REVIEW OF U.  S. ENVIRONMENTAL PROTECTION AGENCY

    ATMOSPHERIC SCIENCES RESEARCH LABORATORY

 DRY DEPOSITION AND MATERIALS EFFECTS PROGRAMS
           Norbert S. Baer,  Chairman
                Pedro Albrecht
              Cliff I. Davidson
             Howard M. Liljestrand
               Ivar H. Tombach
              April 23-25,  1985
                 Prepared by
Research and Evaluation Associates, Inc
     1030 15th Street. N.W., Suite 750
         Washington. D.C. 20005
            (202) 842-2200
       109 Conner Drive. Suite 2101
         Chapel Hill, N.C. 27514
            (919)968-4961

-------
                    TABLE OF CONTENTS

Chapter                                                        Page

   I.  INTRODUCTION                                              1
       1.1  Panel Mission and Responsibilities                   1
       1.2  Materials Reviewed                                   3
  II.  DRY DEPOSITION PROGRAM - Task Group D                     4
       2.1  Background                                           4
       2.2  Task Reviews                                         7
            2.2.1  Parameterization of Pollutant                 7
                   Deposition Velocities
            2.2.2  Development and Evaluation of Methods         8
                   to Measure Deposition Fluxes
                   2.2.2.1  Variance Method                      8
                   2.2.2.2  Similarity/Heat Budget Method        9
                   2.2.2.3  Eddy Accumulation Method             9
                   2.2.2.4  Overall Evaluation of Direct        10
                            Methods Research
            2.2.3  ASRL Concentration Monitor                   11
                   2.2.3.1  Development and Evaluation          11
                            of the ASRL Concentration Monitor
                   2.2.3.2  Development of a Sized Particle     13
                            Collection System
            2.2.4  Research Site Operation and Method           14
                   Evaluation
            2.2.5  Evaluation of a Research Watershed           18
       2.3  Proposed 30-Site Deposition Network                 18
       2.4  General Comments                                    22
            2.4.1  Management Team/Project Staff                22

-------
                TABLE OF CONTENTS  (continued)
Chapter                                                        Page
            2.4.2   Integration of Tasks to Achieve              22
                    Program Objectives
            2.4.3   Related Research: Surrogate Surface          24
                    Methods
            2.4.4   Program Strengths and Weaknesses             25
       2.5  Recommendations  (In Approximate Order of Priority)  25
 III.  MATERIALS EFFECTS PROGRAM - Task Group G                 27
       3.1  Background                                          27
       3.2  Task Reviews                                        28
            3.2.1   Field Studies                                28
                    3.2.1.1   Structural Materials - Metals       28
                    3.2.1.2   Non-Metallics                       31
                    3.2.1.3   Acid Rain Simulator                 33
                    3.2.1.4   Site Monitoring                     34
                    3.2.1.5   Aerometric Data Base                36
            3.2.2   Chamber Studies                              37
            3.2.3   New  Projects  (FY  86)                         38
                    3.2.3.1   Initial  Effects on Painted          38
                             Wood Substrates
                    3.2.3.2   Effects  on Concrete and             39
                             Masonry  Structures
                    3.2.3.3   Effects  on Roofing Asphalt          39
                             Shingles
                    3.2.3.4   Evaluation of Corrosion Monitor     40
                             at Exposure Sites
       3.3  General Comments                                   40
            3.3.1   Management Team/Project Staff                40

-------
                    TABLE OF CONTENTS  (continued)


Chapter                                                        Page

            3.3.2  Integration of Tasks to Achieve Program      41
                   Objectives

            3.3.3  Relationship Between Field Studies,          41
                   Chamber Studies and Damage Functions

            3.3.4  Related Research                             45

            3.3.5  Program Strengths                            45

            3.3.6  Program Limitations                          45

            3.3.7  Quality Assurance and Quality Control        46

        3.4  Recommendations  (In Approximate Order of Priority)  47

   Appendix  A     Peer  Review of the  Acid Deposition, Dry      49
                   Deposition, and Materials Damage Program

   Appendix  B     EPA-ASRL  Peer Review Panel                   51

   Appendix  C     Process Evaluation  Results                   52

   Appendix  D     ASRL  Response to  Reviewers'  Comments         56

   Appendix  E     Review of  the Panel  Report and Responses      65
                   by the Laboratory Director

   Appendix  F     Clarification Comments  by ASRL               67

-------
                              I
                        INTRODUCTION

         1.1  Panel Mission and Responsibilities
     The panel's mandate was to review two U.S. Environmen-
tal  Protection Agency - Atmospheric Sciences Research Lab-
oratory  (EPA-ASRL) programs: (a) Dry Deposition (Task Group
D) and  (b) Materials Effects (Task Group G).   Both efforts
are  part  of the National  Acid  Precipitation  Assessment
Program  (NAPAP).  The specific tasks reviewed are listed in
Table 1.
     The  projects  and  the associated tasks  were  to  be
reviewed from the standpoint of:
      (1)  objectives  and  technical  approach,  including
          quality assurance and quality control;
      (2)  outputs planned;
      (3)  the relationship of the outputs  to the goals of
          the Agency; and
      (4)  the usefulness of the overall programs to the
          scientific community and/or public welfare.
      In  attempting to meet this mandate,   the panel has not
only  commented on the individual tasks listed in Table  1,
but   has also considered their relationship  to the broader
goals of the Agency,  i.e., development of a dry deposition

-------
                                          Table  1
                           Descriptions of Tasks  Reviewed
 Dr   Deosition - Tiak Group D
 Parameterization of Pollutant Deposition Velocities
 Development and Evaluation of Variance Method
 Development and Application of S1n1lar1ty/Heat Budget Method
 Development  and  Application  of  Eddy Accumulator
 ASRL Concentration Monitor
 Development of a Sized Particle Collection System
 Research Site Operation and Methods Evaluation
 Evaluation of a Research Watershed
 H«t«rU1« Effeett - T««fc fit-nun R
Field Studies - Structural  Materials (Metals)
Field Studies - Non-Metalllcs
Field Studies - Add  Rain Simulator
Field Studies - Site  Monitoring
Field Studies - Aeronetrlc  Data Base
Chamber Studies - Chemistry of Acid Precipitation
  Materials Effects
Nev Projects - Initial Effects on Painted Wood  Substrates
               Effects on Concrete and  Masonry  Structures
               Effects on Roofing Asphalt Shingles
               Evaluation of Corrosion  Monitor  at Exposure Sites
C on trie tor/Agency
Argonne
Argonne
NOAA
In-House
In-House
(RFP)
Argonne/NOAA
NOAA
Contractor /A ganev
BOMInes
Northrop
Northrop
Various
CSC
lJUk Kfl*
1452
1452
3014
3016
4037
5059
1452/3014
1014
laxk ILfU
3024
4059
3175
3114
4062
Northrop
3174

-------
network and estimation of economic damage attributable to
acid deposition.

                 1.2  Materials Reviewed
     The  panel has reviewed the information and materials
made  available in scheduled  presentations,  draft  manu-
scripts reporting  results,  interagency agreements,  pro-
ject descriptors, and the most recent peer reviews of both
programs.   It  has visited the chamber facilities of  the
on-site  contractor and the RTF field test site.   At  its
request,  a  conference  call was made to Bruce  Hicks  to
discuss  the status of the NOAA deposition velocity  para-
meterization  project deemed critical to the concentration
monitoring approach.
     The  panel especially appreciated the willingness  of
ASRL  to  arrange  unscheduled meetings at  short  notice.
These  additional meetings,  which were requested  by  the
panel  to augment the review process,  provided  important
information and enabled a more thorough review.

-------
                            II
    DRY DEPOSITION MEASUREMENT PROGRAM - Task Group D

                     2.1  Background
     The  ultimate objective of the dry deposition program
of Task Group D is to determine dry deposition to surfaces
throughout the United States and to monitor future trends.
The dry deposition measurement program of ASRL has several
broad  components  directed  at  meeting  this  objective,
namely
      (a)  Research  on  micrometeorological  methods   for
          determining  dry deposition influences  (the  so-
          called "direct" methods);
      (b)  Determination  of dry deposition velocities over
          various natural land surfaces;
      (c)  Development  of a method to monitor  atmospheric
          concentrations   of  the  chemical  species   of
          greatest interest in dry deposition;
      (d)  Operation  of three field sites for dry deposi-
          tion measurement research; and
      (e)  Evaluation   of   the  utility  of  a   research
          watershed for estimating dry deposition.
The   relationships  among  these tasks are  shown  in  the
diagram in Figure 1.
     A major conclusion of research to date is that direct
measurement  of dry deposition is difficult,  and  is  not

-------
        I  1979 DRY  I
        l DEPOSITION t
        I  WORKSHOP  I
        L--.— -J
              i
R t 0 ON DIRECT METHODS:
1.   EDDY ACCUMULATION (I-H)
2.   SIMILARITY/HEAT BUDGET  (NOAA)
3.
     R 4 0 OF PROTOTYPE CONCENTRATION MONITOR  (C.M.):
V. PARAMETERIZATION
 d   (ANL)
           SUITABLE  \ YES
       FOR SITE CAL1B.,
           RESEARCH,
             QA  7
        HNO,, N03", N02, SOj, 50^ , NHj,
        NH4 , and II* (I-H,  BATTELLE, RTI)
        03 (I-H)
        LARGE PARTICLES (RFP)
        INTERCOMPARISONS (UNC, BATTELLE)
                                                        OPERATING
                                                     DRY DEPOSITION
                                                         NETWORK
                              < _	
 I	-1
 t  NETWORK C.M.  i
^  AND SITES    i
 ,    (EMSL)     I
 t_ _ _ _ _ _ _ J
                                                                                 NOAA: NOAA  - OAK RIDGE
                                                                                 ANL: ARGONNE NAT. LAB
                                                                                 I-H: IN-HOUSE
         Figure  1.   Schematic Representation  of  the  Components of
                       the  ASRL  Dry Deposition Measurement  Program

-------
practical with current technology in a large surveillance
network.   An  alternate approach appears to be estimation
of dry deposition flux from measurements of the  pollutant
concentrations and deposition velocities.   Components  (b)
and  .(c) above address the two variables required for such
an  "indirect"  method,   where  concentrations  would  be
measured and the deposition velocity,  Vj,  would be esti-
mated  by interpolation or extension of measurements  that
have  been  made at a variety of  research  sites.   In  a
separate program, the Atmospheric Turbulence and Diffusion
Laboratory of NOAA is developing the methodology for esti-
mating V(j at sites where measurements are not available.
     Component   (a)  is  concerned  with  development   of
methods  that  could be used in the field  for  deposition
flux  research  and  for validating the estimates  of  the
indirect method.  Component  (d) deals with operating three
field  sites  where basic measurements of deposition  flux
can be made and the indirect methods can be tested.
     An alternative method for determining dry  deposition
fluxes is being evaluated under component  (e).  Here, ASRL
is  exploring  the feasibility of direct determination  of
dry  deposition from the difference between  measured  wet
deposition  input to a watershed and measured  groundwater
and stream fluxes out of the watershed.
     This section reviews the various tasks that  comprise
these components and evaluates their design and  execution

-------
from the perspectives of the program goals.   In addition,
because  the intended use of the methods is in a  national
dry  deposition  trends network,  we discuss briefly  some
issues  raised by our review that have relevance  for  the
implementation of that network.

                    2.2  Task Reviews
2.2.1  Parameterization of Pollutant Deposition Velocities
     The  objective of this task is to express  deposition
velocities  for  each  of  several  pollutant  species  as
functions  of  atmospheric conditions and  surface  condi-
tions.   Various  types  of surfaces will  be  considered,
including  cropland  and  deciduous forests  in  different
seasons.   The  eddy  correlation method will be  used  to
measure V .,  while meteorological data and surface charac-
teristics will be assessed simultaneously.   Such informa-
tion will be used to develop the desired expressions.
     This  is an important task,  since the  link  between
meteorology/surface observations and deposition velocities
is not well understood.  However,  the project description
is unclear as to the details of the parameterization.  The
description  implies that some form of mathematical  model
will be developed.   However, ASRL responses to reviewers'
questions indicated that no such models will be  developed
as part of this particular task.   Rather, the parameteri-
zations   will  consist  of  tables   listing   deposition

-------
velocities for various atmospheric and surface conditions.
     The  panel  feels that more explicit  definitions  of
project  tasks  are needed,  with clear statements of  the
methods  to  be employed and  the  expected  deliverables.
Despite problems with the project description,  the  panel
feels  that parameterized listings of the deposition velo-
cities are worthwhile,  and that the project is  important
to the goals of Task Group D and should continue.

2.2.2  Development and Evaluation of Methods to Measure
Deposition Fluxes
     2.2.2.1    Variance Method.    Measurement   of   dry
deposition  flux  by  the  variance  method  may  be  more
appealing  than  other  techniques if the  method  can  be
developed.   The  primary advantage is that the  necessary
measurements   have   resolution  times  which  are  slightly
relaxed   compared with those needed for eddy   correlation.
Pollutant  monitors  with time  responses of  3-5  seconds,
rather than <1 second as needed for eddy correlation,  may
thus be  usable.   The method was  recommended  for  further
development by the 1979 Dry Deposition Workshop.
      Initial   results of this study of the variance method
indicate  that  ozone  fluxes determined by this  method  agree
reasonably well with those determined by eddy  correlation
when the former method is based  on water  vapor  fluxes.
Agreement is  poorer when heat  fluxes are  used,  possibly
                            8

-------
due  to greater  sensitivity to the upwind  fetch.   Since
fast-response humidity sensors are expensive,  and require
considerable maintenance,  the variance method may be use-
ful  primarily  as a research tool.   Initial work by  the
contractor appears promising, and so continued work on the
project  is worthwhile,  even if the major utility of  the
method  is currently limited to the realm of research  and
special studies.
     2.2.2.2    Similarity/Heat Budget.   Dry   deposition
measurement  by  the similarity/heat budget  (S/HB)  method
offers  the advantage that absolute turbulent fluxes  need
not be measured.   As a  result,  the method is less  site-
specific  than  some of  the other techniques.   The  time-
response requirements are similar to those of the variance
method.  According to presentations made to the committee,
ozone is the only pollutant for which the S/HB method  has
been successfully applied.   However, the method is poten-
tially  applicable  for  estimating the deposition of  fine
particles.
     This investigation  of the S/HB method is in its early
stages of development.   The technique appears to be worth
exploring,   and   continued   funding   is   recommended.
     2.2.2.3   Eddy Accumulation Method.  This method  was
developed  in  an  attempt  to overcome  the  problems  of
measuring small concentrations of certain pollutants  over
short  time scales.   The eddy accumulation method permits

-------
the species of interest to be collected over extended time
periods  in  two separate samples:   one is  activated  by
upward-moving  air and the other by  downward-moving  air.
The  net  flux  to the surface is inferred from  the  dif-
ference of upward and downward fluxes.
     An  eddy accumulator has been built in-house by  ASRL
and tested in the field.   Unfortunately, data reported to
the  panel from initial experiments involving water  vapor
transport  did  not give reliable flux data.   It  is  not
known whether the poor results were caused by data  acqui-
sition  problems  or by fundamental difficulties with  the
method.
     The panel feels that work should continue despite the
disappointing initial results.   The project is relatively
inexpensive, and although considerable additional develop-
ment  work may be needed,  the method offers promise as  a
technique for pollutants which cannot be measured in real-
time.
    2.2.2.4 Overall Evaluation of Direct Methods Research.
Based  on  work  conducted over the  past  few  years,  it
appears that none of the micrometeorological methods  dis-
cussed  in this report (variance,  similarity/heat budget,
and eddy accumulation) are likely to be useful for routine
monitoring  of  dry  deposition  by  unskilled  personnel.
Nevertheless,  development of the methods for use in  site
                           10

-------
calibration,  quality  assurance,  and  research  purposes
justifies continued work as planned by ASRL.

2.2.3  ASRL Concentration Monitor
     2.2.3.1  Development and Evaluation of the ASRL
Concentration Monitor.   In order to  measure  atmospheric
concentrations of the gaseous and particulate species that
are  of greatest interest for acidic dry deposition,  ASRL
has  developed  a prototype device for the  collection  of
samples  of these species.   The design criteria  for  the
sampler include that it be usable reliably in a network to
measure  mean  day-time  and night-time  values  of  these
species  over  about one week periods with 15%  or  better
accuracy  and precision.   It is desired that analysis  of
the  samples  be by the same methods used for wet  deposi-
tion, for maximum comparability.
     The    development  of  this  device  appears  to   be
proceeding  well.   The basic design,  including that of a
transition  flow reactor,  is well thought out  and  seems
sound.   Many of the collection techniques for gaseous and
particulate species are based on accepted methods.
     The  test results shown to us for gaseous HN03,  NH3,
SOpand N0«  and for particulate SO^, N03, and NH^ give con-
fidence that the  basic design is sound  and that the ±15%
precision   goal can be met.  We were not shown  data  that
demonstrate whether  the accuracy specification is  being
                            11

-------
met for all of these parameters.
     The  sampler  should  be tested under  a  variety  of
sampling and shipping temperatures,  with realistic inter-
vals between sampling and sample analysis, to give further
confidence in its accuracy and precision under  conditions
representative   of   network   operation.    Simultaneous
sampling  for oxidants and NOX should demonstrate  whether
or not the NO to N02 ratio is altered as Og is absorbed.
     The  planned  improvements  to the device  to  permit
sampling of H202 and 03 and for improved collection of HNOj
should be developed and tested.  As discussed below, there
is  an urgent need for this sampler for several  purposes.
Therefore,  even  while the above  improvements  are  being
developed  in  the laboratory,  we recommend that  several
such  samplers should be deployed  in the field to  provide
an  operational  field evaluation  before  commitments  are
made for permanent deployment.
       The  sampler  does not yet  have the  capability  of
measuring,  or has not been tested for,  other species  of
interest to NAPAP, namely  HC1, HCOj, Ca4*, PO^7 Na+,  Mg**,
organic   acids,   and   trace   elements.   Although   it
should  ultimately  measure as many of these as practical,
these  species are generally of less interest  than  those
which it already measures, or for  which a sampling capabi-
lity is being planned.   Therefore,  we strongly recommend
                           12

-------
that  deployment of this sampler should not await develop-
ment  or verification of the capability to  measure  these
additional species.
       2.2.3.2  Development of a Sized Particle Collection
System.   A  procurement  request is being  processed  for
development of a method for sampling particles larger than
2 urn aerodynamic diameter with the concentration  monitor
described  above.   The  concentration  monitor includes a
cyclone that excludes  particles larger than 2 urn; the pro-
curement  proposes modification to permit sampling in  two
additional size ranges — 2-10 um and >10 urn.
       The  panel feels that an effort to measure  2-10 um
particles  has merit and should be pursued.   However,  we
seriously  question  the merit of  collecting  the  larger
particles by use of the airborne concentration monitor and
feel  that  the  resources that would be devoted  to  this
purpose  should be used elsewhere.   The reasons  for  our
recommendation  are numerous,  and range from  fundamental
questions  about the scientific merit and validity of such
sampling to questions about the likelihood that such large
particle  sampling can be accomplished at a cost  that  is
reasonable relative to its benefits.
       Specifically,  we  point  out that  the  deposition
velocity  of  such  larger particles varies  greatly  with
particle  size.   Consequently,  the range of sizes >10 um
must consist of several narrow ranges if the  measurements
                            13

-------
are  to have any meaning for dry deposition  characteriza-
tion.  An inlet that can sample particles >10 urn and that
maintains  the  same sampling characteristics over a  wide
range  of  wind speeds would be  relatively  expensive  to
develop  and test.   Finally,  the possibility of gravita-
tional  deposition  of the larger particles  in  the  flow
channels  of the concentration monitor and its  transition
flow reactor has to be considered.
       We  recommend  that  the  resources  available   be
devoted  to  the 2-10 urn collector.   That collector  will
sample  enough  of the coarse mode of ambient  aerosol  to
identify  whether  it  plays  a meaningful  role  in  acid
deposition.  If research using the 2-10 jim collector indi-
cates that coarse particles do play a major role,  then it
might be appropriate to undertake the more difficult  task
of  measuring particles larger than 10 urn.   The issue  of
>10 urn particle deposition may  best be approached through
the use of surrogate surfaces  (see Section 2.4.3).
2.2.4  Research Site Operation and Method Evaluation
       Research   (core)  sites  are  necessary  for   the
measurement  of  parameters to  determine  dry  deposition
flux,  to field test the concentration monitoring  techni-
ques,  to  intercompare methods of  different  theoretical
basis,  and  to gain experience applicable to the develop-
ment of the larger National Trends Network  (NTN).   Three
                            14

-------
sites  are now operational:  Argonne (funded through ANL),
Oak Ridge,  and Penn State (funded through  NOAA).   Addi-
tional  measurements are being taken at cooperative  sites
at  Champaign-Urbana (flat site of uniform surface),  West
Point  (flat  site of broken  vegetation),  and  Whiteface
Mountain  (uniform vegetation on variable terrain).
     One  of  the  planned outputs of the effort  at  the
research  sites is a manual for site planning and  develop-
ment  of  NTN sites.   While the choice of sites  involves
many considerations,  the need for a scientific basis  for
the  decision  is  critical.   A poorly chosen  site  will
provide little useful data, be expensive, and erode confi-
dence  in the ability to quantify the dry  flux.   It   is
important,  in the initial NTN site selection process,   to
make  use of the expertise gained from the  research  site
studies.
     The  choice of the three main research sites has  been
well  made,  with  careful consideration of  the   research
design.   The  initial   emphasis  is on uniform  sites   (the
easiest   micrometeorological situation)  of differing  sur-
face/vegetation type,  with  extension to non-uniform sur-
faces  and  variable terrain.   The  time-integrated  con-
centration  monitoring techniques are being  intercompared
with standard,  realtime monitors as available.    The use-
fulness   of  a 12-hour  (day-night)  sampling  protocol   is
being tested,  as well as  the concentration monitor system
                            15

-------
in general.
     The  concentration measurement and flux parameteriza-
tion protocols have thus far been addressing Oj, S02, HN03,
N02 and particulate matter.   To meet the EPA goals, addi-
tional  species are needed.   Hydrogen chloride,  ammonia,
nitrate aerosol, sulfate aerosol, and aerosol acidity have
all  been previously identified as components of  the  net
acid  flux  which needed to be quantified.   The  alkaline
components,  large particles and ammonia,  have received a
lower priority than the acid species.  However, both acids
and  bases  are needed to determine a net acid  flux,  and
thus  should be considered in the design of research  site
operations.
     This  effort is directed toward quantifying  the  dry
flux, which could be of magnitude equal to that of the wet
deposition.  The uncertainties have been improved, but the
limiting factor is still the estimation of the dry deposi-
tion  velocity  (V
-------
long-term  research.   Improvements  are expected  in  the
routines used to estimate V  from the meteorological para-
meters  as  well as the surface  characteristics  and  the
concentration parameter measurements.
     The  EPA goal is to estimate the dry flux over  large
areas.   However, the actual geographical areas over which
the concentration monitor is applicable are variable  from
region  to region,  and are largely unknown.   These  will
need to be better understood for the NTN sites.
     The dry flux measurement sites have added utility for
other Task Groups.   While the research sites are designed
for method testing and protocol development,  future moni-
toring  sites should be cost-effective in their  location.
In particular,  dry deposition monitoring in the RTF  area
should  be directly coordinated with the materials effects
studies.   A  co-location of research  activities  between
Groups D & G in RTF, North Carolina would be desirable.
     With  respect  to  the  usefulness  of  concentration
measurements and V  estimations,  it is uncertain what can
be expected as to sensitivity and precision.   The concen-
tration measurement approach is anticipated to be accurate
within ±15% or  better, and the V  accuracy under the most
ideal  site  conditions currently ranges from ±30% for SO
to  ±100% for fine particulate matter.  These values imply
significant uncertainty in the final deposition flux esti-
                          17

-------
unavoidable.  Improvements are needed to quantify the flux
within ±30% for all the above species which are  important
to the net dry deposition acidity.

2.2.5  Evaluation of a Research Watershed
     A preliminary field study of the feasibility of using
mass balance over a watershed (Penn State) to quantify the
wet  and dry fluxes is near completion.   The mass balance
approach is well founded, but a precision better than that
for  the dry flux methods seems unlikely.   This  approach
would not have a general applicability.
     The  advantage of a watershed is that it  provides  a
check  on  the  calculated dry flux and the  measured  wet
flux.  It may be useful as an indirect confirmation of dry
flux  measurements.   However, since a watershed covers  a
larger  and  less uniform area than the expected  area  of
representativeness for the dry flux sites, the results may
not be directly comparable.
     The  uncertainties associated with a watershed  study
are large,  and feasibility should be demonstrated  before
future expenditures are made.
         2.3  Proposed 30-Site Deposition Network
     EPA/EMSL intends to deploy a nationwide network  (NTN)
to  measure  trends in both dry and wet  deposition.   The
initial  deployment  of 30 sites is to take place  in  the
next two years,  or so.   Although this network is not the
                           18

-------
responsibility  of ASRL,  several of the  ASRL  activities
have bearing on the implementation of such monitoring, and
therefore we have taken the liberty of noting these cross-
laboratory implications.
     The  tentative plans for the 30 sites include  equip-
ping   each   site  initially  with  a  monitor  for   (X.,
                                                        O
a  device  that  samples  ambient  concentrations  of  key
species of interest to acidic deposition, and meteorologi-
cal instruments.   The sites will not monitor dry  deposi-
tion,  but rather dry deposition will be estimated using a
concentration x V^ method  currently under development  by
the  NOAA Atmospheric Turbulence and Diffusion  Laboratory
(NOAA/ATDL).
     Consequently, the quality of the trend data generated
by  the  network depends strongly on the  quality  of  the
concentration  measurements  and the  deposition  velocity
estimates.   Several  methods could be considered for  the
concentration measurements,  including the ASRL Concentra-
tion  Monitor,  whose  accuracy for the major  species  of
interest is likely to come close to the ±15% goal set as a
design criterion.  The accuracy of the deposition velocity
estimates is likely to be poorer, however, and it could be
much poorer unless careful attention is paid to the design
of the network.
     Bruce Hicks,  the Director of NOAA/ATDL, indicated to
                           19

-------
 the  panel  that the methodology they are  developing  for
 estimating deposition velocity is highly preliminary.   It
 has  only been tested against measurements at a few  loca-
 tions  and  only for S02,  03,  and HN03 (and,  with  much
 uncertainty, to N02); it has not yet been applied to par-
 ticles.  If the parametric method, at its present state of
 development, were to be applied at an ideal site (one with
 topographic  and  vegetative uniformity for at least 1  Km
 around  the measurement location) with properties  similar
 to  those  of the research  sites.  Hicks  estimates  that
 weekly  average  deposition velocities would  be  accurate
 within ±30% for 63 and SC^ (and, possibly N02) within ±50%
 for HNO-J and ±.100% (rough estimate) for particles.
      Hicks indicated that he was uncomfortable with use of
 the  ATDL method without considerable further  development
 and  evaluation.   He  indicated  that  he  had  not  been
 approached  concerning  a site selection  and  measurement
 approach appropriate for use with the parametric method.
      It  is  clear that the success of the trends  network
 depends largely on reasonably accurate estimates of  depo-
 sition velocity.  This places a sizeable responsibility on
 ATDL  for development of a suitable method for  estimating
 the  deposition  velocity.  Major political decisions  may
 ultimately be  made  based on findings that depend on that
 method.
     This  situation concerns the panel from  two  perspec-
tives:
                            20

-------
      (1)  There appears to be little communication between
          ASRL and ATDL, and EMSL and ATDL, concerning the
          deposition velocity method.  Yet,  both ASRL and
          EMSL   are   undertaking  projects   under   the
          assumption  that  such  a  model  will  soon  be
          available.

      (2)  The  selection of sites by EMSL for the  30-site
          network is likely to be severely constrained  by
          the limitations of the deposition velocity esti-
          mating  method.   EMSL  needs to consider  these
          constraints in the network design.  If it cannot
          perform  its  task  within  acceptable  accuracy
          limits  under these constraints,  then  it  will
          have  to  deploy  sites at locations  that  will
          provide lower accuracy (and hope the  estimating
          method  will improve in the  future).   Alterna-
          tively,  it will have to delay the deployment of
          monitors  until  the estimating  method  imposes
          fewer constraints.   In either case, acquisition
          of  acceptably accurate and complete dry deposi-
          tion data would be delayed.

     Considering these factors,  the panel strongly recom-

mends  that  efforts  be  made to increase  the  level  of

technical  interaction between ASRL,  EMSL,  and  ATDL  on

issues  related  to dry  deposition  monitoring.   Regular

meetings  to  review plans and progress are  essential  to

insure the highest possible probability of acquiring mean-

ingful data from the initial 30-site network.

     The  panel  also recommends that the capabilities  of

other qualified organizations be sought to assist develop-

ment of deposition velocity estimation methods.   The cru-

cial role of the dry deposition velocity in obtaining flux

estimates  for  the 30-site network requires that  a  con-

siderable  effort  be devoted to parameterization of  Vj .
                           21

-------
The overall efforts should be peer reviewed by experts  to
maximize  the chances of developing a viable method  which
is of use to NAPAP.

                  2.4  General Comments
2.4.1  Management Team/Project Staff
     Overall,  the  panel  feels that the ASRL  management
team represents a competent and motivated group of indivi-
duals.   The presentations were reasonably well  prepared.
The  panel is pleased that ASRL has been candid and  frank
concerning  perceived  weaknesses in their  program.   All
ASRL members with whom the panel interacted were receptive
to the critical evaluation being provided, and appeared to
be  genuinely  interested in obtaining feedback  from  the
panel to help improve their work.  Similarly, the in-house
technical  staff and external contractors appeared  to  be
competent in the areas being addressed.

2.4.2  Integration of Tasks to Achieve Program Objectives
     The panel recognized the difficulties in coordinating
programs  such as those in Task Group D.   However,  there
are  a  number  of areas in which  efforts  toward  better
coordination would be beneficial.
     The  manner  in  which the written material  sent  to
panel members was prepared is indicative of less-than-
optimal organization.   For example, the in-house projects
                           22

-------
on  the  concentration monitor and eddy  accumulator  were
summarized by inclusion of one manuscript  each,  although
the   bibliographies  of  both  papers  listed  additional
manuscripts authored by ASRL personnel which were relevant
to  the program objectives.   The project descriptors  for
the  external  contracts  were incomplete  in  some  cases
and  provided  little of the technical information  needed
for  proper evaluation.   Without the scheduled  presenta-
tions  to the panel by ASRL members,  as well as  the  un-
scheduled meetings with opportunities for additional ques-
tions,  it  would have been difficult to evaluate the pro-
gram.  The panel feels that the written material submitted
for advance review should have provided enough information
to  allow a reasonably thorough evaluation  by  individual
panel members prior to the meeting.
     Several  specific instances of coordination  problems
were  apparent.   For example,  field measurements in  the
Materials   Effects  Program   (Task  Group  G)  have  been
proceeding  without accurate aerosol NOj or HNO3  monitor-
ing,  in  spite of development of an appropriate sensor by
ASRL  (This is discussed in section 3.2.1.4.).
      It  would  be especially useful for ASRL  members  to
have  a  better understanding of the  deposition  velocity
parameterizations  being  developed  by  NOAA,  since  the
success of the ASRL efforts depend,  to a great extent, on
the success of these parameterizations.
                           23

-------
2.4.3.  Related Research!  Surrogate Surface Methods
     Despite  the sizeable research efforts at  developing
micrometeorological  methods,  no  funding is provided  to
investigate  surrogate surfaces.   The reason for this  is
probably  historical:   there have been sizeable  expendi-
tures  for  acquiring NADP dustfall bucket data  over  the
past several years,  which are now believed to have little
meaning.  This unproductive use of resources has apparent-
ly prejudiced the feelings of many individuals researching
dry deposition against the use of any surrogate surface.
     However,  there  is a growing body of  evidence  that
suggests  surrogate  surfaces may provide  an  inexpensive
means  of acquiring direct dry deposition data  applicable
to  the  goals  of Task Group  D.   For  example,  several
gaseous  species are removed efficiently by surfaces whose
aerodynamic resistance controls overall dry deposition.  A
surrogate surface which reacts readily with these  species
may provide direct estimates of the fluxes.  An example is
the  use  of  nylon surfaces to absorb HNO^.   As  another
example, deposition of large particles (greatly influenced
by sedimentation) may be easiest to measure with surrogate
surfaces rather than with airborne concentration  measure-
ments;   the latter requires accurate knowledge of the size
distribution  to estimate fluxes,  and sampling such large
particles is difficult.
                           24

-------
     The  panel recommends that research be  initiated  in
surrogate  surface  development.   It should  provide  new
techniques  for  reliably and inexpensively measuring  dry
deposition of some species.
2.4.4  Program Strengths and Weaknesses
     The  primary strengths of the program relate  to  the
competence  of  the personnel (in-house and external  con-
tractors) , and the facilities available for conducting the
work.  The objectives of the  program are reasonably clear
and progress is being made toward achieving the objectives
in a way which will be of use to NAPAP.
     The primary weaknesses of the program concern lack of
coordination in certain respects.  Better communication is
needed  among the external contractors,  and  between  the
contractors and ASRL.  Improved communication between Task
Groups is also needed.

 2.5  Recommendations (In Approximate Order of Priority)
       The recommendations that follow represent issues of
high  priority discussed in detail and in context in  this
portion  of the report.   In some cases,  a single  recom-
mendation  synthesizes individual recommendations made for
several tasks.  In addition to those given below, a number
of significant other recommendations appear in the general
discussion.
     •  Technical interaction between ASRL, EMSL and NOAA/
                           25

-------
ATDL  on  issues related to dry  deposition  moni-
toring  and  the proposed 30-site National  Trends
Network should be increased,  perhaps with regular
semi-annual meetings.

The  capabilities of several  qualified  organiza-
tions  should  be sought to assist development  of
deposition  velocity  estimations.    The  overall
efforts  in this area should be reviewed  periodi-
cally by a panel of experts from the international
scientific community.

Research  in micrometeorological methods  for  dry
deposition  measurement appears to  be  worthwhile
and should continue.

Spatial scales, over which the dry flux determina-
tions and concentration monitoring are representa-
tive, need to be identified.

The accuracy and precision of the ASRL  Concentra-
tion Monitor should be evaluated over the range of
conditions  likely  to  be encountered  in  actual
network operations.

Research  into developing surrogate  surfaces  for
routine monitoring should be initiated.   Particu-
lar  emphasis  should  involve  surrogate  surface
design for assessing fluxes of reactive gases, and
of  large  particles  which  may play  a  role  in
neutralizing depositing acids.

The collection of >10 urn airborne particles by the
ASRL Concentration Monitor is unlikely to be  use-
ful, and should not be pursued.

Additional emphasis needs to be put on determining
the  flux  of alkaline species to determine a  net
acid dry flux.
                     26

-------
                           Ill
         MATERIALS EFFECTS PROGRAM - TASK GROUP G

                     3.1  Background
     The  ASRL  program,  Effect  of  Acid  Deposition  on
Materials,  conducted  for NAPAP,  is primarily  concerned
with those materials of construction deemed to be of broad
economic significance.   They are listed  as:   galvanized
steel,  weathering steel, exterior household paints, main-
tenance  paints,   automotive  finishes,  asphalt  roofing
shingles, and concrete/cement.
     The  Materials Effects Program in ASRL is expected to
develop   damage   functions  for   these   materials   to
differentiate  between the normal expected weathering  and
accelerated weathering attributable to wet and dry deposi-
tion.   These damage functions,  the primary output of the
program,  are to be provided to Task Group I (Assessments)
of NAPAP to permit  comprehensive cost-benefit analysis of
acid deposition controls on materials damage effects.
     Two   related  materials  programs  are  the  Gaseous
Air  Pollutants   (GAP) program of ASRL and  the  materials
effects research on stone and marble deterioration  by the

-------
National  Park  Service (NPS).   While the efforts of  the
former  are to a degree integrated with the  ASRL  effort,
that of the NPS appears not to be.

                    3.2  Task Reviews
3.2.1  Field Studies

     3.2.1.1  Structural Materials-Metals (Task 3024).
Damage  functions  for seven metallic materials are  being
developed  from  exposure  data collected  at  five  field
sites.   The corrosion rates of the panels will be  corre-
lated  with data on weather, air pollutants and rain chem-
istry,  all   of  which are being recorded  simultaneously
with the exposure data.  The seven materials are represen-
tative of the most common metals in the inventory.  In our
opinion,  additional metals need  not be considered at the
present time.
     The  four  sites at which measurements are now  being
made  are typical of rural  (North Carolina and New  York),
surburban   (New Jersey) and urban (District  of  Columbia)
environments  with low to medium concentrations of  sulfur
dioxide.   The  opening  of a fifth site in  Steubenville,
Ohio,  is a much needed extension of the program to  cover
industrial environments with high concentrations of sulfur
dioxide.   It would be desirable to open one more exposure
site  preferrably  downstream  from a  combined  urban/in-
dustrial area.
                           28

-------
     .The  exposure racks at the North Carolina,  New  York
and  New  Jersey  sites are installed in open  areas  with
natural  ground cover.   The weathering racks at the  Dis-
trict' of Columbia and Ohio sites are located on the  roofs
of buildings, subjecting the specimens to the  heat radia-
tion  from  the asphalt roofs and emissions from roof  top
vents.   This creates microenvironments not typical of the
macr©environments,  particularly with regard to the forma-
tion of dew.   The researchers should modify the  exposure
conditions  in a manner that eliminates the  microenviron-
ment  effects.  Microenvironments  are also being  created
below  the transparent covers that shelter the  weathering
racks against rainfall at the North Carolina site.   These
racks should, accordingly, be modified.
     Exposure  times of one month are too short for  mean-
ingful studies of the effect of weather and pollution. The
researchers did well to end these tests.  The three-month,
one-year,  and  three-year exposure times are suitable for
studying  the  seasonal and short-term effects  of  atmos-
pheric constituents.   We understand that longer exposures
may  not  be feasible under the time constraints  of  this
project.  However, we strongly recommend the initiation of
long-term exposures.  Most buildings and bridges last more
than 50 years.  Damage functions cannot be reliably extra-
polated  from 3 to 50 years.   If the number of  specimens
                           29

-------
from the same batch is limited, some replicate three-month
or one-year exposure tests could be converted to long-term
exposure tests.
     The simultaneous acquisition of  corrosion,  weather,
and air pollution data is valuable.  A successful correla-
tion of two sets may yield,  perhaps for the first time in
the United States,  the functions needed to estimate, over
a  large  geographical area the corrosion damage to metals
from  acid deposition.   The researchers  should,  without
delay,  begin  to  correlate the exposure  data  with  the
weather  and  air pollution data,  so as to  identify  the
significant  variables  at an early stage of the  program.
An understanding of the phenomena would allow,  if needed,
adjustments in the experimental program.
     We  were not given an indication of the approach that
will be used to construct damage functions.   As a word of
caution,  we  would  like to emphasize  that  multivariate
regression analysis describes the data within the range of
the  variables tested,  but it is not a substitute  for  a
mathematical  model  founded  on an understanding  of  the
physical-chemical mechanisms.   The weight loss data  must
be  carefully  examined  and interpreted in terms  of  the
weather  and  air  pollution data and the results  of  the
microanalysis of the corrosion products.   The development
of damage functions should begin at once.
                            30

-------
     3.2.1.2   Non-Metallics.  Analogous to  the  task  on
metallic materials, Section 3.2.1.1, the objective of this
task  is to derive damage functions for exterior paints by
conducting  field  exposure studies.   The  intent  is  to
characterize  the paint film damage in terms of erosion of
the film,  its surface chemistry,  and changes in  surface
reflectance.   The  damage will be related to the meteoro-
logical, pollution, and rain chemistry data that are being
simultaneously monitored.
     The approach raises questions with regard to the time
of exposure,  type of substrate, and development of damage
functions.   Three years may not be sufficient to signifi-
cantly damage the paint film, especially when that coating
is  applied on a passive substrate such as  the  stainless
steel panels being used in this study.   For example,  the
service  life of paint coatings on highway bridges  fabri-
cated from carbon structural steel varies from about eight
years  in  the moist and chloride-contaminated  microenvi-
ronments that occur in the Northeast and Midwest to  about
30  years  in  the dry environments  in  the  southwestern
states.   The researchers face,  in this  task,  a problem
similar to that in the field study of metals,  namely,  of
having  to  derive damage functions from short-term  expo-
sures  and extrapolating the results to long-term  service
life.  Longer exposures are needed.
                           31

-------
     Our  second comment relates to the type of  substrate
on  which the paint coating is applied.   In our  opinion,
the  paints  should  be  tested  in  combination  with   a
realistic substrate.   We agree with the Paint  Workshop's
recommendation  to test a paint system on hot-rolled steel
substrate.   Advice on the choice of a suitable paint  for
hot-rolled  steel should be sought from specialists,  such
as the staff of the Michigan Department of Transportation,
Lansing,  Michigan,  who are the leaders in the qualifica-
tion of  commercial paint systems.   The proposed tests of
latex paint applied on western red cedar are discussed  in
Section 3.2.3.1.
     The  third  comment  relates to  the  measurement  of
damage.  As noted in the Paint Workshop's recommendations,
failure is to be assessed visually in terms of blistering,
lack of adhesion, substrate rusting, and rust spots on the
film.   We  foresee  difficulties in relating such  visual
observations to weather and air pollution data,  and quan-
tifying  them in damage functions for a  given  paint/sub-
strate system.
     We  recognize that the paint industry may  tradition-
ally have exposed samples at a 90°angle from the  horizon-
tal.   However,  the  choice of the vertical exposure  for
testing  the  paints in this task,  as compared to the 30°
exposure  of the bare metallic samples  (Section  3.2.1.1),
                            32

-------
introduces one additional variable,  without providing any
advantages.   Furthermore, a slanted surface may be better
for studying the effect of acid deposition.
     We  recommend  that the researchers carefully  review
their  current plan for testing and analysis  before  they
continue with the field studies.

     3.2.1.3   Acid Rain Simulator.   The objective of the
study is to quantify separately the effects of wet and dry
acid deposition on steels and household paints exposed  to
ambient  weathering.   Three conditions are being studied:
(1)  dry only;  (2) dry and ambient wet  (4.5 pH average  at
RTF  for  rain);  and  (3) dry and controlled wet   (5.6  pH
spray).
     It  is well known that rain washing of pollutants off
the  surfaces  of weathering steel  structures  helps  the
steel  develop a protective oxide coating.   Based on this
experience,  one would,  for example,  expect higher  cor-
rosion  rates  for  weathering steel exposed to  the  "dry
only"  condition, and lower corrosion  rates when skyward-
facing  surfaces are rinsed by the rain or the  controlled
spray.   The  other materials being tested may behave in a
similar  manner.   Dry deposition will likely be the  con-
trolling  factor,  with  rinsing  being  beneficial.   The
effect  of the average 1.1 pH unit difference between  the
rain and the spray water could be masked by the  generally
                           33

-------
beneficial  effect of rinsing.   The expected results  may
lead  to the incorrect conclusion that the dry  deposition
is damaging to the materials,  the wet deposition is bene-
ficial, and the pH of the rain is not a factor.
     The  interpretation of the results needs to  consider
conditions not modeled by the experiments.   For  example,
rain  water  collecting on horizontal and  poorly  drained
surfaces  in actual structures can increase the effect  of
acid deposition.
     This  experiment  is  elaborate  and  expensive.   We
recommend not placing a simulator at the Steubenville site
until the researchers have fully characterized the results
from  the RTF site and understood the significance of  the
findings.   Snowfall  adds  an uncertainty to the  planned
exposures  at the Steubenville site that needs to  be  re-
solved.   The  pH  of  the rinsing water at the  RTF  site
should be periodically checked.
     3.2.1.4 Site Monitoring.  Air quality and meteorology
are  being  monitored at all five of the  exposure  sites.
Local  agencies  perform the measurements at three of  the
sites,  with the primary purpose of the measurements being
other  than the materials effects program.   ASRL is  sup-
porting  the measurements at the Newcomb,  New  York,  and
Research Triangle Park, North Carolina, sites.
     Each  site measures S02»  NOX,  and O3 continuously.
                            34

-------
Dichotomous  samplers,   specially  modified  to   collect
samples  over a one-week period,  collect fine and  coarse
fractions of PM-10 participate matter.   These filters are
analyzed for  mass, elements  (by XRF), and 804 and NO3 (by
ion  chromatography).   Total  suspended participates  are
also  measured similarly.   Monthly-averaged rain  samples
are analyzed for pH and various ions.   Meteorology  (wind
speed and direction,  temperature, relative humidity, pre-
cipitation  rate,  and solar  radiation) are measured  con-
tinuously.   It  is  expected that all of these  variables
play  a role in the deposition of acidic material  to  the
exposed samples.
     The  monitoring  sites lack the  ability  to  measure
gaseous nitric acid and the nitrate analyses of the dicho-
tomous samples do not properly reflect particulate nitrate
because  of volatilization of some of the material on  the
filters.   We  recommend  adding a  sampling  method  that
properly characterizes NO^ and HNC>3 at all sites; the ASRL
Concentration Monitor would serve this need.
     Relative  humidity  is measured by  hair  hygrographs
that  are calibrated infrequently.   Such instruments  are
known to easily change calibration.   We recommend,  as   a
minimum,  that their calibration be checked weekly by com-
parison  with  a psychrometer.   Installation  of  a  more
reliable  humidity  recording system would be  preferable.
Although  we  recognize that  such systems  are  expensive,
                            35

-------
they are compatible with electronic recording of the data,
which  would  reduce the costs of  handling  and  reducing
recording charts.
     Quality   assurance  procedures  for  the  aerometric
monitoring  at the sites appear to be uneven.   The  sites
operated  by local agencies comply with EPA QA  procedures
for monitoring,  which are focused at concentration levels
near  ambient  air quality  standards.   The  systems  and
performance  audits  that will be performed for the  sites
are desirable, and the planned calibrations at low concen-
tration  levels are absolutely necessary.   We  hope  that
these calibrations will be repeated at regular intervals.
     3.2.1.5  Aerometric Data Base  The compilation of the
aerometric data into a data base that is suited for analy-
sis  is absolutely necessary for the success of the  expo-
sure  program.   The aerometric data base management  task
appears to be accomplishing this successfully.
     However,  the data base would be most effective if it
were  designed for use with the data  analysis  procedures
planned  for the program.   We did not see that a plan for
data  analysis had been formulated,  or that the data base
design had been adapted to such a plan.  This is a serious
shortcoming of the materials exposure program that is
discussed in greater detail in Section 3.3.3 below.
                           36

-------
     There  did not appear to be a procedure for  applying
calibration  factors to datar  which may become necessary.
A  desirable  feature of the data base is the  ability  to
flag  data that deserve special attention because of  poor
quality,  need for special calibration,  or other factors.
We hope that this feature will be utilized.

3.2.2  Chamber Studies
     The panel strongly supports the chamber study effort.
Host significant is the growing interest in  characteriza-
tion of corrosion layers,  mass balances, and reaction ki-
netics. This represents a marked strengthening of the ASRL
materials programs which in the past relied too heavily on
regression  analysis  of simple weight loss data.   It  is
important to recognize that the determination of  reaction
mechanisms  is essential to the development of  mitigative
strategies and to the assignment,  with confidence, of the
proportions  of damage attributable to wet and dry deposi-
tion.   It does not appear that the group has expertise in
corrosion mechanisms,  or has sought the advice of  appro-
priate experts.   An understanding of corrosion mechanisms
will be essential,  not only to interpretation of the data
obtained  in  chamber  studies,  but also for  the  inter-
pretation of field exposure data.  For example, the rever-
sal  in  corrosion rates observed for Cor-Ten  A  and  110
copper  exposed  at the Washington,  DC and RTF  sites  is
                           37

-------
possibly  due to microenvironmental influences on reaction
kinetics and mechanisms.  Yet, no attempt has been made to
explain these data from a mechanistic viewpoint.
3.2.3  New Projects fFY 86)
     Four  new  projects  with  total  support  level   of
$500,000   were  described  briefly.    Two  involved  the
extension  of  materials  effects studies  to  new  areas,
namely concrete and masonry structures and asphalt roofing
shingles.   A  third  proposed an expansion  of  the  non-
metallic  study  to investigate initial effects  on  paint
films   on  wood  substrates.    The  fourth   encompassed
placement  of corrosion rate monitors developed under  the
GAP program at the five field exposure sites.
     3.2.3.1 Initial Effects on Painted Wood Substrates.
The  proposed  extension of the painted surface  study  to
include wood substrates, chamber studies, initial effects,
and the mechanism of paint film failure is based, in part,
on the recommendations of the recent workshop on  exterior
paints held at RTF.   This use of industry expertise in  a
cost-effective manner is to be commended.
     Only  a brief outline was given of the proposed work.
The  panel is concerned that the substantial  complexities
introduced  by  the  new  substrate  (wood)  will   unduly
complicate  the chamber study of paint films.   Similarly,
the  experimental  designs for the FTIR  and  thermometric
                           38

-------
studies are not yet defined.   It is recommended that  the
initial  effort be directed toward the development of  ex-
perimental protocols and test methods.   Field exposure of
paint films on wood substrates is considered premature.
     Consideration   should  be  given  to  the   workshop
observations that: (1) historical field exposure data  may
provide the best means of determining actual service life,
and  (2)  damage measurement techniques  should  correlate
with painting frequency.
     3.2.3.2  Effects on Concretg^and Masonry Structures.
The  extensive  use  and  wide  distribution  of  concrete
structures suggests that this material be investigated for
possible acid deposition effects.  However, the well-known
role of chloride ion in the decay of reinforcing steel and
the  strongly  alkaline  nature  of  the  concrete  matrix
suggests  that  a relatively small proportion of  observed
damage  may  be associated with  acid  deposition.   Since
substantial research and use experience are available,  it
is  recommended  that a workshop be conducted  at  RTF  to
review  the state of concrete structure use experience and
damage mechanism research prior to initiation of the  lab-
oratory  study (Phase I),   or the field survey  involving
core sampling of existing structures  (Phase II).

     3.2.3.3  Effects on Roofing Asphalt Shingles. Asphalt
roofing   shingles  complete  the  list  of   construction
                            39

-------
materials of economic significance under consideration  by
EPA.   Some  sense  of  service life  cycles  and  failure
mechanisms is required before initiation of laboratory and
field investigations.  It appears appropriate to limit the
effort for  FY 86 to the initial literature survey  (Phase
I).
     3.2.3.4 Evaluation of Corrosion Monitor at Exposure
Sites.  The panel was informed of the proposal to evaluate
corrosion  monitors  developed under the  GAP  program  at
materials exposure sites.   As noted in the 1984 review of
the  GAP  Materials Damage Program,  the response  of  the
corrosion monitor to atmospheric pollution and raeterologi-
cal  variables  is not yet fully understood.   It  is  not
obvious that this device will prove useful in the proposed
task  of  comparing  short-term effects  for  the  diverse
materials, e.g., steel, concrete, limestone, marble, paint
films,  and  asphalt,  either in place or eventually to be
placed at the field exposure sites.

                  3.3  general Comments
3.3.1  Management Team/Project Staff
     The   management   and   project   staff   team   has
demonstrated initiative, industry, and the ability to work
together  effectively.   They have  made good use  of  the
workshop  on  paint  damages  to  obtain  guidance  in  an
unfamiliar area.
                           40

-------
     The 1983 review panel for this project observed  that
•the  smog  chamber studies may soon lead to questions  of
corrosion mechanisms,  a field with many experts."  It  is
disappointing that no effort has been made to follow their
recommendation that contact soon be made with such experts
to  obtain  guidance  and  to assure the  success  of  the
studies.

3.3.2  Integration of Tasks to Achieve Program  Objectives
       The  stated  goal  of this program is  to  "develop
materials  damage functions that partition effects of  wet
and dry deposition for use by Task Group I, Assessment, to
develop  a cost benefit model."  The panel  believes  that
inadequate  attention  has been paid to the  uncertainties
associated  with the use of damage functions  derived from
short-term  field  exposure  or  chamber  studies  in  the
development of the cost benefit model (Figure 2) that will
be  used to conduct an economic assessment of acid deposi-
tion  damage  and to prepare an  acid  deposition  control
strategy.   This issue is documented in the following sec-
tion.

3.3.3  Relationship Between Field Studies, Chamber Studies
       and Damage Functions
     For  results  from controlled chamber studies  to  be
related  to  field  studies  and  interpreted  as   damage
                          41

-------
                    COST BENEFIT MODEL

MATERIALS
DAMAGE
FUNCTION
—+

COST
FUNCTIONS
	 »-


MATERIALS
INVENTORY
t
—
ECONOMIC
ASSESSMENT
CONTROL
STRATEGY
• FIELD EXPOSURE  PROGRAM

• CHAMBER EXPOSURE PROGRAM
       Figure 2.   Task Group G  Effects on  Materials
                   and Cultural Resources
                          42

-------
functions, physical-chemical mechanisms must be  proposed.
The -models  are tested statistically, and  best  fits  of
coefficients determined.  The hypothesis is well developed
before  the  use of statistical methods.   The variety  of
possible  mechanisms  and  environmental  conditions  make
engineering approximations appropriate for the simplifica-
tion of the damage functions, but experimental design roust
be  more  than  empirical  measurements  with   regression
relationships.  Otherwise, the response is known only over
the range of test conditions and cannot be extropolated to
new  environmental conditions.
       Once   a  model  is  proposed   and   tested,   the
uncertainty in each bit of data must be carried through to
the   final  step  of  the  determination  of  the  damage
function.   The  propagation of errors must be  determined
for  the  final results.   The  standard  deviation  among
replicate  samples is not adequate for the  interpretation
of overall uncertainty.
     In  the  chamber studies,  concentrations  have  been
measured  for  air species and metal and  anion  corrosion
products. A mass balance approach is being considered.  In
practice,   the  small  differences  in  relatively  large
numbers  makes  the comparison of material lost with  cor-
rosion  products leached impractical.   Once the  products
are collected, linear relationships are considered between
                           43

-------
product species.   Non-linear relationships (e.g., pH con-
trols  of  solubility) have not been  considered/have  not
been  used,  and measurements of carbonates  (a  corrosion
anion product) have not been made.   Approximate linearity
may  exist  between many species,  making it difficult  to
determine the mechanism or cause of damage.   This can  be
partially  alleviated by the use of more  controls.   How-
ever,  it  is doubtful synergistic effects will be  under-
stood  without an experimental design based on the test of
specific physical-chemical mechanisms.  Even the transport
mechanism,  such as thermophoretic effects in the dew con-
densation system, as  compared to the real world, should be
reconsidered.
     With  the  limitations  of the  understanding  of  the
chamber  study  results,   the interpretation of the  field
studies  becomes phenomenological.   The damage  functions
for  metals could be determined from the short-term chamber
study  results.   Expansion  of the testing  in the  chamber
studies  is needed to  generate damage functions.
      In  the  environmental conditions with  the most  rapid
corrosion,    the  surface  resistance  to   the  deposition
velocity should be determined.   This parameter  could  be
related  to the  field  studies,  even with variable meteoro-
logical  conditions.
     The main concern in  the link between  chamber studies,
field  experiments, and damage  functions  is  that the  damage
                            44

-------
is  a long-term process,  yet the best data for the damage
function   will   be   based   on   short-term    studies.
Considerations of long-term mechanisms and failures become
crucial.  Long-term studies are needed, either under NAPAP
or air quality funding sources.

3.3.4  Related Research
     The visit to the RTP field exposure site demonstrated
the  substantial  commitment of the National Park  Service
(NPS) to field exposure studies for marble and  limestone.
These  materials should be considered for parallel chamber
studies  under  the  direction of  the  Materials  Effects
Program  under  review  since  the  NPS  has  neither  the
resources   (smog chamber and associated hardware) nor  the
expertise necessary to conduct such work.

3.3.5  Program Strengths
     The selected experimental designs are being performed
well.   The  chamber  studies are well  controlled.   ASTM
standard  exposure procedures are followed.   The  results
are reproducible.   The right metals have been chosen  for
study.  The simultaneous measurement of air quality in the
field  appears excellent,  with the exception of the omis-
sion of HN(>3.

3.3.6  Program Limitations
     The  materials effects program suffers from  the  in-
herent  limitation of studying long-term processes over  a
                           45

-------
relatively short time.   This is recognized and may not be
avoidable.
     There  is some question  of  representativeness.
Standard methods  have been used for reproducibility,  but
there is a concern that precipitation acidity effects will
be    largest    on   materials    which    collect/retain
precipitation.   The  orientation of test specimens  mini-
mizes contact with wet precipitation and its effects.
     In  particular,  snow  has not been  addressed.   Its
importance  to  material damage in terms of snow  melt  is
unknown, but deserves consideration.
     The  major limitation is in the  characterization  of
the  mechanism  of  corrosion.   While overall  rates  are
empirically  determined,  the mechanism is still  unknown.
With  the high covariance of the depositions  of  chemical
species, the agent of corrosion is obscured.
     An   overall  plan  is  needed  to  determine  damage
functions from the field and chamber measurements.

3.3.7  Quality Assurance and Quality Control
     Quality assurance and quality control are an integral
part of all environmental research.   The researchers have
a  healthy  appreciation of the  limitations  of  specific
analytical procedures.  They have made commendable efforts
at  upgrading  the QA for continuous monitors used in  the
                           46

-------
field and laboratory studies.
     A few limitations were noted.  The metal measurements
need  to  be  confirmed by using split  samples,  and  the
method  of  standard  additions  to  the  liquid  samples.
Hydrogen peroxide is currently not considered in the  cor-
rosion  modeling  of condensation on metals for lack of  a
method of good quality assurance.  This needs to be recti-
fied  to provide a measure of oxidants.   Samples are  not
routinely  subjected  to  a  complete  anion  and   cation
balance,  a standard QA technique which also provides use-
ful   information  about  acidity.    Similarly,   aerosol
measurements  do  not determine ammonium  or  acidity  and
instead  focus on nitrates, sulfates, and elements by XRF.
All chemical determination techniques should be tested for
quality  assurance  by either cross  calibration  with  an
alternate  technique,   or  internal  consistency  between
measurements with respect to mass, charge, or conductivity
for a percentage of the samples.
3.4  Recommendations  fin Approximate Order of Priority!
       The recommendations that follow represent issues of
high  priority  discussed in detail and in context in  the
body  of  this portion of the report.  In  some  cases,  a
single  recommendation synthesizes individual  recommenda-
tions made for several tasks.   In addition to those given
below,  a  number  of  significant  other  recommendations
                          47

-------
appear in the general discussion.

     •  Study  of  physical-chemical mechanisms  for  cor-
        rosion  should be incorporated into  the  research
        design,   and   carried  through  to  the   damage
        functions.

     •  A  plan needs to be developed for the analysis  of
        exposure data,  and  the  development  of   damage
        functions.

     •  The   aerometric   monitoring  at  the   materials
        exposure  sites  should be  augmented  to  include
        nitric  acid and proper measurement of particulate
        nitrate.   The  ASRL Concentration  Monitor  could
        satisfy this need.

     •  A series of workshops is recommended for obtaining
        guidance  for  the proposed new efforts  in  paint
        films   on   wood  substrates,   asphalt   roofing
        shingles,  and concrete before field exposure  and
        laboratory studies are initiated.

     •  A  modification  of the field exposure program  to
        extend measurements to long-term exposures  should
        be  initiated.   Placement  of a second acid  rain
        simulator  at  the  Steubenville  site  should  be
        delayed  until the results from the RTF  prototype
        are characterized and understood.

     •  Propagation  of  errors should be carried  through
        the testing.

     •  Effects  of  snow and irregular  surface  geometry
        should be considered.
                           48

-------
                APPENDIX A

AGENDA - PEER REVIEW OF THE ACID DEPOSITION,
DRY DEPOSITION, AND MATERIAL DAMAGE PROGRAM

-------
                                        AGENDA
                    Peer Review of the Acid Deposition Dry Deposition
                               And Materials Damage Program
                                   April 23 - 25, 1985
      TIME
              TOPIC
  SPEAKER
Tuesday, April 23
   8:00 - 8:15
   8:15 - 8:30 a.m.

   8:30 - 8:35 a.m.
   8:35 - 8:45 a.m.
   8:45 - 9:20 a.m.

   9:20 - 10:15 a.m.
  10:15 - 10:30 a.m.
  10:30 - 11:00 a.m.
  11:00 - 11:30 a.m.
  11:30 -  1:00 p.m.
   1:00 -  1:15 p.m.
   1:15
   2:00
   2:30
   3:00
2:00 p.m.
2:30 p.m.
3:00 p.m.
3:15 p.m.
(Raleigh Inn - Raleigh, NC)
Opening Session
Welcome
Peer Review Program Orientation
 ASRL Peer Review Coordinator
(Closed Session)
Coffee and Donuts in Meeting Room
DRY DEPOSITION MEASUREMENT PROGRAM
Session Called to Order
Dry Deposition Measurement Overview
Direct Dry Deposition Methods
  Extramural Projects
  In-house Project
Concentration Monitor Projects
BREAK
Core Sites and Deposition Velocity
 Parameterization
Questions and Discussion
LUNCH
Overview of Materials  Effect
 Research Program
 MATERIALS FIELD EXPOSURE PROGRAM
Structural Materials
Deterioration of Non-Metal lies
Acid Rain Simulator
BREAK
                                                                           C.  Coley
                                                                           R.  Patterson
R. Patterson
K. Knapp
T. Ellestad
R. Speer
K. Knapp
T. Ellestad
J. Spence

D. Flinn, BOM
J. Spence
E. Edney, NSI
                                        49

-------
                                          - 2 -
      TIME
                                          TOPIC
  SPEAKER
Tuesday.  April 23 (cont.)
   3:15 - 3:45 p.m.
   3:45 - 4:15 p.m.
   4:15 - 5:00 p.m.
   5:00 p.m. - Until
Wednesday, April 24. 1985
   8:30 -  9:30 a.m.
   9:30 - 10:00 a.m.

  10:00 - 11:00 a.m.
  11:00 - 12:00 p.m.
  12:00 -  1:30 p.m.
                            Site Monitoring
                            Data Base Management
                            Questions & Discussion
                            Reviewer Debriefing with Dr. Alfred H.
                            Ellison, Director, ASRL (M-303)
                            (Research Triangle Park, NC and
                            Raleigh Inn, Raleigh, NC)
                            Materials Chamber Exposure Program
                            Chemistry of Acid Precipitation
                            Effects on Materials (EPA: M-303)
                            New Projects
                             Site Visit - Materials Exposure
                            Visit Exposure Chambers
                               Site Visit - Dry Deposition
                            Measurement Methods (ERC Annex)
                            LUNCH  (NIEHS)
                            National Institute of Environmental
                            Health Sciences Cafeteria
   1:30  - 2:30 p.m.
   2:30 -
                            Telephone Call to Bruce Hicks
                            Return to Hotel
                            Executive Session
Thursday,  April  25, 1985  (Research Triangle Park)
   8:30 a.m.  -              Executive Session (EPA/D-102)
J. Spence
J. Graham, CSC

Closed Session
E. Edney
J. Spence
E. Edney
D. Stiles, NSI
K. Knapp
                                                                           Closed Session
                                                                           Closed Session
                                       50

-------
        APPENDIX B



EPA-ASRL PEER REVIEW PANEL

-------
                           REVIEW PANEL
               ACID RAIN (TASK GROUPS D & G) PROGRAM
                        April 23 - 25, 1985
Name:

Work Address:



Area(s) of expertise:
Name:
Work Address:
Area(s) of expertise:

Name:


Work Address:


Area(s) of expertise:



Name:

Work Adress:
Area(s) of expertise:
Name:

Work Address:



Area(s) of expertise:
Norbert S. Baer, Chairman
Professor of Conservation
New York University
14 East 78th Street
New York, New York  10021
212-772-5846
Materials damage effects, preservation of
cultural property

Pedro Albrecht
Department of Civil Engineering
University of Maryland
College Park, Maryland  20742
301-454-5228
Atmospheric corrosion of structural  steels

Cliff I. Davidson
Associate Professor of Civil Engineering and
Public Policy
Carnegie-Mellon University
Pittsburgh, Pennsylvania  15213
412-578-2951
Aerosol deposition, trace elements
characterization and human exposure,
historical air pollution trends

Howard M. Liljestrand
Assistant Professor
Civil Engineering, Environmental Engineering
Group 8.6 ECJ
University of Texas
Austin, Texas  78712-1076
512-471-5602
Acid-base deposition mechanism-tracers of
acid/bases and receptor models, wet & dry
deposition mechanism of aerosols and gases,
reactions upon deposition to materials, box
models of the atmosphere, kinetic models
related to aerosol and acid producation.

Ivar H. Tombach
Vice President, Environmental Programs
Aero Vironment, Inc.
825 Myrtle Avenue
Monrovia, California  91016
818-357-9983
Visibility, aerosol physics, fluid mechanics
                             51

-------
        APPENDIX C



PROCESS EVALUATION REPORTS

-------
            ATMOSPHERIC SCIENCES RESEARCH LABORATORY
                           PEER REVIEW
                Process Evaluation Results for the
           Dry  Deposition and Materials Effects Program
     The Atmospheric  Sciences Research Laboratory (ASRL) of the
U.S. Environmental Protection Agency convened a panel of scientific
experts  on  April 23-25, 1985,  to  review the Dry Deposition and
Material  Effects  Program.  The panel consisted of five scientists.
These reviewers were asked  to  evaluate the process involved in
preparing and  implementing this review.  This report  presents  their
opinions  of  the  process for this  specific meeting.
     The  evaluation  instrument was designed to assess the following
aspects  of  the process:   (1)  Preview Materials;  (2) Process and
Logistical  Information; and (3) the Review Meeting.  A section was
also provided for reviewers to  give their comments  and recommenda-
tions.   The  reviewers were  instructed to respond to 15  items by
circling numbers from  1 to  5  (with  1 representing poor; 2-fair;
3-good;  4-very good; and 5-excellent).
     Only three  of the five panel members have submitted a  process
evaluation  form.  Overall,  the reviewers' evaluation of the peer
review process was positive.  Most of  the areas in the instrument
generated  "very good"  and  "good"  ratings.  Only two items were
rated fair.   One panel member  rated  the utility of the  preview
materials  for outside reviewers as fair, commenting that they were
incomplete.   The adequacy  of  time  for review panel executive
                           52

-------
sessions  also received one fair rating  because the commenting  panel
member felt more  time was needed.  None  of the reviewers rated any
item  as  "poor".  Table  1  presents a summary  of  the reviewers'
ratings, as well  as their specific comments  and recommendations.
                            53

-------
 Preview Materials
                                  Table  1
Review Cateqories

1
I Number of Reviewers
I Rating Each Item
1 Very
I Poor Fair Good Good Excellent
  1.  Written Qua'ity
  2.  Technical  Quality
  3.  Utility for Outside Reviewer
  4.  Adequacy of Time Available
      to Preview
      1
      2
1*

1     4
 Process and Logistical  Information
  5.   Meeting Purpose
  6.   Scheduling of  Meeting:
      Agenda/Format
  7.   Reviewer Responsibilities:
      Time/Preparation Requirement
  8.   Overall  Peer Review Process
  9.   Timeliness of  Meeting
      Notification
 10.   Timeliness of  Logistical
      Information
      1     1
      1     1
      1     1
      1     1

      3
      1     1
Review Meeting
11.  Adequacy of Time for Discussion
     With EPA Staff
12.  Adequacy of Time for Executive
     Session
13.  Quality and Utility of
     Presentations
14.  Quality and Utility of Materials
     Disseminated
15.  Support Services and Activities
1"         2
      1     2
   TOTALS
      17     17
*     "incomplete"
**    "need more time"
***   "very good support by Research and Evaluation
       Associates Staff"
                            54

-------
                       Reviewers' Comments
"The  lion's share of the  useful review materials were distributed
after our arrival  at  RTP, or during (or after) presentations.  This
made  preparation of a  first  draft review report (as suggested in
the introductions)  impossible."

 "The support by the Research and Evaluation Associates staff was
excellent.   The hotel facilities were poorly managed  by the hotel.
The lodging was OK, though."

"The  time  needed  to  prepare a coherent report is somewhat more than
the time we were given,  even though most of  us  gave  up  a  good
portion of the night's sleep to get this job done."

"The  only weakness  in  the review process  is the time constraint on
Review Panel Discussion  as  the report is written.   That a highly
responsible committee averaged 4-5 hours sleep, essentially skipped
dinner (24), skipped breakfast and lunch (25)  during the period
Wednesday (24) 3 PM thru  departure, Thursday  (25) 3 PM and  still
had but little time to discuss the completed draft report indicated
to this reviewer that the schedule is a bit tight."
                     Reviewers' Recommendations
"I would  not eliminate  any aspect of the  review process but  would
suggest that a later  departure (session closing  time)  be scheduled
for the  last day.   I  think we all agreed  that one wants the best
posible draft submitted at  the time of adjustment,  rather than  a
weak draft to be upgraded later."
                              55

-------
             APPENDIX D




ASRL RESPONSES TO REVIEWERS' COMMENTS

-------

                    UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
                            ATMOSPHERIC SCIENCES RESEARCH Ufr.tUiRATORY
                                    RESEARCH TRIANGLE PArtK
                                     NORTH CAROLINA 27711
MEMORANDUM

DATE:     July 8, 1985

SUBJECT:  Response to Peer Review Garments on Materials Effect Program

FROM:     John Spence, Project Manager

TO:
          Ron Patterson
          Technical Planning and Review Office
THRU:     William Wilson
          Acting Director, EMCD
1.
 2.
3.
       Reviewers' Comment

    "The researchers should modify
     the exposure conditions in a
     manner that eliminates the
     microenvironment effects.
     Microenvironments are also
     being created below the trans-
     parent covers that shelter the
     weathering racks against rain-
     fall at the North Carolina
     site.  These racks should,
     accordingly, be modified."
     (p. 29)
     "We understand that longer
      exposures may not be feasible
      under the time constraints of
      this project.  However, we
      strongly recommend the initi-
      ation of long-term exposures.1
      (p. 29)

    "The researchers should, with-
     out delay, begin to correlate
     the exposure data with the
     ASRL Investigators Response

Meteorological instrumentation have
been installed adjacent to the material
exposure racks at the DC and Ohio sites
in an attempt to measure the microen-
vironment associated on the roof.  Air
sampling is also taken at roof top at
both sites.  In addition to these
measurements, a corrosion monitoring
device (developed under GAP), which
is an excellent time-of-wetness sensor,
is proposed for installation at all
material exposure sites in FY86.

The transparent covers installed by
BOM are considered experimental
devices that will likely be modified
during the study.  The BOM has developed
the covering device to insure consistent
corrosion of samples under the device.

Under the NAPAP, the maximum exposure
time of materials is likely 3 to 5
years.  Longer exposure periods (10-
20 years) are not possible within the
existing NAPAP.
Vfe agree.  Correlation studies are
planned in FY86.  The main problem
which could delay these studies is
                                    56

-------
 weather and air pollution
 data,  so as to identify the
 significant variables at an
 early  stage of the program.
 An understanding of the
 phenomena would allow, if
 needed, adjustments in the
 experimental program."
 (p. 30)

"We were not given an indica-
 tion of the approach that will
 be used to construct damage
 functions.  As a word of
 caution, we would like to
 emphasize that multivariate
 regression analysis describes
 the data within the range of
 the variables tested, but is
 is not a substitute for a
 mathematical model founded
 on an  understanding of the
 physical-chemical mechanisms.
 (p. 30)

"The researchers face, in this
 task,  a problem similar to
 that in the field study of
 metals, namely, of having to
 derive damage functions from
 short-term exposures and ex-
 trapolating the results to
 long-term service life.
 Longer exposures are needed."
 (p. 31)

"Advice on the choice of a
 suitable paint for hot-rolled
 steel  should be sought fran
 specialists, such as the staff
 of the Michigan Department of
 Transportation, Lansing,
 Michigan, who are the leaders
 in the qualification of ccm-
 mercial paint systems.  The
 proposed tests of latex paint
 applied on western red cedar
 are discussed in Section
 3.2.3.1." (p. 32)
the computerization and validation
of the aeronetric data base.
Chamber exposures of materials and the
analysis corrosion products of field
exposed materials will provide insight
into the physical-chemical mechanisms
of degradation.  This information will
support the development of damage
functions via multivariate regression
analysis of field exposure data.
In the non-metallics program, exposures
beyond 3-years are planned.  This was
apparently missed in the review.
We agree.  Test panels of an alkyd on
hot-rolled steel are being prepared
by the Steel Structures Painting Coun-
cil for exposure at the materials
sites in FY86.  This study will be co-
ordinated with other exposure programs
of the alkyd paint that the Steel
Structures Painting Council is spon-
soring.
                                    An Interagency Agreement (IAG)  with
                                    the Forest Products Laboratory is
                                57

-------
"We foresee difficulties  in
 relating such visual ob-
 servations to weather and
 air pollution data, and
 quantifying them in damage
 functions for a given paint/
 substrate system." (p. 32)
"We recognize that the paint
 industry may traditionally
 have exposed samples at a 90°
 angle from the horizontal.
 However, the choice of the
 vertical exposure for testing
 the paints in this task, as
 compared to the 30° exposure
 of the bare metallic samples
 (Sect ion 3.2.1.1), introduces
 one additional variable, with-
 out providing any advantages.
 Furthermore, a slanted surface
 may be better for studying
 the effect of acid deposition.'
 (p. 32)

"This experiment is elaborate
 and expensive.  We recommend
 not placing a simulator at the
 Steubenville site until the
 researchers have fully charac-
 terized the results from the
 RTF site and understood the
 significance of the findings.
 Snowfall adds an uncertainty
 to the planned exposures at
 the Steubenville site that
 needs to be resolved.  The
 pH of the rinsing water at
 the KTP site should be peri-
 odically checked."  (p. 34)
planned for late FY85.  This IAG will
involve laboratory testing of latex
paints applied on western red cedar.

The main problem with visual inspec-
tion of the field exposure alkyd
paints is the time it takes to see
film failure (substrate rusting).
The visual observations using color
photographic standards according to
ASTM procedures will be used.  In
conjunction with the field exposures,
chamber exposures of the alkyd paints
are planned for FY86 to assess early
film failure under controlled pollu-
tant environments.  This study will
investigate the use of thermography
and other techniques to detect early
film failure.

All of the painted panels are exposed
at 90° in accordance with ASTM D
1006-73.  If painted wood substrates
are incorporated into the field expo-
sure program, panels will likely be
exposed at the vertical and horizontal
positions.
The placement of a mobile covering de-
vice at the Steubenville site is
scheduled for FY85 and should not be
delayed.  The decision to place mobile
covering devices with simulated rain
at the RIP and Steubenville sites is
based on a two-year study of metals
and coatings using a covering device
at the KTP exposure site.  The Steuben-
ville site has considerably higher
SC>2 levels than the RIP site.  The
covering devices at both locations
will be used to separate the effects
of dry deposition (802) and clean rain
(pH 5.6) on materials degradation.
                                58

-------
10.
11,
12.
13.
14.
"The monitoring sites lack the
 ability to measure gaseous
 nitric acid and the nitrate
 analyses of the dichotoroous
 samples do not properly re-
 flect particulate nitrate
 because of volatilization of
 seme of the material on the
 filters.  We reommend adding
 a sampling method that pro-
 perly characterizes NOo and
 HN03 at all sites; the ASRL
 Concentration Monitor would
 serve this need."  (p. 35)

"Relative humidity is measured
 by hair hygrographs that are
 calibrated infrequently.  Such
 instruments are known to
 easily change calibration.
 We reccnmend, as a minimum,
 that their calibration be
 checked weekly by comparison
 with a psychrometer." (p. 35)

"The systems and performance
 audits that will be performed
 for the sites are desirable,
 and the planned calibrations
 at low concentration levels
 are absolutely necessary."
 (?. 36)

"We did not see that a plan
 for data analysis had been
 formulated, or that the data
 base design had been adapted
 to such a plan,"  (p. 36)
This information is needed for the
FY87 and FY89 assessment reports.

We agree.  A request has been made  to
Task Group D to consider placement  of
dry deposition monitors including mea-
surement of nitric acid, at each of
the materials exposure sites in FY86.
Only the Newcomb site uses the hair
hygrograph to measure relative humidity.
Plans are to replace this unit in FY85.
"It does not appear  that  the
 group has expertise in
 corrosion mechanisms, or
 has sought the advice of
 appropriate experts."  (p. 37)
In addition to the regular calibra-
tions and audits perfomed at the
sites, a complete system's and per-
formance audit of air monitoring and
meteorological instrumentation is
planned at least once a year.
The aerometric data base is being com-
puterized in standard SAPQAD format
for use by various organizations, NFS,
USGS, ANL, BOM, EPA, etc. who are
involved in exposure of materials at
the sites.  Each of these organizations
will use this data base to develop
damage functions.  This project will
not develop a plan for analysis of
their materials damage data.

The review panel fails to acknowledge
the corrosion expertise of F. Haynie,
who has published extensively on pol-
lutant induced corrosion, and D. Flinn,
S. Kramer and J. Carter at the BOM,
                                    59

-------
15.   "Yet, no attempt has been
      made to explain these data
      from a mechanistic view-
      point."  (p. 38)
16.   "It is recommended that the
      initial effort be directed
      toward the development of
      experimental protocols and
      test methods.  Field expo-
      sure of paint films on wood
      substrates is considered
      premature."  (p. 39)

17.   "Since substantial research and
      use experience are available,
      it is reconmended that a work-
      shop be conducted at KIP to
      review the state of concrete
      structure use experience and
      damage mechanism research
      prior to initiation of the
      laboratory study (Phase I),
      or the field survey involving
      core sampling of existing
      structures (Phase II)."
      (p. 39)

18.   "It appears appropriate to
      limit the effort for FY86
      to the initial literature
      survey (Phase I).M  (p. 40)

19.   "It is not obvious that this
      device will prove useful in
      the proposed task of com-
      paring short-term effects
      for the diverse materials,
      e.g., steel, concrete, lime-
      stone, marble, paint films,
      and asphalt, either in place
      or eventually to be placed
      at the field exposure sites."
      (p. 40)

20.   "The panel believes that in-
      adequate attention has been
which is recognized as a leading go-
vernment laboratory on metallic
corrosion.

No interpretation of the corrosion
data was presented because this
phase of study is planned for FY86
when validated aerometric data and
3-year corrosion data are available.

We agree.  A joint research program
involving the exposure of painted
wood substrates in controlled environ-
mental chambers is being planned by
EPA and USDA Forest Products Labora-
tory.
We agree.
We agree.
The evaluation of the corrosion moni-
tor by ASRL was not presented because
the work is within the GAP program.
Based on the results of the lab and
field test of this device, it is our
reccranendation to incorporate the
corrosion monitor into the field ex-
posure study.
The uncertainties associated with the
development of the damage functions,
                                    60

-------
      paid to the uncertainties
      associated with the use of
      damage functions derived
      from short-term field expo-
      sure."  (p. 41)

21.  "With the limitations of the
      understanding of the chamber
      study results, the interpre-
      tation of the field studies
      becomes phenomenological.
      The damage functions for
      metals could be determined
      from the short-term chamber
      study results.  Expansion
      of the testing in the chamber
      studies is needed to generate
     "damage functions."  (p. 44)

22.  "The visit tothe RTF field ex-
      posure site demonstrated the
      substantial commitment of the
      National Park Service (NFS) to
      field exposure studies for
      marble and limetsone.  These
      materials should be considered
      for parallel chamber studies
      under the direction of the
      Materials Effects Program
      under review since the NPS
      has neither the resources
      (smog chamber and associated
      hardware) nor the expertise
      necessary to conduct such
      work."  (p. 45)

23.  "The major limitation is in
      the characterization of the
      mechanism of corrosion.
      While overall rates are
      empirically determined, the
      mechanism is still unknown.
      With the high covariance of
      the depositions of chemical
      species, the agent of corro-
      sion is obscured."  (p. 46)
particularly the uncertainty of the
regression coefficients, will be
available to the users - the Assess-
ment Program (Task Group I).
We agree.  (However, see responses
number 4 and 23.)
We agree.
The chamber studies are providing in-
sight into the role wet and dry acid
deposition play in the corrosion of
materials.  Deposition velocities,
soluble corrosion products, etc. will
provide insight into the mechanism of
corrosion.  However, a detailed charac-
terization of the mechanism of corro-
sion for all materials being studied
is beyond the funds available to this
program.
     We have attempted to respond to the list of recommendations (p.  48)  as
they appeared within the report.  As we have discussed, the covering  device
to be installed at Steubenville, Ohio should be completed in FY85.  The
covering devices at RTP and Steubenville will be used to resolve the  effects
                                   61

-------
of dry deposition of SC>2 on materials within  the timeframe of  the NAPAP.  The
effects of snow, ice, fog, etc.  are  not being addressed  in this program be-
cause of the limitation of resources.  We  agree that  irregular surfaced
materials should be included  in  this study.   Vte are proposing  to include
irregular shaped materials as well as field weathered materials as part of
our chamber program in FY86.
                                      62

-------
DATE:

SUBJECT:
      UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
              ATMOSPHERIC SCIENCES RESEARCH LABORATORY
                      RESEARCH TRIANGLE PARK
                       NORTH CAROLINA 27711


July 8, 1985


Response to Peer Reviewers Comments on Dry Deposition Program
FROM:     Kenneth T. Knapp
          Project Manager-,-Dry Deposition Project

TO:       Mr. Ronald Patterson
           Technical Planning and Review Office
       Reviewers' Comment

1.  "However, the project description
    is unclear as to the details of
    the parameterization." (p. 7)

2.  The ASRL concentration monitor,
    "the sampler should be tested under
    a variety of sampling and shipping
    temperatures, with realistic inter-
    vals between sampling and sample
    analysis,..." "...several such sam-
    plers should be. deployed in the field
    to provide an operational field eval-
    uation before commitments are made
    for permanent development." (p. 12)

3.  "The sampler does not yet have the
    capability of measuring, or has not
    been tested for, other species of
    "interest to NAPAP, namely HCL, HCO'o,
    ..." (p. 12)
4.  "However, we seriously question the
    merit of collecting the larger part-
    icles by use of the airborne concen-
    tration monitor and feel that the
    resources that would be devoted to
    this purpose should be used else-
    where." (p. 13)
                                  ASRL Investigators Response

                                  The PO is revising the project
                                  description of task 1452 on the
                                  IAG with Argonne.

                                  ASRL ran a short field test of 7
                                  week-long test then turned the
                                  systems over to EMSL who is now
                                  doing the recommended evaluation.
                                  The ASRL CM (sampler)  collects
                                  particles that can be  analyzed  for
                                  these species.  However,  neither
                                  time nor funding has permitted  the
                                  development of these measurements
                                  at this time.   While of interest to
                                  NAPAP, they are low priorities.

                                  I do not agree with the panels  as-
                                  sessment of the merit  of collecting
                                  particles larger than  lOum.   In fact,
                                  several good proposed  approaches were
                                  received in the proposals for this
                                  RFP.  In the RFP. we gave a  3 to 1
                                  priority to the smaller 2 to 10 pm par-
                                  ticles but felt that if it were possible
                                  we should start the research for the
                                  greater than 10 urn particles as soon
                                  as possible.
                                    63

-------
10.
II.
     5.  "The  issue >10um  particle  deposition
        may best  be  approached  through  the us
        of surrogate surfaces."  (p.14)
     "One of the  planned outputs  of the
     research sites  in  a manual  for site
     planning and development of  NTN sites."
     "A poorly chosen site  will  provide
     little useful data,  ..."  (p.15)

     "To meet the EPA goals, additional
     species are  needed.  Hydrogen
     chloride, ammonia	"
     "a co-location
     between Groups
     Carolina would
               of research activities
               D & G in RTF, North
               be desireable."  (p.17)
"(1) There appears to be little
communication between ASRL and ATDL,
and EMSL and ATDL, concerning the
deposition velocity method."
"...interaction between ASRL, EMSL
and ATDL on issues related to dry
deposition monitoring.  Regular
meetings to review plans and pro-
jects are essential..." (p. 21)

"The panel also recommends that the
capabilities of other qualified
organizations be sought to assist de-
velopment of deposition velocity
estimation methods." (p. 21)

"The panel recommends that research be
initiated in surrogate surface de-
velopment." (p. 25)
In the 1979 workshop on dry deposition,
many of the attendee felt that there
was no good evidences that surrogate
surfaces is a usable approach, thus it
was given a low priority.  We  are
not working on this since NOAA is
funding research in this area.  We
will withhold judgement of the merit
of this approach until  the NOAA work
is completed

While we donot disagree with this, the
output is not one of ASRL-EPA  but a NOAA/
ATDL output funded by NOAA.  Choosing the
NTN' sites is an EMSL/EPA function not
ASRL.

We intend to go after these but have
had to prioritize the pollutants by
NAPAP needs, research time available
and funding.

This is being worked on by EMSL and
John Spence (PM Group G).
                                                While the Vd project is a NOAA fund-
                                                ded ATDL/NOAA project, Bruce Hicks
                                                (ATDL) has kept ASRL informed.  I
                                                don't know the extent of the communi-
                                                cation between ATDL and EMSL.  I agree
                                                that some regular meeting is a good
                                                idea.  We have had several  such meetings.
We are not in a position  to  recommend
this since it is a NOAA assignment.  We
will be a part of the vigorous  review
of the method.
As stated in response to  number  5, this
is a low priority approach of  question-
able application and we  feel that the
NOAA support is sufficient at  this time.
                                          64

-------
             APPENDIX E

   REVIEW OF THE PANEL REPORT AND
RESPONSES BY THE LABORATORY DIRECTOR

-------
                     UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
        <                    ATMOSPHERIC SCIENCES RESEARCH LABORATORY
   *< PRO^C                            RESEARCH TRIANGLE PARK-
                                     NORTH CAROLINA 2771 1


DATE:     July 24, 1985


SUBJECT:  Review of ASRL Responses to the Dry Deposition and Material  Damage
          Peer Review
FROM:     Jack H. Shreffler
          Deputy Director, AS: (MD-59)

TO:       Ron K. Patterson
          Peer Review Coordinator, TPRO, ASRL (MD-59)

THROUGH:  Alfred H. Ellison
          Director, ASRL  (MD-59"

     I have read the report of the peer reviewers on the Dry Deposition and
Material Effects Programs at ASRL based on the review held April 23-25, 1985.
I have also read the ASRL project managers responses to the various recommenda-
tions and criticisms contained in the review report.  There were some mis-
understandings about the  latitude available to ASRL in choosing methods, time
lines, or budgets and about the exact responsibilities vis-a-vis other lab-
oratories or agencies in  NAPAP.  Generally, the responses are complete and
relevant, although some further questions are highlighted below.  By copy
of this memorandum I am instructing project managers to take some additional
action.

Dry Deposition

     The reviewers were complimentary and highly supportive of the systematic
approach taken to assessing a number of dry deposition methods.  A strong
recommendation was made to increase the interaction and communication between
ASRL, EMSL, and ATDT on issues related to dry deposition monitoring.  Ken Knapp
should take every opportunity to increase the interaction and should be able to
demonstrate the positive  effects of that effort in the future.

Material Effects

     The peer reviewers praised the management and project staff as having
demonstrated initiative,  industry, and the ability to work together.  There
was strong support for the chamber studies.  Field studies were also supported
with some reservations concerning methods.  The response prepared by John
Spenca is complete in addressing all reviewer criticisms, although there are
two areas that I believe  could use further elaboration.  John should prepare
a follow-up memorandum for the report.

     First, the response  to comment 1 (Spence memo) seems to skirt an important
issue.  The microenvironment may require a rather complex set of instruments
for characterization.  Measuring only on the building top or under the canopy
will not answer the real  question:  What is the difference between the site and
a nearby representative location?


                                      65

-------
     Second, the reviewers several  times mention study  of  corrosion'
mechanisms as a missing element (e.g.,  comments 14,  15,  23).   Perhaps
the response to 23 is getting close, but I  think there  should  be  some
elaboration on what "insights" will be  provided and  how that differs
from a "detailed characterization of the mechanism."

     In a conversation with John, I was told that agreements with comments
17 and 18 mean those actions will be taken.

cc:  K. Knapp
     J. Spence
                                      66

-------
          APPENDIX F




CLARIFICATION COMMENTS BY ASRL

-------
;  L.
        '•           UKiTED STATES E.VViRC'-T/.Er.TAL PP.
                                   RESE^-C'-n IKiA\C;i_E F-.-
                                    NO"TH CAROLINA 277" '
     MEMORANDUM

     DATE:      July 30, 1985

     SUBJECT:   Clarification of ASRL Response to Material
               Damage Peer Review

     FROM:      William E.  Wilson
               Acting Director, EMCD '

     TO:        Jack Shreffler
               Deputy Director, ASRL


     Microenvironments (#1)

         There were two questions raised regarding microenvironments.  One has to
     do with the  stationary covering device used by the Bureau of Mines.  Measure-
     ments have shown a higher  relative  humidity under the  shelter and the Bureau
     of Mines plans to modify the covering device so that temperature and relative
     humidity will  not be different.  The second question referred to the increased
     temperature  at rooftop  sites due to  absorption of  light by  asphalt roofs.
     The reviewers  recommended  that "the  researchers  should modify  the  exposure
     conditions in  a manner that eliminates the microenvironment effects."  At one
     site the roof  is covered with a light colored  gravel which should reduce the
     temperature  over  that of  a black  asphalt roof.   We  will  investigate  the
     possibility  of covering the black  asphalt at the other site with light colored
     rock.

         We must emphasize, however, two factors.   One,  in deriving damage func-
     tions the  important  considerations are that  the environment to which  the
     specimens  are  exposed is adequately measured,  not that it  is characteristic
     of any  specific  exposure  condition, and  that  an  adequate  range  of  exposure
     conditions are included in  the  set  of exposure sites.  Therefore, we do not
     believe that the fact that the rooftop environment is different from a ground
     level enviornment will reduce the accuracy or value of the resulting damage
     functions.   However,   since  the time-of-wetness is important  we investigate
     the possibility of modifying the  one  rooftop  site that might have the time-
     of-wetness lowered by the  asphalt roof.  The  second fact  is that because of
     financial  restraints   we must  install  cur exposure  racks  at  pre-existing
     sites.  Due  to  security  problems  many urban  sites  are  placed  on  roofs.

         At one  site a rooftop vent will be  extended to a sufficient height so
     that the emissions will not impact the specimens during most wind conditions.
     The buildings,  a library and an office  building are  not expected to  generate
     emissions  of significant concern.
                                       67

-------
Mechanisms (#23)

     The reviewers emphasized  the need to determine the mechanisms of damage
in order to develop damage functions which could reliably apportion damage to
specific pollutants and which could be used to predict damage under a variety
of environmental  conditions.  We agree  with the  reviewers.   We  have  been
making some progress  in persuading NADAP to  accept this viewpoint but still
do not have sufficient  funds to implement an adequate mechanistic study.  We
feel that the analysis of corrosion on the specimen surface and products in dew
and rain runoff  from both field and laboratory samples will provide a useful
data base  for research  into damage mechanisms.   We will  use  the reviewers
comments in the  next round of program planning  in an effort to obtain funds
to add research on mechanisms to  the program.
                                      68

-------