& 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

-------