&EPA
United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EPA-450/2-78-042d
October 1978
Air
Stack Sampling
Technical Information
A Collection of
Monographs and Papers
Volume IV
-------
EPA-450/2-78-042d
Stack Sampling Technical Information
A Collection of Monographs and Papers
Volume IV
Emission Standards and Engineering Division
U S ENVIRONMENTAL PROTECTION AGENCY
Office of Air, Noise, and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
October 1978
-------
This report has been reviewed by the Emission Standards and Engineering
Division, Office of Air Quality Planning and Standards, Office of Air, Noise
and Radiation, Environmental Protection Agency, and approved for publica-
tion. Mention of company or product names does not constitute endorsement
by EPA. Copies are available free of charge to Federal employees, current
contractors and grantees, and non-profit organizations - as supplies permit
from the Library Services Office, MD-35, Environmental Protection Agency,
Research Triangle Park, NC 27711; or may be obtained, for a fee, from the
National Technical Information Service, 5285 Port Royal Road, Sprinqfield
VA 22161.
Publication No. EPA-450/2-78-042d
n
-------
PREFACE
The Clean Air Act of 1970 requires the Administrator of the
Environmental Protection Agency to establish national emission standards
for new stationary sources (Section 111) and hazardous air pollutants
(Section 112). The development of these emission standards required the
concurrent development of reference test methods and procedures. The
reference test methods and procedures are published in the Federal Register
along with the appropriate regulations.
From time to time, questions would surface concerning the methods and
procedures. In many cases, specific studies would be needed to provide
informed, objective answers. The papers and monographs resulting from these
studies were usually distributed to people involved in emission measurement;
a major method of distribution has been the Source Evaluation Society
Newsletter.
To provide a readily available resource for new and experienced personnel,
and to further promote standardized reference methods and procedures, it has
been decided to publish the papers and monographs in a single compendium.
The compendium consists of four volumes. The Table of Contents for all
four volumes is reproduced in each volume for ease of reference.
Congratulations and sincere appreciation to the people who did the
work and took the time to prepare the papers and monographs. For the most _
part the work was done because of personal commitments to the development
of objective, standardized methodology, and a firm belief that attention
to trie details of stack sampling makes for good data. The foresight of
Mr. Robert L. Ajax, the former Chief of the Emission Measurement Branch and
now the Assistant Director, Emission Standards and Engineering Division, in
providing the atmosphere and encouragement to perform the studies is
gratefully acknowledged. The skill and dedication of Mr. Roger Shigehara,
in providing personal supervision for most of the work, is commended.
Don R. Goodwin
Director
Emission Standards and
Engineering Division
m
-------
VOLUME I
TABLE OF CONTENTS
Method for Calculating Power Plant Emission Rate i
by R. T. Shigehara, R. M. Neulicht, and W. S. Smith
•
Emission Correction Factor for Fossil Fuel-Fired Steam in
Generators (C02 Concentration Approach)
by R. M. Neulicht
Derivation of Equations for Calculating Power Plant Emission 20
Rates (02 Based Method - Wet and Dry Measurements)
by R. T. Shigehara and R. M. Neulicht
Summary of F Factor Methods for Determining Emissions from 29
Combustion Sources
by R. T. Shigehara, R. M. Neulicht, W. S. Smith,
and J. W. Peeler
Validating Orsat Analysis Data from Fossil-Fuel-Fired Units 44
by R. T. Shigehara, R. M. Neulicht, and W. S. Smith
A Guideline for Evaluating Compliance Test Results 55
(Isokinetic Sampling Rate Criterion)
by R. T. Shigehara
-------
VOLUME II
TABLE OF CONTENTS
A Type-S Pi tot Tube Calibration Study 1
by Robert F. Vollaro
The Effect of Aerodynamic Interference Between a Type-S 24
Pi tot Tube and Sampling Nozzle on the Value of the
Pi tot Tube Coefficient
by Robert F. Vollaro
The Effects of the Presence of a Probe Sheath on Type-S 30
Pi tot Tube Accuracy
by Robert F. Vollaro
An Evaluation of Single-Velocity Calibration Technique as 48
a Means of Determining Type-S Pitot Tube Coefficients
by Robert F. Vollaro
Guidelines for Type-S Pitot Tube Calibration 63
by Robert F. Vollaro :
The Effects of Impact Opening Misalignment on the Value of 89
the Type-S Pitot Tube Coefficient
by Robert F. Vollaro
Establishment of a Baseline Coefficient Value for Properly 95
Constructed Type-S Pitot Tubes
by Robert F. Vollaro
A Survey of Commercially Available Instrumentation for the 104
Measurement of Low-Range Gas Velocities
by Robert F. Vollaro
The Use of Type-S Pitot Tubes for the Measurement of Low 122
Velocities
by Robert F. Vollaro
vi
-------
VOLUME III
TABLE OF CONTENTS
Thermocouple Calibration Procedure Evaluation 1
by Kenneth Alexander
Procedure for Calibrating and Using Dry Gas Volume Meters 10
As Calibration Standards
by P. R. Westlin and R. T. Shigehara
Dry-Gas Volume Meter Calibrations 24
by Martin Wortman, Robert Vollaro, and Peter Westlin
Calibration of Dry Gas Meter at Low Flow Rates 33
by R. T. Shigehara and W. F. Roberts
Calibration of Probe Nozzle Diameter 41
by P. R. Westlin and R. T. Shigehara
Leak Tests for Flexible Bags 45.
by F. C. Biddy and R. T. Shigehara
Adjustments in the EPA Nomograph for Different Pitot Tube 48
Coefficients and Dry Gas Molecular Weights
by R. T. Shigehara
Expansion of EPA Nomograph (Memo) 60
by R. T. Shigehara
EPA Nomograph Adjustments (Memo) 63
by R. T. Shigehara
Graphical Technique for Setting Proportional Sampling 65
Flow Rates
by R. T. Shigehara
vi i
-------
VOLUME IV
TABLE OF CONTENTS
Recommended Procedure for Sample Traverses in Ducts Smaller 1
Than 12 Inches in Diameter
by Robert F. Vollaro
Guidelines for Sampling in Tapered Stacks 24
by T. J. Logan and R. T. Shigehara
Considerations for Evaluating Equivalent Stack Sampling 28
Train Metering Systems
by R. T. Shigehara
Evaluation of Metering Systems for Gas-Sampling Trains 40
by M. A. Wortman and R. T. Shigehara
An Evaluation of the Current EPA Method 5 Filtration 49
Temperature-Control Procedure
by Robert F. Vollaro
Laboratory Evaluation of Silica Gel Collection Efficiency 67
Under Varying Temperature and Pressure Conditions
by Peter R. Westlin and Fred C. Biddy
Spurious Acid Mist Results Caused by Peroxides in Isopropyl 79
Alcohol Solutions Used in EPA Test Method 8 (Memo)
by Dr. Joseph E. Knoll
Determination of Isopropanol Loss During Method 8 Simulation 80
Tests (Memo)
by Peter R. Westlin
Comparison of Emission Results from In-Stack Filter Sampling 82
and EPA Method 5 Sampling
by Peter R. Westlin and Robert L. Ajax
EPA Method 5 Sample Train Clean-Up Procedures 93
by Clyde E. Riley
viii
-------
RECOMMENDED PROCEDURE FOR SAMPLE TRAVERSES
IN DUCTS SMALLER THAN 12 INCHES IN DIAMETER
Robert F. Vollaro**
INTRODUCTION
In source sampling, stack gas velocity Is usually measured with a
Type-S pi tot tube. In many field applications, the pi tot tube is attached
to a sampling probe, equipped with a nozzle and thermocouple. This combi-
nation is called a pitobe assembly. Most conventional pitobe assemblies*
have a cylindrical sampling probe of 1-inch diameter, but, occasionally,
an assembly has an external cylindrical sheath of about 2-1/2 inches in
diameter, encasing the probe, pi tot tube and thermocouple. When a pitobe
assembly is used to traverse a duct that is 36 inches or less in diameter,
the pitobe assembly can "block" a significant part of the duct cross section,
as illustrated in the projected-area models, Figures la and Ib. This reduction
in the effective cross-sectional area of the duct causes a temporary, local
increase in the average velocity of the flowing fluid. In most pitobe
assemblies, the impact opening of the Type-S pi tot tube lies in approximately
the same plane as the probe sheath (Figure 2) and, whenever appreciable sheath
blockage exists, velocity head (&P) readings made with the pi tot tube tend
to reflect the local increase in gas velocity, and are not truly representa-
1 2
tive of the mainstream velocity. Recent studies ' have shown that, for
sample traverses in ducts having diameters or equivalent diameters between
12 and 36 inches, blockage effects are not particularly severe, and a simple
*Designed according to the specifications outlined in APTD-0581 (Reference 3),
or allowable modifications thereof.
** Emission Measurement Branch, ESED, OAQPS, EPA, RTP, NQ January 1977
-------
(a)
ESTIMATED
' SHEATH
BLOCKAGE
**"• I
DUCT AREAj
x 100
(b)
PO
Figure 1. Projected-area models for typical pitobe assemblies; shaded area represents approximate
average sheath blockage for a sample traverse.
-------
SAMPLING
PROBE
TYPE-S '
PITOTTUBE
STATIC PRESSURE
OPENING,
I
1
APPROXIMATE
PLANE OF PROBE
SHEATH BLOCKAGE
IMPACT PRESSURE
OPENING
FLOW
DIRECTION
Figure 2. Type-S pilot tube, attached to a sampling probe, showing that the pitot impact
opening and probe sheath lie in approximately the same plane.
-------
adjustment in the value of the Type-S pi tot tube coefficient (C ) can be
made to compensate for the pseudo-high AP readings (Figure 3). When the
duct diameter (Ds) is less than 12 inches, however, probe sheath blockage
effects intensify, and the adjustment technique illustrated in Figure 3 no
longer applies. Therefore, alternative methodology must be used in order
to obtain representative sample traverses in ducts of this size. The
purpose of this paper is to propose a method by which satisfactory sample
traverses can be conducted when DS is between 4 and 12 inches.
PROPOSED METHOD FOR SAMPLE TRAVERSES
WHEN 4 in. <_ DS < 12 in.
METHODOLOGY
To conduct representative sample traverses in ducts having diameters
between 4 and 12 inches, it is recommended that the arrangement illustrated
in Figure 4 be used. In Figure 4, velocity head (AP) readings are taken
downstream of the actual sampling site. The purpose of the straight run of
duct between the sampling and velocity measurement sites is to allow the flow
profile, temporarily disturbed by the presence of the sample probe, to redevelop
and stabilize. The pitot tube and sampling nozzle shown in Figure 4 are
different from those of a conventional pitobe assembly;3 construction details
of these components are discussed below.
A. Pitot tube.
A standard pitot tube shall be used, instead of a Type-S, to monitor
stack gas velocity. When DS is less than 12 inches, a Type-S pitot tube can
begin to block a significant part of the duct cross section and yield
-------
_
2 1/2-in. CYLINDRICAL MODEL.
USE FOR ASSEMBLIES
WITH EXTERNAL SHEATHSv
o
5
u
o
u- O
u£ 3
CO
o
1-in. CYLINDRICAL MODEL.
USE FOR ASSEMBLIES
WITH NO EXTERNAL SHEATHS
i 1234
DECREASE IN PITOT TUBE COEFFICIENT. percent
Figure 3. Adjustment of Type-S pitot tube coefficients to account for sheath blockage
(12 in.
-------
FLOW
DISTURBANCE
-* 2DS-
1 (
1
> i
^ o us ^
^ t '
V\
M- 0 D
I
1 i
s
0,
"s
S
STANDARD1
PITOT
TUBE
TEMPERATURE
SENSOR
•SAMPLING
PROBE
CTl
FLOW
DISTURBANCE
Figure 4. Recommended sampling arrangement, when 4 in. < Ds < 12 in.
-------
7
pseudo-high AP values. Cross-section blockage is not a serious problem
with a standard pi tot tube, however, for two reasons: (1) the impact and
static pressure openings of a standard pi tot tube, unlike those of a
Type-S, follow a 90° bend, and are located well upstream of the stem of
the tube (compare Figures 2 and 5); and (2) when properly aligned, the
sensing head of a standard pitot tube is parallel, not perpendicular, to
the flow streamlines in the duct.
The preferred design for the standard pitot tube is the hemispherical-
nosed design (Figure 5). Pitot tubes constructed according to the criteria
4 5
illustrated in Figure 5 will have coefficients of 0.99 ±0.01 ' . Note,
however, that for convenient tubing diameters (dimension "D" Figure 5), the
static and impact sensing holes of the hemispherical-type pitot tube will
be very small, thus making the tube susceptible to plugging, in particulate
or liquid droplet-laden gas streams. Therefore, whenever these conditions
are encountered, either of the following can be done: (1) a "back purge"
system of some kind can be used to clean out, periodically, the static and
impact holes; or (2) a modified hemispherical-nosed pitot tube (Figure 6),
which features a shortened stem and enlarged impact and static pressure holes,
can be used instead of the conventional hemispherical type. It has recently
been demonstrated that the coefficients of the conventional and modified
hemispherical-nosed tubes are essentially the same.
B. Sampling nozzle.
The sampling nozzle can either be of the buttonhook or elbow design.
The nozzle shall meet the general design criteria specified in Section 2.1.1
of the revised version of EPA Method 5, except that the entry plane of the
-------
T
r~3D
STATIC PRESSURE OPENINGS -
~0.1 D
IMPACT PRESSURE
OPENING
0.4D
16 D
oo
80
Figure 5. Hemispherical-nosed standard pitot tube.
-------
4STATIC HOLES
3/8 D
IMPACT OPENING
1/20
Figure 6. Modified hemispherical-nosed pitot tube.
-------
10
nozzle must be at least 2 nozzle diameters (i.d.) upstream of the probe
sheath blockage plane (see Figure 7).
PROCEDURES
The following procedures shall be used to perform sample traverses
using the arrangement illustrated in Figure 4:
A. Location of sampling site.
Select a sampling site that is at least 8 duct diameters downstream
and 10 diameters upstream from the nearest flow disturbances; this allows
the velocity measurement site to be located 8 diameters downstream of the
sampling location and 2 diameters upstream of the nearest flow disturbance
(see Figure 4). For rectangular stacks, use an equivalent diameter, calcu-
lated from the following equation, to determine the upstream and downstream
distances:
De = L^tf (Equation l)
Where:
D = Equivalent diameter
L = Length of cross section
W = Width of cross section
If a sampling site located 8 diameters downstream and 10 diameters upstream
from the nearest disturbances is not available, select a site that meets
these criteria as nearly as possible. Under no circumstances, however, shall
a sampling site be chosen which is less than 2 diameters downstream and 2.5
diameters upstream from the nearest disturbances; this guarantees a minimum
-------
PLANE OF
SHEATH
BLOCKAGE
Figure 7. Recommended sampling nozzle design for use when 4 in. < Ds < 12 in.
-------
12
of 2 diameters of straight run between the sampling and velocity measure-
ment sites, and 0.5 diameters between the velocity measurement site and
the nearest flow disturbance.
B. Number of traverse points.
The correct number of traverse points shall be determined from
Figure 8. To use Figure 8, proceed as follows: first, determine the three
distances, MA", "B", and "C", and express each distance in terms of duct
diameters; second, read from Figure 8 the number of traverse points
corresponding to each of these three distances; third, select the highest
of the three numbers of traverse points, or a greater number, so that for
circular ducts the number is a multiple of 4; for rectangular ducts, the
number should be chosen so that it is one of those shown in
Table 2.
C. Location of traverse points, circular cross sections.
For circular stacks, locate the traverse points on 2 perpendicular
diameters, according to Table 1 and the example of Figure 9a. Any traverse
point located less than 1/2 inch from the stack wall will not be acceptable
for use as a sampling point; all such traverse points shall be "adjusted"
by relocating them to a distance of 1/2 inch from the wall. In some cases,
this relocation process may involve combining two adjacent traverse points
to form a single "adjusted" point; thus, in some instances, the number of
points actually used for sampling may be less than the number of traverse
points obtained from Figure 8.
D. Location of traverse points, rectangular cross sections.
For rectangular stacks, divide the cross section into as many
equal rectangular elemental areas as traverse points (as
-------
NUMBER OF DUCT DIAMETERS BETWEEN VELOCITY MEASUREMENT SITE AND NEAREST DISTURBANCE "
DISTANCE C
10
NUMBER OF DUCT DIAMETERS BETWEEN SAMPLING SITE AND NEAREST DISTURBANCE
! DISTANCE A
OR
NUMBER OF DUCT DIAMETERS BETWEEN SAMPLING AND VELOCITY MEASUREMENT SITES.
DISTANCE B
Figure 8. Minimum number of traverse points, 4 in. < Ds < 12 in.
-------
Table 1. LOCATION OF TRAVERSE POINTS
IN CIRCULAR STACKS (PERCENT OF STACK
DIAMETER FROM INSIDE WALL TO TRAVERSE POINT)
Traverse
point
number
on a
diameter
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Number of traverse points on a diameter
2
14.6
85.4
4
6.7
25.0
75.0
93.3
6
4.4
14.6
29.6
70.4
85.4
95.6
8
3.2
10.5
19.4
32.3
67.7
80.6
89.5
96.8
10
2.6
8.2
14.6
22.6
34.2
65.8
77.4
85.4
91.8
97.4
12
2.1
6.7
11.8
17.7
25.0
35.6
64.4
75..0
82.3
88.2
93.3
97.9
14
1.8
5.7
9.9
14.6
20.1
26.9
36.6
63.4
73.1
79.9
85.4
90.1
94.3
98.2
16
1.6
4.9
8.5
12.5
16.9
22.0
28.3
37.5
62.5
71.7
78.0
83.1
87.5
91.5
95.1
98.4
18
1.4
4.4
7.5
10.9
14.6
18.8
23.6
29.6
38.2
61.8
70.4
76.4
81.2
85.4
89.1
92.5
95.6
98.6
20
1.3
3.9
6.7
9.7
12.9
16.5
20.4
25.0
30.6
38.8
61.2
69.4
75.0
79.6
83.5
87.1
90.3
93.3
96.1
98.7
22
1.1
3.5
6.0
8.7
11.6
14.6
18.0
21.8
26.2
31.5
39.3
60.7
68.5
73.8
78.2
82.0
85.4
88.4
91.3
94.0
96.5
98.9
24
1.1
3.2
5.5
7.9
10.5
13.2
16.1
19.4
23.0
27.2
32.3
39.8
60.2
67.7
72.8
77.0
80.6
83.9
86.8
89.5
92.1
94.5
96.8
98.9
-------
15
Figure 9a. Cross section of circular stack divided into
12 equal areas, showing location of traverse points.
o o I o
o ! o I o I
—H--i
o o I o ,
I I I
Figure 9b. Cross section of rectangular stack divided
into 12 equal areas, with traverse points at centroid
of each area.
-------
16
determined in Section "B" above), according to Table 2. Locate
a traverse point at the centroid of each elemental area, according to the
example of Figure 9b.
E. Sampling.
Sample at each non^adjusted traverse point for the time interval
specified in the method being used (e.g., Method 5). If two successive
traverse points have been relocated to a single "adjusted" traverse point,
sample twice as long at the adjusted point as at non-adjusted points, taking
twice as many readings, but record the data as though two separate points
had been sampled, each for half of the total time interval. During the
sample run, velocity head (AP) readings shall be taken at points downstream
of, but directly in line with, the sampling points. The sampling rate
through the nozzle shall be set based upon the AP readings; if a nomograph
is used, be sure when setting it to use the correct value (~ 0.99) of the
pi tot tube coefficient.
ALTERNATIVE SAMPLING STRATEGY (STEADY-FLOW ONLY)
If the average total volumetric flow rate in a duct is constant with
time, it is unnecessary to monitor stack gas velocity during a sample run.
Thus, whenever time-invariant flow is believed to exist in a stack (e.g.,
for a steady-state process), the following traverse procedures can be used
in lieu of those outlined in the preceding sections:
A. Location of Sampling-Velocity Measurement Site.
When steady flow is believed to exist in a duct, the sample and
velocity traverses can be conducted non-simultaneously; therefore, the
sampling and velocity measurement sites need not be separate. Rather, a
-------
17
Table 2. CROSS-SECTIONAL LAYOUT FOR RECTANGULAR STACKS
No. of traverse Layout
points
9 3x3
12 4x3
16 4x4
20 5x4
25 5x5
30 6x5
36 6x6
42 7x6
49 7x7
-------
18
single location can be used for both sampling and velocity measurement
(see Figure 10).
Select a sampling-velocity measurement site that is at least
8 duct diameters downstream and 2 diameters upstream from the nearest
flow disturbances. For rectangular stacks, use an equivalent diameter
(Equation 2) to determine the upstream and downstream distances. If a
sampling-velocity measurement site located 8 diameters downstream and
2 diameters upstream from the nearest disturbances is not available,
choose a site that meets these criteria as nearly as possible. Under no
circumstances, however, should a sampling-velocity measurement site be
chosen that is less than 2 diameters downstream and 0.5 diameter upstream
from the nearest disturbances.
B. Number of Traverse Points.
The correct number of traverse points shall be determined from
Figure 11. To use Figure 11, proceed as follows: first, determine the
distances "A" and "B" and express each distance in terms of duct diameters;
second, read from Figure 11 the number of traverse points corresponding to
each of these distances; third, select the higher of these two numbers of
traverse points, or a greater number, so that for circular ducts the number
is a multiple of 4 and, for rectangular ducts, the number is one of those
shown in Table 2.
C. Location of Traverse Points, Circular Cross Sections
For circular stacks, locate the traverse points on 2 perpendicular
diameters, according to Table 1 and the example of Figure 9a. Any traverse
point located less than 1/2 inch from the stack wall will be unacceptable
-------
FLOW
DISTURBANCE
SAMPLING-VELOCIT'i
MEASUREMENT
SITE
FLOW
DISTURBANCE
vo
Figure 10. Recommended sampling arrangement; 4 in. < Ds < 12 in.; steady-flow only.
-------
NUMBER OF DUCT DIAMETERS UPSTREAM (FROM NEAREST FLOW DISTURBANCE).
, . DISTANCE A
2.0
FLOW
\ / DISTURBANCE
A T SAMPLING
AND
ro
o
NUMBER OF DUCT DIAMETERS DOWNSTREAM (FROM NEAREST FLOW DISTURBANCE),
DISTANCE B
Figure 11. Minimum number of traverse points; 4 in. < Ds < 12 in.; steady-flow only.
-------
21
for use, either as a velocity traverse point or as a sample point; all
such points shall be "adjusted" by relocating them to a distance of 1/2
inch from the wall. In some cases, this relocation process may involve
combining two adjacent traverse points to form a single "adjusted" point;
thus, the number of traverse points actually used will sometimes be less
than the number of points obtained from Figure 11.
D. Location of Traverse Points, Rectangular Cross Sections.
For rectangular stacks, divide the cross section into as many
equal rectangular elemental areas as traverse points (as
determined in Section "B" above), according to Table 2.
Locate a traverse point at the centroid of each elemental area, according
to the example of Figure 9b.
E. Preliminary Velocity Traverse.
Perform a preliminary velocity traverse of the duct. Take velocity
head UP) readings at each traverse point, using a standard pi tot tube
(designed as shown in Figure 5 or Figure 6). Calculate the average velocity
o
in the duct, using Equation 2-2 in the December 23, 1971 Federal Register.
F. Sampling
Sample at each non-adjusted traverse point for the time interval
specified in the method being used (e.g., Method 5). If two successive
traverse points have been relocated to a single "adjusted" traverse point,
sample twice as long at the adjusted point as at non-adjusted points, taking
twice as many readings, but record the data as though two separate points
had been sampled, each for half of the total time interval. Time-invariant
-------
22
flow is assumed; therefore, the sampling rate at each point shall be set
based on the AP reading obtained at that point during the preliminary
velocity traverse.
G. Post-Test Velocity Traverse.
Perform a second velocity traverse of the duct, at the end of the
sample run. Calculate the average velocity in the duct (Vg) avg., using
Equation 2-2 of the December 23, 1971 Federal Register.8 If the value of
(V ) avg. is within +. 10 percent of the value obtained in the preliminary
5 /-
traverse, the assumption of time-invariant flow is valid, and the results
are acceptable. If the difference between the pre-test and post-test values
of (V ) avg. is greater than + 10 percent, reject the results and repeat the
run, monitoring velocity during sampling, as shown in Figure 4.
-------
23
REFERENCES
1. Vollaro, R. F. The Effects of the Presence of a Probe Sheath on
Type-S Pitot Tube Accuracy. Environmental Protection Agency, Emission
Measurement Branch. Research Triangle Park, N. C. August 1975.
2. Vollaro, R. F. Guidelines for Type-S Pitot Tube Calibration.
Environmental Protection Agency, Emission Measurement Branch. Research
Triangle Park, N.C. September 1975.
3. Martin, Robert M. Construction Details of Isokinetic Source-
Sampling Equipment. Environmental Protection Agency, Research Triangle
Park, N.C. Publication No. APTD-0851, April 1971.
4. Perry, Robert H., Cecil H. Chilton, and Sidney D. Kirkpatrick (ed.).
Chemical Engineers' Handbook, Fourth Edition. New York, McGraw-Hill Book
Company, 1963.
5. Fluid Meters, Their Theory and Application. New York, Published
by the American Society of Mechanical Engineers. 5th Edition, 1959.
6. Vollaro, R. F. Evaluation of Modified Prandtl-Type Pitot Tube.
Interoffice memorandum. Environmental Protection Agency. Emission Measurement
Branch. Research Triangle Park, N.C. November 28, 1975.
7. Shigehara, R. T. Adjustments in the EPA Nomograph for Different
Pitot Tube Coefficients and Dry Molecular Weights. Environmental Protection
Agency. Emission Measurement Branch. Research Triangle Park, N.C. August 1974,
8. Standards of Performance for New Stationary Sources. Federal
Register. 36 (247). December 23, 1971.
-------
24
GUIDELINES FOR SAMPLING IN TAPERED STACKS
T. J. Logan and R. T. Shigehara*
INTRODUCTION
Tapering of the inside diameter of stacks is occasionally done when
designing natural draft stacks, when there are special flow or structural
considerations, and for pressure recovery. These tapers seldom exceed a
few degrees. Although guidelines for the selection of a sampling site to
aid in the extraction of a representative sample are given in Method 1 of
the December 23, 1971, Federal Register, no mention is made about tapered
stacks. The purpose of this paper is to provide the necessary background
on how to deal with tapered stacks.
BASIC CONSIDERATIONS
In order to obtain a representative sample, the particles must be
extracted at an isokinetic flow rate. The condition of isokineticity de-
mands that the particles and gases flow directly into the sampling nozzle
and that the velocity be accurately measured. Therefore, two factors must
be considered: (1) the effect of the taper on flow conditions within the
stack and (2) the effect of the taper on velocity determination and parti-
culate matter collection.
Effect of Taper on Stack Flow Conditions
About the only information related to this area was the work done with
venturi meters. The American Society of Mechanical Engineers research on
2
fluid meters indicates that beyond a convergent included angle of 21 degrees
and a divergent included angle of 15 degrees, gas separation from the walls
* Emission Measurement Branch, ESED, OAQPS, EPA, RTP, NC, November 1974
-------
25
is expected to occur. This is undesirable as eddies would be formed,
causing particles and gases to flow in undeterminable directions.
From a physical standpoint, convergent angles of 15 or 21 degrees
would not likely occur in stacks due to the tremendous increase in velocity.
If the larger stack diameter D is used, a tapered stack meeting the minimum
2.5 D requirement of Method 1 would cause an increase in velocity of about
8.6 times at the outlet for a 15-degree included angle and 186 times for
the 21-degree included angle. Such an increase would require considerable
additional power and would be impractical and uneconomical.
One builder of chimmeys3 related that convergent stacks generally do
not exceed 0.5 in/ft. This corresponds to an included angle of about 4.8
degrees for convergent stacks. Divergent stacks are normally designed at
about 5 to 15 degrees.
Based on the above, the 15-degree included angle can be considered the
maximum limit for both convergent and divergent stacks, with the under-
standing that the 15-degree angle will be very unlikely in convergent stacks.
The purpose for making this statement is to form the limit and basis for
evaluating the effect of the taper on the velocity determination and the
particulate matter collection.
Effect of 15-degree Included Angle on Velocity and Particulate Concentration
Convergent or divergent stacks with an included angle of 15 degrees
would cause a maximum 7.5-degree angle of attack on the pi tot tube and par-
ticulate sampling probe nozzle. Data presented by Grove and Smith show
that a 7.5-degree angle will result in velocity measurements with a type-S
-------
26
pi tot tube being biased 3.5 percent high. This higher apparent velocity
also causes particulate sampling to be in error because isokinetic sampling
requires that the sample gas velocity be made equal to the stack gas velocity,
which is in error since it is measured by the misaligned pitot tube. In ad-
dition to the sampling rate being over-isokinetic, the misalignment of the
probe nozzle with the stack gas stream results in a reduction of 0.85 percent
in the effective nozzle area.
The magnitude of the effect on the particulate concentration by being
over-isokinetic and having a reduced nozzle area is a function of particle size.
For particles of less than 1 micrometer, the concentration will not be af-
fected. However, with the larger particles of greater than 50 to 75 micro-
meters, the sampled concentration will be low; a bias of about 4.3 percent will
occur (about 3.4 percent from being over-isokinetic and 0.86 percent from the
reduced nozzle area). In a practical case, where there is a distribution of
particle sizes, the error will be considerably less than the 4.3 percent, and
for well-control led sources where the majority of the particles are charac-
teristically small (<2 micrometers), the error will be near zero.
For pollutant mass rates, the error of the higher measured volumetric
flow rates will cancel out the errors of the lower measured concentrations,
with the true concentration being between the maximum limits of +3.5 and -0.8
percent.
RECOMMENDATIONS
Based on the above discussion, the following guidelines, which should not
-------
27
cause maximum errors greater than 4.3 percent in measured concentration or
3.5 percent in mass rate determinations, are recommended (actual errors for
small particle sizes will be from 0 to -0.8 percent):
1. Consider all stacks with the total included angle of <15 degrees
as straight stacks. If this angle is exceeded, consider the taper
to be a flow disturbance and modify the stack with a straight sec-
tion of at least 2.5 D.
2. Use the maximum diameter at point of upstream or downstream dis-
turbance and Method 1 for determining the sampling point location
and number of sampling points.
REFERENCES
1. Standards of Performance for New Stationary Sources. Federal Register.
Vol. 36, No. 247, December 23, 1971.
2. Fluid Meters, Their Theory and Application, Report of ASME Research
Committee on Fluid Meters. (5th Ed.). American Society of Mechanical
Engineers. New York. 1959.
3. Personal Communication with Richard Lohr, Vice-president, International
Chimney Corp., Buffalo, N. Y.
4. Grove, J. D. and W. S. Smith. Pi tot Tube Errors Due to Misalignment and
Nonstreamlined Flow. Stack Sampling News. 1 (5):7-ll, November 1973.
-------
28
CONSIDERATIONS FOR EVALUATINfi EQUIVALENT STACK
SAMPLING TRAIN METERING SYSTEMS
*
R. T. Shigehara
Introduction
The basic purpose of sampling train equipment is to collect a repre-
sentative sample from a point (small area) within a stack cross-section
or, when conducting a sample traverse, to collect a series of such sam-
ples. To accomplish this, the sampling train must (1) maintain either
isokinetic or proportional sampling rate, depending on whether particulate
or gaseous pollutants are being sampled, (2) efficiently collect reprodu-
cible samples of the pollutant at known levels, and (3) accurately mea-
sure the sample gas volume. Thus, conventional sampling trains incor-
porate some means of gas metering to regulate the sampling flow rate and
to measure the sample gas volume.
"Method 5 - Determination of Particulate Emissions from Stationary
Sources, Section 2.1.6" specifies the above requirements. It states,
"Metering system - Vacuum gauge, leak-free pump, thermometers capable of
measuring temperature to within 5° F, dry gas meter with 2% accuracy, and
related equipment, or equivalent, as required to maintain an isokinetic
sampling rate and to determine sample volume."
There are many different workable metering techniques or systems.
Individual stack samplers and control agencies usually have their own
ideas as to which mode is the best. The purpose of this paper is to pro-
pose criteria to evaluate the different stack sampling train metering
techniques or systems.
* Emission Measurement Branch, ESED, OAHPS, EPA, RTP, NC, September 1974
-------
29
Criteria for Isokinetic Sampling - EPA Particulate Test Train
The EPA particulate test train1'2 will be used as a baseline
reference for this discussion on the development of criteria for
evaluating the different stack sampling train metering techniques or
systems. The EPA train uses the pi tot tube-orifice meter-dry gas
meter system for setting isokinetic rates and for determining sample
gas volume. In this system, the pi tot tube is attached to the probe
so that the gas velocity at each of the sampling points can be constantly
monitored. The observed pi tot tube manometer reading is related to the
orifice meter manometer reading by an equation such that the flow rate
through the sampling train can be adjusted to isokinetic conditions.
To perform these calculations, the EPA train utilizes a nomograph, which
requires as little as 5 to 10 seconds to determine and adjust the sampling
rate after a new velocity reading or a change in stack flow has been
observed. The nomograph is only a type of aid. Graphical techniques or
electronic calculators can also be used to yield the same result. The
dry gas meter is used to measure the sample gas volume and to measure
the gas sampling flow rate independently from the orifice meter.
For particulates, it is the "condition" of isokineticity that ensures
the extraction of a representative point sample, not the "means" by which
the desired sampling rate is achieved. However, it is insufficient to
simply state that all metering systems that have at some time demonstrated-
capability of obtaining isokinetic conditions are equivalent. All techniques,
null balance, pi tot tube-rate meter, pi tot tube-volume meter-timer, and
others, rely on the knowledge, experience, and conscientiousness of the
-------
30
operator and require error-free equipment performance during sampling.
Some assurance that isokinetic conditions were maintained throughout the
actual sampling run is needed.
The EPA pi tot tube-orifice meter-dry gas meter system provides
sufficient proof of isokineticity:
1. All the components can be calibrated against a standard. The
dry gas meter and orifice meter can be calibrated against a wet test
meter (secondary standard) or a spirometer (primary standard). With
periodical calibrations, the dry gas meter can maintain an accuracy of about
1% in volume measurement.3'4 The type-S pi tot tube can be calibrated
against a standard type pi tot tube which generally has a calibration
coefficient between 0.98 and l.OO.5 If one would send a standard pitot
tube to the National Bureau of Standards, a certified calibration for
velocity ranges from 6 to 100 fps or 6 to 155 fps can be obtained.
2. The pitot tube attached to the probe allows the velocity to be
monitored and isokinetic sampling to be maintained throughout the entire
test run. In this manner the sampling can be conducted under normal
everyday process conditions; it is not limited to steady-state conditions
of gas velocity.
Some limitations of the pitot tube for monitoring and measuring
velocity should be recognized: (1) It still relies on the operator to
properly orient the pitot tube into the direction of flow, correctly set
up the manometer, and accurately read the velocity pressure head; (2) the
pitot tube also has a lower velocity limit, usually reported at about 10 fps
This limitation is caused mainly by the difficulty in reading the manometer
-------
31
scale; (3) in addition, the pitot tube is dependent on the density of the
flowing gas stream. Thus, "great" changes in temperature, pressure, and
gas composition (particularly moisture) may cause difficulty in determin-
ing the gas velocity and in setting isokinetic rates. Shigehara et al.
show a method of analysis to determine how much variation in the para-
meters can occur before "significant" errors result.
One method of reducing the problem of setting isokinetic rates when
the gas density or composition of the stack effluent changes "significantly"
with time is to place the orifice meter immediately after the filter, which
is heated to stack temperature and does not allow moisture to condense.
This method eliminates the problem of changes in composition, but adds the
variable of total pressure at the orifice. However, this does not solve
the problem of determining velocity.
The problems of low velocities and great changes in gas density have
not yet been adequately solved. Until better means are specified, we
can only attempt to increase the sensitivity of the manometer for low
velocities and for great changes in gas density, to evaluate the source
conditions and use techniques that, in our opinion, would provide adequate
results.
3. The pitot tube-dry gas meter combination allows an overall
average and individual average point deviations from isokinetic condi-
tions to be calculated for each test run. This is helpful in that it
permits acceptance or rejection of a run based on per cent of isokineticity
actually obtained. The Federal Register allows an overall average
deviation of 10% from isokinetic. Smith et al. have shown a calculation
-------
32
method by which a deviation of 20% can be tolerated with the assurance
that the sample concentration will be within 10% of the true concentration.
4. The orifice meter-dry gas meter combination provides a cross-
check of flow rate and sample gas volume.
However, although both the dry gas meter and the orifice meter can
yield accurate results, there is no means for checking against improper
use or malfunctions under the actual operating conditions of the sampling
train if the components are used separately.
Summarizing, it is the condition of isokineticity that produces a
representative point sample. Any means that provides this condition
could theoretically be considered equivalent. However, as improper
uses or errors do occur, "sufficient proof" can be defined as:
1. All components be calibrated against a standard.
2. Velocity be monitored constantly and simultaneously with sampling.
3. A check of isokineticity actually obtained be provided.
Null Balance Probe System
This system is deceptively simple in principle. Also called static
balance, zero pressure, and isokinetic probes, the pressure null balance
probe is a nozzle specifically designed to measure the static pressure of
the stack gases flowing around and within the probe nozzle. When both static
pressures are equal, isokinetic conditions are said to exist. Cooper8
summarized as follows:
-------
33
"However, numerous problems have been observed in attempting to
accurately maintain true isokinetic sampling conditions because the exis-
tence of equal pressures at outer and inner probe walls does not neces-
sarily mean that equal velocities exist at both points. Differences in
frictional flow losses between inner and outer surfaces caused by turbu-
lence and surface nonuniformities, progressive coating and possible plug-
ging of the inner static tap by particles, and possible differences of
static tap location may all produce these conditions. Parker ("Some
Factors Governing the Design of Probes for Sampling in Particle- and
Drop-Laden Streams," Atmospheric Environment 2;477-490, September 1968) found
that null balance systems had limited usage for large probes greater than
3/4 inch diameter. Toynbee and Parkes ("Isokinetic Sampler for Dust Laden
Gases," International Journal of Air and Water Pollution 6_:113-120, 1962)
postulated that by a slight expansion of the rear section of the probe the
inner frictional losses could be reduced inside the nozzle, and the
system could be used over the velocity range from 600 to 2500 fpm. However,
subsequent comments by Nonhebel in the same issue stated that the plug-
ging problems associated with the inner static taps could not be overcome.
Work by Dennis ("Isokinetic Sampling Probes," Industrial and Engineering
Chemistry 49_:294-302, 1957) and Hemeon and Haines ("The Magnitude of
Errors in Stack Dust Sampling," Air Repair 4_:159-164, November 1954)
indicated that it was not always possible to assure isokinetic sampling
conditions, and found the errors at different velocities for two nozzle
sizes when departing from nozzle conditions."
-------
34
This excerpt illustrates that even with careful calibration for
the specific source and conditions, one cannot be positive that iso-
kinetic conditions existed throughout the entire run. Although the
feasibility of such a system has been demonstrated under controlled
conditions, it suffers from the lack of proof of isokineticity for the
actual operating conditions as provided by the EPA metering system. In
order to provide sufficient proof of isokineticity, the null balance
probe system must incorporate a pi tot tube and a dry gas meter. This
is what Wilson and Falgout9 did to show that their null probe design
was workable.
Dry Gas Meter as a Rate Measurement Device
The volume meter (dry gas meter), in addition to measuring the total
sample volume, could serve as a rate meter for setting isokinetic rate by
timing the needle travel. However, since the needle travel must be
observed for one or more whole revolutions to obtain a reasonably accurate
rate value, the rate is only an average, and changes are possibly delayed
one or more minutes past the time they occur. Thus, its application is
limited to sources where velocity is "fairly" constant. There is also
the disadvantage of not having a cross check of volume and rate under
actual operating conditions as with the orifice meter-dry gas meter
combination.
Proportional Samp!ing
The same criteria apply to proportional sampling as to
isokinetic sampling. It is the condition of proportionality that counts,
not the means by which proportional sampling is achieved.
-------
35
Like the EPA participate sampling train, the same pi tot tube-
orifice meter-dry gas meter system can be used to regulate and check
proportionality. But because of the lower sampling rates used for
gaseous trains, a rotameter is normally used instead of an orifice
meter.
Total Gas Sample Volume
The usual means for measuring the qas sample volume are dry na§
or rate meters such as orifices and rotameters. Cyclones, venturi meters,
evacuated containers, critical orifices, and mass flow rates are also
used. Whatever the means, it is the total gas volume that is desired.
Integrating volume meters such as the dry gas meters, when sized properly,
readily provide the desired result. As mentioned previously, the dry
gas meter can maintain an accuracy of about 1% in volume measurement when
3 4
calibrated periodically against a wet test meter or spirometer. '
Rate meters can also be used to measure the sample gas volume.
However, they measure instantaneous flow, which is subject to density
changes of the gas stream. Therefore, other variables such as time,
temperature, pressure, and pressure drop must be carefully recorded
during the test run so that an integrated total volume can be calcula-
ted or obtained graphically. The same is true with dry gas meters if
they are placed before the pump, because pressure could vary considerably,
at times, during the test run as particulate matter builds up on the
filter material.
-------
36
The EPA sampling train places the dry gas meter and orifice meter
behind the gas pump with the orifice meter open to the atmosphere. There
are several practical advantages with this placement, which requires that
the pump be leak proof. The advantages are:
1. The dry gas meter is subjected to a fairly constant pressure--
the only variation coming from the orifice meter pressure drop, which
is no more than 10 in. of water. The orifice meter is at a relatively
constant atmospheric pressure; therefore, there is no need to record
or to observe for all practical purposes pressure and meter readings
extra carefully.
2. The dry gas meter need not be calibrated under the expected
range of negative pressures that would occur if it were placed before the
pump to compensate for the leakage around the meter diaphragm valves,
particularly under high vacuums.
3. It is not necessary to have special gas meters that can
withstand the high vacuums.
Condensers
Condensers are generally an integral part of a metering system. Their
main purpose is to prevent moisture from condensing within the pump and gas
metering devices. They also serve as a means for the determination of the
average moisture content over the sampling duration.
The EPA test method gives a clear procedure for determining moisture
when the gas stream does not contain water droplets. (If liquid droplets
are present, the gas stream is assumed to be saturated). The probe and
filter holder are heated to a minimum of 225°F so that moisture contained
-------
37
in the sample will remain in gaseous form until the gas has passed the
filter. Following the filter is a series of four Greenburg-Smith impingers
which are immersed in an ice bath. The first two impingers each contain
100 ml of water. This chilled water acts to condense and trap the water
vapor contained in the hot gases coming from the filter holder. The third
impinger is empty and acts as a trap to collect any entrained water which
might be carried over from the first two impingers. Finally, the fourth
impinger contains approximately 200 grams of silica gel. The silica gel
adsorbs most of the moisture which remains in the gas stream; for a 1-hour
sampling run, less than 3% passes through if the temperature at the third
impinger is kept below 70°F and less than 15 in. Hg vacuum. The water
collected in the first three impingers is easily measured volumetrically,
and the weight change in the silica gel gives the amount of moisture
collected there. The amount of moisture in the gas stream thus measured,
and the sample gas volume as measured by the dry gas meter are then used
to determine the moisture content.
The choice of equipment is not important as long as the moisture
collected and leaving the condenser and gas sample volume can be measured
accurately. For long sampling runs (3 to 4 hours), condensation coils
may be better than or as effective as the EPA method. Temperature and
pressure must be measured at the exit of the condenser to account for the
moisture still remaining in the gas stream. However, because at 10 in.
Hg. vacuum and 70°F, the amount of moisture at saturation conditions is
about 3.7% by volume, the silica gel should still be used to protect the
pump and metering devices.
-------
38
Summary
Flow rate regulation and sample volume systems have been discussed.
The basic purpose of these systems is to ensure that a representative
point sample is collected and that the sample gas volume is accurately
measured. However, representativeness is not a direct measurement. Thus,
individual measurements that ensure representativeness must be compared
against a standard. In the absence of any standard, the question of which
result is right when two sampling trains yield different values can never
be answered. When a standard is not available dnd if an evaluation is
desired, design and/or performance criteria which have been scientifically
or arbitrarily derived must be used.
Since it is the condition of isokineticity or proportionality that
is important in the extraction of representative point samples, any
technique that provides these conditions can be used. However, since they
are a vital part of obtaining representative samples, checks under actual
operating conditions must be provided. In this regard, the pltot tube-
rate meter-volume meter system offers clear advantages.
-------
39
References
1. "Standards of Performance for New Stationary Sources," Federal
Register, Thursday, December 23, 1971.
Z. W. S. Smith, et al., "Stack Gas Sampling Improved and Simplified
with New Equipment," Presented at the 60th Annual Meeting of APCA,
Cleveland, Ohio, June 11-16, 1967.
3. American Society of Heating, Refrigerating and Air Conditioning
Engineers Handbook of Fundamentals, ASHRAE, Inc., N.Y., N.Y., 1967.
4. "Source Sampling," Institute for Air Pollution Training, Office of
Manpower Development, USDHEW, NAPCA, P. 0, Box 12055, RTP, N.C. 27711.
5. J. H. Perry, C. H. Chilton, & S. D, Kirkpatrick, Chemical Engineering
Handbook. 4th edition, McGraw Hill, N.Y., 1969.
6. R. T. Shigehara, W. F. Todd, and W. S. Smith, "Significance of Errors
in Stack Sampling Measurements," Presented at the Annual Meeting of
APCA, St. Louis, Missouri, June 14-19, 1970.
7. W. S. Smith, R. T. Shigehara, and W. F. Todd, "A Method of Interpreting
Stack Sampling Data," Presented at the Annual Meeting of APCA, St. Louis,
Missouri, June 14-19, 1970.
8. H.B.H. Cooper, Jr. and A. T. Rossano, Jr., Source Testing for Air
Pollution Control, Environmental Science Services, 24 Danbury Rd.,
Wilton, Conn. 06897
9. K. D. Wilson and D.A. Falgout, "A New Approach to Isokinetic Null
Probe Design," presented at the 65th annual meeting of APCA,
Miami Beach, Florida, June 18-22, 1972.
10. "Standard Method for Sampling Stacks for Particulate Matter," ASTM
Designation D 2928-71, American Society for Testing and Materials,
1916 Race St., Philadelphia, Pa., 19103, 1971.
11. W. L. Johnson, Emission Measurement Branch, Emission Standards and
Engineering Division, Office of Air Quality Planning and Standards,
Environmental Protection Agency, Durham, N. C. (unpublished data),
1973.
-------
an
Published in "Stack Sampling Hews," March 1975, Vol. 2, No. 9.
EVALUATION OF METERING SYSTEMS FOR GAS-SAMPLING TRAINS
M. A. Wortman & R. T. Shigehara
U. S. Environmental Protection Agency
INTRODUCTION
In the December 23, 1971, Federal Register, several types of gas-
sampling trains are specified. Each uses basically the same types of
components in its metering system, i.e. flow control valve, diaphragm pump,
rotameter, and dry gas meter, but differs in the sequence in which they are
arranged. The different sequences are summarized as follows:
1. Method 3 (Integrated Gas Sampling Train): Flow control valve,
diaphragm pump, and rotameter. A flexible bag follows the
rotameter in this train.
2. Method 4 (Moisture Sampling Train): Flow control valve, diaphragm
pump, dry gas meter, and rotameter.
3. Method 6 (S02 Sampling Train): Diaphragm pump, flow control
valve, rotameter, and dry gas meter.
2
A recent publication reported an adverse effect on the calibration of
dry gas meters in particulate sampling trains utilizing diaphragm pumps
with bypass valve systems. Although the gaseous sampling train metering
systems do not use a pump bypass valve, questions were raised on whether or
not this same effect would also be present in the smaller gas-sampling trains,
Thus, tests were conducted to determine the effect, if any, of the position
of the control valve in relation to the pump and metering devices on the
calibration of the dry gas meter.
During the course of the test program, certain problems with the leak
check procedure and the diaphragm pump were encountered. The purpose of
this paper is to report these findings and the results of this test.
-------
41
PROCEDURE
Test Equipment
The test train components used were the same as those specified by
Method 6, as published in the December 23, 1971, Federal Register. A wet
3
test meter (0.05 ft /rev.) was connected to the inlet of the metering
system. A drying tube was inserted immediately after the wet test meter
to protect the rotameter, dry gas meter, and pump from moisture condensation.
Schematics of the two sampling train arrangements used to determine the
effect of valve position are shown in Figure 1.
Test Procedure
The test was conducted in the following manner:
1. A leak check was first conducted. This leak check consisted of
plugging the inlet to the metering system (before the drying tube),
leaving the control valve fully open, turning on the pump, and
noting the travel of the dry gas meter dial. If any leaks were
indicated, they were corrected before any test was conducted.
2. Using the rotameter as a flow rate indicator, the following infor-
mation was gathered: rotameter reading, wet test meter reading
and temperature, dry gas meter readings and temperature, barometric
pressure, and running time. From the raw data, two values were
computed: (1) the calibration factor (F), which is the ratio of dry
gas meter volume to wet test meter volume, and (2) average standard
flow rate (Q) obtained by dividing the wet test meter volume, after
being corrected for moisture content, by the running time.
-------
o
Thermometers
Wet test
meter
Rotameter
i
Silica gel tube
Valve
Diaphragm
pump
Dry gas
meter
O
P
O
ro
Figure 1
Initial test sampling train arrangements
-------
43
TEST RESULTS
Four different sampling trains were tested, each with the valve before
and after the pump. During one of the tests, with the valve placed after
the pump and closed completely, movement of the wet test meter dial was
noted. A leak check of the pump with a mercury manometer revealed a leak,
which was not detected by the normal leak check procedure. (This leak was
occurring where the diaphragm was connected by two screws to the connecting
rod.) Plots of the calibration factor, F, versus the flow rate, Q, for
pumps with and without leaks are shown in Figure 2.
During these tests, it was also noted that with the valve placed before
the pump, the rotameter readings were greatly affected due to the pulsating
motion of the diaphragm. But there was less of an effect on the calibration
factor over a wider flow range with this arrangement than with the valve
placed after the pump. The calibration factor was also less affected by
leaks, when present^with this arrangement. Since these were desirable
characteristics, steps were taken to reduce the effect of the pulsations.
This was easily accomplished by placing a surge tank between the pump and
the rotameter or by using the dry gas meter as a surge tank, i.e., inter-
changing the position of the dry gas meter and rotameter. The results are
shown in Fiqure 3. Using the dry gas meter as the surge tank, however,
caused the control response of the rotameter to be sluggish. Therefore,
the surge tank placed before the rotameter was selected, and the final train
shown in Figure 4 was used for subsequent tests.
After ensuring that all systems were leak free, this time using the
manometer or the wet test meter procedure for the leak check, the tests
were rerun. The results are shown in Figure 5.
-------
44
1 ?fl
1.10
1.00
0.90
0.80
1 1
A A TRAIN NO. 10 WITH
—
. A A
A A A A ^
0 0
ft O
O °
O
o
0
o . ,
1 1 — 1
LEAK
—
1.30
1.20
1.10
1.00
o
£ 0.90
o
-------
12
10
LLJ c
I
O
c! 4
O DRY GAS METER AS SURGE TANK
* SURGE TANK
4 6 8
ROTAMETER SETTING
Figures. Rotameter calibration.
10
12
en
-------
Thermometers
Valve
Silica gel tube
CTl
Wet test
meter
Diaphragm
pump
Dry gas
meter
Figure 4
Final test sampling train arrangement
-------
1.00
0.90
0.80
TRAIN NO. 10
o
1.10
1.00
K 0.90
o
| i-io
t-
O£
3 1.00
o
0.90
TRAIN NO. 7
_ o o
o ° 000° °
1 1 1
-o o o ° 0 ° o ° ooo
1 1 1
TRAIN NO. 6
1.00
0.90
0.80
o O
TRAIN NO. 1
Z
10
12
FLOW RATE (Q), dscfh
Figure 5. Calibration factor versus flow rate for Figure 4trains.
-------
48
SUMMARY AND RECOMMENDATIONS
The results of these tests showed that a constant dry gas meter cali-
bration factor could be obtained whether the control valve was placed
before or after the pump. However, the placement of the valve before the
pump provided a constant calibration factor over a wider flow range and
was not as greatly affected by leakages from within the pump. It is
recommended that the metering system shown in Figure 4 be used for gaseous
sampling.
The present leak check procedure was found to be inadequate. It is
suggested that leak checks be conducted by either of the following two
procedures: (1) connect a wet test meter at the inlet of the sampling train,
turn on the pump, pinch off the line after the pump, and note wet test meter
dial (suggested for laboratory), or (2) connect a vacuum gauge (mercury
manometer, bourdon gauge, or similar) at the inlet, turn on the pump, pinch
off the line after the pump, turn off the pump after maximum vacuum is
reached, and note gauge reading (suggested for field use). Any movement
of the wet test meter dial or vacuum gauge reading denotes a leak and
must be corrected.
REFERENCES
1. Standards of Performance for New Stationary Sources. Federal Register.
Vol. 36, No. 247. December 23, 1971. p. 24882-24895.
2. Smith, W. S. When Your Valves Float. Stack Sampling News. Vol. 7,
No. 1. January, 1974. p. 5-8.
-------
AN EVALUATION OF THE CURRENT EPA METHOD 5
FILTRATION TEMPERATURE - CONTROL PROCEDURE
a
Robert F. Vollaro
Introduction
Method 5, promulgated in the December 23, 1971 Federal Register , re-
quires the use of probe and filter holder heating systems during isokinetic
sampling. Prior to sampling, these heating systems are adjusted as follows:
(1) the probe heater is set to provide a gas temperature of about 250°F* at
the probe outlet; probe heater settings are obtained from Figure 21 of the
2
sampling train operations manual , APTD-0576 (Figure 2 of this report); (2)
the sample box thermostat is set to provide a temperature of approximately
250°F* around the filter holder. Although it is not explicitly stated in
Method 5, one of the primary reasons for making these temperature adjustments
is so that filtration will take place at 250°F +_ 25°F*.
Recently, however, some observers have expressed concern over the ade-
quacy of the above filtration temperature control procedure, particularly
whether probe heater setting estimates made from the APTD-0576 reference curves
will actually provide probe outlet temperatures around 250°F under field test
conditions. Among the reasons given are: (1) the curves give no specific
probe heater setting guidelines for sources with temperatures above 250°F or
below 80°F; (2) the temperature of the gases surrounding a sample probe during
an actual traverse will seldom be 80°F, which is the temperature base from
which the curves are derived; and (3) the curves are strictly applicable only
to gas streams of low moisture content. These comments fail to note that the
*Unless otherwise specified by a particular regulation.
a Emission Measurement Branch, ESED, OAQPS, EPA, RTP, NC, July 1975
-------
50
reference curves were not originally intended to provide exact filtration
temperature control; their original purpose was to furnish approximate
guidelines by which moisture could be prevented from condensing ahead of the
impingers. There is, nevertheless, question as to whether probe outlet tem-
peratures around 250°F can be generated with confidence, even with the sample
box set at 250°F.
In light of the above question, experiments were conducted, under a num-
ber of simulated field test conditions, to evaluate the present means of fil-
tration temperature control. This paper reports the results of these experi-
ments.
Experimental Set-up
The Method 5 sampling train configuration used in the experiments is
shown in Figure 1. The components of the train met the design specifications
outlined in the source sampling equipment construction manual, APTD-0581 ,
except for the modifications necessary to facilitate temperature monitoring
at the probe inlet, probe outlet, and inside the back half of the filter holder.
Chromel-alumel thermocouples, insulated from the metal parts of the train,
were used to monitor temperature in these experiments.
Filtration Temperature vs. Probe Outlet Temperature
Preliminary experiments were conducted to establish a relationship between
probe outlet temperature and filtration temperature, at constant sample box
setting. At each of three different box settings* (220, 240, and 260°F), the
* Note that the term "sample box setting," as used in this report, refers to
the average temperature inside the box during a sample run. During sampling,
the box temperature changed continually with time, rising and falling in 5-
minute cycles between thermostatically controlled upper and lower limits.
-------
PROBE INLET
THERMOCOUPLE
SAMPLE BOX
THERMOCOUPLE
PROBE OUTLET
THERMOCOUPLE
FILTRATION TEMPERATURE
THERMOCOUPLE
HEATED SAMPLE
PROBE
THERMOMETER
HEATED
SAMPLE f
BOX
f
»
i:
•- N.
* V t
H
THERMOMETERS ( )
r*T
CALIBRATED
ORIFICE
'100 ml OF WATER J ICE
I BATH
CONTROL
SILICA GEL
VACUUM
GAUGE
MANOMETER '-
Figurel. Sampling train configuration.
-------
52
400
u.
o
K 300
cc
m
| 200
100
ca
o
K
3-ft PROBE (5-minWARMUP)
INLET. 250 °F
INLET. 150 °F
INLET AMBIENT, 80 °F
4-ft PROBE (10-minWARMUP)
— INLET, 250 °FV
INLET. 150 °F
INLET AMBIENT. 80 °F
20 40 60 80 0 20
POWERSTAT SETTING, percent
40 60
80
400
5-ft PROBE (10-min WARMUP)
6-ft PROBE (15-min WARMUP)
INLET. 250 °F>
INLET, 150 °F
20 40
60 80 0 20 40 60 80
POWERSTAT SETTING, percent
oc
in
400
300
200
o 100
I I I
7-ft PROBE (15-min WARMUP)
— INLET. 250 °F
INLET, 150 °F
INLET AMBIENT,
0 20 40 60
POWERSTAT SETTING, percent
80
Figure 2.Probe temperatures.
-------
53
temperature of the gas at the probe outlet was varied from 100 to 450°F while
holding the sampling rate constant at 0.75 cfm. The results of these exper-
iments are presented graphically in Figure 3.
Figure 3 shows that at constant sample box temperature, filtration temper-
ature is a linear function of probe outlet temperature, requiring a 2.3°F
change in probe outlet temperature to effect a 1°F change in filtration temper-
ature.* Figure 3 also shows that, with the sample box set at its customary
250°F, it is necessary for the probe outlet temperature to be maintained be-
tween 230° and 350°F, in order for the filtration to take place at 250 +_ 25°F.
Probe Outlet Temperatures
Further experiments were conducted, under a number of simulated field
test conditions, to determine whether heater setting estimates made from the
APTD-0576 curves (See Figure 2) would provide the necessary probe outlet temper-
atures to keep the filtration temperature between 225 and 275°F. Temperature
was monitored during each run at the probe inlet, at the probe outlet, inside
the sample box, and inside the filter holder, just behind the glass frit (See
Figure 1). A constant sample rate of 0.75 cfm was maintained for all experi-
ments. The following test cases were considered:
Test Case I—Possible Underheating. In this experiment, cold air at 37°F
was drawn through a 3-foot sample probe. The sample box temperature was set
at 260°F, and the temperature of the gases surrounding the probe was 37°F. In
the absence of specific guidelines from APTD-0576 for probe inlet temperatures
below 80°F, the probe heater was set according to the "closest available" probe
*This value will, of course, be a function of sample box design and the
path length that the gas must go through (e.g., if a cyclone is used); however,
a similar relationship should exist for different configurations.
-------
54
500
450
400
u. 350
300
250
2
a
UJ
fe
200
150
100
50
DESIRED OPERATING LIMITS
• 220°F DATA
A 240°F DATA
• 260°F DATA
50 100 150 200 250 300 350 400
STEADY-STATE PROBE OUTLET TEMPERATURE, °F
450 500
Figure 3. Filtration versus probe outlet temperatures (dry air).
-------
55
inlet temperature curve, namely the 80°F curve. Case I was designed to simulate
sampling from an ambient source with a short probe on a cold day. Its purpose
was to determine if the probe heater was capable of heating cold sample gases,
having only a short residence time in a probe set in cold surroundings, to an
acceptable probe outlet temperture.
The results of this experiment are presented in the Appendix (See Table I).
The data show that after a few minutes, the filter temperature had risen above
225°F; it continued to climb slightly thereafter, reaching a steady-state value
of about 235°F. These results indicate that, even when a cold gas stream
(T«80°F) is sampled with a short probe set in cold surroundings, setting the
probe heater by the appropriate 80°F inlet curve of APTD-0576 is satisfactory.
Very little reduction in heater performance occurs, and a steady-state value
of filter temperature safely within the range 250 + 25°F is rapidly established.
Test Case II--Possible Overcooling. Hot sample air at 475°F was drawn
through an 8-foot probe set in 80°F surroundings; the sample box thermostat was
set at 240°F. In the absence of APTD-0576 guidelines for sources hotter than
250°F, the probe heater powerstat was arbitrarily set at 25 percent. Case II
was designed to simulate the testing of a very hot source (T>300°F) with a
long sample probe. More specifically, Case II represents the outset of the
sample traverse, when points close to the near stack wall are tested (i.e., when
a good part of the probe is outside the stack), and overcooling of the sample
gas can occur before it enters the filter box.
After a few minutes of the Case II sample run (See Table II in the Appendix),
it was noted that the sample gases were cooling from 475°F at the probe inlet
-------
56
to 175°F at the outlet. During this same time span, the filtration temperature
reached only 190°F. For the remainder of the test, the probe heater setting
was gradually increased, at 10-minute intervals, until filtration temperatures
consistently above 225°F were obtained. When 225°F was reached, the powerstat
setting was at 75 percent of maximum. Thus, the Case II data indicate the
importance of proper probe heater calibration if the desired level of probe
outlet temperature is to be achieved at the outset of the traverse of a very
hot stack; random guessing at the powerstat setting to be used will not suffice.
Test Case III—Possible Overheating. A 20-inch diameter incinerator
duct in which hot (520°F) combustion gases were flowing was sampled with a
3-foot probe. The probe was inserted as far as it would go into the duct,
leaving about 16 inches of it exposed to the ambient (40°F) air. The sample
box thermostat was set at255°F. Again, in the absence of an APTD-0576 guide-
line, the probe heater was arbitrarily set at 20 percent. Case III was designed
to simulate that stage of the sample traverse of a very hot (T>300°F) stack
when points close to the far wall are tested and a good part of the probe is
inside the stack, surrounded by hot gases. The purpose of this test was to
check for possible overheating.
The Case III data (See Table III in Appendix) show that although the temper-
ature was leveling out, overheating of the filter occurred after 16 minutes of
sampling. After 17 minutes, the probe heater was shut off to try and bring the
filtration temperature back below 275°F. The filtration temperature did drop
to 274°F; however, had the ambient temperature been higher than 40°F or the
-------
57
stack gas temperature higher than 520°F, overheating would most likely have
continued, and to achieve lower temperatures an adjustment in the sample box
temperature would have become necessary.
To determine the severity of filter overheating, had an arbitrary powerstat
setting higher than 20 percent been chosen, Case III was repeated. This time,
the probe heater setting was gradually increased, at 12-minute intervals, from
25 percent to 80 percent of maximum. These data are shown in Table IV (See
Appendix). Filtration temperatures well in excess of 300°F occurred at the
higher powerstat settings.
Test Case IV—Effect of Moisture. During Test Case III, when incinerator
gases were sampled, the sample box was set at 255°F. However, a plot of fil-
tration temperature versus probe outlet temperature (See Figure 4) produced a
data line well above the 225°F region of Figure 3. It was assumed that the
high moisture content of the combustion gases caused the difference. To check
this assumption, four test runs were performed, in which moist air (estimated
at 5 to 10 percent) at different temperatures (228, 270, 293, and 468°F) was
sampled. The sample box was maintained at about 250°F.
The results of these tests are plotted in Figure 4. They confirm that
moisture in the sample stream can alter the relationship between the probe
outlet and filtration temperatures.
Conclusions
An evaluation of the present means of controlling filtration temperature
in the EPA Method 5 train has demonstrated that:
-------
58
500
450
400
S- 350
300
S 250
u 200
§ 150
100
50
CASE III DATA; SAMPLE BOX@25S°F
1
DESIRED OPERATING
LIMITS ~~
i
SAMPLE BOX @ 260°F
.SAMPLE BOX @ 240°F
SAMPLE BOX @220°F
50 100 150 200 250 300 350 400 450 500
STEADY-STATE PROBE OUTLET TEMPERATURE, °F
Figure 4. Filtration versus probe outlet temperatures.
-------
59
1. At constant sample box setting, filtration temperature is a linear
function of probe outlet temperature. High (>5 percent) sample-stream moisture
content (or presence of water droplets) can, however, alter this relationship.
2. The APTD-0576 reference curves provide reasonable estimates of
probe outlet temperature when applied to the sampling of streams similar to
those upon which they are based, i.e., streams of low moisture content with
temperatures between 80°F and 250°F. The use of the 80°F "inlet curves for
ambient streams with temperatures as low as 37°F has been shown to be satisfac-
tory (Case I); by analogy, it can be inferred that the 250°F inlet curves will
apply reasonably well to low-moisture streams with temperatures up to about
300°F.
3. For very hot stacks (T>300°F) and for stack gas streams of high moisture
content (or containing water droplets), it has been demonstrated that the prac-
tical value of the APTD-0576 reference curves diminishes considerably. In
these cases, sample gas overheating or overcooling at the probe outlet can
occur (depending on the probe heater setting and the temperature of the gases
surrounding the main body of the probe) and can cause the filtration temper-
ature to be outside the desired operating limits.
-------
60
REFERENCES
1. "Standards of Performance for New Stationary Sources," Federal Register.
December 23, 1971.
2. Rom, Jerome, J.. Maintenance, Calibration, and Operation of Isokinetic
Source-Sampling Equipment, Environmental Protection Agency. Publication
No. APTD-0576. Research Triangle Park, N. C. 27711. March, 1972.
3. Martin, Robert M.. Construction Details of Isokinetic Source Sampling
Equipment. Environmental Protection Agency. Publication No. APTD-0581.
Research Triangle Park, N. C. 27711. April, 1971.
-------
61
APPENDIX
-------
62
TABLE II: CASE II DATA
Date: 1/7/75
Case: II
Sample Rate: 0.75 cfm
Sample Box Setting: 240 °F
Probe Heater Setting: 25%
Operator: R. Vollaro
Time
(Minutes)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Probe Inlet
Temperature
(°F)
455
457
466
470
472
474
476
477
477
477
477
479
476
476
475
477
477
477
477
477
476
475
475
475
474
Probe Outle
Temperature
(°F)
153
157
161
164
167
170
172
174
177
179
182
188
193
195
197
198
200
201
203
204
208
213
217
221
225
Sample Box
Temperature
(°F)
243
220
241
260
262
231
216
242
267
256
230
245
269
247
226
223
258
265
241
223
234
256
262
240
219
Filtration
Temperature
(°n
181
183
183
185
188
192
193
192
193
196
198
198
199
203
204
204
204
205
208
209
209
208
211
215
216
Powerstat
Setting
(%-)
25
25
25
25
25
25
25
25
25
25
40
40
40
40
40
40
40
40
40
40
60
60
60
60
60
-------
63
TABLE II
(Continued)
Time
(Minutes)
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Probe Inlet
Temperature
r°n
474
475
474
473
473
472
471
472
472
471
471
472
475
472
471
Probe Outlet
Temperature
(°n
228
229
231
234
235
241
244
250
254
257
261
262
265
268
269
Sample Box
Temperature
(°F)
231
257
259
238
222
232
256
257
235
220
235
263
254
234
222
Filtration
(Temperature
f«n
216
216
219
222
223
222
222
225
229
230
230
231
234
236
237
Powerstat
Setting
f*)
60
60
60
60
60
75
75
75
75
75
75
75
75
75
75
-------
64
Date:
Case:
TABLE III: CASE IIIA DATA
12/6/74
IIIA
Sample Rate: 0.75 cfm
Sample Box Setting: £5jj_
Probe Heater Setting: 20%
Operators: R. Vollaro and R. Mobley
Time
(Minutes)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
'robe Inlet
Temperature
(°n
508
509
510
513
514
511
513
514
516
515
514
517
519
521
522
524
529
534
535
532
532
533
532
533
535
Probe Outlet
Temperature
cn
259
272
280
286
291
296
299
302
304
305
306
307
309
310
311
314
314
313
309
306
305
304
304
304
304
Sample Box
Temperature
cn
272
269
245
225
236
266
273
247
226
253
276
283
256
234
251
277
274
249
228
235
265
282
256
235
236
Nitration
Temperature
f«F)
223
235
245
252
255
257
262
266
267
267
269
272
275
275
275
275
276
278
278
276
274
274
275
276
274
Powerstat
Setting
(*)
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
0
0
0
0
0
0
0
0
-------
65
TABLE IV: CASE 11 IB DATA
1/27/75
IIIB
Date:
Case:
Sample Rate: 0.75 cfm
Sample Box Setting: 255 °F
Probe Heater Setting: 25%
Operator: R. Vollaro
Time
(Minutes)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Probe Inlet
Temperature
?°n
450
453
454
455
456
457
456
457
458
462
463
466
469
470
470
467
472
473
476
477
476
478
479
485
486
3robe Outlel
Temperature
ifon
256
277
281
284
288
292
295
299
301
304
305
308
313
316
320
322
326
327
330
332
333
335
336
338
344
Sample Box
Temperature
?-n
279
248
225
243
269
269
243
220
256
278
262
237
234
256
278
264
238
225
262
282
257
231
237
275
266
Filtration
Temperature
7°F)
212
230
243
251
256
262
266
268
269
271
274
277
277
277
279
283
286
286
286
288
291
293
292
292
295
Powers tat
Setting
m
25
25
25
25
25
25
25
25
25
25
25
25
40
40
40
40
40
40
40
40
40
40
40
40
60
-------
66
TABLE IV
(Continued)
Time
(Minutes)
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
'robe Inlet
Temperature
(°F)
493
490
490
492
491
492
492
491
492
496
497
500
502
503
503
509
506
508
507
511
511
511
509
Probe Outlet
Temperature
(°F)
352
357
361
365
367
371
372
373
375
376
379
395
404
413
417
424
428
431
432
434
435
436
438
Sample Box
Temperature
(°F)
244
224
248
275
273
247
227
253
273
274
247
228
261
286
269
243
224
256
275
283
269
251
227
Filtration
Temperature
(°F)
298
300
301
304
307
310
312
312
312
315
316
317
319
325
328
333
335
337
338
339
341
343
344
Powerstat
Setting
w
60
60
60
60
60
60
60
60
60
60
60
80
80
80
80
80
80
80
80
80
80
80
80
-------
67
LABORATORY EVALUATION OF SILICA GEL COLLECTION
EFFICIENCY UNDER VARYING TEMPERATURE AND PRESSURE CONDITIONS
Peter R. West!in and Fred C. Biddy*
Introduction
The impinger section of the EPA Method 5 sampling train is intended
to collect moisture from sample gases for determination of moisture con-
tent. The final stage of the collection train is an impinger with silica
gel. Laboratory experiments were conducted in order to determine the
effectiveness of the silica gel impinger as a moisture collector under
various sampling conditions of temperature and pressure.
Equipment Set-up
Figure 1 shows the sampling train as it was used in the experiments.
The moisture and heat source was a flask of water heated by a rheostat con-
trolled hotplate. The sample entered the train through a flow control-valve
used to simulate resistance through the sampling train. The first and
second impingers of the train were each filled with 100 mill inters (ml) of
water. The third impinger was a dry impinger with wet-bulb and dry-bulb
thermometers attached to the center tube. In order to reach the gas velocity
necessary to obtain correct wet=bulb temperature readings, a 2.2-cm (0.87-in.)
diameter orifice was placed in this impinger, and the thermometer tips were
located in the orifice opening. The fourth impinger contained approximately
200 grams (g) of silica gel for each run. The silica gel was grade 42 and a
6-16 mesh size indicating type. Following the impinger section was another
thermometer for measuring the temperature of the gas stream leaving the silica
gel impinger. A standard EPA Method 5 meter box was used to draw and measure
* Emission Measurement Branch, ESED, OAOPS, EPA, RTP, NC, July 1975
-------
68
FLOW RESTRICTING
VALVE
STEAM
WATER-FILLED DRY IMPINGER SILICA GEL
IMPINGERS WITH IMPINGER
ORIFICE
TO METER
BOX
ICE BATH
Figure 1. Laboratory moisture sampling train.
-------
69
the volume of sample and measure the vacuum in the volume meter.
Procedure
At the beginning of the test run, the flow control valve and the meter
box pump were adjusted to attain the desired meter vacuum and flow rate.
Flow rate was maintained between 1.1 and 1.3 standard cubic meters per hour
(scm/hr)(0.52 to 0.61 scf/min). Temperature in the impingers was controlled
and maintained with an ice bath. Readings of meter volume, meter temperatures,
train vacuum, wet-bulb and dry-bulb impinger temperatures, and exit-gas tem-
perature were recorded at 5-minute intervals during the 2-hour runs. Adjust-
ments to flow rate and train vacuum were made as necessary. The vacuum in
the third impinger was determined so that the moisture content of gas enter-
ing the silica gel impinger could be calculated. This sample vacuum, reported
in Table 1, was varied from 107 to 460 millimeters of mercury (mm Hg)(4.2 to
18.1 in. Hg).
The moisture entering the silica gel impinger, the moisture entering the
meter box, and the moisture collection efficiency of the silica gel were
calculated as follows:
1. Moisture fraction entering silica gel:
PS -0.00066(1 + 0.00116b T^HPjMT^ - Twb)
Bwi fT
Where:
B . = volume fraction of moisture, %/100
-------
Table 1. MEASURED AND CALCULATED VALUES RELATIVE TO
MOISTURE COLLECTED IN SILICA GEL IMPINGER
IN EPA METHOD 5 SAMPLING TRAIN DURING LABORATORY TESTS3
Run
A
B
C
D
E
F
G
H
I
J
K
a i
Sample
vacuum,
mm Hg
107
109
107
107
107
224
226
226
460
460
460
T.
db
Inlet
temp. ,
°C
9.2
15.2
26.6
21.6
32.6
9.9
19.3
33.8
9.7
21.2
33.7
T
'e
Exit
temp. ,
°C
14.2
18.5
30.0
24.1
34.9
14.5
25.0
35.2
16.8
27.0
35.8
V
m
Meter
vol ume
scm
2.490
2.257
2.458
2.494
2.491
2.671
2.496
2.470
2.520
2.546
2.536
e
sg
Collected
H20 vol.,
scm
0.024
0.034
0.068
0.052
0.063
0.035
0.059
0.076
0.037
0.060
0.065
V + e
m sg
Total
volume,
scm
2.51
2.29
2.53
2.55
2.55
2.71
2.56
2.55
2.56
2.61
2.60
B .x 100
Wl
Inlet
H20,
%
1.3
2.0
4.0
3.0
5.7
1.7
3.2
7.5
3.0
6.4
13.2
e.
i
Inlet
H20 vol . ,
scm
0.032
0.046
0.102
0.076
0.145
0.047
0.081
0.190
0.078
0.167
0.343
eo
e
Exit
H20 vol . ,
scm
0.007
0.012
0.034
0.024
0.082
0.012
0.022
0.114
0.041
0.106
0.278
ee
^
m
Exit
H20,
%
0.3
0.5
1.3
0.9
3.2
0.4
0.9
4.5
1.6
4.1
10.7
e
100 x I9-
i
Collection
efficiency,
%
78
74
67
68
43
75
73
40
48
36
19
a Symbols above columns refer to calculations section of text.
-------
71
P_ = saturated vapor pressure at T. , mm Hg
S WD
T .= wet-bulb temperature, °C
T..= dry-bulb temperature, °C
P. = absolute impinger pressure, mm Hg
2. Volume of moisture collected by silica gel:
esg= 1.342 x Iff3 (Msg)
Where:
e = moisture gas volume in silica gel, scm
M = mass of water collected in silica gel, g
Note: Standard temperature and pressure are 21°C (70°F) and 760
mm Hg (29.92 in. Hg)
3. Total moisture volume entering silica gel:
e1 = Bwitesg + VJ
Where:
e. = moisture volume entering silica gel, scm
V = standard dry-gas meter volume, scm
4. Collection efficiency of silica gel:
E = 1
Where:
E = collection efficiency, %
-------
72
5. Moisture exiting the silica gel impinger:
ee = ei esg
Where:
e = moisture volume exiting silica gel, scm
Discussion of Results
Table 1 shows the results of 11 test runs at three different train
vacuums. Note that the calculated values represent averages over each of
the complete 2-hour runs. As can be seen from this table, the moisture
collection efficiency of the silica gel decreased as the inlet temperature
and the exit-gas temperature increased. A result of this relationship
is that the percent of moisture in the sample gas entering the meter box in-
creased from 0.3 percent at 14.2°C (58°F) exit temperature to 3.2 percent
at 34.9°C (95°F) exit temperature at the same train vacuum of 107 mm Hg
(4.2 in. Hg).
Also shown in these results is the effect of sample train vacuum on
collection efficiency. For example, looking at runs A, F, and I, the inlet
temperatures are approximately equal at 9.5°C (49°F) while the train vacuum
varies from 107 mm Hg (4.2 in. Hg) to 460 nm Hg (18.1 in. Hg). The silica
gel moisture collection efficiency decreases from 78 percent at 107 mm Hg
(4.2 in. Hg) to 48 percent at 460 mm Hg (18.1 in. Hg). Moreover, the amount
of moisture exiting the silica gel increases from 0.3 percent to 1.6 percent
over the same conditions. Figure 2 shows graphically the effect of both
exit temperature and sample vacuum on the moisture content in the exit gas
from the silica gel impinger.
-------
73
100
200
300
SAMPLE VACUUM, mm Hg
400
Figure 2. Silica gel exit moisture content versus sample vacuum and exit temperature.
-------
74
Note that the exit gas from the silica gel impinger has a temperature
several degrees higher than the inlet temperature of the silica gel im-
pinger. This temperature difference was somewhat influenced by temperature
of the room, but also by the exothermic reaction that occurred when moisture
was adsorbed by the silica gel.
2
Data from similar EPA experiments reported by Johnson substantiate
the findings of this study. For example, at an impinger temperature of
21 °C (70°F) and a vacuum of about 254 mm Hg (10 in. Hg), Johnson found
that the moisture collection efficiency was from 60 to 70 percent, comparing
favorably with the 70 percent predicted by the curve in Figure 3. Other
values, difficult to compare because temperature conditions and pressures are
different, in general, show trends similar to the results noted here. Table
2 shows the results of Johnson's study. In addition, further experiments by
Johnson showed that adding one or two more silica gel impingers did little
to decrease the moisture content of the final exit gas. This implies that
the moisture collection ability of silica gel in the EPA Method 5 train is
limited by temperature and pressure conditions.
The moisture content in the exit gas of the silica gel also affects
the dry gas meter volume. A 5 percent increase in moisture content produces
a similar increase in volume. In source sampling results, this means a 5
percent error in the isokinetic calculations and an error in the emission
calculations.
Conclusions
The moisture content of the gas entering the meter box can be greatly
-------
75
10
200
400
300
SAMPLE VACUUM, mm Hg
Figure 3. Silica gel moisture collection efficiency versus sample vacuum and exit temperature.
500
-------
76
Table 2. Results of W. L. Johnson's Study of
2
Moisture Collection Efficiency of Silica Gel
Run
A
B
C
D
E
F
G
Vacuum
mm Hg
508
508
508
381
254
152.4
127
Impinger
temp . ,
°C
-
18.3
16.7
21.7
21.7
21.7
16.7
Exit gas
temp. ,
°C
27.8
21.1
20.0
21.7
20.6
22.8
20.0
Collection
efficiency,
%
38
52
61
46
60
52
84
-------
77
affected by the temperature and vacuum of the sampling train during Method
5 testing. The moisture in the sample gas is incorrectly measured as "dry"
gas by the volume meter, and this value is carried through the isokinetic
calculations as well as the concentration calculations. A volume measure-
ment error due to moisture in the sample gas directly affects the isokinetic
calculations; a 3 percent increase in moisture content of the sample gas
produces a similar error in the isokinetic results.
Method 5, as written, stipulates that the sample temperature as it
exits the silica gel impinger exit gas be held below 21°C (70°F) and that
the sample train vacuum be held under 381 mm Hg (15 in. Hg). These tests
show that at these limits the "dry" gas volume error would be less than 2
percent, and a similar error would appear in the isokinetic determination.
It is noted in the text that the temperature of the wet gas in the
third dry impinger was 4° to 7°C (7° to 13°F) less than the temperature of
the exit gas from the silica gel. This difference is influenced by the heat
of adsorption of the silica gel and ambient conditions. These tests were
run under steady-state ambient temperature conditions and therefore do not
reflect results that may be obtained under field conditions. A better field
indicator of acceptable temperature limits for the sample gas would be the
dry-bulb temperature in the dry impinger preceding the silica gel. A limit
of 15.6°C (60°F) in the impinger would meet the intentions of the present
EPA Method 5 specifications.
References
1. Title 40 -- Protection of the Environment, Part 60 -- Standards of Per-
formance for New Stationary Sources. Federal Register. 36_ (247): 24888,
December 23, 1971.
-------
78
2. Johnson, William L., "Moisture Collection Efficiency of Silica Gel in
Stack Sampling Trains," Environmental Protection Agency, National
Environmental Research Center, Research Triangle Park, North Carolina
unpublished report, 1974.
-------
79
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
Research Triangle Park, North Carolina 27711
SUBJECT: spurious Acid Mist Results Caused by Peroxides in DATE: January 22, 1976
Isopropvl Alcohol Solutions Used in EPA Test Method (M-8)
FROM: Dr. josepTT. Knoll, QAB/EMSL (MD#77)
TO.
Mr. Roger T. Shigerhara, EMB/ESED (MD#19)
An evaluation study of EPA Test Method (M-8) for the Determination
of Sulfuric Acid Mist and Sulfur Dioxide Emissions from Stationary Sources
has been carried out in the Quality Assurance Branch. One result of this
study has been the finding that peroxide impurities in the isopropyl alcohol
used for acid mist collection can convert sulfur dioxide to sulfuric acid
and result in erroneously high acid mist values. The quantities of sulfur
dioxide collected as sulfuric acid were of the order of from ten to twenty
five percent of the EPA compliance standard. It was independent of the
quantity or concentration of sulfur dioxide that had passed through the
system and only dependent on the quantity of peroxide, traces of which may
occasionally be found in reagent grade isopropyl alcohol.
The following test is tentatively proposed for detecting peroxides in
isopropyl alcohol:
Shake 10 ml of isopropyl alcohol with 10 ml of freshly
prepared 10% potassium iodide solution. Prepare a blank
by similarly treating 10 ml of distilled water. After
one minute, read the absorbance at 352 nm. If absorbance
exceeds 0.1, reject alcohol for use.
Peroxides may be removed from isopropyl alcohol by redistilling or
by passage through a column of activated alumina. However, it is possible
to obtain reagent grade isopropyl alcohol with suitably low peroxide levels
from commercial sources, so that rejection of contaminated lots may be a
more efficient procedure.
cc: M. R. Midgett
EPA Form 1320-6 (Rev. 6-72)
-------
80
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
SUBJECT: Determination of Isopropanol Loss During Method 8 DATE:
Simulation Tests
FROM: Peter R. Westlin, Test Support Section
Emission Measurement Branch (MD 19)
TO: Roger T. Shigehara, Chief, Test Support Section
Emission Measurement Branch (MD 19)
JUN 2 5 197&
In answer to questions regarding potential loss of isopropanol(IPA)
through evaporation and a subsequent error in moisture determination
when using Method 8, a laboratory program was undertaken at the IRL
during June 17 and 18. A Method 8 sampling train was set up without
the glass filter between impinger 1, the IPA impinger, and number 2, the
first hydrogen peroxide, \\2Q2» impinger. The third impinger in the
train was also a \\2^2 impinger, while the fourth was left dry. The
fifth and last impinger contained silica gel. A standard Method 5 meter
box was used to draw and measure the volume of the sample.
Two test runs were completed. For the first run, 200 milliliters(ml)
of IPA was placed in the first impinger, 100 ml of ^Og in each of the
next two, the fourth impinger was left dry, and about 300 grams(g) of
silica gel were placed in the last impinger. About 1420 liters(l) (50 ft3)
of room air were drawn through the train at a flow rate of about 70 liters
per minute(lpm) (0.8 cfm).
The results showed a loss of 40 ml of solution in the first impinger,
a gain of 15 ml in the second, a 12 ml gain in the third, a negligible
gain in the dry impinger, and 20.5 g gain in the silica gel. The net
change across the train was 7.5 ml (assuming the mass gained on the silica
gel was water). A specific gravity determination showed that the original
IPA solution had been prepared incorrectly and was 67% IPA rather than the
specified 80% IPA. The solution remaining in the IPA impinger after the
test run was shown to be 52% IPA. The 15% loss corrected for total
volume change represented a loss of 51 ml of IPA.
For the second run, 100 ml of IPA solution was placed in the first
impinger and the rest of the sampling train was the same as for run 1.
The IPA solution was prepared as specified in the Federal Register
(December 23, 1971) and a specific gravity check of concentration showed
the solution to be 73% IPA.
<3
After some 1500 1 (54 ft ) of sample were drawn through the train,
the first IPA impinger showed a 50 ml volume loss, the second measured a
75 ml gain, the third showed an 18 ml gain, the third showed a 2 ml gain,
and the silica gel mass total increased 11 g. The net volume change across
EPA Form 1320-6 (Rev. 6-72)
-------
81
the train was a 6 ml increase. The final IPA solution concentration
was 27.4% representing a loss of about 59 ml of IPA.
These test results indicate that measurement of volume gain in the
impingers of the Method 8 for the purposes of calculating sample mois-
ture content is not impaired by any loss of IPA through evaporation.
In neither test run was there a net loss of volume from the sample trains.
The net gain was approximately equivalent to 0.6% moisture or about 25%
relative humidity.
A notable secondary finding of this short study was the great change
in IPA solution concentration during a test run. Approximately 1 ml of
IPA was removed from first impinger per 30 1 of sample gas for each test
run. Initial volume of IPA solution or IPA concentration appear to
have little effect on this ratio. Some IPA may have been evaporated and
condensed farther down the train. More IPA was probably carried through
as a mist and collected later.
This IPA loss may be significant if the concentration of IPA gets
too low to effectively inhibit oxidation of SO* during Method 8 sampling.
Joe Knoll was not aware of this potential problem and could not tell me
what a lower effective limit of IPA may be. He agreed that it could be a
significant problem not only in the possible interference from S02 oxida-
tion, but also by meeting the titration end point analysis.
I suggest further work in this problem area be undertaken. Such a
project may be suitable for one of the co-op students in the next several
months.
-------
COMPARISON OF EMISSION RESULTS FROM IN-STACK FILTER
SAMPLING AND EPA METHOD 5 SAMPLING
Peter R. Westlin and Robert L. Ajax
Abstract
A series of replicate emission tests using in-stack and out-of-stack
sampling trains were conducted at each of four fossil-fuel-fired power
generation stations. The sampling train used for measuring in-stack
particulate included a probe nozzle and an in-stack glass fiber mat
filter, followed by a heated probe extension and an out-of-stack filter
The Environmental Protection Agency Method 5 particulate sampling system
was used as the out-of-stack sampling train. The two sampling trains were
operated simultaneously at approximately the same point in the stack gas
streams with no traversing.
The particulate catch from each sampling train was analyzed for
particulate mass, sulfate content, organic content, and acidity. For
the in-stack train, the results are reported for both the in-stack catch
(the particulate obtained from the nozzle and the in-stack filter), and
the total catch (the in-stack particulate, plus the particulate washed
from the probe extension and the out-of-stack filter).
The tests at two coal-fired units with electrostatic precipitators
and an oil-fired unit with no control device resulted in the out-of-stack
train catch exceeding the in-stack catch, in each case. The difference
varied with the sulfur content of the fuel and ranged from 10 mg/dscm at
the unit firing 0.3% sulfur oil, to 112.6 mg/dscm at the unit firing 3%
sulfur coal. The measured sulfate did not, however, fully account for
this difference.
Opposite results were obtained at a second oil-fired unit with a
wet limestone scrubber. At this unit, which was burning 2.5% sulfur
fuel, the in-stack catch was significantly greater than the out-of-stack
train catch (421.3 mg/dscm versus 217.6 mg/dscm respectively). This
difference was the apparent result of a reaction occurring on the wet
in-stack filter.
Emission Measurement Branch, ESED, OAQPS, EPA, RTP, NC
Presented at the annual APCA Meeting, June 1975
82
-------
75-19.1
Introduction
During the summer of 1973 the Emission Measurement Branch of the
Environmental Protection Agency (EPA) undertook a project in which particu-
late emissions were sampled with in-stack filter and out-of-stack filter
sampling trains.
The purpose of the project was to obtain and compare particulate
emission sampling results using sampling trains in which all components
except filter location were identical. The equipment used for measuring
the in-stack particulate, as shown in Figure 1, consisted of a probe nozzle
and an in-stack filter, followed by a probe extension, and an out-of-stack
filter. The EPA Method 5 sampling train,1 shown in Figure 2, was used to
measure particulate out-of-stack at 120°C. The two trains were operated
simultaneously side-by-side at approximately the same point in the stack
gas streams at each of four fossil-fuel-burning power generating stations.
Two plants were coal-fired with electrostatic precipitators and two plants
were oil-fired, one using a wet limestone scrubber, and the other having no
supplementary emission control. The sulfur content of the fuels ranged
from 0.29 to 3.3 percent.
The particulate catch from each sampling system was analyzed for
particulate mass, sulfate content, organic content, and acidity. The
results are reported for the in-stack catch (the particulate obtained from
the nozzle and the in-stack filter), the total in-stack (the in-stack par-
ticulate plus the particulate washed from the probe extension downstream
and the out-of-stack filter), and the EPA Method 5 catch (the particulate
from the nozzle, the probe, and the out-of-stack filter). The impinger catch
results are not reported in this paper as the dry or front half results were
of concern in this project.
Methods
A special dual-probe sampling box was constructed to house the two
sampling trains and to allow for simultaneous operation of both systems.
Two equal-length sampling probes were employed side-by-side with the probe
tips approximately 10 centimeters apart. No provisions were made for tra-
versing of the stack cross-section as only relative concentrations were
desired. Although only one point was sampled, isokinetic conditions were
maintained. A pi tot tube was attached to the EPA Method 5 sampling probe
to permit velocity head measurements, and adjustments in the sampling rate
of each train were made every five minutes during sampling to maintain iso-
kinetic sampling conditions. Other measurements recorded at regular inter-
vals included stack temperature, dry gas volume, meter vacuum, gas meter
temperatures, orifice pressure drop, and sample box temperature. Sample box
temperature in the enclosure housing the box filters was carefully monitored
and maintained at or above 120°C. The sample box temperature was measured
with a thermocouple located in the downstream half of one of the box filters.
The in-stack sampling train was composed of a button-hook sampling
nozzle; a 5.7 cm diameter glass-fiber mat filter and an in-stack filter
nolder; a heated, glass-lined probe; a second 7.6 cm diameter glass-fiber
83
-------
IIKTACK FILTER HOLDS
CO
AIR-TIGHT PUMP
Figure 1. In-slack paniculate sampling tram.
en
-------
00
en
VACUUM
MAIN LINE
r 1 O VALVE
n L_iH>
-------
75-19.1
mat filter in a heated sample box; and condensation impingers. This arrange-
ment along with the pump and metering equipment is displayed in Figure 1.
The EPA Method 5 train (Figure 2) was described in the December 23, 1971
Federal Register.'
Cleanup procedures were as prescribed for Method 5 in the Federal
Register except that the nozzle tip and trie filter holder upstream of the
in-stack filter were cleaned with acetone and stored separately from the
rest of the probe. Tiie probes were rinsed and brushed with acetone, and
the catch was saved for analyses. The out-of-stack filters and the in-stack
filters were stored in glass petri dishes. The filter holders were rinsed
with acetone as were the impingers after the water condensate was saved.
Each sample was carefully analyzed for particulate mass, sulfate content,
organic content, and acidity, except for one test where only mass and sulfate
were analyzed. The samples were divided into aliquots in order to obtain all
the necessary information. The acetone solutions were divided into three
aliquots: the first was used to determine mass of particulate, the second
was titrated for acidity and for sulfate as S04, and the last was extracted
for organic materials. The filters were first weighed for particulate mass
and then divided in half: one half used to determine organic materials, and
the other half analyzed for acidity and sulfate contents.
Particulate mass was determined gravimetrically after proper dessication.
Sulfate (S0|) content was determined using the thorin titration technique.
Ether-chloroform extraction was used to establish the organic content of each
sample and an acid-base titration was employed to determine the acid content
of each sample._ The analytical results were expressed in milligrams (mg) for
particulate, $04, and organics; and in milliequivalents (meq) of h^SC^ for
acidity. These analytical results were then converted into concentration
units—milligrams per dry standard cubic meter (mg/dscm)--for statistical
analyses and reporting. The sulfate catch was assumed to exist as sulfuric
acid, and the concentration of sulfate was expressed as mg/dscm of H?SO/i +
2H20.2 * *
The statistical significance of differences between the various data
sets was determined by the t-test. For the purposes of this report, a 0.05
percent probability level was set as the minimum of acceptance or rejection
of the hypothesis.
Results
Tables 1 through 4 show emission concentrations as determined by the
emission tests at the four power plant facilities. Comparisons are made
between the EPA Method 5 concentrations and the in-stack concentrations using
data obtained during simultaneous, single point sampling. The pollutant
emission data and the oxygen data supplied by plant personnel were also used
to estimate the emission rate from grams per standard cubic meter to grams
per million calories. The oxygen measurements were made at sampling points
other than the particulate sampling points and as a result the emission rates
are only approximate values.
86
-------
75-19.1
Facility A
Table 1 shows the data obtained at facility A, a coal burning power
plant employing an electrostatic precipitator. The coal being fired had a
reported sulfur content of 3.3 percent. Only particulate mass and sulfate
concentrations were determined for this facility. The average EPA Method 5
particulate emission concentration was 129.8 gm/dscm corresponding to an
approximate emission rate of 0.49 grams per million calories (g/10° cal).J
As shown in Table 5, the average particulate concentration determined from
the EPA Method 5 train was significantly greater than the particulate catch
for the in-stack sampling train, 129.8 versus 17.2 mg/dscm. Adding the probe
wash and the filter catch downstream of the in-stack filter to the in-stack
catch produced a total particulate concentration of 124.6 mg/dscm which was
not significantly different from the Method 5 dry particulate concentrations.
Similarly, the sulfate found in the EPA Method 5 train, 76.3 mg/dscm,
was significantly greater than the 4.7 mg/dscm in-stack catch. The total
catch of the in-stack train indicated a sulfate concentration of 60.6 mg/dscm
which was not significantly different than the 76.3 mg/dscm EPA Method 5
sulfate catch.
Facility B
The emission concentration data obtained at facility B, an oil-fired
power plant with no control devices, are shown in Table 2. Sulfur content
of the oil was reported to be 0.29 percent. The average EPA Method 5 particu-
late emission concentration, 38.8 mg/dscm, was significantly greater than the
in-stack particulate concentration of 30.1 mg/dscm. Compared with facility A,
however, the actual magnitude of this difference is small: 8.7 vs 112.4 mg/dscm.
No significant difference was found between the particulate concentration found
in the total in-stack dry train, 42.0 mg/dscm, and that found in the EPA
Method 5 equipment. The EPA Method 5 particulate concentration corresponded
to an approximate emission rate of 0.05 g/lQo cal for facility B.
The difference between EPA Method 5 sulfate concentration, 13.6 mg/dscm,
and the in-stack sulfate concentration, 8.4 mg/dscm, was small and was not
statistically significant. Similarly, the total in-stack train sulfate con-
centration of 15.7 mg/dscm was not significantly different from the EPA
Method 5 catch, 13.6 mg/dscm.
The organic matter concentration of the EPA Method 5 catch was 12.8
mg/dscm, a level significantly greater than the 7.9 mg/dscm found in the
in-stack filter assembly, but significantly less than the 17.6 mg/dscm
captured in the total in-stack dry sampling train. The acidity concentration
of the EPA Method 5 sampling train, 18.9 mg/dscm of H?S04, was significantly
greater than both the in-stack concentration, 3.4 mg/dscm, and the total
in-stack sampling train concentration, 10.7 mg/dscm.
87
-------
75-19.1
Facility C
Facility C was a coal-fired boiler using coal reported at 0.85 percent
sulfur and controlling emissions with an electrostatic precipitator. Neither
the difference between the dry particulate concentration of the EPA Method 5
train, 226.2 mg/dscm, and the particulate concentration of the in-stack train,
207.9 mg/dscm, nor the difference between the EPA Method 5 particulate con-
centration and the total in-stack train concentration of 223.4 mg/dscm was
statistically significant. It is, however, noteworthy that the difference
of 18.1 mg/dscm between the EPA Method 5 train catch and the in-stack catch,
and the corresponding values of 112.6 and 8.7 mg/dscm fj"nj»t fa"!^?. A
and B respectively, each show a consistent relationship to the fuel sulfur
cSntent-0.85, 3.3 and 0.3% sulfur for facilities C, A, and J respectively.
This relationship is shown graphically in Figure 3. This is in spite of the
fact that the average EPA Method 5 particulate concentration at facility C is
equivalent to an approximate emission rate of 0.34 g/10* cal which differs
from facility B by a factor of 7.
Sulfate emissions for facility C found using the EPA Method 5 train
averaged 5.7 mg/dscm, a level significantly greater than th*^?"!™110"
determined from the catch of the in-stack sampling train, 2.9 mg/dscm. The
sulfate concentration of the EPA Method 5 catch was not significantly different
from the sulfate concentration, 4.7 mg/dscm, of the total in-stack dry
sampling train catch. The average EPA Method 5 organic concentration for
facility C, 12.4 mg/dscm, was not significantly different from the in-stack
organ c concentration of 13.5 mg/dscm. Neither was It different from the
total in-stack train concentration of 16.4 mg/dscm The acidity concentra-
tion (H?S04) of the EPA Method 5 train was small, 3.2 mg/dscm, for facility
but was2significantly greater than the acidity concentration found by the
in-stack sampling train, 2.0 mg/dscm. When the back catch was added tthe
in-stack concentration, the resulting total i n-stack acidity """^ration
was 4.0 mg/dscm. This number was significantly greater tnan the acidity
concentration obtained by the EPA Method 5 sampling assembly. Note that the
actual magnitudes of the components-sulfates, organics. a;J acidity--are
relatively small and are less than about 5 percent of the total particulate
mass for both the EPA Method 5 catch and the in-stack catch.
Facility D
Facility D was an oil-fired steam generating station using oil with a
sulfur content of 2.45 percent. The plant employed a limestone scruboer as
the emission control system. No reheat device was present in the gas stream
prior lo the sampling location. This, along with apparent problems in the
dem?ster resulted in'an exhaust gas stream which was supersaturated with
moisture The mist caused some problems in sampling and may be the source of
tte anomalies in tSe comparison results that follow The particu a e con-
centration found by the EPA Method 5 system was 217 6 mg/dscm, significant^y
less than the particulate catch of the in-stack filter, 421.3 mg/dscm. The
EPA MethSd 5epaPrticulate concentration corresponded to an approximate nass
emission rate of 0.31 g/105 cal for facility D. The total in-stack train
particulate concentration was 727.2 mg/dscm.
C
88
-------
75-19.1
SULFUR CONTENT !N F'JEL.%
Figure 3. EPA method 5 particulate minus in-stack particulate
versus sulfur content in fuel.
89
-------
75-19.1
The results of the comparison of sulfate concentrations showed different
relationships than for the participate concentrations. The sulfate concen-
tration (H?S04 + 2H20) in the EPA Method 5 train averaged 119.2 mg/dscm, a
value significantly greater than the 18.4 mg/dscm found in the in-stack fil-
ter train. Further, the EPA Method 5 value was greater than the total in-stack
train sulfate concentration of 55.5 mg/dscm. The acidity concentration of
the EPA Method 5 train, 36.6 mg/dscm, was also significantly greater than
the concentration found in the in-stack filter train, 10.2 mg/dscm, although
no significant difference was observed between the acidity concentration found
in the total in-stack sampling train, 31.7 mg/dscm, and the EPA Method 5
train. On the other hand, comparisons of the organic catch of the two samp-
ling trains resulted in relationships similar to the particulate concentration
comparisons. The average organic concentration of the EPA Method 5 train,
60.7 mg/dscm, was significantly less than the organic concentration of the
in-stack filter train, 120.9 mg/dscm, for facility D. The same relationship
was true for the total in-stack train concentration of 137.9 mg/dscm.
Evaluation of Results
Various combinations of the different portions of the total particulate
catch of each of the sampling trains were studied in order to determine the
source of the differences between measured concentrations. One combination
studied was designed to determine if sulfates as ^804 + 2^0 make up the
difference between in-stack particulate catch and the EPA Method 5 particulate
catch.3 To do this, the EPA Method 5 particulate concentration was compared
with the sum of the EPA Method 5 sulfate catch plus the in-stack non-sulfate
eaten. If these new, concentrations were found not to be significantly dif-
ferent, then the difference between the in-stack particulate catch and the
EPA Method 5 particulate catch could be attributed to the sulfate caught in
the EPA Method 5 sampling train. A similar analysis was done to determine
if the difference between the in-stack sampling train catch and the EPA
Method 5 catch could be condensible organic matter for those tests in which
organic data were available.
For facility A, the coal-burning power plant with 3.3 percent sulfur
coal, the comparison of the sulfate test showed that a significantly greater
amount of material was caught in the EPA Method 5 train than could be ac-
counted for by the sulfate as H2S04 + 2H20 in the EPA Method 5 catch. In
this case, the difference between in-stack filterable material and EPA
Method 5 catch was apparently not all sulfate matter.
The data from facility B, an oil-fired generator with 0.29 percent
sulfur fuel, showed that no significant difference could be found between
tne in-stack catch plus the EPA Method 5 sulfate catch and the EPA Method 5
dry particulate catch. Thus the sulfate found in the EPA Method 5 train
could have accounted for the difference between the in-stack dry particulate
concentration and the EPA Method 5 concentration. A similar comparison
using the organic catch instead of the sulfate showed, however, that the
90
-------
75-19.1
difference between the in-stack catch and the EPA Method 5 catch could also
have been the organic matter found in the EPA Method 5 dry particulate catch
which could also have accounted for the difference. This indicates that the
variation in the data had as great an influence on the statistical comparison
results as did the sulfate or organic catch of the EPA Method 5 sampling train.
The coal-burning power station, facility C, firing 0.85 percent sulfur
fuel produced a comparison of emissions results similar to that of facility B.
That is, either the sulfate or the organic matter found in the EPA Method 5
particulate catch could have accounted for the differences between the EPA
Method 5 particulate concentrations and the in-stack particulate concentrations.
This is as expected since there was no significant difference between in-stack
particulate collections and EPA Method 5 collection for this site.
Test results of the emissions from site D do not fall into the pattern
set previously by the other three test sites. Comparisons of concentrations
using the sulfate data or the organic data produced no significant results,
as might be expected. The particulate concentrations from the in-stack
filter were significantly greater than the dry particulate concentration
from the EPA Method 5 train and could not be accounted for with either the
sulfate catch or the organic catch.
Conclusions
The in-stack sampling train does not produce results equivalent to the
EPA Method 5 sampling train results at all power plant sites. At two power
plants where samples were collected in dry stack gases, the in-stack filter
tended to collect less material than the EPA Method 5 sampling train. There
was no significant difference between the particulate catch of the two trains
at a third power plant with dry stack gas and low sulfur fuel. At another
site where stack gases were supersaturated with water following a wet scrubber,
the in-stack filter collected considerably more particulate than the EPA
Method 5 train.
The magnitude of the differences in the material collected by the in-
stack filter and the EPA Method 5 train was much greater for the high
sulfur fuel power stations than for the low sulfur fuel power plants and
showed a consistent relationship to the fuel sulfur content. The differences
in the amounts were, however, neither directly attributable to the sulfates
found in the EPA Method 5 catch nor to organic matter. Particulate matter
collected outside the stack, downstream of the in-stack filter made up the
difference between the in-stack catch and the EPA Method 5 catch, but no
definite conclusion as to what this material was and why it passed the in-
stack filter, can be drawn from this study.
As for the cause of the high in-stack filter catch compared to the EPA
Method 5 catch in wet stack gases, chemical reaction between the mineral
scrubbing medium and the sulfur oxides in the gas stream may be occurring.
These reactions may occur in the stack gas streams, or the sulfur oxides may
react with the minerals and the moisture on the wet filter surface of the
91
-------
75-19.1
in-stack filter. These salts would not be collected on the EPA Method 5
filter as this filter is heated above the dew-point of water and is relatively
dry.
Further study in the area of in-stack filters in wet gas streams should
answer tnese questions. Other types of sampling methods may be found more
appropriate under these conditions.
References
ft- IU1? 4°T-prote "^9* of Sulfur Dioxide on Particulate Test Results.
Fnnnf communication with Chief of Combustion and Incineration Section,
Environmental Protection Agency, Research Triangle Park, N. C. August 23,
3. Shigehara, R T. R. M. Neulicht, and W. S. Smith. A Method for Calculat
ing Power Plant Emission Rates. Stack Sampling News. 1 (1), July 1973.
4. Hemeon, W.C.L., and A. W. Black. Stack Dust Sampling: In-stack Filter
or EPA Train. J. Air Pol. Control Assoc. 22 (7): 516-518, July 1972.
92
-------
75-19.1
TABLE 1
SUMMARY OF EMISSION CONCENTRATIONS FOUND
DURING SIMULTANEOUS IN-STACK AND EPA METHOD 5
EMISSION TESTS AT A COAL-FIRED POWER PLANT3
(mg/dscm)
Particulate concentration
Run
1
2
3
4
5
6
EPAD
93.9
155.0
156.2
65.7
175.3
133.7
In-stack
10.1
20.5
30.6
22.7
8.4
10.7
Total c
97.5
190.8
133.1
127.5
76.3
122.2
EPAb Di
43.8
92.7
80.9
33.0
114.6
93.0
H2S04 + 2H20 concentration
EPAb Dry In-stack Total0
1.7 5.0
2.2 26.2
3.6 108.0
14.3 87.2
2.5 52.7
3.9 84.3
Average 129.8
17.2
124.6
76.3
4.7
60.6
Sulfur content of coal =3.3 percent, average stack temperature = 139°C.
Based on catch of EPA Method 5 sampling train.
Based on sum of catches of in-stack filter and probe and dry filter of
in-stack filter.
93
-------
TABLE 2
SUMMARY OF EMISSION CONCENTRATIONS FOUND DURING SIMULTANEOUS IN-STACK AND EPA METHOD 5
EMISSION TESTS AT AN OIL-FIRED POWER PLANT3
(mg/dscm)
Participate Concentration
+ 2H20 Concentration
c
Run
1
' 2
3
4
5
6
7
8
9
10
11
12
EPAb dry
44.8
51.0
43.2
42.5
43.6
26.8
25.5
35.1
49.9
48.6
28.1
27.1
In-stack
26.0
24.7
27.2
25.5
28.2
28.6
22.5
23.7
50.9
45.7
27.8
29.9
Total0
44.7
35.8
50.9
42.6
47.4
34.1
37.1
55.0
34.2
37.8
EPAb dry
30.9
33.8
4.5
3.9
5.8
13.2
4.5
3.4
20.1
18.3
12.0
12.5
In-stack
0.6
0.7
0.9
0.6
0.7
13.9
13.3
13.5
22.6
21.2
11.9
0.7
Tota
12.6
7.9
14.8
9.5
13.6
24.0
22.1
20.1
24.4
23.4
13.2
2.8
Organic Concentration
EPAb dry In-stack Total0
Average 38.8
30.1
42.0
13.6
8.4 15.7
19.5
7.4
20.7
19.1
16.2
10.1
5.6
15.0
17.3
12.1
5.0
6.1
12.8
4.4
6.4
6.4
9.5
9.5
6.5
7.3
4.7
10.3
14.7
4.2
10.9
17.7
17.0
23.8
26.5
24.3
10.3
18.5
14.2
7.4
16.9
Acidity(H^SO.Concentration
EPAb dry In-stack Total0
7.9 17.6
28.2
35.4
27.9
30.2
24.9
12.7
19.4
34.8
3.4
2.3
3.0
4.3
18.9
4.3
3.8
3.2
4.1
9.2
3.6
3.8
3.3
1.2
1.0
1.5
2.0
17.5
11.9
17.0
11.5
20.1
18.4
9.8
8.8
2.9
2.0
5.4
3.1
3.4 10.7
Sulfur content of oil = 0.29 percent, average stack temperature = 168°C.
Based on catch of EPA Method 5 sampling train.
Based on sum of catches of in-stack filter and probe and dry filter downstream of the in-stack filter.
-------
TABLE 3
SUMMARY OF EMISSION CONCENTRATIONS FOUND DURING SIMULTANEOUS IN-STACK AND EPA METHOD 5
EMISSION TESTS AT A COAL-FIRED POWER PLANT3
(mg/dscm)
Particulate Concentration
Run
1
2
3
4
5
6
7
8
9
10
11
12
13
EPAb dry
205.1
159.3
133.0
265.6
147.4
310.4
265.6
279.5
216.9
204.6
275.6
220.1
207.2
In-stack
161.0
170.5
53.6
227.4
131.2
316.9
262.9
321.8
201.1
213.2
235.5
222.8
184.6
Total0
167.6
180.8
170.3
233.8
136.4
320.4
267.0
326.3
206.0
218.6
245.1
235.8
195.8
EPAb
3.3
3.9
5.7
6.9
5.1
5.5
8.0
5.7
4.6
5.7
9.9
5.7
4.2
Average 226.2 207.9 223.4
H2S04 + 2H20 Concentration
dry In-stack Total0
5.7
2.6
1.8
6.8
2.6
1.9
2.8
3.1
1.5
3.1
4.8
3.0
1.9
1.8
2.9
4.8
2.1
9.3
4.4
2.2
4.4
5.1
3.3
5.2
6.6
5.9
4.1
3.6
4.7
Organic Concentration Acidity(H?SOJConcentration
K *+ L« ^ » _
EPA1
dry In-stack Total0
12.7 32.0 34.4
12.7 12.4 16.6
9.5 7.8 18.8
17.1 8.0 10.5
13.6 8.3 10.4
12.5 10.8 12.3
17.0 13.2 15.1
10.7 23.4 25.0
16.1 17.6 19.7
7.4 15.2 17.6
14.3 12.5 13.8
7.2 11.0 13.2
10.7 2.9 4.5
EPA dry In-stack Total
12.4
13.5 16.4
Sulfur content of coal = 0.85 percent, average stack temperature = 199°C.
2.5
1.8
1.8
2.9
3.3
2.5
2.4
3.9
3.3
4.7
6.6
3.5
2.6
3.2
1.8
1.7
1.8
2.0
2.9
1.4
2.4
2.4
2.4
2.4
1.8
1.8
1.8
2.0
Based on catch of EPA Method 5 sampling train.
Based on sum of catches of in-stack filter and probe and dry filter downstream of the in-stack filter.
3.6
3.4
3.7
3.9
5.9
2.9
4.0
4.1
4.0
4.0
4.6
3.5
3.7
4.0
VO
01
-------
TABLE 4
.SUMMARY OF EMISSION CONCENTRATIONS FOUND DURING SIMULTANEOUS IN-STACK AND EPA METHOD 5
EMISSION TESTS AT AN OIL-FIRED POWER PLANT3
(mg/dscm)
Particulate Concentration
Run
1
2
3
4
5
6
7
8
9
10
11
12
Averai
EPAb dry
245.3
101.9
170.9
68.6
203.3
180.6
419.8
171.4
211.3
432.0
279.8
126.9
ge 217.6
In-stack
52.4
314.4
425.7
390.7
482.5
483.8
476.7
525.6
425.7
632.8
644.9
200.7
421.3
Total0
167.8
825.6
868.0
603.1
921.0
618.7
905.7
713.5
788.8
1042.8
930.8
340.7
727.2
H9SOA + 2H90 Concentration
£ *f £•
EPAb dry
102.6
88.0
173.4
53.8
178.8
78.5
110.8
112.6
107.8
127.5
159.5
136.8
119.2
In-stack
7.3
9.0
21.5
8.0
27.2
20.3
16.5
18.0
19.3
33.0
30.0
11.3
18.4
Total0
166.0
25.4
36.2
179.0
30.5
24.9
19.2
21.8
40.4
54.4
48.7
19.3
55.5
Organic
EPAb dry
62.2
37.4
58.8
34.3
70.4
40.0
78.5
63.7
79.9
68.7
65.2
69.1
60.7
Concentration Acidity (FLSOJConcentration
In-stack
6.4
74.5
102.6
124.8
141.6
132.1
209.1
143.5
117.6
162.5
142.6
93.7
120.9
total0
74.9
88.9
111.0
133.2
153.9
139.6
233.7
158.1
128.2
171.2
156.9
101.7
137.6
EPAb dry
30.3
27.5
46.1
21.8
43.6
27.4
53.3
42.1
47.3
17.1
41.9
40.3
36.6
In-stack
11.1
7.8
11.2
8.1
14.3
11.6
12.3
8.7
9.6
14.5
5.5
8.2
10.2
Total0
55.3
30.2
30.2
57.1
33.5
21.1
30.7
19.2
32.7
34.7
20.8
12.4
31.7
a Sulfur content of oil = 2.45 percent, average stack temperature = 60"C.
b Based on catch of EPA Method 5 sampling train.
c Based on sum of catches of in-stack filter and probe and dry filter downstream of in-stack filter.
10
en
-------
TABLE 5
RESULTS OF t-TEST COMPARISONS OF CONCENTRATION DETERMINED
FROM IN-STACK AND EPA METHOD 5 SAMPLING AT FOSSIL FUEL POWER PLANTS
Facility
A
Participate Mass
Method 5a > In-stackb
Method 5 = Total0
Sulfate (H2S04 + 2H2
Method 5 > In-stack
Method 5 = Total
Organic
Acidity(H2S04)
Method 5
Method 5
> In-stack
= Total
Method 5 = In-stack
Method 5 = Total
Method 5 > In-stack
Method 5 < Total
Method 5 > In-stack
Method 5 > Total
Method 5
Method 5
= In-stack
= Total
Method 5 < In-stack
Method 5 < Total
Method 5 > In-stack
Method 5 = Total
Method 5 > In-stack
Method 5 > Total
Method 5 = In-stack
Method 5 = Total
Method 5 < In-stack
Method 5 < Total
Method 5 > In-stack
Method 5 < Total
Method 5 > In-stack
Method 5 - Total
a Based on participate catch of EPA Method 5 dry sampling train.
b Based on participate catch of in-stack filter.
c Based on sum of participate catch of in-stack filter and probe and filter downstream of in-stack filter.
VO
-------
EPA METHOD 5 SAMPLE TRAIN CLEAN-UP PROCEDURES
Clyde E. Riley*
Introduction
In the performance of participate source emission tests, an
important procedure affecting the accuracy is sample recovery. Accurate
results are not possible unless proper procedures are conscientiously
applied in recovering and quantitatively transferring particulate matter
from the sample train to the storage container. Often, however, these
procedures receive only minimum attention. Well-trained and highly
experienced technical staff are normally employed to design and oversee
the performance of a test and the writing of a test report while, in
contrast, the least experienced personnel are often assigned sole
responsibility with limited guidance for the recovery of sample from the
train—a task which includes a high potential for producing significant
errors.
The accuracy of sample recovery procedures are, of course, not only
dependent on the ohysical transfer of sample from the train to the storage
containers; the procedures also involve the selection of proper equipment,
use of proper materials, application of proper cleaning, handling, and
shipping techniques, and an overall awareness of the importance of each
phase of the sample handling procedure. The following guidelines describe
procedures which are employed by the Emission Measurement Branch to assist
in minimizing sources of error in EPA Method 5 sample train cleanup. These
are presented here, both to call attention to the degree of detail which must
be considered in sample recovery, and to make the procedures available to
others engaged in source sampling. These guidelines do not include techniques
* Emission Measurement Branch, ESED, OAOPS, EPA, RTP, NC
Published in Stack Sampling News 3(1): 4-7, July 1975
98
-------
for analysis of the impinger catch or any specific procedures other
than those necessary for analyzing the sample for mass only. Also, it
should be noted that these procedures are not regulatory requirements;
rather, they are procedures to be used by contractors
employed by the Emission Measurement Branch. Although these procedures
reflect the collective experience gained by EMB in the conduct of several
hundred source tests, we recognize that other source sampling groups may use
different cleanup techniques. It is hoped, therefore, that this publication
will provide the impetus to others to publish such alternate or improved
techniques.
Pretest Preparation
1. Brushes and sample recovery support items shall be properly
cleaned and enclosed in dust-free packaging before being used in the
sample recovery operations. This includes the sample containers as well
as the sample collector glassware.
2. Sample containers to be used for the liquid samples shall be
Type I, chemically-resistant, borosilicate narrow-mouth glass bottles
(500 mis. or 1000 mis. size). Screw-cap closures with Teflon rubber-
backed liners shall be used on all such sample containers. Use of any
other type liquid sample container, closure, or liner shall be verified
acceptable prior to use.
3. Glass or plastic petri dishes shall be used to contain the filter
samples, unless otherwise specified by EPA.
4. Pre-weighed indicating silica yel shall be acceptable only if the
containers are completely full and tightly sealed.
Trade Name
99
-------
2
5. Only fresh ACS reagent grade chemicals shall be used for sample
cleanup and recovery.
6. All reagents and samples shall be stored in sealed, non-contaminating
containers. This includes acetone which shall be purchased and stored in
glass containers. Only acetone with blank values less than 0.001 shall be
acceptable for sample recovery operations.
7. If water is required for cleanup of the probe and filter assembly,
it shall be distilled and stored in non-contaminating containers.
Sample Recovery
1. Proper sample recovery procedure begins as soon as the probe is
removed from the stack at the completion of the sampling period. When the
probe can be safely handled, wipe off all external particulate matter near
the tip of the probe nozzle and place a cap over the nozzle tip. Do not
cap off the probe tip tightly while the sampling train is cooling as this
will create a vacuum in the filter holder, thus drawing water from the
impingers into the filter holder.
2. Before moving the sample train to the cleanup site, remove the probe
from the sample train, inspect for condensed water, wipe off the silicone
grease, if used, and cap the open end of the probe. Be careful not to lose
any condensate. Wipe off the silicone grease from the filter inlet where
the probe was fastened and cap it loosely. Remove the umbilical cord from
the last impinger and cap the impinger opening. If a flexible line is used
between eicher the first impinger or condenser and the filter holder, dis-
connect the line at the filter holder and drain any condensed liquid into the
2
American Chemical Society
100
-------
impingers or condenser and remove the line from the impinger. After
wiping off the silicone grease, cap off the filter holder outlet and
impinger inlet and the flexible line, if used. Either ground glass
stoppers or their EPA approved equivalent may be used to close these
openings.
3. Transfer the probe and filter-impinger assembly to the cleanup
area. Exercise care in moving the collection train from the test site to
the sample cleanup area to avoid the loss of collected sample or the gain of
extraneous particulate matter. This area shall be clean and protected from
the wind to minimize the chances of contaminating or losing portions of
the sample.
4. Prior to sample cleanup and during disassembly, an inspection shall
be made of the individual components of the sample collector. This
inspection should reveal whether or not the sample collector was functioning
properly. Also by observing the quantity of sample, it can be estimated if
a sufficient amount of matter has been collected for proper analysis.
Record any items that could possibly affect the results (e.g., cracked
or broken glassware, water in the filter holder, unexpected residue, spent
silica gel). State whether or not the sample is still valid and give basis.
5. A consistent procedure shall be used for the sample collector
disassembly and cleanup. The following order is recommended:
General
a. The sample containers shall be tightly capped after the
sample recovery operation. The closure caps shall be sealed to thp "arrow-
mouth containers with shrink bands, plastic tape, or their equivalent.
101
-------
The glass petri dishes shall be sealed around their circumference with
large rubber bands and secured with plastic tape or its equivalent.
b. All samples including blanks shall be assigned individual
identification numbers by using pre-numbered EPA sample identification labels.
Where more than one container is needed to contain a given sample, each
additional container shall be assigned the same basic identification number.
All such multiple containers shall be further marked to indicate the total
number of containers used for that sample and which container of the series
each represents (examples 1 of 3, 2 of 3, etc.).
c. After the recovery operation, the volume of all liquid
samples including rinses shall be documented either by using graduated sample
bottles and recording the sample volume on the recovery sheet or by permanently
marking the sample container and/or label to indicate the liquid level. By
doing this, the laboratory will be able to determine whether or not sample
leakage occurred during transport.
d. A 200 ml blank reagent sample shall be collected for each
lot of rinse reagents used. Representative blank samples of the acetone or
other solvents, distilled water (if used), and preweighed filters (quantity
three) shall be collected during the test program. The acetone and water
samples shall be analyzed to determine the amount of contamination attributed
to the sample reagents.
102
-------
Filter - Remove the filter holder and inspect the filter mat
for punctures or tears before removing and placing it in an identified
glass or plastic petri dish container. Use a pair of parallel tweezers
and/or clean disposable surgical gloves to handle the filter. If it is
necessary to fold the filter, do so such that the particulate cake is
inside the fold. Quantitatively remove any particulate matter and/or
filter media which may adhere to the filter holder or support by carefully
using a dry nylon bristle brush, rubber policeman, or a sharp-edged blade.
Place this matter into the same container as the filter. Seal the container
as described in the General Section.
Probe - It is recommended that two people be used to clean the
probe to minimize altering the sample. The probe cleanup and disassembly
shall be conducted in the following order. Making sure that dust on the
outside of the probe or other exterior surfaces does not enter into the
sample, quantitatively transfer the particulate matter and condensete from
the probe nozzle, probe fitting, probe liner, and front half of the filter
holder to container No. 2. Rinse these components with acetone, distilled
water (if required), or other appropriate rinsing solvents that have been approved
103
-------
by EPA. In all cases, collect a representative blank of the rinse
solvents. Specific steps are as follows:
a. Carefully remove the probe nozzle and clean the inside
surface by triple rinsing with acetone from a glass wash bottle and brushing
twice with a precleaned nylon brush. Continue brushing until the acetone
rinse shows no visible particles, after which perform a final rinse of the
inside surface with acetone.
b. Brush and rinse with acetone the inside parts of the
probe fitting in a similar way, i.e., until the rinse shows no visible
particles remaining.
c. Rinse the probe liner with acetone by tilting and
squirting acetone into its upper end, while rotating the liner in a 360°
manner so that all inside surfaces will be rinsed. Let the acetone drain
from the lower end into the sample container. A second acetone rinse shall
be performed with the aid of a probe brush. Position the liner as before
and squirt acetone into the upper end while pushing the brush through the
entire length of the liner using a twisting action. Repeat the brushing and
rinsing operation (minimum two times) until no particulate matter remains
in the probe liner upon visual inspection. With stainless steel or other
metal liners, brush and rinse in the above prescribed manner at least six
times; metal liners have small crevices in which particulate matter can be
entrapped. Upon completion of the brushing and rinsing operation, rinse the
brush with acetone and perform a final acetone rinse on the liner. Collect
these rinsings in the same sample container as before.
104
-------
d. After ensuring that the filter holder has been
wipped clean of silicone grease, clean the inside of the front half of
the filter holder by double brushing with a nylon bristle brush while
rinsing with acetone or brush and rinse until all visible particulate is
removed. Make a final rinse of the brush and inside surface of the front
half of the filter holder. Again these rinsings are placed in the No. 2
sample container. (Note: Do not rinse or brush the fritted-glass support.)
Silica Gel - Record the color and condition of the indicating
silica gel in the last impinger and determine if it is completely saturated.
Weigh the used silica gel to the nearest 0.5 gm and determine the amount
of moisture collected. The silica gel shall be transferred to a shipping
container or discarded if contaminated.
Impinger Catch - If analysis of the impinger catch is not required,
discard the liquid after measuring and recording the volume or weight.
105
-------
TECHNICAL REPORT DATA
'/J'i'tjc reau liKlnicnru'i i»» ill
1 REPCPT NO |2
£PA-450/2-78-042d |
4 T'T._E AND SUBTITLE
Stack Sampling Technical Information: A Col
of Monographs and Papers Volume IV
7 AUTHOR(S)
Roger T. Shigehara (Editor)
9 PERFORMING ORGANIZATION NAME AND ADDRESS
U.S. Environmental Protection Agency
i -i iffji be for.; co' '
lection
Emission Standards and Engineering Division
Emission Measurement Branch
Research Triangle Park, NC 27711
12 SPONSORING AGENCY NAME AND ADDRESS
Same as above.
15 SUPPLEMENTARY NOTES
16 ABSTRACT
"Stack Sampling Technical Information" is a four-volume
and papers which have been compiled by the
. ii-if}
3 WtC'PicNT S ACCESSION NO
5 REPOFtf^DATc
October 1978
6 PERFORMING ORGANIZATION CODE
8 PERFORMING ORGANIZATION REPCRT NO
10 PROGRAM ELEMENT NO
11 CONTRACT/GRANT NO
13 TYPE OF REPORT AND PERIOD COVERED
14 SPONSORING AGENCY CODE
collection of monographs
Emission Measurement Branch, ESED, OAQPS.
The information specifically relate to current EPA test
methods and compliance
test procedures. The data presented in some of these documents have
served as
the basis for a number of revisions made in the EPA Reference Methods 1 through 8.
Several of the documents are also useful in determining
procedures .
acceptable al
ternative
17 KEY WORDS AND DOCUMENT ANALYSIS
J DESCRIPTORS
Gas Sampling
Filtered Particle Sampling
Gas Analysis
1R 3 ;-RiB'_|T,r>N ST/l'LMC WT
Unlimited
b IDENTIFIERS/OPEN ENDED TERMS
Stack Sampl
ing
19 SECURITY CLASS !lln\ Report i
Unclassified
20 SECURITY CLASS iTIni pugei
Unclassified
c COSATl 1 icId/Croup
14B
14D
21 NO OF PAGES
118
22 PRICE
£Pi For-, 72?" ! iR- « 4-~
'106
------- |