Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
AIR TOXICS MONITORING
CONCEPT PAPER
Office of Air Quality Planning and Standards
Revised Draft
February 29, 2000
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
AIR TOXICS MONITORING CONCEPT PAPER
Table of Contents
1 The Air Toxics Program and the Role of Ambient Air Monitoring 2
1.1 Background 2
1.2 National Air Toxics Program 2
1.3 National Air Toxics Assessments 3
1.4 The Role of Ambient Monitoring in NATA 4
1.4.1 Air quality characterization, model evaluation and trends assessment 4
1.4.2 Exposure and Risk Assessment 6
2. Current Federal, State and Local Air Toxics Monitoring Activities 8
3 Design of National Ambient Monitoring Network to Support Air Toxics Program .16
3.1 Air Toxics Monitoring Objectives and Principles 16
3.1.1 Measure pollutants of concern to the air toxics program 16
3.1.2 Use scientifically sound monitoring protocols to ensure nationally
consistent data of high quality 17
3.1.3. Collect a sufficient amount of data to estimate annual average
concentrations at each monitoring site 17
3.1.4 Reflect "community-oriented" (i.e. neighborhood-scale ) monitoring
locations 18
3.1.5 Comply with uniform siting guidelines 18
3.1.6 Represent geographic variability in annual average ambient concentrationJ,8
3.1.7 Build upon existing national and State/local monitoring programs 19
3.2 Strategic Air Toxics Monitoring Approach 19
3.2.1 Make use of existing monitoring sites 19
3.2.2 Perform data analysis/assessment 21
3.2.3 Initial Focus on Model Evaluation 22
3.2.4 Longer-term trends network 23
3.2.5 Incorporate long-term and short-term monitoring elements where possible
23
3.2.6 Allow for temporary air toxics monitoring activities 24
3.2.7 Integrate air toxics and other monitoring 24
3.2.8 Utilize standard monitoring methods 24
3.2.9 Enhancement of the PAMS for monitoring toxic VOCs 26
3.2.10 Incorporate measurements for other HAPs when possible (Support research
and development efforts to permit new and better analyses) 27
3.2.11 Review network periodically 28
4 Initial Network Development 29
4.1 Preliminary schedule of monitoring and data analysis activities 29
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
4.1.1 Activities during 3-year rollout: 29
4.1.2 Post Rollout Activities 29
4.2 FY-00/01 activities 30
4.3 Initial Monitoring Protocols 31
4.4 Detailed FY-00/01 Monitoring Plan 32
ATTACHMENTS 33
References 41
1
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
AIR TOXICS MONITORING CONCEPT PAPER
1 The Air Toxics Program and the Role of Ambient Air Monitoring
1.1 Background
There are currently 188 hazardous air pollutants (HAPs), or air toxics, regulated under the
Clean Air Act (CAA) that have been associated with a wide variety of adverse ecosystem and
health effects, including cancer, neurological effects, reproductive effects and developmental
effects. These air toxics are emitted from multiple sources, including major stationary, area, and
mobile sources, resulting in population exposure to these air toxics . While in some cases the
public may be exposed to an individual HAP, more typically people experience exposures to
multiple HAPs and from many sources. Exposures of concern result not only from the inhalation of
these HAPs, but also, for some HAPs, from multi-pathway exposures to air emissions. For
example, air emissions of mercury can deposit in sediments and water where they are
biotransformed into methyl mercury and then bioaccumulate in fish, resulting in exposures to
people who eat the fish.
Our current Government Performance Results Act (GPRA) commitments specify a goal of
reducing air toxics emissions by 75% from 1993 levels to significantly reduce the risk to
Americans of cancer and other serious adverse health effects caused by airborne toxics. Because
of our limited tools to assess the impacts of these emissions on public health and the environment,
we are focusing on reducing emissions to the extent possible. However, as we develop new
assessment tools and begin to address the public health and ecological risks associated with these
emissions as required by the CAA, we will be modifying that goal to one that focuses on risk
reductions associated with exposure to air toxics. In working toward this risk-based goal, we will
first focus on the cumulative effects of air toxics in urban areas, the multi-media effects of air
toxics on water bodies whose water quality and aquatic life are affected by the deposition of
persistent and bioaccumulating air toxics, and the effects on sensitive populations and on
economically disadvantaged communities. Eventually, we have a long-term goal of eliminating
unacceptable risks of cancer and other significant health problems from exposures to air toxics
emissions and to substantially reduce or eliminate adverse effects on our natural environment.
1.2 National Air Toxics Program
To address the concerns posed by air toxics emissions and to meet our strategic goals, we
have developed a national air toxics program designed to characterize, prioritize, and equitably
address the impacts of HAPs on the public health and the environment. The national air toxics
2
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
program seeks to address air toxics problems through a strategic combination of several agencies'
activities and authorities, including regulatory approaches and voluntary partnerships. We
envision four key areas of activities:
• Source-specific standards and sector-based standards, including Section 112
standards, i.e. Maximum Achievable Control Technology (MACT), Generally
Achievable Control Technology (GACT), residual risk standards, and Section 129
standards.
• National, regional, and community-based initiatives to focus on multi-media and
cumulative risks, such as the Integrated Urban Air Toxics Strategy, Great Waters
and National Estuary Program, Mercury initiatives, Persistent Bioaccumulative
Toxics (PBT) and Total Maximum Daily Load (TMDL) initiatives, and Clean Air
Partnerships.
• National Air Toxics Assessment (NATA) activities which will help EPA identify
areas of concern, characterize human health and ecosystem risks and track progress.
These activities include expanded air toxics monitoring, improving and
periodically updating emissions inventories, national- and local-scale air quality
and exposure modeling, and continued research on effects and assessment tools.
These efforts will lead to improved characterizations of air toxics risk and
reductions in risk resulting from ongoing and future implementation of air toxics
emissions control standards and initiatives.
• Public education and outreach.
The NATA activities, as discussed below, will be critical to the success of all the other
major areas of activities within the national air toxics program.
1.3 National Air Toxics Assessments
The success of the national air toxics program critically depends on our ability to
understand and quantify the impacts of air toxics emissions on public health and the environment.
To that end, EPA has initiated numerous NATA activities. All of these activities are aimed at
providing the best technical information regarding air toxics emissions, ambient concentrations,
and health and environmental impacts to support the development of sound policies in the national
air toxics program. These activities include:
• the measurement of air toxics emission rates from individual pollution sources;
• the compilation of comprehensive air toxics emission inventories for local, State,
3
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
• and national domains;
• the measurement of ambient concentrations of air toxics at monitoring sites
throughout the nation;
• the analysis of patterns and trends in ambient air toxics measurements;
• the estimation of ambient and multimedia air toxics concentrations from emission
inventories using dispersion and deposition modeling;
• the estimation of human and environmental exposures to air toxics, and;
• the assessment of human and environmental risks due to air toxics;
• ongoing research in above areas to improve assessments over time.
The wide range of NAT A activities listed above illustrates that emissions data, ambient
concentration measurements, modeled estimates, and health and environmental impact assessments
are all needed to fully characterize air toxics impacts and to determine risk. Specifically,
emissions data are needed to quantify the sources of air toxics impacts and aid in the development
of control strategies. Ambient monitoring data are then needed to characterize air toxics ambient
concentrations and toxics deposition, to better understand the fate and transport of air toxics in the
atmosphere and to help evaluate atmospheric dispersion and deposition models. Since ambient
measurements cannot practically be made everywhere, modeled estimates are needed to
extrapolate our knowledge of air toxics impacts into locations without monitors. A combination of
reliable modeling systems along with well designed ambient networks is the best approach for
estimating ambient concentrations and population/ecosystem exposure across the nation. Exposure
assessment information and health effects information need to be integrated in order to characterize
air toxics risks. Ambient measurements provided from routine monitoring programs together with
personal exposure measurements which currently can be obtained from ongoing research studies
are important to evaluate these air quality and exposure models.
1.4 The Role of Ambient Monitoring in NATA
1.4.1 Air quality characterization, model evaluation and trends assessment
This concept paper focuses on the role of ambient measurement data as one key element of
the full air toxics assessment process. The purposes for collecting ambient monitoring data should
be kept in mind when designing and implementing the measurement networks.
For the initial role of monitoring to support NATA, we anticipate that ambient air toxics data will
be useful to:
• Characterize ambient concentrations and deposition in representative monitoring
areas,
4
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
• Provide data to support and evaluate dispersion and deposition models, and
• Establish trends and evaluate the effectiveness of HAP reduction strategies.
In addition, initial new monitoring together with data analysis of existing measurements will be
needed to provide a sufficient understanding of ambient air toxics concentrations throughout the
country in order to decide on the appropriate quantity and quality of needed data.
As part of the overall NATA process, ambient air quality data are important to help assess
the national toxics inventory and national air toxics modeling and later long-term HAP trends.
Initially, ambient air quality monitoring should focus on characterization (to assess
community-wide concentrations in urban areas and ecosystem impact, and to quantify ambient
conditions in the vicinity of localized hot spots or specific areas of concern like schools) and to
provide data to support or evaluate models. As the monitoring program matures, trend sites can be
established to assess the effectiveness of the air toxics program components.
To achieve our purposes, we must develop a monitoring network which is representative
of air toxics problems on a national scale and which provides a means to obtain data on a more
localized basis as appropriate and necessary. The appropriateness of a candidate monitoring site
with respect to the data uses described above and the timing for needed data will be the key
considerations in identifying sites for the initial data collection activities and for the development
of a longer-term national network.
A key component for the air toxics monitoring network is the HAPs that should be
measured. It is not possible to measure all HAPs at all locations. Recognizing this, the CAA
amendments required EPA to develop a subset of the 188 toxics identified in Section 112 which
are thought to have the greatest impact on the public and the environment in urban areas. This
subset is comprised of the 33 HAPS identified in the Integrated Urban Air Toxics Strategy
(UATS)1 commonly referred to as the "Urban HAP List." Because this list was developed to
reflect a variety of possible exposure periods (acute/chronic), pathways (inhalation, dermal,
ingestion), and types of adverse health effect (cancer/noncancer), the toxics monitoring network
should attempt to address the full list. However, the near-term monitoring discussed in this paper
primarily focuses on ambient air monitoring and does not completely address the pollutants for
which ingestion are the principle route for exposure.
Considering the chemical properties of these HAPs, they can be grouped into several
general categories (Table 1) which include volatile organic compounds (VOCs), metals,
aldehydes, and semi-volatile organic compounds (SVOCs).
'The strategy was finalized on July, 19 1999 (64 FR 38705).
5
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
Table 1 - List of Urban Air Toxics HAPs
VOCs
Metals
(Inorganic Compounds)
Aldehydes
(Carbonyl Compounds)
SVOCs and other HAPs
acrylonitrile
arsenic compounds
acetaldehyde
2,3,7,8-tetrachlorodibenzo-p-dioxin (&
congeners & TCDF congeners)
benzene
beryllium and compounds
formaldehyde
coke oven emissions
1,3-butadiene
cadmium compounds
acrolein
hexachlorobenzene
carbon tetrachloride
chromium compounds
hydrazine
chloroform
lead compounds
poly cyclic organic matter (POM)
1,2 -dibromoethane
(ethylene dibromide)
manganese compounds
polychlorinated biphenyls (PCBs)
1,3-dichloropropene
mercury compounds
quinoline
1,2-dichloropropane
(propylene dichloride)
nickel compounds
ethylene dichloride, EDC
(1,2-dichlore thane)
ethylene oxide
methylene chloride
(dichloromethane)
1,1,2,2, -Tetrachloroethane
tetrachloroethylene
(perchloroethylene, PCE)
trichloroethylene, TCE
vinyl chloride
Given the importance of these pollutants and the impracticality of monitoring for all 188 HAPs on
a national basis, we feel that initial monitoring efforts should focus on a subset of HAPs which
contains most or all of the HAPs on this list. This is not to say that all UATS HAPs should be
measured at all sites and that other HAPs should not be measured in specific situations. This issue
is discussed further in Section 3.1.1.
1.4.2 Exposure and Risk Assessment
To understand and properly quantify the health and environmental risks associated with
ambient emissions of air toxics, it is important to know to what levels of a pollutant people and
ecosystems are actually exposed. In general, ambient air concentrations, produced by fixed station
monitors, do not directly estimate long-term human inhalation exposures (although they may be
appropriate for ecosystem exposure). Such exposures are either measured with personal monitors,
6
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
which follow a human subject through time and space, or are predicted with exposure models, which
simulate long-term human activities. However, ambient monitors indirectly provide information that
is essential to a proper exposure and health risk characterization.
To date, long-term widespread databases of personal exposure monitoring for many pollutants
are limited (and have been developed primarily by organizations outside of the agency). Thus, most
inhalation exposure characterizations currently rely on model predictions of inhalation exposure. A
key component to these models is to properly characterize the concentration in the different flindoor
and outdoor places® where people spend their time (called flmicroenvironments or MEs@). Research
has shown that for many pollutants there is a definitive relationship between the outdoor ambient
concentration and that found in these MEs (i.e., home, vehicles, workplace, park, ...). Thus, in most
exposure models, the outdoor ambient concentration along with ME relationships and human activity
pattern data (an accounting of which MEs people spend their time in) are used to predict human
inhalation exposure concentrations. With adequate temporal and spatial coverage, ambient monitor
data can serve as the required outdoor ambient concentration for these exposure models. Where
adequate coverage does not exist, exposure assessments can rely on air dispersion models to provide
the air quality data at the required temporal and spatial coverage.
When evaluating exposures from criteria air pollutants (ozone, carbon monoxide, etc.), past
regulatory exposure assessments have relied on ambient measurements from fixed-site monitors for
use in exposure models. This could be accomplished because routine long-term ambient monitoring
data for such pollutants were available to a high degree of spatial resolution in many metropolitan
areas. For exposure assessments in support of the ozone national ambient air quality standard
development, 6-16 monitoring sites in 9-10 areas around the country have been used to help estimate
concentrations in MEs. For most air toxics pollutants, a comparable spatial monitoring resolution is
generally not available. As a result, exposure assessment for air toxics are typically driven by
ambient concentration estimates from dispersion models. In addition to filling the void of assuring
adequate spatial coverage, dispersion models also have the ability to predict future concentrations or
evaluate the past effects of various emissions scenarios on ambient concentrations. For example, EPA
is currently performing a national screening assessment which will calculate human exposures based
on modeled ambient levels from a nationwide dispersion model (the Assessment System for
Population Exposure Nationwide, or ASPEN). The ASPEN system calculates these ambient levels
based on a knowledge of meteorology, chemistry, and rates at which air toxics pollutants are emitted
into the atmosphere from all man-made sources in the nation (this information is compiled in EPA's
1996 National Toxics Inventory, NTI). The ambient concentration outputs from ASPEN are then used
to calculate human exposures using the Hazardous Air Pollutant Exposure Model (HAPEM4).
Estimated exposures from HAPEM4 will then be combined with quantitative health impact
information to estimate population health risks estimates. Thus, as noted in Section 1.4.1, the role of
ambient monitoring data in the ASPEN model evaluation process will be an essential step in assuring
the appropriateness of the predicted exposure and health risk estimates.
7
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
2. Current Federal, State and Local Air Toxics Monitoring Activities
Through a combination of Federal, State, local and tribal monitoring activities, there are
approximately 300 sites currently collecting ambient data on hazardous air pollutants.
To support some States with the measurement of air toxics in ambient air, OAQPS initiated
the Urban Air Toxics Monitoring Program (UATMP) in 1988. Since its inception, this voluntary
contractual support program has been used by many State and local agencies to assess the nature and
magnitude of various air toxics problems. The current UATMP supports the collection and analysis
ofVOCs, aldehydes and other HAPs (See attachments 4-7). For this program, one 24-hour integrated
sample is collected every 12 days; the sample is sent to a laboratory for analysis and the resultant
concentration values are reported to EPA's Aerometric Information Retrieval System (AIRS)
database. To participate in the program, States commit funds to the OAQPS contract and receive a
year's worth of monitoring support from the contractor.
Other air quality monitoring programs also provide information on HAPs. The Photochemical
Assessment Monitoring Stations (PAMS) program has been collecting air toxics data for 9 volatile
organic compound (VOC) HAPs since 1993 in more than 20 major urban areas. The HAPs on the
PAMS list are:
Formaldehyde
Toluene
• Benzene
• Ethylbenzene
Xylenes (m,p-xylene and o-xylene separated on PAMS list)
• Hexane
Styrene
2,2,4-Trimethylpentane
Acetaldehyde
These data are stored in the AIRS database, along with all criteria pollutant monitoring data. In the
near future, PM2 5 speciation monitors will provide measurements of 10 HAP metals at over 50 urban
locations in the country. Rural and remote monitoring of these HAP metals takes place as part of
EPA's efforts to assess regional haze. These data are produced by the IMPROVE and CASTNET
networks. The current network of IMPROVE sites will be expanded to 110 Class I monitoring areas
(national parks and wilderness areas) for calendar year (CY) 2000 data collection. CASTNET has
eight rural sites in the eastern US. In addition to these networks, the Mercury Deposition Network
(MDN), a sub-network of the National Atmospheric Deposition Program (NADP), provides regional
measurements of mercury deposition in precipitation at 37 sites, while the Great Lakes and National
Estuary Program networks measure a variety of HAPs at more than a dozen sites. Details about a new
dioxin monitoring network are provided at the end of this section.
8
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
A recent STAPPA/ALAPCO survey identified 252 sites which currently monitor for air toxics
and another 150 sites which plan to monitor in CY-2000. At least 32 States will be involved. The
number of current and planned monitors by State, and the type of HAPs monitored are shown in
Attachments 1 -3. Through a voluntary effort of States, local agencies and private entities to report air
toxics monitoring data, EPA has also inventoried air toxics monitoring data. A summary of the number
of States who have monitored a specific HAP in 1997 and reported the data to EPA is shown in the
table on the following page. Work is underway to integrate the results from the recent
STAPPA/ALAPCO survey with the combined AIRS and archive data bases. As soon is this has been
completed, the summary table will be updated to reflect the expanded air toxics informational data
base2.
2Note that some HAPs do not appear in the current inventory of monitoring information
because there are no demonstrated or routine monitoring methods. (Refer to the discussion in
Section 3.2.8). For other pollutants, the methods can be very expensive (e.g. dioxin and cogeners).
10
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
Table 1. Ambient Measurements of Urban Air Toxics HAPs (States in 1997)*
VOCs
#States
Metals**
#States
Aldehydes
#States
SVOCs and other HAPs
acrylonitrile
arsenic
compounds
11
acetaldehyde
16
2,3,7,8-tetrachlorodibenzo-p-
dioxin (& congeners & TCDF
congeners)
benzene
28
beryllium
and
compounds
formaldehyde
16
coke oven emissions
1,3-butadiene
cadmium
compounds
11
acrolein
hexachlorobenzene
carbon tetrachloride
(tetrachloromelhane)
chromium
compounds
10
hydrazine
chloroform
(trichloromethane)
11
chromium
VI
poly cyclic organic matter (POM)
112(k)
1,2 -dibromoethane
(ethylene dibromide)
lead
compounds
33
polychlorinated biphenyls (PCBs)
1,3 -dichloropropene
manganese
compounds
quinoline
1,2-dichloropropane
(propylene dichloride)
mercury
compounds
ethylene dichloride,
EDC
(1,2-Dichlorethane)
10
nickel
compoiuids
ethylene oxide
methylene chloride
(dichloromethane)
14
1,1,2,2,-
tetrachloroethane
tetrachloroethylene
(perchloroethylene,
PCE)
13
trichloroethylene, TCE
12
vinyl chloride
(chloroethylene)
10
States reporting data to AIRS or captured in the OAQPS Air
oxics Archive Data Base
** Most metal analyses are from TSP. A fewer number are from PM10 or fine particles.
Note: More States reported some specific HAPs in 1996 at the time of this survey. MDN network information
has not been included.
11
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
As examples of the extent of current State and local agency monitoring coverage for urban
HAPs, three maps are presented. First is a map of the 1997 monitoring sites measuring and reporting
benzene (the most commonly measured HAP). The second map shows arsenic data, exemplifying the
lesser coverage for ambient concentrations of HAP metals.
12
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
The third map, presented below provides the most recent information on the Mercury Monitoring
Network for 1999.
National Atmospheric Deposition Program
Mercury Deposition Network
h & WA1G
\ /
-j s
'.CAST i
T \
1
"~T— - 1
CO*7
"i
v\
A, o f
>f
: \l#>:
=005 , ^ ^
<*W WK/
**¦ (iwi
MtSS
MEO!
1
MNie 'w^.' -i
;T I Wl"y I r t^>:
A-'? rJ^* nhm
\ ^ i .PAivJ
V - T1-
\ / r r j '"so
J I - J ( PA3JC. -ffc
I—I
NM10
t
r
rn f—
J '^1.^10,1
J*-"
v
¦'
i V
I QAQ3 ?
FL05 ¦ >
?FL34-.
FL-ll
FL04
13
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
The map below shows the locations of all PAMS areas (which currently provide some daily
monitoring of VOC and aldehyde HAPs) and the locations of all proposed PM25 speciation trend sites
which will provide information on HAP metals. Some of these sites will be good platforms for
expanded air toxics monitoring.
PAMS/PM2j5 Cities
* PAMS
* PM2£
States
Cities with PAMS & PM2.5 Platforms
14
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
The rural IMPROVE and CASTNet Networks currently produce data on trace metal HAPs and
may be useful platforms for other HAP monitoring.
IMPROVE and CASTNet Monitoring Networks
(as of February 2000)
15
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
As seen in Table 1, dioxin is not routinely monitored by State and local agencies. A new Dioxin
Exposure Initiative has identified the need for establishing a nationally-based,
long-term, ambient air monitoring network. Known as the National Dioxin Air
Monitoring Network (NDAMN), it is designed to serve some of the same objectives mentioned
earlier:
S Provide ambient air data useful for calibrating regional-scale long-range transport
models used in estimating air concentrations of dioxin as a function of dioxin
source emissions;
S Provide air monitoring capability for the occurrences and levels of dioxin-like
compounds in areas where animal feeds (used to feed domestic livestock) are
primarily grown;
S Provide for the long-term monitoring of dioxin-like compounds in different regions
of the United States, and over different seasons.
S Provide data on potential transboundary import of CDDs/CDFs into the United
States.
Details about data quality objectives and analytical methods for NDAMN are available elsewhere
(ref: www.epa.gov/nceawwwl/lpage.htm) Phase 1 is a limited deployment of the network in
areas that are west of the Continental Divide where deposition is occurring downwind of source
activities, and in regions where agricultural activity dominates (approximately 10 to 20 air
monitoring stations). Phase 2 will consist of approximately 30 to 40 air monitoring sampling
stations throughout the U.S.
The aforementioned air toxics monitoring stations and related ambient air monitoring
networks can support the development of a new national network. An adequate geographic
distribution of locations and HAP measurements is desired to satisfy the needs of the public health
and environmental components of the air toxics program. The next section will identify the short-
term and long-term objectives and outline a strategic air toxics monitoring approach. This process
will identify how to best utilize the current monitoring stations and networks, what changes or
enhancements are needed, and what activities should be undertaken to satisfy the monitoring
objectives and to develop a national network.
16
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
3 Design of National Ambient Monitoring Network to Support Air Toxics Program
3.1 Air Toxics Monitoring Objectives and Principles
The monitoring network should be incrementally designed to first address the highest
priority needs of the air toxics program: to focus on pollutants and sources which pose the greatest
risk to the largest number of people and the greatest risk to the environment. Over time, the
network will evolve to satisfy the following monitoring objectives:
3.1.1 Measure pollutants of concern to the air toxics program.
Because of our limited knowledge in measuring many of the 188 HAPs, the program should
initially focus on those pollutants that EPA, State and locals agencies have identified as having the
most significant potential health impacts and routine measurement methods. The list of HAPs in the
UATS is a logical starting point. Several UATS pollutants are also important from an ecosystem
risk assessment perspective (POMs and toxic trace elements). As additional priority air toxics are
identified and as monitoring capabilities improve, additional HAPs can be added to the monitoring
program.
This is not to say that all UATS HAPs must be measured at all locations and that non-
UATS HAPs should not currently be measured at some locations. To permit comparisons among
HAPs and to facilitate dispersion and deposition model evaluation, however, some number of
"core" UATS HAPs should be initially measured at a number of locations nationwide. These
HAPs should reflect the practicality of their measurement and include those associated with
highest toxicity-weighted emissions or those that are judged responsible for a large percentage of
the risk associated with exposures to ambient air toxics. Similarly, as many UATS HAPs as
possible should be measured at an agreed upon, albeit initially small, number of
"comprehensive"monitoring locations. Such comprehensive platforms should be selected to
reflect a broad representative mix of UATS HAP emissions. As monitoring capabilities improve
and available resources increase, the list of compounds and locations can change.
For those States and jurisdictions where it can be shown that ambient concentrations are
consistently below the minimum detectable concentrations of current monitoring methods or are
below established health based exposure levels, the target list of HAPs may be modified to
exclude these compounds. As noted later in this paper, the target list of HAPs should be
periodically reviewed and modified accordingly as monitoring methods or health based levels are
revised.
17
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
3.1.2 Use scientifically sound monitoring protocols to ensure nationally consistent data of
high quality.
Appropriate sampling and analytical methods should be followed. Emphasis should first be
placed on standardized methods and analysis and then on supporting quality assurance efforts
between laboratories. Regional or national laboratories could also be helpful to ensure consistent
analysis. The methods must consider risk-based concentrations and must be sufficiently sensitive
to provide an adequate limit of detection. The monitoring protocol must provide for adequate
quality assurance and data management. A Quality Assurance Project Plan (QAPP) is currently
under development by EPA.
3.1.3. Collect a sufficient amount of data to estimate annual average concentrations at each
monitoring site.
The required frequency of sampling is guided by a variety of factors which include the day-
to-day variability in HAP concentrations, the measurement precision of the sampling and analysis
methods, cost, and the intended purpose of the data. Air toxics measurements which are collected
for making comparisons to dispersion and deposition model predictions will require lower
precision and, possibly, less frequent sampling than data needed to detect trends in ambient
concentrations. Sampling frequencies can vary seasonally with more frequent sampling needed in
seasons or time periods with higher source and meteorological variability.
The selection of sampling frequencies will be guided by the data quality objective (DQO)
process. DQOs are currently under development and will provide recommendations on the need to
sample according to the typical UATMP sampling frequency of once every 12 days, the frequency
for VOCs collected at PAMS sites of one in six days or the frequency to be used at PM2.5
speciation trend sites and IMPROVE sites of once in three days. Because of different data
collection objectives, the sampling frequencies may differ during the first few years of the
nationwide air toxics monitoring effort than the frequencies used for the longer-term program.
Sampling frequencies and sampling time periods may also be HAP specific especially in
the case of existing national networks. Consideration of what defines "sufficient data" for a
particular pollutant depends on the day-to-day concentration variability, as well as estimated
precision and accuracy of the monitoring method. While 24-hour samples may be the goal from a
health risk assessment perspective, there should also be flexibility to use alternate frequencies and
sampling times to accommodate different needs. In particular, compounds requiring lower
detection limits may require sampling times greater than 24 hours. Composite samples collected
for stable compounds may offer a better and a more cost effective screening technique. Finally,
staggered sampling schedules across a network (i.e. sites sampling on the same frequency but on
different days) can possibly provide better information on day-to-day or pollutant event behavior.
18
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
3.1.4 Reflect "community-oriented" (i.e. neighborhood-scale ) monitoring locations.
Fixed-site air toxics monitors to support model evaluation and health-based trends
assessments should be sited to be representative of average concentrations within a 0.5 to 4 km
area. In EPA monitoring regulations, this geographic imprint of a monitoring site is termed
"neighborhood scale." Such measurements are intended to be representative of area-wide,
average model predictions. In fact, locations representative of larger areas and populations are
preferred for this purpose. Separate State, local or tribal toxics monitoring efforts may focus on
smaller scale (i.e middle or micro-scale) monitoring sites. All spatial scales of monitoring may be
conducted using fixed-site, or movable monitoring platforms.
3.1.5 Comply with uniform siting guidelines.
Monitoring locations should comply with established siting guidelines to permit unbiased
comparisons among metropolitan areas and geographic regions of the country. General siting
criteria for PAMS site Type-2 and PM2 5 core monitoring stations can be followed when
establishing the desired neighborhood-scale air toxics monitoring sites. These guidelines provide
specifications on set back distances from major roads and stationary sources, inlet heights, and
other siting considerations.
3.1.6 Represent geographic variability in annual average ambient concentrations.
A national network should represent a variety of conditions and environments that will
permit characterization of different emission sources and climatological/meteorological
conditions. Such a network would support dispersion model evaluation, understanding of the
relationships between emissions and air quality under different circumstances and allow for the
tracking of changes in emissions. National assessments should reflect the differences among cities
and between urban and rural areas for selected HAPs. Accordingly, the network should reflect the
following network design goals:
• Include some areas with potentially high anticipated concentration,
• Distinguish differences within and between geographic regions (to describe
characteristics of different climatic areas,
• Determine background concentrations, and
• Reflect relative variability and the spatial patterns of HAPs across communities.
For model evaluation, trends assessments and other data uses, the national network of
ambient air toxics monitoring should reflect the diversity of mobile, area and stationary sources
across the country as well as providing important information about upwind or regional
background locations (in rural areas), and intra- and inter-regional transport of HAPS. National
emission inventories and dispersion and deposition model output can be used to help characterize
20
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
data needs and to help identify inadequacies, important gaps or redundancies among the existing
monitoring sites and data. With the measured concentrations from the monitoring network, the
national emission inventories and dispersion and deposition model output can be evaluated. In
particular, it will help assess the incongruities between singular measurement locations with grid
based modeled outputs. This is an iterative process.
3.1.7 Build upon existing national and State/local monitoring programs.
By building the Nation's air toxics monitoring network upon existing monitoring programs
(adding new sites where necessary), large scale efficiencies of resources can occur. For this
reason, the program should maximize use of appropriate existing platforms (where there is
consistency with air toxics monitoring objectives) and take advantage of mobile monitoring and
saturation monitoring resources. State and local agencies and tribal organizations can use the
NATA ASPEN modeling results to help site new monitors.
3.2 Strategic Air Toxics Monitoring Approach
The national ambient air toxics network will take several years to implement. Two to three
years are needed to collect and analyze available air toxics monitoring data in order to better
understand the sources of spatial and temporal variability of ambient air toxics and to establish the
basis for the long-term monitoring program. These activities include: developing an accurate
inventory of current air toxics monitoring activities {i.e. integrating results of STAPPA/ALAPCO
survey and private monitoring activities with the EPA archive of air toxics monitoring data);
enhancing existing monitoring efforts and initiating new monitoring where needed to
cost-effectively fill initial data gaps and to address network design issues; performing new
monitoring and analyzing new and existing air quality data to better understand air quality patterns
(spatial and temporal relationships); and utilizing dispersion and deposition modeling and other
NATA outputs to assess existing air toxic networks and to better characterize air toxic
concentrations . At the end of the initial network development period, long term monitoring
objectives can be set and then more permanent monitoring stations can be selected to address these
objectives. To facilitate an effective long-term monitoring program, an air toxics network review
process will be conducted periodically.
3.2.1 Make use of existing monitoring sites
To initiate the national air toxics monitoring program, the extent and adequacy of current
and historical HAP measurements among the existing networks must be analyzed before significant
new monitoring is initiated. New air toxics monitoring sites established during FY-00/01 should be
established to address future network design issues and to build upon and fill obvious data gaps
among existing air toxics monitoring sites (State/local/tribal air toxics monitoring sites, PAMS
sites, planned PM chemical speciation sites, the MDN monitoring program, the Great Lakes and
NEP monitoring networks, NDAMN, and the IMPROVE and CASTNET monitoring networks).
21
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
There is a wealth of existing air toxics monitoring data which should be thoroughly analyzed to
address a variety of public health and ecosystem related questions and issues which are discussed
in Section 3.2.2 of this document.
Although State, local and tribal air toxics monitoring sites will be the principal components
of a national air toxics monitoring program, integrations with existing criteria pollutant monitoring
programs will permit the leveraging of data collected for the ozone, PM and regional haze
programs. Regarding the PAMS network, the Type-2 sites are sophisticated monitoring platforms
which already collect information on several urban HAPs. For 8 HAPs, they produce year-round
24-hr measurements and hourly summer-time data . The latter is important to understand temporal
patterns of ambient air toxics. While they are reflective of neighborhood or urban scale monitoring,
Type-2 PAMS are sited within or immediately downwind of the urban core in accordance with the
summer-time wind flows. In many of these large metropolitan areas, these sites are also reflective
of typical year-round urban concentrations and will be appropriate platforms for air toxics
monitoring. Enhancement of these platforms, when appropriate, will be very cost effective. Specific
suggestions for enhancement of the PAMS for monitoring toxic VOCs are discussed in Section 3.2.9
of this document. There are 24 PAMS monitoring areas, most of which have two Type-2 sites.
PM2, speciation trend sites, which are being established in CY-2000, will also reflect
community-oriented monitoring sites; they will provide for fine particle measurement of 10 of the
11 HAP metals (including 7 of the 8 metals in the UATS) and often will contain other air toxics
pollutant-related sampling equipment such as PM,0 and TSP samplers. The latter will provide
opportunity for additional analyses for lead and other HAPs associated with larger particles. The
mature PM2 5 network will have approximately 54 special trend sites. There are also expected to be
up to 250 additional PM2 5 speciation sites which will provide HAP metals and should be
considered, in part, for broader regional air toxics monitoring. Another excellent choice for air
toxics monitoring platforms will be sites in the PM2 5 "supersite" network. This network will
evolve over the next 1-2 years and will provide monitoring stations which will include a variety of
routine and research grade gaseous and PM analyzers, many of which will be directly related to
measurement of HAPs. Additional continuous and integrated 24-hr measurements of VOCs together
with aerosols, semi-volatiles and meteorological data throughout these intensive monitoring areas
will be very useful for more rigorous and scientific model evaluation.
The IMPROVE and CASTNET networks consist of rural or remote monitoring platforms
which provide information on the chemical constituents of particulate matter. As such, they provide
information on trace metals and can serve as platforms for future monitoring of other HAPs. The
national mercury monitoring network (MDN) and the toxics monitoring networks of the Great Lakes
and NEP programs provide important information on the ecosystem risk assessment of HAPs.
22
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
The development of the national monitoring network should allow for flexibility in the
selection of monitoring locations and cities and rural environments that satisfy the stated monitoring
objectives. If existing platforms are not suitable for initial monitoring objectives (model evaluation
and characterization of air toxics), new community-oriented monitoring stations should be
considered. Although PAMS and PM25 sites would permit cost-effective use of existing equipment,
they may not be properly sited to meet the needs of the air toxics program. For example, some
Type-2 PAMS sites may not be the most representative locations for characterizing annual average
air toxics concentrations, because of the relative mix and spatial distribution of point, area and
mobile emission sources in these areas. In addition, the Type-2 sites are located for summertime
meteorological conditions and this may not ideally represent annual average concentrations. On the
other hand, the data from existing monitors must be evaluated to determine the value added or
potential redundancy among currently operated sites. For example, monitors in areas influenced by
similar emissions or climatology may produce similar concentrations and thereby provide similar
information.
As discussed in Section 3.1.6 of this document, data analysis of monitoring data, emission
inventories and model predictions will identify the most useful and the least useful data, and
ultimately guide the design of the national monitoring program. When appropriate, current
monitoring activities across cities and geographic regions should be streamlined to take advantage
of information derived from dispersion model predictions and to eliminate redundant information
derived from existing monitoring systems. Equipment can be redeployed and resources utilized in
other areas or to sample different HAPs in order to make better use of limited monitoring resources.
3.2.2 Perform data analysis/assessment.
Data analysis is an important component of every air monitoring program. Air toxics data
do not have value unless they are quality assured and then analyzed in the context of the overall
NATA. Adequate resources must be allocated for data analysis, implementation of statistical
quality assurance procedures, data management, data assessment and data reporting. This approach
will make best use of previously collected data which can conserve new monitoring resources. It
will also enhance data quality and ensure an effective air toxics monitoring system. Data analysis
and assessment protocols should be developed to define the process by which conclusions may be
developed from the ambient air toxics data. These may include model evaluation, trends analysis,
and source attribution, among other typical purposes of data analysis. The development of data
analysis protocols should be tied to the data quality objectives mentioned earlier.
23
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
3.2.3 Initial Focus on Model Evaluation
In the initial phases of this effort, one of the primary purposes of data analysis is to evaluate
the NTI and ASPEN modeling of ambient concentrations of air toxics. Comparisons between the
observed concentrations collected through the monitoring network and the predicted concentrations
from the model will be developed to better understand both the NTI and ASPEN outputs. As initial
comparison studies will focus on the national-scale ASPEN modeling efforts, long term model-
monitoring comparison efforts may focus on smaller scale studies (e.g., urban, local and hot-spot
studies) or special monitoring programs (e.g., multimedia concerns).
Applicable monitoring data will be used as a "reality check" on model output. These data
should represent sufficient geographic and emission source diversity to determine if the entire
modeling system (model, emissions, meteorology) provides appropriate estimates of ambient
concentrations to assist in assessment of the goals and objectives of the air toxic program. A broad
selection of locations are needed for the model evaluation. These stations must provide good
geographic coverage, represent different climatological regimes and reflect background
concentrations in rural areas as discussed in Section 3.1.6 of this document.
In populated areas, well sited community-oriented locations should be utilized. These
locations should follow established siting protocols and may be selected from the current State and
local monitoring program locations or should be new sites to fill gaps in the model evaluation data
base. This neighborhood-oriented monitoring approach will be analogous to the core network for
PM2.5. Such monitoring sites should not be located in areas with large concentration gradients, and,
as such, should not be very close to large sources. As discussed in Section 3.1.4 of this document,
they should typically have a spatial scale of representativeness of 0.5 to 4 km which is ideally
suited for comparison to county-wide average predictions from a national dispersion model like
ASPEN. Ideally, the network should place a small number of sites in each monitoring area to
assess spatial variability in HAP concentrations. This may be accomplished with fixed sites,
movable platforms or portable monitors. However, the availability of limited monitoring resources
and the need for good geographic coverage will only allow multiple monitors in some areas.
The monitoring should also be standardized in other ways: The sites must monitor
throughout the year and on the same days/ sampling schedule (e.g. 24-hr averages every 6th day or
other appropriate intervals); use consistent sampling, analytical methods and laboratory
procedures; and follow established quality assurance protocols.
It is this initial ambient data set along with the ASPEN national modeling effort that will
help define the spatial locations for future monitor efforts. The development of the network will
therefore be an iterative process, with a continuing assessment of the appropriateness of specific
monitoring locations.
24
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
3.2.4 Longer-term trends network.
The longer-term monitoring network will include permanently sited monitoring stations
which will be used to establish trends and evaluate air toxics program goals. Although the next few
years of data collection and analysis will focus on air toxics characterization and model evaluation,
these formative years will also be important to better define network requirements for the long-term
trends network for specific air toxics. Any new characterization sites should be selected to also
potentially serve as long-term trend sites. Network managers may consider traditional monitoring
and network design principles in choosing new gap filling locations so that they reflect stable
monitoring environments, with potential for long-term monitoring. Similar to the requirements for
model evaluation, the sites selected for trends analysis should:
be neighborhood-oriented and reflective of general population exposure;
comply with established physical siting protocols;
provide good geographic coverage and represent different climatological regimes;
include appropriate numbers of sites with influences by specific emission sources
(mobile and stationary);
represent regional background and transport concentrations (rural areas);
include common sets of HAPs at sufficient numbers of sites;
monitor throughout the year and on the same days/ sampling schedule ;
(e.g. 24-hr averages every 6th day);
ensure sufficient data capture; and
use consistent sampling, analytical methods, laboratory procedures and quality
assurance protocols
3.2.5 Incorporate long-term and short-term monitoring elements where possible
The network may incorporate fixed-station long-term monitoring as well as short-term
monitoring elements. Where possible, this will enable the network to address the multiple
objectives of the air toxics program. The network can be modeled after the existing State and Local
Air Monitoring Station (SLAMS). Such a network includes long-term National Air Monitoring
Stations (NAMS) designed to study trends and pollutant impacts in major metropolitan areas. It also
includes other SLAMS monitoring stations to address State level characterizations and assessments
on a 3-5 year time frame.
26
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
3.2.6 Allow for temporary air toxics monitoring activities.
The national air toxics monitoring program should allow for short-term, local studies. Such
monitoring is desired by many State and local areas. It may utilize temporary or mobile monitoring
stations and be an adjunct to the network of fixed site monitoring locations. These activities can be
useful to facilitate proper assessments of geographic variability, both between and within
metropolitan areas and permit development of hourly ambient concentrations for certain HAPs that
may present health risks from acute exposure. Such short-term data collection activities can also
assist with developing source receptor relationships, identifying contributing sources and
understanding atmospheric processes. The collection of on-site meteorological data would be
useful to assist with these assessments. The aforementioned activities will be particularly useful
during the formative years of national network implementation.
Examples of temporary air quality monitoring activities may also include: characterization
of environmental justice concerns, assessment of ambient concentrations representative of small
geographic areas like schools which may be potentially impacted by specific sources ("hot spots")
or specialized augmentation of routine monitoring to meet State or local needs.
3.2.7 Integrate air toxics and other monitoring.
Collectively, a national toxics network consists of many existing programs including the
sites operated by State and local agencies, the HAPs measured in the PAMS program, metals
measured in particulate matter programs (particularly the PM2 5 speciation and IMPROVE
programs), the NADP/MDN program, the Great Lakes and NEP monitoring programs, MDAMN,
and new initiatives currently under discussion. The new air toxics monitoring program is an
integration across many separate monitoring activities which are envisioned to form a
comprehensive air monitoring network. The long-term goal for a national monitoring program
should also include monitoring of sensitive ecosystems such as the deposition monitoring efforts for
the Great Lakes and coastal waters and support future environmental assessment programs such as
the EPA-ORD Mercury Research Strategy, North American Free Trade Agreement (NAFTA)
related environmental initiatives, and other trans-boundary pollutant assessment programs.
3.2.8 Utilize standard monitoring methods.
Standardized monitoring methods should be used. Currently, there are standard methods that
cover 27 of the 33 UATS HAPs. Methods for the five HAPs not covered are either not generally
considered to be practical or have not been demonstrated and require additional methods
development. The methods are as follows:
Toxic Compendium Method TO-15, "Determination of Volatile Organic
27
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
Compounds in Air Collected in Specially-Prepared Canisters and Analyzed by Gas
Chromatography/Mass Spectrometry (GC/MS)" or TO-14A, "Determination of
Volatile Organic Compounds (VOCs) in Ambient Air Using Specially Prepared
Canisters with Subsequent Analysis by Gas Chromatography" are standard EPA
methods to be used. These methods provides for collection of volatile organic
compound (VOC) HAPs, including most of the 14 VOCs on the 33 UATS HAPs
list3. See attachment 4.
Compendium Method TO-15 is an extension of the Method TO-14 description in the
following ways:
TO-15 incorporates a multisorbent/dry purge technique for water management,
thereby addressing a more extensive set of compounds (oxygenated).
TO-15 using the GC/MS technique versus multiple-detectors as the only means to
identify and quantitate target compounds. The GC/MS approach allows for specific
detection of the target analytes and reduces the chance for misidentification or the
effects of coeluting compounds. The use of non-specific detectors (e.g., FID, ECD)
or a combination of these detectors, used in TO-14 is determined by the required
specificity and sensitivity of the application. While non-specific detectors are in
some cases more sensitive than specific detectors, they vary in specificity and
sensitivity for a specific class of compounds. Trade-offs in sensitivity between non-
selective multiple detectors and GC/MS can be somewhat mitigated by the use of
ion trap technology versus full scan quadrupole MS.
Chemical speciation of filter-based mass collected at particulate matter sites can
provide data on several elements using Compendium Method IO-3, "Chemical
Species Analysis of Filter-collected Suspended Particulate Matter. " This analysis
includes as many as 11 HAP metals and as many as 8 UATS metals. See Attachment
5. Particle sampling can include PM2 5 to focus on the fine fraction of suspended
particles, or total suspended particulate (TSP) for using a high-volume (hi-vol)
sampling system to permit analysis of metals among all suspended particulate
matter. This is consistent with the intended analytical approaches and services
available through EPA contracts. Some valence-specific metals, like chromium VI (
hexavalent chromium) have been identified as having high toxicity and would
require separate chemical analyses. These chemicals would be collected with the
same particulate matter samplers but would be analyzed with more specific
analytical techniques.
The analysis of mercury and speciated forms of mercury (if desired) will require
3Monitoring agencies have reported difficulties with recoverability of acrylonitrile with this method
28
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
the use of technology outside of Compendium Method IO-3. Consideration is should
be given to the use of continuous gas-phase total and speciated mercury monitors for
future, long-term monitoring efforts. Continuous monitoring instrumentation is
currently commercially available and somewhat expensive for both total and
speciated mercury monitoring. More methods research is needed to provide a
standardized, routine monitoring method for mercury.
• Toxic Compendium Method TO-11 A, "Determination of Formaldehyde in Ambient
Air Using Adsorbent Cartridge Followed by High Performance Liquid
Chromatography" provides for the collection of three aldehyde FLAPs including two
UATS aldehydes belong to the family of chemicals known as carbonyls. See
Attachment 6.
Toxic Compendium Method T0-13A, "Determination of benzo(a)pyrene and other
Polynuclear Aromatic Hydrocarbons in Ambient Air Using Gas Chromatography
with Mass Spectrometry" provides for the collection and analysis of PAHs. This
method covers one urban HAP, polycyclic organic matter (POM) (a group of
compounds). With this method, this HAP is being represented by seven polynuclear
aromatic hydrocarbons (PAHs), including (benz(a)anthracene, chrysene,
benzo(b)fluorathene, benzo(k)fluoranthene, benzo(a)pyrene, indeno(l,2,3-cd)pyrene
and dibenz(a,h)anthracene). See Attachment 7.
Toxic Compendium Method T0-4A, "Determination of Pesticides and
Polychlorinated Biphenyls in Ambient Air Using High Volume Polyurethane Foam
(PUF) Sampling Followed by Gas Chromatographic/Multi-Detector (MD)",
provides for the analysis of hexachlorobenzene. See Attachment 7.
• Toxic Compendium Method T0-9A, "Determination of Polychlorinated,
Polybrominated and Brominated/Chlorinated Dibenzo-p-Dioxins and Dibenzofurans
in Ambient Air" provides for the analysis of 2,3,7,8-tetrachlorodibenzo-p-dioxin &
congeners & TCDF congeners. Note that this method is reported to be resource-
intensive/expensive.
The urban HAPs currently without practical or demonstrated methods are acrolein,
acrylonitrile, ethylene oxide, hydrazine, coke oven emissions, and quinoline.
3.2.9 Enhancement of the PAMS for monitoring toxic VOCs
An opportunity to leverage the installed instrument base of the PAMS for monitoring
of toxic VOCs should be considered. The PAMS networks utilize gas chromatography
either for on-site ambient air monitoring at the PAMS monitoring site (auto-GCs) or for
29
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
monitoring of sample air taken from canisters filled at the site and analyzed in a central
laboratory (laboratory GC). These systems consist of a gas chromatograph and a detector
(typically a flame ionization detector or FID at the monitoring stations) and provide
concentration data for the PAMS target list of hydrocarbons. The PAMS list includes the
compounds benzene, toluene, and other aromatic hydrocarbons (for a total of 9 HAPs) that
are on the list of 188 toxics identified in Section 112 of the CAA. Hence there is a
historical database available for these compounds. However, these systems may have the
capability for providing essentially all of the air toxic VOCs. To do this the auto-GCs
would need to be augmented by: (1) adding a supplemental list of target compounds, e.g.
subsets of the TO-15 target list, to the target compound list detected by the FID and (2)
adding an electron capture detector (ECD) for detection of many of the chlorinated VOCs.
The first change can be accomplished by implementing data processing and calibration
procedures which are similar to those used for the PAMS target list. The second change is
possible because the auto-GCs are usually designed to allow for at least one additional
detector system. Another dimension of the auto-GCs is their configuration to allow analysis
of samples taken in canisters and also samples taken in solid sorbent tubes. Thus, samples
taken at various locations could be taken to a centrally-located GC or auto-GC for analysis
so that VOC spatial variability can be determined.
Note that since the PAMS sites have been chosen based primarily on the need to
gather ozone and ozone precursor data, they may not be ideal for the monitoring of Urban
HAPs. In addition, States have noted that changes to the PAMS system such as described
above, may add more complexity and variability to the already complicated PAMS
measurement process. Accordingly, in coordination with the States, the EPA is investigating
the practicality of adding the additional detector to the GCs in the PAMS program before
including recommendations for these changes as part of the air toxics monitoring program.
3.2.10 Incorporate measurements for other HAPs when possible (Support research and
development efforts to permit new and better analyses).
The network should allow for measurements of additional HAPs. For example, there are a
number of urban HAPs without demonstrated routine sampling and analytical methods associated
with them. Examples are acrolein, ethylene oxide, hydrazine, quinoline, and coke oven emissions.
In addition there are other methods which are costly to implement. For example, dioxin (i.e.
2,3,7,8-tetrachlorodibenzo-p-dioxin & congeners & TCDF congeners) and speciated mercury.
Research is needed to develop new or more cost-effective monitoring methods to permit the
measurement for these HAPs on a routine basis. It is also important to modify current measurement
capabilities to improve and streamline the operational aspects of monitoring and make it less labor
intensive. These efforts become important adjunct activities to the overall ambient air toxics
monitoring strategy.
30
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
3.2.11 Review network periodically.
Air toxics data collection activities and a future national network should be reasonably
dynamic and modified as needed. The program should include the following:
Annual or biannual reviews of the network should be designed to eliminate
redundancy of measurement within and across cities, to modify sampling frequencies
and to adjust measurement protocols to ensure that data quality objectives are
achieved.
The number of toxics to be monitored also needs to be evaluated periodically and
may need to be increased or decreased as appropriate
The target list of pollutants should be modified to make cost-effective use of
available resources while still satisfying the multiple goals of the air toxics
program. If a particular analyte requires its own discrete monitoring method and it is
not detectable, then it should be considered for elimination from routine sampling
and laboratory analysis. This does not necessarily apply to compounds that are part
of a suite of compounds that are generated with a particular monitoring method (like
TO-15 or XRF). PCBs or Dioxin might be method specific examples. While air
toxics monitoring efforts may initially include a large set of HAPs, the target list
may be reduced as information about ambient concentrations or predicted
concentrations become available. However, the analysis of once non-detectable
compounds in key monitoring areas should be periodically revisited to ensure that
new emission sources (or better monitoring technologies) have not emerged. It is
suggested that this occur once every 3-5 years.
More sensitive monitoring methods should also be developed and monitoring re-
initiated when the risk based benchmarks and observed concentrations are both less
than current method detection limits.
The network planning process should also make use of surrogate measures
whenever appropriate to help identify areas where more specific monitoring is
needed (e.g. nickel or zinc as a predictor for mercury; and carbon monoxide as a
predictor of mobile source pollutants such as benzene and 1-3 butadiene).
31
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
4 Initial Network Development
The national air toxics monitoring network will be established over a several year period
through a cooperative Federal, State, local and tribal effort. To obtain guidance from EPA's State
and local partners and to take advantage of their air toxics monitoring expertise, EPA has convened
a steering committee which consists of EPA, State and local monitoring agency representatives.
Over the past 6 months, this committee has contributed to the development of this document and has
helped define near-term activities to assist the design and implementation of a national air toxics
network.
The following sections address the activities and protocols anticipated during the initial
rollout and subsequent years. A more detailed, but still tentative approach, for FY-00/01 air toxic
monitoring activities is also presented. Many details must still be developed and will be described
in addendums to this document.
4.1 Preliminary schedule of monitoring and data analysis activities
4.1.1 Activities during 3-year rollout:
(FY-00 ) Select core HAPs
(FY-00) Develop QAPP
(FY-00) Develop DQO's
(FY-00 ) Develop initial monitoring plan with stated objectives
(FY-00) Select FY-00 Pilot Project Areas
(FY-00) Distribute FY-00 grants and guidance
(FY-00) Identify existing sites for continued monitoring
(FY-00 ) Procure equipment
(FY-00/01) Establish new sites and begin monitoring
(FY-00/01) Determine laboratory support process
(FY-00) Utilize Model Predictions to guide selection of FY-01 monitoring sites
(FY-00/02) Analyze data from current State/local toxic monitoring sites
(FY-02) Analyze Data from New Pilot Proj ect Monitors
(FY-02) Utilize Model Predictions
(FY-02) Develop long-term network monitoring plan based on model results/data
analyses/network objectives and insights gained from FY-00 Pilot Projects
(FY-01/02) Determine resource availability and funding status
(Ongoing) Conduct local monitoring projects as needed
4.1.2 Post Rollout Activities
(FY-03) Begin implementation of national network
(FY-03/05) Complete full network deployment
32
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
(Ongoing)
(Every 2 yrs)
(Ongoing)
Conduct data analyses
Conduct network reviews/recommend network modifications as necessary
Consider current state-of-knowledge; current state of monitoring
technologies; need to add/delete HAPs; need to add/delete sites; allocation
of resources; etc.
Conduct local monitoring projects as needed
4.2 FY-00/01 activities
The initial focus should be on meeting the objectives of the initial 3-year rollout of the new
monitoring network.
The principle focus during these formative years will be to:
characterize pollution gradients reflecting diverse areas and a variety of emission sources,
provide information on concentration levels and pollutant variability to compare with
ASPEN model outputs,
obtain data to determine if a small number of monitors operating at minimum sampling
frequencies can adequately characterize the toxics in urban areas,
depict the range of concentrations which may be expected in differing urban/rural environs
and source influences (mobile sources, industrial activity, background, etc.), and
provide a data base sufficient to optimize the eventual implementation of a national air
toxics monitoring effort.
To accomplish these goals, the steering committee has recommended that pilot sampling
programs should be conducted in several urban areas and non-urban rural areas. Two to four urban
areas influenced by diverse emission sources and different climatological conditions would be
selected to produce air toxics measurements for one to two years at multiple neighborhood scale
monitoring sites using the same monitoring protocols. Fewer monitoring sites would be required in
the non-urban areas. These data will be subsequently assessed to better understand the sources and
magnitudes of variability associated with ambient air toxics concentrations within and between
monitoring locations across the U.S.
Where there is consistency between these objectives and existing monitoring sites, efforts
should be directed toward the enhancement of existing State and local air toxic sites (including
PAMS sites where appropriate), enhancement of new PM2 5 chemical special sites (a total of 54
urban PM2 5 monitoring areas have been proposed) and possible enhancement of existing rural
monitoring networks (IMPROVE/CASTNET/MDN/NEP/Great Lakes).
For consistency with long-term air toxics program objectives, it is essential that a number of
long-standing State and local air toxic sites are also maintained. This requires resources to
33
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
refurbish these sites as needed to ensure high data quality and continuation of good data capture.
Resources permitting, new community-oriented ambient air toxics monitoring platforms could also
be established to fill gaps in existing State, local and tribal networks (e.g. currently unmonitored
urban areas with high toxicity weighted HAP emissions). Similarly, new rural monitors where
current background data is lacking or is uncertain could be established. In all cases, the platforms
should be equipped or upgraded to enable an appropriate suite of required urban HAPs.
To maximize the use of resources in gaining new air toxics data, movable or mobile
platforms and portable samplers, may be utilized in addition to fixed sites to help define temporal
and spatial patterns within urban areas. This will help to assess existing monitoring locations, to
identify representative locations for model evaluation or for long-term monitoring. Such activities
will help to characterize air toxics, to evaluate model predictions, to facilitate the derivation of
network design requirements, to reduce the need to relocate established monitors and to address
community concerns. Movable platforms will be particularly useful during the formative years
when the aforementioned pilot studies are conducted. The monitoring activities during the first
years should also reflect appropriate quality assurance, data management, data analysis and data
submission to AIRS.
4.3 Initial Monitoring Protocols.
The first few years should utilize established monitoring protocols. This will include
analysis for VOCs (using TO-14A or TO-15) from year-round 24-hr canister samples collected
intermittently. The TO-14A/15 compounds include benzene, 1,3-butadiene, 1,4- dichlorobenzene,
carbon tetrachloride, trichloroethylene, vinyl chloride, etc. A complete list of these analytes is
attached (Attachment 4). The TO-14/15 compounds were selected as the first priority because the
analysis method has been proven, is cost-effective, and provides data on most of the urban air toxic
VOCs of interest.
Additional 24-hr samples and analyses could include IO-3 for HAP metals (using XRF
analysis on TSP filter media), POMs using the T0-13A method and aldehydes (including
formaldehyde and acetaldehyde) using TO-11 A. Many of the PM2 5 platforms will also include
measurements for PM10 and/or TSP, so comparison of fine to coarse particle urban HAP metals is
also possible. PM25 sampling is planned for a l-in-3-day schedule; for other pollutants (including
TSP), sampling will be likely scheduled once in 6 tol2 days depending on the results of the data
quality objectives process.
As technology develops, continuous and less labor intensive monitoring equipment may
become available. A new continuous formaldehyde analyzer is currently undergoing field
evaluation and may be ready for limited deployment during FY-2000. Continuous total and
speciated mercury monitors are currently commercially available.
34
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
Some other measurements are relatively more expensive, but are very important and worth
including in the initial network. In particular, specific analysis for mercury using new continuous
speciation analyzers could possibly be included at selected sites for comparison to XRF
measurements for particulate mercury and mercury surrogates (particulate zinc and nickel) and wet
deposition measurements. However, because of limited resources, these mercury analyzers may be
utilized on a limited basis. Dioxin measurements are even more expensive, so the NDAMN network
can be used to provide useful information on dioxin. However, monitoring may be considered by
the States on a case-by-case basis for selected sites.
4.4 Detailed FY-00/01 Monitoring Plan
To support the initial data collection goals (model evaluation, air toxics characterization
and network design), a detailed monitoring plan is needed. This plan will identify urban and rural
areas selected for the pilot monitoring programs and other geographic regions or metropolitan areas
for continued, new or expanded monitoring activities. It will also specify the initial core HAPs4 to
be monitored at appropriate number of sites in specific source/population/climatological regimes;
monitoring methods; recommendations for the use of existing PAMS continuous monitors and
deployment of other continuous analyzers; recommendations for sampling frequencies for the
integrated 24-hour sampling methods; and quality assurance protocols. The need for additional
monitoring during the initial phases of data collection and for long-term monitoring will be
addressed. As new priority air toxics are identified, future monitoring activities should address
other pollutants of concern in the air toxics program. The longer term plan will therefore address
the need to produce measurements of additional compounds from the larger set of 188 CAA HAPs .
The detailed monitoring plan will be developed over the next several months. At this point,
we cannot precisely project the size of the national monitoring network. It is noted that monitoring
sites which participate in the FY-00/01 pilot networks or other short-term gap-filling sites may not
be part of a final long term national network. The number of sites for the final network will be
developed from the initial data assessments, revised iteratively as additional monitoring and
modeling information becomes available and also modified as the objectives of the national air
toxics program evolve.
4Recommendations for these core/target HAPs have been developed by a committee of State/local agency
and EPA. technical staff. These pollutants are identified as a part of Attachments 4, 5, 6 and 7 to this document.
The pollutants were chosen based on the availability of standard gas mixtures, available and practical
sampling/analysis methods, and the sensitivities of the available methods.
35
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
ATTACHMENTS
Attachment 1: Existing Air Toxics Monitoring Stations by State (1999)
Attachment 2: New Air Toxics Monitoring Stations Planned for Year 2000
Attachment 3: Types of Air Toxics Monitored
36
-------
0)
3s
to
c
o
IS
O)
©>
CN
U) D)
C .S
S CM
¦r^ « in
¦¦ 1- N
OCD
a ¦
° 3®
Co <
-Q I"
2 T3
a)
o £
x -S
a.
L.
O
<
G)
C
¦
to
¦
X
LU
d
*
-------
CM
ro
0
>-
"O
0
C
C
JS
Q.
(0
c
o
s
55
O)
c
¦
o
¦ HI
c
o
s
o
¦
X
o
<
5
0
o
in
<
I-
o
O)
c
CO
"x
0
o
4-»
c
(0
o
c
'<4—'
_o
T3
"O
03
CO
(0
c
D)
" —
mc
0
' L.
L-
co
o
CO
"c
c
o
O
E
CO
CO
0
CO
0
_C
1-
d
r \
Q
%
-------
Types of Air Toxics Monitored
(Existing or Planned)
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
ATTACHMENT 4
METHOD TO-14A/15 (VOCs)
No.
POLLUTANT
CAS#
UATS a
UATMP
b
Core d
Maxe
1
1,2- Dibromoethane
106934
X
X
T f
2
1,2- Dichloroethane (EDC)
107062
X
X
T
3
1,3-Butadiene
106990
X
X
X
X
4
1,1 -Dichloroethane
75343
X
5
1,1,2,2-T etrachloroethane
79345
X
X
T
6
1,2,4-Trichlorobenzene
120821
X
7
1,1,2-Trichloroethane
79005
X
8
1,2-Dichloropropane (propylene
dichloride)
78875
X
X
X
X
9
1,3-Dichloropropene
542756
X
X
X
10
1,1,1 -Trichloroethane
71556
X
11
1,1 -Dichloroethylene
75354
X
12
1,4-Dichlorobenzene
106467
X
13
2,2,4-Trimethylpentane
540841
14
2-Chloro-l ,3-butadiene
(chloroprene)
126998
X
15
Acetonitrile
75058
X
16
Acrylonitrile (Monitoring agencies
have reported difficulties with
recovery using this method)
107131
X
X
17
Allyl chloride
1070501
18
Benzene
71432
X
X
X
X
19
Benzyl chloride c
100447
20
Bromoform (tribromomethane)
75252
X
21
Bromomethane (methyl bromide)
74839
X
22
Carbon Tetrachloride
56235
X
X
X
X
23
Chlorobenzene
108907
X
24
Chloroethane (ethyl chloride)
75003
X
25
Chloroform
67663
X
X
X
X
26
Chloromethane (methyl chloride)
74873
X
37
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
27
Cumene
98828
28
Ethyl acrylate
140885
X
29
Ethylbenzene
100414
X
30
Hexachlorobutadiene
87683
X
31
Hexachlorocyclopentadiene
77474
32
Hexachloroethane
67721
33
m-Xylene
108383
X
34
Methyl ethyl ketone
78933
X
35
Methyl isobutyl ketone
108101
X
36
Methyl tert-butyl ether (MTBE)
1634044
X
37
Methyl methacrylate
80626
X
38
Methylene Chloride
75092
X
X
X
X
39
n-Hexane
110543
40
o-Xylene
95476
X
41
p-Xylene
106423
X
42
Styrene
100425
X
43
Tetrachloroethylene (PCE)
127184
X
X
X
X
44
Toluene
108883
X
45
Trichloroethylene (TCE)
79016
X
X
X
X
46
Vinyl chloride
75014
X
X
X
X
47
Vinyl acetate
108054
48
Vinyl bromide
593602
49
Xylene (mixed)
1330207
a UATS = Urban Air Toxics Strategy compound
b UATMP = Urban Air Toxics Monitoring Program target list for 1999 and later years,
c May have stability issues over time in cylinders or canisters
d A small subset of HAPs which should be measured at virtually every monitoring site designated as a part of the national HAPs
monitoring network.
e A set of pollutants which includes as many of the Urban Air Toxics Strategy pollutants as possible, measured at selected sites,
f Pollutant temporarily listed.
[Note: The above list of VOCs are those that can be measured by TO-14a or TO-15. However, some of these are not
yet appropriate for routine monitoring because calibration standards are not readily available.]
38
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
ATTACHMENT 5
METHOD IO-3 (ELEMENTS AND METALS BY XRF AND ICP/MS)
POLLUTANT
CAS fta
UATS b
UATMP d
Core e
Maxf
Antimony
7440360
X
Arsenic compounds
7440382
X
X
Ts
T
Beryllium and compounds
7440417
Xc
X
X
X
Cadmium compounds
7440439
X
X
X
X
Chromium compounds
7440473
X
X
X (total Cr)
X (Cr VI)
Cobalt
7440484
Lead compounds
7439921
X
X
X
X
Manganese compounds
7439965
X
X
X
X
Mercury compounds
7439976
X
X
T
Nickel compounds
7440020
X
X
X
X
Selenium
7782492
a Other CAS numbers are defined for the compounds of these metals,
b UATS = Urban Air Toxics Strategy compound
c Beryllium cannot be detected by XRF
d UATMP = Urban Air Toxics Monitoring Program target list for 1999 and later years.
e A small subset of HAPs which should be measured at virtually every monitoring site designated as a part of the national HAPs
monitoring network.
f A set of pollutants which includes as many of the Urban Air Toxics Strategy pollutants as possible, measured at selected sites,
g Pollutant temporarily listed.
39
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
ATTACHMENT 6
METHOD TO-11A (CARBONYL COMPOUNDS) a
POLLUTANT
CAS#
UATS b
UATMP c
Core d
Maxe
Acetaldehyde
75070
X
X
X
X
Formaldehyde
50000
X
X
X
X
Propionaldehyde
123386
X
a These three carbonyl compounds are also known as aldehydes,
b UATS = Urban Air Toxics Strategy compound
c UATMP = Urban Air Toxic Monitoring Program target list for several years
d A small subset of HAPs which should be measured at virtually every monitoring site designated as a part of the national HAPs
monitoring network.
e A set of pollutants which includes as many of the Urban Air Toxics Strategy pollutants as possible, measured at selected sites.
40
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
ATTACHMENT 7
SEMI-VOLATILE ORGANIC COMPOUNDS AND OTHER UATS HAPS
POLLUTANT
CAS#
UATS a
UATMP b
Core f
Max8
Polycyclic organic matter (POM)c
NA
X
X
X
Polychloiinated biphenyls (PCBs) d
NA
X
X
bis (2-chloromethyl) ether
542-88-1
X
bis (2-ethylhexyl) phthalate
117-81-7
X
Hexachlorobenzene d
118741
X
X
Dioxin
[2,3,7,8-tetrachlorodibenzo-p-dioxin (&
congeners & TCDF congeners)]
1746016
X
X
Hydrazine
302012
X
Coke oven emissions e
NA
X
Quinoline
91225
X
a UATS = Urban Air Toxics Strategy compound
b UATMP = Urban Air Toxics Monitoring Program target list for 1999 and later years,
c Method TO-13A; POM represented by 7 PAHs which include:
Benz [a] anthracene
Chrysene
Benzo [b] fluorathene
Benzo [k] fluoranthene
Benzo[a]pyrene
Indeno[l ,2,3-cd]pyrene
Dibenz [a,h] anthracene
d Method TO-4A
e Coke oven emissions are a mixture of benzene, toluene, xylenes, cyanide, naphthalene, phenol and POM.
f A small subset of HAPs which should be measured at virtually every monitoring site designated as a part of the national HAPs
monitoring network.
g A set of pollutants which includes as many of the Urban Air Toxics Strategy pollutants as possible, measured at selected sites.
41
-------
Peer Review Draft for the Science Advisory Committee, Air Toxics Monitoring Strategy Subcommittee FY-00
References
Toxic Compendium Method TO-4A, "Determination of Pesticides and Polychlorinated Biphenyls in
Ambient Air Using High Volume Polyurethane Foam (PUF) Sampling Followed by Gas
Chromatographic/Multi-Detector (MD)".
Toxic Compendium Method TO-11 A, "A Determination of Formaldehyde in Ambient Air Using
Adsorbent Cartridge Followed by High Performance Liquid Chromatography".
Toxic Compendium Method TO-13A, "A Determination of benzo(a)pyrene and other Polynuclear
Aromatic Hydrocarbons in Ambient Air Using Gas Chromatography with Mass Spectrometry".
Toxic Compendium Method TO-15, "A Determination of Volatile Organic Compounds in Air
Collected in Specially-Prepared Canisters and Analyzed by Gas Chromatography/Mass
Spectrometry"
All of the above are available at: http://www.epa.gov/ttn/amtic/airtox.html
Inorganic Compendium Method 10-3.3, "Determination of Metals in Ambient Particulate Matter
Using X-Ray Fluorescence (XRF) Spectroscopy".
Inorganic Compendium Method 10-3.5, "Determination of Metals in Ambient Particulate Matter
Using Inductively Coupled Plasma/Mass Spectrometry (ICP/MS)".
Inorganic Compendium Method 10-2.1, "Sampling of Ambient Air for Total Suspended Particulate
Matter (SPM) and PM10 Using High Volume (HV) Sampler".
"Technical Assistance Document (TAD) for Sampling and Analysis of Ozone Precursors;
EPA/600-R-98/161; September 1998." posted at http://www.epa.gov/ttn/amtic/pams.html
"PAMS Implementation Manual";
Particulate Matter (PM2.5) Speciation Guidance, Final Draft, Edition 1. October 7, 1999.
42
------- |