&EPA
United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
Research and Development
EPA-600/S7-82-036 Sept. 1982
Project Summary
Sampling and Data Handling
Methods for Inhalable
Particulate Sampling
Wallace B. Smith, Kenneth M. Gushing, Jean W. Johnson, Christine T.
Parsons, Ashley D. Williamson, and Rufus R. Wilson, Jr.
The report reviews objectives of a
research program established by the
EPA on sampling and measurement of
particles in the inhalable particulate
(IP) size range in industrial process
emissions, and discusses methods
and equipment that will be required. It
summarizes research at Southern
Research Institute to support the
development of techniques for mea-
suring and characterizing emissions in
the IP size range. Topics studied
include computer techniques for
analyzing cascade impactor data to
recover information on IP emissions
available from existing data; concepts
for maintaining isokinetic sampling
conditions at constant flowrates in
particle-sizing devices; the design and
use of cascade impactors, cyclones,
and elutriators as particle collectors
for IP sampling systems; and a stack
sampling system in which the sampled
gas is diluted under conditions simu-
lating those in stack plumes.
This Project Summary was developed
by EPA's Industrial Environmental
Research 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 U.S. EPA is considering air
pollution standards for emission of
inhalable particulate (IP) matter from
stationary sources. IP matter is defined
in terms of particle size, sincethe extent
of penetration of inhaled particles into
the lungs depends on their size.
Adequate characterization of a pollu-
tion source requires measurement of
stack or fugitive emissions from the
source and background levels in the
ambient atmosphere as well. The
concentration and particle-size distribu-
tion of the suspended particulate matter
and, in some instances, its chemical
composition and biological properties
must be determined.
In December 1978 a workshop was
held at Research Triangle Park, NC, to
discuss the IP sampling and analysis
problem and the EPA research program
on the subject. The workshop was
attended by consultants and other
investigators experienced in aerosol
sampling and characterization.
The consensus of those attending the
workshop was that no methods in use
were adequate for measuring the
amount of material in the IP size range
in process emissions. Thus they recog-
nized the need for information on
methods that could yield such data or
that could be adapted to yield it, and for
the research necessary in the develop-
ment of instruments and procedures for
that purpose.
This report contains recommendations
resulting from that workshop and
summarizes related research at Southern
Research Institute on the problem,
under contract with the EPA.
-------
Recommendations for the
EPA Research Program
The workshop participants recognized
that the greatest problem to be faced in
developing adequate measurement
methods would be the lack of time
available. They concluded that 3 to 5
years would likely be required before
reliable methods would be available for
obtaining the detailed data needed.
They also concluded that the only
practical course would be the use of
available survey techniques to obtain
less comprehensive data as soon as
possible while research programs were
initiated to develop more accurate
methods for future use.
Specific needs and problems recog-
nized at the workshop included the fact
that, in sampling ambient atmospheres
and fugitive emissions, inlets were
needed for the available samplers to
equip them for measuring mass con-
centrations in the IP size range.
The lack of realistic and rational
strategies for sampling fugitive emis-
sions was also seen as a serious
problem.
In sampling stack emissions, the lack
of cyclones and inertial impactors
calibrated for recovery of the IP size
range was noted. It was also recognized
that dilution and cooling sampling
systems were needed that could simulate
the behavior of stack plumes, especially
the behavior of condensable vapors.
Advanced techniques for extrapolation
of particle-size distribution curves to
larger particle sizes to yield additional
information on the IP size range were
noted as conceptually feasible but not
yet in use.
Recommendations
It was recommended that the EPA
adopt a standard mathematical perfor-
mance curve with specified tolerances
for sampling devices used to collect and
measure IP matter. Figure 1 isthecurve
which was actually adopted. Any newly
developed device would have to be
calibrated and shown to agree with the
curve, within the tolerances specified.
The shaded area was constructed by
plotting two log-normal curves through
the (13 /jm, 50 percent) and (17//m, 50
percent) points with geometric standard
deviations (crB)of 1.0 and 1.7, respectively,
and allowing 10 percent for wall losses
of small particles and 10 percent for
penetration of large particles.
Other tasks recommended for prompt
action were:
• Analysis of existing particulate
emission data, using the Southern
Research Institute curve-fitting
procedure and the University of
Minnesota modal analysis.
• Calibration of inlets for the hi-vol
and dichotomous samplers. This
would also include the development
of a total mass sampler (lo-vol)
with the same f lowrate and inlet as
the dichotomous sampler for use
in ambient measurements com-
paring total mass samplers with
the dichotomous sampler.
• A modified EPA Method 5 sampling
train was suggested that would
include in-stack cyclones with
suitable cut points (Dso values) for
covering the IP size range.
Research at Southern
Research Institute on the IP
Problem
Analysis of Cascade Impact or
Data
Because of the lack of information on
the concentration of IP matter at particle
sizes larger than about 10/um (the upper
limit of the particle-size range covered
by an impactor, i.e., the first-stage cut
point or Dso), a computer technique of
extrapolation of the impactor data to 15
A/m has been developed. The full report
discusses improvements in accuracy of
the extrapolation obtained by using a
first-order osculating polynomial for
fitting the cumulative mass curve
between the first stage Dso value andan
assumed maximum particle size.
The function is a third-degree poly-
nomial which uses the known charac-
teristics of the cumulative mass curve
for its solution over the range of particle
sizes. Tests of the technique on a
number of theoretical unimodal and
bimodal size distributions demonstrated
a high degree of accuracy in recovering
the true cumulative particle concentra-
tion. Figure 2 shows an example of the
extrapolation technique applied to
actual data in which the measured
distribution may have been distorted by
losses in a buttonhook nozzle or by an
improperly calibrated cyclone. It is
possible to suppress the fit through
upper stages to compensate for these
errors. From such studies, it was
concluded that the technique could be
used for recovering IP concentrations
from existing impactor data within a
4 6 8 10 20
Aerodynamic Particle, um
60 80 100
Figure 1.
Recommended specifications for collection efficiency of samplers of
inhalable particulate matter.
-------
factor of about three, even when no
information was available .on the type of
nozzle or precollector that was used. If
the effects of these devices are known,
the errors in the IP data are probably
within the experimental error of sampling.
Impactor data were also subjected to
a modal analysis in which the data were
fitted with multi-component log-normal
distributions by a simplex minimization
method. Data were extrapolated to 100
/urn particle diameter by fitting the
portion of the size distribution for which
data were available, Results were
similar to those obtained with the curve-
fitting polynomial procedure. IP concen-
trations could be estimated within a
factor of two or better.
Isokinetic Sampling by
Particle-sizing Devices
Isokinetic sampling of a gas stream is
necessary if a representative sample of
suspended particles, notably those
larger than 2 /urn in diameter, is to be
obtained. This presents the problem of
maintaining isokineticity in a sizing
device such as a cyclone or an impactor,
in which the particle-sizing characteristics
are a function of flowrate. Several
techniques and devices for solving this
problem are suggested in the full report.
-r/o1
ro4-,-
8
10
1
2nd Oso Omitted
•6
-------
EPA Source Assessment Sampling
System (SASS), operated at 185 l/min
and 204°C, and the EPA Fugitive
Ambient Sampling Train (FAST), ope-
rated at 5,282 l/min and 23°C.
For the actual field sampling systems
for IP measurement, it was decided to
develop trains using cyclones for the IP
sampling collector. This decision was
based on the fact that cyclones are
proven field sampling devices without
the damaging problems of particle
bounce or reentrainment. The only
particular design difficulty is the lack of
Thermocouple
Stack
Wall
Temperature
Controller
Remote Actuated
Valve I
Dilution Air
Sizing
Device
^Condenser
.and
1 Nozzle
Figure 4.
[Dry,
•er
Micro Processor
Figure 3. Cool gas recycle concept.
Theory (Rectangular)
Experiment
2 34 6 8 10 20
Aerodynamic Particle Diameter, um
30 40 60 80 100
Theoretical and experimental collection efficiencies for a horizontal
elutriator with rectangular cross-section, plate length 38.1 cm, average
gas velocity 70 cm/sec.
a reliable theory for cyclone performance.
Although many attempts have been
made to predict the behavior of cyclones
from their geometry, or at least to
describe their performance by empirical
mathematical expressions, it is still
necessary to calibrate them experi-
mentally with laboratory-generated
aerosols over a range of temperatures
and flowrates similar to those likely to
be encountered in field use.
Several commercially available cyclones
and some especially fabricated were
calibrated to provide a basis for design
of cyclones for use in an IP sampler. The
individual cyclones of the Southern
Research Institute's five-stage cyclone
system were thus calibrated as part of
this empirical design process.
Under laboratory conditions (temper-
ature 22°C, flowrate 28.3 l/min, and
particle density 1.0 g/cm3) the cut
points of the individual cyclones are 5.6,
2.1,1.4, 0.63, and 0.33um, as shown by
the calibration curve of efficiency vs.
aerodynamic particle diameter in Figure
5.
From design parameters developed in
the previous cyclone calibration studies,
geometries for the IP cyclones were
selected. Two cyclone systems were
designed, and evaluated: a cyclone to be
used as a precollector for cascade
impactors, and a system of two cyclones
(Figure 6) and a filter in series, to be used
as the primary system for measuring IP
and fine particle concentrations. Both
systems were designed for high-
temperature operation in process
streams. The precollector for impactors
and the first cyclone in the series train
had collection efficiency curves satisfying
the specifications for IP samplers
shown in Figure 1. The second cyclone
in the series train was designed to have
a Dso value of 2.5 ±0.5/um aerodynamic
at the flowrate required for the 1 5 /um
cut in the first cyclone.
To measure the paniculate formed by
condensation of volatile material as the
hot stack gases mix with ambient air, a
Stack Dilution Sampling System (SDSS)
was designed. The system was designed
to be used with the in-stack IP dual-
cyclone sampler. Figure 7 shows major
components of the apparatus. In opera-
tion, gas from the stream being sampled
is drawn through the IP cyclone sampler,
in which particles larger than 15/um and
those 2.5-15 um are removed in two
stages. The gas containing the fine
particle fraction (less than 2.5 /urn) and
the condensable vapors passes through
the dilution chamber, and through
-------
100
u
Uj
§
ti
•SJ
§
Q = 1.0ACFM
7 = 22°C
m Cyclone V
Cyclone IV
Cyclone III
Cyclone II
Cyclone I
\\ \
0.2
0.3 0.4 0.60.81.0 2.0 4 6
Aerodynamic Particle Diameter, Mm
8 10.0 20.0 30.0
Figure 5. Calibration curves for the five-stage cyclone system. Flowrate 1.0
ft3/min. temperature 22°C.
Sampling Nozzle
Probe
Cyclone SRI-X
Dso=15±
Cyclone SRI III
050 = 2.5 ±0.6/urn
Figure 6. Sketch of two-cyclone
system.
heated probe and sample line and is
introduced axially into the bottom of the
cylindrical dilution chamber. Here the
gas is mixed with dilution air to form a
simulated plume which flows up through
the dilution chamber, and through a
standard hi-vol filter, which collects the
fine particles and any new particles
formed by condensation. The design of
the dilution chamber simulates the flow
patterns and mixing times observed in
actual plumes.
A field version of the dilution system
was constructed to the following
specifications:
Active length of dilution tube, 1.22 m
(4ft)
Total height of sampler, 1.8-2.1 m (6-
7ft)
I.D. of dilution tube, 21.3 cm (8.4 in.)
I.D. of sample inlet tube, 4.27 cm
(1.68 in.)
Active dilution volume, 43.6 I (1.54
ft3)
Sample flowrate, 17 1/min (0.6
ftVmin), determined by cyclone cut
point
Dilution flowrate, 425 1/min (15
ftVmin)
Dilution factor, 25
Residence time, 6.2 sec Sample
velocity, 25 cm/sec at 150°C Dilution
air velocity, 19.8 cm/sec at 21 °C
The system was tested on flue gas
from a domestic furnace burning fuel oil
under controlled combustion conditions.
The results indicated that organic
chemical vapors condensed to solid
particles within the dilution chamber as
expected, at dilution ratios of 10 to 40.
Higher dilution ratios favored a larger
number of small particles.
Selected Bibliography
Gushing, K. M., J. D. McCain, and W. B.
Smith. Experimental Determination of
Sizing Parameters and Wall Losses of
Five Source-Test Cascade Impactors.
Environ. Sci. Technol. 13:726, 1979.
Johnson, J. W., G. I. Clinard, L G. Felix,
and J. D. McCain. A Computer-Based
Cascade Impactor Data Reduction
System. Report EPA-600/7-78-042
(NTIS PB 285433), Southern Research
Institute, Birmingham, AL, March 1978.
601 pp.
McCain, J. D., G. I. Clinard, L G. Felix,
and J. W. Johnson. A Data Reduction
System for Cascade Impactors. Report
EPA-600/7-78-132a (NTIS PB 283173),
Southern Research Institute, Birming-
ham, AL, July 1978. 44 pp.
Smith, W. B., P. R. Cavanaugh, and R. R.
Wilson, Jr. Technical Manual: A Survey
of Equipment and Methods for Paniculate
Sampling in Industrial Process Streams.
Report EPA-600/7-78-043 (NTIS PB
282501), Southern Research Institute,
Birmingham, AL, March 1978. 280 pp.
Smith, W. B., R. R. Wilson, Jr., and D.
B. Harris. A Five-Stage Cyclone System
for In-Situ Sampling. Environ. Sci.
Technol. 13(11): 1387-1392, 1979.
-------
Hi- Vol Impactor
Filter Assembly
Process Stream
Dilution
Chamber
Exhaust Blower
To Ultrafine
jl Particle Sizing
System (Optional)
Dilution Air
'Heater
Condenser
-To Heaters, Blowers
Temperature Sensors
To Orifice Ice Bath
Pressure Taps
Dilution Air
Blower
Flow, Pressure
Monitors
1 Main Control'
Figure 7. Diagram of stack dilution sampling system (SDSS).
'Wallace B. Smith, Kenneth M. Gushing, Jean W. Johnson, Christine T.
Parsons, Ashley D. Williamson, and Rufus R. Wilson, Jr., are with Southern
Research Institute, Birmingham, AL 35255.
D. Bruce Harris is the EPA Project Officer (see below).
The complete report, entitled "Sampling and Data Handling Methods for
Inhalable Paniculate Sampling," (Order No. PB 82-249 897; Cost: $22.50,
subject to change} will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, v'A 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
. S. GOVERNMENT PRINTING OFFICE: 198^559 -092/0505
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
PS 0000329
U S ENVIR PROTECTION AGENCY
REGION 5 LIBRARY
230 S DEARBORN STREET
CHICAGO IL 60604
------- |