1:R cn
i. (=) i'dce a mi n I iii : ni i a " M-^-I.'I i ;-jK]ii0L^ or
i h:-: ml-irr nn,-ljrt-ci>l metliod. Maki' all com-
putations £CCftrcUri£ 1:n 3i^ (-^nr.'ii irKLliod
v.'ita final rssidra n't ili.r mrxiiud rcpcrting
u:ilfA IT n likirij;
i:AlcrliiLu Oi'j i.ic
isurimcn': is rsquJi'ed tcj
rrd Icvei of ar^lyt*. r,h-
r nc HNC irnTn:>n *
fbj It may
eslj'-.hle hi
b^ [ivcrcigc
fliam tll^ i«±ipm:-
y ;iiid technically
Will: (1) Pli'VtuL J CflKiii.il I tj Wil
i'w wlicicj *Js^ can*1;
^ i• l.i.i
i"i!h!;riijriu tiiat a Itraer cniRi>iiir*ii.^r;ii yf
t.ciiiLyco \v:ll not L'Hs;jll: in M siynilicsinily
Imver merhori cjRuciifjn limit. TEilte two
Klicjix^.t O' tb I sample to be u.--t:u. ut <:Mli:ii1;m;
tl'iB nettled de Eec ti-ij i LiruLt: 4]T5r1 prpro55
t.'-IA f.i-lira rrifl-Tind. 1 icluc'lng
jirujjiU; m; descnaed ahn'/ft in i;v.
these data
(i) If BiEEe mea=;iji'fiirii':iL's indicate the
Is in d^aii'^aijl^! r^r?|j^ Tc?r cctcriiiliui-
of Lit'. MUL, t.n-;^" Flvo nddi liuiuil
?JjCjLSl5 aild jjiuy^cjd. "LTs^ nil £8V^L" oiii^rUi'u
nis-nts fr>pn»li-i-liiiiun of the ,\!DL.
(2) Tf ilisvic mcEisiireii'.ents Jniiii::iH! iht!
sam^lft i^. nin in correct i'£i"fl&, n^^i'iruiilu
ihn "y,T>^.. fi'-Kain r.eiv Eamplft as ir 3 ;ITrn V—1 to IT"
6, [ii CouLULilz; lirac NO3, rjr,
-------
Pt, 134App. B
40 CFR Ch. I (7-1-00
IvJDL - Hie jiii LJJ« J [9uLL-i:ri limi
L>I,I-K .KI - t.hfi Rt.iirf Gilts' t va!ll£
ii g&i'.inifti t- viiltt n-l degrees nf fres-
durri, ira Table.
S = stenriaix-] tie? hit run or tine rsclScati snal-
('j) TIE; OS'S confldaiifH i ml. i?j'vi.! I'ad
f L' IVmjOuTri. Olijtrlbution fjf/cf).
Lt'.L - n.f,i VCDL
UCL-220MDL
whfeit : LOL aiirt t J HI . nrc the lower ia id upper
£151'* confidence liiuii.i, n^[Kir:f fiT HiC ciiirasite of tbe
MDLand s-.jlif.KimjTil MDL dEtsrmliiBtln -A.
(a) if: ill;: fs -.he initial attempt ia> f-.mmn.iit;
MD1 li;i.vvtl an tlia estlmss^i or MTlf. rtirmu-
Inccd ia"i SLiip I. Lafcs 1-hr: MDL OS •vl'.TiomJjiEitca' S"n.
The carnputed ?-i-al1is ij i.ht,'n cornpEii'ed \vi th
cbe i--j-atln rnmnJ in Lhc tibia which is 3.11* ua
Lii:-! hy r.h^
12
if S:.y'-S:n^X'.Ki rDsplke at thfi smjsl rt'.cnt
>:ul';ukil:cd JvIDL ami prfiwus .he; aomplss
Lhroush tha pLTK-fiflur^ suu-lng vdth Sr^p
•:. if rhfi iricwL ri'Ci-nt cslcuJatfi,- ,VJ1)L
ier, nest p^rjllic Ljudljl.-ji.ivn .ItiBtltlflca-
,'Jiesl aiiiiL|jiE:K nrn 5pH(Cii el Liji-L
level. i'fi[>isi-l I h:~ UtJl S3 a COitCuuli-uLien
hetv.iH^ri Uic current snr] |i>'^'ii>iii MDL.
v.'hirii in-scrnics qiic-lltative UitmiriaiLion.
(t) Use the 3r=rtal as calculated In Jb tn
PTnputc The final K-Sf Jl, nceoriiirjfi to the Tol-
ML>L-2.G8" (H,,.w,..j
wJicri! K.tlll 1.^ Rqi.t?; ro t:ii,i _—.«»}•
fr> Tin- a5!Jc coniidenoe limiis fi>r .
i-lvsi* in 7t JTC compute^] F.I :cr.it^l ina tu tiie
fiDli:i\.virg equations defivsd fruin proc-^ntileE
ni' ism i;hi .squarad avei- ^Iftgiwy ul' 1'iGcdom.
' "I.-0.72 MDL
LCL-i.uj txIDL
xvi icre LCL. and UCi_ *r\?. mv lower £iic
35'Ji conr'idEJicE limil
i^ nllquats.
TABUS OF STUDENTS' T V^LUZS AT TI :i 30
I.'DHT'S1 T V/-LUES Al Tr)5i 90
:\HDENCS Lt'-'tL— Co.iLiiiueri
ij(.i ma uf ruNlrrfnn
H ,., , , . ...
TS
:i3
= 1
:1
.;o , _ „
IS
1C
20-
fi1
i
« .H,:.
2 704
2 t-'-'B
Tim :iTi:.ilj"tirnl nittliDil usiuU must lie ipe-
cific-Elly Ltiejiirifpwi by number- or I: Lite nlil Hit
MHiT. TUT ijijich anshtfi s-ejn-tiisud in the sp-
JL-SIJJCJMI.K im-UiocI report Jnji uniis. I" ttic ar.a-
lytJtaL jriKLlmrl permits Optiujis wliiulj Jirj'i.'r;l
"hn m^Jh-DJ d^i;i't:LL(3n liirlr, thc^C Ciij idi'.iiwf^
•rni.'il- h:* .-iprr!Hed IVitb Liu MDI. v^lite, Tile
iaill|ilft nisLrix used 60 flfttei inirin f>io X-TDL
sraist aUo bs Idiiiliritxl v4dl MDL ''Siliiu, Ite-
[>url trie nc:m EiialyfA l-jvul -vich the MDE,
=nd inrilciii: -.f tra: MDL arnt rfi r.-y was
-^or?.".^cd. If a lauoirauir-y st?^ns!zu-d &r a -^iirn
rii.rii-jB.--ifn^leal2s B^'I"-!)
7 a
s $
10 , ,„ g
^i.«i
2.IW1
1VQa i_scxj Rir E
If tl'ie levul
U^3ow the de
.W»<
t:f u
t^i'rl
i'jl imLiuii, M'.SIJ reptrt
rfe in dlt; iiLiiri|]lr X-T.T-,
l \i7~>i . or es;ceeLa ID
aik;
-------
Environmental Protection Agency
Pf. 134 App. C
times the MDL of the amlyrs in ii$cigu-ii
wsier, tin Jint ra^cvr, ra vnliift'"ui' llm.HUL.
(« l-hM-«:in flrt- ).R. IfFM; 60 FR CM 69G, Jai.
4, JDSi, is ameiidtd at 51 ?R 23703. .Turns ill,
JOS3J
rx C TO PAKI- 156
COUPLED PLASMA— ATOMIC ?.MI«SIU.M
SPTXV^pMirjKIi; METHOD ?OR MfcClt
ELEMENT ANALYSTS or IVAVtai AMD
WAyrtt METHOD 5100.7
1, Scape s
i I T'i% lurched mail' be used far uis uu-
TerminQiLLGn sf dJssolvsri. si K| wnciL'd or total
•Ste.TL^JltE 111 df'iill: i;l£j vrtt 1 1! •, SI yfijcc Wi'pC'.T.
an-d domestic and mdns! rir1 1 Vril!j""u'£itcr£.
l.'£ Di-^SOl '.'^1 ttl^rnini i 3 ^i~c determined In
riJtwrtiEl Ttnri rtHrvflcifl 3£inplcs, A[.v^r J^JL iti_-±
strips miiEr >KI taken 1:1 all SUiU^ieis Ujcn:nr-u
thtC. pa~Gjltla] irVfirfeixinn1:- ;irn 1'MIcrn inl-n
account. This i.*, r.^sputiuSlv true whan d!s-
rt-iliils ttx&xd 1SDJ mgi'L. (See Section
1 3 Tutai clanp.-2iit5 are flei:ftr;iirm;M jil'U'r
appj'Opj'.iltjf ciij.;t!i-1.3-nrL r*mccrHt rcis ^fC pvi'-
foi'u tc-il. !^]7i(^ cii i^ifni/in TiaclhnlciU^S ljlCi'[:!a£^
the dlSSOSvoc SQlliE COlltfiJXf. fir 1 1m Nnrnjil;::;.
appropnatfi itsps miii. I in i.-iijijd 60 zorrsct
for potential interfere; i-.:» '.-ffi:t:.-^ (Sec Sec-
tion !i.)
1,1 I'rt'uJti ] ]iM.i K':i-r;ir.'iils for wtll-lh this
tu(RJiu(J npp'inr. nlong viitli i^CuxmuuDiLiii:
v« iviV r'Tirjrhs and typical s^tiuLiLcd inKl"Ti]-
muitBl Elatertiflfl llmi s xiaiiiK tonvcnlional
pneumatic nebuiijMtioi!, AoLuni ,^'nridnR ce-
1*d km. ]"]nilis Lint siHi'.ple depejirtsnf. nnd ys
T.lw ^inylc srmtrm: varies, tiiaaft ofnif:finl.r;H.
tlo'15 JllC;}' t^I^O Veil y. ill UU3IH. ^H.IlR^r
tTt^y be add«J ay mus^c mt'n--n-int'on bwC
avallahta ant. a*. ;v.ipim-.a.
(.;• Because of ^ tku UiutTv-nces befiveen
vaiioui jiialtiA Lif.ii sriuclcls ci' EadsiBctmry fci-
s.7: r.j •!! si i.s. ;KI d^li'.iicj l^sti'Uiiisntai njiLa-
H! Inj! Irutrnttlons can be pixwlc^fl. ft isl ixrf,
is i'tfc±r3'EO un thct tTi
Lhc Tritinnfrac-tU I'wr of the
a technique far
2.1 T.^e met.hfltl
the
in
l3w baulL o:'Lhi mcxlird Ig fho fliCEU'.itej riisriL
i>r ;L rniir r-mrs^'.nn !^v &n api_ic:LiL
si n.'i;ir'jf copse tBcluUijiia. nii:mili!ii ;rs
nuuuiiacd ^nd the £ei--nsrjl l.lifit i.s jjruthicsd is
tri'jispartad to the ^la.^irin mrtti wte-rc cxcl-
taQon Dccu.-fl. Clia[-rti;iwrist!t: atonac-li:iE
utrii?i!iiaii npf^tra are piOduCiiJ. Uy a rtvJiu-l'iu-
nii=nn^- li-xUlCtively Cuupicn pJa.'iinn (TCP).
. I:TIC iLi;eiiEjtle5 of thfi I hire ;TU
tri-' photojuul tijil i«r- tubuu, "J'hi
fram ih« rf
lubus iL%j rJrvrcsstd. ^™i cDncrnUari hy a onnx-
putEf sysam. A iiankp.rf ju i>i I ixn'cvilioiL Lucli-
jirfjiiK irt r»ji;uirwd Lo compensate i'cr variable
~ rti-Alyr^ K'flCS on S£l/il-
plnii rliiTfnj; nrr.lyjls. The (iOEitiO:! iuk'uLcd
for ich5 background iilljraisily mnMMirfT.-irnr,
on sltlifir m- harli .^iilijM uf Liii' inulytical [l^s,
\'i\}(. 'iv. tlk-Kji'iniriuTi -jy (±c tcmpte.xity of the
£;juclrum adjacent Co the e.ial'ytft liiift Th.'
posjUan used must Us frfift i'j " si[:tiu rul iriic;r-
ferencK and riHlnrt fh?- 5,; me cbiUJQC ii'i Utick-
jrrnunrf trtariKlfrj,- r»i ocelli's at Llx' niEilyLt:
h m=SSUfeo. "HArU^riniriri r-rirrrN:-
Is not L'cH|iiii'fi(l Tn ,pt."j'?i. ol' Lire: broad-
enliig v/iei'ft a i ij' i; kyrsmnrj. zorrcction
iner=Ftii\ii-r«i,i vwml;1 iutuiCiy degrade tlie
iiiiMl;-l1:;i.il result. Ths pOSSlJlll^1 Of ;iildi-
rlonrl I ntCi'fCTCHCCa j'uau.K[3 in S.] (anrt tmf'
D bft
li UKUJD
ynd appr3prlaf^ CO
J.I CiBoftwtf-ThoRft Mli.-iMj-il.:, which
pas= tiuougfi ali.-l.i \ r;i irufiilir^nc filter.
Ml !i>
li^.-;l!^ i ;fin {^rrrrnn fl ^) or the- Sum ui Lhi'
i'jls'cd plus auipenced sfjnK^nlrviiiurus. [Sic-
'.ion D.I pLjs !).!).
J.-J Tl^te/ tf.;:u';'jr;tijl?— The conce/itrar.i r>i!
deL^i-iijiiiHi tin ;in tinlUt-ji'Ed aamplft fttl
l(i>*lnjj u'iutrai^iv \vJtli h-ni. flilniei r]iniyp.i]
-S3K1 (Section LM).
S. 0- Jj?£li'ij^.'L.'.'jlj^ fAtv f.vr.lj"f7r7 ?/J7i>V — TT^C Cucj-
Cfcntf;U.if>n nr^isiv^ilr-ir tn PT ^••jnsL dud Lu (Jit
ai jalvi i> , -.vhiijf is equal ta three tiiriK-i (rir
f.rr>:iiiaril nuviii.iuii ol'a aE-i=Bof ten rv>Miii^u
iii^t^irujr.iriiji ui' a reagent tila.ik siyiiu! ill
'Jl the £DEUyc!cal
J.li Asa/cjVjiy~— Tlift H
3.7 InaraiOKJt rfjA* .Wai.'lartt — A mn!r1nlA-
mcnc star.dard iaf ?
-------
TECHNICAL REPORT DATA
(Please read Instructions on reverse before completing)
1. REPORT NO.
EPA-454/R-01-003
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Pilot City Air Toxics Measurement Summary
5. REPORT DATE
2/2001
6. PERFORMING ORGANIZATION CODE
USEPA/OAQPS/EMAD/MQAG
7. AUTHOR(S)
Joann Rice
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Research Triangle Park, NC 27711
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
NONE
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 27711
13. TYPE OF REPORT AND PERIOD COVERED
Measurement summary
14. SPONSORING AGENCY CODE
EPA/200/04
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This summary of guidelines provides agreed-upon measurement procedures needed to provide consistent data for the FY2000
National Air Toxics Pilot Monitoring Program, initiated by EPA through the S103 Grant Process. Although the details of the
procedures herein are not entirely consistent with the Compendium of Methods, items called out are specific to the data quality
goals of the Pilot Monitoring Program.
This document is not policy and does not contain legally binding requirements, nor is it regulation. While it presents
recommendations and suggestions regarding techniques for the measurement of toxic air pollutants for the Program discussed
above, it may not be appropriate for other situations. This document is intended for use by those already familiar with the analysis
of field samples for volatile organic compounds (VOC's), carbonyl compounds and metals.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
VOC Particulates
carbonyls toxic monitoring
metals Replicates, duplicates, field blanks
MDL filters
TSP
Air Pollution control
Air Pollution monitoring
18. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (Report)
Unclassified
21. NO. OF PAGES
65
20. SECURITY CLASS (Page)
Unclassified
------- |