United States
Environmental Protection
Agency
Research and Development
Atmospheric Research and Exposure
Assessment Laboratory
Research Triangle Park NC 27711
EPA/600/S3-90/093 Feb. 1991
Project Summary
Development of Sampling
Methodology for Dilution Air
Sampling of Condensible
Emissions from Stationary
Sources
William E. Farthing
This report describes the initial devel-
opment of a technique using dilution of
stack gas with conditioned ambient air
for measurement of the particulate mass
of condensible emissions from station-
ary sources. The methodology developed
is designed for widespread application
to measure emissions which are in the
vaporphaseattemperaturesgreaterthan
that of the Method 5 filter and which
immediately condense to the particulate
phase upon mixing in a temperature
controlled chamberwithairthathasbeen
cooled, dried, and filtered.
The front half of the condensibles air
dilution train (CADT) is a Method 5 probe
and filter. The portion of the train for
collection of condensibles (back half)
includes a dilution air injection cone and
a mixing chamber followed by a separate
filter for condensibles. The temperature
selected for the filter for condensibles is
20°C, and the dilution factor is 15:1 on a
volume basis, high enough to prevent
condensation of moisture.
The filter for condensibles is desic-
cated for 24 h in a Teflon envelope before
weighing.This limitsthe evaporative loss
of substances condensed during sam-
pling. For sources where H2SO4 is
present, a procedure was selected for
correcting the weight for moisture re-
tained by H2SO4.
The selected protocol was tested in
the field at a site with predominantly
organic condensible emissions. The
CADT was operated simultaneously and
collocated with Method 5 trains which
used another protocol, the impinger
approach, for condensible emissions.
The mass of condensible emissions
measured by the CADT ranged from 25.2
to 27.6 mg/dscm. The average difference
between the two approaches was 2%.
The standard deviation of the differences
was 15%. The precision of the dilution
approach (standard deviation of replicate
sampling runs) was estimated to be 15%.
This Project Summary was developed
by EPA's Atmospheric Research and
Exposure Assessment Laboratory, Re-
search 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 PM10 ambient air particulate stan-
dard has created a need for measurement
methods for PM emissions (particulate
matter of nominally 10-p.m aerodynamic
diameter and smaller) from stationary
sources. Previously, techniques have been
developed for measurement of PM,0 emis-
sions present at process stack conditions.
Some em issions in the vapor phase at stack
conditions are converted to the condensed
phase immediately upon discharge into
ambient air. Substances condensed in
stationary source emissions are believed to
be primarily incorporated with particles less
than 10 jim in diameter. Because of the
emphasis on PM10, the need for a practical
and precise source test method for the
condensible component of emissions has
become important.
The term condensibles is used to referto
materials in the vapor phase at stack condi-
{§y Printed on Recycled Paper
-------
tions which immediately convert to and per-
sist in the solid or liquid phase when sampled
with the source test methodology. This
particulate matterandthat sampled at stack
conditions are referred to as "primary" par-
ticulate emissions. This portion of the am-
bient particulate burden is distinguished
from "secondary" particulate emissions,
which are the results of slower reactions,
such as conversion of SO, to sulfates and
photochemical formation of organic smogs.
Based upon a literature review of current
approachesforsuch source measurements,
two approaches were recommended for
research and development. One recom-
mended approach was dilution cooling of
extracted process exhaust gas with condi-
tioned ambient air. The other was use of the
Method 5 back half (impinger catch) with
standardized procedures for recovery of
condensibles. It was pointed out that for the
definition of condensiblesto be precise (i.e.,
for a consistent definition of emission rates
of condensible material from the source),
conditions, including the sampling tem-
perature at which the sample is obtained,
must be within specified limits. Sampling
conditions should be independent of the
local ambient conditions at the time of the
sampling because, as they vary, these local
conditions may cause varying amounts of
condensation. Examples of local conditions
which vary include background pollutant
levels, adsorption or absorption processes,
and meteorological variables, all of which
may contribute to the actual amount of stack
emissions converted from the vapor to par-
ticulate phase at any given time. Specific
source sampling conditions and hardware
for measurement of condensibles were
selected for the practical and theoretical
aspects described in this report. Therefore,
condensible emissions are defined only in
terms of the recommended sampling meth-
odology.
This report describes the initial develop-
ment effortfor measurement of condensible
emissions by air dilution. Selection of spe-
cific procedures for widespread use with
emphasis on measurement of the mass of
condensible emissions from stationary
sources was the major goal of this project.
Development Approach
Sampling Hardware and
Sampling Procedure
Thecondensiblesairdilution train (CADT),
fabricated and assembled for testing in the
field, Includes the front half of a Method 5
train, nozzle, probe, and filter. Measure-
ments with the CADT provide values for
condensibles relative to current values ob-
tained with Method 5. The condensibles
portion of the train includes a heated sample
orifice meter, dilution air injection cone and
mixing chamber, and the filter for
condensibles. The sample orifice meter is
heated to prevent cooling of the sample gas
prior to entering the dilution air injection
cone.
The sampleorifice meter servesthe same
purpose as that used in Method 5, the
monitoring of sample flow rate required to
maintain isokinetic sampling. In addition, it
serves the purpose of the dry gas meter in
Method 5; the total sample gas volume is
measured atthis point, before dilution of the
sample. This is accomplished with digital
electronic integration of the square root of
the signal from adifferential pressure trans-
ducer. The operator enters a scale factor
calculated directly from the usual sample
and-ambient--gas parameters-and-orifice-
AH_ to provide digital readout of flow rate
anoaccumulated sample volume. The front
half of the CADT is operated by the same
protocol as Method 5. Alternately, it is de-
signed to be operated according to the
constant sampling rate (CSR) procedurefor
PM
me selected temperature at which the
condensibles filter is to be maintained is
20°C and the selected dilution factor is 15:1
on a volume basis (corresponding to near
20 on a mass basis). The dilution factor was
chosen to be high enough to avoid con-
densation of water while limiting the size of
the train components for practical use. The
diluter was based on a 50% scaleup of one
developed previously to provide high tur-
bulence and uniform mixing. The mixing
chamber was sized to provide a residence
time of approximately 2 s.
The dilution air consists of ambient air
conditioned by cooling in an ice bath con-
denser and filtering. The temperature of the
dilution air must be controlled at a tem-
perature low enough to obtain the desired
temperature (20°C)of the combined sample
and dilution gas. Duringrsampling the flow"
rate of the dilution air is varied with the
sample flow rate to maintain the volume
dilution factor, including changes of the
absolute pressure in the diluter due to in-
crease of the differential pressure across
the Method 5 filter.
Exploratory Laboratory
Measurements for Analytical
Procedures
It is expected that sampling for
condensibles will provide particulate
samples with significant differences from
those addressed by Method 5 analytical
procedures; in particular, the volatility of
condensible samples will be higher. Diffi-
culties in obtaining satisfactory sample
weights due to substantial variations have
been attributed to evaporation of substances
collected during sampling runs and to ad-
sorption or absorption of water by hygro-
scopic substances such as condensed sul-
furic acid, tt is relevant that desiccation to
minimize the contribution of adsorbed or
absorbed water may lead to the removal of
a proportionately larger mass of other con-
densed substances than is experienced
with Method 5, samples.
Laboratory measurements of simulated
filter samples were performed to define
procedures for determining weights of
samples of condensibles. Measurements
were performed to evaluate the rate of
desiccating water in presence of other
substances from quartz filters, which were
chosen as the media for collecting
;eondensible samples. —"-•* '*•—-*=—•-
Field Evaluation
The first field test of the condensibles air
dilution train was performed at the stack of
a wood chip dryer used in the production of
particleboard.Theparticulate control device
on this source is an electrically augmented
gravel bed filter. This stack was chosen
because it was expected to have predomi-
nately organic emissions. The CADT was
operated simultaneously and collocated with
multiple Method Strains operated by another
procedure, the impinger approach, for
condensible emissions.
Results
The common analytical procedure of
desiccatingto remove adsorbed or absorbed
moisture and weighing for determination of
sample mass was selected. The rate of
evaporation of substances condensed
during sampling and absorption or adsorp-
tion of moisture during weighing was con-
trolled by keeping the condensible filter in a
Teflon enyelope. The specified1ime_ot2!tJb_
for desiccation of the condensibles sample
provided adequate evaporation of moisture
but limited evaporation of test substances
which were! used to simulate stack
condensible emissions.
Measurements with solutions of sulfuric
acid indicated that the rate of evaporation of
water was sufficient for drying in 24 h at
which time the mass ratio of H O/H2SO4
approached that for water of nydration
causing the evaporation rate to become
very small. For samples containing 10 to 60
mg of H?SO4, the ratio of H O/H2SO4 after 24
h of desiccation was found to be in the range
of 27 to 43%. For sources where H2SO4 is
present, the analytical determination of the
mass of H2SO4 by extraction and t'rtration
can be used to obtain a correction for
-------
moisture retained by H2SO4 after desiccat-
ing the sample for 24 h.
In the field test the condensible emis-
sions measured using the dilution approach
ranged from 25.2 to 27.6 mg/dscm in five
sampling runs performed over three days.
On average the two approaches agreed;
the results from the dilution air approach
were 2% lower than the results from the
impinger approach. The standard deviation
of the difference was 15%. The precision of
the impinger approach in this field test was
3%, expressed as standard deviation of
replicate runs. The estimated precision of
the dilution approach was 15% in this field
test. Blank runs following the protocol for
the dilution approach indicate that 5% of this
variation of the dilution approach was the
result of errors associated with loading and
unloading filters. ••-•• "-«.---^.^-^;.-= j»-^---,",
Conclusions and
Recommendations
In the initial phase of development of the
dilution air approach for measurement of
condensibles, a sampling train was designed
and fabricated. Procedural details were
identified from analysis of the fundamental
nature of the dilution approach. Laboratory
studies of desiccation of simulated
condensible samples were conducted, and
the train and procedures were evaluated in
a field test. The results of the field test at a
site with predominantly organic condensible
substances gave reasonable results. The
close agreement with the impinger approach
and the fundamental differences in the two
approaches suggest that condensible
emissions forthe site tested were composed
of particulate matter that is not subject to
subtle changes.
The laboratory investigation of sample
analytical procedures for mass determina-
tion indicated that some materials may
evaporate significantly during desiccation
of the sample to remove adsorbed or ab-
sorbed water which is in excess of com-
bined water or water of hydration in the
following formula H.SO'2H2O. However,
these are limited to substances with volatility
'or equilibrium vapor concentrations greater
than or comparable to that of water. Also, it
was concluded that for excluding water of
hydration from the measured sample mass,
the most appropriate method was a mod-
erate correction for sample mass determined
gravimetrically based on the amount of
H2SO4 found through titration. EPA is con-
sidering condensible methodology named
the impinger catch (1C) approach in which
water of hydration in addition to H2SO4 will
be included as condensible emissions. If
the 1C approach is adopted, then inclusion
of water of hydration as condensibles will
likely become a part of the gas dilution
methodology. This would eliminate the titra-
tion step from the gas dilution procedures.
ft is recommended that further testing be
performed at other sources, particularly a
source expected to emit sulfates forthe next
test. Extended laboratory measurements
are recommended to determine if the rec-
ommended correction for water retained by
H2SO4 after 24 h of desiccation is accurate
for higher levels of initial water vapor con-
centration. In addition, although other hy-
grosoopicsubstances which may be present
in condensibles are expected to retain
substantially less water than H SQ4, mea-
surements should be performed to quantify
their retention of water underthe conditions
adopted in the procedures.
Multiple filter holders" should be fabri-
cated so that condensibles filters are loaded
and unloaded in the laboratory rather than
at the sampling site to minimize errors due
to handling. Some modifications of the
condensibles air dilution train (CADT) are
recommended. Improved thermal insulation
of the heated sample orifice meter is needed
to better maintain the condensibles filter at
the adopted reference temperature, 20°C.
In addition, improved mobility is needed for
traversing.
-------
William E. Farthing is with Southern Research Institute, Birmingham, AL 35255-5305.
Thomas E. Ward is the EPA Project Officer (see below).
77j0 complete report, entitled "Development of Sampling Methodology for Dilution Air
Sampling of Condensible Emissions from Stationary Sources,"(Order No. PB91-
129 742/AS; Cost: $15.00, cost 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 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
BULK RATE
POSTAGE & FEES PAID
: EPA
; PERMIT No. G-35
Official Business
Penalty for Private Use $300
EPA/600/S3-90/093
------- |