EPA-600/2-76-157
June 1976
SAMPLING INTERFACE FOR THE QUANTITATIVE TRANSPORT OF AEROSOLS
Field Prototype
by
Madhav B. Ranade
I IT Research Institute
Chicago, Illinois 60616
Contract No. 68-02-1295
Project Officer
Kenneth T. Knapp
Emissions Measurement and Characterization Division
Environmental Sciences Research Laboratory
Research Triangle Park, North Carolina 27711
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
;^7
-<..-'.i v
AGENOC
-------
DISCLAIMER
This report has been reviewed by the Environmental Sciences Research
Laboratory, U.S. Environmental Protection Agency, and approved for pub-
lication. Approval does not signify that the contents necessarily re-
flect the views and policies of the U.S. Environmental Protection Agency,
nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.
11
-------
SAMPLING INTERFACE FOR THE QUANTITATIVE TRANSPORT OF AEROSOLS
Field Prototype
ABSTRACT
Deposition of particles in conventional sampling probes
precludes their use with continuous monitors. The transpiration
of air through the probe wall has been shown to be effective in
significantly reducing deposition during the quantitative trans-
port of particles up to 50 ym. The purpose of the present con-
tract was to develop the transpiration concept into a field-
usable prototype.
A sampling nozzle with a 90° bend and incorporating a porous
wall tube through which clean air could be transpired was de-
signed, fabricated, and tested in the laboratory with model
aerosols. A coordinated testing program for the nozzle and the
experimental probe was also conducted.
Based on the laboratory test results, the final prototype
was designed and fabricated. The prototype consisted of a samp-
ling probe and a control box housing necessary air supply and
auxiliary instrumentation.
The prototype was evaluated in the field at two sites at
the Will County Power Station of Commonwealth Edison. The first
site represented a coal-fired power plant emissions following a
hot electrostatic precipitator. The second site represented a
coal-fired power plant emissions following a wet scrubber. The
test results show that the prorotype is efficient in transporting
iii
-------
the particles in these effluents. Transport efficiencies greater
than 95% were obtained with transpiration air to sample flow
ratio of 2 to 1 for the front section of the probe.
iv
-------
TABLE OF CONTENTS
Page No
List of Figures and Tables vi
INTRODUCTION 1
CONCLUSIONS 4
RECOMMENDATIONS 5
DESIGN OF THE 90° BEND 6
LABORATORY TESTING 11
Testing of the 90° Bend 14
Coordinated Testing 16
DESIGN OF THE SAMPLING INTERFACE 21
Sampling Probe 21
Sampling Nozzles 23
90° Bend Based on Transpiration
Principle 23
Straight Section of the Probe 26
Control Box 26
Operation of the Probe 29
FIELD TESTING OF THE SAMPLING INTERFACE 32
Sampling Operation 32
Site No. 1 -- Hot Precipitator Exhaust . . 32
Site No. 2 -- Wet Scrubber Exhaust .... 38
Discussion of Results 39
Effluent Characteristics 39
Sampling Efficiency 39
APPENDIX A -- OPERATION MANUAL, Transpiration
Sampling 45
v
-------
LIST OF FIGURES AND TABLES
Figure Page No
1. Conventional Sampling Probe 7
2. Construction Details of the Porous Bend ... 9
3. Front Section with 90° Bend 10
4. Photograph of the Aerosol Generator 12
5. Laboratory Test Facility 13
6. Facility for Testing the 90° Bend 15
7. Flow Diagram of Prototype Sampling
Interface 22
8. Sampling Nozzles 24
9. Details of the Front Section 25
10. Distribution of Transpiration Air for Rear
Box 27
11. Schematic Diagram of the Control Box .... 28
12. Sampling Port at Site No. 1 33
13. Sampling Arrangement 34
14. Sampling Arrangement Using Transpiration
Probe 37
15. Photomicrographs of Effluent Particulates
(420X) 42
Table
1. Performance Specifications for Sampling
Interface 3
2. Results of Mass Preservation Tests with the
90° Bend 17
3. Coordinated Testings 19
4. Recommended Transpiration Rates 31
5. Sampling Conditions at Site No. 1 ..... 36
6. Sampling Conditions at Site No. 2 40
7. Size Distributions of Collected Samples ... 41
vi
-------
SAMPLING INTERFACE FOR THE QUANTITATIVE TRANSPORT OF AEROSOLS
INTRODUCTION
In the sampling of particulate emissions, the aerosol
must be transported from the source to the sensor with a mini-
mum of deposition, agglomeration, and reentrainment enroute.
Particle losses to and reentrainment from conventional probe
walls can be excessive.
The Fine Particles Research Section of IIT Research
Institute developed a sampling interface which permits trans-
port of aerosols without modification of the total mass rate
and size distribution of the source aerosols. This program
was sponsored by EPA under Contract No. 68-02-0579. The
probe consists of a porous metal tube encased in a manifold
through which transpiration air is passed inward to provide
a moving clean air sheath that minimized particle deposition
on the walls. The efficiency of the probe to transport an
aerosol ranging in size from 0.05 to 50 ym was demonstrated
in a statistically designed test program. The results of this
program are available in the final report.
The purpose of the present program was to apply the
knowledge gained in the development of the experimental probe
to a field operable sampling interface. The experimental
probe consisted of only a straight run of the encased porous
1. Sampling Interface for Quantitative Transport of Aerosols.
Prepared by IITRI for EPA, #EPA-650/2-74-016 (1973).
1
-------
tube. In order to sample out of a stack, the probe must
include a 90° bend. The standard gooseneck nozzle used with
conventional probes suffers from deposition like the conven-
tional probes. Extending the porous tube and transpiration
idea to the bend was considered to be most effective in pre-
venting deposition. Under this program, a 90° bend using the
porous internal tube was designed and fabricated. It was then
tested to determine its efficiency in transporting the particles
in the range 0 to 10 ym. The bend was tested separately, as
well as attached to the experimental probe.
A final design of the prototype sampling interface con-
taining the probe, an air supply system for transpiration, and a
control box was developed to meet the specifications listed in
Table 1. The bend was fabricated and the experimental probe
was modified to withstand temperatures upto 700°F and to meet
the other specifications in Table 1.
The probe was tested in laboratory and in the field. The
sampling sites consisted of two stacks of a coal-fired power
plant. One of the stacks followed a hot electrostatic precipi-
tator; and the other stack followed a wet scrubber.
The laboratory and field tests showed that the sampling
probe could be used for several hours with insignificant
deposition of particles in the lines.
-------
Table 1. PERFORMANCE SPECIFICATIONS FOR SAMPLING INTERFACE
Aerosol concentration range
Aerosol size range
Sampling rate
Sampling temperature
Sampling probe
Sampling nozzles
Sampling requirements
Transpiration rate
102-108 particles/cm3
0.05-10 ym
7.1-28.3 1pm (0.25-1.0 cfm)
Ambient to 300°C (572°F)
4
1.29 cm I.D. (1/2 in. I.D.) x
^180 cm (6 ft) long
0.63, 0.95, and 1.29 cm I.D.
(1/4, 3/8, and 1/2 in. I.D.)
Isokinetic sampling
90° bend
Minimum diameter of sampling
port compatible with the
probe -- 10 cm (4 in.)
Probe to be heated to at least
150°C (300°F) to prevent
water condensation
Up to 142 1pm (5 cfm)
-------
CONCLUSIONS
The concept of the transpiration probe was extended to
design and fabricate a sampling interface that could be used
to transport particulate matter from stacks to a measuring
device with minimal deposition. In this program, the following
significant results were obtained.
1. The transpiration concept was extended to form a
sampling nozzle with a 90° bend. This design significantly
reduced the loss of particles compared to the standard goose-
neck nozzle. The extent of deposition was dependent on the
transpiration rate in the nozzle section and was reduced as the
transpiration rate was increased.
2. Due to the high transpiration rate in the front
section, deposition in the rear section was significantly
reduced. Consequently, less transpiration rate was required
in the rear section.
3_._ The laboratory interface was developed into an inter-
face usable in actual stacks having temperatures of up to 371°C
(700°F).
4. The field prototype was evaluated at two sites and
showed that it can be used to efficiencly transport the particles
uptp 60 ym in size.
5_._ The tests with the scrubber effluent showed that the
interface could be used for sampling in the presence of water
droplets.
-------
RECOMMENDATIONS
Further testing of the sampling interface should be aimed at
adapting the interface with continuous monitors. In addition,
the interface has a potential for conditioning of the parti-
culate samples in a controlled manner to study the interaction
of the particulate matter with the atmospheric components.
Development of a flow measurement device that can be placed
inside the sampling nozzle will avoid errors in calculating the
sample flow rate by difference in the transpiration and the
total flow rates. Such a flow measurement device should not
hinder the flow of the sample. Use of a null type sampling nozzle
should also be investigated to achieve the accuracy in the sample
flow calculations.
-------
DESIGN OF THE 90° BEND
In conventional sampling situations, the sampling probe
is inserted into the stack through ports in the wall. The
ports are usually 10 cm (4 in.) diameter circular openings.
The sample flow must be in the direction of the stack flow.
he sample stream, therefore, must be turned by 90° in the
sampling probe. Many .conventional probes use a gooseneck
nozzle shown in Figure 1. Considerable loss of the particulate
sample occurs in this type of nozzle. Minimizing the deposition
losses in the nozzle should result by extending the idea of the
internal porous tube to the 90° bend.
Fabrication of the 90° bend with the porous inner tube,
such that the entire front end of the probe would pass through
a 10 cm (4 in.) port, was a challenging problem. Attempts at
bending the porous tube were unsuccessful even for a large radius
bend. One approach for obtaining the bend was to join tube
pieces to form an arc. The 90° bend would be formed by welding
the pieces together. This may result in some blind areas around
the joins where the transpiration air would not reach. However,
this arrangement would certainly be an improvement over the
present conventional gooseneck bends. The number of pieces used
to make the bend is also important. If only a few pieces were
used, the aerosol flowing in the bend would have rather sharp
turns. On the other hand, if too many pieces were used, the
area available for the transpiration flow would be greatly
reduced.
6
-------
0)
n
M
•H
-------
We decided to use 10° segments for making the bend as
shown in Figure 2. For preliminary experiments, the sections
could be joined by gluing them. For final prototype, the
sections were welded.
The 90° bend section was designed to have an independent
transpiration air supply. In the earlier program a higher
rate of transpiration in the rapidly established the air
sheath around the sample, and reduced the deposition at the
entrance, To achieve this feature, the separateness of the
front and back section transpiration air was essential. In
addition, the nominal pore size of the porous tube used for
the 90° bend was chosen 5 ym instead of the 2 ym nominal pore
size of the back section porous tube. The assembly of the front
section is shown in Figure 3. The sampling nozzles were inter-
changable. Coupling the front section to the rear section was
achieved with minimum loss of transpiration air, so the clean
air sheath around the sample stream would be preserved.
-------
Porous Tube
5 3/4"-
0.458'^
10£
A
T
Cutting Lines
10C
Figure 2
CONSTRUCTION DETAILS OF THE POROUS BEND
-------
CO
0)
o
O
a
H
M
§
H
H
U
W
to
§
10
-------
LABORATORY TESTING
The sampling probe was tested in the laboratory with a
uranine aerosol. The uranine aerosol was generated according
to the procedure described in the previous final report.
Uranine powder in the 5-50 ym size range was dispersed in an
aerosol form from the generator shown in Figure 4.. Part of
the aerosol was sampled by the probe section in the test set-
up sketched in Figure 5.
For some of the tests, the upper size of the aerosol was
limited to 10 ym with a small cyclone used to remove larger par-
ticles before sampling by the probe section.
As described in the previous final report, the effects
of the particle size and the transpiration air on the deposi-
tion of particles in a 1.29 cm (1/2 in.) I.D. and 176 cm
(70 in.) long porous tube (2 ym nominal pore size) were
studied extensively. However, the main areas where further
experimental verification was needed were the deposition and
effect of transpiration air in the 90° bend front section;
and the effect of varied transpiration rates in the front and
rear sections.
11
-------
Figure 4
PHOTOGRAPH OF THE AEROSOL GENERATOR
12
-------
0)
^
li.
1-3
M
m
0) CO
)-i W
2 H
O
H
PQ
3
13
-------
TESTING OF THE 90° BEND
The set-up for the tests is shown in Figure 6. The test
section was followed by a 10 cm (4 in.) glass fiber filter.
The transpiration air was supplied through rotameters (Fig-
ure 5). The total flow rate through the section was monitored
by noting the pressure drop across the laminar flow element.
The details of the test procedure are described in the pre-
vious report.
To begin the test, the transpiration flow was first
established at the desired level. The aerosol generator was
truned on. Flow through the laminar f]ow element was adjusted
so that
Total flow rate = transpiration air flow rate
+ sample flow rate
The sample flow rate was checked at the beginning of the
experiment with a wet test meter. The aerosol was sampled
for a period sufficient to obtain a sample for analysis. The
probe section and the filter were washed and the amount of
uranine deposited was determined by a colorimetric method.
The percent deposition values were calculated by the
relation
J0 Deposition
_ . mass of uranine in probe wash
mass of uranine in (probewash+filter)
For comparison, tests were also made to determine the
extent of deposition in a standard 1.29 cm (1/2 in.) gooseneck
nozzle and a bent impervious copper tube having the same
14
-------
1
o 3
CD CX<
-------
internal diameter of 1.29 cm (1/2 in.) and having the same
radius of curvature and length as the 90° bend with the porous
tube. The results are summarized in Table 2.
The 90° bend, based on the transpiration principle, was
effective in reducing the deposition significantly compared
to both the impervious 90° bend of the same curvature and the
standard gooseneck nozzle.
The extent of deposition of large particles is controlled
by gravity and, as expected, the deposition increases with
increasing sample velocity for the impervious bend and the
gooseneck nozzle. In the porous bend at low levels of trans-
piration velocity, deposition was fairly high. At higher
levels of transpiration velocity, the deposition was signi-
ficantly reduced at the sample rates of 14.2 and 28.3 1pm
(0.5 and 1 cfm, respectively). However, deposition was still
high at the low sample rate of 7.1 1pm (0.25 cfm), indicating
that the gravitational settling was not completely overcome.
COORDINATED TESTING
Deposition in the entire probe (with both the front and
rear sections) was also experimentally studied. This was
necessary to assess the effect of the varied transpiration
rates in each section on deposition. The tests were conducted
with 0-10 ym uranine powder. Since the data on the straight
section, as obtained in the previous program, were based on
50 ym particles, few tests with 0-10 ym particles were con-
ducted with the straight probe only for comparison. The
16
-------
Q
P^
W
pg
o
0
^
f-T-
D-
ffi
H
M
^
CO
H
CO
£
K^
O
M
EH
<
£>
Pi
r^
CO
t:
CO
CO
<;
t — A
d
CO
H
i-J
!z)
CO
«
CM
CO
i— 1
CO
H
G
O ^>
P CU QJ
6< >H pi G r-l
CO -H <3J N
O co N
ft O O
CU O G
O bO
G W
O D
•H O
4-J "t-UD
>-2 -rl Pi > T3
co -H IM p|
O CU OJ
&1 ft"0
P -H
C
O cc
•H 3
-u o -a
B-3 -H M pj
CO O ^
cO 4-J O
M -H 0)
•H CJ CO
ft O --,
co — i g
fi CO 0
CO >
EH
0
•r-l -s-\
P cu S
CO 4-J M-l
M CO G
ft
CO ^ 0
c o ft
CO r-( rH
$-\ MH
H
O O
.
CM CM
r-
i— 1 vO
CXD <|-
f^v f--^ , — )
CM CM
r — co r^- co
i — | LT)
.
o o o o
\ — • x_x > — ^ v^^
CO CO
OO CM CO CM
.
CM — *
6 ft
CO r-l
CO
CO CO v£> v£>
CT\ O\ CO CO
rH i— I
£" in x-N x-v
CM CM LT> LO
•
O O O O
^~S ^*-S ^^ N '
rH i— 1 CM CM
< c
r- r--,
VO O
,
C*-) t j
CM
r~» co
>d- co
CM
i— 1 LT)
« c
O O
co
OO CM
CM
-------
results are presented in Table 3.
The results for the straight probe show that the deposi-
tion of 0-10 ym particles is low even with a very low trans-
piration air. At all sample rates, the results showed a
similar trend of deposition in the two probe sections.
Deposition in the front section was not affected when the
front transpiration air was increased from 14.2 to 28.3 1pm
(0.5 to 1.0 cfm). Deposition in the rear section was also
similar for transpiration rates of 142 and 283 1pm (5 and
10 cfm).
The particle size of the aerosol in the coordinated tests
were 0-10 ym, as compared to 0-50 ym for the tests with bend
alone. The deposition in the front section at a flow rate of
14.2 1pm (0.5 cfm) was higher than corresponding results for
the bend alone (Table 2). These results are consistent with
the findings of the previous program. Deposition of 50 ym
particles was controlled by gravity and entrainment effects,
and decreased as the sample and transpiration rates were
increased. Deposition of intermediate size particles first
decreased and then increased slightly as the transpiration
rate was increased.
The results of the tests with 90° bend are informative
for mostly the large particles, which contribute to the mass
in the samples. In other words, at 14.2 1pm (0.5 cfm) and
higher transpiration rates, the large particles are trans-
ported better than the intermediate sized particles. As
pointed out from the size distribution data in the previous
18
-------
Table 3. COORDINATED TESTINGS
Sample,
1pm (cfm)
0-10 um Part
7.1 (0.25)
7.1 (0.25)
7.1 (0.25)
14.2 (0.5)
14.2 (0.5)
14.2 (0.5)
28.3 (1.0)
28.3 (1.0)
Transpiration
Front ,
1pm (cfm)
.ides (1-10
14.2 (0.5)
14.2 (0.5)
28.3 (1.0)
14.2 (0.5)
14.2 (0.5)
28.3 (1.0)
14.2 (0.5)
28.3 (1.0)
Test with Straight Probe
14.2 (0.5)
28.3 (1.0)
28.3 (1.0)
Back,
1pm (cfm)
mg/lit):
142 (5)
283 (10)
142 (5)
142 (5)
283 (10)
142 (5)
142 (5)
142 (5)
. Only:
142 (5)
142 (5)
28.3 (1)
Deposition, °/a
90" Bend
4.2
5.1
4.4
6.2
2.3
5.6
4.6
5.0
...
Straight
0.5
0.8
0.4
2.4
1.1
4.3
0.4
1.2
0.86
0.64
0.83
Filter
95.3
94.1
95.2
91.4
96.6
90.1
95.0
93.8
99.14
99.36
99.17
19
-------
program, even this level of deposition for the intermediate
sized particles does not affect the size distribution of the
sample in a significant manner. Moreover, if only intermediate
sized particles are present, the effect on size distribution
will be minimal.
20
-------
DESIGN OF THE SAMPLING INTERFACE
Based on the results of experiments under the present
and the previous programs, the sampling interface was designed
to meet the specifications listed in Table 1, In addition,
the design specifications for larger aerosols (up to 50 ym)
to be transported were also developed so that the fabricated
prototype could be used at higher particle size ranges without
substantial modifications.
In the design of the prototype, attention was given to
simplicity of operation, ease of handling, and adaptability to
coarser aerosols. The sampling interface consists of a
sampling probe and the self-contained control box, A schema-
tic drawing of the interface is shown in Figure 7. These
components are described below.
SAMPLING PROBE
The sampling probe consists of a 1.9 cm (3/4 in.) O.D.
porous walled tube encased in another solid tube through
which the transpiration air is distributed. The probe con-
sists of three parts: sampling nozzle, a front section with
a 90° bend, and a straight section Transpiration air in the
90° bend and the straight section is supplied independently.
A discussion of the components is given below
21
-------
0*
?. g ^
o E
a»
^ s
«--
,P
*.-
0»
^i
'O
(5
I
U
u
H
s;
M
t.o
t~.
[-•I
o
O
U
**
F
o
.4
•*!
"cs
X?
A:.
1
I
O
in
eu
'
-------
Sampling Nozzles
Three 316 S.S. nozzles shown in Figure 8 were fabricated.
Nozzles with 0.65, 0.97, and 1.29 cm (1/4, 3/8, and 1/2 in.,
respectively) diameter openings were chosen for allowing
isokinetic conditions in most stack sampling operations. The
nozzles were designed with a blunt taper so that the probe
assembly could he it.serted into a 10 cm (4 in.) diameter
opening. Nozzles with gentler taper may be fabricated for
situations where the opening is considerably larger. However,
the shortness of the nozzle has the advantage that only a
small area of the probe assembly is without the transpiring air,
90° Bend Based on Transpiration Principle
The porous stainless steel tube could not be bent in a
single piece. In addition, the bend had to be of such a
curvature that the bend and the attached nozzle could be in-
serted through a 10 cm (4 in.) diameter opening. The 90°
bend was accomplished by joining 10° segments of straight
tube, as shown in Figure 9. The assembly of the porous tube
and the outer tube is also shown in Figure 9. Transpiration air
is supplied through a 0.63 cm (1/4 in.) diameter tube. The
individual segments were welded together. The outer case
was welded to the inner tube, as shown in Figure 9. The
sampling nozzle with 0.62, 0.95, and 1.27 cm (1/2, 3/8, and
1/2 in., respectively) openings could be attached simply, as
shown in Figure 9. A metal gasket is used to seal the nozzle-
bend joint. A screw-on socket joint is used to connect the
front and the back section.
23
-------
T
o
X
^
ro
T
ro
1.
CVJ
T
00
txO
•H
CO
W
h4
N
N
O
CO
24
-------
0)
W)
•H
o
M
H
U
w
CO
o
ft
o
oo
H
H
Q
25
-------
Straight Section of the Probe
The straight section consists of a 1.9 cm (3/4 in.) diameter
and 178 cm (70 in.) long porous tube encased in a 3.2 cm
(1-1/4 in.) O.D. tube. At the middle of the section, a 5 cm
(2 in.) diameter by 61 cm (24 in.) long tube is used for dis-
tribution of transpiration air, as shown in Figure 10. A
mating screw-type joint is used to join the front and the
back sections.
CONTROL BOX
A control box for housing the auxiliary components and
controls was designed. A flow drawing is shown in Figure 11.
A combination vacuum pump and compressor unit was chosen to
provide both the supply of the transpiration air and the
suction for sampling.
Transpiration air is filtered at the pump intake. The
air is pumped through a rotameter with a capacity of 0-470 1pm
(0-1000 cfh). A bleed line before the rotameter is used for
varying the flow through the rotameter. The air from the
rotameter flows through an in-line heater. The air leaving
the heater is divided into two transpiration air streams for
the front and the rear sections. Tha air for the front sec-
tion flows through another flowmeter and in supplied to the
front section through Teflon lined flexible hoses with a
glass wool filter at the end. The air for the rear section
is routed through a ball valve and is supplied to the rear
section through another Teflon lined hose.
26
-------
0>
0>
E
*-
0)
U
o
60
•H
w
w
prf
g
Pi
pi
o
o
M
H
en
O
M
H
P
PQ
I—I
Pi
H
C/3
27
-------
(1)
50
•H
g
PQ
O
Pi
H
S
C
U
o
V*
c
o
o>
c
o
c
o
T
I
ffi
U
CO
28
-------
The combined sample and transpiration air is pulled by
the vacuum pump through a laminar flow element. An Inclined
oil manometer is used to measure the pressure drop across the
laminar flow element, which is calibrated for volumetric flow
rate against the pressure drop. A protective glass wool filter
is used to keep the laminar flow element clean. The total flow
rate of the sample and the transpiration air is variable by
means of a needle breed valve.
Dial thermometers and pressure gauges are used at appro-
priate points to characterize the air flow. Details of the
componetics are discussed in the manual, which is enclosed
as Appendix A to this report.
OPERATION OF THE PROBE
The operating procedure of the probe for fixed rate and
isokinetic sampling is given in Appendix A. In general, the
sample flow rate should be chosen first. Experimental results
show that a higher sampling rate is generally preferable.
While sampling isokinetically, the sampling nozzle should be
chosen such that the volumetric flow rate is the highest possible.
A sampling rate of 28.3 1pm (1 cfm) gave good results in our
experimental program. A sample flow rate should be above 7.1 1pm
(0.25 cfm) is recommended whenever possible.
The transpiration flow rate to be used is determined by
the particle size and the sample flow rate. When the particles
are below 10 ym the front transpiration air rate should be above
14. 2 1pm (0.5 cfm) and the rear transpiration air rates
29
-------
should be at least 28.3 1pm (1 cfm). For larger particles
upto 50 ym, the transpiration air rates may be above 14.2 1pm
(0.5 cfm) for the front and above 142 1pm (5 cfm) for the
rear. As the sample rate is increased from 7.1 to 28.3 1pm
(0.25 to 1 cfm), the amount of transpiration air required is
reduced. However, the front transpiration air should always
be kept above 14.2 1pm (0.5 cfm).
Recommended sets of sample and transpiration rates are
listed in Table 4. The total flows are kept below 142 1pm
(5 cfm) whenever possible so that the pump in the prototype
can handle the flow. For larger particles (50 ym), a larger
pump may be used. The laminar flow element may be used for
measuring the total flow rates upto 710 1pm (25 cfm). If
the sample flow rate is kept above 14.2 1pm (0.5 cfm), a front
transpiration rate of 28.3 1pm (1 cfm) and a back transpira-
tion rate of 85 1pm (3 cfm) will be satisfactory for any stack
sampling situation. It is also convenient to use fixed values
of transpiration rates, thus minimizing manipulation.
30
-------
Table 4. RECOMMENDED TRANSPIRATION RATES
Sample flow rates,
1pm (cfm)
<7.1
(0.25)
7.1 - 14.2
(0.25-0.5)
14.2 - 28.3
(0.5-1.0)
Particle size, ym
upto 10
upto 50
upto 10
upto 50
upto 10
upto 50
Transpiration air rate
Front ,
1pm (cfm)
28.3 (1.0)
56.6 (2,0)
28.3 (1.0)
56,6 (2.0)
28.3 (1.0)
56.6 (2.0)
Back,
1pm (cfm)
85 (3,0)
283 (10)
85 (3.0)
142 (5.0)
85 (3.0)
85 (3.0)
31
-------
FIELD TESTING OF THE SAMPLING INTERFACE
The prototype sampling interface was tested at a coal-
fired power plant. Two sites were selected for sampling.
The first site was located at an exhaust duct following a hot
electrostatic precipitator and leading to a stack. The second
site was the exhaust duct following a wet scrubber and leading
co a stack. Both sites were located at the Will County Power
Station* of Commonwealth Edison, who provided cooperation in
the evaluation program. The objective of the field sampling
program was to compare the Method 5 train and the prototype
interface and to study the variation of operating parameters
of the prototype.
SAMPLING OPERATION
Site No. 1 -- Hot Precipitator Exhaust
The No. 3 boiler unit at the plant is equipped with a
^5)
Research Cottrell hot electrostatic precipitator. The pre-
cipitator handles 42000 m3/min (1,400,000 cfm). The operating tem-
perature is 288°C (550°F). The sampling was conducted on a
1.2 meter x 4 8 meter (4 ft x 16 ft) duct through a 10 cm ID
(4 in. nominal pipe) port shown in Figure 12. The sampling
set-up is illustrated in Figure 13. The first series of tests
Located in Romeoville, Illinois
32
-------
O
55
w
H
Pi
•rl O
O
PH
33
-------
-------
consisted of a velocity traverse arid a ninple -point isokineti'-.
sampling by a Hethod 5 train Since the objective of '~.\.<: eval-
uation was only to compare the method 5 train and the proto-
type interface, single point sampling was considered to be
adequate
After the velocity traverse was made, appropriate sampling
conditions were determined by the standard nomograph for the
Jiethod 5 train The sampling point was so chosen that it
reflected the average conditions in the stack Sample was taken
for half an hour. At the end of the sampling period, the probe
sections were washed with acetone. The deposit in the probe
tip (1/2 in. gooseneck nozzle) and the probe were determined
separately by weighing the washings after drying in an oven
The deposit on the 10 cm (4 in.) glass fiber filter was weighed
on a microbalance Results are shown in Table 5. Parts of the
collected samples were dispersed and mounted on slides for
size distribution determination by optical microscopy
The second series of sampling consisted of sampling with
the prototype sampling interface. For the first test, the
sampling point and rate were the same as with the EPA Method 5
tests A transpiration rate of 193.2 1pm (4 cfm) was chosen
The test layout is shown schematically in Figure 14.
To begin the test, the transpiration flow was started arid
maintained at 193.2 1pm (4 cfm) The transpiration air flow
divided equally between the front and the rear sections. The
probe was inserted into the stack and kept with the nozzle
pointing in the reverse direction to the flow for 10-15 minutes.
-------
O
s
w
EH
H
C/3
5
CO
&
O
H
H
H
Q
s
O
U
O
^
co
LO
0)
1—I
-3
H
fc.n
OJJ
4-1
•rH
co
O
P.
4-1
bO
•rH
cd
!H
4-1
0)
,p
O
*
S
O
•H 0
4-1 P,
Ctf rH s~\
V-i &
• r-l •» M [
p, co o
CO CU ^
r> 4_)
cd cd
faO
c
•H -
rH CO
0,0
0-H
CO
bO
C
•H -
i-H
&
4-1
£
O
£
a
•H
0
/^^
0
MH
o
S-X
OCTi OOO
O x-v LO x-s i— 1 x-s rH x-s CO x-s
CM6-S OB-2 OB-2 O B-2 r^ B-2
LOCN Of~ O<1" OvO OcM
O • O • O • O • O •
• OO • O • O • O • rH ^> M K^ JH ^» V-)
~
-------
(U
J-l
m
txO
•H
W
PQ
O
o
H
H
CO
H
M
CO
H
S
W
W—(
W
a
M
i-J
CO
37
-------
The heater was operated so that the transpiration air temperature
leaving the control box was 204°C (400°F). The probe was
straightened, so that it was in line with the flow, and samp-
ling started. The sampling rate was adjusted to the desired
level and sampling continued for 30 minutes.
For the second test, the sampling conditions were kept
the same, but the sampling period was extended to four hours
to evaluate continuous operation. For the remaining tests,
a sampling rate of 28.3 1pm (1 cfm) was chosen and the trans-
piration rate was varied.
The deposits on the front section (with 90° bend) and
the rear section were washed separately by acetone and weighed
after drying. The deposit on the 10 cm (4 in.) filter was
also determined by weighing. Part of each deposit sample was
used for size distribution determination by optical microscopy.
The sampling conditions and results are given in Table 5.
Site No. 2 -- Wet Scrubber Exhaust
The No. 1 boiler unit at the plant is equipped with two
Babcock and Wilcox limestone slurry scrubbers for SOj and
3
particulate removal. The scrubber handles 21,000 m /min
(770,000 cfm); and the temperature is 82°C (180°F).
The scrubber represented a different type of source from
the hot precipitator due to the expected larger particle size
and presence of water. The transpiration air could be used
to evaporate the water by lowering the partial pressure of water,
thereby preventing condensation in the sampling equipment.
-------
Two tests were performed with the prototype interface.
The same conditions were used for each run. The samples were
handled in the same manner as in the other tests. No evidence
of condensation was found on the sampling line or the filter.
The results are summarized in Table 6.
DISCUSSION OF RESULTS
The results of the sampling operations are given in
Tables 5, 6, and 7. Tables 5 and 6 represent the efficiency
of particle transport by mass, while Table 7 reflects the
size selectivity of the sampling operation.
Effluent Characteristics
The samples collected on the glass fiber filter at both
sites were examined by optical microscopy (using polarized
light and ^ 65-600X magnification). At site No. 1, which
followed an electrostatic precipitator, the filter deposit
consisted of glass spheres upto 10 ym, FeaOs particles upto
20 ym, magnetite spheres upto 10 ym, coal particles upto 24 ym,
and metal shavings upto 40 ym. A typical sample is shown in
Figure 15. At site No. 2, which followed a wet scrubber, the
particle size was usually high. The effluent consisted of
coal particles upto 100 ym, partially burnt coal upto 370 ym,
mineral particles upto 90 ym, Fe20s upto 40 ym, magnetite and
glassy spheres upto 30 ym, and small glass spheres less than
1 ym (< 30% by weight).
Sampling Efficiency
The results of the sampling operations show that the
prototype interface can be used for sampling from stacks with
39
-------
CM
CO
cd
E-t
bO
~
4->
•H
CO
O
P,
0)
4J
r|
5-i
CO
4-1
I-l
•H
H-l
bO CO
•H
cd
5-i
4_>
CO
CO
n
o
5-i
fa
PJ
O
•H B
4J Cu
Cd rH X-N
5-4 B
CX CO CJ
en co ^
Pi -&-i
cd cd
t i t i
H
bO
fl
•H -
i-l CO
o S
B-3
cd 4-1
CO
bO
C
H -
-H co S
e>4-> a
B cd i — i
cd 5-1
bO
PI
•H
i — 1
I"
cd
CO
4-1
CO
QJ
H
_O
O
5-1
d
d
•H
4-1
t_|
cd
CO
PH
4J
C
O
5-1
h
C
•H
E
x-s
B
o
^
CO
o
•H
CD
""O
.
O
^
LO x— \ C3> /— s
CN B-S vo B-S
CTi LO CTv VO
0 • 0 •
. v.0 • vO
O O> O ^
^D X"N LO ^™S
i-l B-S 0 B-S
O O O LO
0 • O •
•rH -O
0s—- o v~x
~d" X-N O X~ \
CNS-2 on^S
O LO O CT>
0 • 0 •
• CN • CN
O v-r O "^
vO X~N ^O x™s
• O -0
VO • VO •
LO CN LO CN
V-X V-X
VO x-\ VO x-s
. 0 • O
vo • \^o •
LO CN LO CN
s_x ^s
0 O
VO *sO
CO x^\ CO x*\
• 0 • O
OO • CO •
CN i— 1 CN i— 1
N— ' V_X
CO CO
0 O
co cd co cd
CliMH p,U-)
>, 5-1 >i 5-1
4J OJ 4-1 CO
O 4-J O 4-)
4-1 fi 4-1 P!
OH O H
ti tt
P_j pi
*o r**»
4-)
•H
bO
C
•H
T3
cd
O
CO
4-1
cd
o
•H
4-1
5-1
cd
40
-------
a)
r-l
^Q
cd
H
a
cu
•H
ti
cd
&
4-1
5-1
cu
rH
rH
i
co
6-S
CU
j>
•H
4J
cd
rH
3
O
CU
rH
&
cd
CO
C
O
•H ~x^
4J CU 6
cd 4J M-l
r-l Cd 0
,_J t_i >*_x
o,
w £ 8
coo.
Cd rH rH
r-l LW
H
CO •
rH I-J CU g
0,0 4J CX
B rH Cd iH
cd U-l }-l
CO
VO
LO
-cj-
°
-------
Site No. 1
Site No. 2
Figure 15
PHOTOMICROGRAPHS OF EFFLUENT PARTICULATES (42OX)
42
-------
very little deposition. The Method 5 train transported only
58.7% of the particles to the glass fiber filter at the
site No. 1. The particles that were transported to the glass
fiber filter were extremely fine (less than 3 ym), indicating
that the larger particles were lost in the nozzle and the
probe. Size distributions of the nozzle and probe deposit
show that the larger particles were depo^'^ed on the probe
wall and the nozzle.
In test No. 1, with the Method 5 train, the level of
deposition in the sampling nozzle and the probe were 13.1 and
28.2%, respectively. As seen in Table 5, for the same sample
flow rate in test No. 2, the deposition in the nozzle section
was 11.170 and the deposition in the straight probe was only
0.77o. Test No. 3 showed a similar pattern of deposition.
When the front transpiration flow was increased to 56.6 1pm
(2.0 cfm), the deposition in the front nozzle section was
reduced significantly, while the deposition in the rear sec-
tion remained low. The test No. 5 showed comparable results
to tests No. 2 and No. 3 while sampling for a longer time at
about the same sampling conditions.
Results from site No. 2 (Table 6) showed that sampling
could be performed in the presence of a considerable amount
of liquid water in the effluent of the scrubber. Comparable
results to test No. 4 at site No. 1 were obtained even though
the sample consisted of several large particles. The two
tests showed good reproducibility of the amount of samples.
43
-------
The size distribution results of the samples are con-
sistant with the mass sampling efficiency. For test No. 1,
with the Method 5 train, the glass fiber filter deposit con-
sisted only of particles smaller than a few microns. The
size distributions of the nozzle and probe deposit were
comparable.
For tests No. 2 and No. 3, the nozzle and probe deposits
contained greater proportions of larger particles than the
glass fiber filter. At a higher transpiration rate in the
front section for test No. 4, the size distributions of the
nozzle, probe, and filter deposits were comparable to each
other.
44
-------
APPENDIX A
OPERATING MANUAL
Transpiration Sampling Interface
-------
OPERATING MANUAL
Transpiration Sampling Interface
1. INTRODUCTION
The transpiration sampling interface is based on the
use of transpiring air to prevent the deposition of the par-
ticulates on the walls of the sampling tube. The interface
was developed at IIT Research Institute on two projects
under contracts from the Environmental Protection Agency.
Test results and design principles are reported in the
following final reports for these projects:
1. Sampling Interface for Quantitative Transport of
Aerosols. Prepared by IIT Research Institute,
Chicago, for Environmental Protection Agency,
Research Triangle Park, Contract No. 68-02-0579,
Final Report #EPA 650-2-A016 (December 1973).
2. Sampling Interface for Quantitative Transport of
Aerosols. Prepared by IIT Research Institute,
Chicago, for Environmental Protection Agency,
Research Triangle Park, Contract No. 68-02-1295,
(July 1975).
2. DESCRIPTION
The sampling interface consists of two major components:
the sampling probe, and a control box for the sampling
operation.
The sampling probe is shown in Figure A. It consists of
a front section (FS) with a 90° bend and a rear section (RS)
Each section has an inner tube of 1,27 cm (1/2 in.) I.D. and
1.9 cm (3/4 in.) O.D. 316 stainless steel porous tube. The
inner tube is encased in an outer 316 stainless steel tube
(3.5 cm [1-3/8 in.] O.D.). The two sections are joined by a
screw-on type ball joint. The front section has a provision
to screw on any of the three interchangeable sampling nozzles.
Air is supplied to the front and the back section through
the 0.95 cm (3/8 in.) tubes with Swagelok fittings.
46
-------
cu
w
M
O
&
(Xi
O
53
H
hJ
I
CO
M a
•H O
ELI M
H
47
-------
The probe can be joined to a sampling device through a
3/4 in, NPT male connection or to a 12/25 ball joint through
an adapter,
The control box is shown in Figures B and C. It contains
a combination vacuum and pressure pump rated at 7 cfm air
output and 7 cfm intake capacities. The air from the pump
flows through a flowmeter R, and a heater H, and is divided
into two streams in order to supply air to the front and the
rear sections of the probe. On the vacuum side, the flow is
monitored by observing the pressure drop across a Meriam
laminar flow element (LFE) as measured by a 0 to 3 in. water
inclined tube manometer (M-,) . The output of the pump and
the intake are adjusted by the use of two bleed valves (V-,
and V,, respectively)„ The relative amounts of the trans-
piration air to the front and the back sections of the probe
are adjusted by manipulation of the bleed valve (V-,) and a
ball valve (Vo)• Valve V~ should not be used for flow control.
Full adjustment is possible with the help of the bleed valve
(V,); thus keeping the output of the pump constant for an even
operation„
®
Three Teflon '^ lined flexible hoses are provided to con-
nect the sampling probe to the control box. A thermocouple
is used to monitor stack temperature, The heater with a
variac is used for preheating the transpiration air to temperatures
up to 4000F_ Dial thermometers are used to monitor the temperature
of the airstream,
The transpiration air is filtered by a coarse filter, F,,
before it enters the pump= The heated transpiration air
flows through two glass wool filters, F2 and F~, before enter-
ing the front and back sections.
A list of parts and their sources is given in Table A.
48
-------
2
Figure B
FRONT VIEW OF CONTROL BOX
49
-------
LFE
VP
Figure C
REAR VIEW OF CONTROL BOX
4
50
-------
0)
rH
&
Cfl
H
c
o
•rl
4-1
Cfl
o
•H
MH
•H
rj
cu
ft
oo
rQ
c
Cfl
K**>
4-)
•H
O
ft
cfl
rH
01
T3
O
a
T)
c
^
01
Vl
3
o
cfl
3
C
£
c
o
•rl
4-1
ft
•H
Vi
O
CO
cu
Q
rH
o
J3
s^
m
B
U
o
rH
4J
CO
Cfl 1
CJ 1
05
VI
CU
4->
H
Vl -H
CU Pn
4J
rH rH
•H O
PI* O
£2
0)
CO CO
Vl CO
cfl cfl
0 rH
c_> O
<-*
rH
O 1
rH 0
i i
CO
Vl
Ol
4-1
rH 0)
•rH 60
PH co
rH 3
0
O 01
JS Vi
3
CO CO
CO CO
cfl CU
rH Vl
O (X,
ro
Pn cj
Vl
•H
CO cfl
4-1
4-1 g
CO MH
S 0
O -
*"O *"O
o o
a a
rt «\
a B
cfl cfl
O -rl
rH JH
3 0)
> a
4-1
c
CU
rH
W
|
rH
PH
Vl
Vi cfl
CU C!
4J -r)
cfl B
01 Cfl
« i-4
PH
pn
EC r4
VI
CU
4J 60
cfl M
•
• C
a -H
•rl
CN1
ro rH
1 1
0 O
Vl
cu
Vl
01
01
o
c
cfl Vi
S 01
4->
cu cu
rQ B
3 0
H C
CU
C 01
•rH .^
rH 3
U H
£ 1
I-H p>
rH CM
a a
• *
Q P
M M
• •
C C
•H -H
•o- oo
•^^ -^
rH CO
\ \
1 1
Ol CU
T-H rH
N N
N N
0 0
bO txO
c a
•rl -rl
rH rH
Oj Q^
e s
cfl CO
CO CO
l-H CM
53 53
CO
ft
rH CO
CSJ l/"^ CO rH
^- rH O 1
rH rH rH O
Vl
Vl 01
0) 4->
4J 1H
CO O
Cfl fXi
g
OJ T3
4-1 ti
Cfl Cfl
Vl
„ 0)
Vi J3
0) O
b m
1 1 ? -rl
1 1 Q Pn
cu
rH
N
N
O
53
4-1 Vl Vl
00 cfl 01 Ol
C 4J 4J 4-1
•rH CO O) Ol
rH Vl B B
ft 0) Cfl Cfl
a £s 4-j 4-i
CO O O O
C/5 r1 j p^{ j"y^
CO CO '"H C^l
53 P i P"! P^
^
CU
o
cfl
4-1
Cfl
•H
CO
01
Vi
a
0
d-
(U
C
•H
4J
rj
Cfl
4J
CO
C
O O
cfl U
> 1
I/"! O
rH 1 VI
rH 1 M
1 1 1
1 1 1
0)
rH
ft
3
o
O Ol
O rH
S ft
ft Vi 3
a 01 o
fx, H 0
0
Vi Vi Vi
O O 01
H
O O ,1^
4J 4J O
*H *H ctf
> & -U
O) CO C/3
p-4 CM r-l
C/3 to H
pi-, fTj fTj
o o o
O 0 O
O 0 O
"^" "^d" 00
1 i 1
0 O 0
rH
cu
'O
o
a
A
cfl
toO
0)
1 1 I
1 1 O
Vl
Ol
4-1
01
a
Vl
o
4J
Vl cfl
01 O
4-> -H
01 Vl T3
B 0) C
goi
B cu
O) 0 VI
-c a 3
H ft 4J
0) cfl
VI ,TJ U
Ol E—I Ol
4-1 ft
cfl w a
CU p^ Ol
W r4 H
a'
•
M
CM CO •
H H H
1 1
1 1
H
53
»
C
•H
D nH
ft cfl
>>>
H
O)
01 rH
r"> rrj
O 0)
H Q)
CU 53
Vl
•H
<3
.
ft
CO
c
cfl
Vl
EH
Vi
O 0)
l(H ^
CU cfl
> >
rH
> C
"O M
CU 0)
(U 4-t
iH CU
PQ S
>H CM
> >
PL|
B
3
ft
Vi
•H
cfl
B
MH
O
r^>
I I I
1 1 0
0)
rH
a
s
[ii
H
H PM
FM 53
125 r^*
» Csl
ts o
i~H ^ IS
CO
> Ol CO
i~H cd
i—l 'Q g
rH 0) 0
Cfl CU rC
PO 3 H
01
>
rH
cfl B
> 3
<4H 3
&0 M~l O
C O cfl
cfl 01 C
C Q) o
O rH -H
4J Cfl
• Mac
O 3 'H
ft 3 A
O CJ S
M cd O
Clj > O
CO J° P_|
> > >
o
•H
H
£
vD
rH
*""*^
rH
^
•
n
•
1 — 1
OH
e =
£3 ""^
PH -^-*-
CO
e
3 X
3
U •
cfl P
^ •
O
B
14-4 r.
o ^o
rH
r^ *^^
1 tf)
O rH
1
1
ft
B
3
PH
VI
< CU
rH
Id N
CU N
Cfl O
M 53
0)
Vl Vl
ft^
O
U Vl
01
^ rC
CJ 05
cfl cfl
1s
5
51
-------
3. PREPARATION OF THE PROBE
1. Select sampling nozzle. Suggested nozzles:
Velocity (FPM)
N1 - 1/4" I.D. > 3,000
N2 - 3/8" I.D. 1,000 - 3,000
N3 - 1/2" I.D. < 1,000
2. Screw the nozzle and washer W on the FS section
(as in Figure A).
3. Connect the 90° FS section to the RS section
(Figure 1) at the coupling by hand tightening.
Loosen the clamps to align and connect the FS air
supply tube with the Swagelok® fitting. Tighten
the coupling between FS and RS sections with a
wrench. Tighten the clamps.
4. Connect the air hoses to the front and back trans-
piration air supply (3/8" NPT pipe).
5. Connect the glass wool filters to the other ends
of the hoses.
6. Connect the hoses and filters to the Swagelol
fittings for front and rear transpiration sections
(Figure A).
7. Connect the pump cord to a standard 3 prong 115 volt
outlet.
8. Connect the heater (H) to the powerstat (PS).
9. The probe and the hoses may be preheated to 300°F
by wrapping them with heating tape.
10. Connect the probe to the measuring device or a
sample collector (GF filter holder).
11. If using the pump VP for pulling the flow through
the measurement device or sample collector, it may
be connected using the third hose to the inlet of
the laminar flow element.
12. Level and zero manometer M,.
52
-------
4. OPERATION OF THE PROBE
1. Fixed rate sampling:
A. Convert the desired sample rate to standard
conditions
530 P
SFR (scfm) = SFR (acfm) x ^—x 33 32
s
where SFR = sample flow rate, P = stack
pressure (in Hg), T = stack temperature
°R (460 + °F). s
B. Choose the transpiration flow rate and convert
it to standard conditions
530 P
TFR (scfm) = TFR (acfm) x a
Tx 29.92
a
where TFR = transpiration flow rate, Pa -
ambient pressure (in Hg), and T = ambient
temperature °R (460 + °F) . a
C. Calculate total flow
Qt (scfm) = SFR (scfm) + TFR (scfm)
D. Start the transpiration air flow through the
probe and place the probe tip in the stack
with the nozzle tip in the opposite direction
to the flow (see calibration curves: Figures D & E)
E. Read AP1 across the laminar flow element from
the curve in Figure F corresponding to
1.1 x Qt.
F. Adjust vacuum bleed valve V, until the AP on
manometer M-, across the laminar flow element
is equal to'AP as determined in step E.
G. Read Pf (in Hg) on manometer M2 and T^ on the
dial thermometer.
H. Read ?cf and T f from Tables B and C.
I. Divide Q by P ^ and T f.
53
-------
-
•
1
/
/
f
\
\
s
**•
f
r*
/
•
L
i
"fl
5-
~- t
/7
i
'
/
1 j "1
'/'
>
/
/
/
4
*1
/
^
c
1
1
\
1
\
c
1
/
A
/
L:
Tl
i
n
^
/
/
i
5R
iN
|
A
S
/
r
p-
,
-
j
^
Fi
ION
IRA
/
x
g
T
/
Ul
Cl
K
/
re
JI
D^
1
\
\ D
WE 1
i RO:
1
-
?c
w
\
:
)R
iM
:
^
E
[
FI
n
1C
si
-
"
-
~
)N
I
-
-
-
-
—
T
-
-•
—
-
-
^
-
-
-
-
-
-
-
-
._.
-
-
-
-
-r
-
-
-
-
-
-
-
-
-
-
/ct
54
-------
/-y
l-i
£
o
l.o
.$
• 6
-
\
•
1
'
/
/
/
/
1
\
2
"
/
i j
I
_
•—
/
t.
1.
/
r
S
-
—
~
/
-
C
•-
'
-
/
a
\
-
\i
i
\~
.1
:R
-
B
A
1
Ri
Mi
|
'
VI
31
'I
'I
/-
V
~ •
/
>v
.y
ON
RAr
—
~
...
5U
C
n
I
ir
;u
:o
I
e
R
N
/
_
I
V]
I
/
i;
7
1C
-
-
F
)T
|
-
0
A
/
~
-
-
R
MI
j
/
~
~
F
SI
r
~
j;
'E
"•
-
~
-
A
R
_
_
-
-
-
-
--
-
-
R
—
•
i
—
—
_.
~
i-
E
—
-
-
—
-
—
L
-:
•
-
-
-
—
„.
1-
~
-
;
--
-
—
-
_
-'
t
-
-
-
—
-
-
_
:
-
*
-
-
-
-
55
-------
Figure F
CALIBRATION CURVE FOR LAMINAR FLOW ELEMENT
56
-------
File rto. Oc.u.P:4?0
Table B
NERIAH LAMINAR FLOW ELEMENT
AIR OR GAS PRESSURE CONVERSION MULTIPLluAHON FACTOR
BASE PRESSURE (ASSIGNED STANDARD) 29.92 I.4CHES MERCUt-'Y A3SOL iTt
LAMINAR
INLET
PRESS.INCHES
HG. ABS. C.F.
26. OG
26.05
26.10
26.15
26.20
26.25
Z6.30
26.35
26.4?
26.45
26.30
26.55
26.60
26.65
26.70
26.75
26.80
26.85
26.90
26.95
27:00
27.05
27.10
27.15
27.20
27.25
27.30
27.36
27.40
27.45
27.50
27.55
27.60
27.65
27. 7C
27.75
27.80
27.85
?7.9C
27.95
28.00
23. C5
28.10
28.15
.8689
.8706
.8723
.8739
.8756
.8773
.8790
.8806
.8823
.8840
.3856
.8873
.8890
.8907
.8923
.8940
.8957
.8973
.8990
.9007
.9024
.9040
.9057
.9074
.9090
.9107
.9124
.9141
.9157
.9174
.9191
.9207
.9224
.9241
.9253
.9274
.9291
.9308
.9324
.9341
.9358
.9375
.9391
.9408
LAMINAR
INLET
PRESS.INCHES
Hfi. AftS. C.F.
28.20
28.35
28.30
28.35
28.40
28.45
28.50
28.55
28.60
28.65
28.70
28.75
28.80
28.85
28.90
28.95
29.00
29.05
29.10
29.15
29.20
29.25
29 . 30
29.35
29.40
?9.4C
29.oO
29.55
29 , 6 J
29.65
29.70
29.75
29.80
29.85
29 . 90
29.92
29.95
30.00
30.05
30.10
30.15
30.20
30. 25
30.30
.9425
.9441
.9458
.9475
.9491
.9508
.9525
.9542
.9558
.9575
.9592
.9608
.9625
.9642
. 9659
.9675
.9692
.9709
.9725
.9742
.9759
.9776
.9792
. 9809
,*826
.9842
.9859
.98?i
.9893
.9909
.9926
.9943
.9959
.9976
.9993
1 .0000
1.0010
1.0026
1.0043
1.0060
1.0076
1.0093
1.0110
1.0127
LAMINA*
INLET
PRESS. INCHES
HS. ASS.
30.35
30.40
30.45
30.50
30.55
20.60
30.65
30.70
30.75
30.80
30.85
30.90
30.95
31.00
31.05
31.10
31.15
31.20
31.25
31.30
31.35
31.40
31.45
31.50
31.55
31.60
31.65
31.70
31.75
31.80
31.85
31.90
31.95
32.00
32.05
32.10
32.15
32.20
32.25
32 . 30
32.35
32.40
32.45
32 . bO
C. F .
1.0143
1.0160
1.0177
1.0193
1.0210
1.0227
1.0243
1.0260
1.0277
1.0294
1.03". 3
1.0327
1.0344
1.0360
1.0377
1.0394
1.0411
1.0427
1.0444
1.0461
1.0477
1.0494
1.0511
1.0528
1.05-54
1.U561
1.0578
1.0594
I.Ofll
1.062^
1.0645
1 . 'J66 1
1.0678
1.069s
1.0711
1.0728
1.0745
1.076?
1.0778
1.0795
1.0812
1.W828
1.0845
1.0862
tAMINAR
INLET
PSESS. INCHES
HG. A6S. C.F*
32.55
32.60
3-2.65
32.70
32.75
32.80
32.35
32.90
32.95
33.00
33.05
33.10
3J.15
1-3.20
33.25
'33. 3^
3J.3L
33.40
33.45
33.50
33.55
J3.60
33.65
33.70
33.76
33. bO
33.85
33.90
33.95
34.00
34.05
34.10
34.15
34.20
34.25
34.30
34.35
34.40
34.45
34.50
34.55
34.60
34.65
34.70
1.0879
1 . 0895
1.0912
1.0929
1.0945
i.0962
1.0979
1.0995
1.1012
1.1029
1 . 1 046
1.1062
1.1079
1.1096
i.1112
1 . 11 29
1.1146
1.1163
1 . 1 ! 79
1.1196
1.1213
1.1225
I.i246
1.1263
1 . 1 380
1.1296
1.1313
1.1330
1 . 1 346
; . i 363
1.1 3bO
1.1397
1.1413
1.1430
1.1447
1.1463
1.1480
1.14S7
1.1514
1.1530
1.1547
1.1564
1.1580
1.1597
24-6P
THE MER1AM INSTRUMENT COMPANY
10920 Madison Avenue
Cleveland, Ohio 44102
57
SHf>.
1 OF
3 \ 0
-------
Table B (continued)
HER I AM LAMINAR FLOW ELEMENT
AIR OR GAS PRESSURE CONVERSION MULTIPLICATION FACTOR
BASE PRESSURE (ASSIGNED STANDARD) 29.92 INCHES MERCURY ABSOLUTE
PRESS INCHES
HG. ABS C.F.
34.75
34.80
34.85
34.90
34.95
35.00
35-05
1.1614
1.1631
1.1647
1.1664
1.1681
1 . 1 697
1.1714
PRESS INCHES
HG. ABS C.F.
35.10
35.15
35.20
35.25
35.30
35.35
35.40
1.1731
1.1747
1.1764
1,1781
1.1798
1.1814
1.1831
PRESS INCHES
HG. ABS C.F.
35.45
35.50
35.55
35.60
35.65
35.70
35.75
.1848
.1864
.1881
.1898
.1915
.1931
.1948
PRESS
HG.
INCHES
ABS C.F
35.80
35.85
35-90
35-95
36.00
1.1965
1.1981
1.1998
1.2015
1.2032
For values not shown in table,
P flow P f 1 ow
Po.f. = P Base = 29-92
interpolate or use equation.
Po.f. = Pressure Conversion Factor
P base = Assigned Base Pressure of 29.92 inches mercury absolute
P flow = Laminar Inlet Pressure, inches mercury absolute
Above equation can be used up to and including two atmoshperes absolute.
necessary to calibrate laminars for pressure exceeding above.
It will be
To use: Take the flow value from the flow vs differential pressure curve and
multiply by the pressure conversion multiplication in the table above.
This gives the flow in standard cubic feet per minute referenced to a
pressure base of 29.92 inches mercury absolute.
58
-------
^
r_j
Table
CTORS FOR LAMINAR UNIT
<
CTION 1
w
rv*.
K
O
U
ture = 70°F
CORRECTION FACTOR s4|x^»F
cd
1
H
0)
w
ctf
PQ
W
H
Pd
H
<
CD
00
•f
rv
O
-i-
1
r
LO
CO
ro
O
r— 4
CV
C
r— 4
-)
LO
rr
o
CO
CD
e:
CD
CM
;-- ir§
j j
I
1 "^
•~O
1 ^
o
CD
ro
0
0
CO
o
o
,—4
o
1—4
O
.-^
r— 4
»—4
O
r-(
LO
r--
r— 4
O
i — 4
0
— 4
CM
O
, — i
-o
,_ j LD CM
' 10 ! 0
-t
i
*J '
< '
->• 1
1
i — i
I
i
r
o
+
D.
«S U.
W O
H
0
-T
O
!
1 ,-,
^
f^.
<*O
O
•
^r
r— 1
fs^ *
O
—
o
LO
1—4
0
00
CM
o
,
^
IO
j — 4
CO
O
•
o
LO
CO
o
-1
o
LD
CM
cn
^
^
cn
PN
*
CO
o
CO
CD
i — i
00
CD
CO
CO
CD
O
CD
CD
cn
CO
CD
CD
:M
r^
CT)
c-i
•
LD
O
O
O
r~~ 4
O
rv
O
i — i
(O
CD
CM
*
CM
CD
CD
CO
CM
lf>
CD
•*
CD
00
^o
CT)
, — 1
O
CD
CO
LD
cn
•
O
CO
to
CD
0
00
0
. — i
CO
/— 4
CD
s
. — 1
CD
co
CD
. — l
CD
•M
CM
CM
CD
LO
CM
cn
oc
CM
a>
r—t
<^>
m
•— ^
^*
CO
CD
•
t— 4
f^
ro
CT
O
CD
LD
CO
00
CO
LO
CO
00
00
oo
00
CO
0
CD
CO
CO
CD
00
LD
CO
CM
CD
CD
CO
0
0
CD
.
CD
^J4
"^
CD
•
^
f^
O
CD
0
O
c 1
o
LO
00
00
LO
00
•
CM
LC
oo
00
CO
LD
00
IO
to
CO
CD
LO
CO
OO
1 — I
t^.
00
iTt
XT
rv
CO
,
CM
f-v.
fx.
oo
•
CD
CD
CO
a
—4
r— 4
O
ro
00
-o
-o
CO
LO
00
r— 4
oo
CO
00
*
§
CO
CO
CO
•N
TJ4
00
co
Ttf
oo
,
CO
o
*J*1
CO
•
<3-
ro
LO
CO
o
CM
r-4
LO
o
oo
00
oo
o
co
•
CM
oo
ro
co
c
00
03
i — 4
00
CO
0
CM
oo
CO
CO
CM
oo
t
^
LO
CM
CO
•
CM
oo
CM
CO
o
ro
LTi
CO
r-.
'o
30
•
o
CO
00
o
.
LO
CTl
00
(V
. — 4
r-.
cr.
[**_
'
10 ,
ie
'•o
to
•
oo
LO
LO
o
CO
LO
.
o
^
£
.0
[V, <
cv
co
'0 i
CD 1 l^
O, | [-X
CO
,«4
o
oo
•
*?'
o
CO
o
r— 4
O
CD
tv.
rv
•
CM
rH
00
o
LO
•—4
OJ
00
00
rO
I
M
2 2
u. *
o *
(/) O
5
^•^ *^«
8|
UJ
UJ
59
-------
J. Read AP" corresponding to Qt/P fT f if dif-
ferent than AP1 adjust to AP1 8y means of
valve V,.
K. If the P.p and T^ change significantly, i.e. ,
temperature by more than 10° and pressure
by more than 1 in. Hg, adjust the flow again
repeating steps I and J.
NOTE: If sampling near 70°F the total flow
may be set prior to insertion in the sample
stream by connecting a wet test meter to
the inlet of the sampling nozzle.
2. Isokinetic sampling (single point):
A. Determine the velocity (ft/min) at the sampling
point with a pitot tube.
B. Calculate the sample flow rate in acfm
SFR (acfm) = V x Area of Nozzle
D
2
Nozzle Diameter (in.) Area (ft )
N-l 1/4 0.000347
N-2 3/8 0.000767
N-3 1/2 0.001363
C. Follow procedures described in steps 1A to
IK.
3. Isokinetic sampling (multipoint):
A. Make a velocity traverse with a pitot tube.
B. Find the stack velocity at the sampling points.
C. Choose appropriate nozzle (see Section 3).
D. Calculate SFR.
E. Choose TFR and keep it constant for the entire
sampling operation.
F. Obtain total sample flow rate at each point
QP Q2 .-.
G. Set the AP across M-j^ to I.IQ-^.
H. Adjust for P f and T f.
60
-------
I. Set AP" again.
J. Use the same P ^ and T ^ for all of the points.
K. Sample at each point adjusting the flow through
the laminar flow element.
61
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-76-157
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
SAMPLING INTERFACE FOR THE QUANTITATIVE TRANSPORT
OF AEROSOLS--FIELD PROTOTYPE
5. REPORT DATE
June 1976
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Madhav B. Ranade
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
IIT Research Institute
10 West 35th Street
Chicago, Illinois 60616
10. PROGRAM ELEMENT NO.
1AA010
11. CONTRACT/GRANT NO.
68-02-1295
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Sciences Research Laboratory
Office of Research and Development
Environmental Protection Agency
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final Report
14. SPONSORING AGENCY CODE
EPA - ORD
15. SUPPLEMENTARY NOTES
16. ABSTRACT
A sampling interface for the quantitative transport of aerosols was designed,
fabricated and evaluated. The interface may be used for transporting particles
(up to 50 ym on diameter) from industrial stacks to a collection device or a
monitoring instrument. The interface consists of a porous wall probe with
clean air transpiring inwards to prevent deposition of particles. Laboratory
and field testing of the interface has shown it to be efficient in transporting
particles encountered in industrial stacks.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
* Air pollution
*Aerosols
*Sampling
*Probes
*Design
Fabrication
Evaluation
Tests
Field tests
13B
07D
14B
13H
14B
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
21. NO. OF PAGES
68
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
62
------- |