United States
                     Environmental Protection
                     Agency
Atmospheric Sciences
Research Laboratory
Research Triangle Park NC 27711
                     Research and Development
.EPA-600/S3-84-114 Jan. 1985
&ERA          Project Summary

                     Hazardous  Air  Pollutants:
                     Dry-Deposition  Phenomena
                     G.A. Sehmel, R.N. Lee, and T.W. Horst
                      Dry-deposition rates were evaluated
                    for two hazardous organic air pollutants
                    in order to determine their potential for
                    removal from the atmosphere by three
                    building material surfaces. The organic
                    air pollutants  were nitrobenzene  and
                    perchloroethylene; the building materials
                    used were cement, tar paper, and vinyl
                    asbestos tile. Dry-deposition experi-
                    ments were conducted in two stirred
                    chambers that were originally designed
                    and constructed for determining re-
                    moval rates to vegetation in the environ-
                    ment. Building materials were placed
                    on the bottom of the stirred chambers,
                    and removal rates were evaluated by in-
                    troducing the organic air pollutant
                    along with nondepositing SF6 tracer gas
                    into a stirred chamber. Changes in air-
                    borne concentrations with time were
                    then monitored.
                      Because the HAP removal rates were
                    small, sulfur hexafluoride tracer  gas
                    was  used to  evaluate the rates of
                    leakage  from  the stirred chamber in
                    order to correct the deposition rates.
                    When pollutant wall losses were assumed
                    negligible,  pollutant concentration
                    decreases (corrected using  the  SF6
                    leakage  rate)  ranged from near 0 to
                    3%/h. Dry-deposition velocities were
                    calculated from the rate-of-change over
                    time of the airborne concentrations. All
                    calculated deposition  velocities were
                    approximately 10~4 cm s'1 or less.
                      This Project Summary was developed
                    by EPA's Atmospheric Sciences  Re-
                    search Laboratory, Research Triangle
                    Park, NC, to announce key findings of
                    the research project that is fully docu-
                    mented in a separate report of the same
                    title (see Project  Report ordering
                    information at back).
Introduction

  The role of dry-deposition in removing
potentially hazardous air pollutants
(HAPs) from the atmosphere is not well
defined. This project was conducted to
assess the magnitude of the process,
especially deposition involving surfaces
expected to be found in urban settings.
  The dry-deposition measurements
were made in stirred chambers (SCs). The
SC is an inert-surface box within which
are placed selected deposition surfaces.
The principal advantage of using the SC is
that concentrations  can be controlled
within manageable limits throughout the
duration of each experiment. During the
experiments, the SCs were  sealed to
minimize leakage  and charged with a
preselected atmosphere of HAP and SFe
tracer gas that was mixed continuously
using an open fan. Concentrations within
the SC were monitored intermittently for
specified time periods depending on the
concentration  decay rate. Deposition
rates were then calculated on the basis of
concentration-decay rates.
  Seventeen experiments were conducted
to determine the maximum removal rates
for HAPs to large deposition-surface
areas. Dry-deposition velocities of per-
chloroethylene were evaluated during 12
experiments with cement, tile, and tar
paper substrates. Deposition velocities of
nitrobenzene were evaluated during five
experiments with tile and tar paper sub-
strates. The velocities were calculated
from  airborne concentration  losses of
HAP and SF6 tracer gas with  time. The
nondepositing SFe tracer gas was used to
measure HAP  leakage rates  from  the
stirred chambers. These measurements
of HAP dry-deposition velocities can be

-------
usedjn the general predictive atmospher-
ic transport, diffusion, and deposition
model  MPADD (Multi-Pollutant Atmos-
pheric  Deposition and Depletion),  also
developed as part of this effort.

Procedure
  The experimental design was based on
the assumption that HAP dry-deposition
velocities would be comparable to those
of other  gases.  A literature  review of
reported  dry-deposition velocities for
gases suggested that the deposition rate
for many of the HAPs should range from
1CT4 to 1CT2 cm s"1. The  loss rates of
HAPs, as measured in SCs, were predicted
to vary from 0.5% to >10% h~1 for the
anticipated range  of dry-deposition
velocities. Thus,  a stirred  chamber
measurement approach was selected for
the experimentation.
  Deposition substrates were selected to
represent typical construction materials:
cement, tar paper, and vinyl asbestos tile.
Cement  was selected to  represent
construction  materials  in  buildings,
streets, and highways. Cement  patio
blocks were placed on the floor of the SCs
in two configurations with different
surface areas available for  deposition,
either  6.1  or 10  m2. Similarly,  vinyl
asbestos tile was selected because it is a
floor covering used in  industry and
homes. Tar paper was selected because it
has a high organic content, which might
affect HAP removal rates. Both the vinyl
asbestos tile and  the tar  paper  were
placed  to cover the SC floor  uniformly.
After preliminary characterization tests
with the SFe tracer  were completed,
experiments were begun using both the
HAP and the  SFe. Stirred chamber
experiments were conducted .with either
perchloroethylene (and SF6) or nitro-
benzene  (and SFe). At time zero, the
chamber was charged with a preselected
concentration  of  HAP  and inert tracer
gas. The concentrations within the
chamber were monitored to provide a
time series of the HAP and tracer decay.
In some experiments, both  SCs  were
used to conduct parallel runs to assess
the reproducibility of the experiments.

Results
   Figure  1  shows the concentration
profiles of a typical experiment involving
perchloroethylene and SFe. The data are
normalized for presentation on the same
graph by dividing each measured value by
the starting concentration. Since the
concentrations  in the chamber tend to
decrease exponentially with time, a plot
of the log of the normalized concentration
versus time should be linear. Loss rates of
the HAP and tracer were calculated from
the slope of the least-squares line which
best fit the data.
  Both HAP  and SF6  loss rates were
significant and  ranged from 1.44 to
7.19% h"1. Correcting the HAP loss rates
for  leakage (as measured by the tracer
losses) the  resultant  HAP deposition
losses ranged from near zero to 2.96%
h"1.
  Deposition velocities were then calcu-
lated from the HAP loss rates. For these
calculations, HAP and tracer leak rates
were assumed  to be  identical,  and
deposition in the SC to surfaces other
than the selected deposition  substrate
(e.g., the Teflon walls) was assumed to be
negligible. The results are summarized in
Table 1. The HAP dry-deposition velocities
were  consistently  small, approximately
10~4 cm s~1 or less.
  The observed  HAP dry-deposition
velocities are near the lower limit which
was anticipated, based on values reported
for  other gases. The relatively low HAP
dry-deposition velocities may have been
caused by differences  in substrate (the
                             use of building materials in this study,
                             rather than the more traditional vegetative
                             canopies) or may result because organic
                             gases  deposit less rapidly than the
                             inorganic gases for which most measure-
                             ments have been made. Further research
                             is needed to clarify these possibilities.
                               Neither  substrate area (either 6.1  or
                             10.0 m2 for cement) nor increased fan
                             speed  (experiments 4,  10,  and 11)
                             significantly affected  HAP  loss rates.
                             Only two fan speeds were used: 520 rpm
                             and  a  motor-limited maximum rpm  of
                             600. Nearly all experiments were conducted
                             at a fan speed of 520 rpm.
                               Although losses tended to be exponen-
                             tial  with  time, other  time-dependent
                             effects were apparent. Concentrations
                             increased immediately after gas injection
                             and continued to increase until a maximum
                             concentration was observed. Subsequent-
                             ly, concentrations tended to be exponen-
                             tial with time  as described  by the
                             equations  from which dry-deposition
                             velocities are calculated. Also, deviations
                             from  an exponential decrease were
                             sometimes observed. These  deviations
        10
 a
       10'
       10'
                   Tracer Loss, %/h


                      	  D      SFe 2.04

                      	  A     C2O, 2.40

                  	I            I
                                     I
                                    40          80

                                      Time, Hours
                                                 120
160
Figure 1.
Perchloroethylene andSFe tracer gas concentration decreases for 6.1 rrf of cement
block deposition surface.

-------
Table 1. HAP-Dry Deposition Velocities
Expt. HAP
No. SC Chemical Name
1
2
4
6
7
8
9
10
11
15
16'
17
18
19
1
2
1
2
1
2
1
1
2
1
2
1
1
2
1
2
1
perchloroethylene
perchloroethylene
perchloroethylene
perchloroethylene
perchloroeth ylene
perchloroethylene
perchloroethylene
perchloroethylene
perchloroethylene
perchloroethylene
perchloroethylene
perchloroethylene
nitrobenzene
nitrobenzene
nitrobenzene
nitrobenzene
nitrobenzene
Deposition
Surface
cement
cement
cement
cement
cement
cement
tile
tar paper
tile
tar paper
tile
tar paper
tar paper
tile
tar paper
tile
tar paper
Surface
Area
(m*j
10.0
6.1
10.0
6.1
10.0
6.1
6.0
6.0
6.0
6.0
6.0
6.0
60
6.0
6.0
6.0
6.0
Fan
Speed
(rpm)
520
520
520
520
600
600
520
520
520
520
600
600
520
520
520
520
520
HAP
Deposition]
(%/h)
0.42 + 0.31
0.36 + 0.27
0.34  0.35
0.75 + 0.12
0.72 + 0.17
0.08 + 0. 14
0.41 +0.14
0.48 + 0. 14
-002 + 0.53
0.06  0.42
1.16+0.61
1.57 + 0.98
2.96 +3.60
0.01 1.06
0.70 +1.27
1.19 +0.69
0.68 + 0.72
Deposition
Velocity]
(cm/sec)
5.2 + 3.8 x 10~5
7.3  5.5 x 10~5
4.2 +4.3 x 10~5
1.5 + 0.2 x 10'*
8.9 +2.1 x 10~5
1.6 + 2.8x JO'5
8.4+ 2.9 x 10~5
9.9 + 2.9 x 10's
-0.0+ 1.1 x 70~"
1.2 + 8.6x 10~5
2.4 + 1.3 x 70~5
3.2 +2.0 x 10~4
6.1 +7.4 x 10''
0.0 +2.2 x 10'*
1.4 +2.6 x 10'"
2.5+ 1.4x W"
1.4 +1.5 x 10~4
t Tabulated uncertainties represent the 90% confidence interval.
might have been caused by temperature
effects and sorption of gases on airborne
particulates  rather  than  on  the test
substrates.

Conclusions and
Recommendations
  The  removal of gaseous organic air
pollutants by dry deposition has not been
studied extensively and few experimental
data are available. Hence, obtaining dry-
deposition velocities for nitrobenzene
and  perchloroethylene is a significant
advance in the understanding of organic
removal rates. Measured deposition
velocities  were significantly less than
those  usually reported  for inorganic
gases. Deposition velocities were consis-
tently very low. Many calculated values
were not significantly different from zero.
  Temperature effects were not intensive-
ly investigated since all successful
experiments were conducted  in warm
summer weather (ambient temperatures
from 27 to 36C.) Because  the dry-
deposition velocities measured for the
two HAPs investigated were low, we can
conclude that airborne concentrations of
HAPs are not significantly reduced by dry
deposition at temperatures characteristic
of summer months, or at temperatures
maintained within  buildings.  Vapor
pressure, however, is anticipated to be a
very important property influencing dry-
deposition removal  rates  for  organic
gases.
  The SC approach for evaluating  HAP
dry-deposition velocities  was adequate
for the HAPs and substrates investigated.
Leakage was always observed,  and
leakage rates tended to increase during
the experimental series. To alleviate this
problem,  the SCs were periodically
maintained by replacing critical  seals or
tightening the bolts holding  sections of
the  chambers together to  minimize
exchange with outside air.
  One important  advantage of the SC
technique is that a large substrate
surface  area can be contained for the
deposition substrate. Even though  dry-
deposition velocities might be low,  they
could be a function of spatial variations in
substrate characteristics.  Hence, by
using the relatively large surface areas in
an SC (6.0-m2 floor area), average  dry-
deposition velocities can be evaluated
over these spatial variations.
  The authors recommend that the SC
approach be used to evaluate additional
HAPs and deposition substrates. However,
other measurements techniques should
be considered. Thermogravimetric tech-
niques may be useful as a  screening
method to evaluate  maximum removal
rates for small surface areas on the order
of 1 cm2. This technique may be essential
for evaluating deposition velocities less
than about  10~6  cm  s~1.  Flow-through
chambers may be considered if the
removal rates are significantly greater
than those evaluated  in the current
experiments. Instrumentation with  high
sensitivity and rapid analysis techniques
would be required if flow-through cham-
bers were used.
  The authors also recommend that dry-
deposition velocities  be evaluated for
additional  construction materials, and for
vegetative canopies and water surfaces.
Currently,  no  data exist for  predicting
HAP dry-deposition velocities for vegeta-
tion and soils. Dry-deposition velocities of
HAPs onto construction materials appear
lower than dry deposition velocities for
other gases that  were  evaluated for
vegetative canopies.

-------
    G. A. Sehmel. R. N. Lee, and T. W. Horst are with Pacific Northwest Laboratory,
      Rich/and. WA 99352.
    Larry T. Cupitt is the EPA Project Officer (see below).
    The  complete report,  entitled "Hazardous Air Pollutants: Dry-Deposition
      Phenomena."(Order No. PB 85-138 279; Cost: $11.50, subject to change) will
      be available only from:
           National Technical Information Service
           5285 Port Royal Road
           Springfield VA 22161
           Telephone: 703-487-4650
    The EPA Project Officer can be contacted at:
           Atmospheric Sciences Research Laboratory
           U.S. Environmental Protection Agency
           Research Triangle Park, NC 27711
                                    * U.S. GOVERNMENT PRINTING OFFICE; 1985  559-016/7896
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
       OOOC329   PS
       II  $  ENVIR  PR0TECTIOK  AGENCY
       REGION 5  LIBRARY
       230  S  DEARBORN  STREIT
       CHICAGO               IL

-------