United States
Environmental Protection
Agency
Atmospheric Research and Exposure
Assessment Laboratory
Research Triangle Park NC 27711
Research and Development
EPA/600/S3-90/017 June 1990
 Project  Summary
 Urban  Aerosol  Transformation
 and   Transport Modeling
 Y.Seoand J. R. Brock
   This project has been concerned
with development of an urban aerosol
model which includes the  dynamical
processes shaping  the aerosol size
and composition  distributions. A K-
theory  model, termed AROSOL,  for
the super- and sub-micrometer aer-
osol  mass concentration  has been
developed and tested with urban air
pollution data. This  report deals with
the development and inclusion of
computer modules for description of
the dynamics of  single component
and binary aerosols. Current work
has proceeded along several  lines.
The data bases from the  EPA Phil-
adelphia Aerosol Field Study  (PAFS)
have been examined  and  corrected
for evaluation of AROSOL for sulfate
aerosol modeling.  Modules  have
been included in AROSOL which can
account for chemical conversion, nu-
cleation, condensation, evaporation,
and coagulation as  a  step in the ex-
tension  of AROSOL to multicompo-
nent aerosols. Finally, work has been
done on coupling the nucleation and
condensation/evaporation processes
to a suitable oxidant mechanism for
sulfate and  nitrate formation. Be-
cause the  PAFS data base and other
similar data bases are not sufficiently
complete, it has not been possible to
thoroughly evaluate the  accuracy of
the computer modules.

   This  Project  Summary  was
developed  by EPA's Atmospheric
Research and Exposure  Assessment
Laboratory, Research  Triangle Park,
NC, to announce  key  findings of the
research  project that  is  fully
documented  in a separate  report of
the same  title (see Project  Report
ordering information at back).
Introduction
    The  adverse effects of air pollution
are well  known: increased incidence of
respiratory  diseases, reduced  visibility,
economic  loss, inadvertent  weather
modification, and acid  precipitation, to
name a few. The urban aerosol plays an
important role in these adverse effects
An  understanding of this role  requires,
among other factors,  knowledge  of the
aerosol dynamical processes shaping the
aerosol size and composition  distribu-
tions. These dynamical  processes in-
clude advection, dispersion, deposition,
primary  and secondary  source input,
nucleation, condensation/evaporation, and
coagulation.  Models for the atmospheric
evolution of particle size and composition
distributions reflecting these  dynamic
processes will   ultimately be necessary
for  development of realistic aerosol air
quality regulations and control strategies.
    This project has been concerned
with development of urban  aerosol
models  which  include the dynamical
processes shaping the aerosol size and
composition  distributions. The first part of
this work has involved evaluation of data
bases and  a K-theory model, termed
AROSOL,  for  the super- and  sub-
micrometer aerosol  mass concentration.
Recent and some current  work  has been
devoted to inclusion and development of
computer modules for description of the
coagulation  and  condensation evapora-
tion  processes for single  and multicom-
ponent aerosols. The  major  part of
current  work  has proceeded  along
several lines. The data bases  from the
EPA Philadelphia Aerosol Field Study
(PAFS)  have been  consolidated  and
corrected for evaluation of AROSOL for
sulfate aerosol modeling.  Modules have
been included  in  AROSOL which can
account for chemical conversion, nuclea-

-------
and  aerosol modules'?hatOf the cnernica'
The  modules for seconrS      available.
nitrate formation havPh^ Sulfate ancl
*'th  data  from th- 7no" de.monstrated
                                            surement  data   hese '.fTI? and
                                            analysis optfons 'of mnH° Ude sta«s"cal
                                            recommended  by  EP!  eVa'Uati°n as
                                            spectral analysis.               wel1  as

                                                              9rid  Consists
                                          ""cknes
                                                    rr       7
                                                 very  large and  a  thf
                                                 relative "humTdity  ,'
                                                 conditions are not
                                                  ' ii LI ait/
                                                                             MARS can also oe used tn *      	
                                                                             •mately aqueous phase ox d      approx-

                                                                             ozone  and Mn LnV Pnydr°9en peroxide,
                                                                             MARS  provides  a  rnn  CatlOn  Cata'ysts.
                                                                            estimate of nart^n   en'ent and rap.d
                                                                            species between aeroT? °f  the  Stated
                                                                            However, it must be n   rian  93S  pnase.
                                                                            does not address nn*??.!    tne model
                                                                            this  partitioninVby SSer  -- a'terati°n  Of
                                                                            constituents, and ft -    •
                                            LSODE
                                                                   known
                                                                                   na onen         .-       p^
                                                                                 for this approach  hJfh6 fation-
                                                                             explicitly. The  cnnwh  ,     een  shown
                                                                             Processes for  mese S'°"S to the
                                                                             indicated schemafca^y   6

                                                                                °f
conditions
of data
                     and stability
                      *"
                          ran9e
CBM IV tht isrCUeS (he versi°"  of

              -
                                                                           <3)  (rate of chane due to
                                                                               tion)              to

-------
                   Advection    Dispersion     Dry Deposition     Chemistry     Aerosol Dynamics

Figure 1.     Schematic diagram of structure of urban aerosol model AROSOL
                     Empirical Model
                     Meagher (EMM)


CB - IV




Atkinson-
Uoyd
 Figure 2.    Options  in  AROSOL  for
            calculation of homogeneous
            gas phase reactions.
         AROSOL Dynamics
  Lumped
  Aerosol
   Model
   (LAM)
 Aerosol
Dynamics
 Model
 (ADM)
Model for an
  Aerosol
  Reacting
  System
  (MARS)
Figure 3.    Schematic diagram of aerosol
            dynamics options available in
            AROSOL.
It may  be noted that the coagulation
process does  not  enter  since the total
mass concentration is conserved  by
coagulation;  coagulation does  not alter
the mass concentrations but does change
the  composition  distribution. The  limit-
ation of LAM  appears  in the  aerosol rate
processes,  (3), (4), (5), and (6) which all
require  knowledge of the particle  size
distribution. Current practice  is to treat
such rate processes in  an empirical or
semi-empincal  manner  whose general
validity  is  unknown.  If  heterogeneous
conversion  processes  are  negligible, the
rate of  formation of a species' particulate
mass  concentration is assumed  to be
equal to the  rate of production  of vapor
phase  mass  concentration by chemical
reaction. For  those species  depositing
entirely  in  the  aerosol phase,  this
assumption appears to be  acceptable. Of
course,  for  a  species such as nitrate, the
LAM must be  supplemented  with  an
equilibrium  model such as  MARS.
    For  the  atmospheric aerosol,  the
aerosol dynamics model (ADM) is formu-
lated in  its most  general  form  in terms of
the distribution  function for a  multi-
component  aerosol.  As  yet,  such  a
general  formulation  cannot be validated
owing to lack of field  data. A module  is
included in  AROSOL under  ADM for
binary  homogeneous  nucleation,  con-
densation/evaporation and coagulation.
This module is termed the binary aerosol
dynamics model (BADM) and  is intended
for  simulation of new particle formation
by binary homogeneous nucleation which
may occur, for example, during oxidation
of S02  in urban plumes.  BADM uses  a
moment  method for computational  effic-
iency,  although  very  accurate  modules
have also  been developed for com-
parative  purposes. As this module  was
developed  only  recently  as  part of this
work, it will  be outlined here.
    A binary  aerosol is described by its
distribution,  n(m1,m2,x,y,z,t),  where
ndm1dm2  is  the number of particles
having  masses of species  1 and 2 in the
ranges  pr^dm,  and m2dm2 at spatial
point x,y,z at  time t. The evolution of n  is
followed  through several of its moments,
Mij:
My  =//m1lm2J n(m1,m2) dm1dm2    (1)
where the integration is  over all masses
m-| and  m2.  The moments necessary for
homogeneous  nucleation and particle
growth are M00 (the total particle number
concentration), M10, M0i  , M20, and M02.
    For  simplicity, the  coagulation term
will  be omitted in  this  description; it is,
however, included in the general moment
formulation. When particle growth occurs
by binary nucleation and condensation/
evaporation  processes,  the  evolution
equation for n has the form
an/at
= (a/am1)(G1n) - (a/dm2)(G2n) + R12    (2)
where G-),  G2  are  the  particle  growth
rates for 1 and 2 and R12 is the  binary
nucleation rate. From  equations (1) and
(2) rate  equations for the  moments M10
M0t M20, and M02 are easily derived.
    G!  and  G2  are the  noncontinuuum
growth rates and include the Kelvin effect
and activity  coefficients for the mixture.
The expression  for R12  may  be taken
from  the classical theory  of  binary
homogeneous  nucleation with  its  modif-
ications. These modifications include
nucleation in the presence  of existing
particles.
    The moment  equations  are  not
closed; closure is in general not possible.
Closure  is  achieved here by  using a
parameterized  form for the distribution n.
It is assumed that in the atmosphere, the
mixture  of secondary  and primary  par-
ticles  is  an  intrinsic mixture  in that all
particles will contain both  species 1 and
2. For this case, closure is obtained with
the model:
n(m1,m2)M00f1(m1)f2(m2)          (3)
where the fj are lognormal frequencies:
fi   =(2iO   -  1/2(1/1*1^^)
    exp(- In2(m/mavj)/2ln2nj)
and mavi and  m. are  respectively the
mean mass  and  geometric  standard
deviation for species i. These parameters
are  easily expressed  in  terms of the
moments defined above.
    The validity  of this moment method
has been studied in two ways. First,  \r

-------
f/)e absenr
*r°«th $*

          con V accoun,;/, f^se rZ /Or f/1e







          - QOO/ u 't'lriiQatJn^ _uus r(=lri —
        ^^/ss/nr,^ . ^fe cnn»^.-_ rria'n/'no rea

                  r-» - ^

       9'ven hi/ c.of su/ffl^ n/i 'nc/ur/o. °'
                     — ^er So  are; (7) .a  model
   rer
  no n °rd
   prec'P






                           l
                 Hit
                         mese fw0

-------
models.  The interpretation  of these
various differences is not possible owing
to the lack of detailed information for the
NMOC in the PAFS data base. Also, the
hydrocarbon  data available in PAFS are
known  to  be inaccurate. For this  same
same study period, MARS was  used to
demonstrate secondary sulfate speciation
from the conversions given by CBM;  as
would be expected, the gas sulfate is
very small  and  virtually all the sulfate
resides in the aqueous aerosol. Also from
the CBM calculations,  nitrate speciation
calculated by MARS was also obtained.
Based  on  the  assumed  background
ammonia levels, MARS predicts that  all
the nitrate resides  in the gas  phase with
no  aqueous  or  particulate  nitrate;  the
relative  humidity during this period was
approximately 50%, a value too  small to
shift the equilibrium  toward  aqueous
phase ammonium  nitrate. It is  possible
that  these  findings would need to be
modified  if contributions of nitrate and
sulfate  from background and fumigation
had been considered.
   In all  the  simulations with  CBM for
sulfate,  low  sulfate formation  in  the
modeling domain was found. There are a
number of possible reasons for this. First,
as noted above,  the  hydrocarbon  data
from PAFS are known  to  be  inaccurate;
water interferences are  now known  to
have been present in the analyzers; this
results in hydrocarbon values that are too
low.  Therefore, the  resultant  sulfate
conversion/deposition rates are too small.
Also, with limited  data  on hydrocarbon
speciation available, assumptions accord-
ing  to  the  EKMA procedure  were
necessary;  these assumptions may  not
have been valid for the modeling  domain.
Finally,  the  large differences in sulfate
conversion given by the CBM and  ERT
ALM  models  indicate  difficulties  in
applying current kinetic  models.
   Urban scale  aerosol models of  the
future most likely will need  to deal  more
effectively with some essential problems
characteristic of the urban scale. Two of
these are: (1) Data bases resolved  with
small time  and  length  scales;  (2)
Reaction  and turbulent  mixing  at  small
time and  length scales.  The first of these
will probably require advances in low cost
field measurement technology.   The
second  problem,  in  the context  of
Eulerian models, must   involve develop-
ment of  subgrid  scale models which
account  for the fact that in many urban
areas pollutant transport and transforma-
tion  involve  reaction,  mixing,   and
entrainment  by  plumes over  relatively
small time and length scales.
   Further research should be designed
to  lead  to valid  models for  urban  scale
photochemical reactions and for aerosol
formation applied in simulation of reaction
in  atmospheric turbulence. The current
lumped  parameter (semiempirical)  mod-
els in use, such as CBM and ALM, have
apparently   not been  tested   for
concentration  fluctuations  with  frequen-
cies  characteristic of the urban  scale. In
anticipation of the need,  it is suggested
that it would be useful  to begin  to direct
some smog chamber  and field studies
toward  obtaining the  necessary infor-
mation for semiempirical models suitable
for reactive plume modeling.

-------

-------

-------
   Y. Seo and J.R.Brock are with the University of Texas, Austin, TX 78712.
   H.M. Barnes  is the EPA Project Officer (see below).
   The complete report,  entitled  "Urban Aerosol Transformation  and Transport
      Modeling," (Order No. PB-90 187  1701 AS; Cost: $17.00, subject to change)
      will be available only from:
            National Technical Information Service
            5285 Port Royal Road
            Springfield, VA22161
            Telephone:  703-487-4650
   The EPA Project Officer can be contacted at:
            Atmospheric Research and Exposure Assessment Laboratory
            U.S. Environmental Protection Agency
            Research Triangle Park, NC 27711
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
EPA/600/S3-90/017

-------