& EPA
United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Tidangle Park, NC 27711
BPA-45Q/4-89-Q16
May 1989
Air
PM-10 MONITORING
TASK FORCE REPORT
-------
EPA-450/4-89-016
MONITORING
TASK FORCE REPORT
BY
w. F. HUNT, CHAIR
N, BELOIN, N. BERG, c. BOHNENKAMP, J. DEWEY, N. FRANK,
R. GREGORY, R. KAPICHAK, J. KELLY, M. KEMP, L. LARSON, D. LQHMAN,
T. PACE, j, SCHWEISS, D, STONEFIELD, D. WELLS AND D. WILSON
OF
THE OFFICE OF AIR QUALITY PLANNING AND STANDARDS
AND
u. S. EPA REGIONS I - x
Office Of Air Quality Planning And Standards
Office Of Air And Radiation
U. S. Environmental Protection Agency
Research Triangle Park, NC 27711
November 1989
-------
This report has been reviewed by the Office Of Air Quality Planning And Standards, U, S. Environmental
Protection Agency, and has been approved for publication. Any mention of trade names or commercial products
is not intended to constitute endorsement or recommendation for use.
EPA^50/4-89-016
-------
PREFACE
This report was prepared by the PM,p Monitoring Task Force
which was formed in July 1988. The Task Force was created by
William G. Laxton, Director of the Technical Support Division in
the Office of Air Quality Planning and Standards *in response to
concerns raised at the Air Division Directors meeting in June 1988.
The Task Force was formed to look into the need for evaluating PM1Q
monitoring networks especially in existing Group III areas. The
principal purpose of the Task Force was to address the apparent
disparity in the number of PM,Q nonattainment areas between the
Western and Eastern States. The Task Force is composed of people
from all 10 EPA Regions and the Office of Air Quality Planning and
Standards.
The following people are recogni2e<3 for their contributions
as facilitators of the sections of the report:
Executive Summary - William F. Hunt, Jr., OAQPS, and Task Force
Members
ISSUE 1 - Neil Berg, OAQPS, and Tom Pace, OAQPS
ISSUE 2 - Rudy Kapichak, Region II, and Dale Wells, Region VIII
ISSUE 3 - Ray Gregory, Region IV, and Jon Schweiss, Region X
ISSUE 4 - Norm Beloin, Region I, and Carol Bohnenkamp, Region IX
ISSUE 5 - Mary Kemp, Region VI, and Denis Lohman, Region III
ISSUE 6 - James Dewey, Region V, and James Kelly, Region VII
Other members of the Task Force that deserve recognition for
their contributions are ; Linda Larson of EPA Region V and Neil
Frank, Dave Stonefield and Dean Wilson of OAQPS. Special mention
should also be given to Whitmel Joyner for technical editing and
to Barbara Stroud, Helen Hinton and Kathy Weatherspoon for typing
the report.
111
-------
EXECUTIVE SUMMARY
Introduction • „
The PM,Q Monitoring Task Force was formed in July 1988 by
William G. Laxton, Director of the Technical Support Division in
the Office of Air Quality Planning and Standards. The Task Force
was charged with responding to concerns raised at the Air Division
Directors meeting in June 1988. Specifically, the Task Force was
asked to look into the need for evaluating PM10 monitoring networks
especially in existing Group III areas. The principal purpose of
the Task Force was to address the apparent disparity in the number
of PM1Q areas between the Western and Eastern States.
The Task Force is composed of members from the ten EPA
Regional Offices and the Office of Air Quality Planning and
Standards. The Task Force began its work by designing a
questionnaire and conducting a survey. Six major issues were
identified and are addressed in this report. They are as follows:
1. Does the current monitoring network accurately reflect
the scope and magnitude of the ambient PM,0 problem in
the United States?
2. How were PM10 problems identified?
3. What are the existing tools used for identification and
how -were they used?
4. What was the effect of PM^0 monitoring resources in
identifying the PM1Q problem?
5. Can. EPA do a better job using existing tools and
resources? If so, how?
6. What new tools, resources, policies, authorities, etc.,
are needed?
A workshop was held in Denver during October 1988 to permit
the Task Force members to discuss these issues and to develop the
information for this report.
-------
ISSUE 1. DOES THE CURRENT MONITORING NETWORK ACCURATELY REFLECT
THE SCOPE AND MAGI
THE UNITED STATES?
THE SCOPE AND MAGNITUDE OF THE AMBIENT PM1Q PROBLEM IN
Some people feel it does, some do not.
Adequate in terms of meeting the existing regulation.
Professional judgement says we need to go beyond bare
minimum to properly characterize the magnitude of the
ambient PM|g problem in the United States.
Special studies in suspected high impact areas needed.
- Region X Saturation Monitoring Technique.
Low cost, portable, battery-operated monitors.
Cost effective way to proceed.
- Asheville Study & possible El Paso/Juarez Study
- Investigate nontraditional sources: Residential
wood combustion (RWC), sanding and salting of
roadways, gasoline and diesel exhaust emissions, re-
entrained road dust and mining.
-------
ISSUE 2. HOW WERE PM1Q PROBLEMS IDENTIFIED?
Use of probability guideline.
Applied to TSP data, plus use of limited PM10 data.
- Estimated areas of high, medium and low probability
of violating PM10 standards.
PM10 monitoring network is based on old TSP network.
PMM NAMS located at old TSP NAMS & SLAMS sites,
PM1Q SLAMS tend to be located at former TSP SLAMS,
PM10 Special Purpose Monitoring (SPM) sites are new
sites.
Use of PM10 SIP development guideline - allowed other
criteria.
- Modeling, investigation of SIP requirements, etc,
- Guideline was not used universally.
Region X performed saturation monitoring with portable
monitors to augment existing fixed site networks.
Special studies conducted in Regions I, II, V, ¥11, VIII,
IX and X,
Only Regions VIII, IX and X (and V, to a lesser
extent) actively pursued monitoring RWC.
Only Region IX emphasised agricultural tilling (AT)
and construction activities.
-------
ISSUE 3. WHAT ARE THE TOOLS USED FOR IDENTIFICATION AND HOW WERE
THEY USED?
Probability guideline in combination with existing TSP
data - primary tool.
Limited PM10 monitoring data.
Special PM1Q monitoring studies initiated prior to
promulgation of the NAAQS (see issue 2 ). ,
Other special RO studies investigated nontraditional
sources.
Traffic data and entrapment (canyon) potential used to
identify microscale problems in half of Regions.
Part 58 monitoring regulation requires deployment
of microscale PM^0 monitors in some instances,
- Emphasis on diesel component of traffic.
Receptor modeling used as a data interpretation tool.
-------
ISSUE 4. WHAT WAS THE EFFECT OF PM,0 MONITORING RESOURCES IN
IDENTIFYING THE PM10 PROBLEM?
Resources limited the extent to which PM1Q monitoring
networks were designed.
Air Program allocated 14% of the Region's FTEs (91 of
649) to monitoring.
Depending upon Region, between 10 and 50% of
monitoring resources were devoted to PMjQ.
In total, estimates show that fewer than 18 RO FTEs
were used in PM10 monitoring.
PM10 Priority
In East, ROs rank PM^0 third or fourth,
In West, ROs rank PM1Q first or second.
Given modest resources and lower priority assigned to
PM^, it is not surprising that TSP networks were used
as oasis for designing PM10 NAMS and SLAMS.
Disinvestment of TSP monitoring has allowed for some
resources, mostly labor, to be redirected to PM1Q
monitoring.
Some TSP monitors are being used as surrogate PM.0
monitors, while others are used to measure toxic
compounds.
An estimated 5% of Regional PM10 monitoring resources
were spent on identifying PM-,0 problems. Lack of
resources has hindered PM,Q problem investigations.
-------
ISSUE 5. CAN EPA DO A BETTER JOB WITH EXISTING TOOLS AND
RESOURCES? IF SO, HOW?
Yes
Make air quality monitoring in general, and PM10
monitoring in particular, more visible within Air
Program.
Encourage use of saturation monitoring to determine
adequacy of fixed site monitoring network and to
identify problem areas.
Develop a guideline on use of saturation monitoring.
This was initiated at the Air Monitoring Workshop
in Southern Pines, North Carolina in July 1989.
Asheville, N.C. study and a possible El Paso/Juarez
study
More Headquarters/RO interaction is needed.
Semi-annual meetings with Regional air monitoring
personnel.
Rotational assignments. This has already been
initiated with Norm Beloin, Region I and Mary Kemp,
Region VI participating in a rotational assignment
with OAQPS.
Headquarters visits to the Regional offices. This
has been initiated with Geri Dorosz participating
in an assignment with Region VI.
Provide better training of PM,Q instrument operators.
Elevate importance of Regional NAMS coordinator.
Need to examine current monitoring resources to insure
adequate use.
Involve a RO meteorologist in ambient monitoring network
review and design.
-------
ISSUE 6. WHAT NEW TOOLS, RESOURCES, POLICIES, AUTHORITIES ARE
NEEDED?
New Tools;
- Portable low cost monitors - saturation monitoring.
Technical guidance on saturation monitoring, etc.
Development of a continuous PM10 monitor. ,
- Data quality objectives and guidance on the required
data to use, operate and evaluate screening models
to identify areas of potential high PM^
concentrations and assure consistent model
application.
Cost effective and more accurate stagnation modeling
to deal with mountain valley situations.
WYND Valley is available but there is a lack
of data to test, evaluate and gain experience
with the model's use.
Improve tech transfer through workshops and AMTEC.
New Resources:
Undertake study to determine proper PM|0 resource
allocation and additional resource needs.
Provide additional resources for saturation studies,
etc. to evaluate nontraditional problem areas.
.New Policies:
Develop guidance for saturation monitoring studies*
Develop guidance on use of inferential monitoring.
Reemphasize importance of PM,Q and need for increased
manpower and funding.
Review exceptional events guideline - already
scheduled.
-------
Recommendations
(1) Regions should develop a proactive attitude toward the
design and revision of PM10 networks - Promote this to
State and local agencies.
(2) Umnonitored areas should be prioritized according to high
PM,Q potential - Annual network reviews should consider
allocation of resources to new areas.
(3) OAQPS should support concept of saturation sampling as
a tool in establishing adequate PM^ networks, and with
the Atmospheric Research and Exposure Assessment
Laboratory (ARIALJ/ROS prepare a guideline on its use.
This recommendation is being implemented.
(4) OAQPS and ROs should strongly support development of low-
cost portable PM10 reference methods/survey devices, and
real-time continuous PM10 reference methods.
(5) OAQPS should actively promote the evaluation of an EPA
approved stagnation dispersion model for use in low-wind
speed situations to support SIP development (may also be
of limited use in network design).
(6) OAQPS should promote refinement of emission factors for
area sources such as RWC, agricultural tilling (AT) and
reentrainment/mobile sources for dispersion model use.
(7) OAQPS and ROs should make every effort to identify and
secure resources for saturation sampling studies, etc.
(8) OAQPS should develop a special projects team with capital
and travel resources to promote/perform network design
studies throughout the country.
(9) Within current resource structure, OAQPS and ROs should
develop policies/techniques to promote concept of
inferential monitoring - A set of techniques: saturation
sampling, modeling (dispersion and receptor),
flexible/rotating subnetworks, etc. to extend network
coverage.
-------
ISSUE 1, DOES THE CURRENT MONITORING NETWORK ACCURATELY REFLECT
THE SCOPE AND MAGNITUDE OF THE AMBIENT PM1Q PROBLEM IN
THE UNITED STATES?
It is important to understand how the current PM10 networks
have evolved from an historical perspective. The early pioneers
in air pollution control were usually local agencies (city or
county) that were formed as a result of complaints concerning
obvious pollution. These were usually smoking stacks, visible dust
layers on surfaces, or odors. Since total suspended particulate
(TSP) monitoring was cheap and easy, TSP monitors proliferated,
mainly located in urban core industrial and commercial areas.
Residential neighborhoods devoid of industrial or commercial
sources or major traffic arterials were sparsely monitored, and
rural areas were almost never monitored. At one time, there were
approximately 5,000 high volume samplers, which measured TSP, in
operation with virtually no siting criteria, and monitors were
located at the discretion of the local monitoring agency.
The late 1960s and early 1970s saw consolidation of the TSP
networks due to various factors. The emergence of State-run air
pollution control programs, the intensifying of national oversight
on air pollution programs through EPA, and the expansion of
monitoring in other criteria pollutant areas all contributed to
closing the less desirable TSP sites. With the inevitable
comparisons among cities, it became apparent that some form of
uniform siting criteria was necessary, and the Standing Air
Monitoring work Group (SAMWG) was formed.
As a result of the impetus from SAMWG, uniform monitor siting
regulations were developed and promulgated in 1979. These
regulations provided uniformity in monitor siting, still stressing
point sources, commercial sources and high population density. The
regulations were influenced by two studies. One was a national
assessment of the urban particulate problem which was based upon
a 14 city survey of TSP sites, which identified reentrained roadway
dust as a major problem. The other was a series of siting
guidelines for the criteria pollutants which developed the con-
cept of spatial scales of representativeness for monitoring
sites. As a result of these studies, microscale sites for
National Air Monitoring Stations (NAMS) and State and Local Air
Monitoring Stations (SLAMS) TSP monitoring, as well as middle scale
sites for NAMS, were specifically excluded. The implementation of
these regulations caused further consolidation of the TSP network
to approximately 2800 sites. Another study which has major
ramifications on today's situation was the Portland Aerosol
Characterization Study (PACS) conducted from 1976-1978. This
-------
study showed the effects of industrial point source emissions to
be far less than suspected, and it documented that the major
contribution in urban air sheds is reentrained roadway dust and
other crustal material. Also, this was the first study which used
Carbon 12 - Carbon 14 ratioing technique on ambient particulate
samples to document unequivocally the impact of contemporary carbon
(residential wood combustion) on neighborhood particulate levels.
With the advent of the PM^ standard in 1987, the monitoring
guidance tried to incorporate what was then known about these new
sources; it reinstated micrpscale siting for PMj0, especially in
street canyon settings for diesel particulate; and it specifically
addressed residential wood combustion (RWC.). However, the
conventional thinking was that, since PM10 was a subset of TSP,
given the size and the refinement of the existing TSP monitoring
network over the years, it was natural to characterize the existing
PM,Q problem areas from existing TSP data, using the probability
guideline.
Although not specifically asked to address Issue 1 in the
survey, 5 Regions did. Two felt -that there was adequate coverage,
2 felt that there was not, and 1 felt there was no way to know
based on the TSP network. During the Denver Workshop, a consensus
developed that held that the current PM1Q monitoring network
adequately reflected the effect of traditional sources, and was
adequate in terms of meeting the existing regulations.
Professional judgement says that we need to go beyond the bare
minimum to properly characterize the magnitude of the ambient PM,Q
problem in the United States. The areas of impact, for which
additional PM,n monitoring may be needed, are:
1. Areas affected by RWC
2. Agriculture/silvaculture tilling and burning
3. Roadway/street canyon diesel emissions
4, Roadway sanding and salting
5. Construction activities
6. Mining activities
7. TSP nonattainment areas labeled Category III by
probability guideline
8. Category ill areas with industrial sources
10
-------
ISSUE 1A. Why dogs _"fchere appear to be_ a disparity in the
number of nonattainment areas between the West and the East?
The differences may be philosophical, physical, or both. As
described in detail in subsequent parts of this report, there is
a difference between the East and West in the priority given to
PM,,j relative to the other criteria pollutants. In the East, PM10
ranks third or fourth behind 03 and CO in all cases. In the West,
PM^ tends to be the number one priority, or at least, no lower
than tied for second place. With the current monitoring philosophy
of doing more with a constantly shrinking budget, it is not
surprising that, in the East, the focus is on existing problems
like 03 and CO. The special studies were initiated almost
exclusively in the West, and it is interesting to note that it was
in Portland, Oregon that RCW was first identified as a major
contributor to PM1Q levels.
There still remains a question as to whether there are, in
fact, undiscovered PM10 nonattainment areas in the East or if there
are physical differences between the East and the West that
perpetuate this disparity. Most Group I areas with the highest
probability of violating the PM,0 standard are located in the West
(Figure 1). Also, there may be significant differences in the
amount and type of fuel used in RWC between the two areas, however
wood use surveys presented at the Denver meeting tended to deny
this (see Appendix A).
To answer the question definitely, it was the overwhelming
consensus at the Denver PM^ Task Force meeting that special studies
in suspected high impact areas were needed. The saturation study
technique employed in Region X, with low cost, portable, battery-
operated particulate monitors, appeared to be the most cost
effective and quickest way to go and caught the enthusiasm of the
Denver attendees.
The Technical Support Division (TSD) of EPA's office of Air
Quality Planning and Standards (OAQPS) then funded a study of this
type in Asheville, North Carolina. The results of the study are
reported in Appendix B.
Although the results obtained from a single study at a single
location will not answer the question either way, the results will
be useful in beginning to implement the recommendations of the PM10
Task Force. These would include using the results of a saturation
study to evaluate the effectiveness of the existing fixed site
network and to extend the results by inference to other similar
areas.
11
-------
AREA GROUPINGS FOR PM10 SIP DEVELOPMENT
GROUP
Map: WPF/MRB
Data: K. Woodard, CFDD 4/13/87
RFDA
Figure 1
-------
ISSUE IB. Ho current PM1Q monitoring and implementation
policies and practices march up with the objectives of
protecting public health?
Appendix D to CFR Part 58 specifies the approximate number of
PM,0 NAMS stations per urbanized area, depending on the area's
population and the reported PM10 and/or TSP concentrations. It
further states that NAMS stations shall contain a mix of micro,
middle, and neighborhood scale stations. Locations representative
of these same scales are also applicable to SLAMS stations. If.all
stations are properly sited, and all the planned sites are
operational, the monitoring network should meet EPA's objective of
protecting public health, subject to the following network design
needs and the comments on the identification of exceptional data.
Regarding network design, there are potential deficiencies in
monitoring coverage in Group III areas and in determining the
location of the maximum concentrations in monitored areas. Group
III areas, particularly nonurban and smaller urban areas, may not
receive adequate monitoring unless resources are increased and
policies are clarified to encourage monitoring (at least periodic
screening) in those areas. Priority should be given to monitoring
in those areas where residential wood combustion or industry could
potentially elevate PM,0 concentrations. Also, there is often
little assurance that the maximum concentration sites (Category A
sites) are located at the true area of maximum concentration. In
Region X's experience, the maximum concentration can easily be
several times higher than that measured at a traditional center
city location. Communities with industrial sources or area
sources, such as wood stoves or diesel traffic, are most
susceptible to such concentration variations. Saturation
monitoring studies like the Asheville project would be necessary
to ensure that public health is protected. Such studies could
provide low cost screening to address the apparent lack of
monitoring in Group III areas. Also, they would identify the
locations of ^the maximum concentrations so that permanent sites
could be relocated. Policies must be developed which require
saturation monitoring and which define the proper protocol for
their conduct.
Monitoring practices in many States and EPA Regions are
governed in part by the policy issues discussed above, monitoring
budget constraints, low priority on monitoring, and a lack of
resources to solve new air quality problems. This lack of
resources apparently fosters (in some areas), what could be
interpreted as a lack of "will" to identify new problems. The
resource constraints and low monitoring priorities have been
overcome, at least in part, by some Regions and States, and in some
cases, substantial progress has been made in implementing
saturation monitoring and in providing at least periodic monitoring
in all pertinent areas of the State. This perceived lack of "will"
13
-------
to find new PM,Q problem areas because of insufficient resources
to develop and implement control strategies is not consistent with
our public health objectives. The perceived lack of "will" can be
mitigated by better utilization of existing resources through the
use of new policies and tools (e.g. saturation monitoring). This
should be done in concert with efforts to increase monitoring and
SIP planning resources and to fine-tune the policies.
The exceptional events guideline specifies the criteria and
procedures for treating high concentrations due to certain
activities as exceptional events. The guideline should be modified
to address the following concerns: ' * •
a. Prescribed burning that does not occur regularly or
freguently can be treated as exceptional. Current guidance does
not require any demonstration of the efficacy of the smoke
management plan under which the burn was conducted. Also, there
is no mention that naturally occurring forest fires would likely
have less adverse air quality impact if silvicultural practices
such as prescribed burning or removal of fuels from the forest
floor were practiced.
b. Some events may only be classified as "exceptional" if
"reasonable control measures" or other conditions are met. There
is inadequate discussion in the guidance as to what constitutes
"reasonable control measures11 for sources such as construction/
demolition, sanding and salting (for traction on icy pavement),
sandblasting, and highway construction.
c. The provisions for treating agricultural tilling as
exceptional could be improved through coordination with the
provisions of the Food Security Act's soil conservation
requirements.
d. The guideline should also be revised to consider
industrial data collected during startup/shut downs and air
pollution control device malfunctions. For example, the quarterly
lead standard was violated because of an electrical problem which
rendered an electrostatic precipitator inoperable for relatively
short time period.
REFERENCES
1. Federal Register 44:27558-27604. May 10, 1979.
2. D. A. Lynn, et al. National Assessment of the Urban
Particulate Problem: Volume 1, National Assessment. EPA-
450/3-75-024. U. S. Environmental Protection Agency, Research
Triangle Park, NC. June 1976.
14
-------
3. F. L. Ludwig and J. H. S. Kealoha. Selecting: Sites for Carbon
Monoxide Monitoring. EPA-450/3-75-077. U.S. Environmental
Protection Agency, Research Triangle Park, NC. September
1975.
4. R. J. Ball and G. E. Anderson. Optimum Site Exposure Criteria
for SCU Monitoring. EPA-450/3-77-013. U. S. Environmental
Protection Agency, Research Triangle Park, NC. April 1977.
•^ »
5, F. L. Ludwig, J. H. Kealoha, and I. Shelar. Selecting Sites
for Monitoring Total JSuspended Eartieulates. EPA-450/3-77-
018. U. S. Environmental Protection Agency, Research Triangle
Park, NC. June 1977, Revised December 1977.
6. F. L. Ludwig and E. Shelar. Site Selection, for the Monitoring
of Photochemical Air Pollutants. EPA-450/3-78-013. U. S.
Environmental Protection Agency, Research Triangle Park, NC.
April 1978.
7. J. A. Cooper, et al. "Summary of the Portland Aerosol
Characterization Study." APCA #79-24.4. Presented at the
1979 Annual Air Pollution Association Meeting, Cincinnati, OH.
8. Federal Register 52:24634-24750. July 1, 1987.
9. T. G. Pace, et al. Procedures for Estimating Probability of
Nonattainment oil a PM.0 NAAQS Using Total Suspended Particulate
or PM,Q Data• EPA-450/4-86-017. U. S. Environmental
Protection Agency, Research Triangle Park, NC. December 1986.
15
-------
-------
ISSUE 2. HOW WERE PM1Q PROBLEMS IDEMTIFIED?
ISSUE 2a. Did the States/Regions go beyond the utilization
of the existing TSP data in conjunction with the PM10
Probability Guideline to identify PM,p problems? Please
explain and provide examples, if appropriate.
The guidance to the Regions allowed criteria other than the
probability model to be used in establishing the-PM10 groupings*
The model used data froit existing TSP, PM,* and other particulate
monitoring networks to estimate the probability of exceeding the
PM^g standards.
Region X decided to augment the established monitoring with
"saturation monitoring" in determining monitor placement and in
identifying PM10 problem areas.
In some areas, microscale monitoring sites have been used to
obtain mobile source data. There are five such monitors in New
York, including a special study location on Madison Avenue, and
three in New Jersey. The Madison Avenue site has operated since
January 1988. Readings from these sites are among the highest in
the Region. In another special study, Puerto Rico has oriented
monitoring to landfill and open burning problem areas.
Some difficulties in identifying PM10 problems are:
a. Reliance on the probability guidelines has resulted in
many TSP nonattainment areas being classified as Group III areas.
Such reclassifications may have been inappropriate, because there
is no reguirement for further monitoring in such areas and it is
possible that PM,Q problems may exist there. Most Regions are
conducting some monitoring in these areas. A priority should be
placed on establishing PM,g monitors in such areas.
b. Many of the existing TSP monitors may not have been
located in areas representative of the maximum PM1Q concentrations.
In the future, priority should be placed on PM,n monitoring in TSP
nonattainment areas (which are Group III for FM,O) to obtain more
detailed data on these areas. Resources should be allocated to
make these efforts successful. Specific PM,Q monitoring should
address areas heavily affected by nontraditional sources such as
wood burning, agricultural and silvicultural burning, quarries and
strip mines, and high traffic densities. Saturation monitoring
should be the preferred method of finding new PM10 problem areas and
"hot spots."
16
-------
ISSUE 2b. Before the PM^j SLAMS were established, were
special studies conducted by the States^to determine if they
had problems with residential wood smoke, agricultural
tilling,, diesel particulatesf ajid other emissions? Please
explain and provide examples if appropriate.
Limited studies were conducted in Regions I, II and VIII, and
extensive special effort was used in Region VII. In general,
special studies were not widely used. A 1985 report by the Natural
Resources Defense Council with serious concerns about diesel
vehicle emissions in New York City led to the placement of two
microscale sites there in street canyons with heavy bus traffic.
The most serious problem identified in these regards is the
need for more resources to identify and detail the PM1Q problems.
In general, only very limited use of special studies was possible.
The most important needs in the future are to provide special
studies in the SLAMS network design and to give high priority to
the identification of new problem areas.
ISSUE 2c. Were special studies conducted bv States and EPA
Regions/Office of Research and Development in different
parts of the country which would shed light on _jthis problem?
Regions VIII, IX and X were aware of some applicable and
useful special studies, but Regions I through VII were not. It is
obviously a problem if so many of the involved agencies would have
been helped by such information, but were unaware of it. In the
future, provision should be made for technology transfer and
information exchange on special monitoring studies, among all
affected agencies. This could be in the form of an air monitoring
technology center which would fill the need of a technical report
clearinghouse.
ISSUE 2d. Were the States/EPA Regions/Office of Research
and Development special studies designed to capture worst
case conditions in terms of emissions. seasonality and
sampling freguency? What do the studies show?
These special studies generally did address worst case
situations. Most States did use them and benefitted from their
use. It is helpful to users if such studies give first priority
to worst cases and maximum concentrations. Special studies could
resolve various other problems, such as spatial distribution of
pollutants. A useful future effort would be the development of
national guidance or directives on special studies. Such guidance
should include methodology for performing saturation studies, when
and how to deviate from 40 CFR Part 58 Appendix J requirements, the
difference between a screening study and SLAMS/NAMS, and setting
of priorities on the timing and locations for special studies.
17
-------
ISSUE 2e. Are any special studies planned?
All Regions except Region VII have plans for special studies.
For example, Region II is encouraging mobile source-oriented
monitors in New York and New Jersey. The Regions and some States
have recognized the need for, and the importance of, such special
efforts.
ISSUE 2f. Is there a need for additional special studies?
Beyond the special studies already mentioned, most Regions
have found other matters that could be addressed, preferably with
saturation monitoring in most cases. Some specific proposals for
special studies in the future are:
To identify any new problems that occur.
To place monitors at sites representative of maximum
concentrations.
To determine the spatial distribution of pollutants.
To assess international problems, such as transport of
PM10 from Mexico.
To investigate the effects of nontraditional sources that
may not be correctly represented by existing networks,
such as sanding and salting of roadways, gasoline and
diesel exhaust emissions, re-entrained road dust, wood
burning, and mining.
18
-------
ISSUE 3. WHAT ARE THE EXISTING TOOLS USED FOR IDENTIFICATION
AND HOW WERE THEY USED?
The broad question posed to the Task Force under Issue 3
concerns what tools exist for assisting in the design of PM1Q
monitoring networks (identifying problem areas) and how these were
actually utilized in yielding the networks as we now recognize
them. It is worth noting here at the outset that with few
exceptions, -most Regions have designed their PM10 networks as rather
direct derivatives of their historic TSP networks*-. Justification
for this approach references the presumed adequacy of the old TSP
networks in correctly characterizing the scope and magnitude of all
particulate impacts, certain critical portions of Agency guidance
also indirectly endorsed this perspective, most notably the
probability guidelines, the focus of which centered exclusively on
the interpretation of existing TSP data as a surrogate measure of
PM1Q impact potential. So in the most practical sense, it is quite
immaterial in the vast majority of instances to discuss the linkage
between tools for PM1Q problem identification and their relatively
recent application in yielding the current PM10 networks since they
are now, and have nearly always been, largely decoupled.
This is in no way intended to criticize the design of all
networks. Many surely reflect the product of best professional
judgement which steins from years of experience. But there is
concern that many agencies are unable to support the fundamental
design adequacy of their particulate monitoring networks from a
technical basis other than reliance on the former TSP network.
ISSUE 3A. Did the monitoring guidance address residential
wood burning, agricultural tilling, diesel particulates? Was
it helpful? Please explain and provide examples, if
appropriate.
The Regions concluded that an important general distinction
could be made between the objective adequacy of PM,0 monitoring
guidance and the Regions' actual perception ana consequent
application of it. Although the guidance addressed the above PMW
monitoring situations, it was only viewed as advisory and non-
prescriptive by the Regions and was not generally implemented. The
microscale initiative described in the revised particulate
monitoring regulations endorses and, in some instances requires the
deployment of microscale PM^ monitors, but most Regions did not
see this as emphasizing the need to re-examine network design
adequacy. This underscores the importance the Regions assign to
having network design features articulated in regulatory versus
guidance formats.
19
-------
With the exception of the microscale PM1Q sites, most Regions
were substantially satisfied with the representativeness of their
PMW networks until the task force got underway. Once asked, many
Regions responded that, since the development of PM10 monitoring
guidance was simultaneous with the establishment of pre-
promulgation PM10 monitoring networks, there was little real
opportunity to accommodate new network design initiatives when
negotiating and implementing the original networks.
The Regions also felt that resource constraints (additional
equipment capitalization), political pressures, and overriding
Regional priorities (particularly 0.), often combined as obstacles
to realizing proactive revisions to PM10 network design in the
foreseeable future.
It should be noted that several Regions actively pursued
deployment of microscale monitoring sites to augment the earlier
situation and thereby increase the adequacy of available data.
ISSUE 3B. Did the modeling guidance at that time address
these emission sources? Was it helpful? Please explain.
As discussed above, the Regions felt that, in the vast
majority of instances, the design of PM10 networks borrowed heavily,
if not exclusively, from that of the historic TSP networks. Hence,
agencies tended to view the design guidance as neither prescriptive
nor timely. Because of the time that elapsed before the PM10
standard was finally promulgated, control agencies understandably
sensed the inadequacies in earlier guidance and the existing
emission inventories. Leading up to the PM,0 promulgation, there
was also an appreciable period of time before PM,Q emission factors
were available for many important and pervasive sources. Present
guidance should emphasize that the increasing availability of PM1Q
factors since that time should lead directly to more credible
emission inventories and more reliable dispersion modeling.
Whether or not the Regions would have performed dispersion
modeling as an aid to PMIO network design had these Els been
available to them is an open question. It is again worth noting
that for most Regions it is an uncommon notion to attempt linkage
between dispersion modeling and network design, perhaps partly
because these functions are often performed by different
organizational entities that have little apparent contact.
On a technical plain, there are two substantial difficulties
in modeling particulate sources such as RWC, AT, and DP. These
sources are typically transient and aerially diffuse. When
combined with the fact that their associated emission factors and
Els reflect uncertainties that are orders of magnitude higher than
20
-------
for many (perhaps most) point sources, model estimates are not
perceived as having an acceptable level of accuracy.
The second problem stems from the fact that in many instances,
particularly for RWC, the impacts of most interest occur under
stagnant (
-------
Of all Regions, Region IX has seemingly devoted the most
efforts to monitoring PM,n from agricultural tilling and
construction activities. Efforts to identify prototypical
conditions to be monitored were gleaned from a strict review of
proposed exceptional events, Although Region IX has added monitors
in one urbanized area to determine the impact of construction and
are reevaluating the PM10 network in another urbanized area, their
current networks were not designed specifically to monitor impact
from these activities. While other Regions have" considered these
sources in the design of their networks, most have done so either
inadvertently or with TSP sampler surveillance only. ,
Receptor modeling use is viewed by most Regions as a data
interpretation tool, more suitable in apportioning source
contribution to a site's loading than as a determinant in network
design. In this sense, receptor modeling is descriptive of a
site's immediate representativeness, but it does not speak
effectively to the larger issue of adequate subject
characterization. This requires credible emission inventories,
suitable dispersion models (model reconciliation), and most
importantly, the ability to reference relative impacts at many
other locations throughout the grid. The later ability is needed
in order to render sound judgments on the efficacy of network
design.
ISSUE 3D. How many_PM1Q NAMS/SLAMSare located at former TSP
sites or are collocated with TSP monitors?
The data from the survey were not complete, with Region X not
filling out the table. Although absolute numbers are therefore
not available, the question can be addressed on a percentage basis,
using the data from the other 9 Regions. Table i presents the
results of this part of the survey.
TABLE 1
Distribution Of PM1Q Sites
Former Former
TSP TSP
NAMS SLAMS
PM1Q NAMS 59% 23% 18%
PMU SLAMS 24% 63% 14%
PM1Q SPM 3% 18% 80%
22
-------
As can be seen, there is a strong tendency for PM1Q NAMS to
be located at former TSP NAMS sites, and even stronger tendency for
PM10 SLAMS to be located at former SLAMS sites. Only the PM1Q
special purpose monitoring (SPM) sites are primarily new sites.
ISSUE 3E. How were upmonitored areas dealt with? Was
inferential jonitoring used to establish a link between
gimilar areas, when one area had monitoring and the other did
not? -
The Regions generally endorsed the inferential concept only
when considering the representativeness of mobile source-affected
sites. For example, in Region II, New York and Puerto Rico had
high reading monitors along major highways, although the sites were
fairly open. Additional sites were established in likely mobile
source-affected areas which have corroborated the earlier data.
Since most Regions did not orient networks to look at other types
of area sources, such as RWC and AT, the inferential concept was
not considered.
Region X did make concerted efforts to employ this technique,
particularly in RWC situations, recognizing that resources would
not be sufficient to extend and sustain formal network coverage in
all of the areas projected to have NAAQS-threatening RWC. The
Region attempted to develop credible and compelling technical
rationales to support this approach through establishing the
comparability of characteristics shared by two or more areas,
including topography, meteorology and emissions. This concept has
met with substantial resistance from State and local agencies who
believe that empirical data, preferably generated by PM1Q reference
methods, are needed to establish the formal status of all
physically removed areas. The concept has been increasingly well
received because of mounting expressions of public concern over
exposure and of prospects that technically compelling area links
can be demonstrated by short term saturation studies. Region 10
is pursuing the development of a protocol for implementing this
concept that is acceptable to the affected agencies and
communities. The Region has worked with each of its state agencies
to develop a priority list of heretofore unmonitored areas to which
saturation studies and/or formal network treatment will be applied,
as resources allow. This list is quite extensive.
Region IX has made attempts to generalize the results of data
reflecting AT and construction impacts. Similarly to Region X,
the demand for area-specific empirical data in unmonitored locales
is overriding the general application of this concept.
In summary, most of the Regions have not devoted much time
and effort to addressing impacts in unmonitored areas. This has
been largely a result of limited resources and the lesser
23
-------
importance of PM,0 compared with ozone and carbon monoxide. In
most Regions, the PM10 network is a derivative of TSP network.
There has been a recent shift, though in the national focus, to
East-West disparities in PM,,, problem areas and to the Regions'
awareness of the specific network design inadequacies divulged by
one Region over the past several years.
24
-------
ISSUE 4. WHAT WAS THE EFFECT OF PM,0 MONITORING RESOURCES
IN IDENTIFYING THE PM1Q PROBLEM?
Resources limited the extent to which PM10 monitoring networks
were designed.
ISSUE 4A. What^ percentage of monitoring resources was
. allocated to PM1fl and what is^ the priority of PMiA in
relation to the other criteria pollutants? -
From existing FY '88 data the Regional Air Program full time
employees (FTE) in FY '88 were allocated as follows:
Ambient Air Quality Monitoring, 90.7;
Air Quality Management, 288.2;
Stationary Sources Enforcement, 270;
Total 648.9 FTEs.
The Regions' questionnaire responses indicated that between
10 and 50 percent of the monitoring resources were devoted to PM1Q.
An average of these estimates shows that fewer than 18 FTEs were
used in PM,Q monitoring. This estimate, is high, in all likelihood,
because not all Regions devoted 100 percent of their air monitoring
resources to actual air monitoring work. Also, given the various
tasks in the PM1Q monitoring program, Regions could devote only 5
to 15 percent of their PM1(, monitoring FTEs to identifying PM1Q
problems. The 1988 Air Monitoring Workload Model allocated four
FTEs nationwide for PM1Q network establishment.
It is even more difficult, lacking specific information, to
estimate the State resources dedicated to PM1Q monitoring. The
Regions estimated that the States applied between 5 and 50 percent
of their monitoring resources to particulate monitoring. In some
cases, State numbers for 1988 were inflated because of the capital
cost of procuring a large number of new PM^ samplers. State
budgets usually show between 15 and 30 percent of the total 105
Grant resources being spent on the criteria pollutant monitoring
program.
The Task Force was unable to obtain any hard information on
how air monitoring contract funds were used, or if any 105 grant
money was spent on PM1Q NAMS/SLAMS network design projects. There
is some evidence that some Regions used these funds for PM,Q
projects, but it is estimated that no more than $100,000 was spent
nationwide.
Over the past several years in Region X, both base and special
105 Grant allocations have been dedicated to PM1Q network design
projects—totaling between $40K and $80K.
25
-------
Table 2 summarizes Regional responses to the resource
questionnaire, with Priority Ranking.
In summary, it is estimated that Nationally 5% of Regional
PM10 monitoring resources were spent on identifying PM,0 problems
in Fiscal 1988.
In view of this estimate, and considering the low priority
assigned to PM,0 by both the eastern Regions and the 1988 Air
Monitoring Workload model, it is not surprising that the TSP and
Lead networks were used as the basis for designing the PM10 NAMS
and SLAMS Networks. If PM^ problem areas are to be identified in
the future, the Regions and the States need to apply additional
resources to the issue.
ISSUE 4b. How does PM.Q rank as a Regional problem?
Four regions primarily in the West consider PM10 to be their
first or second priority. The remaining six Regions ranked
as either the third or fourth most important concern.
ISSUE 4c. How was potential TSP disinvestment used in the
resource allocation for PM10 monitoring? Were an other innovative
uses of existing resources considered?
in general disinvestment has allowed for some resources,
mostly labor, to be redirected to PM,Q monitoring, although probably
not as much as originally expected. No Region responded to this
question quantitatively. However, Region V stated that, based on
responses from the States, the savings projected by OAQPS through
disinvestment were overestimated. Region VI stated that, although
the disinvestment of TSP monitors helped in funding PM10 monitoring,
the latter is more resource intensive, because of increases in the
frequency of monitoring and maintenance requirements. Several
group members stated that the small surplus value of a TSP sampler
means little capital to be recovered by disinvestment. Labor
savings from TSP disinvestment varies by site location, as more
remote sites require more labor for filter servicing and selected
activity.
Although most Regions indicated that TSP disinvestment was
actively pursued, a substantial number of TSP sites are still in
operation. Several reasons were given for the continued operation
of the TSP network. In many cases, TSP monitors are being used,
formally or informally, as PM1Q surrogates. Because the TSP
network is already in place, such surrogate use requires little
additional capital. Region X said that, in Washington, PM^
samplers were used in key areas, and TSP samplers were used in less
critical areas, with the understanding that any monitors recording
an exceedance of the PM1Q standard would be the first to be replaced
with PM1Q monitors. When TSP monitors are used as
26
-------
Table 2
PM1Q Resources And Priority Ranking
Survey Workload
Monitoring Estimated Model
Region Resources FTEs* FTEs**
Priority
Ranking
I
II
III
IV
V
VI
VII
VIII
IX
X
Total
10-20
5-10
5-10
10-25
20-25
10-40
5-20
10-50
10-30
30-50
.9
.5
.7
1.9
3.4
2.2
.9
2.7
1.8
2.6
17.6
.04
.15
.16
.11
.77
.43
.32
.58
.90
.61
4.07
4
3
3
3
1-2 (Tie)
o3, co
3
4
1 Tie
o3, co
2 Tie CO
1
*Based on the number of FTEs in Workload Model times the average
percent of monitoring resources devoted to PM1Q. This number
overestimated the actual FTEs since Regions do not devote 100%
of air monitoring resources to that work.
**Allocated by the 1988 Air Monitoring Work-load Model to PM1Q
network development.
27
-------
surrogate PM10 monitors, however, PMW monitoring must be initiated
within 90 days of the end of the quarter in which the PM,Q standard
exceedance was recorded. Several Regions indicated that their
State and local agencies did not have the resources to begin PM10
monitoring at sites where an exceedance occurred. Several group
members stated that, in these cases, the true TSP samplers should
be distinguished from surrogate PM,0 monitors, to avoid penalties
for not converting to PM1Q sampling.
A second reason given for continuing TSP monitoring involves
the measurement of toxic compounds, TSP samplers are preferred
over PM10 samplers for measuring toxic compounds because they
capture all particulate matter that could be ingested or inhaled.
There seems to be an increasing interest in monitoring toxic
pollutants in ambient air, and it is likely that TSP will be
sampled, rather than PM10. California plans to operate 20 TSP sites
for analysis of toxic metal compounds in 1989.
Several other reasons were also given for retention of TSP
monitors. One advantage of using TSP samplers is that operator
retraining can be avoided. However, since the filter changing is
not significantly more difficult for PMW than for TSP, and the
calibration and maintenance can be done by trained technicians,
this is not too great a concern.
Another reason given for retention of the TSP network was that
until PM10 SIPs are approved, TSP data are needed for SIP
enforcement. However, SIP enforcement would not be based on
ambient data in most cases.
Several group members stated that, because the historical data
base for particulate is TSP, continued sampling for TSP would allow
continuity in the data, even though a TSP data base may have
limited use. It may be wiser to start the PM10 data base as soon
.as possible rather than build on an obsolete data base, a point
that may also argue against the use of surrogate monitors.
Some indirect benefits may be gained by TSP sampling. An
increased network gives an increased "presence" to an agency, and
having inspectors travel regularly to a site to change filters
entrances the visibility of the inspection program.
Region X mentioned a mobile network as an innovative use of
resources, and some participants suggested that noon-to-noon
monitoring might require fewer monitors,
28
-------
ISSUE 5. CAN EPA DO A BETTER JOB USING EXISTING TOOLS AND
RESOURCES? IF SO HOW? - ' ' '
Yes. Regions IV, V and VI provided written comments to this
question. Region IV commented that EPA would be putting its tools
and resources to better use by doing all the "homework" (costs of
TSP equipment replacement, etc.) before establishing a new NAAQS.
Region V's reply stressed a strict insistence on satisfactory
placement of monitors, along with a review/revision of older
monitor siting. This Region also recommended the redirection of
unnecessary monitoring toward an examination of nontraditional
sources. The response from Region VI emphasized the importance of
considering specific Regional problems when EPA assigns its
priorities to the various pollutants of concern.
There are no easy answers as to how EPA can do a better job
with existing capabilities. The meeting in Denver saw some
detailed discussion of these points, and below is a listing of the
specific items identified.
Tools Resources
Modeling 105 Grants
Monitoring State matching funds
Technical judgement Contract $ from Work Load
Models
Citizen .complaints outside funding
Clean Air Act Superfund
Audits Existing equipment
A230 & A235 WLM
Over the years, some of these elements and resources have
become overburdened at the State and Regional level due to
increasing activities, the lack of offsetting cuts, and no
substantial increase in funds. Also, these tools and resources
have shortcomings that must be kept in mind.
Modeling
When used for initially identifying the ambient levels of a
new area, modeling can require a resource intensive and time
consuming effort. For some situations, such as stagnation and
urban secondary particle formation, appropriate models may not be
29
-------
available to predict pollutant levels. Moreover, for many PM^
areas the requisite model inputs such as emission inventories and
meteorological data, are either incomplete or nonexistent. All of
these factors make modeling a tenuous choice as the sole basis for
identifying potential nonattainment areas. Nevertheless, where
data bases are available, models may be suitable in many cases for
assessing the adequacy of control strategies designed to ultimately
achieve attainment.
Air Monitoring • . : • '.•'".
PM10 has low priority in most Regions compared to ozone, or
carbon monoxide. Headquarters has also emphasized ozone and carbon
monoxide activities (for both sampling and SIPs) in the A230 and
A235 workload models. In the A23O model an average of 1.00
workyear is given to PM10 activities, while an average 5.00
workyears are given for ozone and carbon monoxide activities. In
the A235 model, an average 0.40 workyears are given to PM10, and
an average l.io workyears are given to ozone and carbon monoxide
monitoring.
Overall, air monitoring is a low priority at the State and
Regional Office levels. State and Regional managers see monitoring
not as the foundation that all air programs are built on but as a
required activity that provides little benefit.
Reluctance to identify the problems is apparent. The States
hesitate to identify problems because of their lack of resources.
The States feel that if they identify a problem they will have to
solve it without a substantial increase in resources.
A lack of training exists. The air monitoring program is
becoming an old program. As in all such situations, original
personnel have moved upward in management or onward into other
careers. The new personnel that replace them are usually
given little training, or on-the-job training only. EPA has
recently held several training programs for handling air monitoring
data. When was the last time there were Regional training programs
on the operation of monitoring equipment?
Other shortcomings are evident as well. No PM10 Quality
Assurance (QA) document had been available to involved personnel.
It is now available. Monitor siting guidance is not prescriptive
enough and is designed to recommend monitoring of traditional
sources. There is a lack of national oversight in agency efforts,
and 1PA Headquarters should visit the Regions regularly to
encourage consistent performance. The cost of lab analysis with
increased frequency of sampling and finally, affecting all the
points mentioned is the fact that key staff people are wearing a
lot of hats.
30
-------
PM10 Equipment - A number of points were noted as shortcomings
of the PMj0 reference method.
- Can only obtain 24-hour concentration
- Does not work well in cold or humid weather
Too expensive
- Not enough monitors
Extensive maintenance is required
- Too noisy
Problems with frequency of sampling
Technical Judgement - Technical judgement is used
predominantly in Regional and State selection of monitoring
sites. The majority of all nonattaimnent areas were found in this
manner. To improve the technical judgement of the Regional and
State air monitoring staffs, additional training, monitor siting
guidance and workshops should be provided.
Risk Assessment Approach - Protection of large population vs.
smaller populations comes under scrutiny.
Accountability of contract funds and PTEs - Some of the
Regions complained at the Denver meeting of not having direct
authority over the contract funds in A230 and A235 workload models.
Recommendations
The general tenor of comments and responses to questions about
the air monitoring programs emphasizes many good and positive
aspects of the programs. Still, the respondents have made some
suggestions for improvement which should be considered and which
may be valuable in the future. In general, the importance of air
monitoring has to be stressed at every level, State, Regional and
Headquarters. The air monitoring programs have aged. They have
become less "visible", and there is a belief that State managers
do not see the importance of air monitoring. It is hoped that the
following suggestions and proposals for future work, accompanied
by a more active and visible campaign at all levels, will be useful
and productive.
Saturation monitoring -
HQ contract to buy monitors for short term
monitoring projects
HQ should write the guidance for monitoring, to
include priority for special studies for PM,0 and
prescribed methods for both reference samplers and
utilizing more economical nonreference samplers
as well.
31
-------
Improved evaluation of monitoring sites
Modeling
Monitoring
Improving spatial and temporal coverage
More frequent meetings for Regional Air Monitoring
personnel -
- One meeting should be held in July at Southern Pines and
a second meeting should be held each January or February
at one of the Eegional offices. Alternating the second
meeting among the Regional offices is advised.
Additional funding would be needed to cover travel.
Training -
Provide training on equipment operation through
Regional workshops
Encourage technology transfer among State, Regions, AREAL
and OAQPS HQ
improve training courses in monitoring procedures and
techniques
More frequent HQ monitoring personnel visits to Regional
Offices
- Provide rotational assignments for Regional
staff to HQ
Also -
Evaluate seasonal monitoring for PM10, O3, pb and CO
Have AREAL develop QA guidance for PM1Q
For Regional NAMS Coordinators
Require that the Regional NAMS coordinator be a
staff person at a GS-13 level
Expand the role of the NAMS coordinator
To accompany the above ideas to improve monitoring programs,
the following recommendations are made regarding overall
procedures.
Improve communication between monitoring personnel
and air programs staff
32
-------
After a problem is identified through monitoring,
provide sufficient time to prepare the SIP remedies if
needed.
Public education programs and better marketing
of air pollution problems are needed.
Address not only the potential risk but also the
potential successes. Maybe a risk should be solved if
the solutions are readily availablei
The answer to the question, "Can EPA do a better job with
existing tools and resources?" is both yes and no. Visibility is
a needed key element of not only air monitoring but the air program
in general. To make the air monitoring program more visible it is
recommended to encourage saturation monitoring, to provide better
training and guidance, to focus the duties of the NAMS Coordinator,
etc. Better guidance from Headquarters before promulgation of new
standards, better review of the air monitoring networks, and a
balancing of all air monitoring activities should all be
productive. The air programs and monitoring staffs continue to
balance their duties, although the A230 and A235 workload model
resources are very limited. There comes a limit to the balancing
that can be done without additional resources. Without increased
resources, ultimately some activities will have to be cut.
33
-------
ISSUE 6. WHAT NEW TOOLS, RESOURCES, POLICIES, AUTHORITIES,
ETC., ARE NEEDED?
The Workshop participants have been deeply involved in the
direct application of the modeling and monitoring guidance
disseminated by EPA. Based upon this substantial collective
experience, the following additional PM10 implementation tools have
been determined necessary: - -
1. The Agency should provide data quality objectives (DQOs)
for modeling applications to be used in monitoring
network design. Guidance on when to use existing
screening models and what data are required to run the
model are required to ensure consistent and correct
design criteria. This is a very critical element of PM,0
SIP control strategy evaluation. Since many Group i
control strategies are due within the next 12 months,
this guidance should have a high priority.
2, Technical guidance on the generation of special studies
is another imminent program need. Such guidance is
necessary to assure that all such special studies
(especially those used for SIP-related purposes) possess
sufficient scientific credibility that their findings can
be the bases of court-defensible, specific, enforceable
and performance-based control strategies. Additionally,
the guidance needs to have an integrated approach, to
coordinate monitoring, modeling and control strategy
development.
3. Recently developed portable low cost monitors appear
.ideally suited to applications such as saturation
studies, collocation (with existing TSP monitors) to
establish TSP/PM^ ratios, finding most sensitive receptor
locations, determining complex terrain particulate
deposition rates, etc. These monitors can be used for
intermittent, continuous and documenting meteorological
conditions of interest applications. A swift Agency
evaluation program on this tool is strongly encouraged.
If evaluation shows this tool to be acceptable, large
scale Agency procurement is recommended.
4. Cost effective and more accurate stagnation modeling is
needed to deal effectively with mountain valley
situations. Evaluation of models like WYND Valley should
be considered if sufficient funding to support this
effort could be procured without affecting high priority
items for the PM,g program (medium to long term effort).
34
-------
5. Development of improved monitors in general should be
initiated. ideally, the finished products would be
portable and quieter and less obtrusive than present
apparatus. Beta and gravimetric continuous PM.* monitors
and intermittent monitors possessing these attributes are
needed.
6. Technology transfer needs to be improved. A technical
report clearinghouse should be established to assure.
broader dissemination of technical reports and
information, especially for the PM10 effort. -Additional
technology transfer through workshops would be very
useful, especially workshops addressing inferential
monitoring and coordination of modeling and monitoring
activities.
Resources
Resource allocation to the various segments of the PM1Q program
is currently inconsistent, with some Regions allocating virtually
no funds specifically for PM™ monitoring. Optimal resource
allocation is a critical component of any program's success. To
achieve optimal resource allocations for the PM^ program, it is
important at this juncture to assess thoroughly both its present
and its intended future status. Workshop participants were
virtually unanimous in the determination that a timely anjd
comprehensive study/evaluation should be undertaken to assure
future success. The study should address, at a minimum, the
following resource-related parameters, and it should incorporate
them into a flexible, coordinated approach to resource allocation.
Regional funding of monitoring activities
- Regional priority of PM10 funding
Special study funding
Section 105 grants
A-235/A-20 contract funds
Rotational Monitoring
Policies
Many policy aspects originally left to the discretion of
Regional Offices (sometimes in an attempt to increase flexibility
to individual State needs) will need national program directives
for Regional use as leverage in Federal/State negotiations. Such
national directives will also help to assure Agency-wide
consistency in critical policy areas such as monitor networks,
35
-------
modeling, and control strategy evaluations. Specifically, the
Regional Office representatives felt that national policy
directives are needed for the following to help them implement the
monitoring regulations:
1. Monitoring network evaluation guidance for existing Group
II monitor sites is needed to assure that only
representative site locations and data are used. Also,
consistent criteria for monitoring network placement,
sampling frequency, and extent of coverage of critical
industrial emitters such as steel mills and powerplants
are necessary for the success of .a . complete and
consistent national program.
2. Many TSP nonattainment areas (both secondary and primary)
were classified as Group III for PM10 based upon the PM1Q
probability guideline. In some cases, historically
polluted areas have been removed from any Federally
enforceable monitoring requirement. Without valid
monitoring data to support such determinations, the
public health may be affected and the Agency risks
citizen litigation. Rectification of this problem may
well be considered a late hit by States; further, they
will need additional monitors to cover affected areas.
Thus, it is critical that the Agency provide funding
and/or monitors to support this increased effort, in
conjunction with a consistent and convincing rationale
for such actions.
3. Several Regions felt that it is important that a national
directive be distributed to warn States of the necessity
for representative monitoring networks in Group I areas.
The directive should require adherence to specific
monitoring guidance.
4. Funding constraints and heavy workloads have put the
search for new problem areas low on the Federal list of
things to do and States have similar problems. However,
unless additional particulate problems are identified,
long term funding for particulate programs may be
adversely affected and the Agency will not be fulfilling
its mandate to protect public health. To overcome State
and intra-agency inertia, a national directive is needed
to reserve 5 to 10 percent of State funding for new site
investigation work.
5. The advent of new and low cost PMW monitors will make
saturation monitoring studies much more affordable.
Since such studies could be employed for a variety of
critical air quality applications, it is important to
employ a nationally consistent and scientifically valid
approach to these studies. As in the previous instances,
36
-------
a national directive presenting specific and detailed
mandatory guidance would be the most effective way to
assure these criteria are met.
6. Clear and comprehensive guidance on the use of
inferential monitoring is vital, given our present
resource constraints. The use of inferential monitoring
would save limited monitoring resources and could be
invaluable in locating new potential problem areas. For
such newly discovered problem areas, inferential
techniques could also -be useful in control strategy
development/evaluation. However, the successful use of
inferred data requires its scientific validity and law
court defensibility. Comprehensive and specific national
criteria are necessary for a valid and defensible
national inferential monitoring program.
7. Rotational and surrogate PM,0 monitoring guidance will
be needed for use in PM10 attainment determinations, to
assure their consistent and valid applications.
8. It may be time for the national program to reemphasize
the high funding and manpower priority of PM1Q, relative
to other programs. Various Regional representatives felt
that their management still did not believe PM10 to be a
high priority item on the national agenda. To be truly
effective, a directive stressing these points should be
signed by the Administrator.
9. An agency policy on rural fugitive dust for PM10 is
needed.
In addition to the aforementioned national policy directive,
the collective experience of the Regions suggests a review of the
exceptional events guidelines. Clarification of how frequently an
event may occur and still qualify as "exceptional" is needed, as
well as more specificity as to the type of event that qualifies as
exceptional.
Under some circumstances, waivers to the standard siting
criteria may be appropriate, if conditions at a monitoring site
under consideration are representative of the area (i.e., if the
monitor can be expected to record a typical ambient concentration
to which the local population would be exposed). Appendix E to
the Part 58 regulations specify the procedures necessary to grant
waivers to the siting criteria.
37
-------
APPENDIX A
CORDS OF WOOD USED BY HOUSEHOLD BY STATE
REGION 1
MAINE
NEW HAMPSHIRE
VERMONT
MASSACHUSETTS
CONNECTICUT
*RHODE ISLAND
REGION 2
NEW YORK
NEW JERSEY
REGION 3
PENNSYLVANIA
MARYLAND
*DELAWARE
WEST VIRGINIA
VIRGINIA
REGION 4
*KENTUCKY
TENNESSEE
NORTH CAROLINA
SOUTH CAROLINA
GEORGIA"
ALABAMA
*MISSISSIPPI
FLORIDA
REGION 5
MINNESOTA
WISCONSIN
MICHIGAN
OHIO
*INDIANA
ILLINOIS
2.13
1.42
2.28
.52
.40
.34
1.18
.33
JJ.
.25
.52
.33
.67
.94
.68
.96
.91
.94
.59
.44
.56
.62
.65
.67
.77
.56
.54
.44
REGION 6
NEW MEXICO
OKLAHOMA
ARKANSAS
TEXAS
*LOUISIANA
REGION ?
*IOWA
*NEBRASKA
MISSOURI
*KANSAS
REGION 8
MONTANA
NORTH DAKOTA
WYOMING
SOUTH DAKOTA
*UTAH
COLORADO
REGION 9
CALIFORNIA
*NEVADA
*ARIZONA
REGION 10
*ALASKA
WASHINGTON
IDAHO
OREGON
.51
.51
1.03
.27
.51
.13
.13
.69
.27
.91
.61
.87
.90
.44
.68
.21
.34
.26
1.19
1.09
1.21
1.56
1.26
.56
38
-------
ESTIMATES OF CORDS/HOUSEHOLD FROM USDA FOREST SERVICE - MOST
ACCURATE FROM CAREFUL RANDOM SAMPLE TELEPHONE SURVEY.
* ESTIMATES OF U.S. DEPARTMENT OF ENERGY
LESS ACCURATE
DERIVED INDIRECTUALLY FROM DATA AND ASSUMPTIONS
NATIONAL TOTALS THE SAME
FOREST SERVICE .51 CORDS/HH
DEPARTMENT ENERGY .50 CORDS/HH
39
-------
APPENDIX B
THE ASHEVILLE PM1Q SATURATION MONITORING STUDY
Introduction
In anticipation of the recommendations of the PM1Q: Monitoring Task
Force, personnel from TSD and Region 4 decided to apply the Saturation Survey
Technique to a mountain valley wintertime situation In the East. OAQPS wanted
to obtain first hand experience in saturation monitoring, already successfully
conducted by Region 10 with their own portable, battery operated PM1Q
samplers, for the following reasons: (1) apply the technology of saturation
monitoring to other Regions/States; (2) to demonstrate the low cost feature of
the saturation monitoring technique; (3) to use saturation monitoring in
site/network validation; (4) to look for residential wood combustion (RWC)
effects in eastern mountain and valley terrain; (5) to investigate the effects
of other types of PM,Q sources; and (6) to validate the portable PM-,0 monitor
by collocating it with a reference PM10 monitor.
Participant Responsibility
It was decided that the Asheville/Black Mountain area of North Carolina
offered the best location for such a study, because of topography, logistics
and reference PM1Q monitor availability. Since it was early in January 1988
when the decision was made to try the study that same winter, an added bonus
would be a demonstration of how quickly a saturation study could be designed
and implemented, especially since it involved the cooperation of five
government entities.
40
-------
The following indicates the various participants in the study and their
responsibilities:
Technical Support Division: Provide guidance and funding for the
project. Participate in initial site selection. Coordinate the individual
efforts of study participants. Develop sampling protocol.
Region 4: Provide coordination among State of North Caroling, EPA
Regional personnel and OAQPS. Participate in initial site selection.
Region 10: Provide samplers and training in their use. Assist in
writing the summary data report.
Lane Regional Air Pollution Authority: Acquire filters and mailers,
tare, number and post weight filters, and calculate flow from data logs. Put
data into AIRS format. Participate in writing data analysis report.
Western North Carolina Regional Air Pollution Control Agency: Obtain
any required municipal authorizations for monitoring sites. Participate in
initial site selection. Assist in obtaining site operator. Operate reference
PM,Q monitor according to agreed schedule.
Site Operator: Perform all field work, sample collection and sampler
maintenance,,sampler storage and shipment.
All participants cooperated, and actual site and operator selection had
occurred by February 3, 1989. The samplers arrived and operators were trained
February 16 through 17. The samplers were deployed on February 18 and
sampling commenced then and continued through March 15. Results of the sample
analysis were available by May. The involved agencies relied heavily on in-
kind service; the total cost of the 26-day study to OAQPS was less than
$7,000.00.
41
-------
Site Selection
The study protocol called for 12 portable PM1Q monitors placed within
the Asheville urbanized area. Two of the monitors were collocated with the
local agency's PM10 monitor, and the rest were located through the area using
the following'considerations:
1. local agency experience and needs
2. area topographical maps
3. census housing density maps
4. average daily traffic maps
5. local emission inventory information
To accommodate the schedule of the college student operator, samples
were taken over a 24-hour period (normally from 4 p.m.-to-4 p.m.). The local
agency ran their reference PM1Q sampler daily 4 p.m.-to-4 p.m. during the
study, except when its normally scheduled samples (every sixth day midnight-
to-midnight) were taken. Over the study period, this provided 14 days of
collocated, concurrent PM1Q. sampling at the reference site. Sampled filters
were stored in a freezer and shipped on ice at the end of the study to Region
10 for analysis.
Selection of sites for the study was undertaken in two steps. First, a
general tour was made to identify candidate areas, then specific sites were
chosen within candidate areas.
The tour went first to the Montreat Black Mountain area some 18 miles
east of Asheville. Montreat is in a small steep valley which opens into the
moderate-sized Black Mountain Valley, which in turn opens into a large valley
running along the Swannanoa River and 1-40 into Asheville.
42
-------
The average elevation of the valley floor is 2200 feet at the town of
Black Mountain, dropping to 2000 feet at Asheville. Mountains rise to
approximately 4000 feet at the sides of the valleys. In Asheville itself, Haw
Creek Valley is located just to the east. The Blue Ridge Parkway offers a
good overlook of the valley from near Asheville. Because of the unseasonable
temperatures, (daytime highs approaching 80°F) there was no visible
residential wood burning. January 31 was classified as an open burning day,
i.e., a day when home owners are allowed to burn yard debris. We observed the
valleys to the east of Asheville gradually filling up with smoke. Also, we
observed a relatively high smoke stack at the mouth of the Haw Creek Valley
emitting a visible plume, this from the Sayles Bleachery, which operates a
coal fired boiler for process steam under a variance to emit an opacity of 40
percent continuously.
The following describes the sites used in the study by number and notes
the major PM1Q effects at each site.
001 - Kerlee Community at McCoy and Ruby Avenue, utility pole #KD-5-4-3.
Residential wood smoke outflow from Ridgecrest at eastern edge of Black
Mountain.
002 - Montreat Road, across the street from GKPS Printing Service,
utility pole #CL07. Outflow from Montreat Valley, residential wood burning.
003 - Lake Tomahawk. Recreational park around lake, outflow of Montreat
Valley and Ridgecrest Valley in Black Mountain, residential wood burning. A
utility pole at the edge of the park was used.
004 - Property of Bussman Industry, utility pole #4734 at intersection of
Old Hwy 70 and Keer Fott Road. Influenced by industrial sources and at the
43
-------
outflow of the three previously described valleys. An asphalt batch plant
immediately northwest, a rock crusher complex to the south and a lumber mill
to the southeast.
005 - Haw Creek Canyon Area, little league ball park near intersection of
New Haw Creek and Bell Road, second utility pole between fence and parking
lot. Residential wood smoke. .
006 and 007 - Health Department Building, city center, elevated
commercial site, PM,0 monitor and two portable samplers (to examine both
precision and accuracy).
008 - K-Mart parking lot, commercial area across from another major
shopping center (Asheville Mall). Site overlooks earth moving (ground
preparation) activity for a large construction site. Site is 1 mile northwest
of Sayles Bleachery coal fired boiler. Utility pole MD13-93,
commercial/construct ion.
009 - Bi-Lo Shopping Center. This site is on a light pole in the center
of a vacant grocery store parking lot. The site is off HWY 81 and Fairway
Drive and is across from a golf course. Also at the outflow of the Haw Creek
Valley, a low drainage point along the Swannanoa River, 3/4 mile northeast of
the Sayles Bleachery stack.
010 - On Winston between Lincoln and Broad, utility pole #KC73-1385. In
an interior neighborhood located south of Haw Creek Valley and 1 mile south of
the Sayles Bleachery stack. At a low point in a densely populated
neighborhood with a substantial tree canopy around it. Primary effect is from
residential wood combustion.
Oil - Biltmore Village on Brook Street by Kitchen Place across the street
from old train station, utility pole #8211. High traffic area, low drainage
44
-------
point along Swannanoi River. Observed high amount of diesel truck traffic.
average daily traffic (ADT) at this site is 7000 vehicles.
012 - Patton Avenue, utility pole #DN44. Highest traffic count in area,
at 49,000 ADT. On a triangle formed by Patton Avenue and access to 1-240.
Primary effect is from traffic.
Results . , -
Before presenting specific results of the study, it may be beneficial to
look at the data collected. Table 1 presents the validated ("nonqualified")
data and the caveats covering questionable data, with "B" indicating battery
problems, "D" indicating damaged filter, "T" indicating timer malfunction, and
"W" indicating problems associated with weather. It can be seen that the
valid data compose 92 percent of the data possible. Only 4 percent of the
data are missing and another 4 percent is questionable. Figure 1 summarizes
these data in the form of a histogram denoting the maximum, the 75th, 50th,
and 25th percentiles, minimum, and the mean. From the data, it is apparent
that sites 004 and 009 behave in a manner similar to the rest of the RWC,
affected sites, hence they will be averaged in with the Black Mountain and
Asheville wood burning categories, respectively.
The averages of the two collocated portable PM,0 monitors are 30.4 ug/m3
and 28.6 ug/m3. This agreement of ± 3 percent during the 26 day study is
excellent. Also, on the 14 days in which the reference PM-,0 monitor was
operated on the 4 p.m.-to-4 p.m. schedule, the averages were 35.8 ug/m3 for
collocated sampler 006, 33.1 ug/m3 for collocated sampler 007 and 34.9 ug/m3
for the reference PM,Q monitor.
The average of the city center commercial elevated site for the study
period, as represented by the average of the two collocated portable monitors,
45
-------
Tabli 1
PKiO
FDR PERIOD! 02/JS/9? - 03/M/B9 •• (24 days)
SITE
1 02/18 02/11 02/20 02/21 02/22 02/21 02/24 02/2S 02/24 02/27 02/28 03/01 03/02 03/03 03/04 03/05 03/04 03/07 03/OB 03/W 03/19 43/11 03/12 03/13 03/14 03/15
001 25 23 15 12 D 7 D 2 35 23 14 J9 42 26 IS 16 IB? 13 31 '42 57 . 41 59 15 60 48 2s?
002 30 25 17 8 20 9 8 '29 33 fl 20 25 D 24 14 16 1 5 • 13 29 28 ' 29 32 42 45 38 45 24
W3 17 24 13 8 14 4 i 43 D 7 27 9 4 17 2 4 f 24 3? 42 SB 5! 74 50 45 24
004 18 32 11 10 14 14 5 35 25 14 15 35 21 9 10 1 1 12 26 *40 33 ' 53 33 75 70 3B 25
005 23 20 i 3 D I 2 42 22 ? 28 13 10 I 10 20 11 43 45 49 72 43 11 23
006 37 23 13 7 2S 12 16 52 14 17 27 54 30 30 12 22 IT 21 28 27 31 53 47 82 <8 41 30.4
HftPHlO 22 19-61 31 55 30 IB 16 IB 26 24 44 80 44
007 23 25 4 9 27 12 13 53 16 15 25 55 29 17 15 I? 10 17 28 25 35 53, 45 78 5! 41 28.6
m 19 B 22 10 !6 6 9 M 17 D 45 8 23 IB 44 104 67 44 2? 31 IB 32 29 50 65 71 117 142 74 42 46
009 4 21 T 4 8 15 9 11 44 14 SI 3? 13 24 7 14 22 14 1! 3i 44 48 49 55 78 55 47 27
010 29 T 18 5 13 16 D 14 44 15 23 36 44 28 S3 14 15 14 20 32 44 38 49 50 79 41 44 JO
Oil 30 15 23 30 D 19 « 47 25 31 57 84 44 J7 If 22 38 44 46 38 42 53 61 104 K 60 IS
t
012 30 14 24 1 T 27 71 21 If D 54 71 44 32 B 18 21 23 30 37 40 58 72 52 89 61 52 40
«s, all site! 24 24 12 11 IB 10 13 48 21 17 35 47 31 20 16 14 13 20 33 S4 45' 51 55 85 55 43 31
t
M, TEHP 29 14 44 44 29 19 23 35 42 38 37 34 43 44 50 43 43 14 40 46 51 60 53 44 40 51
US 1.6 7,8 9,0 IS.I 20,0 23,4 18.3 7,6 12.6 6,9 9,9 5,7 4.5 5,7 9.6 11.S 15.9 10,3 10.0 13.1 9.!' 9.1 9.3 7,6 7,9 ?.Q
M SSSEHNMNNNUNUNNEE£HSSUHNNNNSN S S H
-------
FIGURE 1
ASHEVILLE PM10 SATURATION STUDY
75 percentile to maximum
50 to 75 percent!!
25 to 50 percentile
minimum to 25 percentile
average - ,
001 002 003 004 005 006 007 008 009 010 011 '012 . "Reference all site
Method overage
SITE ID NUMBER .- :
-------
3 3
is 29.5 ug/m The average of all the sites is very similar, at 31 ug/m .
This would seem to indicate that the single reference PM10 instrument in the
city center commercial location does a good job of demonstrating what a
larger, more widespread network would indicate. However, on any given day,
the picture can be different. The city center portable monitor provided the
maximum concentration for the network only one time oat of 26. Wood burning
sites 002 and 004 accounted for the network maximum on two occasions, while
the microscale roadway sites Oil and 012 provided the maximum concentration 11
times and the earth moving ground preparation site, 010, recorded the maximum
12 times and in fact recorded the top three values for the entire study (142,
117 and 106 ug/m3'. The ratios between the maximum site and the city center
site ranged up to 2.63, and on eight different days a value was obtained
somewhere in the network that was over two times higher than the city center
site.
Since one of the driving factors behind the study was to investigate the
effects of residential wood combustion, the average of the four wood burning
sites in Black Mountain (001 through 004) and the three wood burning sites in
Asheville (005, 009 and 010) were plotted, as well as the composite average of
all 7 sites, and Figure 2 presents these data. Although Asheville and Black
Mountain are 18 miles apart, they track each other almost identically. Figure
3 indicates wood burning data in perspective relative to other data from
traffic corridor sites Oil and 012, earth moving/ground preparation site 008,
and city center commercial sites 006 and 007. To reduce clutter, the average
of the sites only in Black Mountain is used to calculate wood burning
conditions. From Figure 3, it can be shown that wood burning is the lowest
contributor during the study period, and that microscale roadway sites and the
48
-------
FIGURE 2
C
o
ro
c_
4->
C
03
O
U
01
ro
c_
01
>
90
\
en
50
40
20
10
0
80 -
70
60
PM 10 Saturation Study
Wood Burning Sites
Black Mountain
Asheville
Average all Woodburning
18 19 30 21 22 23 24 25 26 27 28 01 02 03 04 05 06 07 08 09 10 11 12 13 14 IB
Feb. Mar, :
Date - 1989
-------
FIGURE 3
cn
O
m
e
"x,
Ol
*"""'"
C
o
•r-t
ra
4J
OJ
u
c
o
u
OJ
cn
c_
(U
<
150
140
130
120
110
100
90
80
70
60
50
40
30
20
10
PM 10 Saturation Study
Comparison of Site Type Impact
L
A
Black Mtn. Woodburning
Collocated (Reference)
Mobile
Earth Moving
Average^ All Sites
IB 19 20 21 22 23 24 25 26 2? 28 01 OS 03 04
Feb. Mar.
Date - 1989
08 07 08 03 10 11 12 13 14 15
-------
earth moving site indicate the two highest levels. Also, it is evident that
all the specific categories have are highly correlated PM1Q levels. A
correlation matrix was developed which shows that all of the sites have
correlation coefficients ranging from .7032 to .9713 and all were significant
at the .001 level. Table 2 pulls out the key correlations among the various
readings. It is interesting to note that the thr§e highest correlations in -
the study are between the reference and the collocated portable monitors, and
that the correlations of wood burning sites are equally as good when compared
with sites within Black Mountain, within Asheville, or between Black Mountain
and Asheville.
Factors Influencing PM1Q Levels
To gain some insight into what may be causing the PM,g levels to
increase or decrease, the network average was plotted against wind speed and
temperature, as shown in Figure 4. Also indicated are if there was rain (R),
a trace of rain (T), or dry weather (D) during the sampling period. As seen
from Figure 4, there is little apparent relationship between PM1Q levels and
either temperature or wind speed. It is apparent that the PM10 levels came
down when it rained and went up when it was dry. While this is not in and of
itself a blinding revelation, a nonparametric statistical test was applied to
examine the relationship of the direction of the PM10 levels to the direction
of the temperature and to wind speed changes, as well as to the wet/dry
situation. This led to the hypothesis that, all other things being equal,
when the temperature goes down, more wood is burned and the PM10 levels go
up. Also, when wind speed goes up, there is more dilution/transport, and
PM10 levels should go down. When it rains and continues to rain, PM10 levels
should drop, and conversely, when it is dry and continues to stay dry, the
51
-------
TABLE 2
CORRELATIONS
Black Mountain Wood Burning
Site 1
Site Z
Site 3
Site 4
Site 1
1.0
Site 4
.8839
.8639
.8668
1.0
Asheville Wood Burning
Site 5
Site 9
Site 10
Site10
.8504
.8879
1.0
Black Mountain vs. Asheville Wood Burning
Site 5
Site 9
Site 10
Collocated/Reference Method Sites
Site 4
.8506
.83Z2
.8416
Site 6
Ref
Site 7 -
Traffic Microscale
Stop and go vs. high speed
Ref
.9622
1.0
Site 11
Site 12
Site 11
1.0
Site 12
.8974
1.0
52
-------
FIGURE 4
PM
°F ug/m3
90 90 90
80 80 80
70 70 70
60 60 60 L
Saturation Study
Moisture
All site average (ug/ra )
Average Temperature (°F)
Hindspeed (mph)
50 50 50
40 40 40
30 30 30
20 20 20 -
10 10 10 -
0 0
Feb. 18 20 22 24 26 2BMar.02 04 06 OB 10 12 14
Moisture -DRRRRTDDRRDDRRRRTTDDODTDRR
Date - 1989
-------
PM.Q levels should rise. For the purposes of this test, when there was a
trace of rain (T), i. e., less than ,01 inch in a 24-hour period, that period
included in the dry category. In order not to confound the hypothesis with
other emission changes from Asheville, only the RWC-affected sites in Black
Mountain have been used. Table 3 depicts this test by using a 1 for an
increasing change in PM1Q levels and a predicted increase in PM10 levels for
a change in wind speed, temperature or moisture, as described in the
hypothesis. Conversely, a -1 indicates a lowering of the actual or predicted
PM^Q levels. From Table 3 it is seen that, for both the temperature and the
wind speed column, the predicted change matched the actual change in PM10
levels 15 out of 24 times. For the moisture predictor, however, the agreement
was much better, with 20 out of 24 directional change agreements, the highly
significant probability of .0009 indicates that this agreement was not due to
chance alone.
Meteorological Representativeness
In any short-term study, particularly if it is focused on a specific
impact, it is always necessary to determine how representative of the time
period were the meteorological conditions experienced during the study and how
likely was it that the specific impact or conditions of interest occurred
during the study period. Table 4 documents the meteorological ranges found
during the study period and compares them with the area's 30-year averages.
It is apparent from this table that, during this study, it was slightly
warmer, the wind blew slightly faster, and it was substantially wetter than
average winter conditions in the Asheville area. However, when an event of
interest is associated with infrequent meteorological situations which occur 3
or 4 times per year, average meteorological conditions do not tell the whole
54
-------
TABLE 3
PMio DIRECTIONAL CHANGES VS METEOROLOGICAL CHANGES
OBS
l
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
MONTH
1
l
1
1
1
1
1
1
l
l
2
2
2
2
2
2
2
DAY
19
20
21
22
23
24
25
26
27
28
1
2
3
4
5
DPM10
1
• -1
-1
1
-1
-1
1
-1
-1
4
JL
.
•m^ "¥
-1
— 1
-1
6 1
I
DT
-l
-l
i
l
l
-i
-l
-i
i
i
DWS
-l
-i
. i
•<
i
i
H20
<
l ' "
-1
-1
l
1
i i
i
— i
-1
-1
l
i i 1 l
-i
-l
— i
-l -l
i ] -l
_i | _i
i
i -l 1 i
i
7 111
l i
2 8 ] 1 -1 1 1 1
"2 j • 9 I 1
2
2
2
2
2
2
10
11
12
13
14
15
-i ! -i
l
' *
i -i i i i i
o
i i i
i -i -i ! i
i
_ I
-*
i iji
-i
JL
i
-l -1
i
i -i
-1
Agreements - 15* 15* 20*
(*out of 24 trials)
Ho: P(agree) = P (do not agree) = .5
P (20 or more agreements out of 24) = ,0009
P (15 or more agreements out of 24) — X1537
55
-------
TABLE 4
ASHEVILLE METEOROLOGICAL STATISTICS
Temperature ranged from 13 to 71 °F
Average daily temperature ranged from 19 to 63 °F
Mean, temperature for study period was 41.8°F
Average daily wind speed ranged from 2 to 28 mph
Average wind speed for study period was 10.7 mph,
Total precipitation for study period was 4.26 inches
Comparison to 30—Year Averages
Month Temperature Wind Speed Precipitation
(0F) (mph) (inches)
Dec.
Jan.
Feb.
Mar.
39.2
38.8
40.0
46.7
8.7
9.4
9.9
10.0
2.98
2.84
2.89
3.74
Study 41.8 10.7 5.08*
scaled from. 26 days to 31 days
56
-------
story. Studies in the Northwest have demonstrated that high levels of PM,-
resulting from RWC occur under cold, calm stagnation conditions which have
persisted for two or three days. An indicator of these stagnation conditions
is the hours of wind speed less than 3 knots (recorded as 0 on Local
Climatological Data monthly summaries for an area. With the exception of the
first sampling period,. 4 p.m. February 18 to 4 p.m. February 19, which had 16-
hours of calm, there were only 33 hours, or 5,5 percent, of calm during the
next 600 hours of the sampling period. Although the wind speed (average 1.6
mph) and the temperature (24-hr average. 29°F) were favorable in the first
sampling period for high levels of PM1Q from RWC, little RWC effect was
noted. The previous 2 days it had been mild and then wet, and early the day
sampling commenced, it had snowed four to six inches. Also the RWC sources
(houses) in the northwest are more densely packed or closer together, in the
northwestern U.S. experiencing high levels of PM10 from RWC.
Conclusions
In summary, while the question "Does the Asheville area suffer from a
winter residential wood combustion problem?" cannot be definitively answered,
there are indications that the problem is not nearly as serious in Ashevill.e
'as has been found in the West. OAQPS does believe that the study has been
successful in addressing its other stated objectives stated. The study points
out that Asheville's city center commercial PM,Q monitor adequately reflects
the average of a larger network, but that areas of maximum concentration, in
24-hr levels or maximum monthly averages, are likely being missed. Portable
monitors were again shown to correlate closely with the reference PM10
monitor, and a considerable amount of data can be acquired in a short time for
reasonable costs in both money and effort. Consequently, OAQPS is in the
57
-------
process of acquiring portable monitors of its own to encourage agencies to use
the saturation monitoring approach in previously unmom'tored areas for PM1Q
and/or to investigate the effectiveness of their own PM10 networks.
58
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing}
. REPORT NO.
EPA-450/4-89-Q16
2,
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
PM.JQ Monitoring Task Force Report
S. REPORT DATE
October 1989
6. PERFORMING ORGANIZATION CODE
AUTHORS)
W. F. Hunt, Jr., et al.
8, PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
U. S. Environmental Protection Agency
Research Triangle Park, NC and
Regions I - X
10. PROGRAM ELEMENT NO.
It. CONTflACT/GRANT NO,
12. SPONSORINO AGENCY NAME AND ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
Office of Air Quality Planning and Standards
U. S. Environmental Protection Agency
Research Triangle Park, NC 27711
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This report was prepared by the PM,n Monitoring Task Force which was formed in
July 1988. The Task Force was created by William G. Laxton, Director of the
Technical Support Division in the Office of Air Quality Planning and Standards in
response to concerns raised at the Air Division Directors meeting in June 1988. The
Task Force was formed to look into the need for evaluating PM,n monitoring networks
especially in existing Group III areas. The principal purpose of the Task Force was
to address the apparent disparity in the number of PM,0 nonattainment areas between
the Western and Eastern States. The Task Force is composed of people from all ten
(10) EPA Regions and the Office of Air Quality Planning and Standards. This report
addresses six principal issues raised by the Task Force and offers recommendations
on ways to improve the National PM,_ monitoring program.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTlFIERS/QPEN ENDED TERMS
c. cos AT I Field/Group
PM1Q
Particulate
Monitoring Network Evaluation
Monitoring Task Force
Group III Areas
Monitoring Networks
Nonattainment
18. DISTRIBUTION STATEMENT
19, SECURITY CLASS (Tttis Report)
Unlimited
21. NO. OF PAGES
62
20. SECURITY CLASS (Tills page)
22. PRICE
EPA Form 2220—1 (Rnv. 4-77) PREVIOUS EDITION is OBSOLETE
------- |